Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Development 122, 429-437 (1996) Printed in Great Britain © The Company of Biologists Limited 1996 DEV2041 429 Activation of different myogenic pathways: myf-5 is induced by the neural tube and MyoD by the dorsal ectoderm in mouse paraxial mesoderm G. Cossu1,2,*, R. Kelly2, S. Tajbakhsh2, S. Di Donna1, E. Vivarelli1 and M. Buckingham2 1Istituto Pasteur-Cenci Bolognetti, Istituto di Istologia ed Embriologia generale, Università di Roma ‘La Sapienza’, Via A. Scarpa 14, 00161 Rome, Italy 2C.N.R.S. U.R.A. 1947 Departement de Biologie Moleculaire, Institut Pasteur, 25 Rue du Dr Roux. 75724 Paris Cedex 15, France *Author for correspondence at address1 SUMMARY Newly formed somites or unsegmented paraxial mesoderm (UPM) have been cultured either in isolation or with adjacent structures to investigate the influence of these tissues on myogenic differentiation in mammals. The extent of differentiation was easily and accurately quantified by counting the number of β-galactosidase-positive cells, since mesodermal tissues had been isolated from transgenic mice that carry the n-lacZ gene under the transcriptional control of a myosin light chain promoter, restricting expression to striated muscle. The results obtained showed that axial structures are necessary to promote differentiation of paraxial mesoderm, in agreement with previous observations. However, it also appeared that the influence of axial structures could be replaced by dorsolateral tissues, adjacent to the paraxial mesoderm. To elucidate which of these tissues exerts this positive effect, we cultured the paraxial mesoderm with a variety of adjacent structures, either adherent to the mesoderm or recombined in vitro. The results of these experiments indicated that the dorsal ectoderm exerts a positive influence on myogenesis but only if left in physical proximity to it. In contrast, lateral mesoderm delays the positive effect of the ectoderm (and has no effect on its own) suggesting that this tissue produces an inhibitory signal. To investigate whether axial structures and dorsal ectoderm induce myogenesis through common or separate pathways, we dissected the medial half of the unsegmented paraxial mesoderm and cultured it with the adjacent neural tube. We also cultured the lateral half of the unsegmented paraxial mesoderm with adjacent ectoderm. The induction of the myogenic regulatory factors myf-5 and MyoD was monitored by double staining of cultured cells with antibodies against MyoD and β-galactosidase since the tissues were isolated from mouse embryos that carry n-lacZ targeted to the myf-5 gene, so that myf-5-expressing cells could be easily identified by either histochemical or immunocytochemical staining for β-galactosidase. After 1 day in culture myogenic cells from the medial half expressed myf-5 but not MyoD, while myogenic cells from the lateral half expressed MyoD but not myf-5. By the next day in vitro, however, most myogenic cells expressed both gene products. These data suggest that the neural tube activates myogenesis in the medial half of paraxial mesoderm through a myf-5-dependent pathway, while the dorsal ectoderm activates myogenesis through a MyoDdependent pathway. The possible developmental significance of these observations is discussed and a model of myogenic determination in mammals is proposed. INTRODUCTION muscle cells become determined after gastrulation (Krenn et al., 1988). A large body of evidence points to a key role of the neural tube in promoting myogenesis in vivo (Packard and Jacobson, 1976; Vivarelli and Cossu, 1986; Kenny-Mobbs and Thorogood, 1987; Rong et al., 1992; Christ et al., 1992; Buffinger and Stockdale, 1994, 1995; Stern and Hauschka, 1995; Munstenberg and Lassar, 1995), although there is evidence suggesting that the neural tube promotes survival of already committed cells rather than directly inducing myogenic determination (Teillet and Le Douarin, 1983). It is also clear that the effect of the neural tube is limited to that population of myogenic precursors destined to form dorsal (epaxial) The intracellular mechanisms that lead to skeletal myogenesis are now quite well understood, mainly because of extensive work on the MyoD family of basic helix-loop-helix factors (Weintraub et al., 1991) and, to a minor extent, other transcription factors that activate transcription of muscle genes. In contrast, the extracellular signals that activate myogenesis are still unknown. Studies, mainly conducted in the avian system, leave us with a complex picture (Emerson, 1993). Transplantation of defined regions of Hensen’s node from quail into the wing bud or coelomic cavity of chick suggested that skeletal Key words: myogenic induction, MyoD, myf-5 activation, myogenic lineage, mouse, mesoderm, somite 430 G. Cossu and others muscle, since extirpation of the neural tube has no effect on chemical stain such that presomitic mesoderm from these mice limb and body wall (hypaxial) musculature (Rong et al., 1992). can be used as tissue explants in reconstitution experiments This is in keeping with the idea that two separate myogenic with nontransgenic tissues from the same species. fates are followed by precursor cells in the medial and lateral halves of somites (Ordhal and Le Douarin, 1991). The medial population gives rise to epaxial muscle and requires the neural MATERIALS AND METHODS tube, while the lateral population forms hypaxial muscle independently of axial structures. The importance of surrounding Mouse lines tissues in determining the fate of these cells is demonstrated The MLC3F-nlacZ construct contains 2 