Gene expression pattern

Identi®cation of maverick, a novel member of the TGF-b superfamilyin Drosophila

Minh Nguyen, Louise Parker, Kavita Arora*

Department of Developmental and Cell Biology, University of California, Irvine, CA 92697, USA

Received 21 December 1999; received in revised form 17 March 2000; accepted 21 March 2000

Abstract

The transforming growth factor-b (TGF-b) superfamily of structurally related ligands regulates essential signaling pathways that control

many aspects of cell behavior in organisms across the phylogenetic spectrum. Here we report the identi®cation of maverick (mav), a gene that

encodes a new member of the TGF-b superfamily in Drosophila. Phylogenetic analysis and sequence comparison suggest that Mav cannot be

easily assigned to any one sub-family, since it is equally related to BMP, activin and TGF-b ligands. mav maps to the fourth chromosome and

is expressed throughout development. In situ hybridization experiments reveal the presence of maternally derived mav transcript in

precellular blastoderm embryos. Later in development, mav is expressed in a dynamic pattern in the developing gut, both in endodermal

and visceral mesodermal cells. In adult females, high levels of mav mRNA are present in late stage egg chambers. q 2000 Elsevier Science

Ireland Ltd. All rights reserved.

Keywords: Drosophila; Transforming growth factor-b; Bone morphogenetic proteins; Activin

1. Results and discussion

Based on sequence similarity the TGF-b superfamily can

be subdivided into three main groups: the prototypical TGF-

bs, activins and bone morphogenetic proteins (BMPs). Until

recently the only representatives of the TGF-b superfamily

in Drosophila were the BMP-related ligands, Decapentaple-

gic (Dpp), Screw (Scw) and Glass bottom boat (Gbb;

Padgett et al., 1987; Wharton et al., 1991; Doctor et al.,

1992; Arora et al., 1994). Genetic and functional studies

have established that these ligands are involved in critical

developmental events such as patterning of the body axes

and determination of cell fates, as well as regulation of cell

proliferation and apoptosis (Raftery and Sutherland, 1999).

More recently, an activin B ortholog and Myoglianin, a

ligand related to GDF8-Myostatin, were identi®ed in Droso-

phila, but the biological roles of these ligands are not known

(Kutty et al., 1998; Lo and Frasch, 1999). In this study we

report the identi®cation and expression pattern of Maverick

(Mav), a new BMP/TGF-b related ligand in Drosophila.

A BLAST search of the Berkeley Drosophila Genome

Project (BDGP) EST database for sequences sharing simi-

larity to the carboxyl-terminal ligand domain of Scw iden-

ti®ed two uncharacterized clones, CK00014 and CK00025,

that contained the same 1.3 kb insert (Fig. 1A). Sequence

analysis of CK00014 revealed that it encodes a 378 amino

acid peptide with signi®cant similarity to members of the

TGF-b superfamily. We have named this protein Maverick

(Mav). An approximately 1.1 kb PCR ampli®ed genomic

fragment corresponding to the open reading frame of

CK00014 was used to screen a lgt10 cDNA library derived

from imaginal discs. This resulted in the isolation of a 1.7 kb

mav cDNA, designated 6A1. Sequence analysis revealed

that 6A1 extends 440 nucleotides further 5 0 than

CK00014. BLAST searches with the entire 1.7 kb sequence

identi®ed six additional cDNA clones that originate 5 0 to

6A1, as well as a 17.6 kb genomic clone (AC014858)

containing the mav locus (Fig. 1A). The longest EST clones

LD13618 and LD22618 contain a common 2.7 kb insert.

Comparison of cDNA sequences with the genomic sequence

indicated that the mav locus encodes at least two alterna-

tively spliced transcripts. The transcript corresponding to

cDNAs LD22618, LD13618 and LD46352 is obtained by

splicing out a 108 nucleotide intron (Fig. 1A). In the alter-

natively spliced variant represented by 6A1, an additional

49 nucleotide intron is removed. Both introns are ¯anked by

consensus splice donor and acceptor sites, suggesting that

they represent authentic introns. Northern blot analysis of

adult male and female mRNAs detected an approximately

2.8 kb band (data not shown), suggesting that the cDNA

Mechanisms of Development 95 (2000) 201±206

0925-4773/00/$ - see front matter q 2000 Elsevier Science Ireland Ltd. All rights reserved.

PII: S0925-4773(00)00338-5

www.elsevier.com/locate/modo

* Corresponding author. Tel.: 11-949-824-1087; fax: 11-949-824-4709.

E-mail address: karora@uci.edu (K. Arora).

clones LD13618, LD22618 are close to full length, assum-

ing the presence of a 50±100 nucleotide poly(A)1 tail.

Primer extension analysis using RNA from adult females

supports this assertion, since we detected a prominent tran-

scription start site 18 nucleotides upstream of the ®rst

nucleotide of LD13618 (Fig. 1B). The presence of a single

band on the Northern blot suggests that the alternatively

spliced mav transcripts may be of similar length, or that

the transcript corresponding to 6A1 is of low abundance.

In situ hybridization indicates that mav is located on the

fourth chromosome. Cross-hybridization to a cosmid contig

generated by the Canadian Drosophila Genome Project

allowed a more precise localization of mav to the cytoge-

netic region 102C. Given that chromosome four contains

M. Nguyen et al. / Mechanisms of Development 95 (2000) 201±206202

only 2% of the Drosophila euchromatin, it is striking that

mav is the third member of the TGF-b superfamily to map to

the same region, the other two being myoglianin (102C) and

dActivin (102F; Kutty et al., 1998; Lo and Frasch, 1999).

The longest mav isoform (Mav1) encodes a putative

protein 701 amino acids in length (Fig. 1C). Hydropathy

analysis showed that the ®rst in-frame Methionine is

followed by a stretch of hydrophobic amino acids indicative

of a signal sequence. In isoform 2 represented by 6A1,

initiation of translation would occur at Met 269 and retain

the same reading frame, to yield a predicted protein of 433

amino acids (Fig. 1C). Mav2 lacks a signal sequence

suggesting that this form of the protein may not be secreted.

The consensus multibasic proteolytic cleavage site (RKDK

at residues 586-89 in Mav1), would generate a mature

ligand of 112 amino acids (MassagueÂ, 1998).

Sequence comparisons and phylogenetic analysis of the

ligand domain of Mav suggests that this protein cannot be easily

assigned to either the TGF-b, the BMP or the activin subfami-

lies (Fig. 2). Mav shows the highest sequence conservation with

the BMP related ligands, human BMP3 (32% identity), human

GDF10 (31% identity), and mouse Nodal (29% identity). Inter-

estingly, Mav is only slightly less related to human TGF-b2 and

Activin A (28 and 23% identity respectively). Among the

Drosophila ligands, Mav shares the highest sequence identity

with Myoglianin (28%), and the BMPs, Dpp and Gbb (27%),

followed by Scw (21%). Mav is only 19% identical to dActivin.

Given that phylogenetic analysis places Mav in a cluster of

BMP ligands, it is interesting that the putative Mav ligand

domain contains nine invariant cysteine residues that are typical

of TGF-b and activin ligands (see Fig. 1C; Daopin et al., 1992;

Schlunegger and GruÈtter, 1992). One possible interpretation is

that Mav is part of an emerging group of BMP-like ligands that

contain nine cysteines, such as mammalian GDF8, GDF11 and

GDF15, as well as C. elegans Daf-7 (Ren et al., 1996). Further

investigations of Mav function may help in de®nitively assign-

ing it to a particular subfamily of ligands.

To gain an insight into the role of mav, we determined the

temporal and spatial distribution of the transcript (Fig. 3).

mav is expressed throughout embryonic and larval develop-

ment and persists until adulthood, as determined by RT-

PCR analysis (Fig. 3A). In situ hybridization showed that

mav mRNA is uniformly distributed throughout the embryo

at early syncitial blastoderm (stages 1±4), suggesting that it

is maternally contributed (Fig. 3B). Staining is also

observed in the pole cells. In post-cellular blastoderm

embryos at stage 5, mav mRNA is expressed ubiquitously

(Fig. 3C). During gastrulation (stages 6±8) and germ band

elongation (stages 9±11), mav transcripts can be detected at

low levels throughout the embryo. In addition, mav shows a

dynamic pattern of expression in both germ layers forming

the gut. mav mRNA is ®rst observed in the endoderm at late

stage 9 and early stage 10 in the hindgut and posterior

midgut primordia (Fig. 3D). Stage 11 embryos show stain-

ing in the foregut and anterior midgut primordia, as well as

in the developing posterior midgut and hindgut (Fig. 3E).

However by stage 12, mav message is restricted to a small

subset of cells in the hindgut and a section of the foregut

which eventually becomes part of the pharynx/esophagus

(Fig. 3G). In stage 13 and 14, after germ band retraction,

the visceral mesoderm that surrounds the anterior and

posterior midgut endoderm, expresses mav at high levels

(Fig. 3H). Expression in the gut persists throughout late

embryogenesis, albeit at lower levels (Fig. 3I). In addition

to the gut, mav transcripts are also detected in discrete

segmental patches of cells that are most obvious at stage

11 (Fig. 3F). This staining appears to be restricted to the

dorso-lateral epidermal region. Finally, high levels of mav

mRNA are seen during late stages of oogenesis. In stage 10

egg chambers mav is expressed in the nurse cells and subse-

quently found uniformly distributed in the oocyte (Fig. 3J).

The distribution of mav transcript overlaps signi®cantly

with the expression patterns of the type II receptor Punt,

and the receptor regulated Smad, dSmad2 (Childs et al.,

1993; Letsou et al., 1995; Brummel et al., 1999), raising

the possibility that they are involved in Mav signaling.

In conclusion, we have identi®ed a novel member of the

TGF-b superfamily in Drosophila. Phylogenetic analysis and

sequence comparison precludes the assignment of Mav to a

speci®c subfamily. Thus, Mav may be an orphan molecule or

the prototypic member of a new subfamily of ligands.

2. Materials and methods

2.1. Cloning and sequencing of mav cDNA

Primers QK3 (5 0CCTACTACGTGGCGAA3 0) and QK4

(5 0AGAATAATTCGTTAGAGAAAG3 0), ¯anking the ORF

of the EST CK00014 were used to amplify a genomic DNA

M. Nguyen et al. / Mechanisms of Development 95 (2000) 201±206 203

Fig. 1. (A) Molecular organization of the mav genomic region and cDNAs. A restriction map of the region containing the mav locus is shown. (C, ClaI; E,

EcoRI; H, HindIII; P, PstI). The locations of primers used are marked. The cDNAs analyzed in this study are shown to scale. The closed boxes denote an ORF,

while the stippled regions represent 5 0 and 3 0 untranslated regions. Two alternatively spliced transcripts derive from the mav locus. cDNAs LD13618, LD22618

and LD46352 correspond to a transcript that contains a single 108 nucleotide intron and encodes a protein of 701 residues. An alternatively spliced transcript,

represented by 6A1, contains an additional 49 nucleotide intron and results in a protein of 433 amino acids. (B) Primer extension analysis using the mavPE

primer detects a transcriptional start site (marked with an arrow), 18 nucleotides upstream of the start of the LD13618 cDNA. The ®rst four lanes on the left

indicate a DNA sequencing ladder using the same primer. (C) Sequence of the near full-length mav cDNA and deduced amino acid sequence (Genbank

accession no. AF252386). The ®rst methionine of Mav 1 (Met 1) and the hydrophobic stretch of amino acid residues following it are underlined. The

proteolytic site (RKDK, residues 586±589) is boxed and the nine invariant cysteines are encircled. The ®rst in-frame ATG of the isoform Mav 2 (Met 269) is

highlighted in gray. The 49 nucleotides comprising the alternatively spliced intron are underlined. A potential poly adenylation signal and a poly(A)1 addition

site are in bold letters.

fragment for cDNA library screens. Clone 6A1 as well as two

shorter clones were obtained from 200 000 plaques screened.

cDNA clones LD22618, LD13618 and LD46352, were

obtained from Genome Systems. The entire mav cDNA

sequence was derived from sequencing cDNAs 6A1 and

CK00014 in their entirety, and the 5 0 sequence of LD13618

past the region of overlap with 6A1. All 5 cDNAs have iden-

tical 3 0 ends although they differed in the length of the

poly(A)1 tail. Restriction mapping and sequencing of PCR

ampli®ed DNA was used to con®rm that LD22618,

LD13618 and LD46352, contain sequences corresponding to

the ®rst intron.

2.2. Primer extension analysis

Primer extension analysis was performed as described in

Stathakis et al. (1999) using the primer mavPE

(5 0GATCACAGTTCCTCGAAAATGTG3 0) located 68

nucleotides from the 5 0 end of LD13618.

M. Nguyen et al. / Mechanisms of Development 95 (2000) 201±206204

Fig. 2. (A) Amino acid sequence alignment of Mav. The ligand domains of Mav, human BMP3 (Genbank accession no. 115072), human GDF10 (Genbank

accession no. 1705471), human TGF-b2 (Genbank accession no. 557563), and human Activin A (Genbank accession no. 124279) were compared using the

Clustal program. Identical residues are boxed in black. Residues in gray represent similarities. (B) The dendogram illustrates the phylogenetic relationships

among the ligand domains of proteins in the TGF-b superfamily. The % amino acid identity between Mav and other ligands is listed. The accession numbers

for the ligands are as follows: BMP2 (115068), BMP4 (1070540), Dpp (118409), BMP10 (3873291), BMP9 (5932438), GDF5 (631181), BMP5 (115075),

BMP6 (115076), BMP7 (115078), BMP8 (461635), Gbb (283697), Scw (600163), GDF1 (4503967), Vg1 (137969), BMP3 (115072), GDF10 (4826740),

Nodal (423520), ActivinA (124279), ActivinB (106726), ActivinC (5031795), dActivin (AF054822), GDF11 (5031613), GDF8 (2623582), Myoglianin

(AF132814), TGFb1 (224622), TGFb4 (2501178), TGFb2 (557563), TGFb3 (339552), GDF15 (6753968), Daf7 (1684866), MIS (127109), GDNF (729567).

2.3. RT-PCR analysis

Poly(A)1 RNA was isolated from yw ¯ies using

RNAeasy columns and the oligotex mRNA Midi Kit

(Qiagen), and cDNA prepared with the First Strand cDNA

Synthesis Kit (Amersham Pharmacia Biotech). The Mav

speci®c primers QK3 and QK4 ampli®ed a 1068 bp frag-

ment present in both alternatively spliced forms, and were

used to detect overall levels of mav expression. Primers

PACAI1 (5 0CCAGGGATTTGCGTGCAACTGCTGGT-

GCTATTC3 0) and PACAT2 (5 0ATTTGAAAGGGC-

TCCAACGGCTGCTCCTACACG3 0) ¯anking an intron in

M. Nguyen et al. / Mechanisms of Development 95 (2000) 201±206 205

Fig. 3. (A) RT-PCR analysis demonstrates that mav transcript is expressed throughout development as well as in adults. The top panel shows mav expression

while the bottom panel detects the expression of the a-catenin control. RNA samples corresponding to different stages of development are marked. (B±J)

Spatial and temporal distribution of mav transcript. Lateral views of embryos are shown, oriented anterior to the left, except where noted. (B) Pre-cellular

blastoderm embryo showing maternal contribution of mav. Staining can also be seen in the pole cells. (C) Stage 5 embryo showing uniform distribution of mav

at cellular blastoderm stage. (D) Low levels of zygotic expression are seen throughout a stage 9 embryo, while the hindgut and posterior midgut primordia

express mav at higher levels. (E) In a stage 11 embryo mav is expressed in the foregut and anterior midgut primordia, in addition to the posterior and hindgut

primordia. (F) A segmental-repeat pattern of epidermal expression can be seen in a late stage 11 embryo. (G) At stage 12 the gut speci®c expression is restricted

to a subset of cells in the foregut and hindgut. Epidermal cells continue to express mav in a segmental pattern. (H) Dorsal view of a stage 13 embryo showing

expression in the visceral mesoderm and in the epidermis. (I) In a stage 17 embryo staining is limited to the larval gut. (J) A stage 10 ovariole showing high

levels of expression in the nurse cells and uniform transcript distribution in the oocyte.

a-catenin (Genbank accession no. D13964) were used both

as a positive control and to rule out genomic DNA contam-

ination.

2.4. In situ hybridization

Embryos and ovaries were collected from yw ¯ies. DIG-

labeled SP6 sense and T7 anti-sense riboprobes were gener-

ated from Mav cDNA clone 6A1. In situ hybridization was

performed as described in Tautz and Pfei¯e (1989), except

that hybridization was carried out at 558C for 60 h, followed

by 6£12 h washes in hybridization buffer, also at 558C.

Embryos and ovarioles were mounted in canada balsam/

methyl salicylate (GMM) for observation and photography.

In situ hybridization to salivary gland chromosomes was

carried out using standard protocols.

Acknowledgements

We would like to thank the Canadian Drosophila Genome

Project for help with mapping the mav gene to the fourth

chromosome contig. We are grateful to Dean Stathakis and

Rahul Warrior for their thoughtful contributions to the work

presented in this manuscript. This research was supported

by an NIH-RO1 grant to K.A. (GM55442) and a Searle

Scholar Award from the Kinship Foundation.

References

Arora, K., Levine, M., O'Connor, M., 1994. The screw gene encodes a

ubiquitously expressed member of the TGF-b family required for speci-

®cation of dorsal cell fates in the Drosophila embryo. Genes Dev. 8,

2588±2601.

Brummel, T., Abdollah, S., Haerry, T.E., Shimell, M.J., Merriam, J.,

Raftery, L., Wrana, J.L., O'Connor, M.B., 1999. The Drosophila activin

receptor Baboon signals through dSmad2 and controls cell proliferation

but not patterning during larval development. Genes Dev. 13, 98±111.

Childs, S.R., Wrana, J.L., Arora, K., Attisano, L., O'Connor, M.B., Massa-

gueÂ, J., 1993. Identi®cation of a Drosophila activin receptor. Proc. Natl.

Acad. Sci. USA 90, 9475±9479.

Daopin, S., Piez, K.A., Ogawa, Y., Davies, D.R., 1992. Crystal structure of

transforming growth factor-b2: an unusual fold for the superfamily.

Science 257, 369±373.

Doctor, J.S., Jackson, P.D., Raksha, K.E., Visalli, M., Hoffmann, F.M.,

1992. Sequence, biochemical characterization, and developmental

expression of a new member of the TGF-b superfamily in Drosophila

melanogaster. Dev. Biol. 151, 491±505.

Kutty, G., Kutty, R.K., Samuel, W., Duncan, T., Jaworski, C., Wiggert, B.,

1998. Identi®cation of a new member of transforming growth factor b

superfamily in Drosophila: the ®rst invertebrate activin gene. Biochem.

Biophys. Res. Commun. 246, 644±649.

Letsou, A., Arora, K., Wrana, J.L., Simin, K., Twombly, V., Jamal, J.,

Staehling-Hampton, K., Hoffman, F.M., Gelbart, W.M., Massague, J.,

O'Connor, M.B., 1995. Dpp signaling in Drosophila is mediated by the

punt gene product: a dual ligand binding type II receptor of the TGF-b

receptor family. Cell 80, 899±908.

Lo, P.C.H., Frasch, M., 1999. Sequence and expression of myoglianin, a novel

Drosophila gene of the TGFb superfamily. Mech. Dev. 86, 171±175.

MassagueÂ, J., 1998. TGFb signal transduction. Annu. Rev. Biochem. 67,

753±791.

Padgett, R.W., St. Johnston, R.D., Gelbart, W.M., 1987. A transcript from a

Drosophila pattern gene predicts a protein homologous to the trans-

forming growth factor-(beta) family. Nature 325, 81±84.

Raftery, L.A., Sutherland, D.J., 1999. TGFb family signal transduction in

Drosophila development: from Mad to Smads. Dev. Biol. 210, 251±

268.

Ren, P., Lim, C., Johnsen, R., Albert, P.S., Pilgrim, D., Riddle, D.L., 1996.

Control of C. elegans larval development by neuronal expression of a

TGFb homolog. Science 274, 1389±1391.

Schlunegger, M.P., GruÈtter, M.G., 1992. An unusual feature revealed by the

crystal structure at 2.2 AÊ resolution of human transforming growth

factor-b2. Nature 358, 430±434.

Stathakis, D.G., Udar, N., Sandgren, O., Andreasson, S., Bryant, P.J.,

Small, K., Forsman-Semb, K., 1999. Genomic organization of human

DLG4, the gene encoding post synaptic density 95. J. Neurochem. 73,

2250±2265.

Tautz, D., Pfei¯e, C., 1989. A non-radioactive in situ hybridization method

for the localization of speci®c RNAs in Drosophila embryos reveals

translational control of the segmentation gene hunchback. Chromosoma

98, 81±85.

Wharton, K.A., Thomsen, G.H., Gelbart, W.M., 1991. Drosophila 60A

gene, a new TGF-b family member is closely related to human bone

morphogenetic proteins. Proc. Natl. Acad. Sci. USA 88, 9214±9218.

M. Nguyen et al. / Mechanisms of Development 95 (2000) 201±206206