4971Development 124, 4971-4982 (1997)Printed in Great Britain © The Company of Biologists Limited 1997DEV5169

Regulation of the Twist ta rget gene tinman by modular cis -regulatory

elements during ear ly mesoderm d evelopment

Zhizhang Yin, Xiao-Lei Xu and Manfred Fras ch*

Brookdale Center for Developmental and Molecular Biology, Mount Sinai School of Medicine, Box 1126, One Gustave L. LevyPlace, New York, NY 10029, USA*Author for correspondence (e-mail: frasch@msvax.mssm.edu)

The Drosophila tinman homeobox gene has a major role inearly mesoderm patterning and determines the formationof visceral mesoderm, heart progenitors, specific somaticmuscle precursors and glia-like mesodermal cells. Thesefunctions of tinman are reflected in its dynamic pattern ofexpression, which is characterized by initial widespreadexpression in the trunk mesoderm, then refinement to abroad dorsal mesodermal domain, and finally restrictedexpression in heart progenitors. Here we show that each ofthese phases of expression is driven by a discrete enhancerelement, the first being active in the early mesoderm, thesecond in the dorsal mesoderm and the third in car-dioblasts. We provide evidence that the early-activeenhancer element is a direct target of twist, a gene encodinga basic helix-loop-helix (bHLH) protein, which is necessaryfor tinman activation. This 180 bp enhancer includes threeE-box sequences which bind Twist protein in vitro and areessential for enhancer activity in vivo. Ectodermal mis-expression of twist causes ectopic activation of thisenhancer in ectodermal cells, indicating that twist is the

only mesoderm-specific activator of early tinmanexpression. We further show that the 180 bp enhancer alsoincludes negatively acting sequences. Binding of Even-skipped to these sequences appears to reduce twist-dependent activation in a periodic fashion, thus producinga striped tinman pattern in the early mesoderm. Inaddition, these sequences prevent activation of tinman bytwist in a defined portion of the head mesoderm that givesrise to hemocytes. We find that this repression requires thefunction of buttonhead, a head-patterning gene, and thatbuttonhead is necessary for normal activation of thehematopoietic differentiation gene serpent in the same area.Together, our results show that tinman is controlled by anarray of discrete enhancer elements that are activated suc-cessively by differential genetic inputs, as well as by closelylinked activator and repressor binding sites within anearly-acting enhancer, which restrict twist activity tospecific areas within the twist expression domain.

Key words: Drosophila, tinman, mesoderm, patterning

SUMMARY

in

al.,

yet

alnde

ch,

mhe

INTRODUCTION

The molecular mechanisms that determine patterning adifferentiation of the mesoderm are a major focus of curreresearch. In Drosophila, genetic analysis has shown that thetwist gene occupies a position at the top of a hierarchy of zygically active genes that function in mesoderm development.twist encodes a basic helix-loop-helix (bHLH) protein that expressed in the presumptive mesodermal cells on the ventralside of blastoderm embryos and, in the absence of twistfunction, no mesoderm is formed (Simpson, 1983; Thisseal., 1988; Kosman et al., 1991; Leptin, 1991). In addition to irole in mesoderm formation, twist has a second function instages after gastrulation, where it appears to be requiredmyogenesis of somatic muscles (Baylies and Bate, 199Because twist encodes a putative transcription factor, it isassumed to control mesoderm development through the acti-vation of a large battery of target genes, either in the wholemesoderm or in specific portions of it. Candidates include genes encoding the homeodomain proteins Tinman and Zfh-1

ndnt

ot-

is

etts

for6).

the

(Bodmer et al., 1990; Lai et al., 1991), the MADS-domaprotein MEF-2 (Lilly et al., 1994; Nguyen et al., 1994; Tayloret al., 1995), the FGF-receptor Heartless (Shishido et 1993), the integrin PS2 (Leptin, 1991), the KH-domain proteinStruthio (also named Who or How; Baehrecke, 1997; Zaffranet al., 1997; Lo and Frasch, 1997) and genes with as undefined functions (Casal and Leptin, 1996).

The tinman gene has a key role in early mesoderm pattern-ing and is essential for the formation of all dorsal mesodermderivatives, including the heart, visceral musculature adorsal somatic muscles (Azpiazu and Frasch, 1993; Bodmr,1993). In addition, tinman is required for the formation ofcertain body wall muscles and glia-like cells that are derivedfrom ventral areas of the mesoderm (Azpiazu and Fras1993; Gorczyka et al., 1994). These functions of tin arereflected in its dynamic mesodermal expression pattern, whichcan be subdivided into three major phases. The early, twist-dependent phase includes tin expression from late blastodermuntil after gastrulation in all cells of the trunk mesoder(Bodmer et al., 1990; Azpiazu and Frasch, 1993). During t

4972

e

1,2

eR

R-

nts

e

ir

ir

to.for

chd-reinsn

s

y-

d byn,

s

Z. Yin, X.-L. Xu and M. Frasch

second phase, upon internal spreading of the mesodermtinexpression is restricted to a broad band of cells in the domesoderm. This restricted tin expression domain is controlledby inductive signals mediated through Dpp, which is secreby dorsal ectodermal cells (Frasch, 1995). Most likely, it during this phase when tin fulfills its major role in the deter-mination of dorsal mesodermal derivatives. Finally, tinmanexpression is further restricted to heart precursors and mayinvolved in differentiation and diversification of cardioblastand pericardial cells during this phase.

In this study, we address the question of how mesodeautonomous regulation, involving activation by twist, andinductive inputs, which include those mediated by Dpp, aintegrated molecularly at the promoter level of the tinmangene. Our functional dissection of regulatory regions from ttin locus reveals that tin is controlled by distinct enhancermodules, one of which is responsible for its early and bromesodermal expression, another for its Dpp-mediated domesodermal activation, and a third for cardioblast-specexpression. We further show that the early-acting modulecomposed of positively and negatively acting subsequenthat functionally antagonize each other. We propose that dirbinding of Twist to its target sites within this element caactivate tin expression in the whole mesoderm, but Twiactivity is abrogated in defined mesodermal areas by binding of negative regulators to adjacent sites. As a result,tinis activated in the trunk mesoderm which gives rise to visceand somatic muscles, heart and glial-like cells, but not in arof the head mesoderm that give rise to hemocytes. Texclusion of tinman from the prospective blood mesoderm isat least in part, due to the activity of the head-patterning gebuttonhead. Moreover, negative inputs triggered by the bindinof the pair-rule gene product Eve result in periodicalmodulated levels of tin along the anteroposterior embryo axis

MATERIALS AND METHODS

Isolation of genomic clones and construction of P-transformation plasmidsThe tinmangene from Drosophila viriliswas isolated from a genomiclibrary obtained from W. Hanna-Rose through J.D. Licht (Hanna-Roet al., 1997). The screen was performed with a tin cDNA probe usinghybridization conditions as in McGinnis et al. (1984). Washes wedone three times 30 minutes at 50°C with 1× SSC, 0.5% SDS.

Upstream fragments of tin (D.mel.) were cloned into P-transfor-mation vectors with native orientations relative to the basal promote+1 bp of our map refers to the first base of the longest tin cDNAavailable (Bodmer et al., 1990). Following upstream constructs wmade and tested in vivo: tin-3436/tin1: A fragment from−1716 bp to+287 bp generated by PCR with BamHI-site containing primers andcloned into the BamHI site of pCaSpeR AUG βgal (Thummel et al.,1988). tin-3668/tin1 contains a fragment from−984 bp to +287 bpthat was generated by PCR and cloned as above. tin-Pvu/tin1 containsa fragment from the PvuII site at −384 bp to +287 bp in theEcoRI/BamHI sites of the same vector. tin-3436/3437 was generatedfrom a PCR fragment (−1716 bp to−39 bp), with primers containingBamHI (5′) and XhoI (3′) sites, that was cloned into XhoI/BamHI ofpCaSpeR hs43 βgal (Thummel et al., 1988). tin-3436/3843 wasgenerated similarly from a PCR fragment containing sequences fr−1716 bp to −269 bp. tin-1.7 kb+int was generated from tin-3436/3437 by inserting intron 1 into the NotI site.