kb of the mouse fast myosin by the fact that medial/lateral rotation of somites reverses the light chain 3 (MLC3F) promoter together with 1 kb of 3′ flanking DNA, which includes a muscle-specific enhancer, fused to an nlacZbehavior of the corresponding halves (Ordhal and Le Douarin, SV40 poly(A) sequence. In two independent transgenic lines, the 1991). nlacZ reporter gene is strongly expressed in skeletal muscle from day Recently, it was reported that expression of Pax 3, a marker 9 of embryonic development (Kelly et al., 1995). Heterozygous or of dermomyotome from which skeletal muscle is derived, is homozygous transgenic males were crossed with CD1 outbred female dependent upon the neural tube but also upon the dorsal mice. ectoderm. Furthermore, it was shown that close contact was A positive/negative selection vector containing 5.5 kb and 1.1 kb required between the two tissues (Fan and Tessier-Lavigne, of the 5′ and 3′ flanking homology regions was used to introduce the 1994). Also, microdissection experiments in the chick embryo nlacZ and neomycin-resistance genes into the myf-5 locus. The nlacZ revealed that expression of MyoD in the lateral half of newly was introduced in frame into the first exon of the myf-5 gene such that formed somites is inhibited by the lateral plate, since removal the expression of this reporter gene is under the transcriptional and translational control of the endogenous myf-5 locus. Independent ES of this structure led to immediate expression of this gene (Pourquié et al., 1995). Together, these data argue in favor of the existence of multiple signals acting on the lateral half of newly formed somites: at least one DE activating signal, possibly derived from dorsal PM ectoderm or other closely adjacent structures, and an NT LM DA inhibitory signal, probably originating from the A lateral plate, which would be required to prevent terminal differentiation in situ and allow for migration of committed myogenic cells to the limb bud and the body wall. In the case of mammals, because of the relative B inaccessibility of postimplantation mammalian F embryos to experimental manipulation, even less is known about commitment of myogenic cells, and this information only derives from in vitro experiG ments. We reported that somitic cells from 8.5 day C mouse embryos need to be cultured with cells from the adjacent neural tube in order to complete differH entiation while cells from older embryos do not require this influence (Vivarelli and Cossu, 1986). More recently, we showed a community effect during myogenesis of unsegmented paraxial D mesoderm (UPM) and newly formed mouse somites; I again, myogenic precursors acquire the capacity to differentiate as single cells as somites mature (Cossu et al., 1995). Despite this initial information, several questions remain unanswered: (1) what is the nature of the signals required by a mesodermal cell in order E to become committed to myogenesis?; (2) if the Fig. 1. Scheme of a caudal transverse section of a 20- to 25-somite mouse neural tube is required only for medial cells, what is embryo (9.5 dpc) showing the microdissections (broken lines) performed to the source of an equivalent signal for lateral cells? isolate the paraxial mesoderm with or without various adjacent tissues. (A) The and (3) is myogenesis in the medial and the lateral whole paraxial and dorsolateral structures are cultured together without axial halves of somites activated through the same or structures. (B) The whole paraxial mesoderm is cultured together with axial different pathways? We addressed these questions structures but without dorsolateral tissues. (C) Isolated paraxial mesoderm. by taking advantage of recently developed mouse (D) Isolated paraxial mesoderm cultured with its own adjacent dorsal ectoderm. lines that carry the bacterial enzyme β-galactosidase (E) Paraxial and lateral mesoderm cultured without axial structures and dorsal with a nuclear localization signal (nlacZ) under the ectoderm. Isolated paraxial mesoderm was also recombined in vitro with axial control of promoters of muscle structural or regulastructures (F), dorsal aorta (G), dorsal ectoderm (H) and lateral mesoderm (I). tory genes. Cells that have activated the promoter DA, dorsal aorta; DE, dorsal ectoderm; LM, lateral mesoderm; NT, neural tube; PM, paraxial mesoderm. can be recognized by a simple and sensitive histo- AAA AAAA AAA AA AAAA AA AAAAAAAA AA AAA AA AA AAA AA AAA AA AA AA AAAAAAA AAAA AAA AA A AA AAAA AAA AA A A AAAA AAAA AAAA AAA AA AA A A AAAA AA AAAA AAA AA AA A AAAA AAAA AA AA A AA AA AA AA Activation of Myf-5 and MyoD in mammalian somites 431 Fig. 3. The effect of surrounding tissues on Som I 400 myogenic Som I-III differentiation of paraxial 300 mesoderm. The unsegmented 200 paraxial mesoderm (UPM), the last100 formed somite (Som I) or the 0 last three formed - AS DLS LS DE +AS+DLS +LS +DE +DA somites (Som IIII) were dissected and cultured either in isolation (−), or with the adjacent axial structures (AS, neural tube and notochord), dorsolateral structures (DLS, dorsal ectoderm, lateral mesoderm and dorsal aorta), lateral structures (LS, without dorsal ectoderm), dorsal ectoderm (DE). Alternatively, the same paraxial structures were isolated and then re-aggregated with axial structures (+AS), dorsolateral structures (+DSL), lateral structures (+LS), dorsal ectoderm (+DE) and dorsal aorta (+DA). After 3 days in culture the number of β-galactosidase-positive cells was recorded. Each point is the average ± s.e. of at least five separate experiments, each performed in triplicate. 