A multi-target approach for the discovery of Anti Breast Cancer Agents from Plants Secondary Metabolites

Background: Cancer is a multifactorial disease with multiple complications involving multiple proteins. Breast cancer is the most prevalent form of cancer among women. The pathophysiology of this cancer form has implicated several genetic alterations in its hallmark. Two of the most studied breast cancer molecular pathways are the cell cycle protein kinases and P13/AKT signalling pathway. Objective: Thus, this study identified novel inhibitors through computational screening of a library of medicinal plants compounds against cyclin-dependent kinase 2 (CDK2), phosphoinositide-3-kinase A (PI3Ka) and protein kinase B (AKT1). Methods: Rigid protein docking via Glide algorithm was applied to identify the hits from 3000 plants compounds screened against three drug targets involved in breast cancer pathogenesis. A more accurate and reliable ligand-protein docking called induced fit docking was adopted to extensively improve the scoring function by ranking favourable binding as top-scoring one. Results: Nine hit compounds were identified and found to interact with essential residues at the proteins’ binding sites. Subsequently, the hits pharmacokinetic parameters and toxicity were predicted to determine their potential as drug candidates and minimise toxic effects. The hit compounds were found to be non-carcinogenic, and five of them showed a desirable drug-like property. The built predictive QSAR models with an R2 value of 0.7684, 0.7973 and 0.5649 for CDK2, AKT1 and PI3Ka, respectively were adopted to determine the hits inhibitory activity (pIC50) against the screened proteins; and the predictions revealed compounds with significant activity. Three thousand (3000) compounds from diverse medicinal plants were docked with CDK2, AKT1 and PI3Ka to identify the top-scoring compounds using Glide algorithm scoring function. The identified compounds with low binding energies against the three targets were subsequently subjected to a more accurate and reliable ligand-protein docking called induced fit docking to extensively improve the compounds binding affinity with the proteins. Nine (9) compounds identified as hits were found to form highly stable complexes with the proteins and interacted with essential residues at the proteins’ binding sites. Prediction of the hit compounds drug-likeness, pharmacokinetic and toxicity properties by online web servers showed that the compounds are non-carcinogenic and showed moderate indices for ADMET parameters. The constructed QSAR models with reliable R2 coefficient value were used to predict the pIC50 of the selected compounds. The results revealed potent compounds with significant activity. Concluson: This study thus provides insight into multi-target protein compounds which could be explored as chemotherapeutic alternatives in breast cancer treatment.