Hand, an evolutionarily conserved bHLH transcription factor required for Drosophila cardiogenesis and hematopoiesis

The initial steps in heart formation are remarkably conserved from fruit flies to mammals (Cripps and Olson,2002; Zaffran and Frasch,2002). In both types of organism, mesodermal progenitors become committed to a cardiac fate in response to signals from adjacent tissues and converge along the embryonic midline to form a linear cardiac tube with rhythmic contractility (Cripps and Olson,2002). In Drosophila, the myocardial cell layer of the contractile heart tube, composed of cardioblasts, is surrounded by pericardial nephrocytes, which function as secretory cells(Crossley, 1972; Mandal et al., 2004), and by lymph gland hematopoietic cells that give rise to all the major blood cells in the adult fly (Evans et al.,2003). Cardioblasts, pericardial nephrocytes and lymph gland hematopoietic progenitors – the three major embryonic cell types that comprise the Drosophila circulatory system – arise from the same cardiac mesoderm, specified by signaling pathways involving Decapentaplegic (DPP), Wingless (WG) and FGF(Cripps and Olson, 2002; Evans et al., 2003). In the more complex mammalian cardiovascular system, cardiac and hematopoietic progenitors are also derived from the same mesodermal region – the lateral mesoderm – and are specified by conserved signaling pathways involving bone morphogenetic protein (BMP), WNT and FGF(Cripps and Olson, 2002; Evans et al., 2003),exemplifying the conservation of developmental programs for cardiogenesis and hematopoiesis between Drosophila and mammals.

NK-type homeodomain proteins and the GATA family of zinc-finger transcription factors are required for cardiac and hematopoietic development in Drosophila and mammals(Harvey, 1996; Sorrentino et al., 2005). The Drosophila NK family transcription factor, Tinman, and its mammalian ortholog Nkx2.5, are expressed specifically in the developing heart and are both regulated by the DPP/BMP pathway (Yin and Frasch, 1998; Xu et al., 1998; Liberatore et al., 2002; Lien et al., 2002). Both Tinman and Nkx2.5 play central roles in activation of myocardial genes required for heart development (Bodmer,1993; Lyons et al.,1995). The GATA factors, Drosophila Pannier (Pnr) and its mammalian homologues GATA4, GATA5 and GATA6, are also expressed in the cardiogenic mesoderm and play crucial roles in heart development(Alvarez et al., 2003; Klinedinst and Bodmer, 2003; Molkentin et al., 1997; Gove et al., 1997; Reiter et al., 1999). Pannier and GATA4 function as partners of Tinman and Nkx2.5, respectively, to activate the cardiac gene program in Drosophila and mammals(Gajewski et al., 1999; Lee et al., 1998). Another group of GATA factors, Drosophila Serpent (Srp), and its mammalian homologues GATA1, GATA2 and GATA3 are required for hematopoiesis in Drosophila and mammals, respectively(Lebestky et al., 2000; Mandal et al., 2004; Tsai et al., 1994; Ting et al., 1996; Ferreira et al., 2005). It is likely that the functions of Tinman, Pannier and Serpent in cardiogenesis and hematopoiesis reflect the highly conserved but simplified developmental processes in Drosophila compared with mammals.

The basic helix-loop-helix (bHLH) transcription factor HAND is the only transcription factor known to be specific to the three major embryonic cell types that comprise the Drosophila circulatory system (Kolsh and Paululat, 2002; Han and Olson,2005). In a recent study (Han and Olson, 2005), we showed that cardiac and hematopoietic expression of Hand is controlled by a 513 bp enhancer that integrates the activity of Tinman, Pannier and Serpent, the three central transcription factors that control cardiogenesis and hematopoiesis. Hand expression is activated by Tinman and Pannier in cardioblasts and pericardial nephrocytes in the heart and by Serpent in hematopoietic progenitors in the lymph gland,through evolutionarily conserved DNA-binding sites in this enhancer. These findings place Hand at a nexus of the transcriptional networks that govern cardiogenesis and hematopoiesis, but the potential functions of Hand in these developmental processes have not been explored.

