Developmental dynamics of occipital and cervical somites

Abstract

Development of somites leading to somite compartments, sclerotome, dermomyotome and myotome, has been intensely investigated. Most knowledge on somite development, including the commonly used somite maturation stages, is based on data from somites at thoracic and lumbar levels. Potential regional differences in somite maturation dynamics have been indicated by a number of studies, but have not yet been comprehensively examined. Here, we present an overview on the developmental dynamics of somites at occipital and cervical levels in the chicken embryo. We show that in these regions, the onset of sclerotomal and myotomal compartment formation is later than at thoracolumbar levels, and is initiated simultaneously in multiple somites, which is in contrast to the serial cranial‐ to‐ caudal progression of somite maturation in the trunk. Our data suggest a variant spatiotemporal regulation of somite development in occipitocervical somites.

Introduction

Somites are segmental units within the paraxial mesoderm that provide the basis for the segmental organization of the vertebral column, the trunk musculature and the spinal nerves. Along the craniocaudal body axis, the somites give rise to derivatives that differentiate according to their segmental position, thus identifying somites of occipital, cervical, thoracic, lumbar, sacral and coccygeal (tail) identity. The number of somites within these segmental regions varies between species, as does the total number of somites formed (Burke et al. 1995; Burke, 2000). In the chicken embryo, somites 1–5 are of occipital, somites 5–19 of cervical, somites 19–26 of thoracic, somites 26–30 of lumbar, somites 30–39 of sacral, and somites 39–52 of coccygeal identity (Christ & Ordahl, 1995; Burke, 2000; Christ et al. 2000).

After their formation from the unsegmented paraxial mesoderm at the caudal end of the embryo, somites change their morphology in the process of somite compartmentalization. Initially, somites are epithelial spheres consisting of a pseudostratified epithelial wall and a central cavity, the somitocoel, which is filled with mesenchymal cells (Fig. 1). Subsequently, the ventral half deepithelializes to form a highly proliferative and motile population of mesenchymal cells, the sclerotome, which gives rise to a number of tissues including the vertebral column and the ribs (reviewed in Scaal & Christ, 2004; Scaal, 2016). With the onset of sclerotome formation, these cells express marker genes like Pax1 (Deutsch et al. 1988; Peters et al. 1995) and Pax9 (Stapleton et al. 1993; Wallin et al. 1993; Peters et al. 1995).

Figure 1.

Schematic drawings of early stages of somite maturation proceeding from the right to the left. The sclerotomal cell lineage is shown in yellow. (A) Somites with fully mesenchymalized sclerotome, corresponding for example to somite stage VIII/22. (B) Somites at an intermediate stage of sclerotome formation, showing loss of epithelial integrity in the ventral half but maintenance of the spherical arrangement of cells, corresponding for example to somite stage V/22. (C) Somites at epithelial stage, corresponding for example to stage II/22.

The dorsal somite half, in contrast, remains epithelial and forms a sheet‐like structure, the dermomyotome (Fig. 1). Within the dermomyotome, local epithelio‐mesenchymal transition events recruit motile precursor cells that give rise to skeletal muscle, endothelia and connective tissue. Muscle precursor cells accumulate underneath the dermomyotome and elongate to form a sheet of primary myocytes, the myotome (reviewed in Scaal & Christ, 2004). The myogenic fate of these cells is initiated by the expression of transcription factors like MyoD (Davis et al. 1987) and Myf5 (Braun et al. 1989), and subsequent expression of early muscle‐specific markers like desmin (Granger & Lazarides, 1979) and myosin (Shimizu et al. 1985a,b). Detailed descriptions of the expression patterns of these and other myotomal marker genes during chicken somite development have recently been provided by Berti et al. (2015) and Mok et al. (2015).