Intron fragments from D. melanogaster(Dm) or D. virilis (Dv) were

, rsal

tedis

bes

rm-

re

he

adrsalific iscesectn

stthe raleashe,negly.

se

re

rs.

ere

om

cloned upstream of the basal hsp70 promoter of pCaSpeR hs43 βgal.The final constructs were: tin-A/lac (Dm), a 500 bp HindIII/XhoIfragment cloned into NotI/XhoI; tin-B-374 bp/lac (Dm), a 374 bpfragment cloned into NotI/XhoI, previously generated by PCR from apKS+ clone containing the 2.5 kb HindIII fragment of tin, using theM13 primer (5′) and ATGCGGCCGCCTGCGGGAAA (3′) and cutwith NotI and internally with XhoI; tin-B-180 bp/lac (Dm), a PCRfragment generated with the primer pair AGGAATTCGTCAACAT-GTGT (5′) and AGCTCGAGTCTCGTATATGG (3′) and cloned intoEcoRI/XhoI; tin-B (Dv), a 390 bp PCR fragment generated with thprimer pair AGGAATTCGACAAATCATCG (5′) and AGGGATC-CTCGATTAGTTGGC (3′) and cloned into EcoRI/BamHI. E1,2/lacand E3/lac were made from double stranded oligonucleotides. Ewas flanked by EcoRI (5′) and BamHI(3′) sites, E3 by BamHI (5′) andXhoI (3′) sites and cloned into the corresponding sites of pCaSphs43 β-gal. For E1,2-E3/lac, E3 was cloned into the BamHI/XhoI sitesof E1,2/lac. For E1,2-R-E3/lac and E1,2-2R-E3/lac, one or two PCderived fragments (using the primer BamHI site from the oligonu-cleotide GCGGATCCAGGGGCGTCCTT at 5′ and the endogeneousBamHI site at 3′) were added to E1,2-E3/lac, respectively.

The final versions of constructs containing downstream fragmefrom the tin genes from D. mel.(Dm) and D.vir. (Dv) were made asfollows: tin-C (Dm), a 300 bp PCR fragment generated with thprimer pair AGGCGGCCGCCATGAACAGCTT (5′) andAGGGATCCGAGGCAGGGAAA (3′) and cloned into NotI/BamHI;tin-C (Dv), a 354 bp PCR fragment generated with the primer paAGGCGGCCGCTCAGGCACGGATC (5′) and AGCTCGAG-GCAAAACATTTTACAG (3′) and cloned into NotI/XhoI; tin-D(Dm), a 346 bp PCR fragment generated with the primer paGAGAATTCATGTCAAGTGGCACTA (5′) and ACCTCGAG-GTGGGAGGCTCGCAGCT (3′), cloned into EcoRI/XhoI; tin-D(Dv), a 2.3 kb SalI fragment cloned into XhoI.

In vitro mutagenesis was performed with the TransformerTM muta-genesis kit from Clontech. P-element constructs were injected inembryos from yw flies together with ∆2-3 as the transposase sourceBetween two and five lines from each constructs were analyzed expression.

Embryo stainingsIn situ hybridizations were performed as described in Lo and Fras(1997), and antibody stainings or combined antibody/in situ hybriization stainings as in Azpiazu et al. (1996). Tinman antibodies weraised in a rabbit against bacterially expressed His(6)-tagged protethat were purified on a Ni-column (Quiagen). The expressioconstruct contained a 1.1 kb BamHI (generated at the ATG of the firstMet) to HindIII fragment in pQE11 (Qiagen). Peroxidase antibodiewere obtained from L. Fessler (UCLA) and β-gal antibodies werefrom Sigma (mouse polyclonal) and Cappel (rabbit).

Drosophila stocksFollowing mutant alleles were used: btdXG, Dfd21, ems7099, eve1.27,otdD87. lacZ-balancer chromosomes were used to identify homozgous mutant embryos, except for eve, where homozygous embryoswere identified using Eve-antibodies. For ectopic expression of twist,we used a UAS-twist line (Baylies and Bate, 1996) and an en-GAL4line obtained from M. Baylies.

DNaseI footprinting assaysDNaseI protection assays were performed essentially as describeHeberlein et al. (1985) with following modifications: The reactiomixtures (50 µl) contained 110 mM KCl, 47.5 mM Hepes (pH 7.9)13.75 mM MgCl2, 1 mM DTT, 17% glycerol, 0.05% NP-40, 1 µgpoly[d(I-C)], and 5 ng of 3′ end-labeled probe. Upon addition ofpurified GST/Twist fusion proteins (Ip et al., 1992a) or Eve protein(gift from M. Biggin) and incubation for 60 minutes on ice, 50 µl of10 mM MgCl2/5 mM CaCl2 were added, followed by 1 µl DNaseI(Boehringer) to a final concentration of 0.4 µg/ml. After 2.5 minutes

4973Regulation of tinman during Drosophila mesoderm development

s of

a eingtedmnts

pterns

wn-p),

nd

ity

omtheent,

inasesis,

al

to

H H R R B Xb C

m a n

3'C D

B HR RV

3' C D

Xh RV SB P S

D. mel.

D. vir.

1kb

and location of enhancer elements of the tinmangenes from Drosophilairilis. Exons are shown as black and introns as white boxes. Theription start site of the D. melanogaster tingene was based on the of the longest cDNA (Bodmer et al., 1990). The approximate positions

s and 3′ end of the D. virilis genes were deduced from sequenceogastergene. Enhancer elements A to D, as described in the text, arebbreviations: A, anterior head element; B; broad mesodermal element, C,l mesodermal (dpp-response) element; B, BamHI; C, ClaI; H, HindIII; P,baI; Xh, XhoI; mod(mdg4), modifier of midget4gene.

digestion on ice, the reaction was stopped by adding 90 µl of 1% SDS,20 mM EDTA, 200 mM NaCl, 250 µg/ml yeast tRNA, extracted twicewith phenol/chloroform and ethanol precipitated. Electrophoretic saration was done in 8% polyacrylamide/7.5 M urea gels.

RESULTS

Distinct cis -elements drive each aspect ofendogenous tinman expressionAs an initial step in the identification of enhancer sequencesthe tinman gene, we isolated the tinman homolog and itsflanking sequences from a distantly related species, Drosophilavirilis , with the assumption that essential cis-regulatorysequences would be evolutionarily conserved. The tinmangenes from D. melanogasterand D. virilis share a high degreeof sequence similarity in their open reading frames, aorganized in three exons with similar lengths and displalmost identical patterns of expression in the two species (F1; Z. Y. and M. F., unpublished data). As illustrated in Fig. the temporal and spatial expression of D. melanogaster tinprotein expression can be subdivided into four major aspeDuring invagination and early migration of the mesoderm, tinis expressed in all mesodermal cells, except for a small arethe head mesoderm that is negative (Fig. 2A,J). This eaexpression of tin depends on the function of the twist gene(Bodmer et al., 1990). tin expression then becomes restricteto the dorsal portion of the mesoderm, and both tin mRNA andprotein disappear from ventral cells during elongated geband stages (Fig. 2D). It has been shown that restrictedtinexpression in the dorsal mesoderm is triggered by Dpmediated induction events (Frasch, 1995). In subsequent stof embryogenesis, tin expression is limited to cells of the dorsavessel. As shown in Fig. 2G, tin products are detected ingenerally four out of six pairs ofcardioblasts per segment and inpericardial cells during dorsalclosure (Jagla et al., 1997). Theonly non-mesodermal domain oftin expression is located at theanterior tip of the embryo, withhighest expression levels in cellsthat are fated to become part of thepharynx and esophagus (Fig.2A,D,J).

