A monoclonal antibody specific for Xenopus MyoD (XMyoD) has been characterized and used to describe the pattern of expression of this myogenic factor in early frog development. The antibody recognizes an epitope close to the N terminus of the products of both XMyoD genes, but does not bind XMyfS or XMRF4, the other two myogenic factors that have been described in Xenopus. It reacts in embryo extracts only with XMyoD, which is extensively phosphorylated in the embryo. The distribution of XMyoD protein, seen in sections and whole-mounts, and by immunoblotting, closely follows that of XMyoD mRNA. XMyoD protein accumulates in nuclei of the future somitic mesoderm from the middle of gastrulation. In neurulae and tailbud embryos it is expressed specifically in the myotomal cells of the somites. XMyoD is in the nucleus of apparently every cell in the myotomes. It accumulates first in the anterior somitic mesoderm, and its concentration then declines in anterior somites from the tailbud stage onwards.

MyoD, a member of the helix-loop-helix family of transcription factors, can activate muscle-specific gene expression in non-muscle cells in culture (reviewed by Weintraub et al., 1991). There is increasing evidence that a Xenopus homologue of MyoD is active in the early stages of muscle differentiation in the embryo. A small and unlocalized maternal content of XMyoD mRNA (Hopwood et al., 1989; Harvey, 1990) begins to rise as a result of ubiquitous transcriptional activation at the mid-blastula transition (Rupp and Weintraub, 1991), followed by a more substantial, localized increase in the concentration of XMyoD transcripts from the start of gastrulation, about two hours before the appearance of muscle differentiation markers and more than half a day before myofibrils are first visible (Hopwood et al., 1989; Harvey, 1990; Scales et al., 1990; Muntz, 1975). The XMyoD gene is activated in response to mesoderm induction (Hopwood et al., 1989; Harvey, 1990) in the myotonies (Hopwood et al., 1989), the parts of the somitic mesoderm that form the axial musculature of the tadpole, and from which the skeletal muscles of the trunk develop. Ectopic expression of XMyoD activates muscle-specific gene expression in early embryonic cells of a non-muscle lineage (Hopwood and Gurdon, 1990), in the case of the cardiac actin gene probably by interacting directly with upstream promoter elements (Taylor et al., 1991).

Although a considerable amount is known about the patterns of accumulation of the transcripts of myogenic genes during the development of three vertebrate classes (Hopwood et al,, 1989; Harvey, 1990; Scales et al., 1990; Hopwood et al., 1991; Charles de la Brousse and Emerson, 1990; Sassoon et al., 1989; Ott et al., 1991), there is no equivalent information about the protein products of these genes. In fact, the study of myogenic proteins has so far been restricted to cultured mammalian cells (e.g. Tapscott et al., 1988). In this report we analyze expression of a myogenic protein in a normally developing vertebrate embryo. A monoclonal antibody specific for XMyoD is described, and used to extend our knowledge of the embryonic expression of this factor beyond that deduced from biochemical and in situ hybridization studies of the mRNA. We wished to know, in particular, whether the pattern of expression of XMyoD protein is like that of the mRNA, whether the protein is modified post-translationally, whether XMyoD is located in nuclei from early stages, and whether the same amount of XMyoD accumulates in all cells of the myotomes.

Antibody production

A large pan of the XMyoD protein (amino acid residues 3 to 195) was expressed as a fusion to glutathione S-transferase (GST): pGEX3X-XMyoD3/195 was made by cloning an Alai fragment of XMyoD2-24 (nts 143–723 of Hopwood el al., 1989) into the SmiA site of pGEX-3X (Smith and Johnson, 1988). Fusion protein was purified essentially as described (Smith and Johnson, 1988), except that we used the protease inhibitors, PMSF, EDTA, o-phenanthroline, and p-amino-benzamidinc (all at 2 mM), and protein was recovered from glutathione-agarose beads by column elution with either glutathione or 50 mM NaHCO3, pH 10.8 (K. S. Johnson, personal communication).

Monoclonal antibodies were raised against the GST-XMyoD3/195 fusion protein by standard procedures (Harlow and Lane, 1988). Briefly, CFA x BALB/c Fl mice were injected intraperitoneally with 50 μg fusion protein in Freund’s adjuvant at two-week intervals. After four injections, mice were rested for at least two months, and then given an intraperitoneal injection of 100 μg fusion protein. Mice were killed four days later, and splcnocytes prepared and fused with the mouse myeloma cell line Sp2/0-Agl4 by treatment with PEG 1500 (Boehringer). Hybrids were selected by treatment with azaserine (1 μg ml−1). Clones secreting antibodies to the fusion protein were identified by a solid phase immunoassay, which used fusion protein bound to nitrocellulose as the support, and detected antibody binding by horseradish peroxidase (HRP)-conjugated anti-mouse 1g (Amersham) and an enhanced chemiluminescence (ECL) system (Amersham). Cells from positive wells were single cell cloned by limiting dilution at least three times before use. The D7F2 antibody was typed using a BioRad Typer kit as an immunoglobulin xGl. The source of the antibody for experiments was tissue culture fluid in which the hybridoma cells had been allowed to grow to confluency.

