Cell, Vol. 93, 921±927, J une 12, 1998, Copyright 1998 by Cell Press
A View from Drosophila
signals such as TGF?, Wnt, and Hh family members
(Bate, 1993; Lawrence et al., 1995; Wu etal., 1995; Azpi-
azu et al., 1996; Bate and Baylies, 1996; Tajbakhsh and
Cossu, 1997). Myogenic lineages arise from these seg-
mental units, and genes coding for components of the
myogenic pathways have been remarkably conserved
in many instances. However, there appear to be differ-
ences in the way in which these elements are deployed
in flies and vertebrates. These differences may reflect
a real divergence inthe development ofthe musculature
in two widely separated groups of animals, or alterna-
tively, and perhaps more interestingly, they may repre-
sent variations in a common program that is, as yet,
Mary K. Baylies,*Michael Bate,²
and Mar Ruiz Gomez²
*Molecular Biology Program
Memorial Sloan-Kettering Cancer Center
Sloan Kettering Division
Graduate School of Medical Sciences
1275 York Avenue
New York, New York 10021
²Department of Zoology
University of Cambridge
Cambridge CB2 3EJ
The Origins of the Myogenic Lineage:
Dividing up the Mesoderm
The mesoderm in the fly is formed by the most ventral
cells of the blastoderm stage embryo, which are set
aside by the operation of the maternally acting axial
patterning gene dorsal (Chasen and Anderson, 1993).
As the embryogastrulates, these ventralcells invaginate
and migrate dorsally, coating the inner surface of the
ectoderm (Leptin and Grunewald, 1990). Several different
cell types subsequently arise from these cells, including
the muscles of the gut (visceral muscles), the body wall
muscles (somatic muscles), the heart and the fat body
(Bate, 1993). Thus, the allocation of cells to the somatic
muscle lineage is part of a more general process in
which mesodermal segments, like vertebrate somites,
are divideduptoformthe progenitors ofseveraldifferent
Initially, the internallayerof mesodermalcells is rather
It may be apocryphal but there is a widely held belief
among flyenthusiasts thatthe Drosophilalarva has more
muscles than a human being. Of course, it all depends
on how you count the muscles, but it is undeniable
that the Drosophila larva equips itself with a remarkably
complex muscle pattern (Bate, 1993; Abmayret al., 1995).
This pattern originates during a brief period of early
development when the embryonic fly first subdivides its
mesoderm into the progenitors of different internal
structures and, thereafter, in a remarkable process
whereby individual muscles are seeded and specified
from the cells of the myogenic lineage.Recent work has
fleshed out details of both these processes and reveals
similarities and differences in the development of the
muscles of flies and vertebrates. Like vertebrate em-
bryos, flies divide their mesoderm into segmental units
and the diversification of these units is influenced by
Figure 1. Homologous Signaling Pathways
Operate in Patterning the Drosophila Meso-
dermal Segment and the Vertebrate Somite
(A)Schematic representation of the early sig-
naling events involved in the patterning of a
Drosophilamesodermal segment.The unpat-
terned mesoderm (yellow) receives different
signals from the overlying ectoderm. Dorsal
ectodermal cells express Dpp, a member of
the TGF?family of secreted molecules (green).
Dorsal mesodermal cells that receive Dpp
and Hedgehog (Hh, light blue) from the ecto-
derm will form visceral mesoderm (vm) in the
even-skipped functionaldomain (eve),whereas
ventral cells are committed to generate part
of the fat body (fb) and other mesodermal
tissues. On the other hand, the most dorsal
mesodermal cells in the sloppy paired func-
tional domain (slp) that are under the influ-
ence of Wingless (Wg, graded purple) and
Dpp, will form heart (h). The remaining cells
inthe slpdomainwill formsomatic mesoderm
(sm) under the influence of Wg (dorsal and
ventral cells) and Dpp (dorsal cells).
(B) Scheme of similar interactions that occur
during the patterning of a mouse somite
(modified from Tajbakhsh and Cossu, 1997). The newly formed somites (S) are under the influence of signaling molecules (BMPs, Wnts, and
Shh) from the lateral mesoderm (LM), dorsal ectoderm (DE), neural tube (NT), and notocord (NC). The integration of these signals subdivides
the somite into dermis, the dorsal and ventral myotome, and sclerotome.
uniform. Forexample, allcells express the transcription
factor Twist (Thisse et al., 1988) and its immediate tar-
gets, Tinman (tin) (Bodmer et al., 1990; Yin et al., 1997)
and DMef2 (Lilly et al., 1994; Nguyen et al., 1994; Taylor
et al., 1995). Relatively quickly however, the invaginated
cells are subdivided into a series of units from which
the progenitors of different mesodermal tissues will be
formed (Dunin Borkowski et al., 1995; Azpiazu et al.,
1996; Riechmannetal., 1997).As inavertebrate embryo,
this subdivision in the fly mesoderm is accomplished in
part by signals that are received from adjacent cells, in
this case different regions of the overlying ectoderm
(Bate and Baylies, 1996; Azpiazu et al., 1996; Tajbakhsh
and Cossu,1997;Figure1).Aninductivesignaling mech-
anism of this kind was already indicated by the embryo-
logical experiments done in the 1930s by Bock, which
were influenced by the work of Spemann (Bock, 1941).
