ACE2 Enhances Sensitivity to PD-L1 Blockade by Inhibiting Macrophage-Induced Immunosuppression and Angiogenesis

Anti-PD-L1-based combination immunotherapy has become the first-line treatment for unresectable hepatocellular carcinoma (HCC). However, the objective response rate is lower than 40%, highlighting the need to identify mechanisms of tolerance to immune checkpoint inhibitors and accurate biomarkers of response. Here, we employed next-generation sequencing to analyze HCC samples from 10 patients receiving anti-PD-L1 therapy. Activation of the renin-angiotensin system was elevated in nonresponders compared with responders, and ACE2 expression was significantly downregulated in nonresponders. ACE2 deficiency promoted HCC development and anti-PD-L1 resistance, whereas ACE2 overexpression inhibited HCC progression in immune competent mice. Mass cytometry by time of flight (CyTOF) revealed that ACE2 deficient murine orthotopic tumor tissues featured elevated M2-like tumor-associated macrophages (TAMs), displayed a CCR5+PD-L1+ immunosuppressive phenotype, and exhibited high VEGFα expression. ACE2 downregulated tumor intrinsic CCL5 expression by suppressing NF-κB signaling through the ACE2/angiotensin-(1–7)/Mas receptor axis. The lower CCL5 levels led to reduced activation of the JAK-STAT3 pathway and suppressed PD-L1 and VEGFα expression in macrophages, blocking macrophage infiltration and M2-like polarization. Pharmacological targeting of CCR5 using maraviroc enhanced the tumor suppressive effect of anti-PD-L1 therapy. Together, these findings suggest that activation of the ACE2 axis overcomes the immunosuppressive microenvironment of HCC and may serve as an immunotherapeutic target and predictive biomarker of response to PD-L1 blockade. For ChIP analysis, MHCC97H cells were fixed employing 1% formaldehyde for 30 min, followed by adding 1.25 mol/L glycine for 5 minutes to end the reaction. DNA samples were sonicated in a Biorupter Pico (Diagenode) to acquire nearly 300 bp fragments. Next, the MicroPlex Library Preparation Kit (Diagenode) was employed for library preparation, followed by sequencing through the HiSeq 2000 Illumina platform of the NIG. The quality of the ChIP-seq raw data files were evaluated through FASTQ quality check (FASTQC). FastQ files were evaluated in the public server usegalaxy.org where the sequence reads were aligned to the mouse reference genome (mm9) using Bowtie2 (version 0.4) with default parameters.