Targets of Activity

Dorsal functions as both an activator and repressor of transcription to determine dorsoventral fate in the Drosophila embryo. Repression by Dorsal requires the corepressor Groucho (Gro) and is mediated by silencers termed ventral repression regions (VRRs). A VRR in zerknullt (zen) contains Dorsal binding sites as well as an essential element termed AT2. An AT2 DNA binding activity has been identified (called ZREB) and purified in embryos. It consists of cut (ct) and dead ringer (dri) gene products. dri was isolated as a novel gene encoding a sequence-specific DNA-binding protein. Dri is a founding member of a growing protein family whose members share a conserved DNA binding domain termed the A/T-rich interaction domain. dri is developmentally regulated, being expressed in a restricted set of cells including some neural cells and differentiating cells of the gut and salivary gland ducts. Dri is a member of the recently defined ARID family of DNA binding proteins, a family that includes the B-cell-specific factor Bright and the Drosophila factor Eyelid. Although Bright is thought to function as a transcriptional activator, genetic data suggest that Eyelid functions to repress transcription in response to activation of the wingless pathway (Valentine, 1998 and references).

Studies of loss-of-function mutations in ct and dri demonstrate that both genes are required for the activity of the AT2 site. Dorsal and Dri both bind Gro, acting cooperatively to recruit it to the DNA. Thus, ventral repression may require the formation of a multiprotein complex at the VRR. This complex includes Dorsal, Gro, and additional DNA binding proteins, all of which appear to convert Dorsal from an activator to a repressor by enabling it to recruit Gro to the template. By showing how binding site context can dramatically alter transcription factor function, these findings help clarify the mechanisms responsible for the regulatory specificity of transcription factors (Valentine, 1998).

To determine if cut and dir are required for the activity of the AT2 site in vivo, the effects of mutations in these genes were examined on the activity of the lacZ transgene under control of the minimal zen VRR. For both cut and dir, germ line clones were generated to test the effects of eliminating maternally contributed gene products, and, in addition, the effects of eliminating zygotically produced gene products were examined. A null mutation in ct (which is an X-linked gene) results in strong ventral derepression of the transgene. This ventral derepression is observed in about one-half the embryos derived from a cross between females containing ct germ line clones and hemizygous males. It was never observed in a cross between heterozygous females and hemizygous males, suggesting that derepression requires simultaneous elimination of both maternal and zygotic Ct. A strong hypomorphic mutation in dri (which is an autosomal gene) also results in strong derepression. In contrast to the results observed with ct, this effect is strictly zygotic. It is observed in a cross between heterozygous dri males and females but not in a cross between females carrying dri germ line clones and wild-type males. Most strikingly, in the absence of zygotic Dri, the zen VRR directs strong ventral expression in the blastoderm embryo, reminiscent of the results observed when the AT2 element is mutagenized. These results strongly suggest that, in the context of the minimal zen VRR, Dri plays an essential role in converting Dorsal from an activator into a repressor. The dri mutation results in a significant weakening of the transverse eve stripe (generated by the minimal even skipped (eve) stripe 2 enhancer (MSE) as well as a shift in the position of the stripe toward the anterior pole of the embryo, presumably due to a role for Dri in anteroposterior pattern formation. Despite the strong effects of the cut and dir mutations on the activity of the minimal zen VRR, both genes make only minor contributions to the ventral repression of the endogenous zen gene in the stage 4 embryo. In the absence of both zygotic and maternal Ct or in the absence of zygotic Dri, zen expression in the stage 4 embryo is still largely restricted to the dorsal 40 to 50% of the embryo, although weak ventral patches of zen expression are observed with high frequency. Such patches are never observed in wild-type embryos stained in parallel with these embryos. The contrast between the strong effect observed for the minimal VRR and the weak effect observed for the endogenous zen gene suggests redundancy in the zen locus. In other words, there may be additional unidentified ventral repression regions in the zen locus that function in a Ct- and Dri-independent manner. Although neither Ct nor Dri is essential for ventral repression of the endogenous zen gene in the stage 4 embryo, both factors appear to play essential roles in the refinement of the zen pattern that normally occurs in stage 5 embryos. Normally, zen expression refines during cellularization to a stripe approximately three to five cells in width. However, in the absence of both maternal and zygotic Ct or in the absence of zygotic Dri, a severe refinement defect is observed (Valentine, 1998).

Both Dorsal and Dri bind to the corepressor Gro in vitro, suggesting a possible mechanism for repression in which Dorsal and Dri recruit Gro to the template. This model is strengthened by results showing that Dorsal and Dri bound to DNA can cooperatively recruit Gro to the zen VRR in vitro. However, the magnitude of the cooperativity observed in vitro is small (twofold) and therefore does not completely account for the absolute requirement for the Dorsal and AT2 sites observed in germ line transformation assays. This suggests that factors in addition to Dorsal and Dri are required for the efficient recruitment of Gro in vivo. For example, it is possible that the addition of Ct would enhance cooperative recruitment, an idea that could not be tested due to difficulty obtaining sufficient amounts of recombinant Ct. It is also likely that elements in addition to Dorsal sites and AT2 are required for efficient Gro recruitment and therefore for efficient repression, since previous experiments indicate that, while these sites are required for repression, they are not sufficient for repression. Finally, it is possible that the cooperativity of Gro recruitment would be enhanced in the context of chromatin templates rather than naked DNA templates (Valentine, 1998).

During a screen to identify cut-interacting genes, it was observed that flies containing a hypomorphic cut mutation and a heterozygous deletion of the Antennapedia complex exhibit a transformation of mouthparts into leg and antennal structures similar to that seen in homozygous proboscipedia (pb) mutants. The same phenotype is produced with all heterozygous pb alleles tested and is fully penetrant in two different cut mutant backgrounds. This phenotype is accompanied by pronounced changes in the expression patterns of both cut and pb in labial discs. These experiments implicate cut in the regulation of expression and/or function of two homeotic genes (Johnston, 1998).

The adult mouthparts are produced from the labial imaginal discs. Proboscipedia is expressed in nuclei of labial disc cells in third instar larvae; Cut is expressed in a pattern that substantially overlaps that of Pb and is also nuclear. In wild-type and ctL188 discs, Cut and Pb are expressed throughout the entire disc, however in ctL188; pb5/+ double mutant discs both the level and the pattern of expression for both proteins is altered. The level of expression for both proteins is significantly decreased overall and entirely lost in some of the cells. Where present, Cut expression appears more punctate in comparison to wild-type discs. The mutant discs are morphologically abnormal. Staining of ctL188; pb5/+ labial discs with acridine orange shows no consistent increase in apoptotic cell death relative to that in control discs at this stage. Pb expression is undetectable in pb1/pb5 mutant labial discs and the pattern of Cut expression is altered to resemble that of a leg imaginal disc, in which Cut is expressed in two rows of cells in the position of the future claw organ in the most distal segment (Johnston, 1998).

A significant proportion of cut mutant flies that are heterozygous for certain Antennapedia (Antp) alleles have thoracic defects that mimic loss-of-function Antp phenotypes: ectopic expression of Cut in antennal discs results in ectopic Antp expression and a dominant Antp-like phenotype. Evidence for nonautonomous functions of cut can be found in its interaction with Antp, since ectopic Cut expression in a stripe along the anterior-posterior compartment boundary of both portions of the eye-antennal disc results in uniform activation of Antp throughout the disc. The incidence of outgrowths in the dorsal prothoracic region of flies that are mutant for cut and heterozygous for Antp is highest in the presence of the recessive allele Antp11. However, thoracic outgrowths are consistently observed in combination with dominant alleles, such as AntpWu and and AntpR. Many of the dominant Antp alleles are also recessive lethal, so the outgrowths seen in combination with these alleles can still be reconciled with a loss of Antp function. It has been speculated that the outgrowths may be the product of a homeotic change. Although bristles are generally present on all the outgrowths characterized, they are generally malformed and segment or appendage-specific features could not be identified (Johnston, 1998).

By examining expression of arc in different mutant embryos, it was determined that transcription factors known to be required for patterning and maintenance of various developing epithelia control arc expression in those domains. tll and hkb, which are required to pattern the posterior 15% of the embryo, control arc expression in the posterior midgut primordium. fkh, which appears to act as a maintenance, or permissive, transcription factor, is required for expression of arc throughout the gut. byn, which is required for hindgut development and specifies its central domain (the large intestine), controls expression of arc in the elongating hindgut. Kr and cut, required for evagination and extension of the Malpighian tubule buds control expression of arc in the tubule primordia (Liu, 2000).

