Co-Stimulatory Blockade Permits Transplantation of Human Hematopoitic Stem Cells and HLA Incompatible T Cells in NOD/SCID γ Null (NSG) Mice

Our previous studies have shown the ability of human CD34+ cells to stimulate T cell alloproliferative responses in-vitro.  Here, we investigated the anti-CD34 T cell alloreactivity in-vivo by co-transplanting human CD34+ cells and allogeneic T cells of an incompatible individual into NSG mice.   Human CD34+ cells (2x10⁵/animal) were transplanted with allogeneic T cells at different ratios ranging from 1:50 to 1:0.5, or without as a control.  Engraftment of human CD 45+ cells was analyzed by flow cytometry at 1, 2, 4, or 8 weeks.  Marrow engraftment of human CD45+cells was significantly decreased in mice transplanted at any CD34:T cell ratio compared to control mice that did not receive T cells.  Transplantation of T cells resulted in graft failure since >98% of human CD45+ cells in the marrow and spleen of NSG mice were CD3+.  A normal CD4:CD8 T cell ratio was detected and CD4+ cells were mostly CD45RA+.  Mice transplanted with CD34+ and allogeneic T cells at 1:1 ratio were sacrificed at 1, 2, or 4 weeks to analyze the kinetics of human cell engraftment in the bone marrow and spleen.  At 2 weeks post transplant, the bone marrow showed CD34-derived myeloid cells, whereas the spleen showed only allo-T cells.  At 4 weeks, all myeloid cells had been rejected and only T cells were detected both in the bone marrow and spleen.  Based on our previous in-vitro studies showing that T cell alloreactivity against CD34+ cells is mainly due to B7:CD28 costimulatory activation, we injected the mice with CTLA4-Ig (Abatacept, Bristol Myers Squibb) from d-1 to d+28 post transplantation of CD34+ and allogeneic T cells.  We found that treatment of mice with CTLA4-Ig prevented rejection and allowed CD34+ cells to fully engraft the marrow of NSG mice at 4 weeks with an overall 13 + 7% engraftment of human CD45+ marrow cells (n=5).  These included 53 + 9% myeloid CD33+ cells, 22 + 3% CD14+ monocytes, 7 + 2% CD1c myeloid dendritic cells, and 4 + 1% CD34+ cells while CD 19 B cells were only 3% and CD3+ T cells 0.5 + 1%.   Persistent myeloid engraftment was found at d+56 after stopping CTLA4-Ig on d+28.  Here we demonstrate that costimulatory blockade with CTLA4-Ig can prevent T cell mediated rejection of incompatible human CD34+ stem cells.  We also show that the NSG model can be utilized to test in-vivo immunotherapy strategies aimed at engrafting human stem cells across HLA barriers.  These results will prompt the design of future clinical trials of CD34+ cell transplantation for patients with severe non-malignant disorders, such as sickle cell anemia, thalassemia, immunodeficiencies or aplastic anemia.