These were thenreinforced by evidence that inembryos
where gastrulation is arrested (Baker and Schubiger,
1995) or mesodermal migration inhibited (Gisselbrecht
et al., 1996; Beiman et al., 1996; Maggert et al., 1995),
the fate of the mesoderm that forms is related to the
region of the ectoderm with which it is associated.
More recent work shows that as the mesodermalcells
migrate dorsally, they come under the influence of the
TGF? family member, Decapentaplegic (Dpp) (Figure 1;
Staehling-Hampton et al., 1994; Frasch, 1995). At this
stage, Dpp is expressed in a dorsal band of ectodermal
cells and acts on underlying mesodermal cells to main-
tain the expression of tin and repress the expression of
ventrally expressed genes such as pox meso. Thus, the
inductive effect of Dpp signaling divides the mesoderm
into dorsal and ventral sectors. At the same time, the
mesoderm is partitioned along the anterior±posterior
axis of the embryo by the segmentation genes even-
skipped (eve) and sloppy paired (slp) (Figure 1; Azpiazu
et al., 1996; Riechmann et al., 1997). The requirement
for eve and slp divides each segment of the mesoderm
into two domains that give rise to different progenitor
populations. The eve domain includes the progenitors
of the visceral mesoderm and the fat body, while the
slp domainproduces the somatic muscles and theheart.
eve and slp are expressed in the ectoderm as well as
the early mesoderm, and their effects on mesodermal
patterning are direct (through activation and repression
of genes such as bagpipe and tin; Azpiazu and Frasch,
1993; Yin et al., 1997) and indirect through their activa-
tion in their respective ectodermal domains of down-
streamtargets suchas Engrailed andthe signalingmole-
cules Hedgehog (eve) and Wingless (slp) (Figure 1;
Riechmann et al., 1997).
Because this patterning process depends crucially on
a registration of mesodermal segmentation with ecto-
dermally derived signaling molecules, it requires an or-
derly process ofmesodermalmigrationforits execution.
In normal development, mesodermal cells migrate to
the most dorsal edge of the ectoderm while maintaining
theirrelative positions inthe anterior±posterioraxis.This
ordered movement of cells depends (like the migration
of presomitic mesoderm in vertebrate embryos; Yama-
guchi et al., 1994) on mesodermal expressionof an FGF
receptor, Heartless (htl) (Beiman et al., 1996; Gissel-
brecht et al., 1996). In htl mutant embryos, mesodermal
cells failto reach the dorsalmargin of the ectoderm and
their registration in the anterior±posterior axis breaks
down.The resultis awidespread disruptionofmesoder-
mal differentiation, with, in particular, a loss of those
dorsalmesodermal derivatives (heart, visceral, and dor-
sal somatic muscles) that require activation of the Dpp
signaling pathway for their induction.
It is within this framework that cells of the slp domain
are allocated to form somatic muscles and, as indicated
above, most require the Wingless (Wg) signal to do so
(Figure 1;Baylies et al.,1995).Wg acts to amplifydistinc-
tions between cells of the eve and slp domains and in
the case of the myogenic lineages does so, at least in
part, by maintaining high levels of Twist in the cells of
the slp domain (Bate and Rushton, 1993). Differences
in Twist expression between eve (low levels) and slp
(higherlevels) domains first become apparent aftergas-
trulation as the mesoderm is subdivided by pair-rule
gene expression. Recent work shows that these regu-
lated differences in Twist expression across the seg-
ment are a crucial element in assigning cells to somatic
myogenesis (Baylies and Bate, 1996). The evidence for
Twist's decisive role comes fromexperiments where the
level of Twist expression has been uniformly raised or
lowered across the mesodermas it develops.Forexam-
ple, ectopic expression of abnormally high levels of
Twist inhibits the development of derivatives such as
visceral mesoderm progenitors and promotes the for-
mation of somatic muscle in their place. Gain-of-func-
tion experiments such as these are matched by the
observation that when levels of Twist are reduced, so-
matic myogenesis is deranged (Baylies and Bate, 1996).
Interestingly, homologs of Twistinvertebrates may have
a similar role in partitioning the mesoderm, although
there is no evidence that they are involved in allocating
cells to the myogenic pathway. Indeed, in the mouse,
Twistis excluded from the forming myotome but contin-
ues to be expressed in other parts of the somite (Wolf
et al., 1991). Evidence from cell culture experiments
indicates that Twistcan repress myogenesis in this sys-
tem (Hebrok et al., 1994). Other vertebrate Twist family
members, such as Scleraxis (Cserjesi et al., 1995) and
Dermo-1 (Li et al., 1995), are also excluded from the
myotome butare expressed inother regions suchas the
sclerotome anddermis, respectively, and these proteins
are implicated in the developmentof these mesodermal
subtypes. Thus, ifa comparisonis to be made, the con-
clusion might be that Twist has a conserved function in
subdividing the mesoderm that involves the repression
of a subset of developmental pathways in cells in which
it is expressed. Whether the same mechanism is in-
volved in this repression in vertebrates and in flies re-
mains to be determined.
