Embryonic Development of the Deer Mouse, Peromyscus maniculatus

Deer mice, or Peromyscus maniculatus, are an emerging model system for use in biomedicine. P. maniculatus are similar in appearance to laboratory mice, Mus musculus, but are more closely related to hamsters than to Mus. The laboratory strains of Peromyscus have captured a high degree of the genetic variability observed in wild populations, and are more similar to the genetic variability observed in humans than are laboratory strains of Mus. The Peromyscus Genetic Stock Center at the University of South Carolina maintains several lines of Peromyscus harboring mutations that result in developmental defects. We present here a description of P. maniculatus development from gastrulation to late gestation to serve as a guide for researchers interested in pursuing developmental questions in Peromyscus.


Introduction
Members of the genus Peromyscus are widely distributed throughout North America, and are represented by several species of mice, including deer mice (P. maniculatus), white footed mice (P. leucopus), and oldfield mice (P. polionotus). While superficially resembling Mus musculus, the common laboratory mouse, Peromyscus are more closely related to hamsters than to Mus or Rattus (rats), having shared a last common ancestor with Mus and Rattus approximately 25 million years ago [1]. Peromyscus have many similar characteristics as Mus, including a small size (approximately 25 g) and a short generation time for mammals, which makes them a useful model system for laboratory studies [2]. Peromyscus typically live four to five years in captivity, and remain reproductive for at least two years [3]. Additionally, laboratory strains of Peromyscus have maintained the genetic variability of wild populations and are, therefore, more similar to the genetic variability observed in humans than Mus [4]. Extensive reviews of Peromyscus as a model system expands on these and other topics and serve as an excellent introduction to Peromyscus research [3,5]. The Peromyscus Genetic Stock Center (PGSC) at the University of South Carolina facilitates the use of Peromyscus laboratory research by supplying five species of Peromyscus and several lines of Peromyscus maniculatus harboring unidentified mutations that cause coat color, neurologic, or developmental defects (http:// stkctr.biol.sc.edu). Laboratory strains of three additional species of Peromyscus and four subspecies of P. maniculatus are housed at the University of New Mexico, the University of Illinois Urbana-Champaign, and Harvard University [5].
The natural genetic variation between species and subspecies of Peromyscus is a rich genetic resource that has been instrumental in determining sequence variations that result in phenotypic changes in wild populations. Two species of Peromyscus, P. maniculatus, and P.polionotus, can produce viable, fertile offspring when hybridized. The generation of hybrid offspring has enabled the production of a Peromyscus genomic linkage map and the ability to map genetic loci linked to specific behaviors or phenotypes [6,7]. For instance, burrowing behavior differs between the two species such that P. polionotus dig burrows with longer entrance tunnels and an escape tunnel, while P. maniculatus dig shorter entrance tunnels and do not dig an escape tunnel [8]. F 1 hybrids between the two species dig burrows similar to P. polionotus, indicating that the P. polionotus burrowing behavior is a dominant phenotype. Quantitative trait locus (QTL) analysis of a first backcross generation of F 1 hybrids to P. maniculatus demonstrates that entrance tunnel length is determined by three loci, and the presence of an escape tunnel by a single locus [8]. The natural genetic variation between Peromyscus subspecies has been instrumental in identifying single-nucleotide polymorphisms (SNPs) that result in coat color differences between the subspecies. In P. polionotus a QTL analysis between light colored beach mice and dark colored inland mice identified three QTLs that each contain a candidate gene known to affect coat color in Mus [9]. One candidate gene, melanocortin-1 receptor, contains a SNP in the coding sequences that results in the lighter coat color of some of the P. polinoutus subspecies on the Florida Gulf coast [10]. A second candidate gene, Agouti, contains a SNP in a cis-regulatory sequence that shits the dorsal/ventral boundary of expression of Agouti during embryonic development [11]. The increased expression of Agouti in more dorsal locations results in a lighter pigmentation of beach dwelling P. polinoutus [11]. The application of next-generation sequencing technology in P. maniculatus has also identified 10 SNPs in the Agouti locus, which in combination result in the lighter coat color of deer mice found in the Nebraska Sand Hills [12]. The continued identification of Peromyscus genomic sequence variants will be aided by the genomic sequencing projects of several Peromyscus species. The first assembly of the P. maniculatus genome is now available, and draft sequences of P. polionotus, P. leucopus, and P. californicus are available (https://www.hgsc.bcm.edu/peromyscus-genome-project). Some of the P. maniculatus mutations available at the PGSC, including cataract-webbed (cw), dominant spot (S), and tan streak (tns), potentially harbor developmental defects [3]. Cataract-webbed is a homozygous recessive mutation with syndactyly of the middle digits at birth and cataract formation by eight months of age [3,13]. Dominant spot (S/+) animals have a white blaze on the forehead, similar to piebaldism in humans, and are likely to be embryonic lethal when homozygous [3,14]. The similarity in phenotype between dominant spot in Peromyscus and known spotting mutations in humans and Mus suggests a neural crest defect. Tan streak (tns/tns) is characterized by a pigmented stripe running along the dorsal midline and white fur covering the rest of the body [3,15]. This phenotype suggests a failure in melanocyte migration during development, but remains to be verified. Characterization of these developmental mutations or developmental changes caused by natural variation requires knowledge of the developmental timing of P. maniculatus. P. maniculatus gestation is approximately 24 days, compared to 21 days in Mus. Manceau et al. presented an initial developmental time line of P. maniculatus from e10 to e22 [11]. We have extended this initial report by collecting embryos from earlier time points and presenting additional images that follow the development of external structures throughout development. We also present a comparison of P. maniculatus development to M. musculus development in order to serve as a guide for investigators pursuing Peromyscus developmental biology.

