The Drosophila melanogaster genes midline and H15 encode predicted T-box transcription factors homologous to vertebrate Tbx20 genes. All identified vertebrate Tbx20 genes are expressed in the embryonic heart and both midline and H15 are expressed in the cardioblasts of the dorsal vessel, the insect organ equivalent to the vertebrate heart. The midline mRNA is first detected in dorsal mesoderm at embryonic stage 12 in the two progenitors per hemisegment that will divide to give rise to all six cardioblasts. Expression of H15 mRNA in the dorsal mesoderm is detected first in four to six cells per hemisegment at stage 13. The expression of midline and H15 in the dorsal vessel is dependent on Wingless signaling and the transcription factors tinman and pannier. The selection of two midline-expressing cells from a pool of competent progenitors is dependent on Notch signaling. Embryos deleted for both midline and H15 have defects in the alignment of the cardioblasts and associated pericardial cells. Embryos null for midline have weaker and less penetrant phenotypes while embryos deficient for H15 have morphologically normal hearts, suggesting that the two genes are partially redundant in heart development. Despite the dorsal vessel defects, embryos mutant for both midline and H15 have normal numbers of cardioblasts, suggesting that cardiac cell fate specification is not disrupted. However, ectopic expression of midline in the dorsal mesoderm can lead to dramatic increases in the expression of cardiac markers, suggesting that midline and H15 participate in cardiac fate specification and may normally act redundantly with other cardiogenic factors. Conservation of Tbx20 expression and function in cardiac development lends further support for a common ancestral origin of the insect dorsal vessel and the vertebrate heart (Miskolczi-McCallum, 2005).

A second study by Qian (2005) provides more detailed information on a cellular fate switch accompanying loss- and gain-of-function of this gene pair. Referring to midline and H15 by the alternative name neuromancer (nmr) the Qian study shows that gene function causes a switch in cell fates in the cardiogenic region, in that the progenitors expressing the homeobox gene even skipped (eve) are expanded, accompanied by a corresponding reduction of the progenitors expressing the homeobox gene ladybird (lbe). As a result, the number of differentiating myocardial cells is severely reduced whereas pericardial cell populations are expanded. Conversely, pan-mesodermal expression of nmr represses eve, while causing an expansion of cardiac lbe expression, as well as ectopic mesodermal expression of the homeobox gene tinman. In addition, nmr mutants with less severe penetrance exhibit cell alignment defects of the myocardium at the dorsal midline, suggesting nmr is also required for cell polarity acquisition of the heart tube. In exploring the regulation of nmr, it was found that the GATA factor Pannier is essential for cardiac expression, and acts synergistically with Tinman in promoting nmr expression. Moreover, reducing nmr function in the absence of pannier further aggravates the deficit in cardiac mesoderm specification. Taken together, the data suggest that nmr acts both in concert with and subsequent to pannier and tinman in cardiac specification and differentiation. It is proposed that nmr is another determinant of cardiogenesis, along with tinman and pannier (Qian, 2005).

The insect heart, or dorsal vessel, is a simple muscular tube that pumps hemolymph and hematocytes through the open circulatory system. Heart development in vertebrates and dorsal vessel development have many features in common. During gastrulation, the trunk mesoderm of the fly invaginates along the ventral midline. The mesoderm flattens, forming a single cell layer inside the ectoderm, and migrates dorsally. The fly heart forms from bilateral rows of cells at the leading edge of the migrating mesoderm. The rows fuse along the dorsal midline forming a linear tube. A similar process initiates vertebrate heart development, where bilateral fields of mesoderm migrate and fuse to form a linear heart tube prior to the looping and subsequent morphogenesis that forms a multichambered heart. In addition to these morphological similarities, signaling from other germ layers is required for the cardiac potential of the mesoderm tube and a conserved transcription factor network specifies cardiac fate in both Drosophila and vertebrates (Miskolczi-McCallum, 2005).