Immunoblotling and epitope mapping

Proteins separated by SDS-PAGE (11%; Laemmli, 1970) were transferred to nitrocellulose as described (Towbin et al., 1979). After rinsing in Tris-buffered saline, pH 8.1 (TBS), and blocking for 1 h in TBS, 5% (w/v) nonfat milk powder (TBSM), filters were probed for 1–2 h at room temperature with a 1 in 4 dilution of D7F2-containing tissue culture fluid in TBSM. After washing for 1 h in several changes of TBSM, blots were incubated for 1 h in TBSM containing a 1 in 1000 dilution of HRP-conjugated anti-mouse lg. Blots were washed for 30 min each in several changes of TBSM, then 0.2% SDS, 10 mM Tris, pH 7.5, and finally TBS. Antibody binding was revealed using the ECL system.

The region containing the epitope recognized by D7F2 was approximately located by inspection of the size of the smallest GST-containing proteolytic fragments of GST-XMyoD3/195, and mapped more precisely by testing the binding of D7F2 to fusions of GST and peptides from this region. pGEX2T-XMyoDlfyi35 was made by cloning a PCR fragment derived from XMyoD2-24 between the BumHl and EcoRJ sites of pGEX-2T.

In vitro transcription and translation

The transcription templates pSP64T-XMyoD(b) and pSP64T-XMyf5 have been described (Hopwood and Gurdon, 1990; Hopwood ct al., 1991). XMyoDa mRNA was made from a cDNA (pBS-XMyoD36) isolated by low stringency hybridization to a rat myogenin probe (Hopwood et al., 1991), which was shown by DNA sequencing to have the same coding capacity as the XMyoDa cDNA of Harvey (1990), except that E68 is absent, K.258 is E, and N284 is T, all changes that make XMyoD36 more like the X1mf2S cDNA of Scales et al. (1990). pSP64T-XMRF4 was made from a XMRF4 cDNA clone (pCJM4.2) provided by Charles Jennings, ft was constructed using PCR with the oligonucleotides 5’ CTCCCATGGA-GATGATGGACCTATTTG 3’ and 5’ GAGGGTAACCT-TAATTCTCTACCAGCTCCTG 3’ to make a coding region fragment with 5’ Ncol and 3’ ZB7EII ends, which was inserted between the corresponding sites of pSP64-Xβm (Krieg and Mellon, 1984) to replace the β-globin coding region. pSP64T-XMRF4 was sequenced to show that no mutations had been introduced by the PCR. All pSP64T templates were linearized with Xhal and transcribed with SP6 RNA polymerase; pBS-XMyoD36 was linearized with Hindlll and transcribed with T7 RNA polymerase. Capped mRNAs were synthesized as described (Krieg and Melton, 1987). mRNAs (25 μg ml−1) were translated in rabbit reticulocyte lysate (bought from Tim Hunt, Dept of Biochemistry, University of Cambridge) containing [J5S| me th ioninc (750 μ Ci ml−1) as described by Jackson and Hunt (1983).

Immunoprecipdation

Protein samples were diluted at 4°C into 770 pl RIPA buffer (50mM NaCl, 25 mM Tris, pH 8.2, 0.5% NP40, 0.5% sodium deoxycholate, 0.1% SDS, 0.1% sodium azide, 50 μg ml−1 aprotinin, 50 μg ml−1 leupeptin, and 2 mM PMSF) containing 1% globulin-free BSA, to which were added 80 μ l 50% protein A-sepharose (BioProcessing), and 150 μg antibodycontaining tissue-culture medium. After3 h incubation at 4°C, and 5 × 1 ml washes with RIPA, bound proteins were eluted from the protein A-sepharose beads by incubation in 100 μl 7 M urea in TBS at room temperature for 15 min. The protein A-sepharose was removed, and to the supernatant was added 1 ml TBS containing 50 μg ml−1 aprotinin and 50 μg ml−1 leupeptin. After incubation overnight at 4°C, 80 μl 50% protein A-sepharose was added, incubated for 2 h at 4°C, washed twice with RIPA, and bound protein eluted with SDS sample buffer. Treatment with bacteria! alkaline phosphatase (Tapscolt et al., 1988) preceded the urea elution step, which was omitted when only proteins made in vitro were precipitated (Fig. 2). Proteins were separated by SDS-PAGE (11%), and gels dried and exposed to X-ray film cither directly (Fig. 2), or after fluorography (Fig. 4; Bonner and Laskey, 1974).

Embryos and embryo extracts

Xenopus laevis embryos were cultured as described (Gurdon, 1977) and staged according to Nieuwkoop and Faber (1967). Synthetic mRNA (14 ng in 14 nl per embryo) was microinjected as described (Hopwood et al., 1991), and [:!5S]meth-ioninc (2 μCi in 14 nl per embryo) injected in the same way, and, when appropriate, at the same time. Embryo extracts were prepared essentially as described (Taylor ct al., 1991). Frozen embryos or tissues dissected as described (Hopwood et al., 1991) were homogenized at 4°C in 25% glycerol, 50 mM KCI, 50 mM Tris, pH 8.0, 0.1 mM EDTA, 2 mM DTT, 2 mM PMSF, 50 μg ml−1 aprotinin, and 50 μg ml−1 leupeptin, and centrifuged (4°C, 13,000 g, 8 min). The supernatant was further clarified by a second centrifugation (4°C, 13,000 g, 4 min).