500 no of positive cells/culture UPM Fig. 2. Organ culture of UPM (22-somite embryo) from MLC3F/nlacZ transgenic mice, showing the number of β-galactosidasepositive cells developed after 3 days in vitro. (A) UMP cultured with the adjacent axial structures (neural tube/notochord). (B) UPM cultured in isolation; (C) UPM cultured with the adjacent dorsolateral structures. Bar, 40 µm. Fig. 4. Organ culture of somite I (22-somite embryo) from MLC3F/nlacZ transgenic mice, cultured with the adjacent dorsal ectoderm showing a large number of βgalactosidase-positive cells developed after 3 days in vitro (A). The presence of the ectoderm is confirmed after staining with an anti-keratin monoclonal antibody (C). This antibody only recognizes the ectoderm (arrow) among the dorsal structures of a 22-somite embryo at the level of the unsegmented paraxial mesoderm (D). I, intestine; NT, neural tube; PM, paraxial mesodermPhase contrast shown in B. 432 G. Cossu and others clones containing the mutated myf-5 allele were used to generate heterozygous mice (Tajbakhsh and Buckingham, 1995). Organ and cell cultures Embryos were dated taking day 0.5 postcoitus (p.c.) as the day of the vaginal plug. For most experiments, embryos ranging in age from 20 to 25 somites (9.5 days p.c.) were isolated in PBS. The heart of the MLC3F-nlacZ embryos, where the transgene is also expressed (Kelly et al., 1995), and the cranial myotomes of the myf-5-nlacZ embryos, were stained for β-galactosidase activity (β-gal) and β-gal-positive and -negative embryos were pooled separately. The tissues were then digested with 0.25% pancreatin-0.1% trypsin for 5 minutes at 4°C. After the enzymatic digestion, the various tissues were mechanically separated according to the experimental scheme and then cultured in complete medium (see below). In experiments where re-aggregation of different tissues was required, the tissue fragments were seeded together (under the dissecting microscope) on a layer of 10T1/2 cells, to which they adhered within 10 minutes, and the dishes were then carefully transferred to the incubator. Preliminary experiments had shown that 10T1/2 cells did not alter the extent of myogenic differentiation. In other experiments, dissected halves of UPM (with or without axial structures or dorsal ectoderm) were mechanically dissociated by gently pipetting through the yellow tip of a Gilson pipette to produce a suspension containing single cells and clusters ranging up to approximately 100-200 cells. This suspension was plated in complete medium on collagen-coated dishes. Alternatively, the cell suspensions from the medial and lateral halves of enzymatically dissected UPM were mixed with a similar suspension obtained from axial structures of sibling embryos. All cultures were grown in RPMI medium (GIBCO) supplemented with 10% fetal calf serum (Flow), 300 mM β-mercaptoethanol and 50 µg/ml gentamycin. At the times indicated, cultures were fixed and stained for β-galactosidase activity and/or incubated with different antibodies. Immunocytochemistry Immunocytochemistry on tissue sections and cultured cells was carried out as described (Cusella-De Angelis et al., 1994; Tajbakhsh et al., 1994) using the following antibodies: (1) anti-β-galactosidase monoclonal antibody (from Sigma); (2) FD5 anti-myogenin monoclonal antibody (Cusella De Angelis et al., 1992), donated by W. Wright. (3) MF20 anti-sarcomeric myosin monoclonal antibody (Bader et al., 1982), donated by D. Fischman. (4) a rabbit polyclonal anti-MyoD antibody (Koishi et al. 1995), donated by J. Harris. (5) a rabbit polyclonal anti-β-galactosidase antibody (Tajbakhsh et al., 1994), donated by O. Puijalon. (6) an anti-keratin rat monoclonal antibody, donated by D. Paulin. (7) a rabbit polyclonal anti-sarcomeric myosin antibody (Cusella De Angelis et al., 1994), produced in our laboratory. (8) an anti-PECAM rat monoclonal antibody, which recognizes dorsal aorta, donated by E. Dejana. Polyclonal antibodies and the anti-β-galactosidase monoclonal were diluted 1:100 before use; other monoclonal antibodies were used as undiluted supernatant. RESULTS Myogenic differentiation of mesoderm depends on dorsolateral structures as well as axial structures In order to investigate how axial structures and, possibly, a corresponding dorsolateral structure may influence terminal differentiation in newly formed somites as well as in unseg- mented paraxial mesoderm (UPM), we performed three types of experiments, illustrated in Fig. 1. In the first type of experiment, single somites or streaks of newly formed somites, or UPM (from MLC3F-nlacZ, 22- to 24-somite embryos) were isolated from axial structures (neural tube and notochord) but not from dorsolateral structures of the body (dorsal ectoderm, lateral mesoderm and dorsal aorta); also the same structures from the paraxial mesoderm were isolated from dorsolateral but not from axial structures (A and B in Fig. 1). Finally, UPM or somites were isolated from all surrounding tissues (C in Fig. 1). The explants were cultured for 3 days and the number of β-galactosidase-positive nuclei was then recorded after staining for the enzymatic activity. Preliminary experiments (not shown) had demonstrated that β-gal-positive nuclei invariably localized within myosin-positive cells and that cells expressing neurofilaments or glial fibrillar acidic protein were present only in cultures that included the neural tube (immunocytochemical specific markers for dorsolateral mesoderm are not available and therefore this control was not possible). Fig. 2 shows an example of these cultures: UPM cultured with adjacent axial structures gave rise to a large number of