Machine learning predictions of MHC-II specificities reveal alternative binding mode of class II epitopes.

CD4+ T cells orchestrate the adaptive immune response against pathogens and cancer by recognizing epitopes presented on class II major histocompatibility complex (MHC-II) molecules. The high polymorphism of MHC-II genes represents an important hurdle toward accurate prediction and identification of CD4+ T cell epitopes. Here we collected and curated a dataset of 627,013 unique MHC-II ligands identified by mass spectrometry. This enabled us to precisely determine the binding motifs of 88 MHC-II alleles across humans, mice, cattle, and chickens. Analysis of these binding specificities combined with X-ray crystallography refined our understanding of the molecular determinants of MHC-II motifs and revealed a widespread reverse-binding mode in HLA-DP ligands. We then developed a machine-learning framework to accurately predict binding specificities and ligands of any MHC-II allele. This tool improves and expands predictions of CD4+ T cell epitopes and enables us to discover viral and bacterial epitopes following the aforementioned reverse-binding mode.

Deciphering principles of nucleosome interactions and impact of cancer-associated mutations from comprehensive interaction network analysis.

Nucleosomes represent hubs in chromatin organization and gene regulation and interact with a plethora of chromatin factors through different modes. In addition, alterations in histone proteins such as cancer mutations and post-translational modifications have profound effects on histone/nucleosome interactions. To elucidate the principles of histone interactions and the effects of those alterations, we developed histone interactomes for comprehensive mapping of histone-histone interactions (HHIs), histone-DNA interactions (HDIs), histone-partner interactions (HPIs) and DNA-partner interactions (DPIs) of 37 organisms, which contains a total of 3808 HPIs from 2544 binding proteins and 339 HHIs, 100 HDIs and 142 DPIs across 110 histone variants. With the developed networks, we explored histone interactions at different levels of granularities (protein-, domain- and residue-level) and performed systematic analysis on histone interactions at a large scale. Our analyses have characterized the preferred binding hotspots on both nucleosomal/linker DNA and histone octamer and unraveled diverse binding modes between nucleosome and different classes of binding partners. Last, to understand the impact of histone cancer-associated mutations on histone/nucleosome interactions, we complied one comprehensive cancer mutation dataset including 7940 cancer-associated histone mutations and further mapped those mutations onto 419,125 histone interactions at the residue level. Our quantitative analyses point to histone cancer-associated mutations' strongly disruptive effects on HHIs, HDIs and HPIs. We have further predicted 57 recurrent histone cancer mutations that have large effects on histone/nucleosome interactions and may have driver status in oncogenesis.

Structural studies of WDR5 in complex with MBD3C WIN motif reveal a unique binding mode.

The nucleosome remodeling and deacetylase (NuRD) complex plays a pivotal role in chromatin regulation and transcriptional repression. In mice, methyl-CpG binding domain 3 isoform C (MBD3C) interacts specifically with the histone H3 binding protein WD repeat-containing protein 5 (WDR5) and forms the WDR5-MBD3C/Norde complex. Despite the functional significance of this interaction on embryonic stem cell gene regulation, the molecular mechanism underlying MBD3C recognition by WDR5 remains elusive. Here, we determined the crystal structure of WDR5 in complex with the peptide (residues 40-51) derived from the MBD3C protein at a resolution of 1.9 Å. Structural analysis revealed that MBD3C utilizes a unique binding mode to interact with WDR5, wherein MBD3C Arg43 and Phe47 are involved in recognizing the WDR5-interacting (WIN) site and Tyr191-related B site on the small surface of WDR5, respectively. Notably, the binding induces a ∼91° rotation of WDR5 Tyr191, generating the hydrophobic B site. Furthermore, mutation experiments combined with isothermal titration calorimetry (ITC) assays confirmed the importance of both Arg43 and Phe47 in mediating WDR5 binding affinity. By determining structures of various peptides bound to WDR5, we demonstrated that the WDR5 WIN site and B site can be concurrently recognized by WIN motif peptides containing ''Arg-Cies/Ser-Arg-Val-Phe'' consensus sequence. Overall, this study reveals the structural basis for the formation of the WDR5-MBD3C subcomplex and provides new insights into the recognition mode of WDR5 for the WIN motif. Moreover, these findings shed light on structural-based designs of WDR5-targeted anti-cancer small molecule inhibitors or peptide-mimic drugs.

