Segmentation and specification of the Drosophila...

Segmentation and specification of the Drosophila mesoderm Natal ia Azpiazu , 1,3 Peter A. Lawrence, 2 Jean-Paul Vincent , 2 and Manfred Frasch 1'4

1Brookdale Center for Molecular Biology, Mount Sinai School of Medicine, New York, New York 10029 USA; 2Medical Research Council Laboratory of Molecular Biology, Cambridge CB2 2QH, UK

Patterning of the developing mesoderm establishes primordia of the visceral, somatic, and cardiac tissues at defined anteroposterior and dorsoventral positions in each segment. Here we examine the mechanisms that locate and determine these primordia. We focus on the regulation of two mesodermal genes: bagpipe (hap), which defines the anlagen of the visceral musculature of the midgut, and serpent (srp), which marks the anlagen of the fat body. These two genes are activated in specific groups of mesodermal cells in the anterior portions of each parasegment. Other genes mark the anlagen of the cardiac and somatic mesoderm and these are expressed mainly in cells derived from posterior portions of each parasegment. Thus the parasegments appear to be subdivided, at least with respect to these genes, a subdivision that depends on pair-rule genes such as even-skipped (eve). We show with genetic mosaics that eve acts autonomously within the mesoderm. We also show that hedgehog (hh) and wingless (wg) mediate pair-rule gene functions in the mesoderm, probably partly by acting within the mesoderm and partly by inductive signaling from the ectoderm, hh is required for the normal activation of hap and srp in anterior portions of each parasegment, whereas wg is required to suppress bap and srp expression in posterior portions. Hence, hh and wg play opposing roles in mesoderm segmentation.

[Key Words: Mesoderm; segmentation; visceral musculature; fat body; heart; even-skipped; wingless; hedgehog]

Received September 4, 1996; revised version accepted November 1, 1996.

How does the Drosophila mesoderm become subdivided into different anlagen? The mesoderm is partitioned along the dorsoventral axis: dorsally, cells found the primordia of the cardiac and visceral mesoderm and this subdivision is triggered by inductive inputs from the ectoderm. Dpp is a TGF~ protein that is secreted from dorsal ectodermal cells, and has been identified as the inductive signal (Staehling-Hampton et al. 1994; Frasch 1995). Dpp signaling is required for the formation of the anlagen of the heart and the midgut visceral mesoderm, and the dorsoventral extent of the visceral mesoderm is directly determined by the dorsoventral limits of dpp expression in the ectoderm (Frasch 1995; Maggert et al. 1995). An early response to the Dpp signal in mesodermal cells is the spatial restriction of expression of the homeo-box gene tinman (tin) to dorsal portions of the mesoderm, tin provides the dorsal mesoderm with the competence to form midgut visceral mesoderm, dorsal muscles, and the heart (Azpiazu and Frasch 1993; Bod- mer 1993). In combination with additional regulators that may include Dpp, tin activates the expression of a second homeo-box gene, bagpipe (bap), in dorsal meso-

3Present address: Centro de Biologia Molecular "Severo Ochoa," Facul- tad de Ciencias, Universita Autonoma de Madrid, Madrid 28034, Spain. 4Corresponding author.

dermal areas (Azpiazu and Frasch 1993; Staehling-Hamp- ton et al. 1994). bap is involved in the determination of the anlagen of the midgut visceral mesoderm. Analogous events could activate tin target genes that would specify the anlagen of the heart and dorsal body wall musculature.

We now study patterning of the mesoderm along the anteroposterior axis of the embryo. The morphology of the early mesoderm and the spatial domains of expression of homeotic genes indicate that the mesoderm is organized into parasegmental units (for review, see Lawrence 1992). Previous attempts to identify distinct subdivisions within each mesodermal parasegment that might be homologous to the anterior and posterior compartments of the ectoderm were negative (Lawrence and Johnston 1984). However, more recently, the analysis of expression patterns of genes that control mesoderm development has suggested that mesodermal parasegments are indeed subdivided into anterior and posterior portions with different developmental fates. A clear example is the expression of the homeo-box gene bap which is restricted to metameric clusters of cells in the dorsal mesoderm (Azpiazu and Frasch 1993). Under the control of bap, these cell clusters develop into midgut visceral mesoderm, whereas cells in segmental portions lacking bap expression form other mesodermal derivatives. The expression of several marker genes in the anlagen of the

GENES & DEVELOPMENT 10:3183-3194 �9 1996 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/96 $5.00 3183

Cold Spring Harbor Laboratory Press on April 21, 2018 - Published by genesdev.cshlp.orgDownloaded from

http://genesdev.cshlp.org/

http://www.cshlpress.com

Azpiazu et al.

fat body and studies on the regulatory gene twist also indicate that cells at different anteroposterior positions within each mesodermal parasegment give rise to different mesodermal derivatives (Abel et al. 1993; Hoshizaki et al. 1994; Dunin-Borkowski et al. 1995; Baylies and Bate 1996).

By analogy to dorsoventral patterning, inductive signals from the segmented ectoderm could help segment the mesoderm. Alternatively, the mesoderm could segment itself autonomously. For example, wingless (wg) acts in heart and muscle development and can function either from within the ectoderm or from the mesoderm (Baylies et al. 1995; Lawrence et al. 1995; Wu et al. 1995; Ranganayakulu et al. 1996). However, there must be additional segmental regulators, because both in the absence of wg and in the presence of uniformly distributed Wg protein the mesoderm is still segmentally patterned.

Our experiments show that the mesoderm forms segments autonomously, although induction from the ectoderm also contributes. We show that wg and hedgehog (hh), which are candidates for inductive signals from the ectoderm, play opposite and antagonistic roles in the subdivision of mesodermal parasegments, wg is mainly involved in the formation of derivatives from posterior portions of each parasegment, including somatic musculature and the heart, whereas hh functions mainly in the formation of derivatives in anterior parasegmental portions, in particular the visceral mesoderm of the midgut musculature and the fat body.

R e s u l t s

Segmental organization of mesodermal primordia

The expression patterns of several genes reveal that the early mesoderm is segmented. Notably, the homeo-box gene bap is expressed in 11 metameric clusters of dorsal mesodermal cells that develop into visceral musculature of the midgut (Azpiazu and Frasch 1993). The anterior borders of each of the bap patches coincide with the parasegmental borders of the ectoderm--thus the primordia of the midgut visceral mesoderm are largely po- sitioned below the posterior compartments of the ectoderm (Fig. 1 A; Azpiazu and Frasch 1993). The homeo-box gene even-skipped (eve) marks heart progenitors that will form pericardial cells (Frasch et al. 1987). Double stainings for bap and eve demonstrate that these heart progenitors form between the bap patches at the dorsal crest of the mesoderm (Fig. 1B). In addition, the progenitors of the cardioblasts are formed in close juxtaposi- tion, just anterior to the pericardial progenitors (K. Jagla, pets. comm.). Thus, it appears that the anlagen of the cardiac mesoderm alternate with the anlagen of the midgut visceral mesoderm. There are also primordia of the fat body in each segment; these are marked by the expression of the serpent (srp) gene that is required for fat body formation (srp encodes the GATA transcription fac- tor ABF; Abel et al. 1993; Rehorn et al. 1996). Double stainings for srp and engrailed (en; Fig. 1C) or srp and eve

(Fig. 1D) show that these cells are at the same anteroposterior positions and lie just ventrolateral to the primordia of the midgut visceral mesoderm.