Immunohistochemistry

Staining of whole albino embryos essentially followed the procedure of Hemmati Brivanlou and Harland (1989), except that we detected antibody binding with alkaline phosphatase. Embryos from which the vitelline membranes had been removed by hand were fixed in MEMFA for 1–2 h and, after a 15 min wash in methanol at room temperature, stored overnight or for longer in methanol at —20°C. We used D7F2-containing tissue culture medium diluted 1 in 18, followed after washing by a rabbit anti-mouse lg bridging antibody (1CN, 1 in 200 dilution), and then an alkaline phosphatase-anti-alkaline phosphatase complex (APAAP; Boehringer, 312 U ml−1). Incubations were carried out in 1.5 ml plastic lubes, on a TAAB type N 4 rpm rotator. Substrates were 5-bis-4-chloro-3-indolyl phosphate and nitro blue tétrazolium (BCIP/NBT; blue-purple), and Vector’s substrate kit I (red). Colour reactions were performed al room temperature in 100 mM Tris, pH 9.5, 100 mM NaCI, 50 mM MgCl2, 1 mM levamisole (BCIP/NBT) or 100 mM Tris, pH 8.2, 1 mM levamisolc (Vector Red) for 1–4 h with periodic agitation. Embryos were washed in 10 mM Tris, pH 8.0, 1 mM EDTA, and given two 10–15 min washes in methanol, before clearing in Murray’s solution. For double labelling, embryos stained for XMyoD using BCIP/NBT were rehydrated, treated with 0.2 M glycine, pH 2.6 for 1 h, washed in PBT and the staining procedure repeated with 12/101 (1 in 200 dilution of purified antibody provided by J. Brockes) using the red substrate solution.

Sections (8 μm) were cut from wild-type embryos fixed as above and embedded in Histoplast (Shandon). Rehydrated sections were stained with D7F2 (t in 4, 2 h), rabbit antimouse Ig (1 in 200, 1 h), and APAAP (3 U ml−1, 1 h), with half-hour washes after each layer, and the colour reaction allowed to proceed for 2–4 h. Sections were rapidly dehydrated through an alcohol series for mounting in DPX, or stained with Hoechst 33258 (5 μg ml’), and mounted in PBS-glycerol. Whole-mounts to be sectioned were transferred from Murray’s clear directly into xylene and then Histoplast.

A monoclonal antibody specific for XMyoD, a somitespecific phosphoprotein

Monoclonal antibodies were raised against a GST-XMyoD fusion protein, and the one that bound most avidly to XMyoD was selected for further characterization. This antibody, D7F2, recognizes an epitope near the N terminus of XMyoD: on a Western blot it bound to a GST fusion protein containing only residues 16 to 35 (GSLCAFPTPDDFYDDPCFNT) of XMyoD-de-rived sequence, but not to GST alone (Fig. 1). We showed by immunoprecipitation that D7F2 binds to XMyoD made in a rabbit reticulocyte lysate, but not to the other myogenic factors of Xenopus that have so far been described, XMyf5 (Hopwood et al., 1991) and XMRF4 (C. Jennings, personal communication) (Fig. 2). There are two different kinds of XMyoD transcripts in Xenopus laevis (Harvey, 1990; Scales et al., 1990), probably as a result of the genome duplication that occurred during the evolution of this species: D7F2 recognizes the proteins translated from both of them (Fig. 2).

Fig. 1.

D7F2 recognizes an epitope near the N terminus of XMyoD. Immunoblot of affinity-purified GST and GST-XMyoD fusion proteins, showing that D7F2 binds to the GST-XMyoD3/195 protein used as the immunogen, and to a fusion protein containing only amino acid residues 16 to 35 of XMyoD (GST-XMyoD3/35), but not to GST alone. Protein from 4 μl or (for GST-XMyoD3/195) 30 β2 culture was loaded in each lane. Coomassie staining of a gel on which was loaded 25 times more protein than was blotted showed that similar amounts of protein were loaded in each lane. Numbers to right of this and subsequent figures give M, × 10−3 of marker proteins. Five-second exposure.

Fig. 1.

D7F2 recognizes an epitope near the N terminus of XMyoD. Immunoblot of affinity-purified GST and GST-XMyoD fusion proteins, showing that D7F2 binds to the GST-XMyoD3/195 protein used as the immunogen, and to a fusion protein containing only amino acid residues 16 to 35 of XMyoD (GST-XMyoD3/35), but not to GST alone. Protein from 4 μl or (for GST-XMyoD3/195) 30 β2 culture was loaded in each lane. Coomassie staining of a gel on which was loaded 25 times more protein than was blotted showed that similar amounts of protein were loaded in each lane. Numbers to right of this and subsequent figures give M, × 10−3 of marker proteins. Five-second exposure.

Fig. 2.

D7F2 recognizes both forms of XMyoD, but not other Xenopus myogenic factors, XMyoDa and XMyoDb, but not XMyf5 or XMRF4, are precipitated by D7F2, but not by a control anti-nucleoplasmin monoclonal antibody (PA1C2; Dingwall et al., 1987). 0.8 μl in vitro translate loaded directly; 4 μl added to each immunoprecipitation. Autoradiograph of’’S-labelled protein, overnight exposure.

Fig. 2.

D7F2 recognizes both forms of XMyoD, but not other Xenopus myogenic factors, XMyoDa and XMyoDb, but not XMyf5 or XMRF4, are precipitated by D7F2, but not by a control anti-nucleoplasmin monoclonal antibody (PA1C2; Dingwall et al., 1987). 0.8 μl in vitro translate loaded directly; 4 μl added to each immunoprecipitation. Autoradiograph of’’S-labelled protein, overnight exposure.