By contrast, the functions of the two vertebrate Hand genes, Hand1and Hand2, have been intensively studied. Hand1 and Hand2 are initially expressed throughout the cardiogenic region but later display complementary expression patterns in the left and right ventricular chambers (Srivastava et al.,1995; Srivastava et al.,1997; Thomas et al.,1998). Mice lacking Hand1 die from placental and extra-embryonic abnormalities (Firulli et al., 1998), whereas mice lacking Hand2 die from right ventricular hypoplasia and vascular defects(Srivastava et al., 1995; Srivastava et al., 1997; Yamagishi et al., 2001). Deletion of the Hand1 and Hand2 genes in the heart revealed their dose-sensitive requirement and functional redundancy for myocardial growth (McFadden et al.,2005), and mutation of the single hand gene in zebrafish results in a dramatic reduction in the number of cardiac cells(Yelon et al., 2000). In addition to its cardiac expression, Hand1 is highly expressed in the lateral plate mesoderm (Firulli et al.,1998) from which the intra-embryonic aorta-gonad-mesonephros(AGM), a major source of hematopoiesis, is derived(Medvinsky and Dzierzak,1996). The potential functions of Hand genes in hematopoiesis have not been investigated.

Although HAND factors are essential in vertebrate cardiogenesis, little is known about their mechanism of action. The requisite role of HAND factors in growth of the cardiac chambers during vertebrate heart development also raises interesting questions about the function of the highly conserved Drosophila Hand gene, as the Drosophila heart is thought to be a simple linear tube that does not undergo complex morphogenic changes.

Here, we show that Drosophila Hand functions as a highly potent transcriptional activator, and converting it into a transcriptional repressor blocks heart and lymph gland formation. To explore the possible roles of Hand in cardiogenesis and hematopoiesis, we generated a null mutant in the gene through homologous recombination. Hand mutant embryos and larvae displayed profound cardiac defects, including hypoplastic myocardium, a deficiency of pericardial nephrocytes, and abnormal cardiac morphology,suggesting an essential role of Hand during Drosophilacardiac morphogenesis. Lymph gland hematopoietic progenitors were also dramatically reduced in most Hand mutant larvae, as well as in a subset of Hand mutant embryos, indicating an important role of Hand in Drosophila hematopoiesis. These abnormalities were prevented by cardiac expression of Drosophila or human Hand genes, as well as the caspase inhibitor P35. These findings demonstrate evolutionarily conserved roles of Hand genes in Drosophila and mammalian cardiogenesis, and suggest a possible requirement of Hand genes in mammalian hematopoiesis.

Overexpression of transgenes was accomplished by using the Gal4-UAS system(Brand and Perrimon, 1993). The Hand-GFP and Hand-Gal4 flies were generated using the HCH enhancer identified in previous study (Han and Olson,2005). The UAS-Hand, UAS-Hand-EnR and UAS-Hand-VP16 transgenic flies were generated by cloning the corresponding DNA fragments into the pUAST vector. Transgenic constructs were injected according to standard procedures. Germline-transformed transgenic flies were selected by red eye color(w⁺) and maintained as homozygotes. At least four independent transgenic lines were analyzed for each construct.

Ends-out homologous recombination at the Hand locus was performed as described (Gong and Golic,2003). Briefly, two arms obtained from genomic PCR were subcloned into the pw25 vector to create the construct Hand^ko-pw25 for generating transformants. Transgenic flies bearing this construct on chromosome 3 were crossed to yw; 70FLP, 70I-SceI, Sco/CyO flies and the progeny (F1) were heat-shocked at 37°C for 1 hour on days 3, 4 and 5 after egg laying. Virgins carrying Hand^ko-pw25 and 70FLP,70I-SceI, Sco of the F1 progeny were crossed to w¹¹¹⁸males. About 0.04% of the F2 progeny appeared with red-eyes (8 out of∼20,000 flies). Six independent lines showed translocation of the w⁺ marker from the 3rd chromosome to the 2nd chromosome,where Hand is located. Targeted homologous recombination was verified by genomic PCR and sequencing in four independent lines. Primer information is available on request.

