Current Computer - Aided Drug Design - Volume 9, Issue 3, 2013

Vector Borne Diseases (VBD) present a serious threat to millions of people. In this paper various computational approaches towards new drugs design against some of them are reviewed. Malaria attracts particular attention of computational medicinal chemists. A promising strategy of the fight with VBD is usage of insect repellents. N,N-Diethyl-m-toluamide (DEET) has been the mostly used mosquito repellent for over five decades. Its mode of action is still a matter of intensive studies and debate. A possible mechanism of DEET activity is inactivation of odorant receptor proteins expressed in female mosquitoes, and being critical for finding a prey. In order to check possible interactions of DEET with such a transmembrane protein and to indicate a plausible biophore, we have constructed a hybrid "ab initio" model of Anopheles gambiae Odorant Receptor Protein 1 (AgOR1). The transmembrane regions of AgOR1 were predicted using 10 different bioinformatics algorithms and a consensus approach. A full torsional potential energy surface of DEET was determined using the AM1 method and low energy conformers were further optimized using the HF/6-31G method. DEET and a series of diastereomers of alternative repellent cyclohex-3-enyl 2-methylpiperidin-1-yl ketone (220) was docked to the AgOR1 model using the AutoDock 3.0.5 code, and possible interactions sites inside this GPCR AgOR1 were identified.

PMD (p-menthane-3-8-diol) is an insect repellent that can be synthesized chemically or derived from a steam distillate residue of the leaves of lemon eucalyptus, Corymbia citriodora. It is one of the few natural product endorsed by the Center for Disease Control (USA) for topical application to protect against mosquitoes though it is not as effective as the common repellent DEET (N,N -diethyl-1,3-toluamide). However, DEET has several undesirable side effects and toxicity too. Thus, although PMDs are comparatively safer than DEET, no quantitative structure activity relationship (QSAR) and pharmacophore modeling studies have been reported in literature to improve efficacy and aid further development of more effective PMD analogues. In this study, we report results of quantum chemical analysis of stereoelectronic properties and pharmacophore modeling of PMD and eight of its synthetic derivatives to aid discovery and design of more effective PMD analogues. Stereo-electronic analysis indicates that lower aqueous stabilization (favorable lipophilicity) and larger separation of electrostatic potential energy together with a large localized negative electrostatic potential region by the oxygen atom play important roles for repellent activity. Consistent to these properties, the generated pharmacophore model of the PMDs showed two aliphatic hydrophobic and a hydrogen-bond donor features for potent activity. These results aided us to design more effective PMD repellents which are currently under further investigations.

The study reports the QSAR rationales from multilateral approaches for the antimalarial activities (against W2 and TM90-C2B strains of Plasmodium falciparum) and physical properties (aqueous solubility, permeability and logD) of 1,2,3,4-tetrahydroacridin-9(10H)-one (THA) derivatives. A modification to Free-Wilson matrix (parameter set) is introduced to improve the compound to parameter ratio of the analysis carried out in the study. The models from modified Free-Wilson and physicochemical parameters suggest that 5th (R1) and 8th (R4) positions of THA scaffold should be free from substitution for the antimalarial activity against both the strains. The THA' s 6th (R2) and 7th (R3) positions' substituents with a blend of hydrophobicity and polarity offer scope to modulate the activities of these compounds. The substituent group contributions derived here gives opportunity to examine various combinations possible from these analogues. The models from the topological and other structural indices has suggested that MLOGP, LP1 and JGI6 as three common important features for activity against W2 and TM90 strains. The topological indices participated in the models suggested a preference for longer path lengths in molecules for better activity against both strains. The PLS analysis of the descriptors identified in the feature selection approach, combinatorial protocol in multiple linear regression (CP-MLR), has suggested that MATS3e and nRORPh are best suited to modulate the W2 activity, and Me, T(S..F) and ARR are best suited to modulate the TM90 activity. In these derivatives the analysis of aqueous solubility, permeability and logD has suggested that most of the R1 and R4 substituents have contributed with same arithmetic sign to theses properties. For all other variations in these compounds, the permeability increased with increasing logD and decreased with increasing aqueous solubility. The results suggest ways for the activity-property modulation in these analogues.

