Báo cáo khoa học: docx

MINIREVIEW Retraction: Human telomeric G-quadruplex: targeting with small molecules Amit Arora, Niti Kumar, Tani Agarwal and Souvik Maiti Proteomics and Structural Biology Unit, Institute of Genomics and Integrative Biology, Delhi, India Retraction: The following review from FEBS Journal, ‘Human telomeric G-quadruplex: targeting with small molecules’ by Amit Arora, Niti Kumar, Tani Agarwal and Souvik Maiti, published online on 27 th November 2009 in Wiley InterScience (www.interscience.wiley.com), has been retracted by agreement between the authors, the journal Editor-in-Chief Professor Richard Perham and Blackwell Publishing Ltd. The retraction has been agreed due to overlap between this review and the following reviews: Published in Organic & Biomolecular Chemistry, ‘A hitchhiker’s guide to G-quadruplex ligands’ by David Monchaud and Marie-Paule Teulade-Fichou. Volume 6 Issue 4, 2008, pages 627–636. Published in BioChimie, ‘Targeting telomeres and telomerase’ by Anne De Cian, Laurent Lacroix, Ce ´ line Douarre, Nassima Temime-Smaali, Chantal Trentesaux, Jean-Franç ois Riou and Jean-Louis Mergny. Volume 90 Issue 1, 2008, pages 131–155. Introduction Aberrant cellular proliferation is associated with the infinite extension of telomeric ends, mediated by unusu- ally high telomerase activity or caused by abnormal overexpression of the proto-oncogenes generally required for cellular growth and differentiation [1–3]. As an anticancer strategy, efforts have been invested in targeting and lowering telomerase activity, which is often found to be overexpressed in cancerous cells [4,5]. However, the problem associated with telomerase targeting is that cells can adopt telomerase-independent mechanisms for telomere maintenance called alternative lengthening of telomere [6]. This leads to skepticism about the use of telomerase inhibitors as anticancer agents because cells can quickly switch to alternative mechanisms and hence become resistant to telomerase inhibitors. Therefore, it becomes imperative that the Keywords alkaloids; anticancer agent; click chemistry; ethidium derivative; G-quadruplex; human telomere; metallo-organic complex; N-methylated ligands; proto-oncogenes; telomerase Correspondence S. Maiti, Proteomics and Structural Biology Unit, Institute of Genomics and Integrative Biology, CSIR, Mall Road, Delhi 110 007, India Fax: +91 11 2766 7471 Tel: +91 11 2766 6156 E-mail: souvik@igib.res.in (Received 25 June 2009, revised 1 September 2009, accepted 28 September 2009) doi:10.1111/j.1742-4658.2009.07461.x Over the past few decades, numerous small molecules have been designed to specifically and selectively target the unusual secondary structure in DNA called the G-quadruplex. Because these ligands have been shown to selectively inhibit the growth of cancer cells, they have become a central focus for the development of novel anticancer agents. However, there are many challenges which demand greater effort in order to devise strategies for rational drug design with utmost selectivity. This minireview aims to reflect recent developments in the design of G-quadruplex ligands and also discusses the future outlook for designing more effective G-quadruplex interacting ligands. Abbreviations Dppz, dipyridophenazine; NCQ, neomycin capped quinacridine. FEBS Journal 277 (2010) 1345 ª 2009 The Authors Journal compilation ª 2009 FEBS 1345 mechanisms of telomere maintenance are targeted, by impinging on the structure and function of telomeric ends [7]. Telomeres can organize structurally into different conformations, for example, the G-rich single-stranded DNA overhang can adopt an unusual four-stranded DNA quadruplex structure, and stabilization of this structure by ligands would render the 3¢ overhangs unavailable for hybridization with the telomerase template for the extension of telomeric ends [7]. Further, ligand binding and stabilization of the quadruplex structure affect the recruitment of telomere-associated proteins required for the capping and maintenance of telomeric ends. In addition to telomeric ends, the promoter regions of proto-oncogenes also harbor G-quadruplex motifs. It has been observed that targeting the quadruplex motif in the promoter of proto-oncogenes with quadruplex-interacting ligands decreases the tran- scriptional activity of these proto-oncogene and helps to combat aberrant proliferation. These observations are encouraging many laboratories to synthesize quadruplex-interacting ligands. Given the rich diversity of G-quadruplex scaffolds and their propensity to inter- convert, it will be a challenge to identify small molecules that exhibit recognition selectivity for diverse scaffolds at the cellular level. There is a general notion that binders