Repurposing of the Syk inhibitor fostamatinib using a machine learning algorithm

Area of Science:

• Oncology
• Pharmacology
• Artificial Intelligence in Drug Discovery

Background:

• TAM (TYRO3, AXL, MERTK) receptor tyrosine kinases (RTKs) are implicated in tumor growth, metastasis, and immune evasion.
• Overexpression of TAM RTKs correlates with poor cancer prognosis.
• Targeting individual TAM RTKs shows limited efficacy due to compensatory feedback mechanisms.

Purpose of the Study:

• To identify existing drugs capable of inhibiting all three TAM RTKs using a deep-learning DTI model.
• To evaluate the potential of drug repurposing for pan-TAM inhibition in cancer therapy.

Main Methods:

• Utilized a deep-learning drug-target interaction (DTI) model, molecule transformer-DTI (MT-DTI), for virtual screening.
• Identified potential inhibitors from commercially available drugs targeting TAM RTKs.
• Validated predictions through in vitro studies using cancer cell lines.

Main Results:

• The MT-DTI model predicted fostamatinib, a spleen tyrosine kinase (Syk) inhibitor, as a potent inhibitor of all three TAM RTKs (IC50 <1 µM).
• In vitro experiments confirmed fostamatinib's ability to suppress cancer cell proliferation by inhibiting TAM RTKs.
• Other Syk inhibitors did not demonstrate similar pan-TAM inhibitory activity.

Conclusions:

• Fostamatinib exhibits anticancer activity as a novel pan-TAM inhibitor.
• Artificial intelligence models like MT-DTI are effective for drug repurposing and identifying new therapeutic mechanisms.
• Fostamatinib shows potential for expanded indications in cancer treatment targeting TAM RTKs.