twist

cis-Regulatory Sequences and Functions

There is no obvious TATA box, although there is an AT rich region between positions -40 to -28 (Thisse, 1988). A comparison of the transcriptional regulatory regions reveals a high degree of conservation in the more distal of the two ventral activator regions that have been mapped in the twist 5' flanking region. However, the more proximal VAR is absent at the corresponding position in the D. virilis twist gene. Instead, there is a region in the second intron of the D. virilis gene that resembles the proximal element of the D. melanogaster gene, in that it consists of little more than a series of whole and half binding sites for the Dorsal morphogen. In transformation experiments, the intronic D. virilis element directs an expression pattern that is indistinguishable from that directed by the D. melanogaster proximal VAR. Thus, the twi genes from these two species appear to have evolved enhancer elements with very similar structural and functional properties (Pan, 1994).

Transcriptional Regulation

Four zygotic patterning genes, decapentaplegic (dpp), zerknüllt (zen), twist , and snail are initially expressed either dorsally or ventrally in the segmented region of the embryo, and at the poles. In the segmented region of the embryo, correct expression of these genes depends on cues from the maternal morphogen Dorsal (DL). The DL gradient appears to be interpreted on three levels: dorsal cells express dpp and zen, but not twi and sna; lateral cells lack expression of all four genes; ventral cells express twi and sna, but not dpp and zen. DL appears to activate the expression of twi and sna and repress the expression of dpp and zen. Polar expression of dpp and zen requires the terminal system to override the repression by DL, while that of twi and sna requires the terminal system to augment activation by DL (Ray, 1991).

Dorsal activates twist, and it also functions as a direct transcriptional repressor of a second target gene, zerknüllt. By exchanging DL binding sites between the promoters it can be shown that activator sites from the twi promoter can mediate repression when placed in the context of the zen promoter, and that repressor sites from zen can mediate activation in the context of the twi promoter. This represents the first demonstration that common binding sites for any DNA binding protein can mediate both activation and repression in a developing embryo (Jiang, 1992).

Although CREB-binding protein (CBP: nejire) functions as a co-activator of many transcription factors, relatively little is known about the physiological role of CBP. Mutations in the human CBP gene are associated with Rubinstein-Taybi syndrome, a haplo-insufficiency disorder characterized by abnormal pattern formation. Drosophila CBP is maternally expressed, suggesting that it plays a role in early embryogenesis. Mesoderm formation is one of the most important events during early embryogenesis. To initiate the differentiation of the mesoderm in Drosophila, multiple zygotic genes such as twist (twi) and snail (sna), which encode a basic-helix-loop-helix and a zinc finger transcription factor, respectively, are required. The transcription of these genes is induced by maternal Dorsal protein, a transcription factor that is homologous to the NF-kappa B family of proteins. Drosophila CBP mutants fail to express twi and generate twisted embryos. This is explained by results showing that dCBP is necessary for Dorsal-mediated activation of the twi promoter (Akimaru, 1997).

Subsets of differentiating muscles in the Drosophila embryo express transcription factors, such as NK1/S59 and vestigial. These genes control the development of specific muscle properties. Myogenesis in embryos mutant for wingless is grossly deranged. Mesodermal expression of S59 is lost, whereas some vestigial-expressing muscles develop. wingless dependence and independence of specific muscle subsets correlates with an early derangement of twist expression in wingless mutants (Bate, 1993).

Huckebein (hkb) sets the anterior and the posterior borders of the ventral furrow, but acts through different modes of regulation. In the posterior part of the blastoderm, HKB represses the expression of sna in the endodermal primordium. In the anterior part, HKB antagonizes the activation of target genes by twi and sna. Here, Bicoid permits the co-expression of hkb, sna and twi, which are all required for the development of the anterior digestive tract. Mesodermal fate is determined where sna and twi but not hkb are expressed. Anteriorly hkb together with sna determines endodermal fate, and hkb together with sna and twi are required for foregut development (Reuter, 1994).

