Chronic graft-versus-host disease (cGVHD) is a life-threatening complication following allogeneic hematopoietic stem cell transplantation (allo-HSCT), which causes late non-relapse morbidity and mortality (1, 2). Although recent studies have advanced the understanding of GVHD pathophysiology, the first-line therapy remains corticosteroids, which can achieve a complete response in only 20%–50% of patients (3, 4). For patients that do not respond to steroid therapy, the mortality rate is high (5). Furthermore, allo-HSCT recipients undergoing broad immunosuppressant therapy are more prone to tumor relapse, infections, and drug-related toxicities. Therefore, novel targeted immunomodulatory strategies are highly warranted to improve clinical outcomes in allo-HSCT.

Chronic GVHD patients with either bronchiolitis obliterans (BO), an obstructive pulmonary disease, or scleroderma complications are poor responders to standard available therapies (6–9). Moreover, complete and durable responses in steroid refractory or dependent cGVHD are infrequent. Hence, multiple preclinical cGVHD models have been developed to represent aspects of the spectrum of cGVHD manifestations and further elucidate cGVHD pathogenesis in an attempt to develop therapeutic strategies (10–15). Since cGVHD with BO has a poor prognosis, a cGVHD BO model was developed as a platform to evaluate new therapies. Toward that end, B10.BR (H2^k) recipients were conditioned with high-dose cyclophosphamide and total body irradiation prior to reconstitution with major histocompatibility complex (MHC)-disparate C57BL/6 (H2^b) bone marrow and a low dose of T cells, recapitulating many of the clinical, functional, and pathological manifestations of cGVHD with BO (13, 16). Chronic GVHD pathogenesis with BO was dependent upon alloreactive donor CD4+ T cells that differentiated into T-follicular helper (Tfh) cells to activate germinal center (GC) B cells, resulting in pathogenic antibody production and disposition onto cGVHD target organs. Monocyte and macrophage recruitment into areas of lung injury results in stimulation of fibroblast release of profibrogenic molecules and fibrosis in target organs except the skin (17). In a minor histocompatibility antigen disparate scleroderma model, BALB/c (H2^d) recipients underwent total body irradiation prior to reconstitution with B10.D2 (H2^d) bone marrow and T cells resulting in fibrotic skin disease mediated by donor T helper 1 (Th1) and Th17 cells (18).

Retinoid X receptors (RXRs) are key members of the nuclear receptor (NR) superfamily due to their diverse roles in modulating various physiological processes (19). RXRs form homodimers (RXR-RXR) and heterodimers with other NRs, namely, retinoic acid receptors (RARs), thyroid hormone receptor, liver X receptors (LXRs), vitamin D receptor, farnesoid x receptor, nuclear receptor related 1 protein (Nurr1, Nr4a2), nerve growth factor IB (Nur77, Nr4a1), and peroxisome proliferator-associated receptors (PPARs) (20). Studies in mice have reported the direct role of RXR in controlling Th1 differentiation with loss of RXRα in CD4+ T cells leading to increased Th1 polarization and interferon gamma (IFN-γ) production (21). Since RXRα is also required to maintain the suppressive function of T regulatory cells (Tregs), RXR agonists could have dual benefits as a therapeutic strategy for controlling inflammatory disorders.

Rexinoids are RXR ligands that selectively bind and activate RXRs. Although rexinoids can have high therapeutic value in treating various metabolic disorders and cancers (22, 23), off-target responses due to RXR promiscuity with NRs other than RARα limited their clinical applications (24–26). For instance, in addition to RARα, Food and Drug Administration (FDA)-approved synthetic rexinoid bexarotene binds with liver-X-receptor (LXR) and PPAR (27–29). IRX4204 is a highly selective rexinoid with a potency of 100-fold more than bexarotene in activating RXRs. Specifically, IRX4204 does not transactivate RXR heterodimers of RAR, LXR, or PPAR (30), consistent with the selective biological function of IRX4204 and pointing to its use as a therapeutic agent. Recent studies have demonstrated the therapeutic efficacy of IRX4204 in controlling murine autoimmune disease and acute GVHD (aGVHD) (20, 31). Administration of IRX4204 in a murine model of multiple sclerosis attenuated the severity of the disease (20). We previously showed that IRX4204 treatment ameliorated aGVHD while retaining graft-versus-tumor (GVT) responses against leukemia and lymphoma cells (31). Since the pathophysiology of cGVHD is distinct from aGVHD, we sought to test the prophylactic and therapeutic efficacy of IRX4204 in two pathologically distinct cGVHD models. Herein, we demonstrate that IRX4204 administration prevents and treats both BO and sclerodermatous manifestations of cGVHD. Mechanistically, IRX4204 impairs pathogenic donor Tfh, Th1, and Th17 differentiation leading to protection from cGVHD.