The region containing essentialtin enhancer elements was definedby a genomic construct, encom-passing sequences from−6.2 kb to+4.6 kb with respect to thepresumed transcription start site,that rescued the lethality of nullmutations of tin (Azpiazu andFrasch, 1993). Since the closestgene upstream of tin, mod(mdg4),starts at−1.6 kb (Z. Y. and M. F.,unpublished data), we testedfragments from the entire regionbetween −1.7 kb and +4.8 kb forenhancer activity. Genomicrestriction fragments, or PCR

R C H Xh B

mod(mdg4) t i n

5'A B

C H

5'B

Pv

Fig. 1. Genomic organizationmelanogasterand Drosophila vputative location of the transcposition of the first nucleotideof the 5′ end, exon boundariecomparisons with the D. melanshown as patterned boxes. Acardioblast element; D, dorsaPstI; Pv, PvuII; R, EcoRI; Xb, X

ep-

of

reayig.

2,

cts.

a inrly

d

rm p-

agesl

fragments generated with primer pairs spanning stretchesequences that were conserved between the two Drosophilaspecies, were cloned upstream of a basal promoter and β-galactosidase (β-gal) reporter gene and transformed into thgermline (see Materials and Methods). Fragments showenhancer activity in transgenic embryos were further dissecand tested similarly in additional rounds of analysis. Frothese studies, we identified four distinct enhancer elemefrom D. melanogaster,with lengths varying between 180 band ~500 bp, that activated reporter gene expression in patresembling specific aspects of the tinmanpattern (Fig. 1). Sur-prisingly, all of these elements were located at positions dostream of the transcription start site. Element B (180 blocated in the first intron, activated β-gal expression in thewhole trunk mesoderm during gastrulation and germ-baelongation (Fig. 2B). An element from D. virilis located at asimilar position had an identical temporal and spatial activwhen tested in D. melanogasterembryos (Figs 1, 2C). Theintensity of reporter gene expression with the B elements frboth species was modulated in a pair-rule fashion along anteroposterior axis (see below). A second enhancer elemD (~350 bp), that was located ~2 kb downstream of the 3′ endof tin activated reporter gene expression specifically in cellsthe dorsal portion of the mesoderm (Fig. 2E). Element D wactive between stage 11 and early stage 12 of embryogencorresponding to the period when maintenance of tinmanexpression in the dorsal mesoderm requires dpp. An analogouselement was identified downstream of the D. virilis tinmangene and showed identical activities in transformed D.melanogasterembryos (Figs 1, 2F). A third element from D.melanogaster, C (300 bp; Fig. 1), was active in the dorsvessel. This element activated β-gal expression from stage 12on in four out of six cardioblasts per hemisegment, similarthe cardioblast expression of the endogenous tin gene (Fig.

4974

any

ndin

ofetherm-

Z. Yin, X.-L. Xu and M. Frasch

2H). A cardioblast element with similar activity was agafound at a corresponding position downstream of the D. virilistin gene (Figs 1, 2I). Elements C from both species did exhibit any expression in the pericardial cells of the dorvessel. However, further dissection of the D element fromD.mel. showed it to contain sequences that are indeed abledrive expression in pericardial cells, but this activity obscured by the general dorsal expression with the intacelement (data not shown). Finally, we located an element ~500 bp; Fig. 1) in the 5′ portion of the first intron of the D.mel. tingene that displayed activity in the anterior tip of thhead (Fig. 2K). After invagination of the stomodeum, the buof the cells that are marked with β-gal from this element formthe roof of the pharynx (Fig. 2L).

In summary, the four enhancer elements from the D.melanogaster tingene reflect the four major aspects of edogenous tin expression. At least three of these elements conserved between D.mel and D. vir. with respect to sequencesrelative positions, as well as spatial and temporal activitiImportantly, we were unable to find any enhancer activity

Fig. 2. Comparisons of tinmanprotein expression and the reporter gand D. virilis. (A,D,G,J) tinmanantibody stainings of D. melanogastereembryos transformed with D. mel.-derived enhancer constructs and (vir.-derived enhancer constructs. Anterior is to the left and ventral cells of the trunk mesoderm and in a cap at the dorsal tip of the heD. vir. drive β-gal expression in all cells of the trunk mesoderm. (Dmesodermal cells (d.ms.) and is absent in the ventral mesoderm (and D. vir. are active only during stage 11 and early stage 12 in doshowing tin expression in the majority of cardioblasts (cbs) and in pfrom D. mel.and D. vir. in about four cardioblasts per hemisegmentstage 8 embryo, showing expression in the anterior head cap (arr(bracket). (K) β-gal expression driven by the tin A enhancer from D. meperduring β-gal proteins in pharyngeal structures. The correspondi

in

notsal to

ist D(A,

elk

n-are,es. in

the 5′ flanking and 5′ untranslated regions of tin, in spite oftesting a variety of constructs both with homologous tin andheterologous hsp70 promoters. Neither did we detect enhancement of expression when 5′ flanking sequences werecombined with the first intron that contained elements A aB (see Materials and Methods). These findings are agreement with the virtually normal expression of tin inembryos that are homozygous for deletions of 5′ flankingregions (tin142∆36 and tin142∆32, which delete sequencesbetween−1546 bp and−162 bp or −23 bp, respectively;[Azpiazu and Frasch, 1993]; data not shown).

Three conserved Twist-binding sites in the tin Benhancer are functional in vivoThe early expression of tinman in the whole mesoderm hasbeen shown to depend on twist (Bodmer et al., 1990). Sincethe tin B enhancer elements mimicked this pattern expression, we focused on these elements and tested whthey are direct targets for the Twist protein. Sequence coparisons between the B enhancers from D. melanogasterand

ene expression driven by tinmanenhancer elements from D. melanogastermbryos; (B,E,H,K,L) β-gal antibody stainings of D. melanogasterC,F,I) similar stainings of D. melanogaster embryos transformed with D.is down. (A) In gastrulating embryos, tinmanexpression is observed in allad. (B,C) In embryos of similar stages, the tin B enhancers from D. mel.and

) At stage 11, tin expressing is restricted to an undulating band of dorsalv.ms.). Head expression is enhanced. (E,F) The tin D enhancers from D. mel.rsal portions of the mesoderm. (G) Dorsal view of a stage 14 embryo,ericardial cells (pcs). (H,I) β-gal expression driven by the tin C enhancers

. (J) High magnification view of tin expression in the anterior portion of aowheads) and absence of expression in a region of the head mesoderml.in a head cap (arrowheads). (L) Embryo as in K at stage 12, showingng portion of the intron from D. vir. tin was not tested for enhancer activity.

4975Regulation of tinman during Drosophila mesoderm development

ivityig.dforer 1

y,

ro-nt

ted,w

gs

chedt

al

cksterE1,2g

D. virilis showed strong similarities within multiple stretcheof ~20 bp to ~50 bp length (Fig. 3A), which were candidafor conserved regulatory sequences. Importantly, the sequecorresponding to the minimal tin B enhancer (180 bp) fromD.mel. include three conserved E-box sequences that copotentially bind the bHLH protein Twist. These E-boxecontain the sequences CATGTG or CATATG in both speciand while E-boxes 1 and E-boxes 2 are arranged in tandemboxes 3 are located ~120 bp to ~150 bp further downstre(Fig. 3A). DNaseI footprinting assays with bacterially purifieTwist fusion proteins showed clear protection in the vicinity these sequences, demonstrating that Twist is able to bind tthree E-boxes in vitro (Fig. 4A,C). Binding specificity waconfirmed by using fragments in which the E-boxes have bmutagenized in vitro. Upon altering the sequences of E-boand 2 from D. mel.to TATGTA (E1,2mut I), a strong reductionof Twist-binding was observed, although there was still soprotection of sequences containing the modified E-box 1 athighest concentration of Twist that was tested (Fig. 4B). Tmodified sequence TATATA at the position of E-box 3 (E3mI) lost its ability to bind Twist under these conditions (Fig. 4D

To test whether the in vitro Twist-binding sites arrequired for the activity of the tin B enhancer in vivo, wgenerated a series of reporter constructs with modifi

