CMTM3 is associated with selected unfolded protein response markers and vascular/stromal features in glioblastoma

Background Previous studies have implicated CKLF-like MARVEL transmembrane domain-containing 3 (CMTM3) in glioma progression and have demonstrated effects on glioma-cell proliferation and invasion. However, its relationship with cellular composition of the glioblastoma (GBM) microenvironment and adaptive endoplasmic reticulum (ER)-stress responses remains incompletely characterized. We therefore investigated the cellular context of CMTM3 expression and its association with selected unfolded protein response (UPR)-related signals in GBM, together with complementary functional validation in GBM cells. Methods GEO, TCGA, and Human Protein Atlas datasets were integrated with single-cell RNA sequencing, survival analysis, protein-interaction and correlation analyses, functional enrichment, and immune-infiltration profiling. CMTM3 was silenced in U87 and U251 cells, followed by migration, colony-formation, CCK-8, and Western blot assays. Endoplasmic reticulum (ER) stress was induced with tunicamycin (2 μg/mL, 24 h). Results In the legacy TCGA glioma cohorts, CMTM3 expression was associated with higher WHO grade, IDH-wild-type status, and older age, whereas higher CMTM3 expression in TCGA-GBM was associated with shorter overall survival. In the vascular-enriched single-cell dataset, CMTM3 expression was detected across multiple populations and was prominent in endothelial and mural/perivascular clusters. CMTM3 knockdown impaired migration, reduced clonogenic capacity, and decreased cellular metabolic activity. GO/KEGG analysis highlighted protein processing in the ER. Under tunicamycin treatment, CMTM3-depleted cells showed lower p-PERK, ATF4, and XBP1s protein levels than control cells, whereas BiP and CHOP did not differ significantly between the treated groups. However, two-way ANOVA revealed no significant CMTM3 knockdown × tunicamycin interaction for any of the five examined proteins. Conclusion Our findings extend previous observations on CMTM3 in glioma by identifying associations with selected UPR-related protein levels under tunicamycin exposure and to vascular and stromal features of the GBM microenvironment. These results provide a basis for further investigation of the cellular and stress-response context of CMTM3 in GBM.