Variation in homeodomain DNA binding revealed by high-resolution analysis of sequence preferences

Most homeodomains are unique within a genome, yet many are highly conserved across vast evolutionary distances, implying strong selection on their precise DNA-binding specificities. This study determined the binding preferences of the majority (168) of mouse homeodomains to all possible 8-base sequences, revealing rich and complex patterns of sequence specificity and showing that there are at least 65 distinct homeodomain DNA-binding activities. A computational system was developed that successfully predicts binding sites for homeodomain proteins as distant from mouse as Drosophila and C. elegans, and full 8-mer binding profiles were inferred for the majority of known animal homeodomains. The results provide an unprecedented level of resolution in the analysis of this simple domain structure and suggest that variation in sequence recognition may be a factor in its functional diversity and evolutionary success (Berger, 2008).

It was asked whether the homeodomain monomer binding preferences identified in vitro reflect sequences preferred in vivo. Anecdotally, the highest predicted binding sequences do correspond to known in vivo binding sites. For example, in the predicted 8-mer profile for sea urchin Otx, a previously identified in vivo binding sequence (TAATCC, from the Spec2a RSR enhancer), is contained in the top predicted 8-mer sequence, and, more strikingly, it is embedded in the fifth-highest predicted 8-mer sequence (TTAATCCT). At greater evolutionary distance, three of the four Drosophila Tinman binding sites in the minimal Hand cardiac and hematopoietic (HCH) enhancer are contained within the second (TCAAGTGG), fifth (ACCACTTA), and ninth (GCACTTAA) ranked 8-mers (the fourth overlaps the 428th ranked 8-mer [CAATTGAG], but also overlaps with a GATA binding site and may have constraints on its sequence in addition to binding Tinman) (Berger, 2008).

To ask more generally whether occupied sites in vivo contain sequences preferred in vitro, six ChIP-chip or ChIP-seq data sets in the literature were examined that involved immunoprecipitation of homeodomain proteins that were analyzed, or homologs of proteins analyzed that shared at least 14 of the 15 DNA-contacting amino acids. In all cases, enrichment was observed for monomer binding sites in the neighborhood of the bound fragments, with a peak at the center. Two examples, Drosophila Caudal and human Tcf1/Hnf1 are shown. For Caudal, the size of this ratio peak increased dramatically with E score cutoff, indicating that the most preferred in vitro monomer binding sequences correspond to the most enriched in vivo binding sites (51% of bound fragments have such an 8-mer, versus 17% in randomly selected fragments). For Tcf1/Hnf1, however, the majority of sequences bound in vivo do not contain the best in vitro binding sequences, although most do contain at least one 8-mer with E > 0.45 (53%, versus 27% in random fragments), suggesting utilization of weaker binding sites. Similar results were obtained with PWMs. Thus, the requirement for highest-affinity binding sequences may vary among homeodomain proteins, species, or under different physiological contexts. Nonetheless, a large proportion of the in vivo binding events apparently involve the monomeric homeodomain sequence preferences, which can be derived in vitro (Berger, 2008).

Tinman homologs in C. elegans

Pharyngeal muscle development in the nematode Caenorhabditis elegans appears to share similarities with cardiac muscle development in other species. CEH-22 is an NK-2 class homeodomain transcription factor, similar to Drosophila tinman and vertebrate Nkx2-5, which is expressed exclusively in the pharyngeal muscles. In vitro, CEH-22 binds the enhancer from myo-2, a pharyngeal muscle-specific myosin heavy chain gene. The role played by CEH-22 in pharyngeal muscle development and gene activation was examined by (1) ectopically expressing ceh-22 in transgenic C. elegans and (2) examining the phenotype of a ceh-22 loss-of-function mutant. These experiments indicate that CEH-22 is an activator of myo-2 expression and that it is required for normal pharyngeal muscle development. However, ceh-22 is necessary for neither formation of the pharyngeal muscles, nor for myo-2 expression. These data suggest that parallel and potentially compensating pathways contribute to pharyngeal muscle differentiation. The relationship between ceh-22 and the pharyngeal organ-specific differentiation gene pha-1 was examined. Pha-1 shares weak similarity with basic leucine zipper transcription factors. Mutations in ceh-22 and pha-1 have strongly synergistic effects on pharyngeal muscle gene expression; in addition, a pha-1 mutation enhances the lethal phenotype caused by a mutation in ceh-22. Wild-type pha-1 is not required for the onset of ceh-22 expression but it appears necessary for maintaining the expression of ceh-22 (Okkema, 1997).

