Sox Genes Show Spatiotemporal Expression during Murine Tongue and Eyelid Development

The tongue is a critical organ, involved in functions such as speaking, swallowing, mastication, and degustation. Although Sox genes are known to play critical roles in many biological processes, including organogenesis, the expression of the Sox family members during tongue development remains unclear. We therefore performed a comparative in situ hybridization analysis of 17 Sox genes (Sox1–14, 17, 18, and 21) during murine tongue development. Sox2, 4, 6, 8, 9, 10, 11, 12, and 21 were found to be expressed in the tongue epithelium, whereas Sox2, 4–6, 8–11, 13, and 21 showed expression in the mesenchyme of the developing tongue. Expression of Sox1, 4, 6, 8–12, and 21 were observed in the developing tongue muscle. Sox5 and 13 showed expression only at E12, while Sox1 expression was observed only on E18. Sox6, 8, 9, and 12 showed expression at several stages. Although the expression of Sox2, 4, 10, 11, and 21 was detected during all the four stages of tongue development, their expression patterns differed among the stages. We thus identified a dynamic spatiotemporal expression pattern of the Sox genes during murine tongue development. To understand whether Sox genes are involved in the development of other craniofacial organs through similar roles to those in tongue development, we also examined the expression of Sox genes in eyelid primordia, which also contain epithelium, mesenchyme, and muscle. However, expression patterns and timing of Sox genes differed between tongue and eyelid development. Sox genes are thus related to organogenesis through different functions in each craniofacial organ.


Introduction
e tongue plays a critical role in speaking, swallowing, mastication, and degustation. Malformations of the tongue, including macroglossia, hypoglossia, and aglossia, are seen as congenital defects. erefore, it is crucial to understand the molecular mechanisms involved in tongue development; however, these mechanisms remain unclear. e mammalian tongue is composed of the epithelium, connective tissue, and striated muscle. e connective tissue and vasculature of the tongue are derived from the cranial neural crest, whereas most tongue muscles are formed by myoblasts that migrate from the occipital somites [1]. It is believed that there is an interaction between neural crestderived and myogenic cells during tongue development [2,3].
In mice, tongue development begins at embryonic day (E) 10.5-11.0 with the formation of medial lingual swelling from the first branchial arch. Next, lateral lingual swellings are formed on each side of the median tongue bud. ey overgrow the medial lingual swelling and eventually fuse to form the anterior two-thirds of the tongue. e medial lingual swelling does not form any identifiable part of the adult tongue. e third branchial arch gives rise to the copula and the hypopharyngeal eminence. Subsequently, the hypopharyngeal eminence overgrows the copula, which in turn disappears progressively. Consequently, the rostral part of the hypopharyngeal eminence develops into the posterior third of the tongue [2,3,4]. us, the embryonic origin in the anterior two-thirds and posterior third of the tongue differs. In addition, the anterior twothirds of the tongue is mobile, whereas the posterior third is relatively immobile. Tongue muscles are classified as either intrinsic or extrinsic. ey are bilateral and separated by the lingual septum, which consists of fibrous connective tissue. Four different types of tongue papillae can be distinguished: fungiform, circumvallate, foliate, and filiform.
ey are known to develop through epithelial-mesenchymal interactions. Although the lingual epithelium is histologically homogeneous during the early stages of tongue development, the papillae become distinct and protrude from the dorsum of the tongue on approximately E12. A single circumvallate papilla is located at the center of the terminal sulcus. Fungiform papillae are present on the anterior part of the tongue in a pattern of longitudinal rows bracketing the median furrow. Unlike the other three types of papillae, filiform papillae cover the entire dorsal surface of the tongue. Sox proteins are characterized by a highly conserved DNA-binding motif called high-mobility group domain. To date, 20 Sox genes have been identified in mice. Members of the Sox gene family show dynamic and diverse expression patterns during development, and they play multiple roles during development, as evidenced by mutation analyses in mice [5,6]. Spatiotemporal expression patterns of Sox genes have been reported in tooth and pancreas development [7,8]. In the trunk and limb region, it has been reported that Sox6 is involved in skeletal muscle formation, while Sox7, 15, 17, and 18 have been shown to be related to function of satellite cells [9][10][11][12][13]. Although Sox2 has been reported to be involved in the development of tongue papillae [14], the expression pattern of other Sox genes during the entire development of the tongue remains unclear. erefore, we performed a comparative in situ hybridization analysis of 17 Sox genes (Sox1-14, 17, 18, and 21) during murine tongue development and identified their dynamic spatiotemporal expression.