The specificity of D7F2 for XMyoD was further shown by immunoblotting of embryo extracts, w’hich also allowed us to investigate the tissue- and developmental stage-specificity of the protein. No proteins were bound in extracts of eggs (data not shown) or untreated blastulae (Fig. 3), a result which was anticipated because there is little XMyoD mRNA before gastrulation (Hopwood et al., 1989). However, in extracts of XMyoD mRNA-injected blastulae the antibody did bind to a group of proteins, which migrated with or more slowly than XMyoD translated in vitro (Fig. 3, and see below). Proteins of the same sizes were recognized by D7F2 in extracts of uninjected late gastrulae (data not shown) and neurulae (Fig. 3), thus demonstrating the specificity of D7F2 for XMyoD, and showing that the accumulation of XMyoD protein roughly follows that of the mRNA, the content of which increases about fifteen-fold during gastrulation (Hopwood et al., 1989). Analysis of protein from dissected tissues of late neurulae showed that XMyoD is somitespecific (Fig. 3).

Fig. 3.

D7F2 binds specifically to XMyoD in embryo extracts on an imrnunoblot. Analysis of in vitro translated XMyoD (IVT), XMyoD-injected (inj.) versus uninjected (uninj.) blastulae (st9), and whole and dissected late neurulae (st 18). Five whole embryos or dissected parts of 20 embryos were loaded in each lane. 50 nl in vitro translate or one-eighth injected embryo were mixed with five uninjected blastulae. Ponceau S staining of the filter showed similar amounts of total protein in each lane, as also indicated by re-probing with PA1C2. Neurect + noto, neurectoderm and notochord; vl ect + mes, ventral ectoderm and mesoderm. Seven-minute exposure.

Fig. 3.

D7F2 binds specifically to XMyoD in embryo extracts on an imrnunoblot. Analysis of in vitro translated XMyoD (IVT), XMyoD-injected (inj.) versus uninjected (uninj.) blastulae (st9), and whole and dissected late neurulae (st 18). Five whole embryos or dissected parts of 20 embryos were loaded in each lane. 50 nl in vitro translate or one-eighth injected embryo were mixed with five uninjected blastulae. Ponceau S staining of the filter showed similar amounts of total protein in each lane, as also indicated by re-probing with PA1C2. Neurect + noto, neurectoderm and notochord; vl ect + mes, ventral ectoderm and mesoderm. Seven-minute exposure.

Most of the protein recognized by D7F2 on immunoblots migrates more slowly than XMyoD made in vitro.

It is known that phosphorylation is responsible for similarly retarded migration of most of the mouse MyoD protein in cultured myoblasts derived by 5-azacytidine treatment of C3H10T,/2 fibroblasts (Tap-scott et al., 1988); we wished to find out whether the same is true of XMyoD in normally developing embryonic cells. D7F2-Ímmunoprecipitated protein from XMyoD mRNA-injected blastulae was treated with bacterial alkaline phosphatase. This treatment, but not mock incubation without the enzyme, largely reduced the set of bands to a single one, which comigrated with XMyoD synthesized in vitro (Fig. 4). Whilst we do not know that D7F2 recognizes all of the potentially various forms of XMyoD in embryos, at least it clearly does bind both to phosphorylated and unphosphorylated forms of the protein. The large effect of phosphorylation on its rate of migration suggests that XMyoD is extensively modified in this way, which would mean that there is much potential for regulation of its activity at this level.

We conclude that the monoclonal antibody D7F2 binds only to XMyoD, which is a somite-specific phosphoprotein in Xenopus embryos. Having established the specificity of the antibody, we have used it to examine the distribution of the protein during the early stages of Xenopus development. The location of XMyoD-expressing cells in early embryos was determined by immunostaining sections and whole embryos with D7F2. Whole mounts provide a clearer view of the overall pattern of expression of XMyoD, but it is necessary to look at sections to see accurately which structures contain the expressing cells. Direct staining of sections also controls for possible artefacts of poor penetration of the staining solutions into the centres of whole embryos.

Expression of XMyoD in gastrulae

Il is of particular interest to discover the pattern of XMyoD protein accumulation in gastrulae. XMyoD mRNA is unlocalized in blastulae, but was found by in situ hybridization to accumulate specifically in the mesoderm of early mid-gastrulae, though it was not possible to ascertain in which mesodermal cells (Hopwood et al., 1989). It was detected in only the somitic mesoderm of slightly later gastrulae (stll1/?; Hopwood et al., 1991).

The earliest stage at which we have been able to delect XMyoD protein was in mid-gastrulae (stll). By ?sing immunocytochemistry we could examine the paitem of expression more easily, with higher resolution, and slightly earlier than was achieved for the mRNA with hybridized sections, and so compare it more confidently to the fate maps (Keller, 1976). A collar of labelled nuclei around the closing blastopore (Fig. 5A) was first seen clearly at stll (Fig. 5B), and was weakly detectable in a few specimens at about st 10]/2 (data not shown). The gap in the ring of labelled nuclei (Fig. 5A, B) corresponds to the dorsal cells that will form the notochord. The majority of labelled nuclei, and also those that label most strongly, lie on either side of this unstained region, clearly in the dorsolateral positions occupied by the most anterior cells of the future somitic mesoderm (Keller, 1976).

Fig. 4.

XMyoD is a phosphoprotein. Immunoprecipitation of in vitro translated XMyoD (IVT) or extracts of XMyoD mRNA-injected (i), or uninjected (u) blastulae. Bacterial alkaline phosphatase treatment, but not mock treatment without the enzyme, reduces the set of bands immunoprecipitated from injected embryos largely to a single band that comigrates with XMyoD made in vitro. Fluorograph of 35S-labelled protein. D7F2 was incubated with 0.5 μ l in vitro translate or extract of 9 blastulae (st9). There was as much total labelled protein in the extract of uninjected as in that of injected embryos, and about the same amount of labelled nucleoplasmin was precipitated from each extract by PA1C2. Three-day exposure.

