Sunteți pe pagina 1din 15

Biochimie 93 (2011) 1252e1266

Contents lists available at ScienceDirect

Biochimie
journal homepage: www.elsevier.com/locate/biochi

Mini-review

Molecular modeling of drugeDNA interactions: Virtual screening to structure-based design


Dik-Lung Ma a, *, Daniel Shiu-Hin Chan a, Paul Lee a, Maria Hiu-Tung Kwan a, Chung-Hang Leung b, **
a b

Department of Chemistry, Hong Kong Baptist University, Hong Kong, China Centre for Cancer and Inammation Research, School of Chinese Medicine, Hong Kong Baptist University, Kowloon Tong, Hong Kong, China

a r t i c l e i n f o
Article history: Received 19 January 2011 Accepted 1 April 2011 Available online 19 April 2011 Keywords: DrugeDNA interactions Virtual screening Structure-based design G-quadruplex DNA

a b s t r a c t
Virtual ligand screening (VLS) and structure-based design are strategies that have been routinely used for the development of pharmaceuticals, particularly those targeting enzymes and other protein targets. In recent years, an increased understanding of the role played by nucleic acids in biological systems made DNA an alternative candidate for the development of new drugs. This review highlights some successful applications of molecular modeling in virtual ligand screening and structure-based design of organic and inorganic molecules that target non-canonical nucleic acid structures such as G-quadruplex and triplex DNA. 2011 Elsevier Masson SAS. All rights reserved.

1. Introduction Deoxyribonucleic acid (DNA) is a long polymer consisting of repeating nucleotides (adenine, guanine, cytosine and thymine) that primarily serves as the functional unit for the storage of genetic information in living organisms. Since the seminal discovery of the double helix by Watson and Crick in 1953 [1], other DNA secondary structures such as G-quadruplex [2], i-motif [3,4] and triplex [5] DNA (Fig. 1) have been identied. These DNA secondary structures are putatively involved a number of biological processes such as the regulation of gene expression [6e10]. Thus, the targeting of non-canonical DNA nucleic acid structures such as triplex and G-quadruplex DNA with organic or inorganic compounds has been viewed as an alternative therapeutic approach for the treatment of cancers [6,10e12]. Organic and inorganic molecules can interact with nucleic acids through a number of different modes, including groove binding [13,14], pep stacking [15] and intercalation [16,17]. These interactions can alter the structural and functional properties (i.e. size, charge and stability) of the nucleic acids, affecting their interaction with other biomolecules such as proteins. The stabilization of G-quadruplex DNA found in the promoter regions of a number of
* Corresponding author. Tel.: 852 3411 7075. ** Corresponding author. Tel.: 852 3411 2016. E-mail addresses: edmondma@hkbu.edu.hk (D.-L. Ma), duncanl@hkbu.edu.hk (C.-H. Leung). 0300-9084/$ e see front matter 2011 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.biochi.2011.04.002

proto-oncogenes such as c-myc [18,19], c-kit [20,21], bcl-2 [22,23], hif1 [24] and VEGF [25,26] has been well described in literature. While platinum(II)-based compounds have been widely used to target duplex DNA (e.g. anti-cancer activity [27] or inhibition of topoisomerase activity [28,29]), the selective targeting of non-canonical nucleic acid conformations offers fresh avenues for therapeutic intervention based on the structural heterogeneity of DNA topologies. One of the challenges in the development of organic and inorganic molecules that target nucleic acids is the ability to selectively bind specic nucleic acid structures with high afnity. The selectivity and binding afnity of these compounds toward the desired nucleic acid structure can often be ne-tuned by structural modication of the candidate molecule, especially through the addition of functionalized side chains or substituents. Given the enormous chemical diversity available in the design of therapeutic molecules, to investigate experimentally all the different combinations of side chains, substituents and scaffolds would require vast amounts of chemical resources and manpower. On the other hand, structurebased methods allow the exploration of the chemical space at much lower cost. Exponential advances in computing power and technology, together with continual improvements in docking algorithms, have propelled in silico screening and modeling techniques to the forefront of drug discovery and development research. We discuss in this review the application of molecular modeling methods to the structure-based identication and design of organic and inorganic compounds that selectively target nucleic acid

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266

1253

Fig. 1. Solid state structures of: (a) triplex [104]; (b) G-quadruplex (PDB: 2KKA) [105]; and (c) i-motif (PDB: 1BQJ) DNA [106]. Structures were downloaded from the Protein Data Bank and visualized using Accelrys Discovery Studio.

structures. Molecular modeling allows the rapid screening of a large library of compounds against a nucleic acid target, in order to identify candidate compounds with favorable binding interactions which can be further enhanced through structure-based modication and development. Examples of the successful application of virtual ligand screening and structure-based design will be drawn especially from our own experience as well as that of others, in the hope that the reader will be able to gain an appreciation of the powerful capability of in silico modeling methods in complementing and augmenting the drug discovery and development process. 2. Virtual screening Developments in structural genomics have led to rapid increases in the number of high resolution three-dimensional structures of protein structures, many of which are observed to be bound to small organic or inorganic molecules [30,31]. These advances have been driven by developments in high-throughput X-ray crystallography [32e34] and nuclear magnetic resonance (NMR) spectroscopy [35,36]. They facilitated the application of molecular modeling in the identication and structure-based design of pharmaceuticals. One aspect that has received much attention is the application of virtual ligand screening to hit compound identication [37e39]. Virtual ligand screening involves the docking of a large library of compounds against a three-dimensional model of a biomolecule, typically a protein enzyme active site. Based on the hypothetical binding pose of the small molecule in the binding site, the stability or estimated binding energy is calculated through an analysis of the number of favorable bonding interactions as determined by a predened scoring function. Virtual ligand screening therefore allows a large collection of chemical structures to be screened in silico, and candidate compounds possessing high binding afnity against the desired protein target can be selected for biological testing or structural optimization, possibly in conjunction with structure-based design. Focused libraries

containing natural products or natural product-like structures also received particular attention due to the bioactive scaffolds and high structural diversity. Recent successful applications of virtual ligand screening include the identication of new natural product scaffolds for the inhibition of a number of biologically relevant proteins including tumor necrosis factor-alpha (TNF-a) [40], Nedd8-activating enzyme [41], histone acetyltransfereases [42] and phosphatases [43]. While virtual ligand screening methods have enjoyed success in the identication of new lead compounds for the inhibition of protein targets, the rise of non-canonical nucleic acid structures as potential therapeutic targets has engendered related studies on the application of in silico techniques for discovering novel nucleic acid binders. Nucleic acids potentially represent a better therapeutic target than mRNA or proteins due to their relatively low abundance in cells and their involvement at the start of the amplication cascade [44]. Until recently, the application of virtual ligand screening methods to nucleic acids has been largely unexplored, due to the lack of understanding of the biological importance of different secondary structures such as triplex and G-quadruplex DNA. As new evidence emerges that links the stability and/or formation of triplex and G-quadruplex DNA structures with anticancer activity, the molecular modeling and virtual screening of chemical libraries against nucleic acid targets offer the possibility of identifying novel chemotherapeutics potentially possessing high selectivity and efcacy. 2.1. Triplex DNA intercalation Triplex DNA is a form of DNA secondary structure that is comprised of a single DNA strand occupying the grooves of a DNA duplex, stabilized by Hoogsteen hydrogen bonding [45]. Since its discovery, the manipulation of triplex DNA has been considered as a potential alternative approach in modulating gene transcription [46e48]. The triplex structure can exist in an anti-parallel form or a parallel form. In the anti-parallel form, the third DNA strand is