βgal-positive (i.e. differentiated) cells; in contrast, very few βgal-positive cells were generated from UPM cultured in isolation. However, a good extent of differentiation occurred when UPM was cultured with adjacent lateral structures, in the absence of the neural tube. A quantitative analysis, summarizing more than 20 experiments is reported in Fig. 3. It is clear from the data that the axial structures are required for optimal differentiation of UPM or newly formed somite(s), but that differentiation can also be induced by dorsolateral structures. Occasionally, we noticed more differentiation in isolated somites or UPM, possibly due to incomplete removal of surrounding tissues (see below). The source of the dorsolateral signals In the second series of experiments, we attempted to determine which among the various dorsolateral structures can induce myogenesis in explants of paraxial mesoderm. For this purpose, somites or UPM were isolated from all surrounding tissues and then reaggregated in vitro with either axial or lateral structures isolated from non-transgenic siblings. Specifically, the UPM (or Som I or Som I-III) were recombined with either the dorsal aorta, recognized by an anti-PECAM antibody (G in Fig. 1), or with dorsal ectoderm, recognized by a an anti-keratin antibody (H in Fig. 1), or with adjacent lateral mesoderm, including mesonephron, aorta and coelomic epithelia but not dorsal ectoderm (I in Fig. 1). Reconstitution experiments with axial structures were also carried out as positive controls (F in Fig. 1). However, under the conditions tested, only axial structures reproducibly induced a high extent of differentiation (Fig. 3). In the case of dorsal ectoderm, the experiments were technically difficult, as often the pancreatin-digested ectoderm would roll on itself, making it difficult to recombine it with the mesoderm. Because of the reported observations that dorsal ectoderm will induce dorsal markers only if left in contact with the mesoderm (Kuratani et al., 1994; Fan and TessierLavigne, 1994), we performed a third series of experiments, by culturing UPM or newly formed somites either in isolation or with their own dorsal ectoderm or with their own lateral mesoderm (C, D and E in Fig. 1). Fig. 4A and C shows a very Activation of Myf-5 and MyoD in mammalian somites 300 n° of differentiated cells / culture 250 200 150 100 50 0 d1 d2 d3 d4 d5 days in vitro Fig. 5. Time course of in vitro appearance of β-galactosidasepositive cells from the unsegmented paraxial mesoderm (22-somite embryo) from MLC3F/nlacZ transgenic mice, cultured in isolation (m–m), with adjacent dorsolateral structures (r–r), with dorsal ectoderm only (d–d) or with lateral mesoderm (with dorsal aorta but without dorsal ectoderm: j–j). high extent of differentiation in somite I cultured with its own dorsal ectoderm, revealed by an anti-keratin antibody, while somites cultured in isolation or only with lateral mesoderm showed little or no differentiation (data not shown). Morphological analysis of the unsegmented mesoderm and newly formed somites from mouse embryos at a 22-somite stage, revealed close contact between paraxial mesoderm and the dorsal ectoderm, revealed by the anti-keratin antibody (Fig. 4B,D). In order to demonstrate the existence of both positive and negative signals emanating from dorsolateral structures, we isolated the UPM of 22-somite embryos with or without dorsal ectoderm and with or without lateral mesoderm. The explants were grown in culture and the number of differentiated cells under each condition was monitored daily for 5 days. All cultures were then stained with anti-keratin antibodies to confirm the presence or the absence of ectoderm. In three separate experiments (Fig. 5 shows a representative experiment), no differentiation occurred in cultures of UPM alone or of UPM with lateral mesoderm only. In contrast, rapid differentiation occurred in cultures of UPM also containing dorsal ectoderm. Cultures containing both dorsal ectoderm and lateral mesoderm showed a delayed appearance of differentiated cells which, after longer periods in culture, were more numerous than in the corresponding cultures with dorsal ectoderm only. The simplest interpretation of these data is the release of an inhibitory signal from the lateral mesoderm which delayed differentiation of myogenic cells. If this signal is a growth factor, then the cells would be prevented from differentiating by being forced to divide, thus explaining the final increase in total number. Medial signals activate myf5 while dorsolateral signals activate MyoD We next investigated whether medial and dorsolateral signaling tissues, i.e. neural tube/notochord complex and the dorsal ectoderm, activate the myogenic program in the 433 precursor populations through the same or different members of the MyoD family of basic-HLH transcription factors. Evidence already exists showing earlier dorsal expression of myf-5 and later ventral expression of MyoD in newly formed myotomes (Ott et al., 1991; Smith et al., 1994; Goldhamer et al., 1995). However, to address this point in detail, it is necessary to identify the expression of both genes at a single cell level. We took advantage of mice that carry nlacZ targeted to the myf-5 gene, so that all the cells that express the gene can be identified by staining for β-galactosidase activity or with antibodies against β-galactosidase. In control double-labeling experiments, anti-myf-5 polyclonal (Smith et al., 1994) and anti-β-galactosidase monoclonal antibodies labeled the same nuclei in both cryostat sections and cultured cells. Most of the experiments