In summary, the mesoderm of stage 10-11 embryos appears to be organized into parasegmental repeats that are in exact register with the ectodermal parasegments. Each parasegment is subdivided into two domains along the anteroposterior axis. In keeping with the nomencla- ture used for the ectoderm (Lawrence 1992), we term the mesodermal domains below the anterior compartments "A domains" and those below the posterior compartments "P domains." As shown schematically in Figure 1E, the P domains express bap and srp and give rise to visceral mesoderm derivatives that include midgut musculature and the fat body. The A domains include the primordia of the cardiac mesoderm and the bulk of the somatic mesoderm and give rise to the heart and most of the body wall muscles.

A mesoderm-autonomous role of pair rule genes in segmentation and specification of the visceral mesoderm

The close alignment between mesodermal and ectodermal parasegments suggests that the two germ layers might be segmented by the same genes. Indeed, mutations in pair rule genes have similar effects in both the ectoderm and the mesoderm. For example, fushi tarazu (ftz) mutant embryos retain only six out of normally 13 en stripes between parasegments 2 and 14 in wild-type embryos, and the number of bap patches is similarly reduced from normally 11 to 5 (Fig. 2A; note that parasegments 13 and 14 lack bap expression even in wild- type embryos; see Fig. 1A). srp patches are also reduced in ftz mutant embryos (Fig. 2B; wild-type expression only occurs between parasegments 4 and 12; see Fig. 1C). In spite of their aberrant patterns, the normal spatial relationships between the bap or srp domains and the en stripes are maintained in ftz mutants, indicating that ftz mutations cause identical alterations in the mesoderm and ectoderm. Concomitant alterations of bap and en stripes are observed in mutants for all other pair-rule genes tested, including hairy (h; Fig. 2C), runt (run; Fig. 2D), paired (prd; Fig. 2E), odd-paired (opa; Fig. 2F), and odd-skipped (odd; data not shown). Pair-rule mutations that cause mirror-image arrangements in the ectoderm, such as runt (Nhsslein-Volhard and Wieschaus 1980), result in mirror-image arrangements of the bap patches in the mesoderm (Fig. 2D).

The importance of pair-rule genes to segmentation of the mesoderm is particularly obvious in embryos that are mutant for eve. In embryos lacking eve function, the ectoderm is unsegmented and no en stripes are formed in the trunk region (Nhsslein-Volhard et al. 1985; Harding et al. 1986; Macdonald et al. 1986). Similarly, bap and srp patches are completely lost in the trunk region of eve mutant embryos, indicating that the mesoderm is unsegmented as well (Fig. 2G, H). As a consequence, neither midgut visceral mesoderm nor fat body visceral meso-

3184 GENES & DEVELOPMENT




Drosophila mesoderm segmentation

Figure 1. Intrasegmental and dorsoventral arrangement of mesodermal anlagen. All panels show wild-type embryos with anterior to the left and dorsal sides up. White boxes serve as a reference to indicate the location of the bap domains in parasegments 5. CA) Stage 10 embryo stained for bap mRNA (purple) and En protein (brown). The anterior margins of bap expression coincide with the ectodermal parasegmental borders, but are broader than the posterior compartments marked by en. (B) Early stage 11 embryo stained for bap mRNA (purple) and Eve protein (brown). The heart (pericardial} progenitors marked by eve arise at positions between the bap patches at the dorsal margin of the mesoderm. (C) Stage ll embryo stained for srp mRNA (purple) and En protein (brown). The srp-stained fat body primordia are found at similar intrasegmental, but more ventrally located, positions as compared with the bap-stained visceral muscle primordia. (D) Stage 11 embryo stained for srp mRNA and Eve protein to compare the relative positions of heart and fat body primordia. Bottom right: Schematic summary of the spatial arrangement of mesodermal anlagen (top of panel: dorsal margin; bottom of panel: ventral midline). The primordia of the visceral mesoderm {vm) are located below the posterior compartments of the ectoderm (marked by en in yellow) and extend posteriorly into the first half of each adjacent anterior compart- ment. The primordia of the midgut visceral mesoderm (mgvm; purple) are located in dorsal portions and those of the fat body visceral mesoderm (fbvm; blue-green) immediately adjacent in lateral portions of the mesoderm. The primordia of the cardiac mesoderm are located between the vm primordia at the dorsal margin (cm: cardiac mesoderm; solid red indicates Eve-stained pericardial progenitors, stippled red indicates the presumed location of cardial progenitors). Most of the rest of the mesoderm will give rise to somatic mesoderm (sm; grey). (A,P) Mesodermal A and P domains; (PS) parasegment; (pc) pericardial cell progenitors; (CNS) neuronal progenitors of the central nervous system.

derm is formed in the absence of eve function (Fig. 2I,K, cf. wild-type embryos in Fig. 2J and L}.

It appears that segmentat ion of the mesoderm is al- ready set up by the pair-rule genes during gastrulation, that is, at least 2 hr prior to the t ime when the dorsal bap patches appear. Upon overstaining of embryos hybrid- ized with bap antisense RNA probes, weak stripes of bap expression can be observed throughout the dorsoventral extent of the mesoderm as early as stage 8. Initially these stripes display a pair-rule distribution, but at stage 9, a regular pattern of weak transverse stripes is detected (Fig. 3A). This shows that mesoderm segmentat ion is initiated even before the the mesoderm spreads under the ectoderm and therefore that these early segmentation events occur wi thout any influences from the ectoderm. It has been proposed that the segmental expression of twist is required for visceral mesoderm specification (Baylies and Bate 1996). However, double stainings of stage 10 wild-type embryos for Twist protein and bap m R N A show that twist cannot determine the segmental bap pattern, because Twist protein is still evenly distributed along the anteroposterior axis at the t ime when the bap patches are fully developed (Fig. 3B).

Prior to and during gastrulation, pair-rule genes are

expressed in stripes in both ectoderm and mesoderm. Therefore, they could regulate bap expression and mesoderm segmentat ion through either of two possible routes. The first could be direct and wi th in the mesoder- real cells proper. The second route could be via induction from the overlying ectoderm. To find out which route is used, we made embryos that were genetically mosaic for eve; eve was chosen because it is essential for bap expression and development of the midgut visceral mesoderm. Nuclei from wild-type donor embryos marked with a lacZ reporter gene (Vincent et al. 1994) were transplanted into eve embryos that were stained at stage 10 for bap m R N A and f~-galactosidase. Figure 4A demonstrates that bap expression is rescued near to wild-type cells in an otherwise eve mutan t embryo. The restriction of the bap domains to the area of wild-type cells indicates that signals emanat ing from these cells, if any, are unable to act at a long distance to activate bap in adjacent eve- areas. We then prepared sections of the mosaic embryos. Figure 4B shows an embryo with two separate wild-type patches of cells, one restricted to the ectoderm (left side) and another extending into the mesoderm (right side). Importantly, bap expression is only observed in the mesodermal cells that are wild-type for

GENES & DEVELOPMENT 3185




Azpiazu et al.