The emergence of multidrug resistance of the currently available antimalarial drugs has led to the need of the discovery and development of new antimalarial compounds. In the present study, we have used a novel group based quantitative structure-activity relationship (G-QSAR) approach, which allows to establish a correlation of chemical group variation at different molecular sites of interest with the biological activity, using a series of 53 antimalarial endochin analogs. In our previous work, we developed QSAR models for this data set using different chemometric tools and tried to emphasize on importance of descriptor thinning and noise reduction prior to feature selection step. In the present paper, we have tried to select optimal subset of variables using a new strategy for the development of robust G-QSAR models. Starting with an initial pool of 6395 descriptors, we have finally used 51 descriptors for model development using genetic methods. The best model showed encouraging values for internal.....

Aquilaria agallocha Roxb. family, Thymelaeaceae, is an evergreen plant of South-East Asia, commonly described as aloe wood or agarwood. Traditionally, the bark, root and heartwood are used for their medicinal properties as a folk medicine for hundreds of years. Chemical analyses revealed that the bulk of the oil is constituted by agarospirol (12.5%), jinkoh-eremol (11.8%) and hinesol (8.9%) as major contributor. In the present work, a QSAR model for antiinflammatory activity of 10-epi-γ-Eudesmol, jinkoh-eremol, agarospirol and other compounds has been developed by multiple linear regression method. The r2 and rCV2 of a model were 0.89 and 0.81 respectively. In silico molecular docking study suggests that compound 10-epi-γ-Eudesmol, jinkoh-eremol and agarospirol are preferentially more active than other identified compounds with strong binding affinity to major anti-inflammatory and immunomodulatory receptors. The oil displayed a significant and dose dependent reduction of 12-O-tetradecanoylphorobol-13 acetate (TPA)- induced ear edema and MDA activity when compared with vehicle treated mice. Pro-inflammatory cytokines (IL-1β, IL-6 and TNF-α) were also reduced significantly in a dose dependent manner in all the TPA treated groups as compared to control. The present study indicates that agarwood oil significantly reduced the skin thickness, ear weight, oxidative stress and pro-inflammatory cytokines production in TPA-induced mouse ear inflammation model and contributed towards validation of its traditional use to treat inflammation related ailments.

Present study describes the use of multiple pharmacophore mapping and three dimensional quantitative structure selectivity relationship (3D-QSSR) analysis for human aldose reductase (hALR2) and human aldehyde reductase (hALR1) inhibitors to develop selective molecules against hALR2. Two pharmacophore models one each for highly active hALR2 inhibitors and highly active hALR1 inhibitors were generated. Atom based 3D-QSSR analysis was performed employing docking based alignment of molecules. Sequential multi-step virtual screening was performed to screen PHASE database of approximately 1.5 million molecules. Finally, 22 potent and highly selective hALR2 inhibitors were obtained as potential hits using virtual screening protocol. This study of multiple pharmacophore mapping combined with 3D-QSSR analysis provided deep insight into the structural features requirement for selectivity and may be used as novel tool to design new selective and potent hALR2 inhibitors.

A quantitative structure-activity relationship (QSAR) study has been carried out on acetylcholinesterase (AChE) inhibitors with comparative field analysis (CoMFA), comparative molecular similarity indices analysis (CoMSIA) and hologram quantitative structure-activity relationship (HQSAR). In order to investigate the effect of alignment on modeling and find out the best alignment strategy, three different alignment rules were applied to generate CoMFA and CoMSIA models. Statistical results of the highly significant models (CoMFA q2 = 0.748, r2 =0.996, predicted r2 =0.789; CoMSIA q2 =0.755, r2 =0.973, predicted r2 = 0.706; HQSAR q2 = 0.884, r2 = 0.973, predicted r2 = 0.734) reveal considerable predictive ability. Analysis of the contour maps of CoMFA and CoMSIA models and the atomic contribution maps of HQSAR model may contribute to develop novel and potential AChE inhibitors.

The growing power of central processing units (CPU) has made it possible to use quantum mechanical (QM) calculations for in silico drug discovery. However, limited CPU power makes large-scale in silico screening such as virtual screening with QM calculations a challenge. Recently, general-purpose computing on graphics processing units (GPGPU) has offered an alternative, because of its significantly accelerated computational time over CPU. Here, we review a GPGPU-based supercomputer, TSUBAME2.0, and its promise for next generation in silico drug discovery, in high-density (HD) silico drug discovery.

