Neonatal Bone Marrow Transplantation Following Ex Vivo Gene Transfer and in Vivo Chemoselection for Hereditary Disorders

Transplantation of genetically engineered hematopoietic stem cells (HSC) holds promise for treatment of genetic and other disorders; however, several obstacles restrict the clinical application of this approach. These include leukemogenesis resulting from insertional mutagenesis, the lack of a competitive advantage for transplanted genetically engineered autologous HSCs, and the requirement for the construction and clinical approval of a gene transfer vehicle specific to each disorder. Allogenic transplantation approaches are further limited by a shortage of compatible donors, the toxicity of preparative regimens, immunosuppression, and graft versus host disease (GVHD).

We have established a neonatal transplantation model to develop strategies to overcome these hurdles to HSC transplantation and gene therapy. We have uniquely combined allogenic transplantation during the neonatal period, when tolerance may be more readily achieved, with a positive selection strategy for in vivo amplification of drug-resistant donor HSC. This model system enables the evaluation of tolerance induction mechanisms to neo-antigens, and allogenic stem cell engraftment during immune ontogeny. In this model, bone marrow-derived HSC are transduced ex vivo by lentivirus-mediated gene transfer of P140K-O⁶-methylguanine-methyltransferase (MGMT^P140K) and GFP (MAG vector). The MGMT^P140K DNA repair enzyme confers resistance to benzylguanine, an inhibitor of endogenous MGMT, and to chloroethylating agents such as BCNU. This enables enrichment of donor cells at the stem cell level by in vivo chemoselection. Furthermore, in vivo administration of BG/BCNU may potentially deplete allo-reactive cells of donor and host origin, reducing the need for toxic ablative or immunosuppressive treatment.

The in vivo selection strategy was first tested in neonates in the absence of MHC barriers. Syngenic whole bone marrow (WBM) was transduced with MAG and 5 × 10⁵ cells were transplanted. Five weeks after transplantation and prior to chemoselection, 0.5% of cells expressed GFP in peripheral blood. Following two cycles of in vivo selection we observed 39.5% expression of GFP in peripheral blood. To assess the engraftment of fully MHC-mismatched HSC in neonates without in vivo selection, 10⁷ B10.Br (H-2^k) WBM was transplanted into C57BL/6 (H-2^b) neonates using a non-myleoablative regimen. This resulted in 98% engraftment in adults without development of GVHD. To evaluate in vivo selection in an allogenic setting, 2 × 10⁶ MAG transduced WBM from BALB/c (H-2^d) adult mice were transplanted into C57BL/6 × BALB/c F1 (H-2^b/d) neonatal mice. Despite having a graft versus host MHC disparity, we observed an increase from 4.7% (5 weeks after transplant) to 37.4% GFP expression following two cycles of in vivo chemoselection without evidence of GVHD.

We have demonstrated successful engraftment and enrichment of both syngenic and allogenic HSC expressing MGMT^P140K following neonatal transplantation and in vivo selection. Stable engraftment was achieved without myeloablation or post-transplant immunosuppression. No evidence of GVHD was observed after full MHC-mismatched transplantation. These results hold promise for the development of non-myeloablative allogenic transplant approaches using this in vivo selection strategy.