BIOLOGICAL OVERVIEW

daughterless is so named because of its role in sex determination. It is required for the maturation of follicle cells during egg chamber morphogenesis. The dimerization partner of daughterless in the maturation of follicle cells is unknown (Gonzalez-Crespo, 1993 and Cummings, 1994). In other roles daughterless interacts with Dorsal to bring about the induction of twist and snail, genes required for gastrulation. daughterless is a cofactor for their activation.

However, it is the involvement of daughterless in neural differentiation that is considered of primary developmental importance. Although daughterless is not required for the formation and delamination of "nascent" neuronal precursors from the epidermal layer, it is required for expression of neuron specific genes. Mutation of da blocks transformation of presumptive precursors into true precursors. Since AS-C genes are required for cells to become neuronal precursors, this requirement is fulfilled in the absence of daughterless (Vassin, 1994). This result is paradoxical because it is presumed that DA is the dimerization partner of AS-C proteins. How can Achaete and Scute carry out their proneural function without DA?

The list of genes activated by Daughterless as a cofactor with achaete-scute complex genes will continue to grow. Known targets include prospero, cyclin A and calmodulin (Vaessin, 1994 and Kovalick, 1992).

Proneural gene products like Daughterless and Lethal of scute can bind to promoters of Enhancer of split and achaete genes, and by so doing, activate their transcription. Two proteins of the E(spl)-C (HLH-M5 and Enhancer of split) attenuate the transcriptional activation mediated by the proneural genes. This observation begins to untangle the complicated role of E(spl)-C genes in neurogenesis. Once neuroblasts have segregated, products of proneural genes become restricted to the neuroblasts. Products of the E(spl)-C genes are restricted to cells remaining in the epithelium. Therefore it appears that E(spl)-C functionally antagonizes the proneural proteins and thus silences expression of genes that are activated by the proneural genes (Oellers, 1994).

Daughterless couples the control of differentiation and cell cycle programs in in the developing sensory organ precursor (SOP). Although Daughterless is required for the proper expression of neuronal precursor genes and lineage identity genes in the peripheral nervous system (PNS) of Drosophila embryos, this requirement does not explain the failure of the nascent PNS precursors to undergo a normal cell cycle and divide in da mutants. Four genes whose products are required for various stages of the cell cycle are misexpressed in the PNS of da mutant embryos. Cyclin A, barren, disc proliferation abnormal and Histone H1 transcripts are significantly reduced or undetectable in the precursors of the PNS at stages 11 and 12. Precursors are still present at these stages in da mutants. This suggests that all aspects of PNS precursor differentiation examined so far are under the transcriptional control of da. Sensory organ precursors lacking Da may fail to express and/or accumulate other factors, such as critical differentiation genes, required for SOP entry into the cell cycle. It should be pointed out that these factors are unlikely to be the thus-far described neuronal precursor genes, as mutations in these genes do not result in any obvious cell cycle defects. Thus daughterless controls the expression of cell cycle genes in the PNS sensory organ precursors but nowhere else (Hassan, 1997).

The basic helix-loop-helix transcription factor Twist regulates a series of distinct cell fate decisions within the Drosophila mesodermal lineage. These twist functions are reflected in its dynamic pattern of expression, which is characterized by initial uniform expression during mesoderm induction, followed by modulated expression at high and low levels in each mesodermal segment, and finally restricted expression in adult muscle progenitors. Two distinct partner-dependent functions for Twist were found that are crucial for cell fate choice. Twist can form homodimers and heterodimers in vitro with the Drosophila E protein homolog, Daughterless. Using tethered dimers to assess directly the function of these two particular dimers in vivo, it has been shown that Twist homodimers specify mesoderm and the subsequent allocation of mesodermal cells to the somatic muscle fate. Misexpression of Twist-tethered homodimers in the ectoderm or mesoderm leads to ectopic somatic muscle formation overriding other developmental cell fates. In addition, expression of tethered Twist homodimers in embryos null for twist can rescue mesoderm induction as well as somatic muscle development. Loss of function analyses, misexpression and dosage experiments, and biochemical studies indicate that heterodimers of Twist and Daughterless repress genes required for somatic myogenesis. It is proposed that these two opposing roles explain how modulated Twist levels promote the allocation of cells to the somatic muscle fate during the subdivision of the mesoderm. Moreover, this work provides a paradigm for understanding how the same protein controls a sequence of events within a single lineage (Castanon, 2001).

