Muscle
differentiation Tim Mohun
National
Institute
for Medical
Research,
London,
UK
Recent studies indicate that vertebrate skeletal muscle originates from two distinct populations of muscle precursor cells formed in early embryogenesis. Divergent patterns of expression for the myoD family of myogenic regulatory genes in different vertebrate embryos suggest some functional redundancy amongst the myogenic factors. Initial gene knockout experiments in mice support this view. Systematic mutagenesis has identified a conserved motif in the basic region of these proteins that is necessary for their biological activity. Direct interaction of both MyoD and myogenin with the transcription factor Jun indicates a mechanism for integrating the control of myoblast growth and differentiation. Current
Opinion
in Cell Biology
the medial half of the newly formed somite, whilst the limb musculature can be traced to cells from the lateral half ( [l*=] and references therein). These populations of muscle precursor cells originate early in development, during chick gastrulation. The medial half of the somite, which subsequently forms the myotome and sclerotome, is derived from the lateral portion of Hensen’s node. The lateral half is derived from cells of the primitive streak, caudal to Hensen’s node [ 21. Transplantation experiments, in which the lateral and medial halves of the newly formed somite are interchanged, demonstrate that the two myogenic cell populations initially have similar developmental potential [lo*]. Subsequent restriction to axial and limb muscle fates occurs during somite maturation, indicating that extrinsic influences control the decision to follow one, or perhaps both, developmental pathways. Several lines of evidence suggest that the adjacent notochord and neural tube are the source of such an influence ( [ 3**] and references therein). Ablation of these axial structures in avian embryos results in the disappearance of myotomes, sclerotomes and the consequent loss of axial tissues, including muscle. Somite formation from the unsegmented mesoderm is not itself disrupted in these embryos, rather the newly formed somites rapidly degenerate through cell death. Older somites, which have already been in contact with the adjacent neural tube-notochord complex for a period, are unaffected, and the effects on newly formed somites can be prevented by re-implantation of either neural tube or notochord. These results suggest that signals from adjacent axial structures in the embryo are necessary for the survival of newly formed somites. Strikingly, limb muscles in neuralectomized and notochordectomized embryos are relatively normal, despite the profound morphological disruption. The effects of these treatments are thus
Introduction
The skeletal muscle of vertebrates is derived from the somites of the early embryo. Precursor muscle cells are formed early in embryonic development and differentiate to form both the axial and limb musculature. To begin to elucidate this complex developmental process, researchers have addressed several basic questions. Do all the skeletal muscles differentiate from a single lineage or do different muscles derive from distinct precursor populations? What is the molecular basis for commitment of cells to form muscle? What mechanisms underlie the orderly expression of muscle-specific genes during terminal differentiation?
Origins
of axial and limb
1992, 4:923-928
muscle
Formation of the son-rites along the rostro-caudal axis of vertebrate embryos appears broadly similar in mammals, birds and amphibians. Individual somites arise by progressive segmentation of the presomitic mesoderm on either side of the neural tube. Each somite becomes compartmentalized into dermamyotome and sclerotome. The former organizes into dermatome and myotome, which d8erentiate into dermis and skeletal muscle, respectively. The sclerotome gives rise to the cartilage of the embryonic skeleton. Whilst amphibian embryos have provided a useful experimental system to investigate formation and patterning of the mesoderm, avian embryos have provided a model to study the origins of skeletal muscle. Lineage labelling and transplantation experiments indicate that the newly formed somite comprizes at least two distinct and segregated populations of muscle precursor cells. Axial muscle of the embryo is derived from
Abbreviations HLH-helix-loop-helix.
@ Current
Biology
Ltd ISSN 0955-0674
.
restricted to tissues derived from the medial half of the somite. As both medial and lateral halves of the early somite possess equivalent developmental potential, it seems likely that, as well as ensuring their survival, the neural tube-notochord complex also inlluences the developmental fate of the medial somitic cells. Consistent with this view, differentiation of muscle from explanted early somites is significantIy enhanced by the presence of neural tube-notochord complex, whilst older somites possess the ability to undergo autonomous myogenic differentiation [3**,4].
Expression of the MyoD development
family
during
early
What are the molecular events that underlie the formation of muscle precursor populations? Studies of myogenesis in cultured cells demonstrate that the MyoD family of proteins plays a central role in regulating muscle differ entiation and may also be responsible for the restricted cell fate of myoblasts (reviewed in [ 5-71). The individual roles of these proteins, however, have proved difficult to assess, owing to the differences in the response of individual cell lines to forced expression of the myogenic factors and to the ability of these proteins to cross-activate expression of other myoD gene family members. With the advent of techniques to manipulate gene expression in vertebrate embryos, it should be possible to identify the roles that the myogenic factors play during early development. As yet, we can only infer conclusions by comparing their normal patterns of expression in different vertebrates. In mouse embryo, myf5 is the first myogenic factor gene to be expressed; its transcripts have been detected by in situ hybridization in the dorsomedial lip of the dermamyotome [8,9]. The myogenin, MRF4 and myoD genes are all successively activated later in development and their expression coincides with terminal differentiation of the myotome. Transcription of the myogenin and myoD genes is maintained throughout later development of muscle tissue, whilst that of myJ5 is restricted to the period of somite formation. MRF4 gene transcripts show a biphasic temporal pattern, indicating that the MRF4 gene is re-activated in the myotomes of older embryos. A broadly similar pattern of expression has been found in the limb buds, the only difference being the relatively early detection of MRF4 gene transcripts compared with those from myogenin and myoD genes [9]. In birds, the myogenic factor genes are also activated in a temporal sequence during early development, but their order of expression differs from that found in mouse [lo*]. In quail embryo, expression of the myoD homologue, qmfl, is first detected in medial cells within the newly formed somite and subsequently in the dorsal medial Lip of the dermamyotome. In mature somites, qmfl expression is detected throughout the myotome.