The Drosophila NK-2 class homeodomain transcription factor Tinman is required for the specification of the dorsal mesoderm [heart development is completely abolished in tinman mutants]. After mesoderm involution, tin is initially expressed in all mesoderm cells under the control of the bHLH factor Twist. During stages 9 and 10, tin becomes restricted to a broad domain in the dorsal trunk mesoderm through maintenance by signaling from overlying ectoderm, which secretes the BMP family protein Decapentaplegic (Dpp). By mid stage 11, tin expression is maintained in segmental clusters of dorsal mesoderm cells by ectoderm and mesoderm secreting the Wnt protein Wingless (Wg). These tin-expressing domains contain the precursors for the heart including progenitors of both the contractile cardioblasts and the noncontractile pericardial cells. The dorsal vessel is not specified in the absence of tin function or if Dpp or Wg signal transduction is blocked (Miskolczi-McCallum, 2005 and references therein).

Tin regulates cardiac gene expression in conjunction with the GATA factor Pannier (Pnr). In stage 11, pnr is expressed in the mesoderm under the control of ectodermal Dpp signaling and mesodermal Tin. Null mutations in pnr greatly diminish heart development, and ectopic pnr expression is sufficient to induce excess cardiac fate. Expression of pnr in the mesoderm is also required to maintain proper tin expression. Ectopic expression of tin has little effect on heart formation, but in combination with pnr, ectopic tin has a synergistic effect on the ability of mesoderm to adopt cardiac fate. This synergy is likely to occur at the level of the regulatory sequences of cardiac genes as Tin and Pnr form complexes in vitro and heart-specific enhancers require clusters of Tin and Pnr binding sites (Miskolczi-McCallum, 2005 and references therein).

NK-2 homeodomain and GATA transcription factors are also critical for heart development in vertebrates. Loss of function for Nkx2.5, the vertebrate tin homologue, causes arrested heart development at the linear tube stage. Furthermore, Nkx2.5 activates cardiac gene expression in a complex with GATA4. Genetic analysis of the roles of GATA factors in vertebrate heart development is complicated by potential genetic redundancy of GATA4, 5, and 6, but several lines of evidence suggest that GATA factors are also required to establish cardiac fate (Miskolczi-McCallum, 2005).

More recently, T-box transcription factors have been identified as regulators of vertebrate heart development that interact with both Nkx2.5 and GATA proteins. Tbx5 (Drosophila homolog: Optomotor-blind), the gene responsible for Holt-Oram syndrome, is required for chamber formation. Mutations in Tbx5, Nkx2.5, and GATA4 all cause similar heart defects in humans and Tbx5 interacts physically with both Nkx2.5 and GATA4. Mutations in GATA4 that abrogate the Tbx5-GATA4 physical interaction in vitro cause heart defects similar to those seen in Tbx5 mutants, indicating that the physical interactions are functionally important. All pairwise combinations of Tbx5, Nkx2.5, and GATA4 activate gene expression synergistically in vitro. The cardiac expression of another T-box family member, Tbx20 (also called Tbx12), is conserved in all vertebrates (Ahn, 2000; Carson, 2000; Griffin, 2000; Iio, 2001; Kraus, 2001; Meins, 2000). Zebrafish embryos exhibit heart defects following injection with morpholino oligonucleotides directed against H15-related-T-box (hrT), a zebrafish Tbx20 homolog (Szeto, 2002). Tbx20 shares some of the properties of Tbx5 in vitro, forming complexes with Nkx2.5 and GATA4 (Plageman, 2004; Stennard, 2003) and synergistically activating cardiac gene expression in some assays (Stennard, 2003). However, the relationship between Tbx20 and Tbx5 may be complex: Tbx20 acts to inhibit Tbx5 transcription in vivo (Szeto, 2002; Takeuchi, 2003), and in some assays (Plageman, 2004) can block transcriptional activation by Tbx5 in vitro (Miskolczi-McCallum, 2005).

Clearly, T-box proteins are key regulators of heart development in vertebrates, which raises the question of whether T-box function is conserved in Drosophila dorsal vessel development. T-box genes have duplicated and diverged throughout metazoan evolution, acting in several developmental processes in a wide variety of organisms. In the human and mouse genomes, there are 18 identified T-box genes representing five subfamilies, while the fly genome contains eight T-box genes representing four of the five subfamilies found in mice and humans. There is no clear Tbx5 homolog in Drosophila, and the Drosophila gene most similar to Tbx5, optomotor-blind (omb), is not expressed in the dorsal vessel. Dorsocross 1, 2, and 3 (Doc1, 2, and 3), the Tbx6 homologs, are expressed in a subset of cardioblasts. The Doc genes play a key role in the subdivision of the dorsal mesoderm, but it has not been shown if they play a role specifying cardiac fate and the expression of Tbx6 is not conserved in vertebrate hearts. Therefore, the best candidates for Drosophila cardiogenic T-box genes are the Tbx20 homologs H15 (Brook, 1996) and midline (mid) (Buescher, 2004). H15-lacZ, a lacZ enhancer trap inserted adjacent in the H15 mid locus, is expressed in all cardioblasts (Griffin, 2000), the contractile myocardial cells of the dorsal vessel (Miskolczi-McCallum, 2005).

