Part 2 book “Burger’s medicinal chemistry and drug discovery” has contents: Structure - based drug design, electron cryomicroscopy of biological macromolecules, mass spectrometry and drug discovery, peptidomimetics for drug design, natural products as leads for new pharmaceuticals,… and other cotents.
CHAPTER TEN Structure-Based Drug Design LARRY W HARDY Aurigene Discovery Technologies Lexington, Massachusetts DONALD J ABRAHAM Virginia Commonwealth University Richmond, Virginia MARTIN K SAFO Virginia Commonwealth University Richmond, Virginia Contents Introduction, 418 Structure-Based Drug Design, 419 2.1 Theory and Methods, 419 2.2 Hemoglobin, One of the First Drug-Design Targets, 419 2.2.1 History, 419 2.2.2 Sickle-Cell Anemia, 419 2.2.3 Allosteric Effectors, 421 2.2.4 Crosslinking Agents, 424 2.3 Antifolate Targets, 425 2.3.1 Dihydrofolate Reductase, 425 2.3.2 Thymidylate Synthase, 426 2.3.2.1 Structure-Guided Optimization: AG85 and AG337,426 2.3.2.2 De Novo Lead Generation: AG331,428 2.3.3 Glycinamide Ribonucleotide Formyltransferase, 429 2.4 Proteases, 432 2.4.1 Angiotensin-Converting Enzyme and the Discovery of Captopril, 432 2.4.2 HIV Protease, 433 2.4.3 Thrombin, 442 2.4.4 Caspase-1, 444 2.4.5 Matrix Metalloproteases, 445 2.5 Oxidoreductases, 446 2.5.1 Inosine Monophosphate Dehydrogenase, 447 2.5.2 Aldose Reductase, 448 2.6 Hydrolases, 449 2.6.1 Acetylcholinesterase, 449 2.6.2 Neuraminidase, 450 Burger's Medicinal Chemistry a n d Drug Discovery Sixth Edition, Volume 1: Drug Discovery Edited by Donald J Abraham ISBN 0-471-27090-3 Q 2003 John Wiley & Sons, Inc 417 Structure-Based Drug Design 2.6.3 Phospholipase A2 (Nonpancreatic, Secretory), 452 2.7 Picornavirus Uncoating, 454 2.8 Phosphoryl Transferases, 456 INTRODUCTION Structure-based drug design by use of structural biology remains one of the most logical and aesthetically pleasing approaches in drug discovery paradigms The first paper on the potential use of crystallography in medicinal chemistry was written in 1974 (1)and was presented at Professor Alfred Burger's retirement symposium in 1972 The excerpted last paragraph in the paper, reproduced below, foresaw the integration of X-ray crystallography into the field of medicinal chemistry It is reasonable to assume then that the future of large molecule crystallography in medical chemistry may well be of monumental proportions The reactivity of the receptor certainly lies in the nature of the environment and position of various amino acid residues When the structured knowledge of the binding capabilities of the active site residues to specify groups on the agonist or antagonists becomes known, it should lead to proposals for syntheses of very specific agents with a high probability of biological action Combined with what is known about transvort of drugs through a Hansch-type analysis, etc., it is feasible that the drugs of the future will be tailormade in this fashion Certainly, and unfortunately, however, this day is not as close as one would like The X-ray technique for large molecules, crystallization techniques, isolation techniques of biological systems, mechanism studies of active sites and synthetic talents have not been extensively intertwined because of the existing barriers (1) Since that time there have been numerous successes in advancing new agents into clinical trials by combining crystallography with associated fields in drug discovery Currently, more structures are solved every year than were in the entire Protein Data Bank in 1972 Although almost every major pharmaceutical company has an X-ray diffraction group, Agouron (now Pfizer) was the first biotechnology startup company to make drug discovery based on X-ray crystallography a central and primary theme (2) Other startup companies 2.8.1 Mitogen-Activated Protein Kinase p38a, 456 2.8.2 Purine Nucleoside Phosphorylase, 459 2.9 Conclusions and Lessons Learned, 461 (such as BioCryst and Vertex) were soon founded to apply similar approaches More recent companies, such as Structural Genomix (3) and Astex (41, and the High Throughput Crystallography Consortium, organized by Accelrys (5), have emerged to carry on structure-based drug discovery in a high throughput environment Third-generation synchrotron sources, such as the Advanced Photon Source (APS)at Argonne National Laboratory outside Chicago, and new detectors, have enormously increased the speed of data collection It is now possible to collect high resolution data from protein crystals, solve, and refine the structure in days to a few months This information is covered in an adjacent chapter Simultaneous advances in computing have added to the speed of obtaining threedimensional structural information on interesting drug design targets These developments, coupled with the sequencing of the human genome and the advent of bioinformatics, provide workers in structure-based drug design with a plethora of opportunities for success The utility of any drug discovery tool is measured, in the final analysis, by the output of the tool's use New tools are burdened with unrealistically high expectations As their application begins, the impact is sometimes more limited than was originally envisaged Structure-based design methods have had great utility for the design of enzyme inhibitors, tight-binding receptor ligands, and novel proteins The utility of these methods for the design of drugs is somewhat more limited, simply because there are so many factors that must be balanced in the successful design of a drug Nonetheless, structure-based drug design (SBDD), distinct from the (far easier) structure-based inhibitor design, is now a reality and has had significant impact Aspects of the methods and utility of SBDD have been described in several excellent review articles and monographs (6-12) This chapter focuses on the utility of SBDD in the cases of drugs that have been launched as products, or that have at least entered human clinical trials In some cases, SBDD has been a remarkable suecess In others, it has failed in the sense that, despite its use, the candidate produced did not gain approval to become a marketed drug In the latter cases, this was usually not truly a failure of SBDD, but rather attributed to the complex criteria that drugs must meet, and to the complex regulatory hurdles that candidates and companies face In addition to providing a measure of the impact of SBDD on the creation of actual drugs, these examples will also provide lessons about how to apply SBDD in drug discovery The chapter is not completely encyclopedic, and some significant instances of SBDD will be missed by the informed reader However, the discovery programs with drugs and drug candidates that are discussed will provide sufficient diversity that general trends can emerge In a few cases, relevant results for preclinical compounds that seem likely to enter human trials are described A growing number of the drugs to which structural design methods are applied are themselves proteins (e.g., cytokines, immunomodulators, monoclonal antibodies) However, this chapter is restricted to small organic molecules that are designed by use of the three-dimensional structure of a target protein the biological action with precise structural information It makes good sense at the early stages of design to use lead molecule structural scaffolds that retain low toxicity profiles, given that the latter most often derails successful drug discovery The most active derivative(~)from this cyclic process can be forwarded for in vivo evaluation in animals 2.2 Hemoglobin, One of the First Drug-Design Targets 2.2.1 History Perutz and colleagues de- termined the first three-dimensional structures of proteins Through use of X-ray crystallography Kendrew determined the structure of myoglobin (13), whereas Perutz determined the structure of hemoglobin (Hb)(14-16) At the present time, new protein and nucleic acid structures and complexes are published weekly However, for a long period after the first protein structures were solved, progress was slower Hb was of interest for drug discovery purposes because of the early identification of the mutant Glu -,Val, which causes sickle-cell anemia The crystal structure of sickle Hb (Hbs) was published by Wishner et al (17) and was later solved at a higher resolution by Harrington et al (18) 2.2.2 Sickle-Cell Anemia In 1975, through STRUCTURE-BASED DRUG DESIGN 2.1 Theory and Methods The concept of structure-based drug discovery combines information from several fields: Xray crystallography and/or NMR, molecular modeling, synthetic organic chemistry, qualitative structure-activity relationships (QSAR), and biological evaluation Figure 10.1 represents a general road map where a cyclic process refines each stage of discovery Initial binding site information is gained from the three-dimensional solution of a complex of the target with a lead compound(s) Molecular modeling is usually next applied with the intent of designing a more specific ligandk) with higher affinity Synthesis and subsequent in vitro biological evaluation of the new agents produces more candidates for crystallographic or NMR analysis, with the hope of correlating use of the three-dimensional Hb coordinates, two groups initiated SBDD studies to discover an agent to treat sickle-cell anemia: Goodford et al in England and Abraham et al in the United States Goodford's group was the first to develop an antisickling agent (BW12C), based on structure-based drug design, which reached clinical trials (19, 20) However, Wireko and coworkers were unable to confirm the BW12C binding site proposed by Goodford (21) The second antisickling agent proposed by Abraham et al to advance to clinical trials was the food additive vanillin (compound la) (22) The crystallographic binding site of BW12C (lb)was found to be at the N-terminal amino groups of the a-chains (21), whereas that of vanillin shows binding close to aHisl03 and also at a minor site between PHis117 and PHis117 (22) A recently redetermined binding site of vanillin at a higher resolution shows weak binding to the N-terminal amino group Structure-Based Drug Design Figure 10.1 Schematic of the structure-based drug discovery/design process The figure maps out the iterative steps that make use of X-ray crystallography, molecular modeling, organic synthesis, and biological testing to identify and optimize ligand-protein interactions CHO CHO I (lb) BW12C (la) vanillin of the a-chain (23) A derivative of vanillin has been patented and is a candidate for clinical trials Two marketed medicines, ethacrynic acid (21, a diuretic agent, and clofibric acid (3),an antilipidemic agent, were reported to have strong antigelling activity (24, 25), and through X-ray analyses of cocrystals, the binding sites of these agents to Hb were elucidated (26) Unfortunately, it was found that high Structure-Based Drug Desij d Clinical Trials Figure 10.1 Schematic of the structure-based drug discovery/design process The figure maps out the iterative steps that make use of X-ray crystallography, molecular modeling, organic synthesis, and biological testing to identify and optimize ligand-protein interactions CHO I OH (la) vanillin of the a-chain (23).A derivative of vanillin has been patented and is a candidate for clinical trials Two marketed medicines, ethacrynic acid CHO (lb) BW12C (21, a diuretic agent, and clofibric acid (3),a antilipidemic agent, were reported to hav strong antigelling activity (24, 25), an through X-ray analyses of cocrystals, the bind ing sites of these agents to Hb were elucidate1 (26) Unfortunately, it was found that higl ucture-Based Drug Design (2) ethacrynic acid (3) clofibric acid corm:ntrations of ethacrynic acid were needed to intieract with Hb in deformed red cells (27) Clofil~ r i cacid, when administered in a gm/ day (lose (as the ethyl ester clofibrate), appear6:d to be an ideal potential treatment for sicklc?-cell anemia, but was subsequently founcI to be highly bound to serum proteins and nlot transported in quantities sufficient to inter;~ c with t sickle Hb Furthermore structure-1based derivatives were not found to be effective (28, 29) Tkle major problem with designing a small molec:ule to treat sickle cell anemia is not so much an issue of specificity, but arises from the tr,eatmentof a chronic disease The potential cumulative toxicity from the amount of drug needed to interact with approximately two plounds of hemoglobin S over a homozygous patient's lifetime is the major concern (22) 1:for a review, see Vol 3, Chapter 10 Sicklt? Cell Anemia, by Alan Schecter et al) logical role is to right shift the Hb oxygenbinding curve to release more oxygen The binding site of 2,3-DPG, determined by Arnone (30) lies on the dyad axis at the mouth of the p-cleft (Fig 10.2) interacting with the Nterminal PVall, PLys82, and PHis143 of deoxy Hb A more recent study at a higher resolution, by Richard et al (31), found DPG to interact with the residues PHis2 and PLys82 Goodford and colleagues were the first to design agents that would bind to the 2,3-DPG site (32-34) An effective allosteric effector that can enter red cells might be used to treat hypoxic diseases such as angina and stroke, to enhance radiation treatment of hypoxic tumors, or to extend the shelf life of stored blood Many antigelling agents left shift the oxygen binding curve, producing higher concentrations of oxy-HbS Given to patients with sickle-cell anemia, this should result in less polymerization, and therefore less red blood cell sickling It was a surprise therefore when clofibric acid, which blocks sickle-cell Hb polymerization, was found to shift the Hb oxygen binding curve to the right, in a manner similar to that of 2,3-DPG (25) The clofibric acid binding site was found to be far removed from the 2,3-DPG site (25, 35) The determination of the clofibric acid binding site on Hb was the first report of a tense state (deoxy state) allo- 2.2!.3 Allosteric Effectors $&Diphosphoglycer.ate (2,3-DPG, compound 41, found in most Imammalian red cells, is the naturally occurrir~gallosteric effector for Hb Its physio- Figure 10.2 View of (4) (2,3-DPG) binding site at the mouth of the p-cleft of deoxy hemoglobin See color insert Structure-Based Drug Design steric binding site different from that of 2,3DPG (compound 4) Perutz and Poyart tested another antilipidemic agent, bezafibrate (compound 5), and found that it was an even more (5) bezafibrate potent right-shifting agent than clofibrate (36) Perutz et al (26) and Abraham (35) determined the binding site of bezafibrate and found it to link a high occupancy clofibrate site with a low occupancy site Lalezari and Lalezari synthesized urea derivatives of bezafibrate (37), and with Perutz et al determined the binding site of the most potent derivatives (38) Although these compounds were extremely potent, they were hampered by serum albumin binding (39,40) Abraham and coworkers synthesized a series of bezafibrate analogs (39-42) One of these agents, efaproxaril (RSR 13, compound 6a) is currently in Phase I11 clinical trials for radiation treatment of metastatic brain tu- mors (see, Vol 4, Chapter Radiosensitizers and Radioprotective Agents, by Edward Bump et al) The binding constants and binding sites of a large number of these bezafibrate analogs were measured and agreed with the number of crystallographic binding sites found (42) The degree of right shift in the oxygen-binding curve produced by these compounds was not solely related to their binding constant, providing a structural basis for E J Ariens' theory of intrinsic activity (42) By use of X-ray crystallographic analyses, the key elements linking allosteric potency with structure were uncovered In addition, the computational program HINT, which quantitates atom-atom interactions, was used to determine the strongest contacts between various bezafibrate analogs and Hb residues These analyses revealed that the amide linkage between the two aromatic rings of the compounds must be orientated so that the carbony1 oxygen forms a hydrogen bond with the side-chain amine of aLys99 (41, 43) Three other important interactions were found The first are the water-mediated hydrogen bonds between the effector molecule and the protein, the most important occurring between the effector's terminal carboxylate and the sidechain guanidinium moiety of residue olArgl41 Second, a hydrophobic interaction involves a methyl or halogen substituent on the effector's terminal aromatic ring and a hydrophobic groove created by Kb residues aPhe36, aLys99, aLeu100, aHisl03, and pAsnl08 Third, a hydrogen bond is formed between the side-chain amide nitrogen of Asnl08 and the electron cloud of the effector's terminal aromatic ring (40,41,43).Abraham first observed this last interaction while elucidating the Hb' binding site of bezafibrate (35) Burley and Petsko had previously pointed out this type of hydrogen bond in a number of proteins, indicating that this contact is involved in a number of other receptor interactions (44,451 Perutz and Levitte estimated this bond to be about kcal/mol (46) Figure 10.3 shows the overlap of four allosteric effectors (6a, 6b, 7a and 7b) that bind at the same site in deoxy Hb but differ in their allosteric potency itructure-Based Drug Design Figure 10.3 Stereoview of allosteric binding site in deoxy hemoglobin A similar compound environment is observed at the symmetry-related site, not shown here (a) Overlap of four right-shifting allosteric effectors of hemoglobin: (6a) (RSR13, yellow), (6b)(RSR56, black), (7a) (MM30, red), and [7b)(MM25,cyan) The four effectors bind at the same site in deoxy hemoglobin The stronger acting RSR compounds differ from the much weaker MM compounds by reversal of the amide bond located between the two phenyl rings As a result, in both RSR13 and RSR56, the carbonyl oxygen faces and nakes a key hydrogen bonding interaction with the m i n e of mLys99 In contrast, the carbonyl xygen of the MM compounds is oriented away from the mLys99 amine The aLys99 interaction with ;he RSR compounds appears to be critical in the allosteric differences (b) Detailed interactions ~etweenRSR13 (6a) and hemoglobin, showing key hydrogen bonding interactions that help constrain the T-state and explain the allosteric nature of this compound and those of other related :ompounds See color insert 423 Structure-Based Drug Design 424 Over the course of these studies, an interesting anomaly was solved Allosteric effectors (such as 8a and 8b)can bind to a similar site H3C CH3 (8a)DMHB CHO and yet effect opposite shifts in the oxygenbinding curve Agents such as 5-FSA bind to the N-terminal Val and provide groups for hydrogen bonding with the opposite dimer (across the twofold axis) right shift the oxygen-binding curve In contrast, agents that disrupt the water-mediated linkage between the N-terminal aVal with the C-terminal &gl41 left shift the curve (47) (Fig 10.4) Structure-based stereospecific allosteric effectors for Hb have also been developed and possess activities and profiles appropriate for clinical efficacy (48,49) 2.2.4 CrosslinkingAgents The first crosslink- ing agent that possessed potential as a Hbbased blood substitute was described by Walder et al (50) Bis(4-formylbenzy1)ethane and bisulfite adducts of similar symmetrical aromatic dialdehydes, previously studied by Goodford and colleagues, provided the starting points that led to these compounds Chatterjee et al identified the binding site to deoxy-Hb, and found that the two Lys99 side chains were crosslinked (51) One of the derivatives was proposed as a blood substitute (52), and has been explored commercially (see Vol 3, Chapter Oxygen Delivery and Blood Sub- Figure 10.4 Stereoview of superimposed binding sites for (8b)(5-FSA, yellow) and (8a)(DMHB, magenta) in deoxy hemoglobin A similar compound environment is observed at the symmetry-related site and therefore not shown here Both compounds form a Schiff base adduct with the cvlVall N-terminal nitrogen Whereas the m-carboxylate of 5-FSA forms a salt bridge with the a2Arg141 (opposite subunit), this intersubunit bond is missing in DMHB The added constraint to the T-state by 5-FSA that ties two subunits together shifts the allosteric equilibrium to the right On the other hand, the binding of DMHB does not add to the T-state constraint Instead, it disrupts any T-state salt- or water-bridge interactions between the opposite a-subunits The result is a left shift of the oxygen equilibrium curve by DMHB See color insert stitutes and Blood Products, by Andeas Mozzarelli et al.) Another crosslinked Hb engineered by Nagai and colleagues, at the MRCLMB in Cambridge, was developed into a blood substitute that was clinically investigated at Somatogen, now Baxter (53) Boyiri et al synthesized a number of crosslinking agents (molecular ratchets, such as 9) whose OHC 'Qo~~~ potency was directly related to the length of the crosslink: the shorter the crosslink (three atoms), the stronger the shift of the oxygen binding curve to the right (54) (Fig 10.5) Perutz's hypothesis (55) and the MWC model (56) for allostery, that the more tension is added to the tense (deoxy) state of Hb, the greater the shift to the right of the oxygen- Stru~cture-BasedDrug Design 425 Fi gure 10.5 Stereoview of the binding site for (9) (n = 3, TB36, yellow) in deoxy Hb A similar co:mpound environment is observed at the symmetry-related site, not shown here One aldehyde is CO'valently attached to the N-terminal alVall, whereas the second aldehyde is bound to the opposite subunit, a2Lys99 ammonium ion The carboxylate on the first aromatic ring forms a bidentate hy.drogen bond and salt bridge with the guanidinium ion of a2Arg141 of the opposite subunit The efiTector thus ties two subunits together and adds additional constraints to the T-state, resulting in a shift in the Hb allosteric equilibrium to the right The magnitude of constraint placed on the T-state by the crosslinked aLys99 varies with the flexibility of the linker Shorter bridging chains form tig:hter crosslinks and yield larger shifts in the allosteric equilibrium See color insert bindi~ ig curve, are generally consistent with the biehavior of the allosteric effectors and cross1inking agents TS I C1-Tetrahydrofolate Dihydrofolate - reduced form of folate (tetrahydrofolate) acts as a one-carbon donor in a wide variety of biosynthetic transformations This includes essential st;eps in the synthesis of purine nucleotides 2md of thymidylate, essential precursors to DNIA and RNA For this reason folate-dependent enzymes have been useful targets for the dlevelopment of anticancer and anti-inflamrrlatory drugs (e.g., methotrexate) and anti-irlfedives (trimethoprim, pyrimethamine) During the reaction catalyzed by thymidylate synthiase (TS), tetrahydrofolate also acts as a reducitant and is converted stoichiometrically to dikydrofolate The regeneration of tetrahydrofolate, required for the continuous functioning of this cofactor, is catalyzed by dihydrofolate reductase (DHFR) 2.3 I Dihydrofolate [Purines] I t intifolate Targets DHFR Tetrahydrofolate Reductase The Thymidylate Scheme 10.1 The first crystal structure of a drug bound to its molecular target was provided by the pioneering X-ray diffraction study of the complex between DHFR and methotrexate (57), albeit in this case the target - was a bacterial surrogate for the actual target (the human enzyme) Once X-ray structures of DHFR from eukaryotic sources were also solved, comparisons of the bacterial and eukaryotic DHFR " structures revealed the structural basis for the selectivity of the antibacterial drug trimethoprim for the bacterial enzyme This understanding allowed Goodford and colleagues Structure-Based Drug Design to rationally design trimethoprim analogs with altered potencies (58) Retrospective studies such as those done by David Matthews and others on DHFR (see, for example, Ref 59) set the stage for the iterative process of structure-based inhibitor design as it was later developed at Agouron Pharmaceuticals, targeted against another folate-dependent enzyme, TS (60, 61) 2.3.2 Thymidylate Synthase There are two main modes in which structure-based methods for inhibitor design have been employed The first mode is structure-guided optimization of the design of a previously known chemical scaffold The scaffold could be a known drug or inhibitor, substrate analog, or a hit from screening of a random library The property, which is modified during the optimization, may be, for example, potency, solubility, or target selectivity, or the more challenging aim may be to optimize several properties simultaneously A second and potentially more powerful mode is for the de nouo design of inhibitory ligands, sometimes referred to as lead generation This mode relies strictly on the structure of the target enzyme or receptor as a template A substrate or an inhibitor may be bound to the crystalline target, and deleted to provide the template This is advantageous when, as in the case of TS, a substantial conformational change occurs when ligands bind After a de nouo design process has provided a new inhibitor that is structurally unique, the properties of the new scaffold can be optimized by continued structural guidance Both modes of SBDD have been used to generate TS inhibitors that have entered clinical trials When the design of inhibitors of human TS at Agouron Pharmaceuticals began, the amounts of the human enzyme required for crystallographic study were unavailable Because the active site of the enzyme is so highly conserved, it was assumed that an acceptable surrogate for human TS would be the crystal structure of a bacterial TS (60, 62) Figure 10.6 shows the conformation of the quinazoline folate analog 10 (N10-propynyl-5, 8-dideazafolate), bound within the active site of the Escherichia coli enzyme with the nucleotide substrate, 2'-deoxyuridine-5'-monophosphate (63, 64) This folate analog, designed by classical medicinal chemistry as an analog of the TS substrate, 5,lO-methylenetetrahydrofolate (111, is a potent TS inhibitor Nevertheless, (10) failed as an anticancer drug because of its insolubility and resulting nephrotoxicity (65) 2.3.2.1 Structure-Guided Optimization: AG85 andAG337 In the crystalline complex with E coli TS, the quinazoline ring of compound (10) binds on top of the pyrimidine of the nucleotide, in a protein crevice surrounded by hydrophobic residues (Fig 10.6) The bound molecule bends at right angles between the quinazoline and 4-aminobenzoyl rings (at NlO), with the D-glutamate portion extending out to the surface of the enzyme Hydrogen bonds are made with several enzyme sidechains, the terminal carboxylate, and several tightly bound waters This compound, like folate and most folate analogs, gains entry into cells through a transport system that recognizes its D-glutamate moiety, and intracellular concentrations are elevated because of trap- (10) N10-propynyl-5,8-dideazafolate (also known as PDDF or CB3717) lure-Based Drug Design ping of'the compound as highly charged forms after a1ddition of several additional glutamates by a cellular enzvme " inhibitors were designed by Agouron TS scienti:sts with the aim of providing a drug that could enter cells passively and thus avoid the neted for transport or polyglutamylation The fil.st were designed by structure-guided modific:ation of known antifolates, and others were dlesigned de novo Starting with (12), the glutamlate moiety was deleted from the struc(12), the 2-desamino-2ture [Compound I methyl analog of (lo), had been found to be much more water soluble than (10) This eventually led (65) to AstraZeneca's Tomudex, which is now approved for treatment of colorectal cancer in European markets.] Removal of the glutamate reduced the potency by to orders of magnitude (Table 10.1, 12 versus 13) The crystal structure solved by use of (10) indicated potential interactions that were exploited by substituents such as the m-CF, in compound (14) The phenyl moiety in (15)was added to interact with Phe176 and Ile79 (Fig 10.6) Combining substituents does not necessarily produce the expected sum of binding free energy (compare 16 with 14 and 15) Structures of the complexes with several of these compounds revealed that ideal place- ment of one group does not always accommo- - Figure 10.6 Binding site for (10) (N10-propynyl-5,8-dideazafolate), within the active site of thymirte synthase from Escherichia coli The surface of the inhibitor is shown in the left view The red dyla sphc?resin the left view are tightly bound water molecules See color insert Structure-Based Drug Design 428 Table 10.1 SAR for 2-Methyl-4-0x0-quinazoline Inhibitors of TS a Compound Kim CLM (E.coli TS) R Kim CLM (human TS) - (12) (13) (14) (15) (16) (17) para-CO-glutamate 0.005 H meta-trifluoromethyl para-SO,-phenyl meta-trifluoromethyl, para-SO2-phenyl para-SO2-(N-indolyl) 0.45 0.025 0.037 0.15 0.009 2.2 0.4 0.013 0.05 0.07 "From ref 60 date the best interaction for another (This is a general problem for rigid scaffolds.) Compounds (15-17) had significant activity in in vitro cell-based assays, which could be reversed by exogenous thymidine Compound (17) (AG85)was tested in human clinical trials for treatment of psoriasis (9) The structure shown in Fig 10.6 also suggested another approach to alter the structure of (12) to generate a lipophilic inhibitor of TS The hydrophobic cavity filled by the aromatic ring of the para-aminobenzoyl group could be filled instead by a substituent attached to position of the quinazoline nucleus Four different 5-substituted 2-methyl-4-oxoquinazolines were made to test this idea, and one of inhibitor of human TS these (18)was a (66) The X-ray structure of the bacterial enzyme with (18) confirmed the hypothetical binding mode Two dozen 5-substituted quinazolines were made to explore the SAR for this scaffold However, the eventual clinical candidate (19) was only two steps away from (18).The methyl group at position was incorporated for favorable interaction with Trp80 This also favorably restricted the torsional flexibility for the 5-substituent, and increased the inhibitory potency against human TS by 10-fold The 2-methyl was replaced by an amino group, to create a hydrogen bond to a backbone carbonyl in the protein, and increased potency another sixfold Compound (19) (AG337, also known as nolatrexed, and as the hydrochloride, Thymitaq) advanced into human testing and had progressed into laterstage clinical trials as an antitumor agent by 1996 (67) 2.3.2.2 De Novo Lead Generation: AG331 The de novo design effort was initiated through the use of a computational method, Goodford's GRID algorithm (68,69), to locate a site favorable to the binding of an aromatic system within the TS active site (70) Using computer graphics, naphthalene was visualized and manipulated within this favorable site (Fig 10.7) This facilitated alterations of the naphthalene scaffold to a benz[cd]indole to provide hydrogen-bonding groups to interact with the enzyme and a tightly bound water Elaboration from the opposite edge of the naphthalene core to extend into the top of the Structure-Based Drug Design active site cavity, toward bulk solvent, resulted in (20) The use of an amine for the groups attached to position of the benz- [cdlindole improved the synthetic ease for variation of these groups Compound (20) had value of p M for inhibition of human TS a Ki, and was about 10-fold less potent against the bacterial enzyme The X-ray structure of (20) bound to E coli TS revealed that the compound actually binds more deeply into the active-site crevice than had been anticipated Instead of interacting favorably with the enzyme-bound water indicated in Fig 10.7, the oxygen at position of the benz[cd]indole displaces it This forced the Ah263 carbonyl oxygen to move by about A Replacement of the oxygen at position with nitrogen provided a significant increase in inhibitory potency Structural studies revealed that this also resulted in recovery of the displaced water, and restoration of the original position of the Ah263 carbonyl oxygen The substituents at position 5, on the tertiary arnine nitrogen, and on the sulfonyl group were also varied during the iterative optimization process The process yielded (21) (AG331), which has a Ki,value of 12 n M for inhibition of human TS Compound (21) entered clinical trials as an antitumor agent (71) 2.3.3 Clycinamide Ribonucleotide Formyltransferase Glycinamde ribonucleotide formyl- transferase (GARFT) catalyzes the N-formyla- tion of glycinamide ribonucleotide, through use of N-10-formyltetrahydrofolate as the one-carbon donor Because this is an essential step in the synthesis of purine nucleotides, GARFT is a target for blocking the proliferation of malignant cells Several potent GARFT inhibitors, such as pemetrexed (22, ALIMTA, (22) pemetrexed (23) lometrexol LY231514) and lometrexol (23, 5,lO-dideaztetrahydrofolate, LY-264618), have been shown to be effective antitumor agents in clinical trials (71, 72) These were designed through traditional medicinal chemistry approaches, in which an- Structure-Based Drug Design Figure 10.7 Conceptual design of compound (201, by use of the active site of E coli TS as a template W represents a tightly bound water molecule [Adapted from Babine and Bender (91.1 dogs of folate were synthesized and then tested as inhibitors of tumor cell growth or of the activity of various folate-dependent enzymes (73-75) A recent paper reported the formation in situ of a potent bisubstrate analog inhibitor of GARFT, from glycinamide ribonucleotide and a folate analog, apparently catalyzed by the enzyme itself (76) The substrate analog was designed based on consideration of enzyme structure and the GARFT mechanism This emphasizes the potential to exploit the interplay between binding and catalytic events in the design of new inhibitors The development of GARFT inhibitors at Agouron began with consideration of the structure of the complex between the E coli enzyme and 5-deazatetrahydrofolate (77) An active and soluble fragment of a multifunctional human protein that contained the GARFT activity was provided by recombinant approaches (78), and its structure was also solved (79) in complex with novel inhibitors Comparison of the two structures subsequently validated the use of the bacterial enzyme as a model for the human GARFT The design of novel inhibitors also relied on previous studies of the structure-activity relationships (SAR) for substitutents around the core Structure-Rased Drug Design of (23),including some GARFT inhibitors in which the ring containing N5 was opened (80) Inspection of the structure of the bacterial GARFT-inhibitor complex revealed several important features The pyrimidine portion of the pteridine was fully buried within the GARFT active site, forming many hydrogen bonds with conserved enzymic groups The Dglutamate moiety was largely solvent exposed, with no immediately obvious potential for building additional interactions Retention of the D-glutamate unmodified was also desirable for pharmacodynamic reasons A significant opportunity was presented by the fact that the active site might accommodate a bulkier hydrophobic atom than the methylene group in 5-deazatetrahydrofolate that replaces the naturally occurring N5 in tetrahydrofolate To test this idea, a series of 5-thiapyrimidinones were synthesized, including compound (24) These analogs were more readily prepared than the corresponding cyclic derivatives This compound had a potency of 30-40 nM in both a cell-based antiproliferation assay and a biochemical assay for human GARFT inhibition A crystal structure of human GARFT, complexed with (24) and glycinamide ribonucleotide, confirmed the structural homology between E coli and human enzymes Compounds with one fewer methylene in the linker connecting the thiophenyl moiety to the 5-thia position were much less active Several other analogs, such as (261, were made in attempts to fill the active site more fully, and to restrict the conformational flexibility of the linker Molecular mechanics calculations failed to correctly predict the conformation on the 5-thiamethylene group of (25) bound to GARFT because of unforeseen conformational flexibility of the enzyme revealed by an X-ray structure of this complex This again emphasizes the importance of interative experimental confirmation of molecular designs Several functional criteria in addition to GARFT inhibition and cell-based assays were evaluated during the several cycles of optimization These included the ability of exogenous purine to rescue cells (which indicates selective GARFT inhibition), and the ability of the inhibitors to function as substrates for enzymes involved in the transport and cellular accumulation of antifolate drugs Balancing these criteria has resulted in the choice of compounds (26) and (27) (AG2034 and AG2037, respectively) for clinical development at Pfizer (In 1999, Agouron Pharmaceuticals was acquired by Warner-Lambert, which was subsequently acquired by Pfizer.) It is as yet unclear whether the considerable toxicity of these and other GARFT inhibitors will allow these compounds to be acceptable as anticancer drugs Structure-Based Drug Design (26) X = H (27) X = methyl 2.4 Proteases 2.4.1 Angiotensin-Converting Enzyme and the Discovery of Captopril The design of cap- topril was a landmark in the application of structural models for developing enzyme inhibitors (81,82) This discovery rapidly led to the development of a family of therapeutically useful inhibitors of angiotensin-converting enzyme for the treatment of hypertension (83) The story has been reviewed thoroughly (for a historical perspective, see either Ref 84 or Ref 85), and is briefly summarized here Angiotensin 11, a circulating peptide with potent vasoconstriction activity, is generated by the C-terminal hydrolytic cleavage of a dipeptide from angiotensin I, catalyzed by angiotensin-converting enzyme Therefore, inhibitors of angiotensin-converting enzyme are vasodilators [An important aside: Angiotensin I is generated from a precursor by the action of renin, another exopeptidase that is an aspartyl protease An orally available renin inhibitor remains an elusive goal, although there are still efforts under way that use SBDD methods (86) Renin inhibitors were early tools in the study of the essential aspartyl protease of human immunodeficiency virus (HIV), which is discussed later.] 10.8) This model was based on the already known X-ray structure of bovine pancreatic carboxypeptidase A Both enzymes are C-termind exopeptidases that require zinc ion for activity, but differ in that carboxypeptidase A releases an amino acid, rather than a dipeptide Hence, the binding site for the angiotensin-converting enzyme was postulated to be longer, and to contain groups to interact with the central peptide linkage The suggestion had been made (87) that the inhibition of carboxypeptidase A by benzylsuccinate could be explained by viewing benzylsuccinate as a "byproduct analog" (Fig 10.8, top) The hypothesis was that one of the carboxylates bound into a cationic site, whereas the other interacted with the active site zinc If this were true, then a similar model for angiotensin-converting enzyme predicted that slightly longer diacids, designed with some regard for the sequence preferences of the converting enzyme, should inhibit that enzyme This hypothesis was quickly confirmed by the inhibitory activity of succinyl-proline (28a) Peptide sequences related to those of snake venom peptides had already been used to define the structural requirements for peptide inhibitors of angiotensin-converting enzyme Peptides are unstable in vivo and poorly ab- Asp-Arg-Val-Tyr-Ile-His-Pro-Phe-His-Leu+ Asp-Arg-Val-Tyr-Ile-His-Pro-Phe + His-Leu Angiotensin I A key tool in the discovery of captopril at Squibb was the use of a model for the active site of angiotensin-converting enzyme (Fig Angiotensin I1 sorbed intestinally, and thus are not good drug candidates However, the best peptide inhibitor was 500-fold more potent than (28a) The Structure-Based Drug Design 433 Yc substrate cleavage , 0- -N H L Angi infornlation provided by the peptides, the struct-ural model for the active site of angiotensin-converting enzyme, and biochemical and tissue-based pharmacological assays for the en zyme's function were used to guide an iterative design process to improve the potency, selectivity, and stability of small molecules inhibitors The R1 and R2 substitutents were optimized, and the zinc ligand was changc3d to a thiol, which significantly increased potency (Table 10.2, compare 28a with 18c) This process yielded the orally availald e and stable small molecule captopril (28d) within 18 months of the creation of the model, Thc following quotation [from the original research report (81) on the design of captopril] predicted the great promise of SBDD: "The studie;s described above exemplify the great heuristic value of an active-site model in the design of inhibitors, even when such a model is a hypc~theticalone." 2.4.,2 HIV Protease The aspartyl endoprotease e!ncodedby human immunodeficiency vi- rus (H:IV-P) catalyzes essential events in the Figure 10.8 Active site models for carboxypeptidase A (top) and angiotensinconverting enzyme (bottom) The design of the dipeptidyl derivative that led to the discovery of captopril is shown bound to the latter enzyme maturation of infective virus particles, the cleavage of polyprotein precursors to yield active products After this was demonstrated i n * the mid to late 1980s, HIV-P became a target for the development of antiviral drugs to treat acquired immunodeficiency syndrome (AIDS) Several HIV-P inhibitors have been approved for human therapeutic use in the past 10 years, and the speed with which they were developed is attributed in part to the successful use of SBDD methods There are excellent recent reviews of this area (88, 89) There are numerous reviews of the early work on HIV-P inhibitors (8,9, 90, 91) HIV-P is a symmetrical homodimer of identical 99 residue monomers, structurally and mechanistically similar to the pseudosymmetric pepsin family of proteases (92-941, whose members include renin Because the protease is a minor component of the virion particle, intensive structural studies required overproduction through recombinant DNA methods One of the first structures was determined with material synthesized nonbiologically (through peptide synthesis) As of June 2002, there were over 100 X-ray structures repre- Structure-Based Drug Design 434 Table 10.2 Key Compounds in the Development of Captopril Compound Structure (28a)(succinyl-proline) sented by coordinate sets in the Protein Data Bank, and many hundreds more have been determined in proprietary industrial studies The active site of the enzyme is C2 symmetric in the absence of substrates or inhibitors (Fig 10.9a),and contains two essential aspartic acid residues (Asp25 and Asp25') The entrance to the active site is partly occluded by "flaps" constructed of two beta strands (residues 43-49 and 52-58) from each monomer, connected by a turn In the absence of substrate or inhibitor, the flaps seem to be rather flexible Upon binding of inhibitors and presumably of substrates, the residues within the flaps undergo movements up to several angstroms to interact with the bound ligand (Fig 10.10) A single tightly bound water is observed in the structures of most HIV-P-inhibitor complexes, accepting hydrogen bonds from the backbone amides of both flap residues Ile50 and Ile50' and donating to carbonyls of the bound inhibitors This is referred to as the "flap" water Despite the presence of this water and several tightly bound water molecules on the floor of the active site, the cavity also contains extensive hydrophobic IC,, for inhibition of ACE ( p M l 330 surface area The minor differences between the HIV proteases from two major strains of HIV (HIV-1 and HIV-2) are not addressed here More significant are the HIV-P sequence variants with much reduced sensitivity to existing drugs that have evolved because of selective pressure and the rapid mutation rate of the virus The reader interested in the differences between the proteases from HIV-1 and HIV-2, or in the issues surrounding drug-resistant variants, is referred to Ref 91 and Ref 89, respectively The early work on inhibition of HIV-P was much influenced by previous structural and mechanistic work on pepsin and its inhibitors Both enzymes are thought to catalyze peptide hydrolysis through a tetrahedral transition state, shown below as (29).The previous work ucture-Based Drug Design Figure 10.9 (a) Residues in the active site of H N protease The C2 axis that relates the residues of the two monomers is indicated The carboxylates of Asp25 and Asp25' are the catalytic groups Not shown in this view are several flap residues (Ile47/Ile47', Ile501 Ile501),which move in to interact with inhibitors (b) Active site with bound (31) [saquinavir (PDB code 1HXB)I Note the asymmetry of inhibitor binding The flap water that is shown very close to saquinavir is labeled W See color insert ansition state mimics as pewin - - inhibitors the sequence of some cleavage sites for P led to the discovery at Roche of the R versions of (30)as submicromolar inhibof HIV-P, with the R enantiomer being ?fold more potent (95) These inhibitors oy a hydroxyethylamine moiety to re! the PI-P1' linkage that is normally red (the scissile bond) with a stable group lead molecules were optimized without dedge of the HIV-P crystal structure, to uce (31)(Ro 31-8959, saquinavir, Forto1 Cbz-Asn-N H J?? OH C02-t-Butyl (30) Saquinavir (31)was the first HIV-P inhibitor approved for human use Figure 10.9B ' Structure-Based Drug Design 436 Figure 10.10 Comparison of the structures of HIV-P apoenzyme monomer (top, PDB code 3PH.V) and the complex between HIV-P and (32) (U-85548; bottom, PDB code 8HVP) The inhibitor is shown as a ball and stick structure Note the rearrangement of the flap residues; Ile50 is indicated for reference The van der Wads surface of Asp25 is shown in both structures The flap water (red ball) is also shown between Ile50 and U-85548 In the bottom structure, the locations of theN and C termini of HIV-P are noted See color insert shows the asymmetrical binding mode of the molecule in the HIV-P active site Because the metabolic and pharmacokinetic characteristics of this compound and several other early HIV-P inhibitor drugs are less than ideal, the search for better ones has continued Many of the deficits arise from the large size and peptidic nature of the inhibitors Another early (31) saquinavir, Ro 31-8959 Structure-Based Drug Design Val-Ser -Gln-Asn-N inhibitor was the modified octapeptide (32, U-85548) developed at Upjohn (96) This subnanomolar inhibitor was used to define the extensive hydrophobic and hydrogen bonding interactions available in the HIV-P active site (97) A common feature in the binding of (31)and (32) to HIV-P is the interaction of the central hydroxyl group of the inhibitors with the carboxylates of both Asp25 and Asp25' This hydroxyl group replaces a water molecule that likely binds between these aspartyl side chains during peptide hydrolysis by HIV-P The inhibitors can therefore be seen as mimics of a "collected substrate." The liberation of this water to bulk solvent probably contributes about kcal mol-I to the free energy of inhibitor binding, based on the studies by Rich and his colleagues on similar inhibitors of pepsin (98,991 An interesting difference between (31)and (32) is that (31) has R stereochemistry at the hydroxymethyl center, whereas in (32) this is an S center Part of the reason for this is that when (31) binds to HIV-P, the decahydroquinoline ring system induces a conformational change in the protein, affecting primar- Ile-Val ily site S,' The optimal stereochemistry at the hydroxymethyl center appears to be whichever one will allow the interaction of the hydroxyl with both catalytic aspartates while accommodating the placement of inhibitor moieties in the S,, S,, S,', and S,' sites with minimal conformational strain on the inhibitor (9) Both (31)(Fig 10.9b) and (32) (Fig 10.11) bind to the HIV-P active site asymmetrically However, after the X-ray studies of crystalline HIV-P apoenzyme revealed it to be a symmetrical dimer, C2 symmetric inhibitors were designed to take advantage of this structural feature (Fig 10.12) Both alcohol diarnines and diol diamines were examined For example, the C2 symmetric compound (33) (A-77003) was synthesized at Abbott and entered clinical trials as an antiviral agent for intravenous treatment of AIDS (100) The X-ray structures of complexes between HIV-P and diol diamine derivatives like (33) showed (101) that, although one of the hydroxyl groups bound between the catalytic asparty1 carboxylates and made contacts with both, the second hydroxyl made only one such Structure-Based Drug Design 438 Figure 10.11 Orthogonal views of the complex between HIV-P and (32) (U-85548).The view in panel a is rotated approximately 90" (around the long axis of the protein) from the view in panel b Van der Wads surfaces of Asp25, Asp25', and the flap water (W)are shown In panel b, the solvent-accessible surface of the inhibitor is shown See color insert , diol diamine hydroxyethylene diamine Figure 10.12 Design principle for C2 symmetric inhibitors of HIV-P and the related hydroxyethylene diamine scaffold 2 Structure-Based Drug Design contact Thus the cost of desolvating the second inhibitor hydroxyl upon binding is not compensated by strongly favorable interactions in the complex (8) This led to the deletion of the second hydroxyl, as seen in compound (34), another compound in this program at Abbott Further structural modifications, to enhance solubility and metabolic stability, were guided by the fact that the "ends" of the protease-bound inhibitors were relatively solvent exposed and made fewer contacts with the enzyme (102) Deletion of a d i n e residue (33 34) gave a smaller compound, presumably aiding solubility and absorption The eventual product of this program was ritonavir (35,A-84538, ABT-538, or Norvir), which has been successfully launched Another C2 symmetric HIV-P inhibitor, discovered at Dupont Merck is compound (36) (DMP-450) This was one of a series of cyclic ureas designed to interact with both the aspartyl carboxylates and the Ile50 and Ile50' backbone amides that hydrogen bond with the flap water (103) The compounds interacted with HIV-P in a highly symmetrical fashion, as they had been designed to do, with the urea oxygen replacing the flap water Compound (36) was licensed to Triangle Pharmaceuticals, and the mesylate advanced into Phase I clinical trials Its future is uncertain after the trials were put on hold because of animal toxicity (http://www.tripharm.com/dmp45O.html) One of problems common to many of the HIV-P inhibitors already discussed is their (35) ritonavir + Structure-Based Drug Design (37) indinavir low solubility, which translates to low bioavailability The discovery of (37) (indinavir, L-735,524) was the result of the successful application of SBDD at Merck to directly address this problem During an iterative optimization process, the physicochemical properties of HIV-P inhibitors were modified within constraints that were established structurally (104) Crixivan (the sulfate of 37) was successfully launched for use as an antiviral drug The process leading to indinavir (Fig 10.13) began with (381, a hydroxyethylenecontaining heptapeptide mimic, originally designed as a renin inhibitor (105) The inhibiPhenyl boc tion of HIV-P by (38) was discovered by screening Classical medicinal chemistry methods allowed a reduction in size, and the discovery of an amino-2-hydroxyindan moiety to replace the terminal dipeptide (corresponding to P,', thought to bind into the s,' site) This approach (105, 106) resulted in the generation of (39)(L-685,434).Although (39) had a subnanomolar IC,, for inhibition of HIV-P, it also had very low aqueous solubility, like most peptidomimetics One way to improve solubility is to insert a charged functional group into the molecule The tertiary amino group in the HIV-P inhibitor saquinavir (31) Phenyl boc, OH OH - %Leu- , Phenyl / Phenyl (boc= tert-butyloxycarbonyl) 4" Phenyl \ (41 (cbz = benzyloxycarbonyl) Figure 10.13 Structures of HIV-P protease inhibitors during the optimization process leading to the discovery of (37) (indinavir) 2 Structure-Based Drug Design was already identified Piracy of the decahydroisoquinoline tert-butylamide from (31) provided the idea for the hybrid molecule (40) In addition to the charged group, use of this ring system would partly "preorder" the inhibitor's structure, lessening the entropic cost of binding Molecular modeling was used with known structures of HIV-P-inhibitor complexes to evaluate this idea, and it was judged to be reasonable enough to justify the synthesis of (40) (104) This compound was subsequently shown to have much better pharmacokinetic behavior than its antecedents, consistent with improved solubility and dissolution A convergent synthetic route was devised to generate (40) to improve the accessibility of important analogs Although (40) was an nM inhibitor of the isolated enzyme, better potency was needed for acceptable cell-based activity, and still better solubility characteristics were needed A method for structure-based computational estimation of the interaction energy for HIV protease inhibitors with the enzyme was developed and used to help estimate inhibitor potency before synthesis (107) Variation of the group contributing the tertiary m i n e led to the discovery of the piperazine derivative (41) (L-732,747), which had subnanomolar potency against HIV-P The Xray structure of the HIV-P complex with (41) confirmed the binding mode predicted by molecular modeling, with the molecule filling the S,, S,, S,', and S,' pockets, and the S, pocket occupied by the terminal benzyloxycarbonyl moiety Replacement of the benzyloxycarbony1 with more polar heterocycles, chosen to be accommodated by the S, pocket and to further improve aqueous solubility, yielded (37) Several other approved AIDS drugs that act by inhibition of HIV-P have also been developed through use of SBDD methods Compound (42) (amprenavir, Agenerase, also known as VX-478) is the most recent addition to the HIV-P inhibitors approved for human antiviral treatment, and differs significantly from earlier inhibitors Compound (42) was specifically designed by Vertex scientists to minimize molecular weight to increase oral bioavailability (108) Compound (43) (nelfina- , vir, AG-1343, also known as LY3128571, like the precursors to the earlier drug (37) (indinavir), copied the decahydroisoquinoline tert-butylamide group from the first marketed HIV-P inhibitor (31) (saquinavir) Compound (43) was developed in a collaboration between scientists at Lilly and Agouron (log), and is mar- (42) amprenavir Structure-Based Drug Design keted by Pfizer as Viracept, the mesylate salt of nelfinavir In both (42) and (43), the scientists involved used iterative SBDD methods to alter the physicochemical properties of the drug molecule while maintaining potency by optimizing interactions with the active site of the enzyme An important feature shared by these compounds is the fact that the bound inhibitors appear to be in low energy conformers, so that minimal conformational energy costs must be paid upon binding to the enzyme 2.4.3 Thrombin Thromboembolic diseases such as stroke and heart attack are major health problems, especially in many Western countries This has led to searches for drugs that are effective inhibitors of various serine endoproteases in the blood-clotting cascade, such as factor Xa and thrombin Existing therapeutic agents such as the coumarins (like warfarin), heparin, and hirudin have problems related to their absorption or unpredictable metabolism and clearance Recently, new small molecule inhibitors of thrombin have become available for human use in the United States, including (44) (argatroban, MD-805, developed by Mitsubishi) and (45) (melagatran, developed by AstraZeneca) (110, 111) These nanomolar inhibitors of human thrombin were optimized by classical medicinal chemistry, starting with peptidomimetics similar to the thrombin cleavage site in fibrinogen (see Fig 10.14a) Poor absorption by an oral route requires that they must be administered intravenously or at best subcutaneously At present, the only direct inhibitor of thrombin suitable for oral administration is ximelagatran, a prodrug form of melagatran in late development for various cardiovascular indications by AstraZeneca as of mid-2002 The therapeutic need and the availability of high quality crystal structures for human thrombin bound to inhibitors such as (44) make this an attractive target for SBDD (112) The significant efforts at Merck to use SBDD approaches to develop orally available inhibitors of thrombin, which have yielded compounds that have entered clinical trials, have been reviewed (113,114) For a good overview of this area, see the review by Babine and Bender (9) Compound (46) [NAPAP, N-alpha-(2- naphthylsulfonylglycy1)-4-amidinophenylalanine piperidide] is a moderately potent inhibitor of human thrombin, but was found to have an unacceptably short plasma half-life in animals (115) However, (46) has been a useful experimental tool and a variety of analogs have been made The structures of (44) and (46) bound to human thrombin show that they bind somewhat differently, as shown in Figure 10.14b (112,116) However, both form hydrogen bonds with the backbone at Gly216 (part of the oxyanion hole), and both fill the S, specificity pocket with a permanent cation attached to an extended hydrophobic group Compound (46) was the starting point at Boehringer Ingelheim for the development of the orally bioavailable prodrug (47) (BIBR1048) that generates in vivo a potent inhibitor of human thrombin (117) Compound (47) is currently in human clinical trials Scientists at Boehringer Ingelheim used the crystal structure of the complex between (46) and human thrombin to design a replacement for the central bridging glycine moiety The hypothesis that a trisubstituted benzimidazole could correctly place groups into the S,, S,, and S, pockets was confirmed The first such compound made was (48) The IC,, for thrombin inhibition by (48) was only 1.5 pM, but the compound had an improved serum half-life in rats Determination of the cryst'al structure of t h e thrombin-(48) complex showed that (48) binds in a similar fashion to (46) The N-methyl on the benzimidazole fit into the P, pocket, and the phenylsulfonyl group extended into S, The low affinity is likely attributable to the fact that (48) forms no hydrogen bonds with the backbone of Gly216 An iterative optimization process (Fig 10.15) was used to regain the lost affinity, eventually surpassing the thrombin affinity of the starting point (46) (0.2 Surprisingly, the N-methyl group could not be replaced with larger alkyl substituents, despite what appeared to be room for them in the P, pocket However, replacing phenyl with larger aryl groups such as naphthyl or quinolinyl on the sulfonamide provided favorable interactions in the P, pocket The crystal structure suggested that the increased lipophilicity of such aryl groups could be balanced by appending charged substituents to a) 2 Structure-Based Drug Design (44) argatroban (45) melagatran NH2+ (46) NAPAP Figure 10.14 (a) Sequence in fibrinogen at the thrombin cleavage site (top), and structures of several inhibitors of human thrombin the sulfonamide nitrogen Such substituents appeared likely to extend into solvent and therefore to be tolerated without compromising affinity This was confirmed (i.e., compound 491, and this decreased the undesirable affinity for serum-binding proteins X-ray studies with some of the inhibitors at this point indicated that a longer linker between the central benzimidazole and the benzamidine moiety in the S, pocket might provide some advantage This was confirmed with several analogs, with the methylamino linker pro- viding a 10-fold increase in potency (compound 50) By this point, the structural basis for interaction of this compound series with thrombin was understood sufficiently to suggest that the amidosulfonyl group could be replaced by a carboxamide This was confirmed by use of several compounds, such as (51) Compound (61)(BIBR 953) was quite active as an anticoagulant in animals dosed intravenously, but required conversion to prodrug (compound 47) to mask its charge and allow oral dosing Structure-Based Drug Design Figure 10.14 (b) Schematic comparison of the binding interactions for (44) and (46) in X-ray structural models of crystalline thrombin 2.4.4 Caspase-1 Caspase-1 (interleukin 1-pconverting enzyme, or ICE) is a member of a family of cysteine proteases that catalyze the cleavage of key signaling proteins in such processes as inflammatory response and apoptosis Genetic methods have provided evidence supporting a role for caspase-1 in diseases such as stroke (118) and inflammatory bowel disease (119) The X-ray structure of crystalline human caspase-1 was solved in 1994 by several groups (120,121),and has been a valuable tool in intensive efforts to design potent and bioavailable inhibitors of the enzyme Compound (52) (pralnacasan, VX-740) was developed as a caspase-1 inhibitory therapeutic agent through use of SBDD in a collaboration between Vertex and Aventis Although the details of the discovery process have not been published, (52) probably functions as a prodrug The cleavage of the lactone of (52) would yield a hemiacetal that could hydrolyze to release ethanol and the aldehyde form of the drug, which then can form a covalent thioacetal with the active site thiol of caspase-1, leading to pseudoirreversible inhibition Clinical trials of compound (52) as an anti-inflammatory agent for treatment of rheumatoid arthritis began in 1999 (122) In April 2002, the Structure-Based Drug Design n-hexyl / (47) BIBR 1048 tj (51) (BIBR 953) IC50 = 0.01 pM Figure 10.15 Optimization of structure leading to the discovery of (51) (BIBR 953) companies announced that these trials would continue and would be expanded to include treatment of osteoarthritis 2.4.5 Matrix Metalloproteases Matrix metal- loproteases (MMPs) are a large and diverse family of zinc endoproteases Several members of this family (such as the collagenases and the stromelysins) are thought to have important roles in proliferative diseases, including arthritis, retinopathy, and metastatic in- vasiveness of tumor cells There are publicly available X-ray structures of enzyme-inhibitor complexes for at least seven different MMPs, as of this writing Several detailed reviews of the SAR and binding modes for inhibitors of matrix metalloproteases are available (9, 123) All MMP inhibitors contain a moiety that binds to the active site zinc, such as the hydroxamates of (53) (prinomastat, AG3340) and (54) (CGS-27023) and the carboxylic acid of (55) (tanomastat, BAY 12-9566) These Structure-Based Drug Design (55) tanomastat, Bay 12-9566 (52) pralnacasan (53) prinomastat, AG3340 (54) CGS 27023 compounds each have affinities in the nanomolar to picomolar range for several MMPs The inhibitory profiles and ongoing clinical trials of a variety of drug candidates that inhibit MMPs were reviewed in 2000 (124) Compound (53)was developed at Agouron through use of SBDD (125) and is under clinical investigation by Pfizer as an anticancer drug and as a treatment of proliferative retinopathy Compound (54) is a stromelysin inhibitor discovered at Novartis (1261, without explicit structural guidance However, the lead molecule from which (54) was developed was originally obtained by X-ray structure- based inhibitor design targeted against the bacterial zinc-protease thermolysin Compound (55),with particularly high affinity for the gelatinases, was also developed with consideration of the structures of other MMPinhibitor complexes, but not through use of iterative SBDD (127) The clinical trials of compounds (54) and (55) have been suspended because of their disappointing efficacy (124) It remains somewhat uncertain which MMP is responsible for specific diseases, and the possibility for biological redundancy suggests that inhibition of several MMPs may be required for treatment of some diseases SBDD clearly could have a major impact on the discovery of selective MMP inhibitors These could be useful tools in dissecting the disease relevancy of these targets, as well as providing the selectivity and bioavailability required of effective drugs 2.5 Oxidoreductases Oxidoreductases catalyze the oxidation or reduction of carbon-carbon, carbon-oxygen, or carbon-nitrogen bonds Frequently, nicotinamide cofactors are involved, with the oxidized and reduced forms (respectively, NADt or NADP+ and NADH or NADPH) receiving or donating the equivalent of a hydride during this process Nicotinamide-linked oxidoreductases that have been targeted for the discovery of new therapeutic agents include aromatase, dihydrofolate reductase (mentioned above), aldose reductase, and inosine monophosphate dehydrogenase SBDD methods have been successfully applied recently to the latter two enzymes to discover agents that are currently Structure-Based Drug Design in human testing The efforts with these two targets are described briefly below 2.5.1 lnosine Monophosphate Dehydrogenase Proliferative cells such as lymphocytes have high demands for the rapid supply of nucleotides to support DNA and RNA synthesis, as viruses during their proliferative phase The first dedicated step in the de novo biosynthesis of guanine nucleotides is conversion of inosinate to XMP, catalyzed by inosine monophosphate dehydrogenase (IMPDH) IMP + NAD+ +XMP + NADH A prodrug form of (56) (mycophenolicacid), a noncompetitive inhibitor of IMPDH, is approved for human therapeutic use as an im- sants Other utilities that have been suggested for IMPDH inhibitors are antiviral and anticancer therapies The structure of hamster IMPDH in complex with IMP and (56)was solved at Vertex in the mid-1990s (129) This allowed the visualization of a covalent intermediate, in which a cysteine thiol from the enzyme adds to C2 of the purine ring of the nucleotide substrate An analogous covalent adduct is postulated to be a key catalytic intermediate during normal turnover (130) The structure was a key tool in the discovery of (57) (VX-497,merimepodip),a novel potent inhibitor of human IMPDH suitable for oral administration (131) An experimental screen of a diverse library of commercially available compounds for inhibitors of IMPDH identified molecules with the phenyl, phenyloxazole urea scaffold (58) as weak inhibitors Through use of the compu- (56) mycophenolic acid munosuppressant (mycophenolate mofetil, CellCept).The use of this drug is hampered by gastrointestinal side effects probably related to the metabolism of the drug A second class of IMPDH inhibitors is represented by the nucleoside analog mizoribine (also known as bredinin), a prodrug approved for human use in Japan Such compounds competitively inhibit IMPDH in vivo after phosphorylation (128) These drugs validate the strategy of targeting IMPDH for the discovery of immunosuppres- tational program DOCK (1321, the initial inhibitors were built as models into the ex~erimental structure of the crystalline complex of IMPDH, IMP, and (56) Structural analogs were generated to improve potency in an iterative process, guided by the structural modeling and the observed changes in potency for inhibition of human IMPDH After this process yielded compound (59), with nanomolar potency, an X-ray structure H pRTNy / (O, N (57) merimepodip Structure-Based Drug Design was determined of (59) bound to the hamster enzyme with IMP This revealed both similarities and differences between the binding modes of (56) and (59) Aryl groups of both compounds pack against the covalently tethered purine of the nucleotide Several hydrogen bonding and hydrophobic interactions with the enzyme are also common between the two inhibitors However, there are several hydrophobic and van der Wads interactions seen in the complex with (59) that are not present with (56) Importantly, the urea moiety of (59) forms a network of hydrogen bonds with an aspartyl carboxylate that is not present in the complex with (56).Further modification of the structure was guided by the X-ray study by use of (59), to gain potency in a cell-based assay for inhibition of lymphocyte proliferation This provided compound (57), which Vertex has advanced into clinical trials for treatment of hepatitis C infections (60) tolrestat agents that target aldose reductase should not inhibit the closely related aldehyde reductase, an essential hepatic enzyme The structure of (60) and other inhibitors bound to porcine aldose reductase (136) provided a rich lode of information on the requirements for potent and selective inhibition of aldose reductase This was mined by scientists at the Institute for Diabetes Discovery, in a project that began in 1996 The Institute for Diabetes Discovery filed an IND application for (61) (lidorestat, IDD 676), a potent aldose 2.5.2 Aldose Reductase Aldose reductase has been implicated in many of the pathologies resulting from elevated tissue levels of glucose in diabetes mellitus (133, 134) This nicotinamide-dependent enzyme catalyzes the conversion of glucose to sorbitol, accumulation of which ultimately results in damage to the eyes, the nervous system, and the kidneys Given the enormous damage caused by this disease and the difficulty in regulating blood glucose, selective and potent inhibitors of human aldose reductase offer great potential benefit However, existing drugs that target aldose reductase have unreliable efficacy (135) For example, compound (60)(tolrestat) was withdrawn by Wyeth in 1996 because of poor clinical response Hence, there is still a need to provide an inhibitor of this enzyme that fulfills the potential in the clinic To minimize the risk of undesired toxicities, clinical (61) lidorestat reductase inhibitor, for treatment of diabetic complications, within 30 months of initiating the discovery project on this target The speed with which this was achieved appears in large part because of the use of SBDD methods The X-ray structures showed the cofactor NADP+ buried within the enzyme, with its C4 redox center exposed at the bottom of a deep hydrophobic cleft An anionic binding site is located near NADPf Several potent inhibi- Structure-Based Drug Design tors bind within the hydrophobic cleft and interact with the anionic site The binding of potent inhibitors induces a conformational change, opening an adjacent hydrophobic pocket The conformation induced by (60) differs from that caused by other, less selective inhibitors This "specificity" pocket was thought to offer an opportunity for selective inhibition of aldose reductase while sparing aldehyde reductase Hence, this structural study provided an initial pharmacophore for both potency and selectivity The SAR for this pharmacophore was developed with a series of synthetically accessible salicylic acid derivatives that were scored for potency and selectivity with the purified enzymes, and efficacy in a diabetic rat model (137) One of the most potent and selective of the derivatives was (62), containing the benz- geted in SBDD projects that have produced compounds that are either launched or in clinical trials 2.6.1 Acetylcholinesterase A pronounced decrease in the level of the neurotransmitter acetylcholine is one of the most pronounced changes in brain chemistry observed in the sufferers of Alzheimer's disease (139) Several drugs that are approved for the treatment of the dementia thought to result from this neurotransmitter deficit act by inhibiting acetylcholinesterase These include (63) (tacrine, or (63) tacrine (64) donepezil thiazole heterocycle The SAR was employed, guided by the structures of selected inhibitor complexes, to design a novel indole scaffold to present the pharmacophoric elements (M Van Zandt, personal communication) The optimization of this series provided the clinical candidate (61) (138) 2.6 Cognex, a Pfizer drug that was the first such agent approved for this indication), (64) (donezepil), and (65)(rivastigmine) Several other agents are in clinical trials Disappointing ef- Hydrolases Some other hydrolytic enzymes, in addition to proteases, that are important drug targets include protein phophatases, phosphodiesterases, nucleoside hydrolases, acetylhydolases, glycosylases, and phospholipases Structurebased inhibitor design is currently being applied to a number of these enzymes The last three mentioned have been successfully tar- (65) rivastigmine ficacy is observed with the existing drugs, arising from dose limitations that are likely attributable to the inhibition of acetylcholinesterase Structure-Based Drug Design in peripheral tissues (140) This may be a consequence of the high serum levels required to get these highly cationic molecules to penetrate the blood-brain barrier In a discovery project that is reminiscent of the discovery of captopril, scientists at Takeda created a hypothetical structure for the active site of acetylcholinesterase, based on SAR from previous biochemical and medicinal chemical work (141) The model consisted of (in addition to the serine protease-like catalytic machinery) an anionic binding site separating two discrete hydrophobic binding sites This model was then used to design inhibitors of the enzyme (reviewed in ref 142) One set of analogs examined were based on the N-(wphthalimidylalky1)-N-(w-phenylalky1)-amine (scaffold 66) An iterative process of testing, analysis, design, and synthesis, by use of this and closely related scaffolds, resulted in the production of (67) (TAK-147), which is cur- rently in clinical trials for treatment of the dementia resulting from Alzheimer's disease (142) The design of (66) was partially based on the structures of previously known inhibitors The two aryl substituents were intended to bind to the hydrophobic binding sites, placing the central m i n e cation into the anionic bind- ing site The length of both alkyl linkers was varied, and the effect of adding a third alkyl substituent was examined The phthalimide portion of the structure was chosen to improve the synthetic accessibility of the analogs needed for this exercise The compounds were tested not only for inhibitory potency toward rat cerebral acetylcholinesterase, but also for peripheral response and toxicity in dosed intact rats After the work was under way, Sussman and coworkers solved the atomic structure of acet~lcholinesterase from the electric eel, including complexes with several inhibitors, by X-ray crystallography (143) The availability of this structure made it possible to retrospectively analyze the basis for the S A R in this series of compounds, by use of DOCK (144) 2.6.2 Neuraminidase Influenza virus infections cause severe human suffering throughout the world and economic damage in the billions of dollars annually, although some years are worse than others In 1918 a pandemic caused by this disease killed an estimated 40 million people (145) An important protein in the infectious process is the viral neuraminidase, an integral membrane protein whose catalytic domain is exposed on the viral surface Neuraminidase catalyzes the hydrplytic cleavage of sialic acid (68, N-acetylneuraminic acid) from glycoproteins and extracellular mucin on the surface of the host cell A different viral surface protein tightly binds to terminal sialic acid residues which ~romotes the initial infection, but prevents release of viral progeny from the host cells, unless and until the terminal sialic acids are hydrolytically cleaved by viral neuraminidase Thus, neuraminidase enables the infection to propagate The first X-ray structure of influenza neuraminidase was determined in the early 1980s (146).Ten years later, a landmark paper (147) described a highly efficient drug design project at Monash University in Australia This project yielded antiviral compound (69) (zanamivir, Relenza, or Flunet), which was developed into one of the first drugs to be created through use of SBDD Previous structural work had revealed that the active site of neuraminidase has several rigid pockets and nu- i& Structure-Based Drug Design (68) sialic acid (70) Neu5Ac2en, DANA neuraminidase active site In the case of the guanidine substitution, the binding affinity merous charged groups Electrostatic interactions significantly affect the conformation of for neuraminidase was increased about 5000bound sialic acid, which is deformed into a fold and provided (69), which inhibits viral release in cell cultures and decreases the severhigh energy conformer, attributed in part to ity of influenza virus infections in humans the interactions between the 1-carboxylate and arginine side-chains of the protein This Subsequently, the X-ray structures of neuraminidases from several different influenza deformation may play a key role in catalysis subtypes complexed with (69) were analyzed Synthesis of a sialic acid analog that is dehy(149) Although the positions of protein residrated across the C2-C6 bond of (68)had produes were well conserved, the water structure vided the putative transition state mimic (70) seen in these different complexes was quite (sometimes referred to as Neu5Ac2en, or as 2-deoxy-2,3-dehydro-N-acetylneuraminic variable This may explain the varying poacid, DANA) tency of (69) against different strains of virus One problem with (69) is that it is not well Compound (70) inhibits neuraminidase with micromolar potency (148) Examination absorbed by an oral route, and so must be administered as an aerosolized powder inhaled of the binding mode of (70) in the active site of into the virus-infected lungs Two other neurneuraminidase (Fig 10.16) led to the replaceaminidase inhibitors with nanomolar affiniment of the 4-hydroxyl by cationic groups, ties (71 and 72) have been developed through first an amino and then a guanidino group (147) These groups strongly interact with anthe use of SBDD methods to yield orally bioionic amino acid side chains (corresponding to available drugs The development of these Glu120 and Glu229 shown in Fig 10.16) in the agents was facilitated by the fortuitous discov- Structure-Based Drug Design Figure 10.16 View from above: Polar amino acid side-chains surrounding ('701,bound into the active site of influenza virus neuraminidase (Scheme 10.1 based on PDB code lNNB, the coordinates of an X-ray structure described in Ref 148) ery by scientists at Biocryst, that analogs of (69) in which the cyclic scaffold is a phenyl moiety are much more potent inhibitors if they lack the glycerol side chain! This was subsequently discovered by X-ray structural studies to be attributed to the creation of an unanticipated hydrophobic pocket upon rearrangement of the Glu278 side chain carboxylic acid, which forms several hydrogen bonds with the glycerol portion of (69) (Fig 10.16) Replacement of the permanently cationic guanidine by an m i n e (71) promoted better intestinal absorption, but also greatly decreased the affinity for neuraminidase Structure-guided modification of the carbocycle's substituents was used to recover this lost potency Compound (71) (GS 4071) was developed by Gilead Sciences (150) The ethyl ester of (71) is a prodrug (oseltamivir or GS 4104) that has been approved for oral dosing to treat influenza infection Another amphiphilic carbocycle, compound (72) (peramivir, RWJ270201, or BCX 1812) was developed by BioCryst (151) through use of SBDD, and is in clinical trials The use of clever synthetic routes, biochemical assays for neuraminidase inhibition, a mouse infection model, and X-ray structural information were all valuable tools in the development of both (71) and (72) Optimization of the affmity required the examination of avariety of alkyl substituents in bdth cases, to exploit the new hydrophobic pocket created by the conformational change primarily involving Glu278 The ability of the cyclopentyl ring in (72) to replace the six-membered ring illustrates that differing central scaffolds can display the essential interacting groups in an effective way 2.6.3 Phospholipase A2 (Nonpancreatic, Secretory) Phospholipases A2 (PLA2s) are a di- (72) BCX 1812 verse family of hydrolases that cleave the sn-2 ester bond of phospholipids The fatty acid produced is frequently arachidonate, the precursor to the proinflammatory eicosanoids In several human inflammatory pathologies (e.g., septic shock, rheumatoid arthritis), a nonpancreatic secretory form of PLA2 (hnpsPLA2) is present in extracellular fluids at levels many-fold higher than normal (152) The design of bioavailable inhibitors of this Ca2+dependent isoform of PLA2 as inflammatory ! Structure-Based Drug Design drugs is therefore an attractive goal (153) To be an effective drug, such an inhibitor would also need to be selective for hnps-PLA2 vs the closely related pancreatic PLA2 Whether selectivity is needed against the quite different cytosolic PLA2 is unclear Investigators at an AstraZeneca laboratory (previously Fisons) have used multidimensional NMR and computational techniques to develop an active site model for cytosolic PLA2 (154, 155) Synthesis of compounds based on this model led to (73) (FPL-67047), reported - (75) indomethacin The crystal structures of recombinant hnps-PLA2 bound to (74) and (75) were solved (158), and compared with the previously known structures of PLA2s complexed with substrate mimics (159, 1601, including the phosphonate-containing transition state analog (76) The earlier structures revealed sev- to be a development candidate for treatment of inflammation (156) Investigators at Eli Lilly began a project to develop PLA2 inhibitors by investing the effort to clone, overproduce, purify, crystallize, and determine the structure of hnps-PLA2 (157) This also provided the reagent needed for a massive screening campaign to identify hnps-PLA2 inhibitors They were thus prepared to apply SBDD methods when the screening of Lilly's small molecule collection yielded a weak inhibitor The hit (74) was sur- prisingly similar to indomethacin (751, a nonsteroidal anti-inflammatory drug that acts by inhibiting cyclooxygenase (76) hnps-PLA2 transition-state analog era1 key features These were: (1)the filling of a significant hydrophobic crevice, (2) the displacement (by the sn-2 alkyl moiety) of the His6 side-chain into a solvent-exposed position to create an adjacent cavity, (3) the coordination of the active site calcium, and (4) formation of hydrogen bonds to His48 and Lys69 The polar contacts were provided by the nonbridging phosphate and phosphonate oxygens in the complex with (76) The screening hit (74) bound in the hydrophobic crevice, similarly to the substrate mimics, with the 1-benzyl moiety of (74) bound in the adjacent cavity and displacing the enzvme's His6 imidazole However, there were " two surprising findings First, despite the presence of 10 mM calcium in the crystallization liquor, there was no bound calcium, an essential active-site component, although weakly binding (K,= 1.5 rnM) Second, the carboxylic acid of (74) formed a hydrogen Structure-Based Drug Design bond with another active-site acid, the side chain of Asp49 The latter finding again emphasizes the importance of experimental structures to guide improvements of inhibitor potency, given that placing two presumed anions so close together would likely never have been predicted by a computational model Other slight conformational changes were observed to accommodate the 5-methoxy group of (74) The inhibitor's 3-acetate moiety was converted to an acetamide in a successful attempt to restore the active site calcium, form a hydrogen bond to His48, and increase potency The crystal structure of the complex with the amide version of (74) also revealed a significant reorientation of the indole core and 5-methoxy substituent, resulting in an unanticipated 5-A movement of the terminal methyl Further changes in inhibitor structure were guided by iterative structural studies and functional assays of potency and selectivity These changes involved the use of substituents at positions or to optimize coordination of the metal ion, extension of the van der Wads interaction by lengthening the 2-methyl to an ethyl, and conversion of the 3-acetamide to glyoxamide (159,161) This resulted in the synthesis of (77) (compound LY315920), which has 6500-fold greater ity for hnps-PLA2 than did the original hit molecule (74) LY315920 effectively inhibits hnps-PLA2 in the serum of transgenic mice dosed with the compound orally or i.v., and is undergoing clinical trials in the United States and Japan (162,163) 2.7 Picornavirus Uncoating Picornaviruses, which include the rhinoviruses and enteroviruses are RNA viruses that cause several infectious human diseases These diseases include common colds as well as life-threatening infections of the respiratory and central nervous systems Effective treatments of these diseases would relieve much human suffering, save many lives, and have great economic benefit There are over 100 serotypes of rhinoviruses alone, making it impossible to generate a vaccine effective against infections by all variants of the virus (164) The Achilles heel of ~icornaviruseshas been suggested to be that part of the virus structure that interacts with the cell surface receptor because those structural features must be well conserved (165) The virus particle consists of a positive-strand RNA coated by an icosahedral shell, containing 60 copies of four distinct 0-barrel proteins (166) Thege structural proteins contain the binding site for the cellular receptor and undergo significant conformational changes to liberate the viral RNA genome during infection of the cell A series of isoxazoles that inhibit this picomavirus "uncoating" process were discovered in the early 1980s by scientists at Sterling Winthrop, by use of an in vitro cellular assay for antiviral activity (167-170) One of these, compound (78) (WIN-51711, disoxaril), gave a 50% suppression of viral plaque formation in this assay at 0.3 $ Compound (78) was also effective in animal models (171) and entered phase I clinical trials, but failed to advance because of its toxicity Compound (78) was shown (172) to bind to viral capsid protein (78) WIN-51711,disoxaril ure-Based Drug Design Figure 10.17 Structure of rhinovirus capsid protein VP1 showing the bound conformation of antiviral isoxazole compounds (78) [disoxaril, WIN51711: panel a, top], (79) W N 54954: panel b, middle], and (80) [pleconaril, WIN-63843: panel c, bottom] The PDB codes for the X-ray structural model coordinates used to create these views are: lPIV (for 781, 2HWE (for 79), and 1C8M (for 80) On the left side of each panel, the inhibitors are shown as van der Wads surfaces, and the protein as a ribbon diagram On the right side, the structures of the inhibitor alone are shown, from the same view, as ball and stick representations See color insert ithin a hydrophobic pocket in the floor canyon" that contains the binding site cell surface receptor (Fig 10.17A) ral changes induced in the canyon on binding of such molecules may also receptor binding directly (173) X-ray ographic studies of (78) and analogs ;o the target protein VP1 were an espart of the iterative optimization pro~tled to safer and more effective antiznts (174-176) The goal of the process generate a compound that is potent, dly and metabolically stable, and effecinst as many serotypes of the virus as ! There was therefore a need to bal- ance potency and selectivity, and the structural information helped to guide compound design in pursuit of this balance A second-generation compound, (79) (WIN-54954) also advanced into clinical tests, but had disappointing efficacy in Phase I1 trials, probably because of extensive metabolism Modification of the phenylisoxazole, guided by both structural and metabolic considerations (177), allowed the creation of a stable and potent antiviral, the third-generation compound (80)(WIN-63843, pleconaril, or Picovir) (178) This compound was evaluated in Phase I11 clinical trials and showed efficacy in humans Oral dosing of virally infected patients with * Structure-Based Drug Design (80) VP63843, pleconaril (80) three times daily decreased the average time needed to become free of cold symptoms from 10 days to between and days, and also reduced the duration of severe cold symptoms from 4.5 to 3.5 days (179) During the clinical studies to support the new drug application for (80)' about a quarter of the clinical isolates (of rhinovirus present initially or during the treatment) were resistant to the compound The majority of these resistant viruses had a single mutation at VP1 residue Ile98, which directly interacts with (80) bound to VP1 in wild-type virus The clinical data also showed the elevation in some patients of hepatic cytochrome P450 levels during treatment with (go), raising concerns about potentially hazardous drug-drug interactions ViroPharma sought and failed in early 2002 to gain the approval of the U.S Food and Drug Administration for its new drug application for (80)for treatment of the common cold 2.8 Phosphoryl Transferases Protein kinases and phosphatases play vital roles in intracellular signaling pathways and in the integration and control of major cellular processes Kinases and other phosphoryl group transferases are essential in the metabolism of lipids, nucleotides, and other small biomolecules The use of SBDD methods on such targets has expanded as more of their X-ray structures have been solved, and will continue to grow as more targets are validated for their involvement in human diseases 2.8.1 Mitogen-Activated Protein Kinase p38a Mitogen-activated protein kinase (MAPK) p38a is a member of a family of SerIThr-specific protein kinases that are activated upon exposure of cells to mitogens such as bacterial lipopolysaccharide or environmental stresses such as exposure to W irradiation or chemical oxidants MAPK p38a has a central role in integrating the inputs from a complex signaling network Activation of MAPK p38a requires the dual phosphorylation of conserved threonine and tyrosine residues on a loop near the enzyme's active site (180) The unactivated (nonphosphorylated) enzyme has a very low affinity for ATP, but can bind to pyridinylimidazole inhibitors (181, 182) The activated enzyme in turn phosphorylates numerous substrates, including several transcription factors This leads to activation of the transcription of many genes and causes the release of proinflammatory cytokines, primarily interleukin-lp (IL-1p) and tumor necrosis factor (TNFa) MAPK p38a was identified as a central player in this inflammatory pathway in a key study by scientists at SmithKline Beecham (183) The study involved the molecular cloning of the genes encoding proteins that bind to anti-inflammatory pyridinyl-imidazole compounds already known to block the biosynthesis of IL-1p and TNFa The binding proteins turned out to be members of a known kinase family Since this finding, the enzymes in the MAPK pathway, and especially MAPK p38a, have been attacked by many scientists seeking to discover anti-inflammatory drugs (184) Compound (81) (SB 2035801, a specific inhibitor of MAPK p38a, is a prototype for the pyridinyl-imidazole compounds (185) This compound is active in animal models of several inflammatory diseases (186),but was not itself pursued as a clinical candidate because of its inhibition of other enzymes, including hepatic cytochrome P450 reductases The pyridinylimidazole compounds have dissociation constants for MAPK p38a in the nanomolar range, competing with ATP for binding to the enzyme Because these compounds bind tightly to the unactivated enzyme, which has a Structure-Based Drug Design low affinity for ATP, they are able to compete effectively even i n vivo, where the ATP con:entration is in the millimolar range The Xray structures of (81)and several other pyridinyl-imidazole compounds in complexes with hman MAPK p38a were solved in a collabo.ative effort between scientists at SmithKline 3eecham and the University of Texas (187) several X-ray structures of human MAPK 13thwith and without bound inhibitors have dso been solved by scientists at Vertex (181, 88) The structures of the inhibited enzyme vere useful in understanding what parts of he compounds were responsible for strong binding to MAPK p38a As shown in Fig 0.18, both hydrophobic and hydrogen bondng interactions are important components of the inhibitor binding pocket This structure suggested that ThrlO6 is an important structural determinant of the selective inhibition of MAPK p38a by the pyrimidyl-imidazoles, which have low affinity for other closely related kinases Mutation of ThrlO6 results in the loss of sensitivity to these inhibitors, whereas the replacement of the corresponding residue in another kinase (ERK2) by threonine caused the mutated variant to become sensitive to these inhibitors (189, 190) The X-ray structural models were also used at both SmithKline Beecham (later GlaxoSmithKline) and Vertex to guide the design of new inhibitors For example, both N1 of the central imidazole and the 2-(para-methylsulfony1)-phenyl substituent in enzymebound (81)face a channel that opens to bulk solvent This observation led to the design of (82) (VK19911) at Vertex (181) and (83) (SB242253) at GlaxoSmithKline (191) Compound (83)is fivefold more potent than (81)in vivo, in a mouse disease model, and was advanced into human clinical trials for treatment of rheumatoid arthritis (192) The piperidine on N1 of (82) and (83)was designed to form a salt bridge with Asp168 This interaction, and the preservation of other binding interactions, was directly demonstrated (181) for compound (82) Analysis of the structural information from the X-ray models allowed the design at Vertex of a new scaffold for potent inhibition of MAPK p38a, as shown for compound (84) (VX-745) This design process, SBDD through Figure 10.18 Binding of SB203580 (shown a s a ball and stick structure) in the active site of W K p38a In addition to the side chains of the labeled residues, the protein backbone between Leu104 and Met109 is shown, as well a s several aliphatic side chains and a water molecule (red sphere) Hydrogen bonds (dotted lines) are shown between the backbone amide of Met109 and the inhibitor's pyrirnidinyl nitrogen, and between the €-aminoof Lys53 and the inhibitor's imidazole N3 This figure is based on the PDB coordinate set 1A9U (187) See color insert Structure-Based Drug Design use of a crystal structure of MAPK p38a to design potent inhibitors with potential utility as human therapeutics, is the subject of an international patent application by Vertex, published in 2000 (193) The binding mode for (84) has not been disclosed, but the compound was advanced into clinical trials (194) Vertex has since discontinued the clinical trials of (84) because of the potential for toxicity, based on animal data, but in mid-2002 Vertex began a phase I clinical study of a new compound targeted against MAPK p38a Scientists at Boehringer Ingelheim recently described (195, 196) their discovery of an orally active inhibitor of MAPK p38a, compound (85) (BIRB-7961, that is very different from earlier inhibitors This compound, whose K, for MAPK p38a is 100 picomolar, has entered phase I1 clinical trials for treatment of rheumatoid arthritis The lead compound that led to compound (85) was a diary1 urea originally identified by high throughput screening X-ray structural studies revealed novel modes of binding for both the lead compound and (85) in the active site of MAPK p38a Their binding sites are adjacent to the active site but not directly overlap with that of ATP; rather, their binding mode changes the conformation of MAPK p38a such that ATP cannot bind The optimization of the lead compound to clinical candidate (85) was an iterative process using clever synthetic chemical design, biochemical assays for affinity, Xray crystallographic studies of key complexes, and cell-based and animal models The development of (85)as a MAPK p38a inhibitor with Structure-Based Drug Design efficacy in vivo makes it evident that there are multiple ways to effectively inhibit this enzyme 2.8.2 Purine Nucleoside Phosphorylase Purine nucleoside phosphorylase (PNP) catalyzes the reversible phosphorolysis of purine nucleosides to the purine base and ribosyl or 2-deoxyribosyl-a-1-phosphate The vital role of PNP in the proliferation of T-cells is evident from the fact that people with an inherited deficit in this activity have 30- to 100-fold lower numbers of T-lymphocytes than normal (197) The accumulation of dGTP and the resulting inhibition of ribonucleotide reductase in PNP-deficient T-cells causes the suppression of T-cell proliferation B-lymphocytes are unaffected Hence, small molecule inhibitors of PNP could be used to treat T-cell lymphomas and other T-cell-mediated diseases such as psoriasis Adjunct therapy with PNP inhibitors could also block the catabolism of therapeutically useful nucleoside analogs Human PNP is a homotrimer of 32-kDa subunits The X-ray structures of the apoenzyme and some substrate analog complexes were described in 1990 Each of the three identical active sites, located near the subunit interfaces, are composed primarily of residues from one subunit, with Phe159 participating in the active site of the adjacent subunit (198) Scientists at BioCryst, CIBA-Geigy, Southern Research Institute, and the University of Alabama collaborated to design inhibitors of human PNP (199,200) The project used an iterative process, in which new compound design was guided by synthetic considerations, computer graphics analysis of X-ray structural models, computational (Monte Carlo and energy minimization) methods, and the inhibitory potency of the compounds against PNP in vitro Evaluation of the most potent inhibitors by use of cell-based assays, followed by pharmacokinetic and pharmacological characterization of several inhibitors in animal models, led to the choice of (86)for advancement into clinical trials Compound (86) (BCX-34, peldesine) is being evaluated for treatment of psoriasis and skin cancer (201,202) (86) peldesine + PNP Structure-Based Drug Design In the SBDD project that produced (86), the design work was initiated through use of the X-ray structure of the PNP apoenzyme, but was more successful when the structures of the PNP-guanine complex and other complexes were available (199) The PNP-guanine crystal structure showed no important interactions with N9, and indicated a potential for hydrophobic interaction in the vicinity of the substrate ribose (Fig 10.19) To test this, the 9-deaza compound (87) was synthe- sized This was a weak PNP inhibitor (measured IC,, l fl The X-ray structure of the complex between PNP and (87) showed that the hydrophobic interaction dominated the binding mode, and resulted in the disruption of the hydrogen bonding interactions seen in the guanine complex (i.e., Fig 10.19) To increase the spacing between the hydrophobe and the purine mimetic, compounds (88) and (89) - were made These had affinitiesfor PNP in the nanomolar range X-ray crystallographicanalysis indicated new hydrogen bonding interactions with these 9-deaza compounds (shown for 89 in Fig 10.20),made possible because N7 is protonated Asn243 Lys244 NH3+ - H - Guanine Figure 10.19 Binding interactions in the active site for the complex between guanine and PNP Phe200 Phe156 Structure-Based Drug Design Figure 10.20 Binding interactions in the active site for the complex between PNP and (89) While this work was under way, a Phase I clinical trial was undertaken of PNP inhibitor (90) (PD-119229), which was developed by and superior solubility and pharmacokinetic properties, and so was advanced into human testing 2.9 other workers This led to an exploration of a series of 8-mino, 9-deaza derivatives, although the hydrogen bonding for the simpler 9-deaza compounds turned out to be superior (reviewed in Ref 202) It may be that compounds such as (90) suffer from unfavorable steric interactions between the 8-amino group and the proximal methyl group of Thr242 of the enzyme, or that the energetic cost of dehydrating the 8-amino group cannot be fully repaid by interactions with the enzyme Other chemical series were also explored, but compound (86) had an acceptable safety profile Conclusions and Lessons Learned The projects in which SBDD has been applied to enable the discovery of new drugs and clinical candidates have provided significant lessons for future investigators Some of these' lessons learned are summarized here Much of the credit for the summary presented here belongs to Michael Varney of Agouron, who provided a copy of a presentation that he made in 1998 to a medicinal chemistry symposium concerning the lessons learned in 10 years of use of SBDD methods Experience Matters In every aspect of SBDD, as in all technical fields, there is no substitute for experience Given the variety of different techniques that must be incorporated, this means that experience from several different people will be needed for optimal function of a discovery project team Essential expertise is needed in X-ray crystallographic studies, graphical display of experimental results, initial and iterative design of compounds and synthetic tactics, the creation of databases and database queries, and the analyses of search outputs and of the results of computational simulation experiments Structure-Based Drug Design Combine and Integrate Technologies Dedi- cated molecular biology and protein chemistry personnel and equipment are essential for identifying the right constructs for crystallization and to the assurance of a steady supply of protein Synthetic chemists trained in graphical analysis of protein structures tend to be excellent designers, and will be unlikely to design molecules that they cannot make Early tactical integration of the synthetic approaches is even more important if combinatorial chemistry is part of the program The structural information can be used to design combinatorial libraries as effectively as it can to design molecules one at a time The use of libraries can compensate for the inaccuracies inherent in current computational scoring algorithms More significantly, the integration of orthogonal technologies will stimulate creative thought and yield much more than the sum of the different technologies applied separately Go Big Early and Often Filling active site space as much as possible will maximize the chance that a compound will be a potent inhibitor During compound design, it should also be recognized that proteins are flexible, and that accessible conformations are hard to predict Sometimes, larger functionality can be accommodated than the existing structural model permits A few compounds should be included to probe this These may give rise to an unexpected boon, such as access to a significantly altered new protein conformation with novel sites that can be exploited in new rounds of design and synthesis Aqueous Solubility is Critical to Success Both early in SBDD and later on in clinical development, sufficient aqueous solubility is critical Solubility is important early because the concentrations of compounds must be high during crystallization experiments to saturate the high levels of protein The ratio of the solubility to the inhibition constant of a compound is also critical to the success of the crystallization experiment Once some structural information becomes available, both parameters can be manipulated, but usually, soluble inhibitors must be available before the availability of structural information Solubility matters during animal testing and later in development because compounds with very low solubility have limited or variable bioavailability Binding Sites Can Be Filled Many Ways More than one small molecule scaffold can provide the necessary and sufficient hydrophobic and polar complementarity to generate potent inhibition Sometimes, there are many scaffolds that will work However, the structures of complexes with all the different scaffolds will likely have common features that are distinct from the structure of the apo enzyme, attributed to large-scale conformational changes that occur upon binding any ligand The most useful X-rav " models to use for the design of new compounds will be those that already have some substrate or inhibitor bound There are several ways to design these compounds: modification of existing inhibitors, de novo creation of novel inhibitors, or some combination of these methods Not All Inhibitors Are Drugs Having the Xray structure of the targetprotein, or even having used the solved structure to design a potent inhibitor, is only the beginning of solving the difficult problems of drug design The use of structure to create ~ o t e ninhibitors t can certainly shorten the time to get compounds into human testing, but use of SBDD methods does not guarantee that a potent compound will become a drug This is an old lesson, actually, but is forgotten at great cost Structure of Free Inhibitor Is Important Desolvation of the free ligand and of the protein's active-site groups upon complex formation are both significant Both enthalpic and entropic contributions to the binding energy must be considered Particular attention should be paid to the advantage that can be gained from u preorganization" of the inhibitor before binding, that is, low energy conformers bind with greater apparent avidity Bound Water Is Special, But Not All Hydrogen Bonds Are Created Equal Each of the tightly bound waters present in an X-ray structural model has a uniaue environment and a unique function In some cases, liberation of a bound water molecule by displacing it with an inhibitor's functionality can greatly increase inhibitor affinity, although this is nit globally applicable The entropic advantage of releasing a bound water into bulk solvent does eferences not always exceed the enthalpic cost of the displacement In many situations, the preferred solution will be to retain a water molecule and use it to maximize inhibitor binding For example, a water molecule that donates two hydrogen bonds and accepts one cannot be isosterically replaced Electrostatic interactions that are more complex than hydrogen bonds and simple ion pairs are very difficult to model, anticipate, and exploit in inhibitor design Retain Potency While Addressing Other Issues Structural information can be very use- ful in designing compounds that are not part of a competitor's intellectual property, or that cannot be patented because of information in the public domain Redesign of a compound that is not itself proprietary, by use of structural information obtained with that compound, can yield valuable new proprietary molecules Structural information can also guide the modification of physicochemical, metabolic, or pharmacological properties or target selectivity without compromising the potency against the primary therapeutic target All Models Are Wrong; Some Are Useful At present, it is impossible to calculate an accurate value for a binding constant on an absolute scale However, accurately estimating the relative binding of a series of closely related compounds is possible, and is much more likely to be successful if X-ray structures of target complexes with some of the compounds are available Thus, although there is much room for improvement, local computational models can sometimes be quite useful Even in the absence of an experimentally determined X-ray structure of the target, a hypothetical model can be a powerful tool for the design of useful compounds (e.g., captopril and TAK-147) Iterative SBDD Cycles Are Optimal Small alterations in ligand structure often cause major changes in binding mode, protein conformation, or both These changes can go undetected if the structural effects are not analyzed by X-ray analysis iteratively or too infrequently This can yield confusing or misleading structure-activity relationships, leading to a waste of precious time Moreover, changes in compound structure seldom affect only one variable, so multiple orthogonal methods should be used to assess the effects of changes It is also important during the rational design process to include room for serendipity Do not reject an idea for a new compound that seems to make intuitive sense based on a single crystal structure or computational calculation REFERENCES D J Abraham, Intra-Sci Chem Rep., 8, (1974) http://www.agouron.com/ http://www.stromix.com/ http://www.astex-technology.com http://www.accelrys.com/consortia/htc/ P J Goodford, J Med Chem., 27,557 (1984) C R Beddell, Ed., The Design ofDrugs to Macromolecular Targets, John Wiley & Sons, Chichester, UK, 1992 J Greer, J W Erickson, J J Baldwin, and M D Varney, J Med Chem., 37, 1035-1054 (1994) R E Babine and S L Bender, Chem Rev., 97, 1359 (1997) 10 P Veerapandian, Ed., Structure-Based Drug Design, Marcel Dekker, New York, 1997 11 R T Borchardt, R M Freidinger, T K Sawyer, and P L Smith, Eds., Integration ofPharmaceutical Discovery and Development Case Histories (Pharmaceutical Biotechnology, Band l l ) , Plenum Press, New York, 1998 12 K Gubernator and H.-J Bohm, Eds., Structure-Based Ligand Design, Wiley-VCH, New YorWWeinheim, 1998 13 C L Nobbs, H C Watson, and J C Kendrew, Nature, 209,339 (1966) 14 M F Perutz, Nature, 228, 726 (1970) 15 R C Ladner, E J Heidner, and M F Perutz, J Mol Biol., 114,385 (1977) 16 G Fermi, M F Perutz, B Shaman, and R Fourme, J Mol Biol., 175, 159 (1984) 17 B C Wishner, K B Ward, E E Lattman, and W E Love, J Mol Biol., 98, 179 (1975) 18 D J Harrington, K Adachi, and W E Royer Jr., J Mol Biol., 272, 398 (1997) 19 C R Beddell, P J Goodford, G Kneen, R D White, S Wilkinson, and R Wootton, Br J Pharmacol., 82,397 (1984) 20 M Merrett, D K Stammers, R D White, R Wootton, and G Kneen, Biochem J.,239,387 (1986) Structure-Based Drug Design 21 F C Wireko and D J Abraham, Proc Natl Acad Sci USA, 88,2209 (1991) 22 D J Abraham, A S Mehanna, F C Wireko, E P Orringer, J Whitney, and R P Thomas, Blood, 77, 1334 (1991) 23 M K Safo, S Nokuri, and D J Abraham, Unpublished results 24 P E Kennedy, F L Williams, and D J Abraham, J Med Chem., 27, 103 (1984) 25 D J Abraham, M F Perutz, and S E V Phillips, Proc Natl Acad Sci USA, 80,324 (1983) 26 M F Perutz, G Fermi, D J Abraham, C Poyart, and E Bursaux, J Am Chem Soc., 108, 1064 (1986) 27 E P Orringer, D S Blythe, J A Whitney, S Brockenbrough, and D J Abraham, Am J Hematol., 39, 39 (1992) 28 D J Abraham, A S Mehanna, F Williams, E J Cragoe Jr., and W Woltersdorf Jr., J Med Chem., 32,2460 (1989) 29 D J Abraham, P E Kennedy, A S Mehanna, D Patwa, and F L Williams, J Med Chem., 27,967 (1984) 30 A Arnone, Nature, 237, 146 (1972) 31 V Richard, G G Dodson, and Y Mauguen, J Mol Biol., 233,270 (1993) 32 P J Goodford, J St-Louis, and R Wootton, Br J Pharmacol., 68, 741 (1980) 33 C R Beddell, P J Goodford, F E Norrington, S Wilkinson, and R Wootton, Br J Pharmacol., 57,201 (1976) 34 F F Brown and P J Goodford, Br J Pharmacol., 60,337 (1977) 35 A S Mehanna and D J Abraham, Biochemistry, 29,3944 (1990) 36 M F Perutz and C Poyart, Lancet, 2, 881 (1983) 37 I Lalezari and P Lalezari, J Med Chem., 32, 2352 (1989) 38 I Lalezari, P Lalezari, C Poyart, M Marden, J Kister, B Bohn, G Fermi, and M F Perutz, Biochemistry, 29, 1515 (1990) 39 D J Abraham, R S Randad, M A Mahran, and A S Mehanna, J Med Chem., 34, 752 (1991) 40 D J Abraham, F C Wireko, R S Randad, C Poyart, J Kister, B Bohn, J F Leard, and M P Kunert, Biochemistry, 31,9141 (1992) 41 F C Wireko, G E Kellogg, and D J Abraham, J Med Chem., 34,758 (1991) 42 D J Abraham, J Kister, G S Joshi, M C Marden, and C Poyart, J Mol Biol., 248,845 (1995) 43 M K Safo, C M Moure, J C Burnett, G S Joshi, and D J Abraham, Protein Sci., 10,951 (2001) 44 S K Burley and G A Petsko, FEBS Lett., 201, 751 (1986) 45 S K Burley and G A Petsko, Science, 229,23 (1985) 46 M Levitt and M F Perutz, J Mol Biol., 201, 751 (1988) 47 D J Abraham, M K Safo, T Boyiri, R E Danso-Danquah, J Kister, and C Poyart, Biochemistry, 34,15006 (1995) 48 M P Grella, R Danso-Danquah, M K Safo, G S Joshi, J Kister, S J Hoffman, M Marden, and D J Abraham, J Med Chem., 25, 4726 (2001) 49 A.M Youssef, M K Safo, R Danso-Danquah, G S Joshi, J Kister, M Marden, and D J Abraham, J Med Chem., 45,1184 (2002) 50 J A Walder, R H Zaugg, R Y Walder, J M Steele, and I M Klotz, Biochemistry, 18,4265 (1979) 51 R Chatterjee, E V Welty, R Y Walder, S L Pruitt, P H Rogers, A h o n e , and J A Walder, J Biol Chem., 261,9929 (1986) 52 S R Snyder, E V Welty, R Y Walder, L A Williams, and J A Walder, Proc Natl Acad Sci USA, 84,7280 (1987) 53 N Komiyama, J Tame, and K Nagai, Biol Chem., 377,543 (1996) 54 T Boyiri, M K Safo, R E Danso-Danquah,'J Kister, C Poyart, and D J Abraham, Biochemistry, 34,15021 (1995) 55 M F Perutz, Br Med Bull., 32, 195 (1976) 56 J Monod, J Wyman, and J.