The rnj$5 homologue, qmf?, shows a similar pattern of expression, but is absent in the youngest somites. Transcription of both genes thus precedes the onset of terminal differentiation, Transcripts from the qmJ2 gene (the mq’ogenin homologue) are only detected in the myotome of compartmentalized somites, coincident with myogenic differentiation. The quail homologue of MRF4 has not been described, but in chick, MRF4 expression is detected later than that of myogenin [ill. In the amphibian, Xenopus, expression of genes encoding the myoD gene homologue, XmyoD, can be divided into several, distinct phases (reviewed in [ 12*] >.A low level of XmyoD transcripts is inherited from the unfertilized egg and distributed throughout the cleaving embryo. At the onset of zygotic transcription, XmyoD transcription is activated throughout the embryo, but the transcripts are unstable and by gastrulation they are lost. A third phase of expression begins in the newly forming mesodem-i of early gastrulae. XmjjoDgene transcripts are initially found throughout the mesoderm, but are later restricted to pre-somitic, dorsal regions. Immunocytochemical studies show that the protein is initially present in all regions of the mesoderm, except the most dorsal cells, which will subsequently form the notchord [ 13*]. Staining is most intense in dorsal mesodermal cells and becomes restricted to the premyotomal and myotomal cells in later embryos. Xenopus m$5 expression is also detected in the presomitic mesoderm of gastrulae before somite formation but, unlike XmyoD, it is largely restricted to the posterior region [ 141. In contrast, MW4 gene transcripts are only detected after neurulation is complete and the embryonic musculature has begun to differentiate, whilst no transcripts from the Xenopus myogenin gene have been detected at any stage of development [ 15). What are the implications of these studies? First, the activation of at least some of the myogenic factor genes before myogenic differentiation is consistent with a regulatory role for their proteins in determining a muscle-cell fate. However, much less is known about the distribution of the proteins themselves and the role of modifications in regulating their activity. In the case of the mouse rnJo genin gene, the protein can only be detected in somites two days after the onset of gene transcription, indicating that post-transcriptional controls regulate its expression [16*]. Second, if expression of the myogenic factor genes is necessary for myogenic determination, it may not be sufficient, as XmyoD is initially activated throughout the mesoderm [ 12.1. Furthermore, ectopic expression of the XmyoD or Xmyf5 genes fails to produce ectopic muscle tissue in early embryos, despite the high levels of transcripts injected [14,17]. As muscle-specific genes are activated in non-muscle tissue after mRNA injection, we can assume that myogenic factor protein, with at least some functional activity, is produced at the ectopic site. Third, sequential activation of individual myogenic factor genes suggests that these genes form a regulatory hierarchy [l@] and that particular proteins perform distinct functions during myogenesis. Consistent with this view, MRF4 can be distinguished from the other fac-
Muscle
tars by the target genes that it transactivates in cultured cells [ 11,19-211. However, there is also likely to be redundancy in function between the myogenic proteins, because their order of expression differs between vertebrates. Taken together with cell culture data, it seems likely that myoD and myf5 act early in myogenesis, whilst myogenin and MIW-4 function later during terminal differentiation. Direct evidence for functional redundancy amongst the myogenic factor genes has most recently been provided by transgenic mice in which either the myoD of myf5 genes were inactivated by homologous recombination [22**,23**]. Mice homozygous for a mutant myoD gene are viable and show no apparent morphological abnormalities. However, the level of rnj$5 mRNA is elevated several fold in skeletal muscle during postnatal development, whilst myogenin and MW4 transcripts accumulate to normal levels. These results suggest that the Myf-5 protein can substitute for MyoD in the mutant mice and that expression of the myf5 gene may be negatively regulated by MyoD during normal development [22**]. Inactivation of the my!5 gene also has little effect on normal muscle development, although the mutation results in immediate postnatal death, as the mice are unable to breathe. This is due to a lack of a funtional rib cage, resulting from the absence of the distal portion of each rib and a truncation of the sternum. Skeletal muscle from newborn mutant mice is morphologically normal and the pattern of muscle-specific gene expression appears unperturbed. One notable effect of the mutation is a considerable delay in the first appearance of myotomal cells during embryogenesis. Although this has no apparent effect on the subsequent myogenesis, it may affect differentiation of sclerotomal cells, that are also derived from the somite [ 23=*]. It is clear from these studies that the interplay of the myogenic factors during muscle differentiation is much more complex than a simple hierarchy. Further transgenic studies should indiacte the extent of their functional redundancy and help define their individual roles. A striking difference between amphibians and other vertebrates is the timing of myogenic factor expression relative to somitogenesis. In Xenopus, the stabilization of myoD transcription in the paraxial, dorsal mesoderm is probably a consequence of dorso-ventral signalling, which patterns the newly formed mesoderm before segmentation [12-l. A functionally equivalent event in avian and mammalian embryos might be the inIluence of the neural tube and notochord on the newly formed somites. Lastly, it is intriguing that no expression of the myogenic factors is detected in the lateral half of the early somite in avian embryos, although these cells give rise to limb bud muscle. One possibility is that limb muscle precursors commence expression of the myogenic factors much later than their axial counterparts, perhaps only after they reach the newly formed limb buds [16-l. Consistent with this, no myogenic factor-expressing cells that might constitute migratory limb muscle precursors have been detected in mouse embryos [8].
Transcriptional
activation
differentiation
by the MyoD
Mohun family
The role of myogenic factors as transcriptional activators has been extensively reviewed elsewhere [5-7,241. Recent studies have begun to elucidate the structural domains responsible for their biological activity. In MyoD, the amino-terminal 56 residues can act as a transcriptional activator when linked to a heterologous DNAbinding domain [25-l. In myogenin, both amino- and carboxyl-terminal domains possess this property [26*], although only the amino-terminal domain is necessary to induce myogenesis. In Myf-5, both amino- and carboxylterminal activation domains are required for function of the protein [27]. The myogenic factors bind to their target sites as heteroligomers, (most probably dimers [28,29] > with members of the E-protein family [30**]. Oligomerization requires the helix-loop-helix (HLI-I) domain of the proteins, whilst the adjacent basic region is required for DNA binding. Chime& proteins in which the basic region is substituted by the equivalent portion of the El2 protein, retain the ability to bind E-box sites, but lose their biological activity [31,32]. This suggests that the precise sequence of the basic region is important for function. Systematic mutagenesis has identified two adjacent amino acids, alanine and threonine (Ala-Thr) that are essential both for transactivation of target promoters and myogenic conversion of fibroblasts [ 25*,27,33,34*]. The presence of these distinguishes the myogenic factors from other basic HLH proteins [ 331. One model to account for these observations suggests that the Ala-Thr residues permit a conformational change of the basic HLH heteroligomer, consequent upon binding to the appropriate E-box site. This might unmask transcription activation domains in the myogenic protein that were previously held in an inactive state [260]. Another model invokes an additional factor necessary for transcriptional activation function of the E-box-basic HLH protein complex [ 25*]. Binding of this ‘recognition factor’ requires the Ala-Thr residues of the myogenic factor basic region, as well as appropriate elements of target DNA site, and results in the unmasking of the transcription activation domain. As the myogenic factors can function in a wide variety of cell lines, the recognition factor must be widely distributed, rather than muscle-specific, yet its requirement would ensure specificity of action of the myogenic protein.
Myogenesis
and cell growth
During differentiation of skeletal muscle cells in culture, proliferating myoblasts cease dividing, fuse to form multinucleate myotubes and commence expression of muscle-specific genes. This process depends on depletion of serum growth factors from the culture medium, indicating that proliferation and terminal differentiation are, in some manner, mutually exclusive. Indeed, many studies have shown that purified growth factors and activated oncogenes can block myoblast fusion and terminal
925
926
Cell differentiation
differentiation, supporting the notion that groafh and differentiation are antagonistic [35,36]. In embryonic development, the decision to form muscle precursors is only one of many alternative developmental fates; nevertheless the cell culture model may indicate the molecular mechanisms that underlie such choices. A basis for the antagonism between cell proliferation and muscle differentiation is suggested by studies of the oncogene jun in myogenic cells. The c-~ZUZ gene encodes a transcription factor that can associate with itself, other proteins of the Jun family, the c-fos and related proteins or the cyclic AMP response element binding protein to regulate a wide variety of genes that contain APl or CREbinding sites within their control regions. The c-jun gene is activated as an immediate response to growth factor stimulation and the gene product mediates growth factor regulation of target genes via the API complex [ 371. During myogenesis, c-jun expression declines as myoblasts differentiate. Constitutive expression of the cellular oncogene, or its viral counterpart in myoblasts, blocks cell fusion and terminal differentiation in a dose-dependent manner [38,3~]; expression in differentiated myotubes results in the down-regulation of muscle-specific genes [40]. Similarly, transactivation of muscle-specific reporter constructs by MyoD and myogenin is blocked by Jun in cultured fibroblasts [39**,41=*] even when transcription is driven solely by multimerized E-box sites. As E2A gene products can activate appropriate target genes in cells that constitutively express c-jun [41**], the target of the Jun protein appears to be the myogenic factors themselves. In vitro, MyoD and Jun proteins can associate with each other and a MyoD-Jun complex has also been detected in extracts from cultured myogenic cells [39**]. In rttro, a truncated Jun protein, containing only the carboxylterminal basic and leucine-zipper regions, is sufficient to form a complex with MyoD [39**] and deletion of this region reduces inhibitoly activity in vivo [39-,41-l. However, c-jun deletion mutants lacking the amino-terminal 87 residues also lose their inhibitory activity and the amino-terminal region will confer the ability to repress muscle-specific transcription on heterologous proteins [41**]. These findings raise the possibility that c-Jun may also inhibit myogenesis by a second mechanism, which does not involve direct interaction with MyoD.
for normal muscle development and suggest functional redundancy within the MyoD family. The extent of this redundancy and the distinct functional roles of individual myogenic factors should be clarified by further transgenic studies, which are currently in progress. The interactions of MyoD with Jun provide a first account of how the choice between differentiation and proliferation may be regulated in a dynamic manner. This may serve as a model for investigating the molecular mechanisms that regulate the choice of muscle and non-muscle fates in the early embryo.
