8.18.2.1 MPT
All molybdenum enzymes, with the notable exception of the nitrogenases (Section 8.18.3), involve a single metal atom bound to one or two molecules of a special ligand, originally named
‘‘molybdopterin’’ (MPT)17 (Figure 1). However, as MPT also binds tungsten65–74 (Section 8.18.3) this moiety is better named as the metal-binding pyranopterin ene-1,2-dithiolate, retaining the abbreviation MPT. MPT is comprised of a pterin, with pyrimidine and pyrazine rings, the latter being linked to a pyran ring that carries both an ene-1,2-dithiolate (dithiolene) that coordinates the Mo (or W) as a bidentate ligand and a –CH2OPO32
group. In molybdenum–
MPT enzymes isolated from prokaryotes (bacteria and archaea), the phosphate group may be part of a diphosphate linkage that forms a dinucleotide with adenine, guanine, cytosine, or hypoxanthine as the base. The form of MPT found in tungsten–MPT enzymes varies in a less predictable manner.
In each molybdenum–MPT41–67(and tungsten-MPT69–74) enzyme, the structure of the MPT is remarkably conserved; both the pyrazine and pyran rings are distinctly nonplanar and each of the three chiral carbon atoms of the pyran ring is in the (R)-configuration. The pyran ring adopts a half-chair conformation that deviates significantly from the plane of the pterin system. In the enzymes the best plane defined by the pyran ring is tilted40 from the plane of the pterin ring.
However, the relationship between the rings is not precisely determined since there is some conformational flexibility in the way the pyran ring is tilted out of the plane of the conjugated part of the pterin. Additional conformational flexibility is present, in respect of the phosphorylated
HN
N N
H H N
O H2N
O S
S
OPO(OH)(OR)
R = H (MPT); R = nucleoside
Figure 1 Structure of ‘‘molybdopterin’’17 or preferably the metal-binding pyranopterin ene-1,2-dithiolate (MPT); the phosphate group may be bound to a nucleotide.
hydroxymethyl side chain that leads to a wide variety of positions for the phosphate group relative to the pterin.
The polar (O, N, NH, and NH2) groups of MPT participate extensively in the formation of hydrogen bonds to complementary groups of the polypeptide chain. The extent of this hydrogen bonding is significantly extended when the phosphate is bound to the dinucleotide (seeFigure 1).
This network of hydrogen bonds serves to appropriately locate the catalytically active center at a particular position within the polypeptide chain with respect to:
the ‘‘funnel’’ in the protein ‘‘sheath,’’ through which the substrate enters and product leaves;
the ‘‘pocket’’ that binds the substrate in an orientation suitable for catalytic conversion to the particular product; and
the electron transfer relay, involving the other redox active prosthetic groups, that leads to the site on the exterior of the protein where the redox partner binds.
There are several potential roles for MPT, including: cooperating with the metal–dithiolene center in the two-electron redox change required for catalysis; handling the protons involved in the OAT process (seeEquation (2)); providing a route for electron transfer to or from the metal centre during turnover;72,75fine-tuning the redox potential of the metal center; and/or enhancing the OAT by strong S!M - and -donation.76 With respect to the first of these aspects, it is important to note that in each crystallographic study the structure of MPT is apparently equivalent to the fully reduced state, i.e., a tetrahydropterin. This suggests that the pterin component of MPT is not involved in the redox changes that occur during catalysis. However, it should be noted that the opening of the pyran ring of MPT is likely to be facile and this ring- opened form (Figure 2) shows that, formally, MPT is a dihydropterin.22
The biosynthesis of molybdenum–MPT centers is an evolutionarily conserved pathway in archaea, eubacteria, and eukaryotes, including humans. This process is now reasonably well understood8,77–83 and the structures of several proteins involved in the biosynthesis have been determined.84–87 The metal is incorporated as the final stage and, in Escherichia coli, the mod- ABCD gene products are involved in the uptake of molybdate.88The ModA-C proteins comprise the molybdate transport system.89 ModA is the initial binding protein and has a relatively high affinity for both molybdate and tungstate (4mMand 7mM, respectively);90 the specificity of ModA to bind molybdate and tungstate (tungstate transport) but not sulfate or other anions is determined by the size of the binding pocket, in which the four oxo groups are held by hydrogen bonds to the amide and hydroxyl groups of the polypeptide chain.91
