Differential expression of the Gstp2 gene between the aboriginal species Mus spretus and the laboratory mouse Mus musculus.

Glutathione S-transferases (GSTs) are pivotal phase-II enzymes for detoxification of xenobiotics. Pi-class GSTs play key roles in determining cancer susceptibility. The laboratory mouse Mus musculus (Mm) has two GST-Pi-encoding genes; while MmGstp1 is the counterpart of the unique human and rat Pi-class GST gene, the function of MmGstp2 remains unclear because its expression is almost undetectable in liver and its product lacks activity against typical GST/GST-Pi substrates. Mus spretus (Ms) is an aboriginal mouse species of great interest as a bio-indicator in environmental pollution studies and a reservoir of novel allelic variants and phenotypes. Using absolute real-time RT-PCR, we demonstrate significant differences in the hepatic levels of GST-Pi-encoding mRNAs between both mouse species. Particularly, we found that the Gstp2 gene of M. spretus, unlike its M. musculus counterpart, attains relatively high steady-state level of expression (∼30molecules/pg of total liver RNA in mice dwelling in a non-polluted area). To test whether the interspecies difference in Gstp2 mRNA levels is due, at least in part, to evolutionary divergence in the promoter regions, we (i) sequenced the 5'-flanking regulatory regions of the two Gstp2 genes; (ii) used bioinformatics tools to identify differences in TF binding sites (TFBSs) and cis-regulatory modules; and (iii) extended the in silico results to a cell-based functional assay. We observed high sequence divergence (2.8%) and differences in TFBSs (32.6%) between the two Gstp2 promoters. We also show that constructs harbouring promoter fragments with species-specific cis-regulatory motifs displayed differential luciferase reporter activity, suggesting that these promoter sequence variations may determine, at least in part, the strong difference in Gstp2 mRNA levels between M. musculus and M. spretus. Additionally, the comparative analysis of the coding sequences predicts that the MsGstp2 product may be an active Pi-class GST because of a Pro(12) to Arg(12) substitution. Interestingly, free-living M. spretus mice dwelling at an industrial settlement displayed significantly higher amounts of transcripts for both GST-P1 and GST-P2 than those from a non-polluted area, suggesting that. M. spretus may optimise the response to pollution by co-evolving the expression levels of the two Pi-class GST genes. Overall, our data suggest that MsGstp2 may be one of the genes contributing to the natural resistance of M. spretus, facilitating its adaptation in a wild environment. Further insights into the functional roles of mouse Pi-class GSTs should be gained from the data reported in this work.