Machine learning and biological evaluation-based identification of a potential MMP-9 inhibitor, effective against ovarian cancer cells SKOV3

Abstract

MMP-9, also known as gelatinase B, is a zinc-metalloproteinase family protein that plays a key role in the degradation of the extracellular matrix (ECM). The normal function of MMP-9 includes the breakdown of ECM, a process that aids in normal physiological processes such as embryonic development, angiogenesis, etc. Interruptions in these processes due to the over-expression or downregulation of MMP-9 are reported to cause some pathological conditions like neurodegenerative diseases and cancer. In the present study, an integrated approach for ML-based virtual screening of the Maybridge library was carried out and their biological activity was tested in an attempt to identify novel small molecule scaffolds that can inhibit the activity of MMP-9. The top hits were identified and selected for target-based activity against MMP-9 protein using the kit (Biovision K844). Further, MTT assay was performed in various cancer cell lines such as breast (MCF-7, MDA-MB-231), colorectal (HCT119, DL-D-1), cervical (HeLa), lung (A549) and ovarian cancer (SKOV3). Interestingly, one compound viz., RJF02215 exhibited anti-cancer activity selectively in SKOV3. Wound healing assay and colony formation assay performed on SKOV3 cell line in the presence of RJF02215 confirmed that the compound had a significant inhibitory effect on this cell line. Thus, we have identified a novel molecule that can inhibit MMP-9 activity in vitro and inhibits the proliferation of SKOV3 cells. Novel molecules based on the structure of RJF02215 may become a good value addition for the treatment of ovarian cancer by exhibiting selective MMP-9 activity.

Communicated by Ramaswamy H. Sarma

Work flow for ML-based identification of hit molecules against MMP-9 protein.

Keywords:

• Machine learning
• matrix metalloproteinase 9 (MMP-9)
• ovarian cancer
• support vector machine (SVM)
• molecular docking
• molecular dynamic simulation
• MM/PBSA

Acknowledgments

KS and SP acknowledge CSIR, Govt. of India, for Senior Research Fellowship. SMV acknowledges CSIR, Govt. of India, for Junior Research Fellowship. AY acknowledges DBT, Govt. of India for Junior Research Fellowship. Financial support from CSIR-Pan-Cancer Grant HCP-0040, SERB Grant (EMR/2017/002080) and DBT-Grant GAP0384 (BT/PR40131/BTIS/137/26/2021) are acknowledged. We also acknowledge CSIR-CDRI, Govt. of India for financial and infrastructural support. We acknowledge CSIR-CDRI chemical repository for providing Maybridge library compounds for biological assays. The Institutional (CSIR-CDRI) communication number for this article is 10634.

Authors’ contributions

All authors confirm contribution to the paper as follows: KS and SP designed parts of experiments and performed the relevant experiments. KS and SP wrote the original draft. KS and SP contributed equally. SMV and AY helped in performing the experiments and acquisition of data. MIS and DB conceptualised the experiments, made supervision, data curtain, review and editing. MIS and DB reviewed and edited the manuscript.

Disclosure statement

The authors declare that they have no competing interests.

Additional information

Funding

This work was supported by Council of Scientific and Industrial Research (CSIR)-Pan-Cancer Grant HCP-0040 and Science and Engineering Research Board (SERB) grant (EMR/2017/002080). We deeply acknowledge CSIR-CDRI, Govt. of India for financial and infrastructural support.