The neural selector gene cut, a homeobox transcription factor, is required for the specification of the correct identity of external (bristle-type) sensory organs. Targets of cut function, however, have not been described. bereft (bft) mutants exhibit loss or malformation of a majority of the interommatidial bristles of the eye and cause defects in other external sensory organs. These mutants were generated by excising a P element located at chromosomal location 33AB, the enhancer trap line E8-2-46, indicating that a gene near the insertion site is responsible for this phenotype. Similar to the transcripts of the gene nearest to the insertion, reporter gene expression of E8-2-46 coincides with Cut in the support cells of external sensory organs, which secrete the bristle shaft and socket. Although bft transcripts do not obviously code for a protein product, bereft's expression is abolished in bft deletion mutants, and the integrity of the bft locus is required for (interommatidial) bristle morphogenesis. This suggests that disruption of the bft gene is the cause of the observed bristle phenotype. Attempts were made to determine what factors regulate the expression of bft and the enhancer trap line. The correct specification of individual external sensory organ cells involves not only cut, but also the lineage genes numb and tramtrack. Mutations of these three genes affect the expression levels at the bft locus. Furthermore, cut overexpression is sufficient to induce ectopic bft expression in the PNS and in nonneuronal epidermis. On the basis of these results, it is proposed that bft acts downstream of cut and tramtrack to implement correct bristle morphogenesis (Hardiman, 2002).

In an effort to identify and characterize genes that might integrate information from the selector gene cut and lineage gene ttk, bereft was cloned. bft is expressed in es (mechano- and chemo-receptors), but not in ch (chordotonal) support cells. Analysis of flies with deletions of the bft locus, together with the es support cell-specific expression pattern, suggest that bft function is required for correct morphogenesis of the cuticular structure forming support cells, in particular those of the interommatidial bristles of the eye. Moreover, bft expression in es organs is reduced in cut and ttk mutants, and cut and ttk interact genetically with bft. These data are consistent with the idea that bft is a target for cut and ttk in the implementation of es organ-specific structures (Hardiman, 2002).

Targets of both cut and ttk were sought on the basis of the expression pattern of candidate genes within the PNS. cut is expressed in all the cells of es organs (at higher levels in support cells), whereas ttk is found in three es and two ch support cells, but not the neurons. Thus, the support cells of es organs express both cut and ttk, suggesting that genes responsive to these two pathways (i.e., the pathways leading to organ identity specification and lineage decisions, respectively) should also be expressed in these cells. An enhancer trap line, E8-2-46, was identified in which the lacZ reporter gene is expressed primarily in the support cells of es organs within the PNS, on the basis of position, morphology, and cut expression. Although E8-2-46 is expressed in both es support cells (as identified by high levels of cut expression), the level of expression is lower in one of them. To determine which of the two cell support cells express the reporter gene more strongly, the dorsal-most abdominal es organ (desD) were examined. DesD are aligned in a stereotyped linear fashion: tormogen, trichogen, thecogen, and neuron (from dorsal to ventral). Strong reporter activity is observed in the bristle shaft-forming trichogen cell, whereas cut expression predominates in the shaft-forming tormogen cell (Hardiman, 2002).

By examining both Cut protein and bft transcripts in the same embryo, it has been found that the es precursors express bft transcripts almost coincident with the onset of Cut expression. At later stages, bft transcripts are restricted to the support cells of es organs. Furthermore, bft transcripts are expressed in nonneural tissues that also express Cut, such as in the cephalic segments, and the precursors of both the anterior and posterior spiracles. In the absence of Cut activity, bft expression is reduced or absent. Conversely, the ectopic expression of Cut drives ectopic bft transcription. Moreover, consensus Cux/Cut-binding sites have been identified upstream of the bft transcript: ATC GATTA is found 600 and 660 bp upstream of the transcript start site, and a CCAAT motif, recognized by Cut repeat II, is also found near one of these sites. This, together with the overexpression data, suggests that Cut may activate bft transcription directly. However, Cut is unlikely to be the only factor regulating bft transcription, since in cut null mutants, bft expression is not completely absent (Hardiman, 2002).

Whether cut function activates or modulates bft transcription in the PNS was investigated by examining cutdb7 null mutant embryos. In the absence of cut function, E8-2-46 reporter gene expression is reduced in es support cells. In wild-type embryos, bft is also expressed in the developing posterior spiracles, which is severely reduced or absent in cut mutants (Hardiman, 2002).