1254

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266

hydrogen bonded to the duplex DNA in an anti-parallel arrangement. In the parallel form, the third strand is hydrogen bonded in a parallel orientation with the duplex DNA. The triplex structure can be stabilized or induced by the addition of small molecules or single-stranded oligonucleotides. The ability to recognize chromosomal DNA in cells using triplex-forming oligonucleotides (TFOs) has potential applications for the specic inhibition or activation of gene expression, introduction or correction of mutations, and analysis of chromosome structure and function. For example, the articial introduction of oligonucleotides has been reported to induce the formation of the triplex structure and downregulate c-myc expression [49,50]. However, TFOs tend to exhibit reduced inhibition activity within the cellular environment as a result of poor cellular localization and stability [51e54]. This has subsequently limited the widespread application of TFOs as therapeutic agents for modulation of gene expression. One approach to circumvent this problem has been the use of modied nucleic acid antigenes that are able to more effectively target chromosomal DNA. Corey and co-workers have pioneered the application of peptide nucleic acids (PNAs) and duplex RNA antigenes that are able to enter the nucleus and target chromosomal DNA [55]. Coreys group have also reported the targeting of chromosomal DNA using locked nucleic acids (LNAs). LNAs contain a methylene bridge between the 20 -oxygen of the ribose with the 40 -carbon atom. This reduces the conformational exibility of the polynucleotide which lowers the entropic cost of hybridization, leading to an increased afnity for its complementary sequence [56]. An alternative approach has been to increase the stability of the DNA triplex structure through small molecule triplex-interacting compounds such as naphthylquinoline [57] and dibenzophenanthrolines [58] for the modulation of gene expression. However, their planar aromatic scaffolds allow intercalation into doublestranded DNA, leading to selectivity issues. Through extensive synthetic modications to the naphthylquinoline scaffold, Chaires et al. were able to produce derivatives which displayed increased binding afnity and selectivity toward a parallel dTedAedT DNA triplex [59]. Though this strategy eventually resulted in the generation of an efcient triplex DNA intercalator, signicant resources were expended for the synthesis of the numerous napthylquinoline derivatives in this study. To overcome this issue, Holt et al. proposed an alternative approach to the identication of selective parallel dTedAedT triplex DNA intercalators using virtual ligand screening [60]. The process utilized by Holt et al. involved two distinct steps. The initial screening process identied compounds structurally similar to the known triplex DNA intercalator, MHQ-12 (1) [59]. A score was assigned to each compound in the library based on a comparison with the three-dimensional topology and the alignment of 1 using Surex-Sim. The second part of the strategy involved the docking of the pre-screened compounds against the minoremajor grooves, majoremajor grooves, minor grooves and intercalation sites in SurexeDock. From this two-step virtual ligand screening process, compounds 2 and 3 were identied as potential triplex intercalators (Fig. 2). To validate the results of the virtual ligand screening process, a series of biophysical experiments such as competition dialysis and thermal denaturing studies were undertaken. The competition dialysis experiments showed that compounds 2 and 3 displayed a much higher afnity for the parallel dTedAedT triplex compared to the known triplex intercalators MHQ-15 and OZ-85H [59]. Although compounds 2 and 3 were slightly less selective for the triplex than compound 1 or MHQ-15, their much higher binding afnities meant that compounds 2 and 3 displayed the highest overall afnities and selectivities reported for triplex binders to date. In the thermal denaturing studies, two transitions

Fig. 2. Chemical Structures of the known triplex intercalator MHQ-12 (1), MHQ-15, OZ-85H, compounds 2 and 3 identied from VLS [59,60].

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266

1255

corresponding to the denaturing of triplex DNA (w30  C) and the denaturing of double-stranded DNA (w70  C) were observed in the absence of small molecules. In the presence of compound 2 or 3, the rst transition occurred at w40  C, indicating the stabilization of triplex DNA by the small molecules. It is interesting to note that the intercalation of compounds 2 and 3 does not signicantly alter the denaturation temperature of the DNA duplex, suggesting that these compounds do not interact extensively with duplex DNA. 2.2. Groove binding Small molecules which can bind to nucleic acid grooves have the potential to achieve higher selectivity against their desired targets compared to DNA intercalators, and these compounds have been shown to possess a number of biological activities including anticancer and anti-protozoal properties [61e63]. With the increasing availability of X-ray co-crystal structures of nucleic acids complexed with groove-binding compounds, the application of virtual ligand screening for the identication of new groove-binding compounds is now feasible as demonstrated by Evans et al. [64]. While there are a number of commercially available software programs available for the virtual ligand screening of protein-based targets, there are currently only few examples of the application of these programs for the identication of nucleic acid-interacting compounds, particularly groove binders. Evans et al. examined the ability of DOCK and AutoDOCK to accurately predict the binding modes of known groove-binding compounds as a proof-of-concept. From the Protein Data Bank (PDB), 28 co-crystal structures of small molecules bound to the d(CGCGAATTCGCG)2 sequence were identied and used to evaluate the ability of these programs to predict the binding mode of groove-binding compounds. For the docking studies, a single crystal structure (PDB: 1VZK [65]) that represented an average of the 28 DNA structures was chosen. The ability of the software to correctly predict the binding pose of groove-binding compounds was quantied by the RMSD between the predicted and the experimentally derived structures. The results of the virtual ligand screening of the 28 known groove binders using DOCK and AutoDOCK showed that only 47% and 57% of the compounds, respectively, gave a predicted pose which lied within 2 of the solid state structure. This result was considered by Evans et al. to be unsatisfactory, and was likely attributed to the deciencies of the scoring algorithm used in the respective software programs. Despite this, the programs displayed potential for use in virtual ligand screening for the identication of new groove binders from a library of compounds. From the ZINC database, 9216 compounds with similar electronic properties and molecular weight were chosen as a random set of structures and tested alongside the 28 known groove-binding compounds for virtual screening. This resulted in an enrichment value SE (f 1%) of 86% for AutoDOCK and 57% for DOCK, where the enrichment value is dened to be the ratio of the number of selected known binders to the total number of known binders where the highest-scoring 1% of compounds are selected. These enrichment factors were comparable to those involving similar proteineligand studies. This result suggests that while molecular modeling may not always reproduce the crystallographic binding pose for groove-binding small molecules, the high enrichment factor observed in the virtual ligand screening suggests that these programs are capable of identifying unknown groove binders from chemical databases. However, this study by Evans et al. was only based upon one DNA sequence and did not take into account other DNA topologies. Holt et al. examined the use of AutoDOCK and Surex for virtual ligand screening of nucleic acid targets [66]. The molecular docking programs were tested for their ability to accurately reproduce the binding conformations of small molecule intercalators or groove-

binding ligands. Daunorubicin and ellipticine were selected as intercalators while distamycin and pentamidine were chosen as minor groove binders. A series of docking orientations between the two programs were compared, and some of the potential docking challenges and pitfalls were explored such as the importance of hydrogen treatment on ligands and the scoring functions of the two programs. Importantly, both programs were able to dock daunorubicin, distamycin and pentamidine to within 2 RMSD of the crystallographic structure, and ellipticine to within 3 RMSD. Ellipticine was noted by the authors to represent an especially challenging ligand for docking due to its small binding pocket in the nucleic acid and the tight t associated with the binding of the ligand into the intercalation site. Overall, the docking results show that the docking accuracy and pose ranking of the two programs are quite similar. However, the authors concluded that Surex was superior over AutoDOCK for virtual screening due to its faster performance and more user-friendly interface. This pivotal work by Holt et al. showed that these molecular docking techniques can be used on biologically relevant nucleic acid targets for the potential development of therapeutic small molecules [66]. The studies by Evan et al. and Holt et al. were important proof-ofconcepts that virtual ligand screening could be effectively applied in the identication of groove-binding ligands. Cosconati et al. have recently employed virtual ligand screening to identify G-quadruplex groove binders [67]. G-quadruplex-forming sequences have been located in the promoter regions of oncogenes that are involved in the regulation of cell division and differentiation, cell cycle and apoptosis. The over-expression of such oncogenes have been linked to the proliferation of cancers. Small molecules that can bind to and stabilize the G-quadruplex have the potential to inhibit the expression of oncogenic activity. The rst generation of G-quadruplex ligands discovered typically contained a planar aromatic scaffold that allowed potential pep stacking interactions with the terminal G-quartets. However, due to the differences in the dimensions of the groove regions between different nucleic acids, and between different G-quadruplex topologies, groove binders have the potential to exhibit greater selectivity. While synthetic G-quadruplex groove binders have been described [68,69], the application of large scale virtual ligand screening to identify additional G-quadruplex groove binders has only recently been reported. Using AutoDOCK and a solid state crystal structure of the G-quadruplex d(TGGGT)4 (PDB code: 1S45 [70]), a commercially available library of 6000 structures was screened to identify new groove-binding compounds. The structures were ranked according to their predicted free binding energies, and 137 compounds possessed predicted binding energies lower than the cut-off energy of 6.0 kcal/mol. A second round of selection was enforced by visual inspection of the predicted pose and those compounds that did not display a signicant interaction (electrostatic or hydrogen bonding) with the grooves of the G-quadruplex DNA were discarded, leaving only 30 compounds. These compounds were tested experimentally by 1H NMR spectroscopy titration experiments with d(TGGGT)4 and it was found that 24 of these compounds were false positives as they did not interact signicantly with the grooves of the G-quadruplex. The remaining six compounds 4e9 (Fig. 3) were conrmed as Gquadruplex groove binders. The 1H NMR titration experiments for compounds 4e9 with G-quadruplex d(TGGGT)4 showed that there were signicant perturbations in the nucleotides comprising the grooves of the G-quadruplex, suggesting the binding of the small molecule to these regions. In addition, based on the observed changes in chemical shifts, all of the compounds were determined to interact more closely with the 30 -terminus of the G-quadruplex, consistent with the results of the molecular modeling. This important study from Cosconati et al. demonstrated for the rst time that virtual

1256

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266

Fig. 3. Chemical structures of G-quadruplex groove binders 4e9 identied from virtual ligand screening against d(TGGGT)4 [67].

ligand screening could be a highly effective method for the identication of novel G-quadruplex groove binders from a database of chemical structures. Due to the pre-selection performed in silico, which was based on the molecular modeling of existing structural data, the hit rate achieved in the NMR experiments (20%) was far better than those realized in conventional high-throughput screening campaigns.