Whereas the allocation of mesodermal cells to the
somatic musclelineage in flies requires Twist, this same
process in vertebrates requires the activity of the myo-
genic regulatory factors, in particular, Myf-5 and MyoD.
We refer the reader to the many excellent reviews on
this topic (forexample, Tajbakhshand Cossu,1997; Yun
and Wold, 1996; Rawls and Olson, 1997); however, we
would like to highlight a few points forcomparison. Like
Twist in flies, the critical expression of Myf-5 and MyoD
requires both intrinsic information present in the meso-
derm (i.e., Pax3; Maroto et al., 1997; Tajbakhsh et al.,
Review: Myogenesis in Drosophila
Figure 2. Patterns of Gene Expression inSo-
Muscle diversity is manifested by differences
in patterns of gene expression. Muscle marker
genes canbe expressedinindividualmuscles
(even-skipped and ladybird) or in groups
(Kru Èppel and vestigial). These subpatterns of
expression can be partially overlapping. Some
muscle marker genes encode transcriptional
regulators (S59 and Kru Èppel) whereas others
are membrane-associated proteins that are
also present in a subset of motorneurons
(connectinand Toll). The schemeof the larval
abdominal muscles from an external (left
panel)and an internalview (right panel)repre-
sents the muscles that express each of the
marker genes. Dorsal top and anterior left.
1997)as wellas extrinsic signals emanating from adjoin-
ing tissues (i.e., Wnts and BMPs; Tajbakhsh and Cossu,
1997; Yun and Wold, 1996; Rawls and Olson, 1997). In
addition, embryos lacking both Myf-5 and MyoD do not
produce skeletal muscle (Rudnicki et al., 1993), much
like fly embryos with reduced levels of Twist (Baylies
and Bate, 1996). Also, just as Twist is able to convert
nonsomatic mesoderm and ectoderm into the somatic
muscle fate (Baylies and Bate, 1996), misexpression of
Myf-5 orMyoD in nonmyogenic cells can convert those
cells into skeletal muscle (Tajbakhsh and Cossu, 1997;
Yun and Wold, 1996; Rawls and Olson, 1997). However,
recent analyses suggest that initial activation of Myf-5
orMyoD inmyotomalcells may mark different subpopu-
lations of myoblasts, those giving rise to the epaxial
muscles orhypaxial muscles, respectively (see reviews
above). Interestingly, the MyoD homolog in flies, Nauti-
lus, is only expressed in a subset of muscle progenitors
cells and may contribute to the particularidentity of this
subset of somatic muscles (Abmayr and Keller, 1998,
and see below). It is likely that the way in which these
conserved HLH proteins operate in myogenesis has
been influenced by the rather different organizational
principles that underlie myogenesis in flies and verte-
brates that we describe below.
fibers) in the fly. In both cases muscles are distinct
units with identifiable characteristics, particularly their
attachments and their innervation. In Drosophila these
distinctive characteristics are the attribute of a single
myotube andthe myoblasts thatcontribute toit; inverte-
brates they are the shared property of the population
of myotubes that constitute a muscle.
The diversity of myotubes within the segment is re-
flected in localized patterns of gene expression in indi-
vidual muscles or subsets of muscles (Figure 2; Bate,
1993).These patterns ofexpressionareinadditionto the
generalized expression of the myogenic differentiation
gene DMef2, which is a target of Twist (Nguyen et al.,
1994; Bour et al., 1995; Lilly et al., 1995; Taylor et al.,
1995). DMef2 is required forthe completionof myogen-
esis in all muscles and its role in muscle differentiation
has been extensively reviewed (for example, Taylor,
1995). Here we concentrate on the less well understood
issue of muscle patterning and the recruitment of myo-
blasts to form particular muscles.
