A novel metabolic subtype with S100A7 high expression represents poor prognosis and immuno-suppressive tumor microenvironment in bladder cancer.

BACKGROUND

Bladder cancer (BLCA) represents a highly heterogeneous disease characterized by distinct histological, molecular, and clinical features, whose tumorigenesis and progression require aberrant metabolic reprogramming of tumor cells. However, current studies have not expounded systematically and comprehensively on the metabolic heterogeneity of BLCA.

METHODS

The UCSC XENA portal was searched to obtain the expression profiles and clinical annotations of BLCA patients in the TCGA cohort. A total of 1,640 metabolic-related genes were downloaded from the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. Then, consensus clustering was performed to divide the BLCA patients into two metabolic subtypes according to the expression of metabolic-related genes. Kaplan-Meier analysis was used to measure the prognostic values of the metabolic subtypes. Subsequently, comparing the immune-related characteristics between the two metabolic subtypes to describe the immunological difference. Then, the Scissor algorithm was applied to link the metabolic phenotypes and single-cell transcriptome datasets to determine the biomarkers associated with metabolic subtypes and prognosis. Finally, the clinical cohort included 63 BLCA and 16 para-cancerous samples was used to validate the prognostic value and immunological correlation of the biomarker.

RESULTS

BLCA patients were classified into two heterogeneous metabolic-related subtypes (MRSs) with distinct features: MRS1, the subtype with no active metabolic characteristics but an immune infiltration microenvironment; and MRS2, the lipogenic subtype with upregulated lipid metabolism. These two subtypes had distinct prognoses, molecular subtypes distributions, and activations of therapy-related pathways. MRS1 BLCAs preferred to be immuno-suppressive and up-regulated immune checkpoints expression, suggesting the well-therapeutic response of MRS1 patients to immunotherapy. Based on the Scissor algorithm, we found that S100A7 both specifically up-regulated in the MRS1 phenotype and MRS1-tumor cells, and positively correlated with immunological characteristics. In addition, in the clinical cohort included 63 BLCA and 16 para-cancerous samples, S100A7 was obviously associated with poor prognosis and enhanced PD-L1 expression.

CONCLUSIONS

The metabolic subtype with S100A7 high expression recognizes the immuno-suppressive tumor microenvironment and predicts well therapeutic response of immunotherapy in BLCA. The study provides new insights into the prognostic and therapeutic value of metabolic heterogeneity in BLCA.