-P Changeux, J Mol Biol., 12,88 (1965) 57 D A Matthews, R A Alden, J T Bolin, S T Freer, R Hamlin, N Xuong, J Kraut, M Poe, M Williams, and K Hoogsteen, Science, 197, 452 (1977) 58 L F Kuyper, B Roth, D P Baccanari, R Ferone, C R Beddell, J N Champness, D K Stammers, J G Dann, F E Norrington, D J Baker, and P J Goodford, J Med Chem., 25, 1120 (1982) 59 D A Matthews, J T Bolin, J M Burridge, D J Filman, K W Volz, and J Kraut, J Biol Chem., 260,392 (1985) 60 K Appelt, R J Bacquet, C A Bartlett, C L J Booth, S T Freer, M A Fuhry, M R Gehring, S M Herrmann, E F Howland, C A Janson, T R Jones, C C Kan, V Kathardekar, K K Lewis, G P Marzoni, D A Matthews, C Mohr, E W Moomaw, C A Morse, S J Oatley, R C Ogden, M R Reddy, S H Reich, W S Schoettlin, W W Smith, M D Varney, J E Villafranca, R W Ward, S Webber, S E Webber, K M Welsh, and J White, J Med Chem., 34,1925(1991) 61 S H Reich and S E Webber, Perspect Drug Discov Des., 1,371-390(1993) 62 L W Hardy, J S Finer-Moore, W R Montfort, M 0.Jones, D V Santi, and R M Stroud, Science, 235,448-455(1987) 63 D.A Matthews, K Appelt, S J Oakley, and N H Xuong, J Mol Biol., 214, 923-936 (1990) 64 W R Montfort, K M Perry, E B Fauman, J S Finer-Moore, G F Maley, L Hardy, F Maley, and R M Stroud, Biochemistry, 29, 6964-6977(1990) 65 Y.Takemura and A L Jackman, Anticancer Drugs, 8,3-16(1997) 66 S E.Webber, T M Bleckrnan, J Attard, J G Deal, V Kathardekar, K M Welsh, S Webber, C A Janson, D A Matthews, W W Smith, S T Freer, S R Jordan, R J Bacquet, E F Howland, C L J Booth, R W Ward, S M Herrmann, J White, C A Morse, J A Hilliard, and C A Bartlett, J Med Chem., 36, 733-746(1993) 67 I Niculescu-Duvaz, Curr Opin Invest Drugs, 2,693-705(2001) 68 P.J Goodford, J Med Chem., 28,849(1985) 69 P.Goodford, J.Chemom., 10,107(1996) 70 M.D Varney, G P Marzoni, C L Palmer, J G Deal, S Webber, K M Welsh, R J Bacquet, C A Bartlett, C A Morse, C L Booth, S M Herrmann, E F Howland, R W Ward, and J White, J Med Chem., 35, 663-676 (1992) 71 D R Newell, Semin Oncol., 26 (Suppl 61, 74-81(1999) 72 P Norman, Curr Opin Invest Drugs, 2, 1611-1622(2001) 73 E C Taylor, Adv Exp Med Biol., 338, 387408(1993) 74 G P Beardsley, B A Moroson, E C Taylor, and R G Moran, J Biol Chem., 264,328333 (1989) 75 J R Piper, G S McCaleb, J A Montgomery, R L Kisliuk, Y Gaumont, J Thorndike, and F M Sirotnak, J Med Chem., 31,2164-2169 (1988) 76 S E Greasley, T H Marsilje, H Cai, S Baker, S J Benkovic, D L Boger, and I A Wilson, Biochemistry, 40,13538-13547(2001) 77 R.J Almassy, C A Janson, C C Kan,and Z Hostomska, Proc Natl Acad Sci USA, 89, 6114-6118(1992) 78 C C Kan, M R Gehring, B R Nodes, C A Janson, R J Almassy, and Z Hostomska, J Protein Chem., 11, 467-473(1992) 79 M D Varney, C L Palmer, W H Romines 3rd, T Boritzki, S A Margosiak, R Almassy, C A Janson, C Bartlett, E J Howland, and R Ferre, J Med Chem., 40,2502-2524(1997) 80 C Shih, L S Gossett, J F Worzalla, S M Rinzel, G B Grindey, P M Harrington, and E C Taylor, J Med Chem., 35, 1109-1116 (1992) 81 D.W Cushman, H S Cheung, E F Sabo, and M A Ondetti, Biochemistry, 16,5484 (1977) 82 M A Ondetti, B Rubin, and D W Cushman, Science, 196,441 (1977) 83 M J Wyvratt and A A Patchett, Med Res Rev., 5,483-531(1985) 84 D W Cushman and M A Ondetti, Hypertension, 17,589(1991) 85 D W Cushman and M A Ondetti, Nat Med., 5,1110(1999) 86 J Rahuel, V Rasetti, J Maibaum, H Rueger, R Goschke, N C Cohen, S Stutz, F Cumin, W Fuhrer, J M Wood, and M G Grutter, Chern Biol., 7,493-504(2000) 87 L D Byers and R Wolfenden, Biochemistry, 12,2070-2078(1973) 88 J R Huff and J Kahn, Adv Protein Chem., 56, 213-251(2001) 89 A Wlodawer and J Vondrasek, Annu Rev Biophys Biomol Struct., 27,249 (1998) 90 T D Meek, J.Enzyme Inhib., 6,65(1992) 91 A Wlodawer and J W Erickson, Annu Rev Biochem., 62,543(1993) 92 R Lapatto, T Blundell, A Hemmings, J Overington, A Wilderspin, S Wood, J R Merson, P J Whittle, D E Danley, K F Geoghegan, et al., Nature, 342,299302(1989) 93 M A.Navia, P M Fitzgerald, B M McKeever, C T Leu, J C Heimbach, W K Herber, I S Sigal, P L Darke, and J P Springer, Nature, 337,615-620(1989) 94 A Wlodawer, M Miller, M Jaskolski, B K Sathyanarayana, E Baldwin, I T Weber, L M Selk, L Clawson, J Schneider, and S B Kent, Science, 245,616-621(1989) 95 I B Duncan and S Redshaw, Infect Dis Ther., 25,27-47(2002) 96 A G Tomasselli, M K Olsen, J Hui, D J Staples, T K Sawyer, R L Heinrikson, and C S Tomich, Biochemistry, 29, 264-269 (1990) Structure-Based Drug Design 97 M Jaskolski, A G Tomasselli, T K Sawyer, Staples, R L Heinrikson, J Schneider, " Kent, and A Wlodawer, Biochemistry, 30, S 1600-1609 (1991) 98 M W Holladay, F G Salituro, and D H Rich, J Med Chem., 30,374-383 (1987) 99 F G Salituro, N Agarwal, T Hofmann, and D H Rich, J Med Chem., 30,286-295 (1987) 100 D J Kempf, K C Marsh, D A Paul, M F Knigge, D W Norbeck, W E Kohlbrenner, L Codacovi, S Vasavanonda, P Bryant, X C Wang, N E Wideburg, J.J Clement, J.J Plattner, and J Erickson, Antimicrob Agents Chemother., 35,2209-2214 (1991) 101 M V Hosur, N T Bhat, D J Kempf, E T Baldwin, B Liu, S Gulnik, N E Wideburg, D W Norbeck, K Appelt, and J W Erickson, J Am Chem Soc., 116,847-855 (1994) 102 D J Kempf, H L Sham, K C Marsh, C A Flentge, D Betebenner, B E Green, E McDonald, S Vasavanonda, A Saldivar, N E Wideburg, W M Kati, L Ruiz, C Zhao, L Fino, J Patterson, A Molla, J J Plattner, and D W Norbeck, J Med Chem., 41, 602-617 (1998) 103 C N Hodge, P E Aldrich, L T Bacheler, C H Chang, C J Eyermann, S Garber, M Grubb, D A Jackson, P K Jadhav, B Korant, P Y Lam, M B Maurin, J L Meek, M J Otto, M M Rayner, C Reid, T R Sharpe, L Shum, D L Winslow, and S EricksonViitanen, Chem Biol., 3,301-314 (1996) 104 B D Dorsey, R B Levin, S L McDaniel, J P Vacca, J P Guare, P L Darke, J A Zugay, E A Emini, W A Schleif, J C Quintero, J H Lin, I W Chen, M K Holloway, P M D Fitzgerald, M G Axel, D Ostovic, P S Anderson, and J R Huff, J Med Chem., 37, 34433451 (1994) 105 J P Vacca, J P Guare, S J DeSolms, W M Sanders, E A Giuliani, S D Young, P L Darke, I S Sigal, W A Schleif, J C Quintero, E A Emini, P S Anderson, and J R Huff, J Med Chem., 34,1228-1230 (1991) 106 T A Lyle, C M Wiscount, J P Guare, W J Thompson, P S Anderson, P L Darke, J A Zugay, E A Emini, W A Schleif, J C Quintero, R A F Dixon, I S Sigal, and J R Huff, J Med Chem., 34,1230-1233 (1991) 107 M K Holloway, J M Wai, T A Halgren, P M Fitzgerald, J P Vacca, B D Dorsey, R B Levin, W J Thompson, L J Chen, S J deSolms, N Gaffm, A K Ghosh, E A Giuliani, S L Graham, J P Guare, R W Hungate, T A Lyle, W M Sanders, T J Tucker, 108 109 110 111 112 113 114 115 116 117 118 119 120 121 M Wiggins, C M Wiscount, W Woltersdorf, S D Young, P L Darke, and J A Zuguay, J Med Chem., 38,305-317 (1995) E E Kim, C T Baker, M D Dwyer, M A Murcko, B G Rao, R D Tung, and M A Navia, J Am Chem Soc., 117,1181-1182 (1995) S W Kaldor, V J Kalish, J F Davies 2nd, B V Shetty, J E Fritz, K Appelt, J A Burgess, K M Campanale, N Y Chirgadze, D K Clawson, B A Dressman, S D Hatch, D A Khalil, M B Kosa, P P Lubbehusen, M A Muesing, A K Patick, S H Reich, K S Su, and J H Tatlock, J Med Chem., 40, 39793985 (1997) M Moledina, M Chakir, and P J Gandhi, J Thromb Thrombolysis, 12, 141-149 (2001) J Hauptmann, Eur J Clin Pharmacol., 57, 751-758 (2002) D W Banner and P Hadvary, J Biol Chem., 266,20085-20093 (1991) P E Sanderson and A M Naylor-Olsen, Curr Med Chem., 5,289 (1998) J P Vacca, Curr Opin Chem Biol., 4, 394 (2000) J Hauptmann, B Kaiser, M Paintz, and F Markwardt, Biomed Biochim Acta, 46, 445453 (1987) H Brandstetter, D Turk, H W HoeMren, D Grosse, J Sturzebecher, P D Martin, B F Edwards, and W Bode, J Mol Biol., 226, 1085-1099 (1992) N H Hauel, H Nar, H Priepke, U Reis, J M Stassen, and W Wienen, J Med Chem., 45, 1757-1766 (2002) R M Friedlander, V Gagliardini, H Hara, K B Fink, W Li, G MacDonald, M C Fishman, A H Greenberg, M A Moskowitz, and J Yuan, J Exp Med., 185,933-940 (1997) B Siegmund, H A Lehr, G Fantuzzi, and C A Dinarello, Proc Natl Acad Sci USA, 98, 13249-13254 (2001) N P Walker, R V Talanian, K D Brady, L C Dang, N J Bump, C R Ferenz, S Franklin, T Ghayur, M C Hackett, L D Hammill, L Herzog, M Hugunin, W Houy, J A Mankovich, L McGuiness, E Orlewicz, M Paskind, C A Pratt, P Reis, A Summani, M Terranova, J P Welch, L Xiong, A Moller, D E Tracey, R Kamen, and W W Wong, Cell, 78,343-352 (1994) K P Wilson, J A Black, J A Thomson, E E Kim, J P Griffith, M A Navia, M A Murcko, S P Chambers, R A Aldape, S A Raybuck, and D Livingstone, Nature, 370, 270-275 (1994) 122 R Leung-Toung, W Li, T F Tam, and K Karimian, Curr Med Chem.,9,979-1002 (2002) 123 M R Michaelides and M L Curtin, Curr Pharm Des., 5, 787-819 (1999) 124 P D Brown, Expert Opin Invest Drugs, 9, 2167-2177 (2000) 125 0.Santos, C D McDermott, R G Daniels, and K Appelt, Clin Exp Metastasis, 15, 499-508 (1997) 126 L J MacPherson, E K Bayburt, M P Capparelli, B J Carroll, R Goldstein, M R Justice, L Zhu, S Hu, R A Melton, L Fryer, R L Goldberg, J R Doughty, S Spirito, V Blancuzzi, D Wilson, E M O'Byrne, V Ganu, and D T Parker, J Med Chem., 40, 2525-2532 (1997) 127 G Clemens, B Hibner, R Humphrey, H Kluender, and S Wilhelm in N J Clendeninn and K Appelt, Eds., Matrix Metalloproteinase Inhibitors in Cancer Therapy, Humana Press, Totowa, NJ, 2001, pp 175-192 128 T Kusumi, M Tsuda, T Katsunuma, and M Yamamura, Cell Biochem Funct., 7, 201-204 (1989) 129 M D Sintchak, M A Fleming, Futer, S A Raybuck, S P Chambers, P R Caron, M A Murcko, and K P Wilson, Cell, 85, 921-930 (1996) 130 L Hedstrom, Curr Med Chem., 6, 545-560 (1999) 131 M D Sintchak and E Nimmesgern, Immunopharmacology, 47, 163-184 (2000) 132 D A Gschwend, A C Good, and I D Kuntz, J Mol Recognit., 9, 175-186 (1996) 133 D Dvornik, J Diabetes Complications, 6, 25-34 (1992) 134 D R Tomlinson, E J Stevens, and L T Diemel, Trends Pharmacol Sci., 15, 293-297 (1994) 135 C L Kaul and P Ramarao, Methods Find Exp Clin Pharmacol., 23,465-475 (2001) 136 A Urzhumtsev, F Tete-Favier, A Mitschler, J Barbanton, P Barth, L Urzhumtseva, J F Biellmann, A Podjarny, and D Moras, Structure, 5,601-612 (1997) 137 M C Van Zandt, E Sibley, K J Combs, E E McCann, B Flam, D J Lavoie, D Sawicki, A Sabetta, A Carrington, J Sredy,V Calderone, B Cuevrier, and A Podjarny, Posterpresented at the 218th National Meeting of the American Chemical Society, New Orleans, LA, August 22-26,1999 138 S Borman, Chem Eng News, 80, 35-39 (2002) 139 E K Perry, B E Tomlinson, G Blessed, K Bergrnann, P H Gibson, and R H Perry, Br Med J.,2,1457-1459 (1978) 140 B P Imbimbo, CNS Drugs, 15, 375-390 (2001) 141 Y Ishihara, K Kato, and G Goto, Chem Pharm Bull (Tokyo), 39,3225-3235 (1991) 142 Y Ishihara, G Goto, and M Miyamoto, Curr Med Chem., 7,341-354 (2000) 143 J L Sussman, M Harel, F Frolow, C Oefner, A Goldman, L Toker, and I Silman, Science, 253,872-879 (1991) 144 Y Yamamoto, Y Ishihara, and I D Kuntz, J Med Chem., 37,314143153 (1994) 145 A H Reid, J K Taubenberger, and T G Fanning, Microbes Infect., 3,81-87 (2001) 146 J N Varghese, W G Laver, and P M Colman, Nature, 303,35-40 (1983) 147 M von Itzstein, W.-Y Wu, G B Kok, M S Pegg, J C Dyason, B Jin, T Van Phan, M L Smythe, H F White, S W Oliver, P M Colman, J N Varghese, D M Ryan, J M Woods, R C Bethell, V J Hotham, J M Cameron, and C R Penn, Nature, 363,418-423 (1993) 148 P Bossart-Whitaker, M Carson, Y S Babu, C D Smith, W G Laver, and G M Air, J Mol Biol., 232,1069-1083 (1993) 149 J N Varghese, V C Epa, and P M Colman, Protein Sci., 4, 1081-1087 (1995) 150 C U Kim, W Lew, M Williams, H Liu, L Zhang, S Swaminathan, N Bischofberger, M S Chen, D Mendel, W G Laver, and R C Stevens, J Am Chem Soc., 119,681 (1997) 151 Y S Babu, P Chand, S Bantia, P Kotian, A Dehghani, Y El-Kattan, T H Lin, T L Hutchison, A J Elliott, C D Parker, S L Ananth, L L Horn, G W Laver, and J A Montgomery, J Med Chem., 43, 3482-3486 (2000) 152 J A Green, G M Smith, R Buchta, R Lee, K Y Ho, I A Rajkovic, and K F Scott, Inflammation, 15,355-367 (1991) 153 P Vadas, J Browning, J Edelson, and W Pruzanski, J Lipid Mediat., 8,l-30 (1993) 154 C Bennion, S Connolly, N P Gensmantel, C Hallam, C G Jackson, W U Primrose, G C Roberts, D H Robinson, and P K Slaich, J Med Chem., 35,2939-2951 (1992) 155 S Connolly, C Bennion, S Botterell, P J Croshaw, C Hallam, K Hardy, P Hartopp, C G Jackson, S J King, L Lawrence, A Mete, D Murray, D H Robinson, G M Smith, L Stein, I Walters, E Wells, and W J Withnall, J Med Chem., 45,1348-1362 (2002) Structure-Based Drug Design 156 H G Beaton, C Bennion, S Connolly, A R Cook, N P Gensmantel, C Hallam, K Hardy, B Hitchin, C G Jackson, and D H Robinson, J Med Chem., 37,557-559 (1994) 157 J P Wery, R W Schevitz, D K Clawson, J L Bobbitt, E R Dow, G Gamboa, T Goodson Jr., R B Hermann, R M Kramer, D B McClure, et al., Nature, 352, 79-82 (1991) 158 R W Schevitz, N J Bach, D G Carlson, N Y Chirgadze, D K Clawson, R D Dillard, S E Draheim, L W Hartley, N D Jones, Mihelich, et al., Nut Struct Biol., 2, 458-465 (1995) 159 D L Scott, S P White, J L Browning, J J Rosa, M H Gelb, and P B Sigler, Science, 254,1007-1010 (1991) 160 M M Thunnissen, E Ab, K H Kalk, J Drenth, B W Dijkstra, 0.P Kuipers, R Dijkman, G H de Haas, and H M Verheij, Nature, 347,689-691 (1990) 161 S E Draheim, N J Bach, R D Dillard, D R Berry, D G Carlson, N Y Chirgadze, D K Clawson, L W Hartley, L M Johnson, N D Jones, E R McKinney, E D Mihelich, J L Olkowski, R W Schevitz, A C Smith, D W Snyder, C D Sommers, and J P Wery, J Med Chem., 39,5159-5175 (1996) 162 D W Snyder, N J Bach, R D Dillard, S E Draheim, D G Carlson, N Fox, N W Roehm, C T Armstrong, C H Chang, L W Hartley, L M Johnson, C R Roman, A C Smith, M Song, and J H Fleisch, J Pharmacol Exp Ther., 288, 1117-1124 (1999) 163 D M Springer, Curr Pharm Des., 7,181-198 (2001) 164 C Savolainen, S Blomqvist, M N Mulders, and T Hovi, J Gen Virol., 83 (Pt 2), 333-340 (2002) 165 M G Rossmann, Viral Immunol., 2, 143-161 (1989) 166 M G Rossmann, E Arnold, J W Erickson, E A Frankenberger, J P Griffith, H J Hecht, J E Johnson, G Kamer, M Luo, A G Mosser, R R Rueckert, B Sherry, and G Vriend, Nature, 317, 145-153 (1985) 167 G D Diana, M A McKinlay, M J Otto, V Akullian, and C Oglesby, J Med Chem., 28, 1906-1910 (1985) 168 G D Diana, M A McKinlay, C J Brisson, E S Zalay, J V Miralles, and U J Salvador, J Med Chem., 28, 748-752 (1985) 169 M J Otto, M P Fox, M J Fancher, M F Kuhrt, G D Diana, and M A McKinlay, Antimicrob Agents Chemother., 27, 883-886 (1985) 170 M P Fox, M J Otto, and M A McKinlay, Antimicrob Agents Chemother., 30, 110-116 (1986) 171 B Jubelt, A K Wilson, S L Ropka, P L Guidinger, and M A McKinlay, J Infect Dis., 159, 866-871 (1989) 172 T J Smith, M J Kremer, M Luo, G Vriend, E Arnold, G Kamer, M G Rossmann, M A McKinlay, G D Diana, and M J Otto, Science, 233,1286-1293 (1986) 173 D C Pevear, M J Fancher, P J Felock, M G Rossmann, M S Miller, G D Diana, A M Treasurywala, M A McKinlay, and F J Dutko, J Virol., 63,2002-2007 (1989) 174 K H Kim, P Willingmann, Z X Gong, M J Kremer, M S Chapman, I Minor, M A 01iveira, M G Rossmann, K Andries, G D Diana, F J Dutko, M A McKinlay, and D C Pevear, J Mol Biol., 230, 206-227 (1993) 175 G D Diana, D Cutcliffe, R C Oglesby, M J Otto, J P Mallamo, V Akullian, and M A McKinlay, J Med Chem., 32,450-455 (1989) 176 G D Diana and D C Pevear, Antiviral Chem Chemother., 8,401 (2002) 177 G D Diana, P Rudewicz, D C Pevear, T J Nitz, S C Aldous, D J Aldous, D T Robinson, T Draper, F J Dutko, C Aldi, et al., J Med Chem., 38,1355-1371 (1995) 178 J M Rogers, G D Diana, and M A McKinlay, Adv Exp Med Biol., 458,69-76 (1999) 179 F G Hayden, T Coats, K Kim, H A H q s man, M M Blatter, B Zhang, and S Liu, Antiviral Ther., 7, 53-65 (2002) 180 B DBrijard, J Raingeaud, T Barrett, I.-H Wu, J Han, R J Ulevitch, and R J Davis, Science, 267,682-685 (1995) 181 K P Wilson, P G McCaffrey, K Hsiao, S Pazhanisamy, V Galullo, G W Bemis, M J Fitzgibbon, P R Caron, M A Murcko, and M S Su, Chem Biol., 4,423-431 (1997) 182 B Frantz, T Klatt, M Pang, J Parsons, A Rolando, H Williams, M J Tocci, S J O'Keefe, and E A O'Neill, Biochemistry, 37, 13846-13853 (1998) 183 J C Lee, J T Laydon, P C McDonnell, T F Gallagher, S Kumar, D Green, D McNulty, M J Blumenthal, J R Heys, S W Landvatter, J E Strickler, M M McLaughlin, I R Siemens, S M Fisher, G P Livi, J R White, J L Adams, and P R Young, Nature, 372, 739-746 (1994) 184 J C Lee, S Kumar, D E Griswold, D C Underwood, B J Votta, and J L Adams, Immunopharmacology, 47,185-201 (2000) References 185 A Cuenda, J Rouse, Y N Doza, R Meier, P Cohen, T F Gallagher, P R Young, and J C Lee, FEBS Lett., 364,229-233(1995) 186 A M Badger, J N Bradbeer, B Votta, J C Lee, J L Adams, and D E Griswold, J Pharmmol Exp Ther., 279,1453-1461 (1996) 187 Z Wang, B J Canagarajah, J C Boehm, S Kassisa, M H Cobb, P R Young, S AbdelMeguid, J L Adams, and E J Goldsmith, Structure, 6,1117-1128(1998) 188 K P Wilson, M J Fitzgibbon, P R Caron, J P Griffith, W Chen, P G McCaffrey, S P Chambers, and M S Su, J Biol Chem., 271, 27696-27700(1996) 189 T Fox, J T Coll, X Xie, P J Ford, U A Germann, M D Porter, S Pazhanisamy, M A Fleming, V Galullo, M S Su, and K P Wilson, Protein Sci., 7,2249(1998) 190 R J Gum, M M McLaughlin, S Kumar, Z Wang, M J Bower, J C Lee, J L Adams, G P Livi, E J Goldsmith, and P R Young, J Biol Chem., 273,15605-15610(1998) 191 J L Adams, J C Boehm, T F Gallagher, S Kassis, E F Webb, R Hall, M Sorenson, R Garigipati, D E Griswold, and J C Lee, Bioorg Med Chem Lett., 11, 2867-2870 (2001) 192 T Fullerton, A Sharma, U Prabhakar, M Tucci, S Boike, H Davis, D Jorkasky, and W Williams, Clin Pharmacol Ther., 67, 114 (2000) 193 Pat Appl Vertex Pharmaceuticals, Inc., assignee, PCT WO 00/36096(2000) 194 J J Haddad, Curr Opin Invest Drugs, 2, 1070 (2001) 195 C Pargellis, L.Tong, L Churchill, P F Cirillo, T Gilmore, A G Graham, P M Grob, E R Hickey, N Moss, S Pav, and J Regan, Nut Struct Biol., 9,268-272(2002) 196 J Regan, S Breitfelder, P Cirillo, T Gilmore, A G Graham, E Hickey, B Klaus, J Madwed, M Moriak, N Moss, C Pargellis, S Pav, A Proto, A Swinamer, L Tong, and C Torcellini, J Med Chem., 45,2994(2002) 197 G R Boss and J E Seegmiller, Annu Rev Genet., 16,297-328(1982) 198 S E Ealick, S A Rule, D C Carter, T J Greenhough, Y S Babu, W J Cook, J Habash, J R Helliwell, J D Stoeckler, R E Parks Jr., S Chen, and C E Bugg, J Biol Chem., 265,1812(1990) 199 S E Ealick, Y S Babu, C E Bugg, M D Erion, W C Guida, J A Montgomery, and J A Secrist 3rd, Proc Natl Acad Sci USA, 88,11540-11544(1991) 200 J A Montgomery, S Niwas, J D Rose, J A Secrist 3rd, Y S Babu, C E Bugg, M D Erion, W C Guida, and S E Ealick, J Med Chem., 36,55-69(1993) 201 M Duvic, E A Olsen, G A Omura, J C Maize, E C Vonderheid, C A Elmets, J L Shupack, M F Demierre, T M Kuzel, and D Y Sanders, J Am Acad Dermatol., 44, 940-947 (2001) 202 P E Morris Jr and G A Omura, Curr * Pharm Des., 6,943-959(2000) CHAPTER ELEVEN X-Ray Crystallography in rug Discovery " DOUGLAS A LMNGSTON SEANG BUCHANAN KEVINL D'AMICO MICHAEL V MILBURN THOMAS S PEAT J MICHAEL SAUDER Structural GenomiX San Diego, California Contents Introduction, 472 Methodology, 472 2.1 Theory, 472 2.2 Crystallization, 473 2.3 Data Collection, 474 2.4 Phase Problem, 476 2.5 Computing and Refinement, 478 2.6 Databases, 478 Applications of the Use of Crystallographic Studies in Drug Discovery and Development, 479 Structural Genomics, 481 4.1 Introduction to Structural Genomics, 481 4.2 Genome Annotation, 481 4.3 Pathways, 495 4.4 Protein Structure Modeling, 495 Conclusion, 496 Burger's Medicinal Chemistry and Drug Discovery Sixth Edition, Volume 1: Drug Discovery Edited by Donald J Abraham ISBN 0-471-27090-3 2003 John Wiley &Sons, Inc 471 X-Ray Crystallography in Drug Discovery INTRODUCTION The practice of crystallography is undergoing dramatic change because of the advent of new robotics technologies, orders-of-magnitude improvement in X-ray sources and computational power, and the advances in protein production stemming from the recent revolution in molecular biology This chapter covers these changes in the context of an overview of the techniques of modern crystallography, their application in the identification and characterization of targets and mechanisms for therapeutic intervention, and the nascent field of structural genomics Structure-based drug design applications are covered elsewhere The exponential growth in the rate of determination of new protein structures continues unabated Technologies developed in the late 1980s (1)have now evolved to the point that they have been implemented in highthroughput (HTS) format, driving the rate even higher Super-intense, precise, tunable X-rays are now available from undulator beamlines Three "third-generation" synchrotrons, designed and built for this purpose, are now on line-ESRF in Grenoble, France; Spring-8 in Japan; APS at Argonne National Laboratory in the United States-and others are under construction In a relative sense, this capability has had minimal impact on medicinal chemistry to date, but that will certainly change The companies that have successfully built high-throughput protein crystallography systems (SGX and Syrrx in the United States and Astex in the U.K among others) have all now turned their prodigious capacity to the co-crystallization of small molecules with target proteins for the purpose of drug discovery The capacity to compare, in parallel, the binding modes of a set of hits from HTS, or a given lead series, will be valuable, but an even greater impact will result from the decrease in turnaround time required to generate co-crystal structures This has been the most significant hindrance to realizing the full potential of structurebased drug design A structure is far more use- ful before the chemist has embarked on the synthesis of the next series, rather than after Another important development toward new target identification is the effort in largescale structural annotation of various genomes, the field of structural genomics In classifying proteins by function as a step toward validating them as therapeutic targets, structural homology is perhaps the most important tool available These efforts have been taken up by a number of publicly-funded consortia (2), because the commercial value of genomic databases in general has not been high enough to justify their cost in the private sector Given that medicinal chemists think and communicate largely in structural terms, this recent growth in the influence of structural biology is very important It forms the basis of a powerful link between chemistry and biology, and we have only begun to realize its potential METHODOLOGY 2.1 Theory X-ray crystallography provides atomic or near atomic resolution of matter The periodicity of crystals, reflecting the repeating units of molecular structure, diffracts X-rays according to Bragg's law: nh = 2dsin0, where n is the order of diffraction, h the wavelength of the radiation, d the spacing or distance between a family of lattice planes in the crystal, and e the angle of the diffraction X-radiation is ideal to analyze atomic structure, because the wavelengths used are in the order of 0.1-2.0 A with 0.75 A being about one-half the distance of an aliphatic carbon-carbon bond The images of diffracted crystal lattices can be observed with specialized precession photographic equipment, although the modern day image plate detectors used in most laboratories produce a diffraction image that can be analyzed by computer to provide the indices of the lattice diffraction spots (Fig 11.1, a-c) The X-ray diffraction from the electron clouds surrounding each nucleus is either reinforced or impeded and gives rise to the dif- ethodology 473 Figure 11.1 (a) A look at a two-dimensional crystal lattice diffraction pattern for a small molecule natural product, MW 222 Each diMaction intensity in the lattice is numbered to give a unique three hensional address (identification)for that measurement These numerical addresses are referred to as Uiller indices or hkl values (b) A diffraction pattern from a precession photograph for hemoglobin, MW 55,000 Note the the diffraction lattice spacings are much smaller for the large molecule and reflects the nature of Bragg's law, where the lattice is observed in reciprocal space (lld = 2sinBlnA) (c)An image plate Wfraction pattern for a protein [Adapted with permission from D J Abraham, Computer-Aided Drug Design, Methods, and Applications, Marcel Dekker, Inc., New York, 1989.1 fere:nce in intensities observed in Fig 11.1 The steps that one goes through to solve a crystal structure follow, with the intent of providiiig the non-crystallographer with a simplified and pictorial view of the process Crystallization Cry$ltallization is the critical first and most imp(wtant step, because good single crystals USUiilly provide quality diffraction Linus Paulling once entitled one of his lectures "The Imp1ortance of Being Crystalline" (3) Unfor- tunately, crystallization is still more empirical than scientific It requires closely monitored matrix changes in growing conditions, i.e., pH, salt concentration, temperature, solvents, and crystallization setups Most laboratories now use well-known sparse matrix screens pioneered by Jancarik and Kim (4) and further refined and commercially distributed by Hampton Research (5, 6) Screens will typically employ vapor diffusion experiments (hangingdrops or sitting drops), and occasionally batch and liquid-liquid diffusion methods X-Ray Crystallography in Drug Discovery More recently, batch crystallizations have been rejuvenated by the development of microbatch robots and by the groups of Chayen (7), DeTitta (8), and D'Arcy (9) Although discovering the crystallization conditions for a new protein or nucleic acid can be tedious, relatively inexperienced individuals can usually succeed at growing crystals once the initial conditions are established Some of the most successful crystallization methodologies are based on vapor diffusion methods (Fig 11.2) The general idea behind vapor diffusion crystallization is to dissolve the protein in a buffer, with a non-precipitating amount of the miscible vapor solvent, in a reservoir that is in equilibrium with higher concentration of the vaporizing solvent nearby Another variation is to set up the crystallization cocktail containing salts, buffers, P (poly) ethylene glycols (PEGS), small molecule solvents, etc., where volume is slowly reduced by the equilibrating mixture, which is placed nearby McPherson, Carter, and others have developed more quantitative methods for optimizing crystal growth (10) 2.3 Data Collection Most laboratories have rotating anode sources for production of high intensity X-ray beams These are coupled with an area detector that has made single crystal diffractometers obsolete Mirrors and other technology have also been used to provide a more intense and monochromatic radiation source (11).Radiation from rotating anode sources is at a fixed wavelength, usually from high-voltage electrons impinging on either a copper or molybdenum rotating anode, i.e., radiation at 1.54 (copper) or 0.71 A (molybdenum) Radiation from synchrotron sources can often be tuned to a wavelength of interest for multiwavelength anomalous diffraction (MAD) experiments (see below) X-rays generated by a synchrotron source are typically two orders of magnitude stronger than conventional CuKa radiation generated by a rotating anode Synchrotron sources have greatly extended the ability to solve new protein structures when only weakly diffracting or small crystals are available Another advantage in using the stronger synchrotron radiation is that the crystal exposure time is signif- icantly lower The typical exposure time for home laboratory CuKa sources ranges from to 60 for a range of data, whereas the equivalent set of data at an undulator beamline, i.e., the advanced photon source (APS), requires only about s of exposure time Synchrotron radiation has also allowed the use of MAD, enabling phasing (imaging) of the protein using a derivative with only one heavy element A variety of detectors are in common use to record X-ray data and have the advantage of measuring the intensities of large numbers of diffraction spots simultaneously The most popular detectors are image plates and chargecoupled device (CCD) cameras Image plates are typically the choice for laboratory rotating anode sources and lower flux synchrotron sources (Fig 11.3) CCDs have the distinct advantage of speed at the higher flux synchrotron sources, because they simultaneously measure and record diffraction intensities (amplitudes) Current CCD cameras have readout times on the order of a few (typically 2-8) seconds, a speed not dreamed of when the first protein structure data was recorded from phot&aphs (with intensities measured by eye comparison to standard reference spots on a separate film strip) Speed of data collection can be an important advantage at third generation synchrotron sources, with even shorter exposure times On the other hand, image plates have a greater range of use, being accessible in any X-ray diffraction laboratory, with many of the newer models taking less than to record the intensity data Image plate detectors offen have more than one image plate, so one can be read while the other is exposed, effectively wasting no time during the collection period The image plates also offer a larger surface area for data collection than most CCD cameras and are considerably less expensive X-ray diffraction data from crystals are either collected at room temperature or under cryogenic conditions at liquid nitrogen temperatures [around 100°K (- 170"C)l.For room temperature data collection, crystals are normally mounted in thin-walled glass capillaries, with a small amount of mother liquor about mm from the crystal The mother liquor in the capillary is critical because protein crystals are 40-80% water-dried protein crystals not diffract The nearby mother Drop: 50% protein, 50% cocktail - SCcces " "."-."."-"m""" fur T r i l Y0000750 nn 17 OCT-01 i-,sxxtbn I Drop: 50% protein, 50% cocktail Fi gure 11.2 (a) The drops are typically 1-10 pA total volume, with between 100 and 1000 pA total vo.lume of cocktail in the well The smaller the drop size, the faster the equilibrium occurs, in general Thlere are a variety of plates now available in which to set up these vapor diffusion experiments, the mc1st common being 24-well Limbro plates and 96-well microtiter plates Several robots have been d eveloped to automatically set up the crystallization experiments; although most are no faster than do1ing the same procedure by hand (particularly with a multi-channel pipettor), there can be other ad.vantages (e.g., consistency and reducing repetitive stress syndrome) Once plates are set up, they arc? typically kept at a constant temperature and observed periodically under a microscope (b and c) Prc3gress in automating this aspect of characterization has occurred, and there are now imagers that wil1 take high resolution, digital pictures of each drop in turn and store these for either manual or aut;omated analysis (dl Batch experiments are set up such that the protein is mixed with cocktail and the!re is little concentration or dilution to the sample over time This can now be done in very high thr,oughput and small scale: 50-200 nL drops under oil in 1536-well plates, for example This kind of aPI)roach has been used to screen hundreds of conditions with small amounts of protein, which may all()w for faster optimization later One caveat is that small crystals don't necessarily lead to larger crystals later, and all structures to date have had crystals of greater than 10 microns in a t least one dinlension X-Ray Crystallography in Drug Discovery Figure 11.4 The crystals are manipulated by scooping them up with a small loop of nylon that is glued to the end of a pin Surface tension from the liquid will hold the crystal in the loop, but the crystal can also be held by using a loop that is smaller in size than the crystal of interest This technique will work particularly well with fragile crystals, thin plates for example, that would normally fall apart in a capillary mount Once the crystal is frozen, it is placed on an axis in line of both an X-ray source and a stream of nitrogen set to about 100,000 to keep the crystal frozen The crystal is rotated in increments during the data collection procedure to collect a full data set (typically one or two degrees per frame, depending on the resolution limits, mosaicity of the crystal, unit cell lengths, etc.) Figure 11.3 (a) Area detector showing the configuration of the unit (b) Area detector showing the face liquor ensures that the crystal is bathed in the vapor of the mother liquor and prevents drying during data collection The majority of present day data collections in home laboratories and at synchrotrons are done under cryogenic conditions, which allows high intensity X-radiation to be used without the crystal decay observed in room temperature data collection For cryogenic data collection, crystals are normally mounted in a thin fiber loop with a layer of suitable cryoprotectant solution (Fig 11.4) The cryoprotectant forms alayer of noncrystalline glass around the crystal to protect it from freeze shock Simple freezing of the crystal results in the formation of ice in the interior of the crystal and renders it useless A quick perusal of the literature shows PEG, glycerol, sucrose, and 2-methyl-2,4-pentane diol (MPD) as the most popular cryoprotectants Oils, such as paraffin oil, have also been used successfully as cryoprotectants (12) 2.4 Phase Problem X-ray diffraction measurements as described above only provide the amplitudes of the diffracted waves One must have the phase angles of all measured waves relative to a common origin in order to image the molecule using a Fourier analysis Figure 11.5 illus- Methodology (a) (b) Amplitude t Phase difference k Figure 11.5 Graphs showing the phase relationship of electromagnetic radiation [Adapted with permission from D J Abraham, Computer-Aided Drug Design, Methods, and Applications, Marcel Dekker, Inc., New York, 1989.1 trates the differences in the phases and amplitudes for two reflections The solution of the phase problem that permitted the first image reconstruction of a protein was discovered by Perutz using multiple isomorphous replacement (MIR) (13) The majority of the earliest structures were solved using MIR to phase the maps This requires soaking the crystals or co-crystallizing the protein with two or more heavy atoms and hoping that these heavy atoms bind in a specific way to the protein It also requires that the subsequent crystals are isomorphous with the native protein (i.e., no changes in the unit cell or symmetry of the crystal) Although it is possible to obtain phase information from a single heavy atom derivative using additional information (e.g., anomalous scattering or density modification),one often works diligently to get a second or third derivative to improve the quality of the electron density (Fourier) map Two other common methods are used to estimate phases in protein crystallography: molecular replacement (MR), which uses the structural motif of a homologous protein (14), and MAD from a single heavy element (1) MR methodology requires a structural model that is structurally homologous to the protein that has been crystallized Phasing is accomplished through a six-dimensional search-a three-dimensional rotation search followed by a three-dimensional translation search, using the model against the crystal data Molecular replacement is being employed more frequently as the number of known structures has increased, which has made unique structural motifs available for phasing For highly homologous protein structures, this method is usually straightforward and successful For marginal cases, the addition of some independent phase information, single isomorphous replacement (SIR) or MIR, in combination with MR can enhance the quality of the Fourier map MAD phasing is an alternate methodology for solving the phase problem MAD requires a single heavy atom with anomalous peak scattering at a wavelength where X-rays both at and near the spectral energy are accessible Data sets are collected at different wavelengths to optimize the anomalous and dispersive signals m the heavy atom Certain beamlines have been designed with wavelengths that are tunable "on the fly," and these are often referred to as MAD beamlines MAD has become the method of choice for rapid structure solution when synchrotron radiation is available The advantage of MAD phasing is that one often only needs a single crystal to collect all of the data necessary to solve the structure Although multiple wavelengths are collected (anywhere from two to five sets), data collection is routinely completed in less than a few hours The peak wavelength choice data set is very important to collect first as it contains the greatest anomalous signal and is often used alone to find the heavy atom sites (Shake'n Bake program, anomalous Patterson maps, etc.) If the crystal degrades quickly in the beam, one can also employ single-wavelength anomalous diffraction phasing (SAD) if a full data set at the peak wavelength was successfully collected SAD phasing requires additional information, obtained by density modification, to obtain interpretable electron density maps, but has been X-Ray Crystallography in Drug Discovery proven in many instances to result in very high quality maps (15) Many different heavy atoms have been used for MADISAD phasing, the most popular being selenium Selenium is incorporated into the amino acid sequence of the protein by adding selenomethionine to the growth media when the protein is produced (16) For proteins that bind DNA, 5-bromouracil has been a popular choice for phasing through anomalous scattering Most heavy elements have good anomalous signals (Hg, Pt, U, Au, etc.) Lanthanides have a particularly good signal and can sometimes substitute for divalent metals found naturally in the protein (e.g., Ca) (17) One of the major advantages of MAD phasing is that the signal does not decay at higher resolution with perfectly isomorphic crystals, so the experimentally phased map can be quite good out to the full resolution of diffraction This typically has not been the case when using multiple isomorphous replacement, where the experimentally phased map often only extends to around 2.5 A resolution, because of a lack of isomorphism between the native and heavy metal substituted crystals Anomalous scattering has been useful in the structure determination of very large structures; the 30s ribosome was recently solved using 0s and Lu derivatives (18) 2.5 Computing and Refinement Raw intensity X-ray crystallographic data is next reduced and scaled to provide structure factors (F)that are used to solve and image the structure Two of the most -popular software packages employed to reduce raw date are Mosflm/CCP4 (19) and the HKL suite (20) Both work very well and are very fast with modern computers A variety of programs, such as SOLVE (21), Shake and Bake (22), or SHELX (23), can be employed to find the heavy atom positions, including hand searching methods through Patterson maps Once heavy atom sites are found, they are usually refined with the programs SHARP (24) or MLPHARE (19) The heavy atom positions are next used as phase information input to provide initial phases for electron density maps, which are used to fit the remainder of the protein or nucleic acid Once a model of the structure is obtained it is refined In cases where high resolution data is available, pro- grams such as WARP (24) can automatically provide models of protein structures When high - resolution data is not available a model is most often built in by hand using such graphics programs like (25) or XFIT The models are refined against the data by programs such as REFMAC (19) and CNS (26) All of these programs have become much faster and easier to use because of the incredible increases in speed that new hardware has allowed It is worth mentioning that statistical and probabilistic techniques have had a significant impact in how heavy atoms are found and models are refined (e.g., SHARP, SOLVE, REFMAC) Baysian statistics and maximum likelihood methods are now used instead of least-squares methods One may want to consider how various data collection strategies may affect the later steps in the process by keeping this in mind, i.e., high redundancy in the data makes for better statistics The quality of a structure is measured in many ways: how low the R factor or R,,, is (the fit of observed data to the model), the resolution limit of the data, the ideality of the bonds and angles, etc How well a structure measures up to other structures of about the same resolution also gives a good idea of how "good" a given structure is (PROCHECK program) SFCHECK is a useful program for,assessing the agreement between the atomic model and the experimental X-ray data The level of confidence one expects from a given model will depend on the resolution of the data This can be seen clearly in Fig 11.6, where a residue from a protein structure is shown with three different data cutoffs at different resolution ranges The model from a 3.0-A data set may look the same as one from a 1.3-A data set, but the level of confidence is much higher in the latter A reasonably wellrefined structure will have a crystallographic R factor between 15% and 25% and will have an R,,, of less than 30% under most circumstances 2.6 Databases The Protein Data Bank (PDB) (27,281 is now coordinated by a consortium of several institutions (Rutgers University, the San Diego Supercomputer Center, and National Institute for Standards and Technology) As of this writing, the PDB has over 18,000 structures, lications of the Use of Crystallographic Studies in Drug Discovery and Development 479 gure 11.6 Three density maps at differing resolutions: a, 1.3 A; b, 2.1 A; c, 3.0A See color insert with alver 15,000 of these done by X-ray crystallogeaphy Most of the rest were done by NMR For small molecules, the Cambridge Struct(ural Database (CSD) (29) contains structural information for over 230,000 organic and organometallic compounds All of , these structures have been determined by Xray or neutron diffraction techniques AP PLlCATlONS OF THE USE OF CRYST'ALLOCRAPHIC STUDIES IN DRUG DlSCC)VERY A N D DEVELOPMENT Crystadlization of small molecule compounds with a protein or nucleic acid target followed by X-riay crystallographic determination of the combiined structure is the basis and hallmark of structure-based drug design As structural biology moves into the post-genomic age, many companies and academic laboratories are faced with the challenge of co-crystallization of targets and inhibitors or activators on a scale never before attempted Previously, crystal structure determination of a proteinsubstrate or inhibitor complex in an academic or industrial environment often yielded the structural information desired to understand the mechanism of action or in the design of a more suitable substrate or inhibitor However modern day laboratories are now faced with the daunting challenge of crystallizing hundreds of compounds for clues in further ligand design using standard organic synthesis or combinatorial approaches X-Ray Crystallography in Drug Discovery A variety of methods have been employed to co-crystallize biological molecules with small molecules Discovery of crystallization conditions is still an often tedious task, so newer methods for screening crystallization conditions for proteins include the use of semiautomated robots Two fundamental methods are available for co-crystallizations One method is termed "soaking." This employs the addition of the small molecule directly to saturated solutions containing crystals of the biological macromolecule in hopes that the ligand of interest will soak directly into the crystal and bind to its active or binding site, so that the co-crystal structure can be determined The other method, called co-crystallization, depends on having an ability to add the ligand to the aqueous protein solution in at least stoichiometric amounts, followed by crystallization using either the known crystallization conditions or by setting up a new screen for determining suitable crystallization conditions Both methods have disadvantages and advantages, and it is primarily up to the investigator to decide which method, or both methods, should be employed for their experiments One limitation to the soaking method is that the amount actually dissolved and available to form the complex can often not be easily determined or controlled In general, an excess of ligand, as a solid, is added to the solution with the crystals of protein with the hope that the ligand dissolves completely and will diffuse into the crystal binding site One method that has demonstrated success involves equilibration of the crystal with slightly higher concentration of the crystal mother liquor that contains the ligand solubilized by organic cosolvents (i.e., isopropanol, DMSO, ethanol, etc.) as part of the medium for diffusion However, higher levels of organic solvent often decreases the resolution of diffraction Lowering the level of solvent after the addition of compound has been found to result in better diffracting crystals Another major limitation of this method is that it is necessary to collect an entire X-ray diffraction data set to determine if the small molecule is bound to the protein This trial and error method can be time-consuming or expensive if a highthroughput crystallography approach is the objective Co-crystallization permits highly parallel screening for bound ligands through robotic systems The co-crystallization method is better suited for high-throughput crystallography, because ligand binding can sometimes be determined without the need to solve each structure Faster spectral analyses, or alternatives such as native gel shifts, gel filtration, and mass spectrometry can provide information on which of the crystals should be taken into X-ray studies One difficulty in using the co-crystallization method is the problem of determining the concentration of the protein that is most suitable for complexation For example, if the protein solution is roughly mM in high salt or aqueous buffers, many organic molecules are not as soluble at that level In these cases a lower concentration of protein is usually employed to attain stoichiometric ratios As described above, small percentages of organic solvent can be useful for increasing the concentration of the organic compound in solution, but not without affecting the protein stability or crystal quality In general, lowering the protein concentration sufficiently, followed by addition of the appropriate amount of ligand, and then concentration of the mixture to the desired protein concentration for crystallization is the most successful method Once conditions for obtaining the complexed protein have been obtained, the next step is to decide on which crystallization conditions to use In some cases, those proteins that not undergo large tertiary structure changes when complexed to ligands can be crystallized under similar conditions as for the uncomplexed ligand However, in some instances, proteins will change conformation, depending on the type of ligand that they are complexed with, and a large screening of possible new crystallization conditions is required In many cases, soaking a compound into a crystal is not possible because of low solubility of the compound in the aqueous mother liquor Soaking experiments can also be limited when the conformational space of the binding site is hindered, occupied by adjacent molecules in the crystal lattice, or if there are conformational changes in the binding site because of crystal packing effects On the other hand, co-crystallization of the protein and li- gand, by the nature of the process, usually requires more resources in terms of protein and experimental time, leading to greater expense Soaking crystals and/or growing crystals in the presence of inhibitors or ligands provides an opportunity to directly observe their binding interactions, along with the often subtle conformational effects that can have a profound effect on the mode of binding When it is possible to use this in an iterative fashion to guide the design of the next set of compounds to be synthesized in a lead series, this becomes a potent tool Understanding how and why a compound or series binds to an active site, particularly when the affinity is also known, provides the best understanding, and the highest level at which it is possible to enable drug design As structural biology becomes a more integrated component of drug discovery, better methods of obtaining crystal structures with and without bound ligands will be developed, with lower costs and faster turnaround times Many companies and academic laboratories are now focused on solving these challenges But it is also impressive to see how well we have progressed-Table 11.1 enumerates the structures of known therapeutic targets (with or without ligands) that are available in the public domain This table is based on the Nature Biotechnology "The Usual Suspects" poster, published in 2001, but is almost identical in content to the 1997 version (30) Reference sequences for 300 of the targets were extracted from NCBI Genbank and the nonredundant database was searched with several iterations of PSI-BLAST The resulting profile was used to search the PDB + SGX database of known structures The top hits for each drug target are tabulated below A great many more reside within pharmaceutical and biotech companies as proprietary structures 4.1 STRUCTURAL CENOMICS Introduction to Structural Cenomics Until recently, structure determination by protein crystallography was a time-consuming method accessible to a few privileged skilled practitioners X-ray crystallography was reserved to tackle questions requiring atomic resolution details of a demonstrably important protein, often a drug target Indeed, to this day, crystallography is almost exclusively used in the pharmaceutical industry to study small molecule interactions with drug targets (see Section 3) The development of several new methods (described in Section 2) and the availability of the complete genome sequences of both pathogens and hosts provides an unparalleled opportunity to exploit protein structures for drug discovery research in new ways We can now contemplate using protein structure determination to help annotate genomes, that is, to assess new drug targets as well as provide multiple high-resolution structures that address selectivity issues This emerging science of high-throughput structural biology has been termed structural genomics 4.2 Genome Annotation It is in infectious disease that whole genome information first became available, and it is in this field that structural genomics is having an initial impact (373) A typical approach has been to assess the viability and/or virulence of pathogens by systematic disruption of every predicted gene product As a consequence, a large number of potential new targets have emerged: genes that are essential for pathogenicity of bacteria in a model system Often these new genes have been filtered for those that are conserved in a variety of pathogens and that not have a close human homolog (374) About 30-50% of the genes of a typical pathogen have no reliable functional assignment A similar fraction of the novel targets shown to be essential also fall into this category, which becomes problematic for assay configuration Indeed, in target-based approaches, the number of leads emerging has been disappointing Protein structure can provide the information required to prioritize among these essential genes and to establish assays Co-complex structures with even lowaffinity hits can be used to provide key information for medicinal chemistrv " There are several ways in which structural genomics has promise as a tool for genome annotation and target prioritization For genes of unknown function structure can often -provide clues to biochemical function Sequence homology has become a routine method for functional assignment, but even the most Table 11.1 Known Drug Targets with Published Structures Target and PDB Reference Acetylcholinesterase 1MAHW 1B41(A), 1F8U(A) 1C2B(A), 1C20(A) lMAA(A) Adenosine deaminase lFKX, lFKW 1A4L(A),1A4M(A) 1UI0, lUIP lADD 2ADA Alpha-amylase lJXT(A), lJXK(A) lSMD 1C8Q(A) lCPU(A), 2CPU(A) lBSI lHNY 3CPU(A) 1B2Y(A) lDHK(A) 1J F H lPIF, 1PIG lOSE lHXO(A) lBVN(P) 1PPI Androgen receptor 1E3G(A) 1137(A), 1138(A) Anticoagulant protein C lAUT(C) Aquaporin 1IH5(A) 1FQWA) Resolution Source Green mamba Green mamba Electric eel Mouse Mouse Mouse Mouse Human Human Human Human Human Human Human Human Kidney bean Pig Pig Pig Pig S tendae Human Rat Human Human Human Homology Year Reference P-Amyloid lMWP(A) p-Lactamase[Sal lBTL lFQG(A) lJTD(A) lHTZ(A) lERM(A), lERO(A) lERQ(A) lXPB lESU(A) 1BT5(A) lTEM 1CK3(A) lAXB 0-Tubulin lJFF(B) lTUB(B) lFFX(B) Calcineurin A lTCO(A) lAUI(A) Carbonic anhydrase lHEA, 4CAC, 5CAC 1G6V(A) lCNw, lCNx, 1CNY 1IF4(A), 1IF5(A), 1IF6(A) 1IF9(A) lCA.3, lHEB, lHED lDCA, 1DCB lCRA lCIL, lCIM, lCIN lCAY lRZA, lRZB, IRZC, lRZD,1RZE 2CA2 lBNl,lBN3,1BN4,1BNM lCAH Human Bacteria Bacteria Bacteria Bacteria Bacteria Bacteria Bacteria Bacteria Escherichia coli Escherichia coli Escherichia coli Escherichia coli Bovine Pig Rat Bovine Human Human, HSV-1 Arabian camel Human Human Human Human Human Human Human, HSV-1 Human Human Human Human Human Table 11.1 (Continued) Target and PDB Reference 118Z(A) 1BV3(A) 12CA 1G53(A) 1AM6 lCAN, lCA0 lGOE(A), lGOF(A) lAVN lUGF lHVA 5cA2 lHCA 4CA2,6CA2,7CA2,9CA2 lZNC(A) Catechol methyltransferase lVID Cholecystokinina receptor 1D6G(A) Coagulation factor 10 lEZQ(A), lFOR(A),lFOS(A) 1C5M(D) lxKA(C), lxKB(C) lFAX(A) lFJS(A) lKIG(H) lHCG(A) Coagulation factor 1AI8(H) lMKW(K), lMKX(K) lBTH(H) lHXF(H) 1G30(B) 1A3E(H) 1D3P(B), 1D3Q(B) lHDT(H) 1AD8(H) Source Homology Human Human Human Human Human Human rhinovirus Human Human Human Human Human 100% 100% 99% 100% 100% 100% 99% 100% 99% 99% 99% 100% 100% 100% Resolution 1.93 A 1.85 A 2.40 A 1.94 A 2.10 A 1.90 A 1.60 A 2.00 A 2.00 A 2.30 A 2.10 A 2.30 A 2.10 A-2.80 2.80 A - A Human Human Rat NMR Human Human Human Human Human Soft tick - Hirudo rnedicinalis Bos taurus Bovine Hirudo medicinalis Hirudo medicinalis Hirudo medicinalis Hirudo rnedicinalis Hirudo medicinalis Hirudo medicinalis Year .b % lLHC(H), lLHF(H), lLHG(H) 1JOU(B) lDIT(H) lWS(H) 4THN(H) lTHP(B) lJOU(B) lDIT(H) 1WS(H) 4THN(H) lTHP(B) 1AY6(H) lClU(H), lClV(H) 1A4W(H) 1G37(A) lEOJ(A), lEOL(A) lBBO(B) 1C4U(2), 1D6W(A),1D9I(A), 1DOJ(A) 7KME(H) lQBV(H) 1DM4(B) lUMA(H) lBMM(H), lBMN(H) 1A2C(H) lFPC(H) lNRO(H), lNRR(H) lHAG(E) lHLT(H) 1TMU(H) 4HTC(H) lAK(H), lDWB(H), lDWC(H) 2HPP(H) lABI(H) Coagulation factor lJBU(H) Coagulation factor 7a 1QFWH) lDVA(H) Hirudomedicinalis Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Human Medicinal leech Medicinal leech M aeruginosa Hirudo medicinalis - Bacteria Human Human Table 11.1 (Continued) Target and PDB Reference lDAN(H) lCVW(H) lFAK(H) Coagulation factor lRFN(A) lPFX(C) Cox-1 1DWA) lCQE(A), lPRH(A) 1PTH lEBV(A) 1FE2(A) lEQG(A), lEQH(A), 1HT5(A), lHT\ill\(A) lPGE(A), lPGF(A), lPGG(A) Cox-2 lCW(A), lDDX(A) lCX2,3PGH, 4COX, 5COX, 6COX Cytochrome P450 reductase lBlC(A) lAMO(A) 1J9Z(A), lJAO(A), lJAl(A) Dihydrofolate reductase lBOZ(A) lHFP, lHFQ, 1HFR lOHJ, lOHK lDRl,lDR5,1DR6, 1DR7 1DR2,1DR3 1DR4 lDHF(A), 2DHF(A) lDLR, 1DLS 8DFR 1DRF Dihydroorotate dehydrogenase 1D3G(A), 1D3H(A) Resolution Source Homology Year Reference 2.00 A 2.28 A 2.10 A Human Human Human 100% 100% 100% 1997 1999 1998 (152) (153) (154) 2.80 A 3.00 A Human Pig 100% 88% 1999 1995 (155) (156) 3.00 A 3.10 A, 3.50 A 3.40 A 3.20 A 3.00 A 2.61 A-2.75 A 3.50 A, 4.50 A Sheep Sheep Sheep Sheep Sheep Sheep Sheep 93% 92% 92% 93% 92% 92% 92% 1999 1994 1995 2000 2000 2000 1995 (157) (158) (159) (160) (161) (162) (163) 2.40 A, 3.00 A 3.00 A Mouse Mouse 87% 87% 1999 1996 (164) (165) 1.93 A 2.60 A 2.70 A, 2.60 1.80 A Human 100% 93% 92% 1998 1997 2001 (166) (167) (168) 99% 100% 100% 75% 75% 75% 100% 99% 75% 100% 1998 1997 1997 1992 1992 1992 1989 1995 1989 1990 (169) (170) (171) (172) (173) (174) (175) (176) (177) (178) 100% 1999 (179) Rat Rat 2.10 A 2.10 A 2.50 A 2.20 A, 2.40 A 2.30 A 2.40 A 2.30 A 2.30 A 1.70 A 2.00 A Human Human Human 1.80 A Human 1.60 *, - - Dihydropteroate synthetase[Sal lADl(A), 1AD4(A) DNA helicase pcra[Sa] 1QHHW DNA topoisomerase 1EJ9(A) 1A36(A) 1A31(A), 1A35(A) Estrogen receptor l a lQKT(A), lQKU(A) lHCP 1A52(A) lERR(A), lERE(A) lHCQ(A) 3ERT(A), 3ERD(A) FK506-binding protein lTCO(C) lFKD, 2FKE lFKJ, lFKK, 1FKL lFAP(A) 3FAP(A), 4FAP(A) lNSG(A) lFKR, lFKS, 1FKT lEYM(A) 1BL4(A) 1D60(A), 1D7H(A), 1D7I(A), 1D7J(A) lQPF(A), lQPL(A) 1F40(A) 1B6C(A) 1A7X(A) lBKF 2FAP(A) 1C9H(A) IFKG, lFKH, lFKI(A) lFKF lFKB S aureus B thermophilus Human Human Human 2.20 A, 3.20 A NMR 2.80 A 2.60 A, 3.10 A 2.40 A 1.90 A, 2.03 A Human Human Human Human Human Human 2.50 A 1.72 A 1.70 & 2.20 A 2.70 A 1.85 A, 2.80 A 2.20 A NMR 2.00 A 1.90 A 1.85 A-1.90 A 2.50 A, 2.90 A NMR 2.60 A 2.00 A 1.60 A 2.20 A 2.00 A 2.00 A, 1.95 & 2.20 A 1.70 A 1.70 A Bovine Human Cow Human Human Human Human Human Human Human Human Human Human Human Human Human Human Table 11.1 (Continued) Target and PDB Reference e oa Follicle stimulating hormone 1FL7(B) GABA transferase lGTX(A) Glucocorticoid receptor lLAT(A) lGLU(A) Glutamate receptor lEWK(A), lEWT(A), lEWV(A) Glutathione peroxidase lGPl(A) G-CSF 1CD9(A), lPGR(A) lBGC, IBGD, lBGE(A) lGNC lRHG(A) Granulocyte-macrophage CSF lCSG(A) 2GMF(A) Growth hormone receptor 1A22(B) 1AmB) lHWG(B), lHWH(B) 3HHR(B) HIV reverse transcriptase lDLO 1RT3(B) lHPZ, lHQE, ' ~ H Q U lBQM, lBQN lTVR(B), lUWB(B) lEET lIKv, lIKw, lIKX, lIKY lHW(B) Resolution Source Homology Year Reference 3.00 A Human 99% 2000 (211) 3.00 A Pig 94% 1999 (212) 1.90 A 2.90 A Rat 85% 94% 1995 1992 (213) (214) 98% 2000 (215) 90% J u n 1985 (216) 2.20 - 3.70 A, 4.00 A 2.80 A, 3.50 A 1.70 A, 2.30 $2.20 NMR 2.20 A Rat A - Human Human Human Human - HIV-1 HIV-1 HIV-1 HIV-1 HIV-1 HIV-1 HIV-1 HIV-1 lClB(B) 2HMI(B) 1HYs 1HMV lHNI 1HNV lFKP(B) lJLA, lJLB, lJLC, lJLE, lJLF, 1JLG 1550, lQEl(B) 3HVT(B) Inosine monophosphate dehydrogenase lJRl(A) 1B30(A) Insulin-like growth factor 3LRI(A) lBQT 1IMXA) 1B9G(A) 2GF1,3GF1 Insulin-like growth factor receptor lIGR(A) lGAG(A) 1144(A) 1IR3(A) lIRK Insulin-like growth factor lIGL Integrin alpham lBHQ(l), lIDN(1) lJLM lIDO Intercellular adhesion molecule lIAM lICl(A) 1D3E(I), 1D3I(I), 1D3L(A) HW-1 pol virus Virus Virus Virus virus virus virus HIV-1 - Chinese hamster Human NMR NMR 1.82 A NMR NMR Human Human Human - Human Human Human Human - NMR Human Human Human Human Human Human rhinovirus Table 11.1 (Continued) Target and PDB Reference p rD o Interferon a lITF 1RH2(A) Interferon y 1FG9(A) lFYH(A) lEKU(A) lHIG(A) Interleukin 2ILA Interleukin receptor lGOY(R) 1IPAO lITB(B) Interleukin 10 lVLK 21LK lILK 1J7V(L) lINR Interleukin 12 1F42(A),1F45(A) Interleukin 13 1GA3(A) Interleukin lIRL 3INK(C) Interleukin 1JLI Interleukin lHIJ, lHIK lIAR(A) lHZI(A) lITM lBBN, 1BCN Resolution Source NMR 2.90 A Human Human 2.90 A 2.04 A 2.90 A 3.50 A Human Human Human 2.30 A 3.00 A 2.70 A 2.50 A Human Human Human 1.90 A 1.60 A 1.80 A 2.90 A 2.00 A Epstein-Barr virus Human Human Human Human 2.50 A, 2.80 A Human NMR Human NMR 2.50 A Human NMR Escherichia coli 3.00 2.60 A 2.30 A 2.05 A NMR NMR Human Human Human - Homology Year Reference e lCYL 2CYK lITL 21NT 1RCB lITI Interleukin lHUL(A) Interleukin 1IL6,2IL6 lALU Interleukin lIKL, 1IKM lICW(A) lILP(A), lILQ(A) 1QE6(A) lROD(A) 3IL8 1IL8(A),2IL8(A) Leukotriene A4 hydrolase 1HS6(A) Lipocortin I lAIN 1HM6(A) Luteinizing hormone P lQFW(B) lHCN(B) lHRP(B) Macrophage CSF lHMC(A) Neurarninidase[int B virus] lINF NMR NMR NMR 2.40 A 2.25 A NMR 2.40 A Human NMR 1.90 A Human Human NMR 2.01 A NMR 2.35 A NMR 2.00 A NMR Human Human Human Human Human 1.95 A Human 2.50 A 1.80 A Human Pig 3.50 A 2.60 A 3.00 A Escherichia coli - - 2.50 A 2.40 A 2.20 A 1.90 A 2.40 A 1.70 1.80 A 2.50 A, 2.40 A, 2.35 A 2.40 A 2.20 A Influenza b virus Influenza b virus - Influenza b virus - 94% 99% 94% 99% 99% 94% Table 11.1 (Continued) Target and PDB Reference \O N Neuropeptide Y lRON 1F8P(A) lFVN(A) Parathyroid hormone lHTH lFVY(A) lBWX, lHPY, lZWA, 1ZWC lETl(A) 1HPH lZWB, lZWD, lZWE, lZWF, lZWG PDGF p lPDG(A) Phospholipase A2 lBCI lRLW lCJY(A) Potassium channel shaker 1A68 lEOD(A), lEOE(A), lEOF(A) lTlD(A) lEXB(E) lDSX(A), lQDV(A), lQDW(A) PPAR y 4PRG(A) lPRG(A), 2PRG(A) 1FM6(D), 1FM9(D) 3PRG(A) Progesterone receptor 1E3K(A) 1A28(A) Proladin receptor 1BP3(B) 1F6F(B) Retinoic acid receptor lDSZ(A) lEXA(A), lEXX(A) Resolution Source NMR NMR NMR Human NMR NMR NMR 0.90 A NMR NMR Human Human Human Human Homology - - NMR 2.40 A 2.50 A Human Human Human Sea hare Sea hare Sea hare Rat Rat Escherichia coli Human Human Human Human Human Human Rat Human Human 97% 97% 99% 99% Year Reference P a W 2LBD 3LBD, 4LBD lDKF(B) lFCX(A), lFCY(A), lFCZ(A) lHRA Retinoid X receptor 1FM6(A), 1FM9(A) lDSZ(B) lDKF(A), 1LBD 2NLL(A) lRXR lGlU(A), 1G5Y(A) lFBY(A) 1BY4(A) Serotransferrin p 1JNF(A) Stem cell factor lEXZ(A) lSCF(A) Thymidine k i n a s e [ H W 1KWA) 10HI(A), 2KI5(A) 2.00 A 2.40 A 2.50 A 1.47 A, 1.30 1.38 A NMR Human 2.10 A 1.70 A 2.50 2.70 A 1.90 A NMR 2.50 2.00 A 2.25 A 2.10 A Human Human Human Human Human Human Human Human *, Human Human Human Rabbit Human Human lKIZ,lKI3,1KI4,1KI6,lKI7,1KI8 lWK, 2WK, W K 1E2H(A), 1E2I(A), 1E2J(A) 1E2M(A), 1E2N(A), 1E2P(A) 1E2K(A), 1E2L(A) Tumor necrosis factor receptor lNCF(A) lEXT(A) lTNR(R) Vitamin D receptor 1IE8(A), 1IE9(A) lDBl(A) Xanthine-guanine phosphoribosyltransferase 1A95(A), 1A97(A), 1A98(A) lNUL(A) 1A96(A) Human Human - Human Human Escherichia coli Escherichia coli Escherichia coli X-Ray Crystallography in Drug Discovery sensitive sequence methods [such as ISS (37511 fail to identify many homologous relationships Structure is more conserved than sequence, so structural classification schemes (SCOP, CATH) have been a valuable method to assign proteins to functional groups A now classic example of functional understanding from structural homology was the discovery that the Bcl-2 family of apoptosis proteins are homologous to pore-forming toxins (376) This finding led to the suggestion that Bcl-2 proteins may function by perforating mitochondrial membranes, and has since opened new avenues of fruitful research In addition to structural homology as assessed by global similarities, local structural features can give clues to structure even when proteins are not homologous By identifying surface clusters of polar residues that are well conserved in the sequence family it is possible to identify likely functional sites even when there is no obvious structural homology These three-dimensional motifs can be compared with a structure database to identify similar motifs with known function A classic example is found among the serine proteases Chyrnotrypsin and subtilisin share a similar catalytic triad (His-Asp-Ser) but are otherwise unrelated structurally The PLP-dependent enzymes are famed for the diversity of both structure and function, but even among this group, common structural motifs seem to have evolved convergently (377) Simply searching for large clefts in the protein surface turns out to be an extremely successful method to identify active sites Nucleic acid binding functions can be particularly obvious from an analysis of surface electrostatics (378,379).Mice homozygous for tubby loss-offunction mutations show an obese phenotype, and therefore, the tubby protein has attracted considerable interest However, years after the initial cloning of the tubby gene (380,3811, the molecular function of the protein was still a mystery The structure of the conserved Cterminal domain of tubby was determined by X-ray crystallography, and a large groove of highly positive charge immediately led to the hypothesis that the protein acted as a transcription factor (6) The search is now on for downstream targets of tubby, and further structural work has demonstrated a new role for the tubby protein in G-protein-coupled re- ceptor (GPCR)-mediated signal transduction (382) This on-going story demonstrates the power of structural approaches to determine function In tackling structures of proteins of unknown function bound metal ions, natural substrates or even serendipitously bound small molecules arising from the crystal preparation (e.g., buffers) often suggests the location of an active site If the side-chains contributing to binding are well conserved, then this is good evidence locating an active site and helps assess the "drugability" of the protein The recent structure of LuxS illustrates the power of this approach (373) A number of genes had been identified that are required for quorum sensing in bacteria by system Quorum sensing by the widely conserved system has emerged as an intriguing mechanism by which bacteria monitor their density and seems to be an important component of the progression to virulence, at least in certain pathogens LuxS is the product of one of the genes required for system 2, but nothing was known of the molecular function of LuxS in this pathway Disruption of this pathway has promise in antibacterial drug design, but whether LuxS would be an attractive target for small molecule design was unclear No information was available to develop a biochemical assay, and besides, it was not clear what kind of library to use in high-throughput screening The structure of LuxS was solved a t Structural GenomiX in less than months, and there are representative X-ray structures from three different bacteria The structure showed that LuxS forms a dimer in which each monomer has a zinc ion coordinated by a HisHis-Cys triad and water molecule Non-covalently bound methionine molecules were found to have bound in a pocket formed at the dimer interface and close to the zinc ions (see Fig 11.7, a and b) Methionine was shown to have bound as an artefact of the purification procedure With this information, it became immediately clear that LuxS is likely a zinc metalloenzyme, and a hypothesis for the likely physiological substrate emerged from molecular modeling studies of the methionine binding site This example illustrates how structure could rapidly accelerate an early stage project providing the starting point for assay development and selection of an appropriate Structural Cenomics Figure 11.7 (a) The likely active site of LwrS identified by searching for clusters of polar, conserved residues in the structure (b) Structure of the LuxS monomer highlighting the bound zinc ion (magenta) and methionine (green) See color insert screening library (in this case metalloenzyme inhibitor libraries would be desirable) The model of the likely substrate bound to the active site suggests further experiments to test this hypothesis and even provides a starting point for medicinal chemistry exploration 4.3 Pathways Increasingly in drug discovery, particular molecular pathways are attracting interest in drug design and often manipulation of any of a number of pathway components would achieve the same end Pathways controlling apoptosis, the cell cycle, and inflammation all contain multiple biologically validated targets In microbial disease, several biosynthetic pathways, such as peptidoglycan biosynthesis and translation, are the targets of current drugs and several new pathways are promising targets for the development of novel agents Comprehensive, high-resolution structural information of multiple pathway components provides a basis for the rational design of inhibitors targeting the pathway Interfering with the function of anv " of a number of enzymes of a pathway may have equally beneficial therapeutic value Despite this, some enzymes may be more tractable targets for the design of inhibitors than others Access to high resolution structural information of all the components of a therapeutically relevant pathway enables the rational choice of the best-suited target(s1 to pursue for the design of agonists and antagonists This choice may depend on such pragmatic considerations as the access to libraries targeted to particular enzyme types and available synthetic chemistry expertise Furthermore, comparison of the binding pockets of consecutive enzymes in the pathway that bind similar (or identical) substrates and products may even enable the design of inhibitors of multiple pathway components Such a compound may be particularly desirable in the development of novel anti-microbial and cancer agents where compound re-, sistance can rapidly emerge The evolution of resistance to a drug that inhibits two consecutive enzymes in an essential pathway is theoretically much less probable than evolution of resistance to a single enzyme inhibitor The non-mevalonate isopentenyl pyrophosphate biosynthesis pathway has attracted attention in recent years as a novel target for the design of anti-microbial inhibitors (383) At Structural GenomiX, the structures of three consecutive enzymes in this pathway have been solved There is now a clear understanding of which pathway components may be most tractable to inhibitor discovery, which likely have least structural homology to human proteins, and even how to go about the design of pan-pathway inhibitors 4.4 Protein Structure Modeling An aim of structural genomics efforts, to provide high quality three-dimensional structures for every protein sequence, will not be achieved by experimental approaches alone X-Ray Crystallography in Drug Discovery Many consortiurns are selecting targets for Xray crystallography that would provide the templates for comparative modeling techniques of all other sequences (384, 385) As more structures are determined by NMR and X-ray crystallography, the quality of the models will improve simply because more similar templates will become available but also because and new methods for loop modeling and ab initio structure prediction will undoubtedly emerge (386, 387) Efforts are also underway both in industry and academia to assemble databases of homology models for all sequences that can be reasonably well modeled (388) CONCLUSION Anyone who is involved or interested in drug discovery will recognize the potential of protein crystallography to greatly enhance the process Whether this promise has been met to date is the subject of considerable debate What is certain, however, is that in the very near future the advances in crystallography technology will render this question moot The histograms on the PDB website (27, 28) that show the increasing rate of structures deposited over the last decade are a startling visual indicator of the revolution that is occurring in the field Clearly, the impact will be felt in drug discovery very soon and perhaps very dramatically, and it serves the audience of this series to be well informed of these advances in technology and their subtle limitations It is tempting to draw analogy with the development of other analytical technologies (NMR, FAB-MS) and conclude that protein crystallography will soon leave the incubator of "big machine physics" to become an everyday, routine tool used in the medicinal chemistry laboratory Hopefully, this chapter has shown some of the subtle complexities of sample preparation and handling, data collection, and refinement, etc that temper this vision and will likely keep this a specialized field for some time REFERENCES W A Hendrickson, Science, 254,51-58(1991) T C Tenvilliger, Nat Struct Biol., 7,935-939 (2000) L.Pauling, Lecture presented at the International Congress of X-ray Crystallography at Stonybrook, NY, August 1973 J Jancarik and S H Kim, J Appl Cryst., 24, 409-411(1991) Hampton Research, available online at http:// www.hamptonresearch.com,accessed on October 9,2001 G L.Gilliland, M Tung, D M Blakeslee, and J Ladner, Biological Macromolecule Crystallization Database (BMCD), available online at http://www.bmal.nist.gov:8O8O/bmcdmmal.html, accessedon October9,2001 N.E Chayen, Structure, 5,1269-1274(1997) I Jurisica, et al., IBMSystems J.,40,394-409 (2001) A D'Arcy, et al.,J Cryst Growth, 168, 175180 (1996) 10 C W Carter Jr., Methods Enzymol., 276, 74-99(1997) 11 A C.Bloomer and U.W Arndt, Acta Crystallogr D Biol Crystallogr., 55(Pt lo),1672-1680 (1999) 12 (a) G A Petsko, J Mol Biol., 96, 381-392 (1975);(b) R L.Sutton, J Chem Soc Faraday Trans., 1, 101-105 (1991);(c) D W Rodgers, Structure, 2,1135-1140(1994) 13 D W Green, V M Ingram and M F Perutz, Proc R Soc A, 225,287-307(1954) 14 M G Rossman and D M Blow, Acta Cryst., 15, 24-34(1962) 15 L M Rice, T N Earnest, A T Brunger, Acta Cryst D., 56,1413-1420(2000) 16 W.A Hendrickson, et al., EMBO J.9,16651672(1990) 17 W I Weis, et al., Science, 254, 1608-1615 (1991) 18 W M Clemons, Jr., et al., J Mol Biol 310, 827-843(2001) 19 Collaborative Computational Project, Number 4,Acta Cryst D,50,760-763(1994) 20 Z.Otwinowski and W Minor, available online a t http://www.hkl-xray.com,accessed October 9,2001 21 T.Terwilliger, Automated Structure Solution for MIR and MAD, available online at http:// www.solve.lanl.gov, accessed October 9,2001 22 C M Weeks, S A Potter, J Rappleye, R Miller, available online a t http://www.hwi buffalo edu/SnB, accessed on October 9,2001 23 G M Sheldrick, available online at http:// shekuni-ac.gwdg.de/SHEW, accessed on October 9,2001 References 24 V S Lamzin and A Perrakis, available online accessed at http://www.embl-hamburg.de/ARP, on October 9,2001 25 A Jones and M Kjeldgaard, available online at http://www.imsb.au.dk/-moWo, accessed on October 9,2001 26 A T Brunger, P D Adams, G M Clore, W L Delano, P Gros, R W Grosse-Kunstleve, J.4 Jiang, J Kuszewski, M Nilges, N S Pannu, R J Read, L M Rice, T Simonson, and G L Warren, Crystallography and NMR System, available online at http://cns.csb.yale.edu/vl.0, accessed on October 9,2001 27 H M Berman, D S Goodsell, and P E Bourne, Am Scientist, 90,350-359 (2002) 28 H M Berman, J Westbrook, Z Feng, G Gilliland, T N Bhat, H Weissig, I N Shindyalov, P E Bourne, The Protein Data Bank, Nucleic Acids Res., 28,235-242 (2000) 29 Information on how to obtain this database is available at: http://www.ccdc.cam.ac.uW prods/ 30 J Drews and S Ryser, Nature Biotechnol.15, (1997) 31 Y Bourne, P Taylor, and P Marchot, Cell, 83, 503 (1995) 32 G Kryger, M Harel, M Harel, A Shafferman, I Silman, and J L Sussman,Acta Crystallogr., Sect D, 56, 1385 (2000) 33 Y Bourne, J Grassi, P E Bougis, and P Marchot, J Biol Chem., 274,3370 (1999) 34 Y Bourne, P Taylor, P E Bougis, and P Marchot, J Biol Chem., 274,2963 (1999) 35 V Sideraki, K A Mohamedali, D K Wilson, Z Chang, R E Kellems, F.A Quiocho, and F B Rudolph, Biochemistry, 35,7862 (1996) 36 Z Wang and F A Quiocho, Biochemistry, 37, 8314 (1998) 37 V Sideraki, D K Wilson, L C Kurz, F A Quiocho, and F B Rudolph, Biochemistry, 35, 15019 (1996) 38 D K Wilson, F B Rudolph, and F A Quiocho, Science, 252, 1278 (1991) 39 D K Wilson and F A Quiocho, Biochemistry, 32, 1689 (1993) 40 N Ramasubbu, C Ragunath, and Z Wang, In press 41 N Rarnasubbu, P Venugopalan, Y Luo, G D Brayer, and M J Levine, In press 42 N Ramasubbu, K Sekar, and D Velmurugan, Acta Crystallogr D Biol Crystallogr., 52, 435 (1996) 43 G D Brayer, G Sidhu, R Maurus, E H Rydberg, C Braun, Y Wang, N T Nguyen, C M Overall, and S G Withers, Biochemistry, 39, 4778 (2000) 44 E H Rydberg, G Sidhu, H C Vo, J Hewitt, H C Cote, Y Wang, S Numao, R T Macgillivray, C M Overall, G D Brayer, and S G Withers, Protein Sci., 8,635 (1999) 45 G D Brayer, Y Luo, and S G Withers, Protein Sci., 4, 1730 (1995) 46 G D Brayer, G Sidhu, R Maurus, E H Rydberg, C Braun, Y Wang, N T Nguyen, C M Overall, and S G Withers, Biochemistry, 39, 4778 (2000) 47 M Qian, R Haser, G Buisson, E Duee, and F Payan, Biochemistry, 33,6284 (1994) 48 C Bompard-Gilles, P Rousseau, P Rouge, and F Payan, Structure (Lond), 4, 1441 (1996) 49 M Qian, S Spinelli, H Driguez, and F Payan, Protein Sci., 6, 2285 (1997) 50 M Machius, L Vertesy, R Huber, and G Wiegand, J Mol Biol., 260,409 (1996) 51 C Gilles, J P h t i e r , G Marchis-Mouren, C Cambillau, and F Payan, Eur J Biochem., 238,561 (1996) 52 M Qian, V Nahoum, J Bonicel, H Bischoff, B Henrissat, and F Payan, Biochemistry, 40, 7700 (2001) 53 G Wiegand, Epp, and R Huber, J Mol Biol., 247,99 (1995) 54 M Qian, R Haser, G Buisson, E Duee, and F' Payan, Biochemistry, 33,6284 (1994) 55 P M Matias, P Donner, R Coelho, M Thomaz, C Peixoto, S Macedo, N Otto, S Joschko, P Scholz, A Wegg, S Basler, M Schafer, U Egner, and M A Carrondo, J.Biol Chem., 275,26164 (2000) 56 J S Sack, K F Kish, C Wang, R M Attar, S E Kiefer,Y Ang.Y Wu, J E Schemer, M E Salvati, S R Krystekjr., R Weinmann, and H M Einspahr, Proc Nut h a d Sci USA, 98, 4904 (2001) 57 T Mather, V Oganessyan, P Hof, R Huber, S Foundling, C Esmon, and W Bode, Embo J., 15,6822 (1996) 58 G Ren, V S Reddy, A Cheng, P Melnyk, and A K Mitra, Proc Nut Acad Sci USA, 98, 1398 (2001) 59 K Murata, K Mitsuoka, T Hirai, T Walz, P A g e , J B Heyrnann, A Engel, and Y Fujiyoshi, Nature, 407,599 (2000) 60 J Rossjohn, R Cappai, S C Feil, A Henry, W J Mckinstryd Galatis, L Hesse, G Mul- X-Ray Crystallography in Drug Discovery thaup, K Beyreuther, C L Masters, andM W Parker, Nut Struct Biol., 6, 327 (1999) 61 C Jelsch, L Mourey, J M Masson, and J P Samama, Proteins, 16, 364 (1993) 62 N C Strynadka, H Adachi, S E Jensen, K Johns, A Sielecki, C Betzel, K Sutoh, and M N James, Nature, 359, 700 (1992) 63 D C Lim, H U Park, L Decastro, S G Kang, H S Lee, S Jensen, K J Lee, and N C J Strynadka, Nut Struct Biol., 8,848 (2001) 64 M C Orencia, J S Yoon, J E Ness, W P Stemmer, and R C Stevens, Nut Struct Biol., 8, 238 (2001) 65 S Ness, R Martin, A M Kindler, M Paetzel, M Gold, J B Jones, and N C J Strynadka, Biochemistry, 39, 5312 (2000) 66 E Fonze, P Charlier, Y To'Th, M Vermeire, X Raquet, and A Dubus, Acta Crystallogr., Sect D, 61,682 (1995) 67 E Fonze, P Charlier, Y To'Th, M Vermeire, X Raquet, and A Dubus, Acta Crystallogr., Sect D, 51,682 (1995) 68 L Maveyraud, L Mourey, L P Kotra, J D Pedelacq, V Guillet, S Mobashery, and J P Samama, J Am Chem Soc., 120,9748 (1998) 69 L Maveyraud, I Massova, C Birck, K Miyashita, J P Samama, and S Mobashery, J Am Chem Soc., 118, 7435 (1996) 70 P Swaren, D Golemi, S Cabantous, A Bulychev, L Maveyraud, S Mobashery, and J P Samama, Biochemistry, 38,9570 (1999) 71 L Maveyraud, R F Pratt, and J P Samama, Biochemistry, 37,2622 (1998) 72 J Lowe, H Li, K H Downing, and E Nogales, J Mol Biol., 313, 1045 (2001) 73 E Nogales, S G Wolf, and K H Downing, Nature, 391, 199 (1998) 74 B Gigant, P A Curmi, C Martin-Barbey, E Charbaut, S Lachkar, L Lebeau, S Siavoshian, A Sobel, and M Knossow, Cell, 102, 809 (2000) 75 J P Griffith, J L Kim, E E Kim, M D Sintchak, J A Thomson, M J Fitzgibbon, M A Fleming, P R Caron, K Hsiao, and M A Navia, Cell, 82, 507 (1995) 76 C R Kissinger, H E Parge, D R Knighton, C T Lewis, L A Pelletier, A Tempczyk, V J Kalish, K D Tucker, R E Showalter, E W Moomaw, L N Gastinel, N Habuka, X Chen, F Maldonado, J E Barker, R Bacquet, and J E Villafranca, Nature, 378,641 (1995) 77 A Desmyter, K Decanniere, S Muyldermans, and L Wyns, In press 78 P A Boriack, D W Christianson, J KingeryWood, and G M Whitesides, J Med Chem., 38,2286 (1995) 79 C.-Y Kim and D W Christianson, In press 80 B A Grzybowski, A V Ishchenko, C.-Y Kim, G Topalov, R Chapman, D W Christianson, G M Whitesides, and E I Shakhnovich, Proc Nut Acad Sci USA, 99,1270 (2002) 81 S K Nair and D W Christianson, Biochemistry, 32,4506 (1993) 82 J A Ippolito and D W Christianson, Biochemistry, 32,9901 (1993) 83 S Mangani and A Liljas, J Mol Biol., 232,9 (1993) 84 G M Smith, R S Alexander, D W Christianson, B M McKeever, G S Ponticello, J P Springer, W C Randall, J J Baldwin, and C N Habecker, Protein Sci., 3,118 (1994) 85 K Hakansson, C Briand, V Zaitsev, Y Xue, and A Liljas, Acta Crystallogr D Biol Crystallogr., 50, 101 (1994) 86 K Hakansson, A Wehnert, and A Liljas, Acta Crystallogr D Biol Crystallogr., 50,93 (1994) 87 A E Eriksson, P.M Kylsten, T A Jones, and A Liljas, Proteins, 4,283 (1988) 88 P A Boriack-Sjodin, S Zeitlin, H H Chen, L Crenshaw, S Gross, A Dantanarayana, P Delgado, J A May, T Dean, and D W Christianson, Protein Sci., 7, 2483 (1998) 89 K Hakansson and A Wehnert, J Mol Biol., 228, 1212 (1992) 90 C.-Y Kim, D A Whittington, J S Chang, J Liao, J A May, and D W Christianson, J Med Chem., 45, 888 (2002) 91 F Briganti, S Mangani, A Scozzafava, G Vernaglione, and C T Supuran, J Biol Znorg Chem., 4, 528 (1999) 92 S K Nair, T L Calderone, D W Christianson, and C A Fierke, J Biol Chem., 266, 17320 (1991) 93 C.-Y Kim, J S Chang, J B Doyon, T T Bairdjr., C A Fierke, A Jain, and D W Christianson, J.Am Chem Soc., 122,12125 (2000) 94 L R Scolnick, A M Clements, J Liao, L Crenshaw, M Hellberg, J May, T R Dean, and D W Christianson, J Am Chem Soc., 119,850 (1997) 95 S Mangani and K Hakansson, Eur J Biochem., 210,867 (1992) 96 D Duda, C Tu, M Qian, P Laipis, M Agbandje-Mckenna, D N Silverman, and R Mckenna, Biochemistry, 40,1741 (2001) 97 F Briganti, S Mangani, P Orioli, A Scozzafava, G Vernaglione, and C T Supuran, Biochemistry, 36,10384 (1997) 98 L R Scolnick and D W Christianson, Biochemistry, 35, 16429 (1996) D W Christianson, Biochemistry, 32, 1510 (1993) 100 J F Krebs, C A Fierke, R S Alexander, and D W Christianso, Biochemistry, 30, 9153 (1991) 101 S K Nair and D W Christianson, J Am Chem Soc., 113,9455 (1991) 102 R S Alexander, S K Nair, and D W Christianson, Biochemistry, 30,11064 (1991) 103 T Stams, S K Nair, T Okuyama, A Waheed, W S Sly, and D W Christianson, Proc Nat Acad Sci USA, 93,13589 (1996) 104 J Vidgren, L A Svensson, and A Liljas, Nature, 368, 354 (1994) 105 M Pellegrini and D F Mierke, Biochemistry, 38,14775 (1999) 106 S Maignan, J P Guilloteau, S Pouzieux, Y M Choi-Sledeski, M R Becker, S I Klein, W R Ewing, H W Pauls, A P Spada, and V Mikol, J Med Chem., 43,3226 (2000) 107 B A Katz, R Mackman, C Luong, K Radika, A Martelli, P A Sprengeler, J Wang, H Chan, and L Wong, Chem Biol., 7,299 (2000) 108 K Kamata, H Kawamoto, T Honma, T Iwama, and S H Kim, Proc Nat Acad Sci USA, 95,6630 (1998) 109 H Brandstetter, A Kuhne, W Bode, R Huber, W Vondersaal, K Wirthensohn, and R A Engh, J Biol Chem., 271, 29988 (1996) 110 M Adler, D D Davey, G B Phillips, S H Kim, J Jancarik, G Rumennik, D L Light, and M Whitlow, Biochemistry, 39, 12534 (2000) 111 A Wei, R Alexander, J Duke, H Ross, S Rosenfeld, and C.-H Chang, J.Mol Biol., 283, 147 (1998) 112 K Padmanabhan, K P Padmanabhan, A Tulinsky, C H Park, W Bode, R Huber, D T Blankenship, A D Cardin, and W Kisiel, J Mol Biol., 232,947 (1993) 113 M G Malkowski, P D Martin, J C Guzik, and B F P Edwards, Protein Sci., 6, 1438 (1997) 114 A Vandelocht, W Bode, R Huber, B F Lebonniec, S R Stone, C T Esmon, and M T Stubbs, Embo J.,16,2977 (1997) 115 E Zhang and A Tulinsky, Biophys Chem., 63, 185 (1997) 116 H Nar, M Bauer, A Schmid, J Stassen, W Wienen, H W Priepke, I K Kauffmann, U J Ries, and N H Hauel, Structure, 9,29 (2001) 117 A Zdanov, S Wu, J DiMaio, Y Konishi, Y Li, X Wu, B F Edwards, P D Martin, and M Cygler, Proteins, 17,252 (1993) 118 N Y Chirgadze, D J Sall, S L Briggs, D K Clawson, M Zhang, G F Smith, and R W Schevitz, Protein Sci., 9, 29 (2000) 119 L Tabernero, C Y Chang, S Ohringer, W F Lau, E J Iwanowicz, W.-C Han, T C Wang, S M Seiler, D G M Roberts, and J S Sack, J Mol Biol., 246, 14 (1995) 120 J A Malikayil, J P Burkhart, H A Schreuder, R J Broersmajunior, C Tardif, L W Kutcheriii, S Mehdi, G L Schatzman, B Neises, and N P Peet, Biochemistry, 36,1034 (1997) 121 P C Weber, S L Lee, F A Lewandowski, M C Schadt, C W Chang, and C A Kettner, Biochemistry, 34, 3750 (1995) 122 J A Huntington and C T Esmon, In press 123 R Krishnan, A Tulinsky, G P Vlasuk, D Pearson, P Vallar, P Bergum, T K Brunck, and W C Ripka, Protein Sci., 5, 422 (1996) 124 R A Engh, H Brandstetter, G Sucher, A Eichinger, U Baumann, W Bode, R Huber, T Poll, R Rudolph, and W Vondersaal, Structure (Lond), 4, 1353 (1996) 125 A Lombardi, G Desimone, F Nastri, S Galdiero, R Dellamorte, N Staiano, C Pedone, M Bolognesi, and V Pavone, Protein Sci., 8, 91 (1999) 126 E Guinto, S Caccia, T Rose, K Futterer, G Waksman, and E Dicera, Proc Nat Acad Sci USA, 96,1852 (1999) 127 B E Maryanoff, X Qiu, K P Padmanabhan, A Tulinsky, H R Almondjunior, P AndradeGordon, M N Greco, J A Kauffman, K C Nicolaou, A Liu, P H Brungs, and N Fusetani, Proc Nat Acad Sci USA, 90, 8048 (1993) 128 B A Katz, J M Clark, J.S Finer-Moore, T E Jenkins, C R Johnson, M J Ross, C Luong, W R Moore, and R M Stroud, Nature, 391, 608 (1998) 129 J H Matthews, R Krishnan, M J Costanzo, B E Maryanoff, and A Tulinsky, Biophys J., 71,2830 (1996) 130 B Bachand, M Tarazi, Y St-Denis, J J Edmunds, P D Winocour, L Leblond, and M A Siddiqui, Bioorg Med Chem Lett., 11, 287 (2001) 131 J J Slon-Usakiewicz, J Sivaraman, Y Li, M Cygler, and Y Konishi, Biochemistry, 39,2384 (2000) 132 R Krishnan, E Zhang, K Hakansson, R K Arni, A Tu1inskym.S Lim-Wilby, E Levy, J E Semple, and T K Brunck, Biochemistry, 37,12094 (1998) X-Ray Crystallography in Drug Discovery 133 R Krishnan, I Mochalkin, R Arni, and A Tulinsky, Acta Crystallogr., Sect D, 56, 294 (2000) 134 I Mochalkin and A Tulinsky, Acta Crystallogr., Sect D, 55, 785 (1999) 135 R Bone, T Lu, C R Illig, R M Soll, and J C Spurlino, J Med Chem., 41,2068 (1998) 136 R Krishnan, E J Sadler, and A Tulinsky, Acta Crystallogr., Sect D, 56,406 (2000) 137 M Nardini, A Pesce, M Rizzi, E Casale, R Ferraccioli, G Balliano, P Milla, P Ascenzi, and M Bolognesi, J Mol Biol., 258, 851 (1996) 138 M F Malley, L Tabernero, C Y Chang, S L Ohringer, D G Roberts, J Das, and J S Sack, Protein Sci., 5,221 (1996) 139 J L R Steiner, M Murakami, and A Tulinsky, J Am Chem Soc., 120,597 (1998) 140 I I Mathews and A Tulinsky, Acta Crystallogr D Biol Crystallogr., 51,550 (1995) 141 I I Mathews, K P Padmanabhan,V Ganesh, A Tulinsky, M Ishil, J Chen, C W Turck, and S R Coughlin, Biochemistry, 33, 3266 (1994) 142 J Vijayalakshmi, K P Padmanabhan, K G Mann, and A Tulinsky, Protein Sci., 3, 2254 (1994) 143 I Mathews, K P Padmanabhan, A Tulinsky, and J E Sadler, Biochemistry, 33,13547 (1994) 144 J P Priestle, J Rahuel, H Rink, M Tones, and M G Gruetter, Protein Sci., 2, 1630 (1993) 145 T J Rydel, A Tulinsky, W Bode, and R Huber, J Mol Biol., 221, 583 (1991) 146 D W Banner and P Hadvary, J Biol Chem., 266,20085 (1991) 147 R K Arni, K Padmanabhan, K P Padmanabhan, T-P Wu, and A Tulinsky, Biochemistry, 32,4727 (1993) 148 X Qiu, K P Padmanabhan, V E Carperos, A Tulinsky, T Kline, J M Maraganore, and J W Fentonii, Biochemistry, 31,11689 (1992) 149 C Eigenbrot, D Kirchhofer, M S Dennis, L Santell, R A Lazarus, J Stamos, and M H Ultsch, Structure, 9, 627 (2001) 150 A C W Pike, A M Brzozowski, S.M Roberts, H Olsen, and E Persson, Proc Nat Acad Sci USA, 96,8925 (1999) 151 M S Dennis, C Eigenbrot, N J Skelton, M H Ultsch, L Santell, M L Dwyer, M P O'Connell, and R A Lazarus, Nature, 404, 465 (2000) 152 D W Banner, A D'Arcy, C Chene, F K Winkler, A Guha, W H Konigsberg, Y Nemerson, and D Kirchhofer, Nature, 380,41 (1996) 153 G Kemball-Cook, D J D Johnson, E G D Tuddenham, and K Harlos, J Struct Biol., 127,213 (1999) 154 E Zhang, R Stcharles, and A Tulinsky, J Mol Biol., 285,2089 (1999) 155 K.-P Hopfner, A Lang, A Karcher, K Sichler, E Kopetzkih Brandstetter, R Huber, W Bode, and R A Engh, Structure (Lond), 7,989 (1999) 156 H Brandstetter, M Bauer, R Huber, P Lollar, and W Bode, Proc Nat Acad Sci USA, 92,9796 (1995) 157 M G Malkowski, S L Ginell, W L Smith, and R M Garavito, Science, 289, 1933 (2000) 158 D Picot, P J Loll, and R M Garavito, Nature, 367, 243 (1994) 159 P J Loll, D Picot, and R M Garavito, Nat Struct Biol., 2,637 (1995) 160 P J Loll, C T Sharkey, S J O'Connor, C M Dooley, E O'Brien, M Devocelle, K B Nolan, B S Selinsky, and D J Fitzgerald, Mol Pharmacol., 60, 1407 (2001) 161 E D Thuresson, M G Malkowski, K M Lakkides, C J Riekea.M Mulichak, S L Ginell, R M Garavito, and W L Smith, J Biol Chem., 276, 10358 (2001) 162 B S Selinsky, K Gupta, C T Sharkey, and P J Loll, Biochemistry, 40, 5172 (2001) , 163 P J Loll, D Picot, Ekabo, and R M Garavito, Biochemistry, 35, 7330 (1996) 164 J R Kiefer, J L Pawlitz, K T Moreland, R A Stegeman, J K Gierse, W F Hood, J K Gierse, A M Stevens, D C Goodwin, S W Rowlinson, L J Marnett, W C Stallings, and R G Kurumbail, Nature, 405,97 (2000) 165 R G Kurumbail, A M Stevens, J K Gierse, J J Mcdonald, R A Stegeman, J Y Pak, D Gildehaus, J M Miyashiro, T D Penning, K Seibert, P C Isakson, and W C Stallings, Nature, 384,644 (1996) 166 Q Zhao, S Modi, G Smith, M Paine, P D Mcdonagh, C R Wolf, D Tew, L Y Lian, G C Roberts, and H P Driessen, Protein Sci., 8, 298 (1999) 167 M Wang, D L Roberts, R Paschke, T M Shea, B S Masters, and J J Kim, Proc Nat h a d Sci USA, 94,8411 (1997) 168 P A Hubbard, A L Shen, R Paschke, C B Kasper, and J J Kim, J Biol Chem., 276, 29163 (2001) 169 A Gangjee, A P Vidwans, A Vasudevan, S F Queener, R L Kisliuk, V Cody, R Li, N Gal- References itsky, J R Luft, and W Pangborn, J Med Chem., 41,3426 (1998) 170 V Cody, N Galitsky, J R Luft, W Pangborn, R L Blakley, and A Gangjee, J Mol Biol., 221,583 (1991) 171 V Cody, N Galitsky, J R Luft, W Pangborn, A Rosowsky, and R L Blakley, Biochemistry, 36,13897 (1997) 172 M A McTigue, J F Davies 11,B T Kaufman, and J Kraut, Biochemistry, 31,7264 (1992) 173 M A McTigue, J F Davies 11,B T Kaufman, and J Kraut, Biochemistry, 32,6855 (1993) 174 M A McTigue, J F Daviesii, B T Kaufman, N.-H Xuong, and J Kraut, In Press 175 J F Davies, T J Delcamp, N J Prendergast, V A Ashford, J H Freisheim, and J Kraut, Biochemistry, 29,9467 (1990) 176 W S Lewis, V Cody, N Galitsky, J R Luft, W Pangborn, S K Chunduru, H T Spencer, J R Appleman, and R L Blakley, J Biol Chem., 270,5057 (1995) 177 J F Davies, D A Matthews, S J Oatley, B T Kaufman, N.-H Xuong, and J Kraut, In press 178 C Oefner, A D'Arcy, and F K Winkler, Eur J Biochem., 174,377 (1988) 179 S Liu, E A Neidhardt, T H Grossman, T Ocain, and J Clardy, Structure (Lond), 8, 25 (2000) 180 I C Hampele, A D'Arcy, G E Dale, D Kostrewa, J Nielsenc Oefner, M G Page, H J Schonfeld, D Stuber, and R L Then, J Mol Biol., 268, 21 (1997) 181 P Soultanas, M S Dillingham, S S Velankar, and D B Wigley, J.Mol Biol., 290,137 (1999) 182 M R Redinbo, J J Champoux, and W G Hol, Biochemistry, 39, 6832 (2000) 183 L Stewart, M R Redinbo, X Qiu, W G Hol, and J J Champoux, Science, 279,1534 (1998) 184 M R Redinbo, L Stewart, P Kuhn, J J Champoux, and W G Hol, Science, 279,1504 (1998) 185 M Ruff, M Gangloff, S Eiler, S Duclaud, J M Wurtz, and M Dino, In press 186 J W R Schwabe, L Chapman, J T Finch, D Rhodes, and D Neuhaus, Structure (Lond), 1, 187 (1993) 187 D M Tanenbaum, Y Wang, S P Williams, and P B Sigler, Proc Nut Acad Sci USA, 95, 5998 (1998) 188 A M Brzozowski, A C W Pike, Z Dauter, R E Hubbard, T Bonn, Engstrom, L Ohman, G L Greene, J.-A Gustaffson, and M Carlquist, Nature, 389, 753 (1997) 189 J W R Schwabe, L Chapman, J T Finch, and D Rhodes, Cell, 75,567 (1993) 190 A K Shiau, D Barstad, P M Loria, L Cheng, P J Kushner, D A Agard, and G L Greene, Cell, 95,927 (1998) 191 J W R Schwabe, L Chapman, J T Finch, and D Rhodes, Cell, 75, 567 (1993) 192 A K Shiau, D Barstad, P.M Loria, L Cheng, P J Kushner, D A Agard, and G L Greene, Cell, 95,927 (1998) 193 K P Wilson, M M Yamashita, M D Sintchak, S H R0tsteinm.A Murcko, J Boger, J A Thomson, M J Fitzgibbon, and M A Navia, Acta Crystallogr., Sect D, 51, 511 (1995) 194 J Choi, J Chen, S L Schreiber, and J Clardy, Science, 273,239 (1996) 195 J Liang, J Clardy, and J Choi, Acta Crystallogr., Sect D, 55, 736 (1999) 196 J Liang, J Choi, and J Clardy, Acta Crystallogr D Biol Crystallogr., 55, 736 (1999) 197 S W Michnick, M K Rosen, T J Wandless, M Karplus, and S L Schreiber, Science, 252, 836 (1991) 198 C T Rollins, V M Rivera, D N Woolfson, T Keenan, M Hatada, S E Adams, L J Andrade, D Yaeger, M R Vanschravendijk, D A Holt, M Gilman, and T Clackson, Proc Nut Acad Sci USA, 97, 7096 (2000) 199 T Clackson, W Yang, L Rozamus, J Amara, M H Hatada, C T Rollins, L F Stevenson, S R Magari, S A Wood, N L Courage, X Lu, F Cerasolijunior, M Gilman, and D Holt, Proc Nut Acad Sci USA, 95, 10437 (1998) 200 P Burkhard, P Taylor, and M D Walkinshaw, J Mol Biol., 295,953 (2000) 201 J W Becker, J Rotonda, J G Cryan, M Martin, W H Parsons, P J Sinclair, G Wiederrecht, and F Wong, J Med Chem., 42, 2798 (1999) 202 C Sich, S Improta, D J Cowley, C Guenet, J P Merly, M Teufel, and V Saudek, Eur J Biochem., 267, 5342 (2000) 203 M Huse, Y G Chen, J Massague, and J Kuriyan, Cell, 96,425 (1999) 204 L W Schultz and J Clardy, Bioorg Med Chem Lett., 8, (1998) 205 S Itoh, M T Decenzo, D J Livingston, D A Pearlman, and M A Navia, Bioorg Med Chem Lett., 5, 1983 (1995) 206 J Liang, J Choi, and J Clardy, Acta Crystallogr., Sect D, 55, 736 (1999) 207 C C S Deivanayagam, M Carson, A Thotakura, S V L Narayana, and C S Chodavarapu, Acta Crystallogr., Sect D, 56, 266 (2000) X-Ray Crystallography in Drug Discovery 208 D A Holt, J I Luengo, D S Yamashita, H.-J Oh, A L Konialian, H.-K Yen, L W Rozamus, M Brandt, M J Bossard, M A Levy, D S Eggleston, T J Stout, J Liang, L W Schultz, and J Clardy, J Am Chem Soc., 115, 9925 (1993) 209 G D VanDuyne, R F Standaert, P A Karplus, S L Schreiber, and J Clardy, Science, 252,839 (1991) 210 G D Vanduyne, R F Standaert, S L Schreiber, and J Clardy, J An Chem Soc., 113, 7433 (1991) 211 K M Fox, J A Dias, and P Vanroey, Mol Endocrinol., 15,378 (2001) 212 P Storici, G Capitani, D Debiase, M Moser, R A John, J N Jansonius, and T Schirmer, Biochemistry, 38,8628 (1999) 213 D T Gewirth and P B Sigler, Nut Struct Biol., 2,386 (1995) 214 B F Luisi, W X Xu, Z Otwinowski, L P Freedman, K R Yamamoto, and P B Sigler, Nature, 352,497 (1991) 215 N Kunishima, Y Shimada, Y Tsuji, T Sato, M Yamamoto, T Kumasaka, S Nakanishi, H Jingami, and K Morikawa, Nature, 407, 971 (2000) 216 Epp, R Ladenstein, and A Wendel, Eur J Biochem., 133,51 (1983) 217 M Aritomi, N Kunishima, T Okamoto, R Kuroki, Y Ota, and K Morikawa, Nature, 401, 713 (1999) 218 B Lovejoy, D Cascio, and D Eisenberg, J Mol Biol., 234, 640 (1993) 219 T Zink, A Ross, K Luers, C Ciesler, R Rudolph, and T A Hola, Biochemistry, 33, 8453 (1994) 220 C P Hill, T D Osslund, and D Eisenberg, Proc Natl Acad Sci USA, 90,5167 (1993) 221 D Rozwarski, K Diederichs, R Hecht, T Boone, and P A Karplus, Proteins, 26, 304 (1996) 222 T Clackson, M H Ultsch, J A Wells, and A.M Devos, J Mol Biol., 277, 1111 (1998) 223 S Atwell, M Ultsch, A M Devos, and J A Wells, Science, 278, 1125 (1997) 224 M Sundstrom, T Lundqvist, J Rodin, L B Giebel, D Milligan, and G Norstedt, J Biol Chem., 271,32197 (1996) 225 A M Devos, M Ultsch, and A A Kossiakoff, Science, 255, 306 (1992) 226 Y Hsiou, J Ding, K Das, A D Clarkjunior, S H Hughes, and E Arnold, Structure, 4,853 (1996) 227 J Ren, R M Esnouf, A L Hopkins, E Y Jones, I Kirby, J Keeling, C K Ross, B A Larder, D I Stuart, and D K Stammers, Proc Nut Acad Sci USA, 95,9518 (1998) 228 Y Hsiou, J Ding, K Das, A D Clark, P L Boyer, P Lewi, P A J Janssen, J.-P Kleim, M Roesner, S H Hughes, and E Arnold, J Mol Biol., 309,437 (2001) 229 Y Hsiou, K Das, J Ding, A D Clarkjunior, J P Kleim, M Rosner, I Winkler, G Riess, S H Hughes, and E Arnold, J.Mol Biol.,284, 313 (1998) 230 K Das, J Ding, Y Hsiou, A D Clarkjunior, H Moereels, L Koyrnans, K Andries, R Pauwels, P A Janssen, P L Boyer, P Clark, R H Smithjunior, M B Kroegersmith, C J Michejda, S H Hughes and E Arnold, J Mol Biol., 264, 1085 (1996) 231 M Hogberg, C Sahlberg, P Engelhardt, R Noreen, J Kangasmetsa, N G Johansson, B Oberg, L Vrang, H Zhang, B L Sahlberg, T Unge, S Lovgren, K Fridborg, and K Backbro, J Med Chem., 43,304 (2000) 232 J Lindberg, S Sigurdsson, S Lowgren, H Andersson, C Sahlberg, R Noreen, K Fridborg, H Zhang, and T Unge, Eur J Biochem., 269, 1670 (2002) 233 J Jaeger, T Restle, and T A Steitz, Embo J., 17,4535 (1998) 234 A L Hopkins, J Ren, H Tanaka, B Baba, M Okarnato, D I Stuart, and D K Stammers, J Med Chem., 42,4500 (1999) 235 J Ding, K Das, H Yu, S G Saraf~anos,A D Clarkjunior, A Jacobo-Molina, C Tantillo, S H Hughes, and E Arnold, J.Mol Biol., 284, 1095 (1998) 236 S G Sarafianos, K Das, C Tantillo, A D Clarkjr., J Ding, J Whitcomb, P L Boyer, S H Hughes, and E Arnold, Embo J., 20, 1449 (2001) 237 D W Rodgers, S J Gamblin, B A Harris, S Ray, J S Culp, B Hellmig, D J Woolf, C Debouck, and S C Harrison, Proc Natl Acad Sci USA, 92,1222 (1995) 238 J Ding, K Das, C Tantillo, W Zhang, A D Clarkjunior, S Jessen, X Lu, Y Hsiou, A Jacobo-Molina, K Andries, R Pauwels, H Moereels, L Koymans, P A J Janssen, R H Smithjunior, M K Koepke, C J Michejda, S H Hughes, and E Arnold, Structure (Lond), 3, 365 (1995) 239 J Ding, K Das, H Moereels, L Koymans, K Andries, P A J Janssen, S H Hughes, and E Arnold, Nut Struct Biol., 2, 407 (1995) References 240 J Ren, J Milton, K L Weaver, S A Short, D I Stuart, and D K Stammers, Structure, 8, 1089 (2000) 241 J Ren, C Nichols, L Bird, P Chamberlain, K Weaver, S Short, D I Stuart, and D K Stammers, J Mol Biol., 312, 795 (2001) 242 S G Saraiianos, K Das, C Tantil10,A D Clarkjr., J Ding, J Whitcomb, P L Boyer, S H Hughes, and E Arnold, Proc Nat Acad Sci USA, 96,10027 (1999) 243 J Wang, S J Smerdon, J Jager, L A Kohlstaedt, P A Rice, J M Friedman, T A Steitz, Proc Natl Acad Sci USA, 91, 7242 (1994) 244 M D Sintchak, M A Fleming, Futer, S A Raybuck, S P Chambers, P R Caron, M A Murcko, and K P Wilson, Cell, 85,921 (1996) 245 T D Colby, K Vanderveen, M Strickler, G D Markham, and B M Goldstein, Proc Nat Acad Sci USA, 96, 3531 (1999) 246 L G Laajoki, G L Francis, J C Wallace, J A Carver, and M A Keniry, J Biol Chem., 275, 10009 (2000) 247 A Sato, S Nishimura, T ohkubo,Y Kyogoku, S Koyama, M Kobayashi, T Vasuda, and Y Kobayashi, Znt J Pept Protein Res., 41, 433 (1993) 248 F F Vajdos, M Ultsch, M L Schaffer, K D Deshayes, J Liu, N J Skelton, and A M Devos, Biochemistry, 40, 11022 (2001) 249 E Dewolf, R Gill, S Geddes, J Pitts, A Wollmer, and J Grotzinger, Protein Sci., 5, 2193 (1996) 250 R M Cooke, T S Harvey, and I D Campbell, Biochemistry, 30, 5484 (1991) 251 T P J Garrett, N M Mckern, M Lou, M J Frenkel, J D Bentley, G Lovrecz, T C Elleman, L J Cosgrove, and C W Ward, Nature, 394,395 (1998) 252 K Parang, J H Till, A J Ablooglu, R A Kohanski, S R Hubbard, and P A Cole, Nat Struct Biol., 8, 37 (2001) 253 J H Till, A J Ablooglu, M Frankel, S M Bishop, R A Kohanski, and S R Hubbard, J Biol Chem., 276, 10049 (2001) 254 S R Hubbard, Embo J.,16, 5572 (1997) 255 S R Hubbard, L Wei, L Ellis, and W A Hendrickson, Nature, 372,746 (1994) 256 A M Torres, B E Forbes, S E Aplin, J C Wallace, G L Francis, and R S Norton, J Mol Biol., 248,385 (1995) 257 E T Baldwin, R W Sarver, G L Bryantjunior, K A Currym.B Fairbanks, B C Finzel, R L Garlick, R L Heinrikson, N C Horton, L L Kelley, A M Mildner, J B Moon, J E Mott, V T Mutchler, C S Tomich, K D Watenpaugh, and V H Wiley, Structure (Land), 6,923 (1998) 258 J Lee, L A Bankston, M A Arnaout, and R C Liddington, Structure (Land), 3, 1333 (1995) 259 J Lee, P Rieu, M A Amaout, and R Liddington, Cell, 80,631 (1995) 260 J Bella, P R Kolatkar, C W Marlor, J M Greve, and M G Rossmann, Proc Nat Acad Sci USA, 95,4140 (1998) 261 J M Casasnovas, T Stehle, J H Liu, J H Wang, and T A Springer, Proc Nat Acad Sci USA, 95,4134 (1998) 262 P R Kolatkar, J Bella, N H Olson, C Bator, T S Baker, and M G Rossmann, Embo J.,18, 6249 (1999) 263 W Klaus, B Gsell, A M Labhardt, B Wipf, and H Senn, J Mol Biol., 274,661 (1997) 264 R Radhakrishnan, L J Walter, A Hruza, P Reichert, P P Trotta, T L Nagabhushan, and M R Walter, Structure, 4, 1453 (1996) 265 D J Thiel, M.-H Ledu, R L Walter, A D'Arcy, C Chene, M Fountoulakis, G Garotta, F K Winkler, and S E Ealick, Structure Fold Des., 8,927 (2000) 266 M Randal and A A Kossiakoff, Structure, 9, 155 (2001) 267 A Landar, B Curry, M H Parker, R Digiacomo, S R Indelicato, T L Nagabhushan, G Rizzi, and M R Walter, J.Mol Biol., 299, 169 (2000) 268 S E Ealick, W J Cook, S Vijay-Kumar, M Carson, T L Nagabhushan, P P Trotta, and C E Bugg, Science, 252,698 (1991) 269 B J Graves, M H Hatada, W A Hendrickson, J K Miller, V S Madison, and Y Satow, Biochemistry, 29,2679 (1990) 270 G P A Vigers, D J Dripps, C K Edwards, and B J Brandhuber, Structure, 275, 36927 (2000) 271 H Schreuder, C Tardif, S Trump-Kallmeyer, A Soffientini, E Sarubbi, A Akeson, T Bowlin, S Yanofsky, and R W Barrett, Nature, 386, 194 (1997) 272 G P Vigers, L J Anderson, P Caffes, and B J Brandhuber, Nature, 386, 190 (1997) 273 A Zdanov, C Schalk-Hihi, S Menon, K W Moore and A Wlodawer, J.Mol Biol., 268,460 (1997) 274 A Zdanov, C Schalk-Hihi, and A Wlodawer, Protein Sci., 5, 1955 (1996) X-Ray Crystallography in Drug Discovery 275 A Zdanov, C Schalk-Hihi, A Gustchina, M Tsang, J Weatherbee, and A Wlodawer, Structure, 3,591 (1995) 276 K Josephson, N J Logsdon, and M R Walter, Immunity, 15, 35 (2001) 277 M R Walter and T L Nagabhushan, Biochemistry, 34, 12118 (1995) 278 C Yoon, S C Johnston, J Tang, M Stahl, J F Tobin, and W S Somers, Embo J., 19, 3530 (2000) 279 E Z Eisenmesser, D A Horita, A S Altieri, and R A Byrd, J Mol Biol., 310,231 (2001) 280 H R Mott, B S Baines, R M Hall, R M Cooke, P C Driscoll, M P Weir, and I D Campbell, J Mol Biol., 248, 979 (1995) 281 D B McKay, Science, 257,412 (1992) 282 Y Feng, B K Klein, and C A Mcwherter, J Mol Biol., 259,524 (1996) 283 T Mueller, F Oehlenschlaeger, and M Buehner, J Mol Biol., 247, 360 (1955) 284 T Hage, W Sebald, and P Reinemer, Cell, 97, 271 (1999) 285 M Hulsmeyer, C Scheufler, and M K Dreyer, Acta Crystallogr., Sect D, 57, 1334 (2001) 286 C Redfield, L J Smith, J Boyd, G M P Lawrence, R G Edwards, C J Gershater, R A G Smith, and C M Dobson, J Mol Biol., 238,23 (1994) 287 R Powers, D S Garrett, C J March, E A Frieden, A M Gronenborn, and G M Clore, Science, 256, 1673 (1992) 288 T Mueller, T Dieckmann, W Sebald, and H Oschkinat, J Mol Bio1.237,423 (1994) 289 T Mueller, T Dieckmann, W Sebald, and H Oschkinat, J Mol Biol., 237,423 (1994) 290 L J Smith, C Redfield, J Boyd, G M P Lawrence, R G Edwards, R A G Smith, and C M Dobson, J Mol Biol., 224,899 (1992) 291 M R Walter, W J Cook, B G Zhao, R Cameronjunior, S E Ealick, R L Walterjunior, P Reichert, T L Nagabhushan, P P Trotta, and C E Bugg, J Biol Chem., 267,20371 (1992) 292 A Wlodawer, A Pavlovsky, and A Gustchina, Febs Lett., 309,59 (1992) 293 R Powers, D S Garrett, C J March, E A Frieden, A M Gronenborn, and G M Clore, Biochemistry, 32,6744 (1993) 294 M V Milburn, A M Hassell, M H Lambert, S R Jordan, A E I Proudfoot, P Graber, and T N C Wells, Nature, 363, 172 (1993) 295 G Y Xu, H A Yu, J Hong, M Stahl, T Mcdonagh, L E Kay, and D A Cumming, J Mol Biol., 268,468 (1997) 296 W Somers, M Stahl, and J S Seehra, Embo J., 16, 989 (1997) 297 K Rajarathnam, I Clark-Lewis, and B D Sykes, Biochemistry, 34, 12983 (1995) 298 C Eigenbrot, H B Lowman, L Chee, and D R Artis, Proteins, 27, 556 (1997) 299 N J Skelton, C Quan, D Reilly, and H Lowman, Structure Fold Des., 7, 157 (1999) 300 N Gerber, H Lowman, D R Artis, and C Eigenbrot, Proteins: Struct., Funct., Genet., 38,361 (2000) 301 H Sticht, M Auer, B Schmitt, J Besemer, M Horcher, T Kirsch, J D Lindley, and P Roesch, Eur J Biochem., 235, 26 (1996) 302 E T Baldwin, I T Weber, R St Charles, J.-C Xuan, E Appella, M Yamada, K Matsushima, B F P Edwards, G M Clore, A M Gronenborn, and A Wlodawer, Proc Nut Acad Sci USA, 88,502 (1991) 303 G M Clore, E Appella, M Yamada, K Matsushima, and A M Gronenborn, Biochemistry, 29,1689 (1990) 304 M M G M Thunnissen, P N Nordlund, and J Z Haeggstrom, Nut Struct Biol., 8, 131 (2001) 305 X Weng, H Luecke, I S Song, D S Kang, S-H Kim, and R Huber, Protein Sci., 2, 448 (1993) 306 A Rosengarth, V Gerke, and H Luecke, J Mol Biol., 306,489 (2001) 307 M Tegoni, S Spinelli, M Verhoeyen, P Davis, and C Carnbilla, J Mol Biol., 289, 1375 (1999) 308 H Wu, J W Lustbader, Y Liu, R E Canfield, and W A Hendrickson, Structure, 2, 545 (1994) 309 A J Lapthorn, D C Harris, A Littlejohn, J W Lustbader, R E Canfield, K J Machin, F J Morgan, and N W Isaacs, Nature, 369, 455 (1994) 310 A Bohm, J Pandit, J Jancarik, R Halenbeck, K Koths, and S.-H Kim, Science, 258, 1358 (1992) 311 M J Jedrzejas, S Singh, W J Brouillette, W G Laver, G M Air, and M Luo, J Mol Biol., 267,584 (1997) 312 N R Taylor,A Cleasby, 0.Singh, T Skarzynski, A J Wonacott, P W Smith, S L Sollis, P D Howes, P C Cherry, R Bethell, P Colman, and J Varghese, J Med Chem., 41,798 (1998) 313 M J Jedrzejas, S Singh, W J Brouillette, W G Laver, G M Air, and M Luo, Biochemistry, 34, 3144 (1995) i References 314 W P Burmeister, B Henrissat, C Bosso, S Cusack, and R W H Ruigrok, Structure, 1,19 (1993) 315 J B Finley, V R Atigadda, F Duarte, J J Zhao, W J Brouillette, G M Air, and M Luo, J Mol Biol., 293, 1107 (1999) 316 C L White, M N Janakiraman, W G Laver, C Philippon, A Vasella, G M Air, and M Luo, J Mol Biol., 245,623 (1995) 317 W P Burmeister, R W H Ruigrok, and S Cusack, Embo J.,11,49 (1992) 318 S A Monks, G Karagianis, G J Howlett, and R S Norton, J Biomol NMR, 8,379 (1996) 319 R Bader, A Bettio, A G Beck-Sickinger, and Zerbe, J Mol Biol., 305,307 (2001) 320 C Cabrele, M Langer, R Bader, H A Wieland, H N Doods, Zerbe, and A G BeckSickinger, J Biol Chem., 275,36043 (2000) 321 G Seidel, W Schaefer, A Esswein, E Hofmann, and P Roesch, In Press 322 Z Chen, P Xu, J.-R Barbier, G Willick, and F Ni, Biochemistry, 39, 12766 (2000) 323 U C Mam, K Adermann, P Bayer, W.-G Forssmann, and P Roesc, Biochem Biophys Res Comm., 267,213 (2000) 324 L Jin, S L Briggs, S Chandrasekhar, N Y Chirgadze, D K Clawson, R W Schevitz, D L Smiley, A H Tashjian, and F Zhang, J Biol Chem., 275,27238 (2000) 325 U C Mam, S Austermann, P Bayer, K Adermann, A Ejcharth Sticht, S Walter, F.-X Schmid, R Jaenicke, W.-G Forssmann and P Roesch, J Biol Chem., 270, 15194 (1995) 326 U C Marx, Strukturen VerschiedenerParathormonfragmente in Loesung, University of Bayreuth Thesis, Bayreuth, 1996 327 G Y Xu, T Mcdonagh, H A Yu, E A Nalefski, J D Clark, and D A Cumming, J Mol Biol., 280,485 (1998) 328 Perisic, S Fong, D E Lynch, M Bycroft, and R L Williams, J Biol Chem., 273, 1596 (1998) 329 A Dessen, J Tang, H Schmidt, M Stahl, J D Clark, J Seehra, and W S Somers, Cell, 97, 349 (1999) 330 A Kreusch, P J Pfaffinger, C F Stevens, and S Choe, Nature, 392,945 (1998) 331 S J Cushman, M H Nanao, A W Jahng, D Derubeis, S Choe, and P J Pfaffinger, Nut Struct Biol., 7,403 (2000) 332 K A Bixby, M H Nanao, N V Shen, A Kreusch, H Bellamy, P J Pfaffinger, and S Choe, Nut Struct Biol., , (1998) 333 J M Gulbis, M Zhou, S Mann, and R Mackinnon, Science, 289, 123 (2000) 334 D L Minorjr., Y.