Specifying Individual Muscles
Since each muscle is a syncytium formed by the fusion
of a group of neighboring myoblasts, the general issue
is how such a group of myoblasts can be allocated
to develop the specific characteristics of an individual
muscle. A simple model would assume that there is
some mechanism forselecting myoblasts ingroups and
assigning them to fuse and form a particular kind of
fiber. While this is a straightforward explanation, there
is so far no evidence in favor of it. In fact, studies of
early patterns of gene expression in muscle-forming
mesodermreveal that single cells initiatethe expression
of genes characteristic of individual muscles or muscle
groups, and that neighboring myoblasts are then re-
cruited to these patterns of expression as they fuse with
the originally expressing cells. The firstexample of such
expression was the homeobox containing gene S59,
which is finally expressed in a very few muscles in each
of specific cells in muscle-forming mesoderm (Dohr-
mann et al., 1990). Many such genes have now been
documented (forexamples, see Figure2). LikeS59, their
expression begins in a few cells at predictable locations
in muscle-forming mesoderm, and there is increasing
From Myoblast to Muscle Fiber
An important difference between the larval muscles of
Drosophila and muscles of vertebrates that is likely to
have had profound consequences for the evolution of
the myogenic pathway is that, in Drosophila, each mus-
cle is a single, multinucleated fiber and is therefore the
developmental equivalent of a primary or secondary
myofiber in a vertebrate muscle. Although in the Dro-
sophilalarva allofthese fibers are thoughtto be identical
physiologically, each fiberis unique in terms of its posi-
tion, size, sites of attachment, and patterns of innerva-
tion (Figure 2; Bate, 1993; Bernstein et al., 1993). The
reason for this diversity of fibers is that the individual
myotube in Drosophila is also the functional unit with
which the neuromuscular system operates and hence
the mechanical equivalent of the bundle of fibers that
constitutes a muscle in a higher vertebrate. There are
many different muscles (bundles of fibers) in a verte-
brate, just as there are many different muscles (single
Figure 3. Successive Steps in the Formation of the Drosophila Muscle Pattern
Cartoon of stage 10 embryo showing the modulated pattern of Twist expression. Cells of the high Twist domain (medium blue) will give rise
to all somatic muscles.
(1) Expression of the proneural gene, lethal of scute (l'sc), in the high Twist domain marks groups of mesodermal cells (dark blue circles) that
have the potential to become muscle progenitors.
(2) A process oflateral inhibition,mediated bythe neurogenic genes, restricts l'sc expressionto single cells that become the muscle progenitors
(P). The rest of the cells in the myogenic clusters behave as fusion-competent myoblasts.
(3) Asymmetric division of the progenitors gives rise to pairs of founders or to a founder and an adult muscle precursor (AP). In either case,
the two sibling cells follow different fates (A and B or A and AP).
(4) Each founder will seed the formation of a specific muscle by fusion with fusion-competent myoblasts, which are entrained to patterns of
gene expression characteristic of the particular founder.
evidence (see below) that it is the selective expression
of such genes that endows each muscle in the pattern
with its distinctive properties.
These distinctive patterns of gene expression and the
sequence of muscle formation itself suggest an alterna-
tive model for the organization of muscle development
(Figure 3). In this view, the formation of each muscle is
initiated by the specification of a distinctive founder
myoblast in muscle-forming mesoderm. The founder
then fuses with neighboring fusion-competent myoblasts,
recruiting them to its pattern of gene expression and
forming the syncytial precursor of an individual muscle
in the pattern (Figure 3; Bate, 1990, 1993).
This hypothesis, among otherthings, puts agreat deal
of weight on the founders as special cells that have
privileged access to the information necessary to form
specific muscles.Incontrast,the fusion-competentcells
are seen as ªnaiveº myoblasts that cannotformmuscles
independently of the founder cells. A critical test of this
hypothesis is provided by embryos where myoblast fu-
sion fails. What is the behavior of founders and fusion-
competent cells under these conditions? The answer,
as it turns out, is very clear: in the absence of fusion,
the founders formatappropriate locations, express their
normalcomplementof genes, andthengoontodifferen-
tiate as tiny, mononucleate muscles (Rushton et al.,
1995). These miniature fibers are properly innervated
and contractile; inother words they appearto differenti-
ate perfectly normally. The fusion-competent cells on
the other hand, which are unable to fuse with the found-
ers, express muscle myosin but remain rounded and
otherwise undifferentiated; they develop none of the
specific characteristics of the muscles they would nor-
mally contribute to and eventually many of them degen-
erate (Rushton et al., 1995). It is important to note that
the only defect in such embryos is a block to myoblast
fusionÐ there is no other defect in the myogenic path-
way, which is apparently completed normally by the
founders.Thus, where myoblastfusionfails, the founder
myoblasts are revealed as a special class of cells that
uniquely have access to the information necessary (a)
to complete myogenesis and (b) to execute the specific
program of differentiation characteristic of the muscles
whose formation they seed.
The founder myoblasts are the products of distinct
lineages that are initiated by the segregationofa special
class of muscleprogenitorcellinmuscle-forming meso-
derm.Eachprogenitordivides to giveriseto two founder
myoblasts, or a founder and the precursor of an adult
muscle (Figure 3; Carmena et al., 1995; Ruiz Gomez
and Bate, 1997; and see below). Interestingly, identical
mechanisms are required for the selection of muscle
progenitors inmesodermand neuralprogenitors (neuro-
blasts)inneuralectoderm. Thus, progenitors are singled
out from groups of mesodermal cells that express the
proneural gene lethal of scute (l'sc), and this singling
out depends on a process of lateral inhibition within the
clusterthat is mediated by the Notch signaling pathway
(Figure 3; Carmena et al., 1995).
