Genetic analysis of mouse embryonic stem cells bearing Msh3 and Msh2 single and compound mutations

Abstract

We have previously described the use of homologous recombination and CRE-loxP-mediated marker recycling to generate mouse embryonic stem (ES) cell lines homozygous for mutations at the Msh3, Msh2, and both Msh3 and Msh2 loci (2). In this study, we describe the analysis of these ES cells with respect to processes known to be affected by DNA mismatch repair. ES cells homozygous for the Msh2 mutation displayed increased resistance to killing by the cytotoxic drug 6-thioguanine (6TG), indicating that the 6TG cytotoxic mechanism is mediated by Msh2. The mutation rate of the herpes simplex virus thymidine kinase 1 (HSV-tk1) gene was unchanged in Msh3-deficient ES cell lines but markedly elevated in Msh2-deficient and Msh3 Msh2 double-mutant cells. Notably, the HSV-tk1 mutation rate was 11-fold higher, on average, than that of the hypoxanthine-guanine phosphoribosyl transferase (Hprt) locus in Msh2-deficient cells. Sequence analysis of HSV-tk1 mutants from these cells indicated the presence of a frameshift hotspot within the HSV-tk1 coding region. Msh3-deficient cells displayed a modest (16-fold) elevation in the instability of a dinucleotide repeat, whereas Msh2-deficient and Msh2 Msh3 double-mutant cells displayed markedly increased levels of repeat instability. Targeting frequencies of nonisogenic vectors were elevated in Msh2-deficient ES cell lines, confirming the role of Msh2 in blocking recombination between diverged sequences (homeologous recombination) in mammalian cells. These results are consistent with accumulating data from other laboratories and support the current model of DNA mismatch repair in mammalian cells.

FIG. 1

RT-PCR analysis of Msh3 and Msh2 transcripts. (A) Diagram of wild-type and mutant Msh3 transcripts and the position of PCR primers. Predicted sizes of amplification products are given in base pairs. (B) Agarose gel electrophoresis of amplification products from msh3^−/−, msh2^−/−, and msh2,3^−/− ES cells. M, DNA size standard. (C) Diagram of wild-type and mutant Msh2 transcripts and the position of PCR primers. Predicted sizes of amplification products are given in base pairs. (D) Agarose gel electrophoresis of amplification products from msh3^−/−, msh2^−/−, and msh2,3^−/− ES cells.

FIG. 2

Cytotoxicity of 6TG in mouse ES cell lines. Survival curves of wild-type and mutant ES cells in the presence of various concentrations of 6TG are shown. Datum points represent the average of two experiments. For details, see Materials and Methods.

FIG. 3

Targeting pMUT to the mouse Hprt locus. (A) Schematic diagram of pMUT targeting. Linearization within the Hprt homology region gives rise to an insertion vector. After double selection in G418 and 6TG, a proportion of the clones were targeted with a single copy of pMUT, resulting in the duplication of the target sequences. Hprt exons are represented by numbered boxes. BamHI restriction sites are indicated as Bam. The internal hybridization probe is shown as the striped box. BamHI size fragments are indicated in kilobases. (B) Southern blot analysis of BamHI-digested genomic DNA from wild-type (wt) and G418/6TG^r clones by using the internal pMUT probe. Correctly targeted clones containing a single copy of pMUT display a novel band of the predicted size equal in intensity to the wild-type band. Clones displaying hybridization patterns consistent with multicopy and random integration are included for comparison. M, DNA size marker.

FIG. 4

Targeting frequencies of isogenic and nonisogenic Hprt constructs. Isogenic (I) and nonisogenic (NI) Hprt constructs were electroporated into wild-type, msh3^−/−, msh2^−/−, and msh2,3^−/− ES cells, and their targeting frequencies were calculated as described in Materials and Methods. Targeting frequencies with isogenic constructs are normalized to 100%.