Identification of potential Akt activators: A ligand and structure-based computational approach (preprint)

The Akt pathway plays an important role in cell metabolism, growth, proliferation, and survival. Akt is the central protein whose phosphorylation controls many downstream pathways. In many diseases like Alzheimer’s, Parkinson, and diabetes, there is downregulation of the Akt pathway. It is proven that the binding of small molecules to the PH domain of Akt facilitates its phosphorylation in the cytoplasm. In the current study, to identify Akt activators, ligand-based approaches like fingerprint-based 2D-QSAR, shape, and pharmacophore-based screening were used, followed by structure-based approaches like docking, MM-GBSA, ADME prediction, and MD simulation. Using the 2D-QSAR activity of the Asinex gold platinum database was predicted, and the top twenty-five molecules found to be active using most models were selected for shape-based and pharmacophore-based screening. Later docking was performed using the PH domain of Akt1 (PDB: 1UNQ), and 197105, 261126, 253878, 256085, and 123435 were selected based on docking score and interaction with Lys 14, Arg 23, Arg 25, Asn 53, and Arg 86. The selected molecules were druggable and formed a stable protein-ligand complex. MD simulations of 261126 and 123435 showed better stability and interaction with key residues. To further investigate the SAR of 261126 and 123435, derivates were downloaded from PubChem, and structure-based approaches were employed. The MD simulation of derivates 12289533, 12785801, 83824832, 102479045, and 6972939 was performed in which 83824832 and 12289533 showed interaction with key residues for a longer duration of time. Therefore, 83824832 and 12289533 may act as Akt activators, and further in-vitro and in-vivo experiments must be performed to support the study.

Immunoinformatics approach for a novel multi-epitope vaccine construct against spike protein of human coronaviruses (preprint)

Spike (S) proteins are an attractive target as it mediates the binding of the SARS-CoV-2 to the host through ACE-2 receptors. We hypothesize that the screening of S protein sequences of all the HCoVs would result in the identification of potential multi-epitope vaccine candidates capable of conferring immunity against various HCoVs. In the present study, several machine learning-based in-silico tools were employed to design a broad-spectrum multi-epitope vaccine candidate against S protein of human coronaviruses. To the best of our knowledge, it is one of the first study, where multiple B-cell epitopes and T-cell epitopes (CTL and HTL) were predicted from the S protein sequences of all seven known HCoVs and linked together with an adjuvant to construct a potential broad-spectrum vaccine candidate. Secondary and tertiary structures were predicted, validated and the refined 3D-model was docked with an immune receptor. The vaccine candidate was evaluated for antigenicity, allergenicity, solubility, and its ability to achieve high-level expression in bacterial hosts. Finally, the immune simulation was carried out to evaluate the immune response after three vaccine doses. The designed vaccine is antigenic (with or without the adjuvant), non-allergenic, binds well with TLR-3 receptor and might elicit a diverse and strong immune response.