Fig. 4.

XMyoD is a phosphoprotein. Immunoprecipitation of in vitro translated XMyoD (IVT) or extracts of XMyoD mRNA-injected (i), or uninjected (u) blastulae. Bacterial alkaline phosphatase treatment, but not mock treatment without the enzyme, reduces the set of bands immunoprecipitated from injected embryos largely to a single band that comigrates with XMyoD made in vitro. Fluorograph of 35S-labelled protein. D7F2 was incubated with 0.5 μ l in vitro translate or extract of 9 blastulae (st9). There was as much total labelled protein in the extract of uninjected as in that of injected embryos, and about the same amount of labelled nucleoplasmin was precipitated from each extract by PA1C2. Three-day exposure.

Fig. 5.

XMyoD expression in gastrulae. (A) st 1112 wholemount, showing collar of stained nuclei around the blastopore; (B) higher power view of the area surrounding the yolk plug of a sill embryo, focused to show the gap in dorsal staining, which corresponds to the prospective notochord; (C, D) slightly oblique sections of stll wholemount, cut in approximately the planes indicated in (B). Some of the yolk plug nuclei are faintly stained in (D), a non-specific consequence of rather heavy staining, that we have seen also with other primary antibodies.

Fig. 5.

XMyoD expression in gastrulae. (A) st 1112 wholemount, showing collar of stained nuclei around the blastopore; (B) higher power view of the area surrounding the yolk plug of a sill embryo, focused to show the gap in dorsal staining, which corresponds to the prospective notochord; (C, D) slightly oblique sections of stll wholemount, cut in approximately the planes indicated in (B). Some of the yolk plug nuclei are faintly stained in (D), a non-specific consequence of rather heavy staining, that we have seen also with other primary antibodies.

We are uncertain of the fate of the XMyoD-expressing cells around the ventral side of the blastopore (Fig. 5A, D), because the ventrolateral and ventral marginal zone of the early gastrula (st 10+) gives rise both to posterior somitic mesoderm and to lateral-ventral mesoderm of the neurula, the somitic mesoderm developing from the cells that involute later (Keller, 1976). In sections of sill embryos we sec XMyoD only in cells in the interior of the embryo, that is, only in post-involution mesoderm (Fig. 5C, D), but this information alone is not enough to tell us the fate of the expressing tissue, because the middle stages of gastrulation have not been fate mapped in detail. If XMyoD protein accumulates in cells other than those of the prospective somitic mesoderm of the gastrula, this accumulation must be transient, because by the end of gastrulation we see no staining that is clearly outside the regions fated to form somites (see below).

Staining by D7F2 of gastrula nuclei is less even, from one nucleus to another, than that seen at later stages (see below). This might mean that slight differences are more apparent when the staining is weaker, or more interestingly, that cells in the same part of the embryo begin to accumulate XMyoD at slightly different times.

Distribution of XMyoD protein in neurulae and tailbud embryos: specific and uniform expression apparently in all myotomal cells

From the beginning of neurulation until the tailbud stage, XMyoD protein is found exclusively in the nuclei of premyotomal and myotomal cells (Fig. 6A, B), which in Xenopus, unlike in the amniotes, constitute all, or nearly all, of the overtly segmented paraxial mesoderm (Hamilton, 1969). In later embryos there are other sites of expression, notably in the muscles of the head (N.D.H., unpublished results). We used the 12/101 monoclonal antibody (Kintner and Brockes, 1984) to identify unambiguously the myotomal cells of the somites. Staining sections of tailbud embryos for both XMyoD and the unidentified muscle differentiation marker recognized by the 12/101 antibody showed, for the middle of the embryos where both antigens are expressed strongly (see below), that all XMyoD-expressing cells also contain 12/101 antigen (Fig. 6C). Having established that only myotomal cells express XMyoD, we wished to know what proportion of them do so. Since, at later stages, a nucleus in these elongated cells is not cut by many of the sections that cut the cell containing it, this was seen most easily by staining sections with D7F2 and Hoechst 33258, which binds DNA in all nuclei. This confirmed the nuclear location of XMyoD and, importantly, showed for neurulae and early tailbud embryos that all of the nuclei within the histological boundaries of the myotomes express XMyoD (Fig. 6D) (see Discussion).

Fig. 6.

XMyoD is present only in myotomal nuclei of neurulae and lailbud embryos. Transverse sections of (A) an early neurula (stl2½-13), and (B) an early tailbud embryo (st23) stained with D7F2. (C) Section through the posterior part of the segmented region of a tailbud embryo (st12 ½6) stained for XMyoD (blue-purple) and 12/101 antigen (red), showing that all XMyoD-expressing nuclei are in 12/101-positive cells. (D) Section through mid-neurula (stl5-16) stained with D7F2 for XMyoD and with Hoechst 33258 to show all nuclei. XMyoD non-expressing nuclei show the bright, light blue Hoechst fluorescence; XMyoD-expressing nuclei show a darker blue due to Hoechst fluorescence shining through the dark BCIP/NBT blue purple reaction product. AH clearly myotomal nuclei express XMyoD.

Fig. 6.