Sloppy paired (Slp) and Even-skipped are involved in cell fate determination and segmentation in the Drosophila mesoderm. The primordia for heart, fat body, and visceral and somatic muscles arise in specific areas of each segment in the Drosophila mesoderm. The primordium of the somatic muscles, which expresses high levels of twist, a crucial factor of somatic muscle determination, is lost in sloppy-paired mutants. The effect of slp on Twist levels is probably partly, but not completely mediated by wg. wg mutant embryos show a premature and ectopic decay of Twist, but not to the same degree as seen in slp embryos. Whereas patches of cells expressing high levels of Twist are initially established in wg mutant embryos, no Twist is seen in the trunk region of slp mutant embryos after stage 11. At the same time that twist expression is lost in slp mutants, the primordium of the visceral muscles is expanded (Riechmann, 1997).

bagpipe and serpent expressing mesodermal domains corresponding to the ectodermal even-skipped domains, alternate with the sloppy-paired expressing high-twist mesodermal domains. Ectodermal even-skipped is thought to act through engrailed and subsequently hedgehog to promote bagpipe expression in cardiac and dorsal muscle and serpent in the fat body (Azpiazu, 1996). Ectodermal Dpp is required for the maintenance of mesodermal tinman, which in turn activates bap expression in the eve domain. The visceral muscle and fat body primordia require even-skipped for their development and the mesoderm is thought to be unsegmented in even-skipped mutants. However, it has been found that even-skipped mutants retain the segmental modulation of the expression of twist. Both the domain of even-skipped function and the level of twist expression are regulated by sloppy-paired, and eve serves reciprocally to regulate the slp domain. sloppy-paired thus controls segmental allocation of mesodermal cells to different fates (Riechmann, 1997).

The Drosophila protein DSP1, an HMG-1/2-like protein, binds DNA in a highly cooperative manner with three members of the Rel family of transcriptional regulators (NF-kappaB, the p50 subunit of NF-kappaB, and the Rel domain of Dorsal). This cooperativity is apparent with DNA molecules bearing consensus Rel-protein-binding sites and is unaffected by the presence of a negative regulatory element, a sequence previously proposed to be important for mediating repression by these Rel proteins. The cooperativity observed in these DNA-binding assays is paralleled by interactions between protein pairs in the absence of DNA. In HeLa cells, as assayed by transient transfection, expression of DSP1 increases activation by Dorsal from the twist promoter and inhibits that activation from the zen promoter, consistent with the previously proposed idea that DSP1 can affect the action of Dorsal in a promoter-specific fashion (Brickman, 1999).

DSP1 has opposite effects on the activity of Dorsal assayed with regulatory sequences excised from the twist and zen promoters. These experiments were performed by transiently transfecting mammalian cells in culture. Thus, reporters containing either a 180-bp fragment from zen (a fragment sufficient to mediate repression in Drosophila) or the entire regulatory region of twist (from -1,438 to +38) were activated by cotransfection with DNA encoding Dorsal. Cotransfection with DNA encoding DSP1 has just the opposite effects on this Dorsal mediated activation of the two promoters: activation from the twist promoter is stimulated 4-fold, whereas that from the zen promoter is inhibited 3-fold. DSP1's stimulation of Dorsal-mediated activation from the twist promoter can be mapped to the defined enhancer elements or VARs. Thus, DSP1 also stimulates Dorsal-mediated activation if the template bears, instead of the intact twist promoter, a cassette that contains the two VARs that drive ventral-specific expression of the twist gene in the Drosophila embryo. The two VARs together constitute approximately 300 bp and contain multiple Rel-protein-binding sites (Brickman, 1999).

It is not known what DNA sequences in the zen and twist promoters determine the opposite effects of DSP1 on dorsal-mediated activation. The finding that a negative regulatory element (NRE) has no effect on cooperative binding to DNA of DSP1 and various Rel proteins prompted a reexamination of the earlier claims that DSP1 converts Dorsal, the p50 homodimer, and the NF-kappaB heterodimer into repressors and that this effect requires the NRE. In each case, DSP1 inhibits Rel-protein-dependent activation both in the presence and absence of an NRE. In no case was NRE-dependent conversion of the Rel protein to a repressor by cotransfection with DSP1 observed. It is not understood why the current results differ from those reported previously (Brickman, 1999 and references therein).

Sites of the described protein-protein interactions are found in the conserved Rel domains and in the fragment of DSP1 that bears both HMG domains. The Rel domains of p65 and of Dif differ from those of Dorsal and of p50 in that they lack the HMG-domain-interaction site. The HMG domain of DSP1 also interacts with the TATA-binding protein. Similar interactions have been reported for HMG-1 and HMG-2 with the steroid hormone receptors, for HMG-1 with p53, for HMG-1 with HOXD9, and for HMG-2 with Oct2. Thus, the HMG domain may contain a common structural motif for cooperative DNA binding and interaction with other transcription factors. The interaction between TATA-binding protein and DSP1 also seems to be influenced by the glutamine-rich amino-terminal domain in that the full-length DSP1 interacts more avidly with TATA-binding protein than does the HMG-1 domain. These experiments suggest that the amino-terminal glutamine-rich domain may also potentiate the DSP1-Rel protein interaction as well, because all DSP1-Rel interactions seem stronger with full-length DSP1, particularly the weak interactions seen between DSP1 and p65 or Dif, which are observed only with GST-DSP1 and not with GST-DSP1 (178-393) (Brickman, 1999).