References
Papers of particular interest, published \iew. have been highlighted as: . of special interest . . of outstanding interest 1.
ORDAHL
..
the
CP, LE ~~butlN
Developing
Somite.
reading within
the annual
period
of re-
NM: Two Myogenic Lineages within Detve/opment 1992. 114:339-353.
Elegant demonstration using chickquail grafts that the medial and lat. eml halves of the early avian somite represent distinct myogenic Iin. cages. These gave rise to the body and limb musculature, respectively, in reciprocal grafting experiments, although in early somites the two halves were interchangeable. 2.
SELLECK M, STERN CD: Fate Mapping and Cell ysis of Hensen’s Node in the Chick Embryo. 1991. 112:615-626.
Lineage AnalDel’elopmeM
3.
RONG PM, TEILLET M-A, ZILIJ% C, LE D~LIARJN
. .
Tube/Notochord Complex is Necessary for Vertebral but not Limb and Body Wall Striated Muscle Differentiation. Delalopmen/ 1992. 1151657-672.
NM: The
Neural
The effects of separation or removal of the neural tube-notochord complex on subsequent muscle formation in the chick embryo was assessed using a monoclonal antibody specilic for muscle cells. Ablation of these axial structures resulted in the degeneration of newly formed somites. due to cell death, but limb muscle nevertheless formed in these embryos. Differentiation of somites to give a?tial muscle required a period of contact with these axial structures and this effect was reproduced with cultured somites.
4.
&NNS.MO~BS
5.
EMERSON CP: Myogenesis
T. THOK~C~~~D P: Autonomy of Differentiation in Avian Brachial Somites and the Influence of Adjacent Tissues. Dec~elopment 1987, 100:44H62.
Cut-r
Opin
Cell Rio/
and Developmental 1990,
Control
Genes.
2:1065-1075.
6.
OLIN Getzes
7.
WEIN[RALIB H, DALES R, TAPSCO’II S. THAYER M. KRAUSE M, BENEZRA R, BIACKWIU TK, TURNER D. R~wr R. HOUENBERG
EN: MyoD Family: a Paradigm Decs lW0, 4:1454-1461.
S. ET A.: Specification 251:761-766.
Conclusions
Studies of cultured cells have provided a model for myogenesis that can now be examined in the context of early embryonic development. Expression of the MyoD family in vertebrate embryos from the earliest stage of somitogenesis supports the view that these genes play a central role in terminal differentiation of embryonic muscle. The precise role of the myogenic factors in the commitment of embryonic cells to a muscle fate should be clarified by gene knock-out experiments. Initial results in mice demonstrate that neither MyoD nor Myf-5 is essential
and recommended
8.
The of
BUCKINGHAM
myoD the
M: Making
Gene Muscle
for
Family: Nodal Cell Lineage.
Muscle
in Mammals.
Development
?
Point during Science 1991, Trends
Gene1
1992, 8:144-149. 9.
HANNON
K, SMITH
C, BALES KR, S~vrriw
RF: Temporal
and
Quantitative Analysis of Myogenic Regulatory and Growth Factor Gene Expression in the Developing Mouse Embryo. Dell Eiol 1992, 151:137-144. 10. .
ME,
Myogenic in Quail
Regulatory Genes Embryos. Deu Biol
In S~IU hybridization showed a sequential
EMERSON
CJ: Sequential
POWNAIL
Activation of Three during Somite Morphogenesis 1992. 151:67-79.
shows that three quail pattern of expression
myogenic factors (Qmfl-3) during early development.
Muscle Qmfl was detected in the medial portion of the newly formed somite. consistent with expression in axial muscle precursors (see [ le.1 ). 11.
FUJISAWA-SEHA~~A A. NARESHIMA Y, KOM~YA T, U~tsuto T, A~AKURA A, NA~ESHIMA Y: Differential Trans-activation of Muscle-specific Regulatory Elements including the Mysosin Light Chain Box by Chicken MyoD, Myogenin, and MRF4. / Biol &em 1992, 267:10031-10038,
12. .
FRANK D, GARLAND RM: Transient Expression of X&fyoD in Non-somitic Mesoderm of Xenopus gashtfae. Detpeelopmenr 1991, 113:1387-1393. Expression of XAIyuD was detected by whole-mount in situ hybridization throughout the mesoderm of the early Xenopus gustrulu embtyo. Transcripts were also detected by northern blot analysis in ultra violet light~treated embryos that produce no muscle tissue and in cultured explants that form ventral mesodenn. HOPWG~D ND, PLLICK A, GUFUIION JB, DILWOHTH SM: Expression of XMyoD Protein in Early Xenopus fuevis Embryos. Det~elopmeni 1992, 114:31-38. A detailed study of XMyoD protein distribution in the Xenoplrsembryo using a monoclonal antibody. This gave better resolution than comparable studies of the mRNA and confirmed that XMyoD protein was present in all but the most dorsal cells of the newly formed mesoderm. Expression was localized to the cell nuclei and was later restricted to premyotomal and myotomal cells. The protein was apparently present in every cell of the myotome. 13. .
14.
HOPWO~D ND, PLUCK A, CLIRD~N JB: Early Muscle Cells and can Activate tally in Early Embryos. DetJelopmenl
Xenopus Myf-5 Marks Muscle Genes Ectopi1991, 111:551-60.
15.
JENNINGS CG: Expression of the Myogenic Gene MRF4 ing Xenopus Development. Dell Biol 1992, 151:31’+332.
dur-
16. .
CUSEUA-DE ANGEL& MG, LYONS G, SONNINO C, DE ANGELES AL, V~~ARELU E, FARMER K, Wktctrr WE, MOUNARO M, BOLICHE M, BUCK~NGHAM M, Cossu G: MyoD. Myogenin Independent Dilferentiation of Primordial Myoblasts in Mouse Somites. J Cell Biol 1992, 116:1243-1255. The presence of MyoD, myogenin and myosin heavy-chain proteins was compared in cultured foetal mouse myoblasts. Late somites and limb buds gave rise to cells that expressed the myogenic factors before differentiation; a second population from early somites differentiated in culture without expressing either myogenic factor. These might constitute the migratory limb muscle precursors because they were also de. tected in cultures from limb buds. No antibody was available to identify Myf-5 protein. 17.
BRAUN T, BOBER E, ARNOU) HH: lnhibition of Muscle Differentiation by the Adenovirus Ela Protein: Repression of the Transcriptional Activating Function of the HLH Protein Myf-5. Genes Del! 1992, 6888-902. Expression of adenovirus ElA protein in rat Lb muscle cells blocks the activity of Mfl5 without affecting expression of the m.lfl5gene or DNAbinding ability of the Myf-5 protein. In contrast, myogenin expression is blocked, suggesting that this factor lies downstream of m~f5 in the regulatoty hierarchy. 19.
YUTZEY KE, RHODES SJ, KONIECZNV SF: Differential Transactivation Associated with the Muscle Regulatory Factors MyoDl, Myogenin, and MRF4. I%!o/ Cell Biol 1990, 10:3934-3944.
20.
CHAKRAB~~~TY T, BRENNAN T, ORION E: Differential Transactivation of a Muscle-specific Enhancer by Myogenic Helix-Loop-Helix Proteins is Separable from DNA Biding. J Biol Chem 1991, 266~28782882.