described here were carried out using a polyclonal antibody against MyoD and a monoclonal anti-β-galactosidase antibody to localize the presence of one or both proteins inside the nucleus of myogenic cells both in vivo and in culture. Fig. 6 shows a transverse section (at the level of the forelimb) of a 10.5 dpc heterozygous myf5/nLacZ embryo, double stained with antibodies against MyoD and β-galactosidase. It is clear from the figure that there is a large overlapping area of expression of the two genes, localized to the same nucleus in about half of the cells (double arrowheads in Fig. 6C,D); it also appears that there is a ventrolateral area of the somite muscle where MyoD but not myf-5 is expressed at detectable levels (arrows in Fig. 6A,B), while myf-5 appears to be predominantly expressed in the dorsomedial area. Because MyoD has a later onset of expression in vivo and can be clearly detected only when the myotome is formed (Buckingham, 1992), we performed experiments aimed to clarify the onset of expression of the two genes in the medial and lateral halves of the paraxial mesoderm, by physically separating the medial half of UPM with adjacent inducing tissues (neural tube, notochord) from the lateral half (with its own ectoderm), and growing the two separate halves as high density cell cultures. The tissues were mechanically dissociated to obtain a suspension of small clusters of cells so that inductive interactions among different cells types would not be disrupted, while allowing double immunofluorescence to be performed at different times on the resulting multilayer culture. Fig. 7 shows that after 1 day in vitro, several cells derived from the medial half (roughly from 50 to 100 cells per UPM) would express myf-5 but not MyoD, with rare cells expressing both myf-5 and MyoD. In contrast the cultures derived from the lateral half contained a lower number of positive cells and these cells expressed MyoD but not myf-5. By the next day in vitro, cultures of both medial and lateral halves contained several hundreds of cells that coexpressed both proteins although with variable relative intensity, making quantification very difficult (Fig. 8). A minority of cells expressed predominantly either myf-5 or MyoD, but cells expressing only one of the two gene products at a detectable level were rare, at least in vitro. We next investigated whether myogenic precursor cells in the dorsal portion of the UPM are already determined in their capacity for activating one or the other myogenic program (through myf-5 or MyoD) or whether they are instructed by different signals derived from axial structures and dorsal ectoderm. For this purpose, the medial and the lateral halves of the UPM were dissected free of any adjacent tissues and 434 G. Cossu and others mixed with a similar cellular suspension derived from axial structures. After 1 day in culture, several cells derived from both the medial and the lateral halves of the UPM expressed myf-5 but not MyoD, thus showing that axial structures can activate myf-5 in myogenic precursor cells independently from their previous location in the paraxial mesoderm (data not shown). The complementary experiment could not be performed, since dorsal ectoderm will not induce myogenesis once separated from the underlying mesoderm. DISCUSSION The data reported here show that both dorsal ectoderm and axial structures will induce cells in the paraxial mesoderm to undergo myogenesis. Furthermore, we show that signal(s) emanating from the axial structures, which induce myogenesis in the medial half of paraxial mesoderm, lead to activation of myf-5, while signals derived from the dorsal ectoderm lead to activation of MyoD laterally; in this case, however, differentiation is probably repressed by signals originating from lateral mesoderm. This information is summarized in a model of myogenic induction in mammals, shown in Fig. 9. Dorsal ectoderm induces myogenesis in the paraxial mesoderm In addition to a well-established signal emanating from axial structures (neural tube and notochord), a dorsolateral signal exists that induces myogenesis in the lateral half of the paraxial mesoderm. This signal is derived from the dorsal ectoderm and only acts if a close contact between the two tissues is preserved. Our data showing that newly formed somites or UPM will differentiate if dorsal ectoderm is not removed, are in good agreement with the observation that expression of Pax-3, a marker of dorsal somitic mesoderm, temporally and spatially preceding expression of myogenic markers, takes place in explants of somites in close contact with dorsal ectoderm (Fan and Tessier-Lavigne, 1994). The requirement for a close contact between the basal side of the ectoderm and the paraxial mesoderm would explain why, in different assays where ectoderm is cultured in proximity or across a filter (Stern and Hauschka, 1995; Buffinger and Stockdale, 1995), differentiation is not induced. In our own experience, ectoderm did not induce myogenesis in reconstitution experiments where it did not remain adherent to the mesoderm. In a previous study, Kenny-Mobbs and Thorogood (1987) reported differentiation of isolated brachial somites when reconstituted with adjacent epithelium. Here again, the variability reported might have been due to the more or less intimate reconstitution of the two tissues and to the side of the epithelium facing the mesoderm. It could be argued that neural crest cells leave the neural tube and migrate between the dorsal ectoderm and the somites just at the onset of somitogenesis and may therefore be responsible for the inductive