that stabilize the G-quadruplex structure may pave the way for the discovery of novel anticancer agents. A key feature of appropriate molecules are large flat aromatic systems involved in p-stacking with the G-tet- rad platform and possess reasonable water solubility, i.e. the molecule should display both hydrophobic and hydrophilic characteristics. Various scaffolds have been synthesized to date, and both computational and chemical–biological approaches have been used to understand quadruplex–ligand interaction. N-methylated ligands N-methylated ligands, quaternized on the aromatic ring nitrogens, have been thoroughly exploited because of their low electron density, which in turn leads to increased p-stacking of the aromatic part as well as reasonable water solubility without the need for cationic side chains. TMPyP4 is the pivotal example of this family of ligands (Fig. 1). This tetracationic porphyrin has been shown to have high affinity for G-quadruplex, to efficiently inhibit telomerase and is also known to downregulate the expression of oncogenes such as c-myc or k-ras, along with its potency to convert antiparallel topologies to parallel forms of quadruplexes [8–15]. Despite the nonselectivity of TMPyP4 for the quadruplex structure [16–18], interest in this particular molecule has never declined. TQMP68 and 3,4-TMPyPz (Fig. 1) [19] are two examples of porphyrin-based tetracationic macrocycles, which have been shown to bind efficiently to quadruplex DNA. TMPyP4-related ligands carrying 1–3 N-methylpyridinium arms [20], as well as structurally related corroles [21], have also been described. An important advance in the porphyrin series came with the design of a diselenosapphyrin Se 2 SAP N N N N N N NH N N HN NH N N N N HN N N N N N N TMPyP4 3,4-TMPyPz N HN N N HO OH Se N Se N NH HN N N N N N N N OH HO TQMP Se 2 SAP TQMP Fig. 1. Chemical structures of N-methylated ligands: TMPyP4, 3,4-TMPyPz, TQMP, Se 2 SAP. Quadruplex interacting molecules A. Arora et al. FEBS Journal 277 (2010) 1345 ª 2009 The Authors Journal compilation ª 2009 FEBS (Fig. 1), with an expanded porphyrin core [22,23]. Se 2 SAP showed 50-fold selectivity for quadruplex DNA over duplex DNA and was able to discriminate among the various forms of G-quadruplex DNA. Ligands with protonable side arms These ligands follow the presence of protonable side arms (e.g. amine groups) around an aromatic core which makes the molecule water soluble, with the charge(s) far from the hydrophobic center. Bis- amidoanthraquinone is one example which has been shown to be a G-quadruplex ligand, telomerase inhibitor [24] and possesses IC 50 values in the low lm range [25,26]. To address selectivity problems, Neidle and co-workers modified the core and side arms of the initial ligands from anthraquinone to fluorenone [27], then acridone [28] and acridine [29,30]. A crystal structure of a complex of BSU6039 (Fig. 2), a member of the 3,6-disubstituted acridine series with G-quadruplex was obtained [31]. Based on the concept of p-stacking interactions and electrostatic interactions between the quadruplex and the ligand, an optimized prototype BRACO-19 (Fig. 2) was designed, which was able to interact with the G-quadruplex structure [32,33]. BRACO-19 has also been shown to inhibit cancer cell proliferation [34]. Modification of the 9-amino substi- tuent of BRACO-19 from an aniline to a difluoroben- zylamine group was also carried out to circumvent the problem of cellular uptake of BRACO-19 [35]. There are reports of interaction studies between G-quadruplex DNA and the perylene diimide PIPER (Fig. 2) [36]. This molecule consists of a broader hydrophobic core, with two external amine appendages. This family of compounds displayed moderate telomerase inhibition activity but showed 42-fold selectivity for quadruplex DNA over duplex DNA at pH 7.0 and pH 8.5 [37–39]. Biological experiments showed the cellular uptake of such ligands, thereby suggesting them as potential G-quadruplex binders [40,41]. Well-known duplex-binding ligands like daunomycin [42], distamycin and netropsin [43] have also been tested as G-quadruplex-binding ligands. Several flavonoid [44,45] or steroid derivatives [46] have also been shown to bind quadruplexes with variable efficiency. Another class of ligands known as pentacyclic acri- dines and quinacridines (which possess a crescent shape) likely to maximize overlap with the guanines of the accessible G-quartet, has been developed. RHPS4 (Fig. 2) is one example of a pentacyclic acridine-based G-quadruplex interacting ligand. A number of in vitro and in vivo reports have shown that RHPS4 is a potential molecule for cancer therapy because it targets G-quadruplex structures at the telomeric ends [47]. This molecule has recently