661. Ex Vivo Pre-Selection of Transgenic Mouse Cells Expressing Green Fluorescent Protein-Methylguanine Methyltransferase P144K Fusion Protein Increases Marking Following Transplant into Wild Type Mice

Abstract

Of all drug resistance strategies to enhance numbers of transduced cells in vivo, the most promising studies have used a mutant form of methylguanine methyltransferase (MGMT|[ast]|) which is resistant to inactivation by the irreversible inhibitor, O6-benzylguanine (BG) while still retaining DNA repair activity after alkylation of DNA from 1,3-bis (2-chloroethyl)-1-nitrosourea (BCNU). MGMT|[ast]| allows the use of in vivo selection regimens that combine BG with BCNU or Temozolomide to enhance gene marking. We have previously shown that the GFP-MGMT|[ast]| fusion protein retains MGMT|[ast]| activity and expresses GFP at high fluorescence allowing for easier tracking, and is thus a convenient model system to explore optimization of in vivo selection (Exp. Hematol. 32 (2004) 709). We subsequently engineered a GFP-MGMT|[ast]| transgenic mouse where T-cells, B-cells, granulocytes, and Sca-1+ cells express 40%|[ndash]|95% green fluorescence and used this model to demonstrate in vivo selection with BG/BCNU in wild type C57B6 mice following transplant of transgenic bone marrow (BM). However, we observed significant toxicity to mice associated with the most effective in vivo selection regimens particularly in animals starting out with the lowest percent of GFP-MGMT|[ast]| transgenic cells, suggesting that higher starting levels of primary engraftment of GFP-MGMT|[ast]| cells serves to protect the animals from selection regimen toxicity. To enhance the level of primary engraftment of cells expressing GFP-MGMT|[ast]|, we have begun to explore pre-transplant ex vivo selection methods. Using a 1:4 transgenic to wild-type mixture of BM exposed ex vivo to 20|[mu]|M BG for one hour followed by addition of 40|[mu]|M BCNU for 2 additional hours (cells washed before transplant) we transplanted 1|[times]|10exp6 BM cells into wild type recipients. Different levels of conditioning regimens were tested: myeloablative 900 cGy, subablative 600cGy, and non-myeloablative 300cGy with or without rapamycin (3mg/kg/d for 30 days), given the possibility of rejection of the GFP expressing cells with low dose radiation. Ex vivo selected recipient mice showed 5 to 8 fold enrichment of gene marking compared to the control mice receiving untreated cells (1.90|[plusmn]|0.85%, 4.94|[plusmn]|1.11%, 20.27|[plusmn]|5.59%, and 76.42|[plusmn]|7.29% of GFP positive BM cells vs. 0.38|[plusmn]|0.18%, 1.01|[plusmn]|0.57%, 2.42|[plusmn]|0.79%, 11.13|[plusmn]|1.60% in 300 cGy, 300 cGy with rapamycin, 600 cGy, and 900 cGy irradiated mice respectively). There was no statistically significant difference in peripheral blood counts between the two groups. Notably, the effect of rapamycin was also significant with greater marking in the mice receiving the immunosuppressant in the ex vivo selection recipients (P=0.0003). Finally, there was better survival of the ex vivo selected mice as compared to mice undergoing subsequent in vivo selection. Further experiments are ongoing to determine if combinations of ex vivo and in vivo selection may achieve results superior to either alone.