Single-cell and pseudobulk analyses reveal hidden mitochondrial expression imbalance in gastric cancer

Background Mitochondrial function is essential for biology, particularly in cancer. However, cell-type-specific expression patterns of conserved mitochondrial genes in gastric cancer (GC) remain unclear. We herein raised and tested a novel hypothesis of “mitochondrial conserved gene expression homeostasis imbalance” in GC cohorts with single-cell resolution. Methods This work analyzed an open-accessed scRNA-seq dataset (GSE206785, 24 GC patients, 48 samples) with Seurat and defined 43 clusters grouped into 15 cell subtypes. In parallel, Pseudobulk profiles were generated to simulate bulk RNA-seq. A mitochondrial conserved gene score was computed by Seurat AddModuleScore, GSVA, and AUCell. Mitochondria-related biomarkers were also screened, validated, and incorporated into a mitochondria-dependent prognostic model that was further evaluated. Results Without considering cell-type-specific expression patterns, Pseudobulk analysis showed no significant differences in mitochondrial conserved gene expression between GC and control samples. In contracst, single-cell analysis found a cell-type-specific imbalance, under which tumor-associated epithelial cells displayed relatively elevated mitochondrial conserved gene expression, while non-epithelial cells showed reduced. Notably, survival analyses identified gene KRT7 and KLRC1 as robust prognostic biomarkers for early GC. Conclusion Our findings support a mitochondrial conserved gene expression homeostasis imbalance in GC, which is characterized by compartment-specific mtGene expression imbalance. Also, KRT7 and KLRC1 emerge as prognostic markers for therapies aimed at restoring mitochondrial homeostasis in GC.