Expression of myogenic regulatory factors in chicken embryos during somite and limb development

Abstract The expression of the myogenic regulatory factors (MRFs), Myf5, MyoD, myogenin (Mgn) and MRF4 have been analysed during the development of chicken embryo somites and limbs. In somites, Myf5 is expressed first in somites and paraxial mesoderm at HH stage 9 followed by MyoD at HH stage 12, and Mgn and MRF4 at HH stage 14. In older somites, Myf5 and MyoD are also expressed in the ventrally extending myotome prior to Mgn and MRF4 expression. In limb muscles a similar temporal sequence is observed with Myf5 expression detected first in forelimbs at HH stage 22, MyoD at HH stage 23, Mgn at HH stage 24 and MRF4 at HH stage 30. This report describes the precise time of onset of expression of each MRF in somites and limbs during chicken embryo development, and provides a detailed comparative timeline of MRF expression in different embryonic muscle groups.


Introduction
The myogenic regulatory factors (MRFs) are a group of four helix-loop-helix transcription factors that are critical regulators of muscle cell commitment and differentiation in vertebrates (Pownall et al. 2002;Chanoine et al. 2004;Berkes & Tapscott, 2005;Tapscott, 2005;Sweetman, 2012). The discovery of MyoD, based on its ability to convert fibroblasts into muscle cells (Davis et al. 1987), was soon followed by the identification of other closely related genes that share this transformative property: Myf5, myogenin (Mgn) and MRF4 (Braun et al. 1989(Braun et al. , 1990Edmondson & Olson, 1989;Rhodes & Konieczny, 1989;Wright et al. 1989;Fujisawa-Sehara et al. 1990;Miner & Wold, 1990). Together these genes orchestrate the formation of all skeletal muscle in the vertebrate embryo.
As these genes have such critical roles in the development of muscle, their regulation and expression has been extensively studied. A series of mouse reporter lines have been used to uncover a large number of complex interdigitated enhancers that regulate Myf5 and MRF4 expression, which are distributed across 150 kb around these closely linked loci (Tajbakhsh & Buckingham, 1995;Tajbakhsh et al. 1996;Summerbell et al. 2000;Carvajal et al. 2001Carvajal et al. , 2008Buchberger et al. 2003Buchberger et al. , 2007Teboul et al. 2003). In mammals MyoD expression is primarily controlled by two enhancers, the core enhancer and distal regulatory region, which regulate the onset and maintenance of MyoD, respectively (Goldhamer et al. 1992;Tapscott et al. 1992;Asakura et al. 1995;Faerman et al. 1995;Chen et al. 2001Chen et al. , 2002Chen & Goldhamer, 2004), and similar regulatory sequences have been identified in birds (Pinney et al. 1995). MyoD in turn regulates expression of Mgn and can interact directly with its promoter to recruit chromatin remodelling complexes (de la Serna et al. 2005;Armand et al. 2008;Deato et al. 2008).
Understanding of development has been greatly informed by the use of the chicken embryo as a model organism (Tickle, 2004;Davey & Tickle, 2007), and it has been an exceptionally useful system to understand the developmental signals that regulate myogenesis in different embryonic domains, such as the epaxial myotome (M€ unsterberg et al. 1995;Borycki et al. 1998;Schmidt et al. 2000), the hypaxial myotome (Dietrich et al. 1998;Dietrich, 1999) and the developing limb Delfini et al. 2000;Edom-Vovard et al. 2001;Marics et al. 2002;Geetha-Loganathan et al. 2005;Mok et al. 2014).
However, the existing literature does not comprehensively define the stages and expression patterns of each MRF during early somite and limb development in chicken embryos, although this has been done for the head musculature (Noden et al. 1999). To address this, the expression of each MRF has been compared in somites and limb using in situ hybridisation to delineate the relative timings and expression domains of each of these genes.

Materials and methods
Dig-labelled probes for in situ hybridisation were generated and used as described (Sweetman et al. 2008). Fertilised white leghorn eggs were obtained from Henry Stewart (Norfolk) and incubated to reach the desired stage.

