Altered metabolism is a hallmark of cancer and glycolysis is one of the important factors promoting tumor development. There is however still a lack of molecular characterization glycolysis and comprehensive studies related to tumor glycolysis in the pan-cancer landscape. Here, we applied a gene expression signature to quantify glycolysis in 9229 tumors across 25 cancer types and 7875 human lung cancer single cells and verified the robustness of signature using defined glycolysis samples from previous studies. We classified tumors and cells into glycolysis score-high and -low groups, demonstrated their prognostic associations, and identified genome and transcriptome molecular features associated with glycolysis activity. We observed that glycolysis score-high tumors were associated with worse prognosis across cancer types. High glycolysis tumors exhibited specific driver genes altered by copy number aberrations (CNAs) in most cancer types. Tricarboxylic acid (TCA) cycle, DNA replication, tumor proliferation and other cancer hallmarks were more active in glycolysis-high tumors. Glycolysis signature was strongly correlated with hypoxia signature in all 25 cancer tissues (r > 0.7) and cancer single cells (r > 0.8). In addition, HSPA8 and P4HA1 were screened out as the potential modulating factors to glycolysis as their expression were highly correlated with glycolysis score and glycolysis genes, which enables future efforts for therapeutic options to block the glycolysis and control tumor progression. Our study provides a comprehensive molecular-level understanding of glycolysis with a large sample data and demonstrates the hypoxia pressure, growth signals, oncogene mutation and other potential signals could activate glycolysis, thereby to regulate cell cycle, energy material synthesis, cell proliferation and cancer progression.

Highlights Outgroup was not used due to the unknown source of SARS-CoV-2; 16373 SARS-CoV-2 genomes were included in the evolution analysis; 9 key specific sites of highly linkage and 4 major haplotypes were found; Epidemic trends and possible earlier origins of SARS-CoV-2 were indicated.

T-cell exhaustion is one of the main reasons of tumor immune escape. Using single-cell transcriptome data of CD8+ T cells in multiple cancers, we identified different cell types, in which Pre_exhaust and exhausted T cells participated in negative regulation of immune system process. By analyzing the coexpression network patterns and differentially expressed genes of Pre_exhaust, exhausted, and effector T cells, we identified 35 genes related to T-cell exhaustion, whose high GSVA scores were associated with significantly poor prognosis in various cancers. In the differentially expressed genes, RGS1 showed the greatest fold change in Pre_exhaust and exhausted cells of three cancers compared with effector T cells, and high expression of RGS1 was also associated with poor prognosis in various cancers. Additionally, RGS1 protein was upregulated significantly in tumor tissues in the immunohistochemistry verification. Furthermore, RGS1 displayed positive correlation with the 35 genes, especially highly correlated with PDCD1, CTLA4, HAVCR2, and TNFRSF9 in CD8+ T cells and cancer tissues, indicating the important roles of RGS1 in CD8+ T-cell exhaustion. Considering the GTP-hydrolysis activity of RGS1 and significantly high mRNA and protein expression in cancer tissues, we speculated that RGS1 potentially mediate the T-cell retention to lead to the persistent antigen stimulation, resulting in T-cell exhaustion. In conclusion, our findings suggest that RGS1 is a new marker and promoting factor for CD8+ T-cell exhaustion and provide theoretical basis for research and immunotherapy of exhausted cells.

Graphical abstract