Here we report a new drug design workflow that facilitates the transfer of structure-activity relationships (SARs) and recommends alternative fragments from SAR databases. We first prepare two collections of matched molecular series (MMS) comprising a query set of compounds with their SARs and a set derived from reference SAR databases. The second step detects MMS from the reference SAR sources, which identifies profiles similar to a query MMS according to integrated similarities of scaffold shapes and SAR trends. The third step enumerates new compounds with improved activity profiles compared with a query compound computed using a collaborative filtering algorithm. Our workflow detected direct and latent relationships between a query MMS and those derived from the reference SAR sources. Retrospective application of this workflow to the identification of factor Xa inhibitors yielded recommendations with higher predictive accuracy than a conventional quantitative SAR technique. Moreover, potent S1 binding elements were identified using SAR knowledge independent of information about ligand-protein complexes.
Trifluorothymidine (TFT) has antitumor activity, but it is easily metabolized to inert trifluorothymine by thymidine phosphorylase (TP). Accordingly, TFT alone cannot show satisfactory clinical antitumor effects. Human TP (HTP) is the main enzyme of pyrimidine nucleoside phosphorylase in human. Therefore, it has been necessary to develop a HTP inhibitor to maintain antitumor activity of TFT. Here we reveal the drug design process of HTP inhibitor based on SBDD and classical QSAR analysis. Thymine was selected as a seed compound and then 5-chlorouracil (3) was selected as a lead compound. The introduction of the imino moiety to C6 position of the lead compound (3) enhanced the inhibitory activity of TP. As a result, 5-chloro-6-[1-(2-iminopyrrolidinyl) methyl] uracil hydrochloride (TPI) was chosen as the candidate for the clinical trials. And TAS-102 (the combination of TFT and TPI in a 1:0.5 molar ratio) has been approved as Trifluridine/Tipiracil (Lonsurf) for the treatment of metastatic colorectal cancer in Japan, United States and EU.
CYP3A4 contributes to the metabolism of more than 30% of drugs in clinical use. Predicting the sites of metabolism (SOM) by CYP3A4, as well as the binding modes, for target compounds is important for the design of metabolically more stable drugs. Precisely predicting the structures of CYP3A4–ligand complexes is enormously challenging owing to the high number of conformational possibilities with its numerous binding substrates. We previously described a method for predicting the SOM of carbamazepine by means of docking and molecular dynamics (MD) simulations starting from multiple initial structures. To validate our method, we have now applied it to tolterodine, which is more flexible than carbamazepine. In addition, we evaluated the effectiveness of two methods for selecting the initial structures for MD. In analyzing the MD trajectories, we calculated the frequency with which carbon atoms at each of four groups of the tolterodine molecule approached to within a certain cutoff distance of the heme iron, and we also calculated binding free energy. We found that compared to the other three groups, the position to the experimentally determined SOM was close to the heme most frequently and had the lowest average ∆Gbinding. For selecting the MD initial structures, clustering on the basis of protein–ligand interaction fingerprints (PLIF) was substantially more robust at predicting accessibility compared with clustering based on root-mean-square deviations. These findings demonstrate that our method is applicable for a flexible ligand and that PLIF clustering is a promising method for selecting structures for MD. We succeeded to predict the experimentally determined SOM of tolterodine together with the appropriate binding mode. The predicted binding mode is useful to design metabolically more stable compounds.
Several researchers have focused on the inference of genetic networks as a process for extracting useful information from gene expression data. Their work has led to the proposal of a number of methods for genetic network inference. Yet the genetic networks inferred by these methods often contain large numbers of false-positive regulations along with the true-positives. One effective way to reduce the number of erroneous regulations is to apply inference methods that use a priori knowledge on the properties of the genetic networks. The existing inference methods adopting this approach generally use a priori knowledge and the observed gene expression data simultaneously to determine whether or not the target genetic network actually contains each of the candidate regulations. In this study, we establish a new framework for “using a priori knowledge after genetic network inference.” The framework uses a priori knowledge only to modify the genetic network that has already been inferred by the other inference method. Based on this framework, we propose a new inference method that uses multiple kinds of a priori knowledge about genetic networks. The proposed method effectively combines multiple kinds of knowledge and computes the confidence values of regulations. Here, we confirm the effectiveness of the proposed method by applying it to artificial and actual genetic network inference problems. While only a small improvement is gained from the use of multiple kinds of a priori knowledge, we can improve the performance of many other existing inference methods by combining them with the method we propose here.