Christ & Ordahl (1995) have proposed to stage the maturation level of somites according to the developmental stage of the somitic compartments of a specific segment by using Roman numerals. In this system, the most recently formed thus caudal‐most somite is staged I, the cranially abutting somite formed one segmentation round earlier is staged II, etc. Accordingly, the newly formed, ball‐shaped epithelial somites are staged I–III, the early compartmentalized somites, which are divided into sclerotome and dermomyotome, are staged IV–X, and the late compartmentalized somites showing extensive emigration of sclerotome and dermomyotome derived cells are staged XI–XX. In later stages, the integrity of the somites is lost due to the differentiation of the majority of somite‐derived cells (Christ & Ordahl, 1995). This staging system has been widely accepted and is a useful tool to give unequivocal descriptions of research done in somites at the thoracic and lumbosacral level, which is the vast majority of studies in the field.

However, as frequently not considered, but indicated by Christ & Ordahl (1995), the maturation dynamics of somites differ in regions outside the thoracic and lumbar levels. This means that a somite staged for example VII does not represent the same maturation stage throughout the length of the paraxial mesoderm. For instance, the work by Borman et al. (1994) and Borman & Yorde (1994) has demonstrated that myogenic differentiation in chicken embryos as monitored by desmin expression starts in somite X–XII at occipital and cervical levels, whereas at lumbosacral level it starts as early as in somite V. In other words, the occipital and cervical somites start the myogenic differentiation program approximately 18 h after somite formation, whereas more caudal somites do so already after about 8 h, indicating striking differences in the developmental program between anteriorly and posteriorly located segments. To date, the deviant somite maturation dynamics in the occipitocervical somites has not been thoroughly examined.

The aim of this study is to provide a detailed description of the early development of occipital and cervical somites with respect to the dynamics of the formation of somite compartments, sclerotome and myotome, in order to provide a morphological basis for functional studies on somite development in this region.

Materials and methods

Embryos

Fertilized chicken eggs (Gallus gallus domesticus, White Leghorn) provided by a commercial breeder (LSL, Dieburg) were incubated at 37.5 °C and 50% humidity. Embryos were roughly staged according to Hamburger & Hamilton (1951), and staged to somite stages by careful counting of postotic somites under a dissection microscope (Leica Mz7,5; Leica, Germany).

Semithin sections

Embryos destined for semithin sections were fixed in 4% paraformaldehyde/phosphate‐buffered saline (PBS), treated with 0.5% osmium/PBS, washed in PBS, dehydrated in a graded ethanol series, washed in propylene oxide, and embedded in Durcupan resin (Fluka). Semithin sections were cut with a diamond knife (Leica, Ultracut S) at 0.75 μm and stained with 1% methylene blue and 1% azure II (Fluka) in sodium tetraborate and H₂O.

In situ hybridization on whole‐mount embryos

Operated embryos were fixed in 4% paraformaldehyde/PBS overnight and stored in methanol at −20 °C. In situ hybridizations were performed as described in Nieto et al. (1996). Riboprobes were labeled with digoxigenin RNA labeling kit (Böhringer, Mannheim, Germany) and revealed by BM‐Purple (Roche). To achieve comparability of expression levels between different embryonic stages, embryos of all stages were incubated in staining solution for the same time. Staining reaction was stopped when the somites that showed the strongest expression yielded a robust color reaction. Very faint expression levels observed at this timepoint were not recorded as positive.

The following probes were used in this study: c‐MyoD (kindly provided by Bruce Paterson, Bethesda, CA, USA) and c‐Pax1 (Ebensperger et al. 1995; kindly provided by Bodo Christ, Freiburg, Germany). Specimens were examined and photographed using a Leica MZ 16 F microscope. Selected stained embryos were embedded in 4% Agarose, sectioned with a vibratome (Leica VT 1000S) at 35 μm and mounted in Aquatex (Merck).

Results

In order to study somite development in the occipital and cervical regions, we performed sagittal and transversal semithin sections of chick embryos to examine somite morphology, and subjected embryos to in situ hybridizations for the sclerotomal marker Pax1 and the myotomal marker MyoD to analyse the dynamics of somite compartment formation on a molecular level. We examined embryos between HH‐stage 7, when the first somites are formed, up to HH‐stage 14, corresponding to 22 somite stage, thus allowing us to compare somite development in occipital (somites 1–5), cervical (somites 5–19) and thoracic (somites 19–26) segments.