Production and Analysis of Mice.
CD-1 strain mice were used in this study. e noon time of the day on which the plugs were detected was considered as E0.5. To accurately assess the age of the embryos, somite pairs were counted, and the stage was confirmed using morphological criteria, e.g., relative sizes of maxillary and mandibular primordia, extent of nasal placode invagination, and size of limb buds.

In Situ Hybridization.
Embryo heads were fixed in 4% buffered paraformaldehyde, wax embedded, and serially sectioned at 7 µm. Decalcification using 0.5 M EDTA (ethylenediaminetetraacetic acid, pH 7.6) was performed after fixation of the newborn mice. Sections were split over three to ten slides and prepared for histology or radioactive in situ hybridization. In situ hybridization with [ 35 S]UTPlabeled riboprobes (Table 1) was performed as previously described [15]. Briefly, the slides were pretreated with proteinase K and acetic anhydride to reduce the background. Hybridization was carried out overnight in a humidified chamber at 55°C. e slides were then washed and treated with RNAse A for 30 min at 37°C to remove any nonspecifically bound probe. e high stringency washes were repeated, and the sections were then dehydrated. e slides were air-dried and dipped in Ilford K-5 photographic emulsion. Autoradiography was performed by exposing the sections in a light-tight box at 4°C for 10-14 days. Slides were developed using Kodak D19, fixed in Kodak Unifix, and counterstained with malachite green or hematoxylin.

Results
e expression of Sox3 and 14 was not detected in the developing tongue (data not shown).

E11.
e tongue primordium is firstly recognizable on the mandibular processes as a small protrusion (Figure 1(a)). To identify muscle progenitor cells, Myf5 expression was examined in the developing tongue. Two domains of Myf5 expression were observed in the center of the tongue primordium, which was found to be further divided into four domains in the caudal region ( Figure 1(b)). Sox4, 11, and 21 were ubiquitously expressed in the developing tongue ( Figures 1(d), 1(f ), and 1(g)). Sox2 and 10 expressions were observed in the mesenchyme at the lateral side of the tongue primordium, whereas Sox2 also showed the expression in the epithelium (Figures 1(c) and 1(e)). Sox1, 5, 6, 8, 9, and 12-14 showed no expression in the developing tongue at E11 (data not shown).
Since distinct morphological structures are present in the developing tongue along the anterior-posterior axis, it is likely that the expression patterns differ among the various regions of the developing tongue. erefore, we examined the expression patterns of the Sox gene family members in the anterior, middle, and posterior regions of the tongue.

E13.
e vt, gg, and hg muscles were histologically identifiable at E13 (Figures 3(a)-3(a″)). Myf5 expression in the anterior and middle regions of the tongue was comparable to that at E12, while it was divided into two expression domains in the gg muscles (

E18.
Most tongue muscle progenitor cells had already been differentiated into myoblasts by E18. erefore, only weak expression of Myf5 could be observed in the tongue at E18 (data not shown). Sox1 showed weak expression in the sil, vt, gg, and gh muscles in the middle and posterior regions of the tongue (Figures 5(b)-5(b″)). Sox2 has been shown to be expressed in the epithelium at this developmental stage [14].   . Sox2, 5, 9, and 13 showed no expression in the developing tongue at E18 (data not shown). Sox7, 17, and 18 belong to group F of the Sox gene family (SoxF). Members of SoxF have been shown to regulate blood and lymphatic vascular development [16][17][18]. A punctate expression pattern of Sox7, 17, and 18 was seen throughout the developing tongue, and it probably represents vascularization of the tissue (data not shown).

