Development 121, 651-660 (1995) 651 Printed in Great Britain © The Company of Biologists Limited 1995 Combinatorial signals from the neural tube, floor plate and notochord induce myogenic bHLH gene expression in the somite Andrea E. Münsterberg and Andrew B. Lassar Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, 240 Longwood Avenue, Boston, MA 02115, USA SUMMARY The neural tube, floor plate and notochord are axial tissues plate/notochord, while in more rostral somites (stages IV- in the vertebrate embryo which have been demonstrated to IX) the neural tube without the floor plate/notochord is suf- play a role in somite morphogenesis. Using in vitro co- ficient. By recombining somites and neural tubes from culture of tissue explants, we have monitored inductive different axial levels of the embryo, we have found that a interactions of these axial tissues with the adjacent somitic second signal is necessary to promote competence of the mesoderm in chick embryos. We have found that signals somite to respond to inducing signals from the neural tube. from the neural tube and floor plate/notochord are Thus, we propose that at least two signals from axial tissues necessary for expression of the myogenic bHLH regulators work in combination to induce myogenic bHLH gene MyoD, Myf5 and myogenin in the somite. Eventually expression; one signal derives from the floor somitic expression of the myogenic bHLH genes is main- plate/notochord and the other signal derives from regions tained in the absence of the axial tissues. In organ culture, of the neural tube other than the floor plate. at early developmental stages (HH 11−), induction of myo- genesis in the three most recently formed somites can be Key words: neural tube, floor plate, notochord, myogenic bHLH, mediated by the neural tube together with the floor somitogenesis INTRODUCTION opmental stages somitic cells are committed to particular cell fates and differentiate irrespective of their location within the In all vertebrate embryos, the paraxial mesoderm gives rise to embryo (Aoyama and Asamoto, 1988; Christ et al., 1992; transient structures called somites. The somites are repeated Ordahl and LeDouarin, 1992). Efforts to identify the source of units which are progressively generated during development the cues that dictate somitic cell fate have focused attention on along the anteroposterior axis on either side of the neural tube. the axial tissues of the vertebrate embryo: the developing Somite morphogenesis is outlined in Fig. 1. Initially, somites neural tube and the underlying notochord. It was noted that form as epithelial spheres from the unsegmented paraxial somites that form in the absence of a neural tube do not mesoderm. [Throughout this paper, we refer to the most progress beyond a hollow epithelial sphere morphology remi- recently formed somite as somite I, the next youngest somite niscent of immature somites, and therefore seem to be arrested as somite II, etc., (as proposed in Ordahl, 1993).] As develop- in their maturation (Packard and Jacobson, 1976). Recently, it ment proceeds, a columnar epithelial sheet develops in the was demonstrated that excision of the neural tube and dorsal somite (dermomyotome), while cells in the ventral notochord from early chick embryos results in the striking somite loose their epithelial morphology and give rise to a absence of axial skeletal muscle (Christ et al., 1992; Rong et loosely connected mesenchyme (sclerotome). The sclerotome al., 1992). In vitro experiments have provided evidence that the gives rise to precursor cells of the ribs, vertebrae and interver- neural tube directly promotes skeletal muscle differentiation tebral discs (discussed in Christ and Wilting, 1992). Subse- (Avery et al., 1956; Vivarelli and Cossu, 1986; Kenny-Mobbs quently, cells of the dermomyotome that lie adjacent to the and Thorogood, 1987; Rong et al., 1992; Buffinger and neural tube form an intermediate sheet of cells in the middle Stockdale, 1994; Stern and Hauschka, 1994). In these assays, of the somite (myotome) that gives rise to axial skeletal muscle developmentally immature somites were explanted and (i.e. vertebral and back muscle). Cells at the lateral edge of the cultured under various conditions; muscle differentiation dermomyotome migrate to give rise to the body wall/limb mus- markers (i.e. myoblast fusion or myosin heavy chain synthesis) culature (reviewed in Wachtler and Christ, 1992). Cells from were not activated unless the somites were co-cultivated with the dorsalmost epithelial sheet (dermatome) give rise to cells from the adjacent neural tube. In contrast, developmen- dermis. tally more mature somites could differentiate into muscle when At an early developmental stage, cell fate within the somite explanted and cultured in the absence of neural tube. is plastic and responsive to extrinsic cues, while at later devel- The initiation of the myogenic program is thought to be con- 652 