Twist in C. elegans

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. Analysis of sequences involved in CeTwist regulation suggests that a protein of the HOX family might be an upstream component in this regulatory pathway. The homeotic selector factor MAB-5 plays an essential role in the early activation of CeTwist. MAB-5 protein is prominent in the M mesoblast and in descendants of M for at least four cell divisions; MAB-5 expression is maintained in the sex myoblats and their undifferentiated progeny while expression trails off in M-lineage cells that differentiate as striated muscle. MAB-5 activation cannot, however, account for the entire CeTwist activity pattern, and it is proposed that additional mesoderm-specific components (perhaps interacting with an upstream NF-kappaB/Dorsal control element) cooperate with MAB-5 in the initial activation of CeTwist. The 315-bp promoter-proximal element is required for CeTwist promoter activity in undifferentiated cells of the M lineage. Deletion of 21 bp from the 5' end of the promoter-proximal element results in a complete loss of activity. Within this 21-bp region, the promoter-proximal M lineage element contains a single putative NF-kappaB/Dorsal-binding site and a GAGA motif. Targeted mutations in either site abolish the activity of the 315-bp minimal promoter in the entire M lineage. Interestingly, alteration of a DNA sequence 3' of the NF-kappaB/GAGA motif region also result in inactivation of the hlh-8 promoter. The sequences in this region showed high homology to binding sites for the Antennapedia class of homeodomains.. The 315-bp promoter-proximal element also contains two putative CeTwist-binding sites. In vivo ectopic CeTwist + CeE/DA coexpression experiments reveal no evidence of CeTwist autoregulation (Harfe, 1998).

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 (Harfe, 1998).

egl-15 encodes a member of the FGF receptor (FGFR) family that is required for the proper migration of the sex myoblasts. The egl-15 promoter is active in many early M lineage descendants and later in the four vm1 vulval muscles. Although this promoter activity pattern is distinct from that of ceh-24, the activity of each in the later M lineage suggested the possibility of a common factor specifying M lineage activity for the two genes. An egl-15 promoter fragment of 701 bp is sufficient to drive reporter expression in the M lineage. This fragment contained five matches to the E-box consensus. Three were precise NdE boxes (CATATG), whereas the other two differed from this consensus by a single base pair. Deletion analysis suggests critical roles for the E-box motif and for additional elements in egl-15 promoter activity: specifically, M-lineage activity is eliminated by promoter truncations that remove the first two NdE-like boxes, and by a deletion that removes the three proximal NdE-like boxes. 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).

The basic helix-loop-helix (bHLH) transcription factor Twist plays a role in mesodermal development in both invertebrates and vertebrates. In an effort to understand the role of the unique C. elegans Twist homolog, hlh-8, mesodermal development was analyzed in animals with a deletion in the hlh-8 locus. This deletion was predicted to represent a null allele because the HLH domain is missing and the reading frame for the protein is disrupted. Animals lacking CeTwist function were constipated and egg-laying was defective. Both of these defects were rescued in transgenic mutant animals expressing wild-type hlh-8. Observing a series of mesoderm-specific markers allowed for the thorough characterization of the loss of hlh-8 function. CeTwist performs an essential role in the proper development of a subset of mesodermal tissues in C. elegans. CeTwist is required for the formation of three out of the four non-striated enteric muscles born in the embryo, and is required to develop wild-type enteric defecation muscles. In contrast, CeTwist is not required for the formation of the embryonically derived striated muscles. Most of the post-embryonic mesoderm develops from a single lineage. CeTwist is necessary for appropriate patterning in this lineage and is required for expression of two downstream target genes, but is not required for the expression of myosin, a marker of differentiation. These results suggest that mesodermal patterning by Twist is an evolutionarily conserved function (Corsi, 2000).

The phenotypes that were observed in hlh-8 mutant worms have parallels with null mutations in mice and flies. In M. musculus, twist null homozygotes are embryonic lethal, but these animals survive until embryonic day 11 and have detectable mesodermal development with tissue-specific defects. A complicating issue is that a number of additional mouse twist homologs and related bHLH factors have been identified that may make contributions to other aspects of early mesoderm patterning. To date, the C. elegans genome is approx. 99% complete and no other worm twist homolog has been identified. D. melanogaster twist null mutants die as embryos that completely lack any mesoderm. The severity of the fly mutants may reflect an early developmental requirement for Twist in mesoderm specification that is not found in mice and worms. If Twist is supplied to the early fly embryo, the defects that are seen later in development are restricted to subsets of muscle cells such as executing the choice between somatic and visceral mesoderm and controlling mesodermal patterning. Taken together, the data from mice, flies and worms suggest that a critical subset of twist functions may be conserved evolutionarily (Corsi, 2000).

Another gene product that functions in the M lineage is CeMyoD (encoded by the hlh-1 gene). Twist has been reported to antagonize the function of MyoD through titration of partner proteins and by direct interactions. The results presented here do not support an antagonistic relationship between CeTwist and CeMyoD in the M lineage. In fact, loss of hlh-1 function causes a supernumerary SM phenotype similar to that seen in hlh-8 (nr2061) animals. The division plane defects in the M lineage of hlh-8 (nr2061) animals occurred prior to detectable hlh-1 expression in the M lineage. These division plane defects often misposition early M descendants. One interesting hypothesis is that these misplaced cells may no longer be responsive to positional cues that influence hlh-1 expression and their cell fate. Evidence also exists for Twist acting as an inhibitor to myogenesis in tissue culture systems by antagonizing other bHLH factors such as MyoD both in vivo and in vitro. Such experiments, combined with the pattern of twist expression, suggest that Twist is required to maintain myoblasts in a proliferative state. Consistent with this view is the D. melanogaster expression pattern that shows a decrease in Twist protein prior to differentiation. It has been demonstrated in flies, by the overexpression of twist, that a decrease in protein quantity is not required for normal differentiation. Also, CeTwist is not required for retaining myoblasts in a proliferative state. This conclusion suggests that the observed decrease in hlh-8::gfp prior to differentiation observed in C. elegans is not required to allow normal differentiation to proceed (Corsi, 2000).

The FGF receptor homolog egl-15 and the NK homeodomain transcription factor ceh-24 are likely to be direct targets of CeTwist based on several pieces of evidence: promoter regions upstream of these genes contain canonical Twist-like E boxes (the signature sites for bHLH protein binding), induction of hlh-8 expressed from a heat shock promoter turns on gfp reporters containing egl-15 and ceh-24 promoter sequences, and RNA interference using sequences from hlh-8 causes a reduction in expression from the same gfp reporters. The data presented here confirms and extends these observations by showing that a deletion in hlh-8 causes a lack of reporter expression of these targets that is rescued with the addition of wild-type CeTwist. egl-15 is involved in the migration of SMs to the gonad. The lack of egl-15 expression could contribute to the posterior SM placement and conceivably a subset of developmental defects in hlh-8 (nr2061) animals. ceh-24 mutants have no detectable phenotype and are not constipated. Therefore, in the enteric muscles there must be at least one other gene that requires CeTwist for its expression. The potential discovery of other genes regulated by CeTwist is likely to reveal a clearer picture of how these gene products are cooperating to contribute to mesoderm development (Corsi, 2000).

Twist is a transcription factor that is required for mesodermal cell fates in all animals studied to date. Mutations of this locus in humans have been identified as the cause of the craniofacial disorder Saethre-Chotzen syndrome. The C. elegans Twist homolog is required for the development of a subset of the mesoderm. A semidominant allele of the gene that codes for CeTwist, hlh-8, has defects that occur earlier in the mesodermal lineage than a previously studied null allele of the gene. The semidominant allele has a charge change (E29K) in the basic DNA-binding domain of CeTwist. Surprisingly, the mutant protein retains DNA-binding activity as both a homodimer and a heterodimer with its partner E/Daughterless (CeE/DA). However, the mutant protein blocks the activation of the promoter of a target gene. Therefore, the mutant CeTwist may cause cellular defects as a dominant negative protein by binding to target promoters as a homo- or hetero-dimer and then blocking transcription. Similar phenotypes as those caused by the E29K mutation were observed when amino acid substitutions in the DNA-binding domain that are associated with the human Saethre-Chotzen syndrome were engineered into the C. elegans protein. These data suggest that Saethre-Chotzen syndrome may be caused, in some cases, by dominant negative proteins, rather than by haploinsufficiency of the locus (Corsi, 2002).