At stage 10, in response to transcriptional regulators such as Sloppy paired and Even skipped, as well as signals from the overlying ectoderm such as Wingless, the uniform expression of Twist modulates into regions of high and low expression within each segment. Da is expressed uniformly in the mesoderm at this time. The region that maintains high Twist levels subsequently gives rise to somatic muscles whereas the region that has lower Twist levels gives rise to tissues such as visceral muscle, fat body, gonadal mesoderm and some glia cells. The heart is derived from the region that initially expresses high levels of Twist; however these cells lose Twist expression, an event necessary for the execution of heart fate. Expressing high Twist levels in cells destined to become visceral muscle, for example, blocks visceral muscle differentiation and promotes somatic muscle. Reduction of Twist levels in cells normally expressing high Twist levels blocks somatic myogenesis (Castanon, 2001 and references therein).

Several possible mechanisms are provided to explain these observations and illustrate the in vivo roles for the two opposing activities of Twist homodimers and Twist/Da heterodimers. Regions that normally express lower Twist levels do not form somatic muscles owing to higher concentrations of Twist/Da heterodimers as compared to Twist homodimers. These heterodimers repress transcription of pro-muscle genes, such as l’sc as well as founder cell genes such as Kr, thereby prohibiting somatic muscle development. Other differentiation programs for visceral muscle or fat body development can proceed unaffected. No evidence is found that Twist/Da heterodimers promote visceral mesoderm or fat body fate through the direct activation of targets such as Fas III. Regions that normally express higher Twist levels do form somatic muscle owing to higher concentrations of Twist homodimers as compared to Twist/Da heterodimers. Dimer competition, then, restricts the developmental potential of mesodermal cells, by not allowing Twist homodimers to convert all mesodermal cells into somatic muscle (Castanon, 2001).

These conclusions are consistent with the observations that increasing Twist/Da levels, either by overexpression of Da or the tethered Twist-Da heterodimer, repress the earliest steps in somatic myogenesis. These are the same steps that are activated by Twist homodimers. For example, L’sc expression, which marks clusters of equipotential cells that segregate the muscle founder cells, is drastically reduced or absent upon an increase of Twist/Da heterodimers. This indicates an early failure in the somatic muscle program. Likewise failure in subsequent steps is seen; for example, few founder cells as well as few identifiable muscles are detected. These failures in muscle development are interpreted as an outcome of the initial block in the differentiation pathway. The possibility that overexpression of Da or of Twist-Da could directly repress these subsequent steps is not eliminated. Gal4 lines that drive expression at later stages of muscle development or in particular subsets of muscle cells (i.e., the S59-expressing founder cells) could provide insight into this alternative (Castanon, 2001).

Daughterless dictates Twist activity in a context-dependent manner during somatic myogenesis

Somatic myogenesis in Drosophila relies on the reiterative activity of the basic helix-loop-helix transcriptional regulator, Twist (Twi). How Twi directs multiple cell fate decisions over the course of mesoderm and muscle development is unclear. Previous work has shown that Twi is regulated by its dimerization partner: Twi homodimers activate genes necessary for somatic myogenesis, whereas Twi/Daughterless (Da) heterodimers lead to the repression of these genes. This study examined the nature of Twi/Da heterodimer repressive activity. Analysis of the Da protein structure revealed a Da repression (REP) domain, which is required for Twi/Da-mediated repression of myogenic genes, such as Dmef2, both in tissue culture and in vivo. This domain is crucial for the allocation of mesodermal cells to distinct fates, such as heart, gut and body wall muscle. By contrast, the REP domain is not required in vivo during later stages of myogenesis, even though Twi activity is required for muscles to achieve their final pattern and morphology. Taken together, evidence is presented that the repressive activity of the Twi/Da dimer is dependent on the Da REP domain and that the activity of the REP domain is sensitive to tissue context and developmental timing (Wong, 2008).