Embryos from different lines were collected and stained with various antibodies as previously described (Han et al., 2002). The following primary antibodies were used: mouse anti-β-galactosidase 1:300 (Promega); rat anti-Eve 1:200 (from D. Kosman); rabbit anti-Tinman 1:500 (from R. Bodmer); rabbit anti-Mef2 1:1000(from B. Paterson); rabbit anti-GFP 1:2000 (Abcam); rabbit anti-Srp 1:500(from R. Reuter). Cy2, Cy3, Cy5 or Biotin-conjugated secondary antibodies(from Jackson Lab) were used to recognize the primary antibodies. Images were obtained with a Zeiss LSM510-meta confocal microscope or a Leica DMRXE compound microscope. Three-dimensional reconstruction of the Drosophila heart was carried out using the Volocity 2.6 graphic software from Improvision.

Cell transfection and luciferase assays were performed as described(Han et al., 2004). Reporter plasmid (100 ng) and 100 ng of each activator plasmid were used. The L8E6-luciferase and L8G4-luciferase were generated by cloning six tandem copies of E-box binding sites or four tandem copies of Gal4-binding sites,respectively, into the pGL3 vector (Promega). Luciferase activities are expressed as mean±s.d. from three experiments.

To begin to understand the mechanism of action of HAND, we performed structure/function studies of the HAND protein in Drosophila S2 cells. Members of the HAND family transcription factors share homology in a bHLH domain and a 15 amino acid peptide at their C termini, referred to as the HAND domain, which is unique to this subfamily of bHLH proteins. bHLH transcription factors bind to a conserved DNA-binding site called an E-box(CANNTG). We tested the transcriptional activity of a series of Hand deletion mutants using a luciferase reporter linked to six copies of the E-box sequence(L8E6-luc) (Fig. 1A). Drosophila Hand was a remarkably effective transcriptional activator(Fig. 1A,B). Mutation of the conserved residues in the basic domain (RRR), or deletion of either the N-terminal region or the C-terminal HAND domain, abolished transcriptional activity (Fig. 1A,B),indicating the central bHLH domain cooperates with the latter domains to activate transcription.

The transcription activation domain of Hand was mapped by fusing regions of the protein to the Gal4 DNA binding domain and assaying activity with a luciferase reporter linked to four copies of the Gal4-binding site (L8G4-luc)(Fig. 1C). We found that the transcriptional activity of Hand depends primarily on its N-terminal region(Fig. 1C). Interestingly,mutation of the conserved basic residues in the bHLH domain increased the transcriptional activity of Gal4-Hand dramatically(Fig. 1C, Gal4-Hand-RRR),suggesting that the basic region communicates, directly or indirectly, with the transcription activation domain.

To determine if Hand also functions as a transcription activator in vivo,we converted it to repressor and a super-activator by fusing it to the Engrailed repression domain (EnR) and the VP16 transcription activation domain, respectively. Hand-VP16 functioned as an extremely strong transcriptional activator (Fig. 1B), whereas Hand-EnR, when co-expressed with HAND, efficiently blocked the activity of Hand in Drosophila S2 cells(Fig. 1D).

Using the UAS-Gal4 system, we overexpressed wild-type HAND, Hand-VP16 and Hand-EnR in Drosophila embryos. Pan-mesodermal overexpression of Hand had no effect on embryonic heart or muscle development(Fig. 2A,C, data not shown),although it resulted in lethality at the late larval stage for reasons that are unclear. By contrast, pan-mesodermal over-expression of Hand-EnR dramatically disrupted embryonic heart and lymph gland formation(Fig. 2B,D). The number of cardioblasts (labeled by Mef2 antibody) and pericardial cells (labeled by Even-skipped antibody) was significantly reduced in embryos with ectopic Hand-EnR (Fig. 2B,D). The residual cardiac cells in Hand-EnR-expressing embryos were able to migrate to the dorsal midline at the end of embryogenesis, but their alignment was disrupted (Fig. 2D). Formation of the lymph gland hematopoietic progenitors, labeled by Odd-skipped antibody(Fig. 2C), was also completely blocked by ectopic Hand-EnR (Fig. 2D).

