A Novel Pathway for CD8+ T Cell Activation by Donor Derived Non-Haematopoietic Cells Leading to Acute Allograft Rejection

Abstract

Introduction: Although seminal studies have shown that CD8+ T cells can reject cardiac allografts that lack donor haematopoietic cells (HPCs) through a direct cognate interaction with graft parenchymal cells, the observation from other studies that secondary lymphoid tissue (SLT) is essential for graft rejection creates a paradox yet to be resolved. Here we examine the hypothesis that CD8+ T cell activation by graft parenchymal cells requires acquisition and re-presentation of intact MHC class I alloantigen by recipient antigen presenting cells (APCs) within SLT. If proven, this hypothesis would not only resolve this paradox, but would also represent the first definitive demonstration of the contribution of the semi-direct pathway to graft rejection. Methods: CD8+ T cell-mediated allograft rejection following activation by donor parenchymal cells was examined using a transgenic heterotopic murine heart transplant model. Balb/c cardiac donors were lethally irradiated and treated with depleting antibodies ensuring complete eradication of HPCs, such that parenchymal cells in these grafts were the only source of alloantigen presentation. These HPC-depleted Balb/c (Balb/cHPC-) cardiac allografts were transplanted into: 1) 2C transgenic mice (monoclonal population of CD8+ T cells against Ld MHC class I); 2) Splenectomised aly/aly (aly/alyspl) mice (complete absence of SLT) with adoptive transfer of 2C CD8+ T cells, to investigate the importance of SLT in this pathway; 3) CD11c-DTR transgenic mice, in which host dendritic cells (DCs) are selectively depleted by diphtheria toxin, allowing delineation of the role of these cells in parenchymal alloantigen presentation to CD8+ T cells. Results: 2C transgenic mice rejected Balb/cHPC- grafts as rapidly as non-depleted Balb/c grafts (MST = 5d vs. 4d respectively; p=0.4) suggesting an effective mechanism for parenchymal cell driven CD8+ T cell mediated rejection. Balb/cHPC- allografts showed prolonged survival (>50d) in aly/alyspl mice given 2C CD8+ T cells, whereas in non-splenectomised controls all grafts rejected (MST = 17d; p=0.01) suggesting an essential role for SLT in this pathway. In addition, when aly/alyspl mice were given activated 2C CD8+ T cells (from a 2C recipient of a Balb/c cardiac graft) they rapidly rejected Balb/cHPC- allografts (MST = 7d; p=0.01) indicating that SLT plays a fundamental role in the activation of CD8+ T cells. To examine, therefore, whether recipient DCs were required to present to CD8+ T cells within SLT, CD4+-depleted CD11c-DTR mice given 2C CD8+ T cells were transplanted with Balb/cHPC- allografts. When these mice were also treated with diphtheria toxin to deplete host DCs they demonstrated significantly delayed rejection kinetics when compared with untreated controls (MST = 26d vs. 16d; p=0.02) highlighting a key role for these cells. Conclusion: These results indicate that allorecognition of graft parenchymal cells represents an important mechanism of CD8+ T cell mediated rejection. They support the role of a novel pathway in which recipient DCs acquire intact MHC class I from donor parenchymal cells and present to CD8+ T cells in SLT leading to their activation.