Materials and Methods
All experiments were approved by the University of South Carolina Institutional Committee on the Use and Care of Animals.
Male and female BW P. maniculatus (Peromyscus Genetic Stock Center, University of South Carolina) were housed together on a 16 to 8 hour light/dark cycle, and feed food and water ad libitum. Unlike M. musculus, P. maniculatus do not produce a reliably visible copulation plug. A vaginal lavage was performed on females each morning to assay for the presence of sperm. Noon of the day sperm was detected was designated as embryonic day 0.5 (e0.5). Pregnant females were euthanized using CO 2 asphyxiation on each day of embryonic development between e8.5 and e21.5, and embryos were isolated in 1 x phosphate buffered saline, pH7.4 (PBS). Isolated embryos were fixed in 4% paraformaldehyde in PBS overnight at 4 C. Following fixation, embryos were washed in PBS and then dehydrated through 50% ethanol to 70% ethanol and stored at -20 C. For photography embryos were hydrated through 50% ethanol to PBS. 1% agarose in PBS was melted and poured into standard tissue culture plates and allowed to solidify. The agarose pad was covered with PBS and embryos placed on top of the agarose. Pits dug into the agarose with forceps allowed for positioning of embryos to maintain specific positions for photography. All images were collected using a Leica DFC290 HD camera mounted on a Leica M165FC stereomicroscope, using mainly dark field illumination with additional lighting from above as necessary, and bright field illumination of e8.5 and e9.5 embryos.

Results
P. maniculatus embryos were collected on each day of development between e8.5 to e21.5 to provide an initial atlas of external embryonic development (Fig 1 and Fig 2). As with M. musculus, P. maniculatus displays variability between individual embryos within the same litter and between litters collected on the same day of development (Fig 3). P. maniculatus have litters with between 3 and 5 embryos, compared to 6 to 8 embryos per litter for inbred M. musculus strains. Within one litter, size is the common variability between individuals (Fig 3). Developmental timing can vary with between litters collected at the same gestational time ( Fig  3B). For this atlas only one embryo for each developmental day is presented, with the understanding that significant variation in both size and developmental timing can occur on each day of gestation. Because of the similarity between P. maniculatus and M. musculus development, this survey is not intended to be exhaustive, but rather to highlight key developmental time points that may add future research into specific developmental systems.
The morphology of early P. maniculatus embryos (e8.5) is consistent with the egg cylinder stage of M. musculus (e6.5) (Fig 1A and Fig 2A-2C). The definitive embryonic tissues are located on the ventral side of the egg cylinder with extra embryonic tissues located dorsally. Asymmetry of the egg cylinder is observed with formation of the primitive streak at e8.5 (Fig 2C). A limiting furrow, characteristic of the anterior of the embryo, forms between the extraembryonic and embryonic tissues at e8.5 (Fig 1A). Thickening of the inner layer of embryonic tissue at e9.5 is consistent with formation of the neural plate (Fig 1B), preceding head fold formation. The primitive streak has lengthened and is no longer distinct (Fig 2F). An open neural plate both anteriorly and posteriorly is apparent at e10.5, with the anterior end having folded ventrally (Fig 1C). Formation of the cardiac crescent is observed below the head fold (Figs 1C, 2G and 3A), and endodermal pockets begin formation at both anterior and posterior ends of the embryo, (Fig 2G-2I). The allantois has formed at the posterior end, and is visible when extra-embryonic tissue is dissected away (Figs 1C and 2I). The first somites also form at e10.5 ( Fig 4A).
Embryonic turning occurs in P. maniculatus between e10.5 and e11.5, and embryos at various stages of turning were observed (data not shown); however, the embryo presented has completed turning at e11.5 (Fig 1D). By e11.5 the basic body plan is complete. The anterior neural tube has closed and regionalization of the anterior neural tube into forebrain, midbrain, and hindbrain is observed. The posterior neural pore remains open. The optic and otic placodes are also visible. Somitogenesis, which initiated on e10.5 (Fig 3A), has progressed at e11.5 as somites are formed from the presomitic mesoderm. The heart tube has begun looping, and the swellings characteristic of the forming atria and ventricles are observed. The pharyngeal arches are developing, with the first arch readily apparent. At e12.5 the forelimb bud is visible (Fig 1E), while a distinct hindlimb bud forms at e13.5 (Fig 1F). From e13.5 to e21.5 embryonic development is characterized by the growth of these structures (Fig 1F-1I and Fig 2A-2E), each of which will be described in greater detail.