A new Caenorhabditis elegans NK-2 class homeobox gene (Drosophila homologs tinman and bagpipe) has been identified and designated ceh-24. CEH-24 is one of four NK-2 class homeodomain proteins in C. elegans. Distinct cis-acting elements generate a complex neuronal and mesodermal expression pattern. A promoter-proximal enhancer mediates expression in a single pharyngeal muscle, the donut-shaped m8 cell located at the posterior end of the pharynx. A second mesodermal enhancer is active in a set of eight nonstriated vulval muscles used in egg laying. Activation in the egg laying muscles requires an 'NdE-box' consensus motif (CATATG) that is related to, but distinct from, the standard E-box motif bound by the MyoD family of transcriptional activators. Ectodermal expression of ceh-24 is limited to a subset of sublateral motor neurons in the head of the animal; this activity requires a cis-acting activator element that is distinct from the control elements for pharyngeal and vulval muscle expression. Activation of ceh-24 in each of the three cell types coincides with the onset of differentiation. Using a set of transposon-induced null mutations, it has been shown that ceh-24 is not essential for the formation of any of these cells. Although ceh-24 mutants have no evident defects under laboratory conditions, the pattern of ceh-24 activity is apparently important for Rhabditid nematodes: the related species C. briggsae contains a close homolog of C. elegans ceh-24, including a highly conserved and functionally equivalent set of cis-acting control signals (Harfe, 1998).

Mesodermal development is a multistep process in which cells become increasingly specialized to form specific tissue types. In Drosophila and mammals, proper segregation and patterning of the mesoderm involves the bHLH factor Twist. The activity of a Twist-related factor, CeTwist, was investigated during Caenorhabditis elegans mesoderm development. Within the bHLH domain, CeTwist shares 59%-63% identity to published Twist family members in other species. Outside of the bHLH domain, there is no obvious homology between CeTwist and other Twist family members. Embryonic mesoderm in C. elegans derives from a number of distinct founder cells that are specified during the early lineages; in contrast, a single blast cell (M) is responsible for all nongonadal mesoderm formation during postembryonic development. Using immunofluorescence and reporter fusions, the activity pattern of the gene encoding CeTwist was determined. No activity is observed during specification of mesodermal lineages in the early embryo; instead, the gene is active within the M lineage and in a number of mesodermal cells with nonstriated muscle fates.

A role for CeTwist in postembryonic mesodermal cell fate specification was indicated by ectopic expression and genetic interference assays. These experiments show that CeTwist is responsible for activating two target genes normally expressed in specific subsets of nonstriated muscles derived from the M lineage. In vitro and in vivo assays suggest that CeTwist cooperates with the C. elegans E/Daughterless homolog in directly activating these targets. The two target genes that have been studied, ceh-24 and egl-15, encode an NK-2 class homeodomain (Drosophila homologs Tinman and Bagpipe) and an FGF receptor (FGFR) homolog, respectively. Regulation of the ceh-24 promoter provided a means to study transcriptional activation in mesodermal cells with nonstriated muscle character. Functional analysis of the upstream region had identified a 22-bp fragment that is sufficient (when concatamerized) for enhancer activity in M-derived vulval, uterine, and intestine-associated muscles. This 22-bp sequence contains two E boxes; these motifs have been named NdE boxes for the NdeI restriction enzyme that recognizes this sequence (CATATG). Further characterization of the 22-bp enhancer element was carried out by use of a detailed point mutational analysis. This analysis reveals the following: (1) the first base pairs of two NdE-box motifs are necessary for activity; (2) spacing is important for activity (i.e., an insertion of 21 bp between NdE boxes abolishes activity); (3) activity is abolished by mutations that change several base pairs flanking the NdE boxes, and (4) mutation of both NdE boxes to a consensus MyoD-binding site, CAGCTG, eliminates enhancer activity. CeTwist is shown to be capable of inducing modest expression from the integrated ceh-24 22-bp NdE box enhancer in most C. elegans tissues. Twist is known to activate FGFR and NK-homeodomain target genes during mesodermal patterning of Drosophila; similar target interactions have been proposed to modulate mesenchymal growth during closure of the vertebrate skull. These results suggest the possibility that a conserved pathway may be used for diverse functions in mesodermal specification (Harfe, 1998).

