Clonal Analysis of Hematopoietic Stem Cells after Serial Transplantation of Animals Receiving MGMT^P140K Gene Transduced Cells Following Chemotherapy.

Gene transfer of the DNA repair protein O⁶-methylguanine-DNA-methyltransferase (MGMT) into hematopoietic stem cells has been shown to protect hematopoiesis from the toxic side effects of O⁶-guanine alkylating drugs such as BCNU, ACNU or temozolomide (TMZ). In addition, MGMT gene transfer allows efficient in vivo selection of transduced hematopoietic stem cells and enrichment of genetically corrected cells in the context of gene therapy for monogenetic diseases. We here have analysed the long-term effect of MGMT gene transfer on the hematopoietic stem cell compartment using an in vivo murine transplantation/gene therapy model and a retroviral vector carrying the gene for MGMT^P140K, a mutant resistant to the wtMGMT-specific inhibitor O⁶-benzylguanine (BG). Serial transplants were performed and primary, secondary as well as tertiary recipients were treated with combined BG/ACNU, BG/BCNU or BG/TMZ chemotherapy at doses myeloablative in non-MGMT-protected hematopoiesis. Serial transplantation was performed with 1.8 – 3.0 x 10⁶ mononuclear bone marrow cells and 2 to 3 animals were transplanted per primary or secondary animal.

While initial gene transfer efficiency was low (1–5% of cells engrafted at week four) chemotherapy resulted in efficient selection of transduced cells in primary animals (70–90% transgene expression in peripheral blood). Secondary and tertiary recipients showed 40–80% transgene expression even before CTX. Efficient stem cell engraftment and protection from CTX was demonstrated in > 90% of secondary animals, while tertiary recipients clearly demonstrated compromised engraftment and a substantial number of animals did not survive CTX treatment.

Retroviral vector integration site analysis to study the clonality of hematopoiesis of stem cells by ligation mediated PCR (LM-PCR) was performed in the serially transplanted mice. In three mice of the secondary transplantation cohort we detected 3, 0, and 6 clones, respectively. In three mice of the tertiary transplantation cohort 7, 2, and 2 clones, respectively, were found. Thus, an exhaustion of transduced hematopoiesis following regenerative stress by high dose chemotherapy was not evident. Of the total 20 detected clones one could not be mapped to the mouse genome, while the others could be blasted against the mouse genome (assembly 2004, http://genome.ucsc.edu/; >99.5% identity). It turned out that 5 of 8 integrations landed in RefSeq in the tertiary transplantation cohort, while 3 of 8 integrations occurred in RefSeq genes in the secondary transplantation cohort. This suggests that clones profit from the transcription machinery of their integration site. Thus, our LM-PCR results indicate that the multiclonality of hematopoiesis is conserved after serial transplants which may be considered a safety feature for drug-resistance gene therapy. Furthermore, vector integration in highly resistant stem cells is favored in actively transcribed genomic regions.