Passenger mutations and aberrant gene expression in congenic tissue plasminogen activator-deficient mouse strains

Abstract

Essentials C57BL/6J-tissue plasminogen activator (tPA)-deficient mice are widely used to study tPA function. Congenic C57BL/6J-tPA-deficient mice harbor large 129-derived chromosomal segments. The 129-derived chromosomal segments contain gene mutations that may confound data interpretation. Passenger mutation-free isogenic tPA-deficient mice were generated for study of tPA function.

Summary: Background The ability to generate defined null mutations in mice revolutionized the analysis of gene function in mammals. However, gene-deficient mice generated by using 129-derived embryonic stem cells may carry large segments of 129 DNA, even when extensively backcrossed to reference strains, such as C57BL/6J, and this may confound interpretation of experiments performed in these mice. Tissue plasminogen activator (tPA), encoded by the PLAT gene, is a fibrinolytic serine protease that is widely expressed in the brain. A number of neurological abnormalities have been reported in tPA-deficient mice. Objectives To study genetic contamination of tPA-deficient mice. Materials and methods Whole genome expression array analysis, RNAseq expression profiling, low- and high-density single nucleotide polymorphism (SNP) analysis, bioinformatics and genome editing were used to analyze gene expression in tPA-deficient mouse brains. Results and conclusions Genes differentially expressed in the brain of Plat(-/-) mice from two independent colonies highly backcrossed onto the C57BL/6J strain clustered near Plat on chromosome 8. SNP analysis attributed this anomaly to about 20 Mbp of DNA flanking Plat being of 129 origin in both strains. Bioinformatic analysis of these 129-derived chromosomal segments identified a significant number of mutations in genes co-segregating with the targeted Plat allele, including several potential null mutations. Using zinc finger nuclease technology, we generated novel 'passenger mutation'-free isogenic C57BL/6J-Plat(-/-) and FVB/NJ-Plat(-/-) mouse strains by introducing an 11 bp deletion into the exon encoding the signal peptide. These novel mouse strains will be a useful community resource for further exploration of tPA function in physiological and pathological processes.

Keywords: brain; congenic mice; gene targeting; mutation; tissue plasminogen activator.

Fig. 1

The clustering of genes differentially expressed in brains of congenic tPA-deficient mice in the vicinity of Plat is unrelated to the loss of tPA. (A) RNA was extracted from congenic “Melbourne” C57BL/6J-Plat^−/− mice and age and gender-matched C57BL/6J wildtype controls, and whole-genome transcriptome analysis was performed by using expression arrays. The location of Plat and of differentially expressed genes on chromosome 8 are shown. 11 of 13 differentially expressed genes (excluding Plat) in tPA-deficient brains (85 %) are located within a 27 Mbp interval, flanking Plat. (B) RNA was extracted from congenic “Michigan” C57BL/6J-Plat^−/− mice and wildtype littermates and whole-genome transcriptome analysis was performed by RNAseq. The location of Plat and of differentially expressed genes on chromosome 8 are shown. Eight of 32 differentially expressed genes (excluding Plat) in tPA-deficient brains (25%) are located within a 30 Mbp interval flanking Plat. (C). RNA was extracted from “Michigan” C57BL/6J-Plat^−/− mice and “Bethesda” C57BL/6J-Plat^−/− littermates and whole-genome transcriptome analysis was performed by RNAseq. The location of Plat and of differentially expressed genes on chromosome 8 are shown. 7 of 19 differentially expressed genes (excluding Plat) (37%) are located within a 26 Mbp interval flanking Plat. (D) No anomalous clustering of differentially expressed genes in brains of isogenic “Bethesda” C57BL/6J-Plat^−/− mice around Plat on chromosome 8. RNA was extracted from “Bethesda” C57BL/6J-Plat^−/− mice and wildtype littermates and whole-genome transcriptome analysis was performed by RNAseq. The location of Plat and of differentially expressed genes on chromosome 8 are shown.

Fig. 2

Congenic tPA-deficient mice contain large contiguous segments of 129-derived DNA. High density SNP analysis of congenic “Michigan” C57BL/6J-Plat−/− mice (A), congenic “Melbourne” C57BL/6J-Plat−/− mice (B) and isogenic “Bethesda” C57BL/6J-Plat−/− mice (C). Position of SNPs specific for strain 129 mice versus C57BL/6J mice are indicated with red asterisks, and the position of SNPs specific for strain C57BL/6J mice versus strain 129 mice are indicated with black circles.

Fig. 3

Generation of isogenic C57BL/6J-Plat^−/− and FVB/NJ-Plat^−/− mice. (A) De novo generation of Plat null allele. Schematic structure of the proximal part of the mouse Plat gene (top), and the sequence of the signal peptide-encoding exon 2 (upper case letters) and flanking intron sequences (lower case letters) (bottom). The Zfn binding sites are in bold letters, with the Zfn cleavage site in red. The 11 bp deleted in the C57BL/6J and FVB/NJ strains carried forward for analysis are underlined. This deletion introduces a frameshift resulting in the production of a mRNA encoding amino acid 1-17 of tPA fused to a 19 amino acid nonsense peptide. B. Plasminogen-casein zymography. Lanes 1-4; purified human tPA. Lanes 5-8; protein extracts from brains of a litter of FVB/NJ mice containing 2 wildtype (lanes 5 and 8) and 2 Plat^−/− littermates (lanes 6 and 7). C. tPA Western blot. Lanes 1; purified human tPA. Lanes 2-5; protein extracts from brains of a litter of mice containing 2 wildtype (lanes 2 and 5) and 2 Plat^−/− littermates (lanes 3 and 4). D. Active tPA (middle column) and total tPA (right column) protein extracts from brains of a litter of mice containing 2 wildtype (top and bottom row) and 2 Plat^−/− littermates (middle rows).