Drosophila thoracic muscles are comprised of both direct flight muscles (DFMs) and indirect flight muscles (IFMs). The IFMs can be further subdivided into dorsolongitudinal muscles (DLMs) and dorsoventral muscles (DVMs). The correct patterning of each category of muscles requires the coordination of specific executive regulatory programs. DFM development requires key regulatory genes such as cut (ct) and apterous (ap), whereas IFM development requires vestigial (vg). Using a new vg^null mutant, a total absence of vg is shown to lead to DLM degeneration through an apoptotic process and to a total absence of DVMs in the adult. vg and scalloped (sd), the only known Vg transcriptional coactivator, are coexpressed during IFM development. Moreover, an ectopic expression of ct and ap, two markers of DFM development, is observed in developing IFMs of vg^null pupae. In addition, in vg^null adult flies, degenerating DLMs express twist (twi) ectopically. Evidence is provided that ap ectopic expression can induce per se ectopic twi expression and muscle degeneration. All these data seem to indicate that, in the absence of vg, the IFM developmental program switches into the DFM developmental program. Moreover, the muscle phenotype of vg^null flies can be rescued by using the activity of ap promoter to drive Vg expression. Thus, vg appears to be a key regulatory gene of IFM development (Bernard, 2003).

Thus the absence of Vg leads to IFM degeneration. Some IFM phenotypes have been reported for the vg^83b27R allele, a strong allele of vg. In these flies, the DVMs are absent and some DLMs are missing. It has been shown that this phenotype is fully penetrant in vg^null flies and that apoptosis is involved in loss of IFMs. Since muscle attachment sites are normal in vg^null flies, the process of degeneration is different from that described in ap mutants. Phenotypic analysis shows that degeneration occurs during late metamorphosis (after 48 h APF) (Bernard, 2003).

Thus DLMs degenerate by apoptosis in homozygous vg^null flies. This degeneration could be due to a misprogramming of myoblasts surrounding DLMs during development. The process that leads to apoptosis in these muscles remains to be determined. DLM degeneration is associated with an ectopic expression of Twi transcription factor. During flight muscle development, Twi expression is restricted to myoblasts and that persistent expression in developing muscles leads to muscle degeneration. Thus, Twi expression in vg^null mutants could be responsible for DLM degeneration. Finally, it has been shown that ectopic ap expression induces Twi expression in DLMs. Since AP and twi are known to be, respectively, activator and target of the N pathway, it can be hypothesized that AP activates Twi ectopically in vg^null DLMs through the N pathway. If this hypothesis is confirmed, it can be asked why Ap does not activate Twi during normal DFM development. It is likely that numerous genes, other than vg and ap, are differentially activated during DFM and IFM development. Twi activation by AP could be repressed by one of these genes during DFM development (Bernard, 2003).

Morphogenetic movements are closely regulated by the expression of developmental genes. This study examines whether developmental gene expression can in turn be mechanically regulated by morphogenetic movements. The effects of mechanical stress were examined on the expression of Twist, which is normally expressed only in the most ventral cells of the cellular blastoderm embryo under the control of the Dorsal morphogen gradient. At embryogenesis gastrulation (stage 7), Twist is also expressed in the anterior foregut and stomodeal primordia. Submitting the early Drosophila embryo to a transient 10% uniaxial lateral deformation induces the ectopic expression of Twist around the entire dorsal-ventral axis and results in the ventralization of the embryo. This induction is independent of the Dorsal gradient and is triggered by mechanically induced Armadillo nuclear translocation. Twist is not expressed in the anterior foregut and stomodeal primordia at stage 7 in mutants that block the morphogenetic movement of germ-band extension. The mutants can be rescued with gentle compression of these cells, the stomodeal-cell compression normally caused by the germ-band extension is interpretated as inducing the expression of Twist. Correspondingly, laser ablation of dorsal cells in wild-type embryos relaxes stomodeal cell compression and reduces Twist expression in the stomodeal primordium. The induction of Twist in these cells depends on the nuclear translocation of Armadillo. It is proposed that anterior-gut formation is mechanically induced by the movement of germ-band extension through the induction of Twist expression in stomodeal cells (Farge, 2003).

Therefore, lateral compression of the early embryo induces the ectopic expression of Twist around the entire dorsal-ventral axis and results in the ventralization of the embryo. Despite the probable variations in the direction and amplitude of the deformation of each cell as a function of its location in the embryo, all cells respond to this stress. This suggests that their transcriptional response is triggered by deformation per se and does not depend on the exact geometry and amplitude of the mechanical strain applied to each cell. However, it is unclear how the forces required to artificially deform the embryo lead to embryonic epithelium stresses and strains that are related to endogenous forces and deformations present in the embryo during development (Farge, 2003).

