LAUSR

Background Durable clinical responses to immune checkpoint blockade (ICB) occur in a limited fraction of patients. We thus hypothesized that the characteristic tumor metabolic switch towards aerobic glycolysis could contribute to ICB resistance. High glucose consumption and lactate production by tumor cells can indeed restrict nutrient availability for tumor-infiltrating T cells, which also rely on glycolysis to proliferate and function. Therefore, we investigated whether targeting tumor glucose metabolism potentiates ICB anti-tumor activity. Methods We modeled tumor-selective glycolysis inhibition by knocking down the critical glycolytic enzyme lactate dehydrogenase A (LDHA-KD) in the murine metastatic breast carcinoma 4T1 and melanoma B16, which are known immune-refractory tumor models. Anti-CTLA-4 and anti-PD-1 were tested in immunocompetent mice orthotopically implanted with control vs. LDHA-KD tumor cells. Changes in glucose metabolism were assessed by Seahorse and fluorescent-glucose flow-cytometry staining. Changes in immune cells were measured by multiparameter flow cytometry. Glucose-dependent effects of anti-CTLA-4 in regulatory T cells (Tregs) were tested in standard suppression assays with increasing glucose concentration (0.5–10 mM). Pearson correlations between glycolysis and intra-tumor immune-cell infiltration by CIBERSORT immune-deconvolution method were analyzed in bulk RNA-sequencing data sets from human and murine tumors treated with ICB. Results Comparison of ICB activity in LDHA-KD vs. control tumor-bearing mice revealed improved anti-tumor effects and overall survival in the setting of glycolysis-defective tumors specifically upon CTLA-4 blockade. Anti-tumor CD8+ T-cell responses correlated with Treg phenotypic and functional destabilization in anti-CTLA-4-treated LDHA-KD tumors. CTLA-4 blockade led to CTLA-4 and CD25 downregulation associated with increased IFN-gamma and TNF-alpha production in Tregs from glycolysis-defective vs. control tumors. We next mimicked high- vs. low-glycolysis tumor microenvironment (TME) in vitro using control vs. LDHA-KD tumor co-cultures with Tregs, control vs. LDHA-KD tumor-conditioned media or directly modulating glucose concentrations. In these assays, we observed that CTLA-4 blockade promotes IFN-gamma±TNF-alpha production and glucose uptake by Tregs and more efficiently counteracts Treg suppression and enhances CD28 co-stimulation at higher glucose concentrations. Lastly, by interrogating transcriptomic data from human melanoma and murine 4T1 tumors, we found that CTLA-4 blockade promotes immune-cell infiltration and metabolic fitness especially in glycolysis-defective tumors. Conclusions Our findings indicate that increasing glucose availability in the TME may improve anti-CTLA-4 therapeutic activity and reveal a new mechanism through which CTLA-4 blockade interferes with Treg immunosuppression in a glucose-dependent manner. These results suggest that CTLA-4 blockade can be more effective in tumors with low glycolysis and/or can be best exploited in combination with inhibitors of tumor glycolysis.