Somite development in the avian tail

Abstract

Somites are epithelial segments of the paraxial mesoderm. Shortly after their formation, the epithelial somites undergo extensive cellular rearrangements and form specific somite compartments, including the sclerotome and the myotome, which give rise to the axial skeleton and to striated musculature, respectively. The dynamics of somite development varies along the body axis, but most research has focused on somite development at thoracolumbar levels. The development of tail somites has not yet been thoroughly characterized, even though vertebrate tail development has been intensely studied recently with respect to the termination of segmentation and the limitation of body length in evolution. Here, we provide a detailed description of the somites in the avian tail from the beginning of tail formation at HH‐stage 20 to the onset of degeneration of tail segments at HH‐stage 27. We characterize the formation of somite compartment formation in the tail region with respect to morphology and the expression patterns of the sclerotomal marker gene paired‐box gene 1 (Pax1) and the myotomal marker genes MyoD and myogenic factor 5 (Myf5). Our study gives insight into the development of the very last segments formed in the avian embryo, and provides a basis for further research on the development of tail somite derivatives such as tail vertebrae, pygostyle and tail musculature.

Introduction

During chordate embryogenesis the paraxial mesoderm segments into metameric somites, which contain progenitor cells for the axial skeleton, skeletal muscle and connective tissue. Somites form by rhythmically budding off from the anterior end of the unsegmented paraxial mesoderm, which is called the presomitic mesoderm or segmental plate. The newly formed somites are epithelial spheres and can be specifically addressed by a staging system established by Christ and Ordahl, in which the somites are counted in caudal to cranial direction, with somite I being the latest formed, caudalmost somite, somite II the cranially abutting somite, which has been formed before, and so on (Christ & Ordahl, 1995; Fig. 1A,B). A few hours after their budding off the presomitic mesoderm, the epithelial somites undergo epithelial–mesenchymal transition (EMT) in their ventral half, resulting in the ventral mesenchymal sclerotome and the dorsal epithelial dermomyotome (Fig. 1C–E). The sclerotome gives rise to the vertebral column (Scaal, 2016), whereas the dermomyotome remains an epithelial sheet from which myogenic precursor cells delaminate to form a third compartment, the myotome (Scaal & Marcelle, 2018). The myotome consists primarily of mononucleated primitive muscle fibres, which extend from the posterior to the anterior segment border and give rise to the musculature of the trunk. Dermomyotomal cells, which do not enter the myotome, eventually also undergo epithelial–mesenchymal transition and form the dermal and subcutaneous connective tissue of the back.

Figure 1.

Schematic overview of a chicken embryo at about HH‐stage 21. The dotted frame indicates the tail region shown in more detail in (B). (B) Tail of a chicken embryo corresponding to the region indicated in (A) and to the overviews in Figs 2, 3, 4. The broken lines show the planes of section through somites I–IV according to Christ & Ordahl (1995) as shown Figs 2, 3, 4. (C) Schematic transverse section through a newly formed, spherical epithelial somite. (D) Schematic transverse section through an intermediate stage somite. The somite shows first signs of mesenchymalization in the prospective sclerotomal cells of the ventromedial somite wall but still maintains its overall spherical shape. (E) Compartmentalized somite showing a morphologically distinct epithelial dermomyotome, epaxial and hypaxial myotomes and mesenchymal sclerotome. Dm, dermomyotome; Dml, dorsomedial dermomyotomal lip; EpM, epaxial myotome; HypM, hypaxial myotome; Nc, notochord; Nt, neural tube; Sc, sclerotome; So, somite; Tg, tail gut; Vll, ventrolateral dermomyotomal lip.