Fig. 3. Sequencecomparison andmolecular dissectionof tin B enhancerelements from D.melanogasterand D.virilis . A. BESTFITalignment 374 bp tinB enhancer from D.mel.with the 425 bptin B enhancer fromD. vir. The sequencescorresponding to the180 bp tin B enhancerfrom D. mel.areboxed. E-box sequences and consensus binding sites forhomeodomain proteins (and more specifically, Even-skipped) are shaded in black. It may be significant that, infive of the six E-boxes, the 3′ G is folllowed by a T (E-box 3would be in reverse orientation). B. Shown are the normalsequences of the 29 bp E1,2 and E3 elements, as well as the52 bp ‘R’ element, that were used individually or incombination to test for enhancer activities. The modifiedsequences that were obtained in two steps of in vitromutagenesis and tested in the context of the 374 bp tin Belement are shown below the wild-type sequences.

stesnces ulds

es,, E-amdofo allseenx 1

me theheut). eeed

versions of B enhancer sequences and analyzed their actin transgenic embryos. A B enhancer construct (374 bp; F3A) in which E-box 3 had been mutated to TATATA, showereduced reporter gene expression (Fig. 5A; see Fig. 2B comparison). Similar results were obtained with B enhancconstructs containing an intact E-box 3 but mutant E-boxand E-box 2 sequences (TATGTA; Fig. 5B). Most strikinglthe periodic modulation of β-gal expression, which wasevident with the intact B element, became much more pnounced with the constructs containing one or two mutaE-box sequences. When all three E-boxes were mutaenhancer activity was drastically reduced and only a feresidual mesodermal cells expressed β-gal (Fig. 5C). Theweak, residual activity may reflect the residual bindinaffinity of Twist seen in vitro with these modified sequence(Fig. 4B). Therefore, we tested a reporter construct in whithe sequences of all three E-boxes were further modifi(E1,2mut II and E3mut II; Fig. 3B). As predicted, this mutanversion of B exhibited a complete lack of mesodermenhancer activity (Fig. 5D).

We next tested whether the conserved sequence blospanning Twist-binding sites are sufficient to activate reporgene expression in the mesoderm. When the sequences and E3 (Fig. 3B) were linked with each other, stron

4976

la-n ofhisas

lyx).

sioned

s

hetgn

unk).

Z. Yin, X.-L. Xu and M. Frasch

expression was observed in the whole mesoderm (Fig. 5Thus, a 58 bp element containing three putative Twist-bindsites was able to activate β-gal in all cells of the embryo thatcontain Twist. Even a 29 bp fragment, E1,2, containing tfirst two E-boxes was sufficient to drive reporter genexpression in this pattern, albeit at reduced levels (Fig. 3Fig. 5F). By contrast, the E3 sequence was inactive in trunk mesoderm and showed expression only in small regiof the rostral and caudal mesoderm (Fig. 5G and unpublisdata). In order to provide stronger support for the notion tTwist binds and transactivates gene expression throughthree E-boxes in the B enhancer, we mis-expressed Twist wthe binary GAL4 system (Brand and Perrimon, 1993) undthe control of engrailedenhancers in embryos carrying threporter construct with the combined E1,2 and E3 elemeAs shown in Fig. 5H, ectopic expression of Twist in ectodemal stripes caused striped ectopic expression of β-gal in theectoderm. Taken together, these experiments provide strevidence that the three E-box sequences are in vivo targeTwist and, in the context of surrounding sequences of ~50 are responsible for the early activation of tinman in the trunkmesoderm.

Fig. 4. DNA-binding of Twist and Even-skipped to tin B enhancer seq374 bp tin B element (B1). Lane 1 shows a G + A sequence ladderTwi fusion proteins, respectively, and lane 6 a control experiment and 2 (shadowed black), are shown on the left. (B) Similar footprinboxes 1 and 2 (left column, shadowed sequences). Note that two compared to A. (C) Twist DNA-binding experiment as in A, using t′(shadowed black). (D) Twist DNA-binding experiment as in B, usint(E) DNase I footprinting with Even-skipped (Eve) assaying sequen400 ng, 800 ng and 1600 ng of Eve protein were used for lanes 2 consensus binding site for Eve (shadowed black), are shown in th

A B C

E).ing

heeB;

theonshedhat the

ither

ents.r-

ongts ofbp,

Periodic modulation of tin B enhancer activitydepends on eve and an Eve-binding elementDuring its earliest phase of expression, just prior to gastrution, ventral tin expression is modulated in a pair-rule fashioalong the anteroposterior embryo axis, with highest levelsexpression seen in six mesodermal stripes (Fig. 6A). Tfeature may be reflected in the periodic expression that wobserved with tin B enhancer elements, which was particularobvious upon introducing mutations in individual E-bosequences that diminished Twist binding (Figs 2B,C, 5A,BTo determine the segmental register of reporter gene expresin embryos transformed with these constructs, we performdouble stainings for β-gal and pair-rule gene products. Ashown in Fig. 6B, the pattern of β-gal directed by the E1,2 mutI derivative of the B enhancer is complementary to that of tpair-rule gene even-skipped(Macdonald et al., 1986; Frasch eal., 1987). Peak levels of β-gal were observed in areas lackinEve protein. Since this β-gal pattern could indicate repressioof enhancer activity by Eve, we tested its activity in evemutantembryos. Indeed, in the absence of eve activity, the sameenhancer is able to drive strong expression throughout the trmesoderm without any periodic interruptions (Fig. 6C

uences. (A) DNase I footprinting with Twist using the 5′ portion of the, lanes 2 to 5 protection experiments with 200, 400, 800 and 1600 ng of GST-with the GST fusion moiety only. The protected sequences, including E-box 1ting experiment as in A, but with tin B sequences containing mutated E-times higher GST-Twi protein concentrations were used in each lane ashe 3portion of the 374 bp tin B enhancer (B2) which includes E-box 3g a in B element with in vitro mutated E-box 3 (shadowed grey).ces of the R portion of the 374 bp tin B element (BR). 50 ng, 100 ng, 200 ng,to 7, respectively. The protected sequences, including a sequence matching ae left column.

D E

4977Regulation of tinman during Drosophila mesoderm development

ic

o-ed).tormtive

hat,

e8us in

n-of’ ‘R’n inultsad

Fig. 5. Functional analysis of the in vitro Twist-binding sites from the tin B enhancer in vivo. (A) A374 bp tin B enhancer element with E-box 3mutated to TATATA shows reduced and periodicallymodulated levels of reporter gene expression alongthe anterioposterior axis. Weak anterior expressionindicating the presence of a second ‘head element’in the 374 bp B enhancer is also evident. (B) A 374bp tin B enhancer element with E-boxes 1 and 2mutated to TATGTA, shows similar periodicreductions of expression as with the mutant elementfrom A. (C) Mutation of all three E-box sequences(as in A and B) within the 374 bp element leads to adramatic loss of enhancer activity. Arrows point tomesodermal cells with residual, weak expression.(D) Further modifications of E-box sequenceswithin the 374 bp element (see Fig. 3B) completelyabolish its enhancer activity. (E) A combination ofthe elements E1,2 and E3 with native orientations(see Fig. 3B) exhibits strong mesodermal enhanceractivity. Ectopic expression in the head mesoderm ismarked with arrowheads. (F) The 29 bp elementE1,2 is sufficient to drive mesodermal geneexpression. Note that the expression in E and Fincludes the head mesoderm and anterior andposterior endoderm primordia. (G) The 29 bpelement E3 is not sufficient to drive expression inthe trunk mesoderm, but expresses β-gal ectopicallyin a portion of the head mesoderm. (H) EctopicTwist (expressed under the control of engrailedenhancers in ectodermal stripes) is sufficient to activate ectodermal β-gal expression in embryoswith a reporter construct containing in vitro Twist-binding sequences only (E1,2-E3, as in E).

Among the pair-rule mutant tested, this outcome was uniqto eve. In mutants for fushi tarazu(ftz), hairy (h) and runt (run),we observed interruptions in the β-gal pattern along the anteroposterior axis, however the pattern of the interruptions wconsistent with the previously reported alterations in the evepattern in these mutants (data not shown; Frasch and Lev1987).

