Using the binary language of IL-2 to prevent GVHD

In this issue of Blood, Song et al¹  identified a new means of selective blockade of the interleukin 2 (IL-2) pathway during hematopoietic cell transplantation that prevents graft-versus-host disease (GVHD) and augments donor graft-versus-leukemia (GVL) immunity. Using preclinical murine major histocompatibility complex mismatch models of hematopoietic stem cell transplantation, the group used a monoclonal antibody, JES6-1, that binds to IL-2 and selectively interferes with IL-2 binding to specific subunits of the IL-2 receptor (ie, the IL-2 receptor β and γ subunits but not the IL-2 α subunit). This study is among a number of exciting recent developments in our understanding of IL-2 biology and the potential implementation of whole new classes of clinical interventions targeting IL-2.

The importance of the IL-2 pathway in transplantation is underscored by the pivotal role that calcineurin inhibitors played in advancing transplantation. Calcineurin inhibitors act in large part by blunting T-cell receptor signaling and reducing the ability of immune cells to produce IL-2. This includes CD4⁺ helper T cells, which are a major source of IL-2 that drives both their own autocrine expansion and that of other cells such as effector CD8 T cells. Calcineurin inhibitors such as tacrolimus have known toxicities, but they remain the backbone of most strategies to prevent GVHD.²  As Song et al again demonstrate in their study, calcineurin inhibitors can impede the GVL effects of donor T cells.

IL-2 has been shown to be plieotropic and can sometimes exert apparently contradictory effects.³  This is in part because IL-2 binds to dimeric receptors of 2 general types: the IL-2RαIL2Rγ receptor, which is expressed on activated T cells and regulatory T cells (Tregs), and the IL2RβIL-2Rγ receptor, which is expressed on immunologically active T, B, and natural killer (NK) cells. Administration of IL-2 at certain doses can, for example, cause Tregs to expand. This has been used as a clinical strategy to treat chronic GVHD,⁴  but at other doses, IL-2 can worsen alloimmune responses and drive T cell or NK reactivity. Monoclonal antibodies that block certain IL-2 receptors or designer IL-2 mutant proteins that bind to the IL-2RαIL-2γ receptors preferentially have been shown to increase Treg numbers and are being evaluated in clinical trials to induce immune tolerance.⁵ 

Song et al used the antibody JES6-1 to selectively block IL-2 from binding to IL-2 receptors made from the β and common γ chains but not receptors made from the α and common γ chain I (CD25) that Tregs depend on. This reduced donor T-cell proliferation reduced the number of CD4⁺ T cells in the colon and liver, which are GVHD target organs. Inflammatory cytokines such as granulocyte colony-stimulating factor are also reduced. As experimental controls, they used a rat immunoglobulin G (IgG) isotype antibody, and the anti–IL-2 S4B6 antibody that binds to IL-2 allows for some engagement of IL-2 receptors that use the β chain and common γ chain. Compared with control antibodies, Song et al show that JES6-1–treated mice have significantly less severe GVHD and better survival.

In an extensive mechanistic evaluation, the group showed that multiple inter-related pathways are altered to prevent GVHD using this selective antibody approach. Based on other studies, one would first suspect that conventional CD4⁺CD25⁺FOXP3⁺ Tregs are primarily responsible for GVHD protection because increasing Treg numbers in other contexts protects against GVHD.⁶  However, Treg percentage did not change in JES6-1–treated recipients. Although Tregs might be necessary for protection (which was not tested), Song et al showed that other pathways are involved. In particular, recipient expression of the immune regulatory receptor PD-L1 is required for GVHD protection because PD-L1^−/− recipient mice die from GVHD when treated with JES6-1. Likewise, the group shows that PD-L1 is required for the local colon and liver propagation of CD4⁺ TR1 cells, which make IL-10. CD4⁺ TR1 cells are known to regulate and prevent GVHD.⁷  Song et al elegantly shows that IL-10 produced by donor-derived TR1 cells is needed for GVHD protection by the adoptive transfer of IL-10^−/− donor cells. They also show increased PD-1:PDL-1–dependent Eomes expression by colon and liver infiltrating donor effector T cells. Higher Eomes expression correlates with anergic and exhausted donor T cells likely incapable of mediating GHVD. One consideration is that Eomes expression in T cells after transplantation can reduce GVL.⁸ 

In a series of elegant experiments evaluating GVL, Song et al showed that JES6-1 treatment does not prevent immune clearance of the BCL-1 and BC-CML cancer cell lines, whereas treatment with tacrolimus does. Song at al used single cell RNA-Seq and confirmatory flow cytometric of splenic T cells and showed that JES6-1–treated mice have different T-cell immune reconstitution of CD4 and CD8 populations in the spleen that do not appear anergic vs what is seen in GVHD target tissue. This may promote GVL. Song et al present some evidence that NK cells are not responsible for GVL, but further mechanistic studies are needed to understand the preservation of GVL in this model, including evaluating alternative hypotheses such as the increase in TR1 cells that may be responsible for increased GVL.

Overall, the study by Song et al reinforces the concept that IL-2 and donor CD4⁺ T cells in the local tissue microenvironment play a major role in GVHD pathophysiology through the PD-L1 pathway and the complex interplay of cell populations such as Treg, TR1, and effector T-cell subsets. The authors also make the fundamental observation that targeting the IL-2 pathway more precisely may promote GVL immunity.