XMyoD is present only in myotomal nuclei of neurulae and lailbud embryos. Transverse sections of (A) an early neurula (stl2½-13), and (B) an early tailbud embryo (st23) stained with D7F2. (C) Section through the posterior part of the segmented region of a tailbud embryo (st12 ½6) stained for XMyoD (blue-purple) and 12/101 antigen (red), showing that all XMyoD-expressing nuclei are in 12/101-positive cells. (D) Section through mid-neurula (stl5-16) stained with D7F2 for XMyoD and with Hoechst 33258 to show all nuclei. XMyoD non-expressing nuclei show the bright, light blue Hoechst fluorescence; XMyoD-expressing nuclei show a darker blue due to Hoechst fluorescence shining through the dark BCIP/NBT blue purple reaction product. AH clearly myotomal nuclei express XMyoD.

Apart from the general changes in the concentration of XMyoD during development that are reflected in unequal staining along the anterior-posterior axis (see below), there are no large differences in the intensity of staining between one D7F2-positive nucleus and another (Figs 6, 7), though there is more heterogeneity in the stained regions of gastrulae (Fig. 5). This uniform expression resembles that observed in differentiating mouse myotubes in culture, but is quite unlike the rather wide variation seen in cultured myoblasts before they are stimulated to differentiate (Tapscott et al., 1988).

Fig. 7.

Expression of XMyoD in neurulae and lailbud embryos. Whole-mount immunocytochemistry with D7F2. (A) Early neurula (st ½13); (B) mid-late neurula (st17); (C-E) late neurulae (st 19), (D) same embryo as in C also stained with 12/101 (red), (E) side view of newly formed somites; (F) early tailbud (sl24); (G, H) later tailbud (st27), (H) close-up of developing tail where myotomes are forming, the major site of XMyoD expression at this stage. Note that XMyoD and 12/101 staining in D can be distinguished on the basis of subcellular location, even if the secondary antibodies used to delect 11/101 react with any unremoved D7F2. Some staining of the archenteron contents is visible in F, which we judge to be a result of non-specific sticking of the antibodies, because we have seen it also with several other monoclonal antibodies and when no primary antibody was used, but not when all antibodies were omitted. Colour reactions were allowed to proceed so as to give the highest signal over background, that is for longer for the early ncurula and tailbud embryos, than for the more strongly expressing late neurulae. Anterior of all embryos to the left.

Fig. 7.

Expression of XMyoD in neurulae and lailbud embryos. Whole-mount immunocytochemistry with D7F2. (A) Early neurula (st ½13); (B) mid-late neurula (st17); (C-E) late neurulae (st 19), (D) same embryo as in C also stained with 12/101 (red), (E) side view of newly formed somites; (F) early tailbud (sl24); (G, H) later tailbud (st27), (H) close-up of developing tail where myotomes are forming, the major site of XMyoD expression at this stage. Note that XMyoD and 12/101 staining in D can be distinguished on the basis of subcellular location, even if the secondary antibodies used to delect 11/101 react with any unremoved D7F2. Some staining of the archenteron contents is visible in F, which we judge to be a result of non-specific sticking of the antibodies, because we have seen it also with several other monoclonal antibodies and when no primary antibody was used, but not when all antibodies were omitted. Colour reactions were allowed to proceed so as to give the highest signal over background, that is for longer for the early ncurula and tailbud embryos, than for the more strongly expressing late neurulae. Anterior of all embryos to the left.

Distribution of XMyoD protein in neurulae and tailbud embryos: evolution of the pattern of expression

The mesoderm differentiates from anterior to posterior. At stl2½-13 anterior myotomal nuclei stain more heavily than posterior ones (Fig. 7A), but by st 17 D7F2 staining of posterior cells has become stronger, so that in mid-late neurulae XMyoD is expressed at the same level in nuclei along the length of the anterior-posterior axis (Fig. 7B). XMyoD continues to be expressed after segmentation has begun in late neurulae, in both the anterior, segmented myotomes, and the more posterior prernyotomal mesoderm (Fig. 7C, E). This contrasts with accumulation of the 12/101 antigen only in the anterior, already segmented mesoderm, and the region just posterior to it (Fig. 7D), and reflects the expression of XMyoD much earlier in the process of muscle cell differentiation. The presence of XMyoD in more of the myotomal mesoderm of the late neurula than stains with 12/101 means that in experimental situations D7F2 would be expected to identify a higher proportion of muscle cells than would 12/101 at this stage. As the myotomes differentiate, the nuclei become neatly aligned in a single band in the middle of each myotome (Fig. 7F-H), reflecting their position in the centre of the elongated, mononucleate cells that span them (Muntz, 1975).

In early tailbud embryos (from about st23), D7F2 staining becomes progressively weaker in the anterior of the embryo, the major site of accumulation of XMyoD protein in later embryos being the forming myotomes of the tail (Fig. 7G, H). The greater prominence of XMyoD expression in posterior tissue is due not only to its higher concentration in each nucleus, but also to the higher density of nuclei in less differentiated tissue. Apparently weaker staining of anterior tissue might have arisen artefactually had fixation become relatively less effective as the myotonies differentiated, but this is unlikely, since 12/101 strongly stained anterior sections of later embryos that failed to stain detectably with D7F2, whilst conversely staining only weakly or not at all the posterior sections, which did stain with D7F2 (data not shown). This fall in the level of XMyoD protein follows the pattern of mRNA accumulation observed previously (Hopwood et al., 1989), and is consistent with the idea that a lower concentration of XMyoD protein is required to maintain muscle differentiation than to establish it.