Importantly, the mechanical induction of Twist is independent of the maternal determinants of dorsal-ventral polarity. Instead, this induction depends on the nuclear translocation of Armadillo and its ability to activate transcription. The mechanism that triggers the nuclear translocation of Armadillo in response to mechanical stress is unknown. One possibility is that mechanical strain activates a noncanonical Wingless transduction pathway, which releases the cytoplasmic pool of Armadillo from Axin and allows it to enter the nucleus. Alternatively, mechanical strain might trigger the release and nuclear localization of the pool of Armadillo that is associated with Cadherin at the zonula-adherens. Indeed, this might provide a reason for dual function of Armadillo as an essential component of Cadherin adhesion complex and as a transcription factor (Farge, 2003).

It is interesting to note that the Armadillo homolog, beta-catenin, translocates into the nuclei at the dorsal pole of early frog and fish embryos, where it plays a role in determining dorsal-ventral polarity. Furthermore, the ectopic nuclear localization of beta-catenin induces the dorsalization of vertebrate embryos. Because the dorsal-ventral axis of invertebrates is inverted with respect to that of vertebrates, this corresponds well with the ventralization observed in Drosophila embryos upon the mechanical induction of Armadillo nuclear localization. Thus, mechanical compression may reactivate a conserved and ancient pathway for dorsal-ventral axis formation (Farge, 2003).

The results presented in this study suggest that the expression of Twist in foregut and stomodeal-primordia cells at the onset of gastrulation is mechanically induced by the compression caused by germ-band extension and that this is also mediated by the nuclear translocation of Armadillo. twist is involved in the differentiation and the formation of both the foregut and the anterior midgut. Interestingly, neither the anterior midgut nor the stomodeum invaginate in embryos that lack the mechanical compression and do not express Twist when epithelial dorsal cells have been photo-ablated. It is proposed that, through mechanical induction of twist, the anterior-gut formation is induced by stomodeal cell compression in response to germ-band extension (Farge, 2003).

In addition to its role in dorso-ventral axis formation, Armadillo is thought to induce the differentiation and invagination of the meso-endoderm cells that give rise to the gut in other vertebrate and nonvertebrate embryos. Although the maternal signals that induce the nuclear translocation of beta-catenin in zebrafish and Sea Urchins are not known, they have been shown to be independent of the classical determinant Wingless. These results in Drosophila raise the possibility that the nuclear translocation of Armadillo/beta-catenin in the gut primordia of these embryos might be mechanically induced by morphogenetic movements that are homologous to germ-band extension. Indeed, the nuclear translocation of beta-catenin and the formation of the meso-endodermal gut invagination/involution are concomitant with convergent extension, which tends to compress the meso-endoderm cells (Farge, 2003).

These parallels led to the speculation that mechanical induction may be an ancient mechanism for inducing gut formation. This could have evolved from a primitive reflex response to mechanical deformation. Such a response might have been the phagocytosis of particles in response to physical contact, which has been proposed to be the 'feeding-response' of the earliest organisms. The generation of a permanent gut might have then been stabilized by the Armadillo-induced expression of meso-endodermal genes in response to genetically controlled endogenous morphogenetic movements, such as cell intercalation generating convergent extension. These experiments may have reactivated the genetic pathway of such 'fossil sensorial behavior' in early Drosophila embryos (Farge, 2003).

Tissue deformation modulates twist expression to determine anterior midgut differentiation in Drosophila embryos

Mechanical deformations associated with embryonic morphogenetic movements have been suggested to actively participate in the signaling cascades regulating developmental gene expression. This paper develops an appropriate experimental approach to ascertain the existence and the physiological relevance of this phenomenon. By combining the use of magnetic tweezers with in vivo laser ablation, physiologically relevant deformations were locally control in wild-type Drosophila embryonic tissues. The deformations caused by germ band extension upregulate Twist expression in the stomodeal primordium. Stomodeal compression triggers Src42A-dependent nuclear translocation of Armadillo/beta-catenin, which is required for Twist mechanical induction in the stomodeum. Finally, stomodeal-specific RNAi-mediated silencing of Twist during compression impairs the differentiation of midgut cells, resulting in larval lethality. These experiments show that mechanically induced Twist upregulation in stomodeal cells is necessary for subsequent midgut differentiation (Desprat, 2008).