effect (Bonner-Fraser, 1993). However, if the neural tube is removed at the level of the unsegmented mesoderm as in the experiments described here, then few if any neural crest cells should have already left the neuroepithelium (Serbedzija et al., 1990), making it unlikely that these cells contribute to the inductive effect. Differentiation of myogenic cells in the lateral half of the paraxial mesoderm is delayed by the lateral mesoderm Dorsal ectoderm is adjacent to the dorsolateral portion of the segmental plate and newly formed somites. This region Fig. 6. Double immunofluorescence of a transverse section at the forelimb bud level of a 45-somite (10.5 dpc) myf-5/nlacZ heterozygous embryo, stained with an anti-βgalactosidase monoclonal antibody (A,C) and with anti-MyoD polyclonal antibody (B,D). Staining for β-galactosidase (identifying myf-5-expressing cells) is confined to the dorsal portion of the myotome (D), whereas staining for MyoD extends more ventrally (shown by an arrow in A and B), although overlapping with the majority of the myf-5positive area. Higher magnification of a dorsal area, shown in C and D, reveals nuclei expressing myf-5 but not MyoD (arrowhead), nuclei expressing MyoD but not myf-5 (arrow) and nuclei expressing both proteins (double arrowheads). Bar, 40 µm. Activation of Myf-5 and MyoD in mammalian somites 435 Fig. 7. Double immunofluorescence of a 1-day-old culture derived from the medial half of the UPM with the adjacent neural tube (A,B) and from the lateral half of the UPM with the adjacent dorsolateral structures (C,D) isolated from a myf-5/nlslacZ 22-somite heterozygous embryo, stained with an anti-β-galactosidase monoclonal antibody (A,C) and with an anti-MyoD polyclonal antibody (B,D). Most cells in culture from the medial half express β-galactosidase but not MyoD (shown by arrowhead); the opposite is true for cultures from the lateral half (arrows). Rare double-labeled cells are shown by double arrowheads. Fluorescence in the bottom-left corner in C and D is due to a clump of undissociated tissue and is not specific (i.e. nuclear). Bar, 10 µm. harbors the myogenic precursors of the hypaxial muscle, which will leave the dermomyotome and migrate to the limb and the body wall. These cells do not express any member of the MyoD family either in the somite or during migration (Tajbakhsh and Buckingham, 1994), and yet they are committed irreversibly to myogenesis as shown by classic transplantation experiments (Chevallier et al., 1977; Jacob et al., 1977). It is therefore possible that differentiation of these cells is repressed by signals derived from adjacent tissues. In the explant cultures described here, differentiation of lateral myogenic cells is delayed by the lateral mesoderm which may produce growth factors or factors stimulating cell movement. Fig. 8. Double immunofluorescence of a 2-day-old culture derived from the medial half of the UPM with the adjacent neural tube isolated from a myf-5/nlslacZ 22-somite heterozygous embryo, stained with an anti-β-galactosidase monoclonal antibody (A) and with an anti-MyoD polyclonal antibody (B). Most cells express both antigens although to a variable extent: One cell expressing high levels of β-galactosidase and low levels of MyoD is shown by an arrowhead; cells expressing high levels of MyoD and low levels of β-galactosidase are shown by arrows, double-labeled cells are shown by double arrowheads. Bar, 15 µm. 436 G. Cossu and others D ec orsa tod l erm Neural tube AAAAAAAAAAAA AAAAA AAAAAAA AAAAA AAAAAAAAAAAA AAAAAAA AAAAAAA AAAAAAA AAAAAAA Myf-5 MyoD G.F Celoma Dorsal aorta NC Fig. 9. A model showing the dorsomedial and dorsolateral influences on the paraxial mesoderm, resulting in activation of myogenesis through a myf-5 and a MyoD-dependent pathway, respectively. The possible inhibitory effect of lateral structures is also shown. FGF-5, for example, is expressed is somatic lateral mesoderm at the time that myotome forms (Haub and Goldfarb, 1991) and therefore may stimulate proliferation of lateral myogenic cells. These data are in good agreement with similar conclusions obtained by mechanical separation of the lateral plate from the paraxial mesoderm in the chick embryo (Pourquié et al., 1995). In this case, expression of MyoD is activated in the lateral half of somites where it is not normally observed in control embryos. Myo D is activated before myf-5 in the lateral half of the segmental plate, whereas myf-5 is activated first in the medial half In dissociated high-density cultures obtained from the lateral half of the segmental plate, a certain number of cells, which probably receive signals from overlying ectodermal cells, begin to express one member of the family by the first day in culture: this is invariably MyoD, strongly suggesting that dorsal ectoderm activates myogenesis through a MyoDdependent pathway. By the second day in culture, the great majority of positive cells express both MyoD and myf-5, although with variable intensity. The very large increase in the total number of positive cells makes it impossible to determine whether MyoD activates myf-5 in cultures from the lateral half. This does not occur in myogenically converted muscle cell lines (Montarras et al., 1991). Either this activation takes place in vivo or, alternatively, the majority of newly differentiating cells now express both genes simultaneously. At this time, myogenin and myosin also begin to be expressed in the myogenic population (unpublished observations). Myf-5 is the only member of the MyoD family initially expressed in the large majority of myogenic cells cultured from the