Hybrid identity and distinct methylation profiles of incomplete intestinal metaplasia in the stomach

Abstract

Background Gastric intestinal metaplasia (GIM), particularly the incomplete subtype (Inc IM), is strongly associated with increased gastric cancer (GC) risk. However, its role as a true precursor lesion remains uncertain. Objective We aimed to delineate the molecular identity, differentiation potential and oncogenic relevance of Inc IM. Methods Spatial transcriptomics using a custom lineage-enriched panel was applied to profile GIM and GC tissues. Subtype-specific GIM organoid models were developed for DNA methylation and chromatin accessibility profiling. Single-cell RNA sequencing was performed to evaluate differentiation capacity. Results Spatial transcriptomics revealed that Inc IM potentially originates from the deep antral gland cells and harbours a hybrid transcriptomic signature incorporating gastric, small intestinal and large intestinal lineages across both differentiated and stem/progenitor compartments. DNA methylation profiling of subtype-specific organoids showed that Inc IM exhibits extensive intergenic hypermethylation, resembling native antral mucosa. In contrast, complete subtype was marked by promoter hypermethylation of tumour suppressor genes and displayed a more fully intestinalised epigenetic profile. Organoid models recapitulated subtype-specific traits and demonstrated lineage plasticity. Spatial mapping of GC samples revealed an enrichment of Inc IM-like cells, particularly within microsatellite stable tumours. Approximately 76% of the GCs analysed were linked to GIM, while the remaining (24%) appeared to be associated with deep antral differentiation. Conclusions Inc IM represents a phenotypically unstable and epigenetically deregulated metaplastic state with dual-lineage potential and molecular resemblance to GC. These findings establish Inc IM as a true precursor to GC and underscore the importance of active surveillance and early intervention strategies. Data are available on reasonable request. The raw and processed sequencing data generated in this study have been deposited in the Gene Expression Omnibus (GEO) under accession numbers GSE294729, GSE295640 and GSE295401. This includes scRNA-seq, Xenium, bulk ATAC-seq and DNA methylation microarray data.