In order to further develop and rene novel G-quadruplex groove-binding compounds, a detailed understanding of the interaction between the ligand and the G-quadruplex is required. To that end, Cosconati et al. recently published a study investigating the interactions of the intermolecular G-quadruplex d(TGGGGT)4 with compound 10, which was derived from the known G-quadruplex groove binder distamycin A 11 (Fig. 4a and b) [71]. The d(TGGGT)4 G-quadruplex was chosen as it displays a high level of structural symmetry, which signicantly reduces the number of nucleotide signals in the NMR spectra. Compound 10 differs from distamycin 11 by the replacement of the positivelycharged amidinium group in 11 with a neutral N-methyl amide derivative, allowing the inuence of Coulombic interactions on groove-binding compounds to be investigated. Analysis of the interactions of 10 with d(TGGGGT)4 by 1H NMR spectroscopy revealed that the resonances within the intermolecular G-quadruplex were shifted signicantly upon the addition of 10, implying that 10 was able to interact with the G-quadruplex structure. Detailed analysis of the NOE contacts showed that compound 10 had signicant NOE correlation with residues G4, G5 and T6 (Fig. 4c), suggesting that the compound is bound via the grooves of the intermolecular G-quadruplex. The same result was observed in the interaction between distamycin A and the intermolecular Gquadruplex d(TGGGGT)4 that was reported in an earlier study [72]. To provide additional evidence that compound 10 was bound to the grooves of the G-quadruplex, Cosconati et al. introduced a bromine substituent into the center of the grooves of the Gquadruplex. In this case, it was found that there was no signicant change in the 1H chemical shifts when compound 10 was added. Taken together, the results suggested that compound 10 was bound to the grooves of the G-quadruplex, with a preference for the 30 -terminus. In order to develop a better understanding of the interaction between 10 and d(TGGGGT)4, a three-dimensional model was constructed by performing multidimensional NMR experiments including NOESY, PE-COSY and zTOCSY. Based on

Fig. 4. Chemical structures of: (a) distamycin A derivative 10; (b) distamycin A (11); Schematic representation of: (c) the intermolecular G-quadruplex d(TGGGGT)4.

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266

1257

compound 10 was able to adopt a crescent shape to match the prole of the G-quadruplex grooves in order to maximize the number of favorable interactions. At the 30 -terminus, a number of hydrophobic interactions between the T6 residue and the pyrrole rings of the dimer were also present. Through this study, Cosconati et al. highlighted a number of key interactions which controls both the binding afnity and also the orientation of the ligand with DNA. In addition, the generation of a three-dimensional structure of a groove-binding ligand bound to the G-quadruplex will facilitate the discovery of additional novel groove binders in the future. It will be interesting to see if the molecular modeling methods of Cosconati et al. could be applied to biologically relevant G-quadruplexes, in order to identify and develop groove-binding agents for the potential treatment of cancer. 2.3. G-quadruplex binding ligands With a greater understanding of the link between G-quadruplex DNA and the regulation of gene expression, especially of oncogenic genes that regulate key processes in cell division, cell cycle maintenance and apoptosis, G-quadruplexes have received increased attention as a potential target in the treatment of cancers. Organic molecules such as cationic phorphyrins [73e75], quindolines [76,77], triarylpyridine [78] and isoalloxazines [79] derivatives have been developed which are able to bind and stabilize G-quadruplex DNA to modulate gene expression. Despite these successes, the use of virtual ligand screening to identify novel G-quadruplex-binding ligands has only been recently reported. The use of virtual screening

Fig. 5. Chemical structures of 1H-pyrazole-3-carboxy-4-methyl-5-phenyl-(1H-indol3-ylmethylene)hydrazide (12), quindoline derivative (13) and 2-[[(4-ethoxyphenyl) methylene]-amino]-4,5,6,7-tetrahydrobenzo[b]thiophene-3-carbonitrile (14) [80].

these experiments, the authors postulated that compound 10 existed as an anti-parallel dimer in solution, with a number of hydrogen bonding interactions between the inner ligand (10in) and the nucleotides of d(TGGGGT)4. It was also determined that

Fig. 6. Molecular models showing the interaction of 12 with the intramolecular human telomeric G-quadruplex [80].

1258

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266

to identify G-quadruplex groove-binding compounds has been described in the previous section and will not be discussed here. One of the rst examples of the application of virtual ligand screening to the identication of G-quadruplex-binding ligands was reported by Ma et al. in 2008 [80]. In this work, virtual ligand screening was used in a two-step process to identify compounds binding to the human telomeric G-quadruplex DNA from a library of drug-like structures. One of the chief advantages of using virtual ligand screening to identify active molecules from a library of druglike compounds is the ability to generate and evaluate a large chemical diversity of structures that are potentially druggable without the need for actual synthesis. The authors utilized the X-ray crystal structure of the intramolecular human telomeric sequence (PDB code: 1KF1 [81]) for their virtual screening campaign. In the rst step, a library of 100,000 drug-like compounds were passed through Lipinski lters to remove compounds that would not be predicted to display ideal pharmacokinetic properties [82]. The remaining compounds were docked to a grid representation of the receptor and a score assigned quantifying the stability of the complex according to the ICM method (Molsoft) [83]. From the library of drug-like compounds, the 1H-pyrazole-3-carboxy-4methyl-5-phenyl-(1H-indol-3-ylmethylene)hydrazide (12, Fig. 5) was identied as a top candidate. To investigate the ability of 10 to stabilize the human telomeric G-quadruplex, a FRET melting analysis was performed. In the presence of compound 12, the melting temperature (Tm) of the human telomeric G-quadruplex increased by 18  C, which is in the same order of magnitude as that for the

known G-quadruplex-binding quindoline derivative 13 under comparable conditions. As a control experiment, a low scoring compound 2-[[(4-ethoxyphenyl)methylene]amino]-4,5,6,7-tetrahydrobenzo[b]thiophene-3-carbonitrile (14, Fig. 5) was also added and no change in the melting temperature was also observed, demonstrating that the virtual screening was capable of distinguishing between strong and weak binders to the G-quadruplex. In a parallel experiment, compound 12 did not increase the change of the melting temperature of a duplex sequence, suggesting that the small molecule was selective for G-quadruplex DNA. Additional molecular modeling (Fig. 6) showed that the compound was bound at the 30 terminus of the G-quadruplex with extensive interactions (presumably pep stacking) between the aromatic phenyl and indole scaffolds of 12 and the terminal G-quartet. By comparison, intercalation of 12 into the human telomeric G-quadruplex gave a considerably higher binding energy. Besides drug-like chemical libraries, natural products have also received wide attention as focused databases of biologically active structures. As an extension of the above work, Ma and co-workers have conducted virtual ligand screening of a natural product database in order to identify novel natural product scaffolds that are able to bind and stabilize the c-myc G-quadruplex [84]. Natural products represent a rich source of highly chemically diverse structures that have been rened over evolutionary time-scales for optimal interactions with biomolecules, and they have thus featured prominently in the development of a signicant proportion of new therapeutics [85e88]. The natural product

Fig. 7. Chemical structure of telomestatin (15) and the four lead compounds identied from the VLS 16e19 [84,89].