The work presented above extends to the level of protein our knowledge of the embryonic expression of a vertebrate member of the MyoD family. The pattern of expression of XMyoD protein closely follows that of the mRNA, suggesting that transcripts are not translated differentially. Like that of the mRNA, the concentration of XMyoD protein in anterior myotomes falls from the tailbud stage. We were not able to analyze the anterior-posterior distribution of the mRNA early enough to know if it also accumulated first in anterior somitic mesoderm, but for the protein, probably because it appears slightly later, and staining whole embryos provides a much clearer view, we could see that more was present in anterior than in posterior nuclei of the somitic mesoderm at the start of neurulation. We therefore suggest that expression of XMyoD not only declines first in the anterior cells, but also begins there first.

The presence of XMyoD protein from the mid-gastrula stage in the nuclei of cells that will form the skeletal muscle of the tadpole strengthens the case for its involvement in the early events of muscle differentiation io the embryo. However, the fate map is not sufficiently precise to establish whether the XMyoD-expressing cells around the ventral side of the blastopore of the mid-gastrula will normally make muscle, and therefore continue to express XMyoD, or whether they are fated to form ventrolateral mesoderm of the neurula, and so express XMyoD only transiently. Several experimental approaches could help to identify these cells, but the task is made demanding by their relatively low content of the protein. It is already known that cells can express XMyoD transiently and not make muscle: the low concentration of XMyoD mRNA in eggs and blastulae (Hopwood et al., 1989; Harvey, 1990; Rupp and Weintraub, 1991) does not normally lead to muscle gene activation in all of the cells that contain it; and ectopic expression in animal caps of even rather high concentrations of XMyoD mRNA (Hopwood and Gurdon, 1990) and protein (N.D.H. et al., manuscript in preparation) need not lead to stable myogenesis.

By extending the description of the expression of XMyoD to cellular resolution, we have found that, apart from modulation along the anterior-posterior axis, XMyoD accumulates equally in all myoComal nuclei of neurulae and early tailbud embryos. Staining was more heterogeneous when first detected in mid-gastrulae, though exclusively nuclear even at this stage, but was uniform from the beginning of neurulation. Immunocytochemical studies of the expression of MyoD relatives in two invertebrates have given contrasting results: the CeMyoD protein of Caenorhabduis is expressed in all blastomeres that will give rise uniquely to body wall musculature (Krause et al., 1990), but the nautilus protein of Drosophila has been detected in only about 30% of somatic muscle precursor cells and the muscles derived from them (Michelson et al., 1990). Previous studies of MyoD expression in vertebrate (frog, mouse, and quail) embryos used in situ hybridization with radioactive probes, but these did not give cellular resolution, thus leaving open the possibility that MyoD might be expressed only in a subset of cells, with other myogenic factors acting in the rest (Hopwood et al., 1989; Sassoon et al., 1989; Charles de la Brousse and Emerson, 1990). Now, by using immunocytochemistry, we can see that each myotomal cell expresses XMyoD, at least in neurulae and the posterior parts of tailbud embryos, and can therefore exclude, for example, that the myotomes consist of interspersed expressing and non-expressing cells.

We thank Kevin Johnson for advice on fusion proteins, David Ellar for help with sonication, Charles Jennings for the XMRF4 cDNA, Nigel Garrett for making pGEX2T-XMyoD16/35, and Jonathan Slceman for typing D7F2 and giving much help and advice on antibodies. The manuscript benefitted from the thoughtful comments of Mike Taylor, Jeremy Rashbass, Patrick Lemaire, and Kazuto Kato. We are grateful for the financial support of the Cancer Research Campaign and the Roya] Society.