Demonstrating the role of mechanical deformations in the regulation of developmental gene expression requires an ability to reproduce endogenous deformations by locally controlling tissue deformations within the living embryo. Although tools for measuring and applying global forces had been previously reported for studying Xenopus embryo tissue explants, approaches for locally manipulating tissues within developing embryos were still lacking. In this study, magnetized cells were remotely manipulated to produce a 60 ± 20 nN force necessary to generate deformations similar to those produced endogenously. The magnitude of this force is smaller by a factor of ~20 than the 1 μN force associated with the convergent extension movements in Xenopus explants measured using the deflection of an optical fiber. This is consistent with the fact that the Xenopus embryo is 10 times larger that the Drosophila embryo. This value is also consistent with the 13 nN force developed by a 20 MDCK cell assembly on a soft micropillar surface, noting that the cell colony is five times smaller than the Drosophila embryo length. Importantly, both magnetic and external uncontrolled forces rescued mechano-sensitive Twist expression in the stomodeum. This indicates that Twist expression might not be highly sensitive to the intensity or symmetry of tissue deformations (Desprat, 2008).

The remote manipulation of magnetized cells in the Drosophila embryo enabled demonstration that mechanical compression of stomodeal cells comparable to those induced by endogenous morphogenetic movements upregulates Twist expression in the stomodeal primordium. Arm nuclear translocation is a major instructive step in the mechanical-to-genetic transduction pathway, coupling the macroscopic events of morphogenetic shape changes to the molecular processes regulating developmental gene expression. Moreover, previous studies showed that Src family kinases are involved in mechano-transduction through two distinct modes: either though direct mechanical activation resulting in phosphorylation of Src (Wang, 2005), or through a permissive mode where a mechanically induced conformational change in a Src substrate makes its phosphorylation site accessible to the already activated p-Src (Sawada, 2006). This study found that Src42A acts in the permissive mode in the mechano-transduction pathway upstream of Arm. Because β-catenin is a substrate of Src in mammalian cells, one might speculate that the mechano-sensitive substrate of p-Src42A in Drosophila embryos may be junctional Arm. Further study will be necessary to determine whether this is the case, or if an unknown mechano-sensitive Src42A substrate controls Arm activation (Desprat, 2008).

At later stages of development during organogenesis, mechanical cues generated by organ functions were also suggested to shape the physiological function of specialized organs. For instance, embryonic muscle activity is involved in mouse bone development through β-catenin activation. This study has found that endogenous morphogenetic movements at early stages of development are able to control gene expression, thus identifying a feedback loop of the embryo morphological development onto the genome. Such mechanical cues may mediate long-range effects that coordinate and synchronize differentiation events throughout the whole embryo. Such effects may be especially important under conditions in which dynamical and complex topology prevents the establishment of the long-range morphogen gradients that are efficient at earlier stages, when cells are arranged in simpler, static geometrical patterns (Desprat, 2008).

Dorsal interacting protein 3 potentiates activation by Drosophila Rel homology domain proteins

Dorsal interacting protein 3 (Dip3) contains a MADF DNA-binding domain and a BESS protein interaction domain. The Dip3 BESS domain was previously shown to bind to the Dorsal (DL) Rel homology domain. This study shows that Dip3 also binds to the Relish Rel homology domain and enhances Rel family transcription factor function in both dorsoventral patterning and the immune response. While Dip3 is not essential, Dip3 mutations enhance the embryonic patterning defects that result from dorsal haplo-insufficiency, indicating that Dip3 may render dorsoventral patterning more robust. Dip3 is also required for optimal resistance to immune challenge since Dip3 mutant adults and larvae infected with bacteria have shortened lifetimes relative to infected wild-type flies. Furthermore, the mutant larvae exhibit significantly reduced expression of antimicrobial defense genes. Chromatin immunoprecipitation experiments in S2 cells indicate the presence of Dip3 at the promoters of these genes, and this binding requires the presence of Rel proteins at these promoters (Ratnaparkhi, 2009).

The Drosophila genome encodes three rel homology domain (RHD) containing proteins, Dorsal (Dl), Dorsal-related immunity factor (Dif), and Relish (Rel). The RHD, which is also found in the human NFκB family of transcriptional activators, mediates dimerization and sequence-specific DNA binding. Rel/NFκB family proteins in vertebrates and invertebrates play central roles in the innate immune response by triggering the expression of antimicrobial defense genes in response to signals transduced by Toll and the Immune deficiency (Imd) signal transduction pathways. In Drosophila, Dl also directs dorsoventral (D/V) patterning of the embryo. Specifically, the regulated nuclear localization of maternally expressed Dl in response to Toll signaling in the embryo leads to the formation of a ventral-to-dorsal nuclear concentration gradient of Dl and to the spatially restricted regulation of a large number of genes, including twist (twi), snail (sna), and rhomboid (rho), which are activated by Dl, and zerknullt and decapentaplegic, which are repressed by Dl. This serves to subdivide the embryo into multiple developmental domains along its D/V axis (Ratnaparkhi, 2009).