Figure 2. Regulation of mesoderm segmentation and specification by pair-rule genes, bap and srp mRNAs are stained purple; En, fasciclin III, and [3-galactosidase proteins are stained brown. Em- bryos shown in A - H are in stage 10. {A)ftz mutant embryo stained for hap and En. (B) ftz mutant embryo stained for srp and En. (C) hairy mutant stained for bap and En. {D) runt embryo stained for bap and En. The wedges demarcate examples for segments arranged in a mirror image fashion. (E) paired mutant stained for bap and En. Although the number of fully developed bap patches is reduced from normally 11 to 6, weak bap signals are detected in intervening areas that lack detectable en stripes. (F) opa mutant stained for bap and En. Both the en and bap stripes are reduced in number and irregularly spaced. {G) bap expression in a stage 10 eve ~ mutant embryo. No bap expression in the midgut visceral mesoderm primordia is detected. (H) Stage 11 eve ~ embryo stained for srp (purple) and En (brown). srp is not expressed in the trunk region. (I) eve ~ mutant, and {J) wild-type embryo at early stage 12, both stained for fasciclin III. No midgut visceral mesoderm is formed in eve mutants. (K) eve ~ and {L) wild-type embryo at stage 13, both containing an enhancer trap insertion expressing ~-galactosidase in the fat body. The fat body is missing in the eve mutant embryo. (mgvm) Midgut visceral mesoderm; (fb) fat body.

eve, but not in eve mutan t cells that underly and contact wild-type cells of the dorsal ectoderm (Fig. 4B). The fail- ure of wild-type ectoderm to induce bap in the underlying eve mutan t mesoderm is not attributable to the small size of the wild-type patch in this particular embryo, because embryos with more extensive areas of wild-type cells in the ectoderm also fail to activate bap (Fig. 4C). These results demonstrate that at least one function of eve is required within the mesoderm itself for activating bap and for mesoderm segmentation.

Opposing roles of hh and wg in es tabl ishing m e s o d e r m a l A and P d o m a i n s

Because the pair-rule gene products disappear prior to the stage when bap and srp are fully activated, their function

in regulating these two genes mus t be indirect and involve intermediates. Here we examine whether segment polarity genes, which mediate pair-rule gene functions in the ectoderm, are also required for the segmentat ion of the mesoderm. The spatial overlap between the stripes of en or hh expression and the patches of bap or srp suggests that these two segment polarity genes are candidates (Figs. 1A, C and 5A; note that segment polarity gene expression is predominant ly ectodermal at this stage).

Both hh and en are indeed required for full activation of bap. In hh mutan t embryos, the bap domains are sig- nificantly reduced in size (Fig. 5B). There is also some reduction in embryos deficient for en (Fig. 5C). In Df(en);hh double mutants , there is an even stronger reduction of bap expression than in mutan t s for either hh






Figure 3. Dynamics of bap expression in wild-type embryos. (A) Wild-type embryo at stage 9, overstained with a bap RNA probe. Weak transverse stripes are detected transiently even in ventral portions of the mesoderm. (This ventral expression ceases at stage 10 when dorsal mesodermal expression reaches high levels.) (B) Stage 10 wild-type embryo, stained for bap and Twist. Twist protein is still evenly distributed at the time when the mature bap pattern appears.

or en alone (Fig. 5D), therefore the function of en in al- lowing full bap activation is not mediated soley through hh. The lack of either hh or en function alone has only minor consequences for visceral mesoderm formation (data not shown), whereas the s imultaneous loss of function of both genes results in a clear reduction of visceral mesoderm derivatives. As shown in Figure 5E, the normal ly continuous band of the midgut visceral mesoderm (see Fig. 2J) is interrupted in Df(en);hh double mutants, and resembles that of mutan t embryos with strongly reduced bap function (Azpiazu and Frasch 1993). The num- bers and sizes of cell clusters of the fat body visceral mesoderm are both reduced in the double mutants (Fig. 5E).

The requirement for hh and en for normal bap activation could partially explain the absence of bap expression in eve mutan t embryos, because these embryos lack hh and en expression in their t runk region. To test whether hh or en could rescue bap expression in the absence of eve, we ectopically expressed each of these two genes in the mesoderm of eve mutant embryos (see Materials and Methods). bap is indeed activated upon mesodermal expression of hh in eve mutant embryos, although the number of rescued bap patches is less than in the wild-type (Fig. 5F). The striped expression of bap in these embryos indicates that the mesoderm of eve mutan t embryos retains some periodic properties, which could reflect the contributions of other pair-rule genes to mesoderm segmentation. There is also a partial rescue of midgut visceral mesoderm formation. Ubiquitous expression of en in the mesoderm of eve mutant embryos activates bap expression as well, albeit less efficiently than hh (data not shown). Taken together, these results

Figure 4. Mesoderm-autonomous requirement for eve function to activate bap. Shown are mosaic embryos where host tissues are eve ~ Donor-derived tissues are wild-type and carry an arm-lacZ reporter gene insertion. Embryos are stained for bap mRNA {purplel and ~-galactosidase (brown). (A) bap expression is strictly confined to a wild-type patch of cells {brown) in the otherwise eve mutant embryo. (B,C) Cross-sections of stained mosaic embryos, bap expression is seen only seen when wild-type patches encompass dorsal mesoderm (B, right), but not when eve function is provided in the dorsal ectoderm only (B, lefL, and C).

suggest that hh and en participate in the es tabl ishment of the mesodermal P domains that express both bap and srp and form visceral mesoderm derivatives.

bap is expressed in the mesodermal P domains, whereas the secreted protein Wg is synthesized adjacent in the A domains and, more prominently, in the A compartments of the ectoderm (Fig. 6A). The alternating stripes of expression abut at the parasegmental borders, but small areas posterior to each bap patch lack detectable levels of Wg protein. Wg appears to act negatively on bap, because bap expression is expanded in wg mutant embryos (this interaction may not be direct, however; see Discussion). In posterior regions of these embryos, a continuous rather than periodic pattern of bap expression is observed in the dorsal mesoderm (Fig. 6B). More anteriorly, some periodic bap expression is still ob-





Azpiazu et al.

Figure 5. Positive regulation of bap by hh and en. Embryo in E is at stage 12, all others shown are at stage 10. mRNA detection appears purple and protein detection appears brown. CA) Wild-type embryo stained for bap mRNA and Hh protein. The bap and hh stripes are largely congruent. (B) hh mutant stained for bap mRNA and Wg protein. The bap patches are reduced in size. (C) Df(en) embryo stained for bap mRNA (and En protein). A minor reduction of bap expression is observed. (D) Df(en);hh double mutant embryo, stained as in C. bap expression is strongly reduced. (E) Df(en);hh double mutant embryo, stained for fasciclin III protein and srp mRNA. The visceral mesoderm of the midgut musculature and fatbody are substantially reduced, iF) eve ~ mutant embryo with ectopic mesodermal hh expression (using a twist-GAL4 driver) and stained for bap mRNA and Hh protein. Partial rescue of bap expression is observed.

served. Mutations in genes downstream of wg such as armadillo (arm; Fig. 6C) and dishevelled (dsh; Fig. 6D) have a similar phenotype. The expression of srp is also continuous in wg mutant embryos (Fig. 6E).

