Journal of Computational Biophysics and Chemistry

Human osteoclast-stimulating factor (OSF) induces osteoclast formation and bone resorption in osteoporosis by recruiting multiple signaling complexes with downstream partners. Protein contains a peptide-recognition Src homology 3 (SH3) domain that can recognize and bind class II linear motif $P^{-1}xx{P}^{+2}x{[K∕R]}^{+4}$ to its partner proteins. The motif is defined by two prolines at positions $-$1 and $+$2, which are the primary anchor residues required for the domain recognition, and a positively charged amino acid at position $+$4, which is the secondary anchor residue and determines the binding orientation of the motif peptides on the domain surface. In this study, we systematically examined the intermolecular interaction of OSF SH3 domain with a high-affinity decapeptide segment derived from its partner protein Sam68 at structural and energetic levels. It was found that, in addition to the primary and secondary anchor residues, the residue at peptide position $+$1 is also important, which can form a $\pi$-stacking system (consisting of multiple cation-$\pi$ or $\pi$–$\pi$ stacking interactions) with its vicinal aromatic residues Phe23, Trp49 and Tyr65 of OSF SH3 domain, thus, largely stabilizing the domain–peptide complex. Here, we assigned the position $+$1 as the third anchor residue and investigated the stacking effect by systematically substituting the position $+$1 residue with six charged/aromatic amino acids (Arg, Lys, His, Phe, Tyr and Trp) and one neutral amino acid (Ala), as well as their impacts on the domain–peptide binding. A strong stacking effect was observed in association with charged/aromatic substitutions relative to neutral substitution, conferring substantial stability to the complex formation. A further fluorescence-based assay also substantiated the computational findings; the lysine and tyrosine substitutions ($K^{+1}$ and $Y^{+1}$) were observed to significantly and moderately improve peptide affinity by 4.7-fold and 1.4-fold relative to wild-type Sam68 decapeptide ($R^{+1}$), respectively.

Histone deacetylase 8 (HDAC8) is a zinc-binding metalloprotein being involved in transcriptional regulation, cell cycle progression and cellular development. Its overexpression is associated with various pathologies, including childhood neuroblastoma, leukemia, tumor progression and lung cancer. The FDA-approved HDAC inhibitors are not very safe due to various side effects. Therefore, we implemented an in silico high-throughput screening to identify novel HDAC8 inhibitors from the natural compounds database ($n=220,000$) using computational methods. As a result, some drug-like substances ($n=9$) with nontoxic profiles were identified as hit compounds after using molecular docking, including quantum polarized ligand and induced-fit docking approaches. One hit compound (ZINC02106564) was determined to be highly chemically reactive by using the DFT calculation. Finally, the top binder (ZINC12601598) to HDAC8 together with the reference inhibitor, was studied by 100ns molecular dynamics simulations, confirming previously obtained data. Overall, the proposed computational protocol might be prospective at the early stage of rational design for novel and less toxic HDAC8 inhibitors for the treatment of diverse diseases.

Buruli ulcer (BU), a severe skin disease is caused by Mycobacterium ulcerans. There are concerns of therapeutic inefficacy of existing drugs coupled with chemoresistance. Databases have been shown to augment data mining and integrative systems pharmacology approaches towards the search for novel therapeutic moieties. So far, there is no known integrated database solely dedicated to BU drug discovery. In this work, Buruli ulcer database (BuDb) is a “one-stop-shop” knowledgebase for supporting BU drug discovery. It contains both manually verified literature and database-curated data on BU. The BuDb provides comprehensive information on the various drug targets, tested compounds, existing drugs, ethnopharmacological plants and information on the genome of M. ulcerans. It also contains cross-referenced links to databases including PubMed, PubChem, DrugBank, NCBI, Gene Ontology (GO), UniProt, Prota4u, String database, KEGG Pathway and KEGG genome database. The BuDb has been implemented with accessibility features such as keyword and specific searches as well as browsing. BuDb is the first useful online repository of its kind integrated with enriched datasets that can aid in the discovery of new biotherapeutic entities for BU. BuDb can be freely accessed at http://197.255.126.13:3000/.