Members of the Hox family of homeoproteins and their cofactors play a central role in pattern formation of all germ layers. During postembryonic development of C. elegans, non-gonadal mesoderm arises from a single mesoblast cell M. Starting in the first larval stage, M divides to produce 14 striated muscles, 16 non-striated muscles, and two non-muscle cells (coelomocytes). The role of the C. elegans Hox cluster and of the exd ortholog ceh-20 in patterning of the postembryonic mesoderm has been investigated. By examining the M lineage and its differentiation products in different Hox mutant combinations, an essential but overlapping role was found for two of the Hox cluster genes, lin-39 (Scr homolog) and mab-5 (Antp homolog), in diversification of the postembryonic mesoderm. This role of the two Hox gene products requires the CEH-20 cofactor. One target of these two Hox genes is the C. elegans twist ortholog hlh-8. Using both in vitro and in vivo assays, it has been demonstrated that twist is a direct target of Hox activation. Evidence from mutant phenotypes is presented that twist is not the only target for Hox genes in the M lineage: in particular lin-39 mab-5 double mutants exhibit a more severe M lineage defect than the hlh-8 null mutant (Liu, 2000).

The C. elegans twist ortholog hlh-8 is a direct and critical target of Hox genes and ceh-20 in the postembryonic M lineage. A critical site has been identified in the hlh-8 promoter that is a binding site for the LIN-39/CEH-20 protein complex. The similarity between core binding sequences for Drosophila Antp and Dfd proteins in vitro, and the functional equivalence of mab-5 and lin-39 in activating hlh-8 expression in the M lineage, strongly suggest that this site is also a binding site for MAB-5/CEH-20. Although hlh-8 is a target for Hox/CEH-20 function in the M lineage, it is not the only such target. Several indirect observations demonstrate the existence of additional targets. One line of evidence comes from the observation that forced expression of hlh-8 in lin-39 mab-5 mutants fails to rescue the M lineage defects. An independent line of evidence comes from a comparison of mutant phenotypes: lin-39 mab-5 mutants show a more severe patterning defect in the M lineage than null hlh-8 mutants: (1) while lin-39 mab-5 animals lack both M-derived coelomocytes, the majority of hlh-8 mutants contain normal numbers of M-derived coelomocytes; (2) while lin-39 mab-5 mutants lack all M-derived bodywall muscle, hlh-8 mutants produce variable number of these cells; (3) sex muscles can be produced in hlh-8 mutants, although they are not fully differentiated. The identity of other Hox targets in the M lineage is not known (Liu, 2000).

Proper metazoan mesoderm development requires the function of a basic helix-loop-helix (bHLH) transcription factor, Twist. Twist-containing dimers regulate the expression of target genes by binding to E box promoter elements containing the site CANNTG. In Caenorhabditis elegans, CeTwist functions in a subset of mesodermal cells. This study focuses on how CeTwist controls the expression of its target gene, arg-1. A 385 bp promoter region of arg-1, which contains three different E box elements, is sufficient for maintaining the full CeTwist-dependent expression pattern. Interestingly, the expression of arg-1 in different tissues is regulated distinctly, and each of the three E boxes plays a unique role in the regulation. The first and the third E boxes (E1 and E3) are required for expression in a distinct subset of the mesodermal tissues where arg-1 is normally expressed, and the second E box (E2) is required for expression in the full set of those tissues. The essential role of E2 in arg-1 regulation is correlated with the finding that E2 binds with greater affinity than E1 or E3 to CeTwist dimers. A potential role for additional transcription factors in mesodermal gene regulation is suggested by the discovery of a novel site that is also required for arg-1 expression in a subset of the tissues but is not bound in vitro by CeTwist. On the basis of these results, a model of CeTwist gene regulation is proposed in which expression is controlled by tissue-specific binding of distinct sets of E boxes (Zhao, 2007).

The presence of a specific E box alone may not be sufficient to determine its importance in CeTwist target gene expression. It is possible that the flanking sequences influence the binding activity of CeTwist dimers to specific E boxes. In vitro binding site selection assays identified specific flanking regions of an NdE box, GC/AA CATATG TT/GC, that were preferred for binding to vertebrate Twist-E12 heterodimers. Mutations made with the minimal promoter of ceh-24 also indicated that the immediate flanking sequences of the two NdE boxes are essential for expression. The importance of flanking sequences is further supported by the position of the GT box in the arg-1 promoter. This site was not bound by CeTwist-containing dimers, yet it is important for arg-1 expression in the vm1 muscles. The GT box is positioned halfway between E1 and E2, only 20–30 nucleotides away from each E box. The GT box could be bound directly by another factor that influences the binding or transcriptional activity of CeTwist at E1 or E2. Additionally, the distance of an E box to the ATG and to other E boxes could be a factor as well. In egl-15, mls-1, and arg-1, the minimal promoter regions are all in close proximity to the ATG and the E boxes within those regions are all closely positioned to one another. The combinatorial effect of all of these factors may account for the distinct expression patterns of the four CeTwist target genes (Zhao, 2007).

The temporospatial regulation of genes encoding transcription factors is important during development. The hlh-8 gene encodes the C. elegans mesodermal transcription factor CeTwist. Elements in the hlh-8 promoter restrict gene expression to predominantly undifferentiated cells of the M lineage. hlh-8 expression in differentiated mesodermal cells is controlled by two well-conserved E box elements in the large first intron. Additionally, it was found that these elements are bound in vitro by CeTwist and its transcription factor partner, CeE/DA. The E box driven expression is eliminated or diminished in an hlh-8 null allele or in hlh-2 (CeE/DA) RNAi, respectively. Expression of hlh-8 is also diminished in animals harboring an hlh-8 intron deletion allele. Altogether, these results support a model in which hlh-8 is initially expressed in the undifferentiated M lineage cells via promoter elements and then the CeTwist activates its own expression further (autoregulation) in differentiated cells derived from the M lineage via the intron elements. This model provides a mechanism for how a transcription factor may regulate distinct target genes in cells both before and after initiating the differentiation program. The findings could also be relevant to understanding human Twist gene regulation, which is currently not well understood (Meyers, 2010).

Twist homologs in other invertebrates

Twist plays a major role in mesoderm specification of triploblasts. The presence of a Twist homolog in diploblasts such as the cnidarian Podocoryne carnea raises questions on the evolution of mesoderm, the third cell layer characteristic for triploblasts. Podocoryne Twist is expressed in the early embryo until the myoepithelial cells of the larva differentiate and then again during medusa development. There, the gene is detected first when the myoepithelial cells of the polyp dedifferentiate to form the medusa bud and later Twist is found transiently in the entocodon, a mesoderm-like cell layer that differentiates into the smooth muscle and striated muscle of the bell. In later bud stages and the medusa, expression is seen where non-muscle tissues differentiate. Experimental analysis of in vitro transdifferentiation and regeneration demonstrates that Twist activity is not needed when isolated striated muscle regenerates medusa organs. Developmental roles of Twist are discussed with respect to early animal evolution from a common ancestor of cnidarians and bilaterians (Spring, 2000).