also been used in preclinical trials for solid tumors [48]. MMQ3 (Fig. 2), a compound of the quinacridine family, showed remarkable G-quadruplex stabilization and high telomerase inhibitory activity [49]. An NMR structure for a complex of MMQ1 (dipropylamino analogue of MMQ3) with tetramolecular quadruplex is also available [50]. Another compound BOQ1 (Fig. 2), a dimeric macrocyclic quinacridine, showed improved quadruplex stabilization, better overall selectivity than the monomeric series and efficient telomerase inhibitory activity [51–53]. The crescent-shape particularity of quinacridine has also been found in several other ligands (Fig. 2), such as indoloquinolines [54–56], cryptolepine and its analogues [57], quindolines [58–61] and triazacyclopentaphenantrene [62]. Alkaloid-based ligands Alkaloid-based ligands like berberine (Fig. 3) and its synthetic derivative have been examined for G-quadruplex binding and their ability to inhibit telomerase. Results show that these molecules have selectivity for G-quadruplex compared with duplex DNA, and that their aromatic moieties play a dominant role in quadruplex binding. Our group has also investigated the complete thermodynamic profile of the berberine–telomeric quadruplex interaction using spectroscopic, calorimetric and molecular modeling studies [63]. Fur- thermore, interaction of 9-substituted derivatives with human telomeric DNA indicated that these compounds can induce and stabilize the formation of antiparallel telomeric G-quadruplex in the presence or absence of metal cations [64]. Introduction of a side chain with the proper length of methylene and a termi- nal amino group at the 9-position of berberine resulted in increased binding with G-quadruplex, thus leading to higher inhibitory effects on the amplification of 21-mer telomeric DNA and on telomerase activity. Recently, the interaction of 9-N-substituted berberine derivatives (Fig. 3) with c-myc quadruplex has also been studied and the results indicated that these derivatives may selectively induce and stabilize the formation of intramolecular parallel G-quadruplex in c-myc, thus leading to downregulation of the transcription of c-myc in the HL60 lymphomas cell line [65]. In addition to berberine, other alkaloids such as palmatine and sanguinarine also demonstrate moderate quadruplex-binding abilities; however, the introduction of protonable functional groups might further enhance their recognition and stabilization abilities. Isoindigo- tone (Fig. 3), a naturally occurring alkaloid with a A. Arora et al. Quadruplex interacting molecules FEBS Journal 277 (2010) 1345 ª 2009 The Authors Journal compilation ª 2009 FEBS unique asymmetric chromophore comprising an ali- phatic five-member ring in the middle core has been shown to serve as a new scaffold for unfused aromatic quadruplex ligands. The introduction of at least two cationic side chains to the chromophore resulted in enhanced selectivity and solubility [66]. Interestingly, in a pharmacophore-based virtual screening, two non- planar alkaloid-based G-quadruplex ligands were N O 2 N N N HN Cryptolepine N HN N H 3 C H 3 C O CH 3 y HN R HN O Triazacyclopentaphenanthrene X HN N Quindoline, X= NH or O and R= different groups OO R N N H N N H N O O HH BRACO 19 N N N O O N N N O O PIPER R = H, BSU6039 R = NH-p-C 6 H 4 -N(CH 3 ) 2 , N F N N H N N N N HN N N RHPS4 MMQ3 N N HN NH R 1 NH(CH 2 ) n N(CH 3 ) 2 NH NH NN HN HN HN N R 2 N N BOQ1 Benzoindoloquinoline Fig. 2. Chemical structures of ligands with protonable side arms: BSU6039, BRACO-19, PIPER, RHPS4, MMQ 3 , BOQ1, benzoindoloquinoline, cryptolepine, triazacyclopentaphenanthrene, quindoline. Quadruplex interacting molecules A. Arora et al. FEBS Journal 277 (2010) 1345 ª 2009 The Authors Journal compilation ª 2009 FEBS found. These two ligands exhibit good capability for G-quadruplex stabilization and prefer binding to para- lleled G-quadruplex rather than to duplex DNA. These results have shown that planar structures are not essential for G-quadruplex stabilizers, which may rep- resent a new class of G-quadruplex-targeted agents as potential antitumor drugs [67]. Click chemistry-based ligands Click chemistry is a chemical approach which synthe- sizes new drug-like molecules by joining readily available smaller units together using simple chemical reactions aided by catalysts to reduce large enthalpy hurdles. Neidle and co-workers [68] introduced the concept of click chemistry in designing G-quadruplex binding ligands. The resulting bistriazole derivatives showed good quadruplex stabilization with a high degree of selectivity, although they appeared to be moderate telomerase inhibitors. A compound named neomycin-capped quinacridine (Fig. 4) was developed in which neomycin