Development of pharyngeal muscle in nematodes and cardiac muscle in vertebrates and insects involves the related homeobox genes ceh-22, nkx2.5, and tinman, respectively. To determine whether the nematode and vertebrate genes perform similar functions, activity of the zebrafish nkx2.5 gene was examined in transgenic Caenorhabditis elegans. Ectopic expression of nkx2.5 in the C. elegans body wall muscle can directly activate expression of both the endogenous myo-2 gene, a ceh-22 target normally expressed only in pharyngeal muscle, and a synthetic reporter construct controlled by a multimerized CEH-22 binding site. nkx2.5 also efficiently rescues a ceh-22 mutant when expressed in pharyngeal muscle. Together, these results indicate that nkx2.5 and ceh-22 provide a single conserved molecular function. Further, they suggest that an evolutionarily conserved mechanism underlies heart development in vertebrates and insects and pharyngeal development in nematodes (Haun, 1998).

Wnt signaling regulates many aspects of metazoan development, including stem cells. In C. elegans, Wnt/MAPK signaling controls asymmetric divisions. A recent model proposed that the POP-1/TCF DNA binding protein works together with SYS-1/β-catenin to activate transcription of target genes in response to Wnt/MAPK signaling. The somatic gonadal precursor (SGP) divides asymmetrically to generate distal and proximal daughters of distinct fates: only its distal daughter generates a distal tip cell (DTC), which is required for stem cell maintenance. No DTCs are produced in the absence of POP-1/TCF or SYS-1/β-catenin, and extra DTCs are made upon overexpression of SYS-1/β-catenin. This study reports that POP-1/TCF and SYS-1/β-catenin directly activate transcription of ceh-22/nkx2.5 isoforms in SGP distal daughters, a finding that confirms the proposed model of Wnt/MAPK signaling. In addition, it is demonstrated that the CEH-22/Nkx2.5 homeodomain transcription factor is a key regulator of DTC specification. It is speculated that these conserved molecular regulators of the DTC niche in nematodes may provide insight into specification of stem cell niches more broadly (Lam, 2006).

Thus Wnt signaling and ceh-22/nkx2.5 work together to specify the DTC fate. The common function of DTCs in hermaphrodites and males is that of a stem cell niche. Wnt signaling has emerged as a key regulator of stem cells in many tissues and in many organisms, and that role relies on transcriptional activation by TCF/LEF and β-catenin transcription factors. The current work suggests that one role of Wnt signaling may be to control the stem cell niche. A similar suggestion was recently put forward with respect to osteoblasts, which provide a niche for hematopoietic stem cells. CEH-22/Nkx2.5 and its homologs have not previously been implicated in the control of stem cells. Indeed, the fly and vertebrate homologs, termed Tinman and Nkx2.5, respectively, are best known for their roles in heart specification and differentiation. Nematodes have no heart, but CEH-22 controls development of the rhythmically contracting musculature of the pharynx, and zebrafish Nkx2.5 can functionally replace CEH-22. Therefore, the CEH-22/Nkx2.5 class of homeodomain transcription factors has broadly conserved functions in animal development (Lam, 2006).

A remaining question is whether CEH-22 control of the DTC fate reflects a conserved role for this class of homeodomain transcription factors in regulating stem cell niches. Mouse mutants deleted for Nkx2.5 die with a broad spectrum of defects, including severe defects in vasculogenesis and angiogenesis as well as hematopoiesis in the yolk sac. Intriguingly, endothelial cells appear to function as stem cell niches. It is tempting to speculate that the severe vasculature defects in Nkx2.5 mutants may reflect some role of this conserved regulator in control of a vertebrate niche, much as CEH-22 controls the DTC. Two important challenges for the future are to learn how CEH-22 specifies the DTC niche in C. elegans and to learn whether its homologs specify an analogous stem cell niche in flies and vertebrates (Lam, 2006).