Crystal structure of the GluK1 ligand-binding domain with kainate and the full-spanning positive allosteric modulator BPAM538.

Kainate receptors play an important role in the central nervous system by mediating postsynaptic excitatory neurotransmission and modulating the release of the inhibitory neurotransmitter GABA through a presynaptic mechanism. To date, only three structures of the ligand-binding domain (LBD) of the kainate receptor subunit GluK1 in complex with positive allosteric modulators have been determined by X-ray crystallography, all belonging to class II modulators. Here, we report a high-resolution structure of GluK1-LBD in complex with kainate and BPAM538, which belongs to the full-spanning class III. One BPAM538 molecule binds at the GluK1 dimer interface, thereby occupying two allosteric binding sites simultaneously. BPAM538 stabilizes the active receptor conformation with only minor conformational changes being introduced to the receptor. Using a calcium-sensitive fluorescence-based assay, a 5-fold potentiation of the kainate response (100 µM) was observed in presence of 100 µM BPAM538 at GluK1(Q)b, whereas no potentiation was observed at GluK2(VCQ)a. Using electrophysiology recordings of outside-out patches excised from HEK293 cells, BPAM538 increased the peak response of GluK1(Q)b co-expressed with NETO2 to rapid application of 10 mM L-glutamate with 130 ± 20 %, and decreased desensitization determined as the steady-state/peak response ratio from 23 ± 2 % to 90 ± 4 %. Based on dose-response relationship experiments on GluK1(Q)b the EC50 of BPAM538 was estimated to be 58 ± 29 µM.

fastDRH: a webserver to predict and analyze protein-ligand complexes based on molecular docking and MM/PB(GB)SA computation.

Predicting the native or near-native binding pose of a small molecule within a protein binding pocket is an extremely important task in structure-based drug design, especially in the hit-to-lead and lead optimization phases. In this study, fastDRH, a free and open accessed web server, was developed to predict and analyze protein-ligand complex structures. In fastDRH server, AutoDock Vina and AutoDock-GPU docking engines, structure-truncated MM/PB(GB)SA free energy calculation procedures and multiple poses based per-residue energy decomposition analysis were well integrated into a user-friendly and multifunctional online platform. Benefit from the modular architecture, users can flexibly use one or more of three features, including molecular docking, docking pose rescoring and hotspot residue prediction, to obtain the key information clearly based on a result analysis panel supported by 3Dmol.js and Apache ECharts. In terms of protein-ligand binding mode prediction, the integrated structure-truncated MM/PB(GB)SA rescoring procedures exhibit a success rate of >80% in benchmark, which is much better than the AutoDock Vina (~70%). For hotspot residue identification, our multiple poses based per-residue energy decomposition analysis strategy is a more reliable solution than the one using only a single pose, and the performance of our solution has been experimentally validated in several drug discovery projects. To summarize, the fastDRH server is a useful tool for predicting the ligand binding mode and the hotspot residue of protein for ligand binding. The fastDRH server is accessible free of charge at http://cadd.zju.edu.cn/fastdrh/.

Benchmarking ANI potentials as a rescoring function and screening FDA drugs for SARS-CoV-2 Mpro.