Does cut suffice to activate bft transcription? The UAS-Gal4 system or a heat shock promoter was used to drive cut ectopically in embryos and to examine the resulting pattern of bft transcription. When hairy-Gal4 is used to drive cut expression in the odd-numbered segments, bft is expressed ectopically, most noticeably near the dorsal abdominal PNS clusters, which are the embryonic origins of the lateral chordotonal organs. Ectopic expression of cut causes es-specific gene expression in these chordotonal organs and prevents their lateral migration. Thus, the cell fate changes in the PNS induced by cut result in ectopic bft expression. Furthermore, cells that normally never express cut, in particular ectodermal cells overlying the central nervous system, are induced to express bft when cut is ectopically expressed. Since cut can induce ectopic bft expression outside the PNS, it may participate directly in the regulation of bft. Consistent with this hypothesis, consensus Cut binding sites were identified immediately upstream of the bft 5' RACE products (Hardiman, 2002).

Lozenge causes transcriptional activation of Cut, which then stabilizes a Lozenge repressor complex, which regulates Deadpan

Runx proteins have been implicated in acute myeloid leukemia, cleidocranial dysplasia, and stomach cancer. These proteins control key developmental processes in which they function as both transcriptional activators and repressors. How these opposing regulatory modes can be accomplished in the in vivo context of a cell has not been clear. The developing cone cell in the Drosophila visual system was used to elucidate the mechanism of positive and negative regulation by the Runx protein Lozenge (Lz). A regulatory circuit is described in which Lz causes transcriptional activation of the homeodomain protein Cut, which can then stabilize a Lz repressor complex in the same cell. Whether a gene is activated or repressed is determined by whether the Lz activator or the repressor complex binds to its upstream sequence. This study provides a mechanistic basis for the dual function of Runx proteins that is likely to be conserved in mammalian systems (Canon, 2003).

To understand negative regulation by the Lz protein, regulation of the deadpan (dpn) gene was investigated. In wild-type eyes, Dpn is expressed in photoreceptors R3/R4 and R7. In lz mutants, dpn is also ectopically activated in cone cells, suggesting that Lz either directly or indirectly represses dpn in these cells. Dpn was therefore used as a marker to investigate negative regulation by Lz (Canon, 2003).

The presence of two perfect consensus Runx protein-binding sites (5'-RACCRCA-3') upstream of the dpn-coding region suggested possible direct negative regulation by Lz. Gel-shift experiments showed that Lz specifically binds to both sites. To determine whether these sequences are required for proper dpn regulation, lacZ reporter constructs were made driven by dpn upstream and intronic fragments, and these were transformed into flies. A 4667-bp upstream fragment plus intron I (227 bp) caused expression of lacZ in R3/R4 and R7 faithfully recapitulating the pattern of wild-type dpn expression in the eye. This site is therefore referred to as the dpn eye enhancer (DEE). When the two Lz-binding sites (LBS) in the DEE were mutated (to 5'-RAAARCA-3'; DEE-MutLBS), lacZ expression was also seen in cone cells. Therefore, lack of Lz binding to this enhancer will cause its derepression in cone cells, establishing that Lz directly represses transcription of dpn in cone cells (Canon, 2003).

The homeodomain protein Cut is expressed specifically in the four cone cells in the eye and has been shown previously to bind AT-rich sequences. The ability of Cut to bind the AT-rich sequences next to the Lz sites in the DEE was tested. Electromobility-shift assays were conducted using probes containing the Lz-binding sites and adjacent AT sequences from the dpn enhancer. Nuclear extracts of cells transfected with a Cut-expressing vector bind the two AT-rich sequences, and this binding is specific as established by competition assays. Further, extracts from cells transfected with both lz and cut cause a supershifted band, indicating that Lz and Cut can bind together to the same probe (Canon, 2003).

To address the in vivo relevance of these results, FLP/FRT-mediated clones were made in the eye that were mutant for the cut locus. Strikingly, Dpn was ectopically expressed in cone cells in the absence of Cut. This provides genetic proof that, in vivo, Cut represses dpn expression in cone cells. Cut is therefore required along with Lz for repression of dpn in these cells (Canon, 2003).

Interestingly, D-Pax2, which is directly activated by Lz, is needed to activate cut in cone cells. Therefore, although indirectly, Lz positively regulates cut. This presents an interesting developmental circuit in which Lz, acting as a transcriptional activator, causes expression of a cofactor that then binds with Lz to convert it into a direct repressor of transcription. Both the presence of the cofactor and binding sites for this cofactor in the controlling regions of an Lz target gene are required for Lz-mediated repression (Canon, 2003).