Mohun
HLH Gene muf-5 and Results in Apparently Normal Muscle Development. Cell 1992, 71383390. Homologous recombination in ES cells was used to construct transgenie mice lacking a functional myoD gene. These mice show no apparent abnormalities in muscle dilTerentiation, suggesting that the gene is functionally redundant. my!5 RNA is elevated in the muscles of the mice, indicating that the Myf-5 protein might be substituting for MyoD. BRAUN T, RUDN~CKI MA ARNOU) H-H, JAENISCH R Targeted Inactivation of the Muscle Regulatory Gene myf5 Results in Abnormal Rib Development and Perinatal Death. Cell 1992, 71369-382. rnJf5 is the first myogenic gene to be activated during somitogenesis in the mouse. Transgenic embryos lacking a functional myf5gene show a considerable delay in the first appearance of myotomal cells, but subsequent muscle differentiation is apparently normal. interestingly. the newborn mice die at birth because they lack a normal rib cage and cannot breathe. This suggests that the mu/-‘5 mutation affects normal differentiation of sclerotomal cells. 23. ..
24.
ORDAHI. CP: Developmental Expression. Curr Top Dev
Regulation Biol 1992,
of Sarcomeric 26:145-168.
WEINTRALIB H, Dw~ttto VJ, VE~~MA I, DAMS R, HO~LENEZERG S, SNIDER L, VASSAR A, TAFXOTT SJ: Muscle-specific Transcriptional Activation by MyoD. Genes Deu 1991, 5:1377-1386. The transcriptional-activation domain of MyoD was mapped to the amino-tenninal portion of the protein as assessed by domain swap experiments. Systematic mutagenesis of the basic region identified an Ala-Thr pair of residues to be critical for MyoD function. The authors suggest that another factor that recognizes the basic region is required for MyoD function. 26. .
SCHWARZ JJ, C HAKaABOR’tY T, MARTtN J, EHOU JM, O&ON EN: The Basic Region of Myogenin Cooperates with Two Transcription Activation Domains to induce Muscle-specific Transcription. Mol Cell Biol 1992, 12:266-275. Domainswap experiments demonstrated that the transcriptional activation functions of myogenin map to both amino- and carboxyl-terminal domains of the protein. Activation of reporter genes via E-box sites requires the myogenin basic region, even with chimeric VPlb-myogenin proteins, indicating that the basic region also participates in transctiptional activation Function. 27.
WINTER B, BRAUN T, ARNOU) HH: Domains in the Muscle-specific EMBO J 1992, 11:1843-1855.
Co-operativity Transcription
of Functional Factor myf-5.
28.
ANIH~~-CAHI~ SJ, BENFLELD PA, FA~RMAN R, WASSERMAN ZR, BRENNER SL STAFFORD Wl, ALTENBACH C, HUBBEU Wl, DEGRA~XI WF: Molecular Characterization of Helix-Loop-Helix Peptides. Science 1992, 255979-983.
29.
LJN H, KONIECZNY SF: Identification of MRF4, Myogenin, El2 Oligomer Complexes by Chemical Cross-linking Two-dimensional Gel Electrophoresis. J Biol f&m 26734779780.
and and 1992,
LSAR AB, DAMPS RL, Wtuctrr WE, KADE~cH T, Mutw C, VORONOVA A, BALTIMORE D, WEW~RAUB H: Functional Activity of Myogenic HLH Proteins Requires Hetero-oligomerizacion with E12/E47Rke Proteins in ufuo. Cell 1991, 66305-315. Comprehensive demonstration that E12/E47 proteins associate in vim with MyoD and myogenin. Myogenic conversion of lOT1/2 cells by MOOD results in activation of the muscle-specific regulators, MCAT and MEF-2. The authors discuss the possibility that competition for E-pro. teins provides a mechanism for ensuring that different HLH-dependent developmental decisions are mutually exclusive. 30. ..
31.
DAVIS RL, CHENC PF, LG.%R AB, DNA Binding Domain Contains Muscle-specific Gene Activation.
WEIN~RAUB H: The MyoD a Recognition Code for Cell 1990, 60:733-746.
21.
BLOCK NE, MILLER JB: Expression of MRF4, a Myogenic Helix-Loop-Helix Protein, Produces Multiple Changes in the Myogenic Program of BC3H-1 Cells. MO/ Cell Biol 1992, 12:2484-2492.
32.
CHAKRA~ORIY T, BRENNAN TJ, Lt L, EDMONDSON Inefficient Homooligomerization Contributes dence of Myogenin on E2A Products for Biding. MO/ Cell Biol 1991, 11:3633-3641.
22. ..
RUDNICKI of myoD
33.
BRENNAN TJ, CHAKRAB~~~~Y T, Olson EN: Mutagenesis Myogenin Basic Region Identifies an Ancient Protein
MA, BRAUN B, HINUMA S, JAENISCH R: inactivation in Mice Leads to Up-regulation of the Myogenic
Gene
25. .
HOPWOOD ND, GURD~N JB: Activation of Muscle Genes without Myogenesis by Ectopic Expression of MyoD in Frog Embryo Cells. Nulure 1990. 347:197-200.
18. .
differentiation
D. ORSON EN: to the DepenEfficient DNA of the Motif
927
928
Cell differentiation Critical
USA
for
Activation
of
Myogenesis.
Proc’hfarl
Acad
Sci
1991, &X5675-5679.
DAVIS RL, WE~UB H: Acquisition of Myogenic Specificity by Replacement of Three Amino Acid Residues from MyoD into E12. Science 1992, 256:1027-1030. An El2 protein containing only two replacements (Ala-Thr) in the basic region and a third replacement (Lys) at the junction with helix 1 acquired the ability to convert lOT1/2 fibroblasts into muscle cells.
34.
.
35.
FLOIUNI JR, EW~ON tors and Myogenjc 53:201-216.
36.
OLFON EN, BRENNAN 71, CHAKRAL~OR~ T, CHENC TC. CSERJESI P, EDMOND~ON D, JAMES G, LI L Molecular Control of Myogenesis: Antagonism Between Growth and Differentiation. MoI Cell BicxJxm 1991, 104:7-13.
37.
ANGEL P. KAtuN M: The Role of Jun. Fos and plex in Cell-proliferation and Transformation. /I& AcIa 1991, 1072:12%157.
DZ, MAGIU KA: Differentiation.
Hormones, Growth Annu Ret, P/+0/
the
AP-1
Com-
Bio
SLI HY, BOS TJ, MONTECV\RO FS, VOGT PK: Jun lnbibits genie Differentiation. Oncogene 1991, 6:175+1766.
39. ..
BENGAL E, RAN~oNE L SCHARFMANN WEIN~RAUB H, VERMA IM: Functional Jun and MyoD Proteins: a Direct
1992, 6B:507-519.
40.
Fac1991.
&i&in?
38.
Transactivation of the MyoD promoter, the MCK enhancer and an E. box driven reporter by MyoD were all blocked by cotransfection with a jun expression vector; in reciprocal experiments, MyoD suppressed nansactivation of a reporter containing an AP.1 site by Jun. Direct interaction of the Jun and MyoD was demonstrated both in vitro and in ctfro, requiring the leucine zipper domain of Jun and the HLH region of MyoD.
Myo-
R. DWARKI VJ, TAPSCOIT SJ. Antagonism Between tPhysical Association. Cell
GROW M, CALCONI A, TA-~o F: vjun Oncogene Prevents Terminal Differentiation and Suppresses Muscle-specific Gene Expression in ASVDl7D-infected Muscle Cells. Oncogene 1991, 6:1767-1773.
-11. ..
LI L. C~LAMBARD JC. bRlN M. OLSON EN: Fos and Jun Repress Transcriptional Activation by Myogenin and MyoD: the Amino Terminus of Jun Can Mediate Repression. Genes Del, 1992, 6:67&689. Espression of c.Fos. c.Jun and JunB all suppressed transactivdtion of the MCK enhancer by MyoD or myogenin. Repression activity was targeted at the basic HLH region of the myogenin and mapped to the amino.terminal region of c-Jun. Interestingly, the authors found that the leucine zipper dimeriiJtion domain of cJun was not required for repression (compare with (39**] ).
rJ Mohun. Iabordtov of Developmental tute for Medical Research, The Ridgeway. LIK.