The differential expression of transcription factors
suchas S59insubsets of founders and muscles, carries
withit the strong implicationthat such transcriptionfac-
tors are responsible for the development of some or
all of the characteristics of individual muscles. Indeed,
experiments withtwo suchfactors, Apterous (Bourgouin
et al., 1992) and Nautilus (Keller et al., 1997), showed
that loss of expression or ectopic expression could
cause loss or partial duplication of the muscles that
would normally express them. Other experiments have
demonstrated that if patterns of transcription factor ex-
pression characteristic of one muscle are switched to
those characteristic of anotherduring myogenesis, then
the phenotype of the muscle concerned is correspond-
ingly changed (Ruiz Gomez et al., 1997). For example,
a progenitor cell that expresses Kru Èppel (Kr) generates
a pair of founder cells, one maintaining Kr expression
Review: Myogenesis in Drosophila
Figure 4. Schematic Representation of the
Effects of Lack and Excess of Function of
Numband Kron the Developmentof the Ven-
tral Acute 1±3 (VA1±3) Muscles and the Ven-
tral Adult Precursor (VaP)
The central panel shows the patterns of ex-
pression of Numb, S59, and Kr during the
development of VA1±3 and VaP and the mor-
phology of muscles VA1±3 in wild-type em-
bryos. The divisions of two ventral progeni-
tors that express S59 (red) produce the
founders formuscles VA1±3 and the VaP.The
most dorsal progenitor, which coexpresses
Kr (black; gray indicates low levels), divides
first to generate VA1 and VA2 founders. Asym-
metric distribution of Numb (green outline) in
the progenitors ensures that only one of the
sibling founder cells (VA2 and VA3) will re-
ceive Numb and ultimately leads to the two
cells following alternative fates. Lack of func-
tion of Numb (left panel, top) affects the de-
velopment of all the progenitors. Progenitor
divisions generate duplications of fates asso-
ciated with the absence of Numb: repression
of gene expression (in the case shown: VA1
and VaP). Ectopic expression of Numb (left panel, bottom) in the mesoderm produces the opposite effect: duplications of fates associated
to the presence of Numb (VA2 and VA3). Lack or excess of function of Kr is only manifested in those muscles where Kr is normally expressed.
Maintenance of Kr in VA2 founder confers to muscle VA2 its specific characteristics. In the absence of Kr (right panel, top), VA2 develops as
its sibling, VA1, where Kr expression is normally repressed. On the other hand, ectopic expression of Kr (right panel, bottom) affects the
development of VA1 such that now it is transformed to the sibling VA2 fate that normally maintains Kr expression.
(founder for muscle VA2) and the other not (founder for
muscleVA1; Figure 4).Themuscles that arise fromthese
founders (VA2 and VA1) are distinct from one another
as measured by size, shape, and attachmentto the epi-
dermis. If Kr is lost from both founders, two identical
muscles resembling VA1 are formed. However, if Kr is
maintained in both founders through ectopic expres-
sion, the opposite transformation now occurs:two iden-
tical muscles with the orientation and insertion sites of
VA2 are now formed (Figure 4; Ruiz Gomez et al., 1997).
These experiments not only show that local expres-
sion of factors such as Kr in the myogenic lineage can
regulate the diversification of muscles, but also provide
some insight into the way in which such factors may
interact with the general myogenic pathway. Forexam-
ple, these experiments show that loss of Kr leads to
loss of expression of other founder cell genes such as
S59, yet, together, these losses do not prevent muscle
differentiation (Figure 4; Ruiz Gomez et al., 1997). In-
stead, specific characteristics of the individual muscles
in which Kr is normally expressed are dramatically al-
tered. Thus, Kr acts in concert with the myogenic path-
way to define specific muscle (Ruiz Gomez et al., 1997).
What we need to know now is what aspects of muscle
differentiation are actually controlled by genes such as
these and whether the regulation of these properties is
completely separate from the general pathway of myo-
genesis, which, invertebrates, is considered tobe under
the control of the MyoD family of transcription factors.
The experiments with Kr emphasize a general feature
of the myogenic pathway in Drosophila, namely that
progenitor cells divide to give rise to sibling myoblasts
that initiate the formation of muscles with alternative
fates (Figure 3; Carmena et al., 1995). The different as-
signments of these sibling cells are reflected in the dif-
ferential expression in them of genes such as Kr and
S59 that are expressed in the parent progenitor. Thus,
Kr and S59 are maintained in one founder cell but not
the other, and this determines the alternative fates of
the two muscles concerned. The distinction between
sibling foundercells depends onthe differentialdistribu-
tionofthe cytoplasmic protein, Numb, thatis asymmetri-
cally segregated during the division of the progenitor
cell (Ruiz Gomez and Bate, 1997; Carmena et al., 1998).
Experiments involving loss of Numband ectopic expres-
sion of Numb reveal that the founder to which Numb
is segregated maintains progenitor gene expression
whereas the sibling founder, which lacks Numb, loses
expression ofthese genes (Figure 4).Thus, thereappear
to be two alternative founder cell states, A and B, in
which A represents maintenance of progenitorcell gene
expression and B represents loss. Allmuscles appearto
be formed as a result of such asymmetric cell divisions.