The complete sequence of Podocoryne Twist was compared to representative family members. Only the bHLH domain and the Twist-specific WR motif are conserved in all species. The Podocoryne Twist and the two mouse proteins Twist1-Mm and Twist2-Mm could also be aligned at the N-terminus. The domain structure of Podocoryne Twist is more similar to vertebrate proteins than to proteins in Drosophila, where the domain is more than twice as big as in C. elegans, where the WR motif appears to have deteriorated. Even the bHLH region of C. elegans shares only 60% identical residues with Podocoryne as compared to 80% in vertebrates. The length of the loop in the bHLH domain of Twist is conserved, and is one amino acid shorter than in the most related bHLH proteins such as Ptf1 or the only other known cnidarian relative, CnASH, from Hydra. The Twist proteins share this gap with the MyoD family, which is otherwise not closely related. A phylogenetic analysis of the bHLH domain confirms that Podocoryne Twist corresponds to single orthologs from Drosophila or C. elegans and two paralogs in vertebrates. Mouse sequences (Drosophila homologs in parentheses) of the relevant subfamilies, Ptf (CG10066), Hand, Msc (CG5005), Scx (CG12648), Nhlh (CG3052), and Lyl (CG2655) were compared to Drosophila and C. elegans sequences. The only other known cnidarian bHLH protein, the achaete-scute-like factor Hydra CnASH, is included with the mouse homologs Asc11/2, and mouse Myod1 as an outgroup. Representatives of all Twist-related subfamilies can be found in the Drosophila genome, while C. elegans homologs are usually more derived, as in the case of Twist, or even absent from the complete genome as in the Hand, Msc, Scx, and Lyl subfamilies (Spring, 2000).

To study the original invention of the bilaterian bauplan, investigations must be extended to animals of sister groups of the bilaterians. A large set of data from cnidarians suggests that these animals also make use of the same set of genes, although often for quite divergent functions. Homologs of developmental regulator genes such as Hox, Pax, Brachyury, or Otx have been cloned from cnidarians, but it is not clear how to compare head-specific expression in cnidarians with bilaterian axis specification or brain development. In an alternative classification of animals, bilaterians are called triploblasts and the term diploblast is often used collectively for the leftover phyla. While the bilayered freshwater polyp Hydra consists only of an ectoderm and an endoderm, and therefore can be well described as a diploblast, this is not true for the jellyfish stage, which is present in three out of four cnidarian classes, or the other non-bilaterian phyla Ctenophora, Placozoa, and Porifera. It appears that mesoderm invention predates the split between cnidarians and bilaterians. Therefore, a reevaluation of old descriptions of a third cell layer in medusa bud development is needed in conjunction with studies taking into account knowledge of known mesoderm specification factors from animals with a true mesoderm (Spring, 2000 and references therein).

Sequence comparisons suggest that Podocoryne Twist is a true ortholog of bilaterian cognates. The bHLH region is most similar to real Twist family members, even more similar to vertebrate Twist members than C. elegans Twist, and contains the same loop length, in contrast to other Twist-related subfamilies. In addition to the bHLH domain, the WR motif of 14 amino acids is perfectly conserved in Podocoryne and mammals. A WR motif was not found in any unrelated protein and is barely recognizable in the C. elegansTwist sequence. Furthermore, there is residual sequence similarity at the N-terminus of Podocoryne and vertebrate Twist proteins, where mouse Twist appears to interact with and inhibit the p300 coactivator (Spring, 2000 and references therein).

A phylogenetic analysis of the bHLH domain confirms that the bHLH family can be reliably subdivided into subfamilies with multiple vertebrate paralogs for each invertebrate bHLH gene, reminiscent of the Hox clusters and many other duplicated gene families. In such a natural classification of bHLH proteins Twist belongs to a different subgroup than the MyoD, NeuroD, or the achaete-scute families, from which the only other cnidarian bHLH protein CnASH is known. Still, within its subgroup with the Ptf, Hand, Msc, Scx, Nhlh, and Lyl subfamilies, Twist is the only member with the same loop length as the MyoD family. Whether this has any influence on the direct interaction and competition of Twist and MyoD family members with E proteins (Daughterless in Drosophila) is not known. The robustness of the Twist branch of the phylogenetic tree and the high conservation of Podocoryne Twist suggest that gene duplication within the bHLH family had occurred before the split of cnidarians and bilaterians. Therefore, orthologs for all subfamilies can be expected in both lineages. This is the case in the Drosophila genome, while C. elegans appears to have lost four of the subfamilies. Apparently, genome sequences are needed to detect all the members of gene families since decades of Drosophila studies have only revealed two of the eight Twist-like subfamilies found in the genome, possibly due to partial redundancy of duplicated genes (Spring, 2000 and references therein).

Podocoryne Twist is present from early cleavage stages until myoepithelial cells form, but expression decreases in larva competent to transform into the primary polyp. Feeding polyps are basically made up of the same two myoepithelial cell layers as larvae and also lack Twist expression, even during colonial growth. However, development is only completed with the formation of the medusa, the sexually mature animal, and here again Twist is expressed strongly, but transiently in the proliferating undifferentiated cell mass that will also give rise to the medusa-typical muscles. The formation of the medusa starts with highly proliferative, undifferentiated cells generated by dedifferentiation of epithelial muscle cells of the mother polyp. In a process comparable to gastrulation, migration of tissues and cells, formation of body cavities, and consecutive morphogenesis and differentiation processes finally lead to the formation of the adult stage, the sexual medusa. The expression of Twist is first noticed in the bulging ectodermal and endodermal epithelial muscle cells of the polyp. It is active at the same time when Otx start to be expressed. When muscle differentiation progresses beyond stages 4-5, Twist expression gradually disappears in both muscle tissue layers whereas expression of Otx and Cnox1-Pc is maintained even in the striated muscle of the adult medusa. The disappearance of Twist expression correlates well with the declining rate of DNA replication (Spring, 2000 and references therein).

All the Twist-expressing tissues lack organized contractile myofilament systems and staining is strong where cell proliferation is high. It appears that Twist is expressed either transiently in development, including myogenic tissues, or permanently in the adult medusa at places where cell proliferation continues and nonmyogenic tissues differentiate, like the growth zone at the margin of the bell. Although Twist function cannot be surmised from temporal and spatial expression patterns, the results demonstrate that there exists a correlation between formation of myoepithelial cells in early muscle development in the larva and in the entocodon derived muscle systems, but that its expression clearly correlates with the formation of nonmuscle tissues such as the subumbrellar plate in later bud stages and the medusa. In this context it is interesting to note that epithelial muscle cells of the gonozoid polyp start to express Twist when they dedifferentiate to form the highly proliferative nonmuscle cells of the early medusa buds (Spring, 2000).

In the long-germ insect Drosophila melanogaster, dorsoventral polarity is induced by localized Toll-receptor activation which leads to the formation of a nuclear gradient of the rel/NF-kappaB protein Dorsal. Peak levels of nuclear Dorsal are found in a ventral stripe spanning the entire length of the blastoderm embryo allowing all segments and their dorsoventral subdivisions to be synchronously specified before gastrulation. A nuclear Dorsal protein gradient of similar anteroposterior extension exists in the short-germ beetle, Tribolium castaneum, which forms most segments from a posterior growth zone after gastrulation. Tc-dl mRNA accumulates in the oocytes of mid and late stage egg chambers suggesting that Tc-dl, like Drosophila DL mRNA is maternally provided. In contrast to Drosophila, (1) nuclear accumulation is first uniform and then becomes progressively restricted to a narrow ventral stripe; (2) gradient refinement is accompanied by changes in the zygotic expression of the Tribolium Toll-receptor suggesting feedback regulation and, (3) the gradient only transiently overlaps with the expression of a potential target, the Tribolium twist homolog, and does not repress Tribolium decapentaplegic. The Tribolium Toll homolog however, differs from Drosophila Toll by being transcribed only zygotically. No nuclear Dorsal is seen in the cells of the growth zone of Tribolium embryos, indicating that here dorsoventral patterning occurs by a different mechanism. However, Dorsal is up-regulated and transiently forms a nuclear gradient in the serosa, a protective extraembryonic cell layer ultimately covering the whole embryo (Chen, 2000).

