BIOLOGICAL OVERVIEW

Teashirt plays three essential roles during embryogenesis, each an aspect of determining segmental identity in the trunk (thorax and abdomen excluding the tail). First, tsh is required globally for segmental identity throughout the entire trunk, in contrast to the more specific roles of the "classical" homeotic genes. Second, tsh is a critical requirement for the proper identity of the anterior prothorax. Third, teashirt is required for the subdivision of midgut mesoderm, acting here in partnership with the homeotics.

In its first role, transcripts from the tsh gene will later accumulate in segments destined to acquire trunk identities. It is, in fact, one of the main determinants of trunk identity. tsh is transcribed far too early in embryogenesis to be regulated initially by homeotic genes. It is therefore clear that its early fuction is independent of the homeotics. tsh gene activity is required later for the normal function of the ANTP-C and BX-C genes, which modulate in part the expression of tsh (Röder, 1992).

In its second role, TSH protein is instrumental in establishing the type of segment made by the gene Sex combs reduced(scr). In the presence of TSH, SCR directs cells to a prothoracic identity. In the absence of TSH, SCR directs cells to a labial identity (Röder, 1992). Ectopic expression of tsh in anterior segments results in the replacement of head segment patterns with trunk patterns (de Zulueta, 1994).

Analysis of a tissue specific enhancer element in the tsh gene clarifies its third developmental role. Here, both ANTP and UBX regulate tsh in the epidermis and somatic mesoderm. Separable elements drive tissue specific activation by ANTP and UBX. tsh is necessary for proper formation of anterior and central midgut structures. Earlier studies indicated that the actions of the homeotic genes Ubx and abd-A were mediated by dpp and wingless (Mathies, 1994), but further analysis of the tsh promoter indicates direct action by UBX and ABD-A (McCormick, 1995).

One function of the Wingless signal cascade in Drosophila larvae is to determine the "naked" cuticle cell-fate choice as opposed to the denticled one. Wingless stabilizes cytoplasmic Armadillo, which may act in a transcriptional activator complex with the DNA-binding protein T-cell factor (also known as Pangolin). As these components are critical for all Wingless-dependent patterning events, the problem arises as to how specific outputs are achieved. The Teashirt zinc finger protein is found to be necessary for a subset of late Wingless-dependent functions in the embryonic trunk segments where the teashirt gene is expressed. Loss of wg results in larvae with smaller segments covered with denticles and with no naked cuticle, a phenotype similar to hypomorphic loss-of-function mutations in armadillo. Loss of tsh function affects the identity of the prothorax, which is replaced by labial identity, but segment identity is not affected in the posterior thorax or in the eight abdominal segments. The size of the posterior segments in the ventral epidermis is reduced in the trunk in tsh -null mutants, as compared with wild-type larvae. In tsh -null larvae, abnormal denticle belts differentiate with smaller naked regions in-between, showing that tsh is required for patterning of both the naked and the denticled regions. Notably, denticles differentiate in the naked cuticular regions, especially in the ventral midline, of tsh-null larvae; this phenotype resembles that which arises from late loss-of-Wg signaling in the cell-specification phase (Gallet, 1998).

Armadillo and Teashirt proteins showed similar Wingless-dependent modulation patterns in homologous parts of each trunk segment in embryos, with high levels of nuclear Teashirt and intracellular Armadillo within cells destined to form naked cuticle. Overexpression of Wingless, Tsh and stabilized Arm produce a naked cuticle phenotype. It is concluded that the production of naked cuticle is temporally correlated with high levels of Tsh protein and stabilized, cytoplasmic Arm. Mutations in tsh therefore resemble those of the Wg signaling cascade, but Tsh has an independent homeotic function (Gallet, 1998).

Transcription of the wg gene depends initially on regulation from the pair-rule genes, and then is maintained differently during the early (cell-stabilization) and late (cell-specification phases of embryogenesis. During the cell-stabilization phase (stages 7-10), Wg ensures the regulation of the target gene engrailed, but thereafter, during the cell-specification phase (stage 10 onward), Wg is required to maintain wg but not en expression. Teashirt is found to be required for the maintenance of the late Wingless signaling target gene wingless but not for an earlier one, engrailed (Gallet, 1998).