In order to assess whether the requirement for T-box transcription factors is conserved in heart and dorsal vessel development, the expression, regulation, and function of mid and H15 in the development of the Drosophila dorsal vessel were examined. The expression of the genes was shown to correlate with the specification of cardioblast fate -- the genes are required in a partially redundant fashion in heart morphogenesis, and ectopic expression of mid is sufficient to cause dorsal mesoderm cells to acquire cardiac fate. Based on these results, it is argued that Tbx20 genes represent another component of the conserved transcription factor network acting in heart and dorsal vessel development (Miskolczi-McCallum, 2005).

Therefore, the Tbx20 homologs mid and H15 are expressed in cardioblasts and are required in a partially redundant manner for dorsal vessel morphogenesis. Furthermore, the activation of mid in the dorsal mesoderm depends on cardiogenic factors and mid expression correlates with cardioblast specification. Finally, ectopic expression of mid induces cardiac fate in the dorsal mesoderm in a manner similar to the effects of ectopic tin and pnr expression. Together, these results suggest an important role for mid and H15 in the specification and differentiation of cardiac fate. Vertebrate T-box transcription factors are critical regulators of heart development. Tbx20 is essential for morphogenesis of the vertebrate heart and interacts with the Tin and Pnr homologs Nkx2.5 and GATA4 in gene regulation (Plageman, 2004; Stennard, 2003; Szeto, 2002). Thus, the previously reported conservation of a cardiac regulatory network, including the NK-2 homeodomain proteins and the GATA factors, may now be extended to include T-box transcription factors. These findings offer further support for the proposal that, despite differing greatly in morphology, the vertebrate heart and the insect dorsal vessel may share a common evolutionary ancestor (Miskolczi-McCallum, 2005).

The timing and pattern of mid expression suggest that it is an early specific marker of cardioblasts. mid expression in the dorsal mesoderm is first observed in two cells per hemisegment in early stage 12, increasing to six cardioblasts per hemisegment throughout stages 12 and 13. Analysis of the cardioblast lineage with genetic mosaics and cell cycle manipulation indicates that the six cardioblasts in each hemisegment must derive from two progenitor cells. Furthermore, Notch signaling is required to restrict the number of cardioblasts in the dorsal vessel, and it was found that the number of cells expressing mid in the dorsal mesoderm is also restricted by Notch activation. Thus, these data are consistent with a model where mid-expressing cardioblast progenitors are normally selected from a cluster of competent cells by a signal or signals acting in opposition to Notch. The nature of this signal is unknown, although it is noted that the two mid-expressing cells arise in consistent positions relative to the ectodermal Wg and Engrailed stripes of cells and thus may respond to spatial cues emanating from the overlying ectoderm or arising from within the mesoderm (Miskolczi-McCallum, 2005).

Both mid and H15 are expressed in all cardioblasts from stage 13 onwards and the dorsal vessel is disorganized in H15 mid null embryos. Analysis of mid and H15 mutant embryos suggests that mid has the greater effect, causing some heart defects on its own while H15 is nonessential. However, loss of both genes has a stronger phenotype, suggesting that H15 augments the activity of mid. A similar relative requirement is found for the two genes in the segmentation of the embryonic ectoderm (Buescher, 2004). Normal numbers of cardioblasts, expressing markers of cardiac (tin) and muscle (DMef2) fate specification, as well as muscle differentiation (Myosin Heavy Chain), are formed in H15 mid null embryos. The cardioblasts migrate dorsally but do not always align properly with their contralateral partners. The pericardial cells are often displaced laterally or medially, and the lymph gland is enlarged due to increased cell number. Given these defects and the restriction of mid and H15 expression to cardioblasts, the essential function of mid and H15 may be to regulate the later stages of cardioblast morphogenesis (Miskolczi-McCallum, 2005).