Interestingly, however, not all B fates lead to the forma-
tion of larval muscles. Specific B cells are allocated to
form the precursors of the adult muscles (Ruiz Gomez
and Bate, 1997). These cells, unlike their larval counter-
parts, maintainthe expressionofTwist, donot differenti-
ate, and proliferate during larval life to form pools of
myoblasts from which specific adult muscles will be
formed (Bate et al., 1991). Since Numb appears to act
by blocking the activation of Notch (Frise et al., 1996),
it is likely that the B fate (including the maintenance of
adult precursors ina proliferative,undifferentiatedstate)
requires Notch activation, and that it is the differential
activation of Notch in the two founder cells that is re-
sponsible for their characteristically different patterns
of gene expression.
A Model for Myogenesis in Drosophila
Theview ofmyogenesis thatemerges fromthesestudies
is that while components of a conserved network of
Bate, M. (1990). The embryonic development of larval muscles in
Drosophila. Development 110, 791±804.
Bate, M. (1993). The mesoderm and its derivatives. In The Develop-
ment of Drosophila melanogaster, Vol. 2, M. Bate and A. Martinez-
Arias, eds. (Cold Spring Harbor, NY: CSH Laboratory Press), pp.
Bate, M., and Baylies, M.K. (1996). Intrinsic and extrinsic determi-
nants of mesodermal differentiation in Drosophila. Semin. Cell Dev.
Biol. 7, 103±111.
Bate,M., and Rushton, E.(1993). Myogenesis and muscle patterning
in Drosophila. C. R. Acad. Sci. Paris 316, 1055±1061.
Bate, M., Rushton, E., and Currie, D.A. (1991). Cells with persistent
twist expression are the embryonic precursors of adult muscles in
Drosophila. Development 113, 79±89.
Baylies, M.K., and Bate, M. (1996). twist: a myogenic switch in Dro-
sophila. Science 272, 1481±1484.
Baylies, M.K., Martinez-Arias, A., and Bate, M. (1995). wingless is
requiredfor the formationof a subsetof muscle foundercells during
Drosophila embryogenesis. Development 121, 3829±3837.
Beiman, M., Shilo, B.Z., and Volk, T. (1996). Heartless, a Drosophila
FGF receptor homologue, is essential for cell migration and estab-
lishment of several mesodermal lineages. Genes Dev. 10, 2993±
Bernstein, S.I., O'Donnell, P.T., and Cripps, R.M. (1993). Molecular
genetic analysis of muscle development, structure and function in
Drosophila. Int. Rev. Cytol. 143, 63±144.
Bock, E. (1941). Wechselbeziehungen zwischen den Keimbla Èttern
bei der Organbildung von Chrysopa perla. L. Die Entwicklung des
Ektoderms inmesodermdefektenKeimteilen.Roux' Arch.Entw. Org.
Bodmer, R., J an, L.Y., and J an, Y.N. (1990). A new homeobox-con-
taining gene,msh-2, is transiently expressed earlyduring mesoderm
formation of Drosophila. Development 110, 661±669.
Bour, B.A., O'Brien, M.A., Lockwood, W.L.,Goldstein, E.S., Bodmer,
R., Taghert, P.H., Abmayr, S.M., and Nguyen, H.T. (1995).Drosophila
MEF2, a transcription factorthat is essentialfor myogenesis. Genes
Dev. 9, 730±741.
Bourgouin, C., Lundgren, S.E., and Thomas, J .B. (1992). apterous
is a Drosophila LIM domain gene required for the development of
a subset of embryonic muscles. Neuron 9, 549±561.
Carmena, A., Bate, M., and J ime Ânez, F. (1995). lethal of scute, a
proneural gene, participates in the specification of muscle progeni-
tors during Drosophila embryogenesis. Genes Dev. 9, 2373±2383.
Carmena, A., Murugasu-Oei, B., Menon, D., J ime Ânez, F., and Chia,
W. (1998). inscuteable and numb mediate asymmetric muscle pro-
genitorcelldivisions during Drosophila myogenesis. Genes Dev. 12,
Chasen, R., and Anderson, K.V. (1993). Maternal control of dorsal-
ventral polarity and pattern in the embryo. In The development of
Drosophila melanogaster, Vol. 1, M. Bate and A. Martinez-Arias,
eds. (Cold Spring Harbor, NY: CSH Laboratory Press), pp. 387±424.
Cserjesi, P., Brown, D., Ligon, K.L., Lyons, G.E., Copeland, N.G,
Gilbert, D.J ., J enkins, N.A., and Olson, E.N.(1995). Scleraxis:a basic
helix-loop-helix protein that prefigures skeletal formation during
mouse embryogenesis. Development 121, 1099±1110.
Dohrmann, C., Azpiazu, N., Frasch, M. (1990). A new Drosophila
homeo box gene is expressed in mesodermal precursor cells of
distinct muscles during embryogenesis. Genes Dev. 4, 2098±2111.