Although the regulatory behaviour of DV patterning found in perturbation experiments suggests reduced dependence on a maternal prepattern, it is also possible that interactions between zygotic DV patterning genes play an important role. In this context, it is interesting that the Tc-dl not only differs from the Drosophila Dl gradient, but also the relationship of the gradients to their respective target genes and hence to the cell fates along the DV axis appears to be different. This is apparent for both the mesoderm and the ectoderm. In Drosophila, a ventral stripe of high nuclear Dorsal in the trunk region of the blastoderm embryo is congruent with the mesodermal anlagen since it defines the lateral expansion of twi expression which, together with sna, promotes ventral furrow formation. Drosophila Dl remains present in twi- and sna-expressing cells until the mesoderm has invaginated. In Tribolium, in contrast, the early weak Tc-twi expression domain, which is even along the AP axis and coincides with the highest levels of nuclear Tc-dl, is rapidly replaced by a domain with strong AP asymmetry by becoming repressed anteriorly and broadened toward the posterior pole. When this expression pattern is fully developed, nuclear Tc-dl has disappeared from the germ rudiment. However, this final Tc-twi domain corresponds to the presumptive mesoderm since it presages the position and shape of the ventral furrow. This implies that the shape of the gradient does not fully determine the mesodermal anlagen and that Tc-twi transcription becomes independent of activation by Tc-dl at late blastoderm (Chen, 2000).

Mesoderm formation plays a crucial role in the establishment of the chordate body plan. In this regard, lancelet (amphioxus) embryos develop structures such as the anteriorly extended notochord and the lateral divertecula in their anterior body. To elucidate the developmental basis of these structures, the expression pattern of a lancelet twist-related gene, Bbtwist, was examined from the late gastrula to larval stages. In late-gastrula embryos, the transcripts of Bbtwist are detected in the presumptive first pair of somites and the middorsal wall of the primitive gut. The expression of Bbtwist is then upregulated in the lateral wall of somites and the notochord. At the late-neurula stage, it is also expressed in the anterior wall of the primitive gut, as well as in the evaginating lateral diverticula. No signal is detected in the left lateral diverticulum when it is separated from the gut, while in the right lateral diverticulum, the gene is expressed later, during the formation of the head coelom in knife-shaped larvae, and in the anterior part of the notochord in the same larvae. In 36-h larvae, only faint expression is detected in the differentiating notochordal and paraxial mesoderm in the caudal region. These expression patterns suggest that Bbtwist is involved in early differentiation of mesodermal subsets as seen in Drosophila and vertebrates. The expression in the anterior notochord may be related to its anterior expansion. The expression in the anterior wall of the primitive gut and its derivative, the lateral diverticula, suggests that lancelets share the capability to produce a mesodermal population from the tip of the primitive gut with nonchordate deuterostome embryos (Yasui, 1998).

Ascidian larvae develop mesenchyme cells in their trunk. A fibroblast growth factor (FGF9/16/20) is essential and sufficient for induction of the mesenchyme in Ciona savignyi. Two basic helix-loop-helix (bHLH) genes named Twist-like1 and Twist-like2 have been identified as downstream factors of this FGF. These two genes are phylogenetically closely related to each other, and are expressed specifically in the mesenchymal cells after the 110-cell stage. Gene-knockdown experiments using a specific morpholino oligonucleotide demonstrate that Twist-like1 plays an essential role in determination of the mesenchyme and that Twist-like2 is a downstream factor of Twist-like1. In addition, both overexpression and misexpression of Twist-like1 converts non-mesenchymal cells to mesenchymal cells. The upstream regulatory mechanisms of Twist-like1 are different between B-line mesenchymal cells and the A-line mesenchymal cells called 'trunk lateral cells'. FGF9/16/20 is required for the expression of Twist-like1 in B-line mesenchymal precursor cells, whereas FGF, FoxD and another novel bHLH factor called NoTrlc (for No Trunk lateral cells, because it is essential for the differentiation of TLCs, which give rise to blood cells and body-wall muscles after metamorphosis) are required for Twist-like1 to be expressed in the A-line mesenchymal precursor cells. Therefore, two different but partially overlapping mechanisms are required for the expression of Twist-like1 in the mesenchymal precursors, that triggers the differentiation of the mesenchyme in Ciona embryos (Imai, 2003).

Twist in Spiders

An Achaearanea tepidariorum (Chelicerata, Arachnida) gene related to Drosophila twist (twi), which encodes a basic helix-loop-helix transcription factor required to specify mesoderm fate in the Drosophila embryo, was cloned. The cloned spider gene was designated At.twist (At.twi). Its expression was examined by whole-mount in situ hybridization. At.twi transcripts were first detected in cells located at the polar and equatorial areas of the spherical embryo when the cumulus reached the equator. As the extra-embryonic area expanded, more cells expressed At.twi transcripts. The At.twi-expressing cells became distributed nearly uniformly in the embryonic area. At these stages, some At.twi-expressing cells were found in the surface epithelial cell layer, but other At.twi-expressing cells were at slightly deeper positions from the surface. When the embryo was transformed into a germ band, all At.twi-expressing cells were situated just beneath the surface ectoderm, where they became metamerically arranged. Although little expression was observed in the caudal lobe of the elongating germ band, new stripes of At.twi expression appeared beneath the ectoderm in accordance with the posterior growth. These observations suggested that the cells expressing At.twi were most likely mesoderm. It is proposed that At.twi can be used as a molecular marker for analyzing mesoderm development in the spider embryo. Moreover, comparison of the expression patterns of twi and At.twi revealed divergent aspects of mesoderm development in the fly and spider. In addition, an Achaearanea gene related to snail, which is another mesoderm-determining gene in Drosophila, was cloned, and its expression was shown to be restricted to the ectoderm with no indication for a role in mesoderm development (Yamazaki, 2005; full text of article).

In the development of most arthropods, the caudal region of the elongating germ band (the growth zone) sequentially produces new segments. Previous work with the spider Cupiennius salei suggested involvement of Delta-Notch signaling in segmentation. This study reports that, in the spider Achaearanea tepidariorum, the same signaling pathway exerts a different function in the presumptive caudal region before initiation of segmentation. In the developing spider embryo, the growth zone becomes morphologically apparent as a caudal lobe around the closed blastopore. Preceding caudal lobe formation, transcripts of a Delta homolog, At-Delta, are expressed in evenly spaced cells in a small area covering the closing blastopore and then in a progressively wider area of the germ disc epithelium. Cells with high At-Delta expression are likely to be prospective mesoderm cells, which later express a twist homolog, At-twist, and individually internalize. Cells remaining at the surface begin to express a caudal homolog, At-caudal, to differentiate as caudal ectoderm. Knockdown of At-Delta by parental RNA interference results in overproduction of At-twist-expressing mesoderm cells at the expense of At-caudal-expressing ectoderm cells. This condition gives rise to a disorganized caudal region that fails to pattern the opisthosoma. In addition, knockdown of Notch and Suppressor of Hairless homologs produces similar phenotypes. It is suggested that, in the spider, progressive activation of Delta-Notch signaling from around the blastopore leads to stochastic cell fate decisions between mesoderm and caudal ectoderm through a process of lateral inhibition to set up a functional caudal lobe (Oda, 2007).

Genomic cis-regulatory networks in the early Ciona intestinalis embryo

Precise spatiotemporal gene expression during animal development is achieved through gene regulatory networks, in which sequence-specific transcription factors (TFs) bind to cis-regulatory elements of target genes. Although numerous cis-regulatory elements have been identified in a variety of systems, their global architecture in the gene networks that regulate animal development is not well understood. This determined the structure of the core networks at the cis-regulatory level in early embryos of the chordate Ciona intestinalis by chromatin immunoprecipitation (ChIP) of 11 TFs. The regulatory systems of the 11 TF genes examined were tightly interconnected with one another. By combining analysis of the ChIP data with the results of previous comprehensive analyses of expression profiles and knockdown of regulatory genes, it was found that most of the previously determined interactions are direct. Focus was placed on cis-regulatory networks responsible for the Ciona mesodermal tissues by examining how the networks specify these tissues at the level of their cis-regulatory architecture. Many interactions were found that had not been predicted by simple gene knockdown experiments, and a significant fraction of TF-DNA interactions were found to make major contributions to the regulatory control of target gene expression (Kubo, 2010).