Unlike Dl, Dif and Rel are not required for D/V patterning. Instead, these two rel-family proteins function along with Dl in the innate immune response. Toll signaling in the immune system leads to the translocation of Dl and Dif to the nucleus and the consequent activation of a subset of anti-microbial defense genes, including drosomycin (drs) and Immune induced molecule 1. Dl and Dif are believed to have redundant roles in this process and thus either one alone is sufficient for the induction of drs. Activation of the Imd signal transduction pathway, leads to proteolytic cleavage of Rel. The N-terminal region of Rel, which contains the RHD, then translocates into the nucleus where it activates expression of anti-bacterial genes, such as diptericin (dipt), cecropin-A1 (cec-A), and attacin-A. Dl, Dif, and Rel homo- and hetero-dimerize to activate different subsets of the anti-microbial defense genes in response to signals from the Toll and Imd pathways (Ratnaparkhi, 2009).

Very little is known about the identity of factors that assist the RHD proteins in the activation of the anti-microbial defense genes. Proteins that modulate expression of these genes include transcription factors such as the GATA factor Serpent (Srp), Hox factors, Helicase89B, and an unknown protein that binds region 1 (R1), a regulatory module in cec-A and other anti-microbial defense genes. In addition, a recent screen identified several POU domain proteins as potential regulators of anti-microbial defense genes (Ratnaparkhi, 2009).

To date, about a dozen proteins that interact directly with Dl and modulate its regulatory functions have been identified by genetic and biochemical means. For example, an interaction between Dl and Twist (Twi) enhances the activation of Dl target genes, while an interaction between Dl and Groucho (Gro) is essential for Dl-mediated repression. A yeast two-hybrid screen to identify Dl interacting proteins yielded, in addition to the well characterized Dl-interactors Twi and Cactus, four novel Dl-interactors (Dip1, Dip2, Dip3, and Dip4/Ubc9). Conjugation of SUMO to Dl by Ubc9 was subsequently shown to result in more potent activation by Dl (Ratnaparkhi, 2009).

Dip3 belongs to a family of proteins that contain both MADF (for Myb/SANT-like in ADF) and BESS (for BEAF, Stonewall, SuVar(3)7-like) domains. While MADF-BESS domain proteins are found in both insects and vertebrates, only a few have been characterized and their functions are largely unknown. The Drosophila genome encodes 14 MADF-BESS domain factors. In addition to Dip3, these include Adf-1, which was initially found as an activator of Alcohol dehydrogenase, and Stonewall, which is required for oogenesis. The Dip3 MADF domain mediates sequence specific binding to DNA, while the Dip3 BESS domain mediates binding to a subset of TATA binding protein associated factors as well as to the Dl RHD and to Twi. In addition to functioning as an activator, Dip3 can function as a coactivator to stimulate synergistic activation by Dl and Twi in S2 cells (Ratnaparkhi, 2009).

This study shows that Dip3 assists RHD proteins during both embryonic development and the innate immune response. By stimulating the expression of antimicrobial defense genes, Dip3 improves survival of both larvae and adults following septic injury. The presence of Dip3 near the promoters of antimicrobial defense genes depends upon Rel family proteins suggesting that Dip3 functions as a coactivator at these promoters (Ratnaparkhi, 2009).

It has been shown that Dip3, which binds both Dl and Twi via its BESS domain, synergistically enhances the activation of a luciferase reporter with multiple Dl and Twi binding sites upstream of the promoter. In addition, Dip3 has been implicated as the 'mystery protein' which binds to sites adjacent to Dl and Twi binding sites in a subset of Dl target genes. Therefore the ability of Dip3 to enhance the expression of the Dl target promoters twi, sna, and rho in S2 cell transient transfection assays was examined. All three promoters require both Dl and Twi for full activity. Dip3 was found to synergize with Dl and Twi in the activation of the sna and twi promoters, but not in the activation of the rho promoter (Ratnaparkhi, 2009).