Other invertebrate Tinman homologs

A novel single-sided specific polymerase chain reaction (PCR) strategy inspired by ligation-mediated PCR has been used to clone fragments of divergent homeobox genes from a flatworm, the planarian Polycelis nigra. Eight homeobox-containing fragments were amplified, belonging to the Hox, msh, NK-1 and NK-2 classes. Together with the results obtained from several genomes of platyhelminths, this screening shows the presence of the same array of homeodomain developmental regulators in planarians, traditionally regarded as primitive metazoans in terms of body plan, as in coelomate organisms. However, the presence of a Ubx/abd-A homolog may indicate that platyhelminths are more closely related to protostomes than to deuterostomes and supports the idea that flatworms have inherited an elaborate HOX cluster (seven or eight genes) from their ancestor. Likely homologs of the fly genes tinman, bagpipe and S59 suggest that the mesoderm might be patterned by the same genes in all bilaterally symmetrical animals. Finally, a msh-like gene, a family known to be involved in inductive mechanisms in vertebrates, has been found. These results support the hypothesis that the tremendous diversity of metazoan body plans is specified by a largely conserved array of homeobox-containing developmental genes (Balavoine, 1996).

A full-length cDNA clone of amphioxus AmphiNk2-tin, an NK2 gene similar in sequence to vertebrate NK2 cardiac genes, was isolated, suggesting a potentially similar function to Drosophila tinman and to vertebrate NK2 cardiac genes during heart development. During the neurula stage of amphioxus, AmphiNk2-tin is expressed first within the foregut endoderm, then transiently in muscle precursor cells in the somites, and finally in some mesoderm cells of the visceral peritoneum arranged in an approximately midventral row running beneath the midgut and hindgut. The peritoneal cells that express AmphiNk2-tin are evidently precursors of the myocardium of the heart, which subsequently becomes morphologically detectable ventral to the gut. The amphioxus heart is a rostrocaudally extended tube consisting entirely of myocardial cells (at both the larval and adult stages); there are no chambers, valves, endocardium, epicardium, or other differentiated features of vertebrate hearts. Phylogenetic analysis of the AmphiNk2-tin sequence documents its close relationship to vertebrate NK2 class cardiac genes, and ancillary evidence suggests a relationship with the Drosophila NK2 gene tinman. Apparently, an amphioxus-like heart, and the developmental program directing its development, was the foundation upon which the vertebrate heart evolved by progressive modular innovations at the genetic and morphological levels of organization (Holland, 2003).

Amphioxus AmphiNk2-tin and the vertebrate NK2 cardiac genes, as well as Drosophila tinman, are expressed in the endoderm of the developing foregut. For vertebrates, mouse Nkx2-5 and Nkx2-6 expression is necessary for the development of the pharynx itself, although the pharyngeal transcription of these genes is not known to influence the development of the heart. Similarly, during amphioxus development, the endodermal expression of AmphiNk2-tin, appears to have no direct influence on cardiogenesis: the endodermal transcription of AmphiNk2-tin is at the level of the foregut during the early neurula stage, spatiotemporally separated (as in Drosophila) from the cardiogenic mesodermal expression of AmphiNk2-tin, which is at the late neurula stage at the level of the hindgut (midgut in Drosophila) (Holland, 2003).

From a broader phylogenetic perspective, endodermal expression of NK2 class genes may well be an ancient and conserved feature of animals generally, because a cnidarian member of this gene class is expressed in the gastrodermis of hydra. In the phylogenetic transition from some diploblastic ancestor (perhaps cnidarian-like) to a bilaterian animal, the mesoderm presumably arose from the gastrodermis, which, in cnidarians, comprises epitheliomuscle cells. At least part of the early mesoderm presumably included a layer of muscle cells investing the endoderm and developing under the direction of NK-2 class genes. Thereafter, some of these gut muscle cells could have evolved into pulsatory muscular vessels (i.e., hearts), still specified during development by NK2 class genes (Holland, 2003).

