The role of STING on immune response and overcoming resistance to anti-PD-1 therapy in hepatocellular carcinoma

Abstract

Recently, the FDA approved two PD1 inhibitors(aPD-1) as the second-line therapeutic strategy for patients with hepatocellular carcinoma (HCC). However, an excessive number of patients with HCC lack a response to aPD-1 therapy, which underlies the urgent need to understand the mechanisms of this poor response. This understanding is essential development of targeting immunotherapies. Centrosome abnormalities, both in structure and number, contribute to tumor progression and associated immunosuppression. Our previous research indicated that overexpression of Tax1 binding protein 2 (TAX1BP2) inhibits centrosome overduplication and acts as a tumor suppressor in HCC. However, the role of TAX1BP2 in tumor immunity remains unclear. In this study, we demonstrated that intratumoral TAX1BP2 enhances lymphocyte infiltration and improves the effectiveness of aPD-1 therapy. Clinically, our results showed a positive correlation between the high expression level of TAX1BP2 and the level of tumor-infiltrating lymphocytes (TILs) by analyzing TCGA database, which was further validated in a clinical cohort of 70 untreated patients. Additionally, we established a link between TAX1BP2 inhibition and resistance in HCC patients treated with nivolumab (n=19). By establishing preclinical HCC murine models through hydrodynamic tail-vein injection of oncogenes, we found that genetic re-depression of TAX1BP2 enhances the efficacy of aPD-1 therapy. Mechanistically, we found that the reconstitution of intratumoral TAX1BP2 triggers the type 1 interferon (IFN-I) response, facilitating the recruitment of a specific subset of CD27+CD8+ T cells. Furthermore, we demonstrated that the suppression of intratumoral TAX1BP2 leads to the silencing of STING through the hyperactivation of DNMT1 and EZH2, which is linked to the activity of endogenous LKB1. Additionally, β-catenin (encoded by CTNNB1) mutations occur in around 11-37% of patients with HCC. A hallmark of β-catenin-mutant solid tumors is T-cell exclusion, and other studies have reported that β-catenin mutation is a factor of resistance to aPD-1 immunotherapy in cancer. In this study, we uncovered a cell-autonomous mechanism of β-catenin-driven immune evasion and proposed a novel therapeutic approach to enhance aPD-1 efficacy in β-catenin-mutant HCCs. Clinically, we explored the HCC cohorts, including 70 untreated, CCLE database,19 treated patients with nivolumab, as well as TCGA and CCLE databases. First, we observed that loss of STING signaling may be a potential signature in β-catenin-mutant HCC tumors. Furthermore, β-catenin mutation was associated with significantly reduced CD27/CD8+ T cell infiltration in tumors (n=70), and we confirmed this phenotype in HCC patients treated with nivolumab(n=19),indicating that CD27/CD8+ T cells are related to potentiating aPD-1 efficacy mediated by STING in β-catenin mutant HCCs. Additionally, we demonstrated that reconstitution of STING by pretreatment with decitabine (DAC) remodeled the tumor immune microenvironment (TIME) and potentiated the sensitivity of aPD-1 therapy in preclinical HCC models. Mechanistically, we revealed that β-catenin-mutant HCCs show unique hypermethylation of DNMT1 and silencing of STING, followed by inhibition of IFN-I signaling, subsequently compromising innate immune signaling and CD27+CD8+ T cell infiltration. In conclusion, our results indicated that targeting the intratumor STING pathway is a novel approach to overcome resistance to aPD-1 therapy in hepatocellular carcinoma.