The Tsh protein is detected initially in the cytoplasm soon after the blastoderm stage in a central region of the embryo. By stage 9, Tsh occupies the trunk segments and is modulated, being concentrated in the nuclei in homologous regions of each neuromere. The domains of high nuclear Tsh coincide with high cytoplasmic localization of Arm. Lower levels of intracellular Arm and nuclear Tsh separate the stronger intracellular stripes. By stage 10, Tsh retains a segmentally modulated pattern in dorsal and lateral parts of each trunk segment, with high nuclear Tsh overlapping the En-positive cells, but in the most ventral parts, Tsh becomes highly concentrated in the nuclei of all cells. Therefore, there is a correlation between high levels of Tsh and Arm inside the cells at stage 9 and the production of naked cuticle in wild-type embryos. arm function is required for the nuclear modulation of Tsh in the trunk segments during stage 9-10 (Gallet, 1998).

Teashirt has been found to associate with, and require, Armadillo in a complex for its function. Full-length Arm protein is not easily accessible to Tsh in vitro, indicating that Tsh and/or Arm needs to be modified in order to bind. Arm might act as a cytoplasmic to nuclear transporter for Tsh. It is concluded that Teashirt binds to, and requires, Armadillo for the naked cell-fate choice in the larval trunk. Teashirt is required for trunk segment identity, suggesting that Teashirt provides a region-specific output for Armadillo activity. Further modulation of Wingless is achieved in homologous parts of each trunk segment where Wingless and Teashirt are especially active. These results provide a novel, cell-intrinsic mechanism to explain the modulation of the activity of the Wingless signaling pathway (Gallet, 1998).

Nubbin and Teashirt mark barriers to clonal growth along the proximal-distal axis of the Drosophila wing

The division of the wing imaginal disc into anterior, posterior, dorsal, and ventral compartments is a critical step in Drosophila wing morphogenesis. This study investigated the existence of cell lineage restrictions along the proximal-distal (PD) axis of the wing disc. The existence of classical compartment boundaries in the hinge region was ruled out, but it was demonstrated that there are clonal restrictions corresponding to the expression domains of two transcription factors, Nubbin (Nub) and Teashirt (Tsh), present in distal and proximal cells, respectively. Unlike classical compartments, the Nub and Tsh domains do not define absolute lineage restrictions. Instead, due to regulation by Wingless signaling, the Nub and Tsh expression boundaries shift during development. Once established, the Nub and Tsh domains, and the intervening region in which neither factor is expressed, grow independently, because the progeny of cells present in one domain do not freely populate an adjacent domain. Despite shifting position, the Nub and Tsh domain boundaries, like compartment boundaries, impact the expression of secreted signaling molecules. Thus, like the vein/intervein divisions of the wing and mammalian rhombomeres, the Nub and Tsh domains share some of the attributes of classical compartments, but lack their stringent and immobile boundaries (Zirin, 2007).

The experiments described here investigate whether the Drosophila wing disc is divided by lineage restrictions beyond the well-established anterior, posterior, dorsal, and ventral compartments. This question stemmed from a previous observation that, following the initiation of tsh repression by Wg signaling, tsh is maintained in a repressed state by the PcG genes. If tsh repression was a stably inherited state, as is typically the case for PcG regulation, these cells would be expected to define a distinct lineage. The results presented here are consistent with this view, but also reveal that the tsh repressed state does not define a compartment, using the strict definition of this term. Nevertheless, the experiments suggest that like the vein/intervein divisions in the wing disc, and the rhombomere divisions of the developing vertebrate brain, the Tsh, gap, and Nub domains have many of the properties of compartments (Zirin, 2007).

Three domains were defined whose boundaries restrict the pattern of cell divisions along the PD axis. These three regions are comprised of cells that are Nub+ and Tsh− (the Nub domain), Nub− and Tsh− (the gap domain), and Nub− and Tsh+ (the Tsh domain). In wild type discs, the two boundaries that separate these three domains cannot be considered strict lineage restrictions because examples were found of clones that cross these boundaries even when induced as late as the mid third instar. However, like compartment boundaries, the proximal Nub and distal Tsh expression boundaries clearly affect the patterns of cell division within the disc. Beginning in the 2nd instar, clones tend to grow along these boundaries for many cells. Importantly, this behavior is not typical of most gene expression boundaries in the wing disc. For example, neutral clones readily cross the rn, pnr and Iro-C expression boundaries. Also lineage tracing experiments demonstrate that by the late 2nd instar, these three domains grow largely as independent units, since the progeny of cells from one domain rarely move into the neighboring domain. Thus, it appears that the cells in the Nub, gap, and Tsh domains have a strong, but not absolute, tendency to maintain their gene expression status. Again, this behavior is in contrast to other gene expression domains in the wing disc such as rn, where the progeny of a rn expressing cell can readily turn off rn expression and contribute to a neighboring expression domain (Zirin, 2007).

The primary distinction between a compartment and the three PD domains described in this study is that the boundary of a compartment remains constant during development. The experiments demonstrate that both the Tsh and Nub boundaries, and their associated clonal restrictions, shift during development. The Tsh boundary shifts proximally, due to wg signaling in the second and third instar. Similarly, the experiments suggest that the Nub domain is gradually expanding over time. This conclusion stems from the observation that nub-Gal4 induced mitotic recombination frequently results in small clones at the edge of the Nub domain that sometimes straddle the Nub expression boundary. The small size of these clones suggests that they were generated late in development and contrasts with the typically much larger clones present in the middle of the Nub domain. Unlike tsh repression, which is dependent on Wg signaling, it is not known what signals induce the expansion of the Nub domain in the proximal direction. However, previous work proposed the existence of a diffusible molecule made in vg-expressing cells that is required to turn on nub (Zirin, 2007).

One consequence of the changes in the Nub and Tsh boundaries is that the gap domain expands in size during development. Initially, the gap domain comes from cells that turn off tsh in response to wg signaling. Following this initial tsh repression, the data suggest that the gap domain increases in size primarily due to proliferation of these early Tsh− Nub− cells. This conclusion is based primarily on tsh-Gal4 lineage data, which show that, by the end of the 2nd instar, tsh expressing cells contribute very little to the gap domain. Since this domain continues to grow during the 3rd instar, it is concluded that the domain must expand due to cell proliferation, consistent with earlier reports suggesting that Wg induces cell proliferation within this region of the disc (Zirin, 2007).

The initial motivation for analyzing lineage restrictions along the PD axis was the observation that tsh repression is maintained by PcG silencing. It was possible, therefore, that this initial tsh repression domain established a lineage restricted domain that is maintained for the remainder of development. The experiments suggest that this is not the case but that instead two imperfect clonal restrictions form along the PD axis that correspond to the Nub and Tsh boundaries. For example, in M+ clonal experiments, clear examples were observed of clones that extend across most of the gap domain, arguing that an invisible clonal restriction is not present in this region of the disc. Instead, clones made at the end of the 2nd instar or later have a strong tendency to respect the Nub and Tsh boundaries. These observations suggest that these gene expression boundaries create these clonal restrictions, a conclusion that is supported by the observation that nub− clones fail to respect the Nub boundary, and Tsh+ clones fail to respect the Tsh boundary. Thus, because the Tsh and Nub expression domains are changing during development, so are the barriers to clonal growth that are created by these boundaries. The changing position of these boundaries accounts for the behavior of the clones and lineages characterized in this study: they should largely respect these borders, but there should be exceptions to this rule since these domains change over time (Zirin, 2007).

What might the mechanism be for creating these clonal restrictions? One possibility, which is favored, is that once Nub is activated or once Tsh is repressed, these states of gene expression are heritably maintained even in the absence of the initiating signal. This is known to be true for the repression of tsh in the wing pouch, in which the initial repression by Wg is maintained by PcG silencing. This two-step repression of tsh also appears to hold true for the later phases of tsh repression in the hinge. Hinge clones that cannot transduce the Wg signal show tsh derepression when they are induced in the second instar, but not when they are induced in the third instar suggesting that, as in the pouch, tsh repression is maintained by a wg-independent mechanism. If nub activation is also maintained by a heritable mechanism, then it would be expected that clones generated in any of these three domains would remain in these domains. Thus, although a stable lineage restriction can be observed only when Wg signaling is blocked, it is suggested that the same phenomenon -- the epigenetic inheritance of gene expression states -- is occurring as new cells are recruited to the gap domain due to tsh repression or to the Nub domain due to nub activation (Zirin, 2007).

It is also noted that while cell affinity differences may also play a role in forming the Nub and Tsh domain boundaries, affinity differences appear not to be sufficient to create the clonal restrictions described in this sudy. Previous work has demonstrated, for example, that pnr expressing cells do not readily mix with non-pnr expressing cells, indicating a clear difference in cell affinities. Yet despite this affinity difference, the pnr expression boundary does not influence clonal growth. One possible distinction between these expression domains is that, unlike nub and tsh, the pnr gene expression status is not locked into place by an epigenetic (e.g. PcG-dependent) mechanism (Zirin, 2007).

Why have PD clonal restrictions? For both the D/V and A/P boundaries, it is known that compartmental interfaces create sources of secreted signals that are critical for the subsequent patterning and growth of the disc. Might the same phenomenon be occurring at the Nub and Tsh boundaries? wg expression in the IR is thought to be induced by a non-autonomous signal coming from vg-expressing cells and received by nub-expressing cells, and has been suggested to involve four jointed, fat, dachsous and dachs. At the time the IR is first induced, the Tsh and Nub domains abut each other, raising the possibility that this interface may also be important for IR induction. Although there is no definitive answer to this question, several results reported in this study are consistent with this view. First, in ap>TCF^DN discs it was found, intriguingly, that Wg expression is observed in cells immediately adjacent and distal to the Tsh/Nub interface, in Nub-expressing cells. Second, in flip-out clones that ectopically express Tsh in the hinge, often a non-autonomous induction of wg expression is observed in cells adjacent to the clone. Third, in flip-out clones expressing a tsh RNAi hairpin construct, de-repression of wg is observed in the RNAi-expressing cells. However, an intriguing aspect to this experiment is that wg expression is not only observed at the Tsh+/- interface, but throughout the tsh RNAi clones. Thus, although the Tsh boundary may play a role in inducing wg, the RNAi experiment raises the more conventional possibility that tsh is a repressor of wg expression in this region of the wing disc (Zirin, 2007).

Another role for these domains may be to allow the orientation of cell divisions to differ along the PD axis. In the wing pouch, the predominant pattern of clonal growth is parallel to the PD axis. In contrast, it was found that, by the end of the 2^nd instar, the predominant pattern of clonal growth in the gap domain is perpendicular to the PD axis. There is also a shift in clonal shape from relatively isometric in the mid second instar to long in the third instar. Interestingly, cells divide with a predominantly PD orientation in the wing pouch. Based on the shape of the clones observed, mitoses in the hinge may predominantly orient perpendicular to the PD axis. This orientation may help determine the shape of the adult hinge (Zirin, 2007).

Strikingly, analysis of the PD axis in the wing does not apparently apply to the ventral imaginal discs. Lineage tracing experiments performed in the leg disc indicate that cells in the Tsh domain readily lose tsh expression and contribute to the growth of the distal leg regions. Since there are no rings of Wg associated with the Tsh boundary in the leg disc, perhaps it is the unique relationship between Tsh and Wg in the wing disc that necessitates a more stringent clonal restriction (Zirin, 2007).

The developmental domains defined in this study correspond to the body/hinge and hinge/wing blade anatomical boundaries of the adult fly. As insect wings may have evolved from a proximal outgrowth of a pre-existing multi-branched appendage, it may be that the formation of distinct domains was important for the independent growth and morphological modification of this outgrowth to become a wing (Zirin, 2007).

In closing, it is suggested that the phenomenon of imperfect lineage restrictions defined in this study may be more general in animal development than classical compartments. Although lineage restrictions clearly exist in the vertebrate hindbrain and forebrain, it has also been noted that these boundaries are not absolute. Krox-20, a zinc-finger transcription factor responsible for the specification and segmentation of even-numbered rhombomeres, is expressed in a gradually expanding domain. It is suggested that cell proliferation and changes in the expression patterns of such key genes, as is the case for the Nub and Tsh domains in the wing disc, underlie the leakiness of these boundaries (Zirin, 2007).

GENE STRUCTURE

PROTEIN STRUCTURE

Amino Acids - 993

Structural Domains

teashirt has three distantly spaced C2-H2 zinc finger motifs (Fasano, 1991). The wide spacing of teashirt zinc fingers makes tsh a novel zinc

During the embryogenesis of Drosophila, the homeotic genes are required to specify proper cell fates along the anterior-posterior axis of the embryo. Partial cDNAs of homologs of the Drosophila homeotic gene teashirt and five of the homeotic-complex (HOM-C) genes were cloned from the thysanuran insect, Thermobia domestica (Td/the firebrat), and these genes were assayed for their embryonic expression patterns. The HOM-C genes examined were labial, Antennapedia, Ultrabithorax, abdominal-A and Abdominal-B. Since the expression pattern of these HOM-C genes is largely conserved among insects and since Thermobia is a member of a phylogenetically basal order of insects, the ancestral expression patterns of these genes in insects could be inferred. The expression patterns of the Thermobia HOM-C genes were compared with their expression in Drosophila and other insects; the potential roles these genes may have played in insect evolution are discussed. Interestingly, the teashirt homolog shows greater variability between Thermobia and Drosophila than any of the HOM-C genes. In particular, teashirt is not expressed strongly in the Thermobia abdomen, unlike Drosophila teashirt. It is proposed that teashirt expression has expanded posteriorly in Drosophila and contributed to a homogenization of the Drosophila larval thorax and abdomen (Peterson, 1999).

It is clear that there are marked differences between Td-tsh and Drosophila tsh expression: (1) Td-tsh does not appear to be expressed during early germ band extension and (probably) comes on after Antp; (2) during later germ band extension, Td-tsh appears to have little, if any, abdominal expression and instead appears to be expressed in the gnathocephalon, a region where Drosophila tsh has no embryonic expression. The differences in tsh expression combined with the evidence from Tribolium HOM-C mutational phenotypes suggests that the embryonic expression of Drosophila tsh throughout the trunk, its function as a trunk-specifying gene in general, and as an aT1-specifying gene in particular, may be newly derived, perhaps unique to drosophilids or other higher insects. Accordingly, it may have played a role in the evolution of the dramatic modifications of the Drosophila larval trunk (Peterson, 1999).

In contrast to Td-tsh, Drosophila tsh expression is first observed in the syncitial blastoderm in a large band that maps to the thoracic segments. By the end of germ band extension, the Drosophila Tsh protein is expressed with a dorsal-ventral shift; accumulation is observed in parasegments 3-13 ventrally and in segments T1-A8 dorsolaterally. In late stage embryos, Drosophila tsh is expressed strongly throughout the trunk, but not in the tail. In Drosophila, tsh cooperates with the HOM-C genes to promote development throughout the trunk and is required for the proper specification of PS3 ventrally and aT1 both ventrally and laterally. Eliminating all HOM-C gene expression from the trunk does not prevent trunk segments from forming. In embryos mutant for Scr, Antp, and the bithorax complex (BX-C) genes, all trunk segments have denticle belts with a mixture of T1 and T2 identity. Since distinguishable trunk segments fail to form only when both tsh and HOM expression are removed, tsh has been proposed to be the gene that specifies the 'ground state' of the Drosophila trunk. The fact that tsh and HOM-C function must be removed in Drosophila is strikingly different from what is observed in HOM-C minus embryos of Tribolium. These embryos differentiate all thoracic and abdominal segments as antennae, suggesting that in Tribolium HOM-C genes have greater control over the formation of thorax and abdomen and that either Tribolium tsh is not a trunk-specifying 'ground state' gene or HOM-C genes are required for its expression (Peterson, 1999).

The gnathal expression of Td-tsh is surprising because another function of tsh in the Drosophila embryo is to repress head development. For example, embryos homozygous for tsh null mutations have ectopic labial expression, produce labial sense organs and form procephalic head cuticle in the trunk. However, tsh plays a very different role in the development of the adult Drosophila head, which is morphologically much more similar to the firebrat than its embryonic head. In Drosophila imaginal discs, tsh is expressed in the head primordium, the eye-antennal disc. Adults homozygous for particular hypomorphic alleles of tsh have reduced maxillary palps, which is similar to a phenotype induced by a gain-of-function Antp mutation. Accordingly, tsh is required to repress Antp expression in certain regions of the eye-antennal disc. Although any function ascribed to Td-tsh or to the ancestral tsh is highly speculative, a model is proposed consistent with morphology, expression data and limited genetic analysis from Tribolium. It has been proposed that an ancestral function of early HOM-C gene expression is to divide the segmented embryo into its specific tagmata (procephalon, gnathos, thorax and abdomen), after which modulated HOM-C gene expression modifies the development of particular segments within each tagma to produce unique segment identities. The analysis of HOM-C expression patterns in non-drosophilid insects and mutant analysis in Tribolium is consistent with this idea. For example, late expression of Scr and Ubx in the thorax does not transform the thorax toward labium or abdomen, respectively, but instead modifies the development of thorax. One way this can be accomplished is by the early activation of autonomous subroutines, which provide an irreversible background for continued HOM-C function. The activation of tsh may be one of these subroutines (Peterson, 1999).

