Characterizing tumor nicroenvironment features and identifying potential therapeutic targets in patients with diabetes mellitus and pancreatic cancer based on single-cell transcriptome sequencing

Objective‍ ‍To investigate the changes in the tumor microenvironment of pancreatic cancer (PDAC) complicated with diabetes mellitus (DM) in a mouse model of hyperglycemia and orthotopic pancreatic cancer by analyzing transcriptome and single-cell transcriptome data in order to identify potential therapeutic targets. Method‍ ‍By integrating single-cell transcriptome and bulk transcriptome data, bioinformatics analysis was conducted to compare the characteristics of tumor cells and tumor immune microenvironment between PDAC patients with DM (DM group) and those without DM (non-DM group). Twenty male C57BL/6 mice (6 weeks old, weighing 18~20 g) were randomly divided into a hyperglycemic group [STZ group, continuous intraperitoneal injection of 50 mg/kg streptozocin (STZ) (final concentration of 1%) dissolved in citrate buffer], and a control group (Control group, an equivalent volume of citrate buffer without STZ at the same time points), with 10 mice in each group. Tail-tip blood glucose level was measured to monitor glycemic status. After orthotopic inoculation of pancreatic cancer cells in both Control and STZ groups, tumor-infiltrating immune cells were harvested. Flow cytometry was employed to determine the effects of hyperglycemia on: total CD8+ T cell and Treg cell populations; CD8+ T cell subsets expressing Ki67, TNF-α, granzyme B (GZMB) and IFN-γ; surface expression of PD-1, lymphocyte activation gene-3 (LAG-3) and T cell immunoglobulin and mucin domain-3 (Tim-3) on CD8+ T cells; programmed death-ligand 1 (PD-L1) expression on tumor cells; and tumor-associated macrophage surface expression of major histocompatibility complex class I (MHC-Ⅰ) and cluster of differentiation 206 (CD206). Results‍ ‍Bioinformatics analysis revealed that, compared to the non-DM group, the genes significantly up-regulated in the DM group were associated with poor prognosis (P<0.001). The proportion of type 2 ductal cells was increased in the DM group, exhibiting higher levels of copy number variation (P<0.001). In the tumor immune microenvironment of the DM group, there was an increase in the proportion of Treg cells (P<0.05) and an elevated exhaustion score for CD8+ T cells (P<0.001), accompanied by down-regulated expression of effector molecules, up-regulated expression of inhibitory checkpoints, and a significant increase in the M2 score of M2-like macrophages (P<0.001). Animal experiments and flow cytometry found that, compared to the Control group, the STZ group had a shorter survival time (P<0.001), with decreased proportions of total CD8+ T cells (P<0.01) and CD8+ T cells expressing Ki67, TNF-α, GZMB and IFN-γ (P<0.01), increased proportion of Treg cells (P<0.001), up-regulated expression of PD-1, LAG-3 and Tim-3 on the surface of CD8+ T cells (P<0.001), and up-regulation of PD-L1 on tumor cell surface (P<0.001) and enhanced expression of CD206 on the surface of tumor-associated macrophages, while down-regulated expression of MHC-I (P<0.001). Conclusion‍ ‍High glucose promotes the formation of an immunosuppressive microenvironment in PDAC, and targeting type 2 ductal cells and immunosuppressive cells in the tumor microenvironment, combined with dual immune checkpoint antibody therapy, may improve patient prognosis.