Understanding the essential function of mid and H15 in dorsal vessel development awaits identifying the genes regulated by the two T-box transcription factors. Cardioblasts require specific cell polarity, adhesion, and migration in order to make the proper contacts with other cardioblasts and pericardial cells to form the tube of the dorsal vessel. Many cell adhesion molecules are expressed in cardioblasts and mutations in these genes cause dorsal vessel defects that are similar to those caused by loss of mid and H15. The misalignment of pericardial cells may be a secondary consequence of the defects in cardioblast morphogenesis. Similarly, the increase in lymph gland cell number is likely to be a nonautonomous effect since mid and H15 are not expressed in the lymph gland primordia. Because this defect results from an increase in the lymph gland cell number, it is possible that H15 mid mutant embryos are defective in a signal from the dorsal vessel regulating cell proliferation in the lymph gland (Miskolczi-McCallum, 2005).

The specification of cardiac fate in Drosophila is complex. No single factor has been identified that is both necessary and sufficient to induce mesoderm to form heart cells. Combined Dpp and Wg signaling are essential for heart development and can induce ectopic heart fate, but only in tin-expressing cells. Tin is essential for the specification of dorsal mesoderm and directly regulates many genes expressed in cardioblasts and pericardial cells. However, ectopic expression of tin has little effect on cardiac fate. By comparison, Pnr, a transcriptional cofactor of Tin, is able to induce extensive ectopic cardiac fate. However, some cardioblasts still form in pnr null embryos, albeit in greatly reduced numbers, indicating that pnr is not absolutely essential for heart cells to form. A formally similar situation is found for mid in that it is also sufficient but not necessary (even in the absence of H15) for cardiac fate. Recently, it has been shown that the Xenopus Tbx20 homolog interacts physically with GATA4 and Nkx2.5 (Plageman, 2003; Stennard, 2003). Furthermore, Tbx20 can have a synergistic effect on the activation of cardiac promoters in vitro with Nkx2.5 and GATA4 (Stennard, 2003). Thus, it is possible that Mid and H15 act as cofactors for Tin and Pnr in the regulation of cardiac fate specification in the fly. The expression of mid in the dorsal mesoderm is much more restricted than that of either tin or pnr, thus ectopic expression of mid may promote ectopic cardiac fate through interaction with Tin and Pnr. This interpretation is supported by the restriction of ectopic cardiac fate induced by mid to the dorsal mesoderm (Miskolczi-McCallum, 2005).

An interaction with Tin and Pnr may explain why mid induces ectopic cardioblasts, but it does not explain why mid and H15, or pnr for that matter, are not essential for cardioblast specification. A simple possibility is that another transcription factor may compensate for the absence of mid and H15. This could be either an unrelated factor or another T-box protein, for example, the Doc1, 2, and 3 Tbx6-like proteins, which are restricted to the svp-expressing cardioblasts. Assessing mid and H15 for redundancy with the Doc genes, not a trivial undertaking, because loss of function for Doc blocks embryonic development prior to heart morphogenesis, will be necessary to test this model. Another mechanism that may help compensate for the absence of mid and H15 is redundancy with Tin and Pnr activity. Given the pairwise synergy of GATA4, Nkx2.5, and Tbx20 in transcriptional assays (Stennard, 2003), it may be that removal of any one trans-acting factor would be insufficient to completely block cardiac fate. Either Pnr or Mid could be sufficient to induce some cardiac fate as long as Tin or other factors are present. In this context, it is worth noting that removal of tin, which has the greatest effect, completely blocking cardioblast fate specification, also prevents the expression of pnr, mid, and H15. In contrast, tin, mid and H15 are only partially reduced in pnr mutants, and tin and pnr are widely expressed in the dorsal mesoderm prior to the activation of mid and H15 expression. Analysis of genetic and molecular interactions between tin, pnr, mid, and H15 will be necessary to test these models (Miskolczi-McCallum, 2005).

GENE STRUCTURE

cDNA clone length - 2563 (H15) and 2933 (mid) bp (mid)

Bases in 5' UTR - 885 (H15) and 277 (mid)

Exons - 6 (H15) and 4 (mid)

Bases in 3' UTR - 43 (H15) and 940 (mid)

PROTEIN STRUCTURE

Amino Acids - 660 (H15) and 571 (Midline)

Structural Domains

See InterPro - Transcription factor, T-box for information on H15 and Midline structures.

date revised: 20 March 2005

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.