A second, somewhat similar, idea suggests that muscles of the nematode pharynx, which express an NK2 homolog (ceh-22), might have been the evolutionary source of the myocardium in other bilaterian animals. Nematodes are unusual in that muscle cells are not peripheral to the pharyngeal lining, but instead comprise an integral part of it. One problem with the second proposal is that modern evidence strongly indicates that nematodes are not basal bilaterians; in contrast, they are now thought to be advanced, but rather degenerate ecdysozoan protostomes. Given this, one could even reverse the evolutionary scenario and propose that muscle cells lining the nematode pharynx evolved from mesodermal muscles or cardiomyocytes in some ancestor of the nematodes. Moreover, the phylogenetic treatment of NK2 sequence data consistently places the Caenorhabditis ceh-22 sequence in a Nk2.2-vnd subfamily. Assuming that the observed absence of other NK2 genes in Caenorhabditis is correct, that the placement of nematodes in the Ecdysozoa is appropriate, then the Caenorhabditis lineage has lost three of the four NK2 subfamilies present in Drosophila and chordates. The three absent genes appear to be (1) a scarecrow/ TTF1/Nk2.1 homolog, (2) a bagpipe homolog, and (3) the cardiac/tinman homolog. If these genes are indeed absent in Caenorhabditis, then some compensation in function of the one remaining NK2 gene might be expected. Assessing which of the above perspectives regarding the NK2 gene(s) of Caenorhabditis is correct will require additional data and analysis (Holland, 2003).

NK2 class gene expression has been reported in the presumptive gut musculature of Drosophila and in skeletal muscle lineages of amphioxus and vertebrates. In Drosophila embryos, tinman expression in the visceral mesoderm (destined to become musculature peripheral to the gut lining) is transitory. During subsequent development, tinman in a sense delegates its role in gut muscle specification to bagpipe (another NK2 class homeobox gene) while remaining in control of heart muscle development (Holland, 2003).

In chordates, relatively few of the NK2 class genes are expressed in presumptive or mature skeletal muscle cells. Amphioxus AmphiNk2-tin is transiently expressed in muscle cell precursors in the embryonic somites, whereas mouse Nkx2.6 is expressed in differentiated skeletal muscles. It is reasonable to assume that AmphiNk2-tin is involved in muscle cell specification, while mouse Nkx2.6 is involved in some aspect of maintenance of the differentiated muscle cell state (Holland, 2003).

Another vertebrate NK2 class gene, mouse Nkx2-5, is expressed in striated muscles¬óbut only in the embryonic head. The Nkx2-5 expression in mouse cranial muscles might be comparable to AmphiNk2-tin expression in amphioxus somites, but only if one assumes that vertebrate cranial muscles evolved from somites in the head region of some protochordate ancestor. This is the segmentalist point of view, which is opposed by antisegmentalists, who hold that the proximate invertebrate ancestor of the vertebrates had no somites in the anterior part of the body (Holland, 2003).

While the heart is developing in amphioxus, Drosophila, and vertebrates, the most conspicuous similarities are found at the stage of the single, linear heart tube. In contrast, the developmental events before and after the heart tube stage are, respectively, moderately and strongly divergent. During development leading up to the heart tube stage, Drosophila tinman and vertebrate NK2 cardiac genes are expressed in bilateral, embryologically equivalent regions of mesoderm starting around the time of gastrulation and are the earliest known markers for the cardiac lineage. However, commitment of mesoderm cells as cardiac progenitors occurs soon after gastrulation in vertebrates, but is somewhat delayed in Drosophila. In both vertebrates and Drosophila, the paired heart primordia fuse, respectively, in the ventral and dorsal midline to form the single heart tube. By contrast, in amphioxus, no bilaterally symmetrical cardiac primordia are recognizable before the heart tube stage. In other words, AmphiNk2-tin expression (and presumably commitment to a cardiac fate) occurs relatively late in development in a subset of visceral peritoneal cells already positioned in the embryonic midline ventral to the gut (Holland, 2003).

It is not clear whether a single or a paired heart tube is the ancestral state. The unpaired heart rudiment of amphioxus could have evolved from originally paired heart rudiments (this could have occurred if cardiomyocyte specification became delayed until the after the fusion of the visceral mesoderm from either side of the midline). However, a single median heart rudiment may be the evolutionary precursor of independently evolved bilateral heart primordia in Drosophila and vertebrates. In Drosophila, the bilateral heart primordia might result because formation of median heart tube cannot occur until the extraembryonic amnioserosa has disappeared from the dorsal region of the embryo. In vertebrates, the conspicuous bilateral heart fields may have originated as part of the generally precocious differentiation and subdivision of the mesoderm, possibly as an adaptation to large embryonic size (Holland, 2003).

