Identification of m6A modification patterns and development of m6A-hypoxia prognostic signature to characterize tumor microenvironment in triple-negative breast cancer

BackgroundN6-methylation (m6A) modification of RNA has been found to have essential effects on aspects of the tumor microenvironment (TME) including hypoxia status and mobilization of immune cells. However, there are no studies to explore the combined effect of m6A modification and hypoxia on molecular heterogeneity and TME of triple-negative breast cancer (TNBC). MethodsWe collected The Cancer Genome Atlas (TCGA-TNBC, N=139), the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC-TNBC, N=297), the GSE103091, GSE21653, and GSE135565 series from the Gene Expression Omnibus (GEO-TNBC, N=247), and FUSCCTNBC (N=245) for our study. The non-negative matrix factorization algorithm was used to cluster TNBC samples. Immune cell infiltration was analyzed by the CIBERSORT algorithm. The enrichment scores were calculated by single-sample gene set enrichment analysis(ssGSEA) to characterize TME in TNBC samples. Immunohistochemistry (IHC) and qRT-PCR were performed to detect the gene expression. ResultsBased on the expression of m6A-related genes, we identified three distinct m6A clusters (denoted A, B, and C) in TNBC samples. Comparing the TME characteristics among the three clusters, we observed that cluster C was strongly related to hypoxia status and immune suppression, whereas clusters A and B displayed more immune cell infiltration. Therefore, we combine m6A and hypoxia related genes to classify two m6A-hypoxia clusters of TNBC and screened six prognostic genes by LASSO-Cox regression to construct a m6A-hypoxia signature(MHPS), which divided TNBC samples into high- and low-risk groups. We identified different TME features, immune cell infiltration between the two groups, and a better immunotherapy response was observed in the low-risk group. A nomogram was constructed with tumor size, lymph node, and risk score to improve clinical application of MHPS. ConclusionWe identified distinct TME characteristics of TNBC based on three different m6A modification patterns. Then, we constructed a specific m6A-hypoxia signature for TNBC to evaluate risk and predict immunotherapy response of patients, to enable more accurate treatment in the future.