The formation of somite compartments can be visualized at molecular level by the expression of specific marker genes, which demarcate the prospective fate of the cells. Pax1 is a widely used marker of sclerotomal fate and encodes a paired box transcription factor. Pax1 expression is induced by Shh from the notochord and floor plate of the neural tube. At trunk level, its expression starts in the ventromedial wall of the still epithelial somite at somite stage III (Ebensperger et al. 1995). In both mouse and chick embryos, expression of Pax1 indicates the onset of the deepithelialization of the prospective sclerotomal cells, thus demarcating the overt compartmentalization of the somite into the mesenchymal sclerotome and the epithelial dermomyotome (Brand‐Saberi et al. 1993; Wallin et al. 1994). Interestingly, the epithelio–mesenchymal transition itself does not require the presence of either the notochord or the neural tube (Veini & Bellairs, 1991) nor of Pax1 (Wallin et al. 1994). Thus, Pax1 function in the early somites is not well understood, but it is thought to be involved in sclerotomal cell proliferation.

In late somite stages, prior to sclerotomal chondrogenesis, Pax1 together with the closely related paired box transcription factor Pax9 and the sclerotomal homeobox transcription factors Meox1 and Meox2 activate Nkx3.2 (Bapx1), which promotes chondrogenic differentiation of sclerotomal cells (Rodrigo et al. 2003).

The myogenic program is initiated in the dorsomedial lip (DML) of the dermomyotome by the concerted interaction of Shh signaling from the notochord‐floor plate complex and Wnt1 and Wnt3a signals from the dorsal neural tube (reviewed in Scaal & Marcelle, 2018). Prior to myotome formation, the myogenic fate of the dermomyotomal cells is determined by expression of the closely related bHLH transcription factors Myf5 and MyoD. MyoD and Myf5 have been shown to induce muscle differentiation both in cell culture and in the embryo and act at least in part redundantly (reviewed in Zammit, 2017). MyoD has been reported to be expressed slightly later than Myf5 (Mok et al. 2015), even though earlier studies have found that during somite maturation, MyoD is activated earlier (somite II) than Myf 5 (somite IV) in HH‐stage 12 embryos (Pownall & Emerson, 1992). However, from HH‐stage 20 onward the expression domains of Myf5 and MyoD have been described as being indistinguishable (Berti et al. 2015). The expression of Myf5 and MyoD in the myotome is maintained even when the myotomal cells have formed myotomal fibres (Pownall & Emerson, 1992).

As somites at thoracolumbar levels are most amenable to experimental examination, the progressive steps of somite compartment formation and their molecular regulation have been mostly studied in this region. At thoracolumbar levels, somite stages I–III are epithelial, somite stage IV is an intermediate stage, which shows first signs of ventromedial mesenchymalization, and from somite V on, the somites are compartmentalized to sclerotome, dermomyotome and, with slight delay, to myotome (Fig. 1; Christ & Ordahl, 1995).

However, careful examination of somite development at different axial levels revealed that the dynamics of somite compartment formation is not uniform, but differs at different axial regions (Borman & Yorde, 1994; Borman et al. 1994; Christ & Ordahl, 1995). Notably, in the occipital and cervical regions the formation of sclerotomal and myotomal compartments starts later than at thoracolumbar levels and occurs simultaneously in multiple somites, which is in contrast to the serial cranial‐to‐caudal progression of somite maturation in the trunk (Maschner et al. 2016).

An axial region in which the specifics of somite development have not yet been investigated in detail is the tail region, also termed the coccygeal or caudal region. In the avian embryo, a morphologically distinct tail appears at stage HH‐stage 20 (Schoenwolf et al. 1985). The final number of somites in chicken, which varies between 49 and 52, is reached at about HH‐stage 22 (Bellairs & Sanders, 1986), which implies that about 12 caudal somites are formed before the segmentation machinery and primary axis elongation are terminated (Burke, 2000). Thereafter, the relative length of the tail is progressively reduced, in its distal tip by apoptosis and at its proximal base by integration of tail tissue into the caudal part of the trunk (Schoenwolf, 1981). The caudalmost somites at HH‐stages 25 and later have been reported to decrease in relative size and to become disorganized (Bellairs & Sanders, 1986). The cartilaginous vertebral anlagen of the six to eight caudalmost somites fuse postnatally to give rise to the pygostyle, a small oblong bone at the caudal tip of the shortened avian vertebral column, which is adaptive to flight (Rashid et al. 2018).

