Abstract: β-D-3'-Azido-2',3'-dideoxyguanosine (3'-azido-ddG) is a potent inhibitor of HIV-1 replication with a superior resistance profile to zidovudine. Recently, we identified five novel 6-modified-3'-azido-ddG analogs that exhibit similar or superior anti-HIV-1 activity compared to 3'-azido-ddG in primary cells. To gain insight into their structure-activity-resistance relationships, we synthesized their triphosphate (TP) forms and assessed their ability to inhibit HIV-1 reverse transcriptase (RT). Steady-state and pre-steady-state kinetic experiments show that the 6-modified-3'-azido-ddGTP analogs act as adenosine rather than guanosine mimetics in DNA synthesis reactions. The order of potency of the TP analogs against wild-type RT was: 3'-azido-2,6-diaminopurine >3'-azido-6-chloropurine; 3'-azido-6-N-allylaminopurine > 2-amino-6-N,N-dimethylaminopurine; 2-amino-6-methoxypurine. Molecular modeling studies reveal unique hydrogen-bonding interactions between the nucleotide analogs and the template thymine base in the active site of RT. Surprisingly, the structure-activity relationship of the analogs differed in HIV-1 RT ATP-mediated excision assays of their monophosphate forms, suggesting that it may be possible to rationally design a modified base analog that is efficiently incorporated by RT but serves as a poor substrate for ATP-mediated excision reactions. Overall, these studies identify a promising strategy to design novel nucleoside analogs that exert profound antiviral activity against both WT and drug-resistant HIV-1.
Abstract: Microwave-assisted optimized transglycosylation reactions were used to prepare eleven modified l-3'-azido-2',3'-dideoxypurine nucleosides. These l-nucleoside analogs were evaluated against HIV and hepatitis B virus. The l-3'-azido-2',3'-dideoxypurines nucleosides were metabolized to nucleoside 5'-triphosphates in primary human lymphocytes, but exhibited weak or no antiviral activity against HIV-1. The nucleosides were also inactive against HBV in HepG2 cells. Pre-steady state kinetic experiments demonstrated that the l-3'-azido-2',3'-dideoxypurine triphosphates could be incorporated by purified HIV-1 reverse transcriptase, although their catalytic efficiency (k(pol)/K(d)) of incorporation was low. Interestingly, a phosphoramidate prodrug of l-3'-azido-2',3'-dideoxyadenosine exhibited anti-HIV-1 activity without significant toxicity.
Abstract: Resistance to human immunodeficiency virus type 1 (HIV-1) represents a significant problem in the design of novel therapeutics and the management of treatment regimens in infected persons. Resistance profiles can be elucidated by defining modifications to the viral genome conferred upon exposure to novel nucleoside reverse transcriptase (RT) inhibitors (NRTI). In vitro testing of HIV-1LAI-infected primary human lymphocytes treated with β-D-2',3'-dideoxy-2',3'-didehydro-5-fluorocytidine (DFC; Dexelvucitabine; Reverset) produced a novel deletion of AGT at codon 68 (S68Δ) alone and in combination with K65R that differentially affects drug response. Dual-approach clone techniques utilizing TOPO cloning and pyrosequencing confirmed the novel S68Δ in the HIV-1 genome. The S68Δ HIV-1 RT was phenotyped against various antiviral agents in a heteropolymeric DNA polymerase assay and in human lymphocytes. Drug susceptibility results indicate that the S68Δ displayed a 10- to 30-fold increase in resistance to DFC, lamivudine, emtricitabine, tenofovir, abacavir, and amdoxovir and modest resistance to stavudine, β-d-2',3'-oxa-5-fluorocytidine, or 9-(β-D-1,3-dioxolan-4-yl)guanine and remained susceptible to 3'-azido-3'-deoxythymidine, 2',3'-dideoxyinosine (ddI), 1-(β-D-dioxolane)thymine (DOT) and lopinavir. Modeling revealed a central role for S68 in affecting conformation of the β3-β4 finger region and provides a rational for the selective resistance. These data indicate that the novel S68Δ is a previously unrecognized deletion that may represent an important factor in NRTI multidrug resistance treatment strategies.