This study explores the regulation of Twi activity through mesoderm development and somatic myogenesis in Drosophila . Focus was placed on how Twi is modulated by its dimer partner, Da. The examination of Twi/Da dimers revealed that the activity of these dimers is acutely sensitive to their tissue environment: both between germ layers (the ectoderm versus the mesoderm), and within cell lineages (early mesoderm versus somatic muscle). This sensitivity is determined, in part, by the activity of the Da REP domain, which is critical for Twi/Da activity during mesodermal subdivision and FC specification, but is not required for the later activity of Twi/Da during muscle differentiation. This work provides insight to the mechanism of Twi/Da activity and calls attention to the effect of tissue context and developmental timing on bHLH protein regulation (Wong, 2008).

One of the most striking aspects of this study is the role of the Da REP domain in switching Twi/Da behaviour between a repressor and an activator function. This 'switchable' behaviour of Twi/Da activity was initially observed by its ability to inhibit myogenesis in the mesoderm, but activate myogenesis in the ectoderm. Notably, the deletion of the REP domain from Da has little effect on Da activity in the absence of Twi, as demonstrated by cell culture transcriptional assays. However, the activity of Twi-Da^Δ tethered dimers has a distinct effect on the mesoderm. Overexpression of these dimers had the greatest effect on somatic myogenesis during the process of mesodermal subdivision. The detection of increased numbers of founder cells (FCs), which appear to be specified normally, indicated an increased number of mesodermal cells being allocated to a somatic muscle fate at the expense of cardiac and visceral mesoderm (Wong, 2008).

An outstanding question is how the Da REP domain functions to modulate Twi/Da activity. Since Twi/Da dimers bind DNA and therefore may actively regulate the transcriptional state of a target gene, it was initially postulated that the REP domain must directly interact with transcriptional corepressors or factors that were expressed solely in the mesoderm and therefore were required for the repressive activity of Twi/Da in that tissue context. Exhaustive studies were conducted to identify these factors but no protein that satisfies all necessary criteria has been identified (Wong, 2008).

Deletion analysis of the E protein Rep domain suggested that this domain is required for the repression of the E protein activation domains, AD1 and AD2. Like the Da REP domain, the E protein Rep domain has specific activities depending on its dimer partner and tissue context (Markus, 2002). Informed by this work, the current data was interpreted to suggest that the Da REP domain is a cis-acting repressor, which functions to repress both Da AD1 and AD2 when Da is dimerized to Twi and bound to myogenic enhancers. Moreover, the effect of the Da REP domain is not restricted to the E protein/Da protein family. This work suggests that the Da REP domain also represses Twi's activation domains, Twi-AD1 and Twi-AD2, in Twi/Da dimers. It is proposed that the Da REP domain acts to mask the activation domains in both Twi and Da. Therefore, the net effect of the Da REP domain results in the recruitment of corepressors to myogenic enhancers by Twi/Da dimers. Alternatively, Twi/Da dimers may not actively repress target myogenic genes: instead, these dimers could compete for myogenic E boxes or transcriptional cofactors and machinery. In this model of passive repression, the Da REP domain could function to stabilize interactions with Twi or other factors that are required to properly mediate repression of myogenic target genes. These aspects of Da REP domain repression are currently being evaluated (Wong, 2008).

To date, various transcriptional regulators have been shown to have different activities and target genes in different tissues and be modulated by dimerization partners. Recently, ChIP-on-chip analyses have identified almost 500 direct Twi targets throughout mesodermal development. This study, however, is one of the first that focuses on how Twi activity is dynamically modulated through multiple developmental stages of a specific cell lineage, and how this regulation affects expression of Twi target genes (Wong, 2008).

One gene that is regulated by Twi dimers throughout somatic myogenesis is Dmef2. Dmef2 protein is expressed throughout and necessary for all stages of myogenesis. Dmef2 coordinates multiple processes necessary for proper somatic myogenesis. Moreover, it has been suggested that Dmef2 is required in combination with Twi to regulate the expression of a subset of Twi target genes in a feed-forward mechanism. The current data support these arguments, since mesodermal phenotypes were observed in Twi/Da^Δ (activation) or Twi/Da (repression) overexpressing embryos that mirror those of embryos overexpressing Dmef2 or in Dmef2 mutant embryos, respectively. For example, increased Dmef2 reporter gene expression and increased numbers of FCs were observed in embryos that overexpress Twi/Da^Δ panmesodermally. Consistent with these observations, Dmef2 has been shown to regulate components of the Ras/MAPK and Notch pathways, which are both required for the proper specification of FCs, and the expression of a subset of FC identity genes. Dmef2 has also been shown to regulate a subset of genes that are required for myoblast fusion and muscle attachment, processes required for proper muscle morphogenesis. This study found that Twi/Da and Twi/Da^Δ dimers disrupt myoblast fusion and muscle differentiation, which is likely due to these dimers repressing Dmef2 expression. In agreement with this observation, muscle analysis revealed that embryos overexpressing Twi-Da and Twi-Da^Δ dimers have muscle phenotypes that are similar to those observed in Dmef2⁴²⁴ hypomorph embryos and Dmef2^22.21 null embryos that have been partially rescued by UAS-Dmef2 transgenes. Taken together, these results supported conclusions of the pivotal regulation of Dmef2 activity by Twi dimers throughout myogenesis (Wong, 2008).