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266

1259

Fig. 8. Hypothetical molecular models showing the a) Side view; b) Top view of the interactions of fonsecin B 16 with the c-myc G-quadruplex structure [84].

telomestatin (15, Fig. 7) is one of the most potent G-quadruplexbinding ligands reported to date [89,90]. However, only solutionstate structures of the c-myc G-quadruplex were known, we constructed a molecular model of the c-myc G-quadruplex based on the X-ray crystal structure of the human telomeric G-quadruplex. The loops were modied through insertion and deletion of nucleotides to correspond to the structure of the 1:2:1 loop isomer, which has been reported to represent the dominant topology of the c-myc G-quadruplex in solution. This model has also been previously used to study the interactions of quindoline c-myc G-quadruplex stabilizers [76]. Using this model, a library of 20,000 natural product and natural product-like structures were screened in silico and four hits were identied (16e19, Fig. 7). Of these, the anthraquinone Fonsecin B (16) showed the highest binding afnity toward the c-myc G-quadruplex in in vitro experiments. To verify the results of the virtual screening, both biophysical experiments and in vitro assays were performed to evaluate the binding and biological activity of the hit compound against the cmyc G-quadruplex. Using UVevisible absorption titration, the binding constant of 16 for the c-myc G-quadruplex was found to be an order of magnitude higher compared to that for duplex and single-stranded DNA, suggesting that the lead compound was selective for the G-quadruplex over other DNA secondary structures. To further examine the mode of binding, molecular modeling was performed using the molecular model of the c-myc 1:2:1 loop isomer G-quadruplex structure (Fig. 8). The molecular modeling results showed that compound 16 was bound to the 50 -terminus of the c-myc G-quadruplex presumably through pep stacking interactions with the terminal G-quartet. The binding energies for the end-stacking mode were signicantly lower than those predicted for intercalation. Interestingly, the molecular modeling results showed that the oxygen atoms in compound 16 may interact with the potassium cations in the central ionic channel, potentially contributing favorable Coulombic interactions to the binding energy. The ability of compound 16 to stabilize the c-myc G-quadruplex in a biological system was investigated using a PCR stop assay. The compound was found to inhibit Taq-polymerase-mediated DNA extension of the c-myc G-quadruplex

Fig. 9. Chemical structures of the acridine derivatives (20e22) containing different groove-binding groups (aei) [93].

1260

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266

sequence with a similar potency compared to the well-known G-quadruplex-binding ligand TmPyP4 [73e75] under comparable conditions. 2.4. Summary In this section, we have described successful applications of virtual ligand screening for the identication of natural products or synthetic compounds that can interact with nucleic acids including the G-quadruplex and triplex DNA, as well as groove binders to duplex and G-quadruplex DNA. Despite the concerns presented by Evans et al. regarding the ability of modeling software to accurately predict binding modes, virtual ligand screening still represents a powerful tool in identifying nucleic acid binders from a database of chemical structures. The successful application of virtual ligand screening in identifying novel compounds selectively binding to triplex and G-quadruplex DNA has been demonstrated by several research groups. After identication of the hit compounds, the next step is the hit-to-lead chemical modication and optimization of the candidate molecules to further improve their selectivity and efcacy against the desired target.

2.5. Structure-based design In the previous section, we highlighted a number of examples where virtual ligand screening was successfully employed to identify novel compounds capable of binding to and stabilizing particular nucleic acid structures such as G-quadruplexes and DNA triplexes. After the identication of a hit compound, the next step in the drug development process is optimization through chemical modications in order to improve the potency and selectivity of the compound against the biomolecular target. The nature of the substituents will affect the electronic, lipophilic and steric properties of the small molecules and will thus inuence their interactions with nucleic acids. In the conventional drug optimization process, a diverse range of analogs of the hit compound is required for biological evaluation. However, generating such a large number of compounds with sufcient structural diversity through traditional organic synthesis would require extensive resources in time, chemical reagents, and manpower that are only available to large pharmaceutical companies. Molecular modeling provides a convenient tool to study the effect of varying different substituents on the small molecules on

Fig. 10. Schematic drawings of a G-tetrad and the structures of complexes 23e32 [101].

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266

1261

their interactions with nucleic acids without the need for actual synthesis and testing of all possible compounds. Large libraries of diverse structures can be rapidly analyzed in silico at a vastly lower cost in time and resources compared to traditional screening techniques. Consequently, structure-based design has been increasingly applied to the development of both organic- and inorganic-based G-quadruplex-stabilizing compounds.

2.6. Organic G-quadruplex-stabilizing compounds Organic compounds containing planar polyaromatic ligands have served as the basic scaffold for numerous G-quadruplex ligands. This is due to the ability of the aromatic groups to interact with the terminal G-tetrads through potential pep stacking interactions. However, planar scaffolds may also possess the ability to intercalate into double-stranded DNA, potentially leading to reduced selectivity. One method to overcome this issue for acridine-based compounds has been through the modication of the acridine core. A number of acridine-based human telomeric G-quadruplex ligands including as BRACO-19 [91] and RHSP4 [92] possessing telomerase inhibition and anti-cancer properties have been reported in the literature. However, these compounds also display signicant binding afnity toward duplex DNA and other G-quadruplex structures, such as promoter G-quadruplexes. To address this problem, Sparapani et al. utilized a structure-based approach to design novel acridine-based compounds that were able to target the human telomeric G-quadruplex with superior selectivity [93]. Starting with the acridine core, it was proposed that the addition of heterocyclic rings, such as triazole, at the 3- and 6-positions of acridine would further enhance the potential pep stacking interaction with the terminal G-tetrads. In addition, libraries of acridines containing aliphatic linkers of different lengths and variable end groups [94] (Fig. 9) were designed to further enhance the interactions of the compounds with the G-quadruplex through selective groove-binding interactions. Molecular modeling was employed to examine the binding of these acridine derivatives to the human telomeric G-quadruplex. It has been previously shown that the human telomeric G-quadruplex exists in both the parallel and anti-parallel topologies in the solution state [81,95]. Thus, in order to properly investigate the interaction of the acridine derivatives 20e22 with the human telomeric G-quadruplex, the binding mode to both the parallel and anti-parallel structures was examined using molecular modeling. The results of the molecular modeling showed that the acridine derivative 21 bound to both the parallel and anti-parallel structures of the human telomeric G-quadruplex as indicated by the favorable calculated binding energies. The predicted binding mode revealed that the acridine core and the triazole ring systems were orientated toward the terminal G-tetrads of the human telomeric G-quadruplex, allowing for potential pep stacking interactions. In addition, the side chains of the small molecules interacted signicantly with the groove regions of the human telomeric G-quadruplex. To examine the selectivity of the acridine complexes toward other promoter G-quadruplexes, the acridine derivative 21 was docked against a molecular model of the c-kit 1 [96] and c-kit 2 [21] G-quadruplexes. The molecular modeling results indicated a signicantly reduced binding afnity toward both the c-kit 1 and c-kit 2 G-quadruplexes. Analysis of the predicted low-energy binding pose suggested the presence of steric clashes between the acridine derivative 21 and the 50 -terminus of the c-kit 1 and c-kit 2 G-quadruplexes, leading to reduced interactions with the terminal G-quartet and potentially weaker pep stacking interactions. At the 30 -terminus, in addition to the unfavorable steric interactions, the aliphatic linkers were found to be

insufciently long to bind effectively to the c-kit G-quadruplex grooves. To verify the results of the molecular modeling, the compounds 21aei were synthesized and their interaction with the human telomeric sequence was examined using a FRET melting assay. In the FRET melting assay, the compounds 21a and 21b showed the highest increase in the Tm (16  C) for the human telomeric sequence, while the other compounds in the series only displayed changes of 0e7  C. The acridine derivative 21a and 21b did not show any signicant stabilization of the c-kit 1, c-kit 2 and doublestranded sequences. Taken together, these results indicate that the acridine derivatives 21a and 21b were selective for stabilizing the human telomeric sequence, consistent with the results of the molecular modeling. This is a signicant improvement upon the previously published acridine compound BRACO-19 that lacked selectivity over other double-stranded DNA and promoter Gquadruplexes. 2.7. Inorganic G-quadruplex-stabilizing compounds While organic molecules have been extensively explored, there has been a recent interest in the development of inorganic compounds as G-quadruplex-stabilizing ligands. Metal complexes typically contain a metal center bound to organic ligands in a precise three-dimensional arrangement. Through modications of the steric environment and electronic nature of the organic ligands, the interaction of the resulting complex with a nucleic acid can be ne-tuned [97]. In addition, the synthesis of inorganic compounds is highly modular, contrasting favorably with the

Fig. 11. Molecular modeling of the interaction between 23 and (a) intermolecular G-quadruplex DNA d(T2AG3T)4, (b) the human intramolecular telomeric repeat G4A2-quadruplex, and (c) intramolecular telomeric G4A4-quadruplex [101].