Biochemistry. National Insti. Mill Hill, London NW7 IAA.
differentiation Tim Mohun
National
Institute
for Medical
Research,
London,
UK
Recent studies indicate that vertebrate skeletal muscle originates from two distinct populations of muscle precursor cells formed in early embryogenesis. Divergent patterns of expression for the myoD family of myogenic regulatory genes in different vertebrate embryos suggest some functional redundancy amongst the myogenic factors. Initial gene knockout experiments in mice support this view. Systematic mutagenesis has identified a conserved motif in the basic region of these proteins that is necessary for their biological activity. Direct interaction of both MyoD and myogenin with the transcription factor Jun indicates a mechanism for integrating the control of myoblast growth and differentiation. Current
Opinion
in Cell Biology
the medial half of the newly formed somite, whilst the limb musculature can be traced to cells from the lateral half ( [l*=] and references therein). These populations of muscle precursor cells originate early in development, during chick gastrulation. The medial half of the somite, which subsequently forms the myotome and sclerotome, is derived from the lateral portion of Hensen’s node. The lateral half is derived from cells of the primitive streak, caudal to Hensen’s node [ 21. Transplantation experiments, in which the lateral and medial halves of the newly formed somite are interchanged, demonstrate that the two myogenic cell populations initially have similar developmental potential [lo*]. Subsequent restriction to axial and limb muscle fates occurs during somite maturation, indicating that extrinsic influences control the decision to follow one, or perhaps both, developmental pathways. Several lines of evidence suggest that the adjacent notochord and neural tube are the source of such an influence ( [ 3**] and references therein). Ablation of these axial structures in avian embryos results in the disappearance of myotomes, sclerotomes and the consequent loss of axial tissues, including muscle. Somite formation from the unsegmented mesoderm is not itself disrupted in these embryos, rather the newly formed somites rapidly degenerate through cell death. Older somites, which have already been in contact with the adjacent neural tube-notochord complex for a period, are unaffected, and the effects on newly formed somites can be prevented by re-implantation of either neural tube or notochord. These results suggest that signals from adjacent axial structures in the embryo are necessary for the survival of newly formed somites. Strikingly, limb muscles in neuralectomized and notochordectomized embryos are relatively normal, despite the profound morphological disruption. The effects of these treatments are thus
Introduction
The skeletal muscle of vertebrates is derived from the somites of the early embryo. Precursor muscle cells are formed early in embryonic development and differentiate to form both the axial and limb musculature. To begin to elucidate this complex developmental process, researchers have addressed several basic questions. Do all the skeletal muscles differentiate from a single lineage or do different muscles derive from distinct precursor populations? What is the molecular basis for commitment of cells to form muscle? What mechanisms underlie the orderly expression of muscle-specific genes during terminal differentiation?
Origins
of axial and limb
1992, 4:923-928
muscle
Formation of the son-rites along the rostro-caudal axis of vertebrate embryos appears broadly similar in mammals, birds and amphibians. Individual somites arise by progressive segmentation of the presomitic mesoderm on either side of the neural tube. Each somite becomes compartmentalized into dermamyotome and sclerotome. The former organizes into dermatome and myotome, which d8erentiate into dermis and skeletal muscle, respectively. The sclerotome gives rise to the cartilage of the embryonic skeleton. Whilst amphibian embryos have provided a useful experimental system to investigate formation and patterning of the mesoderm, avian embryos have provided a model to study the origins of skeletal muscle. Lineage labelling and transplantation experiments indicate that the newly formed somite comprizes at least two distinct and segregated populations of muscle precursor cells. Axial muscle of the embryo is derived from
Abbreviations HLH-helix-loop-helix.
@ Current
Biology
Ltd ISSN 0955-0674
.
restricted to tissues derived from the medial half of the somite. As both medial and lateral halves of the early somite possess equivalent developmental potential, it seems likely that, as well as ensuring their survival, the neural tube-notochord complex also inlluences the developmental fate of the medial somitic cells. Consistent with this view, differentiation of muscle from explanted early somites is significantIy enhanced by the presence of neural tube-notochord complex, whilst older somites possess the ability to undergo autonomous myogenic differentiation [3**,4].
Expression of the MyoD development
family
during
early
What are the molecular events that underlie the formation of muscle precursor populations? Studies of myogenesis in cultured cells demonstrate that the MyoD family of proteins plays a central role in regulating muscle differ entiation and may also be responsible for the restricted cell fate of myoblasts (reviewed in [ 5-71). The individual roles of these proteins, however, have proved difficult to assess, owing to the differences in the response of individual cell lines to forced expression of the myogenic factors and to the ability of these proteins to cross-activate expression of other myoD gene family members. With the advent of techniques to manipulate gene expression in vertebrate embryos, it should be possible to identify the roles that the myogenic factors play during early development. As yet, we can only infer conclusions by comparing their normal patterns of expression in different vertebrates. In mouse embryo, myf5 is the first myogenic factor gene to be expressed; its transcripts have been detected by in situ hybridization in the dorsomedial lip of the dermamyotome [8,9]. The myogenin, MRF4 and myoD genes are all successively activated later in development and their expression coincides with terminal differentiation of the myotome. Transcription of the myogenin and myoD genes is maintained throughout later development of muscle tissue, whilst that of myJ5 is restricted to the period of somite formation. MRF4 gene transcripts show a biphasic temporal pattern, indicating that the MRF4 gene is re-activated in the myotomes of older embryos. A broadly similar pattern of expression has been found in the limb buds, the only difference being the relatively early detection of MRF4 gene transcripts compared with those from myogenin and myoD genes [9]. In birds, the myogenic factor genes are also activated in a temporal sequence during early development, but their order of expression differs from that found in mouse [lo*]. In quail embryo, expression of the myoD homologue, qmfl, is first detected in medial cells within the newly formed somite and subsequently in the dorsal medial Lip of the dermamyotome. In mature somites, qmfl expression is detected throughout the myotome.