The subdivision of the mesoderm into A and P domains involves antagonistic interactions between hh and wg

Because the results described above indicate a positive function of hh and a negative function of wg in bap regulation, we tested whether ectopic expression of hh or wg would affect bap expression in opposite directions. Ectopic expression of hh in the Krfippel domain of the ectoderm (Fig. 7A; Frasch 1995), in the whole ectoderm (data not shown), or in the whole mesoderm (Fig. 7B) of wild-type embryos results in a minor expansion of the bap domains only. By contrast, ectopic hh expression in either the ectoderm (Fig. 7C) or mesoderm (Fig. 7D) of wg mutant embryos causes massive overexpression of bap, which comes to resemble the expression pattern of tin at this stage (Bodmer et al. 1990; Azpiazu and Frasch 1993). If hh is ectopically expressed together with dpp in the ectoderm of wg mutant embryos, bap expression expands also ventrally and is observed in the whole mesoderm between parasegments 2 and 12 (Fig. 7F; cf. ectopic expression of dpp in wild-type embryos, Fig. 7E). This clearly illustrates the combinatorial roles of anteroposterior and dorsoventral cues in mesoderm patterning.

Ectopic expression of wg has opposite effects to those observed for ectopic hh. Whereas wg expression either in the ectoderm (within the Krfippel domain, Fig. 7G, or in the whole ectoderm, data not shown) or in the mesoderm of wild-type embryos (Fig. 7H) results only in a slight and partially penetrant reduction of bap expression, similar ectopic expression of wg in hh mutant embryos almost

entirely blocks bap expression in the midgut visceral mesoderm primordia (Fig. 7I, J). These results indicate that overexpression of hh in wg- , and of wg in h h - , produce unsegmented mesoderm. The patterns of bap expression suggest that the mesoderm has opposite identities in these two experimental situations. Ectopic hh in wg- appears to result in a uniform P character, whereas the lack of bap expression upon ectopic wg expression in h h - signifies a uniform A character of the resulting mesoderm. Ectopic hh and wg can produce these alterations only if the other gene is not functional, but are unable to do so in wild-type embryos. Thus, we conclude that wg and hh normally antagonize each other's function. This antagonism is likely to be important for the proper subdivision of the mesoderm into A and P domains.

Analysis of wg- embryos with ectopic hh, and of h h - embryos with ectopic wg, confirmed that the mesodermal fate maps are shifted in opposite directions in these two situations. For example, the fat body in wg-/ec topic hh embryos is expanded as compared with wild-type and wg- embryos (Fig. 8A-D). The midgut visceral mesoderm appears also to be enlarged (Fig. 8C, D) at the expense of heart precursors and somatic musculature (data not shown). The latter phenotypes are also observed in wg- embryos (Baylies et al. 1995; Wu et al. 1995), and it is difficult to judge whether the reduction of somatic muscles is enhanced upon ectopic hh expression. In wg mutant embryos with ectopic ectodermal expression of both hh and dpp, the visceral midgut mesoderm is even more expanded. It occupies the whole mesoderm between parasegment 2 and 12, and no other mesodermal derivatives are formed in this region (Fig. 8E).

In stark contrast to wg-/ec topic hh embryos, hh-/ectopic wg embryos show a massive reduction of visceral mesodermal tissues (Fig. 8G). Ectopic wg expression in wild-type embryos causes a minor reduction of visceral






Figure 6. Negative regulation of bap and srp by the wg pathway. (A) Wild-type embryo stained for bap mRNA and Wg protein. The bap patches abut the wg stripes at their posterior margins but do not appear to fill the entire wg interstripe regions. (B) wg mutant embryo stained for bap mRNA and En protein, bap expression expands and becomes continuous, particularly in the posterior segments. (C) Embryo lacking both the maternal and zygotic contributions of arm, stained as in B. (D) Embryo lacking both the maternal and zygotic contributions of dsh, stained as in B. bap expression in arm and dsh mutants is altered in a similar fashion as in wg-. (E) wg mutant embryo stained for srp mRNA expression, which becomes almost continuous (cf. Fig. 1C).

mesoderm formation only, but the number of heart progenitors is clearly increased (Fig. 8F; Lawrence et al. 1995). The reduction of visceral mesoderm in hh mutants wi th ectopic wg is accompanied by an increased

number of cells with somatic mesodermal cell fates, as shown by the larger number of MyoD expressing cells. Whereas, in stage 13 wild-type embryos, there are segmental clusters of MyoD-expressing cells in the somatic mesoderm (Fig. 8H; Michelson et al. 1990; Paterson et al. 1991 ), hh- /ec top ic wg embryos have uniform MyoD expression in the dorsal mesoderm, and the more ventral clusters appear expanded (Fig. 8I). Taken together, these results show that expansion of the mesodermal P domains at the expense of the A domains, as exemplified by wg /ectopic hh embryos, causes the mesoderm to form mostly visceral organs. By contrast, fate map shifts in the opposite direction in hh- /ec top ic wg embryos result in the exclusive formation of somatic muscles and heart.

D i s c u s s i o n

Anteroposterior subdivisions of mesodermal parasegments and the mesodermal fate map

Our analysis of mesodermal gene expression patterns shows that the early mesoderm is organized into parasegments that are in exact register wi th the ectodermal parasegments. The existence of stable parasegmental mesoderm boundaries is supported by our observation that the stripes of eve-lacZ reporter genes perdure in the mesoderm with sharp anterior borders that coincide with the anterior borders of the bap patches (M. Frasch, unpubl.). Each mesodermal parasegment is subdivided along the anteroposterior axis into a P and an A domain that express different control genes, have different developmental fates, and respond differently to mutat ions in segmentation genes. The P domains comprise the expression domains of bap plus those of srp, genes that mark the anlagen of the transverse midgut muscles and of the fat body, respectively. The A domains contain the anlagen of the heart and the bulk of the body wall muscles. This picture conforms wi th the morphological analysis of Dunin-Borkowski et al. (1995), except we observe that the anlagen of the fat body are located ventrally rather than dorsally to those of the midgut musculature. Our assignment is further supported by an analysis of bap reporter genes showing that all bap expressing cells from the dorsal mesoderm will form midgut visceral mesoderm (Z. Yin and M. Frasch, unpubl.). We are not sure of the fate of those mesodermal cells located between the fat body anlagen and the ventral mid l ine in the P domains of each parasegment (Fig. 1). Although it is possible that some of them give rise to mesodermal glial cells (Gorczyka et al. 1994), others probably generate specific body wall muscles. Thus, al though most of the somatic mesoderm is derived from the A domains, some comes from the P domains and therefore could be subject to different control. This could explain the presence of some residual somatic mesodermal cells and muscles in embryos with genetic backgrounds that appear to produce a uniform "P" character in the mesoderm.