The developmental fates of blastomeres in the Ciona embryo have been determined by the gastrula stage. A comprehensive study has revealed that 53 TF genes are zygotically expressed and regulate one another in complex networks before gastrulation begins. To dissect the architecture of these networks at the level of protein-DNA interactions, focus was placed on 11 TF genes that play core roles in gene regulatory networks for endomesoderm specification: Brachyury, FoxA-a, FoxD, MyoD, Neurogenin, Otx, Snail, SoxC, Tbx6b, Twist-like1 and ZicL. Because the Ciona genome contains multiple copies of FoxD, Tbx6b and ZicL as gene clusters and their precise copy numbers have not yet been determined, these genes are collectively referred to FoxD, Tbx6b and ZicL in this paper. Likewise, there are two copies of Twist-like1, which are highly similar to each other, and these are collectively referred to as Twist-like1 (Kubo, 2010).

Eleven gene-fusion constructs were prepared that encode GFP-tagged TFs expressed under the control of their own promoters (e.g. a fusion gene that encodes GFP-tagged Brachyury driven by the Brachyury promoter). When these constructs were introduced into eggs, the resultant embryos expressed the fusion genes at the same time and in the same blastomeres as the endogenous genes. Exceptions were the Twist-like1 and the Snail constructs. Twist-like1 is normally expressed in three cell lineages (A7.6, B7.7 and B8.5), but the construct drove Twist-like1-GFP expression only in the B7.7 and B8.5 lines. Snail expression in the notochord lineage is normally very weak. The Snail construct did not recapitulate this expression in the notochord lineage but did drive Snail-GFP expression in the remaining lineages (Kubo, 2010).

Expression of these genes did not affect embryonic morphology at the stage when the embryos were fixed. The fixed embryos were subjected to ChIP using anti-GFP antibodies, and subsequently to microarray analysis. To define significant regions, two programs were used employing totally different algorithms. DNA segments regarded as positive by both programs were defined as significant. To confirm that this approach successfully identified TF binding sites, the sequences of ZicL and Tbx6b binding regions defined with three different false discovery rates (FDRs) were analyzed, as the consensus binding motifs of these two TFs are known. The frequencies of matches to the consensus binding sequences for ZicL and Tbx6b around peaks in 0.1% FDR were generally better than in 0.01% and 1% FDRs. As expected, the frequencies of the consensus binding sequences for ZicL and Tbx6b were markedly higher around peaks in the identified regions, suggesting that the method was able to successfully identify the TF binding regions (Kubo, 2010).

Brachyury and Ci-tropomyosin-like are the only known direct targets of ZicL and Brachyury, respectively. As an independent confirmation, the TF binding sites of these genes was expected. The ZicL ChIP profile showed a sharp peak around two known strong ZicL binding sites. The Brachyury ChIP profile also showed a peak around the known Brachyury binding site in the Ci-tropomyosin-like promoter. These peaks were included in significant regions identified with all the FDRs described above. ChIP-qPCRs were performed for these two known interactions. The ChIP-qPCR results showed excellent agreement with the ChIP-chip results (Kubo, 2010).

Next, the promoters were examined of genes that were identified in previous studies as likely direct targets of one of the 11 TFs on the basis of expression assays and gene knockdown assays. Among 29 interactions that had been found in the gene knockdown assays and for which both the source and target genes are expressed in the same cells, 28, 23 and 19 interactions were indicated to be direct under the FDRs of 1%, 0.1% and 0.01%, respectively. The remainder of the interactions were not regarded as direct. Otx expression in the A-line lineage requires a cis-regulatory module that includes Fox binding sites and is suppressed in FoxA-a morphants. The FoxA-a binding to this cis-regulatory element was counted with FDRs of 1% and 0.1%, but not with the most stringent FDR (0.01%). Similarly, several lines of evidence have suggested that MyoD is directly regulated by ZicL. First, MyoD expression is suppressed in ZicL morphants. Second, MyoD and ZicL are both expressed in presumptive muscle cells and the time windows of their expression overlap. Lastly, there is a putative ZicL binding site near to the peaks found in the MyoD upstream region. This putative binding was observed under the FDRs of 1% and 0.1%, but not under the most stringent FDR of 0.01%. On the basis of the above observations, in the following sections the results obtained at an FDR of 0.1% are generally described (Kubo, 2010).

The frequencies of the consensus sequences for ZicL and Tbx6b binding were markedly higher around peaks in the identified regions. Since the consensus binding motifs of the other nine TFs had not been determined previously, similar analyses was performed with motifs of homologs in other animals. The frequencies of the consensus binding motifs for six of the TFs, but not FoxD, SoxC or Twist-like1, were markedly higher around peaks. Because the position weight matrices (PWMs) for FoxD, SoxC and Twist-like1 gave higher background, no significant changes were seen. However, the number of matches to the motifs was markedly higher around peaks than in flanking regions and the background. These observations suggested that the method was able to successfully identify the TF binding regions (Kubo, 2010).

As has been reported in other animals, it was found that the regions bound by Brachyury, MyoD, Neurogenin, Snail, Tbx6b, Twist-like1 and ZicL, especially around the peaks, showed a marked GC bias. This bias is likely to be related to the consensus sequences, because the consensus sequences for these TFs are generally more GC-rich than those of the remaining TFs. The observed enrichment of recognition sequences was unlikely to be an artifact of GC bias because even if background sequences were picked with a base composition comparable to the averaged GC content of the bound regions (the difference between the average GC content of the bound and background regions was less than 0.8%), matches to the PWMs were enriched around peaks versus each of the GC-adjusted backgrounds (Kubo, 2010).

Next, attempts were made to discover overrepresented motifs in the regions (360 bp) around the peaks identified by each ChIP experiment using the Trawler program. It was found that overrepresented motifs were similar to the PWMs that were determined experimentally (Tbx6b and ZicL) or to those of homologs in other animals (the remaining nine TFs). This further supported the conclusion that the results of the ChIP experiments were of high quality (Kubo, 2010).

It is generally believed that TFs tend to bind near promoters, although many examples are known in which TFs bind to enhancers far from promoters. The distributions of peaks in all experiments, except Snail ChIP, were higher around transcription start sites. The reason why Snail binding sites were not enriched around transcription start sites is unclear, but this does not necessarily indicate that the results of the Snail ChIP were of low quality. Altogether, these observations support the conclusion that all of the ChIP experiments revealed in vivo occupancies of the TFs (Kubo, 2010).

TF genes were significantly enriched among the target genes of the 11 TFs. Among 670 potential TF genes in the Ciona genome, at least 607 encode proteins with known TF motifs or proteins with two or more zinc-finger motifs that potentially bind to DNA. A significantly greater number of TF genes were found among the targets than would be expected from random sampling. This enrichment indicates that the TFs examined bind targets selectively and not randomly (Kubo, 2010).

The ChIP data was compared with the results of the comprehensive gene knockdown experiments of a previously study. Among 76 interactions previously found in the early embryo, the ChIP assays indicated that 58 are direct. In addition, 251 novel interconnections were found. Among 121 (11×11) possible interconnections, 84 were observed in the present study. The data indicate that these genes are tightly interconnected with one another (Kubo, 2010).

Because the gene regulatory network model previously constructed from comprehensive expression profiles and comprehensive knockdowns of regulatory genes is of single-cell resolution, the ChIP data was interpred into this network by assuming that the examined TFs bind to the targets wherever their mRNAs are expressed. The reconstructed networks had a complex architecture (Kubo, 2010).