Additional data indicated that Eve may repress B enhanactivity (and tin) through direct binding to sequences in thelement. The conserved sequence block between E-box 2E-box 3 contained a sequence matching homeodomain bindsites, which was particularly close to a consensus binding previously determined for Eve (Fig. 3A; Hoey et al., 1988). vitro DNA-binding experiments confirmed that this sequenis able to bind Eve, both in the case of D. mel.(Fig. 4E) andD. vir. (data not shown). Enhancer constructs that lacked ‘R’ element containing this Eve-binding site showed uniforexpression throughout the mesoderm (Fig. 5E, F). This pattis similar to that of B enhancers with weakened Twist sitesevemutant backgrounds (Fig. 6C). When the 52 bp ‘R’ elemewas added to the E1,2 and E3 elements, with a configurasimilar to that of the native B enhancer, periodic repression wrestored (Fig. 6D). Moreover, addition of two ‘R’ elements tandem resulted in much stronger periodic repression ayielded a pattern of clean stripes of β-gal in the mesoderm.These results indicate that Eve represses B enhancer actby direct binding to recognition sequences in the B enhan

ue

-as

ine,

ceris anding

siteInce

themern innt

tionas

innd

ivitycer

and thereby reduces its activation by Twist in a periodfashion.

Regulation of tinman in trunk versus headmesodermtin is expressed in the whole trunk mesoderm of late blastderm and gastrulating embryos, but expression is excludfrom portions of the mesoderm in the head (Figs 2A,J, 6AThis is in contrast to the expression of its upstream regulatwist, which is seen throughout the length of the mesoder(Thisse et al., 1988), and indicates the presence of negaregulators of tin in the head mesoderm. Lack of headmesoderm expression was also observed with 180 bp tin Benhancer constructs and their derivatives, which suggested tin addition to positively acting Twi-binding sites, there arenegative elements that prevent activation by Twist in thanterior mesoderm. This possibility was confirmed with the 5bp and 29 bp constructs, containing the Twist-binding sites plimmediately adjacent sequences, which lacked repressionthe head mesoderm and showed a pattern identical to twist(Fig.5E,F). Since the major difference between these latter costructs and the tin B enhancer constructs was the absence the ‘R’ element, it was likely that sequences within the ‘Relements mediated the head repression. Indeed, when theelement was added to the E1,2 and E3 elements, repressiothe head mesoderm was restored (Fig. 6D,E). These resshow that the ‘R’ elements mediate repression in the he

4978

tal

Z. Yin, X.-L. Xu and M. Frasch

mesoderm, in addition to their function in Eve binding anperiodic repression of Twist activation. Since head represswas maintained in evemutant embryos (Fig. 6C), additionanegative regulators must act on the ‘R’ element.

The absence of tin in the head mesoderm may be function

u-nedt-

ared

ed

ermet

esfgh- et

allyin

nkunkso-

on byiveti-

hus,larents.g.nspo-tsr et

Fig. 6.Periodic abrogation of B enhancer activity by eve and the ‘Relement. (A)tin mRNA expression in a wild-type embryo atblastoderm, showing periodic reductions of tin levels along theanteroposterior axis. (B) Stage 8 embryo carrying the E1,2mutI/laderivative of the tin B enhancer and stained for β-gal (brown) andEve (black). β-gal staining is observed between the Eve stripes,which have split up into a strong and a weak component at this st(indicated by solid parts of bracket). The absence of β-gal expressionin areas between strong and weak evestripes could be explained bythe repressing activities of more broadly distributed eveproductsprior to this stage (indicated by entire bracket). (C) Stage 8 evemutant embryo (eve1.27) carrying the same transgene and stained aembryo in B. Absence of everesults in a uniform mesodermalactivity of this enhancer. (D) Stage 9 embryo carrying an E1,2-R-E3/lac transgene. The presence of the ‘R’ element leads to periodreductions of mesodermal β-gal expression (compare Fig. 5E).(E) Stage 9 embryo carrying an E1,2-2R-E3/lac transgene. Thepresence of two ‘R’ elements leads to wide periodic gaps of β-galexpression and, thus, a pattern of six mesodermal stripes.

dionl

-

ally important, since these cells have different developmenfates (Tepass et al., 1994) and express serpent(srp), a gene ofthe GATA family. serpentis required for the normal differen-tiation of these cells into hemocytes (Rehorn et al., 1996). srpis expressed in the cells of the head mesoderm that lacktinexpression (Fig. 7A,C, arrowheads; note that srp has a second,more anterior domain that overlaps with endodermal tinexpression). To identify upstream genes involved in the reglation of gene expression in the head mesoderm, we examitin and srp expression in embryos mutant for early head-paterning genes, including buttonhead(btd; Wimmer et al.,1991), Deformed(Dfd; Merrill et al., 1987), empty spiracles(ems; Dalton et al., 1989), orthodenticle (otd; Cohen andJürgens, 1990; Finkelstein and Perrimon, 1990) and ems,otddouble mutants. Interestingly, in gastrulating embryos that mutant for btd, tin expression was expanded into the heamesoderm, whereas the mesodermal domain of serpentwasabsent (Fig. 7B). During blastoderm stages, we observseverely reduced, transient expression of srp in the headmesoderm of btd mutants (Fig. 7D). None of the othermutations tested produced any obvious alterations of the tinand srp domains, indicating a major role of btd in the pattern-ing of the head mesoderm. Similar to tin, expression of β-galfrom B enhancer constructs expanded into the head mesodin the absence of btd function (Fig. 7E,F). Taken together, thesresults suggest that btd acts through the ‘R’ element to preventin activation by twist in the head mesoderm. In addition, btdis required for normal activation of srp in the same area. Sinceboth these activities of btd may be essential for normal bloodcell development, we examined the distribution of hemocytin btd mutant embryos. Normally, a large number ohemocytes, which stain for peroxidasin, are scattered throuout the body cavity of late stage embryos (Fig. 7G; Nelsonal., 1994; Tepass et al., 1994). By contrast, in btd mutantembryos, the number of hemocytes was reduced and virtuall of the residual cells remained near their place of origin the embryonic head (Fig.7H).

DISCUSSION

The mesoderm-patterning gene tinmanhas three major phasesof mesodermal expression, the first occurring in the whole trumesoderm, the second in broad dorsal subdomains of the trmesoderm and the third in heart progenitors at the dorsal medermal crest. This refinement from an initially broad expressidomain to progressively smaller areas could be explainedseveral different molecular mechanisms, including selectmRNA stabilization or successive waves of transcriptional acvation. Our analysis of regulatory regions from the tin genedemonstrates that each of these aspects of tin expression isindeed controlled by a separate enhancer module and, treflects a distinct transcriptional activation event. The moducharacter of these enhancers was also evident in experimwith larger elements containing two different enhancers (eA+B, C+D; data not shown), which produced additive patterof reporter gene expression. Candidates for upstream comnents that trigger tin activation through some of these elemenhave been identified in previous genetic experiments (Bodmeal., 1990; Frasch, 1995). Our present study shows that twistactsthrough the early (B) element (Fig. 8). dpp is likely to act

’

c

age

s

ic

4979Regulation of tinman during Drosophila mesoderm development

ly

o-kle,

rsralre-

hel.,

in tin and srp regulation in the head mesoderm and in hemocyteage 9 wild-type embryo stained for Tin protein (brown) and srpmRNAded from a portion of the head mesoderm (between arrowheads) thatells at the border between the Tin and srpdomains may express bothond domain of srp, which constitute prospective endoderm of the anteriore left arrowhead), co-express Tin. (B) In a stage 9 btdmutant embryo,

panded into the head mesoderm (between arrowheads) and mesodermalent. The endodermal domains of expression are unaltered. (C) srp a wild-type embryo at blastoderm stage. The mesodermal srpdomainytes is marked by arrowheads. (D) srpexpression in a btdmutant embryodermal srpexpression (arrowheads) is severely reduced. (E) The 374 bpt active in the head mesoderm (arrowheads). (F) In a btdmutant embryo,4 bp tin B enhancer is derepressed in the head mesoderm (arrowheads).pe embryo stained with antibodies against peroxidasin to detecty and ring gland are also stained. (H) Stage 15 btdmutant embryo, stainededuced number of hemocytes and all of them remain in the embryoniciations: fb, fat body; hc, hemocytes; rg, ring gland.