To examine further the cell-autonomous requirement of Hand function within the dorsal vessel, we overexpressed wild-type Hand, Hand-VP16 and Hand-EnR using a Hand-Gal4 driver generated by using the Hand cardiac and hematopoietic enhancer (HCH) identified in our previous study(Han and Olson, 2005). Targeted overexpression of wild-type Hand and Hand-VP16 in Hand-expressing cells did not evoke a phenotype(Fig. 2E,G; data not shown),whereas targeted overexpression of Hand-EnR in Hand-expressing cells abolished the formation of lymph gland hematopoietic progenitors, labeled by antibody against the hematopoietic GATA factor Serpent and Hand-GFP, which is a transgene carrying a GFP reporter driven by the Hand cardiac and hematopoietic (HCH) enhancer identified previously(Han and Olson, 2005)(Fig. 2F,H). The number of cardioblasts and pericardial nephrocytes was also diminished and their alignment was disrupted in embryos expressing Hand-EnR(Fig. 2F,H). These data suggest that Hand functions as an essential transcriptional activator during cardiogenesis and hematopoiesis.

To examine the functions of Hand in vivo, we generated a null mutant of the gene by replacing it with a mini-white gene using the ends-out homologous recombination technology(Fig. 3A)(Gong and Golic, 2003). We obtained five independent homozygous lethal lines with a trans-location of the mini-white gene from the 3rd chromosome where it was originally located to the 2nd chromosome where the Hand gene resides. Four out of these five lines failed to complement a deficiency line that deletes the Hand locus (BL-7819). RT-PCR from homozygous mutant larvae from these four independent lines, identified by the absence of a GFP-positive balancer chromosome, showed a loss of Hand transcripts(Fig. 3B). Handtranscripts were also undetectable by in situ hybridization of homozygous Hand mutant embryos, identified by the absence of aβ-Gal-positive balancer chromosome(Fig. 3C), further demonstrating that the Hand mutation resulted in a null allele. Sequencing of genomic PCR products demonstrated that expected homologous recombination occurred identically in these four independent mutant lines.

Most homozygous Hand mutants, identified by the absence of a GFP-positive balancer chromosome, died during late embryonic and early larval stages. About 40% of the homozygous mutant embryos failed to hatch. The remaining 60% of mutant embryos hatched as 1st-instar larvae, but the majority died within 24 hours of hatching. All Hand mutant larvae were less active and smaller than normal. A small number of escapers (∼3%) survived for a few days after hatching, but they were sluggish and remained as small as 1st-instar larvae.

Approximately 20% of Hand mutant embryos showed a range of cardiac morphological defects that included discontinuities and irregularities in the architecture of the heart tube, shown by the misalignment of Mef2-expressing cardioblasts (Fig. 4D,F),reduced numbers of pericardial nephrocytes, shown by Odd-skipped (Odd)expression, and random gaps in expression of the secreted extracellular matrix protein Pericardin (Fig. 4E,F). A small subset of mutant embryos (∼3%) showed more severe cardiac defects characterized by a significant reduction of Mef2-expressing cardioblasts,Odd-expressing pericardial cells and Pericardin expression(Fig. 4G-I). In addition, the number of lymph gland hematopoietic cells was reduced in more than half of Hand mutant embryos. In many of these mutants, the lymph gland cell clusters labeled by Odd antibody were completely absent, whereas the ring gland, which is located anterior to the lymph gland and is labeled by the Pericardin antibody, was intact (Fig. 4D-F).