This model was then tested in R7 cells where both Dpn and Lz are coexpressed. Here, Lz does not repress dpn, presumably because Cut is absent from R7. Consistent with this notion, mis-expression of Cut in R7 cells using lz-Gal4 causes repression of dpn in these cells. This is not a secondary result of a change in cell fate because the expression of the R7 cell-specific marker Prospero remains unchanged in this genetic background (Canon, 2003).

These results add another level of complexity to recent studies demonstrating a combinatorial code whereby a relatively small number of signaling pathways and activated transcription factors work together to generate unique cell fates. In cone cells, the Notch and EGFR pathways are required along with Lz to activate D-Pax2, and therefore cut. In contrast, the combination of these few inputs is not right for activation of cut in the R7 neurons, and therefore dpn is not repressed. The circuit described here demonstrates a higher order of sophistication necessary for a cell to choose between a neuronal and nonneuronal fate using a very limited number of inputs. Using a self-regulated circuit and just two signaling pathways, a single Runx protein is capable of causing opposing effects on different enhancers in the same cell, resulting in a unique fate (Canon, 2003).

Runx proteins have been shown to act as positive and negative regulators. This study, however, is the first to demonstrate that a Runx protein can act as both a direct transcriptional activator and repressor in vivo in the same cell, and that the repressive role requires involvement of the cofactor Cut. The mechanism unraveled here for a Runx protein is similar to that described for a Rel protein, suggesting a common strategy adopted by transcription factors that switch between positive and negative regulation. Furthermore, Cut is conserved in mammals (called CDP or Cux) and has been implicated in the repression of several genes, including osteocalcin (OC). Interestingly, the OC gene is positively regulated by Runx2. These in vitro studies did not investigate a relationship between Runx and Cux. This analysis of dpn repression by Lz and Cut raises the possibility that mammalian Runx proteins may also switch from activation to repression modes through involvement of Cux proteins. If confirmed, such correlations will prove to be important as the mammalian Runx protein AML-1 (Acute Myeloid Leukemia-1) is the most frequent site of translocations that cause leukemia, and human CutL1 is located in a chromosomal region that is often rearranged in cancers, including myeloid leukemia (Canon, 2003 and references therein).

The gene promoter of Drosophila PCNA contains several transcriptional regulatory elements, such as upstream regulatory element (URE), DNA replication-related element (DRE, 5'-TATCGATA), and E2F recognition sites. In the present study, a yeast one-hybrid screen using three tandem repeats of DRE in PCNA promoter was used as the bait allowed isolation of a cDNA encoding Cut, a Drosophila homolog of mammalian CCAAT-displacement protein (CDP)/Cux. Electrophoretic mobility shift assays showed that Cut binds to both DRE and the sequence 5'-AATCAAAC in URE, with much higher affinity to the former. Measurement of PCNA promoter activity by transient luciferase expression assays in Drosophila S2 cells after an RNA interference for Cut or DREF showed DREF activates the PCNA promoter while Cut functions as a repressor. Chromatin immunoprecipitation assays in the presence or absence of 20-hydroxyecdysone further showed both DREF and Cut proteins to be localized in the genomic region containing the PCNA promoter in S2 cells, especially in the Cut case upon induction of differentiation. These results indicate that Cut functions as a transcriptional repressor of PCNA gene by binding to the promoter region in the differentiated state, while DREF binds to DRE to promote expression of PCNA during cell proliferation (Seto, 2006).

Regulation of the Drosophila distal antennal determinant spineless by cut

The transformation of antenna to leg is a classical model for understanding segmental fate decisions in Drosophila. The spineless (ss) gene encodes a bHLH-PAS transcription factor that plays a key role in specifying the identity of distal antennal segments. This report identifies the antennal disc enhancer of ss and then uses enhancer-lacZ reporters to work out how ss antennal expression is regulated. The antennal determinants Distal-less (Dll) and homothorax (hth) are key activators of the antennal enhancer. Dll is required continuously and, when present at elevated levels, can activate the enhancer in regions devoid of hth expression. In contrast, homothorax (hth) is required only transiently both for activation of the enhancer and for specification of the aristal portion of the antenna. The antennal enhancer is repressed by cut, which determines its proximal limit of expression, and by ectopic Antennapedia (Antp). Repression by Antp is not mediated by hth, suggesting that ss may be a direct target of Antp. ss+ is not a purely passive target of its regulators: ss+ partially represses hth in the third antennal segment and lies upstream of Dll in the development of the maxillary palp primordia (Emmons, 2007; full text of article).