Abstract: Microtubules have long been considered an ideal target for anticancer drugs because of the essential role they play in mitosis, forming the dynamic spindle apparatus. As such, there is a wide variety of compounds currently in clinical use and in development that act as antimitotic agents by altering microtubule dynamics. Although these diverse molecules are known to affect microtubule dynamics upon binding to one of the three established drug domains (taxane, vinca alkaloid, or colchicine site), the exact mechanism by which each drug works is still an area of intense speculation and research. In this study, we review the effects of microtubule-binding chemotherapeutic agents from a new perspective, considering how their mode of binding induces conformational changes and alters biological function relative to the molecular vectors of microtubule assembly or disassembly. These "biological vectors" can thus be used as a spatiotemporal context to describe molecular mechanisms by which microtubule-targeting drugs work.
Abstract: Tubulin is an important molecular target in cancer chemotherapy. Antimitotic agents able to bind to the protein are currently under study, commonly used in the clinic to treat a variety of cancers and/or exploited as probes to investigate the protein's structure and function. Here we report the binding modes for a series of colchicinoids, combretastatin A4 and chalcones established from docking studies carried out on the structure of tubulin in complex with colchicine. The proposed models, in agreement with published biochemical data, show that combretastatin A4 binds to the colchicine site of beta-tubulin and that chalcones assume an orientation similar to that of podophyllotoxin. The models can be used to design a new class of podophyllotoxin mimics, the alpha-aryl chalcones, capable of binding to the colchicine-binding site of beta-tubulin with higher affinity.
Abstract: While HIV-1 reverse transcriptase (RT) mutations of M to V at position 184 are commonly observed in the clinic, the double mutation of 65R+74V is rarely seen. It has been demonstrated that rapid R-->K reversion occurs at RT codon 65 during replication of HIV-1 in human peripheral blood mononuclear cells containing 65R+74V mutations and that processivity of the RT is reduced relative to wild type. However, clinical studies show that M184V can be detected after several months of therapy interruption, suggesting more effective processivity. Herein, the in vitro RT processivity of genetically engineered M184V and double mutant 65R+74V was compared. Virion-associated RTs of WT pNL4-3, K65R, L74V, M184V and 65R+74V were used to perform RT processivity assays in the presence of trap, poly(rC)-oligo(dG). Both RTs with 184V and 65R+74V mutations exhibited similar processivity when compared with each other and a significantly decreased processivity as compared to WT RT. Both mutant RTs synthesized shorter cDNA molecules (37-42 nt) relative to WT RT, which made longer (65-70 nt) cDNA molecules. Since these surprising biochemical results cannot explain the clinical phenotype, a hypothesis is presented to explain the discrepancy and suggest new approaches for future testing.
Abstract: Traditional methods for general drug discovery typically include evaluating random compound libraries for activity in relevant cell-free or cell-based assays. Success in antiviral development has emerged from the discovery of more focused libraries that provide clues about structure activity relationships. Combining these with more recent approaches including structural biology and computational modeling can work efficiently to hasten discovery of active molecules, but that is not enough. There are issues related to biology, toxicology, pharmacology, and metabolism that have to be addressed before a hit compound becomes nominated for clinical development. The objective of gaining early preclinical knowledge is to reduce the risk of failure in Phases 1, 2, and 3, leading to the goal of approved drugs that benefit the infected individual. This review uses hepatitis C virus (HCV), for which we still do not have an ideal therapeutic modality, as an example of the multidisciplinary efforts needed to discover new antiviral drugs for the benefit of humanity.
Abstract: This work describes a novel semi-sequential technique for in silico enhancement of high-throughput screening (HTS) experiments now employed at Novartis. It is used in situations in which the size of the screen is limited by the readout (e.g., high-content screens) or the amount of reagents or tools (proteins or cells) available. By performing computational chemical diversity selection on a per plate basis (instead of a per compound basis), 25% of the 1,000,000-compound screening was optimized for general initial HTS. Statistical models are then generated from target-specific primary results (percentage inhibition data) to drive the cherry picking and testing from the entire collection. Using retrospective analysis of 11 HTS campaigns, the authors show that this method would have captured on average two thirds of the active compounds (IC(50) < 10 microM) and three fourths of the active Murcko scaffolds while decreasing screening expenditure by nearly 75%. This result is true for a wide variety of targets, including G-protein-coupled receptors, chemokine receptors, kinases, metalloproteinases, pathway screens, and protein-protein interactions. Unlike time-consuming "classic" sequential approaches that require multiple iterations of cherry picking, testing, and building statistical models, here individual compounds are cherry picked just once, based directly on primary screening data. Strikingly, the authors demonstrate that models built from primary data are as robust as models built from IC(50) data. This is true for all HTS campaigns analyzed, which represent a wide variety of target classes and assay types.
Abstract: Farnesyl transferase inhibitors (FTI) exhibit anticancer activity as a single agent in preclinical studies and show promise in combination with other therapeutics in clinical trials. Previous studies show that FTIs arrest cancer cells in mitosis; however, the mechanism by which this occurs is unclear. Here, we observed that treatment of various cancer cell lines with the FTI lonafarnib caused mitotic chromosomal alignment defects, leaving cells in a pseudometaphase state, whereby both aligned chromosomes and chromosomes juxtaposed to the spindle poles (termed "lagging chromosomes") were observed in the same cell. To determine how this occurs, we investigated the functionality of two farnesylated mitotic proteins, CENP-E and CENP-F, which mediate chromosomal capture and alignment. The data show that lonafarnib in proliferating cancer cells depletes CENP-E and CENP-F from metaphase but not prometaphase kinetochores. Loss of CENP-E and CENP-F metaphase localization triggered aberrant chromosomal maintenance, causing aligned chromosomes to be prematurely released from the spindle equator and become lagging chromosomes, resulting in a mitotic delay. Furthermore, lonafarnib treatment reduces sister kinetochore tension and activates the BubR1 spindle checkpoint, suggesting that farnesylation of CENP-E and CENP-F is critical for their functionality in maintaining kinetochore-microtubule interactions. Importantly, apparently similar chromosomal alignment defects were observed in head and neck tumors samples from a phase I trial with lonafarnib, providing support that lonafarnib disrupts chromosomal maintenance in human cancers. Lastly, to examine how farnesylation could regulate CENP-E in mediating kinetochore-microtubule attachments, we examined possible docking motifs of a farnesyl group on the outer surface of the microtubule. This analysis revealed three hydrophobic patches on the tubulin dimer for insertion of a farnesyl group, alluding to the possibility of an association between a farnesyl group and the microtubule.
Abstract: High throughput screening (HTS) data is often noisy, containing both false positives and negatives. Thus, careful triaging and prioritization of the primary hit list can save time and money by identifying potential false positives before incurring the expense of followup. Of particular concern are cell-based reporter gene assays (RGAs) where the number of hits may be prohibitively high to be scrutinized manually for weeding out erroneous data. Based on statistical models built from chemical structures of 650 000 compounds tested in RGAs, we created "frequent hitter" models that make it possible to prioritize potential false positives. Furthermore, we followed up the frequent hitter evaluation with chemical structure based in silico target predictions to hypothesize a mechanism for the observed "off target" response. It was observed that the predicted cellular targets for the frequent hitters were known to be associated with undesirable effects such as cytotoxicity. More specifically, the most frequently predicted targets relate to apoptosis and cell differentiation, including kinases, topoisomerases, and protein phosphatases. The mechanism-based frequent hitter hypothesis was tested using 160 additional druglike compounds predicted by the model to be nonspecific actives in RGAs. This validation was successful (showing a 50% hit rate compared to a normal hit rate as low as 2%), and it demonstrates the power of computational models toward understanding complex relations between chemical structure and biological function.
Abstract: Development of a pharmacophore hypothesis related to small-molecule activity is pivotal to chemical optimization of a series, since it defines features beneficial or detrimental to activity. Although crystal structures may provide detailed 3D interaction information for one molecule with its receptor, docking a different ligand to that model often leads to unreliable results due to protein flexibility. Graham Richards' lab was one of the first groups to utilize "fuzzy" pattern recognition algorithms taken from the field of image processing to solve problems in protein modeling. Thus, descriptor "fuzziness" was partly able to emulate conformational flexibility of the target while simultaneously enhancing the speed of the search. In this work, we extend these developments to a ligand-based method for describing and aligning molecules in flexible chemical space termed FEature POint PharmacophoreS (FEPOPS), which allows exploration of dynamic biological space. We develop a novel, combinatorial algorithm for molecular comparisons and evaluate it using the WOMBAT dataset. The new approach shows superior retrospective virtual screening performance than earlier shape-based or charge-based algorithms. Additionally, we use target prediction to evaluate how FEPOPS alignments match the molecules biological activity by identifying the atoms and features that make the key contributions to overall chemical similarity. Overall, we find that FEPOPS are sufficiently fuzzy and flexible to find not only new ligand scaffolds, but also challenging molecules that occupy different conformational states of dynamic biological space as from induced fits.
Abstract: Resistance to paclitaxel-based therapy is frequently encountered in the clinic. The mechanisms of intrinsic or acquired paclitaxel resistance are not well understood. We sought to characterize the resistance mechanisms that develop upon chronic exposure of a cancer cell line to paclitaxel in the presence of the P-glycoprotein reversal agent, CL-347099. The epidermoid tumor line KB-3-1 was exposed to increasing concentrations of paclitaxel and 5 micromol/L CL-347099 for up to 1 year. Cells grown in 15 nmol/L paclitaxel plus CL-347099 (KB-15-PTX/099) developed 18-fold resistance to paclitaxel and were dependent upon paclitaxel for maximal growth. They grew well and retained resistance to paclitaxel when grown in athymic mice. Cross-resistance (3- to 5-fold) was observed in tissue culture to docetaxel, the novel taxane MAC-321, and epothilone B. Collateral sensitivity (approximately 3-fold) was observed to the depolymerizing agents vinblastine, dolastatin-10, and HTI-286. KB-15-PTX/099-resistant cells did not overexpress P-glycoprotein nor did they have an alteration of [14C]paclitaxel accumulation compared with parental cells. However, a novel point mutation (T to A) resulting in Asp26 to glutamate substitution in class I (M40) beta-tubulin was found. Based on an electron crystallography structure of Zn-stabilized tubulin sheets, the phenyl ring of C-3' NHCO-C6H5 of paclitaxel makes contact with Asp26 of beta-tubulin, suggesting a ligand-induced mutation. Optimized model complexes of paclitaxel, docetaxel, and MAC-321 in beta-tubulin show a novel hydrogen bonding pattern for the glutamate mutant and rationalize the observed resistance profiles. However, a mutation in the paclitaxel binding pocket does not explain the phenotype completely. KB-15-PTX/099 cells have impaired microtubule stability as determined by a reduced percentage of tubulin in microtubules and reflected by less acetylated tubulin. These results suggest that a mutation in tubulin might affect microtubule stability as well as drug binding and contribute to the observed resistance profile.
Abstract: High-throughput screening (HTS) plays a pivotal role in lead discovery for the pharmaceutical industry. In tandem, cheminformatics approaches are employed to increase the probability of the identification of novel biologically active compounds by mining the HTS data. HTS data is notoriously noisy, and therefore, the selection of the optimal data mining method is important for the success of such an analysis. Here, we describe a retrospective analysis of four HTS data sets using three mining approaches: Laplacian-modified naive Bayes, recursive partitioning, and support vector machine (SVM) classifiers with increasing stochastic noise in the form of false positives and false negatives. All three of the data mining methods at hand tolerated increasing levels of false positives even when the ratio of misclassified compounds to true active compounds was 5:1 in the training set. False negatives in the ratio of 1:1 were tolerated as well. SVM outperformed the other two methods in capturing active compounds and scaffolds in the top 1%. A Murcko scaffold analysis could explain the differences in enrichments among the four data sets. This study demonstrates that data mining methods can add a true value to the screen even when the data is contaminated with a high level of stochastic noise.