1262

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266

lengthy multi-step syntheses required for organic compounds. As a result, inorganic complexes can be prepared in fewer steps and with greater exibility in the variation of the organic ligand during each step of the synthesis. Metal complexes such as platinum(II) bearing at aromatic ligands have been well documented to intercalate into doublestranded DNA and possess anti-cancer properties [98e100]. However, there have been comparatively few reported examples of the interaction of platinum(II) complexes with G-quadruplex DNA. The ability of platinum(II) and ruthenium(II) complexes to interact and stabilize human telomeric G-quadruplex DNA was examined by Ma et al. A series of metal complexes 23e32 bearing the molecular light switch ligand dipyridophenazine (dppz, Fig. 10) or their derivatives were synthesized and their interactions with G-quadruplex DNA were investigated using biophysical techniques [101]. UV melting and UVevisible absorption titration

experiments of the metal complexes also demonstrated the strong binding of complex 23 toward the G-quadruplex structure. Molecular modeling was utilized to investigate the binding mode of the complex 23 with G-quadruplex DNA. The interaction of complex 23 with three related intermolecular or intramolecular human telomeric G-quadruplexes were examined in detail (Fig. 11). The results of the molecular modeling revealed that complex 23 bound to all three human telomeric G-quadruplex structures at the 30 -terminus through extensive interactions (presumably pep stacking) between the G-tetrad and the dipyridophenazine ligand, consistent with the results of the NMR titration experiments. In all three cases, the intercalation of the complex into the G-quadruplex structure was calculated to be highly unfavorable, ruling out an intercalating binding mode. Inspired by the ability of platinum(II) complexes bearing planar aromatic scaffolds to stabilize human telomeric G-quadruplex DNA, Che and co-workers set out to design a series of new platinum(II)

Fig. 12. Chemical structures of the salphen-based platinum(II) complexes 33e34 bearing different functional groups [102].

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266

1263

c-myc G-quadruplex binders using an in silico structure-based approach [102]. As a starting point, two series of platinum(II) salphen-based complexes 33e34 were designed (Fig. 12) and their activity was examined using in vitro assays. From the initial experiments, complex 33a was identied as the most potent compound of the series. Molecular modeling was then used to optimize the lead compound 33a to improve the binding afnity and selectivity of the complexes against c-myc G-quadruplex DNA. Over 60 different derivatives containing 12 different functional groups and linkers of varying length were designed, and their interaction with the c-myc G-quadruplex was examined in silico using the ICM method [83]. The top-scoring compound identied from the in silico docking was complex 35 bearing an appended piperidine moiety (Fig. 13a). Molecular modeling indicated that both the unmodied hit complex 33a and the lead compound 35 were bound at the 50 -terminus with potential pep stacking interactions with the terminal G-tetrad. However, only complex 35 was able to form additional interactions with the grooves of the c-myc G-quadruplex through its longer side chains (Fig. 13b and c). The molecular modeling analysis suggested that the length of the aliphatic linker has a signicant inuence on the ability of the platinum(II) complexes to bind to grooves of the c-myc G-quadruplex, with the ethylene linker (as in 35) yielding the highest binding afnity. To verify the results of the structure-based design process, the most potent derivative was synthesized and its ability to inhibit Taq-polymerase-driven extension of a c-myc sequence through the stabilization of G-quadruplex DNA was examined. The IC50 for the lead platinum(II) complex 35 was estimated to be 4.4 mM, which is an order of magnitude lower compared to the hit complex 33a. This study successfully demonstrates the applicable of structure-based design for the hit-to-lead development of efcacious c-myc G-quadruplex stabilizers. Che and co-workers recently reported another example of the application of structure-based design for the development of platinum(II) complexes as a c-myc G-quadruplex stabilizers and down-regulators. A series of platinum(II) complexes (Fig. 14) were developed from a starting compound 36 that contained a more extended aromatic system [103]. The increased aromaticity was expected to enhance p-stacking interactions with the G-tetrad,

with the possibility of appending side chains for interactions with the groove regions of the G-quadruplex. A total of 550 compounds with different functional groups and varying linker lengths were designed and screened in silico, and complexes 37, 38 and 39 were identied as the lead candidates. The binding afnities of the lead compounds 36e39 were estimated using UVevisible titration experiments, and all the complexes showed the high selectivity toward G-quadruplex over double-stranded DNA. The ability of the compound to stabilize the c-myc G-quadruplex in vitro was examined using a Taq-mediated PCR stop assay. Complexes 37, 38a and 39a showed improved activity against Taq-mediated polymerase extension through stabilization of c-myc G-quadruplex DNA, with IC50 values of 2.2e6.5 mM, compared to 7.4 mM for the hit compound 36. Subsequent RT-PCR experiments also demonstrated that the lead complexes 37, 38a and 39a were able to signicantly inhibit the transcription of c-myc with improved efcacy compared to the parent compound. In this study, the in silico docking of the 550 derivative structures was a key step which allowed the rapid identication of the top candidates, which a dramatic reduction in the number of compounds actually synthesized. 2.8. Summary Virtual ligand screening represents an economical tool for the identication of promising scaffolds for targeting nucleic acid structures. Optimization of the hit compounds by traditional means entails the generation of a large library of chemical derivatives in order to develop a detailed structureeactivity relationship (SAR), necessitating a heavy investment into time, resources and manpower. Molecular modeling has recently emerged as a powerful technique for hit-to-lead optimization by allowing in silico exploration and analysis of the effects of different functional groups on the interactions between small molecules and their cognate nucleic acid targets. We have highlighted several recent studies that have employed molecular modeling for the structure-based design of organic- and inorganic-based compounds as G-quadruplex binders. Computer-aided docking allows the screening of a large range of derivatives in silico, allowing the characterization of optimal

Fig. 13. (a) Chemical structure of the optimized platinum(II) complex 35; (b) schematic diagram showing side view and (c) top view of the interaction of 35 with the c-myc Gquadruplex structure [102].

1264

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266

Fig. 14. Chemical structures of the platinum(II) complexes 36e39 [103].

functional groups that enhance binding afnity and/or selectivity, thus leading to improved biological activity. The net result is a tremendous reduction in the resources required for optimization and biological testing of the lead compounds. Given its signicant advantages, we expect to see many more successful applications of structure-based design for the development and optimization of novel nucleic acid-interacting compounds in the near future. 3. Conclusion and perspectives Molecular modeling has been extensively used for the identication (virtual ligand screening) and lead optimization (structurebased design) of small molecules to regulate the activities of protein-based drug targets. With increasing an understanding of the role of higher order DNA secondary structures in the regulation of gene expression, nucleic acids have emerged as an alternative therapeutic target for the treatment and prevention of a number of diseases. As a result of this interest, there have been a number of reports in the last two years on the application of molecular modeling for the virtual ligand screening and structure-based design of nuclei acid-interacting compounds. The targeting of non-canonical DNA structures such as G-quadruplexes that are related to the proliferation of cancer has attracted particular interest. Through a number of recent examples, we have highlighted successful applications of virtual ligand screening for the identication of small molecules which are able to selectively interact with a specic nucleic acid secondary structure. Despite some on-going concerns about the ability of docking algorithms to accurately predict binding modes of known nucleic acid ligands, virtual ligand screening still represents a powerful tool for the discovery of novel

nucleic acid-interacting scaffolds that can be further rened by lead optimization. Traditional lead optimization is expensive and timeconsuming, involving the synthesis of large libraries of compounds in order to develop a detailed structureeactivity relationship. Molecular modeling provides a more convenient and economical alternative by allowing the exploration of the chemical space in silico. We have highlighted recent examples where molecular modeling was used as a core component of the hit-to-lead developmental process, yielding optimized compounds that displayed superior binding afnity and selectivity to their nucleic acid targets, with a concurrent improvement in their biological activities. In conclusion, this review has attempted to demonstrate the powerful potential of molecular modeling for the identication and optimization of novel organic- and inorganic-based compounds targeting nucleic acids. We hope that this review will encourage the further application of virtual ligand screening and structure-based design methods in order to develop new nucleic acid-targeting small molecules for the prevention and treatment of diseases. Acknowledgments This work was supported by the Hong Kong Baptist University (FRG2/09-10/070 and FRG2/10-11/008), Centre for Cancer and Inammation Research, School of Chinese Medicine(CCIR-SCM, HKBU) and The Hong Kong Anti-Cancer Society (HKACS) Research Grant. References
[1] J.D. Watson, F.H.C. Crick, Molecular structure of nucleic acids. A structure for deoxyribose nucleic acid, Nature 171 (1953) 737e738.