The subdivision of the mesodermal parasegments into P and A domains is reminiscent of the subdivision of ectodermal parasegments into posterior and anterior compartments. At present, we do not know whether





Azpiazu et al.

Figure 7. Effects of ectopic hh and wg expression on bap expression. (A) Wild-type embryo with ectopic ectodermal hh expression (ZKrGAL driver, see Frasch 1995) and stained for bap mRNA and Hh protein. (B) Wild-type embryo with ectopic mesodermal hh expression (twist-GAL4 driver) and stained as in A. In both situations, the bap patches are only slightly expanded. (C) wg mutant embryo with ectopic ectodermal hh expression (E22cGAL4 driver), stained as in A. bap expression is strongly expanded and becomes continuous along the anteroposterior axis. (D) wg mutant embryo with ectopic mesodermal hh expression (twist-GAL4 driver), stained for bap mRNA and En protein. The bap pattern is similar to that seen with ectodermal overexpression of hh. (E) Wild-type embryo with ectopic ectodermal expression of dpp (E22cGAL4 driver), stained for bap mRNA and Eve protein. The bap stripes extend from the dorsal margin (d) to the ventral midline (v). (F) wg mutant embryo with ectopic expression of hh and dpp (driver as in E), showing uniform bap expression between parasegments 2 and 12. (G) Wild- type embryo with ectopic ectodermal expression of wg (ZKrGAL driver as in Fig. 7A), stained for bap mRNA and Wg protein. A weak suppression of bap expression is observed, mostly in the patches #4 and #5. (H) Wild-type embryo with ectopic mesodermal expression of wg (twist-GAL4 driver), stained as in A. Some suppression of bap expression is observed, particularly in more anterior parts of the mesoderm. Note that the examples shown in E and F are among the most strongly affected embryos found, and usually the reduction of bap expression is less conspicuous. (I) hh mutant embryo with ectopic ectodermal expression of wg (E22cGAL4 driver as in C) and {1) hh mutant embryo with ectopic mesodermal expression of wg, stained as in E. In both situations, bap expression in the midgut visceral mesoderm primordia is almost completely abolished.

there are lineage restrictions in the mesoderm prior to or during the stage when this subdivision becomes visible through the expression of genes such as bap and srp. However, it is l ikely that any such restrictions would occur before the onset of the third wave of mitot ic divi- sions in the mesoderm. By that t ime the outer borders of the bap patches form sharp boundaries and all bap-expressing cells start segregating into the interior to form midgut visceral mesoderm (Azpiazu and Frasch 1993). Our observation that most body wall muscles are derived from the A domains may explain why it has not been possible to detect A and P compartments wi th in somatic musculature (Lawrence 1982). If there are lineage boundaries between A and P domains in the gastrulating mesoderm, they would be expected to demarcate visceral

(gut muscle and fat body) P compar tments from somatic / cardiac A compartments.

Regulation of mesoderm patterning

The segmentat ion of the mesoderm and the anteroposterior subdivision of the parasegments are controlled by gene systems that are similar, but not identical, to those that segment the ectoderm. In particular, the pair-rule genes and their upstream regulators have similar roles in both ectoderm and mesoderm and their muta t ions cause identical alterations of the number and polarity of segments in the two germ layers. However, these genes may not act through an identical set of downstream genes. Indeed, previous experiments wi th genetic mosaics for en suggested that this gene is not required for normal






Figure 8. Effects of ectopic hh and wg expression on mesoderm differentiation. (A) Stage 14 wild-type embryo and (B) wg mutant embryo with uniform ectopic hh expression in the ectoderm as in Fig. 7C. Staining for srp mRNA shows enlargment of the fat body in the mutant as compared with the wild-type situation. (C,D) Sec- tioned stage 12 embryos with genotypes as in A and B, stained for fasciclin III protein and srp mRNA. The embryo in D shows enlarged fat body and midgut visceral mesoderm. (E) wg mutant embryos with ectopic expression of hh and dpp as in Fig. 7F, stained for fasciclin III and Eve. All mesodermal cells between parasegments 2 and 12 are specified as midgut visceral mesoderm. (F) Wild-type embryo with ectopic ectodermal wg as in Fig. 7I, stained for fasciclin III (brown) plus Eve (purple).There are supernumerary pericardiat progenitors but visceral mesoderm formation is largely normal. (G) hh mutant embryo with ectopic ectodermal wg expression and stained as in F. Visceral mesoderm formation is largely abolished. (H) Stage 13 wild- type embryo, stained for MyoD (brown) and Eve (black). The muscle precursors expressing MyoD are arranged in segmental clusters. (I) hh mutant embryo with ectopic ectodermal wg, stage and staining as in H. In the dorsal mesoderm, MyoD is expressed in a continuous band of cells, and the more ventrally located MyoD clusters are enlarged. Eve staining of pericardial cells was used to identify embryos with ectopic wg expression.

mesoderm development (Lawrence and Johnston 1984). This is consistent wi th our finding that absence of en causes only minor disruptions of bap expression and has no obvious effects on visceral mesoderm formation. Thus, even though en is expressed in the early mesoderm (Lawrence et al. 1994), the pair-rule genes do not seem to act through it.

More important mediators are wg and hh which are expressed at opposite sides of the parasegmental borders. Whereas wg is involved in defining A (i.e., somatic and cardiac) identities of each parasegment, hh is involved in defining P (i.e., most ly visceral) identities. One function of wg is to prevent the expression of P genes in the A domains. Whereas bap is activated in the dorsal mesoderm by dpp and tin, wg appears to block this activation. Likewise, wg blocks the expression of srp ventrally to the tin domains. These are negative effects, but wg also acts positively to specify the cell fates of the heart progenitors and body wall muscle precursors; ectopic expression of wg can even generate supernumerary cells wi th these identities (Baylies et al. 1995; Lawrence et al. 1995; Wu et al. 1995; D'Alessio and Frasch 1996; Ran- ganayakulu et al. 1996; this paper). The two actions of wg could be related: Specifically, the expansion of vis-

ceral mesoderm primordia (P) in the absence of wg function could cause the loss of somatic and cardiac mesoderm (A). This could explain why wg is required long before these cell types are specified (Wu et al. 1995; Ran- ganayakulu et al. 1996). In addition, wg could continue to provide inputs for the development of muscle and heart precursors after mesoderm segmentat ion has been established. The lack of heart progenitors in hh mutan t embryos, which is a result of the decay of wg expression, could reflect this continued requirement for wg in heart specification (Park et al. 1996).

The experiments involving ectopic expression of wg and hh show that these two genes can antagonize each other during mesoderm segmentation. Uniform expression of either of them does not dramatical ly affect the periodic pattern of bap, provided that the other functions normally. Similar interactions between wg and hh occur during the segmentat ion of the ectoderm (Bejsovec and Wieschaus 1993; Heemskerk and DiNardo 1994; note that these negative, functional interactions are distinct from the positive interactions between wg and hh that serve to maintain the expression of both genes in the ectoderm after stage 9). Because wg and hh are expressed in adjacent cells and encode secreted molecules, this





Azpiazu et al.

functional antagonism could help ensure that wg function is limited to the A domains and hh function to the P domains of the mesoderm.