About 80% of Hand mutant embryos showed normal embryonic heart development and 60% of Hand mutants hatched to become 1st-instar larvae. In order to examine for possible abnormalities in larval cardiac morphology, we crossed the Hand-GFP transgene into the Hand mutant background. Recent work has shown that the Drosophila heart undergoes dramatic cardiac remodeling during late larva and early pupa development(Monier et al., 2005). However, little is known about the cardiac morphological changes during the early larval stages because of the lack of markers of the living heart and the inaccessibility of antibodies at larval stages. The Hand-GFP transgene strongly labels the entire heart from embryos to adults, providing an opportunity to examine the cardiac morphological changes during the late embryo and early larva transition by confocal microscopy. At 18 hours after egg laying (AEL), cardioblasts and pericardial cells were well aligned at the dorsal midline in wild type and a majority of Hand mutants(Fig. 5A,D). The number of lymph gland hematopoietic cells flanking the anterior aorta was largely reduced in most Hand mutants (Fig. 5D). At around 20 hours AEL, cardioblasts and pericardial cells in wild-type larvae no longer aligned in perfect rows, as the cardioblasts started to form the heart tube and the pericardial nephrocytes started to migrate to their final positions around the heart tube(Fig. 5B). A subset of Hand mutants started to show defects around this time with a reduced number of pericardial cells and thinner heart tube(Fig. 5E). The cardiac morphological defects of Hand mutants became more significant around 24 hours AEL, when 1st-instar larvae hatched from the cuticle. In wild-type 1st-instar larvae, a chamber-like structure was seen in the posterior heart and the size of the pericardial nephrocytes was significantly enlarged(Fig. 5C). By contrast, most newly hatched Hand mutant 1st-instar larvae displayed a hypoplastic heart with an abnormally thin heart tube and further reduced numbers of pericardial cells, as well as gaps in the posterior heart tube(Fig. 5F). Higher magnitude confocal scans showed the lymph gland cell clusters flanking the anterior opening of the aorta (Fig. 5G,I), and the three-dimensional structures of the posterior heart(Fig. 5H,J). In wild-type 1st-instar larvae, the posterior heart tube formed two chamber-like structures flanked by two pairs of ostias and the highly organized posterior heart tip(Fig. 5H). By contrast, the lymph gland was completely absent or largely reduced in most Handmutant 1st-instar larvae (Fig. 5I). The three-dimensional chamber-like structure of the posterior heart was also dramatically disrupted in Hand mutant larvae(Fig. 5J). Most pericardial nephrocytes were also missing at 26 hours AEL(Fig. 5J).

To determine whether ectopic cell death might account for the loss of lymph gland hematopoietic progenitors and pericardial nephrocytes in Handmutants, we examined apoptosis in Hand mutant embryos by TUNEL labeling (Fig. 6). Occasional TUNEL-positive cells could be observed around the heart in 16 hour AEL wild-type embryos (Fig. 6A,B). By contrast, ectopic apoptotic cells were found in regions normally occupied by lymph gland hematopoietic progenitors and pericardial cells in more than 30% of Hand mutant embryos (Fig. 6C,D). TUNEL-positive cells were also found among the cardioblasts in a subset of Hand mutant embryos(Fig. 6D). These data suggest that Hand is required for the survival of cardioblasts, pericardial cells and lymph gland hematopoietic progenitors.

To test whether inhibiting apoptosis in the lymph gland and hearts of Hand mutants might rescue the Hand mutant phenotypes, we overexpressed the apoptosis inhibitor P35(Clem et al., 1991), which prevents cell death by inactivating effector caspases (reviewed by Goyal, 2001), in the heart using Hand-Gal4. P35 has been shown to be an efficient caspase suppressor in Drosophila cells (Hay et al.,1994). Targeted expression of P35 in Hand-expressing cells alone did not evoke any phenotypes, whereas targeted expression of P35 in Hand mutant embryos prevented ectopic apoptosis(Fig. 6E), as well as the phenotype of reduced lymph gland hematopoietic progenitors and pericardial nephrocytes in late stage embryos (Fig. 6F). Targeted overexpression of P35 also delayed but did not prevent the larval lethality in Hand mutants. At 18 hours AEL, Hand mutant larvae with targeted P35 expression started to display an abnormal appearance (Fig. 6G). At 24 hours AEL, these larvae developed thin hypoplastic heart and reduced lymph gland hematopoietic progenitors similar to, but less severe than, that of Hand mutant larvae (Fig. 6H, compare with Fig. 5F).