The common staging system after Hamburger & Hamilton (1951) is not exact with respect to somite numbers. Therefore, we staged embryos according to the total number of somites formed, i.e. a HH‐stage 10 embryo with nine somites was termed 9s, and a HH‐stage 10 embryo with 11 somites was termed 11s. Additionally, following the suggestion by Christ & Ordahl (1995), we addressed individual somites unequivocally by Roman numerals to count them in the caudal‐to‐cranial direction reflecting their maturation stage. Thus, for instance, somite 5, representing the occipital–cervical transition, of an early HH‐stage 9 embryo with seven somites is named III/7.

Strikingly, we observed that throughout stages 1s–7s, all somites were epithelial in morphology, and expressed neither Pax1 nor MyoD, indicating that they maintained the undifferentiated and naïve state until somite stage VII (shown exemplarily in Fig. 2A–D). This is in contrast to trunk levels, where the undifferentiated state of epithelial somites is restricted to somite stages I–III (Christ & Ordahl, 1995).

Figure 2.

Progression of somite compartment formation in occipital and cervical somites. Cranial to the left. Scale bar: 100 μm. (A) Sagittal semithin section through the paraxial mesoderm of a HH‐stage 7 embryo with two somites. The first somite (II/2) is continuous with the paraxial head mesoderm, the second somite (I/2) is the first somite separated by two intersomitic clefts. (B) Whole‐mount in situ hybridization to reveal Pax1‐expression in a HH‐stage 7 embryo with two somites. Expression is limited to the pharyngeal endoderm, the somites are devoid of Pax1 expression. (C) Sagittal semithin section through the paraxial mesoderm of a HH‐stage 8 embryo with four somites. All somites show epithelial morphology. (D) Whole‐mount in situ hybridization to reveal Pax1‐expression in a HH‐stage 8 embryo with four somites. Expression is limited to the pharyngeal endoderm, the somites are devoid of Pax1 expression. (E) Sagittal semithin section through the paraxial mesoderm of a HH‐stage 9 embryo with eight somites. All somites show epithelial morphology, only the first somite (VIII/8) shows first signs of mesenchymalization in its cranioventral corner. (F) Whole‐mount in situ hybridization to reveal Pax1‐expression in a HH‐stage 9 embryo with eight somites. Next to the pharyngeal endoderm, expression occurs simultaneously in the cranial‐most five somites VIII/8–IV/8, which correspond to the occipital segments. (G) Sagittal semithin section through the paraxial mesoderm of a HH‐stage 9 embryo with nine somites. All somites show epithelial morphology, only the first somite (VIII/8) is partially mesenchymal cranioventrally. (H) Whole‐mount in situ hybridization to reveal Pax1‐expression in a HH‐stage 9 embryo with nine somites. Pax1 is expressed in the cranial‐most six somites, indicating the onset of craniocaudal progression of Pax1‐expression in the cervical segments. (I) Sagittal semithin section through the paraxial mesoderm of a HH‐stage 10 embryo with 11 somites. The first somite (XI/11) shows a fully mesenchymal sclerotome, somites X/11 and IX/11 show an intermediate stage with a beginning loss of epithelial integrity in the ventral somite half, and all more caudal somites show epithelial morphology. (J) Whole‐mount in situ hybridization to reveal Pax1‐expression in a HH‐stage 10 embryo with 11 somites. Expression in the first somite (XI/11) starts to diminish, somites X/11–IV/11 show strong expression, and the caudal‐most somites I‐III are devoid of expression. (K) Whole‐mount in situ hybridization to reveal MyoD‐expression in a HH‐stage 10 embryo with 11 somites. No expression is detected yet. (L) Sagittal semithin section through the paraxial mesoderm of a HH‐stage 11 embryo with 13 somites. The first three somites (XIII/13–XI/13) show a fully mesenchymal sclerotome, somites X/13 and IX/13 show an intermediate stage with a beginning loss of epithelial integrity in the ventral somite half, and all more caudal somites show epithelial morphology. (M) Whole‐mount in situ hybridization to reveal Pax1‐expression in a HH‐stage 11 embryo with 13 somites. Expression in the first somite (XIII/13) is very weak, and the caudal‐most somites I‐III are devoid of expression. (N) Whole‐mount in situ hybridization to reveal MyoD‐expression in a HH‐stage 11 embryo with 13 somites. Expression is detected simultaneously in the cranial‐most 10 somites, only somites I–III are devoid of expression. (O) Sagittal semithin section through the paraxial mesoderm of a HH‐stage 12 embryo with 18 somites. Somites VIII/18 and VII/18 show an intermediate stage, and somite stages I–VI show epithelial morphology. (P) Whole‐mount in situ hybridization to reveal Pax1‐expression in a HH‐stage 12 embryo with 19 somites. The caudal‐most somite I/19 is devoid of expression, weak expression starts in somite II/19. (Q) Whole‐mount in situ hybridization to reveal MyoD‐expression in a HH‐stage 12 embryo with 19 somites. Weak expression starts at somite stage III, the cranial‐most somites until XII/19 show expression in the sheet‐like myotomal compartment except the first somite (XIX/19), which shows only rudimentary expression. (R) Sagittal semithin section through the paraxial mesoderm of a HH‐stage 14 embryo with 22 somites. Somites VI/22 and V/22 show an intermediate stage, and somites I–IV show epithelial morphology. (S) Whole‐mount in situ hybridization to reveal Pax1‐expression in a HH‐stage 14 embryo with 22 somites. The caudal‐most somites stage I–II are devoid of expression, weak expression starts in somite III/22. (T) Whole‐mount in situ hybridization to reveal MyoD‐expression in a HH‐stage 14 embryo with 22 somites. Expression starts very faintly at somite stage II, and robustly from somite stage III on.

