Contribution of Human T Cell Subpopulations to Xenogeneic Graft-Versus-Host Disease In RAG2^−/−g_c^−/− Mice

Abstract 2542

Graft-versus-host disease (GVHD) and opportunistic infections represent two major causes of morbidity and mortality following allogeneic hematopoietic stem cell transplantation (allo-HSCT). Both of these conditions are respectively mediated or controlled to a large extend by donor-derived T cells. Moreover T cells play a critical role in promoting stem cell engraftment and decreasing the probability of disease relapse. Therefore the separation of the effects of GVHD from the beneficial effects mediated by T cells has been a long-standing challenge of transplantation immunology. The contribution of different subsets of donor T cells to the pathogenesis of GVHD has been studied in different mouse models: both purified naïve CD4⁺ and CD8⁺ T cells were able to induce acute GVHD. On the contrary memory T cells failed to induce GVHD. In order to dissect the contribution of different human T cells, we sought to define a xenogeneic model for GVHD in RAG2^−/−g_c^−/− mice. Newborn RAG2^−/−g_c^−/− mice were intraperitoneally (i.p.) injected with either peripheral blood mononucleated cells (PBMC), immunomagnetically selected CD4⁺ or CD8⁺ cells, or FACS sorted CD45RA⁺CD4⁺ or CD45RO⁺CD4⁺ cells. Mice were weaned 3 weeks after birth and were scored weekly for signs of GVHD and bled for human engraftment evaluation. Both PBMC and CD4⁺ injected groups developed xenogeneic GVHD (x-GVHD) and the survival at 12 weeks was 42% and 53% respectively. The peak of peripheral blood (PB) engraftment was reached between week 6 and 7 and it was about 40% in both groups. Human cells were present in high percentage (>60%) in the spleen and liver of mice that displayed signs of x-GVHD. Moreover in these mice developed detectable thymi and mesenteric lymph nodes (MLN) which are not normally present in the RAG2^−/−g_c^−/− mice and which were populated by human cells. In PBMC injected mice, CD4⁺ and CD8⁺ T cells were the only populations observed with no survival of myeloid, B or NK cells. PBMC or CD4⁺ groups developed GVHD in 55%, while mice adoptively transferred with CD45RA⁺CD4⁺ cells manifested a positive score for x-GVHD (30%) but displayed milder symptoms compared with the PBMC or CD4⁺ groups. The lesser morbidity and mortality correlated with a lower human engraftment in PB (25% at week 7) and in spleen (about 40%). Only 10% of mice injected with CD45RO⁺CD4⁺ cells showed x-GVHD symptoms and the human engraftment level in PB remained below 20%. Finally mice injected with CD8⁺ cells did not develop any x-GVHD symptoms. The human PB engraftment was low (<15%) in these mice and it decreased over time. In summary we established an easy and reliable xenogeneic mouse model suitable for dissecting the contribution of different human T cell populations to x-GVHD. Our results suggest that depletion of CD4⁺CD45RA⁺ naïve T cells and/or adoptive transfer of CD4⁺CD45RO⁺ memory T cells may find use in the allo-HSCT setting, allowing for rapid reconstitution of T cell-mediated immunity while minimizing GVHD.