Absence of ILC2 Recovery Post-Transplant Is Mediated By Epigenetic Conversion of ILC2 to Pro-Inflammatory ILC1-like Cells

Abstract

Although allogeneic hematopoietic stem cell transplantation (alloHSCT) is the preferred treatment for a variety of hematologic malignancies, its use is limited by the development of acute graft-versus-host disease (aGvHD). Type II innate lymphoid cells (ILC2s) are immune cells that play an important role in maintaining homeostasis in mucosal tissues. Previous work has shown that ILC2 cells fail to reconstitute after chemotherapy or stem cell transplantation, though the mechanism for this finding is unclear. Our group has demonstrated that ILC2s in the GI tract are depleted by chemotherapy or radiation prior to alloHSCT, and that the infusion of donor ILC2s improves the clinical score and survival of recipient mice after transplant. Importantly, ILC2s are also lost in human alloHSCT patients following conditioning therapy. Recent clinical evidence indicates that the presence of ILC2s expressing activation and tissue homing markers are associated with decreased susceptibility to therapy-induced mucositis and aGVHD. Thus, delineation of the mechanisms involved in their turnover and reconstitution could have a significant impact on transplant outcomes. Since the existing dogma is that most ILCs are tissue resident, with only a small circulating pool that can be replenished from the bone marrow, we became interested in understanding why these cells fail to recover and repopulate their niche in the GI tract after alloHSCT. Given that current studies investigating ILC biology across multiple mucosal compartments in both mice and human demonstrate significant ILC plasticity, we evaluated the hypothesis that the loss of ILC2 cells post-transplant induced epigenetic changes that convert ILC2 cells to either ILC1 or ILC3-like cells. To address this hypothesis, we used ChIP-, ATAC-, and RNA-sequencing to assess the chromatin landscape and regulation of ILC2 cells expanded ex vivo, as well as those recovered from the GI tract after HSCT in an allo-mismatched murine model. Strikingly, single-cell, multiomic analysis of donor-derived ILC2s after transplantation revealed a previously unreported population of Tbet-expressing, IFN-γ-secreting ILC1-like cells that differentiate from ILC2s in the small intestine lamina propria (exILC2s). To recapitulate this transdifferentiation, we modeled skewing of ILC2s in vitro with IL-12, IL-1β, and IL-18 and observed a reduction in Type 2 lineage-defining regulatory factors (LDRFs) and the acquisition of proinflammatory Type 1 LDRFs consistent with the phenotype and function of IFN-γ-secreting exILC2s. To determine how these regulatory changes occur, we examined transcription factors with motifs present in regions of open chromatin that were differentially expressed in our transcriptome data between 'true ILC2s' and ILC1-like exILC2 cells recovered from the small intestine. Excitingly, this approach confirms the role of Tbx21 as an "on-off" switch for ILC fate in this model and reveals the early immediate response gene Nr4a2 as an additional key driver of the emergent ILC1-like exILC2s after alloHSCT. To test whether these in vitro generated exILC2s contribute to aGVHD-mediated mortality, we infused transdifferentiated donor WT or exILC2s and measured the clinical score and survival of recipient mice after alloHSCT in a mismatched murine model. Unlike their unmanipulated WT ILC2 counterparts, our in vitro-derived exILC2s failed to ameliorate gut aGvHD and instead accelerated morbidity and mortality. These data demonstrate, for the first time, that ILC2s fail to repopulate their protective niche after alloHSCT due to epigenetic reprogramming and pathological transdifferentiation. These findings provide novel insights into the contribution of ILC plasticity to mucosal dysregulation and aGvHD pathogenesis after alloHSCT, and may inform new approaches for modulating innate lymphocytes in human disease.