Abstract
Background: Stromal components of the tumor microenvironment contribute to bladder cancer progression, and Cancer-Associated Fibroblasts (CAFs) were reported to play an important role. Accumulating pieces of evidence indicate that CAFs participate in the crosstalk with tumor cells and have a complex interaction network with immune components. Further study of the role of CAFs in the bladder cancer microenvironment and the search for possible specific markers are important for a deeper understanding of the roles of CAFs in bladder cancer progression and immunomodulation.Methods: In the present study, we examined the abundance of CAFs in the TCGA and GEO datasets using the MCP-Counter algorithm. Additionally, the expression of genes related to CAFs was analyzed through the Weighted Gene Co-expression Network Analysis (WGCNA). The CIBERSORT and ESTIMATE algorithms were used for the correlation analysis between the key CAFs related gene and the tumor microenvironment components.Immunohistochemistry analysis in clinical samples was used to validate the results of bioinformatics analysis.Results: The results showed that CAFs were closely associated with the progression and prognosis of bladder cancer. WGCNA also revealed that CALD1 was a key gene significantly associated with CAFs in bladder cancer. Moreover, the further in-depth analysis showed that CALD1 significantly affected the progression and prognosis of bladder cancer. The CIBERSORT and ESTIMATE algorithms significant correlations between CALD1 and the tumor microenvironment components, including CAFs, macrophages, T cells, and multiple immune checkpoint related genes. Finally, immunohistochemistry results of clinical samples' validated the strong association between CALD1, CAFs, and macrophages.Conclusions: In the present study, we confirmed the cancer-promoting roles of CAFs in bladder cancer. Being a key gene associated with CAFs, CALD1 may promote bladder cancer progression by remodeling the tumor microenvironment. The bioinformatics methods, including the CIBERSORT, MCP-COUNTER and ESTIMATE algorithms, may provide important value for studying the tumor microenvironment.