Molybdate is then delivered to ModB, located on the periplasmic side of the inner membrane; this protein, in concert with ModC, initiates the chemical transformation of the molybdenum and MoeA and MogA are believed to function in the addition of molybdenum to the dithiolene group of MPT.92
‘‘Molybdenum cofactor deficiency’’ is a rare inborn error of metabolism and this can have serious consequences for the organism; this is well documented for humans for sulfite oxidase93(seeSection 8.18.2.5), leading to severe neurological disorders, seizures, mental retardation, and death at an early age.94 Therefore, it is encouraging that an MPT-free form of sulfite oxidase has been reconstituted with a de novo form of the ‘‘molybdenum cofactor’’ synthesized from molybdate and a biosynthethic precursor of MPT, ‘‘compound Z.’’95
8.18.2.2 Classification of Molybdenum–MPT Centers
Structural and spectroscopic studies of the oxidized state of the molybdenum–MPT enzymes have identified four types of molybdenum center (Figure 3) that provide a good basis for the classifica- tion of the molybdenum–MPT enzymes into families.8
Members of the dimethylsulfoxide (DMSO) reductase family have the molybdenum ligated by two MPTs, one oxo group, and the donor atom from the side chain of an amino acid residue—O of a serinyl residue (DMSO reductase or trimethylamineoxide (TMAO) reductase), S of a cysteinyl residue (dissimilatory nitrate reductase), or Se of selenocys- teinyl residue (formate dehydrogenase).
Members of the sulfite oxidase family, including the assimilatory nitrate reductases, possess acis-dioxo {MoO2}2þcenter that is bound to one MPT, one hydroxo group, and a cysteinyl residue.
Members of the xanthine dehydrogenase/xanthine oxidase family, including the alde- hyde oxidases, have acis-oxo, sulfido {MoOS}2þcenter bound to one MPT and a hydroxo group.
Thus, unlike the nitrogenases, which possess an iron–molybdenum cofactor (FeMoco, Chapter 8.22), there is no unique ‘‘molybdenum cofactor’’ (or Moco) for the molybdenum–MPT enzymes.
Rather there is a diverse collection of catalytic centers with the common features of a single molybdenum atom ligated by one or two MPTs, plus additional ligands, which can be an amino acid side chain, and/or one, two, or three small inorganic ligands—such as oxo, sulfido, or hydroxo groups.8,96,97 Molybdenum–MPT enzymes of the same family generally possess a con- siderable homology in the sequence of their polypeptide chains and appear to employ a similar mechanism for catalysis. The former point illustrates that these molybdenum centers are produced under genetic control and the latter point demonstrates that reactivity at a metal center is greatly influenced by its coordination sphere.
There are some general points that can be made concerning the function of the catalytic centers of molybdenum-MPT enzymes. As noted in Section 8.18.1, these enzymes catalyze the two- electron oxidation or reduction of the substrate, often coupled to OAT (Equation (2)). Spectros- copic studies, notably EPR investigations of the Mo(V) state, have shown that the substrate
HN
N N
H H N
O H2N
O S
S
OPO(OH)(OR) HN
N N
H H N
H2N O HN
N N
H N
H2N O HN
N N
N
H2N O
HN
N N
H N
HO H2N
O S
S
OPO(OH)(OR)
Tetrahydropterin Dihydropterin
Fully oxidized pterin
Ring-opened MPT MPT
Figure 2 Oxidation levels of pterins and the opening of the pyran ring of MPT, showing that MPT is formally at the dihydropterin oxidation level.
interacts directly with the metal center98 and a substrate-bound form of the DMSO reductase from Rhodobacter capsulatus has been structurally characterized.44 For oxidases, it is generally considered that catalysis involves the redox change from Mo(VI) to Mo(IV) with reductases utilizing the Mo(IV) to Mo(VI) redox change. In each case, the catalytically active state is regenerated by two, one-electron steps via an Mo(V) intermediate; e.g., reduction of an Mo(VI) center by an Hỵ/e addition converts an MoVIẳO group to MoV–OH; subsequent addition of Hþ/eforms MoIV–OH2.
Although the nature of the catalytic center and the molecular architecture of the various families of the molybdenum–MPT enzymes differ significantly, some general considerations apply.
Each catalytic center is located in the interior of the molecule and the substrate approaches the center through a ‘‘funnel’’ through which the reaction product is removed.