The expression of Td-tsh is consistent with tsh being a specifier of the thorax, rather than the entire trunk. Also, its rather late initiation of expression is compatible with it being downstream of the HOM-C genes, Td-Antp in particular. In this way, loss of HOM-C function would eliminate thoracic identity, as it appears to do in Tribolium. Once activated, tsh may establish a thoracic ground state upon which HOM-C-dependent modifications could be superimposed. In Drosophila, tsh expression has been modified, in part reflecting the homogenization of the thorax and abdomen. HOM-C-dependent modulation of tsh expression is still evident in that the elevated levels of tsh expression in the thorax are dependent on Antp; however, the initiation of tsh expression is HOM-C independent and extends throughout the abdomen. Such independence would allow Drosophila Antp to take on more mosaic expression, having been freed from the requirement of activating tsh, even early in development. The ability of tsh to repress head development may reflect an adoption of Antp function as Drosophila Antp evolved a mosaic pattern. In contrast, the expression of Td-tsh in the firebrat head is consistent with a similar role of Drosophila tsh in adults in the promotion of gnathal development. A genetic analysis of tsh in Tribolium and examination of its expresssion pattern in other insect orders would provide an enlightening test of these ideas (Peterson, 1999).

In Drosophila the teashirt gene, coding for a zinc finger protein, is active in specific body parts for patterning. For example, Teashirt is required in the trunk (thorax and abdomen) tagmata of the embryo, parts of the intestine and the proximal parts of appendages. Vertebrate cDNAs related to teashirt have been isolated. As in Drosophila, human and murine proteins possess three widely spaced zinc finger motifs and an acidic domain. The expression patterns of the two murine genes are described. Both genes show regionalized patterns of expression in the trunk, the developing limbs and the gut (Caubit, 2000).

Mouse ESTs fall into two classes that are referred to as mtsh1 and mtsh2. The vertebrate Tsh proteins share a conserved structural organization with Drosophila Tsh. mtsh1 expression is first detected in 9-9.5 day embryos (20-22 somite stage). Transcripts are only detected in the trunk region. At this stage, mtsh1 is expressed in the neural tube, the somites, the mesenchyme of the developing forelimb buds and in the region of the foregut. In the neural tube, the anterior border lies in the posterior part of the hindbrain whereas the caudal boundary lies adjacent to the last, newly-formed somite. Cross sections at the level of spinal cord reveal that mtsh1 transcripts are restricted to the dorsal part of the neural tube. In the paraxial mesoderm, mtsh1 transcripts are detected in differentiated somites. In the caudal part of the embryo, the expression levels diminish gradually. In transverse sections mtsh1 mRNAs are restricted to the dorsal part of the somite. Transcripts are also detected in the mesenchyme of the forelimb bud as well as in the distal parts of the first branchial arch. In 11 day embryos, mtsh1 expression is maintained in the dorsal part of the neural tube. However, the anterior boundary appears to regress posteriorly. In the somitic mesoderm, the anterior limit is at the boundary between somite 3/4. In transverse sections at the level of the trunk mtsh1 expression is maintained in the dermomyotome. At this stage, mtsh1 expression is also detected in the first and second branchial arches, with stronger levels in the more distal mesenchyme. At 10.5 d.p.c., mtsh1 mRNAs are detected in two proximal regions and in a more distally located posterior zone of the limb bud. At 11 days, mtsh1 expression extends distally through the progress zone. In these regions, mtsh1 expression is restricted to mesenchyme. At 13 days, mtsh1 is expressed in the mesenchyme at the tip of the digits and in a posterior region proximal to digit 5 (Caubit, 2000).