Here, we introduce the use of ANI-ML potentials as a rescoring function in the host-guest interaction in molecular docking. Our results show that the "docking power" of ANI potentials can compete with the current scoring functions at the same level of computational cost. Benchmarking studies on CASF-2016 dataset showed that ANI is ranked in the top 5 scoring functions among the other 34 tested. In particular, the ANI predicted interaction energies when used in conjunction with GOLD-PLP scoring function can boost the top ranked solution to be the closest to the x-ray structure. Rapid and accurate calculation of interaction energies between ligand and protein also enables screening of millions of drug candidates/docking poses. Using a unique protocol in which docking by GOLD-PLP, rescoring by ANI-ML potentials and extensive MD simulations along with end state free energy methods are combined, we have screened FDA approved drugs against the SARS-CoV-2 main protease (Mpro). The top six drug molecules suggested by the consensus of these free energy methods have already been in clinical trials or proposed as potential drug molecules in previous theoretical and experimental studies, approving the validity and the power of accuracy in our screening method.

FDA-approved small molecule kinase inhibitors for cancer treatment (2001-2015): Medical indication, structural optimization, and binding mode Part I.

The dysregulation of kinases has emerged as a major class of targets for anticancer drug discovery given its node roles in the etiology of tumorigenesis, progression, invasion, and metastasis of malignancies, which is validated by the FDA approval of 28 small molecule kinase inhibitor (SMKI) drugs for cancer treatment at the end of 2015. While the preclinical and clinical data of these drugs are widely presented, it is highly essential to give an updated review on the medical indications, design principles and binding modes of these anti-tumor SMKIs approved by the FDA to offer insights for the future development of SMKIs with specific efficacy and safety.

Halogenated bisphenol A derivatives potently inhibit human and rat 11ß-hydroxysteroid dehydrogenase 1: Structure-activity relationship and molecular docking.

Chlorinated bisphenol A (BPA) derivatives are formed during chlorination process of drinking water, whereas bisphenol S (BPS) and brominated BPA and BPS (TBBPA and TBBPS) were synthesized for many industrial uses such as fire retardants. However, the effect of halogenated BPA and BPS derivatives on glucocorticoid metabolizing enzyme 11ß-hydroxysteroid dehydrogenase 1 (11ß-HSD1) remains unclear. The inhibitory effects of 6 BPA derivatives in the inhibition of human and rat 11ß-HSD1 were investigated. The potencies for inhibition on human 11ß-HSD1 were TBBPA (IC50, 3.87 µM) = monochloro BPA (MCBPA, 4.08 µM) = trichloro BPA (TrCBPA, 4.41 µM) > tetrachloro BPA (TCBPA, 9.75 µM) > TBBPS (>100 µM) = BPS (>100 µM), and those for rat 11ß-HSD1 were TrCBPA (IC50, 2.76 µM) = MCBPA (3.75 µM) > TBBPA (39.58 µM) > TCBPA = TBBPS = BPS. All these BPA derivatives are mixed/competitive inhibitors of both human and rat enzymes. Molecular docking studies predict that MCBPA, TrCBPA, TCBPA, and TBBPA all bind to the active site of human 11ß-HSD1, forming hydrogen bonds with catalytic residue Ser170 except TCBPA. Regression of the lowest binding energy with IC50 values revealed a significant inverse linear regression. In conclusion, halogenated BPA derivatives are mostly potent inhibitors of human and rat 11ß-HSD1, and there is structure-dependent inhibition.

Binding Pattern and Structural Interactome of the Anticancer Drug 5-Fluorouracil: A Critical Review.