At 8s, the cranial‐most somite (VIII/8) showed the first loss of epithelial integrity in its cranial–ventral portion (Fig. 2E). This reflects its early disintegration, which has been described in the literature (Williams, 1910; Hamilton & Hinsch, 1956; Huang et al. 1997). Strikingly, all occipital somites, VIII/8 to IV/8, initiated Pax1 expression simultaneously, with the cranial‐most somite 1 (VIII/8) showing weaker expression than the four more caudal somites (Fig. 2F). In some embryos, very faint expression of Pax1 appeared already in III/8, which we did not record as proper expression in this study.

At 9s, the cranialmost somite (IX/9) showed the beginning of mesenchymalization in its cranio‐ventral portion, being still an epithelial sphere, though less cohesive in the ventral half. This stage we defined as intermediate stage, as it represents a transition stage between epithelial and mesenchymal morphology (see Figs 1, 3 and 4). Somites VIII/9 to I/9 were overall epithelial (Fig. 2g). With respect to sclerotomal marker gene expression, somites IX/9 to IV/9 showed robust Pax1 expression, with IX/9 and IV/9 expressing slightly weaker than the somites between. The caudalmost somites I/9 to III/9 did not express Pax1 at all (Fig. 2H).

Figure 3.

Transition from epithelial to compartmentalized somites at thoracic level at HH‐stage 14. Scale bar: 50 μm. Top row: semithin transversal sections through somites III–V. While somite III is clearly epithelial, somite IV shows the first signs of loosening of the ventral epithelial wall, and in V the loss of epithelial integrity in the ventral epithelial wall is evident, even though the spherical shape of the somite is still maintained. This stage is designated as intermediate stage. Middle row: Pax1 expression starts very faintly in the ventromedial aspect of the epithelial somite stage III, and progressively intensifies until it is strong in the mesenchymal sclerotome appearing at stage V. Lower row: MyoD expression is faint in the dorsomedial aspect of somites III and IV, and becomes robust in the forming dorsomedial dermomyotomal lip at stage V.

Figure 4.

Schematic overview of the developmental dynamics of occipital and cervical somites. The horizontal array represent somites I–XXII numbered with Roman numerals as suggested by Christ & Ordahl, 1995, cranial is to the left. The vertical array represents successive developmental stages numbered with Arabic numerals. For details, see the Results section.