Sox Genes in Eyelid Development.
Among the craniofacial organs, the eyelid shows some similarities with the tongue: Both are mobile organs and develop through epithelium, neural crest-derived mesenchyme, and mesoderm [2,3,19,20]. In addition, it has been shown that the lack of Shh signaling pathway and microRNAs result in developmental defects of both the eyelid and the tongue [21][22][23][24][25].
ese features suggest the possibility that tongue and eyelid development are under similar molecular mechanisms. To understand whether Sox genes are also involved in other craniofacial organs through similar roles as those in tongue development, we examined the expression of Sox genes in eyelid development. Murine eyelid initiates by groove formation at approximately E11, and the protrusion of the eyelid primordia occurs from E12 (Figures 6(a), 6(f ), 6(k), and 6(p)) [20]. Fusion of eyelid is observed at approximately E16, since the mammalian eyes require temporary fusion and re-opening between the upper and lower eyelids during their development and growth (Figure 6(u)). Myf5 expression was observed in the eyelid primordia only at E14 and E18 (Figures 6(a′), 6(f′ ), 6(k′), 6(p′), and 6(u′)). At E11, Sox4, 11, and 21 showed ubiquitous expressions in the developing eyelid, whereas Sox6 was only expressed in the eyelid epithelium (Figures 6(b)-6(e)). At E12 and E13, Sox4 and 11 expressions were observed in both the epithelium and mesenchyme, whereas Sox6 and 9 were only expressed in the epithelium (Figures 6(g)-6(j) and 6(l)-6(o)). At E14, Sox4 was ubiquitously expressed in the developing eyelid, with a slightly stronger expression in the muscles (Figure 6(q)). Sox6, 9, and 12 were expressed in the epithelium of eyelid primordia, whereas Sox11 expression was observed in both the epithelium and mesenchyme (Figures 6(r)-6(t)), data not shown). At E18, Sox21 was ubiquitously expressed in the developing eyelid, with a slightly stronger expression in the muscles (Figure 6(y)). Expressions of Sox4 and 6 could be observed in the muscle of eyelid primordia, whereas Sox11 was expressed in the epithelium (Figures 6(v)-6(x)). Sox4, 6, and 11 were also expressed in the extraocular muscle region. Expressions of Sox1-3, 5, 8, 10, 13, and 14 could not be detected in the developing eyelid (data not shown).