5-Fluorouracil (5-FU) stands as one of the most widely prescribed chemotherapeutics. Despite over 60 years of study, a systematic synopsis of how 5-FU binds to proteins has been lacking. Investigating the specific binding patterns of 5-FU to proteins is essential for identifying additional interacting proteins and comprehending their medical implications. In this review, an analysis of the 5-FU binding environment was conducted based on available complex structures. From the earliest complex structure in 2001 to the present, two groups of residues emerged upon 5-FU binding, classified as P- and R-type residues. These high-frequency interactive residues with 5-FU include positively charged residues Arg and Lys (P type) and ring residues Phe, Tyr, Trp, and His (R type). Due to their high occurrence, 5-FU binding modes were simplistically classified into three types, based on interactive residues (within <4 Å) with 5-FU: Type 1 (P-R type), Type 2 (P type), and Type 3 (R type). In summary, among 14 selected complex structures, 8 conform to Type 1, 2 conform to Type 2, and 4 conform to Type 3. Residues with high interaction frequencies involving the N1, N3, O4, and F5 atoms of 5-FU were also examined. Collectively, these interaction analyses offer a structural perspective on the specific binding patterns of 5-FU within protein pockets and contribute to the construction of a structural interactome delineating the associations of the anticancer drug 5-FU.

Unraveling the Binding Mode of Cyclic Adenosine-Inosine Monophosphate (cAIMP) to STING through Molecular Dynamics Simulations.

The stimulator of interferon genes (STING) plays a significant role in immune defense and protection against tumor proliferation. Many cyclic dinucleotide (CDN) analogues have been reported to regulate its activity, but the dynamic process involved when the ligands activate STING remains unclear. In this work, all-atom molecular dynamics simulations were performed to explore the binding mode between human STING (hSTING) and four cyclic adenosine-inosine monophosphate analogs (cAIMPs), as well as 2',3'-cGMP-AMP (2',3'-cGAMP). The results indicate that these cAIMPs adopt a U-shaped configuration within the binding pocket, forming extensive non-covalent interaction networks with hSTING. These interactions play a significant role in augmenting the binding, particularly in interactions with Tyr167, Arg238, Thr263, and Thr267. Additionally, the presence of hydrophobic interactions between the ligand and the receptor further contributes to the overall stability of the binding. In this work, the conformational changes in hSTING upon binding these cAIMPs were also studied and a significant tendency for hSTING to shift from open to closed state was observed after binding some of the cAIMP ligands.

Molecular determinants of tetrahydrocannabinol binding to the glycine receptor.

The recognition of Cannabis as a source of new compounds suitable for medical use has attracted strong interest from the scientific community in its research, and substantial progress has accumulated regarding cannabinoids' activity; however, a thorough description of their molecular mechanisms of action remains a task to complete. Highlighting their complex pharmacology, the list of cannabinoids' interactors has vastly expanded beyond the canonical cannabinoid receptors. Among those, we have focused our study on the glycine receptor (GlyR), an ion channel involved in the modulation of nervous system responses, including, to our interest, sensitivity to peripheral pain. Here, we report the use of computational methods to investigate possible binding modes between the GlyR and Δ9 -tetrahydrocannabinol (THC). After obtaining a first pose for the THC binding from a biased molecular docking simulation and subsequently evaluating it by molecular dynamic simulations, we found a dynamic system with an identifiable representative binding mode characterized by the specific interaction with two transmembrane residues (Phe293 and Ser296). Complementarily, we assessed the role of membrane cholesterol in this interaction and positively established its relevance for THC binding to GlyR. Lastly, the use of restrained molecular dynamics simulations allowed us to refine the description of the binding mode and of the cholesterol effect. Altogether, our findings contribute to the current knowledge about the GlyR-THC mode of binding and propose a new starting point for future research on how cannabinoids in general, and THC in particular, modulate pain perception in view of its possible clinical applications.

Elucidating the significance of molecular interaction between sulphur doped zinc oxide nanoparticles and serum albumin using multispectroscopic approach.