Identification of novel immunomodulatory therapies to prevent and treat cGVHD remains the unmet clinical need in allo-HSCT. In the current study, we demonstrated that a highly potent rexinoid, IRX4204, prevented and treated cGVHD by impairing pathogenic donor T-cell responses in two preclinical models with distinct pathophysiology. Therefore, targeting the RXR pathway with IRX4204 is a promising therapeutic strategy to reduce cGVHD. Furthermore, IRX4204 impairs GC B-cell generation in mice immunized with NP-OVA.

In the MHC mismatched BO model, IRX4204 treatment prevented and reversed established cGVHD as reflected by improved PFTs. Although we found that IRX4204 was effective in reducing the cGVHD-mediated lung and liver pathology, there was no effect on spleen and colon pathology. The lack of effect may be due to contrasting mechanisms of cGVHD in different organs. We previously demonstrated that Tfh, a subset of CD4+ T cells, plays a key role in promoting BO cGVHD pathogenesis by supporting GC B-cell affinity maturation and differentiation to produce pathogenic antibodies (17). An earlier study reported that RXR activation has been shown to negatively affect Tfh differentiation in a chronic model of inflammation (41). A more recent study reported that RARα signaling enhances Tfh differentiation in an airway inflammation mouse model (42), perhaps through inhibiting IL2Rα upregulation on T cells, as IL2R signaling inhibits Tfh differentiation by repressing BCL6. In the current study, we reasoned that IRX4204 would modulate Tfh differentiation in cGVHD mice by directly impairing donor T-cell responses. Indeed, we found that IRX4204 treatment reduced the frequency of pathogenic Tfh cells, consistent with the known direct suppressive effect on the proliferation of murine and human T cells in vitro (31) and in vivo in IRX4204-treated aGVHD mice (31).

Chronic GVHD patients have a reduced frequency of Tregs (43), which leads to the lack of functional tolerance and subsequent immune dysregulation. Restoring immune tolerance through either expansion of Tregs in vivo or adoptive Treg transfer has been shown to be an effective strategy to reduce cGVHD (44–46). Here, we show the increased frequency of Tfrs when recipient mice were treated with IRX4204 compared with vehicle-treated recipients. IRX4204 is a potent activator of Nr4a2, which has been shown to promote Treg generation and stability (47). Consistent with the Nr4a2 activity, IRX4204 fostered in vitro induced Treg (20, 31) and pTreg generation in aGVHD mice (31).

Studies have shown that Tfrs impair the production of pathogenic antibodies by negatively regulating Tfh and GC B-cells responses (38, 48). Therefore, it is likely that both increased Tfr frequency and Tfr/Tfh ratios in IRX4204-treated mice contribute to the alleviated cGVHD. Furthermore, this cGVHD/BO model depends upon Tfh production of IL-21. IL-21 is counterregulatory to pTreg production, as we reported in an aGVHD model (49), and Treg function (50, 51). Taken together, increased Tfr in IRX4204-treated mice may be due to Nr4a2 pathway activation in donor T cells resulting in suppression of Tfh IL-21 production, fostering Treg suppressor function and the generation of Tregs and Tfrs over Tfhs. By employing the NP-OVA immunization model, we observed that prophylactic IRX4204 reduced the frequency of Tfh and GC B cells, although unexpectedly, we did not observe any significant differences in GC B cells between IRX4204 and vehicle-treated mice in cGVHD/BO mice. However, compared to vehicle-treated mice, IRX4204-prophylactically treated mice had a statistical trend toward smaller GC size.

