Related Experiment Video
Updated: Aug 5, 2026
Multiomics Analysis of TMEM200A as a Pan-Cancer Biomarker
Published on: September 15, 2023
From proteome-wide Mendelian randomization and multi-omics integration to functional validation: TGFB3 as a
Luming Zhao1, Chenxi Mao1, Yimeng Xu2
1Department of Oncology, The Affiliated Lianyungang Hospital of Xuzhou Medical University/The First People's Hospital of Lianyungang, Lianyungang, China.
Introduction:
Gastric adenocarcinoma lacks robust circulating biomarkers and tractable targets. We assessed whether genetically predicted plasma protein levels influence disease risk and sought druggable candidates using a proteome-wide Mendelian randomization (MR) framework.
Methods:
We integrated deCODE plasma protein quantitative trait loci (pQTLs) with gastric cancer GWAS in a proteome-wide two-sample MR framework to identify circulating proteins causally associated with gastric cancer. Protein-protein interaction topology was used to prioritize eight hub proteins, which were further evaluated in an eight-gene artificial neural network (ANN) classifier. Single-cell and spatial transcriptomics, together with multiplex immunofluorescence, mapped hub-gene expression across tumor, stromal and immune compartments. Finally, we focused on TGFB3 as a genetically and spatially prioritized target, combining in silico ligand screening with in vitro perturbation of the TGFB3-PI3K survival axis using proflavine hemisulfate as a chemical probe.
Results:
Proteome-wide MR highlighted 29 circulating proteins with putative causal effects on gastric cancer, of which eight (ERBB3, LGR4, BMP4, CD248, MGP, TGFB3, GRP and ETS2) occupied central positions in the interaction network. The corresponding eight-gene ANN showed robust discrimination between gastric cancer and non-tumor tissue across multiple cohorts and improved diagnostic performance beyond clinical variables. Single-cell, spatial and multiplex immunofluorescence analyses localized these hubs to epithelial, fibroblast and endothelial compartments, with TGFB3 enriched at tumor-stroma interfaces and associated with poor survival. Virtual screening nominated TGFB3-binding ligands and proflavine hemisulfate formed a stable complex with TGFB3 in silico while attenuating TGFB3-driven proliferation, migration and PI3K-dependent survival signaling in AGS cells.
Conclusion:
Our proteome-wide MR framework identifies genetically supported circulating proteins that contribute to gastric carcinogenesis and converges on TGFB3 as a tractable stromal signaling axis. Proflavine hemisulfate functions as a mechanistically informative chemical probe of TGFB3-PI3K survival signaling in gastric cancer cells, providing a starting scaffold for future TGFB3-targeted therapeutic strategies.