Dunin Borkowski, O.M., Brown, N.H., and Bate, M. (1995). Anterior-
posterior subdivision and the diversification of the mesoderm in
Drosophila. Development 121, 4183±4193.
Frasch, M. (1995). Induction of visceral and cardiac mesoderm by
ectodermal Dpp in the early Drosophila embryo. Nature 374,
Frise, E., Knoblich, J .A., Younger-Shepherd S., J an, L.Y., and J an,
Y.N. (1996). The Drosophila Numb protein inhibits signaling of the
Notch receptor during cell±cell interactionin sensory organ lineage.
Proc. Natl. Acad. Sci. USA 93, 11925±11932.
Gisselbrecht,S.,Skeath, J .B., Doe,C.Q., and Michelson,A.M.(1996).
genes establish a population of muscle-forming cells,
myogenesis itself cannot proceed without the segrega-
tion of specific subtypes of myoblasts (Figure 3). There
are severalelements to this modelofthe myogenic path-
way. (1) Founders and fusion-competent cells are dis-
tinct cell populations in muscle-forming mesoderm. It
is likely that the differences between them depend on
the activation of distinct genetic pathways as a result
of the lateral inhibition event that segregates muscle
progenitors fromsurrounding cells. However, the nature
of the differences between these cell types and the way
they are implemented are not understood. (2) In the
absence of founders, fusion-competent myoblasts do
not fuse with each other. Founder cells do not fuse
with each other, even when duplicated (e.g., in Numb
overexpressionexperiments).This suggests that thefor-
mation of two distinct types of myoblasts is a prerequi-
site for cell fusion. Whether this is a general feature of
myogenesis in all organisms is not clear. (3) Fusion-
competent myoblasts do not differentiate to form mus-
cle in the absence of founder cells, although founders
alone will form miniature muscles. This suggests that
founders uniquely express genes that are generally re-
quired for myogenesis as well as genes specifically re-
quired for the formation of particular muscles. Taken
together, these points leadto the importantgeneral con-
clusion that because founders are required both for fu-
sion and for completing myogenesis, they gate the pro-
cess of muscle formation. Thus, distinctive patterns of
gene expressionin these cells dictate unique properties
to the muscles whose formation they initiate. It remains
to be shown how these distinctive properties are inte-
grated with the general pathway of myogenesis.
This model provides us with the clearest framework
for thinking about the local control of myogenesis and
muscle patterning in any organism. Even so, our under-
standing is still very incomplete. We need a more com-
plete description of how muscle progenitor cells are
specified at particular locations in somatic mesoderm.
We need to know whetherthe requirement formyoblast
diversification is a general one or a peculiarity of the
Drosophila embryo. We also need to understand how
the function of specific transcription factors in founder
cells is integrated with the general pathway of muscle
differentiation. Most importantly we need to understand
at a molecularlevel how myoblast diversification is con-
trolled and implemented.
Abmayr, S., and Keller, C. (1998). Drosophila myogenesis and in-
sights into the role of nautilus. Curr. Top. Dev. Biol. 38, 35±80.
Abmayr, S., Erickson, M., and Bour, B. (1995). Embryonic develop-
ment of the larval body wall musculature of Drosophila melanogas-
ter. Trends Genet. 11, 153±159.
Azpiazu, N.,and Frasch, M. (1993).tinmanand bagpipe:two homeo-
box genes that determine cell fates in the dorsal mesoderm of Dro-
sophila. Genes Dev. 7, 1325±1340.
Azpiazu, N., Lawrence, P.A., Vincent, J .-P., and Frasch, M. (1996).
Segmentationand specificationofthe Drosophilamesoderm.Genes
Dev. 10, 3183±3194.
Baker, R., and Schubiger, G. (1995). Ectoderm induces muscle-
specific gene expression in Drosophila embryos. Development 121,
Review: Myogenesis in Drosophila Download full-text
heartless encodes a fibroblast growth factorreceptor (DFR1/DFGF-
R2) involved in the directional migration of early mesodermal cells
in the Drosophila embryo. Genes Dev. 10, 3003±3017.
Hebrok, M., Wertz, K., and Fuchtbauer, E.M. (1994). M-Twist is an
inhibitor of muscle differentiation. Dev. Biol. 165, 537±544.
Keller, C.A., Erickson, M.S.,and Abmayr, S.M. (1997).Misexpression
ofnautilus induces myogenesisincardioblasts andalters thepattern
of somatic muscles fibers. Dev. Biol. 181, 197±212.
Lawrence, P.A., Bodmer, R., and Vincent, J .-P. (1995). Segmental
patterning of heart precursors in Drosophila. Development 121,
Leptin, M., and Grunewald, B. (1990). Cell shape changes during
gastrulation in Drosophila. Development 110, 73±84.
Li, L., Cserjesi, P., and Olson, E.N. (1995). Dermo-1: a novel twist-
related bHLHprotein expressed in the developing dermis. Dev. Biol.