Ingenious nanomaterials with improved biocompatibility and multifunctional properties are gaining vital significance in biomedical applications, including advanced drug delivery and nanotheranostics. In a biological system, these nanoparticles interact with serum proteins forming a dynamic corona that affects their biological or toxicological properties producing undesirable effects. Thus, the current study focuses on the synthesis of sulphur-doped zinc oxide nanoparticles (ZnO/S NPs) and characterizing their mechanism of interaction with serum proteins using multispectroscopic approach. ZnO/S NPs were synthesized by employing a co-precipitation approach and characterized using various analytical techniques. The results of interaction studies demonstrated that ZnO/S NPs interact with serum albumins via the static quenching process. Analysis of thermodynamic parameters (ΔG, ΔH and ΔS) revealed that the binding process is spontaneous, exothermic and van der Waals force or hydrogen bonding plays a major role. The interaction of ZnO/S NPs with tyrosine residue in bovine serum albumin was established by synchronous fluorescence spectroscopy. In addition, the results of UV-visible, circular dichroism, Fourier transform infrared, Forster's resonance energy transfer theory and dynamic light scattering spectroscopic studies revealed that the ZnO/S NPs interact with albumin by inducing the conformational changes in secondary structure and reducing the α-helix content.

Discovery of novel and potent InhA direct inhibitors by ensemble docking-based virtual screening and biological assays.

Multidrug-resistant tuberculosis (MDR-TB) continues to spread worldwide and remains one of the leading causes of death among infectious diseases. The enoyl-acyl carrier protein reductase (InhA) belongs to FAS-II family and is essential for the formation of the Mycobacterium tuberculosis cell wall. Recent years, InhA direct inhibitors have been extensively studied to overcome MDR-TB. However, there are still no inhibitors that have entered clinical research. Here, the ensemble docking-based virtual screening along with biological assay were used to identify potent InhA direct inhibitors from Chembridge, Chemdiv, and Specs. Ultimately, 34 compounds were purchased and first assayed for the binding affinity, of which four compounds can bind InhA well with KD values ranging from 48.4 to 56.2 µM. Among them, compound 9,222,034 has the best inhibitory activity against InhA enzyme with an IC50 value of 18.05 µM. In addition, the molecular dynamic simulation and binding free energy calculation indicate that the identified compounds bind to InhA with "extended" conformation. Residue energy decomposition shows that residues such as Tyr158, Met161, and Met191 have higher energy contributions in the binding of compounds. By analyzing the binding modes, we found that these compounds can bind to a hydrophobic sub-pocket formed by residues Tyr158, Phe149, Ile215, Leu218, etc., resulting in extensive van der Waals interactions. In summary, this study proposed an efficient strategy for discovering InhA direct inhibitors through ensemble docking-based virtual screening, and finally identified four active compounds with new skeletons, which can provide valuable information for the discovery and optimization of InhA direct inhibitors.

Chalcone derivatives as xanthine oxidase inhibitors: synthesis, binding mode investigation, biological evaluation, and ADMET prediction.

Xanthine oxidase (XO) is a crucial target for the treatment of hyperuricemia and gout. A series of derivatives based on natural 3,4-dihydroxychalcone, obtained from Carthamus tinctorious and Licorice, were designed and synthesized. Nine derivatives (9a-e, 10b,c, and 15a,b) exhibited apparent XO inhibitory activity in vitro (IC50 values varied from 0.121 to 7.086 µM), 15b presented the most potent inhibitory activity (IC50 = 0.121 µM), which was 27.47-fold higher than that of allopurinol (IC50 = 3.324 µM). The SAR analysis indicated that introducing hydroxyl groups at 3'/4'/5'-position on ring A was more beneficial to the inhibition of XO than at 2'/6'-position; the removal of 3­hydroxyl group on ring B could weaken the inhibitory potency of hydroxychalcones on XO, but it was beneficial to the XO inhibitory potency of methoxychalcones. Molecule modeling studies afforded insights into the binding mode of 15b with XO and supported the findings of SAR analysis. Additionally, kinetics studies demonstrated that 15b presented a reversible and competitive XO inhibitor, which spontaneously combined with XO through hydrophobic force, and finally changed the secondary conformation of XO. Furthermore, the acute hyperuricemia model was employed to investigate the hypouricemic effect of 15b, which could effectively reduce the serum uric acid levels of rats at an oral dose of 10 mg/kg. ADMET prediction suggested that compound 15b possessed good pharmacokinetic properties. Briefly, compound 15b emerges as an interesting XO inhibitor for the treatment of hyperuricemia and gout with beneficial effects on serum uric acid levels regulating. Meanwhile, the XO inhibitors with chalcone skeleton will deserve further attention and discussion.

