Integrative transcriptional characterization of cell cycle checkpoint genes promotes clinical management and precision medicine in bladder carcinoma

BACKGROUND: The aberrant regulation of cell cycle is significantly correlated with cancer carcinogenesis and progression, in which cell cycle checkpoints control phase transitions, cell cycle entry, progression, and exit. However, the integrative role of cell cycle checkpoint-related genes (CRGs) in bladder carcinoma (BC) remains unknown. METHODS: The transcriptomic data and clinical features of BC patients were downloaded from The Cancer Genome Atlas (TCGA), used to identify CRGs correlated with overall survival (OS) by univariate Cox regression analysis. Then, the multivariate and least absolute shrinkage and selection operator (LASSO) Cox regression analyses further developed a prognostic CRG signature, which was validated in three external datasets retrieved from Gene Expression Omnibus (GEO). The receiver operating characteristic curve (ROC) analysis was conducted for evaluating the performance of the CRG signature in prognosis prediction. RNA sequencing (RNA-Seq) was performed to explore the expression difference in the identified CRGs between tumor and normal tissue samples from 11 BC patients in the local cohort. Ultimately, genomic profiles and tumor microenvironment (TME), and the Genomics of Drug Sensitivity in Cancer (GDSC) were investigated to guide precision treatment for BC patients with different CRG features. RESULTS: The novel constructed 23-CRG prognostic signature could stratify BC patients into high-risk and low-risk groups with significantly different outcomes (median OS: 13.64 vs. 104.65 months). Notably, 19 CRGs were the first to be identified as being associated with BC progression. In three additional validation datasets (GSE13507, GSE31684, and GSE32548), higher CRG scores all indicated inferior survival, demonstrating the robust ability of the CRG signature in prognosis prediction. Moreover, the CRG signature as an independent prognostic factor had a robust and stable risk stratification for BC patients with different histological or clinical features. Then, a CRG signature-based nomogram with a better performance in prognostic prediction [concordance index (C-index): 0.76] was established. Functional enrichment analysis revealed that collagen-containing extracellular matrix (ECM), and ECM-related and MAPK signaling pathways were significantly associated with the signature. Further analysis showed that low-risk patients were characterized by particularly distinctive prevalence of FGFR3 (17.03% vs. 6.67%, p < 0.01) and POLE alterations (7.97% vs. 2.50%, p < 0.05), and enrichment of immune infiltrated cells (including CD8+ T cells, CD4+ naïve T cells, follicular helper T cells, Tregs, and myeloid dendritic cells). RNA-seq data in our local cohort supported the findings in the differentially expressed genes (DEGs) between tumor and normal tissue samples, and the difference in TME between high-risk and low-risk groups. Additionally, CRG signature score plus FGFR3 status divided BC patients into four molecular subtypes, with distinct prognosis, TME, and transcriptomic profiling of immune checkpoint genes. Of note, CRG signature score plus FGFR3 status could successfully distinguish BC patients who have a higher possibility of response to immunotherapy or chemotherapy drugs. CONCLUSIONS: The CRG signature is a potent prognostic model for BC patients, and in combination with FGFR3 alterations, it had more practical capacity in the prediction of chemotherapy and immunotherapy response, helping guide clinical decision-making.