Deep learning model enables the discovery of a novel BET inhibitor YD-851

Hongyin Sun1, Guoli Xiong2, Xin Li3

  • 1School of Pharmaceutical Sciences, Southern Medical University, Guangzhou, Guangdong 510080, China; Affiliated Fengxian Hospital, Southern Medical University, Fengxian, Shanghai 201400, China.

Insights

A novel Bromodomain and Extra-Terminal (BET) inhibitor, YD-851, effectively targets solid tumors with low toxicity. This breakthrough offers a promising new strategy for cancer therapy, overcoming limitations of earlier BET inhibitors.

Area of Science:

  • Oncology
  • Medicinal Chemistry
  • Epigenetics

Background:

  • Bromodomain and Extra-Terminal (BET) inhibitors represent a novel epigenetic strategy for tumor therapy.
  • First-generation BET inhibitors have shown limited efficacy and significant toxicity in clinical trials for solid tumors.
  • There is a need for effective, low-toxicity BET inhibitors for solid tumor treatment.

Purpose of the Study:

  • To develop a novel, effective, and low-toxicity BET inhibitor for solid tumor therapy.
  • To identify potent carboline derivatives as BET inhibitors using scaffold hopping and deep learning.
  • To evaluate the preclinical efficacy and safety of a lead compound, YD-851.

Main Methods:

  • Utilized a ring-closure scaffold hopping approach combined with deep learning models for inhibitor design.
  • Synthesized a series of rationally designed carboline derivatives.
  • Evaluated the efficacy of YD-851 in inhibiting tumor cell proliferation and suppressing tumor growth in xenograft models.
  • Assessed toxicity and pharmacokinetic properties of YD-851.

Main Results:

  • Identified YD-851 as a potent BET inhibitor.
  • YD-851 effectively inhibited tumor cell proliferation in vitro.
  • YD-851 demonstrated significant tumor shrinkage and growth suppression in multiple solid tumor xenograft models.
  • YD-851 exhibited favorable toxicity and pharmacokinetic profiles, supporting further development.

Conclusions:

  • YD-851 is a promising preclinical candidate for BET inhibitor therapy in solid tumors.
  • The developed strategy of scaffold hopping and deep learning is effective for discovering novel drug candidates.
  • This approach holds potential for drug discovery targeting other therapeutic areas.