and a quinacridine moiety were conjugated to target the loop and G-quartet of the quadruplex, respectively [69]. Neomycin-capped quinacridine showed preferential binding to loop-containing quadruplexes compared with nonloop-containing quadruplexes, along with efficient quadruplex stabilization and strong telomerase inhibitory activity, thus fully validating the design. The presence of three amino appendages on the same face of the tri-oxazole macrocycles (Fig. 4) resulted in selective stabilization of one form of quadruplex over another, as shown by the preferential binding of tri-oxazole macrocycles to c-kit quadruplex rather than the human telomeric one [70]. Furthermore, isoalloxazines (Fig. 4) have also been shown to bind to c-kit quadruplex with 14-fold selectively over the telomeric quadruplex [71], thereby opening up a possibility for the design of a second generation of ligands capable of selectively altering the expression of a given gene. Recently, copper(I)-cata- lyzed ‘click’ chemistry was used to design a series of diarylurea ligands (Fig. 4). These ligands demonstrated a high degree of selective telomeric G-quadruplex stabilization and were not cytotoxic in several cancer cell lines [72]. Moreover, urea-based nonpolycyclic aromatic ligands with alkylaminoanilino side chains as G-quadruplex DNA interacting agents have been developed (Fig. 4) [73]. Using spectroscopic experiments, it was demonstrated that they have significant selectivity over duplex DNA, and also for particular G-quadruplexes. Preliminary biological studies using short-term cell growth inhibition assays showed that some of the ligands have cancer cell selectivity, although they appear to have low potency for intracel- lular telomeric G-quadruplex structures, suggesting that their cellular targets may be other, possibly oncogene-related, quadruplexes. Balasubramanian and co-workers [74] developed a series of trisubstituted acridine–peptide conjugates (Fig. 4) and explored the ability of these ligands to recognize and discriminate between different quadruplexes derived from the human telomere, and c-kit and N-ras proto-oncogenes. Our group reported the binding properties of 18- and 24-membered cyclic oligopeptides (Fig. 4) developed from a novel furan amino acid, 5-(aminomethyl)-2-furancarboxylic acid, to G-quadruplex. Comparative analysis of the binding data of these ligands with G-quadruplex and double-strand DNA shows that 24-membered cyclic peptides are highly selective for telomeric G-quadruplex structures and thus can be used as a scaffold to target quadruplex structures at the genomic level [75]. Ethidium derivatives as G-quadruplex ligands Mergny and co-workers [76] reported the use of ethidium derivatives as G-quadruplex ligands (Fig. 5). The results showed G-quadruplex stability and telomerase inhibition activity as well as quadruplex over duplex selectivity. However, because of the well-known toxic and mutagenic properties of ethidium bromide, researchers developed a novel and safer series of G-quadruplex ligands, derived from triazine [77–79]. One of the member of the series known as 12459 (Fig. 5) displayed selective stabilization of G-quadruplex and also strongly inhibited telomerase activity. Triazines were followed by a structurally related bis- quinolinium series containing a pyridodicarboxamide X H N R N + O CI - N N O H N (CH 2 ) n R N O (CH 2 ) n 9-N substituted Berberine derivatives, R substituted Berberine derivatives, R represents different groups O O N O O H N O H O Isaindigotone Berberine Fig. 3. Chemical structures of alkaloid-based G-quadruplex ligands: berberine, 9-N-substituted derivatives and isaindigotone. A. Arora et al. Quadruplex interacting molecules FEBS Journal 277 (2010) 1345 ª 2009 The Authors Journal compilation ª 2009 FEBS peptide O O O NH N N H N N H N 9-peptide acridine O O N H X O O X O O O O H H NH HN O O O O HN NH O O O H N O 1: X = NH 3 + 2: X = NH C OC H C H N H + O O H N O 2: X = NH C OC H 2 C H 2 N H 3 Furan based ligands O HO O NH 2 O HO H 2 N OH OH O NH 2 NH 2 O O HO HN OH O 2 HN O O HO N H O N O N H NH HN N O O 2 NH 2 NH 2 N H 2 NH NH 2 NH O N HN O O NH 2 HN NH N Tri-oxazole macrocycles R 2 N NCQ N N H N R 1 O N N R 1 O Trisubstituted isoalloxazines, R1 and R2 represents different groups H N H N O N H N H O HN NH O O O N N N N N N O O HN NH O O N H R N H R O ( ) n ( ) n R R ( ) n O O ( ) n Diarylurea based Ligands NH N N N H N H N H N H O O peptide peptide O O 3,6- bis acridine-peptide conju g ates Fig. 4 Chemical structures of click chemistry-based G-quadruplex ligands: neomycin capped quinacridine (NCQ), tri-oxazole macrocycles, trisubstituted isoalloxazines, diarylurea ligands, 3, 6-bis acridine–peptide conjugates, 9-peptide acridine and 