through the dorsally active (D) element, a notion reinforced our observation that this element is inactive in dpp mutantembryos (data not shown). The upstream activators acting onheart element (element C) are not yet known. Although tinexpression in the heart progenitors, and the formation of hprogenitors proper, have been shown to require wingless(Wu etal., 1995), the critical period of wg requirement occurs earlier(at stage 10-11) than the activation of this enhancer, which dnot begin until stage 12. Furthermore, element C does contain any sequences closely matching those of binding sfor dTCF/Pangolin, which is the Drosophilahomolog of LEF-1and appears to mediate responses to the winglesssignal (Brunneret al., 1997; van de Wetering et al., 1997;Z. Y. and M. F., unpublished results).Therefore we favor the idea that winglessis required early, together with tinmanduring its expression in a broad, dorsaldomain, to determine heart progenitoridentities. In a subsequent step, these cellsactivate tin through as yet undefined mech-anisms that involve element C in the caseof the cardioblasts. In order to obtainfurther insights into the molecular mecha-nisms of mesoderm patterning, Dpp-mediated induction and heart differen-tiation, it will be of major importance tofunctionally dissect the C and D elementsto a similar extent to that done in this studywith the B element, and to identify the cor-responding binding factors. Since tinman-related genes are expressed in the develop-ing heart of vertebrate embryos and theirexpression depends on bone morpho-genetic proteins (BMPs) that are homolo-gous to Dpp (for a review, see Harvey,1996; Schultheiss et al., 1997), it is con-ceivable that some of the molecular mech-anisms involving the C and D elements areconserved. This possibility can now testedthrough sequence comparisons andreporter gene assays in heterologoussystems.

In addition to the regulatory elementsthat are active in the mesoderm, ouranalysis has also revealed the presence ofdistinct elements driving tin expression incells at the anterior tip of the embryonichead. Tagging these cells with β-galshows that they give mainly rise topharynx and anterior endoderm, consis-tent with previous fate map studies for thisarea (Technau and Campos-Ortega,1985). Although this aspect of tinexpression has previously received lessattention, it is interesting to note that thepharynx and anterior endoderm are alsoprominent sites of expression of tinman-related genes from vertebrates (Lints etal., 1993; Tonissen et al., 1994; Evans etal., 1995; Lee et al., 1996; Brand et al.,1997). Thus, some of the upstream regu-

Fig. 7.The role of btddevelopment. (A) St(purple). Tin is excluexpresses srp. A few cgenes. Cells in a secmidgut (anterior to thTin expression is exsrpexpression is absmRNA expression ingiving rise to hemocat blastoderm. Mesotin B enhancer is nothe activity of the 37(G) Stage 16 wild-tyhemocytes. Fat bodas in G. There is a rhead region. Abbrev

by

the

eart

oesnotites

lators acting through element A are likely to be evolutionariconserved as well. In Drosophila, candidates include Bicoidand D-gsc, the Zn-finger protein Tailless, and the Torsdependent Ras pathway (Berleth et al., 1988; Hahn and Jäc1996; Goriely et al., 1996; Sprenger et al., 1989).

The focus of the present study is the regulation of tinexpression in the early trunk mesoderm. Early tin expressionhas several important biological functions. Notably, it appeato be required for the specification of cell fates in the ventportion of the mesoderm, such as those of distinct muscle pcursors and of mesodermally derived glia-like cells at tventral midline (Azpiazu and Frasch, 1993; Gorczyka et a

4980

nth97).the an

theandfent

ngvest,ere

ff’xis. byuslyley,

diva-rsw

ging.a- toenef

en--torsIt

mall

dis-icellstic

ill

tectghing

Z. Yin, X.-L. Xu and M. Frasch

5' 3'

A B C D

HEAD (pharynx,ant.endoderm)

TRUNKMESODERM CARDIOBLASTS

DORSALMESODERM

twi dppevebtd

? wg

Fig. 8. Summary of tinmanenhancer elements and their regulatoryinputs.

1994). It is further required in an autoregulatory fashion allow tin expression at high levels in the second, dpp-dependent phase in the dorsal mesoderm (X.-L. X., Z. Y. aM. F., unpublished data). Our experiments demonstrate tearly mesodermal tin expression is driven by a distinct regulatory module, element B, that is composed of closely linkpositively and negatively acting sequences. The basic heloop-helix protein Twist appears to be the principal activatof this element, which has three E-box-containing bindinsites. Two of these are arranged in tandem, with the E-bobeing only seven base pairs apart. Strikingly, a minimelement, E1,2, containing these two E-boxes and only 17 bpairs of additional sequences is able to activate transcriptiothe mesoderm, suggesting that activation by Twist does require many additional DNA-binding factors. Moreover, ectdermal expression of the reporter gene upon mis-expressioTwist suggests that Twist is the only mesodermally restrictfactor that is required to activate tin through this module,although generally expressed proteins could act as co-factFor example, given that bHLH proteins tend to form heerodimers (Murre et al., 1989b), it is possible that, in vivTwist binds as a heterodimer with ubiquitously expressbHLH proteins to each of these E-box sequences. A candidfor a Twist partner could be the the ubiquitously expressbHLH protein Daughterless (Da; Cronmiller et al., 198Murre et al., 1989a). However, we found that the eaexpression of tin in embryos lacking both the maternal anzygotic functions of da is normal and, thus, Da does not appeto be an essential Twist partner (M. F., unpublished daAlthough Twist could bind as a homodimer, it probabrequires one or more co-factors that bind to immediateflanking sequences for transcriptional activation. This suggested by the inability of other fragments from the tin locusto activate early mesodermal gene expression, even thothey contain E-box clusters that bind Twist in vitro with similaaffinities (Lee et al., 1997; Z. Y. and M. F., unpublished dat

Since vertebrate homologs of tinman are not broadlyexpressed in the early mesoderm, the Twist-dependent action of tin-related genes may not be conserved between insand vertebrates. However, Twist homologs are expressed inearly mesoderm of vertebrate embryos, which suggests thaleast some molecular aspects of Twist function have beconserved. Previous studies have focused on inhibitfunctions of mouse Twist during myogenesis, which appear

to

ndhat-edlix-org

xesalasen innoto-n ofed

ors.t-o,edateed

8;rlydarta).lylyis

ughr

a).

tiva-ects thet aten

ory to

involve competition for binding of myogenic factors, titratioof E proteins, and the formation of inhibitory complexes wipromoter sequences (Spicer et al., 1996; Hebrok et al., 19It will be important to determine whether mouse Twist has ability to also activate target genes, and our identification ofactivating Twist-response element in Drosophilamay providea means to address this question.

The 52 bp ‘R’ element, which is interspersed between E-boxes, counteracts the activating functions of the E1,2 E3 sequences by preventing tin activation in specific areas othe twist expression domain. This arrangement is reminiscof an enhancer module from the rhomboid(rho) promoter thatcontains closely linked positively and negatively acting bindisites (Ip et al., 1992b). In this case, the Dorsal protein seras an activating factor, with a minor contribution of Twiswhile in the mesoderm, Snail repressor molecules interfwith this activation, leading to defined on/off states of rhoexpression along the dorsoventral axis. In the case of tin, Twistis the key activator, and negative regulators determine ‘odomains at specific positions along the anteroposterior aOne repressor molecule, Even-skipped, reduces activationTwist in odd-numbered parasegments. Eve has been previodescribed as a transcriptional repressor (Han and Man1993). In the tin B element, it could interfere with theformation of an active complex, which includes Twist bounto the three E-boxes, or it could reduce transcriptional acttion through interactions with basal transcription facto(Austin and Biggin, 1995; Um et al., 1995). We do not knowhether the pair-rule modulation of tin is functionally relevant,or whether Eve is merely ‘pirating’ a homeodomain bindinsite without any major consequences for mesoderm pattern