Mesoderm-autonomous vs. inductive regulation of mesoderm segmentation

The close contact between ectoderm and mesoderm after gastrulation could allow indirect regulation of mesoderm segmentation by pair-rule genes, through their role in the segmental subdivision of the ectoderm. Inductive signals from ectodermal cells could transmit this posi- tional information to the mesoderm and instruct mesodermal cells to assume either A or P properties. Such a mechanism would be analogous to the dorsoventral subdivision of the mesoderm, which is induced by dorsal ectodermal cells through the secreted signal molecule Dpp (Staehling-Hampton et al. 1994; Frasch 1995). Both wg and hh act in the mesoderm and their ectopic expression in the ectoderm can indeed alter the pattern of mesodermal bap expression and induce changes in visceral mesoderm development. Similarly, ectopic expression of wg in the ectoderm affects the development of somatic muscle precursors (Baylies et al. 1995; Ranganayakulu et al. 1996; this paper). However, embryos that are genetically mosaic for wg show that the function of wg in heart formation can be provided either from the ectoderm or from within the mesoderm (Lawrence et al. 1995). We believe that this function of wg is, at least in part, a consequence of its general role in determining the A domains of each parasegment. Therefore, it is likely that its function to promote somatic mesoderm development and to suppress visceral mesoderm in the A domains is also provided from both ectoderm and mesoderm. Note that the gradual decrease of wg and hh expression in the mesoderm during gastrulation suggests that the ectodermal contributions of both genes, which persist at high levels, are most significant for normal mesoderm development (Baker 1988; Lee et al. 1992; Mohler and Vani 1992; Tabata et al. 1992; Baylies et al. 1995).

Importantly, our results with eve mosaic embryos demonstrate that pair-rule genes act within the mesoderm to segment it. Thus, wg and hh signals provided by ectodermal patches of wild-type cells are not sufficient to mediate normal mesoderm segmentation and bap activation. Probably, the pair-rule genes establish first a prepattern of gene expression within the mesoderm and later inductive inputs from the ectoderm ensure that the mesodermal and ectodermal parasegments remain in exact register.

What are the candidates for genes that are regulated by the pair-rule genes in the mesoderm? We postulate that there exist as yet unidentified downstream genes in addition to hh and wg that are expressed in striped patterns in the early mesoderm. Such genes could be responsible for the residual segmental expression of bap in wg;hh double or wg, en;hh triple mutant embryos (M. Frasch, unpubl.) and the faint stripes of bap expression in the early wild-type mesoderm. A potential candidate is the paired-domain gene pox meso, which is expressed in seg-

mental stripes in the early mesoderm (Bopp et al. 1989). The expression of twist becomes striped, with high-protein levels present in the A domains, and this periodic modulation is required to allow visceral mesoderm development (Baylies and Bate 1996). However, twist cannot be involved in organizing the mesodermal A and P domains, because its periodic expression appears only after the segmental expression of bap has been established. It is more probable that twist could be controlled at a similar level of the regulatory hierarchy as bap and its disappearance in the A domains could be required for the function, rather than the regulation, of bap.

In conclusion, our current view of mesodermal segmentation and patterning can be summarized as follows (see Fig. 9): (1) During gastrulation, pair-rule genes control the expression of target genes in segmentally re- peated stripes in the mesoderm. This process is autonomous to the mesoderm and establishes the A and P domains. Mesodermal targets of the pair-rule genes include hh and en (in P domains) and wg (in A domains), en plays only a marginal role, and therefore we believe that there are as yet unidentified pair-rule target genes, M and N, that are important for the allocation of mesodermal cells to the A and P domains {Fig. 9, left). Gene M could be required to activate bap and srp in P domains, whereas N could activate target genes in A domains that are involved in somatic and cardiac mesoderm formation. Nei- ther M nor N is sufficient to activate their target genes; they both would later require localized coregulators that are not active in the early mesoderm.

(2) After gastrulation, the pair-rule gene products fade away and the maintenance of the striped expression of M and N requires hh and wg function. The mesodermally expressed hh and wg products can only maintain normal A and the P domains (i.e., normal M and N expression) transiently as they are not expressed beyond stage 8 or 9. In the ectoderm, hh and wg maintain each other's expression after this stage. Therefore, the contribution of hh and wg products secreted from the ectoderm is likely

Figure 9. Model of mesoderm segmentation and the specification of mesodermal tissues. We propose that mesoderm segmentation occurs in two phases; an early, mesoderm-autonomous phase (left) and a subsequent phase that involves mesoderm-autonomous as well as inductive regulation (right). During the second phase, developmental control genes are activated in quadrants at the intersections of transverse and lon- gitudinal domains of patterning genes (see text for details).






to become s igni f icant later. Funct iona l an tagon i sm between hh and wg cont r ibutes to the format ion of sharp borders be tween A and P domains (Fig. 9, right).

(3) After the m e s o d e r m has spread below the ectoderm, the cells in the dorsal por t ions of each t ransverse stripe receive Dpp signals from the dorsal ectoderm and therefore express h igh levels of tin. This combina t ion of transverse and long i tud ina l ac t iv i t ies (gene M plus T i n / D p p 1 is suff icient to ac t ivate h igh levels of bap in a dorsal quadrant of each P domain. Similarly, different combina t ions could act ivate deve lopmcnta l control genes in other quadrants and thereby define other anlagen.

(4) bap specifies the midgu t visceral mcsoderm, whereas genes such as srp de te rmine the fat body visceral mesoderm. Both types of visceral mesoderm segre- gate into the in ter ior and their deve lopment becomes independen t from the ectoderm. However, the mesodcr- real cells of the A domains remain in contact w i t h the ec toderm and wg could therefore con t inue to inf luence the deve lopment and pa t t e rn ing of the somat ic and cardiac mesoderm.

Mater ia l s and m e t h o d s

Drosophila mutant strains

The following segmentation mutants were used in this study: eve L27, ftz 9H34, h 7A94, prd 2.45, odd 1.36.1~, o[,Ta Ilp32, rlln LBF',

armHS 6; Df(2R)en c Ideleting both en and invected (inv); Gustav- son et al. 19961, Df(2R)gsb and Cl I)ga-mlA (both of which had normal hap expression, data not shown), dsh 477, hh l'<:>), and Wc~ cx4. Mosaic germlines for arm and dsh were generated by using the yeast recombinase-based FLP-DTS system (Chou and Perrimon 1992). Recombination was induced in femalcs homozygous for an FRT insertion on the X chromosome (FRT mr) and heterozygous for the dominant female sterile mutation ovo TM, as well as the mutation to be studied. The source of FLP recombinase was an autosomal insertion of FLP recombinase under the control of the heat shock promoter (FLp~S). The FLP/ FRT stocks were provided by E. Siegfried (Pennsylvania State University, State College). Homozygous pair-rule mutant embryos were identified through altered e , patterns, hh mutant embryos through the use of a TM3,lacZ balancer or their decreased wg expression, and wg mutants through their decreased en or hh expression.