Abstract: We studied in silico docking of noscapine onto tubulin, combined with calculations of surface charge, pi-pi, van der Waals, and hydrogen bonding interactions, to rationally design a new compound, EM015. This tubulin-binding semisynthetic compound is a selective and potent anti-breast cancer agent and displays a 20-fold lower IC(50) against many tumor cells compared with our founding compound, (S)-6,7-dimethoxy-3-((R)-4-methoxy-6-methyl-5,6,7,8-tetrahydro[1,3]-dioxolo-[4,5-g]isoquinolin-5-yl)isobenzo-furan-1(3H)-one (noscapine). Furthermore, EM015 is also effective against a variety of drug-resistant cells. Surprisingly, the cell cycle profile of nontumorigenic normal cells is not affected. Many antimicrotubule cancer drugs in clinic today, particularly taxanes and Vincas, face challenges including frequent visits to the hospital for prolonged i.v. infusions, toxicities, and tumor recurrences due to drug resistance. EM015, on the other hand, is orally available, regresses breast tumor xenografts in nude mice models, and increases longevity. Furthermore, we have failed to observe any detectable toxicity in tissues, such as liver, kidney, spleen, lung, heart, and brain, as well as neurons, which are common targets of antimicrotubule drug therapy. Thus, EM015 has a great promise in the clinic.
Abstract: Conventional similarity searching of molecules compares single (or multiple) active query structures to each other in a relative framework, by means of a structural descriptor and a similarity measure. While this often works well, depending on the target, we show here that retrieval rates can be improved considerably by incorporating an external framework describing ligand bioactivity space for comparisons ("Bayes affinity fingerprints"). Structures are described by Bayes scores for a ligand panel comprising about 1000 activity classes extracted from the WOMBAT database. The comparison of structures is performed via the Pearson correlation coefficient of activity classes, that is, the order in which two structures are similar to the panel activity classes. Compound retrieval on a recently published data set could be improved by as much as 24% relative (9% absolute). Knowledge about the shape of the "bioactive chemical universe" is thus beneficial to identifying similar bioactivities. Principal component analysis was employed to further analyze activity space with the objective to define orthogonal ligand bioactive chemical space, leading to nine major (roughly orthogonal) activity axes. Employing only those nine activity classes, retrieval rates are still comparable to original Bayes affinity fingerprints; thus, the concept of orthogonal bioactive ligand chemical space was validated as being an information-rich but low-dimensional representation of bioactivity space. Correlations between activity classes are a major determinant to gauge whether the desired multitarget activity of drugs is (on the basis of current knowledge) a feasible concept because it measures the extent to which activities can be optimized independently, or only by strongly influencing one another.
Abstract: Bridging chemical and biological space is the key to drug discovery and development. Typically, cheminformatics methods operate under the assumption that similar chemicals have similar biological activity. Ideally then, one could predict a drug's biological function(s) given only its chemical structure by similarity searching in libraries of compounds with known activities. In practice, effectively choosing a similarity metric is case dependent. This work compares both 2D and 3D chemical descriptors as tools for predicting the biological targets of ligand probes, on the basis of their similarity to reference molecules in a 46,000 compound, biologically annotated chemical database. Overall, we found that the 2D methods employed here outperform the 3D (88% vs 67% success) in correct target prediction. However, the 3D descriptors proved superior in cases of probes with low structural similarity to other compounds in the database (singletons). Additionally, the 3D method (FEPOPS) shows promise for providing pharmacophoric alignment of the small molecules' chemical features consistent with those seen in experimental ligand/ receptor complexes. These results suggest that querying annotated chemical databases with a systematic combination of both 2D and 3D descriptors will prove more effective than employing single methods.
Abstract: The structure of epothilone A, bound to alpha,beta-tubulin in zinc-stabilized sheets, was determined by a combination of electron crystallography at 2.89 angstrom resolution and nuclear magnetic resonance-based conformational analysis. The complex explains both the broad-based epothilone structure-activity relationship and the known mutational resistance profile. Comparison with Taxol shows that the longstanding expectation of a common pharmacophore is not met, because each ligand exploits the tubulin-binding pocket in a unique and independent manner.