At 11s, somite XI/11 had an overtly mesenchymal sclerotome, whereas in somites X/11 and IX/11 sclerotome mesenchymalization had started, but cells were still maintaining a spherical arrangement, thus being classified as intermediate stage (Fig. 2I). All more caudal somites, irrespective of Pax1 expression, showed clear epithelial morphology. Pax1 expression appeared in a regular pattern: Pax1 was expressed in somites IV through XI, again being slightly weaker in somites XI and IV, and the newly formed somites I–III were devoid of expression (Fig. 2J). This pattern was principally conserved over the following stages. It should be noted that the caudal limit of Pax1 expression varied slightly between individual embryos, in some specimens appearing faintly already in stage III, in some specimens only at stage V. MyoD expression was not yet detected at this stage (Fig. 2K).

At stage 13s, the cranial‐most three somites showed clearly mesenchymal sclerotomes, followed by two intermediate stages, whereas all more caudal somites remained epithelial in morphology (Fig. 2L). Importantly, between stages 12s and 13s, MyoD expression was observed for the first time. It appeared simultaneously as a narrow stripe in the medial aspect of somites XII/12 (XIII/13) to IV/12 (IV/13), leaving only the caudal‐most somites I–III devoid of expression (Fig. 2N). Transverse sections revealed that this medial stripe of expression corresponds to the prospective dorsomedial lip of the dermomyotome (DML) (compare with Fig. 3).

In the following stages, this pattern of MyoD expression remained essentially the same. In parallel to the observations with Pax1, we found that the caudal limit of MyoD expression varied slightly between embryos, starting usually at somite stage IV, but sometimes at somite stages V or III, and that MyoD expression in the cranial‐most somite was weaker than at more caudal levels.

At 18s and19s, the cranial‐most 10–11 somites showed a mesenchymal sclerotome, followed by two intermediate stages, and somites I–VI showed epithelial morphology (Fig. 2O).

Pax1 started to be expressed as early as somite stage III or even, more weakly, from somite stage II (Fig. 2P). Both Pax1 and MyoD expression in the cranial‐most somite was only rudimentary in its caudal–medial corner. Cranial of somite stage XI, MyoD expression was not restricted to the DML region any more, but was extending laterally into the now formed myotomal compartment. In more caudal somites, MyoD was still restricted to the DML, and the caudal‐most somites I–III (II, IV) were devoid of MyoD expression (Fig. 2Q).

At 22s and 23s, the epithelial type of somites was restricted to somites I–IV, followed by two somites (V–VI) at the intermediate stage and proper sclerotomes at all stages cranial to VI (Fig. 2R). Both Pax1 and MyoD were found to be expressed from somite stages III, or faintly from stage II (Figs 2S,T and 3). In the cranial‐most somite XXII/22, both Pax1 and MyoD expression were hardly detectable any more. In stage VIII/23 and more cranial somites, MyoD expression appeared in a mediolateral stretch corresponding to the morphological myotome.

In summary, mesenchymalization of the ventral somite half starts as late as somite stage 8s or 9s, and proceeds along the somite array in cranial to caudal progression. At the same time, in contrast, Pax1 expression starts simultaneously in the first five somites representing the occipital segments. With delay of 4–5 somite stages, at 12s, MyoD expression starts simultaneously in the first 8–9 somites, encompassing the occipital segments and the cranial third of the cervical segments. During the following stages of cervical somite formation, sclerotomal and myotomal marker gene expression start at somite stage IV, but the onset of morphological sclerotome and myotome formation occurs relatively late, thus preserving the epithelial character of the somites longer than at thoracolumbar levels. From approximately stage 20s (HH‐stage 13) on, when the formation of thoracic somites starts, the canonical sequence of somite maturation has been reached, with Pax1 and MyoD expression starting at stage III, sclerotomal mesenchymalization starting at stage V (intermediate stage) and overt myotome formation starting at stage VIII. For a graphical summary of these results, see Fig. 4.