UAS/GAL4 strains

Generation of UASwg: wg cDNA c14 (provided by N. Baker, Albert Einstein College of Medicine, Bronx, NY) was cloned as a BamHI-XbaI fragment into BglII-Xbal of pUAST (Brand and Perrimon 1993). An insertion of this P element on the third chromosome (UASwqg2) was used throughout the experiments.

Generation of UAShh: hh cDNA 11 (provided by Phil Beachy, Johns Hopkins University, Baltimore, MD) was made by first cloning a ftindIII-EcoRI insert from hhl l -pNB40 into pBlue- script KS+ and then a KpnI-Xbal insert was recloned into pUAST. The line UAShhl with an insertion on the third chromosome was used in all experiments described.

For ectopic expression in the mesoderm, we used the strains GALSG24 (twist-GAL4 insertion on the second chromosome) or GALSG30 (twist-GALd + 2dB enhancer trap insertion on the third chromosome)(Brand and Perrimon 1993; Greig and Akam 1993), both provided to us by A. Michelson (Brigham and Women's Hospital, Boston, MA). Ectopic expression in the ec-

toderm was achieved with the driver lines ZKrGAL8 (in the Kr0ppel domain; Frasch 1995} or E22c-GAL4 (made by K. Yoffe, Harvard Medical School, Boston, MA). The expression pattern of E22c-GAL4 was tested by crossing it with Bg4-i-2 (Brand and Perrimon 1993). Araldite sections of embryos stained for f3-galactosidase demonstrated that the expression from this driver is restricted to the embryonic ectoderm (M. Frasch, unpubl.). The line UASdpp-5 (third chromosome insertion) was used for dpp misexpression (Frasch 19951. Embryos from UAS- GAL4 crosses were collected at 27~ The enhancer trap line E7-3-63 (provided by V. Hartenstein, University of California, Los Angeles) was used to analyze fat body formation.

Nuclear transplantation and analysis of mosaic embryos

Nuclear transplantations were done as described previously (Lawrence et al. 1994; Vincent and Lawrence 1994), using do- nors marked with arm-lacZ. Hosts were an eve 127 mutant strain maintained over a CyO,hb-lacZ balancer. Mosaic embryos were processed for staining with ~-gal antibodies, fol- lowed by hybridization with digoxygenin-labeled bap cDNA probes as described in Azpiazu and Frasch (1993). Nine of the obtained mosaic embryos were analyzed in serial sections. Of these, three had mesodermally restricted wild-type cells showing hap expression, four had ectodermally restricted wild type cells and lacked bap expression, and in two embryos both ectoderm and the underlying mesoderm was wild-type for eve.

Antibody staining and in situ hybridization of embryos

The following antibodies were used for the phenotypic analysis: Monoclonal mouse-anti En (4D9, provided by C. Doe, Univer- sity of illinois, Urbana), rabbit-anti Eve (Frasch et al. 1987), rabbit-anti Hh (rabbit 1255, provided by J. Knight and T. Korn- berg, University of California, San Francisco), rabbit-anti Wg (provided by R. Nusse, Stanford University, CA), polyclonal mouse-anti 6-galactosidase (Sigma), monoclonal mouse-anti fasciclm III (7G 10, provided Developmental Studies Hybridoma Bank), rabbit-anti Twist (provided by S. Roth, Max Planck In- stitut for Entwicklungsbiologie, Tfibingen, Germany), rabbit- anti Dmyd (1184, provided by B. Paterson, National Institute of Health, Bethesda, MD}. Antibody double stainings and antibody-in situ hybridization double stainings were performed as described in Azpiazu and Frasch (1993).

A c k n o w l e d g m e n t s

We thank Zhizhang Yin for technical assistance and Patrick Lo for bap RNA in situ hybridizations. We thank Esther Siegfried, Andrea Brand, Tom Kornberg, Volker Hartenstein, Alan Mich- elson, Leslie Pick, Jym Mohler, and Ken Yoffe for fly strains, and Chris Doe, Jonathan Knight, Tom Kornberg, Roel Nusse, Bruce Paterson, Siegfried Roth, and the Developmental Studies Hy- bridoma Bank for antibodies. We appreciate the helpful com- ments of Hanh Nguyen on the manuscript. This work was supported by National Institutes of Health grant HD30832 and a Pew Scholarship in the Biomedical Sciences to M.F. and by a fellowship from the Basque Country to N.A.

The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 USC section 1734 solely to indicate this fact.

References

Abel, T., A. Michelson, and T. Maniatis. 1993. A Drosophila GATA family member that binds to Adh regulatory se- quences is expressed in the developing fat body. Develop- ment 119: 623-633.





Azpiazu et al.

Azpiazu, N. and M. Frasch. 1993. tinman and bagpipe: Two homeo box genes that determine cell fates in the dorsal mesoderm of Drosophila. Genes & Dev. 7" 1325-1340.

Baker, N.E. 1988. Localization of transcripts from the wingless gene in whole Drosophila embryos. Development 103: 289- 298.

Baylies, M., A. Martinez Arias, and M. Bate. 1995. wingless is required for the formation of a subset of muscle founder cells during Drosophila embryogenesis. Development 121- 3829- 3837.

Baylies, M.K. and M. Bate. 1996. twist: A myogenic switch in Drosophila. Science 272: 1481-1484.

Bejsovec, A. and E. Wieschaus. 1993. Segment polarity gene interactions modulate epidermal patterning in Drosophila embryos. Development 119:501-517.

Bodmer, R. 1993. The gene tinman is required for specification of the heart and visceral muscles in Drosophila. Develop- ment 118: 719-729.

Bodmer, R., L.Y. Jan, and Y.N. Jan. 1990. A new homeobox- containing gene, rash-2, is transiently expressed early during mesoderm formation of Drosophila. Development 110: 661- 669.

Bopp, D., E. Jamet, S. Baumgartner, M. Burri, and M. Noll. 1989. Isolation of two tissue-specific Drosophila paired box genes, Pox meso and Pox neuro. EMBO J�9 8: 3447-3457.

Brand, A.H. and N. Perrimon. 1993. Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118: 401-415.

Chou, T.-B. and N. Perrimon. 1992. Use of a yeast site specific recombinase to produce female germline chimeras in Droso- phila. Genetics 131: 643-653.

D'Alessio, M. and M. Frasch. 1996. rash may play a conserved role in dorsoventral patterning of the neuroectoderm and mesoderm. Mech. Dev. 58: 217-231.

Dunin-Borkowski, O., N. Brown, and M. Bate. 1995. Anterior- posterior subdivision and the diversification of the mesoderm in Drosophila. Developm en t 121" 4183-4193.

Frasch, M. 1995. Induction of visceral and cardiac mesoderm by ectodermal Dpp in the early Drosophila embryo. Nature 374: 464--467.