The rnj$5 homologue, qmf?, shows a similar pattern of expression, but is absent in the youngest somites. Transcription of both genes thus precedes the onset of terminal differentiation, Transcripts from the qmJ2 gene (the mq’ogenin homologue) are only detected in the myotome of compartmentalized somites, coincident with myogenic differentiation. The quail homologue of MRF4 has not been described, but in chick, MRF4 expression is detected later than that of myogenin [ill. In the amphibian, Xenopus, expression of genes encoding the myoD gene homologue, XmyoD, can be divided into several, distinct phases (reviewed in [ 12*] >.A low level of XmyoD transcripts is inherited from the unfertilized egg and distributed throughout the cleaving embryo. At the onset of zygotic transcription, XmyoD transcription is activated throughout the embryo, but the transcripts are unstable and by gastrulation they are lost. A third phase of expression begins in the newly forming mesodem-i of early gastrulae. XmjjoDgene transcripts are initially found throughout the mesoderm, but are later restricted to pre-somitic, dorsal regions. Immunocytochemical studies show that the protein is initially present in all regions of the mesoderm, except the most dorsal cells, which will subsequently form the notchord [ 13*]. Staining is most intense in dorsal mesodermal cells and becomes restricted to the premyotomal and myotomal cells in later embryos. Xenopus m$5 expression is also detected in the presomitic mesoderm of gastrulae before somite formation but, unlike XmyoD, it is largely restricted to the posterior region [ 141. In contrast, MW4 gene transcripts are only detected after neurulation is complete and the embryonic musculature has begun to differentiate, whilst no transcripts from the Xenopus myogenin gene have been detected at any stage of development [ 15). What are the implications of these studies? First, the activation of at least some of the myogenic factor genes before myogenic differentiation is consistent with a regulatory role for their proteins in determining a muscle-cell fate. However, much less is known about the distribution of the proteins themselves and the role of modifications in regulating their activity. In the case of the mouse rnJo genin gene, the protein can only be detected in somites two days after the onset of gene transcription, indicating that post-transcriptional controls regulate its expression [16*]. Second, if expression of the myogenic factor genes is necessary for myogenic determination, it may not be sufficient, as XmyoD is initially activated throughout the mesoderm [ 12.1. Furthermore, ectopic expression of the XmyoD or Xmyf5 genes fails to produce ectopic muscle tissue in early embryos, despite the high levels of transcripts injected [14,17]. As muscle-specific genes are activated in non-muscle tissue after mRNA injection, we can assume that myogenic factor protein, with at least some functional activity, is produced at the ectopic site. Third, sequential activation of individual myogenic factor genes suggests that these genes form a regulatory hierarchy [l@] and that particular proteins perform distinct functions during myogenesis. Consistent with this view, MRF4 can be distinguished from the other fac-
Muscle
tars by the target genes that it transactivates in cultured cells [ 11,19-211. However, there is also likely to be redundancy in function between the myogenic proteins, because their order of expression differs between vertebrates. Taken together with cell culture data, it seems likely that myoD and myf5 act early in myogenesis, whilst myogenin and MIW-4 function later during terminal differentiation. Direct evidence for functional redundancy amongst the myogenic factor genes has most recently been provided by transgenic mice in which either the myoD of myf5 genes were inactivated by homologous recombination [22**,23**]. Mice homozygous for a mutant myoD gene are viable and show no apparent morphological abnormalities. However, the level of rnj$5 mRNA is elevated several fold in skeletal muscle during postnatal development, whilst myogenin and MW4 transcripts accumulate to normal levels. These results suggest that the Myf-5 protein can substitute for MyoD in the mutant mice and that expression of the myf5 gene may be negatively regulated by MyoD during normal development [22**]. Inactivation of the my!5 gene also has little effect on normal muscle development, although the mutation results in immediate postnatal death, as the mice are unable to breathe. This is due to a lack of a funtional rib cage, resulting from the absence of the distal portion of each rib and a truncation of the sternum. Skeletal muscle from newborn mutant mice is morphologically normal and the pattern of muscle-specific gene expression appears unperturbed. One notable effect of the mutation is a considerable delay in the first appearance of myotomal cells during embryogenesis. Although this has no apparent effect on the subsequent myogenesis, it may affect differentiation of sclerotomal cells, that are also derived from the somite [ 23=*]. It is clear from these studies that the interplay of the myogenic factors during muscle differentiation is much more complex than a simple hierarchy. Further transgenic studies should indiacte the extent of their functional redundancy and help define their individual roles. A striking difference between amphibians and other vertebrates is the timing of myogenic factor expression relative to somitogenesis. In Xenopus, the stabilization of myoD transcription in the paraxial, dorsal mesoderm is probably a consequence of dorso-ventral signalling, which patterns the newly formed mesoderm before segmentation [12-l. A functionally equivalent event in avian and mammalian embryos might be the inIluence of the neural tube and notochord on the newly formed somites. Lastly, it is intriguing that no expression of the myogenic factors is detected in the lateral half of the early somite in avian embryos, although these cells give rise to limb bud muscle. One possibility is that limb muscle precursors commence expression of the myogenic factors much later than their axial counterparts, perhaps only after they reach the newly formed limb buds [16-l. Consistent with this, no myogenic factor-expressing cells that might constitute migratory limb muscle precursors have been detected in mouse embryos [8].
Transcriptional
activation
differentiation
by the MyoD
Mohun family
The role of myogenic factors as transcriptional activators has been extensively reviewed elsewhere [5-7,241. Recent studies have begun to elucidate the structural domains responsible for their biological activity. In MyoD, the amino-terminal 56 residues can act as a transcriptional activator when linked to a heterologous DNAbinding domain [25-l. In myogenin, both amino- and carboxyl-terminal domains possess this property [26*], although only the amino-terminal domain is necessary to induce myogenesis. In Myf-5, both amino- and carboxylterminal activation domains are required for function of the protein [27]. The myogenic factors bind to their target sites as heteroligomers, (most probably dimers [28,29] > with members of the E-protein family [30**]. Oligomerization requires the helix-loop-helix (HLI-I) domain of the proteins, whilst the adjacent basic region is required for DNA binding. Chime& proteins in which the basic region is substituted by the equivalent portion of the El2 protein, retain the ability to bind E-box sites, but lose their biological activity [31,32]. This suggests that the precise sequence of the basic region is important for function. Systematic mutagenesis has identified two adjacent amino acids, alanine and threonine (Ala-Thr) that are essential both for transactivation of target promoters and myogenic conversion of fibroblasts [ 25*,27,33,34*]. The presence of these distinguishes the myogenic factors from other basic HLH proteins [ 331. One model to account for these observations suggests that the Ala-Thr residues permit a conformational change of the basic HLH heteroligomer, consequent upon binding to the appropriate E-box site. This might unmask transcription activation domains in the myogenic protein that were previously held in an inactive state [260]. Another model invokes an additional factor necessary for transcriptional activation function of the E-box-basic HLH protein complex [ 25*]. Binding of this ‘recognition factor’ requires the Ala-Thr residues of the myogenic factor basic region, as well as appropriate elements of target DNA site, and results in the unmasking of the transcription activation domain. As the myogenic factors can function in a wide variety of cell lines, the recognition factor must be widely distributed, rather than muscle-specific, yet its requirement would ensure specificity of action of the myogenic protein.
Myogenesis
and cell growth
During differentiation of skeletal muscle cells in culture, proliferating myoblasts cease dividing, fuse to form multinucleate myotubes and commence expression of muscle-specific genes. This process depends on depletion of serum growth factors from the culture medium, indicating that proliferation and terminal differentiation are, in some manner, mutually exclusive. Indeed, many studies have shown that purified growth factors and activated oncogenes can block myoblast fusion and terminal
925
926
Cell differentiation
differentiation, supporting the notion that groafh and differentiation are antagonistic [35,36]. In embryonic development, the decision to form muscle precursors is only one of many alternative developmental fates; nevertheless the cell culture model may indicate the molecular mechanisms that underlie such choices. A basis for the antagonism between cell proliferation and muscle differentiation is suggested by studies of the oncogene jun in myogenic cells. The c-~ZUZ gene encodes a transcription factor that can associate with itself, other proteins of the Jun family, the c-fos and related proteins or the cyclic AMP response element binding protein to regulate a wide variety of genes that contain APl or CREbinding sites within their control regions. The c-jun gene is activated as an immediate response to growth factor stimulation and the gene product mediates growth factor regulation of target genes via the API complex [ 371. During myogenesis, c-jun expression declines as myoblasts differentiate. Constitutive expression of the cellular oncogene, or its viral counterpart in myoblasts, blocks cell fusion and terminal differentiation in a dose-dependent manner [38,3~]; expression in differentiated myotubes results in the down-regulation of muscle-specific genes [40]. Similarly, transactivation of muscle-specific reporter constructs by MyoD and myogenin is blocked by Jun in cultured fibroblasts [39**,41=*] even when transcription is driven solely by multimerized E-box sites. As E2A gene products can activate appropriate target genes in cells that constitutively express c-jun [41**], the target of the Jun protein appears to be the myogenic factors themselves. In vitro, MyoD and Jun proteins can associate with each other and a MyoD-Jun complex has also been detected in extracts from cultured myogenic cells [39**]. In rttro, a truncated Jun protein, containing only the carboxylterminal basic and leucine-zipper regions, is sufficient to form a complex with MyoD [39**] and deletion of this region reduces inhibitoly activity in vivo [39-,41-l. However, c-jun deletion mutants lacking the amino-terminal 87 residues also lose their inhibitory activity and the amino-terminal region will confer the ability to repress muscle-specific transcription on heterologous proteins [41**]. These findings raise the possibility that c-Jun may also inhibit myogenesis by a second mechanism, which does not involve direct interaction with MyoD.
for normal muscle development and suggest functional redundancy within the MyoD family. The extent of this redundancy and the distinct functional roles of individual myogenic factors should be clarified by further transgenic studies, which are currently in progress. The interactions of MyoD with Jun provide a first account of how the choice between differentiation and proliferation may be regulated in a dynamic manner. This may serve as a model for investigating the molecular mechanisms that regulate the choice of muscle and non-muscle fates in the early embryo.