5-(aminomethyl)-2-furancarboxylic acid (furan)-based G-quadruplex ligands. Quadruplex interacting molecules A. Arora et al. FEBS Journal 277 (2010) 1345 ª 2009 The Authors Journal compilation ª 2009 FEBS core (Fig. 5) [80,81]. Two compounds of the pyridodicarboxamide series, namely 307A and 360A, exhibited a high degree of quadruplex stabilization, exquisite quadruplex over duplex selectivity, and were able to induce efficient inhibition of telomerase. These compounds have also been shown to induce delayed growth arrest and apoptosis in immortalized cell lines. These results are particularly impressive with regard to the structural simplicity of the series and its two-step synthesis [81]. Remarkably, tritiated 360A has also been shown to localize preferentially at the telomeric regions of chromosomes, thus providing new evidence of the existence of quadruplex in a cellular context [82]. Pyridodicarboxamide derivatives have also been shown to induce the formation of tetramolecular quadruplexes and act as molecular chaperones, thereby proving their efficiency as G-quadruplex binders [83]. Moreover, the pyridodicarboxamide family of ligands has been extended with the synthesis of phenanthroline analogues. Phenanthroline-DC (Fig. 5) showed a perfect geometrical match with a G-quartet and was found to be remarkably more selective than telomestatin, thus confirming the great potential of bisquino- linium compounds [84,85]. Metallo-organic complex as G-quadruplex ligand A class of metallo-organic complexes has emerged as highly interesting molecules because of their easy synthetic access and their promising G-quadruplex binding properties. This approach is based on the hypothesis that the central metal core could be posi- tioned over the cation channel of the quadruplex, thereby optimizing stacking interactions between the surrounding chelating agent and the accessible G-quartet [30]. The presence of a cationic or highly polarized nature is also a further advantage in promoting an association with the negatively charged G-quadruplex DNA. The first reported examples described the inser- tion of a metal in the central cavity of TMPyP4 and their use as Cu(II)– [86,87], Mn(III)– [88], etc. A spec- tacular 10 000-fold selectivity for quadruplex over duplex has been measured by SPR for the highly cationic Mn(III)–porphyrin complex [88]. Moreover, Cu(II)– and Pt(II)–terpyridine complexes can also be obtained in one-step or two-step processes and these ligands possess high affinity and high selectivity for the G-quadruplex [89]. Recently, a series of platinum(II) complexes containing dipyridophenazine (dppz) and C-deprotonated 2-phenylpyridine (N-CH) ligands have been developed and their G-quadruplex DNA-binding potential assayed. [PtII(dppz-COOH)(N-C)]CF3SO3 (1; dppz-COOH = 11-carbotxydipyrido [3, 2-a: 2¢, 3¢-c] phenazine) binds G-quadruplex DNA through an external end-stacking mode with a binding affinity in the order of 10 7 m )1 . Using a biotinylated-primer extension telomerase assay, the same molecule was also shown to be an effective inhibitor of human telomerase in vitro, with an IC 50 value of 760 nm [90]. Neutral ligands The category of neutral ligands is not the largest but it includes the paradigm for G-quadruplex recognition, namely telomestatin (Fig. 6). This is isolated from Streptomyces annulatus [91] and has been extensively studied because it appears to be one of the most interesting G-quadruplex ligands [10,91–100]. Indeed, this molecule greatly stabilizes the G-quadruplex and appears to be one of the most selective G-quadruplex ligands: > 70-fold. The total absence of an affinity for N HN H 2 N NH 2 N + NN N + NH 2 N N H 2 N N H N N N N H dnagiL95421desabenizairTsevitav iredmuidihtE N N H HN O O N N N HHN N N NH HN N N O O PDC core PhenDC 3 Fig. 5. Chemical structures of ethidium derivatives as G-quadruplex ligands: ethidium derivatives, 2, 4, 6-triamino-1, 3, 5-triazine derivative (12459), pyridodicarboxamide (PDC) core and phenanthroline analogues (Phen-DC 3 ). S O O O N N N O ON O N O N N R ON N O N NH N N R HN O O O N O N O O N O O R = CH(CH 3 ) 2 R = CH 2 OH Telomestatin Hexa-oxazole macrocycles Fig. 6. Chemical structures of neutral ligands: telomestatin and hexa-oxazole macrocycles. A. Arora et al. Quadruplex interacting molecules FEBS Journal 277 (2010) 1345 ª 2009 The Authors Journal compilation ª 2009 FEBS duplex DNA because of its neutral character and cyclic shape justifies the initial strong bend towards this molecule. It has also been reported to inhibit the proliferation of telomerase-positive cells, by modifying the conformation and length of the telomeres, and the dis- sociation of telomere-related proteins from telomeres. Nevertheless, two major drawbacks of telomestatin