Progress in programmed cell death-1/programmed cell death-ligand 1 pathway inhibitors and binding mode analysis.

Programmed cell death protein 1 (PD-1)/programmed cell death protein ligand 1 (PD-L1) plays an important role in negative regulating immunity. The search for effective PD-1/PD-L1 inhibitors has been at the cutting-edge of academic and industrial medicinal chemistry, leading to the emergence of 16 clinical candidate drugs and the launch of six monoclonal antibodies (mAbs) drugs. However, due to the unclear mechanism of the interaction between drugs and substances in vivo, the screening of preclinical drugs often takes a long time. In order to shorten the time of drug development as much as possible, the binding mode analysis that can simulate the interaction between drugs and substances in vivo at the molecular level can significantly shorten the drug development process. This paper reviews the mechanism of PD-1/PD-L1 signaling pathway at the molecular level, as well as the research progress and obstacles of inhibitors. Besides, we analyzed the binding mode of recently reported PD-1/PD-L1 inhibitors with PD-1 or PD-L1 protein in detail in order to provide ideas for the development of PD-1/PD-L1 inhibitors.

Synthesis, biological evaluation, and molecular modeling of ORM-10921 and its analogs as α2C -adrenoceptor antagonists.

The α2C -adrenoceptor (α2C -AR) is regarded as one of the potential targets for antipsychotics. A few of structurally diverse α2C -AR antagonists have been reported, among which ORM-10921, containing one rigid tetracyclic framework with two neighboring chiral centers, has exhibited remarkable antipsychotic-like effects and pro-cognitive properties in different animal models. Yet the binding mode of ORM-10921 remains elusive. In this study, all of its four stereoisomers and a set of its analogs were synthesized and in vitro evaluated for their α2C -AR antagonist activities. The molecular docking study and hydration site analysis gave a rational explanation for the biological results, which might provide helpful insights into the binding mode and future optimization.

Cembrane Diterpenes Possessing Nonaromatic Oxacycles from the Hainan Soft Coral Sarcophyton mililatensis.

Documents on the chemical composition of the soft coral Sarcophyton mililatensis are sparse. The present investigation of the Hainan soft coral S. mililatensis resulted in the discovery of six new cembrane diterpenes, sarcoxacyclols A-F (1-6) and four known analogs (7-10). Their structures were elucidated by extensive spectroscopic analysis along with a comparison with the data in current literature. The nonaromatic oxacycles in their structures were the rarely found tetrahydrofuran ether across C-1 and C-12 and tetrahydropyran ether across C-1 and C-11, respectively. Moreover, the absolute configuration of compound 4 was established unambiguously by X-ray diffraction analysis using Ga Kα radiation (λ = 1.34139 Å). Based on the biogenetical consideration, the absolute configurations of other five new compounds were tentatively assumed. Assessment of the bioactivity for these secondary metabolites revealed that compound 1 exhibited significant tumor necrosis factor (TNF)-α inhibitory activity (IC50 = 9.5 µmol/L), similar to the positive control dexamethasone (IC50 = 8.7 µmol/L), but no obvious cytotoxicity towards RAW264.7 cells (CC50 > 50 µmol/L). The preliminary molecular docking suggested the crucial roles of the hydroxyl and acetoxyl groups in the computational prediction of the binding mode between the diterpene and the protein.

A General Picture of Cucurbit[8]uril Host-Guest Binding: Recalibrating Bonded Interactions.