References
Papers of particular interest, published \iew. have been highlighted as: . of special interest . . of outstanding interest 1.
ORDAHL
..
the
CP, LE ~~butlN
Developing
Somite.
reading within
the annual
period
of re-
NM: Two Myogenic Lineages within Detve/opment 1992. 114:339-353.
Elegant demonstration using chickquail grafts that the medial and lat. eml halves of the early avian somite represent distinct myogenic Iin. cages. These gave rise to the body and limb musculature, respectively, in reciprocal grafting experiments, although in early somites the two halves were interchangeable. 2.
SELLECK M, STERN CD: Fate Mapping and Cell ysis of Hensen’s Node in the Chick Embryo. 1991. 112:615-626.
Lineage AnalDel’elopmeM
3.
RONG PM, TEILLET M-A, ZILIJ% C, LE D~LIARJN
. .
Tube/Notochord Complex is Necessary for Vertebral but not Limb and Body Wall Striated Muscle Differentiation. Delalopmen/ 1992. 1151657-672.
NM: The
Neural
The effects of separation or removal of the neural tube-notochord complex on subsequent muscle formation in the chick embryo was assessed using a monoclonal antibody specilic for muscle cells. Ablation of these axial structures resulted in the degeneration of newly formed somites. due to cell death, but limb muscle nevertheless formed in these embryos. Differentiation of somites to give a?tial muscle required a period of contact with these axial structures and this effect was reproduced with cultured somites.
4.
&NNS.MO~BS
5.
EMERSON CP: Myogenesis
T. THOK~C~~~D P: Autonomy of Differentiation in Avian Brachial Somites and the Influence of Adjacent Tissues. Dec~elopment 1987, 100:44H62.
Cut-r
Opin
Cell Rio/
and Developmental 1990,
Control
Genes.
2:1065-1075.
6.
OLIN Getzes
7.
WEIN[RALIB H, DALES R, TAPSCO’II S. THAYER M. KRAUSE M, BENEZRA R, BIACKWIU TK, TURNER D. R~wr R. HOUENBERG
EN: MyoD Family: a Paradigm Decs lW0, 4:1454-1461.
S. ET A.: Specification 251:761-766.
Conclusions
Studies of cultured cells have provided a model for myogenesis that can now be examined in the context of early embryonic development. Expression of the MyoD family in vertebrate embryos from the earliest stage of somitogenesis supports the view that these genes play a central role in terminal differentiation of embryonic muscle. The precise role of the myogenic factors in the commitment of embryonic cells to a muscle fate should be clarified by gene knock-out experiments. Initial results in mice demonstrate that neither MyoD nor Myf-5 is essential
and recommended
8.
The of
BUCKINGHAM
myoD the
M: Making
Gene Muscle
for
Family: Nodal Cell Lineage.
Muscle
in Mammals.
Development
?
Point during Science 1991, Trends
Gene1
1992, 8:144-149. 9.
HANNON
K, SMITH
C, BALES KR, S~vrriw
RF: Temporal
and
Quantitative Analysis of Myogenic Regulatory and Growth Factor Gene Expression in the Developing Mouse Embryo. Dell Eiol 1992, 151:137-144. 10. .
ME,
Myogenic in Quail
Regulatory Genes Embryos. Deu Biol
In S~IU hybridization showed a sequential
EMERSON
CJ: Sequential
POWNAIL
Activation of Three during Somite Morphogenesis 1992. 151:67-79.
shows that three quail pattern of expression
myogenic factors (Qmfl-3) during early development.
Muscle Qmfl was detected in the medial portion of the newly formed somite. consistent with expression in axial muscle precursors (see [ le.1 ). 11.
FUJISAWA-SEHA~~A A. NARESHIMA Y, KOM~YA T, U~tsuto T, A~AKURA A, NA~ESHIMA Y: Differential Trans-activation of Muscle-specific Regulatory Elements including the Mysosin Light Chain Box by Chicken MyoD, Myogenin, and MRF4. / Biol &em 1992, 267:10031-10038,
12. .
FRANK D, GARLAND RM: Transient Expression of X&fyoD in Non-somitic Mesoderm of Xenopus gashtfae. Detpeelopmenr 1991, 113:1387-1393. Expression of XAIyuD was detected by whole-mount in situ hybridization throughout the mesoderm of the early Xenopus gustrulu embtyo. Transcripts were also detected by northern blot analysis in ultra violet light~treated embryos that produce no muscle tissue and in cultured explants that form ventral mesodenn. HOPWG~D ND, PLLICK A, GUFUIION JB, DILWOHTH SM: Expression of XMyoD Protein in Early Xenopus fuevis Embryos. Det~elopmeni 1992, 114:31-38. A detailed study of XMyoD protein distribution in the Xenoplrsembryo using a monoclonal antibody. This gave better resolution than comparable studies of the mRNA and confirmed that XMyoD protein was present in all but the most dorsal cells of the newly formed mesoderm. Expression was localized to the cell nuclei and was later restricted to premyotomal and myotomal cells. The protein was apparently present in every cell of the myotome. 13. .
14.
HOPWO~D ND, PLUCK A, CLIRD~N JB: Early Muscle Cells and can Activate tally in Early Embryos. DetJelopmenl
Xenopus Myf-5 Marks Muscle Genes Ectopi1991, 111:551-60.
15.
JENNINGS CG: Expression of the Myogenic Gene MRF4 ing Xenopus Development. Dell Biol 1992, 151:31’+332.
dur-
16. .
CUSEUA-DE ANGEL& MG, LYONS G, SONNINO C, DE ANGELES AL, V~~ARELU E, FARMER K, Wktctrr WE, MOUNARO M, BOLICHE M, BUCK~NGHAM M, Cossu G: MyoD. Myogenin Independent Dilferentiation of Primordial Myoblasts in Mouse Somites. J Cell Biol 1992, 116:1243-1255. The presence of MyoD, myogenin and myosin heavy-chain proteins was compared in cultured foetal mouse myoblasts. Late somites and limb buds gave rise to cells that expressed the myogenic factors before differentiation; a second population from early somites differentiated in culture without expressing either myogenic factor. These might constitute the migratory limb muscle precursors because they were also de. tected in cultures from limb buds. No antibody was available to identify Myf-5 protein. 17.
BRAUN T, BOBER E, ARNOU) HH: lnhibition of Muscle Differentiation by the Adenovirus Ela Protein: Repression of the Transcriptional Activating Function of the HLH Protein Myf-5. Genes Del! 1992, 6888-902. Expression of adenovirus ElA protein in rat Lb muscle cells blocks the activity of Mfl5 without affecting expression of the m.lfl5gene or DNAbinding ability of the Myf-5 protein. In contrast, myogenin expression is blocked, suggesting that this factor lies downstream of m~f5 in the regulatoty hierarchy. 19.
YUTZEY KE, RHODES SJ, KONIECZNV SF: Differential Transactivation Associated with the Muscle Regulatory Factors MyoDl, Myogenin, and MRF4. I%!o/ Cell Biol 1990, 10:3934-3944.
20.
CHAKRAB~~~TY T, BRENNAN T, ORION E: Differential Transactivation of a Muscle-specific Enhancer by Myogenic Helix-Loop-Helix Proteins is Separable from DNA Biding. J Biol Chem 1991, 266~28782882.