are that it is difficult to obtain and it has poor water solubility. The complete synthesis of telomestatin seems to be highly complex and is not compatible with large- scale preparation [101]. There are also a few reports in literature on macrocyclic hexaoxazole ligands and their interaction with telomeric G-quadruplex [102]. HXDV and HXLV-AC (Fig. 6) are two synthetic hexaoxazole-containing macrocyclic compounds which have been characterized for their cytotoxic activities against human cancer cells. Their detailed binding and thermodynamics of interaction with the intramolecular (3+1) G-quadruplex structural motif formed in the presence of K + ions by human telomeric DNA has also been reported [103]. Examples of ligands belonging to different classes are also summarised in Table 1. Mode of action of telomeric G-quadruplex binding ligands The unlimited proliferative potential of cancer cells depends on telomere maintenance, which in turn makes telomeres and telomerase an attractive target for cancer therapy [104]. Most telomere-targeted antitumour strategies address the telomerase-dependent mechanism of telomere maintenance. It is well reported that formation of intramolecular G-quadruplexes by the telomeric G-rich strand inhibits telomerase activity [105]. Therefore, ligand-induced stabilization of intramolecular telomeric G-quadruplexes provides an attractive strategy for the development of anticancer agents. Molecules that target telomeric DNA were initially considered to be telomerase inhibitors [24,106,107]. However, this strategy cannot be considered feasible as a cancer-specific approach because normal cells also have telomeres and bear quadruplex potential. Nevertheless, it is possible that telomeres from normal and cancer cells exhibit differ- ences in structure or accessibility and that a telomere ligand could exhibit selective toxicity. In this minireview, we also discuss the role of G-quadruplex binding ligand on telomerase enzyme, as well as the direct effect on telomeres, thus altering telomere maintenance. Proteins of the telomerase complex More than 30 proteins have been proposed to be associated with the telomerase enzyme complex (Table S1 in [108]). It has been shown that the active complex is composed of three different components, hTERT, hTR and dyskerin, with two copies of each. However, regu- lation of the relative expression levels of hTERT, hTR and dyskerin is poorly understood. It is known that methylation of the promoter and 50 exons may lead to repression of hTERT expression in normal cells [109]. Recently, evidence for a molecular link between choles- terol-activated receptor Ck and hTERT transcription has been reported [110]. Moreover, it has also been shown that the core hTR promoter is activated by Sp1 and is repressed by Sp3 [111,112]. Table 1. Examples of ligands belonging to different classes. Class Compound N-methylated ligands TMPyP4, TQMP68, 3,4-TMPyPz, tetramethylpyridiniumporphyrazines, corroles, Se 2 SAP Ligands with protonable side arms Bisamidoanthraquinone, fluorenone, acridone, acridine, BSU6039, BRACO-19, benzylamino-acridine, perylene diimide PIPER Daunomycin, distamycin and netropsin, flavonoids, steroids RHPS4, MMQ3, MMQ1, BOQ1, indoloquinolines, cryptolepine analogues, quindolines and triazacyclopentaphenantrene Alkaloid-based ligands Berberine derivatives, palmatine sanguinarine, isoindigotone Click chemistry-based ligands Bistriazole derivatives, neomycin-capped quinacridine, tri-oxazole macrocycles, isoalloxazines, diarylurea-based ligands and substituted derivatives, trisubstituted acridine–peptide conjugates, furan-based macrocycles Ethidium derivatives Triazine 12459, pyridodicarboxamide core containing 307A and 360A, tritiated 360A phenanthroline analogues Metallo-organic complex Cu(II)–TMpyP4 complex,Mn(III)–TMPyP4 complex, Cu(II) and Pt(II)–terpyridine complexes, PtII(dppz-COOH)(N-C)]CF 3 SO 3 Neutral ligands Telomestatin, hexaoxazole-containing macrocyclic HXDV and HXLV-AC Quadruplex interacting molecules A. Arora et al. FEBS Journal 277 (2010) 1345 ª 2009 The Authors Journal compilation ª 2009 FEBS Proteins involved in the protection of telomere extremities (shelterin/telosomes) During the last decade, proteins that protect the telomere extremities have been identified and ascertained to make up a complex called the telosome [113] or shelterin [114]. This complex is principally composed of six proteins. Three of these bind directly to the telomeric repeats: TRF1, TRF2 and POT1. TRF1 and TRF2 have Myb-type DNA-binding domains [115], whereas POT1 has two oligonucleotide ⁄ oligosaccharide binding domains and displays a strong preference for the single-stranded telomeric sequence relative to double-stranded DNA [116]. The three other proteins are TIN2, which binds TRF1 and TRF2, TPP1, which binds TIN2 and POT1, and Rap1, which binds TRF2 [117]. TRF2 is shown to be involved in strand invasion and T-loop formation [118,119] and in combination with telomerase deficiency, has been most strongly implicated in carcinogenesis [120]. Overexpression of TRF2 also reduces telomeric, but not genomic, single- strand break repair [121]. Recently, it has been shown that TPP1 ⁄ POT1–telomeric complex increases the activity and processivity of the human telomerase core enzyme [122,123]. Shelterin partners also associate with Apollo, a novel component of the human telomeric complex that, along with TRF2, appears to protect chromosome termini from being recognized and pro- cessed for DNA damage [124,125]. Inactivation of either TRF2 or POT1 [126,127] leads to variation in the overall length of the single-stranded G overhang, aberrant homologous recombination [128] and also induces a specific DNA damage response at most telomere ends [129]. These studies are consistent with the view that telomere ends are engaged in a peculiar structure in order to protect their integrity. Possible role of G-quadruplex binding ligands as telomerase inhibitors G-quadruplex ligands were first evaluated as telomerase inhibitors and could conceivably induce telomere shortening and replicative senescence [130]. Long-term treatment of human cancer cells with subtoxic doses of disubstituted triazines or telomestatin induces progres- sive telomere shortening that correlates with the induction of senescence [77,79,131,132]. This telomere shortening may be the result of inhibition of telomerase and ⁄ or telomere replication. A similar pheno- menon was noticed in human cells treated with telomestatin, a new steroid derivative, and BRACO-19 [46,95,133]. This may intuitively be the result of telomerase inhibition, but as we discuss below, such short- term loss may also be the result of telomere replication inhibition and ⁄ or telomere dysfunction. Some of these ligands were able to downregulate telomerase expression in treated cells [79,131,134,135]. Direct effects of G-quadruplex ligands on telomeres: induction of telomere dysfunction Earlier studies have demonstrated a short-term response (apoptosis) induced by G-quadruplex ligands that could not be explained solely by telomerase inhibition [79,81,131]. After just 15 days of exposure, subtoxic concentrations of the G-quadruplex ligands RHPS4 or BRACO-19 can trigger growth arrest in tumor cells, before any detectable telomere shortening occurs [134,135]. Modifications of hTERT or hTR can interference with the telomere capping function, which in turn leads to short-term and massive apoptosis. Overexpression of either hTERT or a dominant nega- tive of hTERT in a telomerase-positive cell line evidently does not modify the antiproliferative effect of the triazine derivative 12459 (formula shown in Fig. 5) [77]. The observation that BRACO-19 causes chromosome end-to-end fusion marked by the appearance of p16-associated senescence led to the idea that G-quadruplex ligands act primarily to disrupt the telomere structure [136]. Such telomeric dysfunction was also observed in cell lines treated with other quadruplex ligands and in cell lines resistant to a triazine derivative, as evidenced by the typical images of telophase bridges [81,137,138]. These studies suggest that the direct target of these ligands is the telomere rather than telomerase. DNA damage pathways induced by G-quadruplex ligands Telomeres effectively prevent the recognition of natural chromosome ends as double-stranded breaks. It has previously been shown that telomere shortening or the loss of protective factors such as TRF2, TIN2 and POT1 activates a DNA damage response pathway [139]. In addition, initiation of a DNA damage pathway was demonstrated in BCR-ABL-positive human leukemia cells after telomestatin treatment characterized by the phosphorylation of ATM and Chk2 [131]. A similar DNA damage response ensues subsequent to telomestatin treatment in HT1080-treated cells as evidenced by the formation of gH2AX foci that par- tially co-localize at the telomere, thus suggesting the induction of telomeric dysfunction [95]. A similar gH2AX response is elicited in the nucleus of UXF1138L uterus carcinoma cells upon the interaction A. Arora et al. Quadruplex interacting molecules FEBS Journal 277 (2010) 1345 ª 2009 The Authors Journal compilation ª 2009 FEBS of RHPS4 [140]. The triazine derivative 12459 also induces either senescence or apoptosis in the human A549 pulmonary carcinoma cell line, in a concentra- tion- and exposure time-dependent fashion [141]. Future perspectives Although synthesis of small molecules to target the quadruplex is attracting