Phenogenomic resources immortalized in a panel of wild-derived strains of five species of house mice

Abstract

The house mouse, Mus musculus, is a widely used animal model in biomedical research, with classical laboratory strains (CLS) being the most frequently employed. However, the limited genetic variability in CLS hinders their applicability in evolutionary studies. Wild-derived strains (WDS), on the other hand, provide a suitable resource for such investigations. This study quantifies genetic and phenotypic data of 101 WDS representing 5 species, 3 subspecies, and 8 natural Y consomic strains and compares them with CLS. Genetic variability was estimated using whole mtDNA sequences, the Prdm9 gene, and copy number variation at two sex chromosome-linked genes. WDS exhibit a large natural variation with up to 2173 polymorphic sites in mitogenomes, whereas CLS display 92 sites. Moreover, while CLS have two Prdm9 alleles, WDS harbour 46 different alleles. Although CLS resemble M. m. domesticus and M. m. musculus WDS, they differ from them in 10 and 14 out of 16 phenotypic traits, respectively. The results suggest that WDS can be a useful tool in evolutionary and biomedical studies with great potential for medical applications.

Fig. 1

Geographic origin of WDS kept in Studenec; red: musculus, blue: domesticus, green: spretus, brown: macedonicus, light olive: spicilegus; circles: alive WDS, diamonds: extinct WDS. Small green dots indicate the position of the house mouse hybrid zone adapted from^,. The numbers labelling localities correspond with those in Supplementary Table S1. WDS out of the depicted area are KTK (caroli from Thailand), CIM (castaneus from India), TAIG (castaneus from Taiwan), DKN and CKN (castaneus from Kenya), MDG (castaneus from Madagascar), BID, KAK, and TEH (musculus from Iran), KH (musculus from Kazakhstan), MPR (musculus from Pakistan), AH and AH7 (domesticus from Iran), and DOT (domesticus from Tahiti, French Polynesia). The map was created using the Free and Open Source QGIS (https://www.qgis.org/).

Fig. 2

Maximum-likelihood trees for the whole mtDNA dataset (A) and separately for domesticus (B), castaneus (C), and musculus (D) strains. In the domesticus tree, the CLS cluster is shaded. The codes for individual samples indicate group, WDS name, mouse ID, and the number of generations of inbreeding when known.

Fig. 3

Maximum-likelihood tree of the Prdm9 gene; blue: domesticus WDS (including CLS), red: musculus, magenta: castaneus, green: spretus, brown: macedonicus, dark green: spicilegus. The codes for individual samples indicate group, WDS name, mouse ID, and the number of generations of inbreeding.

Fig. 4

Scatterplot of Slx and Sly CN in musculus WDS, domesticus WDS, and CLS. Vertical and horizontal error bars show Poisson distribution-based errors from triplicate measures of Slx and Sly copy numbers, respectively.

Fig. 5

Distribution of phenotypic variation in mouse strains. Morphospaces defined by biplots of body weight vs. relative tail length (100*body length/tail length) in males (A) and females (B); testis weight vs. sperm count (C); and mean litter sizes vs. number of generations delivered per year (D). Small lettered Venn diagrams provide the numbers of overlapping and non-overlapping individuals in domesticus WDS (dom), musculus WDS (mus), and CLS for corresponding graphs labelled with capital letters (A–C).

Fig. 6

Inter-laboratory comparison of Sly CN obtained at the Max-Planck Institute for Evolutionary Biology, Plön, and at CEITEC Masaryk University, Brno. The measures are highly correlated (Pearson’s product-moment correlation = 0.977). A linear model fit with a standard error envelope is depicted in grey. Vertical and horizontal error bars represent Poisson distribution-based errors from triplicate measures of Sly copy numbers in either laboratory.