Canonical interleukin-2 (IL-2) signaling via the high-affinity CD25-containing IL-2 receptor-Janus kinase (JAK)1,3-signal transducer and activator of transcription 5 (STAT5) pathway is essential for development and maintenance of CD4⁺CD25^HiFoxp3⁺ regulatory T cells (Tregs) that support immune homeostasis. Here, we report that IL-2 signaling via an alternative CD25-chemokine receptor pathway promotes the suppressive function of Tregs. Using an antibody against CD25 that biases IL-2 signaling toward this alternative pathway, we establish that this pathway increases the suppressive activity of Tregs and ameliorates murine experimental autoimmune encephalomyelitis (EAE). Furthermore, heparan sulfate, an IL-2-binding element of cell surfaces and extracellular matrix, or an engineered IL-2 immunocytokine can also direct IL-2 signaling toward this alternative pathway. Overall, these data reveal a non-canonical mechanism for IL-2 signaling that promotes suppressive functions of Tregs, further elucidates how IL-2 supports immune homeostasis, and suggests approaches to promote or suppress Treg functions.

Low-dose human interleukin-2 (hIL-2) treatment is used clinically to treat autoimmune disorders due to the cytokine's preferential expansion of immunosuppressive regulatory T cells (Tregs). However, off-target immune cell activation and short serum half-life limit the clinical potential of IL-2 treatment. Recent work showed that complexes comprising hIL-2 and the anti-hIL-2 antibody F5111 overcome these limitations by preferentially stimulating Tregs over immune effector cells. Although promising, therapeutic translation of this approach is complicated by the need to optimize dosing ratios and by the instability of the cytokine/antibody complex. We leverage structural insights to engineer a single-chain hIL-2/F5111 antibody fusion protein, termed F5111 immunocytokine (IC), which potently and selectively activates and expands Tregs. F5111 IC confers protection in mouse models of colitis and checkpoint inhibitor-induced diabetes mellitus. These results provide a roadmap for IC design and establish a Treg-biased immunotherapy that could be clinically translated for autoimmune disease treatment.