attention, there are many challenges which demand greater efforts if we are to devise strategies for rational drug design with high selectivity. The common features that most quadruplex-interacting ligands display are: (a) direct stacking with quartets, (b) loop interactions or external stacking, and (c) interactions between ligand substituents ⁄ side chains and the phosphate backbone of quadruplexes. These properties can be optimized for different quadruplexes, which display variations in loop length, composition and topologies, so as to achieve discriminative quadruplex targeting. A recent study addressed the issue of ligand selectivity by examining the distinct loop geometry in a bimolecular quadruplex of Oxytrichia [142]. However, the limited structures available for quadruplex–ligand complexes retard this exploratory strategy of drug design for biologically relevant quadruplex structures. To circumvent this limitation there is the need to adopt an integrative approach involving molecular dynamics and biophysical techniques to obtain rapid and accurate screening of quadruplex-interacting ligands. Virtual screening of chemical libraries by molecular docking is one attractive approach adopted to identify potential scaffolds. The hits obtained from the in silico search can be validated further through biophysical methods involving spectroscopic and calorimetric measurements, giving a quantitative idea of the thermodynamic stability of the complex. Quadru- plex-forming sequences have an inherent ability to adopt diverse structures, which are influenced by their loop length and composition. Therefore, a systematic study of quadruplex–ligand interaction involving quadruplexes of varying loop length and composition is required. Such an attempt has been made for telomeric, c-myc and c-kit quadruplex–porphyrin interactions, thereby establishing the influence of loop length and composition in perturbing molecular recognition of the quadruplex and its interaction with ligand. However, this dataset needs to be extended for other biologically relevant quadruplexes. Another notable observation is that ligands which demonstrate a promising perfor- mance in vitro usually display poor biological efficacy. This inconsistency between in vitro and in vivo results can be attributed to molecular crowding conditions in the cell. Most in vitro experiments neglect the role of molecular crowding agents, which have a major influence on the structure of nucleic acids and the interaction with their partners. Molecular crowding agents perturb the quadruplex–ligand interaction by influenc- ing the participation of water molecules. Therefore, an important parameter that should be taken into account during molecular dynamics and biophysical studies is hydration and the associated changes in heat capacity upon complex formation. Lastly, the in vitro knowledge generated should be extrapolated to the cellular level to evaluate the therapeutic potential of the ligands inhibiting telomerase and ⁄ or perturbing the molecular recognition of quadruplexes and competing with transcription factor binding. Efficient molecular assays should be developed for the accurate estimation of inhibitory effects and related toxicity with appropriate control experiments [84]. These molecular assays should be combined with glo- bal transcriptomic, proteomic profiling and tumor modeling studies for best candidate ligands to understand their therapeutic efficacy. Because this quadruplex–ligand field is booming, both chemists and biologists in conjunction could provide new molecular principles that may contribute to the emergence of effective anticancer therapies. Acknowledgement Financial support for this work from the Department of Science and Technology (Swarnajayanti project), Government of India, New Delhi to SM is gratefully acknowledged. References 1 Fry M (2007) Tetraplex DNA and its interacting proteins. Front Biosci 12, 4336–4351. 2 Maizels N (2006) Dynamic roles for G4 DNA in the biology of eukaryotic cells. Nat Struct Mol Biol 13, 1055–1059. 3 Ghosal G & Muniyappa K (2006) Hoogsteen base- pairing revisited: resolving a role in normal biological processes and human diseases. Biochem Biophys Res Commun 343, 1–7. 4 Fakhoury J, Nimmo GA & Autexier C (2007) Harness- ing telomerase in cancer therapeutics. Anticancer Agents Med Chem 7, 475–483. 5 Kelland LR (2005) Overcoming the immortality of tumour cells by telomere and telomerase based cancer therapeutics: current status and future prospects. Eur J Cancer 41, 971–979. 6 Lundblad V (2002) Telomere maintenance without telomerase. Oncogene 21, 522–531. Quadruplex interacting molecules A. Arora et al. FEBS Journal 277 (2010) 1345 ª 2009 The Authors Journal compilation ª 2009 FEBS