-F Lin, B C Mobley, A Avelar, Y N Jan1.Y Jan, and J M Berger, Cell, 102, 657 (2000) 335 J L Oberfield, J L Collins, C P Holmes, D M Goreham, J P Cooper, J E Cobb, J M Lenhard, E A Hull-Ryde, C P Mohr, S G Blanchard, D J Parks, L B Moore, J M Lehmann, K Plunket, A B Miller, M V Milburn, S A Kliewer, and T M Wilson, Proc Nut Acad Sci USA, 96,6102 (1999) 336 R T Nolte, G B Wisely, S Westin, J E Cobb, M H Lambert, R Kurokawa, M G Rosenfeld, T M Willson, C K Glass, and M V Milburn, Nature, 395, 137 (1998) 337 R T Gampejr.,V G Montana, M H Lambert, A B Miller, R K Bledsoe, M V Milburn, S A Kliewer, T M Willson, and H E Xu, Mol Cell, 5, 545 (2000) 338 J Uppenberg, C Svensson, M Jaki, G Bertilsson, L Jendeberg, and A Berkenstam, J Biol Chem., 273,31108 (1998) 339 S P Williams and P B Sigler, Nature, 393, 392 (1998) 340 W Somers, M Ultsch, A M Devos, and A A Kossiakoff, Nature, 372,478 (1994) 341 P A Elkins, H W Christinger, Y Sandowski, E Sakal, A Gertler, A M Devos, and A A Kossiakoff, Nut Struct Biol., 7, 808 (2000) 342 F Rastinejad, T Wagner, Q Zhao, and S Khorasanizadeh, Embo J.,19, 1045 (2000) 343 B P Klaholz, A Mitschler, M Belema, C Zusi, and D Moras, Proc Nut Acad Sci USA, 97, 6322 (2000) 344 J P Renaud, N Rochel, M Ruff, V Vivat, P Chambon, H Gronemeyer, and D Moras, Nature, 378, 681 (1995) 345 B P Klaholz, J P Renaud, A Mitschler, C Zusi, P Chambon, H Gronemeyer, and D Moras, Nut Struct Biol., 5, 199 (1998) 346 W Bourguet, V Vivat, J M Wurtz, P Chambon, H Gronemeyer, and D Moras, Mol Cell, 5, 289 (2000) 347 B P Klaholz, A Mitschler, and D Moras, J Mol Biol., 302, 155 (2000) 348 R M A Knegtel, M Katahira, J G Schilthuis, A M J J Bonvin, R Boelens, D Eib, P T Vandersaag, and R Kaptein, J Biomol NMR, 3, l(1993) 349 W Bourguet, M Ruff, P Chambon, H Gronemeyer, and D Moras, Nature, 375,377 (1995) 350 F Rastinejad, T Perlmann, R M Evans, and P B Sigler, Nature, 375, 203 (1995) X-Ray Crystallography in Drug Discovery 351 S M Holmbeck, M P Foster, D R Casimiro, D S Sem, H J Dyson, and P E Wright, J Mol Biol., 281,271 (1998) 352 R T Gampejr., V G Montana, M H Lambert, G B Wisely, M V Milburn, and H E Xu, Genes Dev., 14, 2229 (2000) 353 P F Egea, A Mitschler, N Rochel, M Ruff, P Chambon, and D Moras, Embo J., 19, 2592 (2000) 354 Q Zhao, S A Chasse, S Devarakonda, M L Sierk, and B A Rastinejad, J Mol Biol., 296, 509 (2000) 355 D R Hall, J M Hadden, G A Leonard, S Bailey, M Neu, M Winn, and P F Lindley, Acta Crystallogr., Sect D, 58, 70 (2002) 356 Z Zhang, R Zhang, A Joachimiak, J Schlessinger, and X Kong, Proc Nat Acad Sci USA, 97, 7732 (2000) 357 X Jiang, Gurel, E A Mendiaz, G W Steams, C L Clogston, H S Lu, T D Osslund, R S Syed, K E Langley, and W A Hendrickson, Embo J.,19,3192 (2000) 358 J N Charnpness, M S Bennett, F Wien, R Visse, W C Summers, P Herdewijn, E Declercq, T Ostrowski, R L Jarvest, and M R Sanderson, Proteins, 32, 350 (1998) 359 M S Bennett, F Wien, J N Champness, T Batuwangala, T Rutherford, W C Summers, H Sun, G Wright, and M R Sanderson, Febs Lett., 443, 121 (1999) 360 J N Champness, M S Bennett, F Wien, R Visse, W C Summers, P Herdewijn, E Declerq, T Ostrowski, R L Jarvest, and M R Sanderson, Proteins Struct Funct Genet., 32, 350 (1998) 361 K Wild, T Bohner, G Folkers, and G E Schulz, Protein Sci., 6,2097 (1997) 362 J Vogt, R Perozzo, A Pautsch, A Prota, P Schelling, B Pilger, G Folkers, L Scapozza, and G E Schulz, Proteins Struct Funct Genet., 41, 545 (2000) 363 C Wurth, U Kessler, J Vogt, G E Schulz, G Folkers, and L Scapozza, Protein Sci., 10, 63 (2001) 364 A Prota, J Vogt, B Pilger, R Perozzo, C Wurth, V Marquez, P Russ, G E Schulz, G Folkers, and L Scapozza, Biochemistry, 39, 9597 (2000) 365 J H Naismith, T Q Devine, B Brandhuber, and S R Sprang, J Biol Chem., 270, 13303 (1995) 366 J H Naismith, T Q Devine, H Khono, and S R Sprang, Structure, 4, 1251 (1996) 367 D W Banner, A D'Arcy, W Janes, R Gentz, H-J Schoenfeld, C Broger, H Loetscher, and W Lesslauer, Cell, 73,431 (1993) 368 G Tocchini-Valentini, N Rochel, J M Wurtz, A Mitschler, and D Moras, Proc Nat Acad Sci USA, 98, 5491 (2001) 369 N Rochel, J M Wurtz, A Mitschler, B Klaholz, and D Moras, Mol Cell, 5, 173 (2000) 370 S Vos, R J Parry, M R Burns, J Dejersey, and J L Martin, J.Mol Biol., 282,875 (1998) 371 S Vos and J De Jersey, Biochemistry, 36,4125 (1997) 372 S Vos, R J Parry, M R Burns, J Dejersey, and J L Martin, J.Mol Biol., 282,875 (1998) 373 H A Lewis, Structure (Camb), 9, 527-537 (2001) 374 C Freiberg, Drug Discov Today, 6, S72-S80 (2001) 375 J M Sauder, J W Arthur, and R L Dunbrack, Proteins, 40, 6-22 (2000) 376 S W Muchmore, Nature, 381,335-341 (1996) 377 K A Denessiouk, A I Denesyuk, J V Lehtonen, T Korpela, and M S Johnson, Proteins, 35, 250-261 (1999) 378 M Teplova, Protein Sci., 9,2557-2566 (2000) 379 T J Boggon, W S Shan, S Santagata, S C Myers, and L Shapiro, Science, 286, 2119-2125 (1999) 380 P W Kleyn, Cell, 85,281-290 (1996) 381 K Noben-Trauth, J K Naggert, M A North, and P M Nishina, Nature, 380, 534-538 (1996) 382 S Santagata, Science, 292,2041-2050 (2001) 383 W Eisenreich, Chem Biol., 5, R221-R233 (1998) 384 R Sanchez, Nat Struct Biol., 7, 986-990 (2000) 385 A Sali, Nat Struct Biol., 5,1029-1032 (1998) 386 A Fiser, R K Do, and A Sali, Protein Sci., 9, 1753-1773 (2000) 387 K T Simons, C Strauss, and D Baker, J.Mol Biol., 306, 1191-1199 (2001) 388 R Sanchez, Nucleic Acids Res., 28, 250-253 (2000) AMR and Drug Discovery i DAVID J CFNK RICHARD J CLARK Institute for Molecular Bioscience Australian Research Council Special Research Centre for Functional and Applied Genomics University of Queensland Brisbane, Australia Contents Introduction, 508 1.1 Overview of Drug Development, 509 1.2 Scope of Chapter, 510 1.3 Principles of NMR Spectroscopy, 510 1.4 Instrumentation, 514 1.5 Applications of NMR in Drug Design and Discovery, 516 Ligand-Based Design, 517 2.1 Structure Elucidation, 517 2.1.1 Structure Elucidation of Natural Products, 517 2.1.2 Structure Determination of Bioactive Peptides, 518 2.1.2.1 NMR Structure of Ziconotide: A Novel Treatment for Pain, 518 2.1.2.2 Endothelin as a Lead in LigandBased Design, 523 2.1.3 Instrumental Advances and their Impact on Structure Elucidation, 524 2.2 Conformational Analysis, 525 2.3 Charge State, 526 2.4 Tautomeric Equilibria, 526 2.5 Ligand Dynamics: Line-Shape and Relaxation Data, 528 2.6 Pharmocophore Modeling: Conformations of a Set of Ligands, 531 2.7 Limitations of Analog-Based Design, 532 2.8 Conformation of Bound Ligands: Transferred NOES, 532 Receptor-Based Design, 532 3.1 Macromolecular Structure Determination, 533 3.1.1 Overview of Approach, 533 Burger's Medicinal Chemistry and Drug Discovery Sixth Edition, Volume 1: Drug Discovery Edited by Donald J Abraham ISBN 0471-27090-3 O 2003 John Wiey & Sons, Inc 507 NMR and Drug Discovery 3.1.2 Sample Requirements and Assignment Protocols, 534 3.1.3Recent Developments, 534 3.1.4 Dynamics, 534 3.1.5 Nucleic Acid Structures, 535 3.1.6 Challenges for the Future: MembraneBound Proteins, 535 3.2 Macromolecule-Ligand Interactions, 535 3.2.1 Overview, 535 3.2.2Influence of Kinetics and NMR Timescales, 536 3.2.2.1Slow Exchange, 538 3.2.2.2Fast Exchange, 539 3.2.2.3Intermediate Exchange, 540 3.2.3 NMR Techniques, 543 3.2.3.1Chemical-Shift Mapping, 543 3.2.3.2NMR Titrations, 545 3.2.3.3Isotope Editing and Filtering, 545 3.2.3.4NOE Docking, 545 3.2.4Selected Examples, 547 3.2.4.1 DNA-Binding Drugs, 547 INTRODUCTION NMR spectroscopy has been widely used as a front-line tool in the pharmaceutical industry for several decades In the past, the main use of NMR was in the structural characterization of organic molecules synthesized in the course of medicinal chemistry programs Indeed, medicinal chemists have long regarded NMR as the premier tool to be used in the structure characterization process, to confirm the identity of intermediates or to determine the conformation of lead molecules Over the last decade major developments in both instrumentation and methods have resulted in this traditional use of NMR in the pharmaceutical industry being augmented by a range of exciting new applications Two of the most important of these are the use of NMR in structurebased drug design and in screening for drug discovery Both applications differ from the traditional use of NMR in that now the macromolecular binding partner of the medicinal compound is included in the mixture to be analyzed; that is, contemporary applications of NMR in drug discovery are predominantly focused on the interaction between drug molecules and their macromolecular targets 3.2.4.2Immunophilins: Studies of FK506 Analog Binding to FKBP, 552 3.2.4.3Matrix Metalloproteinases, 555 3.2.4.4 Dihydrofolate Reductase, 557 3.2.4.5 H N Protease, 559 NMR Screening, 562 4.1 Methods, 562 4.1.1 Chemical-Shift Perturbation, 562 4.1.2 Magnetization Transfer Experiments, 568 4.1.3 Molecular Diffusion, 570 4.1.4 Relaxation, 571 4.1.5 NOE, 571 4.1.6 Spin Labels, 573 4.2 Practical Considerations, 574 4.2.1 Screening Approach, 574 4.2.2 Library Design, 574 4.2.2.1 Ligand Properties, 574 4.2.2.2 Mixture Design, 576 4.2.3 Hardware and Automation, 576 Conclusions, 577 Acknowledgments, 577 The aim of this chapter is to describe how NMR spectroscopy is used in modern drug discovery The term discovery is used generically throughout to include processes that involve rational drug design as well as those that involve discovery through NMR screening The latter is a relatively recent development and refers to the use of NMR as a tool to screen a compound library, to identify a molecule or molecules that bind to a chosen macromolecular target Of course, the distinction between "design" and "discovery" is often quite blurred This is nowhere more evident than in the recently developed SAR-by-NMR approach (I),in which the discovery of several weakly bound ligands from a screening program is intimately linked to a design process to chemically join them SAR-by-NMR represents an exciting new technique for lead generation and is described in more detail later in this chapter Drug designtdiscovery represents only the first stage in the whole drug development process As is clear from the other chapters in this volume, there are many other steps that need to be made once a lead molecule has been designed or discovered Although other stages of the process, including lead optimization, tox- Introduction Cycle A Figure 12.1 Overview of the drug development process and summary of various types of NMR experiments that contribute at different stages :ity studies, preclinical investigations, and linical monitoring, not fall within the cope of this chapter, it is worth mentioning hat NMR spectroscopy contributes signifisntly across the whole spectrum of drug deelopment, right through into the clinical omain For example, NMR spectroscopy as been applied for the detection of drug letabolites in biological fluids and magnetic sonance imaging, which is based on the lndamental principles of NMR, plays an nportant role in clinical investigations ~tis lcreasingly being used to monitor the funcanal outcomes of drug therapy We briefly ldress these broader applications of NMR ?forereturning to the main topic of NMR in rug discovery 1.1 Overview of Drug Development To give an overview of the breadth of applications of NMR, Fig 12.1 summarizes the drug development process and indicates the role of NMR at various stages Drug development is an iterative process and can be simplified by representing it with two interconnected cycles of activity Cycle A involves the design or discovery of an initial lead followed by its synthesis and bioassay Based on the initial assay results there may be several loops around this cycle before commencing the in vivo studies represented in Cycle B At this stage consideration of bioavailability, metabolism, and pharmacokinetic profiles must be made and this may involve synthetic modifications of the NMR and Drug Discovery lead molecules to improve their druglike properties Again, several loops around Cycle B may be necessary before one or more development candidates are identified Ultimately one or two of these development candidates are identified for progression through clinical trials As indicated in Fig 12.1, it is convenient to envisage five broad categories of NMR experiments that may contribute to this overall drug development process Small molecule, or ligand-based, NMR This involves studies of drugs and drug leads, typically organic molecules with a molecular weight 8Hz) and downward-pointing arrows indicate a small coupling (100 kDa Another development that is likely to have a significant impact is the increasing number of structural genomics programs being developed The demands arising from such programs will no doubt stimulate new methods for the large-scale production of labeled proteins (77, 78), and for speeding up the rate of structure determination by both NMR and crystallography 3.1.4 Dynamics Proteins exhibit a range of internal motions, from the millisecond to nanosecond timescale, and a full understanding of how small drugs might interact with such a "moving target" requires more than just the time-averaged macromolecular structure Thus, over recent years much effort has been directed toward defining motions within vroteins The most commonly applied approach has been to use 13C or 15N relaxation parameters such as TI, T,, and the heteronuclear NOE to derive correlation times for overall motion, together with rates and amplitudes of internal motions (79) Although the precise interpretation of the NMR relaxation data in terms of motional parameters remains dependent on the appropriateness of the motional model chosen, the results from many studies on the dynamics of proteins are sufficiently clear to confirm that nanosecond timescale motions in proteins are common The functional significance of motions on the nanosecond timescde remains unclear and so far there have been few cases where significant differences in motions on this timescale between ligand-free and ligand-bound forms of proteins have been measured It will be interesting to assess the functional significance of such motions as more data become available However, slower motions have been correlated with function in a number of proteins, with a good example being HIV protease, described in more detail in Section 4.2 Relaxation measurements require a considerable investment of svectrometer time and in some cases it may be possible to derive basic information about molecular dynamics from the structural ensemble alone Although regions of disorder can reflect factors other than dynamics, a recent analysis (55)suggests that ill-defined regions in structural ensembles often reflect slow, large-amplitude motions Even if relaxation measurements are A A Receptor-Based Design done, it is often not necessary to undertake extensive analysis to derive correlation times, given that trends are often apparent from the raw experimental data For example, in the case of tendamistat,described above, it is clear directly from the heteronuclear NOE data that significant internal mobility is present at the N-terminus 3.1.5 Nucleic Acid Structures Most of the discussion on macromolecular targets so far has focused on proteins DNA represents another valuable target in drug design Most studies in which DNA is the target are done using short model oligonucleotides to mimic the binding region of DNA The regular repeating nature of DNA structures makes this a more successful approach than similar attempts to dissect out binding regions of receptor proteins, where often the whole protein must be present to maintain a viable binding site Similar comments apply for RNA, where improvements in synthetic methods have led to an increasing number of structure determinations over recent years The principles involved in structure determination of nucleic acid targets are similar to those of proteins, but in practice nucleic acid structures are somewhat more difficult to solve 3.1.6 Challenges for the Future: MembraneBound Proteins The majority of targets for currently known drugs are membrane-bound receptors, yet this represents the class of proteins for which least structural information is known Membrane proteins are notoriously difficult to characterize at a structural level because they are difficult to crystallize, thus inhibiting X-ray crystallographic studies, and are both too large and too difficult to reconstitute in suitable media for NMR studies Nevertheless, solid-state NMR methods are beginning to show promise that eventually such targets may be structurally characterized (80) Rotational resonance solid-state NMR measurements, for example, allow precise distances to be measured in membrane-bound proteins (81) 3.2 Macromolecule-Ligand Interactions Macromolecule-ligand interactions are integral to a wide range of biological processes, including hormone, neurotransmitter or drug binding, antigen recognition, and enzyrnesubstrate interactions Fundamental to each of these interactions is the recognition by a ligand of a unique binding site on the macromolecule Through an understanding of the specific interactions involved it may be possible to design or discover analogous ligands with altered binding properties that might inhibit the biochemical function of the macromolecule in a highly specific manner The study of macromolecule-ligand interactions thus forms the cornerstone of most structurebased drug design applications The macromolecule of interest may be a protein or a nucleic acid, although the majority of drug design applications have focused on protein-ligand interactions For this reason we will refer mainly to protein-ligand interactions in the following discussion, but will include some examples of drug-DNA interactions 3.2.1 Overview There are several imvor- tant aspects of macromolecule-ligand interactions that have a bearing on structure-based design The simplest question that might be asked is "what is the strength of the binding interaction?," whereas the most detailed task would be to precisely define the atomic coordinates of the complete protein-ligand complex In between these extremes there are many other questions important to the drug design process; these include questions about the binding stoichiometry and kinetics, the conformation of the bound ligand, and about the nature of functional group interactions between the protein and bound ligand These and other important questions were introduced briefly in Table 12.3 and are examined in more detail later in this section Before doing this it is first necessary to consider NMR timescales because the ability of NMR methods to answer questions about macromolecule-ligand complexes depends critically on the kinetics of the binding interaction Section 3.2.2 thus describes how various NMR parameters depend on binding kinetics and in particular how fast- and slow-exchange conditions affect the interpretation of NMR data Having identified the exchange regime, the task then becomes to decide which NMR -parameters can be used to answer the questions NMR and Drug Discovery 536 Table 12.7 N M R Parameters and Their Changes on Binding Parameter - VML Chemical shift VL Coupling constant %I - V M L J I - JML Relaxation rateb Typical Magnitudea Difference 0-1000 s-l 0-500 s-' 0-12 s-l 0-12 s-l 0-50 s-' (for T I ,larger for T,) 0-10 s-l -J M ~ l / T L- l/TML l/TM- l / T M L J~ "Ranges are approximate only and larger effects may be seen in some cases bllT refers to either 1/T, or 1/T, posed above about the complex Many of the NMR parameters that were described earlier for deriving information about ligands are also applicable to studies of complexes These include chemical shifts, NOES, and relaxation parameters However, the presence of two interacting partners means that there are some differences in the way such parameters are measured and this has led to the development of several techniques that are particularly important for the study of macromolecule-ligand interactions, including chemical-shift mapping, isotope editing, and various NMR titrations Section 3.2.3 describes these techniques Finally, illustrative examples of the application of these techniques to specific drug design problems are given in Section 3.2.4 3.2.2 Influence of Kinetics and NMR Timescales Macromolecule-ligand interactions are characterized by an equilibrium reaction that potentially has a wide range of affinities and rates: The rate constant for the forward reaction is referred to as the on rate (k,,), whereas dissociation of the complex is characterized by the reverse rate constant, k,, The equilibrium constant for this interaction, represented in terms of the dissociation constant of the complex KD, reflects a balance of the on and off rates, as shown in Equation 12.2: For many protein-ligand interactions k,, is of the order of 10sM-' s-I, and is typically quite similar for different ligands The observation that KD values may vary over a wide range, typically from millimolar to nanomolar (i.e., K, = lop3-lop9 M) for cases of interest, is a reflection of a variation in k,, for different ligands Consideration of the k,, value above and the range of KD values noted suggests a range in k,,from 10-I to lo5 s-I The lifetime = llk,,) may thus of the bound complex (T, vary from much less than a millisecond to tens of seconds (lop5to 10 s based on the above off rates) The exchange rate for the second-order binding process is given by (82): where pMLand p, are the mole fractions of bound and free ligand, respectively The appearance of an NMR spectrum of a protein-ligand complex is dependent on the rate of chemical interchange between free and bound states In particular, the effects of exchange on an individual NMR parameter (e.g., chemical shift, coupling constant, or relaxation rate) depend on the relative magnitude of the exchange rate and the difference in the NMR parameter between the two states The cases where the rate of interchange is greater than, about equal to, or less than, the parameter difference are referred to as fast, intermediate, and slow exchange, respectively, as indicated in Table 12.7 Table 12.7 shows that the changes in chemical shifts on ligand binding (for signals either from the ligand or from the macromolecule) are in general greater than those for coupling constants or relaxation rates Given that 100 s-' might represent a typical exchange rate between free and bound states, it is clear that Receptor-Based Design Bound Intermediate h Fast individual NMR signals may be found in either slow, fast, or intermediate exchange on the chemical-shift timescale, but it is more likely that couplings or relaxation parameters will be in fast exchange Thus, in most cases where the term "NMR timescale" is used in the literature, it refers to the chemical-shift timescale The table also emphasizes that there are two types of signals that can be monitored, those from the ligand and those from the macromolecule In general, the typical magnitude of changes to chemical shifts or couplings of either type of signal on binding are similar, although the changes to ligand signals may be larger than those from the macromolecule However, changes to relaxation parameters for signals from ligands are much more likely to be greater than those for protein signals Figure 12.14 Schematic illustration of the effects of slow, intermediate, and fast exchange on the appearance of peaks in NMR spectra of macromoleculeligand complexes In the slow exchange case separate peaks are seen for free and bound forms Note the broader peak for the bound ligand because it now adopts the correlation time of the macromolecule In the fast exchange case only an averaged peak is observed This reflects the sensitivity of relaxation parameters to molecular mobility: a ligand undergoes a greater relative change in mobility on binding than does a protein, given that the relative increase in molecular weight in the complex is much greater for the ligand than for the protein The exchange regime (slow, intermediate, or fast) determines how a spectrum of a protein-ligand mixture changes during a titration, or as a function of temperature Figure 12.14 schematically illustrates the various exchange regimes for macromolecule-ligand binding interactions Slow exchange, corresponding to tight binding, is potentially the most useful regime, given that much detailed information on the nature of a complex can be deduced in this case Nevertheless, fast ex- NMR and Drug Discovery change also allows valuable kinetic and thermodynamic parameters to be derived The analysis is more complex for intermediate exchange and few quantitative studies are attempted for this situation 3.2.2.1 Slow Exchange This situation applies when the rate of exchange is much slower than the difference in chemical shifts between the two states (i.e., k 200,000 < 10,000 < 10,000 4,000 > 10,000 Low resolution none Low resolution none MSn, high resolution MS, low resolution High resolution High resolution searchers in the petroleum industry (I), CI became another standard ionization technique for organic mass spectrometry During CI, high energy electrons (as in EI) are used to ionize a gas called a reagent gas at a constant pressure (usually -1 Torr) in the mass spectrometer ionization source The reagent gas in turn ionizes the sample molecules through ion-molecule reactions that usually involve the exchange of protons Less frequently, sample molecule ionization might involve a charge exchange Two of the most common ionization mechanisms in CI are summarized in Equations 13.3 and 13.4 M + RH + -+MH + + R CI through proton transfer, R = reagent gas (13.3) CI through charge exchange During the 1960s, high resolution double-focusing magnetic sector instruments became available and are now standard tools for the determination of elemental compositions using a type of analysis called exact mass measurement In mass spectrometry, resolution is defined as MIAM, where M is the mlz value of a singly charged ion, and AM is the difference (measured in mlz) between M and the next highest ion Alternatively, AM may be defined in terms of the width of the peak High resolution is typically regarded as a value of at least 10,000 At this resolution, the molecular ions of most drug-like molecules (that is compounds with molecular weights less than -500) can be resolved from each other After resolving a sample ion from others in a mass spectrum, an exact mass measurement may be carried out by accurately weighing the unknown ion and comparing its m l value to that of a calibration standard Since the 1960s, other types of mass spectrometers capable of high resolution exact mass measurements have become available as commercial products, including Fourier transform ion cyclotron resonance (FTICR) mass spectrometers, reflectron TOF instruments, and recently, quadrupole time-of-flight hybrid (QqTOF) mass spectrometers (see Table 13.1 for a listing of types of organic mass spectrometers and a comparison of their performance characteristics) By the early 2000s, FTICR and QqTOF instruments became more popular than magnetic sector mass spectrometers for exact mass measurements, high resolution measurements, and drug discovery applications As will be discussed below, exact mass measurements are essential to many types of mass spectrometry-based screening and drug discovery today Biomedical applications of mass spectrometry began during the 1960s both at academic institutions and pharmaceutical companies These applications depended on the volatilization (usually by heating) of pharmaceutical compounds and biochemicals before their gasphase ionization using EI or CI To increase the thermal stability and volatility of these compounds, a variety of derivatization methods were developed to mask polar functional groups and reduce hydrogen bonding between molecules These methods were particularly effective for use with gas chromatographymass spectrometry (GC-MS), which was introduced during the 1960s as a practical and powerful tool for qualitative and quantitative Mass Spectrometry and Drug Discovery analysis of compounds in mixtures Both EI and CI were immediately useful for GC-MS, because both of these ionization methods require that the analytes be in the gas phase When capillary GC was incorporated into GCMS, this technique reached maturity GC-MS may be used to select, identify, and quantify organic compounds in complex mixtures at the femtomole level The speed of GC-MS is determined by the chromatography step, which typically requires several minutes to h per analysis By the 1970s, some organic chemists were announcing that organic mass spectrometry had reached maturity and that no new applications were possible Like the physicists and physical chemists who had pronounced the end of mass spectrometry a generation earlier, this group would soon be proved wrong Although GC-MS remains important for the analysis of many organic compounds, this technique is limited to volatile and thermally stable compounds that comprise only a small fraction of all organic compounds and even fewer biomedically important molecules Therefore, thermally unstable compounds, including many pharmaceutical compounds such as nucleic acid analogs and biomolecules such as proteins, carbohydrates, and nucleic acids, cannot be analyzed in their native forms using GC-MS (For more details regarding GC-MS and its applications, see Watson 1997, Section 4.) Although derivatization facilitates the GC-MS analysis of many of these compounds, alternative ionization techniques were needed for the analysis of the vast majority of polar and non-volatile compounds of interest to drug discovery During the 1970s and early 1980s, desorption ionization techniques such as field desorption (FD), desorption EI, desorption CI (DCI), and laser desorption were developed to extend the use of mass spectrometry toward the analysis of more polar and less volatile compounds (see Watson 1997, Section 4, for more information regarding desorption ionization techniques including DCI and FD) Although these techniques helped extend the mass range of mass spectrometry beyond a traditional limit of mlz 1000 and toward ions of mlz 5000, the first breakthrough in the analysis of polar, non-volatile compounds occurred in 1982 with the invention of fast atom bombardment (FAB) (2) FAB and its counterpart, liquid secondary ion mass spectrometry (LSIMS), facilitated the formation of abundant molecular ions, protonated molecules, and deprotonated molecules of non-volatile and thermally labile compounds such as peptides, chlorophylls, and complex lipids up to approximately mlz 12,000 FAB and LSIMS use energetic particle bombardment (fast atoms or ions from to 30,000 V of energy) to ionize compounds dissolved in non-volatile matrices such as glycerol or 3-nitrobenzyl alcohol and desorb them from this condensed phase into the gas phase for mass spectrometric analysis (see Fig 13.1) Protonated or deprotonated molecules are usually abundant and fragmentation is minimal Introduced in the late 1980s, matrix-assisted laser desorption ionization (MALDI) has helped solve the mass limit barriers of laser desorption mass spectrometry so that singly charged ions may be obtained up to mlz 500,000 and sometimes higher (3) For most commercially available MALDI mass spectrometers, ions up to mlz 200,000 are readily obtained Like FAB and LSIMS, MALDI samples are mixed with a matrix to form a solution that is loaded onto the sample stage for analysis Unlike the other matrix-mediated techniques, the solvent is evaporated before MALDI analysis, leaving sample molecules trapped in crystals of solid phase matrix The MALDI matrix is selected to absorb the pulse of laser light directed at the sample Most MALDI mass spectrometers are equipped with a pulsed UV laser, although IR lasers are available as an option on some commercial instruments Therefore, matrices are often substituted benzenes or benzoic acids with strong UV absorption properties During MALDI, the energy of the short but intense UV laser pulse obliterates the matrix and in the process desorbs and ionizes the sample Like FAB and LSIMS, MALDI typically produces abundant protonated or deprotonated molecules with little fragmentation By the time that GC-MS had become a standard technique in the late 1960s, LC-MS was still in the developmental stages Producing gas-phase sample ions for analysis in a vacuum system while removing the high perfor- Introduction 587 Figure 13.1 Scheme for desorption ionization using FAB or LSIMS from a liquid matrix (0) mance liquid chromatography (HPLC) mobile phase proved to be a challenging task Early LC-MS techniques included a moving belt interface to desolvate and transport the HPLC eluate into an CI or EI ion source or a direct inlet system in which the eluate was pumped at a low flow rate (1-3 pL/min) into a CI source However, neither of these systems was robust enough or suitable for a broad enough range of samples to gain widespread acceptance Because FAB (or LSIMS) requires that the analyte be dissolved in a liquid matrix, this ionization technique was easily adapted for infusion of solution-phase samples into the FAB ionization source in an approach known as continuous-flow FAB Then, continuous-flow FAB was connected to microbore HPLC columns for LC-MS applications (4) Because this method is limited to microbore HPLC applications at flow rates of 1)than the corresponding singly charged species This has the benefit of allowing mass spectrometers with modest mlz ranges to detect and measure ions of molecules with very high masses For example, electrospray has been used to measure ions with molecular weights of hundreds of thousands or even millions of Daltons on mass spectrometers with mlz ranges of only a few thousand (For a review of LC-MS techniques, see Niessen 1999, Section 4.) An example of the C,, reversed phase HPLC-negative ion electrospray mass spectrometric (LC-MS) analysis of an extract of the botanical, Trifoliumpratense L (red clover),is shown in Fig 13.3 Extracts of red clover are used as dietary supplements by menopausal and postmenopausal women and are under investigation as alternatives to estrogen replacement therapy (6) The two-dimensional map illustrates the amount of information that may be acquired using hyphenated techniques such as LC-MS In the time dimension, chromatograms are obtained, and a sample computer-reconstructed mass chromatogram is shown for the signal at mlz 269 An intense chromatographic peak was detected eluting at 12.4 In the mlz dimension, the negative ion electrospray mass spectrum recorded at 12.4 shows a base peak at mlz 269 Based on comparison with authentic standards (data not shown), the ion of mlz 269 was found to correspond to the deprotonated molecule of genistein, which is an estrogenic isoflavone (6) Because almost no fragmentation of the genistein ion was observed, additional charac- terization would require CID and MS-MS as discussed in the next section When analyzing complex mixtures such as the botanical extract shown in Fig 13.3, the use of chromatographic separation before mass spectrometric ionization and analysis is essential to distinguish between isomeric compounds Even simple mixtures of synthetic compounds might contain isomers that would require LC-MS for adequate characterization Another problem overcome by using a chromatography step before mass spectrometric analysis is ion suppression No matter what ionization technique is used, the presence of multiple compounds in the ion source might enhance the ionization of one compound while suppressing the ionization of another Usually, only some of the compounds in a complex mixture can be detected by mass spectrometry without chromatographic separation The presence of salts and buffers in a sample can also suppress sample ionization Therefore, LC-MS has become a powerful tool for analyzing natural products, synthetic organic compounds, and pharmaceutical agents and their metabolites In general, APCI facilitates the ionization of non-polar and low molecular weight species, and electrospray is more useful for the ionization of polar and high molecular weight compounds In this sense, APCI and electrospray are often complementary ionization techniques However, during the analysis of large or diverse combinatorial libraries, both polar and non-polar compounds are usually present As a result, no one set of ionization conditions using APc'I or electrospray is adequate to detect all the compounds contained in the library of compounds Therefore, a UV ionization technique called atmospheric pressure photoionization (APPI) has been developed for use with combinatorial libraries and LC-MS (7) Recently, APPI became a commercially available ionization alternative to APCI and electrospray During APPI, a liquid solution or HPLC eluate is sprayed at atmospheric pressure, as in APCI Instead of using a corona discharge as in APCI, ionization occurs during APPI because of irradiation of the analyte molecules by an intense UV light source Obviously, the carrier solvent must not absorb UV light at the same wavelengths, or interfer- Mass Spectrometry and Drug Discovery Computer-reconstructed mass chromatogram of m/z 269 500 , , , 1 I , I , I 1 , 1 I I , 1 , I I I I I , 1 , , I I / I , I I I 8 , , I .+ # * , , * * I , , , I I I I I I I , I , , , , / , , ,, I , , / I I , I I C I , / , , , I I ~ I , , I , I I , , , , , , , I , , , I I I I , J ! I I I I I I I I I ( I I I I I > I _ I , I ! I I I , I ! I ~ I , I I I I I I I T , I ~ I ., 2r , x; ., ., ., I _ / T , I ~ I I I I I I I i I I , I , , I I I I , I i t , I I , I , I I , I I I , l , , , , , I I , , , I , I , I l I , I , I I I , , , l""', "'l' , , I I I * .* * + ., I ) , , I I I I I I I J - , I / , , I I , , I , , I , , , I I , , I , I , / , I " I # , 1 1 I , , , , , I , '' , ", m O0 u C % 8 , / 8 , , I I I I , I I , I T ' , , I I , I _ , , , I I # , , , , I , I , , * I - , , - I I I , , _ '4 J I , I - I , I L - i I I I ' , , , I I ~ I I I I I I ~ I I I I , I I I , , , I I 8 , I , I I I ' I - I + , I I - , I , - d , L - I I - l , ' , I - - I L I I I - # , - / I - I , J ' , L 1 .> L / - I - I k I- I I I ~ ! , 1: I I I I I _ I ! I , I I , I I I I I ' I - C T - I , I , I I I ~ , I I I , J 8 - I I , , , , I I I 1 I - T i - , _ I I , L I , I I , I I ., : I i J ., I , , J # I I , I I / 1 150 l J , - , J I / I , ! ~ I I , , L , .I7 , .l , L ,, ,, ,, ,, ,, , , """'r'"" T ' , ! : ._* * .+ _ _-,_ , , : * , , , * +- , , , ., .+ * .> ., , , , +:4T ".r: ., , : ; ; ; f : LI-:: : I.-: ,- , v T , , ., , l I , , , , , , T , I l ~ ~ , , , I I , I I 1 , 16 I I / I I I I ' I I 20 / I J , I I I I 24 , I I I I I I I I I , I I t / I , I I I I ~ I I ~ I I I I I 28 Retention time (min) Mass spectrum at 12.4 269 [M-HI- Figure 13.3 Two-dimensional map showing the LC-MS analysis of an extract of red clover under investigation for the management of menopause Reversed phase separation was carried out using a C,, HPLC column in the time dimension and negative ion electrospray mass spectrometry was used for compound detection and molecular weight determination in the second dimension ence would prevent sample ionization and detection The use of APPI as an alternative to APCI and electrospray for drug discovery applications is under investigation Desorption ionization techniques like FAB, MALDI, and electrospray facilitate the molecular weight determination of a wide range of polar, non-polar, and low, and high molecular weight compounds including drugs and drug targets such as proteins However, the "soft" ionization character of these techniques means that most of the ion current is concentrated in molecular ions, and few structurally significant fragment ions are formed To en- hance the amount of structural information in these mass spectra, CID may be used to produce more abundant fragment ions from molecular ion precursors formed and isolated during the first stage of mass spectrometry Then, a second mass spectrometry analysis may be used to characterize the resulting product ions This process is called tandem mass spectrometry or MS-MS and is illustrated in Fig 13.4 Another advantage of the use of tandem mass spectrometry is the ability to isolate a particular ion such as the molecular ion of the analyte of interest during the first mass spec- ~ Current Trends and Recent Developments 100 0 536 C m z M- n m - 50 -9 -m 300 340 380 420 460 mlz 500 trometry stage This precursor ion is essentially purified in the gas-phase and free of impurities such as solvent ions, matrix ions, or other analytes Finally, the selected ion is fragmented using CID and analyzed using a second mass spectrometry stage In this manner, the resulting tandem spectrum contains exclusively analyte ions without impurities that might interfere with the interpretation of the fragmentation patterns In summary, CID may be used with LC-MS-MS or desorption ionization and MS-MS to obtain structural information such as amino acid sequences of peptides and sites of alkylation of nucleic acids, or to distinguish structural isomers such as p-carotene and lycopene Beginning in 2001, TOF-TOF tandem mass spectrometers became available from instrument manufacturers These instruments have the potential to deliver high resolution tandem mass spectra with high speed that should be compatible with the chip-based chromatography systems now under development Over the course of the last century, mass spectrometry has become an essential analytical tool for a wide variety of biomedical applications including drug discovery and development By combining mass spectrometry with chromatography as in LC-MS or by adding another stage of mass spectrometry as in MS-MS, the selectivity of the technique increases considerably As a result, mass spectrometry offers all of the analytical ele- 540 580 Figure 13.4 Scheme illustrating the selectivity of MS-MS and the process by which CID facilitates fragmentation of preselected ions Negative ion electrospray tandem mass spectrum of lycopene CID was used to induce fragmentation of the molecular ion of mlz 536 As a result, the fragment ion of mlz 467 was formed by the loss of a terminal isoprene unit This fragment ion may be used to distinguish lycopene from isomeric a-carotene and p-carotene, which lack terminal isoprene groups ments that are essential to modern drug discovery namely speed, sensitivity, and selectivity CURRENT TRENDS AND RECENT DEVELOPMENTS Since the early 1990s, pharmaceutical research has focused on combinatorial chemistry (8,9) and high-throughput screening (10) in an effort to accelerate the pace of drug discovery The goal has been to produce, in a short time, large numbers of synthetic organic compounds representing a great diversity of chemical structures through a process called combinatorial chemistry and then quickly screen them in vitro against pharmacologically significant targets such as enzymes or receptors The "hits" identified through these high-throughput screens may then be optimized by quickly and efficiently synthesizing and then screening large numbers of analogs called targeted or directed libraries As a result, lead compounds might emerge from such combinatorial chemistry drug discovery programs in a few weeks instead of several years Furthermore, a single organic chemist using combinatorial synthetic methods might synthesize thousands of compounds or more in a single week instead of less than five in the same time using conventional techniques, and a single medicinal chemist might identify hun- t Mass Spectrometry and Drug Discovery dreds of lead compounds per month instead of just one or two in the same period of time Accompanying this new drug discovery paradigm, new scientific journals have been established such as Combinatorial Chemistry & High Throughput Screening, Journal of Combinatorial Chemistry, Journal of Biomolecular Screening, and Molecular Diversity (see list of journal websites in Section 4) The variety of topics published in these journals reflects the multidisciplinary nature of the current drug discovery process and ranges from organic chemistry, medicinal chemistry, molecular modeling, molecular biology, and pharmacology, to analytical chemistry As described below, the most significant analytical component of drug discovery has become mass spectrometry Only mass spectrometry has become an essential element at all stages of the drug discovery and development process Although a variety of spectroscopic and chromatographic techniques, including infrared spectroscopy, nuclear magnetic resonance spectroscopy, fluorescence spectroscopy, gas chromatography, HPLC, and mass spectrometry, are being used to support drug discovery in various capacities, some of them, such as gas chromatography and fluorescence spectroscopy, are not applicable to most new chemical entities, some are not specific enough for chemical identification (e.g., infrared spectroscopy), and other techniques suffer from low throughput (e.g., nuclear magnetic resonance spectroscopy) Unlike gas chromatography, HPLC is compatible with virtually all drug-like molecules without the need for chemical derivatization to increase thermal stability or volatility In addition, mass spectrometry provides a universal means to characterize and distinguish drugs based on both molecular weight and structural features while at the same time providing high throughput With the development of routine LC-MS interfaces and ionization techniques such as electrospray and APCI, mass spectrometry has also become an ideal HPLC detector for the analysis of combinatorial libraries (ll),and LC-MS, MS-MS, and LC-MS-MS have become fundamental tools in the analysis of combinatorial libraries and subsequent drug development studies (12-14) The application of combinatorial chemistry and high-throughput screening to drug discovery has altered the traditional serial process of lead identification and optimization that previously required years of human effort Consequently, neither the synthesis of new chemical entities nor their screening is limiting the pace of drug discovery Instead, a new bottleneck is the verification of the structure and purity of each compound in a combinatorial library or of each lead compound obtained from an uncharacterized library using high-throughput screening Because the number of lead compounds entering the drug development process has increased, in part because compounds are entering development at earlier stages than in the past, the traditional drug development investigations concerning absorption, distribution, metabolism, and excretion (ADME) and even toxicology evaluations of new drug entities have become additional bottlenecks As a solution to the drug development bottlenecks, high-throughput assays to assess the metabolism, bioavailability, and toxicity of lead compounds are being developed and applied earlier than ever during the drug discovery process, so that only those compounds most likely to become successful drugs enter the more expensive and slower preclinical pharmacology and toxicology studies In support of these new combinatorial chemistry synthetic programs and new highthroughput assays, mass spectrometry has emerged as the only analytical technique with sufficient throughput, sensitivity, selectivity, and robustness to address all of these bottlenecks 2.1 LC-MS Purification of Combinatorial Libraries Although combinatorial libraries were originally synthesized as mixtures, today most libraries are prepared in parallel as discrete compounds and then screened individually in microtiter plates of 96-well, 384-well, or 1536well formats To facilitate subseauent structure-activity analyses and to assure the validity of the screening results, many laboratories verify the structure and purity of each compound before high-throughput screening Semi-preparative HPLC has become the most Current Trends and Recent Developments (a) 8e7 m PrepLCMS analysis 50 rng inj (c) Analytical purity assessment Desire1 produc \ Purity = 90.7% \ Time, RIC of desired peak from prepLCMS (b) 6.49 rnin $ 4e6 c Threshold for fraction collection a " 2e6 I I I Time, Time, Figure 13.5 Mass-directed purification of a combinatorial library Chromatographic separation was carried out using gradient elution of 10-90%acetonitrile in water for after an initial hold at 10%acetonitrile for (a) Total ion chromatogram showing desired product and impurities (b) Computer-reconstructed ion chromatogram (RIC) corresponding to the expected product (c) Postpurification analysis of the isolated component with a purity >go% (Reproduced from Ref 15 by permission of Elsevier Science.) - popular technique for the purification of combinatorial libraries on the milligram scale because of high throughput and the ease of automation Typically during semi-preparative HPLC, fraction collection is initiated whenever a UV signal is observed above a predetermined threshold This procedure usually results in the collection of several fractions per analysis and hence creates additional issues such as the need for large fraction collector beds and the need for secondary analysis using flow-injection mass spectrometry, LC-MS, or LC-MS-MS to identify the appropriate fractions When purification of large numbers of combinatorial libraries is required, this approach can become prohibitively time consuming and expensive To enhance the efficiently of this purification procedure, the steps of HPLC purification and mass spectrometric analysis may be combined into automated mass-directed fraction- ation (15-17) Any size HPLC column may be used, and only a small fraction of the eluant (-pL/min) is diverted to the mass spectrometer equipped for APCI or electrospray ionization Because all of the components, including autosampler, injector, HPLC, switching valve, mass spectrometer, and fraction collector, are controlled by computer, the procedure may be fully automated For greatest efficiency, the system may be programmed to collect only those peaks displaying the desired molecular ions, or alternatively, all peaks displaying abundant ions within a specified mass range An example of the MS-guided purification of a compound synthesized during the parallel synthesis of a combinatorial library of discrete compounds is shown in Fig 13.5 Although the crude yield of the reaction product was only 30% (Fig 13.5a), the desired product was detected based on its molecular ion (Fig 13.5b) After MS-guided fractionation, re-analysis us- Mass Spectrometry and Drug Discovery ing LC-MS showed that the desired product was >90% pure (Fig 13.5~) The use of MS-guided purification of combinatorial libraries provides a means for reducing the number of HPLC fractions collected per sample and eliminates the need for post-purification analysis to further characterize and identify each compound as would be necessary when using UV-based fractionation The ionization technique (i.e., electrospray, APCI, or APPI), and ionization mode (positive or negative) must be suitable for the combinatorial compound so that molecular ion species are formed Also, a suitable mobile phase and HPLC column must be selected As an alternative to HPLC, supercritical fluid chromatography-mass spectrometry (SFC-MS) has been used for the high-throughput analysis of combinatorial libraries (18, 19) The advantages of SFC-MS relative to conventional LC-MS for the purification of combinatorial libraries of compounds are the lower viscosities and higher diffusivities of condensed CO, compared with HPLC mobile phases and the ease of solvent removal and disposal after analysis However, SFC instrumentation remains more expensive and less widely available than conventional HPLC systems 2.2 Confirmation of Structure and Purity of Combinatorial Compounds The determination of molecular weights, elemental compositions, and structures of compounds used for high-throughput screening, whether discrete compounds or combinatorial library mixtures, is typically carried out using mass spectrometry, because traditional spectroscopic and gravimetric techniques are too slow to keep pace with combinatorial chemical synthesis In addition, mass spectrometry may be used to assess the purity of compounds being used for high-throughput screening The highest-throughput technique for confirming molecular weights and structures of drug candidates is flow injection analysis of sample solutions using electrospray, APCI, or APPI mass spectrometry Typically, no sample preparation is necessary Although any organic mass spectrometer may be used to confirm the molecular weight of a compound, tandem mass spectrometers provide additional structural information through the use of CID to produce fragment ions As discussed above (see also Table 13.11, tandem mass spectrometers include triple quadrupole instruments, QqTOF mass spectrometers, ion trap mass spectrometers, multiple sector magnetic sector instruments, FTICR instruments, and the new TOF-TOF mass spectrometers In most applications, APCI or electrospray ionization is used In addition to molecular weight and fragmentation patterns, high precision and high resolution mass spectrometers such as QqTOF instruments, reflectron TOF mass spectrometers, double focusing magnetic sector mass spectrometers, and FTICR instruments are necessary for the measurement of exact masses of drugs and drug candidates for the determination of elemental compositions The combination of high resolution and high precision is especially useful for determining the elemental compositions of compounds in combinatorial library mixtures without having to isolate each compound using chromatography or some other separation technique Because FTICR instruments and the hybrid QqTOF mass spectrometers are capable of simultaneously measuring exact masses at high resolution of both molecular ions and fragment ions generated during MS-MS, these instruments are becoming extremely popular within drug discovery programs As an example of the exact mass measurement of a combinatorial library mixture, the FTICR negative ion electrospray mass spectra of a 36- and a 120-compound peptide library mixture are shown in Fig 13.6 The resolution achieved in this experiment was 20,00040,000 Although the exact masses of all components in a small combinatorial library can often be measured during a single infusion experiment, on-line HPLC separation or the analysis of discrete compounds is sometimes required to overcome ion suppression problems However, LC-MS is a relatively slow process because of the slow chromatographic separation step Because LC-MS is required in many instances for the analysis of mixtures and to eliminate interfering salts or buffers, two approaches have emerged to increase the throughput of this technique; parallel LC-MS and fast LC-MS One approach to increasing Current Trends and Recent Developments (a) Pro-X-Asp 621.2817 1.6 ppm Asn-X-Asp 638.2718 1.6 ppm 621.2807 I I I ~ I I I I 620 ~ I I I I ~ I I I I ~ 630 (b) 637.3136 I Asp-X-Asp 639.2559 3.8 ppm I I I Ala-X-TyrMe 657.3181 ASP-X-ASP 639.2559 / 1.4 PPm I ~ I I I I 640 ~ I I I 650 PPm I I ~ I I I I ~ I I I I ~ I I 660 mlz ThrMe-X-Asp 639.2922 2.5 ppm / Figure 13.6 (a) Partial negative ion electrospray mass spectrum of a 36-component library mixture Both the measured mass and the difference between the measured and theoretical values (in ppm) are shown (b) Negative ion electrospray spectrum of the 120-component library showing the resolution of three nominally isobaric peaks (Reproduced from b f 24 by permission of Bentham Science Publishers) throughput of the rate-limiting chromatographic separation has been to simultaneously interface multiple HPLC columns to a single mass spectrometer This approach is called parallel LC-MS Commercial parallel electro- spray interfaces and HPLC systems are now available that can accommodate up to eight HPLC columns simultaneously (20-22) Although the multiple sprays are introduced to the ion source simultaneously, these streams I I J I I Mass Spectrometry and Drug Discovery may be sampled in a time-dependent manner to minimize cross contamination between channels Another solution to increasing the throughput of LC-MS has been to minimize the time required for HPLC separation through an approach called fast HPLC HPLC separations are accelerated by using shorter columns and higher mobile phase flow rates Because coelution of some species is likely to occur during fast chromatographic separations, the selectivity of the mass spectrometer is essential for the characterization and/or quantitative analysis of the target compound However, samples of compounds prepared using combinatorial chemistry are usually simple mixtures of reagents, by-products, and products that require only partial chromatographic purification to prevent ion suppression effects during mass spectrometric analysis In addition to molecular weight determination using conventional MS or high exact mass measurement and structural confirmation using MS-MS, fast LC-MS is also used to assess the purity and yield of combinatorial products (15, 23) Before high-throughput screening, many researchers analyze combinatorial libraries for both purity and structural identity using mass spectrometry to assure the validity of structure-activity relationships that might be derived from the screening data Fast LC-MS and LC-MS-MS may be carried out to satisfy this requirement using gradients (usually a step gradient with a reverse phase HPLC column) with a total cycle time of 1-3 (24) or using an isocratic system requiring less than per analysis A variety of HPLC columns are used for fast LC-MS that include narrow bore (2-mm) and analytical bore (4.6-mm) columns with length typically from 0.5-5 cm The mobile phase flow rate for these fast LC-MS analyses is usually from 1.5-5 mL/min 2.3 Encoding and Identification of Compounds in Combinatorial Libraries and Natural Product Extracts The use of mass spectrometric identification in combinatorial chemistry is not limited to the analysis of synthetic products as a means of quality control, but also for the identifica- tion of active compounds or "hits" during high-throughput screening Although the synthesis and screening of discrete compounds (25) enables them to be followed through the entire process by using partial encoding or bar-coding, it is sometimes advantageous to screen libraries prepared as mixtures (26) and use a technique such as mass spectrometry to rapidly identify the hit(s) in the mixture One approach to the rapid deconvolution of combinatorial library mixtures is to prepare libraries containing compounds of unique molecular weight and then identify them using mass spectrometry However, such libraries are necessarily small because the molecular weight of most drug-like molecules is between 150-400 Da Because of the molecular weight degeneracy of larger combinatorial libraries, several encoding strategies have been devised to rapidly identify active compounds in these mixtures (27-29) Because most combinatorial libraries contain compounds with degenerate molecular weights, various tagging strategies have been devised to uniquely identify library compounds bound to beads Most of these tagging approaches are based on the synthesis of encoding molecules For example, peptide (30) or oligonucleotide (31) labels have been synthesized on the beads in parallel to the target molecules and then sequenced for bead decoding Alternatively, haloarene tags have been incorporated during synthesis and then identified with high sensitivity using electron-capture gas chromatography detection (32) In addition to the increased time and cost for the synthesis of a library containing tagging moieties, the tagging groups themselves might interfere with screening giving false positive or negative results For peptide libraries, one solution to this problem uses matrix-assisted laser desorption ionization (MALDI) mass spectrometry to directly desorb and identify peptides from beads that were screened and found to be hits (33) This technique is called the termination synthesis approach Because the peptide library compounds are analyzed directly, products with amino acid deletions or substitutions, side-reaction products, or incomplete deprotection are readily observed Also, because there are no extra molecules used for chemical Current Trends and Recent Developments tagging, this source of interference is avoided However, this approach is specific to peptide libraries and is not necessarily applicable to other types of combinatorial libraries Another approach that eliminates possible interference from the chemical tags, "ratio encoding," has been developed for the mass spectrometric identification of bioactive leads using stable isotopes incorporated into the library compounds (29,34) Within the ligand itself, the code might be a single-labeled atom that is conveniently inserted whenever a common reagent transfers at least one atom to the target compound or ligand The code consists of an isotopic mixture having one of the many predetermined ratios of stable isotopes and can be incorporated in the linker or added through a reagent used during the synthesis The mass spectrum of the compound shows a molecular ion with a unique isotope ratio that codes for a particular library compound For example, Wagner et al (29) used isotope ratio encoding during the synthesis of a 1000-compound peptoid library and was able to identify uniquely all the components based on their isotopic patterns and molecular weights Because isotope ratio codes are contained within each combinatorial compound, a chemical tag is not required The speed of MS-based decoding outperforms most other decoding technologies, which are time consuming and decode a restricted set of active compounds Although combinatorial synthesis provides rapid access to large numbers of compounds for screening during drug discovery and lead optimization, these libraries are usually based on a small number of common structures or scaffolds There is a constant need for increasing the molecular diversity of combinatorial libraries and finding new scaffolds, and natural products have always been a rich source of chemical diversity for drug discovery The traditional approach to screening natural products for drug leads uses bioassays to test organic solvent extracts for activity If strong activity is detected, then activity-guided fractionation of the crude extract is used to isolate the active compound(s),which is identified using mass spectrometry (including tandem mass spectrometry and exact mass measurements), IR, W M S spectrometry, and NMR Recently, a variety of mass spectrometry- 597 based affinity screening methods have been developed to streamline the tedious process of activity-guided fractionation These approaches are discussed in Section 2.4 Whether lead compounds in natural product extracts are isolated using bioassay-guided fractionation or mass spectrometry-based screening, there is a high probability that the structure of the active compound(s) has already been reported in the natural product literature In such cases, the tedious process of complete structure elucidation using a battery of spectrometric tools should be unnecessary Instead, mass spectrometry alone may be used to quickly "dereplicate" or identify the known compounds based on molecular weight, fragmentation patterns, and elemental composition in combination with natural product database searching (35-39) Commercially available natural products databases include NAPRALERT (40), Scientific & Technical Information Network (STN) (41), and the Dictionary of Natural Products (42) Because some of these databases also contain WIVIS absorbance data, it is also advantageous to use a photodiode array detector between the HPLC and mass spectrometer to obtain additional spectrometric data during LC-W-MS dereplication (36, 37) 2.4 Mass Spectrometry-Based Screening The earliest approaches to combinatorial synthesis used portioning and mixing (26)and enabled the synthesis of combinatorial libraries containing hundreds of thousands to millions of compounds Today, this approach remains the most efficient method for preparing enormous libraries of compounds However, until the mid-1990s, efficient screening techniques did not exist to rapidly identify the "hits" within large combinatorial mixtures Therefore, chemists were motivated to develop ways to prepare large numbers of discreet compounds using massively parallel synthesis, which could be assayed quickly for pharmacological activity using high throughput screening one compound at a time Recently, several mass spectrometry-based screening assays have been developed that are suitable for screening combinatorial library mixtures, and some are even useful for screening natural product extracts which have always been a Mass Spectrometry and Drug Discovery Binding Library + R R R + Affinity column Wash unbound library compounds to waste Isolation Elute bound ligands using pH change Trap ligands on C,, column LC-MS-MS identification Elute trapped ligands onto HPLC column I I Figure 13.7 Affinity chromatography combined with LC-MS-MS for screening combinatorial library mixtures source of molecular diversity for drug discovery All of the mass spectrometry-based screening methods use receptor binding of ligands as the basis for identification of lead compounds 2.4.1 Affinity Chromatography-Mass Spectrometry Since the introduction of affinity chromatography more than 30 years ago, this technique has become a standard biochemical tool for the isolation and identification of new binding partners to specific target molecules Therefore, the coupling of affinity chromatography to mass spectrometry is a logical extension of this technique, and the application of affinity LC-MS to the screening of combinatorial libraries has been demonstrated by several groups (43, 44) During affinity LC-MS screening, a receptor molecule such as a binding protein or enzyme is immobilized on a solid support within a chromatography column The library mixture is pumped through the affinity column in a suitable binding buffer so that any ligands in the mixture with affmity for the receptor would be able to bind Then, unbound material is washed away Finally, the specifically bound ligands are eluted using a destabilizing mobile phase and identified using mass spectrometry This affinitycolumn LC-MS assay is summarized in Fig 13.7 In some applications (43), ligands are eluted from the affinity column and then trapped on a second column such as a reverse phase HPLC column LC-MS or LC-MS-MS identification of the ligands (hits) is then carried out using the trapping column In other systems, ligands are identified directly from the affinity column using mass spectrometry (44) For example, Kelly et al (44) prepared an affinity column containing immobilized phosphatidylinositol-3-kinase and used it for direct LC-MS screening of a 361-component peptide library Electrospray mass spectrometry and tandem mass spectrometry were used to identify the ligands released from the affinity column using pH gradient elution Advantages of affinity chromatographymass spectrometry for screening during drug discovery include versatility and re-use of the column Both combinatorial libraries and natural product extracts can be screened using this approach, and a wide range of binding buffers may be used Mass spectrometry-compatible mobile phases are only required during the final LC-MS detection step Furthermore, a single column may be used multiple times to screen different samples for ligands unless the destabilization solution irreversibly denatures, releases, or inhibits the receptor Despite these advantages, affinity chromatography has numerous drawbacks that have - Current Trends and Recent Developments Binding + L-R R + O 0 0 0 GPC isolation Figure 13.8 GPC followed by LC-MS-MS L-R n identification Reversed phase desaltingldenaturation prompted the development of alternative mass spectrometer screening tools For example, immobilization of the receptor might change its affinity characteristics causing false negative or false positive hits This is particularly problematic for receptors that are solution-phase in their native state Also, developing and then implementing an immobilization scheme is often a slow, tedious, and even expensive process, and this process is unique for each new receptor Finally, false positive hits are often obtained when screeninglarge molecularly diverse libraries, because there are usually compounds in such mixtures that have affinity for the stationary phase or linker molecule instead of the receptor 2.4.2 Gel Permeation ChromatographyMass Spectrometry Another type of chroma- tography that has been combined with mass spectrometry as a screening system for drug discovery is gel permeation chromatography (GPC) (45,461.Also called size-exclusion chromatography, GPC separates molecules according to size as they pass through a stationary phase containing particles with a defined pore size During GPC-based screening, a library mixture is pre-incubated with a macromolecular receptor to allow any ligands in the library to bind, and then GPC is used to separate the large receptor-ligand complexes from the unbound low molecular weight compounds in the mixture Finally, ligands are released from the receptor during reversed for screening mixtures of combinatorial libraries After incubation of a receptor with a library of compounds, the ligand-receptor complexes (L-R) are separated from the low molecular weight unbound library compounds using GPC Next, the L-R complexes are denatured during reversed phase HPLC to release the ligands for MS-MS identification phase HPLC and identified either on-line or off-line using tandem mass spectrometry This screening method is illustrated in Fig 13.8 During the pre-incubation and GPC steps, any binding buffer may be used, because the binding buffer will be removed during reverse phase LC-MS analysis However, the GPC separation step must be carried out quickly, because ligands begin to dissociate from the receptor immediately and can become lost into the size exclusion gel Despite this disadvan-' tage, this approach allows both receptor and ligand to be screened in solution, which avoids some of the problems associated with the use of affinity columns for screening The GPC LC-MS-MS screening method should also be suitable for screening natural product extracts as well as combinatorial library mixtures 2.4.3 Affinity Capillary ElectrophoresisMass Spectrometry Affinity capillary electro- phoresis was originally used for the determination of the binding constants of small molecules to proteins (47-49) This solutionbased technique is rapid and requires only small amounts of ligands Affinity constants are measured based on the mobility change of the ligand on interaction with the receptor present in the electrophoretic buffer (50) By combining affinity capillary electrophoresis with on-line mass spectrometric detection and Mass Spectrometry and Drug Discovery I I I I I I 1 Migration time (min) Figure 13.9 Affinity capillary electrophoresis-UV-mass spectrometry of a 100-tetrapeptide library weened for binding to vancomycin (104 pikf in the electrophoresis buffer) (a) The elution of peptides was monitored with UV absorbance during capiuary electrophoresis, and the elution time increased with increasing affinity for vancomycin 6) Positive ion electmspray mass spedrum with CID of the Tris adduct of the protonated peptide detected at -5 rnin in the electropherogram shown in a (Reproduced from Ref 52 by permission of the American Chemical Society.) identification, affinity constants for multiple compounds can be measured in a single analysis (51) Recognizing that on-line mass spectrometric detection was helpful for the identification of each ligand, Chu et al (52) extended this approach to include the screening of combinatorial libraries as a means of drug discovery The data in Fig 13.9 show the results of screening a 100-tetrapeptide library for affinity to vancomycin using affinity capillary electrophoresis-mass spectrometry Without vancomycin in the electrophoresis buffer, all the peptides eluted within When vancomycin was present, the peptides eluted in order of affinity, with the highest affmity compounds being detected between 4.5 and Positive ion electrospray tandem mass spectrometry was used to identify the highest affinity ligands (see Fig 13.9b) Note that some peptide ligands such as Fmoc-DDFA were detected as adducts with Tris, which was used in the electrophoresis buffer Although the identification of this peptide was not prevented by the formation of this adduct, some buffers used during electrophoresis might interfere with mass spectrometric ionization and detection Also, the types of libraries that have been screened using this approach have contained modest numbers of synthetic analogs such as peptides Libraries exceeding 400 members required preliminary purification using affinity chromatography to reduce the number of compounds (52) As a result, this approach is probably not ideal for screening libraries containing molecularly diverse compounds or for screening natural product extracts However, affinity capillary electrophoresis-mass spectrometry is fast; each analysis requires less than 10 Also, it may be used to measure affinity constants for ligand-receptor interactions 2 Current Trends and Recent Developments 2.4.4 Frontal Affinity ChromatographyMass Spectrometry Like affinity chromatog- raphy-mass spectrometric screening (see Section 2.4.1), frontal affinity chromatography uses an aMinity column containing immobilized receptor molecules (53) The difference between the two screening methods is that the ligands are continuously infused into the column during frontal affinity chromatography and detected using mass spectrometry Compounds with no affinity for the immobilized receptor elute immediately in the void volume, but the elution of the ligands is delayed As compounds compete for binding sites on the affinity column, these sites become saturated until ligands begin to elute from the column at their infusion concentration In this manner, frontal affinity chromatography may be used to measure affinity constants for ligands, and by using a mass spectrometer for on-line identification of ligands, this technique becomes a screening method (54,55) During frontal affinity chromatographymass spectrometry, signals for all compounds eluting from the affinity column are recorded by the mass spectrometer, and the last compounds to elute at their infusion concentrations represent the highest affinity compounds or "hits." An example of the screening of six oligosaccharides with different binding affinities for an immobilized monoclonal carbohydrate-binding antibody is shown in Fig 13.10 Compounds 1-3 eluted immediately (no affinity), whereas compounds 4-6 eluted in order of increasing affinity for the antibody Dissociation constants were determined to be 185, 12.6, and 1.8 p M for compounds 4-6, respectively (54) Because frontal affinity chromatography uses a conventional affinity column, this technique provides additional applications of this type of column to investigators already using affinity-mass spectrometry (See Section 2.4.1) However, the same limitations and disadvantages of using immobilized receptors still apply, such as non-specific binding to the stationary phase, the development time and cost of preparing the affinity columns, and the possibility that immobilizing the receptor might alter its binding characteristics and specificity In addition, mass spectrometric detection creates some additional limitations Because all library compounds must be monitored simultaneously, the compounds must be selected so that they have unique molecular weights Also, one compound in the mixture should not suppress the ionization of another Therefore, this approach is probably restricted to the screening of small combinatorial libraries that are similar in chemical structure and ionization efficiencies Finally, the binding buffer used for affinity chromatography must be compatible with on-line APCI or electrospray mass spectrometry This means that the mobile phase must be volatile and usually of low ionic strength (i.e., typically 1/3.5 kl), even for highly defocused images The use of higher voltages provides potentially higher resolution [greater depth of field (i.e., less curvature of the Ewald sphere) attributed to smaller electron beam wavelength], better beam penetration (less multiple scattering), 11 Selection and Preprocessing of Digitized Images reduced problems with specimen charging that plague microscopy of unstained or uncoated vitrified specimens (go), and reduced phase shifts associated with beam tilt Images are recorded on photographic film or on a CCD camera with either flood beam or spot-scan procedures Film, with its advantages of low cost, large field of view, and high resolution (-10 pm), has remained the primary image recording medium for most cryo-EM applications, despite disadvantages of high background fog and need for chemical development and digitization CCD cameras provide image data directly in digital form and with very low background noise, but suffer from higher cost, limited field of view, limited spatial resolution caused by poor point spread characteristics, and a fixed pixel size (typically between 14 and 24 pm) They are useful, for example, for precise focusing and adjustment of astigmatism [e.g., Krivanek and Mooney (81);Sherman et al (82)l For studies in which specimens must be tilted to collect 3D data, such as with 2D crystals, or single particles that adopt preferred orientations on the EM grid, or specimens requiring tomography, microscopy is performed in essentially the same way as described above However, the limited tilt range (26070") of most microscope goniometers can lead to nonisotropic resolution in the 3D reconstructions (the "missing cone" problem), and tilting generates a constantly varying defocus across the field of view in a direction normal to the tilt axis The effects caused by this varying defocus level must be corrected in high resolution applications 11 SELECTION A N D PREPROCESSING O F D I G I T I Z E D IMAGES Before any image analysis or classification of the molecular images can be done, a certain amount of preliminary checking and normhlization is required to ensure there is a reasonable chance that a homogeneous population of molecular images has been obtained First, good quality micrographs are selected in which the electron exposure is correct, there is no image drift or blurring, and there is minimal astigmatism and a reasonable amount of defocus to produce good phase contrast This is usually done by visual examination and optical diffraction Once the best pictures have been chosen, the micrographs must be scanned and digitized on a suitable densitometer The sizes of the steps between digitization of optical density and the size of the sample aperture over which the optical density is averaged by the densitometer must be sufficiently small to sample the detail present in the image at fine enough intervals (83) Normally, a circular (or square) sample aperture of diameter (or length of side) equal to the step between digitizations is used This avoids digitizing overlapping points, without missing any of the information recorded in the image - The size of the sample aperture and digitization step depends on the magnification selected and the resolution required A value of 114 to 113 of the required limit of resolution (measured in pm on the emulsion) is normally ideal because it avoids having too many numbers (and therefore wasting computer resources), without losing anything during the measurement procedure For a 40,000X image, on which a resolution of 10 A at the specimen is required, a step size of 10 pm { = 1/4 X [(lo A X 40,000)l (10,000 &pm)]) would be suitable The best area of an image of a helical or 2D crystal specimen can then be boxed off using a soft-edge mask For images of single particles, a stack of individual particles can be created by selecting out many small areas surrounding each particle Because, in the later steps of image processing, the orientation and position of each particle are refined by comparing the amplitudes and phases of their Fourier components, it is important to remove spurious features around the edge of each particle and to make sure the different particle images are on the same scale This is normally done by masking off a circular area centered on each particle and floating the density so that the average around the perimeter becomes zero (83) The edge of the mask is apodized by applying a soft cosine bell shape to the original densities so they taper toward the background level Finally, to compensate for variations in the exposure attributed to ice thickness or electron dose, most microscopists normalize the stack of individual particle images so that 624 the mean density and mean density variation over the field of view are set to the same values for all particles (84) Once some good particles or crystalline areas for 1D or 2D crystals have been selected, digitized, masked, and their intensity values normalized, true image processing can begin 12 I M A G E PROCESSING A N D D RECONSTRUCTION Although the general concepts of signal averaging, together with combining different views to reconstruct the 3D structure, are common to the different computer-based procedures that have been implemented, it is important to emphasize one or two preliminary points First, a homogeneous set of particles must be selected for inclusion in the 3D reconstruction This selection may be made by eye, to eliminate obviously damaged particles or impurities, or by the use of multivariate statistical analysis (85) or some other classification scheme This allows a subset of the particle images to be used to determine the structure of a better defined entity All imageprocessing procedures require the determination of the same parameters that are needed to specify unambiguously how to combine the information from each micrograph or particle These parameters are: the magnification, defocus, astigmatism, and, at high resolution, the beam tilt for each micrograph; the electron wavelength used (i.e., accelerating voltage of the microscope);the spherical aberration coefficient (C,) of the objective lens; and the orientation and phase origin for each particle or unit cell of the ID, 2D, or 3D crystal There are 13 parameters for each particle, of which eight may be common to each micrograph and two or three (C,, kV, magnification) to each microscope The different general approaches that have been used in practice to determine the 3D structure of different classes of macromolecular assemblies from one or more electron micrographs are listed in Table 14.2 The precise way in which each general approach codes and determines the particle or unit cell parameters varies greatly and is not described in detail Much of the computer software used in image reconstruction studies is Electron Cryornicroscopy of Biological Macromolecules relatively specialized compared to that used in the more mature field of macromolecular Xray crystallography In part, this may be attributed to the large diversity of specimen types amenable to cryo-EM and reconstruction methods As a consequence, image-reconstruction software is evolving quite rapidly, and references to software packages cited in Table 14.2 are likely to become quickly outdated Extensive discussion of algorithms and software packages in use at this time may be found in a number of recent special issues of the Journal of Structural Biology [volumes 116(1), 120(3), 121(2),and 125(2/3)1 In practice, attempts to determine or refine some parameters may be affected by the inability to determine accurately one of the other parameters The solution of the structure is therefore an iterative procedure in which reliable knowledge of the parameters that describe each image is gradually built up to produce an increasingly accurate structure, until no more information can be squeezed out of the micrographs At this point, if any of the origins or orientations is wrongly assigned, there will be a loss of detail and signal-to-noise ratio in the map If a better determined or higher resolution structure is required, it would then be necessary to record images on a better microscope or to prepare new specimens and record better pictures The reliability and resolution of the final reconstruction can be measured by use of a variety of indices For example, the differential phase residual (DPR) (1331, the Fourier shell correlation (FSC) (134), and the Q-factor (135) are three such measures DPR is the mean phase difference, as a function of resolution, between the structure factors from two independent reconstructions, often calculated by splitting the image data into two halves FSC is a similar calculation of the mean correlation coefficient between the complex structure factors of the two halves of the data as a function of resolution The Q-factor is the mean ratio of the vector sum of the individual structure factors from each image divided by the sum of their moduli, again calculated as a function of resolution Perfectly accurate measurements would have values of DPR, FSC, and Q-factor of O", 1.0, and 1.0 respectively, whereas random data containing no informa- 12 Image Processing and 3D Reconstruction Table 14.2 Methods of Three-Dimensional Image Reconstruction Type Structure (symmetry) Asymmetric (Point group C,) Symmetric (Point groups C,, D,; n > 1) Icosahedral (Point group I) Helical 2D Crystal 3D Crystal Reference(s)to Technicd Theoretical Details Method Random conical tilt *Software package Angular reconstitution *Software package Weighted back projection Radon transform alignment Reference-based alignment Reference free alignment Fourier reconstruction and alignment Tomographic tilt series and remote control of microscopea Angular reconstitution *Software packages Fourier-Bessel synthesis Reference-based alignment and weighted back projection Fourier-Bessel synthesis (common-lines) *Reference-based alignment *Software packages Angular reconstitution Tomographic tilt series Fourier-Bessel synthesis *Software packages and filament straightening routines Random azimuthal tilt *Software packages Oblique section reconstruction *Software package Sectioned 3D crystal "Note: Electron tomography is the subject of an entire issue of J Struct Biol [120,207-395 (199711and a book edited by Frank (132) tion would have values of go0,0.0, and 0.0 The spectral signal-to-noise ratio (SSNR) criterion has been advocated as the best of all (136): it effectively measures, as a function of resolution, the overall signal-to-noise ratio (squared) of the whole of the image data It is calculated by taking into consideration how well all the contributing image data agree internally An example of a typical strategy for determination of the 3D structure of a new and unknown molecule without any symmetry and that does not crystallize might be as follows: Record a single axis tilt series of particles embedded in negative stain, with a tilt range from - 60" to + 60" Calculate 3D structures for each particle by use of an R-weighted back-projection algorithm (93) Average 3D data for several particles in real or reciprocal space to get a reasonably good 3D model of the stain excluding the region of the particle Record a number of micrographs of the particles embedded in vitreous ice Use the 3D negative stain model obtained in (3) with inverted contrast to determine the rough alignment parameters of the particle in the ice images Calculate a preliminary 3D model of the average, ice-embedded structure Electron Cryomicroscopy of Biological Macromolecules Use the preliminary 3D model to determine more accurate alignment parameters for the particles in the ice images Calculate a better 3D model Determine defocus and astigmatism to allow CTF calculation and correct 3D model so that it represents the structure at high resolution 10 Keep adding pictures at different defocus levels to get an accurate structure at as high a resolution as possible For large single particles with no symmetry or for particles with higher symmetry or for crystalline arrays, it should be possible to miss out the negative staining steps and go straight to alignment of particle images from ice-embedding because the particle or crystal tilt angles can be determined internally from comparison of phases along common lines in reciprocal space or from the lattice or helix parameters from a 2D or 1D crystal The following discussion briefly outlines for a few specific classes of macromolecule the general strategy for carrying out image processing and 3D reconstruction (see Fig 14.6) 12.1 2D Crystals For 2D crystals, the general 3D reconstruction approach consists of the following steps: First, a series of micrographs of single 2D crystals are recorded at different tilt angles, with random azimuthal orientations Each crystal is then unbent using cross-correlation techniques, to identify the precise position of each unit cell (1271, and amplitudes and phases of the Fourier components of the average of that particular view of the structure are obtained for the transform of the unbent crystal The reference image used in the cross-correlation calculation can either be a part of the whole image masked off after a preliminary round of averaging by reciprocal space filtering of the regions surrounding the diffraction spots in the transform, or it can be a reference image calculated from a previously determined 3D model The amplitudes and phases from each image are then corrected for the CTF and beam tilt (11,22, 127) and merged with data from many other crystals by scaling and origin refinement, taking into account the proper symmetry of the 2D space group of the crystal Finally, the whole data set is fitted by least squares to constrained amplitudes and phases along the lattice lines (137) before calculating a map of the structure The initial determination of the 2D space group can be carried out by a statistical test of the phase relationships in one or two images of untilted specimens (138) The absolute hand of the structure is automatically correct, given that the 3D structure is calculated from images whose tilt axis and tilt angle are known Nevertheless, care must be taken not to make any of a number of trivial mistakes that would invert the hand 12.2 Helical Particles The basic steps involved in processing and 3D reconstruction of helical specimens include: Recording a series of micrographs of vitrified particles suspended over holes in a perforated carbon support film The micrographs are digitized and Fourier-transformed to determine image quality (astigmatism, drift, defocus, presence, and quality of layer lines, etc.) Individual particle images are boxed, floated, and apodized within a rectangular mask The parameters of helical symmetry (number of subunits per turn and pitch) must be determined by indexing the computed diffraction patterns If necessary, simple spline-fitting procedures may be employed to "straighten" images of curved particles (124), and the image data may be reinterpolated (126) to provide more precise sampling of the layer line data in the computed transform Once a preliminary 3D structure is available, a much more sophisticated refinement of all the helical parameters can be used to unbend the helices onto a predetermined average helix so that the contributions of all parts of the image are correctly treated (123) The layer line data are extracted from each particle transform and two phase origin corrections are made, one to shift the phase origin to the helix axis (at the center of the particle image) and the other to correct for effects caused by having the helix axis tilted out of the plane normal to the electron beam in the electron microscope The layer line data are separated out into nearand far-side data, corresponding to contributions from the near and far sides of each particle imaged The relative rotations and 627 Image Processing and 3D Reconstruction Figure 14.6 Examples of macromolecules studied by cryo-EM and 3D image reconstruction and the resulting 3D structures (bottom row) after cryo-EM analysis All micrographs (top row) are displayed at about 170,000X magnification and all models a t about 1,200,000x magnification (a) A single particle without symmetry: The micrograph shows 70s E coli ribosomes complexed with mRNA and Met-tRNA The surface-shaded density map, made by averaging 73,000 ribosome images from 287 micrographs has a resolution (FSC) of 11.5 A The 50s and 30s subunits and the tRNA are colored blue, yellow, and green, respectively The identity of many of the subunits is known and some RNA double helices are clearly recognizable by their major and minor grooves (e.g., helix 44 is shown in red) [Courtesy of J Frank (SUNY, Albany), using data from Gabashvili et al (86).1 (b) A single particle with symmetry: The micrograph shows hepatitis B virus cores The 3D reconstruction, a t a resolution of 7.4 A (DPR), was computed from 6384 particle images taken from 34 micrographs [From Bottcher et al (441.1 (c) A helical filament: The micrograph shows actin filaments decorated with myosin S1 heads containing the essential light chain The 3D reconstruction, at a resolution of 30-35 A, is a composite in which the differently colored parts are derived from a series of difference maps that were superimposed on f-actin The components include: f-actin (blue), myosin heavy chain motor domain (orange), essential light chain (purple), regulatory light chain (white), tropomyosin (green), and myosin motor domain N-terminal beta-barrel (red) [Courtesy of A Lin, M Whittaker, and R Milligan (Scripps Research Institute, La Jolla, CA).] (dl A 2D crystal, light-harvesting complex LHCII at 4 resolution The model shows the ~ r o t e i nbackbone and the arrangement of chromophores in a number of trimeric subunits in the crystal lattice In this example, image contrast is too low to see any hint of the structure without image processing (see also Fig 14.3) See color insert [Courtesy of W Kuhlbrandt (Max-Planck-Institute for Biophysics, Frankfurt, Germany).] - translations needed to align the different transforms are determined so the data may be me:rged and a 3D reconstruction computed by Fo1~rier-Besselinversion procedures (83) Deterimination of the absolute hand requires conlparison of a pair of images recorded with a Smid l tilt of the specimen between the views (139) lcosahedral Particles Thl3 typical strategy for processing and 3D reconstruction of icosahedral particles consists ofthe following steps: First, a series of micro- graphs of a monodisperse distribution of particles, normally suspended over holes in a perforated carbon support film, is recorded After digitization of the micrographs, individual particle images are boxed and floated with a circular mask The astigmatism and defocus of each micrograph is measured from the sum of intensities of the Fourier transforms of all particle images (140) Autocorrelation techniques are then used to estimate the particle phase origins, which coincide with the center of each particle, where all rotational symmetry axes intersect (141).The view orientation Electron Cryomicroscopy of Biological Macromolecules of each particle, defined by three Eulerian angles, is determined either by means of common and cross-common lines techniques or with the aid of model-based procedures [e.g., Crowther (106); Fuller et al (107);Baker et al (17)l Once a set of self-consistent particle images is available, an initial, low resolution 3D reconstruction is computed by merging these data with Fourier-Bessel methods (106) This reconstruction then serves as a reference for refining the orientation, origin, and CTF parameters of each of the included particle images, for rejecting "bad" images, and for increasing the size of the data set by including new particle images from additional micrographs taken at different defocus levels A new reconstruction, computed from the latest set of images, serves as a new reference and the above refinement procedure is repeated until no further improvements, as measured by the reliability criteria mentioned above, are made Determination of the absolute hand of the structure requires the recording and processing of a pair of images taken with a known, small relative tilt of the specimen between the two views (142) 13 VISUALIZATION, MODELING, A N D INTERPRETATION OF RESULTS Once a reliable 3D map is obtained, computer graphics and other visualization tools may be used as aids in interpreting morphological details and understanding biological function in the context of biochemical and molecular studies and complementary X-ray crystallographic and other biophysical measurements 14 TRENDS The new generation of intermediate voltage (-300 kV)FEG microscopes is now making it much easier to obtain higher resolution images that, by use of larger defocus values, have good image contrast at both very low and very high resolution The greater contrast at low resolution greatly facilitates particle-alignment procedures, and the increased contrast resulting from the high coherence illumina- tion helps to increase the signal-to-noise ratio for the structure at high resolution Cold stages are constantly being improved, with several liquid helium stages now in operation (143, 144) Two of these are commercially available from JEOL and FEIPhilips Finally, three additional likely trends include: (1)increased automation, including the recording of micrographs, the use of spotscan procedures in remote microscope operation (145, 146), and in every aspect of image processing; (2) production of better electronic cameras (e.g., CCD or pixel detectors); and (3) increased use of dose-fractionated, tomographic tilt series, to extend EM studies to the domain of larger supramolecular and cellular structures (102, 147) 15 ACKNOWLEDGMENTS We are greatly indebted to all our colleagues at Purdue and Cambridge for their insightful comments and suggestions; to B Bottcher, R Crowther, J Frank, W Kiihlbrandt, and R Milligan for supplying images used in Figure 14.6; and J Brightwell for editorial assistance T.S.B was supported in part by Grant GM33050 from the National Institutes of Health 16 ABBREVIATIONS OD zero-dimensional (single particles) 1D one-dimensional (helical) 2D two-dimensional 3D three-dimensional CCD charge coupled device (slow scan TV detector) cryo-EM electron cryomicroscopy CTF contrast transfer function EM electron microscope/microscopy FEG field emission gun REFERENCES S Brenner and R W Horne, Biochem Biophys Acta-Prot S t r u t , 34, 103-110 (1959) H E Huxley and G Zubay, J Mol Biol., 2, 10-18 (1960) References A Klug and J E Berger, J Mol Biol., 10, 565-569 (1964) D J DeRosier and A Klug, Nature, 217, 130134 (1968) W Hoppe, R Langer, G Knesch, and C Poppe, Naturwissenschaften, 55,333-336 (1968) H P Erickson and A Klug, Philos Trans R Soc Lond B, 261, 105-118 (1971) P N T Unwin and R Henderson, J Mol Biol., 94,425440 (1975) J Dubochet, J Lepault, R Freeman, J A Berriman, and J.-C Homo, J Microsc., 128, 219237 (1982~) J Dubochet, M Adrian, J.-J Chang, J.-C Homo, J Lepault, A W McDowall, and P Schultz, Q Rev Biophys., 21,129-228 (1988) 10 K A Taylor and R M Glaeser, Science, 186, 1036-1037 (1974) 11 R Henderson, J M Baldwin, T A Ceska, F Zemlin, E Beckmann, and K H Downing, J Mol Biol., 213, 899-929 (1990) 12 W Kuhlbrandt, D N Wang, andY Fujiyoshi, Nature, 367, 614-621 (1994) 13 E Nogales, S G Wolf, and K H Downing, Nature, 391, 199-203 (1998) 14 K Murata, K Mitsuoka, T Hirai, T Waltz, P A g e , J B Heymann, A Engel, and Y Fujiyoshi, Nature, 407,599-605 (2000) 15 L A Amos, R Henderson, and P N T Unwin, Prog Biophys Mol Biol., 39, 183-231 (1982) 16 T Walz & N Grigorieff, J Struct Biol., 121, 142-161 (1998) 17 T S Baker, N H Olson, and S D Fuller, Microbiol Mol Biol Rev., 63,862-922 (1999) 18 J Frank, Three-Dimensional Electron Microscopy of MacromolecularAssemblies, Academic Press, San Diego, CA, 1996,342 pp 19 I Hargittai and M Hargittai, Eds., Stereochemical Applications of Gas-Phase Electron Diffraction, VCH, New York, 1988 20 M Isaacson, J Langmore, and H Rose, Optik, 41,92-96 (1974) 21 R Henderson, Q Rev Biophys., 28, 171-193 (1995) 22 W A Havelka, R Henderson, and D Oesterhelt, J Mol Biol., 247, 726-738 (1995) 23 R M Glaeser, J Ultrastruct Res., 36, 466482 (1971) 24 R Henderson and P N T Unwin, Nature, , (1975) 25 J Frank, Curr Opin Struct Biol., 7,266-272 (1997) 26 J Kenney, E Karsenti, B Gowen, and S D Fuller, J Struct Biol., 120, 320-328 (1997) 27 Y Tao, N H Olson, W Xu, D L Anderson, M G Rossmann, and T S Baker, Cell, 95, 431-437 (1998) 28 L B Kong, A C Siva, L H Rome, and P L Stewart, Structure, 7, 371-379 (1999) 29 A C Bloomer, J Graham, S Hovmoller, P J G Butler, and A Klug, Nature, 276,362368 (1978) 30 R H Jacobson, X.-J Zhang, R F DuBose, and B W Matthews, Nature, 369,761-766 (1994) 31 G P A Vigers, R A Crowther, and B M F Pearse, EMBO J.,5, 529-534 (1986) 32 M Schatz, E V Orlova, P Dube, J Jager, and M van Heel, J Struct Biol., 114, 28-40 (1995) 33 R A Grant, D J Filman, S E Finkel, R Kolter, and J M Hogle, Nut Struct Biol., 5, 294-303 (1998) 34 A Mattevi, G Obmolova, E Schulze, K H Kalk, A H Westphal, A D Kok, and W G J Hol, Science, 255, 1544-1550 (1992) 35 R A Milligan, Proc Natl Acad Sci USA, 93, 21-26 (1996) 36 A Miyazawa, Y Fujiyoshi, M Stowell, and N Unwin, J Mol Biol., 288, 765-786 (1999) 37 K Hirose, W B Amos, A Lockhart, R A Cross, and L A Amos, J Struct Biol., 118, ' 140-148 (1997) 38 K Narnba and F Vonderviszt, Q Rev Biophys., 30,l-65 (1997) 39 T.-W Jeng, R A Crowther, G Stubbs, and W Chui, J Mol Biol., 205, 251-257 (1989) 40 A Cheng, A N van Hoek, M Yeager, A S Verkman, and A K Mitra, Nature, 387, 627630 (1997) 41 V M Unger, N M Kumar, N B Gilula, and M Yeager, Science, 283,1176-1180 (1999) 42 D A Winkelmann, T S Baker, and I Rayment, J Cell Biol., 114, 701-713 (1991) 43 K A Taylor, J Tang, Y Cheng, and H Winkler, J Struct Biol., 120,372-386 (1997) 44 B Bottcher, S A Wynne, and R A Crowther, Nature, 386, 88-91 (1997) 45 A Malhotra, P Penczek, R K Agrawal, I S Gabashvili, R A Grassucci, R Junemann, N Burkhardt, K H Nierhaus, and J Frank, J Mol Biol., 280, 103-116 (1998) 46 R W Horne and I Pasquali-Ronchetti, J Ultrastruct Res., 47, 361-383 (1974) Electron Cryomicroscopy of Biological Macromolecules 47 H Yoshimura, M Matsumoto, S Endo, and K Nagayama, Ultramicroscopy, 32, 265-274 (1990) 48 R Kornberg and S A Darst, Curr Opin Struct Biol., 1,642-646 (1991) 49 B Jap, M Zulauf, T Scheybani, A Hefti, W Baumeister, and U Aebi, Ultramicroscopy,46, 45-84(1992) 50 E W Kubalek, S F J LeGrice, and P Brown, J Struct Biol., 113,117-123(1994) 51 J.-L Rigaud, G Mosser, J J Lacapere, A Olofsson, D Levy, and J.-L Ranck, J Struct Biol., 118,226-235(1997) 52 L.Hasler, J B Heymann, A Engel, J Kistler, and T Walz, J Struct Biol., 121,162-171 (1998) 53.I Reviakine, W.Bergsma-Schutter, and A Brisson, J Struct Biol., 121,356-361(1998) 54.E.M Wilson-Kubalek, R E Brown, H Celia, and R A Milligan, Proc Natl Acad Sci USA, 95,8040-8045(1998) 55 A Polyakov, C Richter, A Malhotra, D Koulich, S Borukhov, and S A Darst, J.Mol Biol., 281,465-473(1998) 56 M Adrian, J Dubochet, J Lepault, and A W McDowall, Nature, 308,32-36 (1984) 57 J.R Bellare, H T Davis, L E Scriven, andY Talmon, J Electron Microsc Technol., 10,87111(1988) 58 E Mayer and G Astl, Ultramicroscopy, 45, 185-197(1992) 59 J Berriman and N Unwin, Ultramicroscopy, 56,241-252(1994) 60 H D.White, M L Walker, and J Trinick, J Struct Biol., 121,306-313(1998) 61 S Subramaniam, M Gerstein, D Oesterhelt, (1993) and R Henderson, EMBO J.,12,l-18 62 D P.Siegel, W J Green, and Y Talmon, Biophys J., 66,402-414 (1994) 63 S Trachtenberg, J Struct Biol., 123,45-55 (1998) (1995) 64 N.Unwin, Nature, 373,37-43 65 S D Fuller, J A Berriman, S J Butcher, and B E Gowen, Cell, 81,715-725(1995) 66 D P.Siegel and R M Epand, Biophys J.,73, 3089-3111(1997) 67 M Walker, X.-Z.Zhang, W Jiang, J Trinick, and H D White, Proc Natl Acad Sci USA, 96,465-470 (1999) 68 M Cyrklaff and W Kiihlbrandt, Ultramicroscopy, 55,141-153(1994) 69 J C H.Spence, Experimental High-Resolution Electron Microscopy, Oxford University Press, Oxford, UK, 1988 70 L Reimer, Transmission Electron Microscopy, Springer-Verlag, Berlin, 1989 71 R H Wade and J Frank, Optik, 49, 81-92 (1977) 72 R H Wade, Ultramicroscopy, 46, 145-156 (1992) 73 C Toyoshima, K Yonekura, and H Sasabe, Ultramicroscopy, 48,165-176(1993) 74 R H Cheng, N H Olson, and T S Baker, Virology, 186,655-668 (1992) 75 B L Trus, R B S Roden, H L Greenstone, M Vrhel, J T Schiller, and F P Booy, Nut Struct Biol., 4,413-420 (1997) 76 F Zemlin, Ultramicroscopy, 46,25-32 (1992) 77 Z H Zhou and W Chiu, Ultramicroscopy, 49, 407-416(1993) 78 F Zemlin, Micron, 25,223-226(1994) 79 E J Mancini, F D Haas, and S D Fuller, Structure, 5,741-750(1997) 80.J.Brink, M B Sherman, J Berriman, and W Chiu, Ultramicroscopy, 72,41-52 (1998) 81 L Krivanek and P E Mooney, Ultramicroscopy, 49,95-108(1993) 82 M B Sherman, J Brink, and W Chiu, Micron, 27,129-139(1996) 83 D J DeRosier and P B Moore, J Mol Biol., 52,355-369 (1970) 84 J L Carrascosa and A C Steven, Micron, 9, 199-206(1978) 85 M van Heel and J Frank, Ultramicroscopy, 6, 187-194(1981) 86 I.S Gabashvili, R K Agrawal, C M T Spahn, R A Grassucci, D I Svergun, J Frank, and P Penczek, Cell, 100,537-549(2000) 87 M Radermacher, T Wagenknecht, A Verschoor, and J Frank, J Microsc., 146,113-136 (1987) 88 M Radermacher, J Electron Microsc Technol., 9,359-394 (1988) 89 J Frank, M Radermacher, P Penczek, J Zhu, Y Li, M Ladjadj, and A Leith, J Struct Biol., 116,190-199(1996) 90 M van Heel, Ultramicroscopy, 21, 111-124 (1987a) 91 M van Heel, G Harauz, and E V Orlova, J Struct Biol., 116,17-24(1996) 92 M Radermacher in D.-P Hader, Ed., Image Analysis in Biology, CRC Press, Boca Raton, FL, 1991,pp 219-246 References 93 M Radermacher in J Frank, Ed., Electron Tomography, Plenum Press, New York, 1992, pp 91-115 94 M Radermacher, Ultramicroscopy, 53, 121-136(1994) 95 P A Penczek, R A Grassucci, and J Frank, Ultramicroscopy, 53,251-270(1994) 96 M Schatz and M van Heel, Ultramicroscopy, 32,255-264(1990) 97 P Penczek, M Radermacher, and J Frank, Ultramicroscopy, 40,33-53(1992) 98 N Grigorieff, J Mol Biol., 277, 1033-1046 (1998) 99 D E Olins, A L Olins, H A Levy, R C Durfee, S M Margle, E P Tinnel, and S D Dover, Science, 220,498-500 (1983) 100.U Skoglund and B Daneholt, Trends Biochem Sci., 11,499503(1986) 101 J C Fung, W Liu, W J DeRuijter, H Chen, C K Abbey, J W Sedat, and D A Agard, J Struct Biol., 116,181-189 (1996) 102 W Baumeister, R Grimm, and J Walz, Trends Cell Biol., 9,81-85(1999) 103 A K Shah and P L Stewart, J Struct Biol., 123,17-21(1998) 104 F Beuron, M R Maurizi, D M Belnap, E Kocsis, F P Booy, M Kessel, and A C Steven, J Struct Biol., 123,248-259 (1998) 105 R A Crowther, L A Amos, J T Finch, D J DeRosier, and A Klug, Nature, 226,421-425 (1970) 106 R A Crowther, Philos Trans R Soc Lond., 261,221-230(1971) 107 S D Fuller, S J Butcher, R H Cheng, and T S Baker, J Struct Biol., 116,48-55(1996) 108 R H Cheng, V S Reddy, N H Olson, A J Fisher, T S Baker, and J E Johnson, Structure, 2,271-282(1994) 109 R A Crowther, N A Kiselev, B Bottcher, J A Berriman, G P Borisova,V Ose, and P Pumpens, Cell, 77,943-950(1994) 110 T S Baker and R H Cheng, J Struct Biol., 116,120-130(1996) 111 J R Castbn, D M Belnap, A C Steven, and B L Trus, J Struct Biol., 125, 209-215 (1999) 112 R A Crowther, R Henderson, and J M Smith, J Struct Biol., 116,9-16(1996) 113 J A Lawton and B V V Prasad, J Struct Biol., 116,209-215(1996) 114 P A Thuman-Commike and W Chiu, J Struct Biol., 116,41-47(1996) 115 I M Boier Martin, D C Marinescu, R E Lynch, and T S Baker, J Struct Biol., 120, 146-157(1997) 116 Z H Zhou, W Chiu, K Haskell, H J Spears, J Jakana, F J Rixon, and L R Scott, Biophys J.,74,576-588(1998) 117 P.L Stewart, C Y Chiu, S Huang, T Muir,Y Zhao, B Chait, P Mathias, and G R Nemerow, EMBO J., 16,1189-1198(1997) 118 J Walz, T Tamura, N Tamura, R Grimm,W Baumeister, and A J Koster, Mol Cell, 1, 59-65(1997) 119 M Stewart, J Electron Microsc Technol., 9, 325-358(1988) 120 C Toyoshima and N Unwin, J Cell Biol., 111, 2623-2635(1990) 121 D G Morgan and D DeRosier, Ultramicroscopy, 46,263-285 (1992) 122 N Unwin, J Mol Biol., 229, 1101-1124 (1993) 123 R Beroukhim and N Unwin, Ultramicroscopy, 70,57-81 (1997) 124 E.H.Egelman, Ultramicroscopy, 19,367474 (1986) 125 B Carragher, M Whittaker, and R A Milligan, J Struct Biol., 116,107-112 (1996) 126 C H Owen, D G Morgan, and D J DeRosier, J Struct Biol., 116,167-175(1996) 127 R Henderson, J M Baldwin, K H Downing, J Lepault, and F Zemlin, Ultramicroscopy, 19, 147-178(1986) 128 J M Baldwin, R Henderson, E Beckman, and F Zemlin, J Mol Biol., 202,585-591(1988) 129 S Hardt, B Wang,andM F Schmid, J Struct Biol., 116,68-70(1996) 130 R A Crowther and P K Luther, Nature, 307, 569-570(1984) 131 H Winkler and K A Taylor, J Struct Biol., 116,241-247(1996) 132 J Frank in J Frank, Ed., Electron Tomography: Three-Dimensional Imaging with the Transmission Electron Microscope, Plenum Press, New York, 1992,399 pp 133 J Frank, A Verschoor, and M Boublik, Science, 214,1353-1355(1981) 134 M van Heel, Ultramicroscopy, 21, 95-100 (198'7b) 135 M van Heel and J Hollenberg in W Baumeister and W Vogell, Eds., Electron Microscopy at Molecular Dimensions, Springer-Verlag, Berlin, 1980, pp 256-260 136 M Unser, B L Trus, J Frank, and A C Steven, Ultramicroscopy, 30,429-434 (1989) Electron Cryomicroscopy of Biological Macromolecules 137 D A Agard, J Mol Biol., 167, 849-852 (1983) 138 J M Valpuesta, J L Carrascosa, and R Henderson, J Mol Biol., 240,281-287(1994) 139 J T.Finch, J Mol Biol., 66,291-294(1972) 140 Z H Zhou, S Hardt, B Wang, M B Sherman, J Jakana, and W Chiu, J Struct Biol., 116, 216-222(1996) 141 N.H Olson and T S Baker, Ultramicroscopy, 30,281-298(1989) 142 D.M Belnap, N H Olson, and T S Baker, J Struct Biol., 120,44-51(1997) 143 Y Fujiyoshi, T Mizusaki, K Morikawa, H Yamagishi, Y Aoki, H Kihara, and Y Harada, Ultramicroscopy, 38,241-251(1991) 144 F Zemlin, E Beckmann, and K D vanderMast, Ultramicroscopy, 63,227-238(1996) 145 N.Kisseberth, M Whittaker, D Weber, C S Potter, and B Carragher, J Struct Biol., 120, 309-319(1997) 146 M Hadida-Hassan, S J Young, S T Peltier, M Wong, S Lamont, and M H Ellisman, J Struct Biol., 125,235-245 (1999) 147 B F McEwen, K H Downing, and R M Glaeser, Ultramicroscopy, 60,357-373 (1995) CHAPTER FIFTEEN Peptidomimetics for Drug M ANGELS ESTIARTE DANIEL H RICH School of Pharmacy-Department of Chemistry University of Wisconsin-Madison Madison, Wisconsin Contents Burger's Medicinal Chemistry and Drug Discovery Sixth Edition, Volume 1: Drug Discovery Edited by Donald J Abraham ISBN 0-471-27090-3 O 2003 John Wiley & Sons, Inc Introduction, 634 Classification of Peptidomimetics,634 Design of Conformationally Restricted Peptides, 636 Template Mimetics, 643 Peptide Bond Isosteres, 644 From Transition-State Analog Inhibitors to NonPeptide Inhibitors: Examples in Protease Inhibitors, 646 6.1 TSA in Aspartic Peptidase Inhibitors, 647 6.2 TSA in Metallo Peptidase Inhibitors, 650 6.3 TSA-Derived Cysteine and Serine Peptidase Inhibitors, 652 Speeding up Peptidomimetic Research, 655 Toward Rational Drug Design: Discovery of Novel Non-Peptide Peptidomimetics, 657 Historical Development of Important NonPeptide Peptidomimetics,659 9.1 HIV Protease, 659 9.2 Thrombin, 660 9.3 Factor Xa,662 9.4 Glycoprotein IIbiIIIa (GP IIbDIIa),662 9.5 Ras-Farnesyltransferase, 665 9.6 Non-Peptidic Ligands for Peptide Receptors, 667 9.6.1 Angiotensin 11,668 9.6.2Substance P, 669 9.6.3Neuropeptide Y,670 9.6.4Growth Hormone Secretagogues, 670 9.6.5 Endothelin, 672 10 Summary and Future Directions, 674 Peptidomimetics for Drug Design INTRODUCTION CLASSIFICATION OF PEPTIDOMIMETICS Protein-protein interactions are central to biology and provide one mechanism to convert genomic information into regulated biological responses Important examples of proteinpeptide interactions include the binding of peptide ligands to proteases, the binding of peptide hormones to peptide receptors, the recruitment of proteins to effect signal transduction, and apoptosis Peptides also act as neurotransmitters, neuromodulators, hormones, and autocrine and paracrine factors Unfortunately, their use as pharmaceutical drugs is made difficult by their poor pharmacokinetic profiles; they are easily proteolyzed, poorly transported, and rapidly excreted Although modern formulation techniques have improved delivery of peptides (e.g., inhalation of insulin), there remains a need for small potent molecules that can be administered orally For these reasons, much effort has been expended to find ways to replace portions of peptides with non-peptide structures, called peptidomimetics, in the hope of obtaining orally bioavailable entities Several types of peptidomimetics have been developed, and the field has emerged as one of the important approaches to drug design and discovery This review will describe the various methods developed to design peptidomimetics Due to space limitations, the biological rationale for each peptidomimetic and its chemical synthesis can not be covered Selected examples of the strategies employed to obtain peptidomimetics are provided to illustrate the breadth of research in this field The term peptidomimetic is often used in the literature to indicate a multitude of structural types that differ in fundamental ways Comparisons between peptidomimetics suffer from the lack of accepted definitions of what a peptidomimetic is (1).The term is often applied to highly modified analogs of peptides without distinguishing how these differ from classical analogs of peptides For example, peptide (2) is derived from the decapeptide LH-RH (1);(2) contains only five amino acids, none of which is present in the parent compound, yet it is a powerful antagonist of the LH-RH receptor (Fig 15.1) (2) Is (2) a peptide analog or a peptidomimetic? In the 19709, Hughes et al were the first to show that two very different chemical structures have similar agonist properties (3) The opioid natural product, morphine (3),was found to resemble the N-terminal structure of the endogenous opioid peptides, enkephalins, (4a) and (4b),and p-endorphin ( )(Fig 15.2) The remarkable similarity between the morphine phenol system and the N-terminal tyrosine residue in the peptide opioids implied that these units reacted with opioid receptors in a similar fashion to elicit comparable responses (4- 6) The realization that a non-peptide natural product was mimicking the action of a natural peptide effector led Farmer to postulate that other non-peptide structures might be found that would mimic other peptide effectors (7) His concept of "peptide mimetic," which later was called "peptidomimetic," proposed that pGIu-His-Trp-Ser-Tyr-Gly-Leu-Arg-Pro-Gly-NH, LH-RH (1 Figure 15.1 Reduced-size antagonist of LH-RH 2 Classification of Peptidomimetics Met-enkephalin Tyr-Gly-Gly-Phe-Met (4a) Leu-Enkephalin Tyr-Gly-Gly-Phe-Leu ,%Endorphin & o\+ (4b) Tyr-Gly-Gly-Phe-Met-Thr-Ser-GIu-Lys-SerGln-Thr-Pro-Leu-Val-Thr-Phe-Lys-Asn-AlaIle-He-Lys-Asn-Ala-Tyr-Lys-Lys-Gly-Glu b~ HO Morphine (3) (5) Figure 15.2 Examples of peptidic and non-peptidic opioid receptor ligands novel scaffolds could be designed to replace the entire peptide backbone while retaining isosteric topography of the enzyme-bound peptide (or assumed receptor-bound) conformation Farmer's definition went beyond simple replacement of amide bonds and the concept of stringing together conformationally restricted amino acid derivatives to mimic the native peptide structure In the intervening years, many non-peptide and partially peptide structures have been found that mimic (or antagonize) the action of the peptide ligand at its receptor; this is particularly true with substances active at G-protein-coupled receptors The pyrrolinone unit (6)designed by Smith and Hirschmann illustrates a modern use of these two concepts (Fig 15.3) (8) Pyrrolinones constrain the peptide-like side-chains into an extended p-structure topography that fits the active sites of most peptidases; pyrrolinones are resistant to normal proteolysis because no a-amino acid units remain, and the units impart sophisticated partitioning properties to the final inhibitor Pyrrolinones, like many peptide-derived peptidomimetics, retain an atom-to-atom correspondence to the parent peptide, especially with respect to the peptide backbone structure Most of these structures contain elements that accomplish one of two objectives: they replace amide bonds with metabolically stable units, and they affect a conformational constraint on peptides or on the peptide replacement In contrast, heterocyclic natural products or screening leads that bind to peptide receptors also have been called peptidomimetics by virtue of their mimicking (or antagonizing) the function of the natural peptide Although structural data confirming mimicry of the designed mimetics are rarely available for receptor bound ligands, ample ev- idence is available from X-ray crystallography that heterocyclic inhibitors are mimicking the extended p-strand of enzyme-bound substrate-derived inhibitors (vide infra) Based on these considerations, four distinct types of peptidomimeticshave been identified to date (9, 10) The first invented were structures that contain one or more mimics of the local topography about an amide bond (amide bond isosteres) Strictly speaking, these are properly classified as pseudopeptides ( l l ) , but in recent years, they have been called peptidomimetics on occasion For historical reasons, we classify the peptide backbone mimetics as type I mimetics (Table 15.1) These Peptide Pyrrolinone analog Figure 15.3 Correlationof pyrrolinone-based peptidomimetics and the parent peptide Peptidomimetics for Drug Design 636 Table 15.1 Peptidomimetic Types Peptidomimetic Examples Type I Type I1 Type I11 Peptide backbone mimetics Functional mimetics Topographical mimetics Substrate-based design Molecular modeling, HTS Structure-based design Type IV Non-peptide peptidomimetics Group Replacement Assisted Binding mimetics often match the peptide backbone atom-for-atom while retaining functionality that makes important contacts with binding sites Some units mimic short portions of secondary structure (e.g., p-turns) and have been used to generate lead compounds Many early protease inhibitors were designed from transition state analog mimetics or from collected substrate/product mimetics, each designed to mimic reaction pathway intermediates of the enzyme-catalyzed reaction These are mimics of the peptide bond in a putative transition state or product state and will be classified here as peptidomimetics The second type of mimetic to emerge was the functional mimetic, or type ZZ mimetic, which is a small non-peptide molecule that binds to a peptide receptor Morphine was the first well-characterized example of this type of peptidomimetic Initially, type I1 mimetics were presumed to be direct structural analogs of the natural peptide, but characterization of both the endogenous peptide and antagonist's binding sites by site-directed mutagenesis has raised doubts about this interpretation (12) The mutagenesis data indicate that antagonists for a large number of receptors seem to bind to receptor subsites different than those used by the parent peptide Consequently, functional mimetics may not mimic the structure of the parent hormone; this remains to be determined Despite this uncertainty, the approach has been quite successful and produced a number of potential drug lead structures Type ZZZ mimetics represent the Farmer definition ofpeptidomimetics in that theypossess novel templates, which appear unrelated to the original peptides but contain the essential groups, positioned on a novel non-peptide scaffold to serve as topographical mimetics Several type I11 peptidomimetic protease inhibitors have been characterized where direct Pseudopeptides GPCR antagonists Non-peptide protease inhibitors Piperidine inhibitors X-ray structural determination of both the peptide-derived inhibitor and the heterocyclic non-peptide inhibitor complexes have been compared These examples demonstrate that alternate scaffolds can display side-chains so that they interact with proteins in fashion closely related to that of the parent peptide Recently, a fourth type of peptidomimetic called a GRAB-peptidomimetic (group replacement-assisted binding) has been identified (10) These structures might share structuralfunctional features of type I peptidomimetics, but they bind to an enzyme form not accessible with type I peptidomimetics Previous reviews on peptidomimetics have addressed pseudopeptides (ll),macrocyclic mimetics (13), natural product mimetics (14), cyclic protease inhibitors (15),mimetics for receptor ligands (16-22), and earlier general overviews (23-29) This review will focus on the design process itself Novel peptidomimetics in which the structural relationship between parent peptide and the peptidomimetic has been established by biophysical methods are used to clarify the principles Successful approaches are highlighted to illustrate how these concepts are currently used DESIGN O F CONFORMATIONALLY RESTRICTED PEPTIDES Peptide derivatives that contain conformationally restricting amino acid units or other conformational constraints were first called conformationally constrained (or restricted) peptide analogs The use of steric hindrance or cyclization to limit rotational degrees of freedom in biologically active molecules has a long history and was originally applied to non-peptide neurotransmitters (30) Subsequently, it was applied to amino acid substituents and to Design of Conformationally Restricted Peptides TRH (7) Figure 15.4 Structure of TRH tripeptide cyclic peptides (31, 32) and to control secondary structure in model proteins Conformational restriction is a very powerful method for probing the bioactive conformations of peptides Small peptides have many flexible torsion angles so that enormous numbers of conformations are possible in solution For example, a simple tripeptide such as thyrotropin-releasing hormone (TRH; 7) (Fig 15.4) with six flexible bonds could have over 65,000 possible conformations The number of potential conformers for larger peptides is enormous, and some method is needed to exclude potential conformers Modern biophysical methods, e.g., X-ray crystallography or isotope edited nuclear magnetic resonance (NMR),(33) can be used to characterize peptide-protein interactions for soluble proteins, but most biophysical methods cannot yet determine the conformation of a ligand bound to constitutive receptors, e.g., G-protein-coupled receptors (34,35) Cyclization is one of the earliest techniques applied to design peptidomimetics Cyclic peptides are more stable to amide bond hydrolysis and allow less conformational flexibility; consequently, the resulting analogs are anticipated to be more selective and less toxic Methods for restricting conformations include peptide backbone cyclization, disulfide bond formation, side-chain cyclization, and metal ion chelation The first successful application of conformational restriction to peptide chemistry was carried out by Veber et al at Merck, (361, who were trying to simplify the structure of somatostatin (8) (Fig 15.5) to produce an orally active derivative Their approach was to introduce conformational restraints into the mac- - rocyclic peptide ring system to reduce the number of conformations available to the analog Not all substitutions were expected to produce biologically active products, but those that retained activity were assumed to be able to adopt conformations close to the normal bioactive conformation This work began from the earlier discovery by Rivier et al (37) that replacement of ~4ryptophanin the position-8 of somatostatin by D-tryptophan produced an analog that retained biological activity This unusual biological result is possible when a D,L-sequence(D-Trp-Lys) replaces an L,L-sequence (Trp-Lys) in a peptide at a type I1 p-turn, because the topography of the amino acid side-chains at these positions is essentially identical in these turns (38) These results led Veber et al to postulate that the amino acid sequence Phe-Trp-Lys-Thr might be part of a type I1 p-turn, and that this tetrapeptide sequence might comprise the active pharmacophore Although this hypothesis was highly speculative for its time, it was shown to be essentially correct by applying the principle of conformational restriction (Fig 15.5) Deletion of the N-terminal dipeptide, followed by insertion of the D-Trp at position-8, and replacement of the disulfide sulfurs with carbons produced analog (9) NMR and other data suggested that the two Phe side-chains were clustered, thus they were replaced by a transannular disulfide bond limiting the available conformation, as in compound (10).After several iterations of this process, a biologically active cyclic hexapeptide (11) was discovered that retained only of the original 14 amino acids in somatostatin yet produced a fully active derivative (31) The work of Veber et al established that valuable information about the bioactive conformation of a flexible peptide could be obtained by applying the principles of conformational restriction, and several additional examples soon were reported that followed this strategy Conformationally restricted enkephalin analogs, e.g., 02-13), were formed by cyclizing between positions and of enkephalins (4a-b) (39) Cyclization of a-melanotropin (14) gave the unusually active analog (15) (40) Small cyclic analogs of endothelin (16) (41) have been discovered by applying these methods, as illustrated by (17) (Fig Design of Conformationally Restricted Peptides H2N-Cys-Ser-Cys-Ser-Ser-Leu-Met \ I HO-Trp-lle-lle-Asp-Leu-His-Cys-Phe-Tyr-Val-Cys-Glu-Lys-~p Endothelin Ac-Ser- (16) Figure 15.6 Cyclic hormone peptide analogs 15.6) Peptide chemists routinely apply conformational restriction in their attempts to determine possible bioactive conformations Flexible peptides can be conformationally restricted by a variety of methods other than macrocyclization of the peptide For example, Marshall et al introduced a-methyl amino acid substituents into peptides as a way to decrease the conformational space available to the resulting peptide (42); these types of approaches led to his "Active Analog" approach for determining bioactive conformations of flexible molecules (43) Some other traditional modifications of the peptide substrate are the replacement of the amino acids of the PI-P,' cleavage site by D-amino acids or the employment of a-C or a-N alkylated amino acids and cyclic or p-amino acids (Fig 15.7) Mimicking the secondary structure of peptides has become one of the most important tools for rational drug design (44-47) These methods induce the synthetic analog to adopt a set of target conformations, which are designed to mimic the bioactive conformation predicted in the native substrate from biophysical techniques Molecular surrogates Peptidomimetics for Drug Design a-Amino acid PAmino acid p-Amino acid a-Alkyl - Cyclic derivatives / N H Figure 15.7 Representative amino acid mimetics have been found that efficiently - mimic turns, strands, sheets, and helices By far, the major efforts have focused on the design of p-turn mimetic~.Some of the templates used to constrain the conformational torsion angles of the peptide chain are summarized in Figs 15.8-15.14 In a very early example, Freidinger et al developed a series of cyclic lactams that stabilized P- and y-turn structures in linear peptides (Fig 15.8) This strategy was applied to determine conformations of LH-RH that are consistent with the turn structure ~ermitted by the constraint For example, the 3-aminolactam (18)was used to mimic a p-turn conformation When inserted in LH-RH, com- - pound (19) retained good biological activity so that the bioactive conformation of LH-RH probably contains a p-turn around residues and (48) Conformational restriction has also been used to determine the bioactive conformation of enzyme-inhibitor systems for which no Xray crystal structure is available Thorsett et al (49) synthesized conformationally restricted bicyclic lactam derivatives of the angiotensin converting enzyme (ACE) inhibitors enalapril (20) and enalaprilat (21) (Fig 15.9) to characterize torsion angles in the bioactive conformation Analog (22) was used to constrain the torsion angle psi (T) Flynn et al Glu-His-Trp-Ser-Tyr Ca(i) I 9v JArg-pro-Gly-NH2 - O \ LH-RH p-turn mimetic (19) Figure 15.8 y-Ladam analog of LH-RH A P-turn mimetic 3 Design of Conformationally Restricted Peptides Enalapril, R = Et, (20) Enalaprilat, R = H, (21) Figure 15.9 Conformationally restricted ACE inhibitors (50) extended this principle to prepare the very tight-binding tricyclic ACE inhibitor (23) (Fig 15.9) Several other y-, 6,and elactam derivatives have been prepared and inserted into receptor antagonists or agonists For instance, the thiazolidine lactam (24) (Fig 15.10) has been shown to induce the desired secondary structure in a gramicidine S analog Later, it was used to prepare a conformationally restricted cyclosporin A analog (51) Several p-turn and y-turn mimetics are shown in Figs 15.10-15.12, and many other examples are available in the recent literature (52-54) Figure 15.10 Lactams as p-turn mimetics Peptidomimetics for Drug Design Figure 15.11 Other p-turn mimetic scaffolds The p-sheet is an important, biologically relevant secondary structure As noted earlier, the pyrrolinones invented by Smith et al (Fig 15.3) adopt a p-strand conformation, which was corroborated by computer modeling and by X-ray crystallography (55).The diacylaminoeindolidinone (25), the dibenzofuran (26), and the N,N-linked oligourea (27) illus- trate other molecular templates designed to stabilize peptides in a p-strand conformation (Fig 15.13) (56, 57) Peptidomimetic structures that support a-helixes (28-30) (Fig 15.14) and loops have been reported less frequently because of the difficulty in presenting the side-chains correctly However, newer approaches have pro- H Figure 15.12 Some y-turn mimetic scaffolds Template Mimetics (27) Figure 15.13 Structures of P-sheet mimetics vided mimetics of multiple discontinuous protein surfaces (56) Over the last few years, the Gellman, Seebach, and Hanessian research groups have invented novel helical structures (e.g., 31,321 by use of P-, y-, and Gpeptides (58) It is important to stress that even a small change in the structure or in a single torsional angle can be sufficient to dramatically modify the conformation of the resulting peptide Numerous additional conformational constraints have been developed, and the reader is encouraged to consult these reviews for additional examples (32,59-63) TEMPLATE MIMETICS Highly functionalized molecular scaffolds have proven to be very successful in mimick- ing specific protein-protein interactions Insertion of the key pharmacophoric groups into a nonpeptidic framework has provided good inhibitors ofa variety of biological targets This technique has been successfully applied in those biological targets where the key structural amino acids of the native peptide for peptide recognition are known Miscellaneous examples are found in glycoprotein GbIIb/IIIa inhibitors (33)that mimic the RGD sequence (64) or in Ras-farnesyltransferase inhibitors (34) that mimic the CAAX sequence (Fig 15.15) (65) An early example of this concept was developed by Hirschmann et al in the design of a somatostatin analog (Fig 15.15)(55).Three of the four crucial amino acid side-chains of the parent peptide (Tyr, Trp, and Lys) were posi- Peptidomimetics for Drug Design OMe \ N-N (28) Figure 15.14 Newer templates found in helical or loop structures tioned on a sugar template (35) Although originally designed as a somatotropin release inhibitory factor (SRIF) antagonist, compound (35)also proved to be a good Substance P antagonist These sugar derivatives, as well as the benzodiazepine, diphenylmethane, and spiropiperidine scaffolds, are elements found in a variety of inhibitors of receptors, and have been designated as "privileged structures" (66) Thus, these common scaffolds can often provide a template for further optimization of a desired activity Evans et al have noted that the essential surface area of biologically active peptides is similar to the surface area of benzodiazepines, one type of non-peptide scaffold known to bind to Gprotein-coupled receptors (67) The quest for functionalized lead structures that effectively mimic the "hot spots" within the biological ligand is not easy (68) Molecular modeling and high-throughput screening (HTS) are techniques that are currently used for this purpose and have been summarized elsewhere The design and synthesis of antifungal analogs of the cyclic peptide rhodopeptin (36) (Fig 15.16) illustrate a recent application of peptidomimetic scaffolding, where the structure of the biological target is not known After structure-activity relationship (SAR) studies, the important side-chains of the peptide ligand were identified; then, NMR and molecular modeling techniques were used to model these side-chains onto known scaffolds and to compare with the original three-dimensional (3D) structure of the native peptide Compound (37) (Fig 15.16) is a potent peptidomimetic derivative with improved solubility in water that functions the same as the cyolic tetrapeptide (69, 70) PEPTIDE B O N D ISOSTERES The replacement of amide bonds by retro-inverso amide replacements (71, 72) and other amide bond isosteres generates pseudopeptides (11) This process was first used to stabilize peptide hormones in viuo, and later to prepare transition state analog (TSA) inhibitors Systematic efforts to convert good in vitro inhibitors into good in viuo inhibitors became the driving force for further development of peptidomimetics Figure 15.17 illustrates some of the peptide backbone modifications that have been made in an effort to increase bioavailability Replacement of scissile amide (CONH) bonds with groups insensitive to hydrolysis (e.g., CH,NH) has been extensively practiced Reviews of this work have appeared (11, 73) Removal of the proton donors and Peptide Bond lsosteres OMe / Somatostatin (35) Figure 15.15 Biologically active template mimetics acceptors in an amide bond also reduces hydration, which improves the ability of the compounds to penetrate lipid membranes (74) These approaches represent important first steps in development of peptidomimetics However, removal of an amide bond also af- fects the geometry and increases the flexibility of the molecule at this position, which decreases ligand binding Effective analogs have been obtained when conformational restriction, transition-state analog design, and amide bond replacements have been applied to Figure 15.16 Rhodopeptin analogs Representative example of scaffolding methodology Peptidomimetics for Drug Design Peptide Peptoid X = NH, 0, S, CH2 Retroinverso X=O,S Figure 15.17 Isosteres that replace peptide backbone amide groups to generate pseudopeptides scaffolds with molecular weights below 500600 (75,76),but at present this process is very labor intensive FROM TRANSITION-STATE ANALOG INHIBITORS TO NON-PEPTIDE INHIBITORS: EXAMPLES IN PROTEASE INHIBITORS Many peptidomimetics derived from the design of TSA inhibitors, molecules designed according to the hypothesis provided by Pauling (77) and implemented by Wolfenden (78, 79) TSA protease inhibitors are stable analogs of the tetrahedral intermediate for peptide bond hydrolysis that inhibit the enzyme (Fig 15.18) The first successful commercial application was the development of captopril (38) by Ondetti et al (80), and many applications have been reported over the past quarter century Figures 15.19-15.32 list examples of analogs of peptidyl transition states that have been employed to develop inhibitors of four classes of peptidases (81, 82) These units are used to replace the scissile amide bond in a substrate sequence with either an amino acid or dipeptide isostere, or with a chelating moiety in the case of metallo peptidases The Tetrahedral intermediate I Bond cleavage Transition state analog I No bond cleavage Figure 15.18 TSA inhibit peptide bond hydrolisis 6 From Transition-State Analog lnhibitors to Non-Peptide Inhibitors: Examples in Protease Inhibitors Statine (Sta) 647 Pepstatin (4'3) Reduced amide (39) Phosphonic Hydroxyethylene (41 Hydroxyethylamine a-Hydroxypamino acid Hydroxyethylurea (42) Hydroxyethyl hydrazide Hydroxyethyl sulfonamide Figure 15.19 TSA used to inhibit aspartic peptidases dipeptide TSA provides the functionality that interacts tightly with the enzyme catalytic groups while the amino acid sequence up- and downstream on the peptide chain provides interactions that lead to selective inhibition of the target enzyme The enzyme active site typically is buried in a cleft capable of accommodating up to three to nine amino acid residues of the substratelinhibitor depending on the minimum amino acid sequence necessary for hydrolysis The inhibitor's exquisite selectivity derives from the interactions of the ligand's p,-P,' residues with the enzyme binding sites (S,S,') (83) Recently, some aspartic and serine peptidase inhibitors have been found that access an additional binding site sub-pocket (S3sP)to increase both inhibitor potency and selectivity (84-86) 6.1 TSA in Aspartic Peptidase lnhibitors The reduced amide isostere (39),developed by Szelke, and the statine (hydroxylmethylene) isostere (40) were early transition-state analogs used to design inhibitors of various aspar- tic proteases, (87-89), and their success led to other tetrahedral intermediate mimics such as the hydroxylethylene (41) and hydroxyethylamine (42) isosteres (Fig 15.19) (90-92).The statine subunit, which mimics the tetrahedral intermediate, represents one of the earlier examples of TSA, although statine is one atom short in backbone length to be a true dipeptide or two atoms too long to be an isosteric replacement for a single amino acid Early work focused on developing inhibitors of renin as potential antihypertensive agents, but these compounds failed to become drugs primarily because of difficulties in obtaining orally active drugs As a result, the first pharmaceutical attempts to develop renin inhibitors for treatment of hypertension through TSA-biased inhibitors failed (93) It was eventually realized after extensive modifications to the ancillary peptide functionality that developing bioavailable peptide-derived inhibitors critically depended on the molecular weight of the inhibitor Developing inhibitors for HIV protease was substantially easier Peptidomimetics for Drug Design Saquinavir (Roche) Ritonavir (Abbott) lndinavir (Merck) Nelfinavir (Agouron) Amprenavir (Vertex) Lopinavir (Abbott) Figure 16.20 Peptide-derived TSA inhibitors used clinically in H W protease-based AIDS therapies than for renin because HIV protease requires a significantly smaller minimum substrate sequence (94) In addition, the principles elucidated to develop renin inhibitors were known and could be applied to the development of HIV protease inhibitors Variations on the hydroxyethyl amine moiety proved to be very successful Some of the highly modified HIV From Transition-State Analog Inhibitors to Non-Peptide Inhibitors: Examples in Protease lnhibitors 649 p-Secretase cleaves APP at: -Ser-Glu-Val-Lys-Met-:-Asp-Ala-Glu-Phe-Arg-Ser-Glu-Val-Asn-Leu+ Asp-Ala-Glu-Phe-Arg- & H2N-Lys-Thr-GIu-Glu-He-Ser-GIu-Val-Asn-HN OH = 30 Val-Ala-Glu-Phe-OH nM (43) Jk/k H2N-Glu-Val-Asn-HN Ala-Glu-Phe-OH OH = K, = 1.6 nM (44) LA BocHN-Val-Met-HN Val-CONHBn OH = K, = 2.5 nM (45) Figure 15.21 Peptide-derived TSA inhibitors as 0-secretase inhibitors protease inhibitors now in clinical use (Fig 15.20) have excellent oral bioavailability and establish that application of the transition state analog design process can be very successful in favorable cases More recently, the principles for designing inhibitors of aspartic proteases have been applied to the design of inhibitors of p-secretase (BACE or Memapsin-2) as potential agents for treating or preventing Alzheimer's disease (95, 96) Both statine-derived inhibitors (43) and hydroxyethylene-derived BACE inhibitors have been reported (Fig 15.21) (97,98).A crystal structure of (44) bound to p-secretase has been reported (99) As expected, the hy- droxyl group is hydrogen bonded to Asp32 and Asp228, like in other hydroxyethylene derivatives, and the inhibitor binds in an extended conformation Because the target p-secretase is within the CNS, successful inhibitors have to penetrate the brain blood barrier readily, a property not yet achieved with any of the peptidomimetic inhibitors currently available With the crystal structure in hand, structure-based modification of the parent lead compound has just started to provide new peptidomimetic structures with lower molecular weight and fewer hydrogen bonds (e.g., 45) (Fig 15.211, opening further avenues to pharmacologically useful compounds (100) Peptidomimetics for Drug Design Captopril Enalapril, R = Et Enalaprilat, R = H, (46) (38) Figure 15.22 Examples of TSA as ACE inhibitors 6.2 TSA in Metallo Peptidase Inhibitors The discovery of the angiotensin converting enzyme inhibitors in the middle 1970s constitutes one of the maior advances in the rational design of drugs, the consequences of which are still being realized The discovery of these metallo peptidase inhibitors was carried out by Ondetti et al as part of a long-term study to develop antihypertensive drugs (80); in 1999 they received the Lasker Prize in Clinical Medicine for their work Angiotensin converting enzyme (ACE) is a carboxy zinc metallo dipeptidase that cleaves His-Leu from the C-terminus of angiotensin-I Ondetti et al reasoned that the product of normal reaction, the carboxyl group, could bind to the active site zinc ion, and that the carboxyl group of a collected-product inhibitor also could bind weakly To im-prove the interaction between inhibitor and enzyme zinc ion, they replaced the carboxyl group with a sulfhydryl group, which binds zinc about 1000 times more tightly This led to captopril (Capoten) (38) (Fig 15.22) (80) Later developments by other companies led to many ACE inhibitors Some are illus- Lisinopril (47) trated by enalaprilat (46) and lisinopril (47) (Fig 15.22) (101, 102) Most metallopeptidase inhibitors append a zinc chelating functionality to a peptide or peptidomimetic that is recognized by the S1'S3' subsites in the target enzyme Successful clinical candidates invariably contain groups that replace the initial di- and tri-peptide moieties to achieve selectivity and orally activity For example, neutral endopeptidase (NEP), another endopeptidase involved in degrading the larger opioid peptides dynorphan and/or endorphan, is inhibited by thiorphan (48) (103) and a variety of NEP inhibitors: retrothiorphan (49) (104) and kelatorphan (50) (Fig 15.23) (105).The hydroxamic acid moiety is used in many inhibitors of metallopeptidases Inhibition of NEP also prevents the degradation of atrial natriuretic factor (ANF),a natural hypotensive peptide Dual inhibitors of NEP and ACE have been designed successfully because both enzymes share significant structural homology, particularly in their active sites Simultaneous inhibition of both peptidases produces a more powerful hypoten- From Transition-State Analog lnhibitors to Non-Peptide Inhibitors: Examples in Protease Inhibitors 651 Thiorphan Omapatrilat ACE IC50 = nM NEP IC50= nM (48) (51) Retrothiorphan I Sampatrilat ACE IC50= nM NEP ICS0= 20 nM Kelatorphan (50) Figure 15.23 Examples of TSA as NEP inhibitors sive response (106, 107) Several dual inhibitors are in phase I11 clinical trial for treating hypertension (Fig 15.24) Omapatrilat (51, BMS-189921) is the furthest along as of late 2001 (105) Matrix metalloproteases (MMP) are also inhibited by hydroxamic acids and/or thiols Over 25 variants of these enzymes are known, and some are involved in diseases ranging from inflammation to metastatic cancer (108) MMPs contain a zinc ion in the active site and function through the metallopeptidases catalytic mechanism already discussed However, subtle differences between enzymes enable selective inhibitors to be developed (109) Fig 15.25 lists some of the reported MMP inhibitors that use carboxylic acid (52-531, a hydroxamic acid (54-55), or thiol groups (56)as metal chelators ACE = 25 nM NEP IC50= nM Figure 15.24 Examples of TSA as dual ACE/NEP inhibitors Other reported zinc binding chelators used in matrix metalloproteinase inhibitors are summarized in Fig 15.26 For instance, one of the oxygens in the phosphonamide (57) binds strongly to the zinc ion, whereas the other one coordinates weakly with the metal (110) More recently, "suicide substrate" MMP inhibitors have appeared (58) (Fig 15.26) (111) The se- Peptidomimetics for Drug Design (54) (55) (56) Figure 15.25 Traditional TSA used to inhibit metallopeptidases lectivity of this type of compound arises from the specific coordination of the thiirane with the active-site zinc ion, which facilitates thiirane ring opening by nucleophilic attack by neighboring Glu-404 This novel mode of binding was assessed by X-ray absorption studies because of the difficulty to obtain a suitable crystal structure (111,112) ADAMs are membrane proteins that contain a disintegrin and a metalloprotease domain Disintegrins are RGD-containing proteins that inhibit cewmatrix interactions (adhesion) and cewcell interactions (aggregation) through the integrin receptors In addition, ADAMs have two other domains that are involved in signaling and transport (113) There are more than 25 ADAMs proteases identified so far ADAM 17 has been shown to be TNF-a converting enzyme (TACE) (114) Inhibition of TACE slows the production of TNF-a, a potent cytokine involved in inflammatory responses to infection Normally TNF-a produces a useful response, but in some cases, too much TNF-a is released and inhibition of TNF-a production would be ther- apeutically useful Synthetic analogs have been synthesized that inhibit this enzyme Clinical candidates like Ro-32,7315 (59) (Fig 15.27) are starting to emerge, and more are expected in the near future (115,116) Aminopeptidases, enzymes that cleave off the N-terminal amino acid from a peptide chain, are bismetallo peptidases, a class of metallopeptidase that contain two metals ions in the catalytic site (117, 118) These can be inhibited by compounds related to bestatin (60) (Fig 15.28), which contains the N-terminal a-hydroxy-P-amino acid residue, sometimes referred to as norstatine In leucine amino peptidase, chelation occurs between both the amide carbonyl group and the adjacent hydroxyl and the hydroxyl and the N-termind amino group (119,120) 6.3 TSA-Derived Cysteine and Serine Peptidase Inhibitors Classical TSA inhibitors of cysteine and serine proteases differ from the metallo- and aspartic protease inhibitors in that they mimic the tet- From Transition-State Analog lnhibitors to Non-Peptide Inhibitors: Examples in Protease Inhibitors Ac-Pro-AalHN, 653 P NHPh Figure 15.26 Novel TSA used to inhibit metallopeptidases Figure 15.27 Example of TSAasan TNF-ainhibitor rahedral intermediates for enzyme-catalyzed amide bond hydrolysis only after a reversible chemical reaction between enzyme and inhibitor takes place Usually this involves the addition of the enzyme catalytic nucleophile (the serine protease hydroxyl group or the cysteine protease thiol group) to an eledrophilicgroup in the inhibitor to generate ketal-like species (121) Some of the serine and cysteine TSA moieties are shown in Fig 15.29 Selective inhibition between these two classes of protease can be achieved easily For example, trifluoromethylketones (61) and peptidyl boronic acids (62) not efficiently inhibit cysteine proteases However, selective inhibition of enzymes within each class can be very difficult Peptidomimetics for Drug Design thepsin B as potential anti-metastatic drugs (124) Cathepsin K was recently discovered and shown to be involved in osteoporosis and bone regulation (125) Inhibitors of cathepsin K illustrate the principles developed to inhibit this class of enzyme This enzyme sequence was detected in 1994 by sequencing of human DNA for the human genome project (126).Cathepsin K was found to be inhibited by leupeptin (63) and by compound (641, which surprisingly binds "backwards" to the active site (Fig 15.30) A hypothesis to develop symmetrical inhibitors of cathepsin K derived from the superposition of both aldehydes on the carbonyl carbon; this led to the diamino ketone TSA (65) The diamino ketone moiety seems to work in several classes of cysteine proteases (127) Based on these results, Marquis et al have recently described the design and synthesis of conformationally constrained cyclic ketones as highly potent and selective cathepsin K inhibitors (66-67) (Fig 15.31) (128) The labile stereogenic group in position a! of the ketone was shown to be important for the binding mode and pharmacokinetic profile of these type of inhibitors The crystal structure of the two epimers showed two alternate directions of binding to the enzyme active site In both structures, the primed region of the enzyhe was occupied by these inhibitors Further investigation, resulted in the azepanone derivative (68) as a configurationally stable template for the selective inhibition of this cysteine protease (Ki = 4.8 pM) (129) Bestatin (60) Figure 15.28 Proposed binding mode of bestatin Many cysteine peptidases are involved in the biosynthesis and degradation of biologically important peptides Most early work was done with papain, a cysteine peptidase isolated from the papaya fruit and used in meat tenderizer many years ago The readily available source of this enzyme led to one of the very first X-ray crystal structures of any peptidase (122, 1231, despite the fact that no cysteine peptidase was then known to be important in human pathology Since then, cathepsins B, H, L, and S were discovered to be involved in biosynthetic steps in human immune response, inflammation, and other biologies For example, cathepsin B is clearly involved in the metastatic process and must act at some stage to permit transformed tumor cells to migrate to other parts of the body; for 20 years, people have sought inhibitors of caR1 gNAB/o~ H I OH Trifluorornethylketone Boronic acid (61) (62) Phosphonic acid $ HN y N H% Diaminoketone or-Ketoamide Figure 15.29 TSA used to inhibit serine or cysteine peptidases Speeding Peptidomimetic Research (.c;!c:"H2 HyHTr H AcHN -zko Leupeptin 'r N - NHCbz 0 Figure 15.30 Structure-based design of cathepsin K inhibitors Caspases are involved in a variety of cell functions, especially in programmed cell death (apoptosis) These enzymes recognize tetrapeptide sequences ending in an aspartic acid recognition point: X-Y-Z-Asp-NHR Much effort has been expended in trying to obtain selective inhibitors of the 14 different types identified to date In this context, selective inhibitors of caspase or of caspase 317 have recently been reported (130) Peptidomimetic modifications of the tetrapeptide sequence have led to the conformationally constrained compound (69)as a selective inhibitor of caspas&l or interleukin-lp converting enzyme (ICE) as potential anti-inflammatory compounds (131) Recently, new non-peptide peptidomimetic diphenyl ether sulfonamides have been described as novel lead structures (70) (Fig 15.32) (132) Finally, researchers from SmithKline Glaxo have identified potent and selective inhibitors of caspases and that lack the required carboxyl group in P, (71) (Fig 15.32) The X-ray co-crystal structure reveals the for- mation of the typical tetrahedral intermediate of the isatin type structures, which may compromise its selective inhibition of proteases (133, 134) These reversible caspase inhibitors differ from inhibitors that form irreversible covalent bonds, the so-called "dead-end" or "suicide" inhibitors of these enzymes, For example, the a-acetoxy ketone (72) in Fig 15.32 is an alkylating irreversible inhibitor; the enzyme cysteinyl group displaces the a-acetoxy group to form an irreversible covalent bond (135) SPEEDING UP PEPTIDOMIMETIC RESEARCH As mentioned before, combinatorial chemistry, high-throughput screening, and analogous techniques have become powerful tools to promote drug discovery in peptidomimetic research It is not the intention of this chapter to summarize all these methods, and excellent Peptidomimetics for Drug Design Figure 15.31 Cyclic ketones in novel cathepsin K inhibitors Figure 15.32 Examples of TSA as caspase inhibitors 8 Toward Rational Drug Design: Discovery of Novel Non-Peptide Peptidomimetics Figure 15.33 Somatostatin receptor agonists found through combinatorial chemistry reviews are available in the literature (136140) However, one successful approach developed at Merck for the rapid identification of selective agonists of the somatostatin receptor through combinatorial chemistry should be highlighted, because it illustrates the evolution of a constrained peptide into a non-peptide peptidomimetic structure (141) A series of combinatorial libraries were constructed on the basis of molecular modeling of known peptide agonists like MK-678 and ocreotide A chemical collection of 200,000 compounds was screened, giving priority to the residues Tyr-Trp-Lys, important pharmacophores in somatostatin determined first by Veber et al (31) This approach yielded five compounds (73-77) (Fig 15.33), each being selective for one of the somatostatin receptor subtypes: sstl (73), sst2 (74), sst3 (75), sst4 (76), and sst5 (77) TOWARD RATIONAL DRUG DESIGN: DISCOVERY OF NOVEL NON-PEPTIDE PEPTIDOMIMETICS Current pharmaceutical research has taken advantage of newer computational methods, the so-called computer-aided drug design, and other physicochemical techniques such as X- Peptidomimetics for Drug Design OMe (79) Rich et al Figure 15.34 Examples of GRAB peptidomimetics ray crystallography and NMR (142) The main goal in rational drug design is to translate the structural information in the native peptide into low molecular weight non-peptidic molecules Over the past years, many 3D structures of biological targets have been solved and have been successfully used to design new, pharmacologically useful compounds (vide infra) Different computer-aided design methods, e.g., 3D pharmacophore model, 3D quantitative SAR (QSAR), docking, and de novo design, have been extensively reviewed elsewhere (75, 143-146) Recently, the importance of generating inhibitors that target receptor conformational ensembles has been pointed out (10) This method goes beyond the current docking of known structures to known active site conformers and can lead to type 111 and GRAB peptidomimetics The concept of Group Replacement Assisted Binding (GRAB) peptidomimetics derives from the discovery at Roche of the piperidine class of renin inhibitors The non- peptide inhibitors of renin (78) (Ki = 26 and (79) (Ki = 31 nM)(Fig 15.34)(84, 147149) stabilize an enzyme active site conformation different than the P-strand binding enzyme conformation typical for other peptidase inhibitors A close analysis of the X-ray crystal structure of the enzyme inhibitor complex shows that the piperidine C4-phenyl group binds to the enzyme to replace Tyr75that has rotated to another position Interestingly, Leu73also rotates to fill some of the vacated Tyr75pocket, and this in turn allows Trp3' to occupy a new site formed in part by the vacated Leu73(Fig 15.35) This cascade of conformational transitions in the renin active site allows the optimized inhibitor to stabilize an enzyme conformation not observed when the classic peptide-derived peptidomimetics bind This stabilization process is defined as group replacement process, and the piperidine inhibitors constitute a new type of peptidomimetic: GRAB peptidomimetics Comparison of (78) and (79) with the structures of other peptide-derived inhibitors re- - Historical Development of Important Non-Peptide Peptidomimetics Figure 15.35 GRAB peptidomimetics in action See color insert veided how the different enzyme active site C01 formation were found Bursavich et al have successfully extended the initial renin )deling to the design of inhibitors of two ier aspartic peptidases: pepsin and R chizsis pepsin (80) (Ki = p&) and (81)(Ki = r-JM) (Fig 15.34) (150) The extended P-strand binding conformation could be changed into the piperidine bindin@ ; conformation by a series of low-energy, me!chanisticallyrelated conformational changes in active site side-chains The discovery of the ROIthe inhibitors and the correlation of these structures with peptide-derived inhibitors are an;dogous to a peptidomimetic "Rosetta Stone." This design strategy has the potential for designing novel types of peptidomimetic structur'es HISTORICAL DEVELOPMENT OF IMIPORTANT NON-PEPTIDE PEF'TIDOMIMETICS HIV Protease TYI?e-I HIV-1 protease inhibitors, Saquinavir, Ritonavir, Indinavir, Amprenavir, Viracept (neflinavir mesilate), and Lopinavir (Fig 15.20) are established drugs for the treatment of 1UDS All these inhibitors employ the centra:I hydroxyl transition state mimetic as a scaffold on which varying functionality was systematically added until the required balanc:ebetween potency, in uiuo activity and oral absorption was achieved In general, the binding interactions were optimized through iternative synthesis and co-crystallization of inhib:itor with enzyme, molecular modeling, and re-clesigning the inhibitor side-chains Pharmac:okinetic properties were addressed only aft€:r the initial inhibitor was identified and opt:imized Compounds (82-83) (Fig 15.36) are highly modified peptidic structures that stabilize the enzyme-bound extended p-conformation (151, 152) Another approach to achieve greater in vivo activity is to start with a molecular template with proven useful pharmacokinetics and oral bioavailability and to selectively modify it to achieve the desired activity Identification of the orally active anticoagulant warfarin (84) (Fig 15.37) as a weak inhibitor (IC,, = 18 p&)of HIV protease was followed by two reports of 4-hydroxycoumarins as possible type I11 HIV inhibitors Subsequent SAR studies led to the more potent 5,6-dihydro-4-hydroxy-3-pyrone inhibitor (85)(IC,, = 2.7 nM), which has good anti-viral activity (EC,, = 0.5 CLM) and is orally bioavailable (153) Upjohn researchers also used a structure-based design approach based on warfarin to obtain (86), their clinical candidate PNU-140690 (154) It should be noted that both inhibitors bind to the extended P-strand binding active site conformation Workers at DuPont used a pharmacophore model and database search to develop the first type I11 mimetic inhibitor of HIV protease, DuP 450 (87) (Fig 15.38) This evolved from a 3D pharnacophore that retained two key interactions: replacement of the flap-bound water and a hydroxyl transition-state isostere (155) Molecular modeling led to a cyclohexanone as a better spacer between these groups, and finally the seven-membered cyclic urea (87) was created (Fig 15.38) The development of these inhibitors illustrates the importance of conformational analysis in the design of constrained analogs Surprisingly, the symmetric cyclic sulfonyl-urea derivative analog (88) (Fig 15.38, Ki = nM) binds differently in the active site and adopts a flipped conformation (156) Peptidomimetics for Drug Design Figure 16.36 @-StrandHIV protease inhibitors Moreover, SAR of the cyclic urea and cyclic sulfamide inhibitors not follow a straightforward pattern These contradictory results clearly illustrate the structural diversity created by a subtle structural modification in two otherwise related peptidomimetic protease inhibitors The peptidase inhibitors, (82) and (83),are actually amino acid and transition-state mimics pieced together to emulate the typical ligand-bound extended p-strand inhibitor conformation The structurally distinct heterocyclic aspartic protease inhibitors (85-86) and (87-88) are non-peptide peptidomimetics because of their remote structural relationship to native peptide substrates Yet these two distinct peptidomimetic classes bind to the same active site topography These structurally distinct peptidomimetics selectively stabilize closely related enzyme conformations 9.2 Thrombin Thrombin and Factor Xa are both serine proteases involved in the blood coagulation cascade Inhibition of these two enzymes is providing novel anticoagulants (157-159) The development of thrombin inhibitors that lack the functionalized TSA highlights a major new approach to type I peptidomimetics In 1995, a Lilly group found that D-PhePro-Agmatine analogs showed increased selectivity for thrombin over other fibrinolytic enzymes despite a 100-fold loss in potency caused by the removal of the aldehyde group (160) These studies paved the way for Merck's development of picomolar thrombin inhibitors (161, 162), which use a similar motif Removal of an a-ketoamide transition state mimic from L-370,518 (89) (Fig 15.39, Ki = 0.09 nM)led to an expected 100-fold drop in potency for (90) (Ki = nM) However, systematic modification of the P, position restored potency and led to an inhibitor (91) with a Ki = 2.5 pM Interestingly, potency seems to be enhanced by a fortuitous hydrophobic collapse into a favorable binding conformation Thrombin inhibitors (92) and (93) illustrate a novel type 111 peptidomimetic Most protease inhibitors bind in an extended p-strand conformation that is stabilized by multiple enzyme ligand hydrogen bonds 9 Historical Development of Important Non-Peptide Peptidomimetics Parke-Davis X-ray structure (85) Pharmacia-Upjohn X-ray structure Warfarin (86) (84) Figure 15.37 Warfarin analogs as non-peptide HIV protease inhibitors Figure 15.38 Cyclic ureas as non-peptide HIV protease inhibitors However, Boehringer Mannheim developed thrombin inhibitors (92) (Fig 15.40) that lack these H-bonds (163) This idea was exploited by researchers at 3D Pharmaceuticals, who were able to crystallize (93) in the active site (164, 165) In this example, the benzene ring acts as a scaffold to display the three different substituents to fill the three principal binding pockets Other type I11 peptidomimetic inhibitors of thrombin have been developed from screening leads (166, 167) such as inhibitor (94) (Fig 15.40) SAR led to the design of (95).Inhibitor (96) was derived from docking studies with the 5-amidino indole nucleus, followed by addition of a lipophilic side-chain to interact with the important S, subsite of thrombin The crystal structures of both (95) and (96) in the active site of thrombin shows that the aromatic core, binds in the S, site as expected, but Peptidomimetics for Drug Design One of the major drawbacks in thrombin inhibitor design was the requirement for a basic side-chain in P, needed to form a salt bridge to enzyme Asp189 However, the other amino acid side-chains in S, are largely lipophilic and neutral This feature suggested that the strongly basic group in P, could be replaced by a weaker base or even with hydrogen-bonding groups Compounds (99-100) are representative of this strategy (Fig 15.40) (171).An X-ray crystal structure of (99) shows a new binding mode in which the formamide group points out of the S, pocket and forms new hydrogen bonds with Gly219 (172) The ability to obtain crystal structures of thrombin inhibitors complexes for many of the inhibitors shown in Figs 15.40-15.41 establishes that most are type 111 peptidomimetics 9.3 Figure 15.39 Non-TSA thrombin inhibitors does not pick up hydrogen bonding from the important active site sequence Ser214Gly216 (168) Both crystal structures showed a similar binding mode; where interaction of the C-2 side-chain with Trp6OD might explain the high thrombin selectivity observed for this series (169) Another type I11 peptidomimetic inhibitor was derived from the crystal structure of a bicyclic [3.1.31 inhibitor (170) complexed to thrombin (97) (Fig 15.41) The X-ray structure revealed that one of the carbonyls was oriented towards the hydrophobic P-pocket (S,) The desolvation necessary to place a carbony1 in a hydrophobic pocket is unfavorable and various alkyl groups were used as possible replacements This led to the potent (Ki = 13 nM) and selective (>760 for thrombin over trypsin) inhibitor (98) Factor Xa New approaches to design inhibitors of Factor Xa as potential anticoagulants have been reviewed (173),and important type I11 mimetics have been described (Fig 15.42) All inhibitors contain amidine or basic groups that bind in the enzyme's S, site; none of the inhibitors contains a classical electrophilic center of the type employed in TSA inhibitors (174-180) Compound (101)(Fig 15.42) was designed from a strategy involving connection of a three-point pharmacophore derived from molecular modeling Beginning with the X-ray structure of the Factor Xa dimer, Gong et al (176) envisioned three important enzymatic contact points: a phenylamidine in the S, subsite, a phenylamidine in the S, site, and a carboxylate moiety postulated by a group at Daiichi to confer selectivity over thrombin through an interaction with Gln192 of Factor Xa Systematic iterative modifications led to = nM), which the potent inhibitor (101) (Ki has 350-fold selectivity over thrombin This approach highlights a truly de novo method where fragments were docked into the active site and an appropriate spacer was chosen to connect them Further SAR data led to modifications that improved both potency and selectivity (176) 9.4 Clycoprotein ilb/llla (GP llb/llla) Some outstanding examples of the use of conformational restriction to characterize the Historical Development of Important Non-Peptide Peptidomimetics Figure 15.40 Non-H-bonding-based thrombin inhibitors bioactive conformations of Arg-Gly-Asp peptidometic antagonists illustrate the present state-of-the-art Members of the integrin family of receptors recognize and bind the peptide sequence, Arg-Gly-Asp, as an important step in platelet aggregation and other physiological processes (181), and competitive antagonists for this process could serve as potential drug candidates Much effort has been directed toward identifying small ligands that might mimic the RGD peptide sequence (182) This drug design concept was supported by the fact that protein antagonists of integrin receptors are known that contain the RGD sequence (183) and that small peptide sequences containing the RGD moiety weakly antagonize the endogenous ligand (184) Consequently, several groups synthesized conformationally restricted derivatives of small peptides as starting points for developing metabolically stable peptides or peptidemimetics Ali et al (185) svnthesized a series of disulfide deriva- tives of the RGD sequence, which were designed by analogy with the somatostatin work (vide supra) Excellent antagonists related to (102) were obtained Further constraint of the peptide system by use of the o-thiol benzene derivatives led to the novel antagonist SKF 107260 (103) (Fig 15.431, a good inhibitor of both platelet aggregation and binding to GPIIbDIIa Barker et al (186) followed a similar strategy but used cyclic sulfides as an additional conformationally restricting element These derivatives had the advantage of being rapidly synthesized by solid phase methods Systematic structure-activity studies with respect to the amino acid preceding the RGD sequence and the chirality of sulfoxide derivatives led to the discovery of (3-4120 (1041, a potent, biologically active derivative The conformation of both (103)and (104) in water was found to be highly constrained, and a single predominant conformation could be characterized in aqueous solution by use of Peptidomimetics for Drug Design / TMS Figure 15.41 Optimized PI and P, thrombin inhibitors NMR methods and computational chemistry (185, 187) This bioactive conformation defined the topographical placement of the arginine guanidine group and the aspartic carboxyl group, and was superimposed onto a conformationally restricted template of a class of compounds with generally suitable pharmacodynamic properties In this case, the benzodiazepine ring system was used, and the strategy generated the low molecular non-peptide RGD receptor antagonist (105-107) (Fig 15.441, which contain at least two conformational restrictions, the bicyclic heterocycle and the acetylene linker The compounds shown in Fig 15.44 represent what can be achieved by applying the principles of conformational restriction to peptides when no X-ray or NMR structural information are available for the complex between ligand and receptor Benzodiazepines (105-107) represent the first type I11 peptidomimetics designed de novo by sys- tematically modifying a natural receptorbinding peptide (187-189) A variety of other scaffolds have been developed by exploiting the idea that glycine represents a spacer between the two important recognition residues Arg and Asp This template-based approach positions the key side functionality, a basic function and an acid one within a distance of 11-17 A, required for presentation to the receptor Several examples of these scaffoldsare shown in Fig 15.45 (190-195) Recent results suggest that the RGD tripeptide can adopt multiple conformations that allow tight binding to the receptor This theory is supported by the fact that nonpeptide RGD peptidomimetics can adopt a range of different topographies such as found in cupped, turn-extended-turn, or p-turn conformations (196) RGD type I peptidomimetics are usually poorly bioavailable compounds because of the presence of multiple hydrogen-bonding sites Historical Development of Important Non-Peptide Peptidomimetics Figure 15.42 Examples of FXa protease inhibitors plus the charged polar functional groups at both ends Esters or coumarin (197) linkers have been used to provide orally available prodrugs, and bioisosteric replacements of the guanidiniurn group by a pyridine, (198) tetrahydronaphtyridine, (199), or aminobenzimidazole (200) moieties provided more bioavailable analogs 9.5 Ras-Farnesyltransferase Inhibitors of Ras-farnesyltransferase have been developed by mimicking the C-terminal CAAX motif (where C is a cysteine residue, A is any aliphatic amino acid, and X is usually Met, Ser, or Ala) This tetrapeptide is the signal for farnesylation of Ras proteins Ras-farnesyltransferase is one of the most promising targets for novel anti-cancer drugs, because at least 30% of the human cancers contain mutated Ras (201,202) Two types of peptidomimetic structures have been used to develop - inhibitors (203) Some typical type I inhibitors were generated by replacing the amide backbone with differ- Peptidomimetics for Drug Design Figure 16.43 Conformationally restricted RGD cyclic peptides 'Bu (107) Figure 15.44 Benzodiazepines RGD analogs ent isosteres like the oxymethylene amide bond in (108)(Fig 15.46, IC,, = 60 nM) (204) The central dipeptide segment of CA,A,X has been replaced with rigid linkers like the 3-arninomethylbenzoic acid (AMBA) in (109) (205) This novel inhibitor was not farnesylated, showing that the two amino acids in the middle of the CAAX tetrapeptide are required for farnesylation An imidazole group has been used to replace the thiol group of the CAM motif to produce compound (110)(206) An outstanding example in peptidomimetic design evolved from these studies Truncation and conformational restriction of a reduced isostere of the parent peptide substrate, followed by systematic replacement of the peptide-like side-chains provided the potent nonpeptidic inhibitor (111) (Fig 15.46) (207) This approach highlights the transition from a peptide-derived structure to a compound with no apparent resemblance to the original peptide Recently, crystal structures of farnesyltransferase complexed with a farnesyl group donor and the native substrate or a type I peptidomimetic show the structural basis for inhibition of this enzyme The X-ray data show that the CAAX motif adopts an extended conformation rather than a p-turn, which is the conformation observed by transferred nuclear Historical Development of Important Non-Peptide Peptidomimetics Figure 15.45 Different scaffolds used in RGD mimetics Overhauser effect experiments; coordination of the cysteine side-chain to the Zn ion promotes the conformational change in the peptide backbone Moreover, differences in the conformation binding mode of peptides and peptidomimetics is one of the bases for selective farnesylation (208) Other type 111 peptidomimetic inhibitors of this enzyme have also been reported Inhibitor (112) (Fig 15.47) was developed by replacing the A,& dipeptidyl sequence with a benzodiazepine scaffold (209) Later, SAR modifications of the benzodiazepine nucleus that included a hydrophilic 7-cyano group and a 4-sulfonyl group provided the potent, orally available and in uiuo active (113)(210) HTS also produced several non-peptide leads typified by inhibitor SCH 47307 (114) (Fig 15.47) (211) Subsequent SAR work led to the potent inhibitor SCH 66701 (115)(Ki = 1.7 which was crystallized within the enzyme active site (212) This series of compounds is completely non-peptidic and also lacks the free sulfhydryl or imidazole seen in the other inhibitors discussed here This is a breakthrough that shows that potency can be achieved even without the "essential" cysteine or sulfhydryl mimic a), Non-Peptidic Ligands for Peptide Receptors 9.6 This section illustrates the successful development of non-peptide peptidomimetics from a screening lead by assuming the inhibitor binds to the receptor in the same way as does the native peptide hormone These assump- Peptidomimetics for Drug Design SMe (109) (111) Figure 15.46 Peptide-like Ras-farnesyltransferaseinhibitors tions actually led to effective inhibitors of the receptor Later, site-directed mutagenesis of target receptors suggested that for many of these compounds, the mimetic was binding to the receptor at ancillary, perhaps overlapping, sites on the receptor Later still, pharmacological studies indicated that peptide receptors adopted multiple states, suggesting that different antagonists might bind to different receptor forms Of course, if compounds not bind to the same receptor site as the endogenous hormone, SAR data collected on the natural peptide substrate is not applicable to these antagonists Most of these peptidomimetics are probably type I1 or functional mimetic~.Yet the success of this approach suggests that at least for some non-peptide antagonists, there may be some congruent structure that interacts with the receptor These issues will only be determined unambiguously when high-resolution structures of the G-protein-coupled receptors (213) and other constitutive receptor systems are determined 9.6.1 Angiotensin II The first non-peptide antagonists of the AT1 receptor were found by HTS The imidazole (116) (Fig 15.48) is a weak (IC,, = 43 pJ4) but quite selective A-I1 receptor antagonist (214) Using this as a lead compound, DuPont and SmithKline Glaxo researchers independently developed potent small molecule A-I1 receptor antagonists The DuPont group used the conformation sug-, gested by Smeby and Fernandjian to guide the design (215) It was speculated that the carboxyl group and the imidazol group of (116) were bound to the A-I1 terminal carboxyl group and to the imidazole group, respectively This rationalization culminated in the synthesis of nanomolar inhibitors, with compound (117) as a clear representative (216) Although workers at SmithKline Glaxo used the same conformation as starting point, they postulated other binding modes to the receptor One of their alternative hypothesis considered compound (116) as a constrained analog in which the benzyl and the carboxyl groups corresponded to the Tyr side-chain and the C-terminal carboxyl group of A-11 Following this hypothesis, modification of lead compound (116)eventually led to compound (118) (Fig 15.48) with an IC,, = 1.45 nM and oral activity of 30% (217) Site-directed mutagenesis studies on the AT1 receptor revealed differences in the bind- Historical Development of Important Non-Peptide Peptidomimetics Figure 15.47 Non-peptidic Ras-farnesyltransferase inhibitors ing site of angiotensin and the small molecule non-peptide compounds (119-120) (Fig 15.49) There is no evidence that the single residues involved in inhibitor binding overlap with endogenous peptide binding Some other non-peptide agonists have also appeared in the literature Surprisingly, their binding mode differs from the binding mode of the peptide agonist (121), as well as that of the structurally similar non-peptide antagonist (122) (Fig 15.49) (218) However, angiotensin and L-162,313 (122) require common critical residues for angiotensin AT1 receptor activation (219) 9.6.2 Substance P The tachykinin receptors (NK-1, NK-2, and NK-3) and their endogenous ligands, the tachykinins, and neurokinins are important neurotransmitters (220- 222) Antagonists of tachykinin receptors ~roducebeneficial effects in several CNS disease states such as pain, asthma, emesis, and depression A general approach for converting a variety of peptide structures into small, type I1 peptidomimetic antagonists was devised by Horwell and colleagues and is illustrated here for antagonists to Substance P An alanine scan of the parent undecapeptide revealed that the Phe4-Phes sequence was required for binding Replacement of one these residues by Trp, followed by introduction of conformational constraints by a-alkylation, provided the subnanomolar inhibitor (123) (Fig 15.50) (223) Improved brain penetration was achieved by amine (124) (224) Chemical screening of corporate compound libraries resulted in the discovery of another Peptidomimetics for Drug Design Asp-Arg-Val-Tyr-He-His-Pro-Phe-OH Angiotensin II Figure 15.48 Angiotensin I1 inhibitors derived from a HTS lead type of non-peptidic NK-1 antagonist, CP96,345 (125) (Ki = 0.66 nM, Fig 15.51) This compound heralded a breakthrough in the design of these potential drugs (225, 226) Replacement of the basic quinuclidine ring with a morpholine core improves duration of action and insertion of an amino triazole unit confers excellent solubility and CNS penetration (126) (Ki = 0.19 nM, Fig 15.51) (227) Dual NK-1/NK-2 inhibitors, e.g., (127) (Fig 15.52), have recently been designed by determining the important sites for maintaining NK-2 selectivity of the lead compound SR48968 and introducing NK-1 pharmacophore groups (228-230) Fewer NK-3 selective receptor antagonists have been described, but a quinoline scaffold previously reported to be a selective NK-3 receptor antagonist, has been converted to a dual NK-2/NK-3 inhibitor (128) (Ki = 0.8 nM NK-2 and 0.8 nM NK-3) The lead optimization was carried out by docking potential structures into a novel receptor model The theoretical model compares closely with the recently published crystal structure of rhodopsin (231) 9.6.3 Neuropeptide Y Neuropeptide Y (NPY) is a 36 amino acid polypeptide that is involved in hormonal, sexual, and cardiac effects (232, 233) In 1994, two "first generation" type I NY-l selective antagonists, BIBP 3226 (129) (234) and SR120107A (130)(235) were reported (Fig 15.53) BIBP 3236 (129) corresponds to a truncated and modified peptide in which the D-Arg is assumed to correspond with in Neuropeptide Y More recently, a series of indole Y1 antagonists discovered by screening (236) led to the benzimidazole (131) (Ki = 0.052 nM) In this type of compound, the diamino moieties are postulated to mimic the two C-terminal arginines of NPY (237) Selective NPY-Y5 inhibitors have been shown to inhibit food intake activity in vivo Most inhibitors found by HTS and lead optimization gave nanomolar and selective antagonists It is not known whether these are functional or topological mimetics (Fig 15.54) (238-240) 9.6.4 Growth Hormone Secretagogues Growth hormone (GH) releasing peptide mimetics have become attractive alternatives to Historical Development of Important Non-Peptide Peptidomimetics Figure 15.49 Examples of angiotensin I1 inhibitors GH replacement therapy (241) The peptidyl GH secretagogue GHRP-6 (242) was used to develop the clinical candidate MK-0677 (132) (Fig 15.55, EC,, = 1.3 nM) (243, 244) After Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-Leu-Met-NH2 Substance P Figure 15.50 Development of a substance P inhibitor identifying the important residues for bioactivity in GHPR-6, the Merck group began searching other receptor libraries for known "privileged structures" in a combinatorial synthetic fashion (see Section 4) (66) The more active derivative contained a spiropiperidine moiety attached to an indoline ring More recently, ghrelin has been isolated and identified as an endogenous ligand of the GHS receptor and some new peptidomimetic structures [e.g., 133 (Fig 15.5511have started to appear (245) In another approach, SAR studies and systematic simplification of GHPR-6 at Novo Nordisk produced the orally bioavailable derivative NN-703 (134) Molecular modeling overlapping of NN703 (134) (Fig 15.56) and MK-0677 (132) (Fig 15.55) showed structural similarities between both compounds Highly potent hybrids of Ipamorelin and NN-703 (e.g., 135) (Fig 15.56) have also been described (246, 247) Peptidomimetics for Drug Design Figure 15.52 Dual NK-1/NK-2 and NK-1/NK-2 inhibitors Figure 15.51 Examples of NK-1 antagonists A common 3D pharmacophore was recently described for peptidic and non-peptidic GH secretagogues by means of computational chemistry After QSAR analysis, four pharmacophoric sites were found: two aromatic rings, a proton acceptor, and a protonated amine Using these strategies, some nanomolar antagonists [e.g., 136 (Fig 15.56)] were discovered (248) 9.6.5 Endothelin The first report of endothelin in 1988 stimulated a huge effort to develop selective and non-selective endothelin receptor (ETA and ETB) antagonists (249, 250) One successful approach derived from the postulate that the phenyl groups of the screening lead might mimic two of the aromatic side-chains (Tyr13, Phe14, or Trp21) of ET-1 (251,252) Knowing that the carboxylic acid was also necessary for good activity, researchers at SmithKline overlaid their inhibitor with the aromatic groups Tyr13, Phe14,and Asp1' in ET-1 After using a conformationally constrained analog of ET-1 to further define their NMR-derived structure of ET-1, the final overlay suggested that a carboxylic acid attached with a linker of two to three atoms on the 2-position of the phenyl ring would provide further binding interaction by mimicking the C-terminal carboxylic acid This led to compound (137) (Fig 15.571, a potent antagonist of both the ETAand ET, receptors with Ki= 0.43 and 15.7 nM,respectively Analogs based on a pyrrolidine scaffold are also effective (e.g., 138) (Fig 15.57) (253) The Kohonen neural network has been used to develop bioisosteres of the methylendioxyphenyl group found in a variety of antag- Historical Development of Important Non-Peptide Peptidomimetics 673 Tyr-Pro-Ser-Lys-Pro-Asp-Asn-Pro-Gly-Glu-Asp-Ala-Pro-Ala-Glu-Asp-Leu-Ala Arg-Tyr-Tyr-Ser-Ala-Leu-Arg-His-Tyr-lle-A~n-Leu-lle-Thr-Arg-Gln-Arg-Tyr-NH~ Neuropeptide Y H N NH2 ~ ,NH Figure 15.53 Examples of neuropeptide Y1 inhibitors onists [e.g., 139 (Fig 15.58)l The benzothiadiazole (140) functions as a bioisostere that retains and sometimes improves binding to the ETAreceptor (254-256) Since the discovery of Ro46-2005 (141) (Fig 15.58), the first orally active ET inhibitor, major efforts have been made to modify arylsulfonamide derivatives An isoxazole as Peptidomimetics for Drug Design Figure 15.54 Examples of neuropeptide Y5 inhibitors the heterocycle attached to the amino fundionality provided selectivity against ETA receptor (257)and led to BMS193884 (142)(Ki= 1.4 nM) (258)and others, e.g., TBC 3214 (143) (Ki = 0.04 nM) (2591, which are potent, selective, and orally available ETAantagonists Different binding modes have been proposed for ET antagonists The acid or sulfonamido groups are needed to interact with a cationic site in the receptor, and an aromatic interaction with Tyr12' is postulated to be responsible for ETA selectivity However, because all these receptors are members of the GPCR, there is no assurance that any bind as modeled Thus, they must be classified as type I1 peptidomimetics until structural data can resolve the issue 10 SUMMARY AND FUTURE DIRECTIONS The "Holy Grail" of peptidomimetic research in drug discovery has been to find ways to transform the structural information contained in peptides into non-peptide structures that have drug-like pharmacodynamic properties Many different strategies have been His-D-Trp-Ala-Trp-D-Phe-Lys-NH2 GHRPB Ghrelin \ / \ TFA S02CH3 Figure 15.55 GHRP-6 and ghrelin non-peptide derivatives as growth hormone secretagogues inhibitors Figure 16.56 Newer approaches to growth hormone secretagogues inhibitors Cys-Ser-Cys-Ser-Ser-Leu-Met \ Figure 15.57 Non-peptide endothelin analogs I Peptidomimetics for Drug Design N-S OMe PD156707 OMe EMD 122946 (139) (140) I BMS 193884 (142) Figure 15.58 Examples of ETAinhibitors employed in the search for useful peptidomimetics-rational design of amide bond replacements, mimics of turn structures, and the like, as well as both designed and discovered scaffolds that replace the amide bond core of peptides The field has a long way to go before rational design of type I11 peptidomimetics can be achieved routinely However, the progress made to date suggests that this goal will be achieved We know that some nonpeptide scaffolds are topographical mimetics of the extended P-strand of enzyme-bound protease inhibitors because we have the biophysical methods for characterizing both types of enzyme-inhibitor complexes Type I11 peptidomimetic inhibitors of peptidases have References been designed from the substrate sequences and they have been revealed by HTS processes and optimized by application of structural biology At this point, we have learned more about the design of inhibitors by studying how screening leads inhibit enzymes than from the design of inhibitors from our current, limited knowledge of enzyme catalysis Probably the most important recent discovery is that some screening leads inhibit proteases by binding to a different enzyme active site conformation that is related mechanistically to the wellcharacterized extended P-strand of enzymebound protease inhibitors This result emphasizes the importance of considering the entire ensemble of protein conformations when designing inhibitors of peptide-protein interactions Our understanding of peptide mimicry for ligands of constitutive receptors, such as Gprotein-coupled receptors (GPCR), is much more primitive because high resolution structural data for agonist- and/or antagonist-receptor complexes are not yet available For this reason, all attempts to rationalize the interactions between ligand and receptor contain a considerable element of speculation It is too early to know whether small non-peptide structures that bind to GPCR are functional or topographical mimetics However, based on the results obtained by studying peptidase inhibitors, it seem likely that at least some of the known functional peptidomimetics receptors ligands will be shown to be topographical mimetics Others may be found to act more like GRAB-peptidomimetics in that they bind to receptor conformations closely related in energy and mechanism to native conformations Still others will no doubt be found that inhibit or stimulate the receptor system by allosteric mechanisms or by interfering with some multi-step binding process preceding the formation of the active ligand-receptor complex In any case, it is clear that successful design of functional mimetics by assuming some structural relationship between a screening lead and the parent peptide can work (see Section 9.6),as can the systematic modification of the parent peptide The application of the principles of peptidomimetic research has become very important to drug discovery Although our present knowledge about protein-protein interactions is still quite limited, the rapid growth of structural information and methods will eventually allow us to design rationally peptidomimetic compounds suitable for use in human therapy REFERENCES M D Fletcher and M M Campbell, Chem Rev., 98, 763 (1998) F Haviv, T D Fitzpatrick, C J Nichols, E N Bush, G Diaz, G Bammert, A T Nguyen, E S Johnson, J Knittle, and J Greer, J Med Chem., 37, 701 (1994) J Hughes, T W Smith, H W Kosterlitz, L A Fothergill, B A Morgan, and H R Morris, Nature, 258,577 (1975) G D Smith and J F Griffin, Science, 199, 1214 (1978) A Aubry, N Birlirakis, M Sakarellos-Daitsiotis, C Sakarellos, and M Marraud, Biopolymers, 28,27 (1989) A F Bradbury, D G Smyth, and C R Snell, Nature, 260, 165 (1976) P S Farmer in E J Ariens, Ed., Drug Design, Academic Press, New York, 1980 A B Smith 111, T P Keenan, R C Holcomb, P A Sprengeler, M C Guzman, J L Wood, P J Carroll, and R Hirschmann, J Am Chem Soc., 114,10672 (1992) A S Ripka and D H Rich, Curr Opin Chem Biol., 2,441 (1998) 10 M G Bursavich and D H Rich, J Med Chem., 45, 541 (2002) 11 A F Spatola in B Weistein, Ed., Chem Biochem Amino Acids, Pept., Proteins, Marcel Dekker, New York, 1983 12 U Gether, Endocr Rev., 21,90 (2000) 13 D P Fairlie, G Abbenante, and D R March, Curr Med Chem., 2,654 (1995) 14 R A.Wiley and D H Rich, Med Res Rev., 13, 327 (1993) 15 J D A Tyndall and D P Fairlie, Curr Med Chem., 8,893 (2001) 16 N R A Beeley, Drug Discov Today., 5, 354 (2000) 17 R M Freidinger, Trends Pharmacol Sci., 10, 270 (1989) 18 R M Freidinger, Curr Opin Chem Biol., 3, 395 (1999) 19 A Giannis and T Kolter, Angew Chem Intl Ed Engl., 32, 1244 (1993) Peptidomimetics for Drug Design 20 G J Moore, Trends Pharmacol Sci., 15, 124 (1994) 21 D C Rees, Cum Med Chem., 1,145 (1994) 22 E E Sugg, Annu Rep Med Chem., 32, 277 (1997) 23 T K Sawyer, Drugs Pharm Sci., 101, 81 (2000) 24 B A Morgan and J A Gainor, Annu Rep Med Chem., 24,243 (1989) 25 G J Moore, Proc West Pharmacol Soc., 40, 115 (1997) 26 A Giannis and F Rubsam, Adv Drug Res., 29, (1997) 27 M Goodman and S Ro in M E Wolff, Ed., Burger's Medicinal Chemistry and Drug Discovery, vol 1, Wiley-Interscience, San Diego, CA, 1995, pp 803-861 28 D Obrecht, M Altorfer, and J A Robinson, Adv Med Chem., 4, (1999) 29 J Gante, Angew Chem Zntl Ed Engl., 33, 1699 (1994) 30 P A Hart and D H Rich, Pract Med Chem., 393-412 (1996) 31 D F Veber, Pept.: Chem Biol., in J A R Smith and E Jean, Eds., Proc Am Pept Symp., 12th, ESCOM, Leiden, (1992) 32 G R Marshall, Tetrahedron, 49,3547 (1993) 33 J W Erickson and S W Fesik, Annu Rep Med Chem., 27,271 (1992) 34 G Muller, Curr Med Chem., 7,861 (2000) 35 D F Mierke and C Giragossian, Med Res Rev., 21,450 (2001) 36 D F Veber, F W Holly, R F Nutt, S J Bergstrand, S F Brady, R Hirschmann, M S Glitzer, and R Saperstein, Nature, 280, 512 (1979) 37 J Rivier, M Brown, and W Vale, Biochem Biophys Res Commun., 65, 746 (1975) 38 G D Rose, L M Gierasch, and J A Smith, Adv Protein Chem., 37, (1985) 39 P W Schiller, The Peptides: Analysis, Synthesis and Biology, Vol 6, Academic Press, Orlando, FL, 1984 40 T K Sawyer, V J Hruby, P S Darman, and M E Hadley, Proc Natl Acad Sci USA, 79, 1751 (1982) 41 K Ishikawa, T Fukami, T Nagase, K Fujita, T Hayama, K Niiyama, T Mase, M Ihara, and M Yano, J Med Chem., 35,2139 (1992) 42 G R Marshall, F A Gorin, and M L Moore, Annu Rep Med Chem., 13,227 (1978) 43 G R Marshall, C D Barry, H E Bosshard, R A Dammkoehler, and D A Dunn, ACS Symp Ser., 112,205 (1979) 44 R M J Liskamp, Recl Trav Chim Pays-Bas., 113, l(1994) 45 G Holzemann, Kontakte (Darmstadt), 1, (1991) 46 G Holzemann, Kontakte (Darmstadt), 2, 55 (1991) 47 M Kahn, Synlett, 821-826 (1993) 48 R M Freidinger, D F Veber, D S Perlow, J R Brooks, and R Saperstein, Science, 210, 656 (1980) 49 E D Thorsett, E E Harris, S D Aster, E R Peterson, J P Snyder, J P Springer, J Hirshfield, E W Tristram, A A Patchett, E H Ulm, and T C Vassil, J Med Chem., 29,251 (1986) 50 G A Flynn, E L Giroux, and R C Dage, J.Am Chem Soc., 109,7914 (1987) 51 U Nagai, K Sato, R Nakamura, a n d R Kato, Tetrahedron, 49,3577 (1993) 52 S Hanessian, G McNaughton Smith, H G Lombart, and W D Lubell, Tetrahedron, 53, 12789 (1997) 53 A J Souers and J A Ellman, Tetrahedron, 57, 7431 (2001) 54 K Burgess, Acc Chem Res., 34,826 (2001) 55 A B Smith 111, M C Guzman, P A Sprengeler, T P Keenan, R C Holcomb, J L.Wood, P J Carroll, and R Hirschmann, J Ani Chem Soc., 116,9947 (1994) 56 K D Stigers, M J Soth, and J S Nowick, Curr Opin Chem Biol., 3, 714 (1999) 57 J S Nowick, E M Smith, and M Pairish, Chem Soc Rev., 25,401 (1996) 58 R P Cheng, S H Gellman, and W F DeGrado, Chem Rev., 101, 3219 (2001) 59 J Venkatraman, S C Shankaramma, and P Balaram, Chem Rev., 101,3131 (2001) 60 D P Fairlie, M L.West, and A K Wong, Curr Med Chem., 5,29 (1998) 61 V J Hruby and P M Balse, Curr Med Chem., 7,945 (2000) 62 V J Hruby, Acc Chem Res., 34,389 (2001) 63 H Nakanishi and M Kahn, Bioorg Chem Pept & Protein, 12,395 (1998) 64 R Hirschmann, P A Sprengeler, T Kawasaki, J W Leahy, W C Shakespeare, and A B Smith 111, Tetrahedron, 49,3665 (1993) 65 Y Qian, A Vogt, S M Sebti, and A D Hamilton, J Med Chem., 39,217 (1996) References 66 B E Evans, K E Rittle, M G Bock, R M DiPardo, R M Freidinger, W L Whitter, G F Lundell, D F Veber, P S Anderson, R S L Chang, V J Lotti, D J Cerino, T B Chen, P J Kling, K A Kunkel, J P Springer, and J Hirshfield, J Med Chem., 31,2235 (1988) 67 B E Evans, K E Rittle, M G Bock, R M DiPardo, R M Freidinger, W L Whitter, N P Gould, G F Lundell, C F Homnick, and D F Veber, J Med Chem., 30, 1229 (1987) 68 T Clackson and J A Wells, Science, 267,383 (1995) 69 H C Kawato, K Nakayama, H Inagaki, and T Ohta, Org Lett., 3,3451 (2001) 70 K Nakayama, H C Kawato, H Inagaki, and T Ohta, Org Lett., 3,3447 (2001) 71 M D Fletcher and M M Campbell, Chem Rev., 98, 763 (1998) 72 M Chorev and M Goodman, Acc Chem Res., 26,266 (1993) 73 R M Williams, Biomed Appl Biotechnol., 1, 187 (1993) 74 C A Lipinski, J.Pharmacol Toxicol Methods, 44,235 (2001) 85 D Banner, J Ackermann, A Gast, K Gubernator, P Hadvary, K Hilpert, L Labler, K Mueller, G Schmid, T Tschopp, H van de Waterbeemd, and B Wirz, Perspect Med Chem., 27-43 (1993) 86 J Rahuel, V Rasetti, J Maibaum, H Rueger, R Goschke, N C Cohen, S Stutz, F Cumin, W Fuhrer, J M Wood, and M G Grutter, Chem Biol., 7,493 (2000) 87 M J Parry, A B Russell, and M Szelke in J Meienhofer, Ed., Chem Biol Pept., Proc Am Pept Symp., 3rd, Ann Arbor Science, Ann Arbor, MI, 1972, p 541 88 M Szelke, D M Jones, B Atrash, A Hallett, and B J Leckie in V J Hruby and D H Rich, Eds., Pept.: Struct Funct., Proc Am Pept Symp., 8th, Pierce Chemical Co., Rockford, 1983, p 579 89 D H Rich and E T Sun, Biochem Pharmacol., 29, 2205 (1980) 90 E M Gordon, J D Godfrey, J Pluscec, D Von Langen, and S Natarajan, Biochem Biophys Res Commun., 126,419 (1985) 91 D H Rich, J Med Chem., 28,263 (1985) 75 G Klebe, J Mol Med., 78,269 (2000) 76 H J Bohm and G Klebe, Angew Chem Intl Ed Engl., 35,2589 (1996) 77 L Pauling, Chem Eng News, 24,1375 (1946) 78 R Wolfenden, Annu Rev Biophys Bioeng., 5, 271 (1976) 79 R Wolfenden, Acc Chem Res., 5, 10 (1972) 92 D H Rich, J Green, M V Toth, G R Marshall, and S B H Kent, J Med Chem., 33, 1285 (1990) 80 M A Ondetti, B Rubin, and D: W Cushman, Science, 196, 441 (1977) 95 E D Thorsett and L H Latimer, Curr Opin Chem Biol., 4,377 (2000) 81 D Leung, G Abbenante, and D P Fairlie, J Med Chem., 43,305 (2000) 82 R E Babine and S L Bender, Chem Rev., 97, 1359 (1997) 96 M S Wolfe, J Med Chem., 44, 2039 (2001) 97 S Sinha, J P Anderson, R Barbour, G S Basi, R Caccavello, D Davis, M Doan, H F Dovey, N Frigon, J Hong, K Jacobson-Croak, N Jewett, P Keim, J Knops, I Lieberburg, M Power, H Tan, G Tatsuno, J Tung, D Schenk, P Seubert, S M Suomensaari, S Wang, D Walker, J Zhao, L McConlogue, and V John, Nature, 402, 537 (1999) 83 I Schechter and A Berger, Biochem Biophys Res Commun., 27, 157 (1967) 84 H P Marki, A Binggeli, B Bittner, V BohnerLang,V Breu, D Bur, P Coassolo, J P Clozel, A D'Arcy, H Doebeli, W Fischli, C Funk, J Foricher, T Giller, F Gruninger, A Guenzi, R Guller, T Hartung, G Hirth, C Jenny, M Kansy, U Klinkhammer, T Lave, B Lohri, F C Luft, E M Mervaala, D N Muller, M Muller, F Montavon, C Oefner, C Qiu, A Reichel, P Sanwald-Ducray, M Scalone, M Schleimer, R Schmid, H Stadler, A Treiber, 0.Valdenaire, E Vieira, P Waldmeier, R Wiegand-Chou, M Wilhelm, W Wostl, M Zell, and R Zell, Farmaco, 56, 21 (2001) 93 E J Lien, H Gao, and L L Lien, Prog Drug Res., 43,43 (1994) 94 F Lebon and M Ledecq, Curr Med Chem., 7, 455 (2000) 98 A K Ghosh, D Shin, D Downs, G Koelsch, X Lin, J Ermolieff, and J Tang, J Am Chem SOC., 122,3522 (2000) 99 L Hong, G Koelsch, X Lin, S Wu, S Terzyan, A K Ghosh, X C Zhang, and J Tang, Science, 290, 150 (2000) 100 A K Ghosh, G Biker, C Harwood, R Kawahama, D Shin, K A Hussain, L Hong, J A Loy, C Nguyen, G Koelsch, J Ermolieff, and J Tang, J Med Chem., 44, 2865 (2001) Peptidomimetics for Drug Design 101 H H Rotmensch, P H Vlasses, B N Swanson, J D Irvin, K E Harris, D G Merrill, and R K Ferguson, A.m J Cardiol., 53, 116 (1984) 102 A A Patchett, E Harris, E W Tristram, M J Wyvratt, M T W u , D Taub, E R Peterson, T J Ikeler, J ten Broeke, L G Payne, D L Ondeyka, E D Thorsett,W J Greenlee, N S Lohr, R D Hoffsommer, H Joshua, W V Ruyle, J W Rothrock, S D Aster, A L Maycock, F M Robinson, R Hirschmann, C S Sweet, E H Ulm, D M Gross, T C Vassil, and C A Stone, Nature, 288,280(1980) 103 B P Roques, M C Fournie-Zaluski, E Soroca, J M Lecomte, B Malfroy,C Llorens, and J C Schwartz, Nature, 288,286(1980) 104 B P Roques, E Lucas-Soroca, P Chaillet, J Costentin, and M C Fournie-Zaluski, Proc Natl Acad Sci USA, 80,3178(1983) 105 R Bouboutou, G Waksman, J Devin, M C Fournie-Zaluski, and B P Roques, Life Sci., 35,1023(1984) 106 J Bralet and J C Schwartz, Trends Pharm Sci., 22,106(2001) 107 S DeLombaert, R Chatelain, C A Fink, and A J Trapani, Curr Pharm Design, 2, 443 (1996) 108 J W.Skiles, L G Monovich, and A Y Jeng, Annu Rep Med Chem., 35,167(2000) 109 M R Michaelides and M L Curtin, Curr Pharm Design, 5,787(1999) 110 M Whittaker, C D Floyd, P Brown, and A J H Gearing, Chem Rev., 99,2735 (1999) 111 S Brown, M M Bernardo, Z H Li, L P Kotra,Y Tanaka, R Fridman, and S Mobashery, J Am Chem Soc., 122,6799(2000) 112 Kleifeld, L P Kotra, D C Gervasi, S Brown, M M Bernardo, R Fridman, S Mobashery, and I Sagi, J Biol Chem., 276,17125 (2001) 113 M L Moss, J M White, M H Lambert, and R C Andrews, Drug Discov Today, 6, 417 (2001) 114 R A Black, C T Rauch, C J Kozlosky, J J Peschon, J L Slack, M F Wolfson,B J Castner, K L Stocking, P Reddy, S Srinivasan, N Nelson, N Boiani, K A Schooley, M Gerhart, R Davis, J N Fitzner, R S Johnson, R J Paxton, C J March, and D P Cerretti, Nature, 385,729(1997) 115 F Anon, Expert Opin Ther Pat., 10, 1617 (2000) 116 H Hilpert, Tetrahedron, 57,7675(2001) 117 H Umezawa, T Aoyagi, H Suda, M Hamada, and T Takeuchi, J Antibiot (Tokyo), 29,97 (1976) 118 T Aoyagi, H Tobe, F Kojima, M Hamada, T Takeuchi, and H Umezawa, J Antibiot (Tokyo), 31,636(1978) 119 H Kim and W N Lipscomb, Proc Natl Acad Sci USA, 90,5006(1993) 120 W T Lowther, A M Orville, D T Madden, S J Lim, D H Rich, and B W Matthews, Biochemistry, 38,7678(1999) 121 For in-depth analysis, see D F Veber and S K Thompson, Curr Opin Drug Discov Dev., 3, 362 (2000); A Krantz, Bioorg Med Chem Lett., 2,1327(1992) 122 J Drenth, J N Jansonius, and B G Wolthers, J Mol Biol., 24,449(1967) 123 J Drenth, J.N Jansonius, R Koekoek, J Marrink, J Munnik, and B G Wolthers, J Mol Biol., 5,398(1962) 124 S Michaud and B J Gour, Expert Opin Ther Pat., 8,645(1998) 125 D S Yamashita and R A Dodds, Curr Pharm Des., 6,1(2000) 126 K Tezuka, Y Tezuka, A Maejima, T Sato, K Nemoto, H Kamioka, Y Hakeda, and M Kumegawa, J Biol Chem., 269,1106(1994) 127 D S Yamashita, W W Smith, B Zhao, C A Janson, T A Tomaszek, M J Bossard, M A Levy, H.-J Oh, T J Carr, S K Thompson, C F Ijames, S A Carr, M McQueney, K: J D'Alessio, B Y Amegadzie, C R Hanning, S Abdel-Meguid, R L DesJarlais, J G Gleason, andD F.Veber, J h Chem Soc., 119,11351 (1997) 128 R W Marquis, Y Ru, J Zeng, R E L Trout, S M LoCastro, A D Gribble, J Witherington, A E Fenwick, B Gamier, T Tomaszek, D Tew, M E Hemling, C J Quinn,W W Smith, B Zhao, M S McQueney, C A Janson, K D'Alessio, and D F Veber, J Med Chem., 44, 725(2001) 129 R W Marquis, Y Ru, S M LoCastro, J Zeng, D S Yamashita, H A Oh, K F Erhard, L D Davis, T A Tomaszek, D Tew, K Salyers, J Proksch, K Ward, B Smith, M Levy, M D Cummings, R C Haltiwanger, G Trescher, B Wang, M E Hemling, C J Quinn, H Y Cheng, F Lin, W W Smith, C A Janson, B Zhao, M S McQueney, K D'Alessio,C.-P Lee, A Marzulli, R A Dodds, S Blake, S.-M Hwang, I E James, C J Gress, B R Bradley, M W Lark, M Gowen, and D F Veber, J Med Chem., 44,1380(2001) References 130 R V Talanian, K D Brady, and V L Cryns, J Med Chem., 43, 3351 (2000) 131 D S Karanewsky, X Bai, S D Linton, J F Krebs, J Wu, B Pham, and K J Tomaselli, Bioorg Med Chem Lett., 8,2757 (1998) 132 A B Shahripour, M S Plummer, E A Lunney, T K Sawyer, C J Stankovic, M K Connolly, J R Rubin, N P C Walker, K D Brady, H J Allen, R V Talanian, W W Wong, and C Humblet, Bioorg Med Chem Lett., 11, 2779 (2001) 133 D Lee, S A Long, J L Adams, G Chan, K S Vaidya, T A Francis, K Kikly, J D Winkler, C.-M Sung, C Debouck, S Richardson, M A Levy, W E DeWolf Jr., P M Keller, T Tomaszek, M S Head, M D Ryan, R C Haltiwanger, P.-H Liang, C A Janson, P J McDevitt, K Johanson, N 0.Concha, W Chan, S S Abdel-Meguid, A M Badger, M W Lark, D P Nadeau, L J Suva, M Gowen, and M E Nuttall, J Biol Chem., 275, 16007 (2000) 134 D Lee, S A Long, J H Murray, J L Adams, M E Nuttall, D P Nadeau, K Kikly, J D Winkler, C.-M Sung, M D Ryan, M A Levy, P M Keller, and W E DeWolf Jr., J Med Chem., 44,2015 (2001) 135 R E Dolle, J Singh, J Rinker, D Hoyer, C V C Prasad, T L Graybill, J M Salvino, C T Helaszek, R E Miller, and M A Ator, J Med Chem., 37,3863 (1994) 136 B K Kay, A V Kurakin, and R Hyde-DeRuyscher, Drug Discov Today, 3,370 (1998) 137 F.Al-Obeidi, V J Hruby, and T K Sawyer, Mol Biotechnol., 9,205 (1998) 138 A E P Adang and P H H Hermkens, Curr Med Chem., 8,985 (2001) 139 K S Lam, M Lebl, and V Krchnak, Chem Rev., 97,411 (1997) 140 A C Good, S R Krystek, and J S Mason, Drug Discov Today., 5,61(2000) 141 S P Rohrer, E T Birzin, R T Mosley, S C Berk, S M Hutchins, D.-M Shen, Y Xiong, E C Hayes, R M Parmar, F Foor, S W Mitra, S J Degrado, M Shu, J M Klopp, S J Cai, A Blake, W W S Chan, A Pasternak, L Yang, A A Patchett, R G Smith, K T Chapman, and J M Schaeffer, Science, 282, 737 (1998) 142 R S Bohacek, C McMartin, and W C Guida, Med Res Rev., 16,3 (1996) 143 Y Kurogi and F Guner, Curr Med Chem., 8, 1035 (2001) 144 J S Mason, A C Good, and E J Martin, Curr Pharm Des., 7,567 (2001) 145 F Ooms, Curr Med Chem., 7,141 (2000) 146 J Wouters and F Ooms, Curr Pharm Des., 7, 529 (2001) 147 E Vieira, A Binggeli, V Breu, D Bur, W Fischli, R Guller, G Hirth, H P Marki, M Muller, C Oefner, M Scalone, H Stradler, M Wilhelm, and W Wostl, Bioorg Med Chem Lett., 9,1397 (1999) 148 G Guller, A Binggeli, V Breu, D Bur, W Fischli, G Hirth, C Jenny, M Kansay, F Montavon, M Muller, C Oefner, H Stradler, E Vieira, M Wilhelm, W Wostl, andH P Marki, Bioorg Med Chem Lett., 9, 1403 (1999) 149 C Oefner, A Binggeli, V Breu, D Bur, J.-P Clozel, A D'Arcy, A Dorn, W Fischli, F Gruninger, R Guller, G Hirth, H P Marki, S Mathews, M Muller, R G Ridler, H Stadler, E Vieira, M Wilhelm, F K Winklier, and W Wostl, Chem Biol., 6, 127 (1999) 150 M G Bursavich, C W West, and D H Rich, Org Lett., 3,2317 (2001) 151 A B Smith 111, R Hirschmann, A Pasternak, W Yao, P A Sprengeler, M K Holloway, L C Kuo, Z Chen, P L Darke, and W A Schleif, J Med Chem., 40,2440 (1997) 152 J D A Tyndall, R C Reid, D P Tyssen, D K Jardine, B Todd, M Passmore, D R March, L K Pattenden, D A Bergman, D Alewood, S.-H Hu, P F Alewood, C J Birch, J L Martin, and D P Fairlie, J Med Chem., 43,3495 (2000) 153 S E Hagen, J V N V Prasad, F E Boyer, J M Domagala, E L Ellsworth, C Gajda, H W Hamilton, L J Markoski, B A Steinbaugh, B D Tait, E A Lunney, P J Tummino, D Ferguson, D Hupe, C Nouhan, S J Gracheck, J M Saunders, and S VanderRoest, J Med Chem., 40,3707 (1997) 154 T M Judge, G Phillips, J K Morris, K D Lovasz, K R Romines, G P Luke, J Tulinsky, J M Tustin, R A Chrusciel, L A Dolak, S A Mizsak, W Watt, J Morris, S L V Velde, J W Strohbach, and R B Gammill, J Am Chem Soc., 119,3627 (1997) 155 G V De Lucca, S Erickson-Viitanen, and P Y S Lam, Drug Discov Today, 2, (1997) 156 W Schaal, A Karlsson, G Ahlsen, J Lindberg, H Andersson, U H Danielson, B Classon, T Unge, B Samuelsson, J Hulten, A Hallberg, and A Karlen, J Med Chem., 44, 155 (2001) 157 J P Vacca, Curr Opin Chem Biol., 4, 394 (2000) 158 P E J Sanderson, Med Res Rev., 19, 179 (1999) Peptidomimetics for Drug Design 159 P E J Sanderson and A M Naylor-Olsen, Curr Med Chem., 5,289(1998) 160 M R Wiley, N Y Chirgadze, D K Clawson, T J Craft, D S Gifford-Moore, N D Jones, J L Olkowaki, A L Schacht, L C Weir, and G F Smith, Bioorg Med Chem Lett.,5,2835 (1995) 161 S F Brady, K J Stauffer, W C Lumma, G M Smith, H G Ramjit, S D Lewis, B J Lucas, S J Gardell, E A Lyle, S D Appleby, J J Cook, M A Holahan, M T Stranieri, J J Lynch Jr., J H Lin, I W Chen, K Vastag, A M Naylor-Olsen, and J P Vacca, J Med Chem., 41,401(1998) 162 T J Tucker, W C Lumma, A M Naylor-Olsen, S D Lewis, R Lucas, R M Freidinger, A M Mulichak, Z Chen, and L C Kuo, J Med Chem., 40,830(1997) 163 R A Engh, H Brandstetter, G Sucher, A Eichinger, U Baumann, W Bode, R Huber, T Poll, R Rudolph, and W Von der Sad, Structure, 4,1353(1996) 164 T B Lu, B Tomczuk, R Bone, L Murphy, F R Salemme, and R M Soll, Bioorg Med Chem Lett.,10,83(2000) 165 T B Lu, R M Soll, C R Illig, R Bone, L Murphy, J Spurlino, F R Salemme, and B E Tomczuk, Bioorg Med Chem Lett., 10, 79 (2000) 166 N Y Chirgadze, D J Sall, V J Klimkowski, D K Clawson, S L Briggs, R Hermann, G F Smith, D S Gifford-Moore, and J.-P Wery, Protein Sci., 6,1412(1997) 167 D J Sall, J A Bastian, S L Briggs, J A Buben, N Y Chirgadze, D K Clawson, M L Denney, D D Giera, D S Gifford-Moore, R W Harper, K L Hauser, V J Klimkowski, T J Kohn, H.-S Lin, J R McCowan, A D Palkowitz, G F Smith, K Takeuchi, K J Thrasher, J M Tinsley, B G Utterback, S.C B Yan, and M Zhang, J Med Chem., 40, 3489(1997) 168 M F Malley, L Tabernero, C Y Chang, S L Ohringer, D G M Roberts, J Das, and J S Sack, Protein Sci., 5,221(1996) 169 N Y Chirgadze, D J Sall, S L Briggs, D K Clawson, M Zhang, G F Smith, and R W Schevitz, Protein Sci., 9, 29(2000) 170 U Obst, D W Banner, L Weber, and F Diederich, Chem Biol., 4,287(1997) 171 A von Matt, C Ehrhardt, P Burkhard, R Metternich, M Walkinshaw, and C Tapparelli, Bioorg Med Chem., 8,2291(2000) 172 U Baettig, L Brown, D Brundish, C Dell, A Furzer, S Garman, D Janus, P D Kane, G Smith, C V Walker, X L Cockcroft, J Ambler, A Mitchelson, M D Talbot, M Tweed, and N Wills, Bioorg Med Chem Lett., 10, 1563(2000) 173 R Rai, P A Sprengeler, K C Elrod, and W B Young, Curr Med Chem., 8,101(2001) 174 J M Fevig, D J Pinto, Q Han, M L Quan, J R Pruitt, I C Jacobson, R A Galemmo Jr., S Wang, M J Orwat, L L Bostrom, R M Knabb, P C Wong, P Y S Lam, and R R Wexler, Bioorg Med Chem Lett., 11, 641 (2001) 175 S I Klein, M Czekaj, C J Gardner, K R Guertin, D L Cheney, A P Spada, S A Bolton, K Brown, D Colussi, C L Heran, S R Morgan, R J Leadley, C T Dunwiddie, M H Perrone, and V Chu, J Med Chem., 41,437 (1998) 176 Y Gong, H W Pauls, A P Spada, M Czekaj, G Y Liang, V Chu, D J Colussi, K D Brown, and J B Gao, Bioorg Med Chem Lett., 10, 217(2000) 177 D K Herron, T Goodson, M R Wiley, L C Weir, J A Kyle, Y K Yee, A L Tebbe, J M Tinsley, D Mendel, J J Masters, J B Franciskovich, J S Sawyer, D W Beight, A M Ratz, G Milot, S E Hall, V J Klimkowski, J H Wikel, B J Eastwood, R D Towner, D S Gifford-Moore, T J Craft, and G F Smith, J Med Chem., 43,859(2000) 178 Y K Yee, A L Tebbe, J H Linebarger, D W Beight, T J Craft, D Gifford-Moore, T Goodson, D K Herron, V J Klimkowski, J A Kyle, J S Sawyer, G F Smith, J M Tinsley, R D Towner, L Weir, and M R Wiley, J Med Chem., 43,873 (2000) 179 M R Wiley, L C Weir, S Briggs, N A Bryan, J Buben, C Campbell, N Y Chirgadze, R C Conrad, T J Craft, J V Ficorilli, J B Franciskovich, L L Froelich, D S Gifford-Moore, T Goodson, D K Herron, V J Klimkowski, K D Kurz, J A Kyle, J J Masters, A M Ratz, G Milot, R T Shuman, T Smith, G F Smith, A L Tebbe, J M Tinsley, R D Towner, A Wilson, and Y K Yee, J Med Chem., 43,883(2000) 180 Z S Zhao, D Arnaiz, B Griedel, S Sakata, J L Dallas, M Whitlow, L Trinh, J Post, A Liang, M M Morrissey, and K J Shaw, Bioorg Med Chem Lett.,10,963(2000) 181 J M Smallheer, R E Olson, and R R Wexler, Annu Rep Med Chem., 35,103(2000) 182 W Wang, R.T Borchardt, and B Wang, Curr Med Chem., 7,437(2000) 183 E Ruoslahti and M D Pierschbacher, Science, 238,491(1987) References 184 D M Haverstick, J F Cowan, K M Yamada, and S A Santoro, Blood, 66,946 (1985) 185 F E Ali, D B Bennett, R R Calvo, J D Elliott, S M Hwang, T W Ku, M A Lago, A J Nichols, T T Romoff, D H Shah, J A Vasko, A S Wong, T Yellin, C K Yuan, and J M Samanen, J Med Chem., 37,769 (1994) 186 P L Barker, S Bullens, S Bunting, D J Burdick, K S Chan, T Deisher, C Eigenbrot, T R Gadek, R Gantzos, M T Lipari, C D Muir, M A Napier, R M Pitti, A Padua, C Quan, M Stanley, M Struble, J Y K Tom, and J P Burnier, J Med Chem., 35, 2040 (1992) 187 R S Mcdowell and T R Gadek, J.Am Chem Soc., 114,9245 (1992) 188 W E Bondinell, R M Keenan, W H Miller, F E Ali, A C Allen, C W De Brosse, D S Eggleston, K F Erhard, R C Haltiwangerc, W F Huffmana, S.-M Hwangd, D R Jakasa, P F Kosterf, T W Kua, C P Leee, A J Nicholsf, S T Rossa, J M Samanena, R E Valocikf, J A Vasko-Moserf, J W Venslavskya, A S Wongd, and C.-K Yuana, Bioorg Med Chem., 2,897 (1994) 189 B K Blackburn, A Lee, M Baier, B Kohl, A G Olivero, R Matamoros, K D Robarge, and R S McDowell, J Med Chem., 40, 717 (1997) 190 J D Prugh, R J Gould, R J Lynch, G X Zhang, J J Cook, M A Holahan, M T Stranieri, G R Sitko, S L Gaul, R A Bednar, B Bednar, and G D Hartman, Bioorg Med Chem Lett., 7,865 (1997) 191 N J Liverton, D J Armstrong, D A Claremon, D C Remy, and J J Baldwin, Bioorg Med Chem Lett., 8,483 (1998) 192 B C Askew, C J McIntyre, C A Hunt, D A Claremon, J J Baldwin, P S Anderson, R J Gould, R J Lynch, C C T Chang, J J Cook, J J Lynch, M A Holahan, G R Sitko, and M T Stranieri, Bioorg Med Chem Lett., 7, 1531 (1997) 193 E J Topol, T V Byzova, and E F Plow, Lancet, 353,227 (1999) 194 C B Xue, J Roderick, S Jackson, M Rafalski, A Rockwell, S Mousa, R E Olson, and W F DeGrado, Bioorg Med Chem., 5,693 (1997) 195 M J Fisher, B Gunn, C S Harms, A D Kline, J T Mullaney, A Nunes, R M Scarborough, A E Arfsten, M A Skelton, S L Um, B G Utterback, and J A Jakubowski, J.Med Chem., 40,2085 (1997) 196 G D Hartman, M E Duggan, W F Hoffman, R J Meissner, J J Perkins, A E Zartman, 197 198 199 200 201 202 203 204 205 206 207 208 209 210 A M Naylor-Olsen, J J Cook, J D Glass, R J Lynch, G Zhang, and R J Gould, Bioorg Med Chem Lett., 9,863 (1999) B Wang, W Wang, G P Camenisch, J Elmo, H Zhang, and R T Borchardt, Chem Pharm Bull (Tokyo), 47,90 (1999) M S Smyth, J Rose, M M Mehrotra, J Heath, G Ruhter, T Schotten, J Seroogy, D Volkots, A Pandey, and R M Scarborough, Bioorg Med Chem Lett., 11, 1289 (2001) M E Duggan, L T Duong, J E Fisher, T G Hamill, W F Hoffman, J R Huff, N C Ihle, C.-T Leu, R M Nagy, J J Perkins, S B Rod m , G Wesolowski, D B Whitman, A E Zartman, G A Rodan, and G D Hartman, J Med Chem., 43,3736 (2000) A Peyman, V Wehner, J Knolle, H U Stilz, G Breipohl, K.-H Scheunemann, D Carniato, J.-M Ruxer, J.-F Gourvest, T R Gadek, and S Bodary, Bioorg Med Chem Lett., 10, 179 (2000) A D Cox, Drugs, 61, 723 (2001) H Waldmann and M Thutewohl, Top Curr Chem., 211, 117 (2001) A Wittinghofer and H Waldmann, Angew Chem Int Ed., 39,4192 (2000) N E Kohl, C A Omer, M W Conner, N J Anthony, J P Davide, S J Desolms, E A Giuliani, R P Gomez, S L Graham, K Hamilton, L K Handt, G D Hartman, K S Koblan, A M Kral, P J Miller, S D Mosser, T J Oneill, E Rands, M D Schaber, J B Gibbs, and A Oliff, Nut Med., 1, 792 (1995) M Nigam, C.-M Seong, Y Qian, A D Hamilton, and S M Sebti, J.Biol Chem., 268,20695 (1993) J T Hunt, V G Lee, K Leftheris, B Seizinger, J Carboni, J Mabus, C Ricca, N Yan, and V Manne, J Med Chem., 39,353 (1996) C J Dinsmore, J M Bergman, D D Wei, C B Zartman, J P Davide, I B Greenberg, D M Liu, T J O'Neill, J B Gibbs, K S Koblan, N E Kohl, R B Lobell, I W Chen, D A McLoughlin, T V Olah, S L Graham, G D Hartman, and T M Williams, Bioorg Med Chem Lett., 11, 537 (2001) S B Long, P J Hancock, A M Kral, H W Hellinga, and L S Beese, Proc Natl Acad Sci USA, 98, 12948 (2001) G L James, J L Goldstein, M S Brown, T E Rawson, T C Somers, R S McDowell, C W Crowley, B K Lucas, A D Levinson, and J C Marsters Jr., Science, 260, 1937 (1993) J T Hunt, C Z Ding, R Batorsky, M Bednarz, R Bhide, Y Cho, S Chong, S Chao, J Peptidomimetics for Drug Design Gullo-Brown, P Guo, S H Kim, F Y F Lee, K Leftheris, A Miller, T Mitt, M Patel, B A Penhallow, C Ricca, W C Rose, R Schmidt, W A Slusarchyk, G Vite, and V Manne, J Med Chem., 43, 3587 (2000) 211 F G Njoroge, R J Doll, B Vibulbhan, C S Alvarez, W R Bihop, J Petrin, P Kirschmeier, N I Carruthers, J K Wong, M M Albanese, J J Piwinski, J Catino, V Girijavallabhan, and A K Ganguly, Bioorg Med Chem., 5, 101 (1997) 212 C L Strickland, P C Weber, W T Windsor, Z Wu, H V Le, M M Albanese, C S Alvarez, D Cesarz, J del Rosario, J Deskus, A K Mallams, F G Njoroge, J J Piwinski, S Remiszewski, R R Rossman, A G Taveras, B Vibulbhan, R J Doll, V M Girijavallabhan, and A K Ganguly, J Med Chem., 42,2125 (1999) 213 M Gurrath, Curr Med Chem., 8,1605 (2001) 214 F Yushiyasu, K Shoji, and N Kohei, US Patent 4,355,040, 1982 215 R R Smeby and S Fermandjian, Chem Biochem Amino Acids, Pept., Proteins, 1978 216 J V Duncia, A T Chiu, D J Carini, G B Gregory, A L Johnson, W A Price, G J Wells, P C Wong, J C Calabrese, and P B M W M Timmermans, J Med Chem., 33, 1312 (1990) 217 J Weinstock, R M Keenan, J Samanen, J Hempel, J A Finkelstein, R G Franz, D E Gaitanopoulos, G R Girard, J G Gleason, D T Hill, T M Morgan, C E Peishoff, N Aiyar, D P Brooks, T A Fredrickson, E H Ohlstein, R R Ruffolo, E J Stack, A C Sulpizio, E F Weidley, and R M Edwards, J Med Chem., 34, 1514 (1991) 218 S Perlman, H T Schambye, R A Rivero, W J Greenlee, S A Hjorth, and T W Schwartz, J Biol Chem., 270, 1493 (1995) 219 B Vianello, E Clauser, P Corvol, and C Monnot, Eur J Pharmacol., 347, 113 (1998) 220 C Swain and N M J Rupniak, Annu Rep Med Chem., 34, 51 (1999) 221 J A Lowe 111, Med Res Rev., 16, 527 (1996) 222 S McLean, Med Res Rev., 16,297 (1996) 223 D C Horwell, J A H Lainton, J A O'Neill, M C Pritchard, and J Raphy, Spec Publ R Soc Chem., 264,95 (2001) 224 V A Ashwood, M J Field, D C Horwell, C Julien-Larose, R A Lewthwaite, S McCleary, M C Pritchard, J Raphy, and L Singh, J Med Chem., 44,2276 (2001) 225 R M Snider, J W Constantine, J A Lowe 111, K P Longo, W S Lebel, H A Woody, S E Drozda, M C Desai, F J Vinick, R W Spencer, and H J Hess, Science, 251,435 (1991) 226 J A Lowe 111,S E Drozda, R M Snider, K P Longo, S H Zorn, J Morrone, E R Jackson, S McLean, D K Bryce, J Bordner, A Nagahisa, Y Kanai, Suga, and M Tsuchiya, J Med Chem., 35,2591 (1992) 227 T Harrison, A P Owens, B J Williams, C J Swain, A Williams, E J Carlson, W Rycroft, F D Tattersall, M A Cascieri, G G Chicchi, S Sadowski, N M J Rupniak, and R J Hargreaves, J Med Chem., 44,4296 (2001) 228 P C Ting, J F Lee, J C Anthes, N Y Shih, and J J Piwinski, Bioorg Med Chem Lett., 11, 491 (2001) 229 P C Ting, J F Lee, J C Anthes, N Y Shih, and J J Piwinski, Bioorg Med Chem Lett., 10,2333 (2000) 230 G A Reichard, Z T Ball, R Aslanian, J C Anthes, N Y Shih, and J J Piwinski, Bioorg Med Chem Lett., 10,2329 (2000) 231 F E Blaney, L F Raveglia, M Artico, S Cavagnera, C Dartois, C Farina, M Grugni, S Gagliardi, M A Luttmann, M Martinelli, G M M G Nadler, C Parini, P Petrillo, H M Sarau, M A Scheideler, D W P Hay, and G A M Giardina, J Med Chem., 44, 1675 (2001) 232 A W Stamford and E M Parker, Annu Rep Med Chem., 34,31(1999) 233 J Wright, Drug Discov Today, 2, 19 (1997) 234 K Rudolf, W Eberlein, W Engel, H A Wieland, K D Willim, M Entzeroth, W Wienen, A G Beck-Sickinger, and H N Doods, Eur J Pharmacol., 271, R11 (1994) 235 R E Malmstroem, A Modin, and J M Lundberg, Eur J Pharmacol., 305,145 (1996) 236 P A Hipskind, K L Lobb, J A Nixon, T C Britton, R F Bruns, J Catlow, D K Dieckman-McGinty, S L Gackenheimer, B D Gitter, S Iyengar, D A Schober, R M A Simmons, S Swanson, H Zarrinmayeh, D M Zimmerman, and D R Gehlert, J Med Chem., 40,3712 (1997) 237 H Zarrinmayeh, D M Zimmerman, B E Cantrell, D A Schober, R F Bruns, S L Gackenheimer, P L Ornstein, P A Hipskind, T C Britton, and D R Gehlert, Bioorg Med Chem Lett., 9, 647 (1999) 238 M H Norman, N Chen, Z D Chen, C Fotsch, C Hale, N H Han, R Hurt, T Jenkins, J Kincaid, L B Liu, Y L Lu, Moreno, V J Santora, J D Sonnenberg, and W Karbon, J Med Chem., 43,4288 (2000) References 239 Dellaz Uana, M Sadlo, M Gerrnain, M Feletou, S Chamorro, F Tisserand, C de Montrion, J F Boivin, J Duhault, J A Boutin, and N Levens, Int J Obes., 25,84 (2001) 240 H Rueeger, P Rigollier, Y Yamaguchi, T Schmidlin, W Schilling, L Criscione, S Whitebread, M Chiesi, M W Walker, D Dhanoa, I Islam, J Zhang, and C Gluchowski, Bioorg Med Chem Lett., 10, 1175 (2000) 241 R P Nargund, A A Patchett, M A Bach, M G Murphy, and R G Smith, J.Med Chem., 41,3103 (1998) 242 C Y Bowers, F A Momany, G A Reynolds, and A Hong, Endocrinology, 114,1537 (1984) 243 M H Chen, M G Steiner, A A Patchett, K Cheng, L T Wei, W W S Chan, B Butler, T M Jacks, and R G Smith, Bioorg Med Chem Lett., 6,2163 (1996) 244 A A Patchett, R P Nargund, J R Tata, M H Chen, K J Barakat, D B R Johnston, K Cheng, W W S Chan, B Butler, G Hickey, T Jacks, K Schleim, S S Pong, L Y P Chaung, H Y Chen, E Frazier, K H Leung, S H L Chiu, and R G Smith, Proc Natl Acad Sci USA, 92,7001 (1995) 245 B L Palucki, S D Feighner, S S Pong, K K McKee, D L Hreniuk, C Tan, A D Howard, L H Y Van der Ploeg, A A Patchett, and R P Nargund, Bioorg Med Chem Lett., 11, 1955 (2001) 246 T K Hansen, M Ankersen, B S Hansen, K Raun, K K Nielsen, J Lau, B Peschke, B F Lundt, H Thogersen, N L Johansen, K Madsen, and P H Andersen, J Med Chem., 41, 3705 (1998) 247 T K Hansen, M Ankersen, K Raun, and B S Hansen, Bioorg Med Chem Lett., 11, 1915 (2001) 248 P Huang, G H Loew, H Funamizu, M Mimura, N Ishiyama, M Hayashida, T Okuno, Shimada, A Okuyama, S Ikegami, J Nakano, and K Inoguchi, J.Med Chem., 44, 4082 (2001) 249 M L Webb and T D Meek, Med Res Rev., 17, 17 (1997) 250 A Ray, L G Hegde, A Chugh, and J B Gupta, Drug Discov Today, 5,455 (2000) 251 E H Ohlstein, P Nambi, S A Douglas, R M Edwards, M Gellai, A Lago, J D Leber, R D Cousins, A M Gao, J Frazee, C E Peishoff, J W Bean, D S Eggleston, N A Elshourbagy, C Kumar, J.A Lee, T L Yue, C Louden, D P Brooks, J Weinstock, G Feuerstein, G Poste, R R Ruffolo, J G Gleason, and J D Elliott, Proc Natl Acad Sci USA, 91,8052 (1994) 252 J D Elliott, M A Lago, R D Cousins, A M Gao, J D Leber, K F Erhard, P Nambi, N A Elshourbagy, C Kumar, J A Lee, J W Bean, C W Debrosse, D S Eggleston, D P Brooks, G Feuerstein, R R Ruffolo, J Weinstock, J G Gleason, C E Peishoff, and E H Ohlstein, J Med Chem., 37, 1553 (1994) 253 H Jae, M Winn, T W von Geldern, B K Sorensen, W J Chiou, B Nguyen, K C Marsh, and T J Opgenorth, J Med Chem., 44,3978 (2001) 254 S Anzali, W W K R Mederski, M Osswald, and D Dorsch, Bioorg Med Chem Lett., , 1 (1998) 255 W W K R Mederski, D Dorsch, M Osswald, S Anzali, M Christadler, C.J Schmitges, P Schelling, C Wilm, and M Fluck, Bioorg Med Chem Lett., 8, 1771 (1998) 256 W W K R Mederski, M Osswald, D Dorsch, S Anzali, M Christadler, C.-J Schmitges, and C Wilm, Bioorg Med Chem Lett., 8, 17 (1998) 257 P D Stein, J T Hunt, D M Floyd, S Moreland, K E J Dickinson, C Mitchell, E C K Liu, M L Webb, N Murugesan, J Dickey, D Mcmullen, R G Zhang, V G Lee, R Serafino, C Delaney, T R Schaeffer, and M Kozlowski, J Med Chem., 37, 329 (1994) 258 N Murugesan, Z X Gu, P D Stein, S Spergel, A Mathur, L Leith, E C K Liu, R A Zhang, E Bird, T Waldron, A Marino, R A Morrison, M L Webb, S Moreland, and J C Barrish, J Med Chem., 43, 3111 (2000) 259 C D Wu, E R Decker, N Blok, J Li, A R Bourgoyne, H Bui, K M Keller, V Knowles, W Li, F D Stavros, G W Holland, T A Brock, and R A F Dixon, J Med Chem., 44, 1211 (2001) CHAPTER SIXTEEN Analog Design JOSEPH G CANNON The University of Iowa Iowa City, Iowa Contents Introduction, 688 Bioisosteric Replacement and Nonisosteric Bioanalogs (Nonclassical Bioisosteres), 689 Rigid or Semirigid (Conformationally Restricted) Analogs, 694 Homologation of Alkyl Chain or Alteration of Chain Branching; Changes in Ring Size; RingPosition Isomers; and Substitution of an Aromatic Ring for a Saturated One, or the Converse, 699 Alteration of Stereochemistry and Design of Stereoisomers and Geometric Isomers, 704 Fragments of the Lead Molecule, 707 Variation in Interatomic Distances, 710 Burger's Medicinal Chemistry and Drug Discovery Sixth Edition, Volume 1: Drug Discovery Edited by Donald J Abraham ISBN 0-471-27090-3 O 2003 John Wiley & Sons, Inc Analog Design INTRODUCTION This chapter is limited to nonprotein therapeutic candidates The subject of peptide analogs and peptidomimetic agents merits separate consideration Contemporary search for new drugs makes extensive use of robotic techniques of combinatorial chemistry and high throughput synthesis, whereby huge numbers of compounds can be prepared for high throughput screening However, this nonselective synthetic method based on a random screening philosophy should not replace the strategy of analog design, but rather it should be considered as a useful prelude to analog design In any strategy aimed at designing new drug molecules or analogs of known biologically active compounds, there are no absolute guidelines or rules for procedure; the knowledge, imagination, and intuition of the medicinal chemist are the most important contributors to success Analog design is as much an art as it is a science The concept of analog design presupposes that a lead has been discovered; that is, a chemical compound has been identified that possesses some desirable pharmacological property The search for and identification of leads is a challenge and is a separate topic It is sufficient for the present discussion to note that lead compounds are frequently identified as endogenous participants (hormones, neurotransmitters, second messengers, or enzyme cofactors) in the body's biochemistry and physiology, or a lead may result from routine, random biological screening of natural products or of synthetic molecules that were created for purposes other than for use as drugs Analog design is most fruitful in the study of pharmacologically active molecules that are structurally specific: their biological activity depends on the nature and the details of their chemical structure (including stereochemistry) Hence, a seemingly minor modification of the molecule may result in a profound change in the pharmacological response (increase, diminish, completely destroy, or alter the nature of the response) In pursuing analog design and synthesis, it must be recognized that the newly created analogs are chemical entities different from the lead compound It is not possible to retain all and exactly the same solubility and solvent partition characteristics, chemical reactivity and stability, acid or base strength, and/or in vivo metabolism properties of the lead compound Thus, although the new analog may demonstrate pharmacological similarities to the lead compound, it is not likely to be identical to it, either chemically or biologically, nor will its similarities and differences always be predictable The goal of analog design is twofold: (1)to modify the chemical structure of the lead compound to retain or to reinforce the desirable pharmacologic effect while minimizing unwanted pharmacological (e.g., toxicity, side effects, or undesired routes of and/or unacceptable rates of metabolism) and physical and chemical properties (e.g., poor solubility and solvent partition characteristics or chemical instability), which may result in a superior therapeutic agent; and (2) to use target analogs as pharmacological probes (i.e., tools used for the study of fundamental pharmacological and physiological phenomena) to gain better insight into the pharmacology of the lead molecule and perhaps to reveal new knowledge of basic biology Studies of analog structure-activity relationships may increase the medicinal chemist's ability to predict optimum chemical structural parameters for a given pharmacological action Analog design is greatly facilitated if the medicinal chemist can initially define the pharmacophore of the lead compound: that combination of atoms within the molecule that is responsible for eliciting the desired pharmacologic effect Analog design may be directed toward maintaining this combination of atoms intact in a newly designed molecule or toward a carefully planned, systematic modification of the pharmacophore If the medicinal chemist is uncertain about the structural features of the pharmacophoric portion of the molecule, a prime initial goal of analog design should be to define the pharmacophore The medicinal chemist should address the fol- Bioisosteric Replacement and Nonisosteric Bioanalogs (Nonclassical Bioisosteres) lowing questions: What change(s)can be made in the lead molecule that permit(s) retention or reinforcement of pharmacological action? and What change(s) can be made in the molecule that diminish, destroy, or qualitatively change the basic pharmacologic action? The ideal program of analog design should involve asingle structural change in the lead molecule with each new compound designed and synthesized An analog in which multiple changes in the structure of the lead molecule have been made simultaneously may occasionally reveal highly desirable pharmacologic effects However, relatively little useful structure-activity information will be gained from such a molecule It cannot be readily determined which change (or combination of changes) was responsible for the change in the pharmacological effect On a practical basis, it is frequently chemically impossible to effect only one discrete change in the lead molecule; one simple molecular structural alteration can influence many structural and chemical parameters Nonetheless, the medicinal chemist should be cognizant of the disadvantages inherent in "shotgun" (nonsystematic, multiparametric) modification of lead molecules In analog design, molecular modification of the lead compound can involve one or more of the following strategies: Bioisosteric replacement Design of rigid analogs Homologation of alkyl chain(s) or alteration of chain branching, design of aromatic ring-position isomers, alteration of ring size, and substitution of an aromatic ring for a saturated one, or the converse Alteration of stereochemistry, or design of geometric isomers or stereoisomers Design of fragments of the lead molecule that contain the pharmacophoric group (bond disconnection) Alteration of interatomic distances within the pharmacophoric group or in other parts of the molecule None of these strategies is inherently preferable to the others; all merit the medicinal chemist's attention and consideration Appli- 689 cation of a combination of these strategies to the lead molecule may be advantageous Considering the possible permutations and combinations of these changes that are possible within a single lead molecule, it is obvious that the number of analogs that can be designed from a lead molecule is potentially extremely large Some structural changes that might be proposed are chemically impracticable (e.g., the molecule is incapable of existence) or the proposed analog may represent an overwhelmingly formidable synthetic challenge These negative factors will diminish the population of possible analogs to be considered for synthesis; nevertheless, the medicinal chemist will always be confronted with a multitude of possible target molecules Rational decisions must be made concerning which compounds should be synthesized, and synthetic priorities must be established for target compounds All other factors being equal, the medicinal chemist should synthesize the less-challenging compounds first Beyond this truism, the medicinal chemist's best resources are intuition and imagination Selection and application of specific molecular modification strategies depend on the chemical structure of the lead compound and, to a certain extent, on the pharmacological action to be studied All of the strategies of analog design as well as subsequent decisions concerning target compounds to be synthesized can be facilitated by the use of computational chemistry (computer-assisted molecular modeling) techniques These may give the medicinal chemist further insights into structural, stereochemical, and electronic implications of the proposed molecular modification BlOlSOSTERlC REPLACEMENT A N D NONISOSTERIC BIOANALOCS (NONCLASSICAL BIOISOSTERES) The concept of bioisosterism derives from Langmuir's (1)observation that certain physical properties of chemically different substances (e.g., carbon monoxide and nitrogen, ketene and diazomethane) are strikingly similar These similarities were rationalized on the basis that carbon monoxide and nitrogen both have 14 orbital electrons and, similarly, Analog Design 690 Table 16.1 Bioisosteric Atoms and Groups Univalent -F -OH -NH2 -CH3 -C1 S H -I -Br Bivalent -0- 4- S e Tervalent -N= -CH= -pH2 t-C4H9 i-C3H7 -CH,- -NH- -P= -As= Quadrivalent -C- S i Ring equivalents -CH=CH- -S-(e.g., benzene, thiophene) H =N-(e.g., benzene, pyridine) -0- S - -CH2- -NH- diazomethane and ketene both have 22 orbital electrons Medicinal chemists have expanded and adapted the original concept to the analysis of biological activity The following definition has been proposed: "Bioisosteres are groups or molecules which have chemical and physical properties producing broadly similar biological properties" (2) This definition might be modified to include the concept that bioisosteres may produce opposite biological effects, and these effects are frequently a reflection of some action on the same biological process or at the same receptor site Bioisosteric similarity of molecules is commonly assigned on the basis of the number of valence electrons of an atom or a group of atoms rather than on the total number of orbital electrons, as was originally specified by Langmuir In a remarkable number of instances, compounds result that have similar (or even diametrically opposite) pharmacological effects compared with those of the parent compound Categories of classic bioisosteres have been described (2) (Table 16.1) A more recent comprehensive review of bioisosterism appeared in 1996 (3) In a short communication, Burger (4) discussed and provided valuable insights into isosterism and bioanalogy in drug design Many compounds have been identified that comply with the "biology" aspect of the bioisostere concept but that not fit the strict chemical (steric and electronic) definition of bioisosteres Floersheim et al (5) proposed that such compounds be designated as nonisosteric bioanalogs, replacing the older term, "nonclassical bioisosteres." However, most of the contemporary literature retains the nonclassical bioisostere terminology Table 16.2 lists representative nonclassical bioisosteres Dihydromuscimol(1) and thiomuscimol(2) are cyclic analogs of y-aminobutyric acid (GABA) (31, in which the C=N moiety of the - heterocyclic ring is considered to be bioisosteric with the of GABA The -S- moiety of thiomuscimol is bioisosteric with the ring -0- of dihydromuscimol Both (1)and (2) are highly potent agonists at GABA, receptors, as determined in an electrophoresis-based assay (6) Because of its bioisosteric similarity to the normal physiological substrate L-dopa (4), L-mimosine (5) inhibits catechol oxidation by the enzyme tyrosinase (7) These compounds exemplify a situation in which bioisosteres display opposite pharmacologic effects at the same receptor The sulfonium bioisostere (6) of N,N-dimethyldopamine (7) retains the dopaminergic agonist effect displayed by (7) (8) The fact that (6) bears a permanent unit positive charge was invoked in support of the hypoth- Table 16.2 Nonclassical Bioisosteres Carbonyl group Carboxylic acid group H -NHCN -CH(CN)2 Catechol H X=O,NR Halogen X CF3 CN N(CN)2 C(CN), Thioether Thiourea NO2 Hydrogen H F R NR3 Analog Design use of the antidepressant dibenzazepine derivative irnipramine (8) as the lead The structural similarity between imipramine and the phenothiazine antipsychotics [typified by chlorpromazine (9)] is apparent Although these two cH2 I H2N- C-H I COOH (4) CH~ I H2N-C-H I COOH bioisosteric molecules have different pharmacological properties and therapeutic uses and likely have different mechanisms and sites of action in the central nervous system (91, they share the property of being psychotropic' agents They illustrate the observation that bioisosteric manipulation of a molecule may change its mode of action In the antidepressant dibenzocycloheptene derivative amitriptyline (lo),the ring nitrogen of imipramine is esis that /3-phenethylamines such as (7)interact with dopamine receptors in their protonated (cationic) form Bioisosteric replacement strategy has been fruitful in design of psychoactive agents, by replaced by an exocyclic olefin moiety Demexiptiline (1l),doxepin (121, and dothiepin (13) represent other bioisosteric modifications of imipramine that possess antidepressant ac- uncertain Apparently, attempts were made (13) to isolate the E- and2-isomers of all of the compounds prepared in the series studied, but no information was provided about the stereochemistry of the dothiepin material used in the pharmacological studies Replacement of the entire indole ring system of melatonin (14) by a naphthalene ring (15)permitted retention of binding affinity in an ovine pars tuberalis membrane assay (14) From a study (15) of a series of muscarinic M, agonists derived from the structure of arecoline (16) and typified by (171, it was con- tivity (10) Variations in the precise nature of psychotropic effects manifested by compounds (8-13)may be ascribed to the marked changes in orientation in space of the two benzene ring components of the tricyclic portion of these molecules, imposed on them by the isosteric moieties (-CH=CH-, -CH,-+Hz-, S-, -CH20-, CH2S-) The 2-isomer of oxepin is a more potent antidepressant than the E-isomer, but the drug is marketed as a mixture of isomers (11).Doxepin is also a potent antagonist at histamine H, receptors The 2-isomer is somewhat more potent than the E-isomer against histamine in the guinea pig ileum (12) The identity of the geometric isomer of dothiepin (13) used in pharmacological testing is Analog Design cluded that the 2-N-methoxyimidoyl nitrile group serves as a stable methyl ester bioisostere The 2-isomer has an l&fold higher affinity than its E-isomer for the rat cerebral cortex tissue used in the binding studies Replacement of the methyl ester moiety of the muscarinic partial agonist arecoline (16) by the putative nonclassical bioisosteric 1,2,4oxadiazole ring system (181, where R = un- branched C,-C, alkyl) permits retention of muscarinic agonism (16) The 1,2,4-oxadiazole ring system of quisqualic acid (19),an agonist at a subpopulation COOH I 165 nM) with respect to potency and effect on two arsenic-resistant strains of the organism Although the strategy of bioisosteric replacement may be a powerful and highly productive tool in analog design, Thornber (2) has emphasized that fundamental chemical and physical chemical changes can be expected to result from these molecular modifications, which may in themselves profoundly affect the pharmacological action of the resulting molecules Contributing factors include change in the size of the atom or group introduced, which may affect the overall shape and size of the molecule; changes in bond angles; change in partition coefficient; change in the pK, of the molecule; alteration of chemical reactivity and chemical stability of the molecule, with accompanying qualitative and quantitative 4teration of in vivo metabolism of the molecule; and change in hydrogen-bonding capacity The chemical and biological results and pharmacological significance of many of these factors are unpredictable and must be determined experimentally R I G I D O R SEMIRIGID (CONFORMATIONALLY RESTRICTED) ANALOGS of glutamate receptors (171, can be considered to be a nonclassical bioisostere of the corresponding carboxyl group of glutamic acid (20) Compounds (21-23) illustrate further examples of nonclassical bioisosteres Compound (21) was reported to display antitrypanosomal activity (18).The analogs (22) and (23) also displayed antitrypanosomal activity (19) Compound (22) demonstrated the most impressive activity (IC,, values of 40 and Imposition of some degree of molecular rigidity on a flexible organic molecule (e.g., by incorporation of elements of the flexible molecule into a rigid ring system or by introduction of a carbon-carbon double or triple bond) may result in potent, biologically active agents that show a higher degree of specificity of pharmacologic effect There are possible advantages to this technique (20): the key functional groups are held in one steric disposition or, in Rigid or Semirigid (Conformationally Restricted) Analogs the case of a semirigid structure, the key functional groups are constrained to a limited range of steric dispositions and interatomic distances By the rigid analog strategy, it is possible to approximate "frozen" conforrnations of a flexible lead molecule that, if an enhanced pharmacological effect results, may assist in defining and understanding structureactivity parameters, including the threedimensional geometry of the pharmacophore These data may be useful in constructing a model of the topography of the receptor Computational chemistry strategies can be a valuable tool in designing rigid analogs The semirigid tetralin congeners (24) and (25) of Nfl-dimethyldoparnine (7) represent the two rotameric conformational extremes of the spatial relationship of the aromatic ring of dopamine to the ethylamine side chain when the ring and the side chain are coplanar Compounds (24) and (25) display effects at different subpopulations of dopamine receptors (211, which have been proposed to reflect dif- 695 ferent conformations assumed by the flexible dopamine molecule at its various in vivo sites of action Restriction of conformational freedom of the acyl moiety in 4-DAMP (261, an antimus- carinic compound displaying higher &nity at ileal M, acetylcholine receptors than at atrial M, receptors) was imposed by the structure of the spiro-compound (27) (22) Analog Design Spiro-DAMP (27) was slightly more potent at M, muscarinic receptors than at M, receptors It was proposed that the geometry of the spiro-molecule might reflect the receptorbound conformation of 4-DAMP (26);this conformation differs from that observed in the crystal structure of 4-DAMP Conformational restriction was introduced into the side chain of a nonclassical serotonin bioisostere (28), a selective 5-HT,, and 5-HT,, receptor agonist) by its incorporation into a fused six-membered ring (29) (23) The conformational restrictions imposed on the indole-3-ethylamine moiety permitted retention of affinity for the 5-HT,, receptor but it diminished affinity for the 5-HT,, receptor by a factor of 1000 In two functional assays, (29) exhibited potency equal to or marginally greater than that of serotonin Com- pound (29) was described as a partial agonist It was concluded that the conformation of the indole-3-ethylamine portion of the fused system (29) reflects the conformation assumed by the flexible system (28) when it binds to the 5-HT,, receptor Imposition of rigidity into the piperidine ring of the opioid analgesic meperidine (30)by introduction of a methylene bridge between carbons and resulted in epimers (31)and (32), representing "frozen" conformations of meperidine (24) The exo-phenyl isomer (32) was six times as potent as the endo-phenyl isomer (31), and it was twice as potent as meperidine itself in a benzoquinone-induced writhing assay for analgesic effect Rigid analogs (331, (341, and (35)of phencyclidine (36) possess a rigid carbocyclic struc- Rigid or Semirigid (Conformationally Restricted) Analogs 697 Incorporation of the choline portion of acetylcholine (37) into a cyclopropane ring system resulted in cis- and trans-1,Bdisubstituted molecules, (38) and (391, in which the ture and an attached piperidine ring that is free to rotate All three rigid analogs showed low to no affinity for the PCP receptor, but they had good affinity in a a-receptor-binding assay (25) These binding data were proposed to be useful in defining a model for the a-receptor pharmacophore This study also provided additional evidence that the a-receptor is independent of the PCP-binding site (cf Ref 26 and references therein) acetylcholine molecule is locked into folded ("cisoid") and extended ("transoid") conformations The (lS),(2S)-(+)-trans-isomer (39) was somewhat more potent than acetylcholine itself in tissue and whole-animal assays for muscarinic agonism (27) and it was an excellent substrate for acetylcholinesterase The (1R),(2R)enantiomer of (39) was exponentially less potent than its (1S),(2S)enantiomer in the assays cited, but it was a good substrate for acetylcholinesterase The (2)-cis-isomer (38) was almost inert at nicotinic and muscarinic receptors and it was a poor substrate for acetylcholinesterase These data were taken as evidence that the flexible acetylcholine molecule interacts with muscarinic receptors in an extended geometry of the chain of atoms (28) When this semirigid analog strategy was applied to a cyclobutane ring system (compound 40), there was a marked loss of pharmacologic effect (29).This result is enigmatic; differences in interatomic distances and bond angles in the pharmacophoric moiety as well Analog Design as differences in extraneous molecular bulk seem insufficient to account for the dramatic difference in pharmacological potencies between the three- and the four-membered ring systems The cyclopropane ring was employed to impart a degree of rigidity to the side chain of dopamine (structures 41 and 42) (30) corpus striatum tissue, but the binding affinities for (43a) and (43b) are much less than that of dopamine (32) Racemic trans-(43a) was more potent than the trans-primary amine (43b), but it was still much less potent, than dopamine The racemic cis-isomer of (44b) demonstrated very low aMinity for the receptor A P-phenethylamine moiety was incorporated into the trans-decalin ring system (45) Neither isomer displayed effects at dopamine receptors, but both were a-adrenoceptor agonists, with the (+)-trans-isomer(41) being approximately five times more potent than the (+)-cis-isomer(42) It was suggested (31)that these findings may contribute to determining the preferred conformation of P-phenethylamines at the a-adrenoceptor The racemic trans-cyclobutane congeners (43a) and (43b) are more potent than their racemic cis-isomers (44a) and (44b) in binding studies on rat and the racemic modifications of all four possible isomers were prepared as "frozen" analogs of possibly significant conformations of the flexible norepinephrine molecule (33) All four compounds displayed approximately equal (extremely low) potency This result il- Homologation of Alkyl Chain or Alteration of Chain Branching lustrates that the achievement of conformational integrity by incorporation of a flexible pharmacophore into a bulky, complex molecule may be at the expense of biological activity Rigidity was introduced into the glutamic acid moiety in a series of bioisosteric congeners (46-48) (34) These systems showed po- tent agonist activity at subpopulations of metabotropic glutamate receptors The geometry of these congeners led to the conclusion that glutamic acid itself interacts with the metabotropic glutamate receptors in a fully extended conformation The rotational orientation of the ester moieties of the myoneural blocking agent succinyldicholine (49) was restricted by introduction of a double bond into the succinic acid portion (501, (51) (35) The E-fumarate ester (51) was approximately one-half as potent as the flexible succinate ester (49), whereas the 2-maleate ester (50) was 1/40 as potent as the 699 succinate These results led to the conclusion that the molecular shape of the E-ester (51) more closely approximates that assumed by succinylcholine when it interacts with myoneural nicotinic receptors Restricted rotation was also introduced into the succinic acid moiety of succinyldicholine by preparation of the choline esters of cisand trans-cyclopropane-1,2-dicarboxylic acids (52) and (53)(36,371 Myoneural blocking activity was assessed in dogs (37) and-cats (36) and, as indicated above for the E- and Z-olefinic esters (51) and (50), the extended transisomer (53) demonstrated much greater potency and a longer duration of action than those of the cis-isomer (52) The cyclobutane congeners (54) and (55)presented unexpected results that are difficult to rationalize: the cisisomer (54) was much less potent than the trans-isomer (55)in a cat assay for myoneural blockade, but it presented a decidedly longer duration of action than that of the trans-isomer (36) H O M O L O G A T I O N O F ALKYL C H A I N O R ALTERATION O F C H A I N BRANCHING; CHANGES IN RlNG SIZE; RING-POSITION ISOMERS; A N D SUBSTITUTION O F A N AROMATIC RlNG FOR A SATURATED ONE, O R THE CONVERSE Change in size or branching of an alkyl chain on a bioactive molecule may have profound (and sometimes unpredictable) effects on physical and pharmacological properties Alteration of the size andlor shape of an alkyl substituent can affect the conformational preference of a flexible molecule and may alter the spatial relationships of the components of the pharmacophore, which may be reflected in the ability of the molecule to achieve complementarity with its receptor or with the catalytic surface of a metabolizing enzyme The alkyl group itself may represent a binding site with the receptor (through hydrophobic interactions), and alteration of the chain may alter its binding capacity Position isomers of substituents (even alkyl groups) on an aromatic ring may possess different pharmacological properties In addition to their ability to affect electron distribution over an aromatic ring sys- Analog Design tem, position isomers may differ in their complementarity to receptors, and the position of a substituent on a ring may influence the spatial occupancy of the ring system with respect to the remainder of a conformationally variable molecule What has sometimes been trivialized as "methyl group roulette" may indeed be an important parameter in the design of analogs Homologation of the N-alkyl chain in norapomorphine (56) from methyl (57) to npropyl(59) produced incremental increases in emetic response in dogs and in stereotypy responses in rodents (38,39) rats It seems likely that the enhanced dopaminergic agonist effects conferred by N-ethyl and N-n-propyl groups on aporphine and p-phenethylamine-derived molecules are not related merely to enhanced lipophilic character or to partitioning phenomena, but rather to the likelihood that the two- and three-carbon chains have a positive affinity for subsites on certain dopamine receptors It may be speculated that these receptor subsites not accommodate longer alkyl chains (e.g., n-butyl or n-pentyl) However, different assays fordopaminergic stimulant effects and different animal species were used in refs 41,42, and 43, and care must be exercised in drawing firm structure-activity relationship conclusions based on these data The alkyl linker between the two heterocyclic ring systems in structure (65) was modi- he next member of the series,the n-butyl homolog (601, demonstrated a tremendous loss in potency and activity compared to that of the lower homologs (39) Studies of NJVdialkyl dopamines (61-64) revealed that some (66) linker = combinations of alkyl groups may impart a high degree of dopamine agonist effects (40) NJV-dimethyldopamine (61) is extremely potent in assays for dopaminergicagonism (pigeon pecking, emesis in dogs, and inhibition of cat cardioaccelerator nerve), as is NJV-di-npropyldopamine (62) (41) N-n-Propyl-N-nbutyldopamine (63) is potent in behavioral assays in nigra-lesioned rats (42) However, NJV-di-n-butyldopamine(64)is virtually inert in these assays (41, 42) N,N-di-n-Pentyldopamine was reported (43) to be inert in a caudectomized mouse behavioral assay and in a rotatory behavioral assay in nigra-lesioned (67) linker = (68) linker = Y =H CH3 Y=H Y =H Y =H (69) linker = CH3 fied in studies of the ability of analogs to bind to the cholecystokinin-B receptor (44) When this linking group was -CH2 CH2-, the compound (structure 66) was extremely potent in radioligand displacement assays on mouse brain membranes Introduction of carbon-car- Analog Design bon unsaturation (E-olefin) into the linker (structure 67) resulted in a 16-fold decrease in binding ability; this suggests that conformational restriction and limitation of molecular flexibility have deleterious effects on biological activity However, no data were reported on the 2-isomer of this olefinic molecule, so that caution should be exercised in drawing conclusions Introduction of a bromine substituent (65, Y = Br) into (66) produced a threefold increase in potency, whereas the same structural modification of the olefin (67) resulted in a threefold decrease in potency Branching the linker chain with a methyl group adjacent to the quinazolinone ring (68) resulted in a 350-fold decrease in affinity However, chain branching with a methyl group in the alternate position on the ethylene chain produced compound (69), whose receptor affhity was of the same order of magnitude as the extremely potent lead compound (66) The exponential difference in receptor-binding ability exhibited by the two isomeric branched-chain linker compounds (68) and (69) was ascribed to unfavorable steric interactions between the receptor and the linker methyl group of (68) (44) This conclusion may be compromised by the fact that both (68) and (69) were evaluated as their racemates A study (45) of 2-(phosphonomethoxy)ethylguanidines (70-73) as antiviral (herpes and (70)R=R'=H (71) R = H ; R'=CH3 (72) R = C H ; R ' = H (73) R = R' = H; R = gem-dimethyl HIV) agents revealed that branching of the ethylene chain by introduction of a methyl group at the 1'-position (as in racemic 72) diminished antiviral activity 25-fold and dimin- ished toxicity 16-fold compared to that of the nonmethylated system (70) In contrast, (R)-(71), the 2'-methyl congener, exhibited only a fivefold decrease in antiviral potency compared to that of compound (70),but it also exhibited a 30-fold lessening of toxicity, to produce a substantial increase in therapeutic index over that of (70) The (S)(71) enantiomer was somewhat less potent than its (R)-enantiomer The gem-dimethyl congener (73) was also somewhat less potent than the (R)-2'-monomethyl compound (71) and it was markedly more toxic The (S)-2'methyl stereoisomer of (71) exhibited a decidedly lower therapeutic index than that of its (R)-enantiomer Closely related to alteration of chain length andlor chain branching is alteration of ring size Compound (74) showed nano- molar-level activity as an inhibitor of 5-lipoxygenase (46) The size of the oxygen-containing ring as well as the position of the oxygen member with respect to the methoxy and aryl substituents was varied The (seven-membered) oxepane ring derivative (79) and the (six-membered) tetrahydropyran ring derivative (78) showed two- to 10-fold enhanced potency over that of the tetrahydrofuran lead compound (74) The other analogs shown demonstrated much weaker enzyme inhibitory activity In a series of spiro-tetraoxacycloalkanes (go), with varying heterocyclic ring sizes, it was found that the compound where n = demonstrated marked antimalarial activity against P bergei and P falciparum, and showed low toxicity (47) The analog in which I Hotnologation of AIkyl Chain w Ateration of Chain Branching n = showed strong activity against P falciparum but it was unimpressive in the P bergei assay In a series of arylsulfonamidophenethanolamines (81) (48), derivatives bearing the sul- fonamido group meta to the ethanolamine side chain displayed properties of a p-adrenoceptor partial agonist, whereas 19 compounds bearing the sulfonamido group in thepara position were p-antagonists Changing the positions of attachment of the two benzene rings linking the quinolinium moieties of the calcium-activated potassium channel blocker (82) reduced activity 10- to 60-fold in a rat superior cervical ganglion assay (49) Other structural variations studied included benzene ring A meta-substituted and benzene ring L meta-substituted; benzene ring A meta-substituted and benzene ring L para-substituted; and benzene ring A parasubstituted and benzene ring L para-substituted All of these variations were much less potent than those of (82) The phenolic group of serotonin (83) was incorporated into a pyran ring (84) (50), thus also introducing an alkyl substituent at position of the indole ring system This tricyclic analog (84) lacked serotoninlike affinity for 5-HT, receptors, but it demonstrated high and selective affinity for 5-HT, receptors Like serotonin, it stimulated phosphatidyl inositol turnover in rat brain slices The low affinity for 5-HT, receptors was rationalized, in part, on the basis of steric interference between the dihydropyran ring and the aminoethyl side chain, which inhibits the tryptamine system from assuming the folded ergotlike conformation, as illustrated in (851, which probably approximates the conforrnation of serotonin at 5-HT, receptors The methyl ether of serotonin exhibits approximately the same affinity for 5-HT,, sites as does serotonin (51) The methyl ether also has marked affinity for 5-HT,, and 5-HT,, receptors, but it has diminished affinity (compared with serotonin) at 5-HT,, receptors It was Analog Design In a study of anticonvulsant agents, the (S)-benzene ring analog (90) was somewhat more potent in a mouse assay than was the (S)-cyclohexane analog (91) (56) There was suggested (50) that the high af'finity for the 5-HT, receptor exhibited by such compounds as (84) demonstrates that the C-5 hydroxyl group of serotonin can function as a hydrogenbond acceptor at the receptor Replacement of the benzene ring of the potent indirect acting central noradrenergic stimulant methamphetamine (86) by a cyclohexane ring (compound 87) results in some only a slight difference in potency between (R)- and (S)-(90) The (R)-enantiomer of (91) was not reported ALTERATION O F STEREOCHEMISTRY A N D D E S I G N O F STEREOISOMERS A N D G E O M E T R I C ISOMERS loss of pressor effect, but the drug, like amphetamine, has been used as a nasal decongestant, and it has CNS-mediated anorexigenic effect (52,53) It is said to have somewhat less central stimulant action than the corresponding aromatic ring derivatives (54a-d) The benzene (88) and cyclohexane (89) congeners have almost identical effects in blocking bronchoconstriction produced by histamine, serotonin, or acetylcholine in the guinea pig in vivo (55).They also showed identical LD,, values in mice The stereochemistry of these compounds was not addressed The earlier, almost universally accepted belief that if one enantiomer of a chiral molecule demonstrates pharmacological activity, the other enantiomer will be pharmacologically inert, is not valid It must be anticipated that all stereoisomers of an organic molecule will exhibit pharmacological effects, frequently widely different and unpredictable Many examples of qualitative and quantitative differences in metabolism of enantiomers are documented (57) ( 2)-3-(3-Hydroxypheny1)-N-n-propylpiperidine (3-PPP, 92) was originally described (58) as having highly selective activity at dopaminergic autoreceptors At high doses (R)-(92) selectively stimulated presynaptic dopaminergic receptor sites, whereas at lower doses it selectively stimulated postsynaptic receptor sites (59) In contrast, the (S)-enantiomer stimulated presynaptic dopamine receptors and at the same dose level, it blocked postsynaptic dopamine recep- Alteration of Stereochemistry and Design of Stereoisomers and Geometric Isomers tors Thus, this enantiomer exhibits a bifunctional mode of doparninergic attenuation: that of presynaptic agonism and postsynaptic antagonism The observed pharmacological effects of the racemic modification are the sum total of the complex activities of the two enantiomers, and the pharmacology of racemic 3-PPP is not an accurate reflection of the pharmacological properties of the individual enantiomers The contemporary literature strongly reflects the philosophy that pharmacological testing only of a racemic mixture is inadequate and may be misleading (R)-(-)-11-Hydroxy-10-methylaporphine (93) is a highly selective serotonergic 5-HT,, agonist (60) Remarkably, the (S)-enantiomer (94) is a potent antagonist at this same subpopulation of serotonin receptors (guinea pig ileum prep- 705 aration) (61) Both enantiomers bind strongly to 5-HT,, receptors from rat forebrain membrane The phenomenon of enantiomers that possess opposite effects (agonist-antagonist) at the same receptor, once considered to be extremely rare, has recently been noted more often, probably because of the increasing recognition by medicinal chemists and pharmacologists that each member of an enantiomeric pair may possess its own unique and unpredictable pharmacology In addition to stereochemistry about a carbon center, other potentially chiral atoms offer possibilities for pharmacological significance A gastroprokinetic compound (95) with serotonergic activity bears a chiral sulfoxide moiety (62) The enantiomers are equiptent, but the (S)-enantiomerdemonstrates a greater intrinsic activity than that of the (R)-enantiomer Casy (63) cited pharmacological differences between stereoisomers of chiral sulfoxide moieties in cholinergic oxathiolane congeners (96-99) of muscarine cis- and trans-4-Aminocrotonic acids (100) and (101) were prepared (64) as congeners of yaminobutyric acid ( G B A ) (6) Analog Design H2C I H N d COOH The folded 2-isomer (100) was inactive in assays for GABA agonism, whereas the extended E-isomer (101) was active These data demonstrate biological differences of geometric isomers, which in turn involve a parameter discussed previously: imposition of a degree of structural rigidity on the molecule A strategy analogous to this El2 olefinic GABA congener design addressed cis- and trans-1,2-disubstituted cyclopropane derivatives (102) and (1031, whose relative effects at GABA receptors paralleled those of the olefinic derivatives (65) The E-isomer of the diethylstilbestrol structure (104) has 10 times the estrogenic potency of the 2-isomer; this effect has been rationalized from the conclusion that the E- COOH geometric isomer is an open-chain analog of the natural estrogen estradiol (105) (66).'In dienestrol(106),the geometric isomerism possible with olefinic moieties has been further F Fragments of the Lead Molecule exploited to achieve a similar kind of openchain analogy to the steroid ring system as in diethylstilbestrol, and a high level of estrogenic activity results Hexestrol(107), the saturated congener of noid X-receptor ligand and it is inactive at the retinoic acid receptor, whereas the (R,R)-enantiomer is an extremely weak agonist at the retinoid X-receptor, although it has some effect a t the retinoic acid receptor Thus, the molecular modifications shown in (109) result in selectivity of action a t these two receptors diethylstilbestrol (104), is the meso-form of the molecule It has the greatest estrogenic potency of the three possible stereoisomers; however, it is less potent than diethylstilbestrol(67) A partial restriction of side-chain flexibility in retinoic acid (108) was achieved by incorporating portions of the side chain into a benzene ring and a cyclopropane ring (109) (68) I COOH Introduction of the cyclopropane ring changes the corresponding trans-olefinic moiety of (108) to a cisoid disposition in (log), thus changing the overall steric disposition of the side chain Moreover, the cyclopropane ring introduces chirality into the molecule The (S,S)-enantiomershown is a potent reti- FRAGMENTS OF THE LEAD MOLECULE Design of fragments of a lead molecule is based on the premise that some lead molecules, especially polycyclic natural products, may be much more structurally complex than is necessary for optimal pharmacologic effect It is hypothesized that a pharmacophoric moiety may be buried within the complex structure of the lead compound and, if this pharmacophore can be clearly defined, it may be possible to "dissect" it out chemically The result may be biologically active, simpler molecules that may themselves be used as leads in further analog design A bond disconnection strategy may be employed in which bonds in the polycyclic structure are broken or removed to destroy one or more of the rings The result may be a valuable drug that is more accessible (through chemical synthesis) than the original lead molecule A possible disadvantage to this strategy of analog design is that the greater flexibility that is introduced into a rigid molecule may compromise or destroy the conformational integrity that may have existed in the pharmacophoric portion, at the expense of activity and/or potency There may be a similar destruction of chiral centers, which may be undesirable Morphine (110) typifies a lead molecule for which fragment analog design has been used Analog Design The analgesic preceptor pharmacophore of morphine has been defined (69) as comprising the basic nitrogen atom, the aromatic ring located three carbon atoms from the nitrogen, and a quaternary carbon adjacent to the aromatic ring, which provides a region of molecular bulk A bond disconnection strategy involved disruption of the hydrofuran ring to give rise to morphinan derivatives [e.g., levorphanol (11I)], whose pharmacologic effects closely parallel those of morphine (70) Further simplification of the morphine ring system led to benzomorphan derivatives, typified by metazocine (112), in which morphinelike analgesic activity is retained Finally, 4-phenylpiperidine derivatives typified by meperidine (113) and the nonheterocyclic system methadone (114) present the putative analgesic pharmacophore with a seemingly minimal number of extraneous atoms These simple compounds retain opioid analgesic activity It must be noted, however, that the discovery of analgesic activity in 4-phenylpiperidine derivatives was not a result of a systematic structure-activity study of the morphine molecule, but was serendipitous (71) Asperlicin (115), a potent cholecystokinin-A antagonist, was subjected to two different bond disconnection strategies, as indicated (72) Path A leads to tryptophan derivatives (1161,some of which are potent cholecystokinin antagonists (73) Some quinazolinone derivatives (117) of disconnection pathway B showed extremely high potency and excellent selectivity as cholecystokinin-B receptor subtype ligands (44) A combination of X-ray crystallography and computational chemistry was used in the decision-making process in the bond disconnection (44) and in the design of the specific target molecules The myoneural-blocking pharmacophore in d-tubocurarine (118)was speculated to include the two cationic heads (the quaternary ammonium group and the protonated tertiary m i n e ) ; the cationic heads are separated by 10 atoms (nine carbons and one oxygen) Based on these parameters, a simple molecule, decamethonium (119),in which two trimethylammonium heads are separated by 10 methylene groups to approximate the internitrogen distance in d-tubocurarine, was designed independently by two groups of investigators (74, 75) This synthetic fragment/ analog of d-tubocurarine exhibits a high degree of potency and activity in production of flaccid paralysis of skeletal muscles, superficially like that of the lead compound However, the myoneural blockade from d-tubocurarine is of the nondepolarizing type, whereas decamethonium produces a depolarizing skel- Fragments of the Lead Molecule eta1muscle blockade This fundamental mechanistic difference is probably attributed, at least in part, to the flexibility of the decamethonium molecule compared with that of dtubocurarine There is a considerable differ- ence in the spectrum and severity of side effects and in the technique of employment of these two drugs in clinical practice In all types of analog design, changes in chemical structure may result in unanticipated changes in Analog Design mechanism of action, even though the chemical nature of the pharmacophore may not be altered VARIATION IN INTERATOMIC DISTANCES Alteration of distances between portions of the pharmacophore of a molecule (or even between other portions of the molecule) may produce profound qualitative and/or quantitative changes in pharmacological actions In a,w-bis-trimethylammonium polymethylene compounds (120-123), maximal activity for blockade of autonomic ganglia (nicotinic N, receptors) resides in those derivatives where n = or (compounds 120 and 121) (76,77) Ganglionic effects drop drastically when n = or These observations have been rationalized as being a reflection of attainment of optimal interquaternary distance in the penta- and hexamethylene congeners, for optimal interaction with ganglionic receptor subsites Remarkably, as the number of methylene groups in (120) is greatly increased, a high level of ganglionic-blocking potency returns The hexadecyl and octadecyl congeners (122) and (123) are approximately four times as potent at autonomic ganglia as the pentaand hexamethylene compounds As was mentioned previously, polymethylene bis-quaternary systems, in which the cationic heads are separated by 10 methylene groups, have potent effects at myoneural junctions (nicotinic N, receptors) and have little ability to affect nerve activity at autonomic ganglia Thus, extension of a bis-quaternary polyalkylene molecule from five or six methylenes to 10 pro- duces a pharmacological change from ganglionic blockade to myoneural blockade, and further extension to 16-18 methylenes results in loss of myoneural effects and a return of ganglionic blocking action Hemicholinium (124) competitively inhibits the high affinity, sodium-dependent uptake of choline into the nerve terminal (the ratedetermining step in acetylcholine synthesis in the nerve terminal), thus depleting stores of acetylcholine and producing slow onset, longduration myoneural blockade (78, 79) In a series of congeners of hemicholinium, the central biphenyl portion of the molecule was changed to terphenyl (125) and to p-phenylene (126) Both changes resulted in profound loss of the myoneural blockade characteristic of hemicholinium (68).This result was ascribed to alteration of the proposed opti- Variation in Interatomic Distances mum interquaternary nitrogen distance of 14.4 A in hemicholinium (124),to 18.4 A in the terphenyl analog, and to 10.2 A in thep-phenylene analog The central biphenyl spacer in hemicholinium was changed to a 2,7-disubstituted phenanthrene (127), trans,trans-4,4'-bicyclohexyl(128), and 2,2'-dimethylbiphenyl(129) In all three of these svstems the 4 inter" quaternary distance found in hemicholinium was maintained; all of these congeners were qualitatively and quantitatively similar to hemicholinium in inhibition of neuromuscular transmission Conformational analysis of the polyalkylene congeners (130) and (131) demonstrated that, when the flexible polyalkylene chain is maximally extended and is in a staggered conformation, the interquaternary distance in the hexamethylene congener (130) is approximately 14 A,and in the heptamethylene congener (131)it is approximately 15 A Both compounds exhibited hemicholiniumlike inhibition of neuromuscular transmission, although they were less potent than hemicholinium (80) This diminution of -DOtency might be ascribed to the compromising of another structural parameter in the hemicholinium molecule: the rigidity of the central biphenyl spacer unit that maintains the internitrogen distance Replacement of the benzene ring linkers of (82) (see above) by alkyl linkers (structure 132) permitted retention of blocking activity on calcium-activated potassium channels (81) The most potent member of the series studied was that in which m = n = In this compound the two respective internitrogen distances closely approximate those in the benzene ringlinked com~ound(82) In a series of phenylalkylenetrimethylammonium derivatives (133-136), nicotinic agonism is maximal when n = (compound 136) It was concluded (82) that a moiety (here, a benzene ring) with high electron density three or four single bond lengths (-6& from the cationic center is a requirement for nicotinic agonism in the series These conclusions may be compromised by the fact that the alkylene series was not extended beyond the three-carbon spacer chain Therefore, it is not known whether the four-carbon homolog would display greater or lesser potency than that of the three-carbon molecule Peculiarly, the first two members of the series have only very weak nicotine-like activity in the presence of atropine A series of compounds, illustrated by (137), was evaluated for in vitro affinity for a, and a,-adrenoceptors by radioligand-binding assays (83) All compounds showed good affinities for the a, adrenoceptor, with Ki values in the low nanomolar range The polymethylene chain spacer between fmylpiperazinylpyradizinone and aryl piperazine moieties was shown to influence the affinity and selectivity of these compounds A gradual increase in affinity for the a, adrenoceptor was observed, by length- Analog Design ening the polymethylene chain, up to a maximum of seven carbon atoms The a,/a, ratio of adrenoceptor-binding affinities for the series of compounds did not parallel the a , adrenoceptor-binding affinities for the series, although all of the seven (C,-C,) congeners of (137)had somewhat more afinity for the a, receptor REFERENCES I Langmuir, J.Am.Chem Soc., 41,868(1919) C W Thornber, Chem Soc Rev., 8,563(1979) G A Patani and E G LaVoie, Chem Rev., 96, 3147(1996) A Burger, Med Chem Res., 4,89(1994) P Floersheim, E Pombo-Villar, and G Shapiro, Chimia, 46,323(1992) P Krogsgaard-Larsen, H Hjeds, E Falk, F S J~rgenson,and L Nielsen, Adv Drug Res., 17, 38(1988) H.Hashiguchi and H Takahashi, Mol Pharmacol., 13,362 (1977) K Anderson, A Kuruvilla, N Uretsky, and D Miller, J Med Chem., 24,683(1981) R J Baldessarini in A G Gilman, L S Goodman, T W Rall, and F Murad, Eds., Goodman and Gilman's The Pharmacological Basis of Therapeutics, t h ed., Macmillan, New York, 1985, pp 393-397,414 10 S I Ankier i n G P Ellis and G B West, Eds., Progress i n Medicinal Chemistry, Vol 23, Elsevier, Amsterdam, 1986, p 121 11 E I Isaacson in J N Delgado and W A Remers, Eds., Wilson and Gisvold's Textbook of Organic Medicinal and Pharmaceutical Chemistry, 10th ed., Lippincott-Raven, Philadelphia, 1998, p 473 12 L Otsuki, J Ishiko, M Sakai, K Shiniahara, and T Momiyama, Pharmacometrics (Tokyo),6, 973(1972) 13 M Protiva, M Rajsner, V Seidlova, E Adlerova, and Z Vejdelek, Experientia, 18,326 (1962) 14 S Yous, J Andrieux, H E Howell, P J Morgan, P Reynard, B Pfeiffer,D Lessieur, and B Guardiola-Lemaitre, J Med Chem., 35, 1484 (1992) 15 S M Bromidge, F Brown, F Cassidy, M S G Clark, S Dabbs, M S Hadley, J Hawkins, J M Loudon, C B Naylor, B S Orlek, and G J Riley, J Med Chem., 40,4265(1997) 16 P Sauerberg, J W Kindtler, L Nielsen, M J Sheardown, and T Honor6, J Med Chem., 34, 687(1991) 17 M Hollmann, A O'Shea-Greenfield, S W Rohers, and S Heinemann, Nature, 342, 643 (1989) 18 F H Bellevue, M L Boahbedason, R H W u , R A Casero Jr., D Rattendi, C J Bacchi, and P M Woster, Bioorg Med Chem Lett.,6,2765 (1996) 19 P M Woster, Annu Rep Med Chem., 36,99 (2001) 20 E Mutschler and G Lambrecht in E Ariens,W Soudjin, and P B M W M Timmermans, Eds., Stereochemistry and Biological Activity of Drugs, Blackwell, Oxford, UK, 1983, p 65 21 J G Cannon in E Jucker, Ed., Progress in Drug Research, Vol 29, Birkhauser Verlag, Basel, Switzerland, 1985, pp 324-334 22 C Melchiorre, A Chiarini, M Gianella, D Giardina,W Quaglia, and V Tumiatti i nV Claasen, Ed., Trends i n Drug Research,Vol 13,Elsevier, Amsterdam, 1990, pp 37-48 23 F D King, A M Brown, L M Gaster, A J Kaumann, A D Medhurst, S G Parker, A A Parsons, T L Patch, and P Raval, J Med Chem., 36,1918(1993) 24 P S Portoghese, A A Mikhail, and H J Kupferberg,J Med Chem., 11,219(1968) 25 R M Moriarty, L A Enache, L Zhao, R Gilardi, M V Mattson, and Prakash, J Med Chem., 41,468(1998) References r 26 C M Bertha, B J.Vilner, M.V Mattson, W D Bowen, K Becketts, H X u , R B Rothman, J L Flippen-Anderson, and K C Rice, J Med Chem., 38,4776 (1995) 27 C Y Chiou, J P Long, J G Cannon, and P D Armstrong, J Pharmacol Exp Ther., 166,243 (1969) 28 J G Cannon and P D Armstrong, J Med Chem., 13, 1037 (1970) 29 J G Cannon, T Lee, V Sankaran, and J P Long, J Med Chem., 18, 1027 (1975) 30 P W Ehrhardt, R J Gorczynski, and W G Anderson, J Med Chem., 22,907 (1975) 31 R R Ruff010 in G Kunos, Ed., Adrenoceptors and Catecholamine Action, Part B, Wiley-Interscience, New York, 1983, pp 10-11 32 H J Komiskey, J F Bossart, D D Miller, and P N Patel, Proc Natl Acad Sci USA, 75,2641 (1978) 33 E E Smissman and W H Gastrock, J Med Chem., 11, 860 (1968) 34 J A Monn, M J Valli, S M Massy, M M Hansen, T J Kress, J P Wepsiec, A R Harkness, J L Grutsch Jr., R A Wright, B G Johnson, S L And&, A Kingston, R Tomlinson, R Lewis, K R Griffey, J P Tizzano, and D D Schoepp, J Med Chem., 42,1027 (1999) 35 J F McCarthy, J G Cannon, and J P Buckley, J Pharm Sci., 52, 1168 (1963) 36 J F McCarthy, J G Cannon, J P Buckley, and W J Kinnard, J Med Chem., 7,72 (1964) 37 A Burger and G R Bedford, J Med Chem., 6, 402 (1963) 38 M V Koch, J G Cannon, and A M Burkman, J Med Chem., 11,977 (1968) 39 E R Atkinson, F J Bullock, F E Granchelli, S Archer, F J Rosenberg, D G Teiger, and F C Nachod, J Med Chem., 18, 1000 (1975) 40 J G Cannon in ref 21, pp 309-310 41 J G Cannon, F.-L Hsu, J P Long, J R Flynn, B Costall, and R J Naylor, J Med Chem., 21, 248 (1978) 42 J Z Ginos a n d F C Brown, J Med Chem., 21, 155 (1978) 43 J Z Ginos, G C Cotzias, and D Doroski, J Med Chem., 21,160 (1978) 44 M J Y u , J R McCowan, N R Mason, J B Deeter, and L G Mendelsohn, J Med Chem., 35,2534 (1992) 45 K.-L Y u , J J Bronson, H Yang, A Patick, M Alam,V Brankovan, R Datema, M J M Hitchcock, and J C Martin, J Med Chem., 35,2958 (1992) 46 G C Crawley, R I Dowell, P N Edwards, S J Foster, R M McMillan, E R H Walker, and D Waterson, J Med Chem., 35, 2600 (1992) 47 H.-S Kim, Y Nagai, K Ono, K Begum, Y Wataya, Y Hamada, K Tsuchiya, A Masuyama, M Nojima, and K J McCullough, J Med Chem., 44, 2357 (2001) 48 R H Uloth, J R Kirk, W A Gould, and A A Larsen, J Med Chem., 9,88 (1966) 49 J Campos-Rosa, D Galanakis, A Piergentili, K Bhandari, C R Ganellin, P.M Dunn, and D H Jenkinson, J Med Chem., 43,420 (2000) 50 J E Macor, C B Fox, C Johnson, B K Koe, L A Label, and S H Zorn, J Med Chem., 35, 3625 (1992) 51 R A Glennon i n J G Cannon, Ed., Advances in CNS Drug-Receptor Interactions, JAI Press, Greenwich, CT, 1991, pp 147-148 52 B B Hoffmanin J G Hardman and L E Limbird, Eds., Goodman and Gilman's The Pharmacological Basis of Therapeutics, 10th ed., McGraw-Hill, New York, 2001, p 218 53 R L Johnson in J N Delgado and W A Remers, Eds., Wilson and Gisvold's Textbook of Organic Medicinal and Pharmaceutical Chemistry, 10th ed., Lippincott-Raven, Philadelphia, 1998, p 493 54 (a) A M Lands, J R Lewis, and V L Nash, J Pharmacol Exp Ther., 83, 253 (1945); ( b ) A M Lands, V L Nash, and B L Dertlinger, J Pharmacol Exp Ther., 89, 382 (1947); (c) D F Marsh and D A Herring, J Pharmacol Exp Ther., 97, 68 (1949); ( d ) E J Fellows, E Macko, and R A McLean, J Pharmacol Exp Ther., 100,267 (1950) 55 M D Mashkovsky, L N Yakhoutov, M E Kaminka, and E E Mikhlina i n E Jucker, Ed., Progress in Drug Research,Vol 27, Birkhauser Verlag, Basel, Switzerland, 1983, pp 35-38 56 P Pevarello, A Bonsignori, P Dostert, F Heidemperger, V Pinciroli, M Colombo, R A McArthur, P Salvati, C Post, R G Fariello, and M Varasi, J Med Chem., 41, 579 (1998) 57 A F Casy, The Steric Factor i n Medicinal Chemistry Dissymmetric Probes of Pharmacological Receptors, Plenum, New YorkiLondon, 1993, pp 52-61 58 S Hjorth, A Carlsson, H Wikstrom, P Lindberg, D Sanchez, U Hacksell, L.-E Arvidsson, Svensson, a n d J L G Nilsson, Life Sci., 28, 1225 (1981) 59 H.Wikstrom, D Sanchez, P Lindberg, U Hacksell, L.-E Arvidsson, A M Johansson, S.-0 Thorberg, J L G Nilsson, K Svensson, S Analog Design Hjorth, D Clark, and A Carlsson, J Med Chem., 27, 1030 (1984) 60 J G Cannon, P Mohan, J Bojarski, J P Long, R K Bhatnagar, P A Leonard, J R Flynn, and T K Chattejee, J Med Chem., 31,313 (1988) 61 J G Cannon, S T Moe, and J P Long, Chirality, 3, 19 (1991) 62 B T Butler, G Silvey, D M Houston, D R Borcherding, V L Vaughn, A T McPhail, D M Radzik, H Wynberg, W Ten Hoeve, E Van Echten, N K Ahmed, and M D Linnik, Chirality, 4, 155 (1992) 63 See ref 57, pp 244-247; see also M Pigini, L Brasili, M Gianella, and F Gualtieri, Eur J Med Chem., 16,415 (1981) 64 G A Johnston, D R Curtis, P M Beart, C J A Game, R M McColloch, and B Twitchin, J Neurochem., 24, 157 (1975) 65 R D Allan, D R Curtis, P M Headley, G A Johnson, D Lodge, and B Twitchen, J Neurochem., 34,652 (1980) 66 D S Fullerton in R F Doerge, Ed., Wilson and Gisvold's Textbook of Organic Medicinal and Pharmaceutical Chemistry, 8th ed., Lippincott, Philadelphia, 1982, p 670 67 R W Brueggemeier, D D Miller, and D T Witiak in W Foye, T L Lemke, and D A Williams, Eds., Principles of Medicinal Chemistry, 4th ed.,Williams & Wilkins, Media, PA, 1995, p 474 68 V Vuligonda, S M Thacher, and R A S Chandraratna, J Med Chem., 44, 2298 (2001) 69 T Nogrady, Medicinal Chemistry, 2nd ed., Oxford University Press, New York, 1988, p 457 70 J H Jaffe and W R Martin in ref 9, p 513 71 J V Aldrich in M E Wolff, Ed., Burger's Medicinal Chemistry and Drug Discovery, 5th ed., Vol 3, Wiley-Interscience, New York, 1996, p 357 72 M J Yu, K J Thrasher, J R McCowan, N R Mason, and L G Mendelsohn, J Med Chem., 34, 1505 (1991) 73 F W Hahne, R T Jensen, G F Lemp, and J D Gardner, Proc Natl Acad Sci USA, 78, 6304 (1981) 74 R B Barlow and H R Ing, Nature, 161, 718 (1948) 75 W D M Paton and E J Zaimis, Nature, 161, 718 (1948) 76 D J Triggle, Neurotransmitter-Receptor Znteractions, Academic Press, New York, 1971, p 360 77 V Trcka in D A Kharkevich, Ed., Handbook of Experimental Pharmacology, Vol 53, SpringerVerlag, New York, 1980, p 138 78 J G Cannon, T M.-L Lee, A M Nyanda, B Bhattacharyya, and J P Long, Drug Des Deliv., 1,209 (1987) 79 J G Cannon, Med Res Rev., 14,505 (1994) 80 J G Cannon, T M.-L Lee, Y.-a Chang, A M Nyanda, B Bhattacharyya, J R Flynn, T Chatterjee, R K Bhatnagar, and J P Long, Pharm Res., 5,359 (1988) 81 J.-Q Chen, D Galanakis, C R Ganellin, P M Dunn, and D H Jenkinson, J Med Chem., 43, 3478 (2000) 82 W C Holland in E J Ariens, Ed., Proceedings of the Third International Pharmacological Meeting, Vol 7, Pergamon, Oxford, UK, 1966, pp 295-303;K.C Wongand J P Long, J.Pharmacol Exp Ther.,137,70 (1962) 83 R Barbaro, L Betti, F Corelli, G Giannacinni, L Maccari, F Manetti, G Straoaghetti, and S Corsano, J Med Chem., 44, 2118 (2001) CHAPTER SEVENTEEN : Approaches to the Rational Design of Enzyme Inhibitors MICHAEL J MCLEISH GEORGE L KENYON Department of Medicinal Chemistry University of Michigan Ann Arbor, Michigan Contents Introduction, 716 1.1 Enzyme Inhibitors in Medicine, 716 1.2 Enzyme Inhibitors in Basic Research, 720 Rational Design of Noncovalently Binding Enzyme Inhibitors, 720 2.1 Forces Involved in Forming the EnzymeInhibitor Complex, 721 2.1.1 Electrostatic Forces, 723 2.1.2 van der Wads Forces, 723 2.1.3 Hydrophobic Interactions, 724 2.1.4 Hydrogen Bonds, 724 2.1.5 Cation-n Bonding, 724 2.2 Steady-State Enzyme Kinetics, 725 2.2.1 The Michaelis-Menten Equation, 725 2.2.2 Treatment of Kinetic Data, 726 2.3 Rapid, Reversible Inhibitors, 728 2.3.1 Types of Rapid, Reversible Inhibitors, 728 2.3.1.1 Competitive Inhibitors, 728 2.3.1.2 Uncompetitive Inhibitors, 729 2.3.1.3 Noncompetitive Inhibitors, 730 2.3.2 Dixon Plots, 731 2.3.3 IC,, Values, 731 2.3.4 Examples of Rapid Reversible Inhibitors, 733 2.4 Slow-, Tight-, and Slow-Tight-Binding Inhibitors, 734 2.4.1 Slow-Binding Inhibitors, 734 2.5 Inhibitors Classified on the Basis of Structure/Mechanism, 740 2.5.1 Ground-State Analogs, 740 2.5.2 Multisubstrate Analogs, 741 2.5.3 Transition-State Analogs, 748 Rational Design of Covalently Binding Enzyme Inhibitors, 754 Burger's Medicinal Chemistry and Drug Discovery Sixth Edition, Volume 1: Drug Discovery Edited by Donald J Abraham ISBN 0-471-27090-3 O 2003 John Wiey & Sons, Inc 715 Approaches to the Rational Design of Enzyme Inhibitors 716 3.1 Evaluation of the Mechanism of Inactivation of Covalently Binding Enzyme Inhibitors, 756 3.1.1 Criteria for the Study of Affinity Labels, 756 3.1.2 Criteria for the Study of MechanismBased Inactivators, 759 INTRODUCTION Many of the top 100 drugs sold worldwide are enzyme inhibitors In recent years, enzyme inhibitors not only have provided an increasing number of potent therapeutic agents for the treatment of diseases, but also have significantly advanced the understanding of enzymatic transformations The aim of this chapter is to present current approaches to socalled rational inhibitor design, which uses knowledge of enzymic mechanisms and structures in the design process Rational inhibitor design is intended to complement laborious and resource-consuming screening processes, which consist of testing large numbers of synthetic chemicals or natural products for inhibitory activity against a chosen target enzyme 1.1 Enzyme lnhibitors in Medicine A human cell contains thousands of enzymes each of which can, theoretically, be selectively inhibited These enzymes constitute the various metabolic pathways that, in concert, provide the requirements for the viability of the cell A selective inhibitor may block either a single enzyme or a group of enzymes, leading to the disruption of a metabolic pathway(s) This will result in either a decrease in the concentration of enzymatic products or an increase in the concentration of enzymatic substrates The effectiveness of an enzyme inhibitor as a therapeutic agent will depend on (1)the potency of the inhibitor, (2)its specificity toward its target enzyme, (3) the choice of metabolic pathway targeted for disruption, and (4) the inhibitor or a derivative possessing appropriate pharmacokinetic characteristics Higher potency will mean less drug is required to obtain a physiological response, whereas high specificity means that the inhibitor will react only with its target enzyme and not with other sites in the body Taken together, low 3.2 Affinity Labels, 760 3.3 Mechanism-Based Inhibitors, 764 3.4 Pseudoirreversible Inhibitors, 771 Conclusions, 774 dosage and high specificity combine to reduce both the toxicity caused by inhibition of other vital enzymes and the problems arising from the formation of toxic decomposition products Further, high specificity will generally avoid depletion of the inhibitor concentrations in the host by nonspecific pathways The areas of potency and specificity will both be addressed in this chapter Clearly, the choice of target enzyme is also of prime importance for chemotherapy, although that is beyond the scope of this review However, there are a number of texts available that provide a good introduction to this subject (1-4) Good bioavailability of the drug is also crucial for the drug to reach its site of action in the body in effective therapeutic concentrations For example, highly polar or charged compounds, such as phosphorylated compounds, frequently cannot readily cross cell membranes and are therefore generally less useful as drugs Physical approaches to facilitate the transport of this class of compounds into the cell include the use of liposomes or nanoparticles (5-7) Chemical approaches may also be employed These include the use of prodrugs, in which functional groups on the inhibitor are modified in such a manner that they are able to be taken up by the cell and, later, metabolically converted to the active drug Prodrugs are discussed in more detail in Volume 2, Chapter 14 As indicated earlier, a wide variety of enzyme inhibitors have found use in the clinic Tables 17.1-17.3 show a number of these compounds and, although they provide by no means an exhaustive list, they give an indication of the range of human disease states that can be ameliorated with the use of enzyme inhibitors The human body, even though its defenses are constantly on guard, is still susceptible to invasion by foreign pathogens Since the de- 717 Introduction Table 17.1 Examples of Enzyme Inhibitors Used in the Treatment of Bacterial, Fungal, Viral, and Parasitic Diseases Clinical Use Antibacterial Antibacterial Antibacterial Antibacterial Antifungal Antifungal Antiviral Antiviral Antiviral Antiviral Antiviral Antiviral Antiprotozoal Antiprotozoal Enzyme Inhibited Inhibitor Dihydropteroate synthetase Dihydrofolate reductase Alanine racemase Transpeptidase Fungal sterol l4a-demethylase Fungal squalene epoxidase Thymidine kinase and thymidylate kinase DNA, RNA polymerases Viral DNA polymerase HIV reverse transcriptase HIV protease Influenza virus neuraminidase Pyruvate dehydrogenase Ornithine decarboxylase Sulphonamides Trimethoprim, methotrexate D-Cycloserine Penicillins, cephalosporins Clotrimazole, ketoconazole Terbinafine, naftifine Idoxuridine Cytosine arabinoside (Ara-C) Acyclovir, vidarabine Dideoxyinosine, zidovudine Saquinavir Zanamavir, oseltamivir Organoarsenical agents a-Difluoromethylornithine velopment of the sulfa drugs (sulfonamides), enzyme inhibitors have played a vital role in controlling these infectious agents Table 17.1 provides a list of enzyme inhibitors that have been used in the treatment of the various diseases caused by these agents All these compounds needed to satisfy the usual requirements for specificity and low toxicity This can be achieved in a variety of ways For instance, it is possible to inhibit an essential pathway in the pathogen that does not exist in the host D-Cycloserine (1) (Fig 17.11, for example, inhibits alanine racemase, an enzyme involved in bacterial cell wall biosynthesis and not found in humans (8,9) D-Cycloserine is active against a broad spectrum of both gram-positive and gramnegative bacteria (lo), but plays its major role in the treatment of tuberculosis (11) Conversely, even if both host and pathogen contain the same enzymes, it may be possi- ble to exploit subtle structural differences between the isozymes to obtain a highly specific inhibitor that preferentially binds to the invader's version Trimethoprim (2) shows this selective toxicity An inhibitor of dihydrofolate reductase, trimethoprim is a potent antibacterial agent because the bacterial enzyme is inhibited at a concentration several thousand times lower than that required for inhibition of the mammalian isozyme (12) Acyclovir ( , an antiviral drug used for the treatment of herpes infections (13, 141, also fits into this category It binds very tightly to the Herpes simplex DNA polymerase with an estimated half-life of about 40 days Acyclovir is a prodrug because it requires transformation by a viral thymine kinase and cellular phosphotransferases to the corresponding triphosphate (3b) to serve in vivo as an inhibitor of the viral DNA polymerase (15) Table 17.2 Examples of Enzyme Inhibitors Used in the Treatment of Cancer Type of Cancer Benign prostatic hyperplasia Estrogen-mediated breast cancer Leukemia, osteosarcoma, head, neck, and breast cancer Colorectal cancer Leukemia Small-cell lung cancer, nonHodgkin's lymphoma Hairy-cell leukemia Enzyme Inhibited Inhibitor Steroid 5a-reductase Aromatase Dihydrofolate reductase Finasteride Arninoglutethimide, fadrozole Methotrexate Thymidylate synthase Glutamine-PRPP amidotransferase Topoisomerase I1 5-Fluorouracil 6-Mercaptopurine, azathioprine Etoposide Adenosine deaminase Pentostatin 718 Approaches to the Rational Design of Enzyme Inhibitors Table 17.3 Examples of Enzyme Inhibitors Used in Various Human Disease States Clinical Use Epilepsy Epilepsy Epilepsy Antidepressant Antihypertensive Cardiac disorders Gout Ulcer Hyperlipidemia Anti-inflammatory Arthritis Glaucoma Glaucoma Enzyme Inhibited Inhibitor GABA transaminase Carbonic anhydrase Succinic semialdehyde dehydrogenase Monoamine oxidase (MAO) Angiotensin converting enzyme Na',K'-ATPase Xanthine oxidase Hf ,K+-ATPase HMG-CoA reductase Prostaglandin synthase, Cyclooxygenase (COX) I and I1 Cyclooxygenase (COX) I1 Acetylcholinesterase Carbonic anhydrase I1 Although their inhibitors are not specifically therapeutic agents in themselves, the p-lactamases are another important target for drug design These are bacterial enzymes and, as with the alanine racemases, are not found in humans Inhibitors of p-lactamases include clavulanic acid (4) (16-20) and sulbactam (penicillanic acid sulfone) (5) (18, 21-24) These two compounds act to prevent the bacterial degradation of penicillins and cephalosporins by p-lactamases, thereby extending their lifetime and effectiveness Accordingly, both clavulanic acid (4) and sulbactam (5) have reached the market as drugs that act synergistically with these commonly prescribed antibacterial agents Even though it has proved possible to selectively inhibit the enzymes of a number of pathogens, the enzymes of cancer cells have proved to be a far more elusive target Indeed, the majority of the currently employed antitumor agents can be described as antiproliferative agents These take advantage of the fact that many, but not all, tumor cells grow and divide more rapidly than normal cells Lymphomas, for example, proliferate more rapidly than solid tumors, whereas, conversely, acute leukemia cells divide more slowly than the surrounding bone marrow cells Most of the enzyme inhibitors used as these antiproliferative agents (Table 17.2) can also be described as antimetabolites (i.e., they inhibit a metabolic pathway), often those involved in DNA biosynthesis, which are important for cell survival or replication 5-Fluorouracil (6), the y-Vinyl GABA Sulthiame Sodium valproate Tranylcypromine, phenelzine Captopril, enalaprilat Cardiac glycosides Allopurinol Omeprazole Atorvastatin, simvastatin Aspirin, naproxen, ibuprofen Celecoxib Neostigmine Acetazolamide, dichlorphenamide prodrug form of an inactivator of thymidylate synthase (25),and methotrexate (71, an inhibitor of dihydrofolate reductase (26, 27), both fit into this category Unfortunately, rapidly dividing normal cells, such as hair follicles, the cells lining the gastrointestinal tract, and the bone marrow cells involved in the immune system are also significantly affected The resultant hair loss, nausea, and susceptibility to infection means that this type of chemotherapy is seldom employed as a first-line defense against cancer The inhibition of enzymes involved in metabolic pathways is not restricted to anticancer agents A variety of diseases have been correlated with either the dysfunction of an enzyme or an imbalance of metabolites A cross section of the disease states treated with enzyme inhibitors is shown in Table 17.3 Practically, these may be treated by the inhibition of an individual enzyme or by using enzyme inhibitors to regulate the metabolite concentration in the body For example, an imbalance of the two neurotransmitters, glutamate and y-aminobutyric acid, is responsible for the convulsions observed in epileptic seizure The latter is metabolized by y-aminobutyric acid aminotransferase (GABA-T) and, consequently, inhibitors of this enzyme offered themselves as potential antiepileptic candidates This led to the development of the GABA-T inhibitor, vigabatrin (8)(28),which clinically results in an increase of the brain concentration of y-aminobutyric acid and cessation of epileptic convulsions As with the anticancer agents, block- Introduction -0 N, H2N ~q~~ I\\ OMe OMe RO (3a) R = H (3b) R = PPP Figure 17.1 Examples of enzyme inhibitors used clinically ade of a metabolic pathway may also have therapeutic benefits The statins, a group of serum cholesterol-lowering drugs, are inhibitors of hydroxymethylglutaryl-CoA (HMGCoA) reductase (29) HMG-CoA reductase catalyzes the irreversible conversion of HMGCoA to mevalonic acid, the rate-determining step in cholesterol biosynthesis (3032) Inhibitors such as simvastatin (9) have been found to be effective in the treatment of hyperlipidemia and familial hypercholesteremia (33,341 and have become some of the world's best-selling drugs Finally, enzyme inhibitors can also be used to induce an animal model of a genetic disease Inactivation of y-cystathionase by propargylglycine, for example, produces an experimental model of the disease state known as cysta- Approaches to the Rational Design of Enzyme inhibitors 720 Table 17.4 Classification of Enzyme Inhibitors Employed in This Chapter Noncovalent Inhibitors Rapid reversible inhibitors (ground-state analogs) Tight, slow, slow-tight binding inhibitors Multisubstrate analogs Transition-state analogs thioninuria (35) Deficiency of this enzyme leads to the accumulation of cystathionine in the urine and has sometimes been associated with mental retardation (36) 1.2 Enzyme lnhibitors in Basic Research In basic research enzyme inhibitors have found a multitude of uses They serve as useful tools for the elucidation of structure and function of enzymes, as probes for chemical and kinetic processes, and in the detection of short-lived reaction intermediates (37) Product inhibition patterns provide information about an enzyme's kinetic mechanism and the order of substrate binding (38) Covalently binding enzyme inhibitors have been used to identify active-site amino acid residues that could potentially be involved in substrate binding and catalysis of the enzyme (39, 40) Reversible enzyme inhibitors are routinely used to facilitate enzyme purification by using the inhibitor as a ligand for affinity chromatography (41, 42) or as eluants in affinity-elution chromatography (43) Immobilized enzyme inhibitors can also be used to identify their intracellular targets (44), whereas irreversible inhibitors can be used to localize and quantify enzymes in vivo (45) In Table 17.4 we have provided the classification of the various types of enzyme inhibitors that we employ in this chapter The classification may appear somewhat arbitrary, in that some inhibitors may fit into more than one category This can arise because these categories are attempting to bring together some nonrelated properties such as structure, mechanism of action, and kinetic behavior Thus, what we have classed as a reversible inhibitor may, simply because it has a slow dissociation rate, be described elsewhere in the literature as being irreversible In each instance we will discuss approaches to the design of that type of inhibitor, as well as indi- Covalent Inhibitors Chemical modifiers Affmity labels Mechanism-based inhibitors Pseudoirreversible inhibitors cating how it may be evaluated The discussion will be accompanied by references to recent, representative examples from the literature Where appropriate, these examples will be of inhibitors of therapeutic interest It should be noted that we will concentrate on inhibitors directed at the active site of the enzyme While recognizing that there are inhibitors that bind to regions other than the active site, such as allosteric effectors, these are not the focus of this chapter and will not be included There are many reviews of enzyme inhibitors available in the literature (37, 46-48) and the reader is referred to them for more detailed analysis RATIONAL DESIGN OF NONCOVALENTLY BINDING ENZYME INHIBITORS As their name indicates, this class of inhibitors binds to the enzyme's active site without forming a covalent bond Therefore the affinity and specificity of the inhibitor for the active site will depend on a combination of the electrostatic and dispersive forces, and hydrophobic and hydrogen-bonding interactions Traditionally, noncovalently binding enzyme inhibitors were analogs of substrates, products, or reaction intermediates More recently, an explosion in the use of combinatorial chemistry and rapid screening techniques has seen the development of large numbers of enzyme inhibitors that bear little or no resemblance to the substrate or products, yet still bind selectively to their target enzyme Computer-aided drug design, in the broadest sense, encompasses both structure-based drug design and quantitative structure-activity relationship (QSAR) methods A complement to the rapid screening techniques, computer-aided methods provide a more focused approach to the Rational Design of Noncovalently Binding Enzyme Inhibitors design and discovery of both substrate and nonsubstrate analog inhibitors In structure-based design, the structure of a drug target interacting with small molecules is used to guide drug discovery Consequently, either the three-dimensional enzyme structure or, at a minimum, t h e pharmacophore structure must be known A pharmacophore represents the nature of the chemical groups of a given ligand and their relative orientation important for inhibitor binding Today, structure-based design, used in conjunction with docking techniques, combinatorial chemistry, and rapid screening not only leads more quickly to novel enzyme inhibitors but also greatly reduces the number of compounds that must be synthesized More information on these approaches may be found in Chapter 10 and some recent monographs (49-52) Traditionally, an increase in inhibitory or biological activity was achieved by synthesizing an analog of the substrate and then making gradual empirical changes in the structure by adding or removing functional groups QSAR methods provide a means of making this empirical testing more focused In this technique there is no need to know the structure of the active site Instead, computer algorithms are employed to correlate the biological activity of a series of inhibitors with their chemical structure, thereby allowing better predictions as to how to change the structure to obtain a more potent inhibitor This topic is discussed further in Chapter 1, and detailed reviews are also available (53-56) Table 17.4 shows the classification of noncovalent inhibitors we use in this chapter Based on their kinetics it is possible to distinguish among rapid reversible, tight-binding, slow-binding, slow-tight-binding, irreversible, and pseudoirreversible inhibitors Conversely, inhibitors classified on the basis of structure, such as ground-state analogs, multisubstrate inhibitors, and transition-state analogs, which mimic the structures of substrates and products, reaction intermediates, and transition states, may fall into any of the kinetic categories However, before introducing these categories, it is important to have an understand- 721 ing of the forces involved in the binding of substrates and inhibitors to an enzyme's active site 2.1 Forces Involved in Forming the EnzymeInhibitor Complex To understand the design concepts of the various types of noncovalently binding enzyme inhibitors, a basic knowledge of the binding forces between an enzyme's active site and its inhibitors is required The forces involved in a substrate or an inhibitor binding to an enzyme's active site are, as with a drug binding to a receptor, the same forces that are experienced by all interacting organic molecules These include ionic (electrostatic) interactions, ion-dipole and dipole-dipole interactions, hydrogen bonding, hydrophobic interactions, and van der Waals interactions A brief overview of the forces involved follows More comprehensive treatments can be found in Chapter and elsewhere (57-60) The binding of an inhibitor is dependent on a variety of interactions, and it is the sum of these interactions that will determine the degree of affinity of an inhibitor for the particular enzyme The reversible binding of an inhibitor to an enzyme's active site can be described as shown in Equation 17.1 There is an equilibrium between the free enzyme (E), inhibitor (I), and the enzyme-inhibitor complex (E I) The affinity of an inhibitor for the enzyme is measured by the inhibition constant Ki, which is the dissociation constant of the enzyme-inhibitor complex, at equilibrium (Equation 17.2) The lower the Ki value, the better the inhibitor, given that the equilibrium lies more in favor of enzyme-inhibitor complex formation The affinity of an inhibitor for an enzyme may be related to the standard free energy (AG") of a system by Equation 17.3 722 where R is the universal gas constant and T the temperature in degrees Kelvin The more negative the value of AG", the more favorable the interaction at equilibrium, and the smaller the Kivalue It should be noted that, from Equation 17.3, at physiological temperature relatively small changes in free energy, only 2-3 k d m o l , will have a significant effect on Ki The standard free energy (A@)can also be expressed in terms of enthalpic ( W and ) entropic (AS") components (Equation 17.4) Equation 17.4 states that the free energy of a system is lowered (i.e., the reaction is made more favorable) by either a decrease in enthalpy or an increase in entropy This is also an important concept because there are both enthalpic and entropic components to the forces that contribute to the strength of the enzyme-inhibitor interaction When discussing the forces involved in the noncovalent binding of a substratelinhibitor to an enzyme, or drug to a receptor, it must be recognized that these interactions will be carried out in an aqueous medium The physical properties of water mean that noncovalent interactions in aqueous solution will be significantly different from those interactions observed in either an organic medium or in the gas phase A water molecule has electronic asymmetry; the strongly electronegative oxygen atom withdraws electron density from the hydrogen atoms This creates partial positive charges on the hydrogens and a partial negative charge on the oxygen As a result a water molecule possesses a permanent dipole moment, facilitating strong interactions with other water molecules as well as with any charged or polar species Water is both a donor and acceptor of hydrogen bonds Consequently, in bulk solvent, water molecules are extensively hydrogen bonded to each other These are relatively weak bonds (-5 kcallmol) and, at physiological temperature, are rapidly broken and reformed However, the hydrogen-bonding network affects many of the properties of water Approaches to the Rational Design of Enzyme Inhibitors For example, water has a higher melting point, boiling point, and heat of vaporization than those of comparable hydrides such as H,S and NH, The heat capacity of water indicates that it is highly structured and its surface tension (73 dyne cm-' at 20°C) is considerably higher than that of most liquids (20-40 dyne cm-l) The dielectric constant of water (80) is also considerably higher than that of most liquids, which are generally less than 30 Ethanol, for example, has a dielectric constant of 24, whereas those of benzene and hexane are 2.3 and 1.9, respectively All told, water is a unique solvent, and one that has a major influence on binding interactions between an enzvme and an inhibitor Hydrogen bonds are readily formed between water and biologically important atoms such as the hydrogen bond acceptors N and and, to a lesser extent, S The conjugate acids NH and OH may act as hydrogen bond donors Molecules containing these atoms have the capacity for many hydrogen-bonding interactions with water and, as a result, are usually soluble in water However, solute-solute hydrogen bonding interactions are less favorable because their formation will require the disruption of favorable solute-water hydrogen bonds Thus, what may be strong hydrogen bonds in the gas phase, or in organic media, are often considerably weaker in aqueous m dia Water's high dielectric constant makes it extremely effective in solvating, dissociating, and dissolving most salts Because of its permanent dipole, water is readily able to interact with ionic species, with the result that ionic solute-solute interactions are less favored The situation is analogous to that observed for hydrogen bonding and again results in a weakening of the normally strong interactions between ions that occur in the gas phase or nonpolar media This is sometimes described as a "leveling effect." Small amounts of many nonpolar substances can also dissolve in water However, these substances not interact verv " well with water and prefer to interact with each other The force driving this interaction, known as the hydrophobic force, is not so much an attraction between hydrophobic molecules as an entropic effect arising from the Rational Design of Noncovalently Binding Enzyme Inhibitors displacement of water Indeed, there are no hydrophobic forces in the gas phase or in nonpolar solvents However, collectively, hydrophobic forces are thought to transcend other types of forces, particularly in the folding of proteins, in all biological systems 2.1.1 Electrostatic Forces Although we re- cognize that, in essence, all forces between atoms and molecules are electrostatic, here we use the term to describe ion-ion, ion-dipole, and dipole-dipole interactions At physiological pH, the side-chains of basic residues such as lysine and arginine and, to a lesser extent, the imidazole ring of histidine will be protonated, whereas the acidic groups on the side chains of aspartic and glutamic acid residues will be deprotonated In addition, the N-terminal amino groups and C-terminal carboxylates will be ionized Therefore, in addition to atoms with permanent and induced dipoles, an enzyme potentially will have several charged groups available for binding to charged or polarized groups on a substrate or inhibitor As described by Equation 17.5, the electrostatic force (F)between the charged atoms (q, and q,) will depend on the distance between the charged groups (r)and the dielectric constant of the surrounding medium (D) 723 polar solvents As discussed above, because of its high net permanent dipole moment, water is very polar and has a large dielectric constant The high polarity of water greatly diminishes the attraction or repulsion forces between any two charged groups giving rise to the leveling effect of water It is somewhat difficult to predict the exact strength of a chargecharge interaction between an enzyme and an inhibitor For example, the formation of a salt bridge (charge-charge) interaction between an enzyme (Enz) and an inhibitor (I) may be described by Equation 17.6 - ~ n z - f i ~ , (H20), + I-COP (HZO), = Both the charged species are initially solvated by water, and to form the salt bridge both ions must be desolvated This comes at some enthalpic cost, but the freeing of water molecules leads to a concomitant, favorable increase in entropy The strength of the ion pair will depend on the stability of the salt bridge vs that of the individual solvated ions If the salt bridge is buried in a relatively hydrophobic active site, it is less solvated and will be more favored than the same interaction in a solvent-exposed active site 2.1.2 van der Waals Forces Also called The strength of an ion-ion interaction is inversely related to the square of distance between the ions, whereas ion-dipole and dipoledipole interactions have llr4 and llr6 relationships, respectively Because the strength of the interaction decreases more slowly - with distance, ion-pair interactions can be thought of as long-range interactions Conversely, interactions involving dipoles are effective over only a short range, although, because they are much more prevalent, dipole interactions may be more significant to the overall binding process Clearly, the dependency of the strength of interaction on the distance between atoms is an important consideration when designing potential enzyme inhibitors Equation 17.5 also leads to the fact that electrostatic interactions are less favorable in nonpolar interactions or London dispersion forces, these are the universal attractive interactions that occur between atoms As two molecules closely approach each other there is an interpenetration of their electron clouds As a consequence, temporary local fluctuations in the electron density occur, giving rise to a temporary dipole in each molecule, even though the molecules may, in themselves, have no net dipole moment Thus there will be an attractive force between the two molecules, with the magnitude of the force depending on the polarizability of the particular atoms involved and the distance between each other Electronegative oxygen has, for example, a much lower polarizability than that of a nonpolar methylene group Accordingly, dispersion forces are considerably stronger between nonpolar compounds than between nonpolar com- 724 pounds and water The optimal distance between the atoms is the sum of each of their van der Wads radii, so these forces come into play only when there is good complementarity between enzyme and inhibitor Although van der Wads forces are quite weak, usually around 0.5-1.0 kcal/mol for an individual atom-atom interaction, they are additive and can make an important contribution to inhibitor binding Approaches to the Rational Design of Enzyme Inhibitors significant in nonpolar solvents, water greatly diminishes their magnitude The energy of the amide-amide N H -0hydrogen bond is about kcdmol, and is typical for hydrogen bonds (60) It should be remembered that, for a hydrogen bond to form between an enzyme and an inhibitor, any hydrogen bonds between the inhibitor and water, as well as those between the enzyme and water, must be broken (Equation 17.7) 2.1.3 Hydrophobic Interactions Hydropho- bic interactions may be described as entropybased forces When a nonpolar compound is dissolved in water, the strong water-water interactions around the solute lead to an effective "ordering" of the structure of the solvent This is entropically unfavorable; that is, there is negative entropy of dissolution When a nonpolar inhibitor binds to a nonpolar region of an enzyme, all the ordered water molecules become less ordered as they associate with bulk solvent, leading to an increase in entropy According to Equation 17.4 any increase in entropy will lead to a decrease in free energy and, through Equation 17.3, stabilization of the enzyme-inhibitor complex It has been calculated that a single methylene-methylene interaction releases about 0.7 kcdmol of free energy Even though this figure is not high, given that enzymes and inhibitors usually have large regions of hydrophobic surface, this type of bonding may also play a significant role in inhibitor binding Overall, the total number of hydrogen bonds remains constant and, provided that the hydrogen bonds between the inhibitor and enzyme are not significantly more favorable than those between water and the inhibitor or those between water and the enzyme, the net change in enthalpy is usually insignificant On the other hand, formation of the enzyme-inhibitor complex usually leads to an overall'increase of entropy because the inhibitor remains bound to the enzyme and the formerly bound water molecules are released 2.1.4 Hydrogen Bonds A hydrogen bond 2.1.5 Cation-.rr Bonding Recently it has occurs when a proton is shared between two electronegative atoms e , -X-H .Y) Electron density is pulled from the hydrogen by X, giving the hydrogen a partial positive charge that is strongly attracted to the nonbonded electrons of Y The bond is usually asymmetric, with one of the heteroatoms, the hydrogen bond donor, having a normal covalent bond distance to the proton The other heteroatom, the hydrogen bond acceptor, is usually at a distance somewhat shorter than the van der Wads contact distance and, for optimal hydrogen bonding, the atoms should be arranged linearly A hydrogen bond is a special type of dipole-dipole interaction and, as we have seen, although these forces can be quite become apparent that there is another important noncovalent binding force that may be exploited when designing enzyme inhibitors Cations, from simple ions such as Lit to more complex organic molecules such as acetylcholine, are strongly attracted to the electronrich (T) face of benzene and other aromatic compounds (61,62).Cation-T bonds, as well as other amino-aromatic interactions, are common in structures in the protein data bank (631, and it has been estimated that more than 25% of tryptophan residues are involved in interactions of this type (64) The finding that the cationic group of acetylcholine was bound primarily by aromatic residues, most especially by a tryptophan residue, not by the ex- Rational Design of Noncovalently Binding Enzyme Inhibitors pected carboxylate anion, provided evidence that cation-.rrinteractions may play an important role in ligand binding (65,66) Model systems suggest that, energetically, the cation-.rr interaction can compete with full aqueous solvation in binding cations (61), and there is now significant effort being expended in studying the contribution of these interactions to molecular recognition (62,661 In summary, the Ki provides an indication of the relative stability of the enzyme-inhibitor complex compared to stability of the enzyme and inhibitor free in solution Moreover, it is clear that entropy, enthalpy, and water will all have a major impact on the binding of an inhibitor to an enzyme 2.2 As can be seen from the following discussion, it is not difficult to carry out a kinetic analysis of a single-substrate reaction such as that described in Equation 17.8 However, as more substrates are added the task becomes more complex Fortunately, kinetic analysis of enzymatic reactions involving two or more substrates can be made easier by varying the concentration of only one substrate at a time By keeping all but one of the substrates at fixed, saturating concentrations, the reaction rate will depend only on the concentration of the varied substrate This permits the use of the kinetic analysis employed for enzyme-catalyzed, single-substrate reactions even for complex multisubstrate reactions In a further simplification, the dissociation of the E P complex is assumed not to be rate limiting, and the reversion of product to substrate is assumed to be negligible The latter assumption is valid under what are known as initial velocity conditions, that is, when less than about 5%of substrate has been consumed Under these conditions, the concentration of P is low, and Equation 17.8 simplifies to Equation 17.9 Steady-State Enzyme Kinetics Just as an appreciation of the forces involved is essential to comprehending the binding of an inhibitor to an enzyme, so is an understanding of the kinetic analysis of an enzyrnecatalyzed reaction essential to any kinetic evaluation of an inhibitor In this section we provide a brief introduction to the study of enzyme kinetics, particularly steady-state kinetics Regardless, the reader is advised to refer to other sources for more in-depth reviews of the kinetic equations and mathematical derivations involved (38, 60, 67-71) 2.2.1 The Michaelis-Menten Equation In the simplest case, an enzyme-catalyzed reaction involves the conversion of a single substrate to a single product, as shown in Equation 17.8 Generally, kinetic analyses are carried out by studying the reaction under steadystate conditions, that is, when the concentration of the enzyme is well below that of the substrate Under those circumstances, following a brief preequilibrium period, the concentrations of the various enzyme-bound species, E S and E P in Equation 17.8, become effectively constant and the rate of conversion of substrate to product will greatly exceed the change in concentration of any enzyme species This is an approximation but, provided the substrate concentration does not greatly change (e.g., under initial velocity conditions), it is a very useful approximation Given steady-state conditions, t h e Michaelis-Menten equation (Equation 17.10) is a quantitative description of the reaction described by Equation 17.9 The free enzyme (E) binds the substrate (S) to form a noncovalent enzyme-inhibitor complex (E S) This is assumed to be a rapid, reversible process, not involving any chemical changes, and with the affinity of the substrate for the enzyme's active site being determined by the binding forces discussed above A chemical transformation of substrate to product (P), initially in complex with enzyme (E P), then takes place Finally, the product (P) is released into the medium with concomitant regeneration of free enzyme (E) e Approaches to the Rational Design of Enzyme Inhibitors The Michaelis-Menten constant KM is a combination of rate constants and is independent of enzyme concentration under steadystate conditions It is equal to the substrate concentration at which half the maximum velocity of the enzyme-catalyzed reaction is reached; that is, when [S] = KM, then v = ?hV , For the reaction illustrated in Equation 17.9, KMis described by Equation 17.11 Figure 17.2 Plot showing dependency of the initial velocity (v)on substrate concentration [S] for an enzyme-catalyzed reaction obeying Michaelis-Menten (saturation) kinetics This implies that the initial velocity (v) is directly proportional to the enzyme concentration [El, and that v follows saturation kinetics with respect to the substrate concentration [S] This is shown graphically in Figure 17.2 and explained as follows: at very low substrate concentrations v increases in a linear fashion, so that v = Vm,[SI/KM As the substrate concentration increases, the observed increase in v is less than the increase in [Sl This trend continues until, at high (saturating) substrate concentrations, u becomes effectively independent of [S] and tends toward the limiting value V., vm, is the maximal velocity that can be achieved at a specific enzyme concentration In the simple Michaelis-Menten mechanism described by Equation 17.9, there is only one E S complex and all binding steps are rapid In this instance, V,, is the product of the enzyme concentration [El and k, (also known as k,,), which is the first-order rate constant for the chemical conversion of the E S complex to free enzyme and product The catalytic constant k, is often referred to as the turnover number because it represents the maximum number of substrate molecules converted to products per active site per unit time In a more complicated reaction, k, is a function of all the first-order rate constants and, effectively, sets a lower limit on all the chemical rate constants * k,, then If, for a given reaction, k-, Equation 17.11 simplifies to KM= Ks, where Ks is the dissociation constant for the enzyme substrate complex It is important to remember that the Michaelis-Menten equation holds true not only for the mechanism as stated above, but for many different mechanisms that are not included in this treatment In summary, KMcan be described as an apparent dissociation constant for all enzyme-bound species and, in all cases, it is the substrate concentration at which the enzyme operates at half-maximal velocity Another parameter often referred to when discussing Michaelis-Menten kinetics is k,,/ KM.This is an apparent second-order rate constant that relates the reaction rate to the free (not total) enzyme concentration As described above, at very low substrate concentrations when the enzyme is predominantly unbound, the velocity (v) is equal to [El [Slk,,I KM.The value of k, JKM sets a lower limit on the rate constant for the association of enzyme and substrate It is sometimes referred to as the specificity constant because it determines the specificity of the enzyme for competing substrates Again, for more detailed treatment of this subject the reader should refer to more specialized texts (38, 60, 67-69) 2.2.2 Treatment of Kinetic Data Analysis of Michaelis-Menten kinetics is greatly facilitated by a linear representation of the data Converting the Michaelis-Menten Equation 17.10 into Equation 17.12 leads to the popular Lineweaver-Burk plot 727 Rational Design of Noncovalently Binding Enzyme Inhibitors Figure 17.3 The Lineweaver-Burk plot Figure 17.4 The Eadie-Hofstee plot If you plot llv against 1/[S1(Fig 17.31,the y-intercept gives a value of l/Vm, and the xintercept gives a value of - l/KM.The slope of the line is equal to KM/Vm, Although very popular, the Lineweaver-Burk plot suffers from the disadvantage that it emphasizes points at lower concentrations and compresses data points obtained at high concentrations (67) As a result it is not recornmended for obtaining accurate kinetic constants A preferable, alternative form of the Michaelis-Menten equation is that of the Eadie-Hofstee plot (Equation 17.13) As shown in Fig 17.4, plotting v against v/[S] results in the y-intercept providing a value of V, whereas the x-intercept provides Vm,/KM, and the slope of the line is equal to -KM Finally, it is possible to directly plot pairs of v, [S] data in such a way as to directly determine KMand V ,, values Of the linear graphical methods, the direct linear plot of Eisenthal and Cornish-Bowden (721, shown in Fig 17.6, is often considered to provide the best estimates of KM and V,, values In this method pairs of v and [Sl values are obtained in the usual manner A v value is plotted on the y-axis and a corresponding negative value of [S] is plotted on the x-axis A straight line is then drawn, passing through the points on the two axes and extending beyond the "point of intersection." This is repeated for each set of u and [S] values Thus, there are n sets of lines for n pairs of values A horizontal line drawn from the point of intersection to the y-axis provides the V, value, whereas a vertical line from the point of intersection to thex-axis provides the KM value Each of these linear plots has its own merits, particularly for plotting inhibition data Another linear representation of the Michaelis-Menten equation is the HanesWoolf plot (Equation 17.14) Thus a plot of [S]/v vs [S] is linear, with a slope of l/Vm, (Fig 17.5) The y-intercept gives KM/Vm, and the x-intercept gives -KM Figure 17.6 The Hanes-Woolf plot Approaches to the Rational Design of Enzyme lnhibitors Michaelis-Menten kinetics and, depending on their preference of binding to the free enzyme and/or the enzyme-substrate complex, competitive, uncompetitive, and noncompetitive inhibition patterns can be distinguished For the purposes of this discussion it will be assumed that the initial equilibrium of free and bound substrate is established significantly faster than the rate of the chemical transformation of substrate to product, that is, k,, kk, k , (Equation 17.9) As discussed in section 2.2.1, this reduces K, to the dissociation constant K, of the E S complex 2.3.1.1 Competitive Inhibitors A competitive inhibitor often has structural features similar to those of the substrates whose reactions they inhibit This means that a competitive inhibitor and enzyme's substrate are in direct competition for the same binding site on the enzyme Consequently, binding of the substrate and the inhibitor are mutually exclusive A kinetic scheme for competitive inhibition is shown in Equation 17.15 * Figure 17.6 Eisenthal-Cornish-Bowdendirect linear plot of enzyme kinetic data fitting the MichaelisMenten equation (38), and its own drawbacks (67) However, the rapid advances in personal computing make it relatively easy to fit kinetic data to the Michaelis-Menten equation (or other appropriate hyperbolic functions) by use of a variety of commercial graphical or spreadsheet packages One simple package, HYPER, which is readily available on the Internet (http:/t www.ibiblio.org/pub/academic/biology/molbio/ ibmpc/hyperl02.zip), simultaneously furnishes Michaelis-Menten parameters obtained using hyperbolic regression analysis, as well as those obtained using three of the plots described here As such, it provides a rapid contrast of these graphical methods but, unfortunately, is not suitable for the study of inhibition kinetics In addition, the recent monograph by Copeland (71) provides a list of useful computer software and Internet sites for the study of enzymes 2.3 Rapid, Reversible lnhibitors This class of inhibitors acts by binding to the target enzyme's active site in a rapid, reversible, and noncovalent fashion The net result is that the active site is blocked and the substrate is prevented from binding Accordingly, in designing inhibitors of this type, optimization of the noncovalent binding forces between the inhibitor and the active site of the enzyme is of paramount importance 2.3.1 Types of Rapid, Reversible Inhibitors Binding of these inhibitors follows simple The enzyme-bound inhibitor may either lack an appropriate functional group for further reaction, or may be bound in the wrong position with respect to the catalytic residues or to other substrates In any event, the enzyme-inhibitor complex E I is unreactive (it is sometimes referred to as a dead-end complex) and the inhibitor must dissociate and substrate bind before reaction can take place Solving this kinetic scheme for simple Michaelis-Menten kinetics leads to Equation 17.16 Here, Ki, sometimes called the inhibition constant, is the equilibrium constant for the Rational Design of Noncovalently Binding Enzyme Inhibitors dissociation of the enzyme-inhibitor complex, and is described by Equation 17.17 [EI[II K = - ‘ (17.17) [E-I] Competitive inhibitors not change the value of V, which is reached when sufficiently high concentrations of the substrate are present so as to completely displace the inhibitor However, the affinity of the substrate for the enzyme appears to be decreased in the presence of a competitive inhibitor This happens because t h e free enzyme E is not only in equilibrium with the enzyme-substrate complex E S, but also with the enzyme-inhibitor complex E I Competitive inhibitors increase the apparent KM of the substrate by a factor of (1 + [Il/Ki) The evaluation of the kinetics is again greatly facilitated by the conversion of Equation 17.15 into a linear form using Lineweaver-Burk, Eadie-Hofstee, or Hanes-Woolf plots, as shown in Fig 17.7 2.3.1.2 Uncompetitive Inhibitors Uncompetitive inhibitors not bind to the free enzyme They bind only to the enzyme-substrate complex to yield an inactive E S I complex (Equation 17.18) / 1/PI (b) v \\k \ vml (C) IS1 Figure 17.7 (a) Lineweaver-Burk, (b) Eadie-Hof- S E E S + E + P A It I (a) (17.18) stee, and (c) Hanes-Woolf plots exhibiting competitive inhibition patterns The dashed line indicates the reaction in the absence of inhibitor, whereas the solid lines represent enzymatic reactions in the presence of increasing concentrations of inhibitor E.S.1 Uncompetitive inhibition is rarely observed in single-substrate reactions but is frequently observed in multisubstrate reactions An uncompetitive inhibitor can provide information about the order of binding of the different substrates In a bisubstratecatalyzed reaction, for example, a given inhibitor may be competitive with respect to one of the two substrates and uncompetitive with respect to the other The linear plots for classical uncompetitive inhibition patterns are described by Equation 17.19 and are illustrated in Fig 17.8 As with a competitive inhibitor, the apparent KMfor the substrate decreases by a factor of (1+ [I]/Ki)because the formation of E S I will use up some of the E S, thereby shifting the equilibrium further in favor of E S formation However, uncompetitive inhibitors also decrease V,, by the same factor because Approaches to the Rational Design of Enzyme Inhibitors Simple Michaelis-Menten kinetics of noncompetitive inhibitors are described in Equation 17.21 From Equation 17.21 it is clear that noncompetitive inhibitors have an effect only on V,, decreasing it by a factor of (1 + [IIIKJ, consequently giving the impression of reducing the total amount of enzyme present As with an uncompetitive inhibitor, a portion of the enzyme will always be bound in the nonproductive enzyme-substrate-inhibitor complex E S I, causing a decrease in maximum velocity, even at infinite substrate concentrations However, because noncompetitive inhibitors not affect substrate binding, the KMvalue of the substrate remains unchanged Linear plots for noncompetitive inhibition are shown in Fig 17.9 Again, this type of inhibition is rarely seen in single-substrate reactions It should also be noted that, frequently, the affinity of the noncompetitive inhibitor for the free enzyme, and the enzyme-substrate complex, are different These nonideally behaving noncompetitive inhibitors are called mixed-type inhibitors, and but also KMfor the they alter not only V,, substrate Further discussion of inhibitors of this type may be found in Segel(38) Sometimes steady-state kinetics are insufficient to analyze the mechanism of inactivation for a given inhibitor For example, irreversible enzyme inhibitors that bind so tightly to the enzyme that their dissociation rate (kOfl) is effectively zero also exhibit noncompetitive inhibition patterns They act by destroying a portion of the enzyme through irreversible binding, thereby lowering the overall enzyme The apconcentration and decreasing V., parent KM remains unaffected because irree Figure 17.8 (a) Lineweaver-Burk, (b) Eadie-Hofstee, and (c) Hanes-Woolf plots exhibiting uncompetitive inhibition patterns The dashed line indicates the reaction in the absence of inhibitor, whereas the solid lines represent enzymatic reactions in the presence of increasing concentrationsof inhibitor some of the enzyme remains in the E S I form, even at infinite substrate concentration 2.3.1.3 Noncompetitive Inhibitors Classical noncompetitive inhibitors have no effect on substrate binding and vice versa, given that they bind randomly and reversibly to different sites on the enzyme They also bind with the same affinity to the free enzyme and to the enzymesubstrate complex Both the enzyme- inhibitor complex E I and the enzyme-substrate-inhibitor complex E S I are catalytically inactive The equilibria are outlined in Equation 17.20 - Rational Design of Noncovalently Binding Enzyme Inhibitors E I + noncompetitive I + irreversible inhibitor Figure 17.10 Plot showing dependency of V,, on the total enzyme concentration,[El,,,, An irreversible inhibitor will titrate a fraction of the enzyme [Elinact situations Such analyses may include more in-depth steady-state kinetics, as well as presteady-state kinetics, and testing for irreversible inhibition Irreversible covalently binding enzyme inhibitors are discussed extensively later in this chapter 2.3.2 Dixon Plots Another linear method Figure 17.9 (a) Lineweaver-Burk, (b) Eadie-Hofstee, and (c) Hanes-Woolf plots exhibiting noncompetitive inhibition patterns The dashed line indicates the reaction in the absence of inhibitor, whereas the solid lines represent enzymatic reactions in the presence of increasing concentrations of inhibitor versible inhibitors not influence the dissociation constant of the enzyme-substrate complex A simple experiment to distinguish between a reversible noncompetitive inhibitor and irreversible inhibitor is shown in Fig 17.10, and a comprehensive review describing the kinetic evaluation of irreversibly binding enzyme inhibitors is available (73) Allosteric effectors may also show noncompetitive kinetic patterns by rendering the enzyme in the E S I complex less active than that in the E S complex Again, additional analyses are often required in these less well defined - for plotting inhibition data, the Dixon plot, is shown in Fig 17.11 (74) In this method the initial velocity is measured as a function of inhibitor concentration at two or more fixed substrate concentrations By plotting llv against [I] for each substrate concentration, the different types of inhibition can easily be distinguished Further, in cases of competitive or noncompetitive inhibition, the value of Ki may be determined from the x-axis value at which the lines intercept Overall, the Dixon plot is probably the simplest and most rapid graphical method for obtaining a Ki value 2.3.3 IC,, Values The potencies of en- zyme inhibitors evaluated using rapid screening techniques are often reported in terms of IC,, values rather than Ki values An value is the inhibitor concentration that is required to halve the activity of the enzyme, that is, that concentration that leads to 50% enzyme inactivation It is important to recognize that an IC,, value is not a constant, except in the case of noncompetitive inhibition, and is dependent on the substrate concentration used in the experiment IC,, values are com- Approaches to the Rational Design of Enzyme Inhibitors hibition It should also be noted that the IC,, value can be no less than half the concentration of the enzyme, a factor that becomes important if the inhibitor is very potent or if high concentrations of enzyme are employed For a competitive inhibitor, a Ki value may be obtained using the relationship described by Equation 17.22 IC,, = ~ g) ~+ ( (17.22) Provided that a reasonable substrate concentration ( K,) is employed for the experiment, the IC,, value may be a reasonable approximation of the true K, Equation 17.22 indicates that substrate concentrations greater than about 0.1-fold of the KM value will lead to an underestimation of the Ki value, an underestimation that becomes quite significant at high substrate concentrations The dependency of the IC,, value on the substrate concentration for uncompetitive inhibitors is given in Equation 17.23 IC,, = ~ Figure 17.11 Dixon plots for (a) competitive, (b) uncompetitive, and (c) noncompetitive inhibitors The solid lines represent enzymatic reactions in the presence of increasing concentrations of substrate The dashed line represents the reaction at infinite substrate concentration monly determined by keeping the concentration of the substrate and the enzyme constant and incrementally varying the concentration of the inhibitor This simple experimental approach makes it relatively easy to screen large numbers of potential inhibitors Industrial high-throughput screens often employ halflog increments, and the value of IC,, provides a ready means of comparing the extent of in- s) ~+ ( (17.23) In this instance it is at high concentrations of the substrate that the Ki value is comparable to the IC,, value, and a significant underestimation will occur at lower substrate concentrations From these two equations it is clear that, for preliminary screening when the type of in- hibition is unknown, substrate concentrations close to the KM value should be used This minimizes the deviation of the IC,, value from the Ki value to, in the cases of competitive and uncompetitive inhibitors, a factor of If necessary, a Dixon plot can be used to provide a quick indication of the Ki and the type of inhibition (38, ) It should be noted that the relationship between IC,, and Ki requires the initial velocity to be linearly dependent on the concentration of inhibitor In the cases of mixed-competitive and irreversible inhibitors, the dependency of the inhibitor concentration and the initial velocity is nonlinear Therefore, in those cases, the use of the IC,, value is limited 2 Rational Design of Noncovalently Binding Enzyme Inhibitors 2.3.4 Examples of Rapid Reversible Inhibitors Competitive inhibitors are often similar in structure to one of the substrates of the reaction they are inhibiting Inhibitors of this type are sometimes called substrate analogs usually approxand their binding affmity (K,) imates that of the substrate One of the first reactions inhibited by a substrate analog was that catalyzed by succinate dehydrogenase (Equation 17.24) - 02C-CH2-CH2-C02 succinate - succinate dehydrogenase fumarate This reaction is competitively inhibited by malonate (-00CCH2C00-) that has, like succinate, two carboxylate groups It is therefore able to bind to the enzyme's active site but, with only one carbon atom between the carboxylates, further reaction is impossible Substrate analogs are rarely useful as enzyme inhibitors, given that large concentra- Norepinephrine PNMT Epinephrine 733 tions are required for inhibition, and their inhibition is readily overcome by any buildup of substrate However, they are often useful probes for determining enzyme specificity and even mechanism Phenylethanolamine N-methyltransferase (PNMT) catalyzes the terminal step in epinephrine (adrenaline) biosynthesis, the conversion of norepinephrine to epinephrine (Equation 17.25), with concomitant conversion of S-adenosyl-L-methionine (SAM, AdoMet) to S-adenosyl-L-homocysteine (SAH, AdoHcy) S-Adenosyl-L-homocysteine (10) (Fig 17.12),the product of the reaction, and 2-(2,5dichlorophenyl)cyclopropylamine (11)are analogs of S-adenosyl-L-methionine and norepinephrine, respectively Using these inhibitors it was possible to ascertain the binding order of the two substrates (75) Kinetic analyses showed that SAH was a competitive inhibitor of SAM and a noncompetitive inhibitor of norepinephrine, whereas (11)was a competitive inhibitor of norepinephrine and an uncompetitive inhibitor of SAM This indicates that the binding of substrates is ordered, with SAM binding first If norepinephrine bound first, it would be expected that SAH would be an uncompetitive inhibitor and (11)would be noncompetitive with respect to SAM If a random Approaches to the Rational Design of Enzyme Inhibitors Figure 17.12 Inhibitors of phenylethanolamine N-methyltransferase binding mechanism were in operation, it would be expected that both inhibitors would be competitive with either substrate More detail on similar uses of reversible inhibitors may be found elsewhere (76) 2.4 Slow-, Tight-, and Slow-Tight-Binding Inhibitors Not all reversible inhibitors have an instantaneous effect on the rate of an enzymatic reaction Some inhibitors, known as slow-binding enzyme inhibitors, can take a considerable time to establish the equilibrium between the free enzyme and inhibitor, and the enzymeinhibitor complex This time period may be on the scale of seconds, minutes, or even longer The enzyme-inhibitor complexes have slow off (dissociation) rates, but the on (association) rates may be either slow or fast Hence, the term slow binding does not necessarily indicate a slow binding of inhibitor to enzyme but rather the fact that reaching equilibrium is a slow process Other inhibitors, known as tight-binding inhibitors, bind their target enzyme with such high affinity that the population of free inhibitor molecules is significantly depleted when the enzyme-inhibitor complex is formed Often, tight-binding inhibitors also have a slow onset of action, and are termed slow-tight-binding inhibitors What these three types of inhibitors have in common is that, generally, the major assumptions of Michaelis-Menten kinetics not hold true As with rapid reversible inhibition, for slow-binding inhibition to take place a significantly larger concentration of inhibitor than enzyme is required However, reaching equilibrium slowly is incompatible with the assumption of Michaelis-Menten kinetics that inhibitors bind much more quickly than the enzyme turns over Unlike rapid reversible and slow-binding inhibitors, both tight-binding and slow-tight-binding inhibitors are effective at concentrations comparable to that of the enzyme At that point, the inhibitor concentration is no longer independent of the enzyme concentration, as assumed for Michaelis-Menten kinetics A summary of the properties of reversible enzyme inhibitors is shown in Table 17.5 Although we give a brief overview of these types of inhibitors, excellent and more in-depth descriptions of slow-, tight-, and slow-tight-binding inhibitors have appeared elsewhere (71, 77-80) 2.4.1 Slow-Binding Inhibitors Two differ- ent mechanisms have been suggested to rationalize the slow-binding behavior of competitive inhibitors (71, 78, 80) In the one-step mechanism A, the direct binding process of the inhibitor to the enzyme is slow (Equation 17.26); that is, the magnitude of k,[I] is small relative to k,[S] and k,, the rate constants for the conversion of substrate to product Table 17.5 Classes of Reversible Inhibitors Inhibitor Class Rapid, reversible Tight binding Slow binding Slow-tight binding Ratio of Inhibitor to Enzyme Necessary for Inhibition Rate at Which Equilibrium is Attained E + I E I %E 1.-E 1%-E Fast Fast Slow Slow I- E Rational Design of Noncovalently Binding Enzyme Inhibitors 735 For slow-tight-binding inhibitors, k-, is very small and formation of the E I complex is essentially irreversible Use of Equation 17.28 ensures that depletion of free enzyme and free inhibitor by formation of the E I complex is taken into account In mechanism B, the more common mechanism for slow-binding inhibition (go),the initial equilibrium between the enzyme, inhibitor, and the E I complex is fast However, there is a subsequent slow rearrangement to form the final, more stable enzyme-inhibitor complex (E I*) (Equation 17.29) The slow on rate (k,) has been attributed to the inhibitor encountering some barrier to binding a t the active site The inhibitor has to overcome this barrier by correct alignment Once aligned properly, it binds so tightly that it is released very slowly from the enzyme, making the overall equilibrium process extremely slow The equilibrium dissociation constant for the E I complex Ki, derived from Equation 17.26, is given by Equation 17.27 This is the same equilibrium as that for a rapid reversible inhibitor (Equation 17.17) From Equation 17.27, it should be noted that, if Ki is very small (as with a tight-binding inhibitor) and [I] is varied in the region of Ki, even if the on rate (k,) is diffusion controlled, both k,[I] and k-, will be very small Thus, the onset of inhibition for a tight-binding inhibitor can appear to be slow, even though k, is in the range expected for rapid reversible inhibitors (78) It is possible to carry out kinetic analyses of tight-binding inhibitors This can be done either by including a preincubation step, to allow sufficient time for the enzyme and inhibitor to reach equilibrium, or by carrying out the reaction at very high concentrations of both substrate and inhibitor More detailed discussion of these methods, with appropriate references, can be found in a recent volume by Copeland (71) If the slow-binding inhibitor described by Equation 17.26 also binds very tightly, it is referred to as a slow-tight-binding inhibitor For inhibitors of this type, Ki is given by Equation 17.28, where [E,] represents the total enzyme concentration (in all forms) present in solution Here the dissociation constant for the initial E I complex is still k-,lk,, but there is also a dissociation constant for the formation of the E I* complex The second dissociation constant is given by Equation 17.30 - To observe the slow onset of inhibition and the E I complex, Ki* must be smaller than Ki and k-, smaller than k, However, if k-, is considerably smaller than k,, then the formation of the E I* complex will be effectively irreversible (i.e., the inhibitor is of the slowtight-binding variety) Under those circumstances it will again be necessary to take depletion of free enzyme and free inhibitor into account when determining Ki and Ki* (78) The slow rearrangement step has been correlated with conformational changes of the enzyme following initial binding of the inhibitor It is possible that the enzyme in its transition state conformation may be better equipped to Approaches to the Rational Design of Enzyme inhibitors A good comparison of rapid reversible and [I] =0 Increasing [I] slow-binding inhibition can be found in a recent study on the inhibition of arginase, an enzyme that catalyzes the hydrolysis of L-arginine to yield L-ornithineand urea (Equation 17.31) Time Figure 17.13 Reaction progress curves in the presence of increasing concentrations of a slowbinding inhibitor accommodate the inhibitor A slow change to reach this optimal conformation will lead to tighter binding of the inhibitor and even slower release from the enzyme An alternative suggestion is that the slow-binding process is linked to a requisite displacement of water molecules from the active site (81).Initially the inhibitor binds loosely to the enzyme, but upon release of water molecules the gain in entropy leads to a more stable E I* complex One way of quickly identifying a potential slow-binding inhibitor is to examine the progress of the reaction at increasing concentrations of inhibitor Under initial velocity conditions (Section 2.2.11, an enzyme-catalyzed reaction will exhibit a linear increase in the amount of product formed over time A reaction progress plot for a reaction carried out in the presence of a rapid reversible inhibitor will also be linear However, a slow-binding inhibitor will initially show a linear relationship, although this will change as the inhibitor binds, resulting in a biphasic plot Typical biphasic progress curves for a reaction in the presence of increasing concentrations of a slow-binding inhibitor are shown in Fig 17.13 The initial burst of the reaction, the linear section of the graphs, can be described by competitive Michaelis-Menten kinetics The higher the concentration of the inhibitor, the shorter the initial linear section of each curve and the slower the subsequent final steady-state rate, as observed in the asymptotes in Fig 17.13 If the inhibitor concentration is small, the substrate might be too depleted to permit observation of steady-state rates Urea Arginase competes with nitric oxide synthase (NOS) for arginine and, in doing so, helps regulate NOS As a consequence, inhibitors of arginase may have therapeutic use in treating NO-dependent smooth muscle disorders, including erectile dysfunction (82) A series of arginine analogs were prepared and tested as inhibitors of arginase (83) Three examples are shown in Fig 17.14 One of these, Nu-hydroxy-L-arginine(12), is a competitive inhibitor of arginase at both pH 7.5 and pH 9.5 with Ki values of and 1.6 pM, respectively The two boronic acid derivatives, 2(S)-amino6-boronohexanoic acid (13) and S-(2-boronoethyl)-L-cysteine (14), were also competitive inhibitors a t pH 7.5 with Ki values of 0.25 and 0.31 pM, respectively However, at pH 9.5, the boronic acid derivatives both became slowbinding inhibitors, apparently binding by mechanism B and with lowered Ki values of 8.5 and 30 nM, respectively It was suggested that, at low pH, the trigonal form of the boronic acid derivative predominates, and that Rational Design of Noncovalently Binding Enzyme inhibitors (14) (16) Figure 17.14 (a) Competitive and (b) slow-binding inhibitors of arginase this species binds with one hydroxyl, coordinating to one of the two requisite manganese ions At pH 9.5 the tetrahedral species is the major form and this initially binds also with one hydroxyl coordinated to a manganese ion Then, in a second, slower step, a water molecule that bridges the two active-site manganese ions is displaced by a second hydroxyl group on the boronic acid (83) Support for this mechanism is provided by crystal structures, showing both (15) and (16)are bound in leueine aminopeptidase the active site of arginase as tetrahedral species at alkaline pH (82) Of course, compound (12) is unable to form the tetrahedral species and is a competitive inhibitor at all times Leucine arninopeptidase (LAP) is a metalloenzyme that has been inhibited in a slowbinding manner This exopeptidase catalyzes the hydrolysis of N-terminal amino acids, particularly those with a leucine at the N-terminus, although it does have a broad specificity (Equation 17.32) Approaches to the Rational Design of Enzyme Inhibitors 738 (17) (18) Figure 17.15 Slow-tight-binding inhibitors of leucine aminopeptidase Bestatin (17) (Fig 17.15) and amastatin (18)have been identified as slow-tight-binding inhibitors of LAP from porcine kidney, with Ki values in the low nanomolar range (84) Later, bestatin was shown to be a slow-bindinginhibitor of LAP employing mechanism B, with a Ki value of 0.11 pit4 and a Ki*value of 1.3 nM Values of 1.5 X lo-' s-' and X lo-, s-' were obtained fork, and k-, (Equation 17.291, respectively (85) It was assumed that the inhibition of bovine lens leucine aminopeptidase (blLAP) by amastatin would also proceed by mechanism B This prediction was supported by an X-ray crystallography study of the amastatin-blLAP complex (86), which suggested that (18)(and, by analogy, 17) initially binds to a Zn2+atom in a groove in the active site The slow step in binding was seen as a subsequent coordination to a second Zn2+ atom located deeper in the active site (86) It is difficult to find clear-cut examples of slow-binding inhibition occurring by mechanism A However, the inhibition of Factor Xa by a peptidyl-a-ketothiazole was found to be unusual because it appeared that the formation of E I was partially rate limiting Factor Xa is a trypsinlike protease found in the blood coagulation pathway, which cleaves prothrombin forming thrombin that, in turn, promotes blood clotting (Equation 17.33) Inhibitors of Factor Xa activity offer potential as anticoagulants and several irreversible inhibitors of Factor Xa have been developed One of the few tight-binding reversible inhibitors of Factor Xa is BnS02-D-kg-Gly-kg-ketothiazole (19) The inhibitor could be displaced from Factor Xa by substrates and, based on steadystate assumptions, the dissociation constant for (19) was found to be 14 pM (87) However, the reaction progress curves indicated a slowbinding process, probably by mechanism B Stopped-flow fluorescence studies, combined with kinetic analysis, showed that the isomerization step (E I + E I*) is unusually fast and that the formation of E I is, at least, partially rate limiting In some instances the type of inhibition has been found to be isozyme specific For example, inducibly expressed isozymes (iNOS) and constitutively expressed isozymes (cNOS) of nitric oxide synthase (NOS) all catalyze the conversion of L-arginine to L-citrulline and nitric oxide (Equation 17.34) 739 Rational Design of Noncovalently Binding Enzyme Inhibitors 0 II 0 II 0 II II II II aHN-CH-C-NH-CH-C-NH-CH2-C-NH-CH-C-NH-CH-C-NH-CH-C-NH~ I I I I I CH(CH3) CH2 CH2 CH(CH3) CH2 I I I I I CH2 CH2 OH CH2 CH2 I I I I CH3 COOH CH2 CH3 I NH Factor Xa 0 0 II II + I1 II II aHN-CH-C-NH-CH-C-NH-CH2-C-NH-CH-COz-+H3N-CH-C-NH-CH-C-NHa I I I I I CH(CH3) I CH2 I CH3 II CH2 I CH2 I COOH CH2 I CH2 CH(CH3) I CH2 CH2 I NH CH3 I I CH2 I OH I fb nitric oxide The inhibition of human iNOS by N43(aminomethyl)benzyl)acetamidine (20) (Fig 17.16) was found to proceed by mechanism B, with an overall Kdof 95% optical purity It has been shown that (-)-Clenbuterol was 100-1000 times more -potent than (+)-Clenbuterol in P-adrenergic agonist bioassays (34) A number of 1,6dihydropyridines (17-20), exhibiting axial chirality (chiralty stemming from the nonplanar arrangement of four groups about an axis), have been separated by small-scale HPLC methods This is an impor- tant class of drugs that are potent blockers of calcium currents and have found use in the treatment of cardiac arrhythmias, peripheral vascular disorders, and hypertension (35) It has been shown that enantiomers of chiral DHP have opposite pharmacological profiles (35) One of the antipodes is a calcium entry activator, while the other is a calcium entry blocker The analytical and semi-preparative separation using chiral HPLC for a number of DHPs of the structures (Fig 18.7) has been described (36) Here a number of different CSP were utilized and their ability to separate the above DHPs determined 2.2 Chromatographic Diastereoisomer Separation Another approach to the separation of enantiomers by chromatography is to prepare a diastereoisomer of the enantiomer to be separated As discussed in the introduction to this chapter, diastereomers exist if there is more than one chiral center, but are not enantiomers of one another As such they not have identical physical properties In chromatography, formation of derivatives such as esters, amides, etc., often leads to better separation of the components In the case of a racemate, if a chiral reagent (i.e., acid or m i n e ) is employed, then a diastereomeric mixture results on treatment with such a derivatizing agent One such example is the derivatization of Pirlindole, which is a racemic anti-depressant drug Here the use of amino acid derivatives as chiral derivatizing agents (CDA) was shown to enable an effective and efficient separation (37) Preparation of the L-phenylalanine methyl ester (21) enabled separation of the Pirlindole enantiomers using a medium liquid pressure (MPLC) method This is highlighted in Fig 18.8, after removal of the CDA the enantiomers of -virlindole were obtained in high optical purity This gave several grams of each enantiomer, which permitted a study of the stereochemical influence at the pharmacological level The interaction with monoamine oxidase A (MAO-A) and B (MAO-B) with Pirlindole racemate and single enantiomers using biochemical techniques (in vitro and ex vivo determination of rat brain MAO-A and hL4O-B activity) was studied In vitro, the MAO-A IC,, of (+)-Pirlindole, R-(-)-Pidin- Chromatographic Separations H DHP (17) I H DHP (19) A DHP (18) I H DHP (20) dole (22), and S-(+)-Pirlindole (23) were 0.24, 0.43, and 0.18 pM, respectively The differences between the three compounds were not significant, with a ratio between the two enantiomers R-(-)IS-(+)of 2.2 in vitro (38) 2.3 Preparative HPLC/SMB In the initial discovery phase of drug research, time is the most important factor where a successful process must be rapidly identified, have a short run time, and have general applicability As the phase of the project changes to full development, the process needs to be established and cost becomes a crucial factor Thus, on scale up of an LC method to the preparative level (100 mg and above), a number of additional important aspects become relevant The selection of a suitable CSP from the plethora available depends on the following factors: CSP availability, loading capacity and selectivity, throughput, and mobile phase The most successful and broadly applied chiral stationary phases comprise the cellulose-and amylose-based phases developed by Okamoto (Chiracel and Chiralpak) (39), brush-type phases developed by Pirkle (40), Figure 18.7 some polyacrylamides (Chiraspher) (4I), cross-linked diallyltartramide (42), and to a lesser extent, cyclo-dextrin based phases Clearly for the larger scale separations, the availability of the CSP in larger quantities is a prerequisite It should also be noted that at the preparative scale, it seems that up to 90% of racemic compounds tested have been resolved with just four different polysaccharidebased phases (43) The degree of separation of the two enantiomers obviously plays an important part in the CSP selection Another equally important parameter is the loading capacity of the stationary phase The higher the loading capacity, the greater the amount of material that can be separated (44) For example the polysaccharide-based CSPs have a saturation capacity of 5-100 mg/g of CSP; this is clearly dependent on the type of racemate that is beingresolved On the other hand, protein-based CSPs have lower saturation capacities, of the order 0.1-0.2 mg/g of CSP For preparative chromatography, throughput can be defined as the amount of purified material obtained per unit of time and per unit Chirality and Biological Activity (21) Formation of diasteroiosrners H Separate diasteroisorners I by HPLC cleave to enantiomers Figure 18.8 mass of stationary phase Several factors affect this including loading capacity, column efficiency, selectivity, column size, temperature, cycle time, flow rate, and the solubility of the racemate The mobile phase plays a crucial role in the separation process for at least three main reasons The selectivity of the separation, retention time and solubilitv of the racemate are directly affected by the kobile phase composition Other parameters such as viscosity, solvent recovery, cost, and solvent handling properties also play a prominent role This brief introduction is also applicable to the criteria for CSP selection for SMB An example of a drug separated by preparative HPLC is cetirizine dihvdrochloride, a ra" cemic drug that is a second generation antihistamine H,receptor antagonist Studies on the effect of racemic and R (25) and S-Cetirizine (26) on nasal resistance indicated that both racemic and the R-enantiomer had similar activity The racemate and R-enantiomer inhibit histamine and induced an increase in nasal resistance, thus indicating the antihistaminic properties of R-Cetirizine(45).The S-enantio- mer was shown not to exhibit these antihistamink effects A n asymmetric synthesis (46), and resolution of an intermediate have delivered the single enantiomer previously However, for various reasons, the development of a preparative HPLC method seems to be the method of choice (47) The main reasons are the rapid scale up and the improved economics of this approach Utilization of the amide(24) (Fig 18.9)gave rise to a highly efficient separation using a Chiralpak AD column in a mixture of acetonitrileliso-propanol60:40.The efficiency of the separation can be measured by the a value(2.76)or the USP resolution(8.54) The a value and USP resolution numbers are measurements of how efficient the separation is; typically the higher the number, the better the separation This enabled the production of 1.6 kg of both the (+) and (-) isomers of high purity Like all methods for separating chiral molecules, chromatographic separations suffer from drawbacks: large quantities of expensive stationary phases are needed and large volumes of mobile phases are used, coupled with the resultant high dilution of separated prod- Chromatographic Separations Separation, HPLC Conversion to acid dihydrochloride Figure 18.9 ucts A number of methods have been introduced in an attempt to improve on this technology, such as recycling (44) Perhaps the biggest advancement in recent times has been the introduction and application of SMB technology in the field of chiral separations (48) This technique was pioneered in the late 1950s by Universal Oil Products in the United States as a useful method for separation of oil derivatives and sugars (49) Initially SMB technology was applied to very large volumes of material For example, xylene isomers are separated in thousands of ton quantities annually The application of SMB to the separation of racemic mixtures has led to downsizing and modifications of this technology, but the main principles remain the same The use of counter-current contact in SMB maximizes the driving force for mass transfer and the contact between the substrate and stationary phase This provides a more efficient use of the adsorbent capacity than that of a simple batch system (50) The separation of racemic mixtures is well suited to SMB technology, because these counter current systems can generally only perform two-component separations at a time (51) A detailed description of this technique is given in an excellent article by Guest (52) The SMB system generally consists of several columns, typically 6-12, which are connected in series An arrangement of pumps and valves are set up to maximize the stationary phase utilization, allowing for better solvent efficiency and adsorbate concentration This leads to two streams coming off the system in solution, one is termed the raffinate, which is enriched in the less adsorbed component, and the other termed extract, which is enriched in the more adsorbed component The complex set of conditions and parameters that are required to optimize SMB chromatography has led to the design and process optimization being done by computer simulations (53) A number of examples will be discussed that highlights this growing area of chiral separa- Chirality and Biological Activity NA H Aminoglutethimide (27) Figure 18.10 tions It should be noted that the scale of operation is dependent on the column size and can lead to a range from tens of grams to tons of separated isomers Clearly, the larger quantities separated imply that this technology has industrial applications The enantiomers of aminoglutethimide (27) (Fig 18.10) have been separated using an SMB approach (48) (see also Section 3, Fig 18.11 for more information on aminoglutethimide) A set of 16 columns (6 X 1.6 cm) containing Chiracel OJ were used The feed concentration was 1.63% in a mixture of hexane:ethanol (15:85), which was used as the mobile phase A feed rate of 0.45 mllmin and a mobile phase rate of mllmin gave rise to a production of 5.27 g of each enantiomer per day The S-(-)-enantiomer was obtained as the extract in solution, in a 99.8% purity, while the R-(+)-enantiomer also in solution as the raffmate, achieved a 99.9% purity This would lead to a productivity of 59.9 g of each enantiomer per kilogram of CSP per day It should be noted that one big advantage of SMB over preparative chromatography are the vast savings on mobile phase consumption; this is generally coupled to thin film evaporators that allow for very high levels of u Tramadol (28) Figure 18.11 recovery of the solvent This becomes even more evident when a poorly soluble compound is used The two isomers of the racemic analgesic drug Tramadol(28) (Fig 18.11) display differing aMinities for various receptors (- )-Tramado1 mainly inhibits the reuptake of noradrendine, whereas the (+)-isomer inhibits the reuptake of serotonin In addition, the (+Iisomer and its primary metabolite, the O-desmethyl derivative, are selective agonists of p opiate receptors (54) Tramadol has been efficiently separated using SMB; in addition, the resolution by crystallization is given in Section of this chapter (55) Comparison between batch chromatography and SMB for the separation of tramadol was made Use of 12 columns (100 x 21.2 mm ID), each packed with 20 g of Chiralpak AD 20-pm phase, and using a mobile phase composition of 2-propanolllight petroleurn/diethylamine (5:95:0.1 V/V/V)with feed concentration of 20 g/L, obtained a very high productivity Thus, 680 g of racemic tramadol could be separated per liter of stationary phase (which equates to 1.2 kg of racemate per kilogram of stationary phase per day) The solvent consumption of 144 L/kg of racemate should also be noted This gives both (+)- and (-)-enantiomers of high optical purity, with the extract of 6.33 g/L and the raffinate of 7.69 g/L Typically, the solvent (mobile phase) is readily recycled by the use of thin film evaporators, which further extends the economic practicality of the process 2.4 Conclusions It should be noted that all the techniques described in this c h a ~ t ecan r be inter-linked In other words, if one technique, i.e., asymmetric synthesis, failed to deliver enantiopure material, then another technique such as crystallization can be used to push through the product to the desired purity As an example of this "double" approach, the application of SMB and crystallization to the separation of mandelic acid is noteworthy (56) When very high levels of enantiopurity are required, the efficiency and cost effectiveness of SMB may not be economical However, if for example, a lower enantiomeric excess can be couded - with an enhancement by crystallization, then the - Classical Resolution SMB approach becomes even more favorable This can lead to substantial increases in the productivity of the SMB process Further examples of coupling of two techniques will be given throughout this chapter In summary chromatographic separations offer an expedient method for the separation of enantiomers on a small scale With the development of more efficient stationary phases and the application of SMB, this may become the method of choice for the separation of racemates Each individual case deserves investigation by all of the techniqueslapproaches described in this section Octhreo (29) Cthreo (30) Oceryfhro (31) Ceryfhro (32) Figure 18.12 3.1 CLASSICAL RESOLUTION Perhaps the most widely used method for the preparation of single enantiomers involves the classical resolution of a racemic mixture, which uses the formation of crystalline diastereomeric salts As discussed in the introduction to this chapter, by converting a racemic mixture of enantiomers to two diastereomeric salts with differing physical properties, one being crystalline and the other remaining in solution, the molecules can be separated and simply converted back to the two separated enantiomers With the advent of automation, the classical resolution approach offers a speedy and through racemate separation methodology This enables the separation of small amounts of material (milligram to gram) and can be directly scaled up to provide an industrial process (kilogram to ton) A number of different approaches to this type of separation are highlighted in the following sections, where it should be noted that diastereomeric salt resolutions have mistakenly been considered to be a mysterious art In fact, there is considerable information in the literature as to how to perform a resolution and the physical chemistry aspects associated with how to define and conduct the resolution to its optimum capability (57-59) We will not go into this in great detail but will highlight some pertinent points; for greater detail, the reader is directed to the monograph by Jacques et al (57) Separation of the Active Pharmaceutical Ingredient A number of single isomer switches, that is, where a drug that was previously sold as a racemate is developed and sold as a single isomer, have been isolated through classical resolution (60) This approach to a single isomer offers several advantages; first, the racemate is freely available and can be purchased to high levels of purity and quality Second, the analytical methods will also be in place Also, no new synthetic development chemistry is required, and hence this is the fastest route to the single enantiomers at the multigram scale Generally, this is the first method to be tried Some of the many available examples demonstrate the different nuances that can be applied in classical resolution to provide the single enantiomers in optimal yields and purities are given in this section An efficient and large scale resolution of methylphenidate (ritalin hydrochloride) using dibenzoyl-tartaric acid has been described (61) Ritalin is marketed for the treatment of children with attention deficient disorder (ADHD).Methylphenidate has two chiral centers and originally was marketed as a mixture of two racemates, 20% DL-threo (29, 30) and 80% DL-erythro (31,32) (see Fig 18.12 for the structures of all four isomers) As introduced previously, the erythro-isomer is defined as the case when the main chain of a molecule (drawn vertically in a Fischer projection) has identical or similar substituents at two adjacent non-identical chiral centers on the same side of the chain, whereas the threo isomer has Chirality and Biological Activity \ i) 4-Me-morpholine MeOH/H2O ii) (D)-(+)-DBTA (D)-(+)-DBTA (D)-(+)-DBTA 4-Me-morpholine i) Aq.NaOH, iPrOAc + ii) Conc HCI/H20 Figure 18.13 the corresponding substituents on opposite sides The racemic drug currently used in therapy comprises only the pair of threo-enantiomers (29, 30) The mode of action in humans is not completely understood, but methylphenidate presumably activates the brain stem arousal system and cortex to produce its stimulant effect In addition, there is no specific evidence that clearly establishes the mechanism whereby methylphenidate produces its mental and behavioral effects in children or conclusive evidence regarding how these effects relate to the condition of the CNS The D-threo (29) enantiomer has, however, been reported to be to 38 times more active than the corresponding L-threo enantiomer (30) (62) The resolution shown in Fig 18.13 uses the racemic hydrochloride salt as input material The HC1 salt is cracked to the free base in situ with 4-Me-morpholine, which then forms a salt with the resolving agent dibenzoyl-tartaric acid (DBTA).The required Dthreo-methylphenidate (29) is removed as the crystalline salt of D-(+I-DBTA,leaving the L- threo enantiomer (30) in solution with 4-methylmorpholine hydrochloride The use of 4-methylmorpholine to effect base release in situ helps to streamline the process and to remove 'a costly free base isolation process The D-threo-methylphenidate, (D)-(+)-DBTA, salt is readily converted into the hydrochloride salt It is interesting to note that recently, Celgene and Norvatis received a FDA approvable letter for the use of dexmethylphenidate for use in ADHD This consists of only the D-threo enantiomer (291, in comparison with the original product, which contained all four isomers (29-32) Chemists at Chiroscience took an alternative approach to the D-threo-methylphenidate (29) single enantiomer (63) An efficient resolution using L-(-)-di-toluoyl-tartaric acid (DTTA) was developed This left the required D-threo diastereoisomer in solution with a diastereomeric excess of 88%yield in 55% chemical yield Conversion of this salt to the free base and subsequent crystallization of the hydrochloride salt gave >98% ee D-threo methylphenidate in high purity in an overall yield of 42% The enhancement of the ee is caused by the eutectic point of methylphenidate hydrochloride, which is at 30% ee A more detailed description of this phenomenon will be discussed later in this section (SJNaproxen (36) is a non-steroidal antiinflammatory drug that was introduced to market in 1976 by Syntex The S-(+)-isomeris about 28 times more effective than the R-(-)isomer (64) The annual sales in 1995 were about $1 billion; thus, a large amount of effort has been spent developing the synthesis of (S)Naproxen (65) The resolution of racemic Naproxen (33), developed by Syntex, approaches the ideal case for a Pope Peachy resolution, that is, resolution using non-stoichiometric quantities of resolving agent (66) Here, a mixture of equivalent (eq) of the racemic acid, 0.5 eq of an achiral amine base, and 0.5 eq of the chiral amine (N-alkylglucaminel are used (Fig 18.13) This results in the formation of two salts: one is the insoluble (S)Naproxen chiral amine (341, obtained in 4547% yield and optical purity of 99% The second salt that remains in solution contains (R)Naproxen and the achiral amine (35) The insoluble salt of (S)-Naproxen (34) is removed Classical Resolution (R,S)-Naproxen(33) + Achiral arnine base + Chiral arnine base / \ Precipitate O / / Mother liquors d H R*NH2 \o / \o (S)-Naproxen.chiralarnine base (34) C / H RNH2 (R)-Naproxen.achiralarnine base (35) Heat &CO2H \o '-0 / / (R,S)-Naproxen.achiralarnine base (37) Recycled to resolution (S)-Naproxen(36) Figure 18.14 by filtration The mother liquors are then heated and the achiral m i n e base catalyzes racemization of the unwanted R-enantiomer The resulting racemic mixture of the acid (R,S)-(37)can then be put back into the resolution loop Using this process, the overall yield of (5')-Naproxen is >95%, based on the input of racemic acid To further highlight the efficiency of this process, the N-alkylglucamine resolving agent is recovered in >98% per cycle Racemic bupivacaine hydrochloride (38, Marcaine) is currently used as an epidural anesthetic during labor and as a local anesthetic in minor operations Clinical studies have shown that levo-bupivacaine (41) is less cardiotoxic in man, making it significantly safer than the racemate (67) Separation of the enantiomers was readily achieved using 0.25 eq of D-tartaric acid This resulted in the isolation of a 2:l (S)-bupivacaine D-tartaric acid salt (39) in 98% de, leaving the (R)-bupivacaine free base (40) in solution Conversion of the tartrate salt to (S)-bupivacaine hydrochloride (39) was obtained in 35-40% overall yield based on racemate input To increase the economics of the process, a racemization of the unwanted R-enantiomer was required Treatment of the liquors containing the enriched (Rbbupivacaine, tartaric acid, propanol, and propionic acid at reflux resulted in complete racemization in h By pertinent processing, the racemic free base thus obtained is isolated by crystallization and can be put back into the resolution cycle (68) Another fine example by chemists from Eli Lilly involves a clever resolution-racemization-recycle (R-R-R) process in the synthesis of Duloxetine (69) As discussed in Section of this chapter, Tramadol is a chiral drug substance that is currently used as a high potency analgesic agent The preparation of Tramadol is shown in Fig 18.16, which results in the formation of all four possible stereoisomers from the Grig- Chirality and Biological Activity +.0.25eq (D)-(+)-Tartaric acid (3-Bupivacaine.