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266 [2] M.F. Gellert, M.N. Lipsett, D.H. Davies, Helix formation by guanylic acid, Proc. Natl. Acad. Sci. U S A 48 (1962) 2013e2018. [3] R. Langridge, A. Rich, Molecular structure of helical polycytidylic acid, Nature 198 (1963) 725e728. [4] K. Gehring, J.L. Leroy, M. Gueron, A tetrameric DNA structure with protonated cytosine-cytosine base pairs, Nature 363 (1993) 561e565. [5] G. Felsenfeld, D.R. Davies, A. Rich, Formation of a three-stranded polynucleotide molecule, J. Am. Chem. Soc. 79 (1957) 2023e2024. [6] K.R. Fox, Targeting DNA with triplexes, Curr. Med. Chem. 7 (2000) 17e37. [7] K.M. Vasquez, J.H. Wilson, Triplex-directed modication of genes and gene activity, Trends Biochem. Sci. 23 (1998) 4e9. [8] D. Praseuth, A.L. Guieysse, C. Helene, Triple helix formation and the antigene strategy for sequence-specic control of gene expression, Biochim. Biophys. Acta Gene Struct. Expr. 1489 (1999) 181e206. [9] J.L. Huppert, S. Balasubramanian, G-quadruplexes in promoters throughout the human genome, Nucleic Acids Res. 35 (2007) 406e413. [10] A. Siddiqui-Jain, C.L. Grand, D.J. Bearss, L.H. Hurley, Direct evidence for a Gquadruplex in a promoter region and its targeting with a small molecule to repress c-MYC transcription, Proc. Natl. Acad. Sci. U S A 99 (2002) 11593e11598. [11] T.-m. Ou, Y.-j. Lu, J.-h. Tan, Z.-s. Huang, K.-Y. Wong, L.-q. Gu, G-quadruplexes: targets in anticancer drug design, ChemMedChem 3 (2008) 690e713. [12] S. Balasubramanian, S. Neidle, G-quadruplex nucleic acids as therapeutic targets, Curr. Opin. Chem. Biol. 13 (2009) 345e353. [13] S. Neidle, DNA minor-groove recognition by small molecules, Nat. Prod. Rep. 18 (2001) 291e309. [14] X. Cai, P.J. Gray Jr., H.D.D. Von, DNA minor groove binders: Back in the groove, Cancer Treat. Rev. 35 (2009) 437e450. [15] E.T. Kool, Hydrogen bonding, base stacking, and steric effects in DNA replication, Annu. Rev. Biophys. Biomol. Struct. 30 (2001) 1e22. [16] B.A.D. Neto, A.A.M. Lapis, Recent developments in the chemistry of deoxyribonucleic acid (DNA) intercalators: principles, design, synthesis, applications and trends, Molecules 14 (2009) 1725e1746. [17] N.J. Wheate, C.R. Brodie, J.G. Collins, S. Kemp, J.R. Aldrich-Wright, DNA intercalators in cancer therapy: organic and inorganic drugs and their spectroscopic tools of analysis, Mini-Rev. Med. Chem. 7 (2007) 627e648. [18] A. Ambrus, D. Chen, J. Dai, R.A. Jones, D. Yang, Solution structure of the biologically relevant G-quadruplex element in the human c-MYC promoter. Implications for G-quadruplex stabilization, Biochemistry 44 (2005) 2048e2058. [19] D. Yang, L.H. Hurley, Structure of the biologically relevant G-quadruplex in the c-MYC promoter, Nucleosides Nucleotides Nucleic Acids 25 (2006) 951e968. [20] A.K. Todd, S.M. Haider, G.N. Parkinson, S. Neidle, Sequence occurrence and structural uniqueness of a G-quadruplex in the human c-kit promoter, Nucleic Acids Res. 35 (2007) 5799e5808. [21] S.-T.D. Hsu, P. Varnai, A. Bugaut, A.P. Reszka, S. Neidle, S. Balasubramanian, A G-rich sequence within the c-kit oncogene promoter forms a parallel Gquadruplex having asymmetric G-tetrad dynamics, J. Am. Chem. Soc. 131 (2009) 13399e13409. [22] J. Dai, T.S. Dexheimer, D. Chen, M. Carver, A. Ambrus, R.A. Jones, D. Yang, An intramolecular G-quadruplex structure with mixed parallel/antiparallel Gstrands formed in the human BCL-2 promoter region in solution, J. Am. Chem. Soc. 128 (2006) 1096e1098. [23] T.S. Dexheimer, D. Sun, L.H. Hurley, Deconvoluting the structural and drugrecognition complexity of the G-quadruplex-forming region upstream of the bcl-2 P1 p.omoter, J. Am. Chem. Soc. 128 (2006) 5404e5415. [24] A.R. De, S. Wood, D. Sun, L.H. Hurley, S.W. Ebbinghaus, Evidence for the presence of a guanine quadruplex forming region within a polypurine tract of the hypoxia inducible factor 1a promoter, Biochemistry 44 (2005) 16341e16350. [25] K. Guo, V. Gokhale, L.H. Hurley, D. Sun, Intramolecularly folded G-quadruplex and i-motif structures in the proximal promoter of the vascular endothelial growth factor gene, Nucleic Acids Res. 36 (2008) 4598e4608. [26] D. Sun, W.-J. Liu, K. Guo, J.J. Rusche, S. Ebbinghaus, V. Gokhale, L.H. Hurley, The proximal promoter region of the human vascular endothelial growth factor gene has a G-quadruplex structure that can be targeted by G-quadruplex-interactive agents, Mol. Cancer Ther. 7 (2008) 880e889. [27] R.A. Alderden, M.D. Hall, T.W. Hambley, The discovery and development of cisplatin, J. Chem. Educ. 83 (2006) 728e734. [28] J. Liu, C.-H. Leung, A.L.-F. Chow, R.W.-Y. Sun, S.-C. Yan, C.-M. Che, Cyclometalated platinum(II) complexes as topoisomerase IIa poisons, Chem. Commun. (Cambridge, U. K.) 47 (2010) 719e721. [29] P. Wang, C.-H. Leung, D.-L. Ma, W. Lu, C.-M. Che, Organoplatinum(II) complexes with nucleobase motifs as inhibitors of human topoisomerase II catalytic activity, Chem.eAsian J. 5 (2010) 2271e2280. [30] H.M. Berman, The protein data bank and the challenge of structural genomics, Nat. Struct. Biol. 7 (2000) 957e959. [31] J. Westbrook, Z. Feng, L. Chen, H. Yang, H.M. Berman, The protein data bank and structural genomics, Nucleic Acids Res. 31 (2003) 489e491. [32] T.L. Blundell, H. Jhoti, C. Abell, High-throughput crystallography for lead discovery in drug design, Nat. Rev. Drug Discov. 1 (2002) 45e54. [33] A. Sharff, H. Jhoti, High-throughput crystallography to enhance drug discovery, Curr. Opin. Chem. Biol. 7 (2003) 340e345. [34] L.W. Tari, D.E. McRee, A.J. Jennings, Use of high-throughput crystallography and in silico methods for structure-based drug design, Methods Biochem. Anal. 45 (2005) 107e129.