Discussion

Our study provides an overview on somite maturation dynamics in the occipitocervical region of the chick embryo. We show that occipital somites stay epithelial until eight somites have been formed in total, and only then start synchronous Pax1 expression in sclerotomal precursor cells of all five occipital segments. Interestingly, the morphologically visible mesenchymalization of sclerotomal cells does not commence simultaneously in the occipital segments, but proceeds stepwise in the cranial to caudal direction. The onset of MyoD expression in myotomal precursor cells occurs after 12 somites have formed, also simultaneously, in the occipital somites and the cranial third of cervical somites. Generally, the onset of morphological somite compartment formation in occipital and cervical somites is delayed when compared with thoracolumbar levels.

Most of the studies on the development of somites have been performed at the thoracic and lumbar level, which is, in the case of the chicken embryo, the part of the embryonic body that is most easily accessible to experimental manipulation. Accordingly, the key publications used by most scientists as the morphological basis of their work on somites are also referring to this region, for example, Christ & Ordahl (1995) in the chick, Gossler & Hrabe de Angelis (1998) in the mouse, and textbooks like Gilbert (2014). However, Williams (1910), Borman et al. (1994), and Borman & Yorde (1994) have pointed out that development of the somites does not proceed at the same pace at all segmental levels along the body axis. This must be considered when working at ‘non‐canonical’ somite levels, like occipital, cervical or coccygeal levels. Still, the developmental dynamics of somites in these axial regions have not yet been thoroughly studied.

The cranial‐most somites, which are the first to form in the postotic paraxial mesoderm, do not give rise to vertebrae as more caudal somites do, but their sclerotomal cells are integrated into the base of the skull, hence representing the occipital segmental level (Couly et al. 1993). In avian embryos, the occipital segments are somites 1–5, the occipito‐cervical transition running through somite 5 (Couly et al. 1993; Wilting et al. 1995; Huang et al. 2000). In spite of the loss of overt metamery, careful studies by Huang et al. (2000) have provided evidence that the sclerotomal contribution to the skull remains segmental.

Our study showed that in the occipital somites, sclerotome formation marked by Pax1 expression starts simultaneously at HH‐stage 9 (8s), indicating that the occipital somites maintain their epithelial state longer than at thoracolumbal levels, and that the regulation of sclerotome induction might be different in occipital vs. truncal levels. What might cause this difference is unclear, as occipital sclerotome formation is induced by a Shh and Ptc1‐dependent pathway similar to more caudal levels, whereas the skeletogenic pathway in the unsegmented (preotic) paraxial head mesoderm is differently regulated (Balczerski et al. 2012). The timing of chicken Pax1 expression found in our study is in line with results by Ebensperger et al. (1995) in quail embryos, who found that Pax1 expression occurs first at 8s and, from about 12s on, Pax1 expression starts in somite III. Interestingly, Wilting et al. (1995) have reported that Pax1 is downregulated in occipital sclerotomes upon fusion of the occipital segments to form the basioccipital bone of the skull, and during fusion of cells of cervical segments 5 and 6 to form the dens axis, which argues for a role of Pax1 during resegmentation. Not only at occipital levels, but also at cervical levels, the latency between Pax1 expression and morphologically visible sclerotome mesenchymalization is longer than at thoracolumbar levels, taking five segmentation rounds vs. two segmentation rounds in the trunk. The molecular reason for this delayed onset of mesenchymalization triggered by Pax1 is not known.