Lilly, B., Galewsky, S., Firulli, A.B., Schulz, R.A., and Olson, E.N.
(1994).D-MEF2:aMADS box transcriptionfactorexpressedindiffer-
entiating mesodermand muscle cell lineages during Drosophila em-
bryogenesis. 91, 5662±5666.
Lilly, B., Zhao, B., Ranganayakulu, G., Paterson, B.M., Schulz, R.A.,
and Olson, E. (1995). Requirements of MADS domain transcription
factor D-MEF2 for muscle formation in Drosophila. Science 267,
Maggert, K., Levine, M., and Frasch, M. (1995). The somatic-visceral
subdivisionof the embryonic mesoderm is initiated by dorsalgradi-
ent thresholds in Drosophila. Development 121, 2107±2116.
Maroto, M., Reshef, R., Munsterberg, A., Koester, S., Goulding, M.,
and Lassar, A. (1997). Ectopic Pax-3 activates MyoD and Myf-5
expression in embryonic mesoderm and neural tissue. Cell 89,
Nguyen, H.T., Bodmer, R., Abmayr, S.M., McDermott, J .C., Spoerel,
N.A., and Nadal-Ginard, B. (1994). D-mef2: A new Drosophila meso-
derm-specific MADS box-containing gene with a bi-modal expres-
sion profile during embryogenesis. Proc. Natl. Acad. Sci. USA 91,
Rawls, A., and Olson, E. (1997). MyoD meets its maker. Cell 89, 5±8.
Riechmann, V.,Irion, U., Wilson,R., Grosskortenhaus, R., and Leptin,
M. (1997). Control of cell fates and segmentation in the Drosophila
mesoderm. Development 124, 2915±2922.
Rudnicki, M.A., Schnegelsberg, P.N.J ., Stead, R.H., Braun, T., and
Arnold, H.H. (1993). MyoD or myf-5 is required for the formation of
skeletal muscle. Cell 75, 1351±1359.
Ruiz Gomez, M., and Bate, M. (1997). Segregation of myogenic
lineages inDrosophilarequiresnumb.Development124, 4857±4866.
Ruiz Gomez, M., Hartmann, C., Romani, S., J a Èckle, H., and Bate,
M. (1997). Specific muscle identities are regulated byKru Èppelduring
Drosophila embryogenesis. Development 124, 3407±3414.
Rushton, E., Drysdale, R., Abmayr,S.M., Michelson, A.M., and Bate,
M. (1995).Mutations ina novelgene,myoblastcity, provideevidence
in support of the founder cell hypothesis for Drosophila muscle
development. Development 121, 1979±1988.
Staehling-Hampton, K., Hoffmann, F.M., Baylies, M.K., Rushton, E.,
and Bate, M. (1994). dpp induces mesodermal gene expression in
Drosophila. Nature 372, 22±29.
Tajbakhsh, S., and Cossu, G. (1997). Establishing myogenic identity
during somitogenesis. Genes Dev. 7, 634±641.
Tajbakhsh, S., Rocancourt, D., Cossu, G., and Buckingham, M.
(1997). Redefining the genetic hierarchies controlling skeletal myo-
genesis: Pax-3 and Myf-5 act upstream of MyoD. Cell 89, 127±138.
Taylor, M.V. (1995). Making Drosophila muscle. Curr. Biol. 5,
Taylor, M.V., Beatty, K.E., Hunter, H.K., and Baylies, M.K. (1995).
Drosophila MEF2 is regulated by twist and is expressed in both the
primordia and differentiatedcells ofthe embryonic somatic, visceral
and heart musculature. Mech. Dev. 50, 29±41.
Thisse, B., Stoetzel, C., Gorostiza-Thisse, C., and Perrin-Schmitt, F.
(1988). Sequence of the twist gene and nuclear localization of its
protein in endomesodermal cells of early Drosophila embryos.
EMBO J . 7, 2175±2183.
Wolf, C., Thisse,C., Stoetzel, C.,Thisse, B., Gerlinger, P., and Perrin-
Schmitt, F. (1991). The M-Twist gene of Mus is expressed in subsets
of mesodermal cells and is closely related to the Xenopus X-Twi
and the Drosophila Twist genes. Dev. Biol. 143, 363±373.
Wu, X., Golden, K., and Bodmer, R. (1995). Heart development in
Drosophila requires the segment polarity gene wingless. Dev. Biol.
Yamaguchi, T.P.,Harpal, K.,Henkemeyer, M.,and Rossant,J .(1994).
fgfr-1 is required for embryonic growth and mesodermal patterning
during mouse gastrulation. Genes Dev. 8, 3032±3044.
Yin Z., Xu, X.-L., and Frasch, M. (1997). Regulation of the Twist
target gene tinman by modularcis-regulatory elements during early
mesoderm development. Development 124, 4971±4982.
Yun, K., and Wold, B. (1996). Skeletal muscle determination and
differentiation: story of a core regulatory network and its context.
Curr. Opin. Cell Biol. 8, 877±889.