medial half of the mouse segmental plate. Here again, by the second day in vitro, most cells express both genes, and this is probably due to myf-5 activation of MyoD given that this occurs under a variety of in vitro conditions (Montarras et al., 1991). Culturing cells from the lateral half of the UPM with cells derived from axial structures resulted in activation of myf-5, thus suggesting that cells in the segmental plate are not already committed to a medial (via myf-5) or to a lateral (via MyoD) fate. This is in agreement with the notion that replacing the medial half of a newly formed somite with the corresponding lateral half does not perturb development (Ordhal and Le Douarin, 1991) indicating persistent plasticity of myogenic precursors at this stage of development. Myf-5 is the factor expressed in MyoD-negative, primordial myoblasts (Cusella De Angelis et al., 1992), which appear to be insensitive to inhibitory signals (growth factors) and initiate myotome formation. However, since the myotome is initially composed of a relatively small number of post-mitotic myocytes, other myogenic precursor cells must continue to proliferate before differentiating subsequently to form the remaining epaxial musculature. In this context, it has been proposed that the dorsal portion of the neural tube inhibits terminal myogenic differentiation through production of growth factors (Buffinger and Stockdale, 1995). Two signals and two pathways for myogenic induction? MyoD and myf-5 are the two members of the family of bHLH myogenic factors present in dividing myoblasts in culture (Buckingham, 1994) and mutations in both genes result in a dramatic impairment of the myoblast cell population in vivo (Rudnicki et al., 1993). During mouse embryogenesis, myf-5 is expressed first, in the medial lip of the dermomyotome before overt muscle differentiation (Ott et al., 1991), in response, as shown here, to signals from axial structures. We now show that MyoD is activated in cells cultured from the lateral part of the paraxial mesoderm in response to signals from the dorsal ectoderm, thus suggesting that in vivo the activation of myogenesis in the lateral part of paraxial mesoderm occurs through a MyoD-dependent pathway, even if its expression is repressed by the lateral mesoderm. Subsequently, many muscle cells coexpress MyoD and myf-5 in vitro and in vivo (this manuscript and also Ott et al., 1991). Independent activation of myf-5 and MyoD accounts for the formation of most skeletal muscle in mice in which only the myf-5 or MyoD gene is mutated (Rudnicki et al., 1992; Braun et al., 1992). This suggests that, although initially precursor cells responding to environmental signals will activate one or the other gene, subsequently the majority of myogenic cells can use either and normally use both factors. It is tempting to speculate that myf-5 and MyoD may have duplicated from a common ancestor with diversification of dorsal and ventral muscles, possibly with the appearance of primitive vertebrates. Whether the myf-5 and MyoD proteins have intrinsically different properties remains to be established and, indeed, the fact that in birds the two genes seem to have exchanged roles (Pownall and Emerson, 1992) in terms of which is activated first might argue against this. However, it is clear that during the diversification of this gene family, the two genes acquired different regulatory sequences that determine their response to signals from axial and dorsolateral structures. The molecules that regulate the expression of myf-5 and MyoD remain to be determined. Activation of Myf-5 and MyoD in mammalian somites This work was supported by grants from Telethon n. A.13, Cenci Bolognetti and from MURST to G. C., and grants from the AFM, ARC and INSERM to M. B. and by an EC grant from the Human Capital and Mobility Program to M. B. and G. C.; R. K. was supported initially by a Wellcome Traveling fellowship and then by an EC fellowship under the HCMP grant. We thank Drs E. Dejana, D. A. Fischman, S. Koniecszny, D. Paulin, O Pujelon and W. Wright for providing the antibodies used in this study. REFERENCES Bader, D., Masaki, T., and Fischman, D.A. (1982). Immunochemical analysis of myosin heavy chain during avian myogenesis in vivo and in vitro. J. Cell Biol. 95, 763-770. Bonner-Fraser, M. (1993) Mechanisms of neural crest cell migration. BioEssays 15, 221-230. Braun, T., Rudnicki, M. A., Arnold, H. H., and Jaenisch, R. (1992). Targeted inactivation of the muscle regulatory gene Myf-5 results in abnormal rib development and perinatal death. Cell 71, 369-82. Buckingham, M. (1992). Making muscle in mammals. Trends Genet. 8, 144148. Buckingham, M. (1994). Muscle: the regulation of myogenesis. Curr. Op. Genet. Develop. 4, 745-751. Buffinger, N. and Stockdale, F. E. (1994) Myogenic specification in somites: induction by axial structures. Development 120, 1433-1452. Buffinger, N. and Stockdale, F. E. (1995) Myogenic specification of somites is mediated by diffusible factors. Dev. Biol. 169, 96-108. Chevallier,A., Kieny, M., Mauger, M. and Sengel, P. (1977). Limb-somite relationship: origin of limb musculature. J. Embryol. Exp. Morph. 41, 245258. Christ, B., Brand-Saberi, B., Grim, M. and Wilting, J. (1992) Local signaling in dermomyotomal cell type specification. Anat. Embryol. 