Cyclohexyl methylphosphonofluoridate (cyclosarin, cyclosin, GF) is a highly toxic organophosphorus (OP) nerve agent considered as potential warfare threats and known to be resistant to conventional oxime antidotal therapy. To aid discovery of novel antidotes for GF toxicity, a three-dimensional in silico pharmacophore model for reactivation efficacy against GF intoxication is presented. The model was generated from published experimental percentage reactivation data on oximes as changes of AChE/BuChE activities in the whole blood after cyclosarin intoxication and administration. The generated pharmacophore model was found to contain a hydrogen bond donor site and two ring aromatic sites as necessary optimal features for reactivation of GF intoxication. Stereo-electronic features of oximes reported by us earlier provided guidance to develop the model and were found to be consistent with the reported structure activity data. Furthermore, from virtual screening of two commercial databases, Maybridge and ChemNavigator using map-fitting of the model led us to identify two new non-oxime compounds showing reactivation efficacy within 10-fold range of 2-PAM for DFP-inhibited AChE. Since GF is a G simulator like DFP (diisopropylfluorophosphate), the model should have the potential for discovery of novel reactivators against GF intoxication.

The discovery and use of antibacterials represents a primary success of modern pharmaceutical industry. However, the pace of antibacterial discovery was heavily hindered by a series of technical difficulties and the unfavorable economics in recent years. The past decade has witnessed rapid progresses in omics and systems biology, which provided an unprecedented opportunity to accelerate the discovery of antibacterials. In this article, we first summarize the successful use of metabolic network analysis in antibacterial discovery. Then, we reveal that metabolite concentration serves as a useful criterion for selecting antimicrobial targets. The essential enzymes with low substrate concentrations (< 0.5 mM) are more druggable antibacterial targets. Besides, we find that the solubility of clinically used competitive antibacterials is at least 100 times higher than the concentrations of the competed substrates. By the new-proposed criterion, we not only identify some promising antibacterial targets but also explain some perplexing experimental observations as well.

Chloroquine resistance is nowadays a great problem. Aurone derivatives are effective against chloroquine resistant parasite. Ligand based validated comparative chemometric modeling through 2D-QSAR and kNN-MFA 3DQSAR studies as well as common feature 3D pharmacophore mapping were done on thirtyfive aurone derivatives having antimalarial activity. Statistically significant 2D-QSAR models were generated on unsplitted as well as splitted dataset by MLR and PLS technique. The MLR model of the unsplitted method was validated by two-deep cross validation and 10 fold cross validation for determining the predictive ability. The PLS technique of the unsplitted method was done to compare the significance of these methods. In the splitted method, model was developed on the training set by Y-based ranking method by using the same descriptors and was validated on fifty pairs of the test and the training sets by k-MCA technique. These models generated by using the same descriptors were well validated irrespective of MLR as well as PLS analysis of unsplitted as well as splitted methods and are showing similar results. Therefore, these descriptors and model generated were reliable and robust. The kNN-MFA 3D-QSAR models were generated by three variable selection methods: genetic algorithm, simulated annealing and stepwise regression. The kNN-MFA 3D-QSAR results support the 2D QSAR data and in turn validate the earlier observed SAR results. Common feature 3D-pharmacophore generation was performed on these compounds to validate both 2D and 3D-QSAR studies as well as the earlier observed SAR data. The work highlights the required structural features for the higher antimalarial activity.

This study investigated 3D quantitative structure–activity relationships (QSAR) for a range of substituted benzimidazole derivatives as AngII-AT1 receptor antagonists by comparative molecular field analysis (CoMFA) and comparative molecular similarity indices (CoMSIA). The alignment strategy was used for these compounds by means of Distill function defined in SYBYL X 1.2. The best CoMFA and CoMSIA models were obtained for the training set compounds was statistically significant with leave-one-out (LOO) validation correlation coefficient (q2) of 0.613 and 0.622, cross validated coefficient (r2 cv) of 0.617 and 0.607, respectively and conventional coefficient (r2 ncv) of 0.886 and 0.859, respectively. Both the models were validated by a test set of 18 compounds giving satisfactory predicted correlation coefficient (r2 pred) of 0.714 and 0.549 for CoMFA and CoMSIA models, respectively. Generated 3D QSAR models were used for the prediction of pIC50 of an external dataset of 10 compounds for predictive validation, which gave conventional r2 of 0.893 for CoMFA model, and 0.774 for CoMSIA model. We identified some key features in substituted benzimidazole derivatives, such as the importance of lipophilicity and H-bonding at 2- and 5, 6, 7- position of benzimidazole ring, respectively, for good antagonistic activity. CoMFA and CoMSIA models generated in this work provide useful information for the design of new compounds and helped in prediction of antagonistic activity.