H. pylori Outer Membrane Protein Antibody Response, MHC/HLA-DMA Methylation, and Gastric Cancer Risk: A Multi-Layer Evidence Integration Study Using Mendelian Randomization.

Background

Helicobacter pylori (H. pylori) infection is a well-established risk factor for gastric cancer; however, evidence on the associations between antigen-specific antibody traits, host methylation, and gastric cancer risk remains limited.

Methods

Using a forward two-sample Mendelian randomization (MR) design, we integrated methylation quantitative trait loci (mQTLs), external cohorts, The Cancer Genome Atlas-Stomach Adenocarcinoma (TCGA-STAD), and functional annotations through a three-stage framework (antibody trait screening, CpG prioritization, and external replication) to evaluate a two-step pathway from host antibody response to gastric cancer via DNA methylation.

Results

Among six H. pylori antigens, antibody response to outer membrane protein (OMP) showed the strongest positive association with gastric cancer risk (OR = 1.186, 95% CI: 1.079-1.304, raw p = 4.08 × 10-4). CpG-level prioritization identified cg24290947 (annotated to HLA-DMA) as a methylation site linked to gastric cancer susceptibility, with an inverse MR association (beta = -0.2347, 95% CI: -0.3980 to -0.0713, raw p = 0.00487). External exploration showed that cg24290947 methylation was elevated in H. pylori-positive gastric mucosa; however, its direct tumor-normal difference in a 12-pair cohort was not significant, although regional methylation perturbations were observed. HLA-DMA and MHC-II expression were consistently higher in tumor versus normal tissues across three cohorts, whereas TCGA-STAD did not support a stable linear relationship between cg24290947 methylation and HLA-DMA/MHC-II expression; three of six composition-adjusted models showed nominal significance but were model-dependent. These signals were enriched in antigen processing and presentation pathways.

Conclusions

cg24290947/HLA-DMA represents a hypothesis-generating regional methylation candidate for further investigation rather than a confirmed linear transcriptional regulator. The MR evidence is based on European-ancestry summary data; its generalizability and any clinical predictive value warrant independent evaluation.