Bonner
,
W. M.
and
Laskey
,
R. A.
(
1974
).
A film detection method for tritium-labelled proteins and nucleic acids in polyacrylamide gels
.
Eur. J Biochem
.
46
,
83
88
Chartes De La Brousse
,
F.
and
Emerson
,
C. P.
(
1990
).
Localized expression of a myogenic regulatory gene, </m/7, in the somite dermatome of avian embryos
.
Genes Dev
.
4
,
567
581
.
Dingwall
,
C.
,
Dilworth
,
S. M.
,
Black
,
S. J.
,
Kearsey
,
S. E.
,
Cox
,
L. S.
and 1
-Askey
,
R. A.
(
1987
).
Nucleoplasmin cDNA sequence reveals polyglutamic acid tracts and a cluster of sequences homologous to putative nuclear localization signals
.
EMBO J
.
6
,
69
74
.
Gurdon
,
J. B.
(
1977
).
Methods for nuclear transplantation in amphibians
.
Meth. Cell Biot
.
16
.
125
138
.
Hamilton
,
L.
(
1969
).
The formation of somites in Xenopus
.
J. Embryol, exp. Morph
.
22
,
253
264
.
Harlow
,
E.
and
Lane
,
D.
(
1988
).
Antibodies, a Laboratory Manual. Cold Spring Harbor: Cold Spring Harbor Laboratory
.
Harvey
,
R. P.
(
1990
).
The Xenopus MyoD gene: an unlocalizcd maternal mRNA predates lineage-restricted expression in the early embryo
.
Development
108
,
669
680
.
Hemmati Brivanlou
,
A.
and
Harland
,
R. M.
(
1989
).
Expression of an engrcf/ed-related protein is induced in the anterior neural ectoderm of early Xenopus embryos
.
Development
106
,
611
617
.
Hopwood
,
N. D.
and
Gurdon
,
J. B.
(
1990
).
Activation of muscle genes without myogenesis by ectopic expression of MyoD in frog embryo cells
.
Nature
347
,
197
200
.
Hopwood
,
N. D
,,
Pluck
,
A.
and
Gurdon
,
J. B.
(
1989
).
MyoD expression in the forming somites is an early response to mesoderm induction in Xenopus embryos
.
EM BO J
.
8
,
3409
3417
.
Hopwood
,
N. D.
,
Pluck
,
A.
and
Gurdon
,
J. B.
(
1991
).
Xenopus Myf-5 marks early muscle cells and can activate muscle genes cctopically in early embryos
.
Development
111
,
551
560
.
Jackson
,
R. J.
and
Hunt
,
T.
(
1983
).
Preparation and use of nucleasetreated rabbit reticulocyte lysates for the translation of eukaryotic messenger RNA
.
Meth. Enzym
.
96
,
50
74
.
Keller
,
R. E.
(
1976
).
Vita) dye mapping of the gastrula and neurula of Xenopus laevis. II. Prospective areas and morphogenetic movements of the deep layer
.
Dev. Biol
.
51
,
118
37
.
Kintncr
,
C. R.
and
Brockes
,
J. P.
(
1984
).
Monoclonal antibodies identify blasiemal cells derived from dedifferentiating muscle in newt regeneration
.
Nature
308
,
67
69
.
Krause
,
M.
,
Firc
,
A.
,
Harrison
,
S. W.
,
Priess
,
J.
and
Weintraub
,
H.
(
1990
).
CeMyoD accumulation defines the body wall muscle cell fate during C. elegans embryogenesis
.
Ceil
63
,
907
919
.
Krieg
,
P. A.
and
Melton
,
D. A.
(
1984
).
Functional messenger RNAs are produced by SP6 in vitro transcription of cloned cDNAs
.
Nucleic Acids Res
.
12
,
7057
7071
.
Krieg
,
P. A.
and
Melton
,
D. A.
(
1987
).
In vitro RNA synthesis with SP6 RNA polymerase. Meth. Enzymol
.
155
,
397
415
.
Laemmll
,
U. K.
(
1970
).
Cleavage of structural proteins during the assembly of the head of bacteriophage T4
.
Nature
227
,
680
685
.
Michelson
,
A. M.
,
Abrnayr
,
S. M.
,
Bate
,
M.
,
Martinez-Arias
,
A.
and
Maniatis
,
T.
(
1990
).
Expression of a MyoD family member prefigures muscle pattern in Drosophila embryos
.
Cenes Dev
.
4
,
2086
2097
.
Muntz
,
L.
(
1975
).
Myogcnesis in the trunk and leg during development of the tadpole of Xenopus laevis (Daudin 1802)
.
J. Embryol. exp. Morph
.
33
,
757
774
.
Nienwkoop
,
P. D
, and
Faber
,
J.
(
1967
).
Normal Table of Xenopus laevis (Daudin), 2nd cd
.
Amsterdam
:
North-Holland
.
Ott
,
M.-O.
,
Bober
,
E.
,
Lyons
,
G.
,
Arnold
,
H.
and
Buckingham
,
M.
(
1991
).
Early expression of (he myogenic regulatory gene, myf-5, in precursor cells of skeletal muscle in the mouse embryo
.
Development III, 1097-1107
.
Rupp
,
R. A. W.
and
Weintraub
,
H.
(
1991
).
Ubiquitous MyoD transcription at the midblastula transition precedes inductiondependent MyoD expression in presumptive mesoderm of X. laevis
.
Cell
65
,
927
937
.
Sussoon
,
D.
,
Lyons
,
G.
,
Wright
,
W. E-
,
Lin
,
V.
,
Lassar
,
A.
,
Weintraub
,
H.
and
Buckingham
,
M.
(
1989
).
Expression of two myogenic regulatory factors myogenin and MyoDl during mouse embryogenesis
.
Nature
341
.
303
307
.
Scales
,
J. B.
,
Olson
,
E. N.
and
Perry’
,
M.
(
1990
).
Two distinct Xenopus genes with homology to MyoDl are expressed before somite formation in early embryogenesis
.
Mol. Cell. Biol. 10,15161524. ’
Smith
,
D.
and
Johnson
,
K S.
(
1988
).
Single-step purification of polypeptides expressed in Escherichia coli as fusions with glutathione S-transfe rase
.
Gene
67
,
31
40
.
Tapscott
,
S. J.
,
Davis
,
R. L.
,
Thayer
,
M. J.
,
Cheng
,
P.-F.
,
Weintraub
,
H.
and
Lassar
,
A. B.
(
1988
).
MyoDl: a nuclear phosphoprotein requiring a Myc homology region to convert fibroblasts to myoblasts
.
Science
242
,
405
411
.
Taylor
,
M. V.
,
Gurdon
,
J. B.
,
Hopwood
,
N. D.
,
Towers
,
N.
and
Mohun
,
T. J.
(
1991
).
Xenopus embryos contain a somite-specific, MyoD-like protein that binds to a promoter site required for muscle actin expression
.
Genes Dev
.
5
,
1149
1160
.
Towbin
,
H
,,
Staehelin
,
T.
and
Gordon
,
J.
(
1979
).
Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications
.
Proc num. Acad. Sci. U.S.A
.
76
,
4350
4354
.
Weintraub
,
H
,,
Davis
,
R.
,
Tapscott
,
S.
,
Thayer
,
M.
,
Krause
,
M.
,
Bcnezra
,
R.
,
Blackwell
,
T. K.
,
Turner
,
D.
,
Rupp
,
R-
,
Hollenberg
,
S.
,
Zhuang
,
Y.
and
Lassar
,
A.
(
1991
).
The myoD gene family: nodal point during specification of the muscle cell lineage
.
Science
251
.
761
766
.