Efficient Treatment of Murine Acute Graft-Versus-Host Disease with In Vitro Expanded CD4⁺CD25⁺ Regulatory T Cells

Abstract 2987

Acute graft-versus-host disease (GVHD) is a frequent complication after allogeneic bone marrow transplantation (BMT). We previously showed that the adoptive transfer of donor-type CD4⁺CD25⁺ regulatory T cells (Treg) at the time of BMT prevents acute GVHD in murine models. However, the therapeutic potential of donor-derived Treg cells for the treatment of established acute GVHD has not yet been examined in detail. In analogy to potential clinical applications we now tested the capacity of in vitro expanded Treg cells to ameliorate acute GVHD after haploidentical BMT (BALB/c→CB6F1). CD4⁺CD25^highCD62L⁺ Treg cells were purified by FACS and stimulated polyclonally using anti-CD3/CD28-coated beads. Cells expanded on average 130±19-fold (n=7) within 2 wks and maintained high levels of FoxP3 expression (96, 8±0, 8% FoxP3+ cells; n=7) as well as potent immunosuppressive activity in vitro. For the induction of acute GVHD CB6F1 recipients were lethally irradiated and transplanted with 2.5×10⁶ BM cells in combination with 5×10⁶ splenocytes. All animals developed severe GVHD by d11, as revealed by an increase of the GVHD severity score (2.3±0.4 in GVHD animals vs 0±0 in BM controls, p<0.001, n=1–11) and by histological analyses of the gut (score: 7.8±0.4 for the GVHD group vs 0.2±0.2 for BM controls, p =0.046, n=3). When animals with acute GVHD were treated with 5×10⁶ expanded CD4⁺CD25^highCD62L⁺ Treg cells on d11 after BMT, they initially developed progressive GVHD comparable to non-treated GVHD animals, as indicated by weight loss and an increase of the GVHD score. However from d44 post BMT onwards, Treg-treated GVHD animals regained body weight (d44: 75±3% vs 67±2% of initial weight; p <0.05; n=9–10) and their clinical GVHD score (d44: 6±0 vs 4.3±0.4; p <0.05; n=9–10) decreased. While all non-treated GVHD animals succumbed to disease by d67 after transplantation, 50% of Treg-treated GVHD animals survived for at least 100d (p =0, 002; n=16–21). As immune reconstitution and in particular reconstitution of the lymphocyte compartment is impaired in animals with GVHD, we analyzed the effect of Treg therapy on the reconstitution of the lymphoid and myeloid compartment. At d21 after BMT spleen and BM of non-treated as well as Treg-treated GVHD animals were completely lymphopenic as compared to control mice and both organs contained exceptionally high numbers of granulocytes. Unlike non-treated GVHD animals, however, Treg-treated recipients by d60 showed a recovery of the lymphocyte compartment in spleen (10±2.6×10⁶ T cells and 23.5±12.5×10⁶ B cells in Treg-treated vs 3.0±0.6×10⁶ T cells and 1.5±0.4×10⁶ B cells in non-treated GVHD animals vs 26.25±2.6×10⁶ T cells and 63.9±9.1×10⁶ B cells in BM controls) and BM (0.7±0.1×10⁶ T cells and 8.6±4×10⁶ B cells in Treg-treated vs 0.3±0.01×10⁶ T cells and 0.7±0.4 ×10⁶ B cells in non-treated GVHD animals vs 0.4±0.03×10⁶ T cells and 11.2±0.6×10⁶ B cells in BM controls), while the number of granulocytes decreased constantly. Successful treatment with Treg cells was finally accompanied by a reconstitution of the lymphatic system comparable to control mice. Furthermore, successfully treated mice showed only mild histological signs of gut GVHD at d100 that was significantly lower then those in non-treated GVHD animals with end-stage disease (score: 4.2±1 vs 9.9±1.5 in treated vs non-treated animals, p =0.006, n=4–6). Taken together, these results indicate that in vitro expanded natural Treg cells may not only be effective for the prevention, but also for the treatment of acute GVHD after allogeneic BMT.