Next to their contribution to the occipital skull, the occipital somites are special compared with more caudal segments in a number of aspects in different vertebrates. Among the occipital somites, somites 1–3 are unique in that they are invaded by head neural crest cells, whereas somites 4 and 5 are populated by trunk neural crest (Ferguson & Graham, 2004). Thus, the head–trunk interface for neural crest cells lies between somites 3 and 4, whereas the skeletal boundary lies within somite 5. This represents an ancient situation, as in the lungfish Neoceratodus the occipitocervical boundary has been described to be within somite 4 (Johanson et al. 2005). In the teleost zebrafish, in contrast, occipital fate has been described only for somites 1 and 2 (Morin‐Kensicki et al. 2002). Comparable to the results presented here for the chicken embryo, the anterior somites in zebrafish embryos all maintain epithelial morphology for a relatively long time (7s), and simultaneously start to differentiate by elongation of adaxial cells in the cranial‐most five somites (van Eeden et al. 1998). Studies of zebrafish mutants have shown that one eyed pinhead:no tail double mutants form only the first five somites, while all more caudal somites are missing (Schier et al. 1997). In mutants of the notch pathway, like deadly seven (des, Notch1a), after eight (aei, DeltaD) and white tail (wit, Mindbomb), the first 7 (± 2) somites form normally and are resistant even to experimental perturbation of the notch pathway, but caudal to this region defective epithelial somite formation and somite boundary formation occurs. On the contrary, mutation of integrinα5 (before eight) disrupts only the formation of the cranial‐most somites, whereas all somites caudal to somites 7–9 are unaffected (van Eeden et al. 1998; Julich et al. 2005). Genes like zebrafish Sox11a (de Martino et al. 2000) and Nanos2 (Julich et al. 2005) are exclusively expressed in the cranial‐most 4–5 somites, but not in more caudal segments. Interestingly, also the kinetics of occipital vs. trunk somite formation is different. In mouse, the cranial‐most somites form every 60 min, whereas more caudal somites form every 2–3 h (Tam, 1981). In Amphioxus, the cranial‐most eight somites form every hour, more caudal ones only every 18 h (Schubert et al. 2001).

Another interesting result of our study is that myotomal differentiation, as monitored by MyoD expression, starts much later than sclerotomal differentiation, but simultaneously at 12s in somites XII/12 to IV/12, so that MyoD expression ‘catches up’, as it were, with Pax1 expression to be coexpressed in all compartmentalized somites. We detected MyoD expression slightly earlier than reported in the literature, as dorsomedial expression of MyoD visualized by radioactive in situ hybridization has been reported to start at 16s in quail and chicken embryos (Piette et al. 1992; Pownall and Emerson 1992). Two years earlier, de la Brousse and Emerson (1990), also using radioactive in situ hybridization, had reported first MyoD expression as late as in HH‐stage 13 quail embryos (19s). A recent study using similar methods as applied here reported MyoD expression starting at 15s (Mok et al. 2015). Not only concerning the developmental stage, but also concerning the somite stage where MyoD expression starts, our study is at variance with these earlier data. We detected the first expression of MyoD at 12s to start in somite stage IV, whereas Mok et al. (2015) described the onset of MyoD expression at 15s to start as late as somite stage VIII–XI. We found that at 16s, MyoD expression starts at somite stage III or IV, whereas Pownall and Emerson (1992) described MyoD expression in 16s quail embryos already at somite stage II, and Piette et al. (1992) in chicken embryos even at somite stage I. de la Brousse and Emerson (1990) detected MyoD at HH‐stage 13–14 (19s–22s) in quail embryos only in the rostral‐most 5 ± 1 somites, with all more caudal somites devoid of MyoD transcripts, which is in stark contrast to our results in chicken showing MyoD expression at this stage already in somite stage III.

These discrepancies remain puzzling, but might be due to differences in the sensitivity of the variant methods applied, and in part to species differences between quail and chicken. In line with our findings on the onset of myotome formation (Fig. 4), Borman & Yorde (1994) and Borman et al. (1994) report the onset of desmin expression in corresponding stages. Interestingly, in zebrafish embryos, MyoD expression also starts simultaneously in the cranial‐most five somites, but already at the 6–9 somite stage (Weinberg et al. 1996).

To conclude, our study highlights the specific somite maturation dynamics in the occipital and cervical region, which is at variance with the ‘canonical’ somite maturation schedule in the commonly known thoracolumbar region. Our study provides a basis for further research on the regulation of the delayed and synchronized onset of sclerotomal and myotomal differentiation and the subsequent morphogenetic events at occipitocervical levels.

Author contributions

AM: acquisition of data, data analysis and interpretation, critical revision of the manuscript. SK: acquisition of data, data analysis and interpretation, critical revision of the manuscript. MD: artwork, data analysis and interpretation, critical revision of the manuscript. FP: data analysis and interpretation, critical revision of the manuscript. MS: concept and design of the study, data analysis and interpretation, drafting of the manuscript.