Donor APCs Promote Gvhd in MHC-Mismatched Transplants by Indirectly Presenting Host Minor Histocompatibility Antigens.

Abstract 689

In MHC-mismatched allogeneic bone marrow/stem cell transplantation (alloBMT), T cells that at least in-part recognize the allogeneic MHC are likely the dominant mediators of acute GVHD. This would explain why far fewer T cells induce acute GVHD in MHC-mismatched as opposed to MHC-matched alloBMT. However, despite this difference, in human recipients of haploidentical allografts, the organ distribution and histologic appearance of GVHD, especially chronic GVHD, are similar. Moreover, late-onset GVHD in MHC-mismatched alloBMT occurs when recipient antigen presenting cells (APCs) have been replaced by donor APCs. We therefore investigated how indirect presentation of host minor histocompatibility antigens (miHAs) by donor APCs contributes to GVHD. To test this, we employed the MHC/miHA mismatched B6 (H-2b) (right arrow) BALB/c (H-2d) strain pairing. We compared GVHD in irradiated BALB/c mice that were reconstituted with B6 CD4 cells and either B6 BM or B6 MHCII-/- BM. We reasoned that if BALB/c miHAs presented by donor B6 APCs were important, GVHD would be reduced in recipients of MHCII-/- BM. As measured by weight loss and survival, this was indeed the case. However, it was possible that receiving MHCII-/- donor BM reduced GVHD for reasons unrelated to the presentation of BALB/c miHAs. To address this we compared GVHD in irradiated B6.C (H-2d) recipients of B6 CD4 cells and either B6 or B6 MHCII-/- BM. This strain pairing shares the same MHC mismatch as does the B6(right arrow)BALB/c model, but lacks miHAs except those encoded by the congenic H-2d region. In contrast to the prior experiments, B6.C recipients of MHCII-/- or wild type BM developed similar GVHD. To further define indirect-presentation of miHAs in MHC/miHA-mismatched alloBMT, we used B6 Marilyn CD4+ T cell receptor transgenic T cells that recognize an IAb-restricted peptide derived from the male DBY protein. Male or female BALB/c mice were irradiated and reconstituted with female B6 BM and female polyclonal B6 CD4 cells along with graded numbers of Marilyn T cells. At day 7 and 14 post transplant Marilyn cells expanded at least 1000-fold greater in male than in female recipients. Marilyn T cell expansion was similar when male antigen was restricted to hematopoietic or nonhematopoietic cells. There was little expansion of Marilyn cells in male BALB/c mice if donor BM was MHCII-/-, confirming that expansion of Marilyn T cells depended on donor APCs presenting DBY. Indirectly primed Marilyn cells were functional as measured by IFN-γ production after restimulation by the DBY peptide in vitro. As we saw reduced GVHD in the B6 (right arrow) BALB/c model in recipients of MHCII-/- BM, we reasoned we should be able to detect polyclonal CD4 cells responding to host miHA presented by donor APCs. That was the case as we found an increase in DBY-reactive polyclonal CD4 cells by ELISPOT in male recipients as compared to female recipients at day 7 (1.5 fold for IFN-γ and TNF-αa). This difference was lost when donor BM was MHCII-/-. In summary, indirect presentation of host miHAs by donor APCs in MHC-mismatched alloBMT is surprisingly efficient and alloreactive CD4 cells that target host miHAs presented by donor APCs contribute to GVHD in MHC/miHA-disparate alloBMT. The generation of T cells that target host miHAs presented by donor APCs allows GVHD to persist even when host APCs that prime T cells that recognize the allogeneic MHC have been eradicated. This provides a mechanistic explanation for why late onset GVHD, including chronic GVHD, in MHC-mismatched alloBMT resembles that in MHC-matched alloBMT as both may target host miHAs indirectly presented by donor-derived APCs. These data also provide further rationale for targeting donor-derived APCs as a method for preventing and treating GVHD.