Somatic mutations in the isoenzymes of isocitrate dehydrogenases (IDHs) account for the pathogenesis of various malignancies including gliomas. To date, ivosidenib is the only FDA-approved drug candidate widely used to target mutated isocitrate dehydrogenase 1 (mIDH1). However, the impotence of the existing drug to evade the blood barrier remains an obstacle to harness ivosidenib as therapeutics against glioma. Thus, in this study, we elucidate a fragment-based drug discovery strategy to design novel lead molecules against the mIDH1 protein. Initially, a fragment library was constructed using 27 known mIDH1 inhibitors from the literature. The vast chemical classes of constructed library consisting of 1109 fragments were then used for breeding. A total of 18,000 breed compounds were generated and the resultant compounds were scrutinized based on the breed score ($>12$) and Tanimoto coefficient ($\mathrm{\text{Tc}}\ge 0.5$). The binding affinity and the energetics of the resultant molecules (2069) were investigated using molecular docking and MM-GBSA calculations. Eventually, the compounds with higher affinity were included in the mutational analysis incorporating the second site mutations namely IDH1^S280F and IDH1^R119P. The bioavailability analysis and toxicity profiling were carried out for screened hybrid molecules. The pipeline of the integrated in-silico approach identified hybrid 209, hybrid 237 and hybrid 504 as the drug-like candidates against the mutational variants of mIDH1 protein. Interestingly, all three compounds exhibited greater binding affinity and better brain penetrating capability. The machine learning-based anti-cancerous sensitivity prediction tool affirmed the inhibitory effect of the resultant hits against various glioma cell lines. In the end, the structural stability of the screened molecules was examined using the molecular dynamic simulation study for a stipulated time of 100ns. Indeed, this evidence speculates that the identified hybrid molecules could serve as important leads for the management of glioma in the near future.

Terminal units’ modification is an effective strategy for designing efficient un-fused nonfullerene acceptors (UF-NFAs) with enhanced power conversion efficiency (PCE). Nowadays, researchers are focused on designing new UF-NFAs that enhance the PCE of organic solar cells. In this line, efforts are being made to design new UF-NFAs for possible application on organic solar cells (OSCs). By doing terminal unit modification of the Cl-4F molecule, we have designed a new series of UF-NFA (ETPJ-1–ETPJ-4). Density functional theory (DFT) and time-dependent density functional theory (TD-DFT) at the B3LYP/6-311G($d,p$) level have been employed for the computation of various geometric and photovoltaic aspects. Energies of highest occupied molecular orbitals (HOMO) and lowest unoccupied molecular orbitals (LUMO) with their band gap suggested that ETPJ-1–ETPJ-4 are effective contributors to the design of the efficient active layer of OSCs. Red-shift (near IR) in the absorption spectrum with easy excitation of exciton has been noted in ETPJ-1–ETPJ-4. Enhanced open circuit voltage with high fill factor percentage (FF%) was also noted for designed systems. Further, the PCE values of the ETPJ-1–ETPJ-4 are better than the reference molecule. So, we recommended a novel kind of unfused nonfullerene acceptors (NFAs) with unique S–O noncovalent interaction for possible application in OSCs.

