Small-molecule BCL6 inhibitor effectively treats mice with nonsclerodermatous chronic graft-versus-host disease

Chronic graft-versus-host disease (cGVHD) is a leading cause of long-term morbidity and mortality after allogeneic hematopoietic stem cell transplantation.¹  Animal models have allowed for greater understanding of the pathology of disease and have been instrumental in developing therapeutic interventions for patients. No 1 model faithfully recapitulates the entire range of clinical, pathophysiological, and immune-mediated events seen in human cGVHD; thus, several preclinical models have been developed to represent various patient characteristics including cGVHD with or without scleroderma (generally antibody mediated).² 

For cGVHD without scleroderma, bone marrow (BM) with low T-cell doses given to conditioned allogeneic recipients can cause chronic T-cell–mediated antigenic stimulation and coordinated interactions of T follicular helper (TFH) cells, germinal center (GC) B cells, and T follicular regulatory (TFR) cells.³  The net effects are GC formation, plasma cell generation with antibody deposition, and subsequent lung, liver, and gut, but not skin, fibrosis with bronchiolitis obliterans (BO) as a prominent manifestation.³  This model simulates active cGVHD patients who have circulating TFH cells with an activated phenotype, increased CXCL13 indicative of TFH cells, and the capacity to promote B-cell maturation.⁴ 

The transcriptional repressor B-cell lymphoma 6 (BCL6) is a master regulator of GC reactions, essential for development and function of TFH, TFR, and GC B cells.^5-10  BCL6 has unique roles in each cell type. BCL6 allows GC B cells undergoing somatic hypermutation and DNA double-stranded breaks during class-switch recombination to better tolerate this stress by suppressing DNA damage responses and checkpoint genes.¹¹  BCL6 also regulates pathways in the B-cell receptor (BCR) and CD40 signal transduction cascades in mature B cells.¹²  In TFH cells, BCL6 represses promoters involved in T-cell function, specifically controlling cell migration and alternative cell-fate inhibition.¹³  Mice deficient in Bcl6 are unable to form GCs and therefore do not produce high-affinity antibodies.¹⁴ 

We assessed the requirement of BCL6 expression in both donor T cells and B cells, as sources of BM-derived GC and splenic-derived TFH precursors, respectively, in a murine BO cGVHD model.³  Furthermore, we used a small-molecule, peptidomemitic BCL6 inhibitor, 79-6, for treating established disease in both BO and sclerodermatous cGVHD models.

C57BL/6 (B6; Charles River) and B10.BR mice were housed in a pathogen-free facility and used with institutional animal care committee approval. B6→B10.BR (BO cGVHD) and B10.D2→Balb/c (scleroderma cGVHD) models, including disease severity assessments, were used as described.^3,15,16  For BO cGVHD, cyclophosphamide-treated (120 mg/kg per day, days −3 and −2), irradiated (8.3 Gy by radiograph, day −1) recipients received, on day 0, B6 T-cell–depleted (TCD) BM ± 0.75 × 10⁵ purified splenic T cells. Where indicated, BM or splenic T cells from BCL6^fl/fl × CD19-Cre or BCL6 knockout (KO) mice was compared with wild-type (WT) cells. For scleroderma cGVHD, irradiated (7 Gy, day −1) recipients received B10.D2 BM ± 1.8 × 10⁶ CD4⁺ and 0.9 × 10⁶ CD8⁺ T cells on day 0.

Pulmonary function tests assessing cGVHD-associated BO were performed as described.³  Flow cytometry, fluorescent microscopy, trichrome staining,^3,15  histopathology,¹⁷  and skin scoring for the scleroderma model¹⁸  were performed as described.

Given the importance of BCL6 in regulating GC reactions in response to foreign antigen exposure, we sought to determine whether BCL6 expression in donor T or B cells is required for the GC reactions in murine cGVHD. B10.BR mice were transplanted with WT BM and WT or BCL6 KO T cells. Recipients of KO T cells did not develop BO pulmonary dysfunction (Figure 1A) and had reduced GC B-cell and TFH cell frequencies (Figure 1B). Pulmonary macrophage infiltration along with antibody deposition in the lung results in pulmonary fibrosis and BO.^19,20  Pulmonary collagen and immunoglobulin deposition were reduced in mice receiving BCL6 KO T cells (Figure 1C-D). These results agree with earlier findings that interleukin-21 (IL-21) KO and ICOS-KO donor T cells do not cause cGVHD,¹⁵  consistent with donor TFH support of GC formation as disease initiators.

To examine the role of BCL6 in donor BM-derived B cells, B10.BR mice were given WT or BCL6^fl/fl × CD19-cre (BCL6 KO) BM plus or minus WT T cells. Mice receiving BCL6 KO BM and WT T cells did not develop pathogenic pulmonary dysfunction (Figure 1E). GC B-cell and TFH frequencies (Figure 1F) and lung collagen deposition (Figure 1G) were significantly decreased. In addition, there was a trend toward statistically reduced (P = .065) immunoglobulin deposition in the lungs (Figure 1H). These data indicate that BCL6 in T and B cells is required for development of BO cGVHD and that inhibiting either population could be beneficial for therapy.

The finding that BCL6 expression is necessary in donor cells for cGVHD development suggests that targeting BCL6 might be therapeutically advantageous. In diffuse large B-cell lymphomas (DLBCLs), BCL6 expression is often upregulated and the gene signature is similar to that of a GC B cell. Computer-aided drug design led to the development of compound 79-6, which binds to the Broad-complex, Tramtrack, and Bric-abrac (BTB) domain of BCL6, the site of corepressor (silencing mediator of retinoid or thyroid hormone receptors, nuclear receptor corepressor, BCL6 corepressor) binding critical for repressor activity and DLBCL cell survival.¹⁸  As expected by highly specific BCL6 vs non-BCL6 BTB domain binding, 79-6 kills DLBCL cells in vitro and in vivo.²¹ 

B10.BR mice received WT BM and WT T cells. On day 28, after cGVHD was established, mice were treated with 79-6 or vehicle. Mice treated with 79-6 had improved pulmonary function (Figure 2A). Treatment with 79-6 also resulted in significantly decreased splenic GC B-cell frequencies (Figure 2B), and plasma cell frequencies in the lung (data not shown). Correspondingly, there was a trend toward decreased immunoglobulin deposition in the lungs of mice treated with 79-6 (P = .08; Figure 2D). Pulmonary collagen deposition was less in 79-6–treated mice (Figure 2E). However, treatment did not result in the improvement in damage observed in liver and colon associated with disease, as assessed by histological scoring¹⁷  (Figure 2F). Splenic TFH frequencies were not significantly reduced in 79-6–treated mice, although TFR-to-TFH ratios were improved in some mice (Figure 2C). Thus, despite the dependency on BCL-6 expression for donor TFH and TFR generation,²²  the major effect of 79-6 appears to be directly on the GC B cells in this model.

To assess 79-6 efficacy in sclerodermatous cGVHD, BALB/c recipients were given B10.D2 BM ± 2.7 × 10⁶ T cells.^16,23  The systemic and cutaneous inflammatory response caused by pSTAT3⁺CD4⁺ T cells results in a Th1-dependent and Th17-associated sclerodermatous cGVHD.²³  The skin is infiltrated with transforming growth factor β1 (TGF-β1)-expressing mononuclear cells, elevated C-C chemokines, and macrophage infiltration.¹⁶  Our findings indicate that a GC reaction appears unnecessary (Figure 1G) for cGVHD development in this model, agreeing with recent findings.²⁴  No improvement in outcome, as determined by cutaneous score, was seen following 79-6 (Figure 2H). IL-17A and interferon-γ (IFNγ), significant contributors to skin cGVHD in this model, remained increased in treated mice (Figure 2I).

In this study, we have shown that mice receiving BCL6 KO T or B cells failed to develop BO cGVHD. Importantly, a small-molecule BCL6 inhibitor (79-6) was effective in limiting cGVHD in this GC-dependent model. In contrast, in a cytokine-dependent scleroderma model with systemic and cutaneous inflammation, 79-6 was ineffective in treating cGVHD. This provides support that BCL6 inhibition could be a novel treatment or prophylaxis for cGVHD and potentially other immune diseases in which the GC has a role.

The authors thank Marc Jenkins for providing BCL6 KO cells used for T-cell studies.

This work was supported by National Institutes of Health grants P01-CA142106 (National Cancer Institute), PO1 AI 056299 (National Institute of Allergy and Infectious Diseases), and T32 CA009138 (National Cancer Institute) as well as grants from the Leukemia & Lymphoma Society of America (Translational Research Grants 6458-15 and 6462-15).

Contribution: K.P. designed experiments and wrote the paper; R.F. designed and performed experiments, discussed results, and edited the paper; J.Q. and J.E.B. were responsible for 79-6 synthesis; J.D., L.L., I.M., K.P.M., G.R.H., J.S.S., W.J.M., P.T.S., A.H.S, D.M., C.S.C., J.K., J.H.A., R.J.S., and J.R. discussed results and edited the paper; A.M.M. provided BCL6^fl/fl × CD19-cre BM; and B.R.B. contributed to experimental design, discussed results, and edited the paper.

Conflict-of-interest disclosure: J.E.B. is now an executive and shareholder of Novartis AG, and has been a founder and shareholder of SHAPE (acquired by Medivir), Acetylon (acquired by Celgene), Tensha (acquired by Roche), Syros, Regency, and C4 Therapeutics. The remaining authors declare no competing financial interests.

Correspondence: Bruce R. Blazar, University of Minnesota, MMC 109, 420 Delaware St SE, Minneapolis, MN 55455; e-mail: blazar001@umn.edu.