The reconstructed regulatory networks allow tracing of development at the single-cell level. Figs S8 and S9 in the supplementary material show the interconnections among the core 11 TFs in A-line and B-line blastomeres, which give rise to endomesodermal tissues, from the 8-cell to the early gastrula stage. Two of the three mesenchymal lineages (B-line mesenchymal cells) and 28 out of 36 muscle cells (B-line muscle cells) in the tadpole larvae are derived from B4.1 blastomeres at the 8-cell stage. Thirty-two and eight notochord cells are derived from A4.1 and B4.1 blastomeres, respectively. Previous studies demonstrated that Twist-like1, MyoD and Brachyury are essential for specification of the mesenchyme, muscle and notochord, respectively (Kubo, 2010).

Twist-like1 is expressed exclusively in the mesenchymal lineage and is regulated by FoxA-a, Otx and ZicL, as indicated by the fact that knockdown of any of these three genes results in loss or reduction of Twist-like1 expression. No direct binding was detected of FoxA-a to the Twist-like1 promoter, but it was found that FoxA-a binds to the upstream regions of Otx and ZicL, and that ZicL and Otx bind to the promoter of Twist-like1. Therefore, it is highly likely that FoxA-a mainly activates Twist-like1 indirectly through activating Otx and ZicL (Kubo, 2010).

Twist-like1 expression begins in B7.7 (the posterior B-line mesenchyme) at the 64-cell stage and in B8.5 (the anterior B-line mesenchyme) at the early gastrula stage. These two mesenchymal lines contribute to distinct adult tissues after metamorphosis. ZicL might be associated with the differences between these two lineages because the contribution of ZicL to Twist-like1 activation is weaker than that of Otx. To confirm this idea, a mutant Twist-like1 promoter was tested, from which a 150 bp segment containing the identified ZicL binding region was deleted. Because the Otx ChIP result indicated that the Otx binding region is distinct from the ZicL binding region, Otx was expected to bind to this mutant promoter. When introduced into fertilized eggs by electroporation, the wild-type promoter (1550 bp) drove reporter expression in 65% of the embryos, whereas the mutant promoter drove reporter expression in 36% of the embryos. In addition to the significant decrease in the number of embryos expressing the reporter, the overall fluorescence was weaker and the posterior B-line mesenchyme did not appear to express the reporter in the mutant construct. To confirm this observation, the experimental embryos were cleavage-arrested at the 110-cell stage. Cells in the arrested embryos cannot divide further, but the developmental programs proceed as in normal embryos. The mutant construct failed to drive reporter expression in the posterior B-line mesenchyme. These results suggest that ZicL contributes to the difference between these two lineages (Kubo, 2010).

A previous study showed that nine mesenchyme-specific non-regulatory genes are under the control of Twist-like1. None of these genes was identified as a direct target in the present study. Even when applied with an FDR of 1%, only one gene was identified as a direct target. Therefore, it is likely that Twist-like1 regulates the expression of mesenchyme-specific genes through its downstream regulatory gene circuit, although there is a possibility that Twist-like1 binds to the regulatory elements of these genes at later stages (Kubo, 2010).

The B6.2 and B6.4 cell pairs in the 32-cell embryo have the potential to give rise to mesenchyme and muscle. At the 64-cell stage, these cells divide, and one of the daughter cells becomes specified to give rise to the muscle cells. Previous functional assays showed that ZicL, Tbx6b and MyoD are essential for specification of muscle cells. Tbx6b begins to be expressed at the 16-cell stage, and cells expressing Tbx6b give rise not only to muscle cells but also to mesenchyme cells. Tbx6b expression declines to undetectable levels before the tailbud stage. ZicL starts to be expressed at the 32-cell stage in a variety of cells, including those with developmental fates of muscle, mesenchyme, notochord and neurons. ZicL expression in the muscle lineage disappears before the late gastrula stage. MyoD expression begins at the 44-cell stage exclusively in the muscle lineage under the control of Tbx6b and ZicL. The present study showed that ZicL, Tbx6b and MyoD constituted a tightly interconnected gene circuit that is responsible for this specification: (1) ZicL bound to the promoters of MyoD and Tbx6b; (2) Tbx6b bound to the promoters of MyoD and ZicL; and (3) MyoD bound to the promoter of Tbx6b and to its own promoter. All of these interactions, except MyoD binding to the Tbx6b promoter, have been confirmed by functional assays (Kubo, 2010).

To understand how this gene circuit regulates downstream muscle-specific genes, the promoters were examined of 13 muscle structural genes that are well annotated and known to be expressed in the larval tail muscle. Of these, ten were directly bound by MyoD and Tbx6, one by MyoD and ZicL, one by Tbx6b and ZicL, and one by MyoD alone (Kubo, 2010).

Both MyoD and Tbx6 bound to the promoters of more than three-quarters of the muscle genes examined. To test the action of this feed-forward loop comprising MyoD and Tbx6b in the regulation of muscle-specific gene expression, the expression patterns of genes under the control of this circuit were examined. Of the 155 genes under the direct control of MyoD and Tbx6b, 50 (including the above ten) were already known to be expressed in muscle cells. From the remaining 105 genes, 20 were randomly chosen, and 15 were found to be expressed in muscle cells, suggesting that this circuit is widely used for the regulation of genes expressed in muscle cells, and also that this circuit might not necessarily be sufficient for driving expression of the target (Kubo, 2010).

Brachyury is activated at the 64-cell stage exclusively in the notochord lineage, and this expression specifies the notochord fate. ZicL directly binds to the Brachyury promoter and activates its expression. It has also been shown that FoxD and FoxA-a are required for Brachyury expression, probably through regulating ZicL expression, and that FGF signaling is also required for Brachyury expression. The present assays showed that not only ZicL, but also FoxD binds to the Brachyury promoter. Although FoxD mRNA is not present in the notochord lineage at the 32-cell and 64-cell stages, when ZicL and Brachyury are activated, respectively (FoxD is expressed in the ancestors of cells in which ZicL and Brachyury are expressed), the ChIP assay indicated that FoxD binds to the promoters of ZicL and Brachyury. Because knockdown of FoxD eliminates ZicL and Brachyury expression and because the FoxD-GFP fusion protein exists in the notochord lineage at the 32-cell stage, it is likely that FoxD protein exists in these cells and binds to the promoters of ZicL and Brachyury when these two genes begin to be expressed (Kubo, 2010).

FoxA-a binding to the Brachyury promoter was not identified under 0.1% FDR. There was, however, a small peak that was counted as significant under 1% FDR. The possibility could not be ruled out that FoxA-a binds weakly to the Brachyury promoter. It is also possible that FoxA-a could bind weakly to a FoxD binding site because the FoxA-a binding peak coincided with that of FoxD. Even if this weak binding occurs in vivo, the regulation of Brachyury by FoxA-a would largely be achieved indirectly through FoxD and ZicL, since strong binding was found of FoxA-a to the promoters of FoxD and ZicL (Kubo, 2010).

Next, 14 non-regulatory genes were examined that are known to be expressed in the notochord under the control of Brachyury. Among them, 11 were identified here as direct targets of Brachyury. The present results suggest that the remaining three genes are regulated indirectly through a gene circuit under the control of Brachyury, although it cannot be ruled out that Brachyury binds to the regulatory elements of these three genes at later stages (Kubo, 2010).

The present study found many interactions between TFs and genomic DNA that were unexpected from preceding gene knockdown assays. Similar observations were also reported in preceding ChIP studies. To estimate what proportion of the binding makes a major contribution to gene regulation in Ciona embryos, MyoD mRNA or an MO against MyoD was injected into eggs, and their effects were analyzed on the expression of the same targets that were analyzed at the gastrula stage or at the tailbud stage, respectively. The mRNA levels of 14 targets, ten of which were expressed in muscle, were significantly increased (>2-fold) in embryos injected with MyoD mRNA, and MyoD MO injection significantly reduced the mRNA levels of three of these targets. The mRNA level of one target (KH.C12.38), which was weakly expressed in muscle at the tailbud stage, was significantly decreased in embryos injected with MyoD mRNA, whereas the mRNA level of one target (KH.C9.27), which was expressed in muscle at the gastrula stage, was significantly increased in embryos injected with the MyoD MO. In total, the mRNA levels of 16 targets were significantly altered by MyoD mRNA overexpression or gene suppression. The remaining four were not significantly affected, although three of these were expressed in muscle, implying that MyoD binding makes a relatively small contribution to activating these target genes. It was also found that eight of 15 Brachyury targets and seven of 12 Twist-like1 targets were significantly affected in the embryos by overexpression or knockdown of Brachyury or Twist-like1, respectively. Therefore, it is estimated that more than half of TF binding makes a major contribution to the regulatory control of gene expression (Kubo, 2010).