Prior to reaction, the substrate binds in a ‘‘pocket’’ immediately adjacent to the catalytic center in an orientation that facilitates the subsequent catalysis.
In addition to the catalytic center, almost all of the molybdenum–MPT enzymes involve one (or several) other prosthetic group(s). These enzymes are organized in a manner analogous to an electrochemical cell: the site where oxidation occurs is the anode and the site where reduction occurs is the cathode; the redox active prosthetic groups are arranged to conduct electrons from the anode to the cathode; the external medium that supports the enzyme is analogous to a salt bridge.
For many of the molybdenum–MPT enzymes, the information available allows a mechanism to be proposed for the catalytic action; however, for every enzyme, further investigations are required to establish the precise details of the catalytic process.
8.18.2.3 The DMSO Reductase Family
DMSO reductases are common enzymes that catalyze the reduction (Equation (3)) of DMSO to DMS in bacteria and fungi that are present in the oceans and salt marshes. These enzymes are located in the periplasm and function in a respiratory chain that uses DMSO as the terminal electron acceptor and involves the transfer of two electrons from a cytochrome, to form DMS.99 The activity of these enzymes can be linked to photosynthesis, for which they serve as a terminal electron acceptor. Although DMSO respiration (G=92 kJ mol1) does not provide as much
Mo O
S S S S
HN
N N
H H N
O H2N
O S
S
OPO(OH)(OR) S
S
L
DMSO reductase family
R = H (MPT); R = nucleoside
=
L = OSer for DMSO reductase, TMAO reductase
L = SCys for dissimilatory nitrate reductase
L = SeCys for formate dehydrogenase Mo
O
OH S
S
CO dehydrogenase family
S Cu
SCys O
Glu O
Mo S
O OH S S
Xanthine oxidase family
Mo O
O S Cys S
S
Sulfite oxidase family
Figure 3 Structural basis for classifying the major families of the molybdenum–MPT enzymes.8
energy as O2 respiration (G=218 kJ mol1), it is more efficient than alcohol fermentation.
These enzymes play a significant role in the global sulfur cycle since DMS is volatile and, in the atmosphere, is converted to methylsulfonate that acts to nucleate cloud formation.100 Further- more, the distinctive smell of DMS acts as a guide to certain seabirds who use it to locate productive regions of the ocean:101
DMSO + 2e– + 2H+ DMS + H2O ð3ị DMSO reductases have been purified and characterized from a range of sources.25 The majority of investigations have been accomplished for the enzymes isolated from the purple phototrophic bacteria Rhodobacter capsulatus99 and Rhodobacter sphaeroides.102 These are the simplest of the known molybdenum–MPT enzymes, in that they are monomeric with a molecular weight of 85 kDa and the molybdenum center is the sole prosthetic group. Thus, in contrast to the other molybdenum–MPT enzymes, these DMSO reductases lack Fe/S, heme, or flavin centers, although E. coliDMSO reductase103,104does contain an Fe4S4center.
Several crystal structure determinations of the Rhodobacter DMSO reductases have been determined.41–46 The polypeptide chain of 780 residues binds two MGD molecules (pterins P and Q) and one molybdenum. The polypeptide chain is folded into four, noncontiguous domains that link to form an ellipsoidal molecule with a ‘‘funnel-like’’ depression on one side. The molybdenum is located at the bottom of this funnel, close to the center of the molecule; the funnel is lined by aromatic residues, creating a hydrophobic pocket that facilitates the passage of DMSO and DMS to and from the active site, respectively. Both P-and Q-pterins are locked into the polypeptide by an extensive series of hydrogen bonds and there are significant conformational differences between the two pterins, notably in the deviation of the pyran ring from the plane of the pterin (P-pterin &30, Q-pterin 40–60).
