Methylguanine Methyltransferase-Based In Vivo Selection Results in Only Transient Improvement in Long-Term Marking after Autologous Transplantation of Transduced Hematopoietic Stem Cells in Rhesus Macaques.

In vivo selective survival advantage of transduced cells contributed to clinically beneficial levels of genetic correction of lymphocytes following X-SCID gene therapy. For most blood disorders there will be no constitutive selective advantage of the gene-corrected cells. Alternatively, a selectable gene incorporated into the vector may provide selective survival advantage. The P140K mutant of human O6-methylguanine-DNA methyltransferase (MGMT*) is a candidate mammalian selectable gene for hematopoietic stem cell (HSC) gene therapy. AMD3100-mobilized CD34+ cells from 5 rhesus macaques were transduced daily from day 2 to 4 of culture using oncoretroviral (n=2 animals) or lentiviral (n=3 animals) vectors encoding the gp91phox-IRES-MGMT* cassette or the GFP-MGMT* fusion protein, respectively. Transduced CD34+ cells were selected after (in vivo, n=4) or before (ex vivo, n=1) autologous transplantation in rhesus macaques using the BG (120mg/m2)/TMZ 400 mg/m2 combination for in vivo selection and the BG (5uM)/BCNU (7.5uM) combination for ex vivo selection. Marking of peripheral blood (PB) cells was evaluated by FACS and/or real-time PCR. Bulk CD34+ cells were marked at 27–58% after transduction with oncoretroviral or lentiviral vectors. Four animals were transplanted with transduced non-selected CD34+ cells. Small fractions of cultured cells not transplanted were exposed to BG/BCNU resulting in an increase of marking to 88–97% in each case, confirming the in vitro survival advantage. Cells from animals #1 and #2 were transduced with oncoretroviral vectors and steady-state marking of 3.5% was obtained in PB. Animal #1 received BG/TMZ infusions at 3 and 6 months post-transplant. Marking declined to 3.3% and 1.1% after BG/TMZ treatment 1 and 2, respectively. Animal #2 received one cycle of BG/TMZ at 4 months post-transplant. Full hematopoietic recovery was not achieved and the animal died of infectious complications one month after treatment. Marking of 2% was detected in the PB at the time of death. Cells from animals #3 and #4 were transduced with lentiviral vectors. Animal #3 received 4 monthly infusions of BG/TMZ starting 5 months after transplantation. Marking increased from 0.1% at steady-state to 1.8% in PB after the first cycle but rapidly declined to 0.2%. Despite significant myelosuppression, additional cycles of BG/TMZ resulted in no significant improvement in marking. Animal #4 received 4 monthly infusions of BG/TMZ starting 3 months after transplantation. Marking increased from 3.3% at steady-state to 29.2% after the first cycle but rapidly declined to 6.2%. Each additional cycle of BG/TMZ resulted in a transient increase in marking with a peak increase gradually declining with each cycle. Animal #5 was transplanted with CD34+ cells transduced with lentiviral vector expressing GFP-MGMT* and exposed to BG/BCNU ex vivo before transplantation. At the time of reinfusion, 55% of the cells were vector positive. Stable hematopoietic recovery required one month, compared to an average recovery of 2 weeks in animals transplanted with transduced cells without ex vivo selection. Steady state marking in PB of only 0.7% was detected. These data combined with the theoretic concern that the use of cytotoxic drugs could increase the risk of leukemogenesis in the setting of drug-resistance gene therapy, raise concerns for the clinical applicability of this approach.