Another notable question is how the Da REP domain is required for Twi/Da mediated transcriptional repression during mesodermal subdivision, but not during muscle morphogenesis. One possibility is that during somatic muscle differentiation, the repressive activity of Twi/Da relies on a different protein domain. Another possibility includes the changes in Twi/Da target genes through the course of somatic myogenesis. Studies conducted on chromatin remodeling have emphasized the specificity involved with the transcriptional regulation of a single gene. Therefore, it is likely that the regulation of multiple sets of genes through time would rely on the modular nature of transcriptional regulators. The Da REP domain may be required for the repression of a subset of Twi/Da target genes, whereas other target genes are unresponsive to this domain's repressive activity (Wong, 2008).

In summary, these results suggest that the regulation of Dmef2 by Twi/Da throughout myogenesis and the subsequent feed-forward mechanism by which Dmef2 and Twi regulate myogenic genes is critical for the coordination of the various disparate processes-mesodermal subdivision, FC specification, and muscle differentiation-necessary for somatic myogenesis (Wong, 2008).

Twi proteins are conserved across species [mouse, chicken, C. elegans, and jellyfish] and have been shown to dimerize with Da homologs, suggesting that REP domain regulation of Twi activity is conserved. Similarly to flies, Mouse Twi1 (MTwi1) heterodimerizes with E proteins to compete with MyoD/E proteins for binding sites on myogenic enhancers. In this manner, MTwi1/E protein heterodimers act like Twi/Da dimers to repress myogenesis. In other tissues, however, MTwi1/E protein heterodimers have been identified as an activator of targets, such as thrombospondin-1 during cranial suture formation. Therefore, like Twi/Da, MTwi1/E protein heterodimers are sensitive to tissue contexts. Of particular interest would be the examination of the E protein Rep domain in vivo. The function of this domain has been studied in mammalian cell culture, but not yet investigated in developmental processes. Moreover, the function of the E protein Rep domain has not been addressed in MTwi1/E protein dimers (Wong, 2008).

Notably, Twi proteins have also been implicated in a variety of tumourigenic processes, such as the inhibition of apoptosis and the coordination of metastasis. Mouse models and correlative data from human tumour samples suggest that MTwi1 and human Twi1 (HTwi1), respectively, direct epithelial-to-mesenchymal transitions (EMT) during breast cancer metastasis. The involvement of Twi1 in the complex process of cancer has many similarities to the developmental processes that Twi directs in the fly mesoderm, which include cell proliferation and cell migration, processes that have been recently revealed to be directly regulated by Twi. The role of the Da REP domain in directing Twi/Da transcriptional repression, and the tissue specificity of this domain's activity has illuminated various aspects of Twi regulation. It is anticipated that these findings will shed light on mammalian Twi1 activity and the Twi family of proteins in development and disease (Wong, 2008).