Abstract: Nb-containing polyoxometalates (POMs) of the Wells-Dawson class inhibit HIV-1 protease (HIV-1P) by a new mode based on kinetics, binding, and molecular modeling studies. Reaction of alpha(1)-K(9)Li[P(2)W(17)O(61)] or alpha(2)-K(10)[P(2)W(17)O(61)] with aqueous H(2)O(2) solutions of K(7)H[Nb(6)O(19)] followed by treatment with HCl and KCl and then crystallization affords the complexes alpha(1)-K(7)[P(2)W(17)(NbO(2))O(61)] (alpha(1)()1) and alpha(2)-K(7)[P(2)W(17)(NbO(2))O(61)] (alpha(2)()1) in 63 and 86% isolated yields, respectively. Thermolysis of the crude peroxoniobium compounds (72-96 h in refluxing H(2)O) prior to treatment with KCl converts the peroxoniobium compounds to the corresponding polyoxometalates (POMs), alpha(1)-K(7)[P(2)W(17)NbO(62)] (alpha(1)()2) and alpha(2)-K(7)[P(2)W(17)NbO(62)] (alpha(2)()2), in moderate yields (66 and 52%, respectively). The identity and high purity of all four compounds were confirmed by (31)P NMR and (183)W NMR. The acid-induced dimerization of the oxo complexes differentiates sterically between the cap (alpha(2)) site and the belt (alpha(1)) site in the Wells-Dawson structure (alpha(2)()2 dimerizes in high yield; alpha(1)()2 does not). All four POMs exhibit high activity in cell culture against HIV-1 (EC(50) values of 0.17-0.83 microM), are minimally toxic (IC(50) values of 50 to >100 microM), and selectively inhibit purified HIV-1 protease (HIV-1P) (IC(50) values for alpha(1)()1, alpha(2)()1, alpha(1)()2, and alpha(2)()2 of 2.0, 1.2, 1.5, and 1.8 microM, respectively). Thus, theoretical, binding, and kinetics studies of the POM/HIV-1P interaction(s) were conducted. Parameters for [P(2)W(17)NbO(62)](7)(-) were determined for the Kollman all-atom (KAA) force field in Sybyl 6.2. Charges for the POM were obtained from natural population analysis (NPA) at the HF/LANL2DZ level of theory. AutoDock 2.2 was used to explore possible binding locations for the POM with HIV-1P. These computational studies strongly suggest that the POMs function not by binding to the active site of HIV-1P, the mode of inhibition of all other HIV-1P protease inhibitors, but by binding to a cationic pocket on the "hinge" region of the flaps covering the active site (2 POMs and cationic pockets per active homodimer of HIV-1P). The kinetics and binding studies, conducted after the molecular modeling, are both in remarkable agreement with the modeling results: 2 POMs bind per HIV-1P homodimer with high affinities (K(i) = 1.1 +/- 0.5 and 4.1 +/- 1.8 nM in 0.1 and 1.0 M NaCl, respectively) and inhibition is noncompetitive (k(cat) but not K(m) is affected by the POM concentration).
Abstract: The chemotherapeutic drug Taxol is known to interact within a specific site on beta-tubulin. Although the general location of the site has been defined by photoaffinity labeling and electron crystallography, the original data were insufficient to make an absolute determination of the bound conformation. We have now correlated the crystallographic density with analysis of Taxol conformations and have found the unique solution to be a T-shaped Taxol structure. This T-shaped or butterfly structure is optimized within the beta-tubulin site and exhibits functional similarity to a portion of the B9-B10 loop in the alpha-tubulin subunit. The model provides structural rationalization for a sizeable body of Taxol structure-activity relationship data, including binding affinity, photoaffinity labeling, and acquired mutation in human cancer cells.
Abstract: Numerous structure-activity studies combining synthesis and bioassay have been performed for the anti-cancer drug Taxol. The four-membered D-ring, an oxetane, is one of four structural features regarded to be essential for biological activity. This proposition is examined by application of a Taxol-epothilone minireceptor, K(i) estimation for microtubule binding and docking of Taxol analogues into a model of the Taxol-tubulin complex. In this way, we evaluate the two characteristics considered responsible for oxetane function: (1) rigidification of the tetracyclic Taxol core to provide an appropriate framework for presenting the C-2, C-4, C-13 side chains to the microtubule protein and (2) service as a hydrogen-bond acceptor. An energy decomposition analysis for a series of Taxol analogues demonstrates that the oxetane ring clearly operates by both mechanisms. However, a broader analysis of four-membered ring containing compounds, C- and D-seco derivatives, and structures with no oxetane equivalent underscores that the four-membered ring is not necessary for Taxol analogue bioactivity. Other functional groups and ligand-protein binding characteristics are fully capable of delivering Taxol biobehavior as effectively as the oxetane D-ring. This insight may contribute to the design and development of novel anticancer drugs.