Frasch, M., T. Hoey, C. Rushlow, H.J. Doyle, and M. Levine. 1987. Characterization and localization of the even-skipped protein of Drosophila. EMBO J. 6: 749-759.

Gorczyka, M.G., R.W. Phyllis, and V. Budnik. 1994. The role of tinman, a mesodermal cell fate gene, in axon pathfinding during the development of the transverse nerve in Droso- phila. Development 120: 2143-2152.

Greig, S. and M. Akam. 1993. Homeotic genes autonomously specify one aspect of pattern in the Drosophila mesoderm. Nature 362: 630-632.

Gustavson, E., A.S. Goldsborough, Z. Ali, and T.B. Komberg. 1996. The Drosophila engrailed and invected genes: Part- ners in regulation, expression and function. Genetics 142: 893-906.

Harding, K., C. Rushlow, H.J. Doyle, T. Hoey, and M. Levine. 1986. Cross-regulatory interactions among pair-rule genes in Drosophila. Science 233: 953-959.

Heemskerk, J. and S. DiNardo. 1994. Drosophila hedgehog acts as a morphogen in cellular patterning. Cell 76: 449-460.

Hoshizaki, D.K., T. Blackburn, C. Price, M. Ghosh, K. Miles, M. Ragucci, and R. Sweis. 1994. Embryonic fat-cell lineage in Drosophila melanogaster. Development 120: 2489-2499.

Lawrence, P.A. 1982. Cell lineage of the thoracic muscles of Drosophila. Cell 29: 493-503.

�9 1992. The making of a fly: The genetics of animal de- sign. Blackwell Scientific Publications, Oxford.

Lawrence, P.A. and P. Johnston. 1984. On the role of the engrailed + gene in the internal organs of Drosophila. EMBO J. 3: 2839-2844.

Lawrence, P.A., P. Johnston, and J.-P. Vincent. 1994. Wingless can bring about a mesoderm-to-ectoderm induction in Drosophila. Development 120" 3355-3359.

Lawrence, P.A., R. Bodmer, and J.-P. Vincent. 1995. Segmental patterning of heart precursors in Drosophila. Development 121: 4303-4308.

Lee, J.J., D.P. yon Kessler, S. Parks, and P.A. Beachy. 1992. Se- cretion and localized transcription suggest a role in posi- tional signalling for products of the segmentation gene hedgehog. Cell 71: 33-50.

Macdonald, P.M., P. Ingham, and G. Struhl. 1986. Isolation, structure, and expression of even-skipped: A second pair- rule gene of Drosophila containing a homeo box. Cell 47: 721-734.

Maggert, K., M. Levine, and M. Frasch. 1995. The somatic-visceral subdivision of the embryonic mesoderm is initiated by dorsal gradient thresholds in Drosophila. Development 121: 2107-2116.

Michelson, A.M., S.M. Abmayr, M. Bate, A. Mart/nez-Arias, and T. Maniatis. 1990. Expression of a MyoD family member prefigures muscle pattern in Drosophila embryos. Genes & Dev. 4: 2086-2097.

Mohler, J. and K. Vani. 1992. Molecular organization and embryonic expression of the hedgehog gene involved in cell- cell communication in segmental patterning of Drosophila. Development 115:957-971.

Nfisslein-Volhard, C. and E. Wieschaus. 1980. Mutations affecting segment number and polarity in Drosophila. Nature 287: 795-801.

Nfisslein-Volhard, C., H. Kluding, and G. Jiirgens. 1985. Genes affecting the segmental subdivision of the Drosophila embryo. Cold Spring Harbor Symp. Quant. Biol. 50: 145-154.

Park, M., X. Wu, K. Golden, J.D. Axelrod, and R. Bodmer. 1996. The Wingless signaling pathway is directly involve in Droso- phila heart development. Dev. Biol. 177: 104--116.

Paterson, B.M., V. Walldorf, J. Eldrigde, A. Dfibendorfer, M. Frasch, and W. Gehring. 1991. The Drosophila homologue of vertebrate myogenic-determination genes encodes a transiently expressed nuclear protein marking primary myogenic cells. Proc. Natl. Acad. Sci. 88: 3782-3786.

Ranganayakulu, G., R.A. Schulz, and E.N. Olson. I996. wingless signaling induces nautilus expression in the ventral mesoderm of the Drosophila embryo. Dev. Biol. 176: 143-148.

Rehorn, K.-P., H. Thelen, A. Michelson, and R. Reuter. 1996. A molecular aspect of hematopoiesis and endoderm development common to vertebrates and Drosophila. Development 122:4023-4031.

Staehling-Hampton, K., F.M. Hoffmann, M.K. Baylies, E. Rush- ton, and M. Bate. 1994. dpp induces mesodermal gene expression in Drosophila. Nature 372: 783-786�9

Tabata, T., S. Eaton, and T.B. Kornberg. 1992. The Drosophila hedgehog gene is expressed specifically in posterior compart- ment cells and is a target of engrailed regulation. Genes & Dev. 6: 2635-2645.

Vincent, J.-P. and P.A. Lawrence. 1994. Drosophila wingless sustains engrailed expression only in adjoining cells: Evi- dence from mosaic embryos. Cell 77: 909-915.

Vincent, J.-P., C. Girdham, and P. O'Farrell. 1994. A cell-autonomous, ubiquitous marker for the analysis of Drosophila genetic mosaics. Dev. Biol. 164: 328-331.

Wu, X., K. Golden, and R. Bodmer. 1995. Heart development in Drosophila requires the segment polarity gene wingless. Dev. Biol. 169: 619-628.





10.1101/gad.10.24.3183Access the most recent version at doi: 10:1996, Genes Dev.

N Azpiazu, P A Lawrence, J P Vincent, et al. Segmentation and specification of the Drosophila mesoderm.

References

http://genesdev.cshlp.org/content/10/24/3183.full.html#ref-list-1

This article cites 41 articles, 23 of which can be accessed free at:

License

ServiceEmail Alerting

click here.right corner of the article or

Receive free email alerts when new articles cite this article - sign up in the box at the top

Copyright © Cold Spring Harbor Laboratory Press


http://genesdev.cshlp.org/lookup/doi/10.1101/gad.10.24.3183

http://genesdev.cshlp.org/content/10/24/3183.full.html#ref-list-1

http://genesdev.cshlp.org/cgi/alerts/ctalert?alertType=citedby&addAlert=cited_by&saveAlert=no&cited_by_criteria_resid=protocols;10.1101/gad.10.24.3183&return_type=article&return_url=http://genesdev.cshlp.org/content/10.1101/gad.10.24.3183.full.pdf

http://genesdev.cshlp.org/cgi/adclick/?ad=41173&adclick=true&url=https%3A%2F%2Fipaper.ipapercms.dk%2FEXIQON%2FMarketing%2Ftechnote%2FNGS-comparison-of-methods-for-microrna-profiling-from-plasma-and-plasma-derived-exosomes%3Futm_source%3DCSHL%26utm_medium



Date post:	05-Feb-2018
Category:	Documents
Upload:	hangoc
View:	224 times
Download:	1 times

Segmentation and specification of the Drosophila...

Documents