Atomic-level understanding of the dynamic feature of host-guest interactions remains a central challenge in supramolecular chemistry. The remarkable guest binding behavior of the Cucurbiturils family of supramolecular containers makes them promising drug carriers. Among Cucurbit[n]urils, Cucurbit[8]uril (CB8) has an intermediate portal size and cavity volume. It can exploit almost all host-guest recognition motifs formed by this host family. In our previous work, an extensive computational investigation of the binding of seven commonly abused and structurally diverse drugs to the CB8 host was performed, and a general dynamic binding picture of CB8-guest interactions was obtained. Further, two widely used fixed-charge models for drug-like molecules were investigated and compared in great detail, aiming at providing guidelines in choosing an appropriate charge scheme in host-guest modelling. Iterative refitting of atomic charges leads to improved binding thermodynamics and the best root-mean-squared deviation from the experimental reference is 2.6 kcal/mol. In this work, we focus on a thorough evaluation of the remaining parts of classical force fields, i.e., the bonded interactions. The widely used general Amber force fields are assessed and refitted with generalized force-matching to improve the intra-molecular conformational preference, and thus the description of inter-molecular host-guest interactions. The interaction pattern and binding thermodynamics show a significant dependence on the modelling parameters. The refitted system-specific parameter set improves the consistency of the modelling results and the experimental reference significantly. Finally, combining the previous charge-scheme comparison and the current force-field refitting, we provide general guidelines for the theoretical modelling of host-guest binding.

A reinforcement learning approach for protein-ligand binding pose prediction.

Protein ligand docking is an indispensable tool for computational prediction of protein functions and screening drug candidates. Despite significant progress over the past two decades, it is still a challenging problem, characterized by the still limited understanding of the energetics between proteins and ligands, and the vast conformational space that has to be searched to find a satisfactory solution. In this project, we developed a novel reinforcement learning (RL) approach, the asynchronous advantage actor-critic model (A3C), to address the protein ligand docking problem. The overall framework consists of two models. During the search process, the agent takes an action selected by the actor model based on the current location. The critic model then evaluates this action and predict the distance between the current location and true binding site. Experimental results showed that in both single- and multi-atom cases, our model improves binding site prediction substantially compared to a naïve model. For the single-atom ligand, copper ion (Cu2+), the model predicted binding sites have a median root-mean-square-deviation (RMSD) of 2.39 Å to the true binding sites when starting from random starting locations. For the multi-atom ligand, sulfate ion (SO42-), the predicted binding sites have a median RMSD of 3.82 Å to the true binding sites. The ligand-specific models built in this study can be used in solvent mapping studies and the RL framework can be readily scaled up to larger and more diverse sets of ligands.

Discovery of selective covalent cathepsin K inhibitors containing novel 4-cyanopyrimidine warhead based on quantum chemical calculations and binding mode analysis.

Cathepsin K (Cat K), mainly expressed by osteoclasts, plays an important role in bone resorption. Covalent Cat K inhibitors will show great potential in the future treatment of osteoporosis. It has been reported that the selectivity of covalent cathepsin K inhibitors was related to the drug's safety. The type of warhead has a crucial influence on the enzyme bioactivity and selectivity of covalent inhibitors. In order to develop novel covalent inhibitors with the selective new warhead, quantum chemical calculations were performed to estimate the reactivity of the nitrile warheads. Moreover, binding mode analysis between ligands and high homology Cat K, S and B revealed differences in non-covalent interactions. Novel covalent Cat K inhibitors containing 4-cyanopyrimidine warhead (11) were determined for the first time. Among them, compound 34 significantly inhibited Cat K (IC50 = 61.9 nM) with excellent selectivity compared to Cat S (>810-fold) and Cat B (>1620-fold), respectively. Binding mode analysis of Cat K-34 complex provided the basis for further optimization. Compound 34 could be a valuable lead compound for further research on safe and effective Cat K inhibitors.