There has been some difficulty in defining the coordination geometry of the molybdenum in the Rhodobacter DMSO reductases, and detailed structural studies of oxidized DMSOR from R. sphaeroides at 1.3 A˚45 and R. capsulatus46 have revealed a ‘‘plasticity’’ of the active site. In the former, the molybdenum center is discretely disordered and exists in both hexacoordinate and pentacoordinate environments. The hexacoordinated site (Figure 4) involves the metal bound to one oxo group (that is hydrogen bonded to Tyr114), the -oxygen of Ser147, and four sulfur atoms from the two MPTs; the six donor atoms form an irregular trigonal prism centered on the molybdenum. The pentacoordinated site involves the metal bound to two oxo groups, the -oxygen of Ser147, but only two sulfur atoms—those of pterin P, with the sulfurs of pterin Q (at 3.62 A˚ and 4.53 A˚) being too far away to be considered bonded to the molybdenum. Molybdenum K-edge EXAFS105,106and resonance Raman spectra107–109are consistent with all four sulfur atoms being bound to the metal in the native enzyme. Thus, the structure shown inFigure 4is generally accepted as the form of the molybdenum center in the oxidized state of the DMSO reductases. The results of crystallographic studies45,46 demonstrate unequivocally that Hepes buffer leads to a significant change in the nature of the active site of the oxidized enzyme.45,46
A protein crystal structure has been reported for a form of theR. capsulatusDMSO reductase generated by addition of DMS to crystals of the enzyme.44 The DMS binds to the oxo group, apparently producing DMSO bound to the reduced molybdenum center. This is depicted in Figure 5 as a des-oxo species, as formed by the addition of DMS to the mono-oxo center shown in Figure 4. The structures of the oxidized and reduced forms of the catalytic centers of the molybdenum DMSO reductases have stimulated an extensive range of coordination chemis- try, directed at the production of their chemical analogues with some considerable success.110–116 Spectroscopic measurements on the Rhodobacter DMSO reductases are consistent with the oxidized state containing Mo(VI). The absence of prosthetic groups, other than the molybdenum–
MPT center, facilitates spectroscopic studies of this center; e.g., the UV/visible spectrum of the oxidized enzyme has absorptions at 280 nm, 350 nm, 470 nm, 550 nm, and 720 nm and the dithionite reduced form has absorptions at 280 nm, 374 nm, 430 nm (sh), and 645 nm.103 The Mo(V) state of the R. capsulatusenzyme, generated by titration of dithionite or reduced methyl viologen, has been investigated by MCD spectroscopy and the transitions observed assigned to ligand-to-metal charge transfer from the dithiolene sulfurs.117Reduction by dithionite produces a surprising number of Mo(V) EPR signals, suggesting some structural flexibility at the molyb- denum center. Further reduction with dithionite produces an EPR-silent form, considered to contain Mo(IV).118 ESEEMspectroscopy has been used to investigate the environment of the proton of the MoV–OH center of the DMSO reductase ofR. capsulatus.119
EPR potentiometric titrations ofR. sphaeroidesDMSO reductase at pH 7.0 and room tempera- ture gave redox potentials of þ144 mV for the Mo(VI)/Mo(V) couple and þ160 mV for the Mo(V)/Mo(IV) couple.102 The first cyclic voltammetric investigation of a molybdenum–MPT enzyme has been accomplished forR. capsulatus DMSO reductase.120 In this study a reversible Mo(VI)/Mo(V) couple was observed at þ161 mV and a reversible Mo(V)/Mo(IV) couple at 102 mV. The former couple exhibits a pH dependence of ca.59 mV/pH unit consistent with an e/Hỵtransfer (5<pH<9) and is taken to indicate the conversion of the MoVIẳO group to a MoV–OH unit. The potential of the Mo(V)/Mo(IV) couple was constant for 5<pH<10, indicating that reduction of the Mo(V) does not involve protonation. This latter behavior contrasts with that of theE. coliDMSO reductase, for which each reduction involves coupled e/Hþtransfer.121
Figure 4 Structure of the molybdenum center of native, oxidized DMSO reductases.45
Figure 5 Structure of the molybdenum center of DMS-treated native DMSO reductase fromRhodobacter capsulatus(McAlpineet al.44but see text).
Catalytic and single-turnover experiments with the R. sphaeroides DMSO reductase,
18O-labeled DMSO and 1,3,5-triaza-7-phosphatricyclo[3.3.1.1]decane as an oxygen atom acceptor have been used to demonstrate that the enzyme is an oxotransferase.122Complementary reson- ance Raman studies have been interpreted on the basis of a direct mechanism for OAT, with the active site cycling between mono-oxo-Mo(VI) and des-oxo-Mo(IV) forms via a DMSO-bound Mo(IV) intermediate. Both MPT dithiolene groups stay firmly attached to the molybdenum throughout the catalytic cycle.108 However, EPR and UV/visible spectroscopic evidence has been interpreted on the basis of the species formed upon the addition of DMS to oxidized R. capsulatusDMSO reductase involving a Mo(V) center bound to a sulfur radical.44
A membrane-bound DMSO reductase has been identified inE. coli.123This enzyme is composed of three subunits: one containing the molybdenum–MPT catalytic center, which is essentially the same as those of the DMSO reductases ofR. capsulatusandR. sphaeroides; a second containing four Fe4S4clusters, to conduct electrons to the catalytic center; and a third that anchors the enzyme in the membrane and also contains the binding site for the external reductant, menaquinol.