1265

[35] R. Powers, Advances in nuclear magnetic resonance for drug discovery, Expert Opin. Drug Discov. 4 (2009) 1077e1098. [36] G. Liu, Y. Shen, H.S. Atreya, D. Parish, Y. Shao, D.K. Sukumaran, R. Xiao, A. Yee, A. Lemak, A. Bhattacharya, et al., NMR data collection and analysis protocol for high-throughput protein structure determination, Proc. Natl. Acad. Sci. U S A 102 (2005) 10487e10492. [37] B.O. Villoutreix, R. Eudes, M.A. Miteva, Structure-based virtual ligand screening: recent success stories, Comb. Chem. High Throughput Screen. 12 (2009) 1000e1016. [38] M. Pellecchia, Fragment-based drug discovery takes a virtual turn, Nat. Chem. Biol. 5 (2009) 274e275. [39] C.N. Cavasotto, A.J.W. Orry, Ligand docking and structure-based virtual screening in drug discovery, Curr. Top. Med. Chem. 7 (2007) 1006e1014. [40] D.S.-H. Chan, H.-M. Lee, F. Yang, C.-M. Che, C.C.L. Wong, R. Abagyan, C.-H. Leung, D.-L. Ma, Structure-based discovery of natural-product-like TNF-a inhibitors, Angew. Chem. Int. Ed. 49 (2010) 2860e2864. [41] C.-H. Leung, D.S.-H. Chan, R. Abagyan, S.M.-Y. Lee, G.-Y. Zhu, W.-F. Fong, D.-L. Ma, A Natural Product-Like Inhibitor of NEDD8-Activating Enzyme, Chem, Comm 47 (2011) 2511e2513. [42] E.M. Bowers, G. Yan, C. Mukherjee, A. Orry, L. Wang, M.A. Holbert, N.T. Crump, C.A. Hazzalin, G. Liszczak, H. Yuan, et al., Virtual ligand screening of the p300/CBP histone acetyltransferase: identication of a selective small molecule inhibitor, Chem. Biol. (Cambridge, MA, U. S.) 17 (2010) 471e482. [43] M. Montes, E. Braud, M.A. Miteva, M.-L. Goddard, O. Mondesert, S. Kolb, M.P. Brun, B. Ducommun, C. Garbay, B.O. Villoutreix, Receptor-based virtual ligand screening for the identication of novel CDC25 p.osphatase inhibitors, J. Chem. Inf. Model. 48 (2008) 157e165. [44] M.M. Dailey, C. Hait, P.A. Holt, J.M. Maguire, J.B. Meier, M.C. Miller, L. Petraccone, J.O. Trent, Structure-based drug design: from nucleic acid to membrane protein targets, Exp. Mol. Pathol. 86 (2009) 141e150. [45] M.D. Frank-Kamenetskii, S.M. Mirkin, Triplex DNA structures, Annu. Rev. Biochem. 64 (1995) 65e95. [46] A. Kalota, S.E. Shetzline, A.M. Gewirtz, Progress in the development of nucleic acid therapeutics for cancer, Cancer Biol. Ther. 3 (2004) 4e12. [47] J.Y. Chin, E.B. Schleifman, P.M. Glazer, Repair and recombination induced by triple helix DNA, Front. Biosci. 12 (2007) 4288e4297. [48] M.M. Seidman, N. Puri, A. Majumdar, B. Cuenoud, P.S. Miller, R. Alam, The development of bioactive triple helix-forming oligonucleotides, Ann. N. Y. Acad. Sci. 1058 (2005) 119e127. [49] H.-G. Kim, J.F. Reddoch, C. Mayeld, S. Ebbinghaus, N. Vigneswaran, S. Thomas, D.M. Miller, Inhibition of transcription of the human c-myc protooncogene by intermolecular triplex, Biochemistry 37 (1998) 2299e2304. [50] M.M. Seidman, P.M. Glazer, The potential for gene repair via triple helix formation, J. Clin. Invest. 112 (2003) 487e494. [51] G.M. Carbone, E. McGufe, S. Napoli, C.E. Flanagan, C. Dembech, U. Negri, F. Arcamone, M.L. Capobianco, C.V. Catapano, DNA binding and antigene activity of a daunomycin-conjugated triplex-forming oligonucleotide targeting the P2 p.omoter of the human c-myc gene, Nucleic Acids Res. 32 (2004) 2396e2410. [52] J. Lacoste, J.-C. Francois, C. Helene, Triple helix formation with purine-rich phosphorothioate-containing oligonucleotides covalently linked to an acridine derivative, Nucleic Acids Res. 25 (1997) 1991e1998. [53] C. Marchand, C. Bailly, C.H. Nguyen, E. Bisagni, T. Garestier, C. Helene, M.J. Waring, Stabilization of triple helical DNA by a benzopyridoquinoxaline intercalator, Biochemistry 35 (1996) 5022e5032. [54] A.A. Moraru-Allen, S. Cassidy, J.-L.A. Alvarez, K.R. Fox, T. Brown, A.N. Lane, Coralyne has a preference for intercalation between TA-T triples in intramolecular DNA triple helixes, Nucleic Acids Res. 25 (1997) 1890e 1896. [55] B.A. Janowski, J. Hu, D.R. Corey, Silencing gene expression by targeting chromosomal DNA with antigene peptide nucleic acids and duplex RNAs, Nat. Protoc. 1 (2006) 436e443. [56] R.L. Beane, R. Ram, S. Gabillet, K. Arar, B.P. Monia, D.R. Corey, Inhibiting gene expression with locked nucleic acids (LNAs) that target chromosomal DNA, Biochemistry 46 (2007) 7572e7580. [57] M. Keppler, O. Zegrocka, L. Strekowski, K.R. Fox, DNA triple helix stabilization by a naphthylquinoline dimer, FEBS Lett. 447 (1999) 223e226. [58] O. Baudoin, C. Marchand, M.-P. Teulade-Fichou, J.-P. Vigneron, J.-S. Sun, T. Garestier, C. Helene, J.-M. Lehn, Stabilization of DNA triple helixes by crescent-shaped dibenzophenanthrolines, Chem.eEur. J. 4 (1998) 1504e1508. [59] J.B. Chaires, J. Ren, M. Henary, O. Zegrocka, G.R. Bishop, L. Strekowski, Triplex selective 2-(2-naphthyl)quinoline compounds: origins of afnity and new design principles, J. Am. Chem. Soc. 125 (2003) 7272e7283. [60] P.A. Holt, P. Ragazzon, L. Strekowski, J.B. Chaires, J.O. Trent, Discovery of novel triple helical DNA intercalators by an integrated virtual and actual screening platform, Nucleic Acids Res. 37 (2009) 1280e1287. [61] A.H. Fairlamb, Chemotherapy of human African trypanosomiasis: current and future prospects, Trends Parasitol. 19 (2003) 488e494. [62] J. Mann, A. Baron, Y. Opoku-Boahen, E. Johansson, G. Parkinson, L.R. Kelland, S. Neidle, A new class of symmetric bisbenzimidazole-based DNA minor groove-binding agents showing antitumor activity, J. Med. Chem. 44 (2001) 138e144.

1266

D.-L. Ma et al. / Biochimie 93 (2011) 1252e1266 [84] H.-M. Lee, D.S.-H. Chan, F. Yang, H.-Y. Lam, S.-C. Yan, C.-M. Che, D.-L. Ma, C.H. Leung, Identication of natural product fonsecin B as a stabilizing ligand of c-myc G-quadruplex DNA by high-throughput virtual screening, Chem. Commun. (Cambridge, U. K.) 46 (2010) 4680e4682. [85] D.J. Newman, G.M. Cragg, Natural products as sources of new drugs over the last 25 years, J. Nat. Prod. 70 (2007) 461e477. [86] F.E. Koehn, G.T. Carter, The evolving role of natural products in drug discovery, Nat. Rev. Drug Discov. 4 (2005) 206e220. [87] J.W.H. Li, J.C. Vederas, Drug discovery and natural products end of an era or an endless frontier? Science 325 (2009) 161e165. [88] M.S. Butler, Natural products to drugs: natural product-derived compounds in clinical trials, Nat. Prod. Rep. 25 (2008) 475e516. [89] K. Shin-ya, K. Wierzba, K.-i. Matsuo, T. Ohtani, Y. Yamada, K. Furihata, Y. Hayakawa, H. Seto, Telomestatin, a novel telomerase inhibitor from Streptomyces anulatus, J. Am. Chem. Soc. 123 (2001) 1262e1263. [90] M.-Y. Kim, M. Gleason-Guzman, E. Izbicka, D. Nishioka, L.H. Hurley, The different biological effects of telomestatin and TMPyP4 can be attributed to their selectivity for interaction with intramolecular or intermolecular Gquadruplex structures, Cancer Res. 63 (2003) 3247e3256. [91] A.M. Burger, F. Dai, C.M. Schultes, A.P. Reszka, M.J. Moore, J.A. Double, S. Neidle, The G-quadruplex-interactive molecule BRACO-19 inhibits tumor growth, consistent with telomere targeting and interference with telomerase function, Cancer Res. 65 (2005) 1489e1496. [92] M. Gunaratnam, O. Greciano, C. Martins, A.P. Reszka, C.M. Schultes, H. Morjani, J.-F. Riou, S. Neidle, Mechanism of acridine-based telomerase inhibition and telomere shortening, Biochem. Pharmacol. 74 (2007) 679e689. [93] S. Sparapani, S.M. Haider, F. Doria, M. Gunaratnam, S. Neidle, Rational design of acridine-based ligands with selectivity for human telomeric quadruplexes, J. Am. Chem. Soc. 132 (2010) 12263e12272. [94] M. Read, R.J. Harrison, B. Romagnoli, F.A. Tanious, S.H. Gowan, A.P. Reszka, W.D. Wilson, L.R. Kelland, S. Neidle, Structure-based design of selective and potent G quadruplex-mediated telomerase inhibitors, Proc. Natl. Acad. Sci. U S A 98 (2001) 4844e4849. [95] J. Dai, M. Carver, C. Punchihewa, R.A. Jones, D. Yang, Structure of the hybrid-2 type intramolecular human telomeric G-quadruplex in K solution: insights into structure polymorphism of the human telomeric sequence, Nucleic Acids Res. 35 (2007) 4927e4940. [96] H. Fernando, A.P. Reszka, J. Huppert, S. Ladame, S. Rankin, A.R. Venkitaraman, S. Neidle, S. Balasubramanian, A conserved quadruplex motif located in a transcription activation site of the human c-kit oncogene, Biochemistry 45 (2006) 7854e7860. [97] S.N. Georgiades, K.N.H. Abd, K. Suntharalingam, R. Vilar, Interaction of metal complexes with G-quadruplex DNA, Angew. Chem. Int. Ed. 49 (2010) 4020e4034. [98] C.X. Zhang, S.J. Lippard, New metal complexes as potential therapeutics, Curr. Opin. Chem. Biol. 7 (2003) 481e489. [99] P.C.A. Bruijnincx, P.J. Sadler, New trends for metal complexes with anticancer activity, Curr. Opin. Chem. Biol. 12 (2008) 197e206. [100] R.W.-Y. Sun, C.-M. Che, The anti-cancer properties of gold(III) compounds with dianionic porphyrin and tetradentate ligands, Coord. Chem. Rev. 253 (2009) 1682e1691. [101] D.-L. Ma, C.-M. Che, S.-C. Yan, Platinum(II) complexes with dipyridophenazine ligands as human telomerase inhibitors and luminescent probes for G-quadruplex DNA, J. Am. Chem. Soc. 131 (2009) 1835e1846. [102] P. Wu, D.-L. Ma, C.-H. Leung, S.-C. Yan, N. Zhu, R. Abagyan, C.-M. Che, Stabilization of G-quadruplex DNA with platinum(II) Schiff base complexes: luminescent probe and down-regulation of c-myc oncogene expression, Chem.eEur. J. 15 (2009) 13008e13021. [103] P. Wang, C.-H. Leung, D.-L. Ma, S.-C. Yan, C.-M. Che, Structure-based design of platinum(II) complexes as c-myc oncogene down-regulators and luminescent probes for G-quadruplex DNA, Chem.eEur. J. 16 (2010) 6900e6911. [104] S. Rhee, Z.-j. Han, K. Liu, H.T. Miles, D.R. Davies, Structure of a triple helical DNA with a triplexeduplex junction, Biochemistry 38 (1999) 16810e16815. [105] Z. Zhang, J. Dai, E. Veliath, R.A. Jones, D. Yang, Structure of a two-G-tetrad intramolecular G-quadruplex formed by a variant human telomeric sequence in K solution: insights into the interconversion of human telomeric G-quadruplex structures, Nucleic Acids Res. 38 (2010) 1009e1021. [106] J. Weil, T. Min, C. Yang, S. Wang, C. Sutherland, N. Sinha, C. Kang, Stabilization of the i-motif by intramolecular adenine-adenine-thymine base triple in the structure of d(ACCCT), Acta Crystallogr. D Biol. Crystallogr. 55 (1999) 422e429.