mtsh2 mRNA can be first detected in the presumptive forelimb buds in 9-9.5-day-old embryos. In contrast to mtsh1, mtsh2 is not detected in paraxial mesoderm nor in the neurectoderm at this stage. In 10.5 day embryos, strong expression is observed in the limb buds. In the forelimb, mtsh2 expression is detected throughout the mesenchyme underlying the apical ectodermal ridge; expression is more extensive in the ventral, compared to the dorsal part. At this stage, mtsh2 mRNA is detected in the somites just posterior to the hindlimb and in all somites up to the forelimb bud. Expression levels decrease gradually in the anteriorly located somites adjacent to the forelimb bud. mtsh2 is also expressed in part of the hindgut. At 12.5 days, mtsh2 transcripts are detected in two dorso-lateral regions of the neural tube, the limb buds and whisker pads. In 13.5 days, in the limb bud, mtsh2 mRNAs are localized in a proximal domain of the interdigital regions and at more proximal levels. mtsh2 expression is stronger near the lateral margin of digit 1. These results support the hypothesis that Teashirt function may have been conserved for patterning the primary axis and specification of limb structures (Caubit, 2000).

Three putative murine Teashirt orthologues specify trunk structures in Drosophila in the same way as the Drosophila teashirt gene

Drosophila teashirt functions as a region-specific homeotic gene that specifies trunk identity during embryogenesis. Based on sequence homology, three tsh-like (Tsh) genes have been identified in the mouse. Their expression patterns in specific regions of the trunk, limbs and gut raise the possibility that they may play similar roles to tsh in flies. By expressing the putative mouse Tsh genes in flies, evidence is provided that they behave in a very similar way to the fly tsh gene: (1) ectopic expression of any of the three mouse Tsh genes, like that of tsh, induces head to trunk homeotic transformation; (2) mouse Tsh proteins can rescue both the homeotic and the segment polarity phenotypes of a tsh null mutant; (3) following ectopic expression, the three mouse Tsh genes affect the expression of the same target genes as tsh in the Drosophila embryo; (4) mouse Tsh genes, like tsh, are able to induce ectopic eyes in adult flies; (5) all Tsh proteins contain a motif that recruits the C-terminal binding protein and contributes to their repression function. As no other vertebrate or fly protein has been shown to induce such effects upon ectopic expression, these results are consistent with the idea that the three mouse Tsh genes are functionally equivalent to the Drosophila tsh gene when expressed in developing Drosophila embryos (Manfroid, 2004).

Denticles are replaced with naked cuticle when the dose of Tsh is increased by combining two insertions. In addition, Tsh genes rescue the segment polarity phenotype of tsh⁸ null mutants. These results indicate that Tsh genes can operate in the gene network involved in formation of naked cuticle. In tsh⁸ null mutants, wg expression is not maintained in the trunk and the reduced naked cuticle domains resemble late wg loss-of-function phenotype. The rescue of the tsh⁸ cuticular phenotype by Tsh genes suggests that these genes are sufficient to ensure wg transcription and/or signalling in the posterior part of each segment as seen in wild-type flies. Although the maintenance of wg in the tsh⁸ trunk was not assessed upon Tsh and tsh expression, Tsh genes, like tsh, are able to maintain wg expression in the gnathal segments. This would suggest that, like tsh, Tsh genes could control wg expression in the trunk as well as in the head. An autoregulatory loop involving Tsh in the Wg signalling pathway has been postulated for the maintenance of wg expression, suggesting that in addition to regulating wg expression, mouse Tsh genes, like Drosophila Tsh, might modulate Wg signalling. Given the striking conservation of the Wg and Wnt signalling components between species, one could hypothesize that at least some aspects of tsh activity in Wg/Wnt signalling may be conserved from Drosophila to vertebrates. However, further investigation is required to assess the role of Tsh genes in Wnt signalling in vertebrates (Manfroid, 2004).