Introducing Pitt-Hopkins syndrome-associated mutations of TCF4 to Drosophila daughterless

Pitt-Hopkins syndrome (PTHS) is caused by haploinsufficiency of Transcription factor 4 (TCF4), one of the three human class I basic helix-loop-helix transcription factors called E-proteins. Drosophila has a single E-protein, Daughterless (Da), homologous to all three mammalian counterparts. This study shows that human TCF4 can rescue Da deficiency during fruit fly nervous system development. Overexpression of Da or TCF4 specifically in adult flies significantly decreases their survival rates, indicating that these factors are crucial even after development has been completed. da transgenic fruit fly strains with corresponding missense mutations R578H, R580W, R582P and A614V found in TCF4 of PTHS patients were generated and the impact of these mutations was studied in vivo. Overexpression of wild type Da as well as human TCF4 in progenitor tissues induces ectopic sensory bristles and the rough eye phenotype. By contrast, overexpression of DaR580W and DaR582P that disrupt DNA binding reduces the number of bristles and induces the rough eye phenotype with partial lack of pigmentation, indicating that these act dominant negatively. Compared to the wild type, DaR578H and DaA614V are less potent in induction of ectopic bristles and the rough eye phenotype, respectively, suggesting that these are hypomorphic. All studied PTHS-associated mutations introduced into Da lead to similar effects in vivo as the same mutations in TCF4 in vitro. Consequently, these Drosophila models of PTHS are applicable for further studies aiming to unravel the molecular mechanisms of this disorder (Tamberg, 2015).

This study shows that Da, the only E-protein in Drosophila with highly conserved bHLH domain, functions as human TCF4 orthologue. As the overall identity of a protein sequence between Drosophila and mammals is usually around 40% between homologues and 80-90% within conserved functional domains, Da can be considered the orthologue for all three human E-proteins. In all experiments conducted in this study, TCF4 was found to act very similarly as Da, proving the possibility of modelling PTHS in the fruit fly. The two human TCF4 isoforms, TCF4-A and TCF4-B, were able to activate E-box dependent lacZ expression in Drosophila, and more importantly, to induce ectopic bristle formation in the adult thorax, to rescue embryonic nervous system development in da null embryos, and to induce the rough eye phenotype when overexpressed in the nervous system identically to Da. Altogether these results show that TCF4 has comparable activity in the fruit fly as Da (Tamberg, 2015).

To further study the PTHS-associated mutations in Drosophila, four mutations found in TCF4 (R580W, R578H, R582P and A614V) and two control mutations (D515G and R580L) were introduced into Da. The mutants were analysed by luciferase assay in a mammalian cell line in order to compare the results of Da directly to results obtained with human TCF4. Subsequently PTHS-associated Da mutants were studied in vivo in E-box lacZ reporter assay, and in both rescue and overexpression experiments (Tamberg, 2015).

PTHS-associated arginine mutations R580W, R580L, and R582P abolish Da transactivation capability in luciferase reporter assays in HEK293 cells. DaR580W, DaR580L and DaR582P behave similarly to each other in both overexpression and rescue experiments in the fruit fly. The rescue by daG32-GAL4 driver of da null embryonic nervous system phenotype fails when using Da proteins with these arginine mutations. When Da carrying one of above mentioned mutations was overexpressed in flies under the control of the nervous system specific driver GMR12B08-GAL4, the strongest eye phenotype was observed. These flies have rough and partially unpigmented eyes with fused ommatidia consistent with Da having an important role in Drosophila eye development. In addition, overexpression of these arginine mutants under pnr-GAL4 causes malformation of the thorax. Altogether, these results indicate that mutations R580W, R580L and R582P abolish the Da transactivation capability resulting in dominant-negative effects. This is in line with the previous data about the corresponding mutations in TCF4 having dominant-negative effects in vitro (Tamberg, 2015).

R578H was found to differ from the other three arginine mutations (R580W, R580L and R582P) in in vivo experiments. Although DaR578H was unable to activate reporter gene expression in luciferase assay carried out in mammalian cell line HEK293 and in lacZ assay in vivo, it causes rough eye phenotype similar to Dawt when overexpressed by GMR12B08-GAL4. Furthermore, DaR578H rescues da null embryonic neuronal phenotype when expressed using daG32-GAL4. Also DaR578H shows weak induction of ectopic bristles. Taken together these results indicate that transactivation capability of DaR578H probably depends on its dimerisation partners, which could be lacking in mammalian cell line and weakly presented in the wing disc notum. Similarly, it has been previously shown that while TCF4 carrying the R578H mutation is unable to bind to E-box in vitro as a homodimer or in complex with either ASCL1 or NEUROD2, it does not act in dominant negative manner in reporter assays in mammalian cells (Tamberg, 2015).

