Machine learning-aided discovery of T790M-mutant EGFR inhibitor CDDO-Me effectively suppresses non-small cell lung

Rui Zhou1, Ziqian Liu1, Tongtong Wu2

  • 1International Research Centre for Food and Health, College of Food Science and Technology, Shanghai Ocean University, Shanghai, 201306, China.

Abstract

Insights

Machine learning identified CDDO-Me as a potent inhibitor for EGFR T790M-mutant non-small cell lung cancer. This compound induces apoptosis and suppresses tumor growth, offering a promising strategy against drug resistance.

Area of Science:

  • Oncology
  • Pharmacology
  • Computational Biology

Background:

  • Epidermal growth factor receptor (EGFR) T790M mutations confer resistance to erlotinib in non-small cell lung cancer (NSCLC).
  • Traditional screening methods struggle to identify novel, selective EGFR-T790M inhibitors.
  • Machine learning (ML) accelerates the identification of biologically active molecules with improved targeting efficiency.

Purpose of the Study:

  • To leverage ML for identifying selective inhibitors of EGFR T790M mutations.
  • To validate the efficacy of identified compounds against NSCLC harboring EGFR T790M.
  • To explore novel therapeutic strategies for overcoming erlotinib resistance in NSCLC.

Main Methods:

  • An integrated ML approach employing Bayesian inference screened 70,413 molecules.
  • In vitro cellular assays, molecular dynamic simulations, and EGFR knockout cell lines validated candidate binding and activity.
  • An in vivo xenograft mouse model assessed the antitumor efficacy of CDDO-Me.

Main Results:

  • CDDO-Me, an oleanolic acid derivative, was identified as a potent inhibitor of NSCLC cells with EGFR T790M mutations.
  • Selective binding of CDDO-Me to T790M-mutant EGFR was confirmed via cellular thermal shift assay and molecular dynamics simulations.
  • CDDO-Me induced apoptosis and cell cycle arrest by inhibiting the PI3K-Akt-mTOR pathway, suppressing tumor growth in a xenograft model.

Conclusions:

  • CDDO-Me effectively targets EGFR T790M, inducing apoptosis and cell cycle arrest via the PI3K-Akt-mTOR pathway.
  • In vivo studies demonstrated CDDO-Me's dose-dependent tumor growth suppression in a NSCLC xenograft model.
  • ML-aided drug screening offers a promising avenue for developing novel compounds to overcome NSCLC drug resistance.