A polyclonal antibody against recombinant Dip3 was generated, and used to determine where and when Dip3 is present in the embryo. Maternally expressed Dip3 is observed in all nuclei as early as nuclear cycle 7. It was detected in subsequent nuclear cycles during formation of the Dl nuclear concentration gradient. In interphase embryonic as well as S2 cell nuclei, Dip3 localizes to nuclear speckles of unknown identity. During mitosis Dip3 is enriched on chromosomes. It associates with the centrosome proximal portion of the anaphase chromatids and the inside ring of the polar body rosette suggesting a predominant pericentromeric location at this stage of the cell cycle and hinting at a possible role of Dip3 in centromeric function. Confirming the specificity of the antibodies, the immunoreactivity is absent from Dip3¹ embryos in which the Dip3 transcriptional and translational start sites as well as a large segment of the Dip3 coding region have been deleted. Weak Dip3 expression is also detected in the fat body (Ratnaparkhi, 2009).

Homozygous Dip3¹ flies are viable and fertile, indicating that Dip3 cannot have an essential role in embryonic D/V pattern formation. However, a small proportion (7±4%) of the embryos fail to hatch and exhibit D/V patterning defects. Embryos produced by females transheterozygous for Dip3¹ and a deficiency that removes a portion of the second chromosome containing the Dip3 gene (Df(PC4) exhibit similar embryonic lethality (10%) and D/V patterning defects. Also, maternal overexpression of Dip3 using the Gal4-UAS system leads to 54±9 % embryonic lethality with cuticles of the dead embryos showing both anteroposterior and D/V patterning defects, indicating that Dip3 may have a role in embryonic pattern formation (Ratnaparkhi, 2009).

Consistent with a non-essential role for Dip3 in D/V patterning, a Dip3 mutation enhances the temperature sensitive dl haploinsufficieny phenotype. The degree of dorsalization is often quantified by categorizing embryos on a scale from D0 (completely dorsalized, lacking all dorsoventral pattern elements other than dorsal epidermis) to D3 (inviable, but with little or no apparent defect in the cuticular pattern). At 29°, about half the dead embryos produced by dl¹/+ females exhibit detectable D/V patterning defects and the majority of these fall into the D2 category (moderately dorsalized, exhibiting mildly expanded ventral denticle belts and a twisted germ band). Removal of maternal Dip3 increases the proportion of dorsalized embryos to about 75% with most of the increase being due to an increase in the number of D2 embryos. The effect seems to be strictly maternal as the paternal genotype does not modulate the dl haploinsufficiency phenotype (Ratnaparkhi, 2009).

Dip3 is present in the fat body, the organ in which RHD factors activate antimicrobial defense genes in response to infection. Since Dip3 binds the Dl RHD, the role of Dip3 in the innate immune response was examined by assessing the sensitivity of Dip3¹ flies to bacterial and fungal infection. Wild-type and Dip3¹ adults and larvae were injected with gram positive bacteria (M. luteus), gram negative bacteria (E. coli), and fungi (B. brassiana). For comparison, flies were infected that contained mutations in known components of the Toll (spz^rm7) and Imd (Rel^E20) pathways. Wild-type, Rel^E20, spz^rm7, and Dip3¹ adults showed little lethality (<15%) 30 days after mock infection. However, the Dip3¹ adult flies exhibited 55% lethality one month after injection with a 1:1 mixture of M. luteus and E. coli, compared to 10% lethality after 30 days for wild-type flies and 98% after 30 days for Rel^E20 flies. In contrast, wild-type and Dip3¹ adults were equally sensitive to fungal infection, both showing 55-70% lethality after 30 days compared to 100% lethality after 22 days for Rel^E20 adults and 100% lethality after 7 days for spz^rm7 adults. Similar results were seen in larvae in which Dip3¹, Rel^E20 and spz^rm7 mutations resulted in reduced rates of eclosion following septic injury compared to wild-type. The effectiveness of the immune challenge was further verified by an experiment showing that septic injury leads to translocation of Dl into the nucleus (Ratnaparkhi, 2009).

To determine if the sensitivity of Dip3¹ flies to infection results from reduced induction of antimicrobial peptides, the expression of dipt, drs and cec-A was monitored as a function of time following septic injury. Relative to uninfected flies, the levels of expression of drs and dipt were reduced by the Dip3¹ mutation, especially at the 2 and 4 hr time points, while the levels of cec-A expression were not significantly altered. Thus, some, but not all, antimicrobial defense genes that are regulated by RHD family proteins exhibit dependence on Dip3. At the 4 hr time point, relative to infected, wild type flies, the spz^rm7 mutation reduced drs expression to basal levels while the Rel^E20 mutation reduced dipt expression ten fold (Ratnaparkhi, 2009).