During early cardiogenesis, Drosophila tinman and vertebrate Nk2 cardiac genes play key roles, but they comprise only part of the upstream genetic circuitry involved. Numerous additional cardiogenic genes are known, and their transcriptional networks are being worked out for Drosophila and vertebrates. In amphioxus, by contrast, few of the upstream genes directing cardiogenesis have yet been studied. One important exception is AmphiBMP2/4, a member of the TGFß superfamily, which includes Drosophila decapentaplegic and vertebrate BMP genes. Heart formation in Drosophila and vertebrates requires signaling by Dpp and BMP proteins, respectively, to tinman and NK2 cardiac genes in cardiogenic mesodermal precursor cells. The Drosophila ectoderm produces these Dpp proteins, and in amniote vertebrates, the cardiogenic BMP proteins evidently originate, at least in part, from the endoderm. However, in anamniote vertebrates, BMP is expressed in the mesodermal cells of the heart primordium itself. In amphioxus, AmphiBMP2/4 is transcribed in the late neurulae in a row of visceral peritoneal cells ventral to the midgut and hindgut. This expression is approximately contemporaneous with AmphiNk2-tin expression in the same or in closely adjacent mesoderm cells. Both of these amphioxus genes are presumably transcribed in progenitor cells of the myocardium. In sum, although the available evidence is fragmentary, it is tempting to speculate that the overall transcriptional network for early cardiogenesis (through the median heart tube stage) is well conserved in amphioxus and vertebrates (Holland, 2003).

Following the heart tube stage, the amphioxus heart elongates rostrocaudally, but always remains a linear tube composed exclusively of a single layer of cardiomyocytes. Presumably, this histologically simple amphioxus heart develops and is maintained under the direction of correspondingly simple genetic programs. In comparison to amphioxus, Drosophila has a somewhat more complex heart, comprising not only cardiomyocytes, but also specialized ostial cells, associated pericardial cells of several subtypes, and a cardioaortic valve in later larvae. Expectedly, relatively complex genetic programs direct the formation and maintenance of these specialized heart cell components in Drosophila. A definitive answer about the relative genetic complexity of the amphioxus and Drosophila hearts must await more information on the genetic program responsible for amphioxus cardiogenesis (Holland, 2003).

Fish and frog Tinman homologs

Nkx2.5, a zebrafish tinman homolog, demarcates the heart field and initiates myocardial differentiation. Zebrafish Nkx2.5 is associated with cardiac precursor cells throughout development This association occurs even in the early gastrula, where the level of zebrafish Nkx2.5 is in a gradient which spatially matches the regional propensity of ventral-marginal cells to become heart. Overexpression of Nkx2.5 causes formation of disproportionally larger hearts in otherwise apparently normal embryos. Transplanted cells expressing high levels of Nkx2.5 express cardiac genes even in ectopic locales. Fibroblasts transfected with myc-tagged Nkx2.5 express cardiac genes. These effects require the homeodomain. Thus, Nkx2.5 appears to mark the earliest embryonic heart field and to be capable of initiating the cardiogenic differentiation program. Because ectopic cells or transfected fibroblasts do not beat, Nkx2.5 is likely to be but one step in the determination of cardiac myocyte cell fate. Its overexpression increases heart size, perhaps by bringing cells on the edge of the field to a threshold level for initiation of cardiac differentiation (Chen, 1996).

A Xenopus homeodomain sequence, XNkx-2.5 shows significant similarity to mouse Nkx-2.5 and to the Drosophila tinman gene product. XNkx-2.5 is expressed in the heart region during early Xenopus development and later is also expressed in gut tissue. The observed similarity of sequences and expression patterns suggests that the regulatory mechanisms underlying heart formation may be conserved between distant species (Tonissen, 1994).