To confirm that the phenotypes of the Hand null mutant were due solely to the absence of Hand, we specifically overexpressed wild-type Hand in Hand mutants using Hand-Gal4, and were able to completely rescue the phenotype and lethality of Hand mutants(data not shown). We also expressed human HAND2 in Drosophila Hand mutants using Hand-Gal4. Control experiments showed that transgenic expression of human HAND2 in wild-type flies caused no abnormalities. Remarkably, expression of human HAND2 in the Hand mutant background effectively rescued the cardiac and lymph gland defects, such that almost all mutant embryos hatched and developed to 1st-instar larvae with nearly normal hearts and lymph glands (Fig. 6I). Hand mutant larvae rescued by targeted expression of human HAND2 survived up to 6 days and developed a fairly normal heart and lymph gland at 24 hours AEL (Fig. 6J), suggesting an evolutionary conserved role of HAND factors in cardiogenesis and hematopoiesis.

The simple but highly organized Drosophila heart provides a powerful model with which to study the genetic network that controls cardiogenesis. The conservation of this genetic network between Drosophila and mammals has facilitated exploration of signaling pathways and transcription factors that are involved in this delicate process. However, such conservation has generally been thought to be limited to the early steps in cardiac cell fate specification and differentiation. Using a Hand-GFP transgene, we were able to visualize the detailed structure and function of the heart in living Drosophila embryos and larvae,revealing that the larval heart undergoes extensive morphological changes and forms a chamber-like structure. These findings suggest possible conservation between Drosophila and mammalian cardiogenesis beyond the initial steps of cardiac specification and differentiation. Consistent with this notion, the Drosophila Hand gene is required for cardiac growth and morphogenesis, functions reminiscent of those of the vertebrate Hand genes.

HAND1 and HAND2 have been shown to play essential roles the processes of cardiac remodeling and chamber specification during mammalian cardiogenesis. As the Drosophila heart has generally been considered to function as a linear tube, without a defined chamber, the function of the single highly conserved HAND factor in Drosophila has been a source of curiosity. Our results show that a substantial fraction of Hand mutant larvae display cardiac morphological defects, including a thin hypoplastic heart tube and dramatically reduced pericardial nephrocytes, as well as disruption of the chamber-like structure. Hand mutant larvae also displayed abnormal cardiac function, reflected by their sluggish heart rate and more frequent discontinuities between continuous periods of heart beating, which could be the cause of lethality after hatching (data not shown). These findings suggest that Hand plays an essential role in Drosophila heart development.

HAND transcription factors are expressed during heart development in human,mouse, chick, frog, zebrafish, ciona and Drosophila embryos (Cserjesi et al., 1995; Srivastava et al.,1995; Angelo et al.,2000; Yelon et al.,2000; Davidson and Levine,2003; Han and Olson,2005). Mouse Hand2 and Drosophila Hand are both regulated by GATA factors during heart development(McFadden et al., 2000; Han and Olson, 2005). Functional studies have suggested that Hand genes are essential for cardiogenesis in mouse, chick, zebrafish and Drosophila(Srivastava et al., 1995; Srivastava et al., 1997; Yelon et al., 2000; McFadden et al., 2005) (this study). The finding that cardiac expression of human HAND2 can rescue the early larval cardiac and hematopoietic phenotype of the Drosophila Hand mutant provides strong evidence that Hand genes play evolutionarily conserved roles in cardiogenesis.

Mouse embryos lacking HAND2 exhibit hypoplasia of the right ventricle and pharyngeal arches and associated apoptosis(Srivastava et al., 1997; Thomas et al., 1998; Yamagashi et al., 2001). Loss of the apoptosis protease-activating factor 1 (Apaf1), a downstream mediator of mitochondrial-induced apoptosis, partially rescues the ectopic apoptosis in Hand2-null embryos and delays embryonic lethality (Aiyer et al., 2005),suggesting that HAND2 acts, at least in part, to inhibit apoptosis.

