Background:Mitochondrial fission and fusion play important roles in tumorigenesis,
progression and therapy. Dysregulation of these processes may lead to tumor progression,
and regulation of these processes may provide novel strategies for cancer therapy. The involvement
of genes related to mitochondrial fission and fusion (MD) in gastric cancer (GC) remains
poorly understood.Objective:The aim of this study was to establish an MD gene signature for GC patients and to
investigate its association with prognosis, tumor microenvironment and treatment response in
GC.Methods:We use the TCGA-GC database as the cohort, focusing specifically on genes associated
with MD. We conducted identification and consistency clustering analysis of differentially
expressed genes in MD, conducted MD gene mutation and copy number variation analysis,
as well as correlation and functional enrichment analysis between MD gene cluster classification
and immune infiltration. TCGA-GC and GSE15459 were used to construct training and
validation cohorts for the model. We used various statistical methods, including Cox and Lasso
regression, to develop the model. We validated the model using bulk transcriptome and single-
cell transcriptome datasets (GSE13861, GSE26901, GSE66229, and GSE13450). We
used GSEA enrichment, CIBERSORT algorithm, ESTIMATE, and TIDE to gain insight into
the annotation of MD signature and the characterization of the tumor microenvironment. OncoPredict
was used to analyze the relationship between the PRG signature and the drug sensitivity.
We validated the expression of several key genes in MD signature on GC cell lines using
quantitative real-time PCR (qRT-PCR).Conclusion:The MD signature has the capacity to significantly contribute to the prediction of
personalized outcomes and the advancement of novel therapeutic strategies tailored for GC patients.Results:These MDs-related subtypes exhibited different prognosis and immune filtration patterns.
A five-gene signature, comprising AGT, HCFC1, KIFC3, NOX4, and RIN1, was developed.
There was a clear distinction in overall survival between low- and high-risk patients.
The analyses showed further confirmation of the independent prognostic value of the gene signature.
There was a notable correlation between the MD signature, immune infiltration and
drug susceptibility. The expression levels of AGT, HCFC1, KIFC3, NOX4 and RIN1 mRNA
were all increased in these GC cells.