An Nk-homeobox gene from Xenopus laevis, XNkx-2.3, appears by sequence homology and expression pattern to be a homologue of tinman. The expression pattern of XNKx-2.3 both during development and in adult tissues partially overlaps with that of another tinman homologue, Csx/NKx-2.5/XNkx-2.5. Embryonic expression of both XNkx-2.3 and XNkx-2.5 is induced at a time when cardiac specification is occurring. XNkx-2.3 is expressed in early cardiac primordia before the expression of a marker of cardiac differentiation. XMLC2, as well as in pharyngeal endoderm. In adult tissues, XNkx-2.3 is expressed in the heart and several visceral organs. As the helix-loop-helix factor Twist is thought to regulate tinman expression in Drosophila, the expression of XNkx-2.3 and Xtwist were compared during embryonic development in Xenopus. There appears to be no overlap in expression patterns of the two RNAs from the neurulae stages onward, the first time at which the RNAs can be visualized by in situ hybridization. The overlapping expression patterns of XNkx-2.3 and mNkx-2.5/XNkx-2.5 in conjunction with evidence presented here that other Nk-homeodomains are expressed in adult mouse and Xenopus heart suggests that tinman may be represented by a family of genes in vertebrates (Evans, 1995).

The tinman homeobox gene of Drosophila is absolutely required for development of the insect heart. This observation prompted the isolation of tinman-related genes from vertebrates, in the hope that the developmental function of the gene would be conserved between evolutionarily distinct species. The first vertebrate tinman gene, Nkx2-5, was isolated from mouse and subsequently, orthologs of Nkx2-5 have been isolated from a number of different species. In all cases, a conserved pattern of Nkx2-5 expression is observed in the developing heart, commencing prior to differentiation. Genetic ablation of Nkx2-5 in the mouse results in embryonic lethality due to heart defects, but most myocardial genes are expressed normally and a beating heart tube forms. This observation raises the possibility that additional genes related to Nkx2-5 are partially rescuing Nkx2-5 function in the null mouse. Recently, additional members of the tinman-related gene family have been discovered and characterized in a number of different species. Somewhat surprisingly, orthologous genes in different organisms can be rather divergent in sequence and may show completely different expression patterns. In at least some organisms, expression of the tinman-related genes is not observed in the heart. Due to the increasing number of family members and the somewhat divergent expression patterns, the precise role of the tinman-related genes in cardiac development remains an open question. In a search for additional tinman-related genes in the frog Xenopus laevis, Nkx2-9, a novel member of the tinman-related gene family, has been identfied. Preliminary characterization reveals that Nkx2-9 is expressed in the cardiogenic region of the embryo prior to differentiation, but in the heart transcript levels decrease rapidly, at about the time that differentiation commences (Newman, 1998).

Nkx2.5 is expressed in the cardiogenic mesoderm of avian, mouse, and amphibian embryos. To understand how various cardiac fates within the cardiogeneic mesoderm are apportioned, Xenopus XNkx2.5 expression within this domain was examined. The lateral regions of the XNkx2.5 expression domain in the neural tube stage embryo (stage 22), form the dorsal mesocardium and roof of the pericardial cavity while the intervening ventral region closes to form the myocardial tube. XNkx2.5 expression is maintained throughout the period of heart tube morphogenesis and during the differentiation of myocardial, mesocardial, and pericardial tissues. A series of microsurgical experiments has shown that myocardial differentiation in the lateral portion of cardiogenic mesoderm is suppressed during normal development by signals from the prospective myocardium and by tissues located more dorsally in the embryo, in particular the neural tube. These signals combine to block myogenesis downstream of XNkx2.5 either at or above the level of contractile protein gene expression. It is proposed that the entire XNkx2.5/heart field is transiently specified as cardiomyogenic. Suppression of this program redirects lateral cells to adopt dorsal mesocardial and dorsal pericardial fates and subdivides the field into distinct myogenic and nonmyogenic compartments (Raffin, 2000).

Advantage has been taken of a transient transgenic strategy in Xenopus embryos to demonstrate that BMP signaling is required in vivo for heart formation in vertebrates. Ectopic expression of dominant negative Type I (tALK3) or Type II (tBMPRII) BMP receptors in developing Xenopus embryos results in reduction or absence of heart formation. Additionally, blocking BMP signaling in this manner downregulates expression of XNkx2-5, a homeobox gene required for cardiac specification, prior to differentiation. Notably, however, initial expression of XNkx2-5 is not affected. Mutant phenotypes can be rescued by co-injection of mutant with wild-type receptors or co-injection of mutant receptors with XSmad1, a downstream mediator of BMP signaling. Whole-mount in situ analyses indicate that ALK3 and XSmad1 are coexpressed in cardiogenic regions. Together, these results demonstrate that BMP signaling is required for maintenance of XNkx2-5 expression and heart formation and suggest that ALK3, BMPRII, and XSmad1 may mediate this signaling (Shi, 2000).