Expression of vertebrate Twist homologs

Neural patterning occurs during early development, soon after neural induction. In Xenopus, several caudalizing factors transform anterior neural tissue to posterior neural tissue at the open neural plate stages, while other factors are responsible for setting up mediolateral polarity, which becomes the dorsoventral (D-V) axis after neural tube closure. Many Wnt ligands are expressed in the neural tube in distinct anteroposterior (A-P) and D-V domains, implying a function in neural patterning. Xwnt7B induces neural crest markers Xslug and Xtwist in ectodermal explants coinjected with neural inducer noggin and in ectodermal cells neuralized by dissociation. In vivo, Xwnt7B expands the Xtwist expression domain when injected in the animal pole. These results suggest that Wnt members are involved in dorsoventral patterning of the neural tube (Chang, 1998).

The markers Xslug, Xsnail, and Xtwist all are expressed in the presumptive neural folds and are thought to delineate the presumptive neural crest. However, their interrelationship and relative spatiotemporal distributions are not well understood. A detailed in situ hybridization analysis of the relative patterns of expression of these transcription factors from gastrulation through neurulation and post-neural crest migration is presented. The three genes mark the prospective neural crest and roof plate, coming on sequentially, with Xsnail preceding Xslug preceding Xtwist. By combining gene expression analysis with a fate map of the same region using DiI labeling, the correspondence between early and late domains of gene expression has been determined. At the beginning of gastrulation, Xsnail is present in a unique domain of expression in a lateral region of the embryo in both superficial and deep layers of the ectoderm, as are Xslug and Xtwist. During gastrulation and neurulation, the superficial layer moves faster toward the dorsal midline than the deep layer, producing a relative shift in these cell populations. By early neurula stage, the Xsnail domain is split into a medial domain in the superficial ectoderm (fated to become the roof plate) and a lateral domain in the deep layer of the ectoderm (fated to become neural crest). Xsnail is down-regulated in the most anterior neural plate and up-regulated in the posterior neural plate. These results show that changes in the expression of Xsnail, Xslug, and Xtwist are a consequence of active cell movement in some regions coupled with dynamic changes in gene expression in other regions (Linker, 2000).

The murine homolog of the Drosophila twist gene is essential for head mesenchyme formation; it acts as an inhibitor of muscle differentiation. M-twist expression patterns from day 7 post coitum to day 18, indicate a more general function of the M-twist gene. At day 7, M-twist is expressed in the mesoderm outside the primitive streak. Later M-twist message is found predominantly in the somites, the head mesenchyme, the branchial arches, the limbs, and in the mesenchyme underneath the epidermis. Beginning at day 8, M-twist is mainly expressed in undifferentiated cells committed to muscle and cartilage development: this expression is consistent with a suggested role for M-twist in inhibiting overt muscle and cartilage differentiation. However, during organogenesis, M-twist is expressed in several areas of mesenchyme-epithelia interactions, suggesting additional tissue specific functions (Fuchtbauer, 1995).

M-twist is the murine homolog of the Drosophila twist gene. Before gastrulation, M-twist transcripts are detected in morulae and blastocysts, then in extra-embryonic tissues of early implanted mouse embryos before the onset of gastrulation. M-twist is expressed in the embryo proper at the egg cylinder stage, within some embryonic ectodermal cells of the primitive streak. Slightly later, scattered cells within the amniotic cavity apparently detached from the primitive streak also express the gene. M-twist transcripts gradually accumulate in head mesenchyme, the first aortic arches, somites and lateral mesoderm and, as development proceeds, successively the second, third and fourth branchial arches, the anterior limb buds and, finally, the posterior limb buds.

Thus M-twist expression in implanted embryos occurs first along a dorso-ventral gradient pattern until the headfold stage, then is gradually observed along the rostro-caudal axis of the embryo as development proceeds in the mesodermal cell layer and in neural crest cell derivatives (Stoetzel, 1995). M-twist knockout mice have deficits in head mesenchyme, somites and limb buds. Head mesenchyme is required for cranial neural tube closure. M-twist acts in a cell-autonomous manner here (Chen, 1995).

The genes in the twist family of bHLH transcription factors are essential for embryogenesis of both vertebrates and invertebrates, as demonstrated by the embryonic lethal phenotypes of the Drosophila and mouse twist-null mutants. The appearance and presence in the course of early embryogenesis of murine Twist protein (MTwist) was followed with a monoclonal antibody, alphaTwiMab-1, the first antibody generated against any of the vertebrate Twist proteins. The specificity for MTwist was demonstrated in comparative Western blot experiments, reticulocyte lysate assays, and immunohistochemistry of embryonic mouse samples. Consistent with its probable role as a transcription factor, MTwist localizes to the nuclei. MTwist signal first appears in 8- to 10-somite embryos at 8.25 dpc in the cranial neural crest cells and branchial arches, in the limb-bud mesenchyme and the somatic lateral plate, and in the sclerotome and the dermatome of the somites. The presence of MTwist protein in these tissues corroborates the reported MTwist RNA distribution and the phenotype of the MTwist-null mutants. There also emerges, however, an unexpected difference between MTwist RNA and protein expression. No protein can be detected prior to 8.25 dpc, despite the reported high levels of transcripts as early as 7.0 dpc. The presomitic mesoderm, epithelial somites, and anterior mesoderm express abundant MTwist RNA, but no protein. These results suggest posttranscriptional downregulation of MTwist in these regions. A proposed role of MTwist in somite formation and maturation is the inhibition of myogenic bHLH and MEF2 genes and thus the prevention of premature and/or ectopic differentiation in the presomitic mesoderm and epithelial somites. The absence of MTwist protein from these areas indicates that its role in somitogenesis must be reevaluated (Gitelman, 1997).

The human Twist protein consists of 201 amino acids with 96% amino acid sequence identity to mouse Twist, and 100% sequence conservation in the DNA-binding region among all species in which it has been characterized. The protein has six putatitive protein kinase C (See Drosophila PKC) phosphorylation sites. H-twist is expressed in placenta, skeletal muscle, kidney and pancreas but not in brain (Wang, 1997).

During vertebrate embryogenesis, the paraxial mesoderm becomes segmented into somites that form as paired epithelial spheres with a periodicity that reflects the segmental organization of the embryo. As a somite matures, the ventral region gives rise to a mesenchymal cell population, the sclerotome, that forms the axial skeleton. The dorsal region of the somite remains epithelial and is called the dermomyotome. The dermomyotome gives rise to the trunk and limb muscle and to the dermis of the back. Epaxial and hypaxial muscle precursors can be attributed to distinct somitic compartments that are laid down prior to overt somite differentiation. Inductive signals from the neural tube, notochord, and overlying ectoderm have been shown to be required for patterning of the somites into these different compartments.

Paraxis is a basic helix-loop-helix transcription factor expressed in the unsegmented paraxial mesoderm and throughout epithelial somites before becoming restricted to epithelial cells of the dermomyotome. Paraxis is one of a group of mammalian Twist related proteins that do not have overlapping patterns of expression with the characterized vertebrate Twist homologs. For example, the mouse Paraxis protein, called bHLH-EC2, has localized to the paraxial mesoderm and the dermomyotome by the time that Mouse Twist expression begins, after the evolution of the mesoderm into somites. It is expressed in the sclerotome where Paraxis is absent. paraxis is expressed as early as the primitive streak stage (Sosic, 1997).