Here, we provide a detailed description of somite development in the tail of the chicken embryo from the beginning of overt tail formation at HH‐stage 20 to the onset of caudal tail reduction at HH‐stage 27. We characterize the somite maturation stages in the tail by analysis of semithin sections in combination with the expression pattern of the crucial marker genes Pax1 and MyoD/Myf5 for sclerotome and myotome, respectively. Our systematic description of tail somite development thus provides a basis for further work investigating tail development and the termination of somitogenesis in the chicken embryo.

Materials and methods

Embryos

Fertilized chicken eggs (Gallus gallus domesticus, White Leghorn) provided by commercial breeders (Bronner, Freiburg, Germany; LSL, Dieburg, Germany) were incubated at 37.5 °C and 60% humidity. Embryos were staged according to Hamburger and Hamilton (Hamburger & Hamilton, 1951) 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

Embryos were fixed in 4% paraformaldehyde/PBs‐Tween (PBT) overnight and stored in methanol at −20 °C. In situ hybridizations were performed as described (Nieto et al. 1996). Riboprobes were labelled with digoxigenin RNA labelling kit (Böhringer, Mannheim, Germany) and revealed by BM‐Purple (Roche). The following probes were used in this study: c‐MyoD (Lin et al. 1989), kindly provided by Bruce Paterson, Bethesda, MD, USA, c‐Myf5 (Geetha‐Loganathan et al. 2005) and c‐Pax1 (Ebensperger et al. 1995), kindly provided by Bodo Christ, Freiburg. Specimens were examined and photographed using a Leica MZ16F microscope.

Vibratome sections

Selected stained embryos of HH‐stages 20–22 were embedded in 4% agarose. Older embryos of HH‐stages 23–27 were embedded in a gelatin/albumin substrate and postfixed with glutaraldehyde. After embedding, specimens were sectioned with a vibratome (Leica VT 1000S) at 35 μm and mounted in Aquatex (Merck).

Results and Discussion

In the following we systematically describe the development of somites at caudal levels for each embryonic stage from HH‐stages 20–27. To identify the different somite maturation stages, we used semithin sections both in the sagittal and transverse plane to analyse somite morphology (Fig. 2), Pax1 expression as marker for sclerotomal differentiation (Fig. 3) and MyoD/Myf5 expression as marker for myotomal differentiation (Fig. 4). The planes of the transverse sections used in Figs 2, 3, 4 are schematically shown in Fig. 1B, and the results are graphically summarized in Fig. 5.

Figure 2.

Semithin sections through the tail somites of embryos at different stages. Here and in the following figures, cranial is to the left and dorsal to the top, as illustrated in Fig. 1. On the left, semithin paramedian sagittal section through the tail somites; somite stages are indicated. On the right, transverse semithin sections through the indicated somite stages. HH‐stage 20, 21 and 22: somites I, II and III are epithelial, and somite IV is an intermediate stage. HH‐stage 23: somite I and II are epithelial and somites III and IV are intermediate stages. HH‐stage 24: Somites I, II and III are intermediate stages and in somite IV the dermomyotome and sclerotome are fully formed. HH‐stage 25: somite I is an intermediate stage and somites II, III and IV are fully compartmentalized. HH‐stage 26: all somites including the caudalmost somite are fully compartmentalized. HH‐stage 27: all somites are fully compartmentalized with the caudalmost somites showing the first signs of disintegration.

Figure 3.

