We identified ammonia as a byproduct of amino acid metabolism as an important factor limiting the activity of effector immune cells. We have documented that ammonia can be accumulated in tumor cell cultures and tumor interstitial fluid (TIF) isolated from mice inoculated with tumor cell lines. We have established a method of isolation of TIF from murine tumors that enables the analysis of tumor metabolome. To elucidate the mechanisms responsible for the immunosuppressive activity of ammonia, we optimized a vast array of methods to study the activity of NK cells and CAR-T cells. We observed that ammonia inhibits the natural cytotoxicity of NK cells, as well as CAR-mediated killing. To further identify the mechanisms of inhibition, we optimized and employed methods to study degranulation, cytokines production, and synapse formation. We determined that ammonia, as a lysosomotropic agent, impairs the process of perforin production and maturation and, in this mechanism, inhibits the cytotoxic potential of effector cells. Altogether, we have identified a unique, novel and never described before mechanism by which ammonia interferes with cytotoxic activity of T cells and NK cells. The experiments on the role and mechanism of impairment of cytotoxic cells by ammonia were published in Cancer Research journal (doi: 10.1158/0008-5472.CAN-24-0749).
In the screening experiments, we also observed that liver cancer cell lines produce low amounts of ammonia. To better mimic the tumor microenvironment and ammonia distribution within the liver, we optimized more advanced orthotopic models of hepatocellular carcinoma (HCC) in xenograft model of NSG mice with tumor cells administered directly into the portal vein. In these orthotopic in vivo models of HCC, we observed no accumulation of ammonia in TIF. Therefore, we subsequently selected a liver cancer model characterized by low ammonia concentrations to determine the efficacy of CAR-based immunotherapy. Subsequently, we developed PD-L1-CAR T cell therapy administered through the splenic vein to directly target liver neoplasms and reshape TME. PD-L1-CAR T cells effectively eliminated liver cancer cell lines in vitro, however, systemic delivery in vivo resulted in lethal toxicity. To improve the safety profile, liver locoregional administration was employed which helped to avoid major side effects, preserve organ integrity, and maintain anticancer efficacy. Additionally, PD-L1-CAR T cells demonstrated high efficacy against patient-derived liver tumor cultures from diverse origins. Altogether, we showed that locoregional administration of PD-L1-CAR T cell is effective against liver cancer.
Using the optimized method of TIF isolation, we also quantified other immunosuppressive metabolites in TME. We demonstrated that hydrogen peroxide concentrations are elevated in tumor interstitial fluid isolated from murine breast cancers in vivo. We also observed that NK cells were the most susceptible to hydrogen peroxide and identified peroxiredoxin-1 (PRDX1) as a lacking element of NK cells' antioxidative defense. Finally, we generated PD-L1–CAR NK cells overexpressing PRDX1 that displayed potent antitumor activity against breast cancer cells. These experiments were published in Cancer Immunology Research journal (doi: 10.1158/2326-6066.CIR-20-1023).