The A614V mutation positioned in the second helix of the bHLH domain shows the mildest effects. DaA614V was able to activate E-box-specific transcription in vitro and in vivo. Expressing DaA614V using daG32-GAL4 rescues da null embryonic neuronal phenotype. Overexpression using GMR12B08-GAL4 results in the rough eye phenotype only when both of the transgenes are homozygous, indicating that this mutation causes hypomorphic effects. This is consistent with a recent study which shows that the A614V mutation leads to lower levels of TCF4 because of reduced protein stability (Tamberg, 2015).

The control mutation generated in this study, D515G, does not reduce Da transactivation capability in vitro and behaves similarly to Dawt in vivo. This shows that D515 positioned outside of the conserved bHLH is not required for Da transcriptional activity. The other control mutation generated in this study, R580L, where the same arginine is mutated as in DaR580W, leads to dominant-negative effects in vivo similarly to R580W. At least in the case of R580, the mutation specificity, whether it was mutated into tryptophan or leucine, was found to have no affect (Tamberg, 2015).

In rescue experiments with tested driver strains (69B-GAL4, tub-GAL4, ubi-GAL4, GMR12B08-GAL4, daG32-GAL4) all Da transgenes fail to rescue da null embryonic lethality. Apparently the successful rescue of da null lethality closely mimics the endogenous Da expression. daG32-GAL4, comprising of 3.2 kb of da gene covering the promoter, the first intron, and the upstream noncoding region, is widely used as a ubiquitous driver line. Most probably the expression of this driver line is far too strong compared to the native expression of da gene as Da has been shown to positively autoregulate its own expression via a transcriptional feedback loop. If daG32-GAL4 expression is regulated by Da itself, then Da overexpression might drive even stronger GAL4 expression, resulting in a positive feedback loop. Furthermore, it has been hypothesised that daG32-GAL4 lacks putative regulatory repressor elements since using a 15 kb genomic da transgene that has an additional 12 kb of downstream sequence rescues da null embryonic lethality (Tamberg, 2015).

Little is known about the role of E proteins in adult nervous system. This study shows that exact temporal and spatial expression of Da/TCF4 remains vitally important during adulthood of fruit flies. It was shown that overexpression of Da/TCF4 in adults leads to lethality within 2-3 days. Surprisingly, TCF4 isoforms A and B lead to strikingly different outcomes when overexpressed in adult fruit flies. While the long isoform TCF4-B behaves identically to Da, TCF4-A affects the survival only slightly compared to the control group. This could be related to the lack of interaction capability of much shorter N terminus of isoform A in fruit fly or different regulation of subcellular location and dimerisation of the alternative TCF4 isoforms. Analysis of survival divides the PTHS related mutations into severe (R580W, R580L and R582P) and milder (R578H and A614V) according to survivorship. The severe mutants lead to lethality within 3-4 days and the milder ones in 10-11 days. Further experiments with cell type or tissue-specific drivers would help to understand the role of E-proteins during adulthood in more details (Tamberg, 2015).

The fact that overexpression of wild type as well as dominant negative forms of Da causes comparable reduction in survival and induction of the rough eye phenotype raises the possibility that overexpression of wt protein is also eliciting dominant negative effects as suggested earlier. One explanation for this phenomenon could be that excess homodimers outcompete transcriptionally more potent heterodimers at various promoter sites. Intriguingly, recent studies suggest that in addition to TCF4 haploinsufficiency, increased TCF4 dose is also a risk factor for disturbed cognitive development as a TCF4 duplication has been described in a patient with developmental delay and a partial duplication in a patient with major depressive disorder. Nevertheless, in case of induction of ectopic bristles, opposite effects were observed for Dawt and dominant negative Da mutants, indicating that in addition to its dominant negative effects, excess wt protein also has specific effects during development (Tamberg, 2015).

In patients with PTHS just one copy of TCF4 is mutated or deleted. Seemingly the most relevant way to model PTHS in animal models would be to use the appropriate heterozygotes of the orthologous protein. However, in Drosophila there is a sole E-protein Da corresponding to all three mammalian E-proteins. In a way the heterozygous Da null mutation corresponds to the heterozygous deletion of all three E-proteins in mammals. Accordingly, Da as the only binding partner of class II bHLH proteins has a large variety of roles outside nervous system. As TCF4 is highly expressed in the nervous system, the mutated alleles were expressed specifically in the nervous system in a wild type background. Overexpression of DaPTHS under the nervous system specific GMR12B08-GAL4 leads to viable flies and stocks with each mutation generated in this study were created. An alternative tactic to model PTHS and to mimic dosage loss by TCF4 deletions would be to slightly downregulate Da expression nervous system specifically by RNAi. Additional studies are needed to generate and compare different PTHS models and to perform behavioural tests that would give valuable information about cognition and social behaviour of the PTHS model flies (Tamberg, 2015).