Dip3 was over expressed in the larvae using the Cg-Gal4 driver to examine the effect of increasing levels of Dip3 on the expression of antimicrobial defense genes in the fat body. Cec-A and drs levels were unaffected, while dipt levels increased two-fold in infected flies. Thus, both loss-of-function and over expression data are consistent with the conclusion that Dip3 makes the immune response more robust by elevating the expression of a subset of antimicrobial defense genes (Ratnaparkhi, 2009).

Radiolabeled Dip3 interacts with FLAG-tagged Dl and Rel immobilized on anti-FLAG beads. Similarly, immobilized FLAG-Dip3 binds Dl (Bhaskar, 2002) and Rel (Residues 1-600). Dip3 binds to DNA via its MADF domain and to the RHD via its BESS domain, and can thus function either as an activator or as a coactivator (Bhaskar, 2002). To determine if Dip3 is present at the promoters of antimicrobial defense genes, ChIP assays were carried out in S2 cells transfected with FLAG-Dip3. FLAG antibody was used to immunoprecipitate Dip3 crosslinked to chromatin. Compared both to mock-transfected cells and to the transcribed region of a ribosomal protein-encoding gene (rp49), Dip3 was highly enriched at the drs, dipt and cecA promoters. As expected, dsRNA directed against Dip3 eliminated the ChIP signal verifying antibody specificity. The association of Dip3 with the promoters of the anti-microbial defense genes depended on Rel family proteins, since knockdown of these proteins by dsRNAi significantly reduced association of Dip3 with the promoters. Similar results were observed with an anti-GFP antibody and cells expressing a Dip3-GFP fusion protein (Ratnaparkhi, 2009).

These results suggest that Dip3 may synergize with RHD proteins in multiple developmental contexts possibly through contact with the Dl rel homology domain. Dip3 is expressed maternally and present in cleavage stage nuclei at the time that Dl is functioning to pattern the D/V axis. Furthermore, Dip3 can potentiate Dl-mediated activation of the twist and snail promoters in S2 cells. These observations suggest that Dip3 might have a role in D/V patterning. Consistent with this possibility, it was found that removal of maternal Dip3 results in occasional D/V patterning defects and significantly enhances the dl haploinsufficiency phenotype suggesting the Dip3 renders D/V patterning more robust perhaps by assisting in Dl-mediated activation (Ratnaparkhi, 2009).

An important aspect of the immune response is activation in the fat body of genes encoding antimicrobial peptides by the Rel family transcription factors Dl, Dif, and Rel. This study found that synergistic killing of flies by a mixture of E.coli and M. luteus is enhanced in Dip3¹ flies. This suggests roles for Dip3 in the Imd and/or Toll pathways, which mediate the response to microbial infection. In accord with this idea, it was found that activation of the Imd pathway target dipt and the Toll pathway target drs are compromised in Dip3 mutant larvae (Ratnaparkhi, 2009).

To determine if the role of Dip3 at antimicrobial defense gene promoters is direct, ChIP assays were carried out demonstrating that this factor associates directly with the drs, dipt, and cec-A promoters in S2 cells. Since Dip3 contains a DNA binding domain, it is possible that it binds to these promoters through a direct interaction with DNA. However, with one exception in the drs promoter, these promoters lack matches for the consensus Dip3 binding sites. Thus, Dip3 may be acting as a coactivator at these promoters consistent with its ability to bind the rel homology domain. In support of this idea, it was found that simultaneous knockdown of all three rel family proteins significantly reduced recruitment of Dip3 to the promoters (Ratnaparkhi, 2009).

The mechanism of Dip3 co-activation remains unclear. The finding that the Dip3 BESS domain binds TAFs (Bhaskar, 2002) suggests a role for Dip3 in the recruitment of the basal machinery. In addition, the MADF domain is closely related to the SANT domain, which binds histone tails and may have a role in interpreting the histone code. While analysis of RHD targets suggests roles for Dip3 in activation, Dip3 also associates with pericentromeric heterochromatin during mitosis, consistent with a possible role in silencing. Other heterochromatic proteins including a suppressor of position effect variegation (Su(Var)3-7) also contain BESS domains. However, the loss of Dip3 does not appear to modify position effect variegation (Ratnaparkhi, 2009).

In flies, additional roles for RHD-mediated activation have been demonstrated in haematopoesis, neural fate specification, and glutamate receptor expression. Antimicrobial defense genes are also expressed constitutively in barrier epithelia and in the male and female reproductive tracts. It will be interesting to determine if Dip3 is involved in rel protein-dependent and independent gene activation in some or all of these tissues. One tissue in which Dip3 appears to have clear rel-independent functions is in the developing compound eye, where Dip3 overexpression results in conversion of eye to antenna, while Dip3 loss-of-function leads to mispatterning of the retina (Ratnaparkhi, 2009 and references therein).

Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.

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