Rodents of the genus Peromyscus are among the most common North American mammals. Crosses between natural populations of two of these species, P. maniculatus (BW) and P. polionotus (PO), produce parent-of-origin effects on growth and development. BW females mated to PO males produce growth-retarded offspring. In contrast, PO females mated to BW males produce overgrown but dysmorphic conceptuses. Variation in imprinted loci and control of genomic imprinting appear to underlie the hybrid effects. Prior morphological and genetic analyses have focused on placental and post-natal growth. Here, we assess the frequency and scope of embryonic defects. The most frequent outcome of the PO x BW cross is death prior to embryonic day 13. Conceptuses lacking an embryo proper are also observed as in gestational trophoblast disease. Among the common embryonic phenotypes described and tabulated are edema, blood vessel enlargement/hemorrhaging, macroglossia, retention of nucleated erythrocytes, placentomegaly. We investigate expression of loci known to be mis-regulated in human growth/placental disorders and/or mouse knockouts with similar phenotypes. These loci are Igf2, Cdkn1c, Grb10, Gpc3, Phlda2 and Rb1. All exhibited significant differences in either placental or embryonic expression levels at one or more of the three timepoints examined. The data underscore the importance of placental gene expression on embryonic defects. We suggest that the hybrid defects offer a novel system to understand how natural allelic combinations interact to produce disease phenotypes. We propose that such interactions and their resulting epimutations may similarly underlie the phenotypic and causal heterogeneity seen in many human diseases.

Deer mice (Peromyscus) are the most common native North American mammals, and exhibit great natural genetic variation. Wild-derived stocks from a number of populations are available from the Peromyscus Genetic Stock Center (PGSC). The PGSC also houses a number of natural variants and mutants (many of which appear to differ from Mus). These include metabolic, coat-color/pattern, neurological, and other morphological variants/mutants. Nearly all these mutants are on a common genetic background, the Peromyscus maniculatus BW stock. Peromyscus are also superior behavior models in areas such as repetitive behavior and pair-bonding effects, as multiple species are monogamous. While Peromyscus development generally resembles that of Mus and Rattus, prenatal stages have not been as thoroughly studied, and there appear to be intriguing differences (e.g., longer time spent at the two-cell stage). Development is greatly perturbed in crosses between P. maniculatus (BW) and Peromyscus polionotus (PO). BW females crossed to PO males produce growth-restricted, but otherwise healthy, fertile offspring which allows for genetic analyses of the many traits that differ between these two species. PO females crossed to BW males produce overgrown but severely dysmorphic conceptuses that rarely survive to late gestation. There are likely many more uses for these animals as developmental models than we have described here. Peromyscus models can now be more fully exploited due to the emerging genetic (full linkage map), genomic (genomes of four stocks have been sequenced) and reproductive resources.

Although laboratory-reared species of the genus Peromyscus - including deer mice - are used as model animals in a wide range of research, routine manipulation of Peromyscus embryogenesis and reproduction has been lagging. The objective of the present study was to optimize conditions for oocyte/embryo retrieval and for in vitro culturing. On average, 6.4 oocytes per mouse were recovered when two doses of 15 IU of PMSG was given 24 h apart, followed by 15 IU of hCG 48 h later. Following this hormone priming, females mated overnight with a fertile male yielded an average of 9.1 two-cell stage embryos. Although two-cell stage embryos developed to 8-cell stage in KSOM media in vitro, but not further, embryos recovered at the 8- to 16-cell stages developed into fully expanded blastocysts when cultured in M16 media in vitro. These blastocysts had full potential to develop into late stage fetuses and possibly into live pups. As a result of the present work, all stages of Peromyscus pre-implantation development are now obtainable in numbers sufficient for molecular or other analyses. These advances provide the opportunity for routine studies involving embryo transfer (e.g. chimeras, transgenics), and preservation of genetic lines by cryopreservation.

BackgroundCrosses between natural populations of two species of deer mice, Peromyscus maniculatus (BW), and P. polionotus (PO), produce parent-of-origin effects on growth and development. BW females mated to PO males (bw×po) produce growth-retarded but otherwise healthy offspring. In contrast, PO females mated to BW males (PO×BW) produce overgrown and severely defective offspring. The hybrid phenotypes are pronounced in the placenta and include PO×BW conceptuses which lack embryonic structures. Evidence to date links variation in control of genomic imprinting with the hybrid defects, particularly in the PO×BW offspring. Establishment of genomic imprinting is typically mediated by gametic DNA methylation at sites known as gDMRs. However, imprinted gene clusters vary in their regulation by gDMR sequences.Methodology/Principal FindingsHere we further assess imprinted gene expression and DNA methylation at different cluster types in order to discern patterns. These data reveal PO×BW misexpression at the Kcnq1ot1 and Peg3 clusters, both of which lose ICR methylation in placental tissues. In contrast, some embryonic transcripts (Peg10, Kcnq1ot1) reactivated the silenced allele with little or no loss of DNA methylation. Hybrid brains also display different patterns of imprinting perturbations. Several cluster pairs thought to use analogous regulatory mechanisms are differentially affected in the hybrids.Conclusions/SignificanceThese data reinforce the hypothesis that placental and somatic gene regulation differs significantly, as does that between imprinted gene clusters and between species. That such epigenetic regulatory variation exists in recently diverged species suggests a role in reproductive isolation, and that this variation is likely to be adaptive.

Crosses between two species of deer mouse (Peromyscus) yield dramatic parent-of-origin effects. Female P. maniculatus (BW) crossed with male P. polionotus (PO) produce animals smaller than either parent. PO females crossed with BW males yield lethal overgrowth that has been associated with loss-of-imprinting (LOI). Previously, we mapped two loci influencing fetal growth. These two loci, however, do not account for the LOI, nor for the dysmorphic phenotypes. Here we report that maternal genetic background strongly influences the LOI. Analyses of crosses wherein maternal genetic background is varied suggest that this effect is likely due to the action of a small number of loci. We have termed these putative loci Meil. Estimation of Meil loci number was confounded by skewed allelic ratios in the intercross line employed. We show that the Meil loci are not identical to any of the DNA methyltransferases shown to be involved in regulation of genomic imprinting.