In Vivo Prevention of Murine GVHD

Abstract
Our overall goals are focus on the prevention and treatment of graft-vs-host disease (GVHD) via innate lymphoid
type 2 cell (ILC2) anti-inflammatory and tissue reparative properties. We found that host ILC2s in are eliminated
by total body irradiation {TBI} or chemotherapy and remain depleted for at >/=90 days. This finding is highly
relevant as there is an inverse correlation between peripheral blood activated ILC2s and GVHD. As such. we
sought to determine whether supplemental infusion of mature donor ILC2s could be used to prevent murine
GVHD. We showed for the first time that donor ILC2s could prevent or partially treat GVHD in an amphiregulin
(AREG) dependent process. Whether AREG or ILC2 direct contact with intestinal stem cells {ISC) supports small
intestine epithelial cell repair in TBI treated mice or organoids is unknown. Additionally, third-party ILC2 infusion
also significanUy reduced murine GVHD lethality. lmportanUy for translational purposes, we found ILC2s to be
relatively steroid resistant. Peri-BMT {bone marrow transplant) IL-33 increased ST2/IL33R+ ILC2s at BMT day
0 and reduced GVHD. Ko mice had accelerated GVHD; IL-33 given pre-BMT prevented the full lLC2 loss. IL-33
ko recipients have hypo-proliferative epithelial cells, reduced ISCs and Paneth cells, and smaller crypt height
and numbers. Ex vivo intestine organoid culture modeling revealed that IL-33 coordinated regeneration by
inducing epidermal growth factor (EGF), significantly reduced by TBI. EGF restored ISC deficiency, uncovering
a gut repair IL-33/EGF loop between ISCs and Paneth cells. Donor IL-13 ko ILC2s or host IL 13Ra ko mice had
reduced GVHD. ILC2 IL 13 supports both ST2+ tuft cells and goblet cells. Tuft cells produce IL-25 driving ILC2
production and survival. TBI markedly reduced tuft cells for :!:38 days and ko recipients had a striking increase
in GVHD. When given to wildtype mice exogenous IL-25 significantly reduced GVHD. Consequences of ko of
tuft cells (and ILC2s) on donor T cell expansion, trafficking and function are unknown. The role of IL-25 and
IL 17RB has not been examined. We will address the dynamics and interplay between ILC2, IL-33 and
host intestinal cells (tuft cells, ISCs, Paneth cells) after TBI and during GVHD.
Aim 1 will test the hypothesis that: Donor ILC2 repopulation fails due to destruction of ILC2 BM niche
that supports ILC2s. In vitro pre-lLC2 differentiation, maturation and expansion ± proinflammatory
cytokines and ILC2 supporting cytokines will be studied. GATA3-GFPhi pre-lLC2s/mature ILC2s transfer
into lethally irradiated congenic BMT recipients will provide data on the differential ability to repopulate the
BM. If the BM cannot support pre-lLC2s/lLC2s, we will study stem cell deficient mice. Aim 2 will test the
hypothesis that: Pre-lLC2s/lLC2s and their secreted products have direct effects on intestinal cell subsets.
Intestinal organoid cultures from wild-type and IL-33 ko mice with syngeneic Tregs and ILC2s will measure
organoid size, number, and gene expression related to proliferation, cell cycle regulation, and specific
epithelial lineage markers under homeostasis or after TBI. Anti-AREG mAbs or co-cultures with AREG ko
lymphocytes will be characterized for promoting epithelial regeneration. Aim 3 will test the hypothesis that: Tuft
cells are essential for ILC2 development and survival via an ILC2 release of IL-13 that stimulates tuft cells to
release IL-25 that causes ILC2 proliferation (aim 3). Tuft cell and ILC2 ko hosts have accelerated GVHD. We
hypothesize that IL-25 effects are due to direct stimulation of ILC2s or alternatively, with donor IL-17RB+ T
cells.
Significance. Studies in the R37 extension phase will provide fundamental information as to the mechanisms
by which peri-BMT IL-33 diminish GVHD lethality via effects on ST2+ host ILC2s and regulatory T cells,
both of which produce AREG, and the EGF/IL-33 loop that occurs between ISCs and Paneth cells resulting in
small intestine repair. Further, the key role of host ILC2s in subduing GVHD, the nature of post-BMT ILC2
deficiency that occurs after pre-BMT conditioning regimens and predisposes patients to GVHD, and the
essential requirement for tuft cells or their product, IL-25 will be elucidate. LasUy, critical insights will be
gained as to the inability of pre-lLC2s generation, gut migration or differentiation.
Translational Impact: The efficacy of donor and third-party ILC2 infusion in preventing and treating GVHD
support our planned human ILC2 clinical trial, funded via other auspices, that will infuse "off-the-shelf' third
party ILC2s to treat steroid refractory gut GVHD, that portends a particularly poor prognosis, in a 2-
institutional study at the University of Minnesota and UNC-CH.

Grant Number: 5R37AI034495-33
NIH Institute/Center: NIH
Principal Investigator: Bruce Blazar