(D)-(+)-Tartaric acid (39) 1) NaOH 2) HCl(g) in IPA Figure 18.15 nard reaction (70) The trans isomers (42,431 form over the cis isomers (44,45) in a ratio of : ; the currently marketed racemate consists of only the trans isomers It is possible to take this crude reaction mixture and selectively isolate either the (+)-trans isomer (421, by using di-p-toluoyl->tartaric acid [D-(+)DTTA] resolving agent or the (-)-trans isomer (43) using L-(-)-DTTA This highlights the high selectivity that can be achieved when using certain resolving agents In the case of Tramadol, the cis isomers (45,46) not form crystalline salts with DTTA and therefore remain in solution This results in a highly efficient process, where the chiral acid not only separates the single enantiomers (42 or 43) but also removes other impurities (i.e., cis isomers 44 and 45) at the same time (71) Another drug that is sold as a racemate is Etodolac (46),which is used as a non-steroidal anti-inflammatory agent (NSAID) that also has analgesic properties; it has the ability to retard the progression of skeletal changes in rheumatoid arthritis (72) It has been shown that the majority of therapeutic activity lies in the S-(+)-isomer (73) D-(-)-N-Methylglucamine (meglumine) is obtained by ring opening of D-glucose with methylamine, and hence it is readily available and inexpensive Scientists at Chiroscience have described the use of meglumine to separate the enantiomers of Etodolac (74) It was shown that the meglumine salt possessed suitable properties to enable its use as a salt for pharmaceutical administration Therefore, in the case of Etodolac, meglumine can not only be used to separate Classical Resolution H (Nl + N H / Precipitate 2(S)-CSA C02H (47)yer liquors / \ Precipitate Mother liquors Figure 18.16 Figure 18.18 the enantiomers, but it can also be used as the pharmaceutical salt form of choice In addition to the racemic drugs discussed in this section, resolutions are also used in the isolation of key building blocks for the pharmaceutical industry An important class of these intermediates are amino acids, many of which are available as the single isomer from natural sources (see INTRODUCTION) The use of unnatural amino acids and D configured ones are expected to have a greater influence at the biological level In the drive for molecular diversity and metabolic stability, a number of unnatural amino acids such as the non-proteinogenic piperazine carboxylic acid (47) (Fig 18.18) have been developed Specifically, this amino acid has found use as an intermediate compound of the HIV proteinase inhibitor L-735,525 (75) The racemic cyclic amino acid (47) has been resolved with S-camphorsulfonic acid (CSA), which yields the Sisomer as the double CSA salt (48) as the precipitate (76) Retained in the mother liquors is the R-isomer (49) This can neatly be racemized to the S-isomer by mixing with S-CSA in a suitable solvent On seeding with pure (S,S)diastereomeric salt, a further quantity of the desired (S,S)product (48) is obtained, leaving the R-isomer (49) once more in the liquors The whole cycle can be repeated and has been demonstrated with four complete cycles To complete the whole process, the resolving agent is also readily recovered and recycled 3.2 Separation of Intermediates to Single Enantiomer Active Pharmaceutical Ingredient Etodolac (46) Figure 18.17 The previous examples given for diastereomeric salt resolution have all involved separation of the active pharmaceutical ingredient (API) or late stage intermediate Whereas this Chirality and Biological Activity (4-Verapamilicacid (50) (R)-Verapamil(51) (a-Verapamilicacid (52) (a-Verapamil(53) Figure 18.19 does offer several advantages from the point of view of time and quality aspects, there are also a number of drawbacks If, for example, a racemization of the unwanted isomer cannot be found, there would be a waste of 50% of material Therefore, it can often be advantageous to conduct the separation at an earlier stage in the synthesis of the drug This leads to better atom efficiency compared with resolution of the final product, resulting in a reduction of the overall amount of waste and cost One such example is Verapamil, which is a well-established treatment of cardiovascular ailments (77) S-(-)-Verapamil (51) has specific transmembrane calcium channel antagonist activity, whereas its antipode (53) influences a wider range of cell pump actions, including those for sodium ions (78) Verapamil has been separated into its single enantiomers by resolution with expensive resolving agents, which required multiple recrystallizations to effect complete separation (79) Looking into the synthetic sequence of Verapamil, several intermediates seemed to be attractive alternatives to Verapamil(80) The intermediate verapamilic acid (Fig 18.19) was efficiently separated using a-methylbenzylamine (a-MBA), which is an extremely cheap resolving agent (81) Subsequent transformation of the easily obtained R- or S-verapamilic acid (50 or 52), required a further three to four synthetic steps to yield the active pharmaceutical ingredient The racemate aminoglutethimide (27) has been shown to be effective in the treatment of hormone-dependent breast cancer (Fig 18.20) Further studies have shown that the R-enantiomer is more potent than its antipode as an aromatase inhibitor (82) The resolution of aminoglutethimide itself has been reported in the literature, using tartaric acid This resolution suffers from the formation of solid solutions (83),which require endless crystallizations to deliver the single enantiomer (84) Use of a suitable precursor (54) enabled separation of the intermediate (55),by treatment with the alkaloid resolving agent (-)-cinchonidine This chiral acid was then cyclized to nitroglutethimide, which on reduction, gave the desired R-aminoglutethimide (56) (85) It is noteworthy that in the case of aminoglutethimide, the m i n e functionality is an aniline moiety Because of the low pK, associated with this amine (2.5-4.6), the number of acidic resolving agents that can be employed are reduced, because they need to be of relatively high acidity to form a salt 3.3 Crystallization-Induced Asymmetric Transformation A number of amino acids have been separated by resolution, in certain cases the yield of the required diastereoisomer has been greater than 50% (86) p-Chlorophenylalanine is of considerable pharmacological interest, because of its ability to inhibit serotonin forma- Nonclassical Resohtion steps I Figure 18.20 tion in laboratory animals (87) Both the Rand S-enantiomers have also been used as building blocks in the synthesis of other drugs An ingenious approach to R-p-chlorophenylalanine methyl ester, which is based on a one-pot resolution-racemization sequence, is highlighted in Fig 18.21 Here, treatment of racemic p-chlorophenylalanine methyl ester (57) with 0.5 eq of D-tartaric acid and 0.1 eq of salicylaldehyde in methanol gave a 68% yield of 98% enantiomeric purity of the 2:l R-pchlorophenylalanine D-tartaric acid salt (58) The reason that the absolute yield is greater than 50% is caused by the S-enantiomer being racemized in situ The 2:l tartrate salt is crvs" talline and is therefore removed from the system by virtue of its insolubility This drives the equilibrium further in favor of the 2:1R-pchlorophenylalanine D-tartrate salt (88) While the common goal remains to be the rational design of resolving agents (89), it is clear that we are still away from this actually happening An alternative "family" approach to classical resolution has been demonstrated by Vries et al (90) A group of similar resolving agents are mixed simultaneously with the racemate This was done to shorten the time required to complete the resolving agent screen Note should be made that the families of resolving agents are very similar and that the crystalline species obtained by this method contained more than one of the resolv- ing agents As with all screens, analysis of the data is often time consuming and laborious Bruggink et al have shown that differential scanning calorimetry (DSC) of the isolated salts can help to quickly determine whether the isolated salt will provide a through resolution (91) However, with a methodical and precise screening protocol, it is nearly always possible to find a suitable resolving agent that effects separation of the enantiomers (92) 4.1 NONCLASSICAL RESOLUTION Preferential Crystallization A brief description of the type of "racemic" compounds is necessary for the reader to better understand the principles behind the application of crystallization methods to the separation of enantiomers Three fundamental types of crystalline racemates exist In the first, the crystalline racemate is a conglomerate, which exists as a mechanical mixture of crystals of two pure enantiomers The second, which is the most common, consists of the two enantiomers in equal proportions in a welldefined arrangement within the crystal lattice; this is termed racemic compound The third possibility occurs with the formation of a solid solution between the two enantiomers that coexist in an unordered manner in the crystal This kind of racemate is called a pseu- Chirality and Biological Activity (R)-pchlorophenylalanine.0.5eq(D)-tartaricacid (58) Figure 18.21 doracemate and is rather rare Conglomerates have been estimated to be approximately 10% of all racemates (93) Diagrammatic representation of the first two types of racemate are shown in Fig 18.22 By understanding the appropriate phase diagrams, which describe the melting behav- ior of the two enantiomers (binary melting point phase diagram) or their solubility behavior in the presence of a solvent (ternary solubility phase diagram), separation of enantiomers can be reproduced Phase diagrams for the three types of racemate are shown in Fig 18.23 For a full and detailed explanation of this topic refer to the monograph of Jacques et al (57) ( +x.>+x+ z.> 4.2 Enrichment of Enantiomeric Excess by Crystallization Racemic mixture (conglomerate) The attainment of high levels of enantiopurity is not always possible by enzymatic or diastereomeric resolutions or by asymmetric syntheses alone It is however frequently possible to prepare a pure enantiomer from a partially resolved sample by simple recrystallization For this process to proceed successfully it is necessary that the initial enantiopurity of the mixture is greater than that of the eutectic point in the phase diagram By utilization of the phase diagram, the optimal quantity of solvent required can be calculated It is also possible to calculate the maximum expected yield x 9+x.)+z+ Racemic compound Figure 18.22 4 Nonclassical Resolution Conglomerate (-1 Racemic compound High EE eutectic point (+I u (-1 Figure 18.24 Pseudo racemate Figure 18.23 Note should also be made that in some cases recrystallization reduces the enantiomeric excess, which can lead to crystallization of the racemate (94) In these cases the mother liquors contain moderately to highly enriched material It is therefore important to plan the strategy at which point the enantiomer is recrystallized to optical purity This may be from an enzymic resolution, or in the event that an asymmetric synthesis has failed, to deliver enantiopure product As discussed in Section 3, the liquors from the diastereomeric resolution with DTTA of 88%de can be cleaved to the free base, and crystallization of the hydrochloride salt gives >98% ee This is because of the fact that methylphenidate hydrochloride has a eutectic point of 30%ee Davies et al (95) and Winkler et al (96) have prepared single enantiomer methylphenidate (29) Their approaches use an enantioselective synthesis; the enantiomeric excesses are 86% and 69%, respectively, thus requiring recrystallization to deliver enantiopure product Another example of this type of compound is Warfarin (13).Chemists at Dupont (97) developed an asymmetric hydrogenation approach, which gave Warfarin in -80% ee Simple crystallization in an appropriate solvent yielded optically pure Warfarin, thus indicating that the eutectic point is below 80% ee (See earlier section on the metabolism and binding properties of the Warfarin enantiomers) The phase diagrams below highlight two typical cases, the first where the eutectic point E is close to the racemate, and the second where the eutectic approaches the single enantiomer as shown in Fig 18.24 In the first case, it would be preferable to crystallize the enriched enantiomer to optical purity, e.g., methylphenidate However, in the second case, a very stable racemic compound exists, giving rise to a high eutectic point Here crystallization of enriched enantiomer mixture will only be successful at high ee For example, verapamil hydrochloride requires that the ee be greater than 98% for crystallization to yield Chirality and Biological Activity enantiopure product Below this, the enantiopurity is reduced In this case, it is advantageous to recrystallize the diastereomeric salt precursor to optical purity before proceeding to final product 4.3 Resolution by Direct Crystallization It is important to show how conglomerates are identified We have already seen that they have specific phase diagrams as shown in Fig 18.23 Other such data that support identification of a conglomerate are IR, X-ray data, and observation of a spontaneous resolution or resolution by entrainment Note should be made that in 1848, Louis Pasteur separated the dextrorotatory and levorotatory crystals of sodium ammonium tartrate This manual sorting of crystals is also known as triage, and by its very nature is time consuming and laborious The readers are again directed towards the Jaques et al monograph, which lists over 250 known examples of conglomerates (57) There are two possibilities for separation of enantiomers by direct crystallization The first uses spontaneous resolution, which occurs when a conglomerate crystallizes This crystallization may be followed by the mechanical separation of the crystals of the two enantiomers Various techniques have been developed that aid this separation The second type of resolution by direct crystallization is known as entrainment Here, the differences in the rate of crystallization of the enantiomers in a supersaturated solution give rise to a separation Strict control of the conditions for the crystallization are required, with the system of crystals and solution not being allowed to come to equilibrium and time playing an important role The occurrence of conglomerates has been estimated to be approximately 10% of all racemic compounds We will now illustrate this phenomenon with some pertinent examples An example of use of the conglomerate Narwedine (59) in the synthesis of a natural product Galanthamine (61) which is an Amarylliduceae alkaloid and has been used clinically for 30 years for neurological illnesses (98) More recently it has been approved for the use in the treatment of Alzheimer's disease (AD) (99) Galanthamine acts to inhibit acetylcholinesterase (AChE), thus increasing the levels of acetylcholine An increase in the level of acetylcholine in patients with AD has been shown to improve their cognitive performance Galanthamine has been extracted from botanical sources; however, several tons of daffodil bulbs are needed to produce kg of product A synthetic route has been developed that uses a crystallization-induced chiral transformation (Fig 18.25) This crystallization was first reported by Barton and Kirby (100) and further developed by Shieh and Carlson (101) The success of this transformation is based on two phenomena: narwedine (591, which crystallizes as a conglomerate, and (-)-namedine (60), which equilibrates with (+)-namedine through a retro-Michael intermediate This process has now been developed so that (-)-narwedine (60) is routinely obtained in 80% yield from the racemate input, as shown in Fig 18.25 (102) Recently a number of potent 5-HT, receptor antagonists such as Ondansetron have been reported to be clinically effective for the blockade of chemotherapy-induced nausea and emesis (103) The structurally novel compound (62) has also been shown to be a highly potent 5-HT, antagonist (104); specifically, the R-(-)-(62) enantiomer was shown to be the most active Comparison of the physical data of the racemate and single enantiomer indicated that this structure (62) exists as a conglomerate (104) By careful experimentation, the best concentration, temperature, and time for crystallization were discovered Table 18.1 highlights the results obtained for the entrainment The initial concentration of the solution was 10.0 g of (2)-(62)in 50 g of acetone In all runs, 10 mg of seed crystals were used From the 10 runs highlighted in the 18.1, 21.0 g of R-(-1462) of >92.O% ee and 21.4 g of (S)-(+)(62) of >90% ee are obtained from an input of 50.4 g of racemate The table also nicely illustrates the continuous nature of the process, which coupled with the fact that no resolving agent, chiral auxiliary, enzyme, or catalyst is needed, underlines the economic advantages of this type of process The importance of amino acids as building blocks for asymmetric synthesis is well documented (105) A number of amino acids have been shown to exist as conglomerates Shi- Nonclassical Resolution 0 Entrainment NMe Me0 Me0 Figure 18.25 raiwa et al have described the preferential crystallization of racemic methionine hydrochloride (106) The obtained D- or L-methionine hydrochloride was, however, only -75% optically pure, requiring a further recrystallization to furnish enantiopure product Shiraiwa et al have also recently disclosed the resolution of (2RS, 3SR)-2-amino-3-chlorobutanoic acid HC1 again using entrainment (107) Here it was shown to be necessary to conduct the crystallization in an ethanol15 M hydrochloric acid solvent mixture for optimal results By careful control of the conditions, high levels of enantiomeric excess were obtained in the crystalline salt Chemists in Japan have developed an excellent approach to (+)-Diltiazem, which is a coronary vasodilator (108) An intermediate Figure 18.26 is successfully resolved using preferential crystallization The glycidic acid-substituted phenylesters were prepared; of the 30 synthesized, only one exhibited conglomerate properties (109) This was the 3-(Cmethoxypheny1)glycidic acid 4-chloro-3-methylphenyl ester (63) Table 18.2 summarizes the physical data collected, which is illustrative of the conglomerate nature of this compound The obtained single enantiomer (- )-epoxide (64) is then converted into the required (+)-isomer of Diltiazem (65) in several steps, as highlighted in Fig 18.27 Taxol is a natural product isolated in very low yield from Taxus brevifolia and is used in the treatment of cancer (110) The extreme chemical complexity of Taxol makes production by total synthesis uneconomical However, a semisynthetic approach using the naturally derived 10-deacetylbaccatin I11 (66) condensation with N-benzoyl-(2R, 3s)-3-phenylisoserine (67) does provide an alternative and economic approach (111) N-benzoyl-(2R, 3s)-3-phenylisoserine (67) is also commonly known as the Taxol side-chain and has been prepared in optically active form using chiral auxiliaries or resolving agents (112) It has been shown that the Taxol side-chain is a conglomerate and can therefore be cheaply and Chirality and Biological Activity 804 Table 18.1 Resolution of (62) by Preferential Crystallization Run Added (g) Seed Time (minutes) EE of Solution (%EE) Amount of Crystals (g) Rotation %EE of Solid Reprinted from H Harada, Tetrahedron Asymmetry, vol 8, T Marie, Y Hirokawa, and S Kato, 1997, pp 2367-2374 Reproduced with permission from Elsevier Science efficiently entrained to the single required enantiomer (113) ENZYME-MEDIATED ASYMMETRIC SYNTHESIS Enzymes have found frequent use in the synthesis of single isomer drugs from racemic or prochiral compounds at the larger manufacturing scales The use of enzymes to effect chiral transformations in the medicinal chemistry laboratory has been far less frequent; however, the increasing availability of immobilized and stabilized forms of enzymes has made their use easier and the resultant transformations more predictable By virtue of their complex macromolecular structure, including a highly defined active site, enzymatic transformations generally proceed with a high degree of chemical selectivity and stereospecificity Reactions are typically conducted under mild conditions of temperature, pressure, and pH, thus minimizing losses caused by unwanted side reactions or partial racemization The use of extremophiles or cross-linked enzymes such as CLECs enable the use of higher temperatures, pressures, and organic solvents Enzymes can be utilized to affect a number of transformations; the broad spectrum of reactions, including amide bond formation, hydrolysis, esterification, reduction, oxidation, and carbon-carbon bond formation, has been reviewed elsewhere (114) 5.1 Amide Bond Formation The use of enzymes to stereospecifically form amide bonds has been described in many texts (115); however, the commercial availability of cross-linked enzyme crystals (CLECs),for example, PeptiCLEC-TR, which is an immobilized form of Thermolysin protease, has been used in the synthesis of D2163 (68), a novel matrix metalloproteinase inhibitor (116) In vitro enzyme screening identified the all-natural SSS-isomer as the active product The elegant CLEC (117) technology used in this example makes the enzyme stable to typical organic reaction conditions and enables facile removal of the enzyme at the end of the reaction by simple filtration On this basis, it is Table 18.2 Properties of (63) Indicating Conglomerate Nature Compound MP PC) Solubility (g1100 mL) THF (21-4.2 (-h4.2 123-124 139-141 14.0 6.7 Solubility (g/100 mL) DMF IR Spectrum 13.0 6.9 Identical Identical Enzyme-Mediated Asymmetric Synthesis Figure 18.27 anticipated that medicinal chemists will more commonly use these enzymes in the future The coupling of dipeptide (69) to the protected a-thio carboxylic acid (70) was conducted in organic solvent at high concentration with the desired product produced in a few hours with high enantiospecificity the role of cyclooxygenase-independent properties of the R-enantiomers in the gastrointestinal toxicity of the racemates and the likelihood that the use of racemates increases the propensity of profens to alter the pharmacokinetics of other drugs has been described (118) Whereas not all profens are sold as single isomers, Naproxen is sold as the single Senantiomer (36) where various strategies including crystallization, chromatographic separation, asymmetric hydrogenation and enzymatic hydrolysis, and esterification have been used to prepare the single isomer (65).Specific examples include the use of Candida cylindracea lipase to enantioselectively prepare single isomer naproxen ester with trimethyl silyl methanol (119)and the use of Candida rugosa lipase in an enantioselective continuous hydrolysis of Naproxen methyl ester (120) Pipecolic acid is a component of a number of active drugs, including bupivacaine (38) and thioridazine (72) (Fig 18.30), which has been efficiently resolved as the racemic n-octyl pipecolate with Aspergillus niger The S-isomer is obtained as the free acid in a 40% yield based on the available enantiomer with a 97% ee (121) Propanolol(14) is a broadly used P-adrenergic receptor blocking agent that is sold as the racemate However, the majority of the activity is associated with the S-enantiomer (74) (see Section 2) (122) The asymmetric 5.2 Transesterification and Hydrolysis A widely used technique for separating racemic mixture is the use of enzyme mediated transesterification or hydrolysis One important example is the separation of Naproxen (331,which is a member of the 2-arylpropionic acid class of profens that are broadly used as NSAIDs (see Section for the separation of enantiomers using a crystallization approach) The important association between chirality and biological activity of this class of drugs has been extensively researched, where Figure 18.28 Chirality and Biological Activity SCOPh I Figure 18.29 synthesis of the desired S-enantiomer has been achieved by the selective acylation of the R-enantiomer of the key intermediate (73) as shown in Fie 18.30 - Thioridazine (72) Bupivacaine (38) Figure 18.30 5.3 Oxidation and Reduction In addition to the widely reported techniques of amide bond formation, transesterification,and hydrolysis,enzymic enantioselectiveoxidation is also used in the synthesis of single isomer drugs Pate1 described the efficient oxidation of benzopyran (751,an intermediate in the synthesis of potassium channel openers (123).The transformationwaseffected with a cell suspension ofMortie~llaramanniana with glucose over a 48-h period, the isolated product (77) was obtained in a 76% yield with an optical purity of 97%and a chemical purity of 98%, as shown in Fig 18.32 Reduction with a variety of enzymes has been reported (114), including bakers yeast for the reduction of a-methyleneketones to the corresponding a-methylalcohol (124),a functionality that is present in a number of drugs The reduction of an azidoketone (78) using Pichia angusta enzyme has been used in the synthesis of S-salmeterol (79) (125) Salmeterol (Serevent) is a potent, long-acting P2-adrenoreceptor used as a bronchodilitor in the treatment of asthma Recently, Sepracor claimed that the Senantiomer had a higher selectivity for P2 receptors and that it did not cause certain adverse effeds associated with the administration of ( ) -or (R)-salmeterol (126) The synthesis of 6')-salmeterol(79) is shown in Fig 18.33 6 Asymmetric Synthesis Figure 18.31 ASYMMETRIC SYNTHESIS Synthetic organic chemists have a vast array of tools at their disposal when faced with the challenge of preparing a chiral compound as a single enantiomer The purpose of this section is to introduce the reader to some asymmetric approaches toward chiral drugs and medicinal compounds, highlighting examples where the stereoisomers behave differently in biological systems There are many excellent books and reviews covering methods for asymmetric synthesis and their application to the preparation of pharmaceutical agents and complex natural products (127) 6.1 Chiral Pool The use of enantiopure starting materials from nature in the synthesis of chiral drugs is not only of great historical significance but remains of critical importance to the pharmaceutical industry Consideration of the current biggest-selling single enantiomer drugs shows how important this approach is (8of the top 10 in 1996 were obtained from chiral pool Figure 18.32 starting materials or by synthetic manipulation of fermentation products) (127) The optically pure starting materials that have been used in drug synthesis include amino acids, hydroxy acids, terpenes, alkaloids, carbohydrates, and many more structurally diverse compounds There are many syntheses involving clever manipulation of chiral pool starting materials and use of these chiral centers to induce further asymmetry (i.e., by diastereoselective reactions) We will briefly " consider some examples in which all or most of the chiral centers in the target molecule originate directly from nature Angiotensin-converting enzymes (ACE) inhibitors are used mainly for the treatment of cardiovascular disorders and are among the biggest selling drugs worldwide (128) Enalapril (80) is synthesized from the natural amino acids L-alanine and L-proline(129) Lisinopril (81) incorporates a lysine derivative (130) One of the chiral centers in Captopril (82) is derived from proline, but the other is generated by chemical or enzymatic resolution (131) Cilazapril(83) is a conformationally restricted second generation ACE inhibitor developed by Roche, and the core is synthesized from a glutamic acid derivative and an amino acid-derived pyridazine (128, 132) There are many other examples of drugs based on an amino acid backbone Stoner et al recently reported a synthesis of the HIV protease inhibitor ABT-378 (Lopinavir) (84) (133) In a similar synthesis to that of the related compound, Ritonavir, key intermediate (85)is prepared by stepwise diastereoselective reduction of enaminone (86).This means that the existing chiral center, derived from natural L-phenylalanine (protected to 87),controls the formation of the two new stereocenters as Chirality and Biological Activity /// Br(CH2)60(CH2)4Ph,DMF OH - H AcOH, water Figure 18.33 discussed for chiral auxiliaries below Two acylations then complete the synthesis, with the final chiral center clearly derived from Lvaline The stereospecificity of binding at the histamine H3-receptor was investigated by preparing a series of ligands from D- or L-histidine (88)(134) It was found that compounds such as (5')-(89)had greater affinity for the receptor than their R-enantiomers In addition, replacing the aromatic moiety with a cyclohexyl group (e.g., 90) switched the activity to agonism for compounds with an amino group in the chain Hydroxy acids are important c h i d starting materials in the synthesis of many biologically active compounds (135) (S)-3-Hydroxyy-butyrolactone (91)is a very useful synthetic unit available from D-pyranoses (136) Workers at Schering-Plough used this as the key starting material in a concise synthesis of Sch 57939 (92), a P-lactam-based cholesterol absorption inhibitor (137).The condensation between the dianion of (S)-3-hydroxy-y-butyrolactone and an appropriate diary1 imine proceeded with very high diastereo- and enantioselectivity, generating azetidinone (93) with a trans:& ratio of >95:5 Researchers at Abbott have been investigating the use of pyrrolidinyl isoxazoles as nicotinic cholinergic channel activators (138) Until recently, ABT-418 (97) was undergoing clinical trials as a potential treatment for cognitive impairment and decline and for Alzheimer's disease A short synthesis of ABT-418 was devised starting from commercially avail- Asymmetric Synthesis L-histidine methyl ester (88) Enalapril R=CH3 (80) Lisinopril R=(CH2)4NH2 (81) Captopril (82) Figure 18.36 Figure 18.34 able (8)-pyroglutamic acid methyl ester (94) (139).Acetone oxime dianion was added to the methyl ester (94) to generate an intermediate (95) Racemization of the chiral center was found to occur under basic conditions; however, this was avoided by immediate treat- \ (i) NaBH(TFA)3 (i) NaBH4,TFA Lopinavir (87) (86) (89-93%) Figure 18.35 Chirality and Biological Activity LDA (2 eq.)/DMF/DMPU Ar 'CH=NA? L (S)-3-Hydroxy-y-butyrolactone (91) Sch 57939 (93) (99.5%ee) 'F Figure 18.37 ment with concentrated sulfuric acid resulting in cyclization and dehydration to amide isoxazole (96) Redudion and N-rnethylation yielded ABT-418 (97) The binding affinity of ABT-418 at neuronal cholinergic channel receptors was measured to be one order of magnitude greater than the corresponding R-enantiomer (Ki = 4.2 versus 44 nM)(138) 6.2 Chiral Auxiliary In this approach the substrate is attached to a chiral, non-racemic unit that controls the formation of one or more new chiral groups Reaction of the coupled unit with a reagent or prochiral substrate is designed to produce one diastereomeric product in excess The auxil- Me ABT-418 (97) (ee >99%) Figure 18.38 6 Asymmetric Synthesis BuLi, THF, -78°C HN i THF, -78" to -50°C (100) LiOOH, THFM20 NaHMDS, BrCH2C02Bu eq,dO t-Bu02C * 6.12 Figure 18.39 iary is then removed (and preferably recovered), providing the product in high enantiomeric excess This process is most attractive when both isomers of the auxiliary are readily available in enantiomerically pure form, and where the reaction leads to high levels of stereoselectivity in a predictable manner Attaching and removing the auxiliary should be straightforward and proceed without loss of stereochemical integrity Many auxiliaries currently in use are derived from 1,Parninoalcohols (140) These are readily available from natural sources with little or no synthetic manipulation and can react in a variety of ways to form conformationally well-defined (usually cyclic) auxiliary systems The use of oxazolidinones in asymmetric synthesis was developed by Evans et al.,and these oxazolidinones have been used extensivelv in a variety of different reactions (140, 141) The use of this chiral auxiliary in the preparation of pharmaceuticals is widespread, and there are several large-scale processes using such chemistry (142) Abbott reported an improved synthesis of ABT-627 (98)involving an asymmetric alkylation of the valine-derived acyl oxazolidinone (99) (143).ABT-627 (Atrasentan)is a selective endothelin ETAreceptor antagonist under development for the treatment of cancer, particularly prostrate cancer Acid (100) was activated as a mixed anhydride and treated with the lithium anion of the oxazolidinone to give (101) Both of the following deprotonation and alkylation steps must be controlled to give high levels of stereoselectivity The (Z)-enolate (102) is favored, both kinetically and thermodynamically, by the bulky iso-propyl group 812 Chirality and Biological Activity Table 18.3 Stereochemical Variation (3a, ) Ki (nMversus HIV-1) IC,, (a, cell HIV-1) R, R Tipranavir R,S S, R s, s and is held rigid by chelation to the carbonyl oxygen of the oxazolidinone The major stereoisomer then results from alkylation of this chelated enolate anion from the least hindered "upper" face to yield (103) as the major product There are many strategies for removal and recovery of an oxazolidinone auxiliary (141) In this case, hydrolysis with lithium peroxide provides the acid that is transformed into Atrasentan through a cyclization-ring contraction strategy controlled by the chirality present in (103) Tipranavir (PNU-140690)is a potent thirdgeneration HIV protease inhibitor in clinical development by Boehringer Ingelheim (under license from Pharmacia) The biological activity of such 5,6-dihydro-4-hydroxy-2-pyrone sulfonamides shows considerable stereochemical variation (Table 18.3) (144).The R-configuration is preferred at both chiral centers (3cr and 6), and Tripanavir is more than 50 times as potent as its enantiomer in a cell culture assay using HIV-lI,I,-infected H9 cells An asymmetric synthesis (145) begins with the Michael addition of an aryl cuprate (derived from commercially available Grignard reagent 105) to the unsaturated oxazolidinone imide (1041, yielding the adduct as a single diastereomer (106) The nitrogen protecting group was changed and an acetyl group introduced to give ketone (107), which undergoes a stereoselective aldol reaction with an acetylenic ketone (108) The highest diastereoselectivity was obtained for this reaction using Ti(OnBu)C1, as the Lewis acid Both of the critical asymmetric steps to form new chiral centers are controlled by the (R)-phenyl oxazolidinone The chiral auxiliary is removed when (109) is treated with base to form the lactone ring This is followed by two further steps that generate PNU-140690 (110) as a single enantiomer The enantioselective synthesis of dopaminergic benzyltetrahydroisoquinolines and their binding to Dl and D, dopamine receptors was investigated by Cabedo et al (146) The synthetic route, illustrated by the preparation of the (1s)-isomer involves stereoselective reduction of the isoquinolinium salt (114) with (R)-phenylglycinol (introduced in protected form as 112) as the chiral auxiliary The (1R)enantiomer of (115), prepared in an analogous fashion using (S)-phenylglycinol,binds to dopamine receptors with considerably less f i n ity (>lo0 p N versus Dl and 61.2 pM versus D In contrast, stereochemical differentiation was not observed at the dopamine uptake site for these compounds Two different chiral auxiliary approaches have been applied to the synthesis of NPS 1407 and it's enantiomer (119) (147) NPS 1407 is an antagonist of the glutamate NMDA receptor that has in vivo activity in neuroprotection and anti-convulsant assays The R-enantiomer was synthesized in four steps from (116) with the chiral center introduced by a completely stereoselective alkylation of hydrazone (117) The chiral auxiliary, S-(-)-1-amino-2-(methoxylmethyl)pyrrolidine (SAMP), was introduced by condensation with aldehyde (116) and removed by catalytic hydrogenolysis In the second method, the S-enantiomer was formed in a four-step sequence with the chiral center installed by the Michael addition of chiral amine (121) (formed in one step from the readily available cr-methylbenzylarnine) to benzyl crotonate (120) NPS 1407 (123) was found to be 12 times more potent than it's enantiomer (119) at the NMDA receptor in an in vitro assay An example of the use of a terpene as a chiral auxiliary is provided by the synthesis of the anti-viral reverse transcriptase inhibitor Lamivudine (148) The nucleoside analog is marketed by Biochem Pharma (now Shire Pharmaceuticals) and Glaxo Wellcome (now GlaxoSmithKline) for the treatment of HIV and hepatitis B virus infection In the Asymmetric Synthesis CuBrIDMSrrHF O°C/l hour * N(TMS)z (105) MgBr N(W2 Ti(0 " B U ) C I ~ / C H ~ C I ~ / - ~ ~ ~ C OH / - (109) (de= 92%) / - 30 Tipranavir (110: 3~i=R,6=R) Figure 18.40 production route, the glycolate derived from (-)-menthol(124) is coupled with thioacetyl dimer (125) The chiral auxiliary directs reaction to install the desired (%)-stereochemistry in (126) In situ formation of chloro compound (127) is followed by a stereoselective coupling reaction with trimethylsilyl cytosine again directed by the (-)-menthy1 carboxylate Reductive removal of the auxiliary yields Lamivudine (129) as a single isomer that was found to have favorable toxicological and pharmacokinetic properties to the racemate 6.3 Chiral Reagent In this approach, asymmetry is induced in a prochiral molecule or functional group by reaction with a stoichiometric amount of an en- antio-enriched reagent system The reaction proceeds through diastereomeric transition states, resulting in the preferential formation of one enantiomer or diastereomer Current reagents can lack generality and may be difficult to prepare in both chiral forms At least one equivalent of the chiral component is required, which can present economical and practical difficulties Many examples are provided by the reduction of double bonds, especially ketones Ketone (130) was reduced enantioselectively using either (+) or (-)-bchlorodiisopinocampheylborane (149) Reduction with (-)-b-chlorodiisopinocampheylborane generated the alcohol (8)-(131), which was transformed into the (1R,3S)-isochroman compound, (lR,3S)-(132), through a ste- 814 Chirality and Biological Activity OTBDMS - Ph Ph Ph 9- 95% eel using opposite enantiomers of the chiral oxaziridine (-)-(I341 was found to be a better activator of a cloned BK channel than the generating a current (+)-isomer at 20 increase of 141% compared with 124% for (+)-(134) a, 6.4 Chiral Catalyst The use of a chiral catalyst represents the ideal method for asymmetric synthesis because only small amounts of the chiral mediator are required and no modifications of the prochiral substrate are necessary In many systems both enantiomers of the product cgn be prepared in a predictable and reproducible manner The pharmaceutical industry is particularly interested in the capability of new catalyst systems to operate as reliable manufacturing processes on a large scale (127,152) Substantial effort continues to be expended by the synthetic organic community with the goal of extending the number of efficient and broadly applicable catalyst systems capable of generating high levels of enantiomeric excess in a wide range of substrates (127) The reduction of ketones by borane catalyzed by chiral oxazaborolidines such as (136), derived from the enantiomeric amino alcohols, has been applied to the synthesis of several drug candidates (127) This system is known as the CBS (Corey, Bakshi, Shibata) reduction (1531, and Corey himself has applied it to the synthesis of pharmaceutical compounds (154) A further example is provided by the synthesis of MK-499 (1371, a Asymmetric Synthesis -78"C, 89% MeLi, THF H NH2.HCI < HCI, 23% i-' F H2, Pt02.H20 OMe (119) (IC50= 1.11 1M) (118) (de > 99%) NPS 1407 (123) (ICSo= 0.089pM) (122) Figure 18.42 potassium channel blocker that was developed for the treatment of cardiac arrhythmia by Merck (155) Asymmetric hydrogenations with transition metal catalysts have been applied to single enantiomer synthesis in the pharmaceutical industry with considerable success ChiroTech and Pfizer developed an improved synthesis of glutarate derivative (1391, an intermediate required for the synthesis of Candoxatril (140) (156).The drug, a neutral endopeptidase inhib- itor, was in clinical development for the treatment of hypertension and congestive heart failure, and its enantiomer does not possess the same biological activity Several catalysts and conditions were screened before arriving at optimized conditions using cationic rhodium(R,R)-MeDuPHOS (141), which provided the product with complete enantioselectivity and avoided previously observed problems associated with isomerization of the enone starting material The reaction could be conducted at a Chirality and Biological Activity Lamivudine (129) Figure 18.43 high substrate-to-catalyst ratio of 3500:l without a detrimental effect on enantiomeric excess or reaction rate In catalytic asymmetric reactions, it is clearly economically advantageous to minimize the amount of catalyst that may comprise expensive chiral material and transition metals A method for the asymmetric dihydroltylation of alkenes to yield cis-diols was developed by the research group of Sharpless using chiral ligands derived from the cinchona alkaloids dihydroquinidine (DHQD) and dihydroquinine (DHQ) with a catalytic amount of osmium tetroxide (157) Although they are diastereomers, the phthalazine ligands act as "pseudo-enantiomeric" ligands, i.e., they give opposite asymmetric induction in a predictable manner This procedure was recently used to prepare both isomers of combretadioxolane (144), a chiral analog of the natural product Combretastatin A-4 (146) (158) Com- bretastatin A-4 displays antitubulin activity and cytotoxicity to tumor cells and is therefore an interesting lead structure for new anticancer drugs The asymmetric synthesis of (S,S)combretadioxolane (144) involved treatment of the trans-stilbene (142) with "AD-mix-a" [containing (DHQ),-PHAL] (145), whereas the enantiomer (R,R)-combretadioxolane resulted from use of AD-mix-& which contains (DHQD),-PHAL as the chiral ligand The tubulin polymerization-inhibitory activity of (S,S)-wmbretadioxolane was comparable with combretastatin A-4 (IC,, = 4- CL2M) in an in vitro assay, whereas (R,R)-combretadioxolane was essentially inactive (IC,, > 50 CLM) In addition, (23,s)-combretadioxolane was 20 times more potent than vincristine as an in vitro growth inhibitor of the multidrug-resistant cell line PC-12 Workers at SmithKline Beecham reported the stereoselective synthesis of inhibitors of Asymmetric Synthesis Figure 18.44 Figure 18.45 the cysteine protease cathepsin K (159) A procedure was sought to allow preparation of either enantiomer of azido alcohol (148) This was readily achieved by Jacobsen asymmetric desymmetrization of the meso-epoxide (147) using azidotrimethylsilane catalyzed by chromium salen complex (149) (160) Use of the (R,R)-salen catalyst shown generated (3S, 4R)-(1481, whereas the (S,S)-catalyst provided the (3R,4S)-azidosilyl alcohol, both with very high enantioselectivity After removal of the silyl group and reduction of the azido moiety, the resultant enantiomeric amino alcohols were transformed into diastereomers (4s)-and ( a ) - ( )by reaction with leucine, amide formation, and oxidation The cathepsin K activity for the diastereomers showed the (4s)-isomers to be up to 40-fold more potent than the corresponding (a)-(150)in an enzyme assay A large scale synthesis of the drug Nelfinavir, an HIV protease inhibitor developed by Agouron (now Pfizer) was reported with the amino alcohol derived from (1481, prepared using the Jacobsen procedure described above (161) A similar approach uses the chromiumSalen complex (149) to open racemic terminal epoxides in a highly efficient resolution pro- Chirality and Biological Activity 3.136 (10mol%) MeOH Ph MK 499 (137) (98%de; 92% yield) Figure 18.46 cess that has been applied to the synthesis of biologically active compounds (162) As with any such resolution process, the maximum yield of enantiopure material is 50%based on starting material Terminal epoxides are easy to prepare in racemic form, and conversely, difficult to prepare as single enantiomers by epoxidation of the corresponding alkene (R)9-[2-(phosphonomethoxy)propyl]adenine (RPMPA) is a nucleotide reverse transcriptase """P t-BuO C02Na (138) [((RR)-Me-DuPHOS)Rh(COD)]BF4 Hz (5 atm)/MeOH (COD = 1,5-cyclooctadiene) Me ,,, t-BuO Meow* C02Na (139) (>99%ee, 95% yield) mO\ Me (R,R)-Me-DuPHOS (141) Figure 18.47 Candoxatril (140) C02H Asymmetric Synthesis 819 OMe OMe (S,S)-(143)(s99%ee;89% yield) OMe Cornbretastatin A-4 (146) Figure 18.48 inhibitor being developed by Gilead Sciences and a collaborative group from the University of Washington for the treatment and prevention of HIV infection (163) The compound can be prepared through kinetic resolution of propylene oxide using (S,S)-(149) and the resultant (R)-1-amino-2-propanol (153)was transformed into (R)-PMPA (154) in five steps (162) In 1997, Tokunaga et al reported the hydrolytic kinetic resolution of racemic terminal epoxides using a Co(II1)-Salen catalyst (164) This remarkably general process uses only water as the nucleophile and provides the synthetically useful chiral epoxides and diols in highly enantioenriched form The catalyst can be recycled and the reactions conducted under solvent-free conditions The process has been used by academic and industrial groups and is operated by Rhodia ChiRex on a large scale (165) A wide variety of synthetic processes have been rendered asymmetric through the use of a chiral catalyst In addition to the types of reaction described above, chiral transition metal catalysts have been used to influence the stereochemical course of isomerization, cyclization, and coupling reactions As an example, an approach towards the natural product (-)-epibatidine (158) was recently reported by Namyslo and Kaufmann (166) Epibatidine is a potent analgesic and a nicotinic receptor agonist The synthesis involves an asymmetric Heck-type hydroarylation between the bicyclic alkene (155) and pyridyl iodide (156) A number of bidentate chiral li- Chirality and Biological Activity (147) (3S,4R)-(148) (98%ee;93% yield) H ++H t-Bu t-Bu Figure 18.49 gands were investigated with BINAP (1591, which were observed to give the highest enantioselectivity By using the (R)-or (8)-BINAP ligand, both enantiomers of (157) were accessible with about the same level of enantioselection The continuing development of efficient and practical asymmetric processes will be one of the major driving forces in the future of drug discovery and development In particular, the design of new general and practical catalytic processes will help explore the link between chirality and biological activity CONCLUSIONS The ultimate focus of the endeavors of medicinal chemists is to develop a successful drug that will cure patients However, with the increased regulatory requirements within the competitive biotechnology and pharmaceutical industry, the initial research to achieve this objective must be conducted in a rapid and thorough manner During the drug research and development process, the important and subtle relationship between chirality and biological activity should be carefully considered TMSN3 (0.5 equiv.) Steps O y ' Me Me (152) (97%ee; Figure 18.50 NH2 Me (153) (84%yield) References (157) (81%ee;53% yield) Epibatidine (158) Figure 18.51 Enantiomers frequently display markedly different biological activity; however, the fact that a large and adaptable toolbox of chemical and biological techniques to obtain single isomers are available allows the medicinal chemist to avoid working with mixtures of stereoisomers As reviewed in this chapter, there are numerous synthetic strategies available to the medicinal chemist that offer their own particular drawbacks and advantages In the early stages of research it may be preferable to separate isomers by chromatography, thus providing both single enantiomers for biological testing - It should be noted that all the techniques described in this chapter can be used in conjunction with one another That is to say, if one technique such as asymmetric synthesis failed to deliver enantiopure material, then another technique such as crystallization can be used to push through the product to the desired purity As an example of this "double" approach, the use of SMB and crystallization in the separation of mandelic acid is worthy of note (56) The use of asymmetric hydrogenation followed by asymmetric enzymic transformation to obtain single isomer products has also been described by Taylor et al at ChiroTech (167) In conclusion, if a chiral center is present in a molecule designed and synthesized by a medicinal chemist, there are a broad number of methods available to prepare or isolate either isomer From the examples given in this chapter, stereoisomers frequently display markedly different biological properties where the desirable properties associated with one isomer may not be apparent when the corresponding racemic mixture is tested either in vivo or in vitro REFERENCES A Michaels and J Fuller, Financial Times (Lond.),23,21(2001) R B Raffa, E Friderichs, W Reimann, R P Shank, E E Codd, J L Vaught, H I Jacoby, and N Selve, J Pharmacol Exp Ther., 267, 331-340(1993) M B Smith and J March, March's Advanced Organic Chemistry, 5th ed., Wiley, New York, 2001,pp 125-217;E.L Eliel, S H Wilen, and L N Mander, Stereochemistry of Organic Compounds, Wiley, New York, 1994; E L Eliel, S H Wilen, and M P Doyle, Basic Organic Stereochemistry, Wiley, New York, 2001 T D Stephens, Chem Br., 37,38(2001) J J Baldwin and W B Abrams in I W Wainer and D E Drayer, Eds., Stereochemically Pure Drugs: An Industrial Perspective, Marcel Dekker, New York, 1988,p 311 G C Cotzias, P S Papavasiliou, R Gellene, N Engl J Med., 280,337(1969) H.K.Kroemer, J Turgeon, R A Parker, and D M Roden, Clin Pharmacol Ther., 46,584 (1989) Chirality and Biological Activity R A O'Reily, Clin Pharmacol Ther., 16, 348 (1974) A Breckenbridge, M Orme, H Wesseling, R J Lewis, and R Gibbons, Clin Pharmacol Ther., 15,424 (1974) 10 T Walle, J G Webb, E E Bagwell, U K Walle, H B Daniell, and T E Gaffney, Biochem Pharamacol., 37, 115 (1988) 11 W Lindner, M Rath, K Stoschitzky, and H J Semmelrock, Chirality, 1, 10 (1989) 12 D E Drayer in I W Wainer and D E Drayer, Eds., Drug Sterochemistry-Analytical Methods and Pharmacology, Marcel Dekker, New York, 1988, p 209 13 K J Fehske, W E Muller, and U Wollert, Biochem Pharmacol., 30,687 (1981) 14 R H McMenamy and J L Oncley, J Biol Chem., 233,1436 (1958) 15 W E Muller in I W Wainer and D E Drayer, Eds., Drug Sterochemistry-Analytical Methods and Pharmacology, Marcel Dekker, New York, 1988, p 227 16 S Toon, L K Low, M Gibaldi, W F Trager, R A O'Reily, C H Motley, and D A Goulart, Clin Pharmacol Ther., 39, 15 (1986) 17 M A Campanero, B Calahorra, M Valle, I F Troconiz, and J Honorato, Chirality, 11, 272 (1999) 18 R Stevenson, Chem Br., 37,24 (2001) 19 N M Maier, P Franco, and W Lindner, J Chromatogr A, 906,3 (2001) 20 L Miller, C Orihuela, R Fronek, D Honda, and Dapremont, J Chromatogr A, 849,309 (1999) 21 V M Meyer, Chirality, 7, 567 (1995);0 P Kleidernigg, M Lammerhofer, and W Lindner, Enantiomer, 1,387 (1996) 22 V Schurig, J Chromatogr.441,135 (1988);K Watabe, S C Chang, E Gil-Av, and B Koppenhofer, Synthesis, 3,225 (1987) 23 E R Francotte in S Ahuja, Ed., Chiral Separations, Applications and Technology, chap 10, American Chemical Society, Washington DC, 1997, p 271 24 K D Altria, N W Smith, and C H Turnbull, Chromatographia, 46,664 (1997);K D Altria, M A Kelly, and B J Clark, Trends Anal Chem., 17,214 (1998) 25 K L Williams, L C Sander, and S A Wise, J Pharm Biomed Anal., 15, 1789 (1997);N Bargrnann-Leyder, A Tambute, and M Claude, Chirality, , 1 (1995) 26 M Juza, M Mazzotti, and M Morbidelli, Trends Biotechnol., 18, 108 (2000) 27 J T F Keurentjes and F J M Voermans in A N Collins, G N Sheldrake, and J Crosby, Eds., Chirality and industry 11 Developments in the Manufacture of Optically Active Compounds, chap 8,Wiley, NewYork, 1997, p 157 28 S C Stinson, Chem Eng News, 73,44 (1995) 29 E Francotte, J Chromatogr A, 666, 565 (1994) 30 CHIRBASE, available online at http:// chirbase.u-3mrs.fr, accessed on July 29,2002 31 Daicel Chemical Industries, Ltd., available online at http://www.daicel.co.jp/chiral,accessed on July 29, 2002 NOVASEP, available online at http://www.novasep.com, accessed on July 29, 2002 Astec, available online at http:ll www.astecusa.com, accessed on July 29, 2002 32 D Boyd, M O'Keeffe,and M R Smyth, Analyst, 119, 1467 (1994) 33 D A von Deutsch, I K Abukhalaf, L E Wineski, H Y Aboul-Enein, S A Pitts, B A Parks, R A Oster, D F Paulsen, and D E Potter, Chirality, 12,637 (2000) 34 B Waldeck, E Widmark, Acta Pharmacol Toxicol., 56,221-227 (1985) 35 D J Triggle, D A Langs, and R A Janis, Med Res Rev., , 123 (1989);V C Njar and A M H Brodie, Drugs, 58,233 (1999) 36 S Visentin, P Amiel, A Gasco, B Bonnet, C Suteu, and C Roussel, Chirality, 11, 602 (1999) 37 P Tullio, A Ceccato, J-F Liegeois, B Pirotte, A Felikidis, M Stachow, P Hubert, J Crommen, J Geczy, and J Delarge, Chirality, 11, 261 (1999) 38 J Bruhwyler, J F Liegeois, J Gerardy, J Damas, E Chleide, C Lejeuns, E Decamp, P De Tullio, J Delarge, A Dresse, and J Geczy, Behav Pharmacol., , 731 (1998) 39 T Shibata, I Okamoto, and K Ishii, J Liq Chromatogr., 9,313 (1986);E Yashima andY Okamoto, Bull Chem Soc Jpn., 68, 3289 (1995) 40 W H Pirkle, D W House, and J H Finn, J Chromatogr., 192, 143 (1980) 41 J N Kinkel in A Subramanian, Ed., A Practical Approach to Chiral Separations by Liquid Chromatography, VHC, New York, 1994 42 S G Allenmark, S Andersson, P Moller, and D Sanchez, Chirality, 7,248 (1995) 43 M Meurer, U Altenhoner, J Straube, and H Schmidt-Traub, J Chromatogr A, 769, 71-79 (1997) 44 M Schulte, R Ditz, R M Devant, J N Kinkel, and F Charton, J Chromatogr A, 769, 93 (1997) References 45 D Y Wang, F Hanotte, C De Vos, and P Clement, Eur J Allerg a n d Clin Immunol., 56, 339 (2001) 46 E J Corey and C J Helal, Tetrahedron Lett., 37,4837 (1996) 47 D A Pflum, H Scot Wilkinson, G J Tanoury, D W Kessler, H B Kraus, C H Senanayake, and S A Wald, Org Process Res Dev., , 110 (2001) 48 E R Francotte and P Richert, J Chromatogr A, 769, 101 (1997);M Negawa and F Shoji, J Chromatogr., 590, 113 (1992) 49 G Ganetsos, P E Barker, J A Johnson, R G Kabza, K Hashimoto, S Adachi, Y Shirai, M Morishita, B Balannec, G Hotier, and H Makai in G Ganetsos and P E Barker, Eds., Preparative and Production Scale Chromatography, chaps 11-15, Marcel Dekker, New York, 1993, p 233; D B Broughton and C G Gerhold, inventors, Universal Oil Prod Co., assignee, US patent 2,985,589, May 23, 1961 50 D M Ruthven, Principles of Adsorption and Adsorption Processes, chap 12, Wiley, New York, 1984, p 380 51 E R Francotte, Chim Nouvelle, 53, 1541 (1996) 52 D W Guest, J Chromatogr A, 760, 159 (1997) 53 M Schulte and J Strube, J Chromatgr A, 906, 399 (2001) 54 K E Goeringer, B K Logan, and G D Christian, J Anal Toxicol 21, 529 (1997) 55 E Cavoy, M.-F Deltent, S Lehoucq, and D Miggiano, J Chromatogr A, 769,49 (1997) 56 H Lorenz, P Sheehan, and A Seidel-Morgenstern, J Chromatogr A, 908,201 (2001) 57 J Jacques, A Collet, and S H Wilen, Enantiomers, Racemates a n d Resolutions, Krieger: Malabar, Florida, 1994 58 A Collet, M J Brienne, and J Jacques, Chem Rev., 80,215 (1980) 59 S H Wilen in E L Eliel, Ed., Tables ofResolving Agents and Optical Resolutions, University of Notre Dame Press, Notre Dame, IN, 1972; P Newman, Optical Resolution Procedures for Chemical Compounds, vol 1-3, Optical Resolution Center, Manhattan College, New York, 1978-1984 60 M J Cannarsa, Chimica Oggi, 17,28 (1999) 61 M Prashad, D Har, Repic, T J Blacklock, and P Giannousis, Tetrahedron Asymmetry, 10,3111 (1999) 62 R A Maxwell, E Chaplin, S B Eckhardt, J R Soares, and G Hite, J Pharmacol Exp Ther, 173,158 (1970) 63 S Faulconbridge, H S Zavareh, G R Evans, and M Langston, inventors, Medeva Europe Ltd (GB), assignee, World patent W0981 25902, June 18, 1998 64 A S C Chan, Chemtech., , (1993) 65 P J Harrington and E Loderwijk, Org Process Res Dev., 1, 72 (1997) 66 W J Pope and S J Peachey, J Chem Soc., 75, 1066 (1899) 67 R Gristwood, H Bardsley, H Baker, and J Dickens, J Exp Opin Invest Drugs, , 1209 (1994) 68 M Langston, U C Dyer, G A C Frampton, G Hutton, C J Lock, B M Skead, M Woods, and H Zavareh, Org Process Res Dev., 4, 530 (2000) 69 B A Astleford and L Weigel in A N Collins, G N Sheldrake, and J Crosby, Eds., Chiralty in Industry I, chap 6, Wiley, New York, 1997, p 99 70 K Flick and E Frankus, inventors, Gruenenthal Chemie, assignee, US patent 3,652,589, March 28, 1972 71 G R Evans, inventor, Darwin Discovery Ltd., (US) assignee, World Patent W000132554, June 8,2000; G R Evans, J A Henshilwood, and J O'Rourke, Tetrahedron Asymmetry, 12, 1663 (2001) 72 L G Humber, Med Res Rev., , l (1987) 73 L G Humber, J Med Chem., 29,871 (1986) 74 M Woods, U C Dyer, J F Andrews, C N Morfitt, R Valentine, and J Sanderson, Org Process Res Dev., 4,418 (2000) 75 D Askin, K K Eng, R M Purick, K M Wells, R P Volante, and P J Reider, Tetrahedron Lett., 35,673 (1994) 76 M Kottenhahn, K Stingl, and K Drauz, inventors, Degussa (DE), assignee, US patent 6,093,823, July 25,2000 77 M Eichelbaum, Federation Proc., 43, 2298 (1984);M Eichelbaum, Biochem Pharmacol., , (1988) 78 H Echizen, T Brecht, S Neidergesass, B Volgelgesang, and M Eichelbaum, Am Heart J.109,210 (1985) 79 Ehrmann, H Nagel, and W Karau, inventors, Knoll Aq (DE), assignee, US patent 5,457,224, October 10, 1995 and World patent WO94/08950,April 14,1994 80 E J Trieber, M Raschack, and F Dengel, inventors, Knoll Aq (DE), assignee, German patent 2059923,1972 81 R M Bannister, M H Brookes, G R Evans, R B Katz, and N D Tyrrell, Org Process Res Dev., 4,467 (2000) ' Chirality and Biological Activity 82 P E Graves, H A Salhanick, Endocrinology, 105, 52 (1979) 83 See ref 57, pp 299-301,382-383 84 N Finch, R Dziemian, J Cohen, and B G Steinetz, Experientia, 31, 1002 (1975) 85 M J Bunegar, U C Dyer, G R Evans, R P Hewitt, S W Jones, N Henderson, C J Richard, S Sivaprasad, B M Skead, M A Stark, and E Teale, Org Process Res Dev., 3, 442 (1999) 86 See ref 57, pp 374-375 87 K P Datla and G Curzon, Neuropharmacology, 35, 315 (1996); S Salvadori, C Bianchi, L H Lazurus, V Scaranari, M Attila, and R Tomatis, J Med Chem., 35,4651 (1992) 88 C A Maryanoff, L Scott, R D Shah, and F J Villani Jr., Tetrahedron Asymmetry, 9, 3247 (1998) 89 A Bruggink in A N Collins, G N Sheldrake, and J Crosby, Eds., Chirality in Industry 11, chap 5, Wiley, New York, 1997, p 81 90 T Vries, H Wynberg, E van Echten, J Koek, W ten Hoeve, R M Kellogg, Q B Broxterman, A Minnaard, B Kaptien, S van der Sluis, L Hulshof, and J Kooistra, Angew Chem Int Ed., 37,2349 (1998) 91 E Ebbers, G J A Arianns, B Zwanenburg, and A Bruggink, Tetrahedron Asymmetry, 9, 2745 (1998);U C Dyer, D A Henderson, and M B Mitchell, Org Process Res Dev., 3, 161 (1999) 92 S H Wilen, A Collet, and J Jacques, Tetrahedron, 33,2725 (1977) 93 Z J Li, M T Zell, E J Munson, and D J W Grant, J Pharm Sci., 88,337 (1999) 94 R Tamura, T Ushio, H Takahashi, K Nakamura, N Azuma, F Toda, and K Endo, Chirality, 9, 220 (1997) 95 H M L Davies, T Hansen, D W Hopper, and S A Panaro, J Am Chem Soc., 121, 6509 (1999) 96 J M Axten, R Ivy, L Krim, and J D Winkler, J Am Chem Soc., 121,6511 (1999) 97 A Robinson, H Y Li, and J Feaster, Tetrahedron Lett., 37,8321 (1996) 98 W Goppel, W Betram, Psychiatr Neurol Med Psychol., 23, 712 (1971) 99 H A M Mucke, Drugs Future, 33,259 (1997) 100 D H R Barton and G W Kirby, J Chem Soc., 806 (1962) 101 W Shieh and J A Carlson, J Org Chem., 59, 5463 (1994) B Kuenburg, L Czollner, J Frohlich, and U Jordis, Org Process Res Dev., 3,425 (1999) G L Plosker and K L Goa, Drugs, 42, 805 (1991) H Harada, T Morie, Y Hirokawa, and S Kato, Tetrahedron Asymmetry, 8,2367 (1997) R M Williams in J E Baldwin and P D Magnus, Eds., Synthesis of Optically Active '61Amino Acids, vol 7, Pergamon Press, Oxford, UK 1989; R Duthaler, Tetrahedron, 50, 1539 (1994) T Shiraiwa, H Miyazaki, T Watanabe, and H Kurokawa, Chirality, 9,48 (1997) T Shiraiwa, H Miyazaki, A Ohta, K Motonaka, E Kobayashi, M Kubo, and H Kurokawa, Chirality, 9, 656 (1997) K Abe, H Inoue, and T Nagao, Yakugaku Zasshi, 108, 716 (1988) S Yamada, K Morimatsu, R Yoshioka, Y Ozaki, and H Seko, Tetrahedron Asymmetry, 9, 1713 (1998) E K Rowinsky, L A Cazenav, and R C Donehower, J Natl Cancer Inst., 82,1247 (1990) J N Denis, A E Green, D Guenard, F Gueritte-Voegelein, L Mangatal, and P Potier, J Am Chem Soc., 110, 5917 (1988) J Kearns and M M Kayser, Tetrahedron Lett., 35, 2845 (1994) R P Srivastava, J K Zjawiony, J R Peterson, and J D McChesney, Tetrahedron Asymmetry, 5, 1683 (1994) R N Patel, Adv Appl Microbiol., 47, 33 (2000); V Gotor, Biocat and Biotrans 18, 87 (2000);W A Loughlin, Bioresource Technology, 74, 49 (2000); S M Roberts, J Chem Soc., Perkin Trans., 1, (1999) C.-H Wong and G M Whitesides, Enzymes in Synthetic Organic Chemistry, Pergamon, New York, 1994 A D Baxter, J B Bird, R Bannister, R Bhogal, D T Manallack, R W Watson, D A Owen, J Montana, J Henshilwood, and R C Jackson in N Clendeninn and K Appelt, Eds., Matrix Metalloproteinase Inhibitors in Cancer Therapy, Humana Press, Totawa, NJ, 2000, pp 193-221 117 N L St Clair and N L Nathrough, J Am Chem Soc., 114,7314 (1992);N Khalaf, C P Govardan, J J Lalonde, R A Persichetti, Y F Wang, and A L Margolin, J Am Chem Soc., 118, 5495 (1996) 118 A M Evans, J Clin Pharmacol.36 (Suppl 12), 7s (1996) References 119 S.-W Tsai and H J Wei, Enzyme Microb Technol., 16,328(1994) 120 J.-Y Xin, S.-B Li, Y Xy, J.-R Chui; and C.-G Xi, J Chem Tech Biotechnol., 76,579(2001) 121 M E Swarbrick, F Gosselin, and W D Lubell, J Org Chem., 64,1993-2002(1999) 122 W L Nelson and T R Burke, J Org Chem., 43,3641(1978) 123 R M Patel, Stereosel Biocatal., Marcel Dekker, Inc., New York, 2000, pp 87-130 124 E P Siqueira Fihlo, J A R Rodrigues, and P J S Moran, Tetrahedron Asymmetry, 12, 847(2001) 125 P A Procopiou, G E Morton, M Todd, and G Webb, Tetrahedron Asymmetry, 12, 2005 (2001) 126 T P Jerusi, inventor, Sepracor Inc (US),assignee, World patent W099113867,March 25, 1999 127 J D Morrison, Ed., Asymmetric Synthesis, Academic Press, San Diego, CA, 1983;G Procter, Asymmetric Synthesis, Oxford University Press, Oxford, UK, 1996; M N6gradi, Stereoselective Synthesis, 2nd ed., VCH, Weinheim, Germany, 1995; D J Ager and M L East, Eds., Asymmetric Synthetic Methodology, CRC Press, Boca Raton, FL, 1995; D J Ager, Ed., Handbook of Chiral Chemicals, Marcel Dekker, New York, 1999; P O'Brien, J Chem Soc., 1, 95-113 (2001); H Tye and P J Comino, J Chem Soc., 1, 1729-1747 (2001); P I Dalko and L Moisan, Chem Int Ed., 40, 3726-3748 (2001); K C Nicolaou and E J Sorensen, Classics in Total Synthesis, VCH, Basel, Switzerland, 1996 128 S Redshaw in C R Ganellin and S M Roberts, Eds., Medicinal Chemistry, 2nd ed., Academic Press, San Diego, 1993, pp 163-186 129 A A Patchett, E Harris, E W Tristram, M J Wyvratt, M T W u , D Taub, E R Peterson, T J Ikeler, J ten Broeke, L G Payne, D L Ondeyka, E D Thorsett, W J Greenlee, N S Lohr, R D Hoffsommer, H Joshua, W V Ruyle, J W Rothrock, S D Aster, A L Maycock, F M Robinson, R Hirschmann, C S Sweet, E H Ulm, D M Gross, T C Vassil, and C A Stone, Nature, 288,280(1980) 130 T J Blacklock, R F Shuman, J W Butcher, W E Shearin, J Budavari, and V J J Grenda, J Org Chem., 53,836(1988) 131 T Ohashi and J Hasegawa in A N Collins, G N Sheldrake, and J Crosby, Eds., Chirality i n Industry II: Developments in the Commercial Manufacture andApp1ication.s of Optically Active Compounds, Wiley, Chichester, UK, 1997, p 269 132 M R Attwood, C H Hassall, A Krohn, G Lawton, and S Redshaw, J Chem Soc., 1, 1011-1019(1986) 133 E J Stoner, A J Cooper, D A Dickman, L Kolaczkowski, J E Lallaman, J.-H Liu, P A Oliver-Shaffer,K M Patel, J B Paterson Jr., D J Plata, D A Riley, H L Sham, P J Stengel, and J.-H J Tien, Org Process Res Dev., 4, 264-269(2000) 134 J T Kovalainen, J A M Christiaans, S Kotisaari, J T Laitinen, P T Mannisto, L Tuomisto, and J Gynther, J Med Chem., 42, 1193-1202(1999) 135 G M Coppola and H F Schuster, Chiral a -Hydroxy Acids in Enantioselective Syntheses, VCH,Weinheim, Germany, 1997 136 G.Wang and R Hollinsworth, J Org Chem., 64,1036-1038(1999) 137 G G W u , Org Process Res Dev., 4, 298-300 (2000); G W u , Y S Wong, X Chen, and Z Ding, J Org Chem., 64, 3714-3718(1999) 138 D.S Garvey, J T Wasicak, M W Decker, J D Brioni, M J Buckley, J P Sullivan, G M Carrera, M W Holladay, S P Arneric, and M Williams, J Med Chem., 37, 1055-1059 (1994) 139 N.-H Ling, Y He, and H Kopecka, Tetrahedron Lett 36,2563-2566(1995) 140 D J Ager, I Prakash, and D R Schaad, Chem Rev., 96,835-875(1996) 141 D A Evans, J M Takacs, L R McGee, D J Mathre, and J Bartroli, Pure Appl Chem., 53, 1109 (1981); D A Evans, M D Ennis, and D J Mathre, J Am Chem Soc., 104, 1737 (1982);D.A Evans, Aldrichimica Acta, 15,23 (1982) 142 D R.Schaad i n ref 127, pp 287-300 143 S J Wittenberger and M A McLaughlin, Tetrahedron Lett., 40,7175-7178(1999) 144 S R.Turner, J W Strohbach, R A Tommasi, P A Aristoff, P D Johnson, H I Skulnick, L A Dolak, E P Seest, P K Tomich, M J Bohanon, M.-M Horng, J C Lynn, K T Chong, R R Hinshaw, K D Watenpaugh, M N Janakirarnan, and S Thaisrivongs, J Med Chem., 41,3467-3476(1998) 145 T M Judge, G Phillips, J K Morris, K D Lovasz, K R Romines, G P Luke, J Tulinsky, J M Tustin, R A Chrusciel, L A Dolak, S A Mizsak, W Watt, J Morris, S L Vander Velde, J W Strohbach, and R B Gammill, J Am Chem Soc., 119,3627-3628(1997) 826 146 N Cabedo, I Andreu, M C Ramirez de Arellano, A Chagraoui, A Serrano, A Bermejo, P Protais, and D Cortes, J Med Chem., 44, 1794-1801 (2001) 147 S T Moe, D L Smith, E G DelMar, S M Shimizu, B C Van Wagenen, M F Balandrin, Y Chien, J L Raszkiewicz, L D Artman, H S White, and A L Mueller, Bioorg Med Chem Lett., 10,2411-2415(2000) 148 M D Goodyear, P 0.Dwyer, M L Hill, A J Whitehead, R Hornby, and P Hallett, inventors, Glaxo Group Ltd (GB), assignee, World patent WO-09529174,April 18,1995 149 M P DeNinno, R Schoenleber, R J Perner, L Lijewski, K E Asin, D R Britton, R MacKenzie, and J W Kebabian, J Med Chem., 34,2561-2569 (1991) 150 P Hewawasam, N A Meanwell, V K Gribkoff, S I Dworetzky, and C G Boissard, Bioorg Med Chem Lett., 7,1255-1260(1997) 151 F A Davis and B.X Chen, Chem Rev., 92, 919(1992) 152 M K O'Brien and B Vanase, Curr Opin Drug Discovery Dev., 3,793-806(2000);R.A Sheldon, Ed., Chirotechnology: Industrial Synthesis ofOptically Active Compounds, Marcel Dekker, New York, 1993 153 E J Corey, R K Bakshi, and S Shibita, J.Am Chem Soc.109,5551(1987);E J Corey, R K Bakshi, S Shibita, C.-P Chen, andV K Singh, J Am Chem SOC.,109,7925(1987) 154 E.J Corey and J 0.Link, Tetrahedron Lett., 31,601 (1990); E J Corey and J Link, J Org Chem.,56,442(1991) 155 Y.-J Shi, D Cai, U.-H Dolling, A W Douglas, D M Tschaen, and T R Verhoeven, Tetrahedron Lett., 35,6409-6412(1994) Chirality and Biological Activity 156 M J Burk, F Bienewald, S Challenger, A Derrick, and J A Rarnsden, J.Org Chem.,64, 3290-3298(1999) 157 H C Kolb, M S VanNieuwenhze, and K B Sharpless, Chem Rev., 94,2483(1994) 158 R Shirai, H Takayama, A Nishikawa, Y Koiso, and Y Hashimoto, Bioorg Med Chem Lett., 8,1997-2000 (1998) 159 A E Fenwick, A D Gribble, R J Ife, N Stevens, and J Witherington, Biorg Med Chem Lett., 11,199-202 (2001) 160 L E Martinez, J L Leighton, D H.Carsten, and E N Jacobsen, J Am Chem Soc., 117, 5897 (1995);S E Schaus, J F Larrow, and E N Jacobsen, J Org Chem.,62,4197-4199 (1997) 161 S E Zook, J K Busse, and B C Borer, Tetrahedron Lett., 41,7017-7021 (2000) 162 J F Larrow, S E Schaus, and E N Jacobsen, J Am Chem Soc., 118, 7420-7421 (1996); S E Schaus and E N Jacobsen, Tetrahedron Lett., 37,7937-7940 (1996) 163 C.-C Tsai, K E Follis, A Sabo, T W Beck, R F Grant, N Bischofberger, R E Benveniste, and R Black, Science,270,1197-1199 (1995) 164 M Tokunaga, J F Larrow, F Kakiuchi, and E N Jacobsen, Science, 277,936 (1997) 165 J M Keith, J F Larrow, and E N Jacobsen, Adv Synth Catal., 343,5-26 (2001) 166 J H Namyslo and D E Kaufmann, Synl@t., 804-806(1999) 167 S J C Taylor, K E Holt, R C Brown, P A Keene, and I N Taylor in R N Patel, Ed., Stereoselective Biocatalysis, Marcel Dekker, New York, 2000,p 397 CHAPTER NINETEEN Structural Concepts in the Prediction of the Toxicityof Therapeutical Agents HERBERT S ROSENKRANZ Department of Biomedical Sciences Florida Atlantic University Boca Raton, Florida Contents Introduction, 828 1.1 Development of Database, 828 1.2 Model Building, 829 1.3 Model Characterization, 831 1.4 Model Validation, 834 1.5 Applications and Mechanistic Studies, 836 Conclusions, 844 Acknowledgments, 844 Burger's Medicinal Chemistry and Drug Discovery Sixth Edition, Volume 1: Drug Discovery Edited by Donald J Abraham ISBN 0-471-27090-3 O 2003 John Wiey & Sons, Inc Structural Concepts in the Prediction of the Toxicity of Therapeutical Agents INTRODUCTION The increased acceptance and availability of various structure-activity relationship (SAR) approaches in health hazard identification (1, 2) is accompanied by many opportunities and some pitfalls The latter are derived from the availability of various computer-based SAR platforms whose basis and performance characteristics are not transparent to the user Such programs, in the hand of the non-expert, may be misused SAR models and associated technologies on the other hand, while not crystal balls, provide the expert toxicologist with meaningful information regarding the putative toxicological profile of candidate agents They can guide in the design of agents with decreased or without unwanted side effects and yet retain or even enhance therapeutic effectiveness Finally, the SAR technology can provide insights into the mechanism whereby a chemical exerts its toxic effects and thereby provide a better understanding of the risk that the agent poses to humans However, to achieve these aims, it is essential that the performance characteristics and the basis of the SAR model be known This involves several critical steps in the SAR model development These are listed below, each of which will be amplified: Development of database Model building Model characterization 'Model validation Model application to individual agents and for mechanistic evaluations Although the present review focuses on the MULTICASE SAR methodology (3-51, the concepts discussed herein apply to all generally available SAR techniques used to study toxicological phenomena Basically SAR approaches that have been used fall into two categories: those that are based on statistical automated algorithms not dependent on prior expert judgment and those that are a priori Table 19.1 Some SAR Approaches Used in Toxicology Designation MULTICASE TOPKAT COMPACT DEREK ONCOLOGIC Hazard Expert PROGOL Structural Alerts Approacha References I I I I1 I1 I1 I I1 3-5 8-10 11,12 13 14 15 "Approach I indicates statistical automated algorithms not dependent on prior expert judgment Approach I1 indicates a rule-based technique that requires prior expert judgment rule-based requiring prior expert knowledge (Table 19.1) However, as will be stressed herein, even the approach not requiring prior expert input is very much dependent on human expertise at various stages of the model development and interpretation process Reviews and assessments of the various SAR methodologies used to analyze toxicological phenomena are available (16-21) I Development of Database Most experimental data compilations of to%cological effects both in the public domain as well as in proprietary databases were not developed for SAR purposes Thus, with respect to some toxicological phenomena, the database may be rich with certain chemical classes such as chloroarene and lacking in data relating to others, e.g., aminoarenes Yet, unlike the SAR models developed for drug discovery, SAR models of toxicological phenomena must be able to handle databases composed of noncongeneric chemicals Additionally, as a consequence of how toxicological data are generated, there may be a paucity of data altogether Yet, for optimal SAR models of toxicological phenomena, the "learning set" should include at least 300 non-congeneric chemicals (3, 22) Accordingly, the human expert may suggest, that for certain purposes, the results of certain assays be pooled, e.g., rat and mouse Introduction carcinogenicity or the results of the Salmonella and E coli WP uvrB mutagenicity assays (23,24) Obviously, such data pooling must be based on a sound scientific basis as well as data that show extensive concordance between the experimental results of the systems to be pooled, i.e., that a substantial number of chemicals must give identical results in the two systems, thereby indicating that results obtained with one system can be amalgamated with those obtained in the other (25) For example, when the same chemicals were tested for their ability to induce sister chromated exchanges and chromosomal aberrations in cultured Chinese hamster ovary (CHO) cells, they showed divergent results (26).Hence, the results of the two assays cannot be amalgamated into a single database to develop an SAR model of cytogenetic effects Similarly, even using the same indicator system, results cannot be merged if different criteria are used to interpret the significance of the results That situation prevails with respect to the induction of mutations at the thymidine kinase locus of mouse lymphoma cells vis-a-vis the criteria used by the U.S National Toxicology Program versus those employed by the U.S Environmental Protection Agency's GeneTox Program In fact, each data set gives rise to a distinct SAR model (27-29) On the other hand, the consensus database of potential developmental toxicity in humans, based on experimental results in animals, observations in exposed humans, and expert judgment, yields a coherent SAR model of developmental risks to humans (30) That model is distinct from SAR models of developmental toxicity to individual rodent species (31) 1.2 Model Building Once a "learning set" (i.e., database) satisfying preset criteria for acceptance (3) has been developed, the model building phase can begin In general, this is a straightforward process that is specific for the SAR method em-ployed Here, I will exemplify the various stages with the MULTICASE SAR system ( ) Thus once the structures of the chemicals and an indication of their potency (i.e., either ac- tive, marginally active, and inactive, or a continuous scale of potencies) are entered, the program identifies the chemical substructures significantly associated with the toxicological phenomenon under investigation (Table 19.2) Each of these structural determinants (''toxicophore") is associated with a base potency and a probability of activity (see Fig 19.1) The latter is derived from the distribution of active and inactive molecules that contain the toxicophore The program also identifies the chemicals that give rise to the toxicophore (Table 19.3 and Fig 19.7) This enables the human expert (see below) to ascertain whether the structures of the chemicals giving rise to the toxicophores are germane to the test chemical whose toxicity is predicted In addition to the toxicophores, the program also identifies modulators for specific toxicophores (Table 19.4) These are substructures or physicochemical parameters that determine whether the specific toxicity inherent in the toxicophore will be expressed or whether it is augmented further Thus, when faced with a chemical of unknown activity, the program uses the presence or absence of toxicophores and of modulators to predict its toxicity (Figs 19.1-19.3) Thus, (a the presence of the toxicophore OH+ phenol) endows a chemical with an 87.5% probability of being a contact allergen and a potency of 51 (moderate activity, see Table 19.3) That basal activity is modulated by -25.8 x electronegativity (see Table 19.3) For the example in Fig 19.1, this results in a further increase in potency The total potency of 55 units corresponds to a moderately strong activity (Table 19.3) A chemical with that toxicophore may also contain a structural modulator that augments the basal activity further (Fig 19.2) On the other hand, the chemical may contain a modulator which completely abolishes a chemical's potential to be an allergen (Fig 19.3) Additionally, the MULTICASE SAR program will identify substructures that are absent from the learning set and therefore may introduce an element of uncertainty in the prediction, i.e., the "unknown" substructure could represent either potential toxicophore or a modulator that alters a rec- Table 19.2 Major Toxicophores Associated with Allergic Contact Dermatitis in Humans Fragment N* Inactives* Marginals* Actives* Toxicophore No The database and derivation of the SAR model have been described (33) *N indicates the number of chemicals in the database that contain that toxicophore "Inadives," "marginals," and "actives" indicate the distribution of that toxicophore among activity groups Toxicophore No is shown embedded in chemicals in Figs 19.1-19.3 and No is shown in Fig 19.5 C indicates a carbon atom shared by two rings; (3-NH,) indicates an amino group attached to the third non-hydrogen atom from the left In toxicophore No 17, the last carbon to the right is shown as unsubstituted This means that it can be substituted with any atom except a hydrogen On the other hand, in toxicophore No 8, the penultimate carbon is shown unsubstituted; it can only be substituted by an amino group (i.e., (SNH,) However, the last carbon of that toxicophore is shown with an attached hydrogen It cannot be substituted by any other atom 1 Introduction Table 19.3 Derivation of Toxicophore: The 19 Molecules Containing Fragment SH CH, ~- Chemicals 2,3-Dimercapto-1-propanol 2-Mercaptoethanesulfonic acid 2-Mercaptoethyl methyl sulfone 2-Mercaptoethyl urea 2-Methoxyethyl mercaptoacetate N-(1,l-dimethylolethyl) mercaptoacetamide N,N-dimethyl mercaptoacetarnide N-(2-mercaptoethyl)acetamide N42-mercaptoethyl) pyrolidone N-(mercaptoacetyl) urea N4mercaptoacetyl) glycine N-(mercaptoethyl) morpholine N-methyl mercaptoacetamide N-trimethylolmethyl mercaptoacetamide Cysteine Mercaptoacetamide Mercaptoacethydrazide Mercaptoacetic acid Thioglycerol Potency" 55 55 45 35 35 45 45 55 45 45 55 The program identifies the chemicals that are responsible for toxicophore No of Table 19.1 (see also Fig 19.7) The toxicophore is shown embedded in a molecule in Fig 19.5 "The allergenic potencies were defined based on the percent responders in the human maximization test as follows (33):10, Non-sensitizer; 25, "marginal" (4-7% responders); 39, "weak" (8-23% responders); 49, "moderate": (2445% responders); 59, "strong" (56-83% responders); 69, "extreme" (84-100%responders) ognized toxicophore or a noninformative structure unrelated to toxicity (Fig 19.4) It should be stressed that not every experimental data set gives rise to a coherent SAR model Failure to construct a model may be caused by the fact that the experimental data are invalid or that they not reflect a specific toxicological phenomenon Additionally, the phenomenon under investigation may be so complex or be the result of so many different mechanisms that the experimental database is not sufficiently large to describe it With this in mind, it should be stressed that the predictivity of the SAR model will be a reflection of the complexity of the phenomenon, the size of the database (i.e., the number of chemicals for which experimental data are available), and the ratio of activesJinactivesin the data set (3,221 In view of the above considerations, once an SAR model has been developed, it requires extensive validation and characterization 1.3 Model Characterization As mentioned above, the nature of the SAR model that is derived is a reflection of the complexity of the toxicological phenomenon that it describes, as well as of the size of the learning and the extent to which it includes chemical classes and/or substructures that are representative of the chemical species to which it will be applied Thus, the chemical substructures present among therapeutics are much greater and diverse than, for example, those used or generated in the chemical or agricultural industries This means that SAR models used to examine pharmacologically active substance must contain a greater variety of chemical substructures This may well translate into a requirement for a larger experimental data set (i.e., one containing an increased number of chemicals) In evaluating the SAR model, it is of importance to determine the relationship between its predictivity and the size of the database to determine whether the model is ovtimal This can be ascertained by first determining the model's predictivity (see below), and then systematically decreasing the size of the database by random deletion of chemicals to determine the predictive parameters of the model derived from the reduced data set Doing this iteratively will allow a determination of the relationship between database size and concordance between predicted and experimentally derived results (22) If the relationship, including the value for the SAR model derived from the total database is linear, then the model will not be optimally predictive and consideration should be given to obtaining additional experimental data and deriving a further model On the other hand, if the relationship including the data for the SAR model derived from the total database is no longer linear, the size of the data set may be satisfactory Incremental data may not yield a correspondingly significant increase in the model's performance Thus, the predidivity of the SAR model of mutagenicity in Salmonella improves linearity until a database size of 350 chemicals is reached, and then it plateaus (22) m Table 19.4 List of MODULATORS Related to Toxicophore OH -c Constant = 51.0 QSAR Fragment o W N 2D [N-I ( A - ) [NH-I CO -4H2 H 2cH ===c h OH -C ===c cH =c -cH CH2 H -c cH =cH -cH cH =cH -cH OH -c =c OH -c =cH OH -c (HOMO + LUMO)/2 = Toxicophore No 3H4 H ==xH ===c ==xH -cH -cH Modulators associated with toxicophore No of Table 19.2 Each of the modulators augments or decreases the activity inherent in the toxicophore (i.e., 51.0 units; see Figs 19.1-19.3) (HOMO + LUM0)I.Zdescribes the electronegativity of the molecule That value is multiplied by -25.8 Modulator No and No are shown embedded in chemids in Figs 19.2 and 19.3, respectively Modulator No describes a 2D distance descriptor of 6.5 A between two atoms For interpretation of the structures see legend to Table 19.2 1 Introduction (nr.occ.= ) : The molecule contains the Toxicophore *** 76 out of the known 86 rnolec~le~ ( BBO) containing such a Toxieophore are Contact Allergen8 with a n average accivity of 49 (conf.level=lOO%) +** Constant is QSAR Contribution : 51.04 ** The following Modulator is also preeent: Electronegativity ** = -0.15 It8 contribution is Total projected QGATC activity 3.83 -54.87 *** The probability that this molecule is a Contact Allergen is 87.5% f f ** The projected Allergic Potency is CASE units ** Figure 19.1 Prediction of the contact allergenicity of 2-methyl-1,4-benzenediol The prediction is based on the presence of the toxicophore (shown in bold) The potency is modulated fkrther by the electronegativity (see Table 19.4) A potency of 55 units indicates a moderately strong allergen (see Table 19.3) Another concern relates to the effect of the ratio of active to inactive chemicals in the data set Some SAR models are most predictive when that ratio is unity (3, 22) Hence, for a model that will be widely used for hazard identification and risk assessment purposes, it would be of importance to determine whether its performance is optimal Thus, if the number of inactives exceeds the number of actives, the number of inactives can be decreased by randomly removing the appropriate number of inactives and determining the performance of the resulting SAR model The random deletion of inactives and the model derivation should be repeated several times to ascertain that a robust model has been derived We found that because the nature of the toxicophores is determined primarily by the actives and because the "quality" of the toxicophores is a function of the size of the database (22,34, 35), it follows that if the number of actives exceeds the number of inactives that removal of actives to achieve a ratio of unity is not the optimal solution Rather, we have found that supplementing the database with randomly selected chemicals from a "pool" of normal physiological chemicals (amino acids, sugar, lipids, purines, pyrimidines, etc., but excluding hormones, prostaglandins, and vitamins), assuming these chemicals to be inactive, is a viable alternative (36,37) This is based on the recognition that the biological and/or toxicological phenomena being modeled occur in a milieu that is rich in these physiological chemicals Finally, the "informational content" of an SAR model determines its coverage Thus, if a test molecule contains a substructure un- Structural Concepts in the Prediction of the Toxicity of Therapeutical Agents 834 The molecule contains the Toxicophore +** 76 out of the known 86 (nr.occ.= 2): molecules ( 88%) containing such a ~oxicophoreare contact allergiee with an average activity of *** QSAR ** 49 Constant is Contribution r 51.04 The following Modulators are also present: ( 1) cH =c -cH -cH -c Electronegativity ** Total projected *** The c2-OH > = o QSAR -0.17 Activating Its contribution is ; 8.33 4.29 -63 - activity probability chat thie molecule is a Contact Allergen is ** The projected Allergic Potency ie 63.7 CASE units +* Figure 19.2 Prediction of the contact allergenicity of 4-chloro-1,3-benzenediol In addition to the probability of activity and the basal potency derived from the toxicophore (shown in bold in A),the chemical also contains an activating modulator (shown in bold in B),which further augments the potency A potency of 64 units indicates a very strong potency (Table 19.3) known to the model, this introduces a measure of uncertainty into the SAR prediction In the MULTICASE SAR program, such an "unknown" moiety is flagged (Fig 19.4) We have found that a satisfactory approach to determining informational content is to challenge an SAR model with a panel of 10,000 chemicals representative of the "universe of chemicals" and determining the frequency with which the SAR predictions are accompanied by " a "warning" - of the presence of "unknown" substructures An enumeration of the frequency with which the individual unknown moieties are present will allow a determination of their importance and thereby identifies chemicals that should be tested and the results included in the model to improve the predictive performance This is based on the observation that the greater the informational content (i.e., the fewer warnings of "unknown" moieties), the greater the model's predictivity (22,34,35) 1.4 Model Validation In its application to toxicology, SAR can serve two functions: (1)to predict a specific toxico- Introduction The molecule contains the Toxicophore OH *** *** -cU 76 out of the known Toxicophore are (nr.occ.= 1): 86 molecules ( 80%) containing such a Contact Allergen8 with an average activity of 49 W A R Contribution Constant is : - 04 ** The following Modulator8 are also present: ( 1) CO -CHZ-CH2- Electronegativity Inactivating -53.33 = -0.10 ; It6 contribution is 2.51 ** Total projected QSAR activity ** The molecule contains rhe following DEACTIVATING Fragment: *r* 0.22 The probability that thim molecule is a Contact Allergen in 63.6% ** Figure 19.3 Prediction of the lack of contact allergenicity of zingerone.Whereas the presence of the toxicophore (A) is associated with a probability of activity and a potency, the presence of the inactivating modulator (B) abolishes the potency Moreover, the presence of a deactivating moiety (C), which is present in five chemicals in the database that are devoid of allergenicity (Table 19.2, No 19), further decreases the likelihood that the zigerone is a contact allergen logical effect based on the identification of substructures significantly associated with that activity and (2) to gain insight into the mechanistic basis of that effect To be useful in its predictive mode, the performance of a model does not need to be perfect, but it must be known The predictivity of an SAR model is defined by the concordance between the predictions of chemicals external to the SAR model and the experimentally determined toxicities The predictivity is governed by the sensitivity (number of correct positive predictions/total number of positive chemicals) and the specificity (number of correct negative predictionsltotal number of negative chemicals) (22) Moreover, because the basic function of SAR applied to toxicological phenomena is the prevention, reduction, or elimination of harmful chemicals from the home, the environment, and the workplace, risk averse prediction models are preferred That is achieved by the development of SAR models that yield a low frequency of false negative predictions, i.e., high specificity Obviously, ideally the model should have high sensitivity as well as high specificity (38) Structural Concepts in the Prediction of the Toxicity of Therapeutical Agents O*** WARNING *** *** *** CO -0 The following functionalities are UNKNOWN to me; -C =C -C - ** The molecul e does not contain any known Biog it is therefore preaumed to be INACTIVE Figure 19.4 Prediction of the lack of contact allergenicity of of dehydroalantolactone The chemical contains no toxicophore; therefore, it is presumed to be inactive However, it contains two structures (shown in bold) that are "unknown" to the model That introduces an element of uncertainity in the prediction The simplest way to determine predictivity parameters is to remove initially from the data set a random representative sample (e.g., 5%) to be used as a "tester set," to develop the SAR model on the remaining chemicals (i.e., 95%), and then challenge the model with the "tester set" and ascertain the predictivity However, as has been demonstrated on a number of occasions, the predictivity of an SAR model is determined by the size of the database (221, and as in most instances, the size of the available data set is not optimal, therefore, further decreasing the size of the learning set by sequestering the "tester set" is not optimal To overcome this limitation, a cross-validation approach has been used (39) In that procedure, a portion of the database (e.g., 5%) is randomly selected and removed, and a model is developed from the remaining 95% That model is challenged with the "tester set" (5%) That procedure is repeated 20 times, and the cumulative predictivity is determined The final SAR model includes the complete database i , 100%) Because the predictive performance is a function of the size of the database, the performance of the final model will be better than that based on 95% of the data When, however, the learning set consists of less than 150 chemicals, a more tedious procedure may be required, wherein one to two chemicals (i.e., n-1 or n-2) are removed at a time to serve as the "tester set" and the process is repeated n or n/2 times 1.5 Applications and Mechanistic Studies As has been mentioned earlier (Table 19.11, SAR methodologies can be divided into two general non-mutually exclusive approaches: (1) hypothesis driven and (2) knowledge based The former is rule driven, wherein specific properties or chemical substructures are looked for, e.g., mutagens are electrophiles and hence one would look for electrophilic or proelectrophilic moieties This approach assumes that mutations are caused solely by covalent binding of electrophiles to DNA Agents that induce mutations by a nonelectrophilic (i.e., non-DNA damaging) mechanism will not be detected Thus, agents that mutagenize purely as a result of intercalation between DNA base pairs (e.g., acridine orange, ethidium bromide) will not be identified Such rules are based on prior knowledge and/or in- Introduction The molecule contains the Biophore SH -a2 *=* 18 out of che known 19 molecule8 ( 95%) containing such a toxicophore are Contact ALlergene with an average activity of 42 (conf.levrl=100%) *** ** QSAR Contribution ; Constant i a 52.50 Inactivating -7.6 The following M0d~lator6are also present: (2D) [ C O - ] Electronegativity ** ISH-1 e- - A - - > = 0.10 ; Ite contribution is -0.67 Total projected QSAR activity *** The probability that this molecule is a Contact Allergen is 95.0% ** ** The projected Allergic Cont rativity i e 4 CASE units +* Figure 19.5 Prediction of the contact allergenicity of N-acetyl-~qsteine.The prediction is based on the presence of the toxicophore (shown in bold), which is present in 19 chemicals in the database (18allergens and marginal allergen; see Table 19.3) The arrow indicates the 5.2 A distance described by the inactivating modulator tuition and not necessarily require adherence to strict statistical criteria The approach illustrated herein, exemplified by MULTICASE (31, is knowledge based The input consists of the structures and toxicological activities of the chemicals in the learning set The program then identifies structural descriptors (toxicophores) that are significantly associated with activity (see Table 19.3) The human expert participates in setting criteria for the inclusion of experimen- tal results in the database (3) as well as in examining the plausibility of the final model based on exact knowledge of the toxicological phenomenon under investigation The human expert again also determines the acceptability of individual predictions (see below) Once an SAFt model has been developed and validated, it can be applied in a number of fashions SAFt methodologies, such as MULTICASE (3-5), which document predictions (Table 19.21,are obviously preferable to those Structural Concepts in the Prediction of the Toxicity of Therapeutical Agents The molecule contains the Toxicophore S -C \\ N / C" *** out of the known molecules (100%) containing such Biophore are Mouse carcinogens with an average activity of 62 (conf lcvel=W%) ** This Biophore exists in a significantly different environment: than in che data baee (i-e ) ; *** ** QSAR Contribution Total projected QSAR activity *** The ** : It may not be relevant Constant is 64.00 64.00 probability that this molecule is a Mouse carcinogen is 05.7% The projected Mouse carcinogenic potency is 64.0 CASE units ** Figure 19.6 Prediction of the carcinogenicity in mice of epitholone A The structure of epitholone A (toxicophore shown in bold) is given in Fig 19.7 that operate like a "black box." The latter simply provides a likelihood that a test chemical is active or inactive When, however, the SAR prediction is accompanied by documentation of the basis of that forecast, the human expert can determine whether it is justified and whether it applies to the specific test chemical Thus, the mucolytic agent N-acetyl-L-cysteine is predicted to have a potential to induce allergic contact dermatitis by virtue of the biophore SH CH, (Fig 19.5) Moreover, examination of the chemicals that contribute to that toxicophore reveals that indeed they all have the substructure in an environment that is similar to the one found in N-acetyl-L-cysteine (Table 19.3) On the other hand, the tubulin polymerization perturber (and potential antineoplastic agent) epitholone A (Fig 19.6) is predicted to be a mouse carcinogen by virtue of the toxicophore units shown in bold That toxicophore is present in five molecules in the learning set The presence of that toxicophore is associated with an 89% probability of carcinogenicity and a potency of 63 units, which corresponds to a TD,, value of 0.039 mmoljkg per day (40) However, the program flags the toxicophore because its environment in epitholone A is significantly different from that of the molecules in the learning set (Fig 19.6) In fact, examination of the structures of the molecules that contribute to the biophore (Fig 19.7) indicates that indeed the molecules are quite different from that of epitholone A, and hence, the prediction of carcinogenicity can be disregarded (however, also see below) Moreover the molecules that contributed to this toxicophore (Fig 19.7), even though they moiety (Fig 19.61, contain the W N - C ! = also contain functionalities (i.e., "structural alerts") that are associated with carcinogenicity/genotoxicity such as nitro, amino, and hydrazino groups In fact, these could be responsible for the murine carcinogenicity of these chemicals Obviously, these latter functionalities are absent in epitholone A 1 Introduction H' Epitholone A Figure 19.7 Structures of epitholone A and of chemicals that contain the toxicophore The toxicophore (Fig 19.6) is shown in bold Table 19.5 SAR Predictions Related to the Potential Carcinogenicity of Epitholone A SAR Model Mutagenicity: Salmonella Error-prone DNA repair Unscheduled DNA synthesis Mouse MTD Rat LD,, Cell toxicity Inhibition GJIC Prediction Negative Negative Negative Positive Positive Positive Negative References 22,47 48 49 50 SAR model based on RTECS 51 52 A positive response indicates a potential for maximum tolerated dose of less than 0.9 mmolkg; an LD,, value of less than 7.2 mmolkg or a toxicity (IC,,) for cultured BALBl3T3 cells of less than p M GJIC, gap junctional intercellular communication; RTECS, Registry of Toxic Effects of Chemical Substances 840 Structural Concepts in the Prediction of the Toxicity of Therapeutical Agents Table 19.6 Predicted Toxicological Profile of N-Acetylcysteine Multicase SAR Model Structure alerts Salmonella mutagenicity SOS chromotest umuISOS repair Carcinogenicity: rodents-NTP Carcinogenicity: mice-NTP Carcinogenicity: rats-NTP Carcinogenicity: rodent-CPDB Carcinogenicity: mice-CPDB Carcinogenicity: rats-CPDB Inhibition gap junction intercell comm Binding to Ah receptor Mutations in mouse lymphoma (NTP) Mutations in mouse lymphoma (GenTox) Sister chromatic exchanges in vitro Chromosomal aberrations in vitro Unscheduled DNA synthesis in vitro Cell transformation Drosophila somatic mutations Sister chromatic exchanges in vivo Induction of micronuclei in vivo Yeast malsegregation Inhibition of tubulin polymerization Sensory irritation Eye irritation Respiratory hypersensitivity Allergic contact dermatitis Rat lethality (LD50) Mouse MTD Rat MTD Cellular toxicity (3T3) Cellular toxicity (HeLa) Nephrotoxicity: male rats (a2pglobulin) Inhibition human cyt P4502D Developmental toxicity: hamster Developmental toxicity: human Aquatic toxicity (minnows) Water solubility: 3.88 Electronegativity: 0.10 Probability (%) Potency (units) 0 0 0 0 0 0 0 0 0 0 0 89 72 95 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 72 52 44 0 0 0 0 0 log P (Octanol: water): -1.79 NTP and CPDB refer to the US.National Toxicology Program carcinogenicity assays (45) and to the Carcinogenic Potency Data Bases (461,respectively Based on all of these considerations, the "human expert" would overrule the prediction of rodent carcinogenicity Additionally, in overriding the computer-based prediction, cognisance was also taken of the understanding that the vast majority of recognized human carcinogens are genotoxicants, i.e., "genotoxic carcinogens" (41-44) Epitholone A, on the other hand, was not predicted to be genotoxic (i.e., a DNA-damaging agent), evidenced by its lack of potential to induce mutations in Salmonella, error-prone DNA repair, or unscheduled DNA synthesis in rat hepatocytes (Table 19.5) Thus, even if the potential for murine carcinogenicity were accepted, in view of the fact that the vast majority of rec- Introduction The molecule contains the Biophore *** 38 out of the (nr.occ.