Combining PD-L1 blockade with a second-generation HSP110 inhibitor enhances antitumor immunity in colorectal cancer

In a previous work we demonstrated that pharmacological inhibition of wild type heat shock protein-110 (HSP110) could mimic an HSP110-inactive mutated phenotype reported to be associated with good prognosis in colorectal cancer (CRC) patients. We also demonstrated that HSP110 favored the formation of anti-inflammatory macrophages. We developed a second-generation HSP110 inhibitor called compound 7 (C7) through a hit-to-lead approach. Here, we deciphered its anti-cancer immune effect alone and associated with an anti-PD-L1/PD-1 targeting therapy. The study (FACS, IHC, immunoblots, qPCR, shRNA approaches) included syngeneic CRC mouse models (CT26/BALB/c and MC38/C57BL/6), 3D human CRC (HCT 116, HT-29) spheroids and primary human macrophages and lymphocytes isolated from buffy coats, and tumor biopsies from 134 CRC patients of the Prodige-13 trial. C7 specifically inhibited HSP110, leading to significant tumor growth suppression in both mice bearing CT26 and MC38 tumors and our human spheroids models incorporating immune cells. C7 reshaped the tumor microenvironment by promoting a pro-inflammatory phenotype. Studies in 2D and 3D co-cultures of human macrophages and lymphocytes indicated that C7's effect involved a direct action on macrophages. C7 also induced the expression of the immune check point PD-L1 both in macrophages and tumor cells. Combining C7 and an anti-PD-L1 antibody resulted in a more effective tumor regression both in the immune check point inhibitor resistant CT26- and in the non-resistant MC38- tumor-bearing mouse model. Finally, we established a clinical relevance of HSP110 effect on macrophages by showing a correlation between HSP110 and the anti-inflammatory biomarker CD163 in a CRC patients' cohort. These findings demonstrate that pharmacological inhibition of HSP110 alters pro-tumoral macrophages and can overcome resistance to immune check point inhibitors.