We also observed ectopic apoptosis in Hand mutant Drosophila embryos, accompanied by a dramatic reduction in pericardial nephrocytes and gaps in the cardiac tube (indicative of missing cardioblasts). Interestingly, both the ectopic apoptosis and the early cardiac and hematopoietic defects could be rescued by targeted expression the apoptosis inhibitor P35 in Hand-expressing cells, indicating that one of the important roles of Hand is to inhibit apoptosis.

To determine if Hand can generally inhibit apoptosis, we tested if overexpression of Hand in transfected Drosophila S2 cells could block apoptosis induced by genes that induce apoptosis, such as Reaper and HID (Vucic et al., 1998),or with drugs that induce apoptosis, such as Etoposide and Taxol(Fang et al., 1998). However,Hand failed to inhibit apoptosis in response to these stimuli (data not shown), suggesting that it does not function as a general inhibitor of apoptosis. The fact that targeted overexpression of P35 could not completely rescue the cardiac morphological defects in Hand mutant larvae also suggests that Hand performs functions in addition to inhibiting apoptosis. It is possible that Hand could control differentiation of the cardiac and lymph gland cells and the absence of Hand would lead to apoptosis indirectly as a result of its role in some differentiation event.

Although Hand family genes have been identified for a long time, their mechanism of action has not been fully elucidated. The results of this study demonstrate Drosophila Hand to be a potent transcriptional activator in vitro and during heart and lymph gland development in vivo. Converting Hand into a transcription repressor evokes more severe cardiac and hematopoietic defects than simply removing it, suggesting that its function depends on the activation of its downstream target genes. Based on the phenotypes resulting from Hand mutants and from overexpression of Hand-EnR, we predict that these target genes participate in cell growth and survival and in maintaining cardiac and hematopoietic cell fates. Given the functional redundancy among Hand genes in mammals, Drosophila offers a powerful system with which to uncover conserved functions and mechanisms of action of this gene family in both cardiogenesis and hematopoiesis.

In Drosophila, adult blood cells originate from the lymph gland hematopoietic progenitors, which are derived from cardiac mesoderm. The lymph gland dissociates at the pupal stages to release all the adult blood cells. Hand is the only transcription factor identified to date that is expressed in all hematopoietic progenitors and the entire heart. The dramatic reduction of lymph gland hematopoietic progenitors in Hand mutants suggests that Hand is essential for Drosophila hematopoiesis.

In mammals, the adult hematopoietic system originates from the yolk sac and the intra-embryonic aorta-gonad-mesonephros (AGM) region(Medvinsky and Dzierzak,1996). Previous studies have suggested a close relationship between the Drosophila cardiac mesoderm and the mammalian cardiogenic and AGM region (Evans et al.,2003; Mandal et al.,2004). In both Drosophila and mammals, the specification of these regions requires the input of BMP, WNT and FGF signaling from the neighboring germ layer and function of NK and GATA factors in the mesoderm(Cripps and Olson, 2002; Evans et al., 2003). Although the possible role of HAND factors in mammalian hematopoiesis has not been explored, mouse Hand1 is expressed at high levels in the lateral plate mesoderm, from which the cardiogenic region and the AGM region arise(Firulli et al., 1998). Our study provides the first evidence for the requirement of Hand in Drosophila hematopoiesis, suggesting similar functions for its mammalian orthologs.

We thank the Bloomington stock center for fly stocks, B. Paterson and the University of Iowa Hybridoma Bank for antibodies, J. Wu for technical help and A. Tizenor for graphics. Z.H. was supported by a National Scientist Development Grant from The American Heart Association and E.N.O. was supported by grants from The National Institutes of Health and the Donald W. Reynolds Cardiovascular Clinical Research Center (Dallas, Texas), and the Robert A. Welch Foundation.