Expression in somites from HH stages 9 to 14
Myf5 expression is first detected at HH stage 9 (6-8 somites) in the medial somite adjacent to the neural tube. Expression was observed in somites 1-6 in embryos with 7 and 8 somites, but Myf5 expression was not observed in embryos earlier than this. At HH stage 10, expression is seen in the anterior somites 1-6 ( Fig. 1A), while the other MRFs are not expressed ( Fig. 1B-D). At HH stage 12 (14-16 somites), Myf5 is detected in all somites and, in some cases, paraxial mesoderm (Fig. 1E).
MyoD was not detected in 14-somite embryos but was in somites 1-8 in 2/5 15-somite embryos and 16-somite embryos (Fig. 1F), while Mgn and MRF4 are expressed in anterior somites, respectively (Fig. 1G,H). The staining of Mgn and MRF4 observed at these stages is variable. Mgn is detected in 5/10 16-somite embryos: in two of these expression is seen in the two most anterior somites, in the other three of the four most anterior somites are stained. Similarly, in 16-somite embryos MRF4 is seen in 7/11 embryos: two have expression in somites 1-4, four in somites 1-6 and one in somites 1-8.
At HH stage 14 (20-22 somites), strong Myf5 expression is seen in all somites and in somite position I, the region where a new somite is about to form. Posterior somites show Myf5 restricted to the medial somites, while in more anterior somites expression is expanded through the whole medial-lateral extent. MyoD expression at this stage is seen in somites 1-18 and, like Myf5, is restricted to the medial region in more posterior somites but is more widespread in anterior somites (Fig. 1J), while Mgn and MRF4 are now detected in somites 1-10/12 (Fig. 1K,L). Again there is some heterogeneity in the expression of Mgn and MRF4 at these stages and, at a given somite number, differences in the staining pattern of up to two somites either more or less are observed.
At HH stage 16 (26-28 somites), Myf5 is seen in all somites and in the paraxial mesoderm that is about to form a new somite, and MyoD is detected in all but the most recently formed somites (Fig. 1M,N), while Mgn (Fig. 1O) and MRF4 are expressed in somites 1-20 (Fig. 1P). As is seen at earlier stages, there is a degree of variability in this staining and some embryos will have one additional somite expressing either Mgn or MRF4.
Transverse sections of HH stage 16 embryos show distinct localisation of MRF transcripts. In anterior somites Myf5 is expressed in the dorso-medial and ventero-lateral lips of the dermomyotome, and throughout both the epaxial and hypaxial myotome that lies directly under the dermomyotome ( Fig. 2A-C). In more posterior regions (i.e. somite 15), Myf5 is only detected in the dorso-medial lip of the dermomyotome and forming epaxial myotome (Fig. 2D). In the segmental plate Myf5 expression is found in the medial unsegmented paraxial mesoderm immediately adjacent to the neural tube (Fig. 2E).
MyoD in anterior somites is expressed throughout the myotome but not the dermomyotome (Fig. 2F-H), and in more posterior somites this is restricted to the dorso-medial region of the myotome. In contrast to Myf5, MyoD is not detected in the dorso-medial or ventro-lateral lips of the dermomyotome.
Mgn is also expressed only in the myotome and not the dermomyotome, although it is restricted to the central domain in anterior somites and does not extend as far into the ventro-lateral myotome as MyoD (Fig. 2J-M).
MRF4 is also only expressed in anterior somites, but is only seen in the dorso-medial part of the myotome ( Fig. 2N-P) and does not extend into the ventro-lateral myotome as far as either MyoD or Mgn at equivalent stages.

Expression in somites from HH stage 20 to HH stage 26
Sections through older embryos (HH stage 20 to HH stage 26) at interlimb levels (i.e. between somites 22 and 25) show the myotome extending ventrally to form the trunk muscles. At HH stage 20, both Myf5 and MyoD are expressed throughout the myotome (Fig. 3A,E), while Mgn and MRF4 are expressed in a more restricted dorsal myotomal domain (Fig. 3I,M). At HH stage 22 and 24, Myf5, MyoD and Mgn are expressed in the ventrally extending myotome (Fig. 3B, C,F,G,J,K), while MRF4 is still only expressed in the dorsal myotome. By HH stage 26, all four MRFs are expressed throughout the entire myotome (Fig. 3D,H,L,P).

Expression in limb muscles
The muscles of the limb are derived from somitic migratory muscle precursors that express the transcription factors Pax3 and Lbx (Dietrich, 1999;Otto et al. 2006). These cells delaminate from the ventro-lateral lip of the dermomyotome then migrate into the developing limb where they form the dorsal and ventral muscle masses. Only at this point do they begin to express MRFs.
Whole-mount in situ hybridisation shows Myf5 expression in the forelimb at HH stage 22 (Fig. 4B), with expression in the hindlimb detected soon after at HH stage 24 (Fig. 4C). In sections, Myf5 can be seen in both dorsal and ventral muscle masses in the forelimb at HH stage 22 (Fig. 5B), and weak expression can also be seen in hindlimb myogenic cells (Fig. 6B). Myf5 expression is then maintained throughout the developing limb muscles at stages HH stage 24 to HH stage 30 (Figs 4D-F 6C,D). At HH stage 30, Myf5 staining becomes less intense, especially in distal muscles, as differentiation proceeds (Fig. 4f).
In contrast to Myf5, MyoD expression is first detected in dorsal and ventral muscle masses in the hindlimb at HH stage 23 (Fig. 4I), and then in the forelimb at HH stage 24 (Figs 3J, 5G and 6G). Expression is then maintained in the developing muscles throughout the limb up to HH stage 30 (Figs 4K,L, 5H and 6H).
Mgn is first seen in both fore-and hindlimbs at HH stage 24 (Figs 4P, 5K and 6K), and is then maintained to HH stage 30 (Figs 4Q,R, 5L and 6L). MRF4 is not detected in limb muscles until muscle differentiation is well advanced at HH stage 30 (Fig. 4W), although some weak staining can be seen in sections through hindlimbs at HH stage 26 (Fig. 6P).