To determine whether paraxis might be a target for inductive signals that influence somite patterning, the influence of axial structures and surface ectoderm was studied on paraxis expression by performing microsurgical operations on chick embryos. These studies revealed two distinct phases of paraxis expression: an early phase in the paraxial mesoderm that is dependent on signals from the ectoderm and independent of the neural tube, and a later phase that is supported by redundant signals from the ectoderm and neural tube. Under experimental conditions in which paraxis fails to be expressed, cells from the paraxial mesoderm fail to epithelialize and somites were not formed. Surface ectoderm is sufficient to induce paraxis expression in segmental plate mesoderm. These results demonstrate that somite formation requires signals from adjacent cell types and that the paraxis gene is a target for the signal transduction pathways that regulate somitogenesis. In paraxis-null mouse embryos, there is no morphological evidence for the dermomyotome, but Pax-3, which is a marker for the dermomyotome, is expressed in the correct region of the presumptive somite. Similarly, the myotome and sclerotome are specified in paraxis mutant embryos. It is therefore believed that paraxis controls epithelialization of the paraxial mesoderm, but it is not involved in specifying the fates of somitic cell lineages. It is tempting to speculate that paraxis functions in a pathway that relays inducive signals (which are required for the epithelialization steps in somitogenesis) from the ectoderm and neural tube to the genes encoding cell adhesion, cytoskeletal, or matrix molecules: for example, N-cadherin (see Drosophila Cadherin-N), N-CAM, Fibronectin and Laminin (Sosic, 1997).

Vertebrate Twist homologs and limb development

Mouse Twist is essential for cranial neural tube, limb and somite development. To identify the molecular defects disrupting limb morphogenesis, expression of mesenchymal transcription factors involved in patterning and the cell-cell signaling cascades controlling limb bud development were examined. These studies establish that Twist is essential for maintenance and progression of limb bud morphogenesis. In particular, the SHH/FGF signaling feedback loop operating between the polarizing region and the apical ectodermal ridge (AER) is disrupted. These defects in epithelial-mesenchymal signaling are most likely a direct consequence of disrupted fibroblast growth factor (FGF) signaling in Twist-deficient limb buds. In early limb buds, down-regulation of Fgf receptor 1 and Fgf10 expression in the mesenchyme occurs concurrent with loss of Fgf4 and Fgf8 expression in the AER. Finally, Twist function, most likely by regulating FGF signaling, is required for cell survival since apoptotic cells are detected in posterior and distal limb bud mesenchyme (Zuniga, 2002).

The activation of 5'HoxD genes in the limb bud mesenchyme precedes SHH signaling by the polarizing region. The early domains of 5'HoxD genes are posteriorly restricted and their nested domains provide good indicators of early limb bud mesenchymal patterning. While Hoxd11 expression is initially restricted to the posterior half of wild-type limb buds, expression is lower and anteriorly expanded in Twist-deficient limb buds. Hoxd13 is the 5'-most HoxD gene activated and thereby most posteriorly restricted (temporal and spatial co-linearity principle. In Twist-deficient hindlimb buds, Hoxd13 expression remains similar to wild-types, while its posterior restriction is lost and expression down-regulated in mutant forelimb buds. In contrast, the spatial expression of 5'HoxD genes along the rostro-caudal body axis is not affected in Twist-deficient embryos. These results show that while Twist is dispensable for activation of 5' HoxD genes, it is required for maintaining their spatial expression during progression of limb bud outgrowth (Zuniga, 2002).

The SHH/FGF feedback loop is required for regulating 5'HoxD expression during progression of limb bud development. Therefore, the observed alterations in 5'HoxD gene expression point to possible defects in epithelial-mesenchymal signaling. A signal relay mechanism involving the BMP antagonist Gremlin is critical for establishment of the SHH/FGF feedback loop between the polarizing region and the AER. In Twist-deficient hindlimb bud mesenchyme, Shh expression is reduced and its domain expanded more distal-anteriorly in comparison to wild-type controls. Accordingly, the expression domain of Gremlin in the mesenchyme is also expanded anteriorly in mutant hindlimb buds. In mutant forelimb buds, only low Shh levels are detected and Gremlin expression is displaced distal-anteriorly. In younger limb buds, expression of the SHH transcriptional target Patched is initially up-regulated in posterior mesenchyme, suggesting that activation of SHH signaling is not affected, while its maintenance is disrupted. Indeed, Fgf4 expression is initially also rather normal in mutant hind limb buds, while no expression is detected in developmentally more advanced mutant forelimb buds. These results establish that while Shh and other signals acting in the SHH/FGF feedback loop are activated in Twist-deficient limb buds, their maintenance and timely distal-anterior propagation is disrupted (Zuniga, 2002).

Loss of Twist gene function arrests the growth of the limb bud shortly after its formation. In the Twist-/- forelimb bud, Fgf10 expression is reduced, Fgf4 is not expressed, and the domain of Fgf8 and Fgfr2 expression is altered. This is accompanied by disruption of the expression of genes (Shh, Gli1, Gli2, Gli3, and Ptch) associated with SHH signaling in the limb bud mesenchyme, the down-regulation of Bmp4 in the apical ectoderm, the absence of Alx3, Alx4, Pax1, and Pax3 activity in the mesenchyme, and a reduced potency of the limb bud tissues to differentiate into osteogenic and myogenic tissues. Development of the hindlimb buds in Twist-/- embryos is also retarded. The overall activity of genes involved in SHH signaling is reduced. Fgf4 and Fgf8 expression is lost or reduced in the apical ectoderm, but other genes (Fgf10, Fgfr2) involved with FGF signaling are expressed in normal patterns. Twist+/-;Gli3+/XtJ mice display more severe polydactyly than that seen in either Twist+/- or Gli3+/XtJ mice, suggesting that there is genetic interaction between Twist and Gli3 activity. Twist activity is therefore essential for the growth and differentiation of the limb bud tissues as well as regulation of tissue patterning via the modulation of SHH and FGF signal transduction. The finding of the down-regulation of the Gli genes in the Twist mutant limb mesenchyme is concordant with the observation that the expression of a Gli-related gene (lame duck) is also altered by the loss of Twist function in the Drosophila embryo (O'Rourke, 2002).

The basic helix-loop-helix transcription factor Twist1 is essential for normal limb development. Twist1-/- embryos die at midgestation. However, studies on early limb buds found that Twist1-/- mutant limb mesenchyme has an impaired response to FGF signaling from the apical ectodermal ridge, which disrupts the feedback loop between the mesenchyme and AER, and reduces and shifts anteriorly Shh expression in the zone of polarizing activity. This study combined Twist1 null, hypomorph and conditional alleles to generate a Twist1 allelic series that survives to birth. As Twist1 activity is reduced, limb skeletal defects progress from preaxial polydactyly to girdle reduction combined with hypoplasia, aplasia or mirror symmetry of all limb segments. With reduced Twist1 activity there is striking and progressive upregulation of ectopic Shh expression in the anterior of the limb, combined with an anterior shift in the posterior Shh domain, which is expressed at normal intensity, and loss of the posterior AER. Consequently limb outgrowth is initially impaired, before an ectopic anterior Shh domain expands the AER, promoting additional growth and repatterning. Reducing the dosage of FGF targets of the Etv gene family, which are known repressors of Shh expression in anterior limb mesenchyme, strongly enhances the anterior skeletal phenotype. Conversely this and other phenotypes are suppressed by reducing the dosage of the Twist1 antagonist Hand2. These data support a model whereby multiple Twist1 activity thresholds contribute to early limb bud patterning, and suggest how particular combinations of skeletal defects result from differing amounts of Twist1 activity (Krawchuk, 2011).

Vertebrate Twist homologs: Protein Interaction with Daughterless homologs

Continued: see twist Evolutionary homologs part 2/2

twist: Biological Overview | Regulation | Targets of Activity and 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.