= 1) : known 41 molecules (93%) containing such a Biophore are perturbers of Tubulin Polymerization *** QSAR Contribution : ** Total projected QSAR activity ** The probability that this molecule inhibits Tubulin Polymerization is 93% ** ** The projected Tubulin Polymerization Inhibitory activity is CASE units ** Figure 19.8 Prediction of the ability of colchicine to inhibit tubulin polymerization The structure of colchicine is shown in Fig 19.9 The biophore is shown in bold (a) in Fig 19.9 ognized human carcinogens are mutagens1 genotoxicants or are hormones and epitholone A is neither, it would not represent a human risk If, based on the above, it were accepted that epitholone A is not genotoxic, and if the human expert examining the documentation wished not to override the prediction of carci- Figure 19.9 Structure of colchicine The biophore A (bold, see Fig 19.8) is responsible for the therapeutic effectiveness Toxicophore B (see Fig 19.10; shown in bold) is responsible for the induction of sister chromatid exchanges (SCE) in vivo Removal of toxicophore B or its replacement be isopropoxy groups abolishes the induction of SCEs without affecting the therapeutic potential nogenicity in mice based on the differences in chemical environments between epitholone A and the molecules responsible for the toxicophore (Figs 19.6 and 19.7), he could examine mechanisms of non-genotoxic carcinogenicity, even though its relevance to human may not be applicable One of the mechanisms of nongenotoxic carcinogenicity is inhibition of intercellular communication (53) Epitholone A does not possess such a potential (Table 19.5) Another mechanism for non-genotoxic rodent carcinogenesis may involve systemic or cell toxicity followed by mitogenesis (54-56) This may occur as a consequence of including the maximum tolerated dose (MTD) in the cancer bioassay protocol When this is done, up to 50% of chemicals tested are found to be rodent carcinogens (54) Obviously, this MTD situation rarely, if ever, applies to humans Still, epitholone A has the potential for inducing cellular as well as systemic toxicity (Table 19.5), which may explain its potential carcinogenicity in mice, were we to discount the difference in chemical environment Obviously, the availability of a number of characterized and validated SAR models allows the development of a putative toxicologi- Structural Concepts in the Prediction of the Toxicity of Therapeutical Agents 842 The molecule contains the toxicoophore (nr.0cc.s 3): *** *** ++ out of the known molecules (100%) containing such a toxicophore are Mouse SCE inducers with an average activity of 57 QSAR Contribution Constant is ; The following Modulators are also present: ( 3) -3-0 ( 1) CH3-0 -c = -c =cH - Log partition c0eff.r ** Total projected +* 73.17 3.19 ; Inactivating -7.41 Inactivating -7.41 LogP contribution is QSAR activity 50.70 The probability that this molecule induces Mouse SCEe ie +* The -7.65 projected Mouse SCE inducing activity i o 90.0% ** CASE units ** Figure 19.10 The potential of colchicine to induce sister chromatic exchanges in vivo The structure of colchicine and of the toxicophore B is given in Fig 19.9 One of the inactivating modulators (c) is also shown in bold in Fig 19.9 cal profile (Table 19.6) This can be used as a guideline in the product developmental phase to select lead compounds least likely to induce unwanted side effects However, the SAR approach can also be used to optimize beneficial effects and decrease or eliminate unwanted toxic effects Thus, let us examine colchicine (CH), an anti-inflammatory agent that has been in use for several centuries for the treatment of gout The anti-inflammatory potential of CH is understood to derive from its ability to inhibit tubulin polymerization (iTP) (57).That is also the basis of the anticancer activity of paclitaxel (Taxol) (58-60) The structural basis of that activity derives from the presence in CH of the N H - C H - C = moiety (Figs 19.8 and 19.9), which endows the molecule with a 93% probability of activity However, colchicine also has the potential for inducing sister chromatid exchanges (SCEs) in vivo (Fig 19.10) This SCE-inducing ability may endow it with genotoxic and developmental toxicity poten- tials However, the potential for inducing SCEs i n vivo is associated with the methoxy moiety (Figs 19.9 and 19.10) Removal of that moiety or replacing it with an isopropoxy group abolishes the SCE-inducing ability of CH without affecting its potential for iTP (i.e., the basis of its anti-inflammatory action) Finally, SAR approaches can also be used to provide a basis for making intelligent risk assessments Thus, it has been shown that the similarity in biophores/toxicophores present in different SAR models of toxicological phenomena provides a measure of mechanistic similarity (3) The SAR models of mutagenicity in Salmonella and of error-prone DNA repair (SOS Chromotest) show significant overlaps (Table 19.7) This is not unexpected because DNA is the target of both phenomena, and the tester strain used for the Salmonella mutagenicity assays contains a plasmid that codes for error-prone DNA repair (61) In fact there is a substantial (though not complete) overlap among chemicals that cause the two Introduction Table 19.7 Structural Commonalities among SAR Models SAR Models Percent Salmonella mutagenicity and SOS chromotest Salmonella mutagenicity and iGJIC Salmonella mutagenicity and iTP Salmonella mutagenicity and Mnt Mnt and iTP 57 10 53 71 iGJIC, inhibition of gap functional intercellular communications; iTP, inhibition of tubulin polymerization; Mnt, induction of bone marrow micronuclei in uivo phenomena (48,62) On the other hand, there is little overlap between Salmonella mutagenicity and inhibition of gap junctional intercellular communication (Table 19.7), which is considered the epigenetic (non-genotoxic) phenomenon par excellence (53) Nor the SAR models for Salmonella mutagenicity and inhibition of tubulin polymerization overlap significantly (Table 19.7), which is further support for the fact that genotoxicity and inhibition of tubulin polymerization can be dissociated (see above) With respect to the in vivo induction of micronuclei (Mnt), a different situation prevails There is considerable overlap between the toxicophores associated with Mnt and those with the induction of mutation is Salmonella (Table 19.7) This is not surprising, because the induction of Mnt is known to involve a genotoxic mechanism (63, 64) Indeed, when attempting to identify potential genotoxic carcinogens, when a chemical is found to induce mutations in Salmonella, this result is frequently followed by a Mnt test to determine OH Discodermolide whether the chemical is genotoxic in vivo as well (43, 65) and thus represent a risk to humans However, it was found that there is also substantial overlap between Mnt and iTP, the latter being a non-genotoxic phenomenon (Table 19.7) (66) This finding suggests that Mnt can occur by genotoxic as well as non-genotoxic mechanisms Thus, a positive Mnt response by a chemical that does not induce mutations in Salmonella does not necessarily represent a carcinogenic risk to humans Discodermolide (Fig 19.11) is a promising antineoplastic agent, which like paclitaxel, inhibits tubulin polymerization (671, but being considerably more water-soluble than paclitaxel, discodermolide may present certain therapeutic advantages while also being effective against paclitaxel-resistant cells (67) Neither discodermolide nor paclitaxel are mutagenic in Salmonella (and in fact neither is predicted to be a rodent carcinogen) However, both of these agents have a potential (determined by SARI to induce Mnt i n vivo In fact, for paclitaxel that potential has been determined experimentally This has led to the suggestion that paclitaxel, because of its ability to induce Mnt, presented a carcinogenic risk (68) However, based on the above findings (Table 19.71, it can be assumed that the ability of discodermolide and of paclitaxel to induce Mnt is independent of genotoxicity, and in fact, derives from iTP Thus, it does not represent an unreasonable risk to humans who are treated with those antineoplastic agents In fact, the biophores in discodermolide responsible for the induction of Mnt and iTP are identical (Fig 19.11) Figure 19.11 Structure of discodermolide The circled biophore is responsible for the inhibition of tubulin polymerization Structural Concepts in the Prediction of the Toxicity of Therapeutical Agents 844 CONCLUSIONS SAR methodologies, in their present state, coupled with human expertise, can be used to determine and to understand the potential toxicity of therapeutic agents In fact, this approach can be used to engineer molecules devoid of the moieties associated with these unwanted side effects It must be understood, however, that while SAR techniques can be used to accelerate the identification and development of safe therapeutic agents, it is to be used as an adjunct to experimental determinations ACKNOWLEDGMENTS The support of the Vira Heinz Endowment is gratefully acknowledged REFERENCES National Research Council, Science and Judgment in Risk Assessment, National Academy Press, Washington, DC, 1994 I D McKinney, A Richard, C Waller, M C Newrnan, and F Gerberich, Toxicol Sci., 56, 8-17 (2000) H S Rosenkranz, A R Cunningham, Y P Zhang, H G Claycamp, 0.T Macina, N B Sussman, S G Grant, and G Klopman, SAR QSAR Environ Res., 10,277-298 (1999) G Klopman and H S Rosenkranz, Mutat Res., 305,33-46 (1994) G Klopman and H S Rosenkranz, Toxicol Lett.,79,145-155 (1995) K Einslein, V K Gombar, and B W Blake, Mutat Res., 305,47-61 (1994) D F V Lewis, C Ioannides, and D V Parke, Environ Health Perspect., 104, 1011-1016 (1996) D M Sanderson and C G Earnshaw, Human Exp Toxicol., 10,261-273 (1991) N Greene, J Chem Znf Comput Sci., 37,148150 (1996) 10 J E Ridings, M D Barratt, R Cary, C G Earnshaw, C E Eggington, M K Ellis, P N Judson, J J Langowski, C A Marchant, M P Payne, W P Watson, and T D Yih, Toxicology, 106, 267-279 (1996) 11 Y T Woo, D Y Lai, M F Argus, and J C Arcos, Toxicol Lett.,79,219-228 (1995) 12 Y T Woo, D Y Lai, M F Argus, and J C Arcos, Environ Carcin Ecotoxicol Rev C., 16, 101-102 (1998) 13 F Darvas, A Papp, A Allerdyce, E Benfenati, G Fini, M Tichy, N Sobb, and A Citti, A M Spring Symposium on Predictive Toxicology of Chemicals: Experiences and Impact of A1 Tools, Technical Report SS-99-01, AAAI Press, Mento Park, CA, 1999 14 R D King, S H Muggleton, A Srinivasan, and M J E Sternberg, Proc Natl Acad Sci., 93, 438-442 (1996) 15 J Ashby, Environ Mutagen., 7, 919-921 (1985) 16 A M Richard, Mutat Res., 305,73-77 (1994) 17 A M Richard, Knowledge Engineer Rev., 14, 307-317 (1999) 18 R D Combes and P Judson, Pestic Sci., 45, 179-194 (1995) 19 A M Richard, Toxicol Lett.102-103, 611-616 (1998) 20 C Helma, E Gottmann, and S Kramer, Stat Methods Med Res., 9, (2000) 21 A M Richard and R Benigni, SAR QSAR Environ Res., 13, 1-19 (2002) 22 M Liu, N Sussman, G Klopman, and H S Rosenkranz, Mutat Res., 358,63-72 (1996) 23 D M Maron and B N Ames, Mutat Res., 113, 173-215 (1983) 24 D Brusick, V F Simmon, H S Rosenkranz, V Ray, and R S Stafford, Mutat Res., 76, 1%9190 (1980) 25 J Pet-Edwards, H S Rosenkranz, V Chankong, and Y Y Haimes, Mutat Res., 153, 167-185 (1985) 26 H S Rosenkranz, F K Ennever, M Dimayuga, and G Klopman, Environ Mol Mutagen., 16, 149-177 (1990) 27 A D Mitchell, A E Auletta, D Clive, P E Kirby, M M Moore, and B C Myhr, Mutat Res., 394,177-303 (1997) 28 B Henry, S G Grant, G Klopman, and H S Rosenkranz, Mutat Res., 397,313-335 (1998) 29 S G Grant, Y P Zhang, G Klopman, and H S Rosenkranz, Mutat Res., 465,201-229 (2000) 30 M Ghanooni, D R Mattison, Y P Zhang, 0.T Macina, H S Rosenkranz, and G Klopman, Am J Obstet Gynecol., 176, 799-806 (1997) 31 J Gbmez, 0.T Macina, D R Mattison, Y P Zhang, G Klopman, and H S Rosenkranz, Teratology, 60, 190-205 (1999) 32 H S Rosenkranz, Y P Zhang, and G Klopman, Altern Anim Test., 26, 779-809 (1998) References 33 C Graham, R Gealy, T Macina, M H Karol, and H S Rosenkranz, Quant Struct Activ Relat., 15, 224-229 (1996) 34 N Takihi, Y P Zhang, G Klopman, and H S Rosenkranz, Mutagenesis, 8,257-264 (1993) 35 N Takihi, Y P Zhang, G Klopman, and H S Rosenkranz, Qual Assur Good Pract Regul Law, 2,255-264 (1993) 36 H S Rosenkranz and A R Cunningham, Mutat Res., 476, 133-137 (2001) 37 H S Rosenkranz and A R Cunningham, SAR QSAR Environ Res., 12,267-274 (2001) 38 V Chankong, Y Y Haimes, H S Rosenkranz, and J Pet-Edwards, Mutat Res., 153, 135-166 (1985) 39 Y P Zhang, N Sussman, G Klopman, and H S Rosenkranz, Quant Struct Activ Relat., 16, 290-295 (1997) 40 A R Cunningham, H S Rosenkranz, Y P Zhang, and G Klopman, Mutat Res., 398,l-17 (1998) 41 F K Ennever, T J Noonan, and H S Rosenkranz, Mutagenesis, 2, 73-78 (1987) 42 H Bartsch and C Malaveille, Cell Biol Toxicol., 5, 115-127 (1989) 43 J Ashby and R S Morrod, Nature, 352, 185186 (1991) 44 M D Shelby, Mutat Res., 204,3-15 (1988) 45 J Ashby and R W Tennant, Mutat Res., 257, 229-306 (1991) 46 L S Gold, N B Manley, T H Slone, G B Garfmkel, L Rohrbach, and B N Ames, Environ Health Perspect., 100,65135 (1993) 47 H S Rosenkranz and G Klopman, Mutat Res., 228,51-80 (1990) 48 V Mersch-Sundermann, G Klopman, and H S Rosenkranz, Mutat Res., 340,81-91 (1996) 49 Y P Zhang, A van Praagh, G Klopman, and H S Rosenkranz, Mutagenesis, 9, 141-149 (1994) 50 H S Rosenkranz and G Klopman, Environ Mol Mutagen., 21, 193-206 (1993) 51 H S Rosenkranz, E J Matthews, and G Klopman, Altern Anim Test., 20, 549-562 (1992) 845 52 M Rosenkranz, H S Rosenkranz, and G Klopman, Mutat Res., 381, 171-188 (1997) 53 J E Trosko and C C Chang in R W Hoerger and F D Hoerger, Eds., Banbury Report 31: Carcinogen Risk Assessment: New Directions in Qualitative and Quantitative Aspects, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1988, pp 139-174 54 B N Ames and L S Gold, Proc Natl h a d Sci USA, 87,7772-7776 (1990) 55 S M Cohen and L B Ellwein, Science, 249, 1007-1011 (1990) 56 S Preston-Martin, M C Pike, R K Ross, P A Jones, and B E Henderson, Cancer Res., 50, 7415-7421 (1990) 57 E ter Haar, H S Rosenkranz, E Hamel, and B W Day, Bioorg Med Chem., 4, 1659-1671 (1996) 58 E Hamel, Med Res Rev., 16,207-231 (1996) 59 P B Schiff and S B Horwitz, Proc Natl h a d Sci USA, 77,1561-1565 (1980) 60 P B Schiff, J Fant, and S B Horwitz, Nature (Lond.),277,665-667 (1979) 61 J McCann, N E Spingarn, J Kobori, and B N Ames, Proc Natl Acad Sci USA, 72,979-983 (1975) 62 V Mersch-Sundermann, U.Schneider, G Klopman, and H S Rosenkranz, Mutagenesis, 9, 205-224 (1994) 63 J A Heddle, M C Cimino, M Hayashi, F Romagna, M D Shelby, J D Tucker, Ph Vanparys, and J T MacGregor, Environ Mol Mutagen., 18,277-291 (1991) 64 K H Mavournin, D H Blakey, M C Cimino, M F Salamone, and J A Heddle, Mutat Res., 239,29-80 (1990) 65 H Tinwell and J Ashby, Environ Health Perspect., 102, 758-762 (1994) 66 E ter Haar, B W Day, and H S Rosenkranz, Mutat Res., 350, 331337 (1996) 67 E ter Haar, R J Kowalski, E Hamel, C M Lin, R E Longley, S P Gunasekera, H S Rosenkranz, and B W Day, Biochemistry, 35, 243250 (1996) 68 H Tinwell and J Ashby, Carcinogenesis, 15, 1499-1501 (1994) CHAPTER TWENTY Natural Products as Leads for New Pharmaceuticals A D Buss MerLion Pharmaceuticals Singapore Science Park, Singapore B Cox Medicinal Chemistry Respiratory Diseases Therapeutic Area Nvvartis Pharma Research Centre Horsham, United Kingdom R D WAIGH Department of Pharmaceutical Sciences University of Strathclyde Glasgow, Scotland Contents Burger's Medicinal Chemistry and Drug Discovery Sixth Edition, Volume 1: Drug Discovery Edited by Donald J Abraham ISBN 0-471-27090-3 O 2003 John Wiley & Sons, Inc Introduction, 848 Drugs Affecting the Central Nervous System, 849 2.1 Morphine Alkaloids, 849 2.2 Conotoxins, 851 2.3 Cannabinoids, 852 2.4 Asperlicin, 855 Neuromuscular Blocking Drugs, 856 3.1 Curare, Decamethonium, and Atracurium, 856 Anticancer Drugs, 858 4.1 Catharanthus Winca) Alkaloids, 858 4.2 Camptothecin, 860 4.3 Paclitaxel and Docetaxel, 861 4.4 Epothilones, 864 4.5 Podophyllotoxin, Etoposide, and Teniposide, 865 4.6 Marine Sources, 867 Antibiotics, 868 5.1 p-Lactams, 868 5.2 Erythromycin Macrolides, 874 5.3 Streptogramins, 876 5.4 Echinocandins, 877 Cardiovascular Drugs, 878 Natural Products as Leads for New Pharmaceuticals 848 6.1 Lovastatin, Simvastatin, and Pravastatin, 878 6.2 Teprotide and Captopril, 881 6.3 Adrenaline, Propranolol, and Atenolol, 881 6.4 Dicoumarol and Warfarin, 882 Antiasthma Drugs, 883 7.1 Khellin and Sodium Cromoglycate, 883 7.2 Ephedrine, Isoprenaline, and Salbutarnol, 884 7.3 Contignasterol, 886 Antiparasitic Drugs, 886 8.1 Artemisinin, Artemether, and Arteether, 886 8.2 Quinine, Chloroquine, and Mefloquine, 888 8.3 Avermectins and Milbemycins, 891 Conclusion, 891 would now call "bioactive" substances was a mystery A modern view is that these compounds have a role in protecting the otherwise defenseless, stationary plant from attack by mammals, insects, fungi, bacteria, and viruses Taking morphine as an example of a secondary metabolite whose value to the plant is not entirely obvious, 14 steps are required from available amino acids, including at least one step that is highly substrate specific (2) The presence of morphine in the tissues of Papaver somniferum must therefore confer a selectional advantage on the plant (3): genetic code is required for each,of the enzymes involved in the biosynthesis, valuable amino acids are utilized in forming the enzymes, and a relatively scarce nutrient (nitrogen) is locked up in the compounds produced If the morphine did not continue to have value for the plant, mutants would have arisen with the advantage of not having a drain on their metabolic resources We can only guess a t the ecological functions of morphine Perhaps a mammalian herbivore that consumed too many poppies would become drowsy and itself fall prey to a carnivore It may be significant that the cannabinoids, produced in greatest abundance in the nutritious growing tips of the plant, also induce mental effects that would compromise a herbivore's ability to escape a predator Whatever their natural protective functions, natural products are a rich source of biologically active compounds that have arisen as the result of natural selection, over perhaps 300 million years The challenge to the medicinal chemist is to exploit this unique chemical diversity The following account illustrates how natural products have been used as what are called lead compounds, or templates for the development of important medicines INTRODUCTION Of the 520 new pharmaceuticals approved between 1983 and 1994,39% were derived from natural products, the proportion of antibacterials and anticancer agents of which was over 60% (1) Between 1990 and 2000, a total of 41 drugs derived from natural products were launched on the market by major pharmaceutical companies (Table 20.11, including azithromycin, orlistat, paclitaxel, sirolimus (rapamycin), Synercid, tacrolimus, and topotecan In 2000, one-half of the top-selling pharmaceuticals were derived from natural products, having combined sales of more than US $40 billion These included the biggest selling anticancer drug paclitaxel, the "statin" family of hypolipidemics, and the immunosuppressant cyclosporin During 2001 we have seen the launch of caspofungin from Merck and galantamine from Johnson & Johnson, with rosuvastatin, telithromycin, daptomycin, and ecteinascidin-743 due to follow in 2002 Despite the figures, the popularity of natural products, particularly those from higher plants as leads for new pharmaceuticals, tends to fluctuate At the time of writing, several of the world's biggest pharmaceutical companies have reined back their natural product drug discovery programs and have placed great faith in combinatorial chemistry, coupled to very high throughput screening Time will tell whether this is a wise stratagem, or whether the unique features of compounds that are themselves derived from living organisms will once again see renewed acceptance The abundance of plant and microbial secondary metabolites and their value in medicine are undisputed, but one question that is only partly answered concerns the reasons for this abundance of complex chemical substances In the past, the production of what we Drugs Affecting the Central Nervous System 849 Table 20.1 Drugs Derived from Natural Products (1990-2000) Name Acarbose Artemisinin Azithromycin Carbenin Cefetamet pivoxil Cefozopran Cefpimizole Cefsulodin Clarithromycin Colforsin daropate Docetaxel Dronabinol Galantamine Gusperimus Irinotecan Ivermectin Lentinan LW-50020 Masoprocol ~e~artricin Miglitol Mizoribine Mycophenolate mofetil Orlistat Paclitaxel Pentostatin Podophyllotoxin Policosanol Everolimus Sirolimus Sizofilan Subreum Synercid Tacrolimus Teicoplanin Tirilazad mesylate Topotecan Ukrain Vinorelbine Voglibose 2-100 Originator IndicationDJse Bayer Kunming & Guilin Pliva sankyo Takeda Takeda Ajinomoto Takeda Taisho Nippon Kayaku Aventis Solvay Intelligen Nippon Kayaku Yakult Honsha Merck & Co Ajinomoto Sankyo Access SPA Bayer Asahi Chemical Hoffman-LaRoche Hoffman-LaRoche Bristol-Myers Squibb Warner-Lambert Nycomed Pharma Dalmer Novartis American Home Prod1lcts Taito OM Pharma Novartis Fujisawa Aventis Pharmacia & Upjohn GlaxoSmithKline Nowicky Pharma Pierre Fabre Takeda Zeria Diabetes Malaria Antibiotic Antibiotic Antibiotic Antibiotic Antibiotic Antibiotic Antibiotic Asthma Cancer Alzheimer's disease Alzheimer's disease, arthritis Arthritis Cancer Parasiticide Cancer Immunomodulation Cancer Benign prostatic hyperplasia Diabetes Arthritis Arthritis Obesity Cancer Leukemia Human papillomavirus Hyperlipidaemia Immunomodulation Immunomodulation Cancer, hepatitis-B virus Arthritis Antibiotic Immunomodulation Antibiotic Subarachnoid haemorrhage Diabetes Cancer, HIVIAIDS Cancer Diabetes, obesity Immunomodulation DRUGS AFFECTING THE CENTRAL NERVOUS SYSTEM 2.1 Morphine Alkaloids The history of the opium alkaloids is too well known to warrant repetition here, but the analgesics based on morphine (1)are too important to be left out of an account of natural products as leads Thus we shall summarize the clinically more important developments that have occurred since the isolation of morphine in 1803 Codeine (2) continues to be used widely for the treatment of moderate pain and, although present in the opium poppy (Papaver somniferum), it is normally synthesized in higher yield from morphine (4) Other than codeine, the earliest significant semisynthetic derivative of morphine is the diacetate heroin (31, which is still widely used in terminal cancer where its addictiveness is ir- Natural Products as Leads for New Pharmaceuticals (1)morphine R1 = Rz = H (2) codeine Rl = CH3, R2 = H (3) heroin Rl = Rz = COCH3 (5) naloxone relevant Acetylation masks the polar hydroxy groups, so that penetration into the central nervous system (CNS)is enhanced; hydrolysis then occurs to liberate the phenolic hydroxyl, giving an active analgesic, and ultimately regenerates morphine (5) Heroin was thus one of the first prodrugs Modifications to the C-ring of morphine are legion, but none of the derivatives is free from addictive liability, though many have been used clinically N-Demethylation and realkylation yield more interesting analogs, notably N-allylnormorphine and nalorphine (4), which is a morphine antagonist (6) Further modification leads to naloxone (51, which unlike nalorphine has very little agonist activity (7) and has retained a place in therapy for treatment of opiate-induced respiratory depression Naloxone will also precipitate withdrawal symptoms in opiate addicts, thereby facilitating diagnosis give the morphinans (6) The system may be simplified even further (9),to give the benzomorphans (7), although neither these nor the morphinans have provided the long-sought analgesic without addictive properties (6) morphinan (7) benzomorphan (4) nalorphine Total synthesis of morphine is difficult, but analogs lacking the dihydrofuran ring are accessible (8) from 1-benzylisoquinolines, in analogy with the biosynthesis of morphine, to A semisynthetic route to morphine analogs was found (10) from thebaine (8) using Diels-Alder reactions in the C-ring Adducts such as (9) have the distinction of enormous potency (1I), sufficient to immobilize rhinoceroses at moderate dose levels! Unfortunately, the addictive liability runs parallel to the increase in analgesic potency, a tendency that was partly overcome (12) in the analog buprenorphine (10) 2 Drugs Affecting the Central Nervous System (8) thebaine vation that meperidine (pethidine) (12) unexpectedly produced a reaction in mice known as Straub tail, normally characteristic of the morphine series (15) Meperidine itself is still used widely in childbirth in the belief that there is a lower incidence of respiratory depression in the fetus The realization that 4-phenylpiperidines, which are not obvious structural analogs of morphine, could give rise to useful analgesic effects, led to the synthesis of many thousands of derivatives (161, many with far greater potency than that of meperidine Unfortunately, as potency increases so addiction liability and respiratory depression (9) etorphine (11) atropine (10) buprenorphine All this work was carried out in ignorance of the nature of the natural transmitter(& which subsequently proved to be the peptides known as endorphins and their pentapeptide fragments, the enkephalins (13) It is perhaps significant that vastly improved understanding of the biochemical basis for analgesia and the characterization of a family of related receptors (14), known as 8, K , and p, have so far failed to yield any better drugs for the treatment of pain A series of analgesics that were discovered initially in an attempt to obtain smooth muscle relaxants based on another natural product, atropine ( l l ) ,started with the obser- (12) pethidine 2.2 Conotoxins Elan Pharmaceuticals is developing SNX-111 (Ziconotide), the synthetic equivalent of w-Conopeptide-MVIIA, found in the venom of the predatory marine snail Conus magus, for the treatment of severe pain and ischemia by the intrathecal or intravenous routes The peptide has the structure H-'Cys-Lys-Gly- Lys-Gly-Ala-Lys-'Cys-Ser-Arg-Leu-Met-TryAsp-15Cys-16Cys-Thr-Gly-Ser-20Cys-Arg-SerGly-Lys-25Cys-NH,cyclic(1-16),(8-20),(1525)-tris(disulfide), which does not make it an Natural Products as Leads for New Pharmaceuticals OH (13) conotoxin analog easy target for synthesis and gives it poor distribution properties in vivo (17) SNX-111 blocks N-type calcium channels, which are located throughout the CNS on neuronal somata, dendrites, dendritic spines, and axon terminals, where they play a major role in the regulation of the neurotransmitters associated with pain transmission and stroke The drive is to discover an orally active, selective, small-molecule modulator of N-type calcium channels to overcome the disadvantages of administration of SNX-111 High-throughput screening campaigns have resulted in a number of leads being identified; whereas others have chosen to modify known drugs shown to block N-type channels Workers at Parke-Davis, however, employed a ligand-based approach using the three-dimensional solution structure of the peptide (18) Compounds such as (13)were designed where key binding motifs are attached to an alkylphenyl ether scaffold The compound had an IC,, value of 3.3 pit4 in a human N-type channel assay but showed no selectivity over the L-type channel Structure-activity work on the conotoxins has shown that other regions of the peptide, absent in these synthetic ligands, are responsible for channel family selectivity (17, 18) 2.3 drocannabinol(14) (THC), which has a multiplicity of actions In animals the effects include sedation and apparent hallucinations (19), which are similar to the major effects in the CNS in humans There are also cardiovascular effects, notably tachycardia and postural hypotension, that can be separated from the CNS action, as in the synthetic analog A,,,,,dimethylheptylTHC (151, which has minimal CNS activity (20) (14) THC Cannabinoids The plant Cannabis sativa has been used by humans for thousands of years, both for the effects when ingested and for making rope from the fibers in the stem The major constituent of pharmacologicalinterest is A,-tetrahy- Given the widespread illicit use of C sativa, it was perhaps inevitable that eventually one Drugs Affecting the Central Nervous System or two cancer patients receiving chemotherapy would dose themselves with their own sedative in the form of marijuana An unexpected blessing from this uncontrolled combination was a reduction in the nausea experienced during chemotherapy A variety of anticancer agents cause severe nausea and vomiting, including nitrogen mustard, adriamycin, 5-azacytidine, cyclophosphamide, and methotrexate: the unique situation arose in which the remedy was discovered by the patients themselves (21) Although smoking reefers gives rapid absorption and close control of the effects, smoking is itself carcinogenic and cannot be recommended to those who are unaccustomed to it; thus, when the physicians in charge were made aware of their patients' discovery, they devised a controlled clinical trial in which measured doses of THC were dissolved in sesame oil and administered in gelatin capsules A placebo was similarly prepared for use in a randomized, double-blind, crossover experiment (21).The results left no doubt that a majority of patients benefited from THC pretreatment, even those who had previously been refractory to the effects of the standard antiemetics such as prochlorperazine There remained the problem of tachycardia associated with THC treatment The multiplicity of effects of THC have led to the synthesis of large numbers of analogs (221, particularly in the hope of finding non-morphinelike analgesics without addictiveness and without the other CNS effects of THC The analog nabilone (16) had been shown to exert less effect than that of THC on the cardiovascular system, while retaining the mixture of CNS actions, including analgesic, antianxiety, and antipsychotic properties (23) When tested as an antiemetic, nabilone proved to be superior to THC (24) and has been used for this purpose for more than 30 years The f i s t 10 years of clinical experience was reviewed (25) After the demonstration of THC binding sites in the CNS (261, a search for an endogenous ligand produced the long-chain ethanolamine derivative (17) of arachidonic acid, known as anandamide (27) Subsequently, the glycerol ester of arachidonic acid (la), known as 2-AG, was shown to be a more abundant endogenous ligand in the brain than anandamide (28) Further development has tended (16) nabilone to concentrate on analogs of the natural ligands, notably the methyl derivative of anandamide (19)' which is resistant to the amide hydrolase that terminates the action of anandamide itself and the dimethylheptyl analog (20) that is traceable to the earlier modifications to THC (29) Such analogs tend to have activity similar to that of THC (17) anandamide An interesting twist in the tail is provided by the observation that anandamide is also a ligand for the so-called enigmatic vanilloid re- Natural Products as Leads for New Pharmaceuticals plus a hydrolase and a transport protein, interference with any or all of which might provide new drugs (22) resiniferatoxin ceptors, previously characterized through their interactions with two other natural products, capsaicin (21) and resiniferatoxin (22) (30) and responsible for the "hot" sensation caused by compounds in, for example, chilies A functional vanilloid receptor was cloned in 1997 and is activated by heat and acid as well as the chemical ligands (30) A combination of the anandamide structure with a vanilloid motif, as in AM404 (23), enhances the anandamide transport inhibitory properties (29) The situation is complex from the viewpoint of drug design, not least because there are two cannabinoid (CB) receptors, The cannabinoid acids, which are devoid of psychotropic activity, are promising anti-inflammatory agents (31) and it is possible that the next useful therapeutic agent will come from this direction, rather than the soughtafter analgesic (21) capsaicin Drugs Affecting the Central Nervous System 855 Asperlicin is moderately potent, poorly soluble in water, and not bioavailable by the oral Cholecystokinin (CCK) is a peptide hormone, route (41) When discovered it was also, with present in the gut and CNS; it is one of the morphine, one of the very few nonpeptides most abundant peptides in the brain (32,331 with affinity for a peptide receptor (peptoids The whole peptide is composed of 33 amino are discounted in this assessment) It was an acids, but the C-terminal octapeptide H-Aspinteresting target for synthetic modification, Tyr(S0,H)-Met-Gly-Trp-Met-Asp-Phe-NH, particularly viewed as a benzodiazepine derivpossesses the full range of activities, sufficient ative with potential CNS activity for it to be classed as a neurotransmitter (34) Based on the benzodiazepine nucleus, and Specific, high-affinity binding sites have been an overt mimic of diazepam, one of the first found on mammalian CNS cell membranes successful synthetic analogs was L-364,286 and in other organs such as pancreas, gall (25), which had potency on CCK-A receptors bladder, and colon (35) The latter have been similar to that of asperlicin Better receptor classed as CCK-A receptors, but the majority affinity was achieved with 3-amide-substiof CNS receptors were classed as CCK-B, tuted benzazepines: the 2-indolyl derivative based on affinity differences for various agoL-364,718, also known as MK-329 (261, is five nists and antagonists (36) To confuse the isorders of magnitude more potent than aspersue slightly, the gastrin receptor in the stomlicin (42) at CCK-A receptors and is a valuable ach is closely related to the CCK-B (now pharmacological tool known as the CCK,) receptor (37) and is stimulated by the C-terminal tetrapeptide of CCK: in the periphery, gastrin receptors are the same as CCK, receptors (38) The effects of CCK on intestinal smooth muscle and pancreas are easy to demonstrate pharmacologically, unlike the role in the CNS, which is a matter for conjecture It was assumed that the CNS activity must be significant, given the abundance of the peptide in the brain, and that the discovery of antagonists might lead to new drug treatments, as yet unspecified (39) Fishing in microbial broths, using radioreceptors as bait, produced asperlicin (241, the first potent, competitive and selective CCK-A (CCK,) antagonist, from a culture medium of Aspergillus alliaceus (40) 2.4 Asperlicin (24) asperlicin Modification of the 3-amide to give a urea linkage as in (27) led to a reduction in CCK-A receptor affinity Importantly, discrimination between CCK-A and CCK-B receptors by (27) 856 Natural Products as Leads for New Pharmaceuticals is governed by the stereochemistry at C3, the (S)-enantiomer showing greater affinity for CCK-A receptors The (R)-enantiomer,known as L-365,260, prefers CCK-B receptors, antagonizes gastrin-stimulated acid secretion in animal models, and, among other CNS effects, induces analgesia in primates and displays anxiolytic properties (32) NEUROMUSCULAR BLOCKING DRUGS 3.1 Curare, Decamethonium, and Atracurium The development and use of muscle relaxants, to allow a reduction in the level of anesthesia during surgery, follows entirely from studies of South American arrow poisons (44)and particularly from the isolation by King (45) of pure D-tubocurarine (29) in the 19309, from tube curare Another of the South American blowpipe poisons, calabash curare, was used for similar purposes and developed (46,47), to give alcuronium (30) from the alkaloid C-toxiferine (31) Both types of curare paralyze skeletal muscle by a similar mechanism, antagonizing the effect of acetylcholine at the neuromuscular junction (48) Further development in this series has very substantially improved receptor affinity: YM022 (28) has IC,, 0.05 nM/kg (38) Clinical trials of compounds in this series have been disappointing because of poor bioavailability, but the general concept of finding a therapeutic agent through antagonism of CCK, receptors is still viable and it is reported that the number of patents in this area has increased in the last years (43) (29) tubocurarine R = H (32) metocurine R = CH3 The muscle-paralyzingcurare alkaloids are quaternary salts that are not absorbed when taken orally For surgical procedures they must be administered by intravenous injection, which results in onset of paralysis in at most a few minutes: anesthesia is normally induced before administration of the muscle relaxant (44), which is followed by artificial respiration Although the neuromuscular blocking agents are potentially lethal when administered alone, in the environment of an operating theater they are truly life-saving Neuromuscular Blocking Drugs (31) C-toxiferine R = CH3 (30) alcuronium R =CH2CH=CH2 drugs that have made a major impact on survival rates during surgery At the time of King's work in the 1930s there were no spectroscopic aids to structure elucidation, and it is not surprising that he made a small error in the structure assigned to D-tubocurarine, believing it to have two quaternary nitrogens, a mistake that was not corrected (49) until 1970 The methylation product of D-tubocurarine, known as metocurine (32) is a more potent muscle relaxant It was known for a long time as dimethyltubocurarine because of the error in the structure allocated to compound (29) King's error, in assigning a bisquaternary structure to a molecule with one quaternary and one protonated tertiary nitrogen, led to a large number of highly active synthetic bisquaternaries The simplest of these was decamethonium (331, which was nothing more than two trimethylammonium end groups connected with a decamethylene chain As one of a series with different chain lengths (50), decamethonium became the prototype for many more complex structures with 10 atoms between the quaternary centers, which appeared to be optimal for (33) decamethonium 857 binding to the acetylcholine receptor at the neuromuscular junction Unlike tubocurarine, decamethonium depolarizes the muscle endplate, rendering the membrane insensitive to acetylcholine (48) The action of tubocurarine is competitive and can be overcome with increased concentrations of acetylcholine, brought about by administration of an anticholinesterase: the latter is thus an antidote to tubocurarine, but not to decamethonium Despite the lack of an antidote, decamethonium was used very widely for over two decades One of its disadvantages is an overlong duration of action, during which time the patient has to be maintained on artificial respiration, because the muscle of the diaphragm is also susceptible to the actions of the drug An early and highly successful attempt (51) to shorten the action of decamethonium gave suxamethonium (34), a diester formed between succinic acid and two molecules of choline, which hydrolyzes rapidly in the presence of pseudocholinesterase Tubocurarine suffers from cardiovascular side effects induced by direct interactions with ganglionic acetylcholine receptors and from stimulation of histamine release, so analogs have been well worth pursuing The macrocyclic structure of tubocurarine is a difficult synthetic target, but fortunately ring-opened analogs, such as laudexium (351, have high potency and relatively few side effects (52) The main problem with (35)is the duration of action, which at about 40 is too long for many operations Two approaches have been used to shorten the duration of action The concept of pH-controlled Hofmann elimination was employed successfully (53) in the design of atracurium (36),which in clinical use (54) has the big advantage that the drug disappears at a constant rate, irrespective of liver or kidney function Some ester hydrolysis contributes to the destruction of atracurium in vivo, as might be expected A slightly later development (55) centered on an empirical search for structures that would undergo ester hydrolysis more rapidly, resulting in mivacurium (37), which has a slightly shorter duration of action than that of atracurium, the latter being about 15-20 Natural Products as Leads for New Pharmaceuticals 0 /\/COOCH2CH2N(CH3)3 (H3C)3NCH2CH20C0 pseudocholinesterase "4-COOH (H3C)3NCH2CH20C0 HOCH2CH2N(CH3)3 decomposition of suxamethonium (34) (35) laudexium 4.1 ANTICANCER DRUGS Catharanthus (Vinca) Alkaloids In 1949 Canadian researchers at the University of Western Ontario began investigating the medicinal properties of the rosy periwinkle (Catharanthus roseus), a plant that had been used for many years to treat diabetes mellitus in the West Indies Despite finding that the plant extract when given orally had no effect on blood sugar levels in rats or rabbits, the researchers noted that when given intravenously, the extract caused the animals to succumb to bacterial infection and die This curious observation prompted further studies, which showed that the plant extract reduced levels of white blood cells, causing granulocytopenia and bone marrow damage, toxic effects that are encountered with many antitumor drugs (56) These findings led the Canadian group to isolate an alkaloid fraction with potent cytotoxic activity The active principle was eventually purified and became known as vinblastine (38), a dimeric indoledihydroindole alkaloid Concurrently, researchers at the Lilly Research Laboratories had been investigating extracts of C roseus and they too had detected cytotoxic activity, specifically against acute lymphocytic leukemia (57,581.The U.S group isolated several alkaloids, including vinblastine and another closely related alkaloid, vincristine (39) Although many other alkaloids have been isolated from C roseus, only vinblastine and vincristine have been developed for clinical use The antiproliferative activity of the two compounds is related to their specific interaction with tubulin, thus preventing assembly of tubulin into microtubules and arresting cell division (59) However, despite this apparent identical mechanism of action and their clear chemical similarities, vinblastine and vincristine display very different clinical effects Vinblastine, for example, is used to treat Hodgkin's disease and metastatic testicular tumors, whereas vincristine is used mainly in combination with other anticancer drugs for the treatment of acute lyrnphocytic leukemia in children Toxicity profiles are also different, in that vinblastine causes bone-marrow depression, whereas peripheral neuropathy often proves to be dose-limiting in vincristine therapy 4 Anticancer Drugs (36) atracurium pH 7.4 (37) mivacurium Lilly introduced vinblastine and vincristine into the clinic in 1960 and 1963, respectively, but this did not preclude the search for improved derivatives A chemical modification program aimed at improving antitumor activity and reducing toxicity was initiated in 1972 (60) Concern about the neurotoxicity displayed by vincristine, its chemical instability, and low natural abundance (0.03 glkg dried plant material) led to vinblastine's being chosen as a template for semisynthetic modifica- tion Selective ammonolysis of the ester function at C-3 and hydrolysis of the adjacent acetyl group yielded the desacetyl vinblastine amide, vindesine (40) Better yields of vindesine were obtained from the hydrazide (41) on treatment with nitrous acid and reacting the resultant azide (42) with ammonia The azide (42) proved to be a useful intermediate for the preparation of a range of substituted amides, although vindesine proved to be the derivative of choice, with significant differ- Natural Products as Leads for New Pharmaceuticals 860 (38) R = CH3 (39) R = CHO ences in the spectrum of antitumor activity and toxicity compared to that of the naturally occurring alkaloids Phase I clinical trials commenced in 1977 and vindesine has been used for the treatment of non-small cell lung cancer, lymphoblastic leukemia, and nonHodgkin's lymphomas In combination with cisplatin, vindesine ranks among the foremost treatments for non-small cell lung cancer with respect to response rate and survival (61) Back in the 1950s, the U.S researchers could not have guessed that 30 years on, the demand for Catharanthus alkaloids would necessitate the processing of around 8000 kg of plant material per year (62)! 4.2 Camptothecin Camptothecin (43) was first isolated by Monroe Wall and Mansukh Wani in 1966, after ethanolic extracts of Camptotheca acuminata, a tree native to China, showed unusual and potent antitumor activity (63) Starting with 19 kg of dried wood and bark, Wall and Wani painstakingly purified the principal active component with a combination of hot solvent extraction, an ll-stage Craig countercurrent partition process, silica gel chromatography, and crystallization Camptothecin was characterized as a novel pentacyclic alkaloid, present as just 0.01% wlw of the stem bark of C acumi- Anticancer Drugs nata Of particular note was the unusual activity that camptothecin displayed in L1210 and P388 mouse leukemia life-prolongation assays The compound also inhibited the growth of solid tumors in vivo and the watersoluble sodium salt was progressed to phase I1 clinical trials before being withdrawn because of severe bladder toxicity (43) Camptothecin: R1 = R2 = R~ = H (44) 10-hydroxycamptothecin: R1 = R2 = H, ical activity, but the 10-hydroxy analog (44) showed greater activity than that of (43) (65) Wall and Wani successfully deployed the Friedlander reaction between substituted 2-aminobenzaldehydes and the tricyclic intermediate (45), to synthesize a variety of ringA-substituted analogs These studies may have prompted SmithKline Beecham (now GlaxoSmithKline)to synthesize the water-soluble 10-hydroxycamptothecin analog topotecan (46) that was first approved in 1996 for the treatment of recurrent ovarian cancer and, years later, for small cell lung cancer (66) Irinotecan (471, developed by Daiichi and Yakult Honsha in Japan and marketed by Pharmacia, was also approved in 1996 for the treatment of advanced colorectal cancer Irinotecan is inactive as a topoisomerase I inhibitor, but acts as a prodrug of the active 7-ethyl-10-hydroxycamptothecin (48) (67) R~ = OH (48) 7-ethyl-10-hydroxycamptothecin: R1 = C2H5, R~ = H, R3 = OH (46) Topotecan: R1 = H, R2 = CH2-N(CH3)2, R~ = OH (47) Irinotecan: R1 = CzH5,R2 = H, 1I - N ' J - N ~ R3 = O - C Interest in camptothecin gained new impetus in 1985, when it was discovered that the compound exerts its antitumor activity through a novel mechanism of action (64) Camptothecin binds to the covalent complex formed by topoisomerase I and DNA, which initiates DNA replication and thus stabilizes the enzyme-DNA complex and prevents cell proliferation The elucidation of the mechanism of action provided a means of evaluating camptothecin analogs as topoisomerase inhibitors in vitro and efforts then focused on synthesizing water-soluble analogs with broadspectrum antitumor activities The a-hydroxy lactone (ring E) and, in particular, the 20(S)form proved essential for maintaining biolog- 4.3 Paclitaxel and Docetaxel Regarded as the tree of death by the Greeks and used to prepare arrow poison by the Celts, the yew tree has been associated with death and poisoning for centuries (68, 69) The English yew, Taxus baccata, was used to make funeral wreaths and it was believed that one could die by merely standing beneath the boughs of the tree Yew certainly contains highly toxic metabolites and their potency and fast duration of action has often made extracts of yew the poison of choice for numerous murders and suicide attempts It is thus ironic that extracts from the Pacific yew, T brevifolia, after being Natural Products as Leads for New Pharmaceuticals tested in the National Cancer Institute's (NCI) screening program during the 1960s, yielded what was described (70) as the most exciting anticancer compound discovered in the previous 20 years; that is, paclitaxel (49) (originally given the name tax01 by Wall and Wani) The initial isolation and characterization of paclitaxel proved particularly difficult because of (1) its very low natural abundance in T breuifolia bark (although this was the best known source, the isolated yield was only 0.02% w/w, equivalent to 650 mg per tree), (2) the poor analytical data obtained from the purified compound, and (3) the failure of paclitaxel to give crystals that were suitable for X-ray analysis (71) The structure of paclitaxel was published in 1971 (72),but further biological testing continued to be troubled by difficulties The compound showed only modest in viuo activity in various leukemia assays, which was no better than that displayed by a number of other new compounds at the time In addition to the limited supplies of paclitaxel (the complexity of the molecule precluded chemical synthesis), the compound was very poorly soluble in water, which made formulation difficult However, various new assays were developed in the 1970s, including the murine B16 melanoma model, in which paclitaxel showed very good activity, and another boost came when Horwitz et al (73) discovered that the compound prevented cell division by a unique mode of action In contrast to the antimitotic vinblastine and podophyllotoxin analogs (q.v.), which prevent microtubule assembly, paclitaxel inhibits cell division by promoting assembly of stable microtubule bundles, which leads to cell death Phase I clinical trials were initiated in 1983, but these were to proceed at a slow and tortuous pace and proved all but disastrous when the high levels of oil-based adjuvant used to formulate paclitaxel caused severe allergic reactions in many volunteers Undaunted by the formulation problem and spurred on by paclitaxel's novel mechanism of action, clinicians were able eventually to minimize the allergic events and demonstrate useful activity Phase I1 clinical trials began in 1985 despite continuing supply problems, and years later the program received a significant boost when McGuire et al (74) reported good responses from patients suffering from refractory ovarian cancer, a disease that kills some 12,500 women a year in the United States alone In many ways, the development of paclitaxel mirrored that of the camptothecin analogs, both being dogged for many years by supply issues, poor pharmacokinetics, and toxicity, but the subsequent uncovering of novel mechanisms of action fueled renewed efforts to develop these leads into important new anticancer agents (75) In 1991 Bristol-Myers Squibb in conjunction with the NCI agreed to manage the supplies of paclitaxel and were granted a licence to further develop the compound The following year the U.S Federal Drug Administration approved paclitaxel for the treatment of ovarian cancer in patients unresponsive to standard treatments, and in December 1993 approval was given for the treatment of metastatic breast cancer The sourcing of paclitaxel from T brevifolia was a major problem (76) because to treat just the groups of patients suffering ovarian cancer in the United States would require about 25 kg of compound per year, necessitating the felling of some 38,000 trees (70)! Although the Pacific yew is not a rare tree, it is extremely slow growing and such harvesting could not be sustained indefinitely It has been estimated that there were enough trees available to maintain a supply of paclitaxel for only 2-7 years (77) The isolation of paclitaxel from other Taxus species has been investigated at length and reasonable quantities have been obtained from the needles of several species including T baccata Using the needles has Anticancer Drugs 863 alleviated the supply problem because they can be harvested without damaging the tree However, the needles contain much higher quantities of several biosynthetic precursors of paclitaxel and two of these, baccatin I11 (50) and 10-desacetylbaccatin I11 (51) have been used to prepare paclitaxel semisynthetically One approach, developed by Potier et al (781, involved acylation of the sterically hindered C-13 position of baccatin 111 with cinnamic acid and subsequent double-bond functionalization through hydroxyamination, to give paclitaxel together with various regio- and stereoisomers A better approach involved protection of 10-desacetylbaccatin I11 as the triethylsilyl ether, followed by direct acylation with the phenylisoserine derivative (521, giving paclitaxel in 38% overall yield (79) Further improvements were made using less sterically demanding acylating reagents; for example, acylation with the p-lactam (53) gave paclitaxel in up to 90% yield (80) and this may be the preferred method for commercial production in the future the C-13 ester side-chain can be tolerated Thus, the N-t-(butoxycarbonyl)derivative, docetaxel(54), which appears to be more potent than paclitaxel (81) and has better solubility characteristics, has been developed and launched by Aventis for the treatment of ovarian, breast, and lung cancers Various "protaxols," designed to release paclitaxel in situ under physiological conditions, have been prepared by acylating the C-2' hydroxyl group Nicolaou et al (82) reported the synthesis of the sulfone (551, which is soluble and stable in aqueous media, but is able to release paclitaxel rapidly in human blood plasma T ~ ~ ; EtO Ph \\ \ COPh These semisynthetic approaches also provide access to analogs with potential advantages over paclitaxel itself Structure-activity studies have shown that, although the oxetane ring appears to be essential for activity, wide variation in the nature and stereochemistry of Plant tissue culture (70),microbial fermentation (83),and total synthesis (84,85)provide other possibilities for the production of paclitaxel and its derivatives, although it is far from certain whether any of them will be commercially viable Natural Products as Leads for New Pharmaceuticals OCOC~H~ 4.4 Epothilones Epothilones A (56) and B (57), 16-membered macrocyclic polyketide lactones, were first isolated from the cellulose-degrading myxobacterium Sorangium cellulosum by Hoefle, Reichenbach, and coworkers (86) as narrowspectrumantifungal and cy~otoxicmetabolites The compounds were then tested by the National Cancer Institute in the United States and found to be highly active against breast and colon cancer cell lines (87) Subsequently, 0.42 -0u R ,,,,\\OH / h,,,, 11 X OH (56) epothilone A: X = 0,R = H (57) epothilone B: X = 0,R = CH3 (59) BMS 247550: X = NH, R = CH3 Bollag et al (88)at the Merck Research Laboratories discovered that the epothilones stabilize microtubule assembly and thus inhibit cell division by the same mechanism as that of paclitaxel (see above) This observation, together with their less complex chemical structure, increased water solubility, more rapid action in vitro, and effecti~enessagainst multidrug-resistant tumor cells, has prompted significant interest in the epothilones as anticancer agents On learning the absolute stereochemistry of (56) and (571, three academic research groups embarked on the total synthesis of the epothilones Nicolaou, Danishefsky, and Schinzer independently adopted successful, elegant synthetic approaches involving olefin metathesis, macrolactonization, Suzuki coupling, or ester-enolate-aldehyde condensation (89) Within years of the disclosure of their absolute stereochemistry, 17 different total syntheses of the natural products were reported These syntheses paved the way for the generation of a large number of epothilone Anticancer Drugs analogs for biological evaluation, including the use of solid-phase combinatorial approaches The academic groups focused on modifications around the core macrocyclic lactone, establishing important structure-activity relationships, but not improving on the in vitro biological activity of the most active natural product, epothilone B (57) In vivo biological data were comparatively scarce and, although one group reported that epothilones B (57) and D (58) showed activity in murine tumor models, researchers at Bristol-Myers Squibb have shown that (58)lacks in vivo activity as a result of rapid metabolic inactivation (90) It was postulated that esterase-mediated hydrolysis of the macrocyclic lactone formed an inactive ring-opened species and, therefore, efforts were focused on replacing the lactone with a more stable macrocyclic lactam moiety Several macrocyclic lactam derivatives were synthesized from (57) and (58).Of note was the preparation of BMS-247550 (59) in a three-step synthesis from epothilone B (571, utilizing a novel Pd(0)-catalyzed ring-opening reaction followed by reduction and macrolactamization BMS-247550 (59), which is in phase I clinical trials, retains its activity against human cancer cells that are naturally insensitive to ~aclitaxelor that have developed resistance to paclitaxel, both in vitro and in vivo (91) - (58) epothilone D 4.5 Podophyllotoxin, Etoposide, and Teniposide The development of the natural constituents of Podophyllum Resin into effective semisynthetic and, ultimately, totally synthetic compounds for the treatment of various kinds of cancer provides one of the most sustained and intriguing stories of drug discovery (92, 93) 865 The story has all the classic ingredients, starting with observation and reasoning, extending through chance into new areas, and characterized throughout by persistence and determination, particularly when biological activity had to be traced to very minor constituents in the crude plant extract Podophyllum peltatum (may apple, or American mandrake) and P emodi are respectively, American and Himalayan plants, widely separated geographically but used in both places as cathartics in folk medicine (94) An alcoholic extract of the rhizome known as podophyllin was included in many pharmacopoeias for its gastrointestinal effects; it was included in the U.S.P., for example, from 1820 to 1942 At about this time the beneficial effect of podophyllin, applied topically to benign tumors known as condylomata acuminata, was demonstrated clinically (95) This usage was not inspirational, given that there are records of topical application in the treatment of cancer by the Penobscot Indians of Maine and, subsequently, by various medical practitioners in the United States from the 19th century (96) The crude resinous podophyllin is an irritant and unpleasant mixture unsuited to systemic administration The first chemical constituent was isolated from podophyllin in 1880 and named podophyllotoxin (97) A structure was proposed in 1932 and after some fine-tuning (98) was shown to be the lignan (60) As might be expected, the crude resin contains a variety of chemical types, including the flavonols quercetin and kaempferol (99) Although these other constituents undoubtedly have biological activity, it is the lignans that have received most attention and to which we shall devote the remainder of this section Chemists at Sandoz in the early" 1950s reasoned that crude podophyllin might contain lignan glycosides with anticancer activity, which might be more water soluble and less toxic than podophyllotoxin (92) The reasoning for the latter is not entirely clear, but in the event they proved to be correct in both respects careful isolation gave podophyllotoxin p-D-glucopyranoside (61) its 4'-desmethyl analog (62) and some less important lignans lacking the B-ring hydroxy group (100-102) Unfortunately, the sugar deriva- Natural Products as Leads for New Pharmaceuticals (60) podophyllotoxin tives were less active as inhibitors of cell proliferation than were the aglycones, as well as less toxic; however, as expected, they were much more water soluble (92) While continuing work to isolate more natural lignans, a substantial program of structural modification of the known compounds was undertaken, with a view to protecting the glucosides from hydrolytic enzymes and also to improve cellular uptake Most of these changes were ineffective: the per-acylated derivatives, for example, were insoluble in water and had inferior cytostatic effects (103) this, water solubility was a problem with the podophyllotoxin derivatives (63) Gastrointestinal absorption was greatly improved, however, as was chemical stability (1041, and positive effects were observed in a few cancer patients with the benzylidene derivative (64) It was at this point that luck played a hand, backed up by a good deal of determination A crude podophyllin fraction, which was simpler and cheaper to prepare than pure podophyllin glucoside, was also treated with benzaldehyde to give a mixture of benzylidene derivatives, about 80% of which was compound (64) The crude product was found to be more potent than compound (64) and subsequently to possess a different mode of action from that of the lead compounds: rather than arresting cells in metaphase, cells were prevented from entering mitosis altogether (105) The crude mixture was marketed for cancer treatment as Proresid (63) R1 = H,CH3 R2 = various alkyl, aryl (64) R1 = CH3 R2 = C6H5 Condensation of the glucosides with a variety of aldehydes was more useful, in that not all the hydroxy groups were blocked Despite Improved biological assay methods (106) indicated the presence of an unknown, highly active constituent of Proresid For example, Proresid prolonged the life of mice inoculated with L1210 leukemia cells (93), an effect that was not observed with the known major constituent In the early 1960s chromatographic and spectroscopic techniques were not as highly developed as they are now and more Anticancer Drugs than years' work was required to isolate and identify the unknown component of the mixture, which proved to be the 4'-desmethoxy-1epi analog (65) of the podophyllotoxin glucoside adduct (92) Present only in very small amounts in the derivatized extract, it was necessary to devise a synthesis from readily available materials It was fortunate that the desired 10 configuration was readily secured from la-hydroxy-4'-desmethylpodophyllotoxin, itself obtained by selective demethylation of podophyllotoxin: the remainder of the synthesis would now be considered fairly routine (107) Given a large supply of the key intermediate (66), it was straightforward to prepare a number of aldehyde derivatives, resulting in analogs with up to a 1000-fold increase in potency (108) The selected adducts were those prepared from thiophen-2-aldehyde,giving teniposide (67), and from acetaldehyde, giving etoposide (68) Both drugs are of value, etoposide in the treatment of small-cell lung cancer and testicular cancer, teniposide in the treatment of lymphomas and leukemias The thiophene derivative is also of use in the treatment of brain tumors (93) The natural products, podophyllotoxin and its congeners, are "spindle poisons" that inhibit cell proliferation by binding to tubulin and preventing formation of microtubules (105) Presumably this effect is sufficient to account for the success of podophyllin in the treatment of condylomata acuminata, although the crude extract contains many other candidates for a contribution to the biological activity As has been described, a very minor component of the natural mixture, missing the 4' hydroxy group, having the lp- instead of the la-hydroxy configuration and with this hydroxy group conjugated with p-D-glucose, must be treated with an aldehyde to produce the highly active and most important derivatives These derivatives not bind to tubulin, but have been shown to be inhibitors of topoisomerase 11, which may account for most of the observed biological effects, including DNA strand breaks, that lead to anticancer activity (109) 4.6 Marine Sources Cytosine arabinoside (69), a synthetic analog of the C-nucleosides spongouridine (70) and spongothymidine (71) from the sea sponge Cryptotheca cripta, was the first and, so far, the only marine-derived compound used routinely as an anticancer agent (110) However, a number of chemically diverse natural products from marine sources have been progressed to clinical trials The three most advanced compounds are in phase I1 trials; ecteinascidin-743 (721, a tetrahydroisoquino- Natural Products as Leads for New Pharmaceuticals 868 line alkaloid isolated from the mangrove ascidian Ecteinascidia turbinata, bryostatin-1 (731, a macrolide isolated from the bryozoan Bugula neritina, and dolastatin-10 (74), a linear peptide from the sea mollusk Dolabella auricularia Ecteinascidin-743 (72) was first isolated by Rinehart's group at the University of Illinois (111)and has been licensed to PharmaMar (Zeltia), which plans, together with Ortho Biotech, to launch the compound in late 2002 initially for the treatment of soft tissue sarcoma (112) (72) Ecteinascidin 743 tunately, supplies of ecteinascidin-743 (72) and also bryostatin-1 (73) have been met by aquaculture techniques (115),and more viable synthetic routes are now available for (73) (116) and dolastatin-10 (74) (117) (69) Cytosine arabinoside (70) Spongouridine: R = H (71) Spongothymidine: R = CH3 Yields of marine-derived natural products are invariably low and supply problems have delayed their development as useful pharmaceutical agents For example, over 3000 kg of the sea squirt E turbinata is required to produce g of ecteinascidin-743,sufficient forjust one cycle of treatment (113) and 1000 kg of B neritina yields 1.5 g of bryostatin-l(114) For- ANTIBIOTICS In 1929 Alexander Fleming published the results of his chance finding that a Penicillium mold caused lysis of staphylococcal colonies on an agar plate (118) He also showed that the culture filtrate, named penicillin, possessed activity against important pathogens including Gram-positive bacteria and Gram-negative cocci However, it was not until 1940 that the true therapeutic efficacy of penicillin was revealed, when Chain et al (119) successfully tested the material in mice that had been previously infected with a lethal dose of streptococci Several years later the precise chemical structure of the main active component, benzylpenicillin (75),was determined and efforts to synthesize the compound were initiated (120) Benzylpencillin proved to be an elusive target because of the instability of the p-lactam ring: it was unstable under acid conditions and was deactivated by p-lactamase enzymes produced by various Gram-positive and Gram-negative bacteria 5 Antibiotics (73) Bryostatin I v (74) Dolastatin 10 The discovery that the fused p-lactam nucleus, 6-aminopenicillanic acid (6-APA) (76), could be obtained from cultures of Penicillium chrysogenum led to the preparation of new, semisynthetic derivatives with improved stability to gastric acid and p-lactamases, and with activity against a wider range of pathogenic organisms (121) Sheehan (122) showed that compound (76) would react readily with acid chlorides to form new penicillin derivatives with novel substituents at the 6-position Methicillin (77), with a sterically demanding 2,6-dimethoxybenzamide side-chain, was the first semisynthetic penicillin to show resistance to staphylococcal p-lactamases, although the compound was still acid labile Ampicillin (78) has an a-aminophenylatamido side-chain and displays good activity against Gram-negativeorganisms, it is stable to acid and thus can be administered orally, although it is susceptible to degradation by p-lactamases Arnoxycillin (79) differs from ampicillin by the addition of a single hydroxy group, but the compound is better absorbed by the gastrointestinal tract Clavulanic acid (SO), isolated from Streptomyces clavuligerus, is similar in structure to the penicillins, except oxygen replaces sulfur in the five-membered ring (123) Clavulanic acid has weak antibacterial activity, but is a potent inhibitor of p-lactamases (124) A mixture of clavulanic acid and the p-lactamasesensitive amoxycillin was introduced in 1981 as Augmentin and has proved to be an effective combination to combat P-lactamase-producing bacteria (125) In 2001, 20 years after its launch, Augmentin is the best-selling antibacterial worldwide The clinical introduction of the penicillin group of antibiotics prompted an intensive search for novel antibiotic-producing organisms and Selman Waksman demonstrated the value of actinomycetes in this role, discovering the aminoglycoside streptomycin (81) from Streptomyces griseus in 1943 (126) Pharma- Natural Products as Leads for New Pharmaceuticals (78) R = COCHPh I ceutical companies also embarked on large programs of screening soil samples for antibiotic-producing microorganisms (127) Chloramphenicol(82) was isolated from Streptomyces venezuelae in 1948 and other clinically important antibiotics followed: chlortetracycline (83), neomycin (84), oxytetracyclin (85), erythromycin (86), oleandomycin (87), kanamycin (88),and rifamycin (89) In 1948 Giuseppe Brotzu isolated the fungus Cephalosporium acremonium from a water sample collected off the coast of Sardinia The culture showed significant antimicrobial activity, but Brotzu could not interest the Italian authorities in his discovery He then turned to a friend in England for help, who arranged for Howard Florey at Oxford to receive a sample of the producing culture Eventually, an antibacterial substance was isolated and named cephalosporin C (90) (128) The compound, which had a structure similar to that of the penicillins, except it had a dihydrothiazine ring fused to the p-lactam core, Antibiotics showed good resistance to p-lactamases and was less toxic than benzylpenicillin However, plans to market the compound were terminated with the introduction of methicillin (see above) The discovery that the basic structural building block of cephalosporin C, that is, 7-aminocephalosporanic acid (7-ACA) (911, could be synthesized led to the preparation of numerous cephalosporin derivatives in a similar way to the synthesis of penicillins from 6-aminopenicillanic acid (129, 130) Modification of the substituent at the 7-position, while retaining the 3-acetoxymethyl group, gave cephalothin (92),cephacetrile (93),and cephapirin (94), so-called first-generation cephalosporins with good activity against Gram-posi- tive bacteria, although the acetyl ester was susceptible to degradation by esterases and thus limited the duration of action Replacement of the acetoxy group by other substituents rendered the products less prone to esterase attack For example, the pyridinium derivative, cephaloridine (95),has a longer duration of action than that of cephalothin The first orally active cephalosporin was cephaloglycin (96), which possessed a phenylglycine substituent in the C-7 side-chain, although the labile 3-acetoxymethyl group was retained Replacing the acetoxy group with a proton or chlorine, for example, cephalexin Natural Products as Leads for New Pharmaceuticals (89) (97), cefadroxil (98), cephradine (991, and cefaclor (loo),extended the duration of action of these orally active products Cefaclor has been classified as a second-generation cephalospo- rin because it has a wider spectrum of activity, which includes Gram-negative bacteria such as Haemophilus influenme Cephamandole (101) and cefuroxime (102) are parenterally administered cephalosporins with similar activities against clinically important Gramnegative bacteria and are also resistant to many types of p-lactamases The newer third-generation cephalosporins, including ceftazidime (103),ceftizoxime (104),and ceftriaxone (105),which all contain an a-arninothiazolyl group in the C-7 sidechain, have been developed for treating specific pathogens such as Pseudomonas aeruginosa Thienamycin (106), isolated from Streptomyces cattleya in 1976, represented a new class of P-lactam antibiotics produced by bacteria where the sulfur of the penicillin nucleus was replaced by a methylene group (131) An N-formylimidoyl derivative, imipenem (107),was the first example from this Antibiotics COOH qlH2 R1= COCHPh R2 = C1 NH2 I R1 = COCHPh R1 = COC /n S NOC(CH3)2 I COOH new class of carbapenem antibiotics to become available for clinical use (132) Imipenem has a very broad spectrum of activity against most Gram-positive and Gram-negative aerobic and anaerobic bacteria Screening bacteria such as Pseudomonas acidophila and Chromobacterium uiolacium for production of p-lactam antibiotics resulted in the discovery of naturally occurring monobactams, which had moderate antimicrobial activity (133-135) Side-chain varia- tions, as developed for the penicillins and cephalosporins, led to compounds with improved activity against both Gram-positive and Gram-negative bacteria A derivative containing the a-aminothiazoyl group, a sidechain component common to the third-generation cephalosporins (see above), showed specific activity against Gram-negative aerobic bacteria, including Pseudomonas spp., and was stable to most types of p-lactamases The compound aztreonam (108) became the first Natural Products as Leads for New Pharmaceuticals 874 RHN 'rfS\l COOH I (90) R = COCH2CH2CH2CHNH2 (91) R = H (92) R = COCH2 (93) R = COCH2CN (94) R = COCHz (108) (formerlyStreptomyces erythreus) As a broadspectrum antibiotic erythromycin has proved invaluable for the treatment of bacterial infections in patients with p-lactam hypersensitivity and is also the drug of choice in the treatment of infections caused by species of Legionella, Mycoplasma, Campylobacter, and Bordetella (137) (109) Erythromycin A, R = H (114) Clarithromycin, R = CH3 commercial1y monobactam and showed a mode of action similar to that of the other p-lactam antibiotics by blocking bacterial cell wall synthesis (136) 5.2 Erythromycin Macrolides Although safe and effective, erythromycin is not a perfect antibacterial The presence of hydroxy groups suitably disposed with respect to the keto function at C-9 leads to the ormation of a tautomeric mixture of hemiketals (138) The 6.9-hemiketal (110) may be dehydrated in stomach acid to give the inactive A, analog (111),which may undergo further ring closure to give the 9,12-tetrahydrofuran (112) that is also inactive (139) The A, derivative f - Erythromycin (109) was isolated, in 1952, from a strain of Saccharopolyspora erythraea Antibiotics (111)may be responsible for some gastrointestinal disturbance (140) To avoid these problems by increasing the stability to acid, the 2'-stearate, estolate, and ethylsuccinate esters have been prepared (141), but even when the tablets are enteric-coated the bioavailability is erratic and relatively frequent dosing is required (137) An understanding of the acid-catalyzed decomposition of erythromycin has led to a variety of semisynthetic derivatives with improved oral bioavailability (142) Reductive amination of the 9-keto function gives erythromycylamine, which reads with (2-methoxyethoxy)acetaldehyde (143) to give dithromycin Beckmann rearrangement of the 9-oximefollowed by reduction and methylation (144) gives azithromycin (1131, which shows good activity against Gram-negative bacteria, including Haemophilus influenzae An alternative for prevention of cyclization between the 9-keto and 6-hydroxy is to mask the 6-hydroxy group If the 6-hydroxy is methylated (145), the result is clarithromycin (114), which like (1131, has an improved pharmacokinetic profile compared with that of the parent molecule Natural Products as Leads for New Pharmaceuticals dised to a ketone (147) The loss of potency that would ensue is compensated by two further modifications, which improve binding, formation of a carbamate at positions 11/12, and extension with a heterocycle-substituted side-chain In ABT 773 a similar side-chain is placed at position 6, with comparable results (147) 5.3 (113) Azithromycin Both azithromycin and clarithromycin have been used for various bacterial infections for a number of years Within the last decade, resistance has emerged to a range of antibacterials, including the macrolides, arising from methylation of an adenine in the 23s ribosomal RNA target site, which prevents binding (146) The invention of the ketolides [e.g., telithromycin (115)l overcomes MLS, resistance by removing the L-cladinose moiety at position 3: the exposed hydroxyl is also oxi- Streptogramins The streptogramins are produced by Streptomyces species and have been classified into two groups: Group A are polyunsaturated macrocyclic lactones and Group B are cyclic hexadepsipeptides Both groups bind bacterial ribosomes and inhibit protein synthesis at the elongation step and they act synergistically against many Gram-positive microorganisms However, the naturally occurring streptogramins are poorly soluble in water and this, until recently, has limited their use to treat bacterial infections New, water-soluble derivatives have been developed and the semisynthetic dalfopristin (116) and quinupristin (117) mixture (Synercid) has been approved for the treatment of Gram-positive infections, including multidrug-resistant strains of Enterococcus faecium, Staphylococcus aureus, and S pneumoniae (148) (1 15) Telithromycin Antibiotics 877 water soluble (150), despite the hydrogenbonding ability of the polyhydroxylated hexapeptide I (116) Dalfopristin 5.4 Echinocandins The fungal metabolite echinocandin B (118)is one of the lipopeptides, in which a cyclic hexapeptide is combined with a long-chain fatty acid Echinocandin B inhibits p-1,3-glucan synthesis and as a result has anti-Candida and anti-Pneumocystis carinii activity (149) As a group, the echinocandins are not orally bioavailable, are haemolytic, and are not very (118) echinocandin B, R = linoleyl Synthesis of the cyclic hexapeptide is unattractive for the purpose of securing analogs with improved biological activity because of the unusual nature of the amino acids used (117) Quinupristin Natural Products as Leads for New Pharmaceuticals and the complex stereochemistry generated by the high degree of hydroxylation However, echinocandin B can be produced efficiently by fermentation of a culture of Aspergillus nidulans and then deacylated by fermentation with Actinoplanes utahensis (151) The free amino group thus exposed can be derivatized with a number of active esters Synthesis of the amide from 4-octylbenzoic acid gives cilofungin (119), which has specifically high potency against Candida albicans and some other Candida species (151) now in clinical trial and has the major advantage of oral bioavailability (153) Many other antifungal peptides are under investigation (152) The member of this series that is furthest advanced is caspofungin (MK-991, L-743,872) (121),following its approval by the FDA, early in 2001, for the treatment of aspergillosis The two analogs, LY-303366 and caspofungin, have been compared against clinical fungal isolates i n vitro (154) and the latter has been evaluated in immunosuppressed mice (155) (119) cilofungin, R - 6.1 For systemic use cilofungin had to be given intravenously and unfortunately ran into problems associated with the cosolvent (PEG) (152) A better derivative, LY-303366 (120) is CARDIOVASCULAR DRUGS Lovastatin, Simvastatin, and Pravastatin One of the most significant natural product discoveries in the last 25 years has been a fungal secondary metabolite called lovastatin (122) Heralded as a major breakthrough in OH (121) caspofungin Cardiovascular Drugs the treatment of coronary heart disease (1561, lovastatin was introduced onto the market by Merck in 1987 for the treatment of hypercholesterolemia, a condition marked by elevated levels of cholesterol in the blood Lovastatin works by inhibiting 3-hydroxy3-methylglutaryl coenzyme A (HMG-CoA) reductase, a key rate-limitingenzyme in the cholesterol biosynthetic pathway However, the first specific inhibitors of this enzyme were discovered several years earlier by Endo et al at Sankyo (157) The compounds, which are structurally related to lovastatin, were isolated from Penicillium citrinum and shown to block cholesterol synthesis in rats and lower cholesterol levels in the blood Development of the most active compound, designated ML-236B (123),is believed to have been curtailed because of toxicity problems (158) Brown et al a t Beechams also reported the isolation of (1231, but as a metabolite from Penicillium brevicompactum (159) The 879 group, naming the compound compactin, reported its antifungal activity but failed to reveal its mode of action as an inhibitor of HMGCoA reductase The search for naturally occurring inhibitors of HMG-CoA reductase gained pace and after spending several years developing appropriate screens, Merck found during only the second week of testing a culture of Aspergillus terreus that displayed interesting inhibitory activity (160) In February 1979 the active component, lovastatin (mevinolin), was isolated and characterized (1611, and in November the following year Merck was granted patent protection in the United States Although lovastatin proved to be identical to monocolin K, a metabolite isolated earlier from Monasus ruber (162), the chemical structure of the latter compound had not been reported, whereas Merck filed for patent protection giving complete structural details for lovastatin The discovery of compactin and lovastatin prompted efforts to develop derivatives with improved biological properties (163, 164) Modification of the methylbutyryl side chain of lovastatin led to a series of new ester derivatives with varying potency and, in particular, introduction of an additional methyl group a to the carbonyl gave a compound with 2.5 times the intrinsic enzyme activity of lovastatin (165) The new derivative, named simvastatin (124), was the second HMG-CoA reductase inhibitor to be marketed by Merck Both lovastatin and simvastatin are prodrugs and are hydrolyzed to their active open-chain dihydroxy acid forms in the liver (166) A third compound, pravastatin (125), launched by Sankyo and Squibb in 1989, is the open hydroxyacid form of compactin that was first identified as a urinary metabolite in dogs Pravastatin is produced by microbial biotransformation of compactin The HMG-CoA reductase inhibitors described above bind to two active sites on the enzyme: the hydroxymethylglutaryl binding domain and an adjacent hydrophobic pocket to which the decalin moiety binds (167).The recognition that the ring-opened hydroxy acids resemble mevalonic acid and that the decalin moiety could be replaced by 4-fluorophenylsubstituted heterocycles led to the launch of several new products including fluvastatin Natural Products as Leads for New Pharmaceuticals (127) Cerivastatin (126), the ill-fated cerivastatin (127), and the so-called turbostatin atorvastatin (128) Although cerivastatin was withdrawn from the market in 2001 because of fatal adverse drugdrug interactions, the "statins" remain one of the fastest growing segments of the pharmaceutical industry The latest member of this group of cholesterol-lowering drugs, Astra- (126) Fluvastatin (128) Atorvastatin (129) Rosuvastatin Cardiovascular Drugs Zeneca's rosuvastatin (129), is due to be launched in 2002 and is forecast to achieve sales of US $2.8 billion by 2005 (168) 6.2 881 way for other ACE inhibitors, such as enalapril(132) and lisinopril, which have had a major impact on the treatment of cardiovascular disease (173) Teprotide and Captopril While studying the physiological effects of snake poisoning, Ferreira (169) discovered that specific components in the venom of the pit viper Bothropsjararaca inhibited degradation of the peptide bradykinin and potentiated its hypotensive action The "potentiating factors" proved to be a family of peptides that worked by inhibiting the dipeptidyl carboxypeptidase, angiotensin-converting enzyme (ACE) (170,171) In addition to catalyzing the degradation of bradykinin, ACE also catalyzes the conversion of human prohormone, angiotensin 1, to the potent vasoconstrictor odapeptide, angiotensin 11.However, the significance ofACE in the pathogenesis of hypertension was not fully appreciated until the 1970s after Ondetti et al (172) had first isolated and then synthesized the naturally occurring nonapeptide, teprotide (130) The compound proved to be a specific potent inhibitor of ACE and showed excellent antihypertensive properties in clinical trials, although its use was limited by the lack of oral activity 6.3 Adrenaline, Propranolol, and Atenolol The true clinical potential of P-adrenoceptor blocking agents for treating angina, atrial fibrillation, and tachycardias was first recognized by James Black and colleagues at ICI (174).Black noted a report from Neil Moran of Emory University in 1958, showing that dichloroisoprenaline antagonized the effects of Pyr -Trp-Pro-Arg -Pro-Gln-Ile-Pro-Pro adrenaline on heart rate and muscle tension The first effective P-adrenoceptor blocker, pronethalol (133), was synthesized years later by the ICI group and marketed for limThe discovery of teprotide led to a search ited use in 1963 Toxicity problems soon led for new, specific, orally active ACE inhibitors pronethalol to be replaced by the 1-naphthyl Ondetti et al (172) proposed a hypothetical analog, propranolol (134), which became the model of the active site of ACE, based on analfirst P-adrenoceptor antagonist approved for ogy with pancreatic carboxypeptidase A, and general use, being more potent and yet devoid used it to predict and design compounds that of the partial agonist or intrinsic sympathomiwould occupy the carboxy-terminal binding metic activity shown by many other analogs site of the enzyme Carboxyalkanoyl and merCompounds with improved selectivity for the captoalkanoyl derivatives of proline were P-adrenoceptor of cardiac muscle (P-l-adrenofound to act as potent, specific inhibitors of ACE and 2-~-methyl-3-rnercaptopropanoyl-~proline (131) (captopril) was developed and launched in 1981 as an orally active treatment for patients with severe or advanced hypertension Captopril, modeled on the biologically active peptides found in the venom of the pit viper, made an important contribution to the (133) Pronethalol understanding of hypertension and paved the 882 Natural Products as Leads for New Pharmaceuticals ceptor blockers) were to follow, including atenolol (135), which became the most frequently prescribed P-blocker and one of the best-selling drugs of the time (134) Propranolol olite of coumarin (137), itself a common component of Melilotus sp Soon after the compound had been identified, trials were initiated that confirmed the oral anticoagulant activity in humans and in 1942 it was marketed under the name dicoumarol (177) The compound had a slow, erratic onset of action and efforts were initiated to prepare synthetic analogs that acted faster and had longer duration of action A 4-hydroqcoumarin residue, substituted at the 3-position, proved essential for biological activity and in 1948, after synthesizing over 150 compounds, a Chydroxycoumarin derivative that was longer acting and more potent than dicoumarol was selected not for clinical use, but as a rodenticide for development by the Wisconsin Alumni Research Foundation! The compound (138), (135) Atenolol 6.4 Dicoumarol and Warfarin Sweet clover has a long history of medicinal use, often as an antiflammatory or analgesic preparation in the form of ointments and poultices Melilotus officinalis (yellow sweet clover, or ribbed melilot) was reputed to have been a favorite herbal treatment used by King Henry VIII of England and the plant is still referred to as King's Clover in some publications (175) The plant flourishes in poor soil &d was cultivated extensively in Europe for cattle fodder and for soil improvement In the early 1920s M officinalis was planted on the prairies of North Dakota and Alberta, Canada, but with disastrous consequences Soon cattle and sheep throughout these regions began literally bleeding to death The mysterious hemorrhagic disease was traced to clover fodder that had not been stored properly and had become "spoiled," or moldy However, the insolubility of the anticoagulant component and the difficulty of assaying extracts for biological activity made the task of isolating the active principal component intractable (176) It took almost 20 years before the compound was identified as 3,3'-methylenebis(4-hydroxycoumarin) (136), an oxidative degradation metab- named warfarin (an acronym derived from the name of the institute coupled with "arin" from coumarin), became a household name for rat poison Concern over the use of oral anticoagulants and the inherent risk of hemorrhage inhibited the development of warfarin as a therapeutic agent However, in 1951, a Antiasthma Drugs U.S.Army cadet unsuccessfully attempted to commit suicide by taking massive doses of the compound The incident prompted further clinical trials that resulted in warfarin being used as the anticoagulant of choice for prevention of thromboembolic disease (177) The mode of action of the coumarin anticoagulants involves blocking the regeneration of reduced vitamin K and induces a state of functional vitamin K deficiency, thus interfering with the blood-clotting mechanism (178) ANTIASTHMA DRUGS 7.1 Khellin and Sodium Cromoglycate The toothpick plant, Ammi visnaga, had been used for centuries in Egypt as an antispasmodic agent to treat renal colic and ureteral spasm In 1879 one of the plant's main constituents was isolated, crystallized, and named khellin (139) (179) Subsequently, the pure compound was shown to relax smooth muscle and in 1938 the chemical structure was characterized as a chromone derivative (180) In 1945 a medical technician took khellin to treat renal colic and found instead that it acted as a potent coronary vasodilator and relieved his angina (181) This chance discovery, together with earlier observations, led to khellin being used as a coronary artery vasodilator and for treating bronchial asthma (182) However, its clinical use was severely limited by some unpleasant gastrointestinal side effects Five years later, a small British pharmaceutical company, called Benger Laboratories, initiated a program to synthesize khellin analogs as potential bronchodilators for treating asthma, and had prepared a series of compounds that relaxed guinea pig bronchial smooth muscle and protected the animals against allergen-induced bronchospasm (183) A clinical pharmacologist on Benger's staff, who suffered from chronic asthma, questioned the validity of the animal model and decided instead to test the compounds on himself He then prepared a "soup" of guinea pig fur, inhaled the vapors to induce a reproducible asthma attack, and assessed the effects of the synthesized khellin derivatives Many of the compounds first prepared were insoluble in water and caused nausea and other unpleasant side effects when taken orally This led to the test compounds being formulated as aerosol sprays and in 1958, an aerosol preparation of a chromone-2-carboxylic acid derivative (140) was found to exert a protectant effect, albeit short lived, against bronchial allergen challenge without showing the bronchodilator activity seen with other compounds The compound was completely inactive in the guinea pig asthma model and afforded its protectant effect in humans only when inhaled as an aerosol About two new compounds were tested each week and in 1965, after synthesizing some 670 analogs, a bischromone was prepared that gave good protection, even when inhaled up to h before bronchial allergen challenge (184) The compound sodium cromoglycate (141) was obtained by condensing diethyl oxalate with the bis(hydroxy acetophenone) (142) and cyclizing the resultant bis(2,4-dioxobutyric acid) ester (143) under acidic conditions (185) The essential chemical features required for activity appeared to be the coplanarity of the chromone nuclei, the flexible dioxyalkyl link, and the carboxyl groups in the 2-positions It is believed to act by stabilizing tissue mast cells against degranulation, thereby preventing release of inflammatory mediators (186) Natural Products as Leads for New Pharmaceuticals Sodium cromoglycate entered clinical trials in 1967 and emerged to become a first-line prophylactic treatment for bronchial asthma The coronary dilator properties of khellin have not been ignored and at least one successful program was initiated to prepare analogs for testing as potential antiangina drugs (187, 188) Benziodarone (144) was the first useful compound to emerge from the Labaz laboratories in Belgium based on the benzofuran ring system However, the compound caused hepatotoxicity in man and was soon superseded by amiodarone (145), a more potent coronary dilator for treating angina In 1970 the first report of antiarrhythmic activity in the clinic was published (189) and amiodarone became established for prophylactic (144) benziodarone control of supraventricular and ventricular arrhythmias during the 1980s (188) 7.2 Ephedrine, Isoprenaline, and Salbutamol The Chinese have been using a plant extract known as ma huang to treat asthma and hay Antiasthma Drugs (145) amiodarone fever for thousands of years The extract is prepared from several species of Ephedra, a small leafless shrub found in China Following experiments at the Peking Union Medical College and then at the University of Pennsylvania and the Mayo Clinic in the United States, the active ingredient, ephedrine (1461, was introduced into Western medicine in 1926 as an orally active bronchodilator for the treatment of acute asthma (190,191) Ephedrine is related to another natural product that has been used to treat asthma, that is, the adrenal hormone adrenaline (147) (epinephrine) Adrenaline is a potent agonist of both a-and P-adrenoceptors and thus produces arterial hypertension as an undesirable side effect In 1951 a synthetic alternative, isoprenaline (148), was introduced and for almost 20 years it was considered the drug of choice for treating bronchospasm associated with acute asthmatic attack (191) Isoprenaline is a specific P-adrenoceptor agonist and, although it has no vasoconstrictor activity, the compound does have marked cardiac stimulant properties and a short duration of action Ahlquist's concept (192) of two types of adrenoceptor was developed further by Lands et al (193),who established the existence of PI- and P,-adrenoceptor subtypes Clear structure-activity relationships emerged with the preparation of compounds related to adrenaline and ephedrine; the basic requirement for p-adrenoceptor agonist activity was an aromatic ring substituted by an ethanolamine side-chain The branched methyl substituent on the sidechain was associated with prolonged duration of action (i.e., ephedrine), whereas aromatic hydroxylation (in isoprenaline) prevented penetration across the blood-brain barrier and thus prevented stimulation of the CNS (191) However, 1,2-dihydroxy substituents were found to promote enzymic degradation, and replacement of the 3-hydroxy group by a hydroxymethyl substituent was required to extend the duration of action In 1969 salbutarn01 (149) was launched by Glaxo as a longer-lasting, selective &-adrenoceptor agonist for the treatment of bronchial asthma (194) and, recently, a lipophilic ether analog, salmeterol(150), was introduced with an even longer duration of action that has potential advantage in the prevention of nocturnal asthma Despite the many chemical alterations that have been carried out on the phenylethanolamine "template," the key chemical features associated with modern P-agonists can be seen Natural Products as Leads for New Pharmaceuticals 886 8.1 to have originated from the naturally occurring compounds, adrenaline and ephedrine 7.3 Contignasterol The use of inhaled corticosteroids such as fluticasone propionate to treat asthma and rhinitis has been well documented and will not be repeated here Less well known is an unusual, highly oxygenated marine-derived steroid isolated from the sponge Petrosia contignata that possesses a unique cyclic hemiacyl sidechain (151) The compound was isolated by Andersen and coworkers (195) at the University of British Columbia and found to possess anti-inflammatory properties in vivo Contignasterol is being developed by Inflazyme, in collaboration with Aventis, for the treatment of asthma and other inflammatory diseases and has progressed to phase I1 clinical trials ANTIPARASITIC DRUGS Artemisinin, Artemether, and Arteether Artemisia annua (sweet wormwood, qing ha01 has been used in Chinese medicine for well over 1000 years The earliest recommendation is for the treatment of hemorrhoids, but there is a written record of use in fevers dated 340 A.D Modern development dates from the isolation of a highly active antimalarial, artemisinin (qinghaosu), in 1972, and has been carried out almost entirely in China Much of the original literature is therefore in Chinese, but there is an excellent review on qinghaosu by Trigg (196) and an account of the uses ofA annua (197).This section is largely a summary of these two articles Artemisinin (152) is a sesquiterpene lactone with an unusual peroxide bridge One of the earliest modifications involved catalytic reduction of the peroxide, resulting in loss of one oxygen and total loss of antimalarial activity (196) in the adduct (153) The role of the peroxide bridge in producing antimalarial effects was not fully understood, but it appeared essential for activity, so much of the early work on analogs conserved this structural feature as an empirical finding The mechanism (152) artemisinin HO"' (151) Contignasterol (153) Antiparasitic Drugs of action of artemisinin has since been elucidated (198, 199), although it is not without controversy (200, 201) The drug has a high affinity for hemozoin, a storage form of hemin that is retained by the parasite after digestion of hemoglobin, leading to a highly selective accumulation of the drug by the parasite Artemisinin then decomposes in the presence of iron, probably from the hemozoin, and releases free radicals, which kill the parasite The peroxide bridge is therefore a crucial part of the drug molecule, as was suspected from structure-activity studies Elucidation of the mechanism of action has led to the synthesis of a range of simple analogs capable of iron-catalyzed decomposition, some of which have good antimalarial activity (202) In retrospect, it is not surprising that the peroxide-bridged compound (154), isolated from Artabotrys uncinatus, also has antimalarial activity (197) Because peroxides of this kind are likely to be formed from a variety of precursors in dried plant material (see below), there may well be many more antimalarials of this kind to be found Artemisinin is an excellent antimalarial, approximately equal in potency to chloroquine, with a good therapeutic index except on the fetus The preparation of semisynthetic derivatives has been stimulated primarily by a requirement for improved solubility because artemisinin is relatively insoluble in both water and oil Reduction of (152) with sodium borohydride occurs at the lactone carbonyl, leaving the peroxide intact (196, 197) The resulting cyclic hemiacetal, dihydroartemisinin (155), which is a more potent antimalarial than the parent compound, shows typical acetal reac- 887 tivity In the presence of acid, a highly reactive carbocation intermediate allows S,1-type substitution with a variety of nucleophiles For example, boron trifluoride catalyzes reactions with methanol and ethanol to give artemether (156) and arteether (1571, respectively, two of the most important derivatives (196) Both are more potent than the parent compound and have improved solubility in oil Artemether has been chosen for development in the West under the name Paluther (155) (156) (157) (158) R=H R = CH3 artemether R = CH2CH3arteether R = COCH2CH2COONasodium artesunate Water solubility can be greatly improved by the standard ploy of esterification with succinic acid and conversion to the sodium salt Applied to compound (155), this technique gives sodium artesunate (158),a water-soluble prodrug that may be given intravenously (196) It may be assumed that hydrolysis occurs in vivo to give back (155) as the active antimalarial because (156) has been shown to be unstable in aqueous solution and because analogous carboxylic acids with a nonhydrolyzable ether link are relatively inactive There are two reasons for the great interest being shown in artemisinin and its derivatives First, there is little cross resistance with Plasmodium falciparum between the members of this series and the quinoline-based antimalarials like chloroquine (203) On the contrary, significant potentiation of effect is observed in combination with chloroquine analogs such as mefloquine (204) Second, the high lipid solubility of, for example, artemether ensures rapid penetration into the CNS, so these sesquiterpene lactones are first- Natural Products as Leads for New Pharmaceuticals line drugs for the treatment of cerebral malaria caused by P falciparum (197), which is otherwise fatal It seems highly likely (205)that most of the artemisinin found in dried plant material is formed by autoxidation after the death of the plant From the medicinal chemist's point of view this is unimportant, but some plant biochemists might have doubts about the description of artemisinin as a "natural product." In our view, air drying in sunlight is a natural, although not a botanical, process It is probable that many other plant-derived peroxides are formed in a similar way Whole plant extracts often show promising activity that may not be traceable to single components This is obviously not true of Artemisia annua extracts, but it is interesting to note that other constituents, notably methoxylated flavones, have potentiating effects on the antimalarial activity of artemisinin (206) The reported effect of artemisinin on systemic lupus erythematosus (196) is intriguing, given the history of use of quinine-type antimalarials in this disease 8.2 Quinine, Chloroquine, and Mefloquine The use of Cinchona bark (e.g., Cinchona succirubra) by South American indians to treat fevers and the subsequent importation of the bark into Europe by Jesuit priests in the 17th century is well known (207) At that time malaria was widespread, even as far north as eastern Scotland, and there was no effective treatment for "the ague." Although quinine (159) is not very potent or long acting, a good sample of Cinchona bark contains about 5% of the alkaloid (208) This high concentration permitted genuinely therapeutic doses of bark to be given and allowed the pure alkaloid to be isolated (209) as early as 1820 During the next 100 years quinine was the only effective treatment for malaria known to Europeans Without quinine, life in the tropics was impossible for those without natural immunity to malaria "One thing that was compulsory was the taking of five grains of quinine a day And if you didn't take it and got ill your salary was liable to be stopped" (210) Supplies of quinine to Europe were threatened during World War I, stimulating a major program of research into synthetic analogs (159) quinine The chemical techniques available to chemists in the period 1820-1920, although improving rapidly, did not allow a structure to be proposed for quinine with any confidence: the first completely correct proposal (211) came in 1922 and was finally confirmed by total synthesis (212) as late as 1945 However, part structures were known, such as the 6-methoxyquinoline moiety, from long before, and were sufficient to allow the synthesis of mimics The first clinically successful mimics were the 8-aminoquinolines In the early years of the 20th century, synthetic organic chemistry was a young disdpline, largely governed by empirical rules Progress toward synthetic analogs of complex natural structures was governed as much by synthetic feasibility as by a desire for close mimicry The first quinine analogs were, therefore, a combination of the accessible 6-methoxyquinoline part of the quinine structure, with elements of the first successful antimicrobial agents, such as 9-aminoacridine Nitration followed by reduction could be used to generate a number of new molecules from a variety of parent heterocycles It is recorded (213) that 4-, 6-, and 8-aminoquinolines have antimalarial properties and, quite extraordinarily, two of these chemical classes are still used today, have quite different uses as antimalarials, and quite possibly have different modes of action to be inThe first of the 8-aminoauinolines traduced into medicine was pamaquine (1601, not long after World War I (214) Despite greater toxicity than that of quinine, this class - Antiparasitic Drugs of drugs was found to have radical curative ability against the relapsing malarias Several hundred analogs were tested during World War I1 and of these, primaquine (161) survives to the present day for short-term use as a radical curative (215) (160) pamaquine As has been explained, the major stimulus for research into synthetic antimalarials was not so much the therapeutic inadequacy of quinine as the potential lack of availability in times of social upheaval During World War 11, the United States encouraged the planting of Cinchona in Costa Rica, Peru, and Ecuador (216) The total synthesis of quinine was too difficult in the 1940s and is unlikely to become economically viable even in the new millennium This problem was partly overcome with quinacrine, which was used widely in World War 11, although quinacrine has the defects described above The conceptual derivation of chloroquine (163) from quinacrine is obvious and apparently happened twice, in Germany and the United States, the latter about 10 years after the Germans had discarded the drug as being too toxic! The story of the rediscovery of chloroquine is fascinating, as an account of human muddle and misjudgment, finally leading to an extraordinarily valuable drug (216) (161) primaquine Quinacrine (162) is an obvious embodiment of the principle outlined above; as a derivative of both quinine and 9-aminoacridine it combined a known antimalarial with a known antimicrobial The result was a useful, relatively nontoxic antimalarial, although it stained the skin and eyeballs yellow (216) Despite this side effect and a high incidence of gastrointestinal disturbance, quinacrine was widely used during World War I1 by European troops in East Asia The availability of the results of medicinal chemistry research to both sides in wartime is a curious feature of antimalarial development, highlighted below (162) quinacrine (mepacrine) (163) chloroquine Over decades of sublethal exposure the resistance of all types of malaria has increased to a point where chloroquine no longer offers certain protection (217) With the partial exception of quinine and dihydroquinine (218), resistance to antimalarials had reached the stage at the time of the Vietnam war where more research was required The development of mefloquine (164) was a continuation of the World War I1 effort, with a gap of about 20 years Resistance to chloroquine had developed widely during that period, but surprisingly less so to quinine, given the obvious similarities in structure This observation stimulated a reappraisal of quinolines, known as quinoline methanols, which bear a hydroxy group on the a-carbon of a substituent at- Natural Products as Leads for New Pharmaceuticals tached to the 4-position (219) Up to 1944, a total of 177 quinoline methanols had been synthesized and tested, resulting in one compound (165) with activity superior to that of quinine In human volunteers there was a high incidence of phototoxicity associated with (165), so research on quinoline methanols in 1944 had ceased in favor of the 4-amino series, which included chloroquine Reappraisal of about 100 of the World War I1 compounds confirmed the high activity and phototoxicity of (165) and also showed the high potency of an analog (1661, which had reduced phototoxicity (219) These data, together with results from about 200 newer compounds, fostered the belief that phototoxicity was separable from antimalarial activity Extensive evaluation of (166) in humans with chloroquine-resistant Plasmodium falciparum infections showed promise, but with a significant incidence of toxic reactions; the dose required was also inconveniently large Two hypotheses concerning the effect of the 2-phenyl substituent were proposed One was that metabolic oxidation was blocked at - (164) mefloquine this position, so that duration of action was prolonged, which was considered desirable Second, the W chromophore was enlarged, which would increase the likelihood of druginduced photosensitivity The phenyl substituent was thus replaced by trifluoromethyl in the 2-position (220) Before the first such derivatives were tested, further analogs were prepared with an additional trifluoromethyl group on the benzene ring This was serendipitous because the first series of 2-trifluoromethyl analogs had low potency and were also photosensitizing The series with two trifluoromethyl groups, one at position and another in the 6-, 7-, or 8-position were all potent and free from phototoxicity (221) The most potent was mefloquine (164), a very successful drug but one that produces unacceptable CNS effects in a small proportion of users (222); parasite resistance has also been observed in parts of Southeast Asia (217) There is now a serious attempt by the World Health Organization to find new antimalarials Physicians are pragmatic when choosing therapy for patients whose suffering is not alleviated by accepted methods A drug that has been shown to be toxicologically safe may be utilized in a new area for the flimsiest of reasons Thus Page (223) described his use of quinacrine in two cases of lupus erythematosus as being based on "[a] chance observation ," although he did not describe the observation that led to his decision He did, however, record that quinine had been tried previously and "prevented extension of the lesions," so this may have been the basis for his rationale In any event, the beneficial effects of quinacrine were remarkable and ap- Conclusion peared to be related to the degree of yellowing of the skin that, as described earlier, is a common side effect of the use of quinacrine in malaria Among Page's group of patients with lupus erythematosus were two with rheumatoid arthritis, whose symptoms also responded to treatment with quinacrine The following year, other physicians (224) conducted a trial of quinacrine on a larger group of patients with rheumatoid arthritis; the results encouraged Haydu (225) to test chloroquine on similar patients, again with positive results A year later, two more physicians (226) compared quinacrine with chloroquine and found the latter to be better tolerated, the majority of patients gaining some benefit Both quinawine and chloroquine caused gastrointestinal disturbances, which led to a trial (227) of hydroxychloroquine (167), an unsuccessful antimalarial but with less effect on the gut, thus allowing larger doses to be given Hydroxychloroquine has remained part of the standard drug therapy for rheumatoid arthritis ever since (167) hydroxychloroquine So far, the choice of quinine-like drugs to treat rheumatoid arthritis has been based on preliminary selection as antimalarials Because the two types of action are presumably unconnected, there might be some value in a screening program aimed directly at rheumatoid disease 8.3 Avermectins and Milbemycins There is no major distinction between the avermectins and milbemycins, which are based on the same complex polyketide macrocycle (168): the avermectins are oxygenated at (2-13 and bear a disaccharide on this oxygen They have been isolated from cultures of a 891 number of Streptomyces species, obtained from all over the world (228) The avermectins, particularly, have been the subject of intense commercial interest because they possess potent activity against both nematode and arthropod parasites of livestock (229) A full discussion of structure-activity relationships would be out of place here, not least because the data are voluminous so we shall concentrate on the development of ivermectin, which has been a major success Structural designation of avermectins is quaintly based on three series: A, B; a, b; and 1, These are illustrated diagrammatically Greater activity resides in the B series, with a free OH at position There is little difference in potency between the a and b series In the more potent B series there are important differences between the series and the series; Bl is the more active orally, whereas B, is the more potent by injection There are also differences in their spectrum of activity (230) The spectrum of activity was kept as broad as possible by hydrogenation of a mixture of avermectins Bla and Blb to give ivermectin (169), which contains at least 80% of 22,23-dihydroavermectin Bla and not more than 20% of 22,23-dihydroavermectin Blb Ivermectin was developed for, and has been highly successful in, the treatment and control of parasites in cattle, horses, sheep, pigs, and dogs Following studies in humans with river blindness (onchocerciasis) (231-233), the developers of ivermectin (Mectizan) have participated in a major program aimed at eradication of the disease The sufferers inhabit some of the poorest parts of Africa and cannot pay for their treatment, so the drug has been donated by Merck and Co Since 1996 more than 20 million treatments have been given (234) The drug does not kill the adult worms that cause onchocerciasis (2351, but is useful in interrupting the life cycle (236) Ivermectin is also of value in treatment of scabies (237) A great deal of information on the biological aspects of the use of ivermectin has recently been summarized (238) CONCLUSION Natural product research has been the single most successful strategy for discovering new Natural Products as Leads for New Pharmaceuticals milbemycins R = H In the avermectins the series are designated a s follows (Y = CH3): A, Z = CH3 B,Z=H a, X = CH(CH3)CH2CH3 b, X = CH(CH3)2 1, V-W =CH=CH 2, V-W = CH2CH(OH) For further details of these descriptors, in the milbemycins, see Ref 228 I n ivermectin (169),V-W = CH2CH2,X = CH(CH3)CH2CH3(major) or CH(CH3)2 (minor), Y = CH3 and Z = H pharmaceuticals and has contributed dramatically to extending human life and improving clinical practice As long as Nature continues to yield novel, diverse chemical entities possessing selective biological activities, natural products will play an important role as leads for new pharmaceuticals An interesting recent example is the alkaloid galantamine (Ni- valin, Reminyl) (170),originally isolated from the bulbs of the Arnaryllidaceae family (snowdrops, daffodils, etc.), which has found use in the symptomatic treatment of Alzheimer's Disease (239) It is a reversible and competitive inhibitor of acetylcholinesterase that also interacts allosterically with nicotinic acetylcholine receptors to potentiate the action of Conclusion ' J H3C =063 H3C OH (169) X = CH(CH3)CH2CH3 (major)or CH(CH& (minor) agonists By acting to enhance the reduced central cholinergic function associated with this disease, significant improvements in cognition and behavioral symptoms have been observed in patients In this case it is the alkaloid itself that is used as the active compound and it will be interesting to see whether development leads to better drugs There are as yet relatively few publications in this area, although Sanochemia is interested (240,241) Over 90% of bacterial, fungal, and plant species are still waiting to be investigated (242) High throughput screening methods will allow even greater numbers of samples to be tested against more biological targets (243, 244, although this approach sometimes produces more data than can be conveniently - integrated into a research program An alternative view is that the elucidation of the biological effects of chosen compounds, in some detail, will yield insight into biological processes that may open avenues for medicinal chemistry research that is not based on pure chance This view is based on the recognition that secondary metabolites have been produced and ruthlessly selected, by evolution, over a long period of time Either way, the medicinal chemist has a wonderful opportunity to continue utilizing the rich chemical diversity offered by nature, as is shown in two recent reviews that explore this topic in some detail (245,246) The best approach for the identification of natural product leads is a matter of debate Some very inventive techniques have been used in the bioassay-guided method; for example, by spraying TLC plates with reactive media that respond by producing a color change in the presence of an active compound An alternative is to use an ethnobotanical or ethnopharmacological technique, whereby the accumulated wisdom of many generations of native plant users may be harnessed in the 894 Natural Products as Leads for New Pharmaceuticals search for better medicines for all These two techniques may be combined, so that the native people describe the uses to which they put the plant and the researchers devise a bioassay that is used to find the active components The problem with any bioassay-guided technique, however, is that the inactive constituents are not identified This represents a considerable waste, given that the plant has had to be collected, preserved, and identified An alternative view is that it is best to extract all the constituents, with a view to screening in whichever way is appropriate, at that time or in the future With modern high-performance liquid chromatography facilities it is possible to reduce a plant to its secondary metabolites, as single compounds, in a few days: the products are then able to be screened in a high throughput manner in an equally short time and the compounds can be reevaluated when new screens become available One thing is certain: the variety of natural product structures, after perhaps 300 million years of natural selection, far exceeds the bounds of human imagination, unlike the typical output from combinatorial chemistry! 12 J W Lewis, Adv Biochem Psychopharmacol., 8, 123 (1974) 13 J Hughes, T W Smith, H W Kosterlitz, L A Fothergill, B A Morgan, and H R Morris, Nature, 258,577-579 (1975) 14 J A H Lord, A A Wakerfield, J Hughes, and H W Kosterlitz, Nature, 267,495-499 (1977) 15 0.Schaumann, Arch Exp Pathol Pharmacol., 196,109-136 (1940) 16 See Fkf.7, pp 209-301 17 B Cox, Curr Rev Pain, 4,448-498 (2000) 18 B Cox and J C Denyer, Expert Opin Ther Pat., 8, 1237-1250 (1998) 19 A G Gilman, T W Rail, A S Nies, and P Taylor, Goodman and Gilman's The Pharmacological Basis of Therapeutics, 8th ed., Pergamon Press, New York, 1990, p 550 20 L Lemberger, Clin Pharmacol Therap., 39, 1-4 (1986) 21 S E Sallan, N E Zinberg, and E Frei, N Engl J Med., 293, 795-797 (1975) 22 R K Razdan, in P Krogsgaard-Larsen, S Brogger Christensen, and H Kofod, Eds., Natural Products and Drug Development, Munksgaard, Copenhagen, 1984, pp 486-499 23 L Lemberger and H Rowe, Clin Pharmacol Ther., 18, 720-726 (1976) 24 T S Herman, L E Einhorn, S E Jones, C Nagy, A B Chester, J C Dean, B Furnas, S D Williams, S A Leigh, R T Dorr, and T E Moon, N Engl J Med., 300,1295 (1979) 25 A Ward and B Holmes, Drugs, 30, 127-144 (1985) 26 W A Devane, F A Dysarz, R M Johnson, L S Melvin, and A C Howlett, Mol Pharmacol., 34,605-613 (1988) 27 W A Devane, L Hanus, A Breuer, R G Pertwee, L A Stevenson, G Griffin,D Gibson, A Mandelbaum, A Etinger, and R Mechoulam, Science, 258, 1946-1949 (1992) 28 N Stella, P Schweitzer, and D Piomelli, Nature, 388, 773-778 (1997) 29 A D Khanolkar and A Makryannis, Life Sci., 65,607-616 (1999) 30 A Szallasi and V Di Marzo, Trends Neurosci., 23,491-497 (2000) 31 S H Burstein, Pharmacol Ther., 82, 87-96 (1999) 32 M G Bock, Drugs of the Future, 16,631-640 (1991) provides a succinct summary 33 R S L Chang,V J Lotti, R L Monaghan, J Birnbaum, E Stapley, M A Goetz, G Al- REFERENCES G M Cragg, D J Newman, and K M Snader, J Nut Prod., 60, 52 (1997) R Gerardy and M H Zenk, Phytochemistry, 32,79-86 (1993) M J Stone and D H Williams, Mol Microbiol., 6,29-34 (1992) R J Bryant, Chem Znd., 146-153 (1988) C E Inturissi, M Schultz, S Shin, J G Umans, L Angel, and E J Simm, Life Sci., 33 (Suppl 11, 773 (1983) W Sneader, Drug Discovery: The Evolution of Modern Medicine, John Wiley & Sons, Inc., New York, 1985, pp 78-80 summarizes the confusion surrounding the early work A F Casy and R T Pariitt, OpioidAnalgesics, Plenum, New York, 1986, p 407 R Grewe and A Mondon, Chem Ber., 81,279 (1948) See Ref 7, p 153 10 K W Bentley and D G Hardy, Proc Chem Soc., 220 (1963) 11 G F Blane, A L A Boura, A E Fitzgerald, and R E Lister, Br J Pharmacol., 30, 11 (1967) References bers-Schonberg, A A Patchett, J M Liesch, D Hensens, and J P Springer, Science, 230,177-179 (1985) 34 P R Dodd, J A Edwardson, and G J Dockray, Regul Pept., 1, 17 (1980) 35 R B Innis and S H Snyder, Proc Natl Acad Sci USA, 77,6917-6921 (1980) 36 D R Hill, N J Campbell, T M Shaw, and G N Woodruff, J Neurosci., 7, 2967-2976 (1987) 37 R A Gregory, Bioorg Chem., 8, 497-511 (1979) 38 J Dunlop, Gen Pharmacol., 31, 519-524 (1998) 39 P S Anderson, R M Freidinger, B E Evans, M G Bock, K E Rittle, R M Dipardo, W L Whitter, D F Veber, R S L Chang, and V J Lotti, Znt Cong Ser Excerpta Med., 766 (Gastrin Cholecystokinin), 235-242 (1987) 40 M A Goetz, M Lopez, R L Monaghan, R S L Chang, V J Lotti, and T B Chen, J Antibiot., 38,1633-1637 (1985) 41 B E Evans, 2.Gastroenterol Verh.,26, 269271 (1991) 42 B E Evans, K E Rittle, M G Bock, R M Dipardo, R M Freidinger, W L Whitter, G F Lundell, D F Veber, and P S Anderson, J Med Chem., 31,2235-2246 (1988) 43 I M McDonald, Expert Opin Therap Pat., 11, 445-462 (2001) I G Marshall and R D Waigh in A L Harvey, Ed., Drugs from Natural Products, Ellis Horwood, Chichester, UK, 1993, pp 131-151 H King, J Chem Soc., 1381 (1935) P Karrer in D Bovet, F Bovet-Nitti, and G B Marini-Bettolo, Eds., Curare and Curare-like Agents, Elsevier, Amsterdam, 1959, pp 125136 P G Waser, Helv Physiol Pharmacol ActaII (Suppl VIII) (1953) W C Bowman, Pharmacology of Neuromuscular Function, J Wright, Bristol, 1990 A J Everett, L A Lowe, and S Wilkinson, J Chem Soc Chem Commun., 1020-1021 (1970) R B Barlow and H R Ing, Br J Pharmacol Chemother., 3,298 (1948) D Bovet, F Bovet-Nitti, S Guarini, V Longo, and R Fusco, Arch Intern Pharmacol Ther., 88, 1-50 (1951) E P Taylor and H 0.J Collier, Nature, 167, 692 (1951) 53 J B Stenlake, R D Waigh, G H Dewar, R Hughes, D J Chapple, and G G Coker, Eur J Med Chem., 16,515-524 (1981) 54 R Hughes in J Norman, Ed., Clinics i n h a e s thesiology, Vol 3, W B Saunders, London, 1985, pp 331-345 55 J E Caldwell, T Heier, J B Kitts, D P Lynam, M R Fahey, and R D Miller, Br J Anaesth., 63, 393-399 (1989) 56 R L Noble, C T Beer, and J H Cutts, Ann N Y.Acad Sci., 882-894 (1958) 57 I S Johnson, H F Wright, and G H Svoboda, J Lab Clin Med., 54,830 (1959) 58 G H Svoboda, Lloydia, 24, 173 (1961) 59 A C Sartorelli and W A Creasey, Annu Rev Pharmacol., 9,51 (1969) 60 K Gerzon, "Dimeric Catharanthus Alkaloids," in J M Cassady and J D Douros, Eds., Anticancer Agents Based on Natural Product Models, Academic Press, New York, 1980, pp 271317 61 J B Sorensen and H H Hansen, Znvestigational New Drugs, 11,103-133 (1993) 62 J Mann, Murder Magic and Medicine, Oxford University Press, Oxford, 1992, pp 213-214 63 M E Wall, M C Wani, C E Cook, K H Palmer, H T McPhail, and G A Sim, J Am Chem SOC.,88,3888 (1966) 64 Y.-H Hsiang, R Hertzberg, S Hecht, and L Lui, J Biol Chem., 260,14873-14878 (1985) 65 M E Wall andM C Wani, "Camptothecin and Analogues" in Human Medicinal Agents from Plants, ACS Symposium Series 534, American Chemical Society, Washington, DC, 1993, pp 149-169 and references therein 66 W D Kingsbury, J C Boehm, D R Jakas, K G Holden, S M Hecht, G Gallagher, M J Caranfa, F L McCabe, L F Faucette, R K Johnson, and R P Hertzberg, J Med Chem., 34,98-107 (1991) 67 S Sawada, S Okajima, R Aiyama, K Nokata, T Furuta, T Yokokura, E Sugino, K Yamaguchi, and T Miyasaka, Chem Pharm Bull., 39,1446-1450 (1991) 68 J Caesar, The Battle for Gaul, Book 6, Section 31, A Wiseman and P Wiseman translators, Chatto and Windus, London, 1980, p 126 69 T Bryan-Brown, Q J Pharm Pharmacol., 5, 205-219 (1932) 70 D G I Kingston, "Tax01and Other Anticancer Agents from Plants" in J D Coombes, Ed., New Drugs from Natural Sources, IBC Technical Services, 1992, pp 101-108 Natural Products as Leads for New Pharmaceuticals 71 M Suffness, "Taxol: From Discovery to Therapeutic Use" in J A Bristol, Ed., Annual Report of Medicinal Chemistry, Vol 28, Academic Press, New York, 1993, pp 305-314, provides a good review of the discovery and development of tax01 and related derivatives 72 M C Wani, H L Taylor, M E Wall, P Coggon, and A T McPhail, J.Am Chem Soc., 93, 2325-2327 (1971) 73 P B Schiff, J Fant, and S B Horwitz, Nature, 277,665-667 (1979) 74 W P McGuire, E K Rowinsky, N B Rosenhein, F C Grunbine, D S Ettinger, D K Armstrong, and R C Donehower, Ann Intern Med., 111,273-279 (1989) 75 M E Wall and M C Wani, Cancer Res., 55, 753-760 (1995) 76 G M Cragg, S A Schepartz, M Suffness, and M R Grever, J Nut Prod., 56, 1657-1668 (1993) 77 D G I Kingston, Pharmacol Ther., 52, 1-34 (1991) 78 L Mangatal, M.-T Adeline, D Guenard, F Gueritte-Voegelein, and P Potier, Tetrahedron, 45,4177-4190 (1989) 79 J N Denis, A E Greene, D Guenard, F Gueritte-Voegelein, L Mangatal, and P Potier, J Am Chem Soc., 110, 5917-5919 (1988) 80 C Palomo, A Arrieta, F Cossio, J M Aizpuma, A Mielgo, and N Aurrekoetxea, Tetrahedron Lett., 31, 6429-6432 (1990) 81 F Gueritte-Voegelein, D Guenard, F Lavelle, M.-T Le Goff, L Mangatal, and P Potier, J Med Chem., 34,992-998 (1991) 82 K C Nicolaou, C Riemer, M A Kerr, D Rideout, and W Wrasidlo, Nature, 364, 464-466 (1993) 83 A Stierle, G Strobel, and D Stierle, Science, 260,214-217 (1993) 84 K C Nicolaou, Z Yang, J J Liu, H Ueno, P G Nantermet, R K Guy, C F Claibome, J Renaud, E A Couladouros, K Paulvannan, and E J Sorensen, Nature, 367, 630-634 (1994) 85 R A Holton, H B Kim, C Somoza, F Liang, R J Biediger, P D Boatman, M Shindo, C C Smith, and S Kim, J Am Chem Soc., 116, 1597-1600 (1994) 86 G Hofle, N Bedorf, K Gerth, and H Reichenbach, Ger Pat DE 91-4138042 (1993);Chem Abstr., 120, 52841 (1993) 87 M R Grever, S A Schepartz, and B A Chabner, Semin Oncol., 19,622-638 (1992) 88 D M Bollag, P A McQueney, J Zhu, Hensens, L Koupal, J Liesch, M Goetz, E Lazarides, and C M Woods, Cancer Res., 55, 2325-2333 (1995) 89 For an excellent review of the "Chemical Biology of Epothilones," see K C Nicolaou, F Roschangar, and V Dionisios, Angew Chem Znt Ed Engl., 37, 2014-2045 (1998) and references therein 90 R M Borzilleri, X Zheng, R J Schmidt, J A Johnson, S.-H Kim, J D DiMarco, C R Fairchild, J Z Gougoutas, F.Y F Lee, B H Long, and G D Vite, J.Am Chem Soc., 122,88908897 (2000) 91 F Y F Lee, R Borzilleri, C R Fairchild, S.-H Kim, B H Long, C Reventos-Suarez, G D Vite, W C Rose, and R A Kramer, Clin Cancer Res., 7, 1429-1437 (2001) 92 H Stahelin and A von Wartburg in E Jucker, Ed., Progress in Drug Research, BirkhauserVerlag, Basel, Vol 33, 1989, pp 169-266 93 H Stahelin and A von Wartburg, Cancer Res., 51, 5-15 (1991) present a shorter and more readable account 94 M G Kelly and J L Hartwell, J.Natl Cancer Znst., 14,967-986 (1954) 95 I W Kaplan, New Orleans Med Surg J., 94, 388 (1942) 96 J L Hartwell and A W Schrecker in L Zechmeister, Ed., Progress in the Chemistry of Organic Natural Products, 1958, pp 83-166 provide a detailed review of the earlier developments and background 97 V Podwyssotzki, Arch Exp Pathol Pharmacol., 13,29 (1880) 98 J L Hartwell and A W Schrecker, J Am Chem Soc., 73,2909-2916 (1951) 99 K S Pankajarnani and T R Seshadri, Proc Ind Acad Sci., 36A, 157 (1952) through Chem Abstr., 48,2702 (1954).See Ref 77 for a wider discussion 100 A Stoll, J Renz, and A von Wartburg, J.Am Chem Soc., 76,3103-3104 (1954) 101 A Stoll, A von Wartburg, E Angliker, and J Renz, J Am Chem Soc., 76, 6413-6414 (1954) 102 A von Wartburg, E Angliker, and J Renz, Helv Chim Acta, 40, 1331-1357 (1957) 103 I Jardine in J M Cassady and J D Douros, Eds., Anticancer Agents Based on Natural Product Models, Academic Press, New York, Vol 16, 1980, pp 319-351 provides a useful review of the middle years References 104 H Emmenegger, H Stahelin, J Rutschmann, J Rertz, and A von Wartburg, Drug Res., 11, 3274333,459-469 (1961) 105 H Stahelin, Planta Med., 22,336-347 (1972) 106 H Stahelin, Med Exp (Basel), 7, 92-102 (1962) 107 M Kulm and A von Wartburg, Helv Chim Acta, 52,948-955 (1969) 108 C KellerJuslen, M Kuhn, A von Wartburg, and H Stahelin, J Med Chem., 14, 936-940 (1971) 109 B H Long and A Minocha, Proc Am Assoc Cancer Res., 24,321 (1983) 110 J McConnell, R E Longley, and F E Koehn, in V P Gullo, Ed., The Discovery of Natural Products with Therapeutic Potential, Butterworth-Heinemann, Boston, p 109 (1994) 111 K L Rinehart, T G Holt, N L Fregeau, J G Stroh, P A Keifer, F Sun, L H Li, and D G Martin, J Org Chem., 55,4512-4515 (1990) 112 Reuters News Service, 11 July (2001) 113 J Adams and P J Elliott, Oncogene, 19,66876692 (2000) 114 G R Pettit, F Gao, D Sengupta, J C Coll, C L Herald, D L Doubek, J M Schmidt, J R Camp, J J Rudloe, and R A Nieman, Tetrahedron, 47,36013610 (1991) 115 M Jaspars, Chem Znd., 51-55 (1999) 116 E J Corey, D Y Gin, and R Kania, J Am Chem Soc., 118,9202-9203 (1996) 117 Y Hamada, K Hayashi, and T Shiori, Tetrahedron Lett., 32,931-934 (1991) 118 A Fleming, Br J Exp Med., 10, 226-236 (1929) 119 E Chain, H W Florey, A D Gardner, N G Heatley, M A Jennings, J Orr-Ewing, and A G Sanders, Lancet, 2,226-228 (1940) 120 Ref 6, pp 298-315, provides a good review of the discovery and development of penicillin antibiotics 121 Ref 19, pp 1065-1085, summarizes the pharmacological properties of the more important commercial penicillins 122 J C Sheehan, "Molecular Modification in Drug Design," in Advances in Chemistry Series, No 45, American Chemical Society, Washington, DC, 1964, pp 15-24 123 T F Howarth, A G Brown, and T J King, J Chem Soc Chem Commun., 266-267 (1976) 124 C Reading and P Hepburn, Biochem J.,179, 67-76 (1979) 125 A P Ball, A M Geddes, P G Davey, I D Farrell, and G R Brookes, Lancet, 1,620-623 (1980) 126 Ref 6, pp 321-324, provides an interesting account of this discovery 127 See Ref 6, pp 324-300 128 G G F Newton and E P Abraham, Nature, 175,548 (1955) 129 H J Smith, "Design of Antimicrobial Chemotherapeutic Agents," in Smith and William's Introduction to the Principles of Drug Design, 2nd ed., Wright, London, 1988, pp 285-288 130 E H Flynn, Ed., Cephalosporins and Penicillins Chemistry and Biology, Academic Press, New York, 1972 131 G Albers-Schonberg et al., J Am Chem Soc., 100,6491-6499 (1978) 132 W J Leanza, K J Wildonger, T W Miller, and B G Christensen, J Med Chem., 22, 1435-1436 (1979) 133 A Imada, K Kitano, K Kintaka, M Muroi, and M Asai, Nature, 289,590-591 (1981) 134 R B Sykes, C M Cimarusti, D P Bonner, K Bush, D M Floyd, N H Georgopapadakou, W H Koster, W C Liu, W L Parker, P A Principe, M L Rathnum, W A Slusarchyk, W H Trejo, and J S Wells, Nature, 291,489491 (1981) 135 R B Sykes, D P Bonner, K Bush, N H Georgopapadakou, and J S Wells, J Antimicrob Chemother., (Suppl E), 1-16 (1981) 136 R B Sykes and D P Bonner, "Monobactam Antibiotics: History and Development," in J D Williams and P Woods, Eds., Aztreonam, The Antibiotic Discovery for Gram-negative Znfections, Royal Society Medicine International Congress Symposium Series No 89, Royal Society Medicine, London, 1985, pp 3-24 137 N Bahal and M C Nahata, Ann Pharmacother., , (1992) 138 J Barber, J I Gyi, G A Morris, D A Pye, and J K Sutherland, J Chem Soc Chem Commun., 1040-1041 (1990) 139 P Kurath, P H Jones, R S Egan, and T J Perun, Experientia, 27,362 (1971) 140 S Omura, K Tsuzuki, T Sunazuka, S Marui, H Toyoda, N Inatomi, and Z Itoh, J Med Chem., 30,1941-1943 (1987) 141 L D Bechtol, V C Stephens, C T Pugh, M B Perkal, and P A Coletta, C u r Ther Res., 20, 610 (1976) 142 H A Kirst and G D Sides, Antimicrob Agents Chemother., 33, 1413-1418 (19891, provide a useful, brief review Natural Products as Leads for New Pharmaceuticals 143 P Luger and R Maier, J C ~ s tMol Struct., 9, 329 (1979) 144 G M Bright, A A Nagel, J Bordner, K A Desai, J N Dibrino, J Nowakowska, L Vincent, R M Watrous, F C Sciavolino, A R English, J A Retsema, M R Anderson, L A Brennan, R J Borovoy, C R Cimochowski, J A Faiella, A E Girard, D Girard, C Herbert, M Manousos, and R Mason, J Antibiot., 41,1029-1047 (1988) 145 S Moromoto, Y Takahashi, Y Watanabe, and S Omura, J Antibiot., 37, 187-189 (1984) 146 For an overview, see R Leclercq, J Antimicrob Chemother., 48,9-23 (2001) 147 S Douthwaite and W S Champney, J Antimicrob Chemother., , l - (2001) 148 G Bonfiglio and P M Furneri, Expert Opin Invest Drugs, 10,185-198 (2001) 149 J S Tkacz i n J Sutcliffe and N H Georgopapadakou, Eds., Emerging Targets in Antibacterial and Antifungal Chemotherapy, Chapman and Hall, New York, 1992, pp 504-508 150 J M Balkovec, R M Black, M L Hammond, J.V Heck, R A Zambias, G Abruzzo, K Bartizal, H Kropp, C Trainor, R E Schwartz, D C McFadden, K H Nollstadt, L A Pittarelli, M A Powles, and D M Schatz, J Med Chem., 35,194-198 (1992) 151 R Gordee and M Debono, Drugs of the Future, 14,939 (1989) 152 A J De Lucca, Expert Opin Invest Drugs, 9, 273-299 (2000) 153 S Y Ablordeppey, P Fan, J H Ablordeppey, and L Mardenborough, Curr Med Chem., 6, 1151-1195 (1999) 154 M A Pfaller, F Marco, S A Messer, and R N Jones, Diagn Microbiol Infect Dis., 30, 251255 (1998) 155 G K Abruzzo, C J Gill, A M Flattery, L Kong, C Leighton, J G Smith,V B Pikounis, K Bartizal, and H Rosen, Antimicrob Agents Chemother., 44,2310-2318 (2000) 156 E E Slater and J S McDonald, Drugs, Suppl 3,72-82 (1988) 157 A Endo, M Kuroda, and Y Tsujita, J Antibiot., 29, 1346-1348 (1976) 158 D J Gordon and B M Riffind,Ann Int Med., 107,759-761 (1987) 159 A G Brown, T C Smale, T J King, R Hasenkamp, and R H Thompson, J Chem Soc Perkin Trans 1, 1165-1170 (1976) 160 P R.Vagelos, Science, 252,1080-1084 (1991), gives a brief, chronological account of the discovery o f lovastatin 161 A.W Alberts et al., Proc Natl Acad Sci USA, 77,39573961 (1980) 162 A Endo, J Antibiot., 32,852-854 (1979) 163 A W Alberts, Am J Cardiol., 62, 105-155 (1988),and references therein 164 S M Grundy, "HMG Co A Reductase Inhibitors: Clinical Applications and Therapeutic Potential," in B M Rifkind, Ed., Drug Treatment of Hyperlipidemia, Marcel Dekker, New York, 1991, pp 139-167 165 W F Hoffrnann, A W Alberts, P S Anderson, J S Chen, R L Smith, and A K Willard, J Med Chem., 29,849-852 (1986) 166 E E Slater and J S MacDonald, Drugs, 36 (Suppl 3), 72-82 (1988) 167 C E Nakamura and R H Abeles, Biochemist ~24,1364-1376 , (1985) 168 Reuters News Service, 24 Sept (2001) 169 S H Ferreira, Br J Pharmacol., 24,163-169 (1965) 170 S H Ferreira, L J Greene, V A Alabaster, Y S Bakhle, and J R.Vane, Nature, 225,379380 (1970) 171 S H Ferreira, D C Bartelt, and L J Greene, Biochemistry, 9,2583-2593 (1970) 172 M A Ondetti, B Rubin, and D W Cushman, Science, 196, 441-444 (1977) 173 R A Maxwell and S B Eckhardt, Drug Discovery: A Casebook and Analysis, Humana Press, Clifton, N J , 1990, pp 174 Ref 6, pp 105-114, chronicles the develol;ment of the P-adrenoceptor blocking agents 175 D Potterton, Ed., Culpeper's Colour Herbal, Foulsham, London, 1983, p 123 176 K P Link, Harvey Lectures, Series 39, 1944, pp 162-216 177 K P Link, Circulation, 19,97-107 (1959) 178 See Ref 19, pp 1317-1322 179 Mustafa,C R Acad Sci Paris, 89,442 (1879) 180 E Spath and W Gruber, Ber Dtsch Chem Ges., 71, 106 (1938) 181 G V Anrep and G Misrahy, Gaz Fac Med Cairo, 13,33 (1945) 182 G VAnrep,G S Barsourn, M R Kenawy, and G Misrahy, Lancet, 557-558 (1947) 183 G B Kauffman, Educ Chem., 21, 42-45 (1984) 184 See Ref 62, p 192 185 H Cairns, C Fitzmaurice, D Hunter, P B Johnson, J King, T B Lee, G H Lord, R Minshull, and J S G Cox, J Med Chem., 15, 583-589 (1972) References 186 See Ref 19, pp 630-632 187 B N Singh, Am Heart J., 106, 788-797 (1983) 188 B N Singh, N Venkatesh, K Nademanee, M A Josephson, and R Karman, Prog Cardiovasc Dis., 31,249-280 (1989) 189 J van Schepdael and H Solvay, Presse Med., 78,1849-1855 (1970) 190 See Ref 62, pp 189-191 191 See Ref 6, pp 98-105 192 R P Ahlquist, Am J Physiol., 153, 586-600 (1948) 193 A M Lands, F P Luduena, and H J Buzzo, Life Sci., 6,2241-2249 (1967) 194 See Ref.173, pp 333-348 195 D L Burgoyne, R J Anderson, and T M Allen, J.Org Chem., 57,525428 (1992) 196 P I Trigg, in H Wagner, H Hikino, and N R Farnsworth, Eds., Economic and Medicinal Plant Research, Academic Press, London, Vol 3, 1989, pp 19-55 197 W Tang and G Eisenbrand, Eds., Chinese Drugs of Plant Origin, Springer-Verlag, Berlin, 1992, pp 161-175 198 S R Meshnick, A Thomas, A Ram, C.