A. E. Münsterberg and A. B. Lassar trolled by a set of basic-Helix-Loop-Helix (bHLH) transcrip- g/l) and pinned down with their ventral side facing up using insect tion factors, which include MyoD, myogenin, Myf-5 and pins. The endodermal epithelium was removed from the region to be MRF-4 (reviewed in Buckingham, 1992; Emerson, 1993; dissected, and the following tissues were explanted: somites alone; Sassoon, 1993; Weintraub, 1993; Lassar and Münsterberg, somites with the adjacent neural tube, floor plate and notochord; 1994; Olson and Klein, 1994). Gene knockout by homologous somites with the adjacent neural tube excluding floor plate; or somites recombination in mice has demonstrated that either MyoD or with floor plate and notochord (see figure legends for details and stages of embryos). Following dissection with micro feather scalpels Myf-5 is necessary for the determination or survival of (Oasis, CA) the tissues were transferred into Tyrodes buffer contain- myogenic precursor cells and that myogenin is necessary to ing Dispase (Boehringer Mannheim, 1 mg/ml) and incubated for 5 execute the differentiation program (Hasty et al., 1993; minutes at room temperature to facilitate the removal of the ectoder- Nabeshima et al., 1993; Rudnicki et al., 1993). Detailed in situ mal epithelia. After washing in medium, the tissues were cultured on hybridization studies have shown that the myogenic bHLH gelatin-coated 24-well dishes. Tissues that had been separated during genes are expressed early during embryogenesis (Hopwood et the dissection were aggregated on the tissue culture dish. The cultures al., 1989a; Sassoon et al., 1989; Bober et al., 1991; Charles de were maintained in α-MEM (Gibco) supplemented with 15% heat- la Brousse and Emerson, 1990; Hinterberger et al., 1991; Ott inactivated horse serum (Gibco), 2.5% chick embryo extract and 1% et al., 1991; Pownall and Emerson, 1992) and, thus, they penicillin/Streptomycin solution, in vitro for 5 days in 5% CO2 at provide molecular markers for myogenic precursor cells in 37°C. Chick embryo extract was prepared as follows: 9- to 11-day chick vivo. In different species, the order in which the myogenic embryos were rinsed in PBS and then forced through a 60 ml syringe. bHLH factors are expressed differs; however, in all species An equal volume of α-MEM was added and incubated for 1 hour on examined, at least one member of this gene family is expressed ice. After removing the debris by centrifugation at 25,000 in the early somite prior to its morphological segregation into revs/minute, the supernatant was sterile filtered and kept frozen at dermatome, myotome and sclerotome (discussed in Bucking- −20°C until used. ham, 1992). Interestingly, in both mouse and quail, myogenic RNA isolation and reverse transcription bHLH transcripts first accumulate in the medial portion of the somite adjacent to the neural tube (Ott et al., 1991; Pownall At the end of the culture period, medium was aspirated and RNA was prepared from the explants using the method described by Chom- and Emerson, 1992, Barth and Ivarie, 1994). czynski and Sacchi (1987). Tissues were lysed in 100 µl lysis buffer In this report, we focus on the inductive interactions that are (25 g guanidinium thiocyanate, 1.76 ml 0.75 M NaCitrate, pH 7, 2.64 necessary to activate high level expression of the myogenic ml 10% Sarkosyl, 38 µl β-mercaptoethanol, 29.3 ml H2O) and trans- bHLH genes and induce skeletal muscle differentiation in ferred to a 1.5 ml centrifuge tube, 10 µl 2 M sodium acetate, pH 4 somites. To study these interactions, we have dissected and and 100 µl water saturated phenol were added and mixed, then 20 µl cultured embryonic somites in the absence or presence of various axial tissues (neural tube, floor plate and notochord) and Dermatome monitored the expression of skeletal muscle- Nt specific genes by RT-PCR. This approach has revealed that myogenesis of epithelial XI somites (stages I-III), isolated from stage Myotome Nc 11− or younger chick embryos, requires Sclerotome interaction with both neural tube and floor plate in vitro. In contrast, myogenesis of Dermomyotome more rostral somites (stages IV-IX), from similar stage embryos, requires interaction V only with the neural tube in the absence of floor plate. This finding suggests that two III Sclerotome signals are necessary for somite myogenesis II I in vitro. We propose a model in which inter- Epithelial 5 day culture; action with the floor plate/notochord Somite RT-PCR analysis mediates the competence of the somite to psm respond to a muscle-promoting signal from the neural tube. Fig. 1. Somite maturation and experimental scheme. The diagram on the left depicts the axis of a stage 11+ chick embryo with rostral at the top and caudal at the bottom. The MATERIALS AND METHODS roman numerals indicate the somite stage according to Ordahl, 1993; the last somite formed from the presegmented mesoderm (=psm) being somite I.