Neural Development
Frontal images reveal an open neural plate at e10.5 ( Fig 5A). One day later, the anterior neural tube has closed and the divisions between the telencephalon, diencephalon, and mesencephalon are apparent (Fig 5B). The paired swellings of the telencephalon are observed at e12.5 ( Fig  5C), and they expand significantly from e13.5 to e15.5, forming the cortex (Fig 5D-5F). The formation of the skull, facial structures, and thickening of the skin then obscure observation of the neural tissue at e16.5 (Fig 5G). An open neural plate is also observed in dorsal images at e10.5 (Fig 6A), and the division between mesencephalon and rhombencephalon is apparent at e11.5 (Fig 6B). At e12.5 the rhombencephalon expands forming the characteristic rhombus shape of the hindbrain, and the division with the midbrain is sharpened (Fig 6C). The rhombic lip thickens at e13.5 and is observable through e15.5, before becoming obscured at e16.5 ( Fig  6D-6G).

Eye Development
The optic cup is visible at e11.5, encircling the optic placode (Fig 7A). These structures become more refined by e12.5, with the first appearance of the retinal pigmented epithelium (RPE) at e13.5 (Fig 7B & 7C). Pigmentation of the RPE continues at e14.5 and the division between the optic cup and the forming lens is more apparant (Fig 7D). These tissues continue to develop in unison through e15.5, with the lens becoming more opaque at e16.5 (Fig 7E & 7F). The eye lids start to form at e16.5, becoming thicker and narrower at e17.5 and e18.5, before closing over the eye at e19.5 (Fig 7F-7I).

Ear Development
The otic placode is visible and has begun invagination by e11.5, and completes invagination by e12.5 (Fig 8A & 8B). The paired placodes are visible adjacent to the hindbrain in dorsal views (Fig 7B & 7C). At e13.5 and e14.5 the otic placode is internalized, and is difficult to observe externally (Fig 8C & 8D). At e15.5 the external ear is first visible as the pinna extends away from the head (Fig 8E). The auditory meatus is visible at e16.5, and the pinna begins to curve rostrally (Fig 8F). The pinna continues to grow, covering more of the auditory meatus from e17.5 to e19.5 until the auditory meatus is completely obscured at e20.5 (Fig 8G-8J). The thickening skin is apparent in the pinna at e21.5 (Fig 8K).

Pharyngeal arch development
The first pharyngeal arch is visible at e11.5 (Fig 9A). It continues to grow at e12.5, and a cleft appears separating the first arch from the swelling of the forming maxilla (Fig 9B). The second pharyngeal arch is distinct at e12.5, followed by a swelling indicating the third pharyngeal arch at e13.5 (Fig 9C). At e14.5 the divisions between the arches are becoming obscured and the third pharyngeal arch is no longer distinct (Fig 9D). By e15.5 the pharyngeal arches are no longer apparent, as craniofacial formation progresses and the upper and lower jaws form (Fig 9E).

Craniofacial development
Lateral views of the pharyngeal arch region highlight the formation of the maxilla and mandible, forming from the first pharyngeal arch, between e12.5 and e15.5 (Fig 9B-9E). Frontal views highlight the forming frontonasal mass. The first indention of the nasal pits occurs at e12.5 and deepen at e13.5 (Fig 5D). The growing frontonasal mass and maxilla narrow the nasal pits to distinguishable nares at e15.5 (Fig 1H). The snout continues to lengthen and rows of vibrissae are visible beginning at e16.5 (Fig 5G), becoming more prominent by e18.5 ( Fig 5I). Whisker growth from the vibrissae is visible at e19.5 with lengthening through e21.5 (Fig 5J-5L).

Discussion
The gross morphology of P. maniculatus development from gastrulation to birth is very similar to the development of M. musculus [16]. The high degree of similarity between the two suggests that P. maniculatus researchers can use the numerous resources of M. musculus development as a general guide for P. maniculatus, using an appropriate time point correction. Table 1 compares select developmental time points in M. musculus with P. maniculatus. Both rodent species have a range of variability in developmental timing for embryos collected on each day of development. Therefore, this table is presented as a guide and not as strict staging criteria. P. maniculatus trails M. musculus development by roughly two days at gastrulation, and is caused in part by the increased length of time spent at the 2-cell stage for P. maniculatus [3]. This two day difference in developmental timing is roughly maintained through limb bud formation. However, the rate of development appears faster in M. musculus, such that an additional half day of development separates P. maniculatus and M. musculus at e14.5 (Mus e12), and an additional full day of development separates the two rodent species at e17.5 (Mus e14.5). Ultimately, a difference of four days separates P. maniculatus and M. musculus at birth, which occurs at e24 and e20, respectively. Postnatal development is also slower in P. maniculatus, as developmental hallmarks such as skin pigmentation, open eyes, weaning, and sexual maturity all occur later than in M. musculus [3].