Mohun
HLH Gene muf-5 and Results in Apparently Normal Muscle Development. Cell 1992, 71383390. Homologous recombination in ES cells was used to construct transgenie mice lacking a functional myoD gene. These mice show no apparent abnormalities in muscle dilTerentiation, suggesting that the gene is functionally redundant. my!5 RNA is elevated in the muscles of the mice, indicating that the Myf-5 protein might be substituting for MyoD. BRAUN T, RUDN~CKI MA ARNOU) H-H, JAENISCH R Targeted Inactivation of the Muscle Regulatory Gene myf5 Results in Abnormal Rib Development and Perinatal Death. Cell 1992, 71369-382. rnJf5 is the first myogenic gene to be activated during somitogenesis in the mouse. Transgenic embryos lacking a functional myf5gene show a considerable delay in the first appearance of myotomal cells, but subsequent muscle differentiation is apparently normal. interestingly. the newborn mice die at birth because they lack a normal rib cage and cannot breathe. This suggests that the mu/-‘5 mutation affects normal differentiation of sclerotomal cells. 23. ..
24.
ORDAHI. CP: Developmental Expression. Curr Top Dev
Regulation Biol 1992,
of Sarcomeric 26:145-168.
WEINTRALIB H, Dw~ttto VJ, VE~~MA I, DAMS R, HO~LENEZERG S, SNIDER L, VASSAR A, TAFXOTT SJ: Muscle-specific Transcriptional Activation by MyoD. Genes Deu 1991, 5:1377-1386. The transcriptional-activation domain of MyoD was mapped to the amino-tenninal portion of the protein as assessed by domain swap experiments. Systematic mutagenesis of the basic region identified an Ala-Thr pair of residues to be critical for MyoD function. The authors suggest that another factor that recognizes the basic region is required for MyoD function. 26. .
SCHWARZ JJ, C HAKaABOR’tY T, MARTtN J, EHOU JM, O&ON EN: The Basic Region of Myogenin Cooperates with Two Transcription Activation Domains to induce Muscle-specific Transcription. Mol Cell Biol 1992, 12:266-275. Domainswap experiments demonstrated that the transcriptional activation functions of myogenin map to both amino- and carboxyl-terminal domains of the protein. Activation of reporter genes via E-box sites requires the myogenin basic region, even with chimeric VPlb-myogenin proteins, indicating that the basic region also participates in transctiptional activation Function. 27.
WINTER B, BRAUN T, ARNOU) HH: Domains in the Muscle-specific EMBO J 1992, 11:1843-1855.
Co-operativity Transcription
of Functional Factor myf-5.
28.
ANIH~~-CAHI~ SJ, BENFLELD PA, FA~RMAN R, WASSERMAN ZR, BRENNER SL STAFFORD Wl, ALTENBACH C, HUBBEU Wl, DEGRA~XI WF: Molecular Characterization of Helix-Loop-Helix Peptides. Science 1992, 255979-983.
29.
LJN H, KONIECZNY SF: Identification of MRF4, Myogenin, El2 Oligomer Complexes by Chemical Cross-linking Two-dimensional Gel Electrophoresis. J Biol f&m 26734779780.
and and 1992,
LSAR AB, DAMPS RL, Wtuctrr WE, KADE~cH T, Mutw C, VORONOVA A, BALTIMORE D, WEW~RAUB H: Functional Activity of Myogenic HLH Proteins Requires Hetero-oligomerizacion with E12/E47Rke Proteins in ufuo. Cell 1991, 66305-315. Comprehensive demonstration that E12/E47 proteins associate in vim with MyoD and myogenin. Myogenic conversion of lOT1/2 cells by MOOD results in activation of the muscle-specific regulators, MCAT and MEF-2. The authors discuss the possibility that competition for E-pro. teins provides a mechanism for ensuring that different HLH-dependent developmental decisions are mutually exclusive. 30. ..
31.
DAVIS RL, CHENC PF, LG.%R AB, DNA Binding Domain Contains Muscle-specific Gene Activation.
WEIN~RAUB H: The MyoD a Recognition Code for Cell 1990, 60:733-746.
21.
BLOCK NE, MILLER JB: Expression of MRF4, a Myogenic Helix-Loop-Helix Protein, Produces Multiple Changes in the Myogenic Program of BC3H-1 Cells. MO/ Cell Biol 1992, 12:2484-2492.
32.
CHAKRA~ORIY T, BRENNAN TJ, Lt L, EDMONDSON Inefficient Homooligomerization Contributes dence of Myogenin on E2A Products for Biding. MO/ Cell Biol 1991, 11:3633-3641.
22. ..
RUDNICKI of myoD
33.
BRENNAN TJ, CHAKRAB~~~~Y T, Olson EN: Mutagenesis Myogenin Basic Region Identifies an Ancient Protein
MA, BRAUN B, HINUMA S, JAENISCH R: inactivation in Mice Leads to Up-regulation of the Myogenic
Gene
25. .
HOPWOOD ND, GURD~N JB: Activation of Muscle Genes without Myogenesis by Ectopic Expression of MyoD in Frog Embryo Cells. Nulure 1990. 347:197-200.
18. .
differentiation
D. ORSON EN: to the DepenEfficient DNA of the Motif
927
928
Cell differentiation Critical
USA
for
Activation
of
Myogenesis.
Proc’hfarl
Acad
Sci
1991, &X5675-5679.
DAVIS RL, WE~UB H: Acquisition of Myogenic Specificity by Replacement of Three Amino Acid Residues from MyoD into E12. Science 1992, 256:1027-1030. An El2 protein containing only two replacements (Ala-Thr) in the basic region and a third replacement (Lys) at the junction with helix 1 acquired the ability to convert lOT1/2 fibroblasts into muscle cells.
34.
.
35.
FLOIUNI JR, EW~ON tors and Myogenjc 53:201-216.
36.
OLFON EN, BRENNAN 71, CHAKRAL~OR~ T, CHENC TC. CSERJESI P, EDMOND~ON D, JAMES G, LI L Molecular Control of Myogenesis: Antagonism Between Growth and Differentiation. MoI Cell BicxJxm 1991, 104:7-13.
37.
ANGEL P. KAtuN M: The Role of Jun. Fos and plex in Cell-proliferation and Transformation. /I& AcIa 1991, 1072:12%157.
DZ, MAGIU KA: Differentiation.
Hormones, Growth Annu Ret, P/+0/
the
AP-1
Com-
Bio
SLI HY, BOS TJ, MONTECV\RO FS, VOGT PK: Jun lnbibits genie Differentiation. Oncogene 1991, 6:175+1766.
39. ..
BENGAL E, RAN~oNE L SCHARFMANN WEIN~RAUB H, VERMA IM: Functional Jun and MyoD Proteins: a Direct
1992, 6B:507-519.
40.
Fac1991.
&i&in?
38.
Transactivation of the MyoD promoter, the MCK enhancer and an E. box driven reporter by MyoD were all blocked by cotransfection with a jun expression vector; in reciprocal experiments, MyoD suppressed nansactivation of a reporter containing an AP.1 site by Jun. Direct interaction of the Jun and MyoD was demonstrated both in vitro and in ctfro, requiring the leucine zipper domain of Jun and the HLH region of MyoD.
Myo-
R. DWARKI VJ, TAPSCOIT SJ. Antagonism Between tPhysical Association. Cell
GROW M, CALCONI A, TA-~o F: vjun Oncogene Prevents Terminal Differentiation and Suppresses Muscle-specific Gene Expression in ASVDl7D-infected Muscle Cells. Oncogene 1991, 6:1767-1773.
-11. ..
LI L. C~LAMBARD JC. bRlN M. OLSON EN: Fos and Jun Repress Transcriptional Activation by Myogenin and MyoD: the Amino Terminus of Jun Can Mediate Repression. Genes Del, 1992, 6:67&689. Espression of c.Fos. c.Jun and JunB all suppressed transactivdtion of the MCK enhancer by MyoD or myogenin. Repression activity was targeted at the basic HLH region of the myogenin and mapped to the amino.terminal region of c-Jun. Interestingly, the authors found that the leucine zipper dimeriiJtion domain of cJun was not required for repression (compare with (39**] ).
rJ Mohun. Iabordtov of Developmental tute for Medical Research, The Ridgeway. LIK.
Biochemistry. National Insti. Mill Hill, London NW7 IAA.
