Very High Expression of MGMT(P140K) Paradoxically Impairs Proliferation, In Vivo Selection and Protection of Transduced Hematopoietic Stem Cells.

The drug resistance gene MGMT(P140K) mutant has been extensively studied as a means to protect/select hematopoietic stem cells (HSC) following treatment with O6-alkylating agents such as temozolomide (TMZ) in tandem with the inhibitor of endogenous MGMT, O6-benzylguanine (6BG). Since the repair of O6-alkyl adducts by MGMT is stoichiometric, it has been suggested that higher levels of MGMT(P140K) will confer better protection upon HSC. To test this hypothesis, we developed a panel of SIN γ-retroviral vectors which express low to very high levels of MGMT(P140K) via different internal promoters. The MGMT activity in bone marrow cells (BMC) transduced with a viral LTR-derived promoter, SF(MGMT), was approximately 7-fold higher than in cells transduced with either cellular promoters PGK(MGMT) or EFS(MGMT) which were in turn >300-fold higher than in control transduced cells (CON). Mice transplanted with transduced BMC were treated with 6BG (30mg/kg) and TMZ (80mg/kg) on 3 consecutive days starting 7 weeks post-transplant. Surprisingly, we found that selection/protection was not evident in mice engrafted with SF(MGMT) transduced BMC, while robust long term in vivo selection/protection was observed in mice transduced with PGK(MGMT) and EFS(MGMT) BMC (Table 1). We next sought to explore the mechanism of these findings. 2 hours after 6BG/TMZ treatment, BMC harvested from SF-, PGK- and EFS(MGMT) groups showed equivalent repair of the O6-alkylguanine lesion. Thus, lower MGMT expression is adequate for O6-alkylguanine lesion repair, and the lack of selection by SF(MGMT) is not related to defective repair. In the absence of chemoselection, we found a modest but statistically significant decrease in the in vivo reconstitution of SF(MGMT)-transduced BMC compared to EFS- or PGK- transduced BMC in a competitive repopulation assay, while secondary recipients showed a more pronounced engraftment defect (18- and 32-fold lower engraftment of SF(MGMT) vs EFS(MGMT), (p<0.05) or PGK(MGMT), (p<0.01) respectively). To further examine this repopulation defect, 32D cells were transduced with SF(MGMT) and demonstrated a growth defect in vitro (Table 2), have >40% reduced colony forming ability (p<0.01) and show >30% reduced 3H thymidine uptake (p<0.01), but do not demonstrate an elevated rate of apoptosis.These data strongly suggest that very high expression of MGMT(P140K) has a deleterious effect upon cellular proliferation, engraftment and chemoprotection. Since clinical gene therapy studies with MGMT(P140K) are underway these studies have direct translational relevance. Table 1. Chemoselection/Protection in vivo mean ± SEM % PBC pre-treatment % change PBC 15 weeks post-treatment WBC (×10e6/ml) pre-treatment WBC (×10e6/ml) 7 weeks post-treatment (**p<0.01 compared to non-treated) CON 49.5±3.8 −11.4±5.1 12.3±1.0 5.4±0.3** SF-MGMT 57.3±2.4 −18.3±7.3 11.7±0.8 8.6±1.1** EFS-MGMT 39.3±3.2 +18.8±4.4** 12.9±1.1 13.5±1.3 PGK-MGMT 46.8±4.4 +27.9±2.9** 12.2±0.8 12.5±1.1 Table 2. Growth of transduced 32D cells in the absence of treatment CON EFS-MGMT PGK-MGMT SF-MGMT (**p<0.01 for SF vs. CON, EFS and PGK) % GFP+ after 28d culture 101.8±14.5 91.6±2.8 98.4±19.4 31.2±3.2**