Answers are grounded in this paper and cite their sources.
Chimeric antigen receptor regulatory T cells (CAR-Tregs) offer a strategy to concentrate immune regulation within diseased tissues while limiting systemic immunosuppression. Their application to autoimmunity, however, introduces design requirements that differ from those in transplantation. Autoimmune lesions are heterogeneous, dynamic, and often shaped by epitope spreading, inflammatory disruption of Treg identity, and a narrow tolerance for off-target immune suppression. Here, we examine CAR-Treg development through four interconnected design requirements: target architecture, context-dependent function, lineage stability and persistence, and clinical control. Target selection must consider not only antigen specificity, but also native presentation, accessibility, tissue distribution, and the ability to support productive CAR engagement. Functional programming is increasingly extending beyond constitutive suppression toward inducible and locally restricted regulatory outputs. Durable activity further depends on maintaining Treg identity under inflammatory and repeated antigenic challenge, requiring coordinated control of FOXP3 expression, receptor signaling, cytokine support, and tissue fitness. Finally, therapeutic activity cannot be inferred from blood persistence alone; tissue localization, local phenotype, target engagement, manufacturability, and reversibility are integral to the pharmacology of these products. Using type 1 diabetes as a benchmark and drawing on evidence from neuroinflammation, inflammatory bowel disease, antibody-mediated autoimmunity, and early clinical development, we show how different targeting and engineering strategies address distinct biological constraints. The emerging principle is that successful autoimmune CAR-Tregs will depend less on maximizing suppressive potency than on matching recognition, function, stability, and persistence to the biology of the affected tissue.
No takes yet. Share an insight, caveat, or question.
Noyan et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: