Previous Article | Next Article ![]()
Molecular and Cellular Biology, June 2007, p. 3962-3969, Vol. 27, No. 11
0270-7306/07/$08.00+0 doi:10.1128/MCB.00093-07
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Department of Biochemistry and Molecular Biology, Program in Genes and Development, The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030
Received 16 January 2007/ Returned for modification 27 February 2007/ Accepted 12 March 2007
|
|
|---|
|
|
|---|
Drosophila has also emerged as an important model for the study of the genetic and cellular bases of hematopoiesis. Mature hemocytes arise during embryogenesis from the cephalic mesoderm and during larval development from the lymph glands, tissues associated with the anterior region of the dorsal vessel (7, 13, 17). A close connection exists between certain genes controlling cardiogenesis and hematopoiesis, as tin and pnr functions are required for the specification of the cardiogenic mesoderm, and lymph gland primordia are derived from a subset of these cells. Additionally, hemanogioblast-like progenitors that divide asymmetrically to generate daughter cells that are cardioblasts contributing to the anterior dorsal vessel and prohemocytes contributing to the lymph glands have been identified (17). Thereafter, hematopoietic factors of the GATA, Friend of GATA, Runx, and Glial Cells Missing classes control the differentiation of hemocyte precursors into defined hematopoietic cell lineages (7, 8). As with the cardiogenic regulators, functional homologues of these genes control various aspects of hematopoiesis in vertebrates.
While important insights into genes controlling heart and blood cell development in Drosophila have been gained, the story remains complex, with the need to discover additional regulatory components of these processes. With this goal in mind, we generated transgenic fly strains that express cardiac enhancer-green fluorescent protein (GFP) constructs that mark cardioblasts of the dorsal vessel. One characterized enhancer includes a cardioblast regulatory module of the Toll gene, which is positively activated by the functions of Tin and Doc (27). Such a reagent has been used in screens of the Drosophila genome for additional genes required for dorsal vessel morphogenesis.
In this report, we identify tailup (tup) as being a gene that is essential for heart formation in Drosophila. Tup is a LIM homeodomain transcription factor (26) expressed in several cell types of the dorsal vessel, and its mutation causes severe cardioblast, pericardial cell, and lymph gland deficiencies. Our studies also establish Tup as being a direct transcriptional activator of the Hand gene, which encodes a heart-expressed basic helix-loop-helix transcription factor (11). The importance of these findings is twofold. First, they provide new mechanistic insights into the transcriptional network controlling heart and blood cell development in Drosophila. Second, as Tup- and Hand-related proteins function in vertebrate heart development, these results further substantiate the evolutionary conservation of genes controlling cardiogenesis and hematopoiesis and implicate a regulatory relationship for the vertebrate genes.
|
|
|---|
Mutant phenotype analyses. Wild-type and mutant embryos were collected and stained with various antibodies as previously described (27). The following primary antibodies were used in these studies: mouse anti-ß-galactosidase at a 1:300 dilution (Promega, Madison, WI), rabbit anti-D-MEF2 at a 1:1,000 dilution (H. Nguyen), mouse anti-Tup at a 1:100 dilution (Developmental Studies Hybridoma Bank), mouse anti-GFP at a 1:200 dilution (BD Biosciences, Palo Alto, CA), mouse anti-Col at a 1:100 dilution (M. Crozatier), rabbit anti-Srp at a 1:500 dilution (D. Hoshizaki), and rabbit anti-Odd at a 1:500 dilution (J. Skeath). GFP expression in living embryos was detected as previously described (27).
Tup and Hand enhancer analyses. Four partially overlapping DNAs spanning 13 kb of sequence upstream of the tup gene were generated by PCR amplification of the bacterial artificial chromosome (BAC) clone RP98-3E19 obtained from the BACPAC Resource Center (Children's Hospital, Oakland Research Institute). PCR products were initially cloned into the pCRII-TOPO vector using the TA cloning kit from Invitrogen (San Diego, CA) and subsequently moved into the P-element vector pH-Stinger (1). A similar strategy was used to generate Hand-GFP transgenes. DNAs were obtained by PCR amplification of the BAC clone RP98-30M19 with primers designed based on the previously published Hand cardiac and hematopoietic enhancer (HCH) sequence (11). The compositions of primers used to generate wild-type tup and wild-type or mutant Hand DNAs are available upon request. Transgenic lines carrying the tup-GFP or Hand-GFP DNAs were established after the injection of yw67c23 embryos using standard transformation techniques (10). At least five lines were established and analyzed for each DNA tested.
Additional molecular analyses. To determine the molecular mutations present in the tup1 and tupisl-1 alleles, genomic DNA was isolated from wild-type and heterozygous adults by using a protocol described previously by McGinnis and Beckendorf (18). Primer pairs were designed to generate six overlapping tup DNAs that spanned the complete exon/intron organization of the gene. Specific sequences of the primers used for PCR amplification are available upon request. PCR products were generated using Taq DNA polymerase, amplification reagents, and protocols obtained from Roche Applied Science (Indianapolis, IN). Wild-type and mutated tup DNA sequences were determined by the Macrogen Corporation (Rockville, MD).
For electromobility shift assays with Tup and Hand, the Tup protein was generated using the TnT Quick Coupled Transcription/Translation system (Promega). These assays were performed as described previously by Ranganayakulu et al. (21). Complementary oligonucleotides containing the consensus Tup-1 and Tup-2 binding sites of the Hand HCH enhancer were used. The oligonucleotide sequences for wild-type and competitor probes for the Tup-1 site were Tup-1A (5'-AATCAAACCCCTAATGGATTAAAATG-3') and Tup-1B (5'-CATTTTAATCCATTAGGGGTTTGATT-3') (the Tup binding site is underlined). Oligonucleotides with an altered sequence, previously shown to disrupt Isl1 DNA binding (6), served as the mutant competitor DNA. These included Tup-1A mutant (5'-AATCAAACCCCTAggtGATTAAAATG-3') and Tup-1B mutant (5'-CATTTTAATCaccTAGGGGTTTGATT-3'). The oligonucleotide sequences for wild-type and competitor probes for the Tup-2 site were Tup-2A (5'-CCGCACTTCCATTAGGAATATATCT-3') and Tup-2B (5'-AGATATATTCCTAATGGAAGTGCGGT-3') (the Tup binding site is underlined). Oligonucleotides with altered sequences served as the mutant competitor DNA. These included Tup-2A mutant (5'-ACCGCACTTCaccTAGGAATATATCT-3') and Tup-2B mutant (5'-AGATATATTCCTAggtGAAGTGCGGT-3') (lowercase type indicates the base changes within the Tup binding sites).
|
|
|---|
![]() View larger version (28K): [in a new window] |
FIG. 1. Dorsal vessel phenotypes observed in tup mutant embryos. (A to D) Toll-nGFP transgene expression in (A) wild-type (wt), (B) homozygous Df(2L)OD15, (C) homozygous tup1, and (D) homozygous tupisl-1 embryos. Embryos are at stage 16 of development. (E and F) H15-lacZ transgene expression in (E) wild-type and (F) homozygous tup1 embryos. Embryos are at stage 14 of development. (G and H) D-MEF2 protein expression in (G) wild-type and (H) homozygous tup1 embryos. Embryos are at stage 15 of development. Open arrowheads point to missing cardioblasts and pronounced gaps in dorsal vessels of mutant embryos. All embryos are oriented with the anterior to the left. Abbreviations: as, amnioserosa cells; cb, cardioblasts; sm, somatic body wall muscles.
|
tup expression in multiple cell types of the dorsal vessel. tup has been studied mostly in the context of its role in patterning the central nervous system. Thus, limited information was available on its expression in the heart tube other than the detection of tup mRNA and protein in undefined cells of the dorsal vessel (FlyBase report on tup) (26). We used an anti-Tup monoclonal antibody to determine the Tup-positive cells within the dorsal vessel. The Tup protein is initially observed in the dorsal mesoderm around stage 10, where it could function in the specification of various dorsal vessel cell types (data not shown). By stage 16, four specific cell types were identified as expressing Tup in the mature linear organ: prohemocytes of the lymph glands, pericardial cells, all cardioblasts, and alary muscles that attach the heart tube to the overlying epidermis (Fig. 2A and B). Such observations were consistent with the requirement of tup for normal dorsal vessel morphogenesis.
![]() View larger version (42K): [in a new window] |
FIG. 2. tup expression in cells within and associated with the dorsal vessel. (A) Low-magnification (x20) image of a wild-type (wt) embryo immunostained for the Tup protein. Tup expression in the dorsal vessel (dv) is indicated. (B) High-magnification (x40) image of a wild-type embryo immunostained for Tup protein. Tup expression is observed (from anterior to posterior) in lymph glands (lg), pericardial cells (pc), cardioblasts (cb), and alary muscles (am). (C) Mapping the location of regulatory modules controlling tup dorsal vessel expression. Boxes within the tup gene correspond to exon sequences. (D) Expression of the tup-F4-GFP transgene in all four types of Tup-positive cells of the dorsal vessel. All embryos are at stage 16 of development and oriented with the anterior to the left.
|
The tup gene produces two mRNA transcripts due to alternative splicing, which result in two versions of Tup that differ solely in their C-terminal sequences (Fig. 3A). We sought to determine the molecular mutations present in the tup1 and tupisl-1 alleles so as to define potential alterations in the Tup proteins. For the tup1 allele, a missense mutation that changes Cys-57 to Tyr-57 was found. This Cys residue is present in the first zinc finger of the LIM domain and is fully conserved among LIM domain proteins (14). Mutation of this critical amino acid resulted in highly unstable Tup proteins, as tup1 mutant embryos accumulated very low levels of Tup (Fig. 3C). Such an observation was consistent with the strong dorsal vessel phenotype observed in homozygous tup1 embryos. We failed to find a mutation in tup exon or intron sequences in DNA obtained from tupisl-1 mutants, suggesting that this alteration may reside in transcriptional regulatory sequences of the gene.
![]() View larger version (13K): [in a new window] |
FIG. 3. Determination of the molecular mutation present in the tup1 allele. (A) Two very similar forms of Tup are generated by the alternative splicing of pre-mRNA. The LIM and homeobox domains of the proteins are indicated. A single nucleotide mutation present in the tup1 allele results in the change of an amino acid crucial for zinc finger formation within the first LIM domain; specifically, the codon for Cys-57 is mutated to one encoding Tyr-57. (B and C) Double staining of (B) wild-type (wt) and (C) homozygous tup1 embryos for Tup and D-MEF2 reveals an absence (open arrowhead) of the Tup protein in mutant animals. The embryos are at stage 13 of development and oriented with the anterior to the left. Abbreviations: as, aminoserosa cells; cb, cardioblasts; sm, somatic body wall muscles.
|
![]() View larger version (39K): [in a new window] |
FIG. 4. High-resolution analyses of tup dorsal vessel and lymph gland phenotypes. (A and B) tup-F4-GFP transgene expression in (A) wild-type (wt) and (B) homozygous tup1 embryos. Lymph glands (lg), pericardial cells (pc), and cardioblasts (cb) of the normal dorsal vessel are indicated in the wild-type embryo. (C and D) Col protein and tup-F4-GFP transgene expression in (C) wild-type and (D) homozygous tup1 embryos. Arrows point to Col expression in the posterior signaling centers of the lymph glands associated with the normal dorsal vessel, and comparable Col staining is observed along the defective heart tube of the mutant embryo. Col expression is also observed in bilateral posterior structures in the wild-type embryo, which are out of focus in the tup1 embryo due to a defect in dorsal closure. (E and F) Hand-GFP transgene expression in (E) wild-type and (F) homozygous tup1 embryos. Prominent GFP expression in the lymph glands, pericardial cells, and cardioblasts in the normal dorsal vessel is highlighted. (G and H) Srp protein and Hand-GFP transgene expression in (G) wild-type and (H) homozygous tup1 embryos. Arrows point to the coincident expression of Srp and Hand in the lymph glands of the wild-type embryo. (I and J) Odd protein and Toll-nGFP transgene expression in (I) wt and (J) tup1 embryos. Prominent Odd expression in the lymph glands and pericardial cells is highlighted, while Toll expression in cardioblasts is also indicated in the wild-type embryo. (B, D, F, H, and J) All five of the markers, or marker combinations, reveal the absence of most lymph gland (closed arrowheads) and pericardial (open arrowheads) cells in defective dorsal vessels found in tup1 mutant embryos. Mutant embryos also exhibit defects in dorsal closure, with prolonged GFP expression in cells of the amnioserosa using the Toll-nuclear GFP (nGFP) marker (J). All embryos are at stage 16 of development and oriented with the anterior to the left.
|
Direct regulation of the Hand dorsal vessel enhancer by Tup. Since tup is expressed and functions in three cell types that also express the Hand gene, we considered the possibility that Tup may be a transcriptional regulator of Hand through its cardiac and hematopoietic enhancer. Previous studies defined the presence of required Tin and GATA recognition elements in the Hand 513-bp HCH enhancer (11), and a scan of this sequence also identified two regions with consensus Isl1 binding sites (6). Electromobility shift assays were undertaken with the Tup protein and the putative binding elements, with the finding that Tup can selectively interact with both of the wild-type recognition sequences (Fig. 5). These molecular data were consistent with the possibility of Tup functioning as a transcriptional activator of Hand dorsal vessel expression through this defined regulatory module.
![]() View larger version (36K): [in a new window] |
FIG. 5. Two Tup DNA recognition sites are present in the Hand 513-bp HCH enhancer. The locations of Tup CTAATG binding elements, relative to previously defined Tin and GATA protein binding sites (15), are indicated in the Hand DNA. At the bottom, electromobility gel shift assays reveal that Tup can selectively bind to double-stranded oligonucleotides containing the Tup-1 or Tup-2 sequence. Abbreviations: MT, Tup-site-mutated oligonucleotide; WT, wild-type oligonucleotide.
|
![]() View larger version (32K): [in a new window] |
FIG. 6. Tup is a direct transcriptional regulator of the Hand dorsal vessel enhancer. (A, left) Schematic of the Hand 513-bp HCH enhancer and three Tup site mutant versions thereof. (A, right) Function of wild-type and mutated Hand-GFP DNAs in transgenic embryos. Activities of the various enhancers in Tin or Svp/Doc cardioblasts (cb), lymph glands (lg), or pericardial cells (pc) are indicated as positive (+), negative (), or greatly reduced (+/). (B and C) Normal cellular expression profile of GFP in embryos harboring the wt-Hand-GFP or mTup1-Hand-GFP transgenes. (D and E) Representative embryos expressing the mTup2-Hand-GFP or mTup1/2-Hand-GFP construct. With both mutated DNAs, GFP is observed in Tin cardioblasts but absent from the lymph glands (closed arrowheads) and pericardial cells (open arrowheads). Reporter expression is greatly diminished in Svp/Doc cardioblasts (arrows) with both mutations. (F) Twi-Gal4>UAS-tup embryo expressing the wt-Hand-GFP marker. This forced expression condition results in an expanded population of prohemocytes in the lymph glands (horizontal bar and asterisk) and a disorganized heart region (arrow). (G) Expression of the tup-F4-GFP transgene in a homozygous Handko mutant embryo. tup expression appears to be normal in cardioblasts and pericardial cells but reduced in the lymph glands. All embryos are at stage 16 of development and oriented with the anterior to the left.
|
|
|
|---|
While the degree of cardioblast hypoplasia is variable in mutant embryos, the severe reduction in prohemocytes of the lymph glands and pericardial cells surrounding the contractile tube is fully penetrant. The Col protein serves as an excellent marker for lymph gland primordia and the posterior signaling centers of lymph glands associated with the mature dorsal vessel (5). Since Col expression is normal in tup mutants, Tup function is not required for the early specification of lymph gland primordia within the dorsal mesoderm. However, the severe reduction of several mature lymph gland markers such as tup-GFP, Hand-GFP, Srp, and Odd suggests that either prohemocytes are present within lymph glands with Tup activity essential for expression of all four of these indicator genes or the cells are absent due to defects in prohemocyte proliferation and/or programmed cell death. The latter is an attractive possibility since Hand knockout embryos show ectopic apoptosis among lymph gland progenitor cells (12).
A function for the Hand basic helix-loop-helix transcription factor has been reported for cardioblast, pericardial, and lymph gland cells. This is the same set of dorsal vessel and hematopoietic cells that require Tup function. Through our analysis of the Hand cardiac and hematopoietic enhancer, we demonstrated that Tup is a direct transcriptional regulator of Hand in these cell types. Specifically, mutation of the single Tup-2 element in the HCH regulatory module resulted in a dramatic loss or reduction of Hand enhancer activity in prohemocytes, pericardial cells, and the Svp/Doc cardioblast subtype (9, 16, 22, 31). These findings invoke two possibilities. First, tup phenotypes may be due to the lack of Hand expression and function in cardioblasts, pericardial cells, and lymph gland progenitors. However, Tup function is likely to be even more critical for cardiogenic and hematopoietic events, as the forced expression of tup results in the production of excess prohemocytes, while the ectopic expression of Hand does not. Thus, Tup can be considered to be a seminal upstream regulator of genetic and cellular events controlling lymph gland formation. Second, Tin and GATA factors have been shown to regulate the HCH enhancer. Thus, it is possible the Hand cardiac and hematopoietic transcription occurs due to combinatorial control, specifically via Tup and Doc cofunction in Svp/Doc-expressing cardioblasts and Tup and Srp coactivity in lymph gland progenitors. Ample evidence exists for the function of multiple interacting transcription factors in the regulation of heart and blood cell gene expression in Drosophila (23, 25). To summarize regulatory aspects of its function, the data showing that Tup is a direct transcriptional activator of Hand expression in lymph glands, pericardial cells, and Svp/Doc-positive cardioblasts through the HCH enhancer module are compelling (Fig. 7). Likewise, Tup serves as either a direct or indirect regulator of srp expression in lymph gland cells and odd expression in lymph gland and pericardial cells.
![]() View larger version (23K): [in a new window] |
FIG. 7. Proposed functions for Tup in regulating lymph gland, pericardial cell, and Svp/Doc cardioblast gene expression.
|
In summary, Drosophila has served as a powerful system to discover signaling pathways and transcriptional regulators governing heart development. Through the generation of a sensitive cardiac enhancer-GFP transgene and its use in a screen for additional genes required for normal dorsal vessel morphogenesis, we identified an additional evolutionarily conserved component of the gene regulatory network controlling cardiac development. The further detailed study of Tup in terms of its functions in cardiogenesis and hematopoiesis and its molecular interactions with other proven heart and blood cell regulators, should provide needed insights into mechanistic functions of LIM homeodomain class proteins in these vital developmental processes.
This research was supported by grants to R.A.S. from the National Institutes of Health (HL059151 and HL071540).
Published ahead of print on 19 March 2007. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»