186, 505-510. Cossu, G., Kelly, R., Di Donna, S., Vivarelli, E. and Buckingham, M. (1995). Myoblast differentiation during mammalian somitogenesis is dependent upon a community effect. Proc. Natl. Acad. Sci. USA 92, 22542258. Cusella-De Angelis, M.G., Lyons, G., Sonnino, C., De Angelis, L., Vivarelli, E., Farmer, K., Wright, W.E., Molinaro, M., Bouche, M., Buckingham, M. and Cossu, G. (1992). MyoD, Myogenin independent differentiation of primordial myoblasts in mouse somites J. Cell Biol. 116, 1243-1255. Cusella-De Angelis, M.G., Molinari, S., Ledonne, A., Coletta, M., Vivarelli, E., Bouche’, M., Molinaro, M., Ferrari, S. and Cossu, G. (1994). Differential response of embryonic and fetal myoblasts to TGFβ : a possible regulatory mechanism of skeletal muscle histogenesis. Development 120, 925-933. Emerson, C. P. J. (1993). Embryonic signals for skeletal myogenesis: arriving at the beginning. Curr. Opin. Cell Biol. 5, 1057-1064. Fan, C. and Tessier-Lavigne, M. (1994). Patterning of mammalian somites by surface ectoderm and notochord: evidence for sclerotome induction by a hedgehog homologue. Cell 79, 1175-1186. Goldhamer, D. J., Brunk, B. P., Faerman, A., King, A., Shani, M. and Emerson, C. P.Jr. (1995) Embryonic activation of the myoD gene is regulated by a highly conserved distal control element. Development 121, 637-649. Haub, O. and Goldfarb, M. (1991) Expression of the fibroblast growth factor5 gene in the mouse embryo. Development 112, 397-406. Jacob, M., Christ, B. and Jacob, H.J. (1977) The migration of myogenic cells from somites into the leg region of avian embryo. Anat. Embryol. 157, 291-309. Kelly, R., Alonso, S., Tajbakhsh, S., Cossu, G. and Buckingham, M. (1995) Myosin light chain 3F regulatory sequences confer regionalised cardiac and skeletal muscle expression in transgenic mice. J. Cell Biol. 129, 383-396. Kenny-Mobbs, T. and Thorogood, P. (1987) Autonomy of differentiation in avian brachial somites and the influences of adjacent tissues. Development 100, 449-462. 437 Koishi, K., Zhang, M., McLennan, I. S. and Harris, A. J. (1995) MyoD protein accumulates in satellite cells and is neurally regulated in regenerating myotubes and skeletal muscle fibers. Develop. Dynam. 202, 244-254. Krenn, V., Gorka, P., Wachtler, F., Christ, B. and Jacob, H. J. (1988). On the origin of cells determined to form skeletal muscle in avian embryos. Anat. Embyol. 179, 49-54. Kuratani, S., Martin, J. F., Wawersik, S., Lilly, B., Eichele, G. and Olson, E. N. (1994). The expression pattern of the chick homeobox gene gMHox suggests a role in patterning of the limbs and face and in the compartmentalization of somites. Dev. Biol. 161, 357-369. Montarras, D., Chely, J., Bober, E., Arnold, H., Ott, M. O., Gros, F. and Pinset, C. (1991). Developmental patterns in the expression of Myf-5, MyoD, myogenin and MRF4 during myogenesis. New Biol. 3, 592-600. Münsterberg, A. E. and Lassar, A. B. (1995). Combinatorial signals from the neural tube, floor plate and notochord induce myogenic bHLH gene expression in the somite. Development 121, 651-660. Ordahl, C. P. and Le Douarin, N. (1991). Two myogenic lineages within the developing somite. Development 114, 339-353. Ott, M. O., Bober, E., Lyons, G., Arnold, H. and Buckingham, M. (1991). Early expression of the myogenic regulatory gene, myf-5 in precursor cells of skeletal muscle in the mouse. Development 111, 1097-1107. Packard, D. S. Jr. and Jacobson, A. G. (1976). The influence of axial structures on chick somite formation. Dev. Biol. 53, 36-48. Pownall, M. E. and Emerson, C. P.Jr. (1992). Sequential activation of myogenic regulatory genes during somite morphogenesis in quail embryos. Dev. Biol. 151, 67-79. Pourquié, O., Coltey, M., Breant, C. and Le Douarin, N. (1995). Control of somite patterning by signals from the neural tube. Proc. Natl Acad. Sci. USA 92, 3219-3223. Rong, P. M. Teillet, M. A., Ziller, C. and Le Douarin, N. M. (1992). The neural tube/notochord complex is necessary for vertebral but not limb and body wall striated muscle differentiation. Development 114, 339-353. Rudnicki, M. A., Braun, T., Hinuma, S. and Jaenisch, R. (1992). Inactivation of myoD in mice leads to up-regulation of the myogenic HLH gene myf-5 and results in apparently normal muscle development. Cell 71, 383-390. Rudnicki, M. A., Scnegelsberg, P. N. J., Stead, R. H., Braun, T., Arnold, H. H. and Jeanisch, R. (1993) MyoD or Myf-5 is required for the formation of skeletal muscle. Cell 75, 1351-1359. Serbedzija, G., Fraser, S. E., and Bonner-Fraser, M. (1990). Pathways of trunk neural crest migration in the mouse embryo revealed by vital dye analysis. Development 108, 605-612. Smith, T. H., Kachinsky, A. M. and Miller, J. B.(1994). Somite subdomains, muscle cell origins and the four regulatory factor proteins. J. Cell Biol. 127, 95-100. Stern, H. M. and Hauschka, S. D. (1995) Neural tube and notochord promote in vitro myogenesis in single somite explants. Dev. Biol. 167, 87-103. Tajbakhsh, S. and Buckingham, M. E. (1994). Mouse limb muscle is determined in the absence of the earliest myogenic factor myf-5. Proc. Natl Acad. Sci. USA 91, 747-751. Tajbakhsh, S., Vivarelli, G., Cusella-De Angelis, G., Rocancourt, D., Buckingham, M. and Cossu, G. (1994). A population of myogenic cells derived from the mouse neural tube. Neuron 13, 813-821. Tajbakhsh, S. and Buckingham, M. E. (1995). Lineage restriction of the myogenic conversion factor myf-5 in the brain. Development 121, 4085-4094. Teillet, M. A. and Le Douarin, N. M. (1983) Consequences of neural tube and notochord excision on the development of the peripheral nervous system in the chick embryo. Dev. Biol. 98, 191-211. Vivarelli, E. and Cossu, G. (1986) Neural control of early myogenic differentiation in cultures of mouse somites. Dev. Biol. 117, 319-325. Weintraub, H., Davis, R., Tapscott, S., Thayer, M., Krause, M., Benezra, 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. (Accepted 2 November 1995)