The DMSO reductases are capable of catalyzing the reduction of a wide variety of substrates, including a range of sulfoxides. An interesting aspect of these reductions, doubtless a consequence of the chiral nature of the substrate-binding pocket, is the enantioselective reduction of a racemic mixture of chiral sulfoxides. Also, these enzymes catalyze the loss of an oxygen atom from a variety of N-oxides, including TMAO and pyridine-N-oxides, and chlorate, [ClO3]. TMAO reductases possess a structure that is very similar to those of the DMSO reductases of R. sphaeroides andR. capsulatus, including an essentially identical catalytic center.47,48
A Tyr114 residue is present in the active site of DMSO reductase and biotin sulfoxide reductase,124,125 but not TMAO reductase. This difference has been implicated in different substrate specificity of the enzymes, i.e., the DMSO reductases catalyze the reduction of both sulfoxides and amine oxides, but TMAO reductases will not catalyze the reduction of sulfoxides.
To test this hypothesis,E. coliTMAO reductase was cloned and expressed at high levels and site- directed mutagenesis used to generate the Tyr114 Ala and Phe variants of R. sphaeroidesDMSO reductase and to insert a Tyr residue into the equivalent position in TMAO reductase. Although all of the mutants turned over in a manner similar to their respective wild-type enzymes, mutation of Tyr114 in DMSO reductase resulted in: (i) a decreased specificity for S-oxides; (ii) an increased specificity for TMAO; and (iii) a loss in the ability of the oxidized enzyme to oxidize DMS. Also, insertion of Tyr into TMAO reductase resulted in a decreased preference for TMAO relative to DMSO.126 Furthermore, a system for expressing site-directed mutants of the molyb- denum DMSO reductase fromR. capsulatushas been constructed and used to replace Tyr114 by Phe. The mutant has increased kcat andKm towards both DMSO and TMAO, compared to the native enzyme. Direct electrochemistry showed that the mutation did not affect the potential of the Mo(V)/Mo(IV) couple, but the midpoint potential of the Mo(VI)/Mo(V) couple was raised by ca. 50 mV.127 Thus, the Tyr114 residue plays a critical role in the redox properties and OAT capability of the catalytic center of the DMSO reductases. As for the earlier mutagenesis,126it was found that the oxidation of DMS by the Phe114 mutant was significantly impaired, as compared to the native enzyme.127A DMS dehydrogenase, an enzyme that catalyzes the oxidation of DMS to DMSO, has been isolated from Rhodovulum sulfidophilum.128
8.18.2.4 Formate Dehydrogenases
Formate dehydrogenases catalyze the oxidation of formate to carbon dioxide (Equation (4)) and play an important role in the global fixation of carbon dioxide. Unusually for molybdenum–MPT enzymes,Equation (4)is not OAT but involves the cleavage of a C–H bond with the formation of a CẳO bond. The majority of the formate dehydrogenases occur in aerobic microorganisms and they contain no molybdenum and depend on pyridine nucleotide for their activity. However, some formate dehydrogenases occur in anaerobic bacteria and involve molybdenum and other redox-active prosthetic groups.25,129–134The considerable similarity in the amino acid sequences of these formate dehydrogenases and members of the DMSO reductase family strongly suggests that these enzymes possess the same class of metal center8and this has been confirmed by the determin- ationof the crystal structure of formate dehydrogenase H from E. coli.49This enzyme contains a single molybdenum center bound to two molybopterin guanine dinucleotide (MGD) cofactors and a selenocysteine (SeCys140), together with a Fe4S4 cluster. The structure was determined for the oxidized enzyme, with and without the inhibitor nitrite, and for the formate-reduced enzyme:
HCO2– CO2 + H+ + 2e– ð4ị