-M X u , and H.-Z Pan, Mol Biochem Parasitol., 49, 181-190 (1991) .99 S R Meshnick, Y.-Z Yang, V Lima, F Kuypers, S Kamchonwongpaisan, and Y Yuthavong, Antimicrob Agents Chemother., 37, 1108-1114 (1993) 00 P L Olliaro et al., Trends Parasitol., 17, 122126 (2001) 01 G H Posner and S R Meshnick, Trends Parasitol., 17, 266-267 (2001) 02 For example: J Cazelles et al., J Chem Soc Perkin Trans 1, 1265-1270 (2000); M D Bachi et al., Bioorg Med Chem Lett., 8,903-908 (1998) 33 J Karbwang, K N Bangchang, A Thanavibul, D Bunnag, T Chongsuphajaisiddhi, and T Harinasuta, Lancet, 340, 1245 (1992), report some clinical experience to support the data in Refs 196 and 197 A N Chawira, D C.Warhurst, B L Robinson, and W Peters, Trans R Soc Trop Med Hyg., 81,554-558 (1987) G D Brown, personal communication B C Elford, M F Roberts, J D Phillipson, and R J M Wilson, Trans R Soc Trop Med Hyg., 81,434-436 (1987) A I White in C Wilson, Gisvold, and R F Doerge, Eds., Textbook of Organic, Medic- inal and Pharmaceutical Chemistry, 7th ed., J B Lippincott, Philadelphia, 1977, pp 247268 208 F A Fluckiger and D Hanbury, Pharmacographia, A History of the Principal Drugs of Vegetable Origin, Met With in Great Britain and British India, Mamillan, London, 1879, pp 361-362 J Pelletier and J Caventou, Ann Chim Phys., XV, 292 (1820) Anonymous, quoted by C Allen, Tales from the Dark Continent, Warner, London, 1992, p 30 P Rabe, Berichte, 55,522 (1922) R B Woodward and W E Doering, J Am Chem Soc., 67,860 (1945) F Schonhofer et al., 2.Physiol Chern., 274, (1942) P Miffdens, Naturwissenschaften, 14, 11621166 (1926) See Ref 19, pp 988-991 G R Coatney, Am J Trop Med Hyg., 12, 121-128 (1963) P Winstanley and P Olliario, Expert Opin Invest Drugs, 7,261-271 (1998) Anonymous, Bull World Health Org., 61, 169-178 (1983) L H Schmidt, R Crosby, J Rasco, and D Vaughan, Antimicrob Agents Chemother., 13, 1011-1030 (1978) R M Pinder and A Burger, J Med Chem., 11, 267 (1968) C J Ohnmacht, A R Patel, and R E Lutz, J Med Chem., 14,926 (1971) 222 J E Rosenblatt, Mayo Clin Proc., 74, 11611175 (1999) 223 F Page, Lancet, 755 (1951) 224 A Freedman and F Bach, Lancet, 321 (1952) 225 G 0.Haydu,Am J Med Sci., 225,71(1953) 226 J Forestier and A Certonciny, Rev Rhum Mal Osteoartic., 21, 395 (1954) 227 A L Scherbel, S L Schuchter, and J W Harrison, Cleve Clin Q., 24, 98 (1957); see also A L Scherbel, Am J Med., 75, (1983) 228 H G Davies and R H Green, Chem Soc Rev., 20,211-269 (19911, provide structural details of a large number of analogs 229 H G Davies and R H Green, Nut Prod Rep., 3,87-121(1986) 230 W C Campbell, M H Fisher, E 0.Stapley, G Albers-Schonberg, and T A Jacob, Science, 221,823-828 (1983) Natural Products as Leads for New Pharmaceuticals 231 K Awadzi, K Y Dadzie, H Schulzkey, D R W Haddock, H M Gillies, and M A Aziz, Ann Trop Med Parasitol., 79,63 (1985) 232 B M Greene, H R Taylor, E W Cupp, R P Murphy, A T White, M A Aziz, H Schulzkey, S A Danna, H S Newland, L P Goldschmidt, C Auer, A P Hanson, S.V Freeman, E W Reber, and P N Williams, N Engl J Med., 313, 133-138 (1985) 233 F A Drobniewski, Microbiology Europe, 24-28 (1993) 234 F Richards, E S Miri, M Katabanva, A Eyamba, M Sauerbrey, G Zea-Flores, K Korve, W Mathai, M A Homeida, I Mueller, E Hilyer, and D R Hopkins, Am J Trop Med Hyg., 66, 108-114 (2001) 235 K Awadzi, S K Attah, E T Addy, N Opoku, and B T Quartey, Trans R Soc Trop Med Hyg., 93,189-194 (1999) 236 B A Boatin, J M Hougard, E S Alley, L K B Akpoboua, L Yameogo, N Dembele, A Seketeli, and K Y Dadzie, Ann Trop Med Parasitol., 92, S47S60 (1998) 237 B Leppard and A E Naburi, Br J Dermatol., 143,520-523 (2000) 238 C N Burkhart, Vet Hum Toxicol., 42,30-35 (2000) 239 L J Scott and K L Goa, Drugs, 60, 10951122 (2000) 240 M A H Mucke, J Froehlich, and U Jordis, WO 0032199 (2000) 241 U Jordis, J Froehlich, M Treu, M Hirnschall, L Czollner, B Kaelz, and S Welzig, WO 0174820 (2001) 242 J D Coombes, Ed., New Drugs from Natural Sources, IBC Technical Services, London, 1992, pp 59-62,93-100 243 G G Yarbrough, D P Taylor, R T Rowlands, M S Crawford, and L Lasure, J.Antibiot., 46, 535-544 (1993) 244 W H Moos, G D Green, and M R Pavia, "Recent Advances in the Generation of Molecular Diversity," in J A Bristol, Ed., Annual Report of Medicinal Chemistry, Vol 28, Academic Press, New York, 1993, pp 315-324 245 Y.-Z Shu, J Nut Prod., 61,1053-1071 (1998) 246 D J Newman, G M Cragg, and K M Snader, Nut Prod Rep., 17, 215-234 (2000) Index Terms that begin with numbers are indexed as if the number were spelled out; e.g., "3D models" is located as if it were spelled "ThreeD models." structure-based design, 437-438 A-80987 structure-based design, 438, 439 A-306552,675 Absolv program, 389 Absorption, distribution, metabolism, and excretion (ADME) as bottleneck in drug discovery, 592 estimation systems, 389-390 molecular modeling, 154-155 Absorption, distribution, metabolism, excretion, and toxicity (ADMET),389 and druglikeness screening, 245 and virtual screening filter cascade, 267 ABT-418,808-810 ABT-627,811-812 ABT-773,876 Academic databases, 387-388 Acarbose, 849 Accelrys databases, 384-385 Accord, 377,385 Accord Database Explorer, 385 Acebutol renal clearance, 38 ACE inhibitors, 718,881 asymmetric synthesis, 807,809 comparative molecular field analysis, 151-153 conformationally restricted peptidomimetics, 640- 641 molecular modeling, 131, 132-133, 145 multisubstrate analogs, 746-748 receptor-relevant subspace, 204 structure-based design, 432-433 transition state analogs, 650-651 Acetic acid CML representation, 372 Acetylcholine, 772 cation-T bonding, 724-725 conformationally restricted analogs, 697-698 muscarinic agonist analogs, 143-144 Acetylcholinesterase inhibitors, 718 CoMFA study, 58-59 pseudoirreversible, 772-774 substrates from acetylcholine analogs, 697-698 target of structure-based drug design, 449-450 volume mapping, 140 X-ray crystallographic studies, 482 N-Acetylcysteine toxicological profile prediction, 838,840 Acetylsalicylic acid (aspirin), 762-763 Acquire database, 386 Actimomycin D thermodynamics of binding to DNA, 183 Actinoplanes utahensis, 878 Active Analog Approach, 58,60, 639 flow of information in, 146 and molecular modeling, 134, 151 and systematic search, 144-145 Active-site directed, irreversible inhibitors, 755 Activity binding affinity contrasted, 131-135 Activity-guided fractionation, 597 Activity similarity, 255 Acyclovir, 717, 719, 756 Acyl halides filtering from virtual screens, 246 901 ADAM geometriclcombinatorial search, 295 ADAMs, 652 ADAPT, 53 Adenosine deaminase inhibitors, 717 mass-spectrometric screening for ligands to, 604 transition-state analogs, 750-752 X-ray crystallographic studies, 482 S-Adenosyl-L-homocysteine, 733,740 S-Adenosyl-L-methionine, 733 ADEPT suite, 225,226,237 ADME studies, See Absorption, distribution, metabolism, and excretion (ADME) ADMET studies, See Absorption, distribution, metabolism, excretion, and toxicity (ADMET) ADP analogs, 763-764 Adrenaline, 885,886 p-Adrenoreceptor antagonists, See p-Blockers Afferent, 387 Affinity calculation, 118-122 Affinity capillary electrophoresis-mass spectrometry, 599-600 Affinity chromatography-mass spectrometry, 598599 Affinity grids, 292-293 Affinity labels, 756-759, 760-764 Affinity NMR, 571 2-AG, 853 AG31 structure-based design, 428-429 AG85 structure-based design, 428 AG331 structure-based design, 428-429 AG337 structure-based design, 428 Index structure-based design, 431-432 AG2037 structure-based design, 431-432 Agenerase structure-based design, 440, 441 Agent, 398 Aggregate concept, in molecular modeling, 90-91 AIDS, See HIV protease inhibitors AIDS database, 386 AIMB, 255 ALADDIN, 259,363 in molecular modeling, 111, 113 Alaninates binding to chymotrypsin, 35 Alanine racemase inhibitors, 717 Alcohol dehydrogenase QSAU studies, Alcohols pharmacophore points, 249 QSAU studies, 27-29 Alcuronium, 856,857 Aldehydes filtering from virtual screens, 246 Aldose reductase inhibitors novel lead identification, 321 target of structure-based drug design, 447-449 Aldosterone, 746 Alkyl amines polarization energy, 173 protonation, 179-180 Alkyl chain homologation analogs, 699-704 Alkyl halides filtering from virtual screens, 246 Allinger force field, 80 Allosteric effectors of hemoglobin, 421-424 and lock-and-key hypothesis, N-Allylmorphine, 850 Almond program, 202 Alogp, 390 Alpha-amylase, 482 N-Alpha-(2-naphthylsulfonylglycyl)-4-amidinophenylal[NAPAP] piperidide structure-based design, 442, 444 Alprenolol renal clearance, 38 Altracurium lead for drugs, 856-858 AM1 wave function, 15, 102 AM404,854 Arnaryllidaceue (snowdrops, daffodils, etc.), 892 Amastatin, 738 AMBER energy function, 264, 307-308 performance in structure prediction, 315 seeding experiments, 319 AMBERIOPLS force field, 80, 81,103 in molecular modeling, 118 American mandrake, drugs derived from, 865 Amides enzyme-mediated asymmetric bond formation, 804-805 exchange ratesltemperature coefficients, in NMR, 512 pharmacophore points, 249 Amines pharmacophore points, 249, 250 Aminidine pharmacophore points, 249 Amino acid mimetics, 640 Amino acids chemical modification reagents, 755 chirality, 784 classical resolution, 797 conglomerate racemates, 802-803 Aminoacyl-tRNA synthetases binding of alkyl groups to, y -Aminobutyric acid aminotransferase (GABA-T) inhibitors, 488, 718, 766-767 y -Aminobutyric acid (GABA), 766 analogs, 690 geometric isomer analogs, 705-706 molecular modeling, 149 7-Aminocephalosporanic acid (7-ACA),871,874 2-Amino-3-chlorobutanoic acid nonclassical resolution, 803 Aminoglutethimide, 717 chromatographic separation, 792 classical resolution, 798, 799 6-Aminopenicillanicacid (6-HA),869,870 Aminopeptidases transition state analog inhibitors, 652 8-Aminoquinolines, 888-889 Amiodarone, 884,885 Amitriptyline, 692 Ammi visnaga (toothpick plant), 883 Amoxycillin, 869,870 AMP analogs, 764 Ampicillin, 869, 870 Amprenavir, 648,659 structure-based design, 440, 441 cu-Amylase X-ray crystallographic studies, 482 P-Arnyloid X-ray crystallographic studies, 482 Analog design, 687-689 alkyl chain homologation, 689, 699-704 bioisosteric replacement and nonisosteric bioanalogs, 689 - 694 chain branching alteration, 689 fragments of lead molecule, 689,707-710 interatomic distances variation, 689, 710-712 limitations of, 532 rigid or semirigid analogs, 689,694-699 ring-position isomers, 689 ring size changes, 689 stereochemistry alteration and design of stereoisomers/geometric isomers, 689, 704-707 substitution of aromatic ring for saturated, or the converse, 689 Anandamide 853 Anchor and grow algorithm, 294,296 AND logical operator, 406 Androgen receptor X-ray crystallographic studies, 482 Angiotensin-converting enzyme (ACE),650,881 See also ACE inhibitors Index active site molecular modeling, 131,132-133 target of structure-based drug design, 432-433 Angiotensin I, 432-433,746,881 Angiotensin 11, 432, 746, 881 non-peptide antagonists, 668-669,670 Anhydrides filtering from virtual screens, 246 Anomalous Patterson maps, 477 Antiasthma drugs natural products as leads, 883-886 Antibacterial enzyme inhibitors, 717 Antibiotic drugs natural products as leads, 868-878 Antibiotic resistant pathogens, 770 Anticancer drugs enzyme inhibitors, 717, 718 molecular modeling, 151 natural products as leads, 858-868 Anticoagulant protein C X-ray crystallographic studies, 482 Anticoagulants, 882-883 Antifolate targets structure-based design, 425-432 Antifungal enzyme inhibitors, 717 Antiparasitic drugs natural products as leads, 886-891 Antiprotozoal enzyme inhibitors, 717 Antisickling agents, 419-421 Antiviral enzyme inhibitors, 717 Aparnin molecular modeling, 124 AP descriptors, 55, 56 Apex, 256,387 Apex-3D, 60 Application tier, 392,398,406 Aquaporin X-ray crystallographic studies, 482 Aqueous solubility and structure-based design, 408 Arabidopsis thaliana genome sequencing, 344 Arabinose binding protein genetic algorithm study of docking, 88-89 Arachidonic acid, 762, 763 Arecoline analogs, 693-694 Argatroban structure-baseddesign, 442,444 Arginase inhibitors, 736-737 Arginine chemical modification reagents, 755 Aromatase inhibitors, 717, 770 Aromatic-aromatic interactions, 286 Aromatics analogs based on substitution of aromatic for saturated ring; or the converse, 699-704 growth inhibition by, 38 molecular comparisons, 139 ArrayExpress, 345 Artabotrys uncinatus, 887 Arteether, 887 Artemether, 887-888 Artemisia annua (sweet wormwood), 886,888 Artemisinin, 849,886-888 Artificial intelligence, 398 Artificial neural networks for druglikeness screening, 247-248,250 in molecular modeling, 126 in QSAR, 53,62,67 for structural genomics study, 353 Arylsulfonamidophenethanolamine analogs, 703 p-Arylthio cinnamide antagonists, 566-567 ASCII (American Standard Code for Information Exchange), 398 Ascomycin binding to FKBP, 552, 553-554 Asinex catalog, 385 Aspartate transcarbamoylase (ATCase)inhibitors, 743-744 Aspartic acid chemical modification reagents, 755 Aspartic peptidase inhibitors transition state analogs, 647-649 virtual screening, 315 Aspergillus alliaceus, 855 Aspergillus nidlans, 878 Aspergillus terreus, 879 Asperlicin fragment analogs, 708 lead for drugs,855-856 Aspirin, 762-763,764 Assay Explorer, 387 Association thermodynamics drug-target binding, 170-171, 177-179 Asymmetric centers, 784,785 Asymmetric synthesis, 784, 804-820 enzyme-mediated, 804-807 Asymmetric transformation crystallization-induced, 798-799 Atenolol, 882 renal clearance, 38 Atom-atom mapping, 380,398 Atom-centered point charges, 101-102 Atomic counts, 54 Atom list, 398 Atom-pair interaction potentials, 120 Atom stereochemistry, 365,398 Atom-type E-State index, 26 Atorvastatin, 744, 880 ATP analogs, 763-764,765 Hf ,KC-ATPase inhibitors, 718 Na+,K+-ATPaseinhibitors, 718 Atracurium, 857,859 Atrasentan (ABT-6271,811-812 Atrial natriuretic factor, 650 Atrolactate, 762 Atropine lead for drugs, 851 Augmentin, 869 Aura-Mol, 388 AUSPYX, 387 AutoDock affinity grids, 293 explicit water molecules, 303 flexible ligands, 263 Lamarckian genetic algorithm, 299 Monte Carlo simulated annealing, 297 protein flexibility, 301 Automap, 398 Available Chemicals Directory (ACD), 385,386 virtual screening application, 254 Avermectins, 891,892 Index Azathioprine, 717 Azithromycin, 848,849, 875-876 Aztreonam, 873-874 Babel, 372 Baccatin 11,863 Backtracking, 382,399 in virtual screening, 67 Bacterial luminescence inhibition by ROH, 27 Bacterial natural products, 848, 893 Bacteriorhodopsin electron cryomicroscopy, 612 homology with GPCRs, 123, 150 Barnard Chemical Information, 388 Bcl-xL target of spin-label NMR screening, 573-574 BCUT descriptors defined, 399 estimation systems, 388 for molecular similarityldiversity methods, 193-194, 203-204 with pharmacophore fingerprints, 223-224 for target class-focused approaches, 228-229,232-233 BCX 1812 structure-based design, 452 BDE index, 11 Beilstein Database, 362,385 Beilstein Online, 385 Benign prostatic hypertrophy expression probabilities, 343 Benzamidines inhibitors of trypsin, 120 Benzene electron density, 135 intermolecular interactions, 174 Benziodarone, 884 Benzomorphans, 850 Benzopyran enzyme-mediated oxidation, 806 Benzylpenicillin,868,870 (R)S-Benzylsuccinic acid, 746 Bestatin, 652, 654, 728 p-Blockers, 881-882 enantiomers, 786 nonrenal clearance, 39 renal clearance, 38 Bezafibrate, 422 BIBP 3226,670,673 BIBR 953 structure-based design, 443, 448 BIBR 1048 structure-based design, 442, 443 BIBU52,213 Bifenthrin, 41, 42 Binary data, 373,399 Binary QSAR, Binding afiinity, 286 activity contrasted, 131-135 calculation, 118-122 Binding constant, 286 Binding property classes, 192 Binding site models, 130 Bioactivity databases, 386 Bioafiinity screening by electrospray FTICR MS, 601-603 Bioavailability, 716 BioCatalysis database, 384 Biochemical force fields, 175-176 Bioinformatics, 333-337 databases, tools, and applications, 345-349 defined, 399 and functional genomics, 338340 future developments, 354 and sequences, 364 standardization, 337 and structural genomics, 352-354 for target discovery, 335, 338-345 Bioinformatics knowledge model, 349-352 Bioisosteric replacement analogs, 689-694 Biological data, 399 Biological evaluation in structure-based design, 419 BioScreen NPISC, 385 BIOSTER database, 202,384 Biosterism, 689-690 Biotin genetic algorithm study of docking to streptavidin, 89 interaction with streptavidin, 181-183 Biotransformationsdatabase, 384 Biphenyls privileged structures, 252 Biphenyl tetrazole, 231 BIRB-796 structure-based design, 458-459 Bisubstrate analog enzyme inhibitors, 741-742 Bitset, 399 BLAST (Basic Local Alignment Search Tool), 347 BLEEP potential, 312 performance in structure prediction, 314 BLOB (Binary Large Object) data type, 399 Blood substitutes structure-based design, 424 BMS-193884,674,676 BMS-247550,864,865 Bohm scoring function, 264 Boltzmann law, 93,310311 Boltzmann probability, 94 Boltzmann weighted average, 97,98 Bond stereochemistry, 365, 399-400 Born-Oppenheimer approximation, 79 Born-Oppenheimer surface, 85 Bothropsj a r m a (pit viper), 881 Bovine liver DHFR, QSAR inhibition studies, 34 Bovine pancreatic trypsin inhibitor long range electrostatic effects, 177 Bradykin, 746,881 targeted library design, 69, 70 BRIDGE, 113 BRN (Beilstein Registry Number), 378, 400 Bromoaspirin, 763 Bromobutide, 42 Bromperidol HIV protease inhibitor, 111, 112 Bryostatin-1,868,869 Bufuralol renal clearance, 38 Bugula neritina, 868 Building block hypothesis, 88 Bupivacaine hydrochloride, 805, 806 classical resolution, 795, 796 Buprenorphine, 850-851 Business rules, 378,400,403 BW12C antisickling agent, 419-420 BW1476U89,742-743 Index CACTVS, 254 Calabash curare, 856 Calcineurin A X-ray crystallographic studies, 483 Cambridge Crystallographic Database, 110 Cambridge Structural Database, 354,387 X-ray crystallography application, 479 CAMP molecular property visualization, 137 D-Camphor connection table, 367 tabular molecular file formats, 370-371 CAMP phosphodiesterase I1 inhibitors molecular modeling, 130 Camptotheca acuminata, 860 Camptothecin drugs derived from, 860-861 Candoxatril, 815,818 Cannabinoids lead for drugs, 852-854 Cannabis sativa, 852 Canonical numbering, 378,400 Capillary electrophoresis for enantiomer separation, 787 Capsaicin, 854 Captopril, 646,650,746-747, 881 asymmetric synthesis, 807, 809 structure-based design, 432-433,434 Carbamates pharmacophore points, 249 Carbenin, 849 Carbo index, 202 Carbonic anhydrase inhibitors, 718 4-fluorobenzenesulfonamide binding to, 538 molecular modeling, 120 virtual screening for, 316 Carbonic anhydrase I1 inhibitors, 718 novel lead identification by virtual screening, 320 X-ray crystallographic studies, 483-484 Carboxylic acids pharmacophore points, 249 privileged structures, 252 Carboxypeptidase A similarity of active site to ACE, 433,746-747 Carboxypeptidase inhibitors conformational changes on binding, 261 genetic algorithm study of active site, 89 molecular modeling, 116 Cardiovascular drugs natural products as leads, 878-883 CART, 67 Cascade clustering with molecular similarityldiversity methods, 205 CASETox, 246 CAS Numbers, 378,400 CAS ONLINE, 361 similarity searching, 383 Caspase-1inhibitors target of structure-based drug design, 443 transition state analogs, 655 Caspase-3 inhibitors transition state analogs, 655 Caspase-7 inhibitors transition state analogs, 655 CASP experiment (Critical Assessment of techniques for protein Structure Prediction), 123,353 Caspofungin, 848,878 CASREACT, 385 CAS Registry, 363,385 Catalyst (program),259 3D shape-based searching, 199,201 for novel lead identification, 321 Catalyst/Hip/Hop, 60 Catalyst/Hypo, 60 Catechol methyltransferase X-ray crystallographic studies, 484 CATH, 353 X-ray crystallography application, 494 Catharanthus roseus, vinca alkaloids from, 858 Catharanthus (vinca)alkaloids, 858-860 Cathepsin B transition state analog inhibitors, 654 Cathepsin D combinatorial docking studies, 318 non-peptide inhibitors, 227 Cathepsin K transition state analog inhibitors, 654-655, 656 Cation-P interactions, 286, 313 enzyme inhibitors, 724-725 CATS descriptors, 192 CAVEAT, 111,113 CAVITY, 106-107,108 CBS reduction, 814 cDNA clone libraries, 341-342 cDNA microarray chips, 344-345 Cefaclor, 872 Cefadroxil, 872 Cefetamet pivoxil, 849 Cefozopran, 849 Cefpimizole, 849 Cefsulodin, 849 Ceftazidime, 872,873 Ceftizoxime, 872,873 Ceftriaxone, 872,873 Cefuroxime, 872,873 Cell-based partitioning methods, 203 CellCept structure-based design, 446-447 Central Library, 378,387 Central nervous system (CNS) drugs, See CNS drugs Cephacetrile, 871,874 Cephalexin, 871-872 Cephaloglycin, 871-872 Cephaloridine, 871,872 Cephalosporin C, 870471,874 Cephalosporins, 717,870-871 preventing bacterial degradation, 718 Cephalosporium acremonium, 870 Cephalothin, 871,874 Cephamandole, 872,873 Cephapririn, 871, 874 Cephradine, 872 Cerivastatin, 880 Cetirizine, 783 Cetirizine dihydrochloride chromatographic separation, 790-791 cGMP molecular property visualization, 137 Index CGS 27023 structure-based design, 444, 446 Chain branching alteration analogs, 699-704 Chapman databases, 387 Charge-charge interactions, 82 Charge-coupled devices (CCDs) for electron cryomicroscopy, 623 for X-ray crystallography, 474 Charge-dipole interactions, 82 Charge parameterization, 101-102 Charge state determination NMR spectroscopy for, 526 Charge transfer energy, 173 CHARMM, 298,299,307-308 in molecular modeling, 118, 126 CheD, 387 ChemBase, 362 CHEMCATS, 385 CHEMDBS3D, 260,363 ChemDraw, 362 ChemEnlighten, 387 ChemExplorer, 384 ChemFinder for Word, 384,388 ChemFolder, 388 Chemical Abstracts (CAS) registry file, 50 Chemical Abstracts Service databases, 254,361,385 Chemical business rules, 378, 403 Chemical information computing systems, 357-363 chemical property estimation systems, 388-390 chemical representation, 363-373 databases, 384-388 data warehouses and data marts, 390393,402-403 future developments, 393-397 glossary of terms used, 397-412 registering chemical information, 377-379 searching chemical structures1 reactions, 379-384 storing chemical information, 373-377 Chemical information management databases, 384 Chemical information management systems, 384 Chemical libraries, See Libraries Chemical Products Index, 391-392 Chemical property estimation systems, 388390 Chemical reactions, 366 searching, 379-384 Chemical representation, 363-373 Chemical shift, in NMR, 511, 512 changes on binding, 536-537 perturbations as aid in NMR screening, 562-568 Chemical-shift mapping, in NMR, 543-545 Chemical similarity, 382 Chemical space, 244, 383,400 exploring with molecular similarity/diversity methods, 188, 191 reduction by virtual screening, 244-245 Chemical Structure Association, 360 Chemical structures file conversion, 372-373 searching, 379 -384 Chemical suppliers searching, 384 CHEM-INFO, 360 ChemInform, 386 Cheminformatics, 359,400 Cheminformatics Glossary, 360 ChemPort program, 385 Chemscape, 387 ChemScore consensus scoring, 266 empirical scoring, 310 ChemSpace, 199 ChemText, 362 ChemWindow, 388 Chem-X, 60, 111 CheflChemDiverse 3D pharmacophores, 195-196, 206 optimization approach, 217 and property-based design, 234 Cherry picking, combinatorial libraries, 216-217,237 Chesire, 378,387 Chicken liver DHFR, QSAR inhibition studies, 31-32 Chilies, capsaicin in, 854 Chime, 369,371,387 Chiral auxiliary, 810-813 Chiral catalysts, 814-820 Chiral centers, 783-785 Chiral derivatizing agents, 788 Chiral flags, 365,366 Chirality, 781-787,820-821 asymmetric synthesis, 804-820 chromatographic separations, 787-793 classical resolution, 793-799 enzyme-mediated asymmetric synthesis, 804-807 nonclassical resolution, 799-804 Chiral pool, 807-810 Chiral reagent, 813-814 Chiral stationary phase, 787-788,790-791 Chlopromazine, 692 Chloramphenicol, 870 molecular modeling, 150 4-Chloro-l,3-benzenediol allergenicity prediction, 834 Chloromethyl ketones protease inhibitors, 761-762 p-Chlorophenylalanine classical resolution by crystallization, 798-799, 800 Chloroquine, 889,890 Chlortetracycline, 870 Cholchicine toxicological profile prediction, 841,842 Cholecystokinin, 855 X-ray crystallographic studies, 484 Chorismate mutase inhibitors transition state analogs, 753-754 Chromatographic separation of chiral molecules, 787-793 Chromobacterium violactum, 873 Chromosomes in genetic algorithms, 87 Chymotrypsin inhibitors affinity labels, 761, 762 molecular modeling, 118 QSAR studies, 5, 35-36 CICLOPS, 223 Cilazapril asymmetric synthesis, 807, 809 Cilofungin, 877 Cinchona bark, quinine from, 888 Index CIP (Cahn-Ingold-Prelog) stereochemistry, 365,400 Cisplatin vindesine with, 860 cisltrans stereochemistry, 399 Clarithromycin, 849,874, 875-876 Classical resolution, of c h i d molecules, 793-799 Clavulanic acid, 718,869,870 Cleaning and transforming data, 400 Clenbuterol chromatographic separation, 787,788 Client-server architecture, 400-401 Clipping, 378,401 Clique search techniques, 262 CLOB (Character Large Object) data type, 401 Clofibric acid antisickling agent, 421, 422 CLOGP, 18,389 Clog P, 17-18,36 Cloning, 127 cDNA clone libraries, 341-342 Clotrimazole, 717 Clustering methods, 379,401 for combinatorial library design, 220 in molecular modeling, 90-91 with molecular similarityldiversity methods, 205 CML (Chemical Markup Language), 371372,401,405, 412 CNS drugs complementarity, 134 natural products as leads, 849-856 pharmacophore point filters, 250 polar surface area, 245 CNS program, 478 Coagulation factor X-ray crystallographic studies, 484-485 Coagulation factor X-ray crystallographic studies, 485 Coagulation factor 7a X-ray crystallographic studies, 485-486 Coagulation factor X-ray crystallographic studies, 486 Coagulation factor 10 X-ray crystallographic studies, 484 COBRA, 255 R-Cocaine dopamine transporter inhibitor, 268 Codeine, 849,850 Coformycin, 750-752 Cognex structure-based design, 449 Colforsin daropate, 849 Collagenase NMR binding studies, 555, 556 target of structure-based drug design, 443 CombiBUILD, 227 CombiChem Package, 386 CombiDOCK, 217,227 combinatorial docking, 318 CombiLibMaker, 378,387 Combinatorial chemistry, 283, 358,591-592 defined, 401 and molecular modeling, 155 and natural product screening, 848 Combinatorial chemistry databases, 387 Combinatorial docking, 317318 Combinatorial libraries, 214 comparisons, 221-223 design for molecular similarity methods, 190,214-228 encoding and identification with mass spectrometry, 596-597 integration, 224-225 LC-MS purification, 592-594 optimization, 217-221 peptidomimetics, 657 screening for ligands to two receptors simultaneously, 601-602 structure-based design, 225-228 structure/purity c o n h a t i o n with mass spectrometry, 594-596 with virtual screening, 317 S,S-Combretadioxolane, 816, 819 Combretastatin A-4,816,819 CoMFA, See Comparative molecular field analysis Compactin, 744,879 Comparative binding energy analysis (COMBINE), 53 and docking methods, 304-305 Comparative molecular field analysis (CoMFA), 53-54 assessment of predictability, 151-153 3D, 58-60 and docking methods, 304 field mapping, 107 molecular field descriptors, 56-57 and molecular modeling, 138, 147 Comparative quantitative structure-activity relationships database development, 39 database mining for models, 39-41 Competitive inhibitors, 728-729 Complementarity, 134 Comprehensive Medicinal Chemistry database, 379, 386 Computational Chemistry List, 360 Computational protein-ligand docking techniques, 262-264 Computing technologies, 334-335, 337 See also Chemical information computing systems COMSiA, 53,60 CONCORD, 363,366,387,401 3D coordinate generation, 267 3D descriptors, 55, 110 virtual screening application, 254 Concordance, 390,401 Conformational analysis in molecular modeling, 87, 93-94 NMR spectroscopy for, 525-526 and systematic search, 89-93 Conformational clustering, 92-93 Conformational flexibility, 288 Conformationally restricted analogs, 694-699 Conformationally restricted peptides, 636-643 Conformational mimicry, 140-142 Index Conformational mimicry index, 142 Conglomerate racemates, 799-800,801,802-803 Connection tables, 365-368,371, 401 file conversion, 372373 ISIS database, 376 Connectivity, See Molecular connectivity o -Conotoxins lead for drugs, 851-852 NMR spectroscopy, 518-523 ConQuest search program, 387 Conscore constraint score, 218 Consensus scoring, 265-266, 291,319-320 and molecular modeling, 117-118 Consistent force fields, 102 Constrained minimization, 143-144 Contact matrix, 125-127 Contignasterol, 886 CONTRAST, 361 Conus magus, conotoxins from, 851 Convertases homology modeling, 123 CONVERTER, 366,402 CoQSAR, See Comparative quantitative structure-adivity relationships CORINA, 366,402 3D coordinate generation, 267 3D descriptors, 55 virtual screening application, 254 Cosine coefficient, 68 COSMIC force field, 80 Coulomb's law, 80,82,285 and dielectric problem, 83 Coumarin, 882 Counting schemes in druglikeness screening, 245-246 Coupling constants, in NMR, 511,512 changes on binding, 536-537 for conformational analysis, 525 COUSIN, 361,373,387 and combinatorial library integration, 224 Covalent bonds, 6,170 Covalently binding enzyme inhibitors, 720, 754-756 inactivation of, 756-760 Cox-1 inhibitors, 718 X-ray crystallographic studies, 486 COX-2 inhibitors, 718 mass-spectrometric binding assay screening, 604 seeding experiments, 319 X-ray crystallographic studies, 486 CP-96,345, 670,672 C-QSAR database, 39 Crambin molecular modeling, 124 Crixivan structure-based design, 438-439 CROSS-BOW, 361 Crossfire Beilstein, 385 Cross-linked enzyme crystals, 804 Crosslinking agents, 424-425 Cross validation, 57, 64 Cryoprobes in NMR screening, 577 in NMR spectroscopy, 515 Cryptotheca cripta, 867 Crystallization for asymmetric transformation of enantiomers, 798-799 for enhancing chromatographic separation of enantiomers, 792-793 in nonclassical resolution, 799-804 CScore, 117 Curare lead for drugs, 856-858 Cyclic lactarns conformationally restricted peptidomimetics, 640-642 Cyclic protease inhibitors, 636 Cyclin-dependent kinase (CDK2) H717 inhibitor pharmacophore, 253 Cyclo(Gly6) genetic algorithm exploration of conformational space, 88 Cycloheptadecane potential smoothing study, 86 Cyclooxygenase 112 inhibitors, See COX-1 inhibitors; COX-2 inhibitors Cyclophilin, 552 D-Cycloserine, 717, 719 Cyclosporin, 848 molecular modeling, 106 NMR spectroscopic binding studies, 539 Cyclosporin A binding to FKBP, 552-553 y -Cystathionase inhibitors, 719-720 Cysteine chemical modification reagents, 755 Cysteine peptidase inhibitors transition state analogs, 652-655 Cysteine protease inhibitors affinity labels, 762 Cytochrome P450 homology modeling, 123 Cytochrome P450 reductase X-ray crystallographic studies, 486 Cytosine arabinoside, 717, 867-868 D2163,804,806 Daemon, 392,402 Daffodils, drugs derived from, 892 Dalfopristin, 876-877 Spiro-DAMP, 696 4-DAMP semirigid analogs, 695-696 Daptomycin, 848 DARWIN, 299 explicit water molecules, 303 Databases for bioinformatics, 345349 cDNA microarray chips, 345 chemical information management, 384 commercial systems for drugsized molecules, 8 comparative QSAR, 39-41 comparing expressed sequence tags with, 342 history of, 360-363 knowledge discovery in, 393395 natural products, 387,597 for pharmacophore screening, 254-255 proprietary and academic, 387-388 sequence and 3D structure, 387 storing chemical information in, 373-377 Index for X-ray crystallography, 478-479 Database tier, 393,403,407 Data cartridges, 395,402 Data compression, 402 Data dictionary table, 375 Data marts, 391393,402 Data mining, 402,410,411,412 future prospects, 394-395 with QSAR, 66-67 Data warehouses, 390-393, 402-403 Dative bonds, 170,365 Daunomycin thermodynamics of binding to DNA, 183 DayCart, 386 DayCGI, 386 Daylight Chemical Information Systems databases descriptors, 192 in virtual screening, 254 10-Deacetylbaccatin11,803 Decamethonium, 58 fragment analogs, 708-710 lead for drugs, 856-858 Deceptive fitness functions, 88 Decision support systems, 403 Decision tree approach, 247-248 Deconvolution, 401 Deduplication, 378,403 Dehydroalantolactone allergenicity prediction, 836 Demexiptiline, 692-693 DENDRAL, 393 De novo design, 113 (R)-Deoxycoformycin, 750-752 (S)-Deoxycoformycin, 751 DEREK, 246 Dement Information databases, 386 Dement Selection database, 386 Dement World Drug Index (WDI),379, 386, 387 Dement World Patents Index (WF'I), 386 10-Desacetylbaccatin111,863 Descriptor pharmacophores, 60- 63 Design in Receptor (DiR)approach, 236 DHFR, See Dihydrofolate reductase Diamino, 5Y, 6-Z-quinazolines QSAR studies of DHFR inhibition, 34-35 Diastereomers, 784 chromatographic separation, 788 Dice coefficient, 68 Dicoumarol, 882 Dictionary of Natural Products, 597 Dideoxyinosine, 717 Dielectric problem, 83-84 Dienestrol, 706-707 Diethylstilbestrol stereoisomer analogs, 706-707 Diffusion-filtered NMR screening, 570-571 a-Difluoromethylornithine, 717 Digital Northern, 342 Dihydroartemisinin, 887 Dihydrofolate reductase inhibitors, 545, 717 chemical-shift mapping of binding, 545 comparative molecular field analysis, 153 genetic algorithm study of active site, 89 genetic algorithm study of docking, 88-89 interaction with methotrexate, 120 interaction with trimethoprim, 151,183 interaction with trimetrexate, 531,557459 mass-spectrometric binding assay screening, 604 molecular modeling, 114, 115, 116,147,151 QSAR studies, QSAR studies of inhibition by diamino, 5Y, 6-Z-quinazolines, 34-35 QSAR studies of inhibition by diamino-5X-benzyl pyrimidines, 39 QSAR studies of inhibition by triazines, 31-33 target of structure-based drug design, 425-426 volume mapping, 140 X-ray crystallographic studies, 486 Dihydromuscimol, 690 Dihydroorotase transition state analogs, 752 Dihydroorotate dehydrogenase inhibitors X-ray crystallographic studies, 486 Dihydropteroate synthetase inhibitors, 717 X-ray crystallographic studies, 486 1,4-Dihydropyridines chromatographic separation, 788,789 Dihydroquinine, 889 Dihydrotestosterone, 36, 768 Diller-Merz rapid docking approach, 292,295 assessment, 303 combinatorial docking, 317 Diltiazem nonclassical resolution, 803, 805 Dimension tables, 390,403 N,N-Dimethyldopamine alkyl chain homologation analogs, 701 bioisosteric analogs, 690, 692 semirigid analogs, 695 A6,,,,-DimethylheptylTHC, 852 Dimethyl sulfoxide (DMSO) force field models for, 176 Dimethyltubocurarine, 857 Diphenylmethane, 231 privileged structures, 252 2,3-Diphosphoglycerate (2,3-DPG),104,421 2,3-Diphosphoglycerate (2,3-DPG)analogs, 103, 104 Dipolar electrostatic forces, 172 Dipole-dipole interactions, 6,82 Dipole-induced dipole interactions, 173 Directed tweak algorithm, 260 Directionality, 140 DISCO, 58,60,256 and molecular modeling, 147 Discodermolide genotoxicity prediction, 843 Disintegrins, 652 Disoxaril structure-based design, 454-455 Dispersive interactions, 82, 174 See also van der Wads forces Dissimilarity approaches, 189-190,206-208 Dissociation constant, 286 Distamycin binding perturbations, 544 Index Distance geometry methods in molecular modeling, 126, 142,147 in QSAR, 60 in virtual screening, 263 Distance geometry QSAR technique, 53 Distance matrix, 135 Distance measures molecular modeling, 135-137 molecular similarityldiversity methods, 201-202 Distance range matrix, 135-136 Dithromycin, 875 DiverseSolutions, 387 for molecular similarityldiversity methods, 193-194, 203-204 Diversity analysis, 358 Diversity methods, See Molecular similarityldiversity methods Diversity-property derived (DPD) method, 201,203 Dixon plots, 731-732 D,L descriptors, for chiral molecules, 783 DMB-323 NMR binding studies with HIV protease, 560-562 DMHB structure-based design, 424 DMP 450,659 structure-based design, 438-439 DNA molecular modeling, 154 NMR structural determination, 535 noncovalent bonds in, 170 supercoiling modeling, 95 synthesis inhibition by phenols, 40 DNA-binding drugs chemical-shift mapping of binding to, 544-545 molecular modeling, 116 NMR spectroscopy, 547-552 thermodynamics of binding, 183 DNA gyrase inhibitors novel lead identification, 321 DNA helicase pcra X-ray crystallographic studies, 487 DNA polymerase inhibitors, 342, 717 DNA topoisomerase X-ray crystallographic studies, 487 Docetaxel, 849,863 DOCK anchor and grow algorithm, 296 assessment, 303,304 combinatorial docking, 318 consensus scoring, 266 empirical scoring, 310 force field-based scoring, 308 force-field scoring, 264 geometriclcombinatorial search, 295 ligand handling, 293 molecular modeling, 112, 113, 115,116 molecular modeling of small cavity, 106, 107 penalty terms, 313 performance in structure prediction, 314 protein and receptor modeling, 267 protein flexibility, 301 receptor representation in, 291 rigid docking, 262-263 sampling/scoring methods used, 261 seeding experiments, 319 with site-based pharmacophores, 236 DOCK4.0 PMF scoring, 265 weak inhibitors, 319 Dockcrunch project, 317 Docking methods See also Scoring functions; various docking programs; Virtual screening assessment, 303-304 basic concepts, 289-290 combinatorial, 317-318 flexible ligands, 293-294,322 and homology modeling, 305-306 and molecular modeling, 113-118 and molecular size, 312313 NOE docking in NMR, 545-546 penalty terms, 313 protein flexibility, 300-302, 322 protein-ligand docking software, 261 and QSAR, 304-305 searching configuration and conformation space, 294-300 seeding experiments, 318-319 special aspects, 300-306 in structure-based virtual screening, 260-267 as virtual screening tool, 266-267 water's role, 302-303, 313-314 Docking problem, 289 DockIT, 261 Dockvision, 261 Dolabella auricularia, 868 Dolastatin-10, 868,869 DoMCoSAR approach, 305 Donepezil structure-based design, 449 L-Dopa, 785 analogs, 690 Dopamine semirigid analogs, 697 Dopamine-transporter inhibitors pharmacophore model, 256, 258 virtual screening, 267-269, 270 D-Optimal designs, 65-66 Dose-response curves, Dothiepin, 692-693 Doxepin, 692-693 DragHome method, 305 DRAGON, 388-389 DREAM++, 318 Drill-down, 391,403 Dronabinol, 849 Drug databases, 385-386 See also Databases Drug Data Report, 379,386 Drug development, 509-510 serial design costs, 359 Druglikeness screening See also Lipinski's "rule of 5" molecular similarity/diversity methods, 191 similarity searching, 383 virtual screening, 245-250 Drug-receptor complexes, 170-179 low energy state of, Drug resistance antibiotic resistant pathogens, 770 Index essential pathways versus single enzyme inhibitor, 495 Drugscore function, 311,312 assessment 303 performance in structure prediction, 314 seeding experiments, 319 in virtual screening, 315 Drug screening, See Screening Drug-target binding forces, 170-171 association thermodynamics, 170-171,177-179 energy components for intermolecular noncovalent interactions, 171-174 example drug-receptor interactions, 181-183 free energy calculation, 180-181 molecular mechanics force fields, 174-177 Drug targets and bioinformatics, 351-352 estimated number of, 50 x-ray crystallography of published structures, 482-493 DYLOMMS, 107 E Coli mutagenicity prediction, 829 Eadie-Hofstee plot, 727, 729, 731 ECEPP force field, 118 Echinocandins, 877-878 Ecteinascidia turbinata, 868 Ecteinascidin-743, 848, 867-868 ECTL (Extracting, Cleaning, Transforming, and Loading) data, 377-379,403 Edman sequencing, 518 Edrophonium, 58 Efaproxaril (RSR-13),422 Eflornithine, 768, 769 Eigenvector following method, 292,301 Einstein-Sutherland equation, 24 Elan, 387 Electron cryomicroscopy, 611-628 image processing and 3D reconstruction, 624-628 image selection and preprocessing, 623-624 three-dimensional, 615-616 Electron-donating substituents, 12-15 growth inhibition by, 41 Electronic parameters in QSAR, 11-15,50 Electron lenses, 612 Electron probability distribution, 101 Electron-topological matrix of congruence, 147 Electron-withdrawing substituents, 11-15 growth inhibition by, 41 Electrospray FTICR mass spectrometry, 601-603 Electrostatic interactions, 171-172,285 charge parameterization, 101-102 and docking scoring, 308 enzyme inhibitors, 721, 723 long range, 177 molecular modeling, 81-85, 108-110,140 and molecular property visualization, 137 and QSAR, 6-7,52 Electrotopological indices, Elimination algorithms, 207 EMBL Nucleotide Sequence Database, 335 Embryo tail defects, 40 EMD 122946,676 Empirical scoring, 264, 307, 308-310 Enalapril, 650,747,881 asymmetric synthesis, 807, 809 conformationally restricted peptidomimetics, 640-641 Enalaprilat, 650, 747 conformationally restricted peptidomimetics, 640-641 Enantiomeric excess, 784 enrichment by crystallization, 800-802 Enantiomers, 365, 366 See also Chirality with agonist-antagonist properties at same receptor, 705 chromatographic separations, 787-793 defined, 783-785 Enantioselective metabolism, 786-787 Enantioselectivity, 784 Encoding and genetic algorithm, 88 natural products with mass spectrometry, 596-597 Encryption, 403 Endorphins, 634,850-851 model receptor site, 149 Endothelin antagonists, 211, 672-674, 675,676 conformationally restricted peptidomimetics, 637, 639 NMR spectroscopy, 523-524, 526-527 ENERGI approach, 127 Energy driven/stochastic search strategies, 292,296-2300 Energy of association, 177 English yew, paclitaxel from, 861-862 Enkephalins, 634,850-851 conformationally restricted peptidomimetics, 129, 637, 639 model receptor site, 149 Ensemble, 94 Enthalpy of association drug-receptor complexes, 170-171 Entoviruses target of structure-based drug design, 454-456 Entrainment, 802 Entropy, 94 Entropy of association drug-receptor complexes, 170-171 Enumerated structure, 368 Enumeration, 401,403 Enzyme-induced inactivators, 756 Enzyme-inhibitor complexes, 721-722 Enzyme inhibitors, 715-720 See also specific Enzymes affinity labels, 756-759, 760-764 design of covalently binding, 720, 754-756 design of noncovalently binding, 720-754 examples used in disease treatment, 717 ground-state analogs, 720, 740 -741 inactivation of covalently binding, 756-760 mechanism-based, 759-760, 764-771 multisubstrate analogs, 720, 741-748 Index Enzyme inhibitors (Continued) pseudoirreversible, 771-774 rapid, reversible, 720, 728-734 slow-, tight-, and slow-tightbinding, 720,734-740,749 transition-state analogs, 720, 748-754 Enzyme-mediated asymmetric synthesis, 804-807 Enzymes as drug targets, kinetics, 725-728 pathways and inhibitor design, 495 and structural genomics, 352 Ephedra, 885 Ephedrine, 884-886 (-)-Epibatidine,819-820,821 Epothilone A,864 toxicological profile prediction, 838-839 Epothilone B, 864-865 Epothilone D, 865 Epoxides filtering from virtual screens, 246 Equivalence class, 403 Erythro-9-(2-hydroxy-3-nonyl) (EHNA) high-affinity adenosine deaminase ligand, 604 Erythromycin, 870,871 Erythromycin macrolides, 874-876 Erythro- prefix, 784 Erythrose enantiomers, 784 E, constant (Taft), 23-24 E-selectin NMR screening binding studies, 572 E-State index, 26,54 Esters pharmacophore points, 249 Estradiol, 706,771 Estrogen receptor l a X-ray crystallographic studies, 487 Estrogen receptors mass-spectrometric binding assay screening, 604 Ethacrynic acid antisickling agent, 421 Ethidium bromide thermodynamics of binding to DNA, 183 Etodolac classical resolution, 796-797 Etoposide, 717,867 Etorphine, 851 Euclidean distance, 68,202 EUDOC assessment, 303 ligand handling, 293 European Bioinformatics Institute (EBI),335 sequence databases, 387 Everolimus, 849 Evolutionary algorithms, 299 with QSAR, 53-54,61 Exact match search, 378, 379-381,403 Exchange repulsion energy, 172-173 Exemestane, 770,771 Exhaustive mapping, 398 Expert Protein Analysis System, 335 Expressed sequence tags, 338 expression level significance, 342-344 profiling, 341-342 Expression analysislprofiling, 334 genome-wide, 344345 for target discovery, 340-345 Extended stereochemistry,365, 404 External registry number, 379, 404 Extrathermodynamic relationship, 26 E,Z system, 365,399 Factorial designs, 65-66 Factor Xa inhibitors, 103,738 3D pharmacophores, 199 non-peptide peptidomimetics, 662.665 site-based pharmacophores, 235-236 target of structure-based drug design, 442 Fact tables, 390,401,404 Failed Reactions database, 385 Families, 93 Family competition evolutionary algorithm, 299 FASTA, 347 Fast ion bombardment, 586,587 Fastsearch index, 376377,399, 404 FBSS, 202 FeatureTrees, 316,321 Fibonacci search method, 11 Fibrinogen virtual screening studies, 212-213 Field-based descriptors, 201 Field effects, 140 Field mapping, 107 Fields, 404 Filtercascade, 267 Filters, for searching, 315-316, 376,380,392,404 Finasteride, 717,768-770 Fingerprint Generation Pack, 388 Fingerprints, 376,378,399,404 molecular similarity methods, 188 FIRM, 67 Fitness functions, 87-88 FK506 binding to FKBP, 552-555 NMR spectroscopic binding studies, 539 FK506 binding protein inhibitors, 552-555 de novo design, 113 flexible docking studies, 265 hydrogen bonding in, 288 target of NMR screening studies, 565-566,571 weak inhibitor screening, 319 X-ray crystallographic studies; 487 Flat database storage, 362-363, 404 Flat file storage, 360-362,404 Flecainide enantiomers, 786 FlexE, 301 FlexibaseFLOG, 263 Flexible ligands in docking methods, 293-294, 322 and geometric/combinatorial search, 295 in virtual screening, 263-264 Flexmatch index, 376 Flexmatch search, 404 FlexS, 316 novel lead identification, 320, 321 PlexX assessment, 303,304 consensus scoring, 266,320 empirical scoring, 264,310 explicit water molecules, 302 Index hydrogen bonding, 319 incremental construction, 295-296 molecular modeling, 115 novel lead identification, 320 performance in structure prediction, 314 protein and receptor modeling, 267 receptor representation in, 291 sampling/scoring methods used, 261 seeding experiments, 319 FlexXc extension, 318 Flickering cluster model, of hydrophobic interactions, 15 Flo, 256 Flobufen, 41-42,42 FLOG explicit water molecules, 303 ligand handling, 293 and molecular size, 312313 seeding experiments, 318 Flunet structure-based design, 451 Fluorescence spectroscopy, 592 4-Fluorobenzenesulfonamide binding to carbonic anhydrase, 538 5'-p-Fluorosulfonylbenzoyl adenosine (5'-FSBA), 763764 5-Fluorouracil, 717, 718 Flurbiprofen, 763 Fluvastatin, 744,879-880 FOCUS-2D method, 68-69 Fold compatibility methods, 353 Fold patterns limited number of, 353 Follicle stimulating hormone X-ray crystallographic studies, 488 Force fields drug-target binding forces, 170-183 molecular modeling, 79-81 parameter derivation, 102-103 Force-field scoring, 264, 306-308 Formestane, 770, 771 Formula table, 376 FOUNDATION, 112-113 Fourier transform ion cyclotron resonance (FTICR) mass spectrometry, 585,601-603 4-Point pharmacophores, 408 molecular similarity methods, 189,196-198,205 privileged, 231 virtual screening, 210 FPL-67047 structure-based design, 453 Fractional factorial designs, 66 Fragment analogs, 707-710 Fragment-based ligand docking, 294 FRED ligand handling, 293 sampling/scoringmethods used, 261 Free energy of association, 286 calculating, 180-181 drug-receptor complexes, 5, 170-171 enzyme-inhibitor complexes, 722 Free energy perturbation, 307, 308 Free-Wilson approach, in QSAR, 4,29-30 Frontal affinity chromatography-mass spectrometry, 601 5-FSA structure-based design, 424 FTDOCK, 115 Ftrees-FS algorithm, 221 Fujita-Ban equation, Fujita-Nishioka analysis, 13 Functional genomics, 338-340 Functional group filters in druglikeness screening, 246-247 Functional mimetics (peptidomimetics), 636 Fungal natural products, 848, 893 Fungal squalene epoxidase inhibitors, 717 Fungal sterol l4a-demethylase inhibitors, 717 Fuzzy bipolar pharmacophore autocorrelograms, 197 Fuzzy clustering technique with molecular similarityldiversity methods, 205 Fuzzy distance, 57-58 Fuzzy searches, 376 G-4120,663 GABA, See y -Aminobutyric acid GABA aminotransferase (GABA-T)inhibitors, 718, 766-767 X-ray crystallographic studies, 488 p-D-Galactoside saturation transfer difference in binding to agglutin I, 569 Galantamine (galanthamine), 848,849,892-893 nonclassical resolution, 802, 803 GALOPED, 218 Gambler consensus scoring, 266 flexible ligands, 263 seeding experiments, 319 Gas chromatography, 592 for enantiomer separation, 787 Gas chromatography-mass spectrometry (GC-MS), 585-586 GASP (Genetic Algorithm Similarity Program), 256 in molecular modeling, 147 Gas phase association, 177, 178 G-CSF X-ray crystallographic studies, 488 Gelatinase NMR binding studies, 555 Gel permeation chromatography-mass spectrometry, 599 GenBank, 335 growth of, 339 X-ray crystallography application, 481 GeneChips, 344 Gene expression, 351 See also Expressed sequence tags; Expression analysislprofiling Gene family approaches, 188, 244 subset selection, 190-191 Gene family databases, 347-349 Gene nomenclature, 337 Gene Ontology project, 337 Generic structures, 367, 368, 404-405 Geneseq database, 346 Genetic algorithms and combinatorial library design, 217,218 with docking methods, 292, 298-299 Index Genetic algorithms (Continued) with FOCUS-2D method, 68- 69 inverse folding and threading, 124-125 Larnarckian, 299 in molecular modeling, 87-89, 117 with QSAR, 53,61 in virtual screening, 263 GeneTox, 829 Genie Control Language, 378 Genitoxants, 840 Genome annotation, 481,494 Genome-wide expression analysis, 344345 Genomics, See Functional genomics; Phylogenomics; Structural genomics GEOCORE, 124 Geometric atom-pair descriptors, 210 Geometric/combinatorialsearch strategies, 292,295 Geometric hashing, 262 Geometric isomer analogs, 704-707 Ghost membranes EPR signal changes by ROH, 27 Ghrelin, 671, 674 GHRP-6,671,674 Gibbs-Helmholtz equation, 286 Gigabyte, 405 GLIDE sampling/scoring methods used, 261 Global stereochemistry, 398 Glucocorticoid receptor X-ray crystallographic studies, 488 Glucose, 784 Glutamate dehydrogenase, 764 Glutamate NDMA agonists, 150 Glutamate NDMA antagonists, 150 Glutamate receptor X-ray crystallographic studies, 488 Glutamic acid chemical modification reagents, 755 nonclassical bioisostere analog, 694 rigid analogs, 699 Glutamine-PRPP amidotransferase inhibitors, 717 Glutathione peroxidase X-ray crystallographic studies, 488 Glycinamide ribonucleotide formyltransferase inhibitors, 742-743 target of structure-based drug design, 429-432 Glycophorin A potential smoothing study of TM helix dimer, 86 Glycoprotein IIb/IIIa (GpIIb/ IIIa) inhibitors non-peptide peptidomimetics, 662-665 template mimetics, 129, 643, 645 Glycopyrrolate stereoisomers, 784-785 Gmelin database, 386 GOLD assessment, 303, 304 empirical scoring, 309 flexible ligands, 263 genetic algorithm, 299 protein flexibility, 300 sampling/scoring methods used, 261 GOLEM, 67 GOLPE, 54,60 GPCR libraries 3D pharmacophore fingerprints for, 205 GPCR-likeness, 251,252 GPCRs (G-protein-coupledreceptors), 668 focused screening libraries targeting, 209,250 homology modeling, 123, 150 molecular modeling, 122 peptidomimetics, 644, 677 7-transmembrane, 229-234 GRAB-peptidomimetic (GroupReplacement Assisted Binding), 636,658-659,677 GRAMM (Global Range Molecular Matching), 115 Granulocyte-macrophage CSF X-ray crystallographic studies, 488 Graphical representation, 371 Graph isomorphism problem, 380,405 GREEN force-field scoring, 264 GRID, 58,315 3D pharmacophores, 198 empirical scoring, 309 explicit water molecules, 303 hydrogen bonding, 107 and molecular modeling, 138 Gridding and Partitioning (Gap) approach, 199,200 GRIDIGOLPE analysis, 304305 Grid tyranny, 91, 144 GRIND, 60 Groove-binding ligands, Ground-state analog enzyme inhibitors, 720, 740-741 Growth hormone receptor X-ray crystallographic studies, 488 Growth hormone secretagogues, 671-672,675 GS 4071 structure-based design, 452 Guanidine pharmacophore points, 249 Gusperimus, 849 Hall databases, 387 Haloperidol HIV protease inhibitor, 111, 112 Halopyrimidines filtering from virtual screens, 246 Hammett constants, 11,50 Hammett equation, 12, 13, 26, 50 Hamming distance, 202 Harnmond postulate, 748 Hanes-Woolf plot, 727, 729, 731 Hansch approach, to QSAR, 26-27,30 Hansch-Fujita-Ban analysis, 31 Hansch parabolic equation, Hansch-type parameters, 54 HARPick program, 218,221 Hash code, 376,380,405 and combinatorial library design, 223 molecular fragment based, 54 Hemicholinium interatomic distance analogs, 710-711 Hemoglobin molecular modeling of crystal structure, 105, 107 target of structure-based drug design, 419-425 Hepatocyte growth factor activator inhibitor mariptase inhibitor, 269,271 Index Heroin, 849-850 HE-State index, 26 Heterochiral molecules, 782 Heteronuclear multiple bond correlation spectroscopy natural products, 518 Heteronuclear single quantum correlation spectroscopy, 512 Hexestrol, 707 Hierarchical clustering, 220, 401,405 with molecular similarityldiversity methods, 205 High density oligonucleotide arrays, 344 High performance liquid chromatography (HPLC), 586-589 for combinatorial library screening, 592-596,598, 599,607 fast, 596 for hydrophobicity determination, 16-17,23 for separation of chiral molecules, 783,788-792 High-throughput chemistry, 358,405 chemical libraries for, 367 and natural product screening, 848 High Throughput Crystallography Consortium, 418 High-throughput screening, 283 mass spectrometry applications, 591,592-596 and molecular modeling, 155 molecular similarityldiversity methods, 191 raw data points obtained by companies, 50 with virtual screening, 316 X-ray crystallography application, 472 HIN file format, 369 HINT descriptors, 56 Hint!-LogP, 389 HipHop, 60,256 Histamine antagonists molecular modeling, 143 Histidine chemical modification reagents, 755 Hit list, 380,405,411 HlV protease inhibitors, 717 binding-site molecular models, 130 comparative molecular field analysis, 153 consensus scoring study, 266 3D CoMFA, 59 de novo design, 113 force field-based scoring study, 307 homology modeling, 123 knowledge-based scoring study, 311 molecular modeling, 103-104, Hydrofinasteride, 768,770 Hydrogen bonds, 285286,365 acidic protons and sr -systems, 313 and empirical scoring, 309-310 enzyme inhibitors, 722,724 hydrophobic interactions contrasted, 319 molecular modeling, 81, 107-108 and QSAR, and structure-based design, 409 105,108,109,111,117,120,Hydrolases 122 target of structure-based drug NMR binding studies, 533, design, 449-454 559-562 Hydrolysis non-peptide peptidomimetics, 659- 660 novel lead identification by virtual screening, 320 seeding experiments, 318-319 target of structure-based drug design, 433-442 transition state analogs, 647-649 and water, 302 HIV reverse transcriptase inhibitors, 717 X-ray crystallographic studies, 488-489 H+,K+-ATPaseinhibitors, 718 HKL suite, 478 Hoechst 33258 binding to DNA, 544,547-552 Homochiral molecules, 782 HOMO (Highest Occupied Molecular Orbital) energy, ll, 14,54 Homology, 348 and X-ray crystallography,494 Homology modeling, 261-262 and docking methods, 305306 molecular modeling, 123 Homo Sapiens genome sequencing, 344 HQSAR,4 HTML (HyperText Markup Language),371,405 HUGO Gene Nomenclature Committee, 337 Human Genome Database protein classes, 262 Human genome sequencing, 344 Human serum albumin, See Serum albumin enzyme-mediated asymmetric, 805-806 Hydrophobic bond, 15 Hydrophobic effect, 178,182 Hydrophobic interactions, 50, 286,287-288 discovery of importance of, and empirical scoring, 310 enzyme inhibitors, 724 hydrogen bonding contrasted, 319 molecular modeling, 85, 108-110 and QSAR, 6,7,1519,23,52 Hydrophobicity, 16-17 determination by chromatography, 17-18,23 (S)-3-Hydroxy-y-butyrolactone, 808,810 Hydroxychloroquine,891 6R-Hydroxy-1,6-dihydropurine riboside, 752 Hydroxyethylurea, 153 R-(-)-11-Hydroxy-10-methylaporphine, 705 Hydroxymethylglutaryl-CoA (HMG-CoA) reductase inhibitors, 718,719,744-746 (f)-3-(3-Hydroxypheny1)-N-npropylpiperidine (3-PPP), 704-705 D,L-3,5-Hydroxyvalerate, 745 HYPER, 727 HypoGen, 256 Hypothetical descriptor pharmacophore, 63 Iceberg model, of hydrophobic interactions, 15 Index ICM affinity grids, 293 homology modeling, 305 ligand handling, 294 Monte Carlo minimization, 298 novel lead identification, 320 ICGovalues, 731-732 QSAR, ID3,67 iDEA, 390 IDEALIZE, 255 Identification See also Lead identification combinatorial library compounds, mass spectrometry application, 596497 mass spectrometry application, 594-596 Idoxuridine, 717 Ifosfamide, 783 Imines filtering from virtual screens, 246 Iminobiotin binding to avidin, 181, 182 Imipenem, 872-873,874 Imipramine analogs, 692-693 Immobilized enzyme inhibitors, 720 Immunophilins chemical-shift mapping of binding to, 545 FK506 binding to FKBP, 552-555 Importance sampling, 98 Incremental construction in docking, 292,295-296 in virtual screening, 262, 317-318 Indexes, 376-377,405 Indinavir, 648,659 structure-based design, 438-439,440,441 Indomethacin, 453 Inductor variables, 25-26 Influenza virus neuraminidase inhibitors, 717 InfoChem ChemReactIChemSynth database, 386 InfoChem SpresiReact database, 386 Infrared spectroscopy, 592 Inhibitors See also Enzyme inhibitors; specific inhibition targets: i.e., Dihydrofolate reductase inhibitors finding weak by virtual screening, 319 not all are drugs,408 structure of free, and structure-based design, 409 In-house databases, 387-388 Inosine monophosphate dehydrogenase consensus scoring study, 266 seeding experiments, 318-319 target of structure-based drug design, 446-447 Inosine monophosphate dehydrogenase X-ray crystallographic studies, 489 In silico screening, 191,244 See also Virtual screening Insulin-like growth factor X-ray crystallographic studies, 489 Insulin-like growth factor X-ray crystallographic studies, 489 Insulin-like growth factor receptor X-ray crystallographic studies, 489 Integrin alphaM X-ray crystallographic studies, 489 Interatomic distance variant analogs, 710-712 Intercellular adhesion molecule X-ray crystallographic studies, 489 Interferon a X-ray crystallographic studies, 490 Interferon y X-ray crystallographic studies, 490 Interleukin X-ray crystallographic studies, 490 Interleukin X-ray crystallographic studies, 490 Interleukin X-ray crystallographic studies, 490 Interleukin X-ray crystallographic studies, 490-491 Interleukin X-ray crystallographic studies, 491 Interleukin X-ray crystallographic studies, 491 Interleukin X-ray crystallographic studies, 491 Interleukin 10 X-ray crystallographic studies, 490 Interleukin 12 X-ray crystallographic studies, 490 Interleukin 13 X-ray crystallographic studies, 490 Interleukin-lp- converting enzyme (ICE) transition state analog inhibitors, 655 Interleukin receptor X-ray crystallographic studies, 490 Intermolecular forces, Interpro, 349 InterProScan, 349 Intracellular adhesion molecule (ICAM-1) target of NMR screening studies using SAR-by-NMR, " 566-567 Inventory data, 405 Inverse folding, 123-125 Inverse QSAR, Inverted keys, 405 Ionic bonds, 6, 170,365 Ion-induced dipole interactions, 173 Ipconazole, 41 Irinotecan, 849,861 Irreversible inhibitors, 755 ISISBase, 377,387 ISIS databases, 373,376-377, 387 descriptors, 192 exact match searching, 380 similarity searching, 382-383 substructure searching, 382 ISIS/Direct, 387 ISIS/Draw, 387 Isomer search, 388,405-406 Isoprenaline, 885 Isotope editing and filtering, in NMR, 545,546 Index Iterative cyclic approaches and combinatorial library design, 217 Ivermectin, 849,891,892 Japanese Patent and Trademark Documents database, 386 Jarvis-Patrick algorithm, 205,401 and combinatorial library design, 220,222 Java, 396,406,407 JG-365,121,122 Joins, 390,406 Journal content searching, 383 Kaempferol, 865 Kanomycin, 870,871 Karplus relationship, 525 Kelatorphan, 650,651 Kennard-Stone method, 65,66 Ketoconazole, 717 8-Ketodeoxycoformycin, 751 Ketol-en01tautomerization NMR spectroscopy, 527-528 Ketolides, 876 Ketones pharmacophore points, 249 Key-based similarity searching, 382-383 Key field, 406 Keys, 363 encryption, 403 molecular similarity methods, 188 Khellin, 883, 884 Kinases focused screening libraries targeting, 250 King's Clover, drugs derived from, 882 Kitz-Wilson plots, 757 K-means clustering, 401,406 K-medoids clustering, 406 K-Nearest Neighbors, 53,62-63 KNI-272, 562 Knowledge-based scoring, 264-265,307,310-312 Knowledge-bases, 352,379 Knowledge Discovery in Databases (KDD),394,406 Kohonen's Self-organizing Map method, 65,66 KOWWIN, 389 Kubinyi bilinear model, 3,31 L-365,260,856 L-370,518, 660 L-685,434 structure-based design, 439 L-732,747 structure-based design, 440, 441 L-735,525, 797 L-746,072,211 L1210 growth inhibition, 37,40-41 inhibition of DHFR, 32,34 L major DHFR, QSAR inhibition studies, 33 Laboratory information management systems, 377 p-Lactamase inhibitors, 718, 868-874 X-ray crystallographic studies, 483 p-Lactams, 868-874 Lamarckian genetic algorithm, 299 Lamivudine, 812-813,816 LASSO0 algorithm, 217 Latent inactivators, 756 Latent semantic structure indexing, 255 Laudexium, 857,858 Lead generation, 426 Lead identification, 244 focused screening libraries for, 250-252 virtual screening for novel, 320-321 Lead molecule fragment analogs, 707-710 Leaf nodes, 377 Leave-one-out cross validation, 57,64 Leave-some-out cross validation, 64 Legion, 387 Lennard-Jones potential, 285 Lentinan, 849 Leucine aminopeptidase inhibitors, 737-738 Leukocyte function-associated antigen (LFA-1) target of NMR screening studies using SAR-by-NMR, 566-567 Leukotreine A4 hydrolase X-ray crystallographic studies, 491 Leveling effect, 722 Levorphanol, 708 LH-RH antagonist, 634 LH-RH peptidomimetic analog, 640 LibEngine, 221 LiBrain, 220 Libraries, 367, 400 See also Combinatorial libraries focused screening libraries for lead identification, 250-252 for NMR screening, 574576 QSAR for rational design of, 68-69 Lidorestat structure-based design, 448-449 Ligand-based design NMR screening for, 564-577 NMR spectroscopy for, 510, 517-532 Ligand-based virtual screening, 188 Ligand design, 110-118 LigandFit sampling/scoringmethods used, 261 Ligand flexibility, See Flexible ligands Ligands macromolecule-ligand interactions, NMR spectroscopy, 510,517,535-562 non-peptidic ligands for peptide receptors, 667-674 visually assisted design, 110 Ligand strain energy, 308 LIGSITE, 291 Linear free energy relationship, 12,14 Linear interaction energy method, 120 Linear notation, 368-369,406 Linear QSAR models, 26-28 descriptor pharmacophores, 61-62 Linear regression analysis in QSAR, 8-11,50,53,67 Line-shape, in NMR, 512 and ligand dynamics, 528-531 Lineweaver-Burk plot, 727, 729, 731 Link nodes, 381,397 LINUS (Local Independent Nucleating Units of Structure), 124 Linux, 396,406,411 Index Lipinski's "rule of 5" and combinatorial library design, 214-215,216 in druglikeness screening, 245 for molecular similarityldiversity methods, 193,208 and NMR screening, 575 Lipocortin I X-ray crystallographic studies, 491 Lipophilic interactions, 286 Liquid chromatography for enantiomer separation, 787 Liquid chromatography-mass spectrometry (LC-MS), 586-591 affinity screening, 598-599 fast, 596 future developments, 607-608 gel permeation chromatography screening, 599 pulsed ultrafiltration screening, 603-606 for purification of combinatorial libraries, 592-594 structure/purity confirmation of combinatorial libraries, 594596 Liquid chromatography-NMRMS, 608 Liquids force field models for simple, 176 Liquid secondary ion mass spectrometry (LSIMS),586,587 Lisinopril, 650, 881 asymmetric synthesis, 807, 809 Literature content searching, 383 LitLink, 387 Local stereochemistry, 398 Lock-and-key hypothesis, 251, 252 deformable models, Locus maps, 140 Log 1/C, 25,27-29 Log CR, 25 Logic, in query features, 406 Logic and Heuristics Applied to Synthetic Analysis (LHASA), 379 Log MW,24 Log P chloroform-octanol, 17 chromatographic determination, 17-18,23 estimation systems, 388,389 for molecular similarityldiversity methods, 193,208 and polarity index, 26 Log Perm, 25 Log TA98,25-26 Lomerizine, 41,42 Lometrexol structure-based design, 429-430 London forces, See van der Wads forces Lopinavir, 648,659 asymmetric synthesis, 807, 809 Lorentz-Lorenz equation, 24 Lovastatin, 878-879 Low Mode Search, 292,301 LUDI, 259,295-296,315 combinatorial docking, 318 empirical scoring, 310 in molecular modeling, 112, 113 for novel lead identification, 321 LUMO (Lowest Unoccupied Molecular Orbital) energy, ll, 14,26, 54 Luteinzing hormone P X-ray crystallographic studies, 491 LwrS X-ray crystallographic structure elucidation, 494-495 LW-50020,849 LY-303366,877 LY-315920 structure-based design, 454 Lycopene positive ion APCI mass spectrum, 588 tandem mass spectrum, 591 Lymphomas, 718 Lysine chemical modification reagents, 755 MACCSSD, 259,363 in molecular modeling, 111 MACCS (Molecular Access System), 254,361362 Machine learning techniques in molecular modeling, 151 in QSAR, 62 Macrocyclic mimetics, 635-636 MACROMODEL, 94 Macromolecular structure determination, 334 NMR spectroscopy applications, 533535 Macromolecule-ligand interactions, See Protein-ligand interactions Macrophage CSF X-ray crystallographic studies, 491 MACROSEARCH, 94 Magnetization transfer NMR, 568-570 Ma huang,884-885 Mandelate racemase inhibitors, 762,763 Manhattan distance, 68 Marcaine classical resolution, 795 Marijuana, 853 Marine source drugs antiasthma, 886 anticancer, 867-868 Markup languages, 371-372,405 Markush feature, 381 Markush structures, 367,368, 373,406 MARPAT, 385 Masoprocal, 849 Mass spectrometry, 583-592 affinity capillary electrophore* sis-mass spectrometry, 599-600 affinity chromatography-mass spectrometry, 598-599 bioaffinity screening using electrospray FTICR MS, 601- 603 encoding and identification of combinatorial compounds and natural product extracts, 596-597 frontal affinity chromatography-mass spectrometry, 601 future directions, 607-608 gel permeation chromatography-mass spectrometry, 599 LC-MS purification of combinatorial libraries, 592-594 MS-based screening, 597-598 pulsed ultrafiltration-mass spectrometry, 603-606 solid phase mass spectrometric screening, 606-607 Index structurelpurity confirmation of combinatorial compounds, 594-596 types of mass spectrometers, 585 Material Safety Data Sheets database, 386 Material Safety Data Sheet searching, 384 Matriptase virtual screening of inhibitors, 269-271,272 Matrix-assisted laser desorption ionization (MALDI), 586, 596,606-607 Matrix metalloprotease inhibitors, 227,555457 chemical-shift mapping of binding to, 545 target of NMR screening studies using SAR-by-NMR, 566 target of structure-based drug design, 443-445 transition state analog inhibitors, 651-652 virtual screening, 315 Maximum Auto-Cross Correlation (MACC),202 Maxmin approach, 208 May apple, drugs derived from, 865 Maybridge catalog, 385 MCDOCK Monte Carlo simulated annealing, 297 MD Docking (MDD) algorithm, 298 MDL Information Systems, Inc databases, 386-387 Mechanism-based enzyme inhibitors, 759-760,764-771 Mefloquine, 889-890 artemisinin potentiates, 887-888 Meglumine, 796-797 Melagatran structure-based design, 442, 444 a-Melanotropin conformationally restricted peptidomimetics, 637 Melatonin analogs, 693 antagonists, 211-212 Melilotus officinalis (ribbed meMot), 882 Melittin molecular modeling, 124 Members, of Rgroups, 368,406 Membrane-bound drug targets, 351 Membrane-bound proteins molecular modeling, 154 NMR structural determination, 535 Membrane-bound receptors, Mepartrican, 849 Meperidine, 708,851 rigid analog, 696 6-Mercaptopurine, 717 Mercury search program, 387 Merged Markush Service, 386 Merimepodip structure-based design, 447 MERLIN, 39,386 Messenger RNA, See mRNA Metabolism See also Absorption, distribution, metabolism, and excretion (ADME) enantiomers, 786-787 Metabolism databases, 385,386 Metabolism screening, 591 pulsed ultrafiltration application, 605 Metabolite database, 386 Metadata, 375,376,406 Meta-layer searching, 395 Metallopeptidase inhibitors transition state analogs, 649-652 Metamitron, 42 Metazocine, 708 Metconazole, 41,42 Methadone, 708 Methamphetamine ring substitution analogs, 704 R-Methanandamide, 852,853 Methanol force field models for, 176 Methicillin, 869, 870, 871 Methionine:adenosyl transferase, 148 Methionine hydrochloride nonclassical resolution, 803 Methods in Organic Synthesis database, 385 Methotrexate, 717, 718, 749 interaction with dihydrofolate reductase, 120 structure-based design, 425 N-Methyl-acetemide force field models for, 176 2-Methyl-1,4-benzenediol allergenicity prediction, 833 a-Methyldopa, 785 5,lO-Methylene-tetrahydrofolate, 426, 427 Methyl group roulette, 700 Methylphenidate (Ritalin) classical resolution, 793-794 nonclassical resolution, 801 Metocurine, 856,857 Metoprolol renal clearance, 38 Metropolis algorithm, 94,98 D,L-Mevalonate, 745 Mevastatin, 744-745 MHC I receptor homology modeling, 123 molecular modeling, 117 Michaelis-Menten constants, 725-728 use in QSAR, , Michaelis-Menten kinetics, 725-728 Microarray chips, 334,344-345 Microbial secondary metabolites, 848 Micropatent, 386 Microsoft Access, 373 Middle tier, 392, 406-407 Miglitol, 849 Milbemycins, 891,892 L-Mimosine analogs, 690 MIMUMBA, 255 Mini-fingerprints, 255 Minimum topological difference (MTD) method, 4,147 Mining minima algorithm, 292, 299-300 Mitogen-activated protein kinase target of structure-based drug design, 456-459 Mivacurium, 857,859 Mixtures, 367-368 Mizoribine, 849 MK-329,855 MK-383,213 MK-499,814-815,818 MK-0677,671,674 MK-678,657 ML-236B, 879 MLPHARE, 478 MM-25 structure-based design, 423, 424 Index structure-based design, 423, 424 MM2 force field, 80, 307 MM3 force field, 80, 118 MM-PBSA method, 315 Modeling, See Molecular modeling Model mining, Model receptor sites, 149-150 Molar refraction, 24, 54 MOLCONN-Z, 55,192,389 Molecular Biolom Database Collection, 345 Molecular comparisons, 138-142 Molecular connectivity, 192,407 estimation systems, 388 in QSAR, 26,55,56,61 Molecular docking methods, See Docking methods Molecular dynamic simulations See also Monte Carlo simulations barrier crossing, 98 with docking methods, 292, 298 and force field-based scoring, 308 hydrogen bonds, 107 in molecular modeling, 85, 93, 95-100,116-117,142 and non-Boltzmann sampling, 100 protein flexibility, 301-302 statistical mechanical, 94,95 of temperature, pressure, and volume, 96 thermodynamic cycle integration, 99 in virtual screening, 263 water's role in docking, 302303 Molecular eigenvalues, 54 Molecular electrostatic potential, 102 Molecular extensions, 130-131 Molecular field descriptors, 54, 55-57 Molecular Graphics and Modeling Society, 360 Molecular holograms, 54 Molecular mechanics, 79-100 force fields, 174-177 Molecular modeling, 77-79, 153-154,358 affinity calculation, 118-122 and bioinformatics, 351 common patterns, 142-150 conformational analysis, 87, 93-94 and electrostatic interactions, 81-85 and force fields, 79-81 known receptors, 103-127 ligand design, 110-118 molecular comparisons, 138-142 and molecular mechanics, 79-100 pharmacophore versus binding site models, 127-135 potential surfaces, 85-89 protein structure prediction, 122-127 and QSAR, and quantum mechanics, 100-103 similarity searching, 135-138 site characterization, 105-110 and statistical mechanics, 94-95 in structure-based design, 419, 420 systematic search, 89-94, 116 unknown receptors, 127-153 and virtual screening, 244 Molecular multiple moments, 54 Molecular property visualization, 137-138 Molecular recognition, 283 and hydrophobic interactions, 15 physical basis of, 284-289 Molecular replacement, 477 Molecular sequence alignment, 353 Molecular sequence analysis bioinformatics for, 335-336 Molecular shape analysis, 53 Molecular shape descriptors, 54 Molecular similarityldiversity methods, 54, 188-190 analysis and selection methods, 203-209 combinatorial library design, 190,214-228 descriptors for, 191-203 example applications, 228-237 future directions, 237 and molecular modeling, 135-138 virtual screening by, 188,190, 209-214 Molecular structure descriptors in QSAR, 26 Molecular targets, See Drug targets Molecular weight for molecular similarityldiversity methods, 193,208 and QSAR, 24-25 MOLGEO, 255 Molinspiration, 390 MOLPAT, 110-111 Monasus ruber, 879 Monoamine oxidase inhibitors, 718 Monobactams, 873 Monocolin K, 879 Monomer Toolkit, 377378 Monte Carlo simulated annealing and combinatorial library design, 218 with docking methods, 292, 297 with virtual screening, 263 Monte Carlo simulations See also Molecular dynamic simulations barrier crossing, 98 and combinatorial library design, 217 de novo design, 113 with docking methods, 292, 297-298 in molecular modeling, 85, 86, 93,96-99,116-117,142 and non-Boltzmann sampling, 100 statistical mechanical, 94,95 thermodynamic cycle integration, 99 in virtual screening, 263 Moore's Law, 393 Morgan algorithm, 378,407 Morphiceptin, 144,145 Morphinans, 850 Morphine, 634 ecological function, 848 fragment analogs, 707-708 Morphine alkaloids, 849-851 Mosflm/CCP4,478 Most descriptive compound (MDC) method, 207-208 mRNA and expression profiling, 340-341 MSDRLICSIS, 361 Index MS-MS, See Tandem mass spectrometry (MS-MS) Mulliken population analysis, 101-102 MULTICASE SAR method toxicity prediction application, 828- 843 Multidimensional databases, 390,407 Multidimensional NMR spectroscopy, 512-514 Multidimensional scaling, 201 Multidimensional scoring, 291 Multilevel chemical compatibility, 249 Multiple-copy simultaneous search methods (MCSS), 298 Multiple isomorphous replacement (MIR) phasing, 477 Multiple regression analysis in QSAR, 8-11,50,52,53 Multisubstrate analog enzyme inhibitors, 720, 741-748 Multi-tier architecture, 392, 407 Multiwavelength anomalous diffraction (MAD) phasing, 474,477-478 Munich Information Center for Protein Sequences (MIPS), 335 Muscarinic receptors distance range matrices, 136 stereoisomer analogs, 705-706 Mutation in genetic algorithms, 87,88 MVIIA (Ziconotide) NMR spectroscopy, 518-523, 526,534 MVT-101,103-104,105,117 Mycophenolate mofetil, 849 Mycophenolic acid structure-based design, 446-447 Myoglobin, 419 Nabilone, 853 Nadolol renal clearance, 38 Naftifine, 717 Na+,K+-ATPaseinhibitors, 718 Nalorphine, 850 Naloxone, 850 NAPFMLERT, 597 Naproxen classical resolution, 794-795 enzyme-mediated asymmetric synthesis, 805 Narwedine, 802,803 National Cancer Institute database, 222,254,385486,387 National Center for Biotechnology Information (NCBI), 335 sequence databases, 387 National Toxicology Program, 246,829 Natural product mimetics, 636 Natural products antiasthma drug leads, 883-886 antibiotics drug leads, 868-878 anticancer drug leads, 858- 868 antiparasitic drug leads, 886-891 cardiovascular drug leads, 878-883 CNS drug leads, 849-856 drugs derived from, 1990-2000,849 extract encoding and identification, 596-597 leads for new drugs, 847-894 neuromuscular blocking drug leads, 856-858 NMR structure elucidation, 517-518 Natural products databases, 387, 597 Nearest neighbors methods, 53, 62-63,67 Neighborhood behavior, 211 Nelfinavir, 648 asymmetric synthesis, 817-818 structure-based design, 440, 442 Neomycin, 870,871 Netropsin binding perturbations, 544 Neu5Ac2en structure-based design, 451 Neural networks, See Artificial neural networks Neuraminidase inhibitors, 717 flexible docking studies, 265 PMF function application, 314 Screenscore application, 319 target of structure-based drug design, 450-452 X-ray crystallographic studies [int B virus], 491 Neuroleptics molecular modeling, 150 Neuromuscular drugs natural products as leads, 856-858 Neuropeptide Y X-ray crystallographic studies, 492 Neuropeptide Y inhibitors, 671, 673,674 Neutral endopeptidase (NEP), 650-651 Nitric oxide synthase, 736 Nitric oxide synthase inhibitor, 738-739 Nivalin, 892 NK receptor antagonists, 669-670,672 NMR, See Nuclear Magnetic Resonance (NMR) spectrosCOPY NMR timescale, 537 NN-703,671,675 NOE, See Nuclear Overhauser effects (NOE),in NMR Nolatrexed, 428 Non-Boltzmann sampling, 100 Nonclassical bioisosteres, 690-694 Nonclassical resolution, of chiral molecules, 799-804 Noncompetitive inhibitors, 730-731 Noncovalent bonds, 6,170 energy components for intermolecular drug-target binding, 171-174 Noncovalently binding enzyme inhibitors, 720-754 Nonisosteric bioanalogs, 689-694 Nonlinear QSAR models, 28-29 descriptor pharmacophores, 62- 63 Nonlinear regression, 67 Non-overlapping mapping, 398 Non-peptide peptidomimetics, 636,657-674 Nonpolar interactions, See van der Wads forces Nonstructural chemical data, 373 Norapomorphine alkyl chain homologation analogs, 701 Index Norfloxacin, 41,42 Norstatine, 652 Norvir structure-based design, 438, 440 Nostructure, 410 NOT logical operator, 406 NPS 1407,812,815 Nuclear hormone receptors focused screening libraries targeting, 250 Nuclear Magnetic Resonance (NMR) imaging, 510 Nuclear Magnetic Resonance (NMR) screening methods, 510,562-577 capacity issues, 190 Nuclear Magnetic Resonance (NMR) spectroscopy, 351, 507-514,592 See also SARby-NMR approach applications, 516417 chemical shift mapping, 543-545 instrumentation, 514-516 with LC-MS, 608 ligand-based design, 510, 517532 macromolecule-ligand interactions, 510, 517, 535-562 metabolic, 510 and molecular modeling, 78 multidimensional, 512-514 for pharmacophore modeling, 531-532 receptor-based design, 510, 532-562 in structure-based drug design, 419,516-517 and structure-based library design, 225 structure determination of bioactive peptides, 517-518 structure elucidation of natural products, 517-518 and virtual screening, 244 Nuclear Magnetic Resonance (NMR)titrations, 545 Nuclear Overhauser effect (NOE)pumping, 573 Nuclear Overhauser effects (NOE),in NMR, 511,512 for conformational analysis, 525 and distance range matrix, 136 for macromolecular structure determination, 533 and NMR screening, 571-573 NOE docking, 545-546 transferred NOE technique, 532 Nucleic acid receptors, Nucleic acids See also DNA, RNA biochemical force fields, 175-176 NMR structural determination, 535 Nucleotide intercalation, 183 (graphics program), 478 Object-oriented language, 407 Object relational database, 407 Ocreotide, 657 Octanol/water partitioning system, 16-17 OLAP (OnLine Analytical Processing), 390,408 Oleandomycin, 870,871 OLTP (OnLine Transaction Processing), 390,408 Omapatrilat, 651 OMEGA, 255 Ondanetron nonclassical resolution, 802 OpenBabel, 372 Open Molecule Foundation, 360 Open reading frames housing in DNA databases, 338 Opium poppy, 848,849 Optimization approaches for combinatorial library design, 217-220 OptiSim method, 207 and combinatorial library design, 220 Oracle, 373 Organic structure databases, 385 Organoarsenical agents, 717 Orientation map ( O W ) ,131, 144,146 Oriented-substituent pharmacophores, 224 Orlistat, 848,849 OR logical operator, 406 use in molecular similarity1 diversity methods, 194 Ornithine decarboxylase inhibitors, 717, 766, 768, 769 Oseltamivir, 452, 717 OSPPREYS (Oriented-substituent Pharmacophore PRopErtY Space), 199,224 Overlapping mapping, 398 OWFEG (one window free energy grid) method, 308,315 Oxidation enzyme-mediated asymmetric, 806 Oxidoreductases target of structure-based drug design, 445-449 Oxprenolol renal clearance, 38 Oxytetracyclin,870 P carinii DHFR, QSAR inhibition studies, 32-33 Pacific yew, paclitaxel from, 861-862 Paclitaxel, 843,848,861-863 Pairwise interactions, 79-80 PALLAS System, 389 Paluther, 887 Pamaquine, 888-889 Pancreatic polypeptide molecular modeling of avian, 124 Papain QSAR studies, transition state analog inhibitors, 654 Papaver somniferum (opium POPPY), 848,849 Parallel chemistry, 283 Parallel library, 214 Parallel processing, 408 Parathion, 774 Parathyroid hormone X-ray crystallographic studies, 492 Parent structure, 368,404 Pareto optimality, 220 Partial charge, 366,373 Partition coefficients, 16-17, 54 Partition function, 94-95 Partitioning algorithms, 67 PASS, 291,390 Patent Citations Index, 386 Patent databases, 386 Patent searching, 383-384 Pathways, 495-496 X-ray crystallographic analysis, 495-496 Pattern recognition, 408 and cluster analysis, 401 with QSAR, 53 PC cluster computing, 283-284 PCModels, 386 PD-119229 structure-based design, 460-461 PDB file format, 369 PDGF beta X-ray crystallographic studies, 492 Peak intensities, in NMR, 512 Peldesine structure-based design, 460 Pemetrexed structure-based design, 429-430 Penicillins, 717,868-870 preventing bacterial degradation, 718 Penicillipepsin inhibitors molecular modeling, 116 Penicillium brevicompactum, 879 Penicillium chrysogenum, 869 Penicillium citrinium, 879 Pentostatin, 717, 750-751,849 PeptiCLEC-TR, 804 Peptide backbone mimetics, 636, 644- 645 Peptide bond isosteres, 644-646 Peptides, 634 biochemical force fields, 175-176 NMR structural determination of bioactive, 517-518 non-Boltzmann sampling of helical transitions, 100 Peptidomimetics, 128-129, 633-634 classification, 634-636 conformationally restricted peptides, 636- 643 future directions, 674-677 molecular modeling, 154 non-peptide, 636,657-674 peptide bond isosteres, 644-646 protease inhibitors, 646-655 speeding up research, 655-657 template mimetics, 643-644 Peramivir structure-based design, 452 Personal chemical databases, 387-388 Petabyte, 408 Pethidine, 851 PETRG 390 Petrosia contignata, 886 Pfam, 349 Pharmacophore keys, 376, 408-409 Pharmacophore mapping, 255 Pharmacophore point filters, 196,249-250 Pharmacophores, 368 with BCUT descriptors, 223-224 binding site models contrasted, 127-135 defined, 252-253,408 descriptor, for QSAR,60-63 3D searching, 366-367 in molecular modeling, 110 for molecular similarityldiversity methods, 194-201, 204-206 NMR-based modeling, 531-532 NMR spectroscopy-based modeling, 531532 oriented-substituent, 224 site-based, 235-237 virtual screening, 252-260 PharmPrint method, 223 Phase problem in X-ray crystallography, 476-478 Phencyclidine rigid analogs, 696-697 P-Phenethylamines, 697-698 Phenols DNA synthesis inhibition by, 40 growth inhibition by, 38, 40-41 Phenylacetic acids ionization of substituted, 12-14 PhenylethanolamineN-methyltransferase (PNMT) inhibitors, 733-734,740 (R,S)-a-Phenylglycidate,762 (S)-a-Phenylglycidate, 762 N'-(R-Pheny1)sulfanilamides antibacterial activity, 10 Phosphatidylcholine monolayers penetration by ROH, 27 Phosphocholine docking to antibody McPC603, 298 Phosphodiesterases alignment of catalytic domains in gene family, 349 Phos~holi~ase * - A2 homology modeling, 123 target of structure-based drug design, 453-454 X-ray crystallographic studies, 492 Phosphonoacetate, 740 N-Phosphonoacetyl-L-aspartate (PALA), 743-744 Phosphonoformate, 740 2-(Phosphonomethoxy)ethylguanidines chain branching analogs, 702 (R)-9-[2-(Phosphonomethoxy)propylladenine (RPMPA), 818-819 Phosphoryl transferases target of structure-based drug design, 456-4561 Phylogenomics, 347-349 Physicochemical descriptors, 54 estimation systems, 389 for molecular similarityldiversity methods, 193 for virtual screening, 255 Physicochemical properties, 373, 409 Physostigmine, 774 Picornaviruses target of structure-based drug design, 454-456 Picovir structure-based design, 455-456 Pigeon liver DHFR, QSAR inhibition studies, 31-32 Pipecolic acid, 805 Pirlindole chromatographic separation, 788-789,790 Pit viper, drugs derived from, 881 Pivoting data, 409 Plant natural products, 848,893 Plant secondary metabolites, 848 Pleconaril structure-based design, 455-456 PLOGP, 389 PLP function, 266,309 consensus scoring, 320 hydrophobic interactions, 319 performance in structure prediction, 314 seeding experiments, 319 PLUMS, 225 p38 MAP kinase consensus scoring study, 266 seeding experiments, 318319 PMF function, 265,311,312 performance in structure prediction, 314 seeding experiments, 319 PMML (Predictive Model Markup Language), 405 PNU-107859 NMR binding studies, 555 PNU-140690,659,812,813 PNU-142372 NMR binding studies, 555-556 POCKET, 259 Podophyllin drugs derived from, 865-867 Podophyllotoxin, 849,865-866 Podophyllum emodi, 865 Podophyllum peltatum (May apple), 865 Poisson-Boltzmann equation, 83,84 Polarity index, 26 Polarizability, 85 Polarizability index, 11 Polarization energy, 173 Polar surface area in druglikeness screening, 245 Policosanol, 849 Pomona College Medchem, 385 Potassium channel shaker X-ray crystallographic studies, 492 Potential smoothing, 86 Potential surfaces, 85-89 PPAR y X-ray crystallographic studies, 492 Pralnacasan structure-based design, 443, 446 Pravastatin, 879,880 Preliminary screening, 111-112 Pressure molecular dynamic simulation, 96 Primaquine, 889 Principal components analysis with molecular similarityldiversity methods, 192,201 in QSAR, 15 Principal components regression, 53 Prindolol renal clearance, 38 Prinomastat structure-based design, 444, 446 PRINTS, 335,349 Privileged structures in molecular similarity/diversity methods, 209 template mimetics, 644 in virtual screening, 251-252 PROBE, 126 PROCHECK program, 478 ProDock affinity grids, 293 Monte Carlo minimization, 298 ProDom, 349 Proflavin thermodynamics of binding to DNA, 183 Progesterone receptor antibody FAB fragment, 128 X-ray crystallographic studies, 492 PROGOL, 67 Project Library, 387 Prolactin receptor X-ray crystallographic studies, 492 PRO-LEADS, 299 assessment, 303,304 flexible ligands, 263 PRO-LIGAND genetic algorithm with, 89 Pronethalol, 881 Propargylglycine, 719-720 Property-based design, 234-235 Propranolol, 881-882 enantiomers, 786 enzyme-mediated asymmetric synthesis, 805-806 renal clearance, 38 N10-Propynyl-5,8-dideazafolate, 426,427 Proresid, 866-867 PRO-SELECT combinatorial docking, 318 PROSITE, 348349 Prostacyclin, 762-763 Prostaglandin synthase inhibitors, 718,762-763,764 Protaxols, 863 Protease inhibitors See also HIV protease inhibitors affmity labels, 762 QSAR studies, structural genomics, 353 target of structure-based drug design, 432-445 transition state analogs, 646-655 Protein classes, 262 Protein Data Bank, 110,353 sequence database, 387 X-ray crystallography application, 478-479 Protein Database and virtual screening, 261-262 Protein families targeting in libraries for virtual screening, 251 Protein interactions, 334 Protein-ligand docking programs, 292 Protein-ligand docking techniques, 262-264 Protein-ligand interactions, 284-289,322 NMR spectroscopy, 510,517, 535-562 QSAR studies, scoring, 264-267 scoring in virtual screening, 264-266 Protein-protein interactions, 634 characterizing,637 Proteins See also Macromolecular structure determination binding and chirality, 786-787 flexibility and docking, 300-302,322 phylogenetic profiling, 347-348 Protein structures prediction, 122-127 in structure-based virtual screening, 261-262 X-ray crystallographic analysis, 496 Proteome, 352 Proteomics 409 Pseudoirreversible enzyme inhibitors, 771-774 Pseudomonas acidophila, 873 Pseudopeptides,635-636 isosteres replacing peptide backbone groups, 646 Pseudoracemate, 799-800,801 Pseudo-receptor models, 261 PSI-BLAST, 335,347 X-ray crystallography application, 481 Pulsed ultrafdtration-mass spectrometry, 603-606 Index Purine biosynthesis inhibitors, 752 Purine nucleoside phosphorylase target of structure-based drug design, 459-461 Purine ribonucleoside, 751-752 Purity verification as bottleneck in drug discovery, 592 LC-MS-based purification, 592-594 mass spectrometry application, 594-596 Pyridoxal phosphate-dependent enzymes mechanism-based inhibitors, 765-768 Pyrimidine biosynthesis inhibitors, 752 Pyrrolinones peptide-like side chains, 635, 642 Pyruvate dehydrogenase inhibitors, 717 Pyruvate kinase, 764 Qinghaosu (artemisinin), 886 Q-jumping MD, 298 QSAR, See Quantitative structure-activity relationships QSAR and Modeling Society, 360 QSDock, 295 QSiAR, 53,60 QSPR, See Quantitative structure-property relationships Quadratic shape descriptors, 295 Quadrupole time-of-flight hybrid (QqTOF) mass spectrometry, 585,607 QUANTA, 258 Quantitative strudure-activity relationships, 1-4,49-52, 358 See also Comparative quantitative structure-activity relationships; 3D quantitative structure-activity relationships applications with interactions at cellular level, 37-38 applications with interactions in vivo, applications with isolated receptor interactions, 30-37 2D, 52,53 data mining, 66-67 defined, 409 descriptor pharmacophore concept, 60-63 and docking methods, 304-305 Free-Wilson approach, 4, 29-30 guiding principals for safe, 66 and library design, 68-69 linear models, 26-28, 51, 61-62 model validation, 63-66 and molecular similarity/diversity methods, 194 multiple descriptors of molecular structure, 54-58 nonlinear models, 28-29, 51, 62-63 parameters used, 11-26 problems with Q2, 64-65 receptor theory development, 4-7 standard table, 51 in structure-based design, 419 substituent constants for, 19-23 taxonomy of approaches, 52-54 tools and techniques of, 7-11 training and test set selection, 65-66 variable selection 60-63 as virtual screening tool, 66-69 Quantitative structure-property relationships, 53 and molecular similarityldiversity methods, 194 Quantum chemical indices, 11, 14-15,54 Quantum mechanics, 100-103 Quercetin, 865 Query features, 381 logical operators, 406 Query structures, 368 mapping, 380 Quinacrine, 889,890-891 Quinine, 888-891 Quinolines, 889-890 Quinupristin, 876-877 Quisqualic acid, 694 Qxp Monte Carlo minimization, 298 Rabbits narcosis induction by ROH, 27 Racemates, 782 types of, 799-801 Racemization, 783-784 Radiation damage in electron cryomicroscopy, 612-613,614-615,616 Raffinate, 791 Ramachandran plot, 92 and conformational mimicry, 141 Ramipril, 746 Ramiprilat, 746-747 Random searching in virtual screening, 263 Rapamycin, 848 binding to FKBP, 552,554 Rapid, reversible enzyme inhibitors, 720,728-734 Rapid sequence screening, 334 Rare gas interactions, 174 Ras-farnesyltransferase inhibitors non-peptide peptidomimetics, 665-667,668,669 template mimetics, 643,645 Rats ataxia induction by ROH, 29 liver DHFR, QSAR inhibition studies, 34 REACCS, 398 reaction searching using, 383 Reactant-biased, product-based (RBPB) algorithm, 215,216, 219 Reacting centers, 366,383, 398, 409 Reaction Browser~Web,387 Reaction databases, 386 Reaction field theory, 83 Reaction indexing, 383 Reaction Package, 386 Reactions, See Chemical reactions Reaction scheme, 409 Reagent Selector, 387,391-392 RECAP (Retrosynthetic Combinatorial Analysis Procedure), 249 Receptor-based design NMR spectroscopy for, 510, 532-562 pharmacophore generation, 259 Receptor-based 3D QSAR, 304 Receptor-ligand complexes, 78 Receptor-ligand mimetics, 636 Receptor mapping, 148-149 Index Receptor-relevant subspace, 204, 222 Receptor theory, 4-7 Reciprocal nearest neighbor, 220 Recursive partitioning, 247-248 Red clover extract LC-MS mass spectrum, 589, 590 Reduction enzyme-mediated asymmetric, 806 Refining, search queries, 409 REFMAC, 478 Registration, of chemical information, 377-379 Registry number, 378-379,409 Relational databases, 363, 373, 409 Relative diversitylsimilarity, 209 Relaxation parameters, in NMR, 511,512 changes on binding, 536-537 and ligand dynamics, 528-531 and NMR screening, 571-573 in receptor-based design, 534 Relenza structure-based design, 451 Relibase, 315 Reminyl, 892 Renin inhibitors, 432 molecular modeling, 123, 153 transition state analogs, 647 REOS filtering tool, 225 RESEARCH Monte Carlo simulated annealing, 297 Resiniferatoxin, 854 Restrained electrostatic potential, 102 Result set, 409,411 Retigotine, 783 Retinoic acid docking and homology modeling, 305 stereoisomer analogs, 707 X-ray crystallographic studies, 492-493 Retinoid X receptor X-ray crystallographic studies, 493 Retrosynthetic analysis, 409 Retrothiorphan, 650, 651 Reverse nuclear Overhauser effects pumping, 573 Reversible enzyme inhibitors, 720 RGD peptide sequence mimics, 129,643,645,662-665 Rgroups, 368,373,397,405, 409-410 and combinatorial library design, 221 Rhinoviruses comparative molecular field analysis, 153 molecular modeling of antiviral binding to HRV-14, 120, 122 target of structure-based drug design, 454-456 Rhodopeptin template mimetics, 644,645 Ribbed melilot, drugs derived from, 882 Rifamycin, 870,872 Rigid analogs, 694-699 Rigid body rotations, in molecular modeling, 90-91 Rigid docking, 262-263,293 Rigid geometry approximation, in molecular modeling, 89 Ring-position isomer analogs, 699-704 Rings in druglikeness screening, 245 molecular comparisons, 139 in molecular modeling, 91 Ring-size change analogs, 699-704 Ritalin classical resolution, 793-794 nonclassical resolution, 801 Ritonavir, 648,659 asymmetric synthesis, 807-808,809 structure-based design, 438, 440 Rivastigmine, 774 structure-based design, 449-450 RNA molecular modeling, 154 NMR structural determination, 535 RNA polymerase inhibitors, 717 Ro-31-8959,121 Ro-32-7315, 652, 653 Ro-46-2005,673, 676 ROCS, 256,259 shape-based superposition, 260 Roll-up, 410 Root structure, 368,404,410 ROSDAL notation, 368,410 Rosuvastatin, 848,880-881 Rosy periwinkle, vinca alkaloids from, 858 Rotatable bonds in druglikeness screening, 245 in molecular modeling, 90-91 Royal Society of Chemistry Chemical Information Group, 360 RPR109353,211 R,S descriptors, for chiral molecules, 365, 783 RS Discovery System, 377,385 RSR-13,422,423 RSR-56,422,423 RTECS, 246 RUBICON, 386 virtual screening application, 254 "Rule of 5," See Lipinski's "rule of 5" S-37435,675 Saccharomyces cerevisiae genome sequencing, 344 Saccharopolyspora erythraea, 874 Salbutamol, 885,886 Salmeterol, 885,886 S-Salmeterol enzyme-mediated reduction, 806,808 Salmonella mutagenicity prediction, 829, 831-832,840,842-843 Salt bridges, 285 and virtual screening, 272 Salts definitions, 376 Salts search, 388 Sampatrilat, 651 Saquinavir, 648,659, 717 structure-based design, 435-437,440 SAR-by-NMR approach, 508, 516 in NMR screening, 564468, 576 Sarin, 774 Saturated rings analogs based on substitution of aromatic for saturated ring; or the converse, 699-704 Saturation diversity approach, 223 Index Saturation transfer difference NMR, 568-570 SB203580 structure-based design, 457 SB209670,211,675 SB214857,213 SB242253 structure-based design, 458 Scaled particle theory, 84 SCH 47307,667 SCH 57939,808,810 SCH 66701,667 Schrodinger equation, 79,363 Scientific and Technical Information Network, 597 SciFinder, 385 SCOP, 353 X-ray crystallography application, 494 ScoreDock assessment, 303 Scoring functions, 261,264-266, 306-312,322 assessment, 312-315 basic concepts, 289-290 and molecular modeling, 115-116 overview of, 307 penalty terms, 313 Screening See also Combinatorial chemistry; Highthroughput screening; Virtual screening mass spectrometry-based, 597-598 solid phase mass spectrometric, 606-607 Screenscore, 319 Sculpt, 387 SEAL, 316,321 Search queries, 368 p-Secretase inhibitors transition state analogs, 649 Selector, 387 SELECT program, 218-219,221 SELECT statement, 404,406 Self-organizingMap method, 65,66 Semirigid analogs, 694-699 Sequence assembly, 342 Sequence comparison, 334 bioinformatics for, 346-347, 352-353 Sequence databases, 387 Sequences, 363364 Sequential docking, 317 Sequential simplex strategy, 11 Serevent, 806 Serial analysis of gene expression (SAGE), 344 Serine chemical modification reagents, 755 Serine peptidase inhibitors transition state analogs, 652-655 Serine protease inhibitors affinity labels, 762 common structural motifs, 494 QSAR studies, Serotonin conformationally restricted analog, 696 ring position analogs, 703-704 Serotransferrin p X-ray crystallographic studies, 493 Serum albumin binding of enantiomers, 786 mass-spectrometric binding assay screening, 604 target of NMR screening studies, 567-568,573 SFCHECK program, 478 Sgroups, 373,397,405,410 Shake and Bake, 477,478 SHAPES NMR screening libraries, 575 and SAR-by-NMR,568 SHARP, 478 SHELX, 478 Sialic acid, 450-451 Sialidase genetic algorithm study of docking, 88-89 Sickle-cell anemia, 419-425 Side chains of known drugs, and druglikeness screening, 248-249 peptide-like, 635, 642 Signature molecular similarity methods, 188 Similarity searching, 379, 382483,410 See also Molecular similarityldiversity methods in molecular modeling, 135-138 and QSAR, 67-68 SQL for, 395 Simulated annealing See also Monte Carlo simulated annealing and combinatorial library design, 217 with FOCUS-2D method, 68 hydrogen bonds, 107 in molecular similarityldiversity methods, 205 with QSAR, 53,61 in virtual screening, 263 Simulated moving bed chromatography for enantiomer separation, 787,789-793,821 Simvastatin, 719, 744,879,880 Single nucleotide polymorphism (SNP) maps, 338-340 Single-wavelength anomalous diffraction phasing (SAD), 477-478 Sirolimus, 848,849 Site-based pharmacophores, 235-237 Size-exclusion chromatography, 599 Sizofilan, 849 SKI? 107260,663 SLIDE anchor and grow algorithm, 296 combinatorial docking, 317 explicit water molecules, 302 geometric/combinatorial search, 295 ligand handling, 293 protein flexibility, 301 receptor representation in, 291 SLN (Sybyl Line Notation), 369, 410 Slow-binding enzyme inhibitors, 720,734-740,749 Slow-tight-binding enzyme inhibitors, 720, 734-740 SMART, 349 functional group filters, 246 SMILES notation, 254,410 and canonical renumbering, 378 described, 368-369,371 use with comparative QSAR, 39 SmoG, 311 SN-6999,544 Snowdrops, drugs derived from, 892 Index SNX-111,851-852 SOCRATES, 361 Sodium cromoglycate, 883-884 Soergel distance, 68 Solid Phase Synthesis database, 385 Solution molecular dynamics, 528 Solvation effects and docking scoring functions, 307,308,310 drug-receptor complexes, 177-179 molecular modeling, 83-85 SOLVE, 478 Somatostatin conformationally restricted peptidomimetics, 129,637, 638 receptor agonists found through combinatorial chemistry, 657 template mimetics, 643-644, 645 Sorangium cellulosum, epothilones from, 864 SPC model, 175 Specific structure, 368,403 Sphere coloring, 296 Sphere-exclusion, 207 Spindle poisons, 867 Spin-label NMR screening, 573-574 SPLICE, 89,113 Spongothymidine, 867-868 Spongouridine, 867-868 SPRESI, 254 SPRESI'95,385 SQL (Structured Query Language), 395,410 SR-48968,670 SR-120107A,670,673 SRS (Sequence Retrieval System), 335 Standardization bioinformatics, 337 Star schema, 390,391,410 Statins, 719,848 multisubstrate analogs, 744-746 Statistical mechanics, 94-95 Stem cell factor X-ray crystallographic studies, 493 Stereoisomer analogs, 704-707 Stereoisomers, 365-366, 783-785 Stereoplex, 387 Stereoselective synthesis, See Asymmetric synthesis Steric parameters in QSAR, 23-25,52 STERIMOL parameters, 24, 50 Steroid 5a-reductase inhibitors, 717,768-770 QSAR studies, 37-38 Steroids affinity for binding proteins, 147 biosynthesis inhibition, 770 STN Express, 385 STN International, 385 STO-3G basis set, 175 Storage, of chemical information, 373-377 Streptavidin free energies of binding, 286 genetic algorithm study of biotin docking to, 89 interaction with biotin, 181-183 Streptogramins, 876-877 Streptomyces, 876,891 Streptomyces cattleya, 872 Streptomyces clavuligerus, 869 Streptomyces erythreus, 874 Streptomyces griseus, 869 Streptomyces venezuelae, 870 Streptomycin, 869-870 Stromelysin flexible docking studies, 265 NMR binding studies, 555-557 target of NMR screening studies using SAR-by-NMR, 566 target of structure-based drug design, 443-444 Structural data mining, 410 Structural frameworks of known drugs and druglikeness screening, 248-249 Structural genomics, 283 and bioinformatics, 352354 and X-ray crystallography, 481,494-496 Structural homology, See Homology Structural similarity, 255 Structure-activity relationships See also Quantitative structure-activity relationships and data mining, 66-67 and molecular modeling, 134 nonlinear, 62 pharmacophore searching for generating, 255,272-273 and toxicity prediction, 828- 843 Structure-based drug design, 358,417-419,467-469 antifolate targets, 425-432 and combinatorial chemistry, 227 combinatorial library design, 225-228 and docking studies, 282, 321-322 hemoglobin, 419-425 hydrolases, 449-454 iterative cycles, 282,463 NMR spectroscopy for, 419, 516-517 oxidoreductases, 445-449 phosphoryl transferases, 456-461 picornaviruses, 454-456 proteases, 432-445 and virtual screening, 244 Structure-based inhibitor design, 418 Structure-based virtual screening, 260-267 Structure elucidation NMR spectroscopy for, 517-525 Structure table, 376 Structure verification as bottleneck in drug discovery, 592 mass spectrometry application, 594-596 Subgraph isomorphism, 67,405, 410 Subreum, 849 Substance P antagonists, 669-671 Substances, 368,410 Substituent constants, for QSAR, 19-23 Substrate analog enzyme inhibitors, 733 Substructure searching, 255, 379,381-382,410-411 and QSAR, 67 SQL for, 395 Substructure search keys, 375, 376,378,410 molecular similarity/diversity methods, 189,221 Index Subtilases homology modeling, 123 Succinate dehydrogenase, 733 Succinic semialdehyde dehydrogenase inhibitors, 718 Succinyldicholine conformationally restricted analogs, 699 Sugars chirality, 784 Suicide substrate MMP inhibitors, 651-652 Suicide substrates, 756 Sulbactam, 718 Sulfonamides pharmacophore points, 249 Sulfones pharmacophore points, 249 Sulfonyl halides filtering from virtual screens, 246 Sulphonamides, 717 Supercritical fluid chromatograP ~ Y for enantiomer separation, 787 Supercritical fluid chromatography-mass spectrometry (SFC-MS) for combinatorial library purification, 594 Superstar, 315 Superstructure search, 255,257, 411 Supervised data mining, 66-67, 411 Suxamethonium, 857 Sweet clover, drugs derived from, 882 Sweet wormwood, drugs derived from, 886 SWISS-PROT, 335,345-346 SYBYL, 130 Sybyl Programming Language, 378,410 Synercid, 848,849, 876 SYNLIB, 361 SYSDOC ligand handling, 293 Systematic search and Active Analog Approach, 144-145 and conformational analysis, 89-93 in docking methods, 292 in molecular modeling, 89-94, 116 T gondii DHFR, QSAR inhibition studies, 33 Tabular storage, 369-371 Tabu search with docking methods, 292, 299 in virtual screening, 263 Tachykinin receptors, 669 Tacrine, 58 structure-based design, 449 Tacrolimus, 848,849 Tadpoles narcotic action of ROH, 28-29 Tagging approaches, 596-597 TAK-029,213 TAK-147 structure-based design, 450 Tandem mass spectrometry (MS-MS),590-591 of combinatorial libraries, 592 for structure determination of bioactive peptides, 518 types of mass spectrometers, 585 Tanimoto coefficient, 68,202, 411 cluster-based methods with, 206 and similarity searching, 382, 410 for virtual screening, 210 Tanimoto Dissimilarity, 220 Tanomastat structure-based design, 444-445,446 TargetBASE, 348 Target class approach, 188, 228-234 Target discovery See also Drug targets bioinformatics for, 335, 338-345 TAR RNA inhibitors, 103 Tautomenzation NMR spectroscopy, 526-528 Tautomers, 366 Tautomer search, 388,405-406 Taxol, 843,848,861-863 HMBC spectroscopy, 518 NMR spectroscopy, 525-526, 531 Taxol side-chain, 803-804 Taxus baccata (English yew), 861-862 Taxus brevifolia (Pacific yew), 861-862 TB36 structure-based design, 424-425 TBC 3214,674,676 Team Works, 377 Teicoplanin, 849 Telithromycin, 848,876 Temperature molecular dynamic simulation, 96 Template mimetics (peptidomimetics), 643-644 Tendamistat NMR relaxation measurements, 528-529,535 Teniposide, 867 Teprotide, 746,881 Terabyte, 411 Terbinafine, 717 Testosterone, 36, 768, 771 A,-Tetrahydrocannabinol (THC), 852-853 Tetrahydrofolate, 425 Tetrahymena pyriformis growth inhibition, 27,37-38 spiro-Tetraoxacycloalkanes ring-size analogs, 702-703 Tetrazoles, 135 as surrogates for cis-amide bond, 141-142 Thalidomide, 783-784,785 Thebaine, 850,851 Theilheimer/Chiras/Metalysi database, 386 Therapeutic area screening molecular similarityldiversity methods, 191 Thermodynamic cycle integration, 99-100,120-121 Thermolysin inhibitors genetic algorithm study of active site, 89 molecular modeling, 117, 120, 121,151-153 novel lead identification, 321 transition-state analogs, 749-750 Thick clients, 400-401,411 Thienamycin, 872,874 Thin clients, 363,392,401,411 Thiobiotin binding to avidin, 181, 182 Thioesters filtering from virtual screens, 246 p-ThioGARdideazafolate (P-TGDDF), 742-743 Index Thiol proteases QSAR studies, Thiomuscimol, 690 4-Thioquinone fluoromethide, 770 Thioridazine, 805,806 Thiorphan, 650,651 Thor database manager, 386 Thor system, 377 exact match searching, 380381 Threading, 123-125 3D descriptors molecular similaritytdiversity methods, 55-58,191-201 validation, 211-213 Three-dimensional electron cryomicroscopy, 615616 3D models, 363,366-367, 397-398 3D pharmacophores filter cascade, 267 for molecular similarityldiversity methods, 194-201 for searching, 381-382 similarity searching, 189,383 for virtual screening, 210, 255-259 3D quantitative structure-activity relationships (3DQSAR), 52,53,58-60 and molecular modeling, 115, 138 3D query features, 368,381382, 398 3DSEARCH, 111,259 3D structure databases, 387 3-Point pharmacophores, 376, 408 molecular similarity methods, 189,195196,198 for virtual screening, 210 Threo- prefix, 784 Threose enantiomers, 784 Thrombin inhibitors, 227 combinatorial docking, 318 force field-based scoring study, 307 molecular modeling, 116 non-peptide peptidomimetics, 660-662,663,664 seeding experiments, 319 site-based pharmacophores, 235-236 target of structure-based drug design, 442-443 Thromboxane 4,762-763 Thymidine kinase inhibitors, 717 role of water in docking, 303 X-ray crystallographic studies, 493 Thymidylate synthase inhibitors, 227, 717 target of structure-based drug design, 425,426-429 Thymitaq, 428 Thyroid hormones NMR spectroscopy, 529-531 Thyroid receptor beta, 263 Thyroliberin peptidomimetics, 129 Thyrotropin-releasinghormone, 637 Thyroxine NMR spectroscopy, 529-531 Tight-binding enzyme inhibitors, 720, 734-740, 749 Time-of-flight mass spectrometry, 585,607 Timolol renal clearance, 38 TIP3P model, 175 Tipranavir, 812, 813 Tirilazad mesylate, 849 Titrations NMR application, 545 TNF-a converting enzyme (TACE),652 Tolamolol renal clearance, 38 Tolrestat structure-based design, 448 Tomudex, 427 Toolkits, 386,411 Toothpick plant, drugs derived from, 883 TOPAS, 192 TOPKAT, 246 Topographical data, 411 Topographical mimetics (peptidomimetics), 636 Topoisomerase I1 inhibitors, 717 Topological descriptors for druglikeness screening, 247-249 estimation systems, 388-389 with QSAR, 54-55 Topotecan, 848,849,861 Torsional potential, 80 Toxicity databases, 246,386 development, 828-829 Toxicity prediction, 827-843 Toxicity screening as bottleneck in drug discovery, 592 and functional group filters, 246-247 pulsed ultrafdtration application, 605 Toxicophores, 829- 831 associated with allergic contact dermatitis, 830 C-Toxiferine 1,856,857 TPCK, 760-761,762 Tramadol, 782 chromatographic separation, 792 classical resolution, 795-796 metabolism, 786-787 Transesterification enzyme-mediated asymmetric, 805-806 Transferred NOE technique, 532 and NMR screening, 572-573 Transition-state analog enzyme inhibitors, 720, 748-754 Transition state analog inhibitors, 646 peptide bond isosteres, 644 7-Transmembrane G-proteincoupled receptors, 229-234 Transpeptidase inhibitors, 717 Transverse relaxation-optimized spectroscopy (TROSY), 515 for macromolecular structure determination, 533,534 Trees, 376377,411 TrEMBL, 335,346 Triazines QSAR studies of cellular growth inhibition, 37-38 QSAR studies of DHFR inhibition, 31-33 Trimethoprim, 717, 719 interaction with dihydrofolate reductase, 151,183 structure-based design, 425 a,@-bis-Trimethylammonium polymethylene compounds, 710 Trimetrexate interaction with dihydrofolate reductase, NMR spectroscopy, 531,557-559 Triple resonance spectra, 514 Tripos, Inc databases, 387 Tripos force field, 80 t-RNA guanine transglycosylase inhibitors novel lead identification, 321 Trojan horse inactivators, 756 Trypsin inhibitors molecular modeling, 120 QSAR studies, 5, 25 site-based pharmacophores, 235-236 Trypsinogen inhibitors molecular modeling, 116 Tryptophan chemical modification reagents, 755 TSCA database, 386 Tubby gene X-ray crystallographic function elucidation, 494 Tube curare, 856 D-Tubocurarine drugs derived from, 856,857 fragment analogs, 708-710 p-Tubulin X-ray crystallographic studies, 483 Tumor necrosis factor receptor X-ray crystallographic studies, 493 2D descriptors molecular similarityldiversity methods, 191-194 with QSAR, 54-55 validation, 211-213 2D pharmacophore searching, 383 filter cascade, 267 virtual screening, 255 2D quantitative structure-activity relationships (2DQSAR), 52,53 2D query features, 397 2D structures, 364366,397 conversion of names to, 373 2-Point pharmacophores, 376 Tyrosine chemical modification reagents, 755 Tyrosine kinase inhibitors molecular modeling, 130 U-85548 structure-based design, 436-437,438 Ugi reaction, 229,231, 232, 236 UK QSAR and Cheminformatics Group, 360 Ukrain, 849 Uncompetitive inhibitors, 729-730 Unicode, 411 UNITY, 259,377 descriptors, 192,201 in molecular modeling, 111 novel lead identification, 320 UNITY 2D, 212 UNITY 3D, 363,387 University of Manchester Bioinformatics Education and Research site (UMBER), 335 Unix, 396, 411 Unsupervised data mining, 66-67,412 Urea pharmacophore points, 249 Ureido resonance, 182 USEPA Suite, 390 VALIDATE, 116,310 Vancomycin, 770 Vancomycin-peptide complex binding affinity, 119 van der Waals forces, 174,285 and docking scoring, 308 enzyme inhibitors, 723-724 and molar refraction, 24 molecular modeling, 79-80, 81, 82, 89 and QSAR, , van der Waals radius, 79,81, 173 Vanillin antisickling agent, 419-420 Vanilloid receptors, 853-854 VanX inhibitors, 770-771, 772 VARCHAR data type, 412 VARCHAR2 data type, 412 Vector maps, 140-142 Verapamil classical resolution, 798 Verapamilic acid, 798 Vidarabine, 717 Vigabatrin, 718,766, 767,782 Vinblastine, 858-859,860 Vinca alkaloids, 858-860 Vincristine, 858-859,860 Vindesine, 859-860 Vinorelbine, 849 Viracept, 659 structure-based design, 440, 442 Viral DNA polymerase inhibitors, 717 Virtual chemistry space, 67 Virtual libraries, 237, 283,315 handling large, 220-221 and QSAR, 61 Virtual rings, 91 Virtual screening, 244-245, 271-274,315-317,412 See also Docking methods; Scoring functions applications, 267-271 basic concepts, 289-290 combinatorial docking, 317-318 consensus scoring, 265-266, 291,319-320 docking as virtual screening tool, 266-267 druglikeness screening, 245-250 filter cascade, 267 focused screening libraries for lead identification, 250-252 hydrogen bonding and hydrophobic interactions, 319 ligand-based, 188,209-214 molecular similarityldiversity methods for, 188, 190, 209-214 novel lead identification, 320-321 pharmacophore screening, 252-260 QSAR as tool for, 66-69 seeding experiments, 318-319 structure-based, 260-267 weak inhibitors, 319 Vista search program, 387 Vitamin D receptor X-ray crystallographic studies, 493 VK19911 structure-based design, 458 Voglibose, 849 VolSurf program, 202 Volume molecular dynamic simulation, 96 Volume mapping, 139-140 Voronoi QSAR technique, 53 VRML (Virtual Reality Markup Language), 405 VX-497 structure-based design, 447 VX-745 structure-based design, 458 Warfarin, 882-883 enantiomers, 786 Warfarin(Continued) HIV protease inhibitor, 659, 661 nonclassical resolution, 801 WARP, 478 Water gas phase association thermodynamic functions, 178 importance of bound in structure-based design, 409 molecular modeling, 85 octanoVwater partitioning system, 16-17 and protein-ligand interactions, 288 role in docking, 302-303, 313-314 solvating effect in enzyme inhibitors, 722-723 Wellcome Registry, 222 White-Bovill force field, 80 WIN-35065-2 dopamine transporter inhibitor, 268 WIN-51711 structure-based design, 454-455 WIN-54954 structure-based design, 455 WIN-63843 structure-based design, 455-456 Wiswesser line notation, 368-369 WIZARD, 255,260 World Drug Index (WDI),379, 386,387 World Patents Index (WPI), 386 Xanthine-guanine phosphoribosyltransferase X-ray crystallographic studies, 493 Xanthine oxidase inhibitors, 718 XFIT graphics program, 478 Ximelagatran structure-based design, 442 XML (extensible Markup Language), 371,405,412 XMLQuery, 412 X-ray crystallography, 351, 471-473,612 applications, 479-481 crystallization for, 473-474, 480-481 databases for, 478-479 data collection, 474-476 drug targets with published structures, 482-493 and molecular modeling, 78 phase problem, 476-478 and QSAR, and structural genomics, 481, 494-496 in structure-based drug design, 418,419,420 and structure-based library design, 225 and virtual screening, 244 X-ray diffraction, 472-473, 614 X-ray lenses, 612 Yellow sweet clover, drugs derived from, 882 Yew tree, paclitaxel from, 861-862 YM-022,856 Yukawa-Tsuno equation, 14 Z-100,849 Zanamivir, 717 structure-based design, 451 Ziconotide NMR spectroscopy, 518-523, 526,534 Zidovudine, 717 Zingerone allergenicity prediction, 835 "An essential addit to the libraries o a n outstanding work highly information in d r u g studies and research Toozzmal o f Medicinal Chenzi This new edition of Dr Alfred Burger's internationally celebrated classic helps researchers acquaint themselves with both traditional and state-of-the-art principles and practices governing new medicinal drug research and development Completely updated and revised to reflect the many monumental changes that have occurred in the field, this latest edition brings together contributions by experts in a wide range of related fields to explore recent advances in the understanding of the structural biology of drug action, as well as cutting-edge technologies for drug discovery now in use around the world This Sixth Edition of Burger's Medicinal Chemistry and Drug Discovery has been expanded to six volumes: Volume 1: Drug Discovery Volume 2: Drug Discovery and Drug Developme1 Volume 3: Cardiovascular Agents and Endocrines Volume 4: Autocoids, Diagnostics, and Drugs from New Biology Volume : Chemotherapeutic Agents Volume 6: Nervous System Agents UUNALDA ABRAHAM, PHD, is Proiessor and Chairman of the Department of Medicinal Chemistry at the Virginia Commonwealth University School of Pharmacy A world-renowned leader in medicinal chemistry and biotechnology, he is the author of more than 140 journal citations and twenty-five patents H e was selected by the AACP Board of Directors as the recipient of the 2002 Ar? ? P ' D2xxrqon Biotechnology Award ... (22 , ALIMTA, (22 ) pemetrexed (23 ) lometrexol LY231514) and lometrexol (23 , 5,lO-dideaztetrahydrofolate, LY -26 4618), have been shown to be effective antitumor agents in clinical trials (71, 72) ... in drug discovery The chapter is not completely encyclopedic, and some significant instances of SBDD will be missed by the informed reader However, the discovery programs with drugs and drug candidates... involved in the transport and cellular accumulation of antifolate drugs Balancing these criteria has resulted in the choice of compounds (26 ) and (27 ) (AG2034 and AG2037, respectively) for clinical