Vertebrate homologs of the Drosophila tinman transcription factor have been implicated in the processes of specification and differentiation of cardiac mesoderm. In Xenopus three members of this family have been isolated to date. The XNkx2-3, Xnkx2-5, and XNkx2-10 genes are expressed in increasingly distinctive patterns in endodermal and mesodermal germ layers through early development, suggesting that their protein products (either individually or in different combinations) perform distinct functions. Using amphibian transgenesis, it has been found that the expression pattern of one of these genes, XNkx2-5, can be reproduced using transgenes containing only 4.3 kb of promoter sequence. Sequence analysis reveals remarkable conservation between the distalmost 300 bp of the Xenopus promoter and a portion of the AR2 element upstream of the mouse and human Nkx2-5 genes. Interestingly, only the 3' half of this evolutionarily conserved sequence element is required for correct transgene expression in frog embryos. Mutation of conserved GATA sites or a motif resembling the dpp-response element in the Drosophila tinman tinD enhancer dramatically reduces the levels of transgene expression. Despite its activity in Xenopus embryos, in transgenic mice the Xenopus Nkx2-5 promoter is able to drive reporter gene expression only in a limited subset of cells expressing the endogenous gene. This intriguing result suggests that despite evolutionary conservation of some cis-regulatory sequences, the regulatory controls on Nkx2-5 expression have diverged between mammals and amphibians (Sparrow, 2000).

A direct comparison of the DNA site preference of all three Xenopus tinman homologs demonstrates that at least in vitro, these proteins all recognize a common 8-bp consensus, TC/TAAGTGG/C. Identical data were obtained using only the homeodomains of these proteins. The common consensus sequence is similar to that (TCAAGTG/T) identified for the Drosophila NK2 protein, the nematode ceh-22 protein, and the mammalian thyroid-specific transcription factor TTF-1 (Nkx2-1) and is consistent with the TNAAGTG motif identified in 7 of 15 sequences selected by the murine Tinman homolog Nkx2-5. Subtle differences have been suggested for the sequence preference in bases flanking the invariant AAGT core and similar results were obtained with XeNK2, and NK2 protein expressed in the developing brain of the frog embryo. Together, these results suggest that conservation of the homeodomain sequence across phyla has maintained a largely similar DNA site preference for NK2 proteins (Sparrow, 2000).

The atrial natriuretic factor (ANF) gene is initially expressed throughout the myocardial layer of the heart, but during subsequent development, expression becomes limited to the atrial chambers. Mouse knockout and mammalian cell culture studies have shown that the ANF gene is regulated by combinatorial interactions between Nkx2-5, GATA-4, Tbx5, and SRF; however, the molecular mechanisms leading to chamber-specific expression are currently unknown. The Xenopus ANF promoter was isolated in order to examine the temporal and spatial regulation of the ANF gene in vivo using transgenic embryos. The mammalian and Xenopus ANF promoters show remarkable sequence similarity, including an Nkx2-5 binding site (NKE), two GATA sites, a T-box binding site (TBE), and two SRF binding sites (SREs). Transgenic studies show that mutation of either SRE, the TBE or the distal GATA element, strongly reduces expression from the ANF promoter. However, mutations of the NKE, the proximal GATA, or both elements together, result in relatively minor reductions in transgene expression within the myocardium. Surprisingly, mutation of these elements results in ectopic ANF promoter activity in the kidneys, facial muscles, and aortic arch artery-associated muscles, and causes persistent expression in the ventricle and outflow tract of the heart. It is proposed that the NKE and proximal GATA elements serve as crucial binding sites for assembly of a repressor complex that is required for atrial-specific expression of the ANF gene (Small, 2003).

Chicken Tinman homologs

Continued: Evolutionary Homologs part 2/3 | part 3/3

tinman: Biological Overview | Regulation | Targets of Activity | Protein Interactions | Developmental Biology | Effects of Mutation | References

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.