Concluding remarks
The current data show a clear temporal progression in both somites and limbs where Myf5 is expressed first followed by MyoD. Subsequently, Mgn and MRF4 are expressed with similar dynamics. This is consistent with current models in which Myf5 and MyoD are required for myogenic commitment while Mgn and MRF4 regulate differentiation. Although MRF4 expression has been reported earlier or contemporaneously with Myf5 in mouse somites (Summerbell et al. 2002) and can act to specify muscles in the absence of both Myf5 and MyoD (Kassar-Duchossoy et al. 2004), the current data imply that this function is not conserved in chicken embryos where MRF4 is only expressed later in myogenic development.
Expression of Myf5 in the paraxial mesoderm in HH stages 8-14 in the À1 somite position was also observed. This is consistent with previous reports that have also observed Myf5 expression prior to somite formation (Kiefer & Hauschka, 2001). However, it is apparent when comparing larger numbers of embryos at these stages that there is heterogeneity in these samples with some embryos showing this expression and others not. This is observed even in embryos harvested, processed and stained in a single batch. One pos- sible explanation is that the in situ protocol produces variable results; however, staining in other tissues, such as the limbs, is far more reproducible. Therefore, we believe that a more likely explanation is that they are capturing embryos at subtly different stages and that those that do not show this expression have just formed the newest somite and Myf5 expression is yet to be initiated in the paraxial mesoderm prior to somite formation. The regulation of Myf5 during development is under the control of a large number of diverse enhancer elements reviewed in Francetic & Li (2011), which may help to explain this complex expression during somitogenesis. A degree of heterogeneity in the staining of Mgn and MRF4 in somites between HH stages 12 and 16 was also observed. This may simply reflect variable efficiency of staining in different embryos. However, this variability is still observed in embryos with the same somite number stained in a single batch, and is not apparent in limb buds where staining of all probes is highly consistent and reproducible. There are two possibilities for this: either the current protocol does not reliably detect weaker staining in more posterior somites in all cases or there is some variability in the precise timing of onset of these later MRFs between different embryos, possibly because these embryos do not come from an isogenic inbred line and the difference in time between stages is small; as new somites are added every 90 min the time between one HH stage and the next is only 4.5 h.
One notable difference between limb and somite staining is relative timing of Mgn and MRF4 expression. In somites these genes are expressed at the same point, although the domain of MRF4 within the myotome is more restricted than that of Mgn. In contrast, Mgn expression in limb muscles precedes that of MRF4. It is clear that different muscle groups have different regulatory networks driving their differentiation (Mok & Sweetman, 2011), and this may explain the distinct staining dynamics observed. Also clearly different shapes in myotomes at different axial levels are observed. At limb levels the myotome does not extend dorso-medially to the same extent as it does in the interlimb region. This is presumably to facilitate the migration of limb muscle precursors, although limb level somites do also produce the pectoral muscles by myotomal extension (Beresford et al. 1978), so this difference in shape does not preclude this mode of myogenesis.
It is also apparent that the data presented do not entirely agree with some previous reports that have detected Myf5 expression in primitive streak at HH stage 3 and paraxial mesoderm and HH stage 5 (Kiefer & Hauschka, 2001). Despite extensive staining, expression has not been seen in these early embryos. This is presumably due to differences in the staining protocols and probes used, and it may be that the current approach sacrifices some sensitivity for specificity.
One intriguing question that remains is to what extent the myoblasts in a particular region of the embryo are homogeneous or consist of distinct subpopulations with different characteristics. Previous reports have already identified differences in the expression domains of Myf5 and MyoD within the dorsal and ventral muscle masses of the limb (Delfini et al. 2000), and ablation experiments have also suggested that there are Myf5-independent (Gensch et al. 2008;Haldar et al. 2008) but not MyoD-independent (Wood et al. 2013) myogenic lineages within the embryo, although this view has been challenged (Comai et al. 2014;Haldar et al. 2014). However, the current in situ hybridisations do not provide single cell resolution and distinct populations cannot reliably be identified, so the resolution of this question will require further work.