This study explored a series of reported 5-lipoxygenase-activating protein (FLAP) inhibitors to understand their structural requirements and identify potential new inhibitor scaffolds through automated unbiased procedures. Docking studies have revealed that inhibitor binding affinity can be influenced by several key binding interactions with Phe114 and Lys116 from chain B and Val21, Phe25, His28 and Lys29 from chain C in the FLAP-binding site. A ligand-based alignment three-dimensional (3D)-quantitative structure–activity relationship (QSAR) was adopted, resulting in a robust model with a statistically significant noncross-validated coefficient ($r^2=0.992$), a cross-validated correlation coefficient ($q^2=0.681$) and a predictive squared correlation coefficient ($r_{\mathrm{\text{pred}}}^2=0.736$). Overall, the analysis revealed the important electrostatic and steric attributes responsible for the FLAP inhibitory activity, which appeared to correlate well with the docking results. In addition, two statistically significant two-dimensional (2D)-QSAR models ($r^2=0.9369$, $q^2=0.889$ and $r^2=0.9679$, $q^2=0.655$) were developed by a genetic function approximation (GFA). HypoGen 1, a proposed pharmacophore model, was used for database mining to identify potential new FLAP inhibitors. The bioactivity of the retrieved hits was then evaluated in silico based on the validated QSAR models, followed by pharmacokinetics and toxicity predictions.

Prostate carcinoma is one of the most commonly diagnosed visceral malignancies and the fifth leading cause of cancer-related mortality in males. Reportedly, a series of dietary lipids are identified as 1-cis-4-cis-pentadiene polyunsaturated fatty acids (PUFAs), which play a dominant role in prostate carcinogenesis. Four species of human lipoxygenases (LOXs), a family of nonheme iron-containing enzymes, mediate the deoxygenation of the aforementioned PUFAs. 15-LOX-1 in particular metabolizes the $\omega$-6 lipids and generates certain metabolites (e.g., 13-(S)-hydroxyoctadecaenoic acid) which results in vascular homeostasis, cell proliferation and tissue differentiation in the prostate. Furthermore, in prostate cancer (PCa), the expression of 15-LOX-1 is elevated and positively correlated with the Gleason score of the tumor (an indicator of the disease severity). As membrane receptors, kinases and transcriptional factors are all affected by carcinogenic signals of 15-LOX-1, therapeutic agents that directly inhibit this enzyme can be advantageous in the treatment of PCa. To our knowledge, there are limited effective treatments for PCa, and there is no therapy for its metastatic condition. In this respect, 15-LOX-1, as an appropriate candidate for drug development, was subjected to homology modeling, phylogenic assessment, cross-docking analysis and molecular dynamics (MD) simulation to identify an eligible inhibiting agent amongst a library of 30 potential targeting compounds for PCa management.

Bone morphogenetic proteins (BMPs) are multi-functional growth factors that initiate, promote and maintain cartilage and bone morphogenesis, differentiation and regeneration in both the developing embryo and adult. The proteins have a conformational wrist epitope and a linear knuckle epitope responsible for, respectively, type-I and type-II receptor binding, as well as a hybrid armpit epitope targeted by natural BMP antagonists. In this study, the recognition and interaction between human BMPs and their pan-antagonist Crossveinless was investigated systematically at molecular level. It is revealed that the armpit epitope shares a roughly common region over different BMPs, which consists of a loop segment and a turn segment that are sequentially discontinuous but spatially vicinal on these BMP protein surfaces. Turn segment is the primary binding site that can be bound effectively by Crossveinless using a tightly packed mode. The segment was further extended at its two termini to cover a complete double-stranded sheet of BMPs, which was then split from the interfacial context of BMP–Crossveinless complexes to derive a series of osteogenic peptides; they exhibit moderate intrinsic disorder in free state, but can be constrained into a native-like conformation by stapling a disulfide bridge across two strands of the sheet. The disulfide bridge was rationally designed and optimized to avoid disrupting the native interaction of BMP sheet peptides with the active pocket of Crossveinless. Biophysical assays substantiated that the binding affinities of resulting cyclic peptides were improved by 2–6-fold relative to their linear counterpart upon the stapling, in which the cyclic peptide Bmp7-sb1 (S[CLYFDDNSNVILC]K) derived from the double-stranded sheet region of BMP7 armpit epitope was determined to have the highest affinity to Crossveinless in all tested samples. These rationally designed epitope-derived peptides can be used as osteogenic agents to activate the human BMP signaling by competitively targeting their natural antagonist.