Msh2 deficiency attenuates but does not abolish thiopurine hematopoietic toxicity in msh2-/- mice.

The amount of MSH2 protein, a major component of the mismatch repair system, was found to differ >10-fold in leukemia cells from children with newly diagnosed acute lymphoblastic leukemia, with a subgroup of patients (17%) having undetectable MSH2 protein. We therefore used a murine Msh2 knockout model to elucidate the in vivo importance of MSH2 protein expression in determining thiopurine hematopoietic cytotoxicity. After mercaptopurine (MP) treatment (30 mg/kg/day for 14 days), there was a significantly greater decrease in circulating leukocytes in Msh2+/+ and Msh2+/- mice when compared with Msh2-/- mice (p < 0.002). Likewise, the decrease in erythrocyte counts was more prominent in mice with at least one functional Msh2 allele. MP doses of more than 50 mg/kg/day for 14 days resulted in treatment-related deaths, but Msh2-/- mice had a significant survival advantage (p = 0.02). Murine embryonic fibroblasts (MEFs) from Msh2+/+ mice also exhibited increased sensitivity to MP when compared with MEFs from Msh2-/- mice (IC50, 3.8 +/- 0.1 microM versus 11.9 +/- 1.3 microM, p < 0.001). After MP treatment, deoxythioguanosine incorporation into DNA was similar in mice and MEFs with each of the Msh2 genotypes. Electromobility shift assay experiments identified an Msh2-containing GT- or GST-DNA-nuclear protein complex in Msh2+/+ but not Msh2-/- MEFs. Together, these findings establish that hematopoietic toxicity in vivo after treatment with mercaptopurine is attenuated but not abolished by MSH2 deficiency.