Discussion
e mammalian tongue is composed of the epithelium, connective tissue, and striated muscle. Our comparative in situ hybridization analysis demonstrated a dynamic spatiotemporal expression pattern of Sox genes during murine tongue development. Sox2 and 11 were expressed in the tongue epithelium during all five stages investigated in this study, whereas Sox8 and 9 were expressed in the tongue epithelium only at E14 and E12, respectively. Sox4, 6, 12, and 21 showed expressions in the epithelium at several stages ( Table 2). Sox4 and 21 were expressed in the mesenchyme at all stages, whereas Sox5, 6, and 13 were only detected in the mesenchyme at E12 (Sox5, 13) and E18 (Sox6). Sox2, 8,9,10, and 11 showed detectable expressions in the mesenchyme at several stages (Table 2). Unlike the epithelium and mesenchyme, none of the Sox genes were expressed in the tongue muscle during all five stages. Sox1 was expressed in International Journal of Dentistry the tongue muscle only at E18. Sox6 and 10 were not detected in the muscle at E11 but were expressed during the other four developmental stages (E12-E18). Additionally, Sox11 was not expressed in the muscle at E18 but was expressed during the other four developmental stages (E11-E14). Sox4, 8,9,12, and 21 showed detectable expressions in the muscle at several stages (Table 2). Sox1 was only expressed in the muscle, whereas Sox5 and 13 were expressed only in the mesenchyme. Sox2 could not be detected in the muscle at any stage, whereas Sox12 showed no expression in the mesenchyme at any stage. Although the expressions of Sox2, 4, 10, 11, and 21 were detected in the developing tongue during all five stages investigated in this study, their expression patterns differed among the stages. Our results also revealed a different expression pattern of Sox genes between the anterior, middle, and posterior region of the developing tongue (Figure 7). Sox genes thus showed a dynamic spatiotemporal expression pattern during murine tongue development.
Fungiform, circumvallate, and foliate papillae contain taste buds and are referred to as taste (or gustatory) papillae, whereas filiform papillae do not contain taste buds and are thus regarded nongustatory. Although Sox2 has been shown to play a critical role in regulating the formation of gustatory papillae, its expression has also been detected in the nongustatory tongue epithelium [14]. Our results revealed that Sox2, 4,6,8,9,10,11,12, and 21 were expressed in the epithelium of the entire tongue. Although none of these genes showed a restricted expression pattern in the tongue papillae, it is possible that they are involved in regulating the formation of all papillae types.
Skeletal muscle fibers are classified as "slow-twitch fiber" (type I) and "fast-twitch fiber" (type II), which display marked differences in contraction strength, metabolic strategies, and susceptibility to fatigue. Slow-twitch fibers are rich in mitochondria, have increased contraction endurance with lesser strength potential, and use predominantly oxidative phosphorylation for energy production. Fast-twitch fibers contain comparatively less numbers of mitochondria and rely more heavily on anaerobic glycolysis for energy production, which allows considerable strength and contraction speed, but only for short anaerobic bursts of activity before the muscles fatigue. It has been reported that Sox6 is involved in fast-twitch muscle fiber differentiation in the trunk region [9,10]. It is known that rodent tongue is composed of only fast-twitch fiber, and we also found Sox6 expression in the muscles of the developing tongue [26,27]. ese suggest that Sox6 is likely to be involved in fast-twitch muscle fiber differentiation during tongue development. Sox8, 9,10,12, and 21 were observed in the tongue muscles where Sox6 was expressed, suggesting the possibility that these Sox genes also regulate fast-twitch muscle fiber differentiation. On the other hand, unlike the mouse tongue, the human tongue shows high proportion of slow-twitch fibers, suggesting that the role of Sox genes differs between the human and mouse tongue [28]. e developmental origin of the masticatory muscles is the cardiopharyngeal mesoderm, while the tongue and other muscles, including the trunk and limb muscles, are derived from the somites [29]. Although the embryonic origins are the same in the tongue, limb, and trunk muscle, it has been shown that the program controlling tongue myogenesis, including the specification and migration of muscle progenitor cells, differs from those of the trunk and limb myogenesis [2,3,30]. erefore, we could not exclude the possibility that the role of Sox genes expressing in developing tongue muscle differ from that in trunk and limb muscle development.
e tongue consists of many types of muscles, including vt, sil, and gg muscles [2,30,31]. Although our results indicated that many Sox genes were expressed in these tongue muscles, there was no specific Sox gene expressing in the particular muscles (Figure 7). Sox genes are therefore unlikely to play a critical role in tongue muscle specification. Different expression patterns between Sox genes in each tongue muscle type suggested that each muscle type is controlled by different molecular mechanisms during development.
Skeletal myogenesis consists of six phases: specification, migration, proliferation, determination, differentiation, and  International Journal of Dentistry maturation [1,2,[29][30][31]. It is unlikely that each phase of myogenesis proceeds at the same time in all types of tongue muscle since the timing of appearance differs between muscle types in the tongue [1,2,[29][30][31][32]. It is possible that the different expression patterns of Sox genes in tongue muscles could be correlated with the onset of these stages, although the molecular mechanisms regulating the initiation of each phase during tongue development remain unclear. It has been shown that Sox7, 15, 17 and 18 are involved in the function of satellite cells after birth, although these are not exerted at embryonic stages [11][12][13]. erefore, a punctate expression pattern of Sox7, 17, and 18 observed in the embryonic tongue are likely to represent vascularization of the tissue. Furthermore, these also indicated that function of Sox genes differ between embryonic and postnatal stages in tongue formation. Previous studies have reported that neural crestderived cells give rise to tissues surrounding the skeletal muscles, including the perimysium, epimysium, endomysium, and tendon, during tongue development [33,34]. Sox genes have been shown to regulate the functions of neural crest-derived cells including their differentiation, collagen synthesis, and melanocyte formation [35,36,37]. In our study, Sox9 displayed a restricted expression pattern in the lingual septum, which is composed of neural crestderived cells [2,3]. It has been reported that Scleraxis is expressed in a similar pattern to that of Sox9 during tongue development [2,32]. Interestingly, Sox9 shows coordinated expression with Scleraxis during digital tendon development [38]. erefore, it is possible that Sox9/Scleraxisexpressing cells are involved in the formation of tongue muscle tendons. e origin of the tongue is known to be hybrid, neural crest-derived cells, the mesoderm, and endoderm. It has been shown that there is an interaction between neural crestderived and myogenic cells during tongue development [2,3]. We detected the expressions of Sox1, 4,6,8,9,10,11,12,13, and 21 in the muscles of the developing tongue and of Sox2, 4,5,6,8,9,10,11, and 21 in the mesenchyme. It is possible that these genes are involved in the interaction between neural crest-derived and myogenic cells during tongue development. In fact, some Sox genes have been proven to be involved in the interaction between the epithelium and mesenchyme in digestive tract development [39,40].
Members of the Sox gene family play multiple roles during development [5,6]. e dynamic spatiotemporal expression pattern of Sox genes discovered in this study indicates their critical role during murine tongue development. We also examined the expression of Sox genes in eyelid development (Table 2). Both organs consist of connective tissue, muscle, and epithelium, and there is an interaction between these tissues during their development [2,3,19,20]. In addition, both tongue and eyelid are mobile organs. Although there are thus numerous similarities between the eyelid and tongue, our results showed that the expression patterns and timing of Sox genes significantly differed between tongue and eyelid development. Sox genes are thus likely to be involved in organogenesis through different functions in each organ.

Conclusion
Sox genes show a dynamic spatiotemporal expression pattern during murine tongue development.

Data Availability
e data used to support the findings of this study are included within the article.

Conflicts of Interest
All authors declare no potential conflicts of interest with respect to the authorship and/or publication of this article.