Naive and Ex Vivo Activated Human T Cells Generate Consistent Engraftment and Lethal Graft-Versus-Host Disease (GvHD) in NOD SCID β 2M Null Mice: A New Xenogeneic Model for GvHD.

GvHD remains a major cause of morbidity and mortality following allogeneic hematopoietic stem cell transplantation and donor lymphocyte infusion. The human GvHD pathophysiology includes recipient tissue destruction and proinflammatory cytokine production associated with the conditioning regimen; donor T cells become allo-activated, proliferate, and mediate tissue injury in various organs, including the liver, skin, and gut. Modern therapeutic strategies to control GvHD while maintaining the beneficial graft-versus-leukemia effects require ex vivo T cell stimulation and expansion. Multiple studies have demonstrated that these ex vivo expanded T cells exhibit decreased survival and function in vivo, including reduced alloreactivity and GvHD potential. Unfortunately no in vivo models exist to consistently examine the impact of ex vivo manipulation of human T cells (HuT) on T cell function. Naive HuT were compared to HuT activated using CD3/28 beads (XcyteTMDynabeads) with 50 U/ml IL-2 for 4 days (Act). We initially evaluated the HuT engraftment and GvHD potential of naive and Act in RAG2γ null mice (n=22) conditioned with clodronate liposomes on day −1 and 350cGy on day 0, as previously described by others. We injected 10⁷ and 1.5x10⁷ naive or Act HuT intravenously (iv). All mice exhibited low HuT engraftment and no lethal GvHD. NOD SCIDβ 2M null mice (β 2M) were next conditioned with 250cGy on day −1 (n=34), or 300cGy on day 0 (n=21). 10⁷ naive vs Act HuT were injected retroorbitaly (ro). Lower HuT doses or iv injection resulted in no expansion or GvHD. Engraftment of HuT in peripheral blood of recipient mice was evaluated weekly by FACS and euthanasia was performed if mice lost > 20% body weight. 60% of the mice conditioned with 250cGy that received naive HuT developed lethal GvHD, in comparison to 75% of mice that received 300cGy and nave HuT, and 100% of mice that received 300cGy and Act HuT.

All mice receiving 300cGy had well preserved CD4/CD8 ratios (1–1.5). Tissue infiltration was greatest in mice that had received 300cGy and Act HuT (spleen, liver, lung, kidney: 50–70%). Of interest, serum levels of hu IFNγ dramatically increased over time in all mice who went on to develop lethal GvHD (day 3=270 ug/ml and day 15=36,000 ug/ml) compared to mice that did not develop lethal GvHD (day 10=40 ug/ml and day 17=1,020 ug/ml)(p<0.05). Interestingly, the up-regulation of the activation markers CD25 and CD30 in HuT, and IFNγ production predicted lethal GvHD in β 2M null mice. In summary, we developed a xenogeneic model of lethal GvHD where naive or ex vivo Act HuT injected ro in sublethaly irradiated β 2M not only engraft, expand in vivo, but also infiltrate and damage different mouse target organs. HuT are allo-activated against mouse antigens and damage the target tissues, sharing the major characteristics of human GvHD and causing the death of mice. This model will allow us to study the effects of specific ex vivo T cell manipulation including transduction, selection, expansion, and the depletion or addition of various T cells and other cellular subsets on the outcome of GvHD, to determine improved therapeutic interventions.