Spatial regulation of expanded transcription in the Drosophila wing imaginal disc

Growth and patterning are coordinated during development to define organ size and shape. The growth, proliferation and differentiation of Drosophila wings are regulated by several conserved signaling pathways. This study shows that the Salvador-Warts-Hippo (SWH) and Notch pathways converge on an enhancer in the expanded (ex) gene, which also responds to levels of the bHLH transcription factor Daughterless (Da). Separate cis-regulatory elements respond to Salvador-Warts-Hippo (SWH) and Notch pathways, to bHLH proteins, and to unidentified factors that repress ex transcription in the wing pouch and in the proneural region at the anterior wing margin. Senseless, a zinc-finger transcription factor acting in proneural regions, had a negative impact on ex transcription in the proneural region, but the transcriptional repressor Hairy had no effect. This study suggests that a complex pattern of ex transcription results from integration of a uniform SWH signal with multiple other inputs, rather than from a pattern of SWH signaling (Wang, 2018).

The development, differentiation and growth of cells and tissues require precisely regulated patterns of gene expression, depending on the time and spatial location of its activation, and its crosstalk with other signaling pathways. This study used the Drosophila wing disc as a model system to investigate how ex transcription is regulated. Ex is important as a negative growth regulator that acts through the SWH pathway. It restricts wing size in normal development, so that mutants have larger, 'expanded' wings, and can also be induced to block growth of cells with developmental perturbations, including those with emc mutations that over-express Da. In normal development, ex-LacZ is expressed most highly in the hinge region that surrounds the wing pouch, and expression decreases in a gradient until none is detected at the wing margin. Unlike expression patterns of ex-LacZ, a relatively ubiquitous distribution of Ex protein is seen in the wing imaginal discs. The discrepancy between ex reporter and Ex protein might be due to the post-translational control of Ex protein stability. Alternatively, Ex protein might be strongly influenced by expression patterns in earlier developmental stages. Regardless, because ex is itself a transcriptional target of the SWH pathway, acting through Sd and Yki, ex-LacZ is extensively used as a transcriptional readout of SWH signaling activity, if not of Ex protein distribution. This transcription pattern of ex could be interpreted to indicate a proximal-to-distal gradient of Sd/Yki activity, which would be consistent with certain models of wing growth regulation that propose that SWH activity represses growth in central regions of the disc whereas Yki activity is higher in proximal regions. Another mechanism predicted to repress growth in central regions of the wing disc is activation of Notch signaling there. Notch signaling induces expression of Vg, a protein that binds Sd in competition with Yki and is therefore predicted to reduce Yki activity (and ex transcription) in central regions of the wing pouch. Since this study has identified an ex enhancer whose activity reflects the ex-LacZ reporter in the endogenous gene, this enhancer can be explored to understand how these and other signals are integrated at the ex locus (Wang, 2018).

A deletion analysis (see Analysis of Intron3 enhancer)outlines several major features of ex regulation. First, the core of the enhancer, centered around the 'C' element and probably including contributions from the flanking 'B' and 'D' elements, is active throughout the wing disc. All this activity depends on both Sd and Yki, suggesting that Yki is active throughout the wing disc. This is consistent with the previous finding that yki is required for growth throughout the wing disc. ChIP-seq analysis revealed that Yki binding peaks over Ex^Intron3. Although Sd binding has not been mapped yet in Drosophila, it is likely that Sd binding strongly correlates with Yki, as seen for the mammalian YAP1 and TEAD1 proteins. Expression of the Sd/Yki-dependent BC and CD elements provided no evidence of a proximal-distal gradient of Sd/Yki activity. Instead, the overall reduction of ex expression in the central, wing pouch region of the wing disc requires both the B and D regions together, suggesting that two other inputs are both required to achieve this silencing. It does not seem that either of these silencing inputs corresponds to Notch signaling, because although it was confirmed that Notch is required to silence ex enhancer activity in the wing pouch, it is not required to silence the BCD element (Wang, 2018).

