The etiology of graft-versus-host disease (GVHD) is rooted in the alloreactive response of donor T cells present in the hematopoietic graft resulting in the destruction of the patient's tissues, particularly the gastrointestinal tract. Gut GVHD affects up to 50% of patients, is a leading cause of death, and has overlapping features with inflammatory bowel disease (IBD). Increased gut permeability, alterations of the gut microbiome and intestinal stem cell niche damage predispose the gut to the local and systemic effects of GVHD, and is exacerbated by the inability of the gut to adequately regenerate.

Severe shifts in metabolism and reduced oxygen (O2) availability in the inflamed gut are major underlying factors in the pathogenesis of IBD. Two related transcription factors, hypoxia-inducible factor-1 (HIF-1) and HIF-2, originally discovered as master regulators of the adaptive response to hypoxia, have been recently shown to be gut protective and promote mucosal healing in response to injury. Using a MHC mismatched B10.BR→B6 bone marrow transplant (BMT) model, we previously found that loss of intestinal epithelial (IE) HIF-1α or HIF-2α worsened survival compared to wild-type mice and exhibited increased GVHD-induced-histopathology. Thus, we hypothesized that HIF-1 and HIF-2 protects and repairs the gut from conditioning and GVHD-related damage.

HIF-1 and HIF-2 are heterodimers consisting of an O2-labile HIF-1α or HIF-2α subunit respectively and a constitutively expressed HIF-1β subunit. The recent discovery that iron-dependent prolyl hydroxylase enzymes (PHD1-3) can trigger the O2-dependent proteasomal degradation of the HIF-1α subunit has led to the development of pan-PHD inhibitors (PHDi) that activate HIF-1 and HIF-2. PHDi such as dimethyloxallyl glycine (DMOG) and AKB-4924 have been shown to attenuate colitis and radiation-induced gut toxicity in animal models. We thus sought to determine whether PHDi could also ameliorate gut GVHD in the B10.BR→B6 BMT model. B6 mice were lethally irradiated (10Gy, split dose) and transplanted with 5x106 T-cell depleted bone marrow (BM) cells from B10.BR donors with 2x106 enriched T cells from spleens and lymph nodes (BM+T). B6 mice transplanted with only T cell depleted BM cells served as our non-GVHD control group. B6 mice were treated with AKB-4924 (AKB, 5 mg/kg) or vehicle (β-cyclodextrin) via intraperitoneal delivery, beginning 1 day prior to BMT and for 6 consecutive days post-BMT. Treatment with AKB prevented significant weight loss, compared to vehicle-treated mice, 7 days post-BMT (n=6, p<.002). Histopathologic assessment of the jejunum of AKB treated mice after 7 days revealed fewer apoptotic cells and an increased number of Paneth and goblet cells compared to vehicle-treated mice. Immunohistochemical staining for lysozyme showed that AKB prevented GVHD-induced Paneth cell ablation compared to their vehicle BMT counterpart (2.8 Lyz+ cells/crypt vs 0.84 Lyz+ cells/crypt, n=3, p<.003). Additionally, we evaluated the effects of AKB on T cell alloreactivity. In mixed lymphocyte reaction Elispot assays, AKB and DMOG, reduced the alloreactive interferon (IFN)-γ response in B10.BR anti-B6 (p<0.001) and MHC matched, minor antigen-mismatched B6 anti-BALB.B cultures (p<0.05). Furthermore, AKB reduced the number of gut-infiltrating CD3+ T cells compared to vehicle-treated BMT mice (39 vs 155.1 CD3+ cells/high-power field, n=3, p<.0003). Our results provide the framework to validate the therapeutic potential of PHDi as an intestinal regenerative strategy in mitigating GVHD.


Peters:Aerpio Therapeutics: Employment, Equity Ownership.

Author notes


Asterisk with author names denotes non-ASH members.