A Murine Xenograft Model for Human T Cell Mediated Graft Versus Host Disease.

Murine xenograft models of human T cell (HuT) mediated graft-versus-host-disease (GvHD) are of potential value but limited by poor engraftment and low and variable incidence of clinical GvHD even after injection of >10⁸HuT cells. The NOD SCID β2M null mice (β2 mice) lack macrophage activity, T, B and NK cells and represent an improved target for HuT cell expansion and activation compared to other immunodeficient mouse models. To induce GvHD, sublethally irradiated β2 mice were injected intravenously via the tail vein (iv) or retroorbitally (ro) with human peripheral blood mononuclear cells (huPBMC) or purified HuT (98% purity). β2 mice conditioned with 250cGy and injected iv with huPBMC (10⁷T cells;n=4) or HuT (0.5–2x10⁷T cells;n=28) failed to engraft and did not develop GvHD. In contrast, β2 mice conditioned with 250cGy and injected ro with huPBMC (10⁷T cells;n=11) or HuT cells (10⁷;n=14) exhibited 19% HuT engraftment 2–3 weeks post-infusion and developed weight loss (>20%) consistent with lethal GvHD, with an overall survival of 82% and 21%, respectively, at 5 weeks (p=0.006). Addition of IL-2 (3x10⁵ IU IP/TIW) had no effect on T cell expansion or GvHD. FACS analysis demonstrated HuT infiltration in the spleen (46%), liver (60%), lung (49%), kidney (40%), and bone marrow (11%). Histological analysis showed an extensive and diffuse accumulation of immature lymphocytes in the spleen, thymus and lymph nodes, and a perivascular infiltration in the lung, liver, kidney but not in the skin or gut. The immunohistochemestry confirmed that these cells were HuT (human CD45+ and CD3+). Furthermore, we observed a 10–15 fold increase in the expression of T cell activation markers CD25, CD30, and CD69 in both the peripheral blood and tissues, compared with naive T cells or T cells from mice that did not develop GvHD. We also evaluated the levels of various human cytokines in the serum of the β2 mice using a cytometric bead array multiplex assay. On day 10 after the injection of HuT and before the start of any clinical sign of GvHD, mice that went on to develop lethal GvHD had 90 times higher levels of IFNγ in serum (>5000pg/ml) compared to mice that did not develop GVHD (<62 pg/ml) (p=0.003). Interestingly both had nearly identical numbers of HuT/ul in blood (32+39 and 33+41 HuT/ul) on day 10. We also observed a significant increase in human IL-10 levels and TNFα in mice that developed GvHD. Mice that developed lethal GvHD had a 70 fold increase of HuT/ul in the 3^rd week (1550 versus 22/ul p<0.003). We improved this model by depletion of murine macrophages using clodronate-containing liposomes (clod) administered iv before the HuT injections. Mice injected with 5x10⁶HuT with clod developed lethal GvHD (3/3) on day 15.7+1.5, with 10⁷HuT (3/3) on day 10.3+5.4 and mice injected with 10⁷HuT without clod on day 13.4+5.4 (8/12)(p<0.05). In contrast, RAG2 γ −/ − mice (RAG2) treated in identical fashion to the β2 mice failed to engraft HuT after both iv and ro injection (350cGy). Both increasing radiation doses (350 to 600cGy) and/or the addition of clod iv resulted in significantly enhanced engraftment of HuT and lethal GvHD. CD4/CD8 ratio of HuT cells expanding in RAG2 mice was <1 in sharp contrast to the β2 mice where the ratio was >2.5. Conclusion: NOD-SCID-β2M null xenotransplant model is uniquely permissive for human T cell expansion after sublethal radiation and may be used as a preclinical platform to study the impact of ex-vivo manipulation and genetic modification of human T cell as GvHD.