In the wing margin proneural region (high SWH activity; low Yki activity), ex is negatively regulated by inputs acting through elements A and F. The E-box site #2 (E2) is required for expression in the wing margin proneural cells while Da acts through E1 and E3 to regulate ex transcription. Sd/Yki regulates ex transcription through element BCD in wing pouch and hinge. Notch acting through element E and other inputs acting through elements B and C repress ex transcription in wing pouch (Wang, 2018).

An unexpected aspect of ex transcription regulation is independent regulation of ex enhancer activity in the proneural cells of the anterior wing margin. Activity here was encoded by the BC and CD elements, and depended on an E-box within C (E-box #2). Surprisingly, given the normal role of Notch signaling in lateral inhibition to prevent neural differentiation, the wing margin activity depended positively on N signaling. Anterior wing margin activity was dependent on Da, the obligate heterodimer partner of all proneural genes, indicating that it is possibly encoded by proneural genes of the AS-C. This regulation was unexpected because there was no evidence for ex-LacZ or Ex^intron3-GFP reporter activity at the anterior wing margin. This is because such expression was silenced by either one of two flanking elements, A and F, acting redundantly. The hypothesis was tested that Sens, which is expressed in the anterior wing margin proneural region and can act as a transcriptional repressor, was responsible for blocking the activity. Although Sens was sufficient to inhibit Ex^intron3-GFP activity at some ectopic locations, it was not required for Ex^intron3-GFP repression at any location, including at the anterior wing margin (Wang, 2018).

The ex enhancer was originally identified through its role in mediating ex transcription in response to elevated Da activity. The Da response was mapped to E-boxes #1 and #3 within elements B and D, respectively. Since Da can bind to DNA as a homodimer, and as no other Class I or Class II bHLH gene is known to be widely expressed in the wing pouch, it is possible that Da homodimers activate this site. The E1 site within element B matches the consensus CACCTG sequence that is preferred by Da-Da homodimers and by homologous E-protein homodimers in mammals. Endogenous Da protein levels are normally elevated at the anterior wing margin, because Emc levels are reduced there, so it is interesting that Ex^intron3 is activated at the anterior wing margin through distinct E-boxes that differ in sequence from E2 site. One possibility is that it is heterodimers of Da with proneural proteins encoded by AS-C that activates these E-boxes at the anterior wing margin, since it was previously shown that ectopically-expressed AS-C proteins can activate the intact enhancer. The consensus E-box site for Sc/Da has been described as GCAGC/GTG and the 5' flanking base G is essential for the binding specificity of Sc. The E2 site matches both the core sequence and the 5'4 flanking base. In part this could explain why emc loss is detrimental for proliferating progenitor cells in the rest of the wing disc and leads to their ex-dependent loss during growth, in contrast to the proneural cells that normally tolerate elevated Da levels, if Da proteins contribute to distinct dimers in these two situations (Wang, 2018).

This analysis shows that multiple regulatory inputs are integrated by the Ex^intron3 enhancer. Although this study confirms previous conclusions that Sd/Yki and Notch signaling regulate ex transcription, these studies indicate that the full enhancer, and hence the widely-used ex-LacZ line, are not straightforward reporters of Yki activity. Instead the smallest sequence that responds to Sd and Yki is active almost uniformly throughout the wing disc, suggesting that this may be the pattern of Yki activity. This may be a useful reporter for SWH activity in future studies. The final ex-LacZ pattern is strongly influenced by additional sequences that silence transcriptional activity in the wing pouch and at the anterior wing margin, in the latter case apparently preventing ex transcription in response to proneural gene activity while preserving sensitivity to ectopic Da expression (Wang, 2018).

GENE STRUCTURE

Bases in 5' UTR - 229

Exons - two

Bases in 3' UTR - 992

PROTEIN STRUCTURE

Amino Acids - 710

Structural Domains

One-third of the way into the protein there is a tandem repeat of a histidine-rich region. This is followed by a PEST sequence, a "myc similarity region," (a basic HLH domain) and a C terminal lysine repeat region (Caudy, 1988 and Cronmiller, 1988).

date revised: 15 November 2001  

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.