IT-18 * GLIOBLASTOMA STAT3 AND PD-L1 EXPRESSION DRIVES ACCUMULATION OF IMMUNOSUPPRESSIVE LEUKOCYTES

Abstract

BACKGROUND: Glioblastoma (GBM) is a devastating disease with a median survival of 14 months. Immunotherapies hold promise but are compromised by tumor-mediated immunosuppression including accumulation of regulatory T-cells (TREG's) and myeloid-derived suppressor cells (MDSC's). Mechanisms leading to TREG and MDSC expansion in GBM patients are unclear. Tumor-derived STAT3 and PD-L1 have been implicated in direct GBM-mediated immunosuppression. We hypothesized that they might also be critical for accumulation of TREG's and MDSC's. METHODS: STAT3 and PD-L1 expression was knocked down (KD) in primary human GBM cell cultures using lentiviral constructs expressing shRNA and confirmed by western blot. Wild type, PD-L1-KD, or STAT3-KD GBM cells were co-cultured with healthy donor human CD4+ lymphocytes or CD14+ monocytes. TREG (CD4 + /CD25 + /FoxP3+), monocytic MDSC's (CD14 + /HLA-DR-), and granulocytic MDSC's (CD11b + /CD14-/CD15+/-/CD33+) frequencies were measured by flow cytometry. RESULTS: Lentiviral shRNA produced efficient STAT3 and PD-L1 KD in primary human GBM cell cultures. STAT3-KD did not alter PD-L1 expression, suggesting they are independently regulated. GBM co-culture produced TREG expansion from naive CD4+ cells (2.01% + /-1,03% vs. 0.77% + /-0.10% of CD4+) but this was markedly reduced with GBM cell STAT3-KD and PD-L1-KD (1.15% + /-0.04%, 1.35% + /-0.04%; P < 0.05). Monocytic MDSC's were 1.24% of naive CD14+. This increased with GBM co-culture (7.61% + /-1.87%) but decreased to 0.48% + /-0.34% with STAT3-KD and 4.43% + /-0.64% with PD-L1-KD (P< 0.05). Similar results were seen for granulocytic MDSC's. CONCLUSION: These findings provide clear evidence that expression of STAT3 and PD-L1, two important immunosuppressive signals, are independently regulated in human GBM cells. Furthermore, GBM cell STAT3 and PD-L1 expression are critical drivers of immunosuppressive leukocyte (TREG and MDSC) accumulation from naive CD4+ and CD14+ cells in addition to their more widely appreciated direct immunosuppressive effects on activated T cells. These findings suggest that targeting both STAT3 and PD-L1 may have synergistic immunomodulatory effects of potential benefit to GBM patients.