Expression of the sclerotomal marker gene Pax1 at different stages. On the left whole mount, in situ hybridizations are shown, and on the right, transverse vibratome sections of the somites I–IV. HH‐stages 20 and 21: somites I and II are devoid of Pax1 expression, somite III is epithelial with onset of Pax1 expression in the epithelial and somitocoel cells of the prospective sclerotome, and somite IV is an intermediate stage with Pax1 expression in the forming sclerotome. HH‐stage 22: somite I is lacking Pax1 expression, somite II and III show Pax1 expression in the cells of the prospective sclerotome, and somite IV shows Pax1 expression in the forming sclerotome. HH‐stage 23: somites I and II show Pax1 expression in the cells of the prospective sclerotome, somites III and IV show Pax1 expression in the sclerotome. HH‐stage 24: somite I is epithelial with Pax1 expression in the ventromedial prospective sclerotomal cells, somites II, III and IV show Pax1 expression in the forming sclerotome. HH‐stage 25–27: all somites show Pax1 expression in the sclerotome.

Figure 4.

Expression of the myotomal marker genes MyoD/Myf5 at different stages. On the left, whole mount in situ hybridizations are shown, and on the right, transverse vibratome sections of the somites I–IV. HH‐stage 20: somites I–III are devoid of Myf5 expression, somite IV shows Myf5 expression in the myogenic precursor cells of the forming dorsomedial dermomyotomal lip (DML). HH‐stage 21 and 22 show no difference to HH‐stage 20 in MyoD expression. HH‐stage 23: somites I and II are lacking Myf5 expression and somites III and IV show Myf5 expression in the myogenic precursor cells of the forming DML. HH‐stage 24: somite I is still devoid of MyoD expression, somites II and III show MyoD expression in the myogenic precursor cells of the forming DML, and somite IV MyoD expression in the epaxial and hypaxial myotome. HH‐stage 25: no MyoD expression in somite I. somites II, III and IV are fully compartmentalized with MyoD expression in the epaxial and hypaxial myotome. HH‐stage 26: all somites including the caudalmost somite are fully compartmentalized with MyoD expression in the epaxial and hypaxial myotome. HH‐stage 27: all somites show Myf5 expression in the epaxial and hypaxial myotome.

Figure 5.

Schematic summary of the maturation stages of the caudalmost somites I–IV from HH‐stages 20–27.

HH‐stage 20

At this stage, somites 40–43 have formed and sensu stricto tail development commences: the tail becomes morphologically distinct as a caudal appendix of the body axis (Schoenwolf et al. 1985). Semithin sections showed that somites I–III are typical epithelial somites. In somite IV the ventral somitic wall loses its epithelial coherence as the cells start EMT, but the sclerotomal cells still largely maintain their spherical arrangement. We address this stage as intermediate stage (Maschner et al. 2016). From somite V on, the sclerotomal cells form a proper mesenchyme (not shown) and disperse to migrate to their destinations in the forming anlagen of the vertebral column (Scaal, 2016). This sequence of somite stages is largely in line with the classical standard sequence in the trunk (Christ & Ordahl, 1995). Accordingly, faint Pax1 expression starts at somite stage III (Ebensperger et al. 1995) in the ventromedial somite wall and in the mesenchymal cells of the somitocoel. This coincides with the initial loss of the basement membrane in the nascent sclerotome and its ‘ventromedial bulging’ in the words of Christ and Ordahl (Christ & Ordahl, 1995), representing a typical intermediate stage at somite stage IV. According to Christ and Ordahl, the dermomyotome and sclerotome are clearly separated from somite stage IV on. The onset of MyoD/Myf5 expression in the myotome precursor cells, which has been assigned to the dorsomedial wall of the epithelial somite at stage II by Christ & Ordahl (1995) with reference to the data of Pownall & Emerson (1992), we detected two somite stages later, starting in the DML of the early dermomyotome at somite IV. This might reflect a difference to the standard sequence of somite development at caudal levels, but might as well be due to variances in sensitivity of the different techniques used here and by Pownall and Emerson.

HH‐stage 21

This stage with 43–44 somites shows the same dynamics of somite development as described for HH‐stage 20.

