Single-nucleus RNA-sequencing analysis reveals transcriptome differences in TIGIT+ exhausted T-cell associated with poor prognosis in muscle-invasive bladder cancer (MIBC).

Abstract

e16544 Background: Compared with single-cell RNA-sequencing (scRNA-seq), Single-nuclei RNA-sequencing(snRNA-seq) can detect more genes with higher sensitivity, and it is more robust in identifying cellular markers. TIGIT+ exhausted T cells were reported significantly associated with poor prognosis in muscle invasive bladder cancer(MIBC) by flow cytometry. We aimed to evaluate unique markers of TIGIT+ exhausted T cells in MIBC via snRNA-seq and to construct a potential prognostic survival model. Methods: We collected snRNA-seq data from 25 MIBC and 4 non-MIBC patients in GSE169379, including 57 cancer tissues and 4 distant normal tissues. Meanwhile, expression profiling data and clinical information of 402 Bladder Urothelial Carcinoma(BLCA) patients were collected in The Cancer Genome Atlas (TCGA). Firstly, the snRNA-seq data was preprocessed and filtered. Next, We re-clustered T-cell subsets and abstracted TIGIT+ exhausted T cells (TIGIT+Tex), then screened out the genes uniquely expressed in TIGIT+Tex subsets by differential expression analysis. We further evaluate the correlations between uniquely expressed genes and the exhausted T cell markers, identified significantly correlated differential expressed genes, constructed a prognostic model based on multivariate cox regression analysis and exploited the patient's overall survival and disease-free survival to confirm our findings. Results: Top 10 genes (CNTN1, CTPS1, EYA1, GRIK2, FOXP2, YBX1, DACH1, SYT1, NOL4, TRIT1) of unparalleled differential up-regulation were identified from TIGIT+ exhausted T cells via snRNA-seq data. Based on the correlation analysis of TCGA expression profile data, six genes (CNTN1, CTPS1, EYA1, GRIK2, FOXP2, YBX1) that were significantly positively correlated with exhausted T cells were obtained. Next, we constructed a regression prediction model consisting of 5 genes (CNTN1, CTPS1, EYA1, GRIK2, FOXP2) through multivariate cox regression analysis based on patients’ clinical information. The model was associated with significantly poor prognosis overall survival (OS) (P < 0.01,HR = 2) and disease-free survival (DFS) (P < 0.01,HR = 1.6) in BLCA patients. Conclusions: In our study, Single-nuclei RNA expression profiling data revealed transcriptome differences between TIGIT+ exhausted T cells and other types of T cells. The regression model for potential prognostic survival was constructed using TCGA expression profiling data. These results demonstrated that distinctive differentially high expressed genes associated with TIGIT+ exhausted T cells may act as potential markers for predicting prognosis outcomes in bladder cancer patients.