The major role of the ‘R’ element may be to prevent activtion of tin in the portion of the head mesoderm that is fatedform hemocytes. We have identified the head-patterning gbtd as a negative regulator of tin and as a positive regulator othe hematopoietic differentiation gene srp in this region. Thisand additional genetic experiments with tin and srpmutants (Z.Y. and M. F., unpublished data) suggest that the complemtary patterns of srp and tin are not achieved by mutual inhibition, but rather by an overlapping set of upstream regulathat affect tin and srp expression in an opposite manner. should be noted that other regulators in addition to btdmust beinvolved, because we observe some residual activation ofsrpin a small mesodermal domain in btd mutants that disappearsprematurely during gastrulation. Similarly, tin mRNAexpression is initially absent in the head mesoderm of btdmutants, but expands into this area after blastoderm. The samount of srpproducts in btd− appears to be sufficient to allowthe formation of some hemocytes. However, the premature appearance of srp expression and perhaps the ectopexpression of tinmanseem to interfere with normal hemocytdifferentiation, as indicated by the failure of the remnant ceto migrate into the embryonic body cavity. There is geneevidence for a distinct enhancer driving srp expression in theblood mesoderm (Rehorn et al., 1996). Once available, it wbe interesting to compare this enhancer element to the tin Benhancer and to determine how reciprocal patterns of tin andsrpare achieved on the molecular level. Since we did not deany specific binding of Btd protein to the R element (althouit bound well to control sequences containing a SP1-bindsite; data not shown), it is possible that tin repression by btd is

4981Regulation of tinman during Drosophila mesoderm development

l in

al

the

e

nd

indirect and is mediated through gene products downstreambtd.

In conclusion, our experiments have uncovered two distimolecular stategies to achieve temporally and spatially restrictin expression during mesoderm patterning. The first makes of an array of discrete enhancer modules that are targets forferential regulatory inputs and function independently of oanother to activate tin expression in progressively smalledomains of the mesoderm. The second strategy is exemplifiethe early-acting enhancer and employs a combination of closlinked binding sites of activating and repressing molecules wita module. Functional competition of activators and repressrestrict the activity of this enhancer to defined areas within Twist domain, thus contributing to the subdivision of the mesdermal germ layer into blood and trunk mesoderm.

We thank N. Azpiazu for making the Tin expression construct, Biggin for Eve proteins, L. Fessler for antibodies, J. Licht for genomlibraries, T. Ip and A. Michelson for plasmids, M. Baylies, C. Cromiller, J. Mohler, E. Wimmer for fly stocks and H. Nguyen focomments on the manuscript and for helping to produce tin antibodThis work was supported by grants from the National Institute Health (HD30832), the American Heart Foundation and a Pew Awto M. F.

REFERENCES

Austin, R. and Biggin, M. (1995). A domain of the even-skippedproteinrepresses transcription by preventing TFIID binding to a promotrepression by cooperative blocking. Mol. Cell. Biol.15,4683-4693.

Azpiazu, N. and Frasch, M.(1993). tinman and bagpipe: two homeo boxgenes that determine cell fates in the dorsal mesoderm of Drosophila. GenesDev.7,1325-1340.

Azpiazu, N., Lawrence, P., Vincent, J.-P. and Frasch, M.(1996).Segmentation and specification of the Drosophilamesoderm. Genes Dev.10,3183-3194.

Baehrecke, E.(1997). whoencodes a KH RNA binding protein that functionsin muscle development. Development124,1323-1332.

Baylies, M. K. and Bate, M.(1996). twist: A myogenic switch in Drosophila.Science272,1481-1484.

Berleth, T., Burri, M., Thoma, G., Bopp, D., Richstein, G., Frigerio, G.,Noll, M. and Nüsslein-Volhard, C.(1988). The role of localization of bicoidRNA in organizing the anterior pattern of the Drosophilaembryo. EMBO J.7,1749-1756.

Bodmer, R. (1993). The gene tinmanis required for specification of the hearand visceral muscles in Drosophila. Development118,719-729.

Bodmer, R., Jan, L. Y. and Jan, Y. N.(1990). A new homeobox-containinggene, msh-2, is transiently expressed early during mesoderm formationDrosophila. Development110,661-669.

Brand, A. H. and Perrimon, N. (1993). Targeted gene expression as a meaof altering cell fates and generating dominant phenotypes. Development118,401-415.

Brand, T., Andree, B., Schneider, A., Buchberger, A. and Arnold, H.(1997).Chicken NKx2-8, a novel homeobox gene expressed during early heart foregut development. Mech. Dev.64,53-59.

Brunner, E., Peter, O., Schweizer, L. and Basler, K.(1997). pangolinencodesa Lef-1 homologue that acts downstream of Armadillo to transduce Wingless signal in Drosophila. Nature385,829-833.

Casal, J. and Leptin, M.(1996). Identification of novel genes in Drosophilareveals the complex regulation of early gene activity in the mesoderm. Proc.Natl. Acad. Sci. USA93,10327-10332.

Cohen, S. and Jürgens, G.(1990). Mediation of Drosophilahead developmentby gap-like segmentation genes. Nature346,482-485.

Cronmiller, C., Schedl, P. and Cline, T. W.(1988). Molecular characterizationof daughterless, a Drosophilasex determination gene with multiple roles indevelopment. Genes Dev.2, 1666-1676.

Dalton, D., Chadwick, R. and McGinnis, W. (1989). Expression and

of

nctteduse dif-nerd byely

hinorstheo-

M.ic

n-ries.ofard

er:

t

of

ns

and

the

embryonic function of empty spiracles: a Drosophilahomeo-box gene withtwo patterning functions on the anterior-posterior axis of the embryo. Gen.Dev.3,1940-1956.

Evans, S., Yan, W., Murillo, M., Ponce, J. and Papalopulu, N.(1995).tinman, a Drosophila homeobox gene required for heart and visceramesoderm specification, may be represented by a family of genesvertebrates: XNkx-2.3, a second vertebrate homologue of tinman.Development121,3889-3899.

Finkelstein, R. and Perrimon, N.(1990). The orthodenticlegene is regulatedby bicoid and torsoand specifies Drosophilahead development. Nature346,485-488.

Frasch, M. (1995). Induction of visceral and cardiac mesoderm by ectodermDpp in the early Drosophilaembryo. Nature374,464-467.

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

Frasch, M. and Levine, M.(1987). Complementary patterns ofeven-skippedand fushi tarazuexpression involve their differential regulation by a commonset of segmentation genes in Drosophila. Genes Dev.1,981-995.

Gorczyka, M. G., Phyllis, R. W. and Budnik, V.(1994). The role of tinman, amesodermal cell fate gene, in axon pathfinding during the development oftransverse nerve in Drosophila. Development120,2143-2152.

Goriely, A., Stella, M., Coffinier, C., Kessler, D., Mailhos, C., Dessain, S.and Desplan, C. (1996). A functional homologue of goosecoid inDrosophila. Development122,1641-1650.

Hahn, M. and Jäckle, H.(1996). Drosophila goosecoidparticipates in neuraldevelopment but not in body axis formation. EMBO J.15,3077-3084.

Han, K. and Manley, J.(1993). Transcriptional repression by the Drosophilaeven-skippedprotein: definition of a minimal repression domain. Genes Dev.7,491-503.

Hanna-Rose, W., Licht, J. and Hansen, U.(1997). Two evolutionarilyconserved repression domains in the Drosophila Krüppelprotein differ inactivator specificity. Mol. Cell. Biol.17,4820-4829.

Harvey, R. (1996). NK-2 homeobox genes and heart development. Dev. Biol.178,203-216.

Heberlein, U., England, B. and Tjian, R. (1985). Characterization ofDrosophila transcription factors that activate the tandem promoters ofthalcohol dehydrogenase gene. Cell 41,965-977.

Hebrok, M., Füchtbauer, A. and Füchtbauer, E. (1997). Repression ofmuscle-specific gene activation by the murine Twist protein. Exp. Cell Res.232,295-303.

Hoey, T., Frasch, M. and Levine, M.(1988). Sequence-specific DNA bindingactivities of even-skippedand other homeobox proteins in Drosophila.Nature332,858-861.