HH‐stage 22

During this stage, the final number of about 52 somites forms (Bellairs & Sanders, 1986). In the tail, the epithelial morphology of somites I–III and the intermediate stage in somite IV is identical to HH‐stages 20 and 21, but expression of Pax1 is detectable already in somite II. MyoD expression still starts in the DML of the early dermomyotome at somite stage IV, even though in some specimens very faint expression can be detected already in somite III. Thus, even though the morphological sequence of somite development is still standard, the expression of marker genes of sclerotome and myotome starts slightly earlier.

HH‐stage 23

From this stage on, somitogenesis ceases and no further somites are added caudally. Consequently, the normal sequence of somite maturation stages is disrupted as the last formed somites, which continue their developmental programme, acquire features of increasingly advanced somite stages. At HH‐stage 23, the caudalmost somites I–II show epithelial morphology, but display clear expression of Pax1 in the ventromedial cells. Somites III–IV are intermediate stages, and somite V has proper dermomyotome and sclerotome (not shown). Myf5 is clearly detectable in somite III, and very faintly in some cases as early as somite II.

HH‐stage 24

The morphology of intermediate stage somites is now already seen in the caudalmost somites I–III, which show strong Pax1 expression in the forming sclerotome, whereas overt MyoD expression starts in somite II, in some cases even as early as somite I. From somite IV on, dermomyotome and sclerotome are fully formed.

HH‐stage 25

At this stage the chicken tail reaches its maximum length (Schoenwolf, 1981). The caudalmost somite I still shows intermediate stage morphology and Pax1 expression, but as yet no MyoD expression. In contrast, all other somites starting with somite II have completed somite compartment formation, with robust Pax1 expression in the ventromedial sclerotome, and MyoD expression not any more limited to the myogenic precursor cells in the DML, but now evident in the myotome including both epaxial and hypaxial domains.

HH‐stage 26

With all somites, including the caudalmost somite I, showing distinct morphology of dermomyotome, myotome and sclerotome and robust expression of Pax1 in the sclerotome and MyoD in the epaxial and hypaxial myotome, somite compartment formation has reached all segments.

HH‐stage 27

Bellairs & Sanders (1986) have observed smaller and disorganized segments in the caudalmost somites from HH‐stage 25/26 on. In our study, we observed small size and degeneration of somite I only from HH‐stage 27 on, whereas somite II and the more cranial somites still maintain their regular morphology. The small‐sized somite I presumably reflects the onset of tail regression, which initiates at this stage (Schoenwolf, 1981).

Conclusion

Even though there has been some recent progress in understanding the termination of somitogenesis in the tail region (Gomez et al. 2008; Robinton et al. 2019), the specifics of somite development in the tail have not been addressed so far. Our study gives an overview of the characteristics of caudal somite development in the chicken embryo (Fig. 5). We show that somite development in the tail proceeds in the ‘canonical’ fashion known from classical literature until HH‐stage 21. Between HH‐stages 22 and 23, somitogenesis is completed, and henceforth more mature somite stages are progressively found in more caudal segments. At HH‐stage 26 all somites show full compartment formation including dermomyotome, myotome and sclerotome, but from HH‐stage 27 degeneration of somite morphology starts in caudal to cranial progression. This might reflect the onset of tail regression (Schoenwolf, 1981) and the subsequent fusion of segments during pygostyle formation, even though recent data have implied that overt vertebral fusion during pygostyle formation is a postnatal process (Rashid et al. 2018). Hence, due to the termination of somitogenesis at HH‐stage 22, the maturation stages of tail somites at later stages no longer follow the ‘canonical’ sequence described by Christ & Ordahl (1995). These data provide a structural basis for further research on the development of somites and their derivatives in the avian tail.

Author contributions

M.D., F.P. and M.S. contributed to concept and design of the study, data analysis and interpretation, drafting and writing of the manuscript. K.H., R. B. and M.W. contributed to acquisition of data, data analysis and interpretation, and critical reading of the manuscript.