Fragment molecular orbital (FMO) calculations were performed for explicitly solvated single-stranded DNA (ssDNA), ssDNA binding protein, and their complex in order to assess the solvent effects on the solutes and thereby to find optimal solvation conditions for FMO calculation. A series of solvated structures were generated with different solvent thicknesses. The structures were subjected to FMO calculation at MP2/6-31G* to obtain the net charges and internal energies of the solutes and the solute–solvent interaction energies as functions of the solvent thickness. In all cases, the properties showed complete or marginal convergence at ca. 6 Ǻ, regardless whether or not the system charge was neutralized. This suggested that the first and second solvent shells mainly determine the electronic structure of a solute while the outer solvent including ions has only minor effects, consistent with several preceding reports. In light of this, and considering safety as a factor, we conclude that a solvent shell thickness of ca. 8 Ǻ suffices for FMO calculation of the solutes.
In peptide vaccine therapy, a peptide with high affinity for human leukocyte antigen (HLA), is important to stimulate the immune system to kill cancer cells. Several methods to predict HLA–peptide binding have been reported, but most of them rely on informatics to analyze the amino acid sequence of the peptide. Although intermolecular-interaction-based analysis is expected to improve prediction accuracy, such a method generally involves a high computational cost. Therefore, comparative binding energy (COMBINE) analysis, a 3D-quantitative structure–activity relationship method, combined with a rapidly implemented protein modeling method, was applied to solve this problem. The new method enabled quick evaluation of peptide affinity predictions with accuracy beyond a statistical method. In addition, several amino acid residues of HLA, which are known to be important for peptide binding, could be identified.
Pseudohypoaldosteronism type II has been known as a rare autosomal dominant disorder caused by WNK1 [with no K (lysine) protein kinase-1] or WNK4. These serine/threonine kinases have unusual structures with a back pocket located just behind the ATP binding site. Moreover, a lysine residue (Lys 233 in WNK1) in a glycine-rich loop plays a key role in their activity. In this work, we performed docking simulations of about 9,000 compounds from a fragment library with the back pocket of WNK1 in order to discover candidate lead compounds for development of specific inhibitors. Based on binding energy index, we selected β-tetralone (compound 5) as a lead structure that interacts with the back pocket, but not with the hinge region of WNK1. Guided by the four predicted docking patterns of β-tetralone with the back pocket, we designed four derivatives A-D that were expected to form hydrogen bonds with Lys 233. Docking studies indicated that these derivatives interact selectively with Lys 233, but not with the hinge region. These compounds are considered potential lead compounds for developing selective WNK inhibitors.
Pseudohypoaldosteronism type II is a rare, familial, autosomal-dominant hypertensive disease that is caused by mutations of WNK (with no lysine [K]) protein kinases 1 and 4. WNKs lack a lysine residue in β3 strand that is generally conserved in protein kinases. WNK 1 and WNK4 share 87% homology, and possess an unusual back pocket just behind the catalytic lysine residue (Lys233 in WNK1). Therefore, compounds interacting with both the back pocket and catalytic lysine residue could be selective inhibitors. Here, we screened a fragment library for inhibitors of WNK1-mediated phosphorylation by means of mobility shift assay and surface plasmon resonance (SPR)-based binding assay. Among the identified inhibitors, some interacted with the back pocket rather than the hinge region of WNK1, as determined by SPR competitive binding assay. The results of kinase profiling suggest these compounds are promising leads for development of selective inhibitors of WNK 1 and 4.
Widespread arguments of the hydrogen bond are criticized with respect to theories of protein folding and interactions of proteins with their specific ligands. Contrary to popular belief, by no means does the hydrogen bond play an important role in determining the conformation of proteins and the interactions of proteins with specific substrates. Stereotypical thinking on the hydrogen bond constitutes a myth in modern science and seriously restricts the success of structure-based drug design.