Identification of succinylation-related genes in bladder cancer: integration of single-cell and transcriptomic data

Background Bladder cancer (BLCA) exhibits a poor prognosis, highlighting the urgent need for reliable prognostic genes. Although succinylation is linked to tumor progression, its role in BLCA remains understudied. This study aimed to identify and validate prognostic succinylation-related genes (SRGs) in BLCA and elucidate their impact on the tumor microenvironment (TME). Methods SRGs were initially identified through transcriptomic sequencing of 15 paired BLCA and adjacent normal tissues (Soochow-BLCA cohort) and further refined by integrating single-cell and bulk transcriptomic data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) datasets. A prognostic risk model was developed using LASSO and multivariable Cox regression, incorporating clinical factors for nomogram construction. Mechanistic insights were obtained through functional enrichment, immune profiling, somatic mutation, and drug sensitivity analyses. The Scissor algorithm mapped bulk transcriptome risk signatures to single-cell resolution, enabling identification of high-risk cell subpopulations. Pseudotime and cell-cell communication analyses characterized dynamic expression patterns of the core genes. Finally, the expression profiles and functional roles of the core genes were validated using RT-qPCR and CCK-8 proliferation assays in vitro . Results KCTD16, CD3D and GSDMB were identified as prognostic genes. The risk score derived from these genes, in combination with age and N stage, was incorporated into a risk model that exhibited robust predictive accuracy (AUC 0.7). Expression of these genes differed significantly between non-muscle-invasive and muscle-invasive subtypes. The high-risk group displayed enhanced immune evasion (higher TIDE score, p 0.001), while the TME appeared to be in an “inflamed yet dysfunctional” state. Single-cell analysis indicated epithelial cells as key subpopulations, with additional involvement of T cells and fibroblasts. Scissor + cells were correlated with the high-risk phenotype and exhibited pseudotime-dependent expression patterns. RT-qPCR revealed significant upregulation of GSDMB and downregulation of KCTD16 and CD3D in BLCA ( p 0.05). Functionally, knockdown of GSDMB and KCTD16 significantly promoted T24 cell proliferation, supporting their tumor-suppressive roles. Conclusions The succinylation-related prognostic model accurately predicts outcomes in BLCA, reveals links to immune escape and TME, and highlights the pivotal role of epithelial cells, offering potential targets for individualized therapy.