Transfer learning algorithm assisted in the discovery of novel gp130 inhibitors and their application in colorectal cancer treatment

The gp130 protein is the common receptor for IL-6 family cytokines, and its aberrant activation significantly promotes tumor progression. However, research on inhibitors targeting gp130 remains limited. Here, we employed an artificial intelligence (AI)-assisted drug design strategy and identified evodiamine as a candidate gp130-targeting scaffold. Based on this, we developed a series of novel indolopyridine-based small molecules. The representative compound 8a effectively downregulates the JAK2/STAT3 signaling pathway by directly targeting the gp130 protein (K_D = 2.17 µM), disrupting the binding of STAT3 to DNA, and thereby inducing apoptosis. In HT-29 xenograft tumor models, 8a demonstrated 56.20% tumor growth inhibition (20 mg/Kg). In summary, the compound 8a can be used as a gp130-targeted compound for the treatment of colorectal cancer. Furthermore, the method applied in the research can provide a feasible case for the development of drugs targeting related pathways.