[63] R.K. Arafa, R. Brun, T. Wenzler, F.A. Tanious, W.D. Wilson, C.E. Stephens, D.W. Boykin, Synthesis, DNA afnity, and antiprotozoal activity of fused ring dicationic compounds and their prodrugs, J. Med. Chem. 48 (2005) 5480e5488. [64] D.A. Evans, S. Neidle, Virtual screening of DNA minor groove binders, J. Med. Chem. 49 (2006) 4232e4238. [65] S. Mallena, M.P.H. Lee, C. Bailly, S. Neidle, A. Kumar, D.W. Boykin, W.D. Wilson, Thiophene-based diamidine forms a super at binding minor groove agent, J. Am. Chem. Soc. 126 (2004) 13659e13669. [66] P.A. Holt, J.B. Chaires, J.O. Trent, Molecular docking of intercalators and groove-binders to nucleic acids using autodock and surex, J. Chem. Inf. Model. 48 (2008) 1602e1615. [67] S. Cosconati, L. Marinelli, R. Trotta, A. Virno, L. Mayol, E. Novellino, A.J. Olson, A. Randazzo, Tandem application of virtual screening and NMR experiments in the discovery of brand new DNA quadruplex groove binders, J. Am. Chem. Soc. 131 (2009) 16336e16337. [68] J. Dash, P.S. Shirude, S.-T.D. Hsu, S. Balasubramanian, Diarylethynyl amides that recognize the parallel conformation of genomic promoter DNA Gquadruplexes, J. Am. Chem. Soc. 130 (2008) 15950e15956. [69] Q. Li, J. Xiang, X. Li, L. Chen, X. Xu, Y. Tang, Q. Zhou, L. Li, H. Zhang, H. Sun, et al., Stabilizing parallel G-quadruplex DNA by a new class of ligands: two non-planar alkaloids through interaction in lateral grooves, Biochimie 91 (2009) 811e819. [70] C. Caceres, G. Wright, C. Gouyette, G. Parkinson, J.A. Subirana, A thymine tetrad in d(TGGGGT) quadruplexes stabilized with Tl/Na ions, Nucleic Acids Res. 32 (2004) 1097e1102. [71] S. Cosconati, L. Marinelli, R. Trotta, A. Virno, T.S. De, R. Romagnoli, B. Pagano, V. Limongelli, C. Giancola, P.G. Baraldi, et al., Structural and conformational requisites in DNA quadruplex groove binding: another piece to the puzzle, J. Am. Chem. Soc. 132 (2010) 6425e6433. [72] L. Martino, A. Virno, B. Pagano, A. Virgilio, M.S. Di, A. Galeone, C. Giancola, G. Bifulco, L. Mayol, A. Randazzo, Structural and thermodynamic studies of the interaction of distamycin A with the parallel quadruplex structure [d(TGGGGT)] 4, J. Am. Chem. Soc. 129 (2007) 16048e16056. [73] F.X. Han, R.T. Wheelhouse, L.H. Hurley, Interactions of TMPyP4 and TMPyP2 with quadruplex DNA. Structural basis for the differential effects on telomerase inhibition, J. Am. Chem. Soc. 121 (1999) 3561e3570. [74] C.L. Grand, H. Han, R.M. Munoz, S. Weitman, H.D.D. Von, L.H. Hurley, D.J. Bearss, The cationic porphyrin TMPyP4 down-regulates c-MYC and human telomerase reverse transcriptase expression and inhibits tumor growth in vivo, Mol. Cancer Ther. 1 (2002) 565e573. [75] T.M. del, P. Bucek, A. Avino, J. Jaumot, C. Gonzalez, R. Eritja, R. Gargallo, Targeting the G-quadruplex-forming region near the P1 p.omoter in the human BCL-2 gene with the cationic porphyrin TMPyP4 and with the complementary C-rich strand, Biochimie 91 (2009) 894e902. [76] Y.-J. Lu, T.-M. Ou, J.-H. Tan, J.-Q. Hou, W.-Y. Shao, D. Peng, N. Sun, X.-D. Wang, W.-B. Wu, X.-Z. Bu, et al., 5-N-methylated quindoline derivatives as telomeric G-quadruplex stabilizing ligands: effects of 5-N positive charge on quadruplex binding afnity and cell proliferation, J. Med. Chem. 51 (2008) 6381e6392. [77] X.-D. Wang, T.-M. Ou, Y.-J. Lu, Z. Li, Z. Xu, C. Xi, J.-H. Tan, S.-L. Huang, L.-K. An, D. Li, et al., Turning off transcription of the bcl-2 gene by Stabilizing the bcl-2 promoter quadruplex with quindoline derivatives, J. Med. Chem. 53 (2010) 4390e4398. [78] Z.A.E. Waller, P.S. Shirude, R. Rodriguez, S. Balasubramanian, Triarylpyridines: a versatile small molecule scaffold for G-quadruplex recognition, Chem. Commun. 2008 (2008) 1467e1469. [79] M. Bejugam, S. Sewitz, P.S. Shirude, R. Rodriguez, R. Shahid, S. Balasubramanian, Trisubstituted isoalloxazines as a new class of G-quadruplex binding ligands: small molecule regulation of c-kit oncogene expression, J. Am. Chem. Soc. 129 (2007) 12926e12927. [80] D.-L. Ma, T.-S. Lai, F.-Y. Chan, W.-H. Chung, R. Abagyan, Y.-C. Leung, K.Y. Wong, Discovery of a drug-like G-quadruplex binding ligand by highthroughput docking, ChemMedChem 3 (2008) 881e884. [81] G.N. Parkinson, M.P.H. Lee, S. Neidle, Crystal structure of parallel quadruplexes from human telomeric DNA, Nature (London, U. K.) 417 (2002) 876e880. [82] C.A. Lipinski, F. Lombardo, B.W. Dominy, P.J. Feeney, Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings, Adv. Drug Deliv. Rev. 23 (1997) 3e25. [83] M. Totrov, R. Abagyan, Flexible protein-ligand docking by global energy optimization in internal coordinates, Proteins(Suppl. 1) (1997) 215e220.

S-ar putea să vă placă și