Scleroderma is a serious and severe fibrosing disorder that occurs in the majority of cGVHD patients affecting the skin, subcutaneous tissue, and fascia (52). In an MHC-matched, multiple minor antigen mismatched scleroderma mouse model (32, 53), IRX4204 treatment as either a prophylactic or therapeutic significantly reduced sclerodermatous cGVHD with improved clinical outcomes. A previous study showed that in a scleroderma cGVHD model, transplantation of either donor IL-17^−/− or IFN-γ^−/− T cells significantly ameliorated the disease (18). In the current study, we observed a reduced frequency of Th1 and Th17 cells in IRX4204-treated recipients, consistent with previous studies showing that IRX4204 impaired Th1 and Th17 differentiation in aGVHD and experimental autoimmune encephalomyelitis, respectively (20, 31). RA signaling is required to maintain the stability of Th1 cells (54). We and others demonstrated that heightened RA synthesis during allo-HSCT exacerbated aGVHD lethality by enhancing Th1 differentiation of donor T cells (55–57). IRX4204 may directly suppress Th1 differentiation via Nr4a2 activation, as Nr4a2 has been shown to repress Th1 lineage commitment (47) or as a result of inadequate RA signaling in donor T cells.

Although RA signaling has been shown to exacerbate aGVHD (55–57), Nishimura and colleagues (18) demonstrated that the in vivo administration of synthetic retinoid attenuated scleroderma cGVHD by reducing the differentiation of Th1 and Th17. Since IRX4204 preferentially activates RXR homodimers, the endogenous RA signaling pathway that is heightened in GVHD may be impaired in cGVHD mice due to the competitive binding of RXRs to the agonist or other receptors, reducing binding with RARs. Whereas the RXR agonist tributyltin inhibited Th17 differentiation that mechanistically may be the consequence of LXR-RXR pathway activation (58), IRX4204 inhibition of Th17 differentiation in vitro and in vivo (20) does not require LXR activation, suggesting that RXR homodimers or other RXR binding partners are involved in controlling Th17 differentiation. Notably, IL-17 is known to support GC formation, ectopic lymphoid follicles, and antibody class switching in mouse B cells (59–61). In cGVHD/BO model, the in vivo administration of neutralizing anti-IL-17 antibody or small molecule RORγt inhibitors given as a therapeutic markedly alleviated cGVHD (62).

Taken together, our results suggest that targeting the RXR pathway with IRX4204 represents a novel immunomodulatory strategy to prevent or treat cGVHD. IRX4204 is currently in phase I and II clinical trials (NCT0154007 and NCT02991651) to treat cancer. Having shown the beneficial effects of IRX4204 in two distinct cGVHD preclinical models, our studies support consideration for clinical testing of IRX4204 in patients who do not respond to FDA-approved drugs for steroid refractory cGVHD.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding authors.

The animal study was reviewed and approved by University of Minnesota.

GT designed and performed research, provided, and analyzed the data, and wrote the paper. MZ, FM, RF, JD, SR, MR and EA performed experiments, analyzed the data and edited the manuscript. APM performed histopathological analysis. MS provided IRX4204 and edited the manuscript. BB designed, organized, and supervised research and edited the paper. All authors contributed to the article and approved the submitted version.

BB receives remuneration as an advisor to Kamon Pharmaceuticals, Five Prime Therapeutics, Regeneron Pharmaceuticals, Magenta Therapeutics, and BlueRock Therapeuetics; research support from Fate Therapeutics, RXi Pharmaceuticals, Alpine Immune Sciences, Abbvie, the Leukemia and Lymphoma Society, the Children’s Cancer Research Fund, and the KidsFirst Fund and is a cofounder of Tmunity. Author MS is shareholder, officer, and director of the company Io Therapeutics, Inc., which is developing IRX4204 for commercialization.The reviewer DW has declared a past co-authorship with one of the authors BB at the time of review.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.