Engineered CAR-T cells offer potential rheumatoid arthritis cure

For the 18 million people living with rheumatoid arthritis, disease management currently means a lifetime of injections-a reality that new research from Tsinghua University may soon begin to change. Current first-line biologics targeting tumor necrosis factor (TNF)-including adalimumab (Humira), long among the world's best-selling drugs-require injections every two weeks, potentially for decades. Prolonged use drives anti-drug antibody (ADA) formation in many patients, eroding therapeutic benefit and causing secondary treatment failure. Chronic TNF blockade also heightens infection risks, including latent tuberculosis reactivation. Emerging targeted protein degradation (TPD) technologies, such as LYTACs and KineTACs, offer promising alternatives. Yet, all rely on the host's endogenous degradation machinery and require repeated dosing-making them poorly suited to chronic disease management.

A research team led by Professor Min Peng at Tsinghua University has now developed a fundamentally different approach. Published in hLife, the study reports engineered CAR-T cells-termed TNFR1TIF cells-that specifically capture and degrade soluble TNF through receptor-mediated endocytosis, functioning as a self-renewing, living cellular drug. A single infusion of TNFR1TIF cells prevented and treated all stages of RA in human TNF-transgenic mice, lowering serum human TNF to near wild-type levels and matching the efficacy of repeated high-dose adalimumab-without redosing or ADA risk.

Existing anti-TNF biologics rely on neutralization rather than degradation, while all current targeted protein degradation tools borrow the host's endogenous protein degradation machinery-an approach that can introduce unpredictable side effects and fundamentally limits durability. By recruiting an exogenous cellular platform-CAR-T cells-to perform the degradation, we bypass this limitation entirely. Our vision is a one-time intervention that provides years of sustained benefit for patients with chronic inflammatory disease."

The TNFR1 CAR was constructed by fusing the mouse TNFR1 ectodomain (amino acids 1–212) with CD28 co-stimulatory and CD3ζ signaling domains. To overcome a key translational barrier-conventional CAR-T cells' failure to expand without lymphodepleting preconditioning-the team co-deleted Bcor and Zc3h12a via CRISPR-Cas9. The resulting TNFR1TIF cells expanded robustly and maintained stable levels in peripheral blood for more than one year in fully immunocompetent mice, with no conditioning regimen required. Their persistence proved strictly TNF-dependent, as expansion was abolished in TNF-knockout animals. Twelve-month safety monitoring revealed no disruption of endogenous T cell homeostasis, organ integrity, or antibacterial immunity. A pharmacological safety switch-anti-Thy1.1 depleting antibody-efficiently eliminated TNFR1TIF cells in vivo, providing a controllable fail-safe mechanism for potential clinical translation.