Ip, T. Y., Park, R. E., Kosman, D., Yazdanbakhsh, K. and Levine, M.(1992a). dorsal-twist interactions establish snail expression in thepresumptive mesoderm of theDrosophilaembryo. Genes Dev. 6,1518-1530.

Ip, Y. T., Park, R. E., Kosman, D., Bier, E. and Levine, M.(1992b). Thedorsalgradient morphogen regulates stripes of rhomboidexpression in thepresumptive neuroectoderm of the Drosophilaembryo. Genes Dev.6, 1728-1739.

Jagla, K., Frasch, M., Jagla, T., Dretzen, G., Bellard, F. and Bellard, M.(1997). ladybird, a new component of the cardiogenic pathway in Drosophilarequired for diversification of heart precursors. Development, in press.

Kosman, D., Ip, Y. T., Levine, M. and Arora, K.(1991). Establishment of themesoderm-neuroectoderm boundary in the Drosophilaembryo. Science252,118-122.

Lai, Z., Fortini, M. and Rubin, G. (1991). The embryonic expression patternsof zfh-1 and zfh-2, two Drosophila genes encoding novel zinc-fingerhomeodomain proteins. Mech. Dev.34,123-134.

Lee, K., Xu, Q. and Breitbart, R.(1996). A newtinman-related gene, nkx2.7,anticipates the expression of nkx2.5 and nkx2.3 in zebrafish heart andpharyngeal endoderm. Dev. Biol.180,722-731.

Lee, Y., Park, T., Schulz, R. and Kim, Y.(1997). Twist-mediated activation oftheNK-4 homeobox gene in the visceral mesoderm of Drosophilarequirestwo distinct clusters of E-box regulatory elements. J. Biol. Chem.272,17531-17541.

Leptin, M. (1991). twist and snail as positive and negative regulators duringDrosophilamesoderm formation. Genes Dev. 5,1568-1576.

Lilly, B., Galewsky, S., Firulli, A., Schulz, R. and Olson, E.(1994). D-MEF2:a MADS box transcription factor expressed in differentiating mesoderm amuscle cell lineages during Drosophila embryogenesis. Proc. Natl. Acad.Sci. USA91,5662-5666.

Lints, T. J., Parsons, L. M., Hartley, L., Lyons, I. and Harvey, R.(1993).

4982

d

re

Z. Yin, X.-L. Xu and M. Frasch

Nkx-2.5: A novel murine homeobox gene expressed in early heart progencells and their myogenic descendants. Development119,419-431.

Lo, P. and Frasch, M. (1997). A novel KH domain protein mediates celadhesion processes in Drosophila. Dev. Biol., in press.

Macdonald, P. M., Ingham, P. and Struhl, G.(1986). Isolation, structure, andexpression of even-skipped: a second pair-rule gene of Drosophilacontaininga homeo box. Cell 47,721-34.

McGinnis, W., Levine, M. S., Hafen, E., Kuroiwa, A. and Gehring, W. J.(1984). A conserved DNA sequence in homoeotic genes of the DrosophilaAntennapediaand bithoraxcomplexes. Nature308,428-33.

Merrill, V. K., Turner, F. R. and Kaufman, T. C. (1987). A genetic anddevelopmental analysis of mutations in the Deformedlocus in Drosophilamelanogaster. Dev. Biol.122,379-95.

Murre, C., McCaw, P. S. and Baltimore, D.(1989a). A new DNA binding anddimerization motif in immunoglobulin enhancer binding, daughterless,MyoD, and mycproteins. Cell 56,777-783.

Murre, C., McCaw, P. S., Vaessin, H., Caudy, M., Jan, L. Y., Jan, Y. N.,Cabrera, C. V., Buskin, J. N., Hauschka, S. D., Lassar, A. B., Weintraub,H. and Baltimore, D. (1989b). Interactions between heterologous helixloop-helix proteins generate complexes that bind specifically to a commDNA sequence. Cell 58,537-544.

Nelson, R., Fessler, L., Takagi, Y., Blumberg, B., Keene, D., Olson, P.Parker, C. and Fessler, J.(1994). Peroxidasin: a novel enzyme-matrixprotein of Drosophiladevelopment. EMBO J.13,3438-3447.

Nguyen, H., Bodmer, R., Abmayr, S., McDermott, J. and Spoerel, N.(1994).D-mef2: a Drosophilamesoderm-specific MADS box-containing gene with biphasic expression profile during embryogenesis. Proc. Natl. Acad. Sci.USA91,7520-7524.

Rehorn, K.-P., Thelen, H., Michelson, A. and Reuter, R.(1996). A molecularaspect of hematopoiesis and endoderm development common to verteband Drosophila. Development122,4023-4031.

Schultheiss, T., Burch, J. and Lassar, A.(1997). A role for bonemorphogenetic proteins in the induction of cardiac myogenesis. Genes Dev.11,451-462.

Shishido, E., Higashijima, S., Emori, Y. and Saigo, K.(1993). Two FGF-receptor homologues of Drosophila: one is expressed in mesodermaprimordium in early embryos. Development117,751-761.

Simpson, P.(1983). Maternal-zygotic gene interactions during formation of thdorsoventral pattern in Drosophilaembryos. Genetics105,615-632.

Spicer, D., Rhee, J., Cheung, W. and Lassar, A.(1996). Inhibition ofmyogenic bHLH and MEF2 transcription factors by the bHLH protein TwisScience272,1476-1480.

itor

l

-on

,

a

rates

l

e

t.

Sprenger, F., Stevens, L. M. and Nüsslein-Volhard, C.(1989). TheDrosophilagene torso encodes a putative receptor tyrosine kinase. Nature338,478-483.

Taylor, M., Beatty, K., Hunter, H. and Baylies, M.(1995). DrosophilaMEF2is regulated by twistand is expressed in both the primordia and differentiatecells of the embryonic somatic, visceral and heart musculature. Mech. Dev.50,29-41.

Technau, G. M. and Campos-Ortega, J. A.(1985). Fate-mapping in wild-type Drosophila melanogasterII. Injections of horseradish peroxidase incells of the early gastrula stage. Roux’s Arch. Dev. Biol.194,196-212.

Tepass, U., Fessler, L., Aziz, A. and Hartenstein, V.(1994). Embryonic originof hemocytes and their relationship to cell death in Drosophila. Development120,1829-1837.

Thisse, B., Stoetzel, C., Gorostiza-Thisse, C. and Perrin-Schmitt, F.(1988).Sequence of the twist gene and nuclear localization of its protein inendomesodermal cells of early Drosophilaembryos. EMBO J.7,2175-2183.

Thummel, C. S., Boulet, A. M. and Lipshitz, H. D.(1988). Vectors forDrosophila P-element-mediated transformation and tissue cultutransfection. Gene74,445-450.

Tonissen, K., Drysdale, T., Lints, T., Harvey, R. and Krieg, P.(1994). XNkx-2.5, a Xenopusgene related to Nkx-2.5and tinman: evidence for a conservedrole in cardiac development. Dev. Biol.162,325-328.

Um, M., Li, C. and Manley, J. (1995). The transcriptional repressor even-skippedinteracts directly with TATA-binding protein. Mol. Cell. Biol. 15,5007-5016.

van de Wetering, M., Cavallo, R., Dooijes, D., van Beest, M., van Es, J.,Loureiro, J., Ypma, A., Hursh, D., Jones, T., Bejsovec, A., Peifer. M.,Mortin, M. and Clevers, H. (1997). Armadillo coactivates transcriptiondriven by the product of the Drosophilasegment polarity gene dTCF. Cell88,789-799.

Wimmer, E., Jackle, H., Pfeifle, C. and Cohen, S.(1991). A Drosophilahomologue of human Sp1 is a head-specific segmentation gene. Nature366,690-694.

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

Zaffran, S., Astier, M., Gratecos, D. and Semeriva, M.(1997). The held outwings (how) Drosophila gene encodes a putative RNA-binding proteininvolved in the control of muscular and cardiac activity. Development124,2087-2098.

(Accepted 24 September 1997)