Advanced gastric cancer is frequently refractory to therapy because of multidrug resistance and apoptotic escape. BCL2 and SIVA1 are core apoptosis regulators, and their direct interaction suggests a potentially druggable BCL2–SIVA1 signaling axis that remains insufficiently explored in advanced gastric cancer. We integrated bulk transcriptomics, single-cell RNA sequencing, spatial transcriptomics, and immune infiltration analyses from GEO/TCGA to characterize the BCL2–SIVA1 axis and applied a bioinformatics-guided drug-repurposing workflow (2D-QSAR machine learning, molecular docking, molecular dynamics, and MM/GBSA) followed by in vitro validation and mechanistic rescue experiments in NCI-N87 and HGC-27 cells. Multiomic analyses identified BCL2 as a progression-associated factor and revealed a strong positive correlation between BCL2 and SIVA1. Single-cell analysis showed that BCL2 was enriched and heterogeneous in malignant cells; pseudotime analysis placed BCL2-high states toward terminal branches, and hallmark scoring indicated increased proliferative and metastatic potential. Spatial transcriptomics demonstrated localized enrichment of BCL2 and SIVA1 in advanced gastric cancer tissues, and coimmunoprecipitation confirmed a direct BCL2–SIVA1 interaction, supporting a functional signaling axis. In silico screening prioritized dihydroergotamine as a high-affinity BCL2 binder, and dihydroergotamine suppressed proliferation and promoted apoptosis in both cell lines, accompanied by downregulation of BCL2/SIVA1 and upregulation of BAX and cleaved caspase 3. Rescue experiments further revealed that BCL2 overexpression during dihydroergotamine exposure partially restored SIVA1 protein levels and attenuated apoptosis, whereas SIVA1 overexpression did not restore BCL2 expression, supporting BCL2-linked coupling within this axis. By combining multiomic analyses with structure-based screening and in vitro validation, we identified the BCL2–SIVA1 axis as a potential apoptosis-related vulnerability in advanced gastric cancer and highlighted dihydroergotamine as a candidate modulator. Dihydroergotamine showed in vitro antiproliferative and proapoptotic effects consistent with the involvement of BCL2-associated apoptotic signaling, providing a preclinical rationale for further target dependency and translational feasibility studies.
Huang et al. (2026) studied this question.