Identification of Zinc-Binding Inhibitors of Matrix Metalloproteinase-9 to Prevent Cancer Through Deep Learning and Molecular Dynamics Simulation Approach

Shalini Mathpal; Priyanka Sharma; Tushar Joshi; Veena Pande; Shafi Mahmud; Mi-Kyung Jeong; Ahmad J Obaidullah; Subhash Chandra; Bonglee Kim

doi:10.3389/fmolb.2022.857430

Identification of Zinc-Binding Inhibitors of Matrix Metalloproteinase-9 to Prevent Cancer Through Deep Learning and Molecular Dynamics Simulation Approach

Front Mol Biosci. 2022 Mar 31:9:857430. doi: 10.3389/fmolb.2022.857430. eCollection 2022.

Authors

Shalini Mathpal¹, Priyanka Sharma², Tushar Joshi¹, Veena Pande¹, Shafi Mahmud^{3

4}, Mi-Kyung Jeong⁵, Ahmad J Obaidullah⁶, Subhash Chandra⁷, Bonglee Kim^{5

8}

Affiliations

¹ Department of Biotechnology, Kumaun University Uttarakhand, Bhimtal, India.
² Department of Botany, DSB, Campus, Kumaun University, Nainital, India.
³ Genetic Engineering and Biotechnology, University of Rajshahi, Rajshahi, Bangladesh.
⁴ Department of Genome Science, John Curtin School of Medical Research, Australian National University, Canberra, Australia.
⁵ KM Convergence Research Division, Korea Institute of Oriental Medicine, Daejeon, South Korea.
⁶ Drug Exploration and Development Chair (DEDC), Department of Pharmaceutical Chemistry, College of Pharmacy, King Saud University, Riyadh, Saudi Arabia.
⁷ Computational Biology and Biotechnology Laboratory, Department of Botany, Soban Singh Jeena University, Almora, India.
⁸ Department of Pathology, College of Korean Medicine, Kyung Hee University, Seoul, South Korea.

Abstract

The overexpression of matrix metalloproteinase-9 (MMP-9) is associated with tumor development and angiogenesis, and hence, it has been considered an attractive drug target for anticancer therapy. To assist in drug design endeavors for MMP-9 targets, an in silico study was presented to investigate whether our compounds inhibit MMP-9 by binding to the catalytic domain, similar to their inhibitor or not. For that, in the initial stage, a deep-learning algorithm was used for the predictive modeling of the CHEMBL321 dataset of MMP-9 inhibitors. Several regression models were built and evaluated based on R2, MAE MSE, RMSE, and Loss. The best model was utilized to screen the drug bank database containing 9,102 compounds to seek novel compounds as MMP-9 inhibitors. Then top high score compounds were selected for molecular docking based on the comparison between the score of the reference molecule. Furthermore, molecules having the highest docking scores were selected, and interaction mechanisms with respect to S1 pocket and catalytic zinc ion of these compounds were also discussed. Those compounds, involving binding to the catalytic zinc ion and the S1 pocket of MMP-9, were considered preferentially for molecular dynamics studies (100 ns) and an MM-PBSA (last 30 ns) analysis. Based on the results, we proposed several novel compounds as potential candidates for MMP-9 inhibition and investigated their binding properties with MMP-9. The findings suggested that these compounds may be useful in the design and development of MMP-9 inhibitors in the future.

Keywords: MD simulation; MMP-9; cancer; deep learning; drug bank compounds.