Comparison of the organization of Tsh with Tsh-related proteins in mouse and humans suggests that common functional features are probably defined by the region encompassing the three zinc-finger motifs and by the presence of a motif known to interact with CtBP. Interestingly, mouse and Drosophila Tsh proteins display intrinsic transcriptional repression activity. The repression ability of Tsh proteins is partly due to their interaction with the co-repressor CtBP. In the visceral mesoderm, Tsh is recruited to the Ubx enhancer in a repressor complex containing Brk and CtBP, wherein Tsh does not seem to bind directly to DNA, but rather Brk is the DNA-binding partner. In the ectoderm, however, Tsh directly binds to the modulo enhancer and represses transcription in vivo. The association of CtBP with Tsh is dependent on the CtBP-interacting motif (PLDLS) located in the N-terminal part of Tsh, and this CtBP/Tsh complex contributes to the observed repression. An analogous motif (PIDLT) is found in the C-terminal part of the three mouse Tsh proteins. Despite the different context encompassing the PIDLT motif in the mouse proteins (C-terminal), this motif is functional and essential for the repressor function of mouse Tsh1. Although the role of this motif was addressed only for only mouse Tsh1, repression activity of Tsh2 and Tsh3 is equally potentiated by mouse Ctbp1, suggesting that mouse Ctbp1 is a co-repressor acting with all mouse Tsh proteins. Interestingly, the PIDLT motif lies within a region of the three mouse Tsh proteins where the sequence similarity is low and thus appears to be a highly conserved functional domain in a variable region. In addition, it is worth noting that, in mammalian cells, some repression activity persists in mouse Tsh1 after deletion of the CtBP-interacting motif, implying that other mechanisms of transcriptional repression are used by mouse Tsh. In contrast to Tsh, which contains a repressor domain rich in Ala, analysis of the mouse Tsh protein sequences fail to reveal a comparable feature or any known motif that could account for the mouse Tsh1DeltaPIDLT repressor activity (Manfroid, 2004).

Two related Xenopus homologues of the homeotic zinc finger protein Teashirt1 (Tsh1), XTsh1a and XTsh1b, were isolated. While Drosophila teashirt specifies trunk identity, the developmental relevance of vertebrate Tsh homologues is unknown. XTsh1a/b are expressed in prospective trunk CNS throughout early neurula stages and later in the migrating cranial neural crest (CNC) of the third arch. In postmigratory CNC, XTsh1a/b is uniformly activated in the posterior arches. Gain- and loss-of-function experiments reveal that reduction or increase of XTsh1 levels selectively inhibits specification of the hindbrain and mid/hindbrain boundary in Xenopus embryos. In addition, both overexpression and depletion of XTsh1 interfere with the determination of CNC segment identity. In transplantation assays, ectopic XTsh1a inhibits the routing of posterior, but not of mandibular CNC streams. The loss of function phenotype could be rescued with low amounts either of XTsh1a or murine Tsh3. These results demonstrate that proper expression of XTsh1 is essential for segmentally restricted gene expression in the posterior brain and CNC and suggest for the first time that teashirt genes act as positional factors also in vertebrate development (Koebernick, 2006).

A single ortholog of teashirt and tiptop regulates larval pigmentation and adult appendage patterning in Bombyx mori

Two paralogous genes, teashirt (tsh) and tiptop (tio), encode zinc-finger transcription factors and play important roles in insect growth and development. In the fruit fly, Drosophila melanogaster, tsh promotes trunk segmental identities and contributes to the patterning of other tissues during the embryonic stage. During the adult stage, tsh contributes to the specification and patterning of appendages, including the leg, wing and eye. While tio acts redundantly with tsh, flies lacking tio function are viable without deleterious phenotypes. This gene pair is present in the genomes of all Drosophila species but only as a single homologue in several other insect species. In Oncopeltus fasciatus and Tribolium castaneum, tsh/tio has been functionally characterized as specifying the identity of the leg during the adult stage. However, in lepidopteran insects which include large numbers of pests in agriculture and forestry, as well as the major silk producer silkworm Bombyx mori, the biological functions of tsh/tio are still poorly understood. This study performed functional analysis of tsh/tio by using both CRISPR/Cas9-mediated mutagenesis and transposon-mediated ectopic expression in B. mori. The results show that loss of tsh/tio function affected pigmentation during the larval stage and appendage pattering during the adult stage. RNA-seq analysis and subsequent q-RT-PCR analysis revealed that depletion of tsh/tio significantly elevated the expression of the kynurenine 3-monooxygenase gene, as well as melanin synthase-related genes during the larval stage. Furthermore, ubiquitous ectopic expression of tsh/tio induces developmental retardation and eventually larval lethality. These data reveal evolutionarily conserved functions of tsh/tio in controlling adult appendage patterning, as well as the novel function of regulating larval pigmentation in B. mori, providing novel insights into how tsh/tio regulates insect growth and development (Zhang, 2020).

date revised: 7 July 2020

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.