Targeting tumoral GLI1 in triple-negative breast cancer: A novel strategy to reverse the immunosuppressive microenvironment

Triple-negative breast cancer (TNBC) comprises 15-20% of breast cancer cases, is highly aggressive, lacks effective treatment options and represents a disproportionately high number of BC-related deaths. Despite the clinical success of immune checkpoint blockade (ICB) in many malignancies, ICB has failed to demonstrate a similar response in TNBC, where most cases are highly infiltrated by tumor-associated macrophages (TAMs), which can dampen CD8+ cytotoxic T-lymphocyte (CTL) function and promote resistance to ICB. The hedgehog (HH) signaling pathway is highly activated in TNBC and higher expression of GLI1 is associated with worse overall survival. Recent pan-cancer analyses highlight a significant correlation between activated HH signaling and characteristics of immune evasion. Results: To examine how activated HH signaling regulates the tumor immune microenvironment (TIME) in TNBC, murine KBP (K14- cre ; Brca1 fl/fl ; P53 fl/fl ) TNBC cells were transfected with CRISPR/Cas9 + sgRNA targeting GLI1 or GLI2 (GLI1/2). KBP and KBP-GLI1/2-KO cells were orthotopically injected into the mammary fat pads of FVB mice and allowed to grow for 6-8 weeks. Compared to wild-type KBP tumors, KBP-GLI1-KO tumors, but not KBP-GLI2-KO tumors, were smaller. Flow cytometric analyses showed that KBP GLI1-KO tumors were infiltrated with fewer F4/80+ CD11b+ TAMs, particularly the CD206+ pro-tumoral M2-like subtype. This was accompanied by increased CD8+ and CD4+ T cells. Similar results were observed in the modified murine TNBC KBP transgenic mouse model where either GLI1 or GLI2 was genetically deleted in tumor cells. In GLI1 (but not GLI2) KBP mice reduced delayed tumor onset and growth were accompanied by reduced M2-like TAM and increased CD8+ and CD4+ T cells infiltration in TIME. Moreover, compared with the low responsiveness of KBP tumors to anti-PD1, the therapeutic benefit of anti-PD1 in KBP-GLI1-KO tumors was markedly improved and was accompanied by increased activated, anti-tumoral CD8+GrB+ T cell infiltration. Conclusion: here we demonstrate that tumoral GLI1, but not GLI2, promotes tumor growth and supports an immunosuppressive TIME in TNBC. Targeted deletion of GLI1 reduces TAMs, increases CTLs, and re-shapes TIME towards a pro-inflammatory ‘hot' tumor phenotype, potentiating anti-PD1 ICB in immunocompetent mouse models of TNBC. Overall, these preliminary findings identify GLI1 as a potential therapeutic target to improve ICB response in TNBC.