Antigen–IL-2 CAR-enhancer drives CAR-T fate and stemness, enhancing antitumor efficacy across models independent of IL-2Rα

Abstract

BACKGROUND: Limited durability of clinical responses remains a major challenge in chimeric antigen receptor (CAR)-T therapy. CAR-enhancers (CAR-Es), which fuse tumor antigens to interleukin (IL)-2 muteins, provide a targeted strategy to enhance CAR-T persistence and function. It remained unclear whether CAR-Es are effective across distinct tumor contexts, when using patient-derived T cells, or in preventing exhaustion and sustaining persistence. It was also unknown whether CAR-Es can selectively expand CAR-Ts in humanized mice with pre-existing T cells, and to what extent their efficacy depends on IL-2Rβγ vs IL-2Rα engagement. While IL-2Rα (CD25) has been classically linked to potent antitumor responses and memory formation, it also drives IL-2-associated toxicities, including vascular leak and preferential regulatory T cell expansion. METHODS: We systematically dissected CAR-E signaling requirements by engineering IL-2 variants with selective receptor affinities. Multiple CAR-E constructs were developed and tested across a range of in vitro and in vivo models. RESULTS: We demonstrate that CAR-E activity is entirely independent of IL-2Rα and critically dependent on IL-2Rβγ signaling. A next-generation IL-2Rα-sparing CAR-E maintained full potency, driving robust CAR-T expansion, persistence, and tumor clearance, even at low doses and when using CAR-T cells derived from previously treated multiple myeloma patients. These CAR-T cells not only resisted exhaustion but also re-expanded months later to eradicate tumor rechallenges. In humanized mice with pre-established T cells, CAR-Es selectively expanded CAR-Ts to dominate the circulating T-cell pool. CAR-E exerted a dominant influence on CAR-T fate, overriding tumor-derived cues and enforcing consistent phenotypes across diverse preclinical models. CONCLUSIONS: These findings nominate a lead B-cell maturation antigen (BCMA)-IL-2 CAR-E candidate with strong translational potential for clinical development and establish IL-2Rβγ as a key driver of CAR-E activity. The results also identify IL-2Rα as dispensable and provide a mechanistic framework for designing safer, IL-2Rα-sparing CAR-Es.