This study used lacZ reporters to identify the enhancers responsible for most aspects of ss expression during embryonic and imaginal development. Antennal expression is driven by two large fragments from the ss 5' region, B6.9 and EX8.2. Both of these fragments drive expression in the antennal segment of the embryo and in the distal portion of the pupal antenna. B6.9 is also expressed in the antennal disc through most or all larval development. Dissection of B6.9 allowed localization of the larval antennal enhancer to a fragment of 522 bp. The B6.9 and 522 reporters were used as a proxy for ss expression in experiments to determine the effects of potential upstream regulators of ss. This strategy has its strengths and weaknesses, but has been made necessary by an inability to generate antisera against Ss. A major strength of the approach is that it was possible to assess the effects of regulators on individual enhancers. It is likely that monitoring endogenous ss expression would give results that are less clear cut since both the antennal and tarsal enhancers of ss are active within the antenna. A potential weakness is that the reporters may not faithfully reproduce the normal expression of ss. However, as far as is possible to tell, the antennal reporters reproduce ss expression very well. The expression of B6.9 and EX8.2 in the embryonic antennal segment and the pupal antenna corresponds very closely to that of endogenous ss. Expression of B6.9 and 522 in the larval antennal disc appears very similar or identical to that of ss+, and the transient requirement for hth+ in the activation of these reporters corresponds well to the transient requirement for hth+ in aristal specification. The tarsal enhancer P732 likely also reproduces the spatial pattern of ss+ expression as its tarsal expression domain corresponds well to the region deleted in ss mutants (Emmons, 2007).

The results of this dissection of the B6.9 fragment were surprising. Removal of the left-hand 2 kb of B6.9 to produce S4.9 resulted in the loss of antennal specificity; S4.9 reporters are expressed in both antennal and leg discs. The E2.0 subfragment of S4.9 shows a similar expression pattern, and expression of this fragment in both leg and antennal discs is independent of Hth, but requires Dll continuously. On further subdivision of the E2.0 fragment, it was found that antennal and leg expression are separable; the 522 fragment is largely specific for the antenna, whereas the 531 fragment drives expression primarily in leg discs. To summarize, antennal specificity is present in B6.9, lost in S4.9 and E2.0 and regained in 522. How can sense be made of this? The region deleted from B6.9 to produce S4.9 clearly plays an important role in enforcing antennal specificity. Since this region contains a PRE, one might suspect that it functions in larval stages to maintain repression of the enhancer outside of the antennal segment. However, that the E2.0 fragment has lost the requirement for Hth in both the antenna and leg (S4.9 has not been tested) suggests that the PRE-containing region might function in both locations. One possibility is that this region represses the enhancer in both antennal and leg discs. In the antenna, this repression can be overcome by the combined action of Hth and Dll, while in the leg Dll alone is not sufficient for activation. When the PRE-containing region is deleted, repression is absent or reduced, so that Dll can activate the enhancer without assistance from Hth, and expression is seen in both antennal and leg discs. Why then is antennal specificity restored in the 522 subfragment? Perhaps this fragment is lacking a subset of Dll interaction sites so that it can no longer be activated by Dll alone, but requires combined activation by Hth and Dll. Although this model is consistent with many of the results, it does not provide a ready explanation for the leg specificity of the 531 fragment (Emmons, 2007).

In addition to activation by combined Hth and Dll, the ss antennal disc enhancer is repressed by Cut and by ectopic Antp. Each of these regulators will be discussed separately. It was found that hth+ is required only transiently for activation of the B6.9 reporter. hth clones induced in the embryo or first instar lose expression of B6.9 autonomously in both A3 and the aristal primordia. However, some time in the second of early third instar. Regulatory instar expression of B6.9 becomes independent of hth. Consistent with this transient requirement, it is shown that hth+ is required only early in larval development for specification of the arista. hth clones induced in the first and second instars show a transformation of the entire antenna to a leg-like appendage. However, clones induced after this time show normal aristal development. These temporal requirements are reflected in the expression pattern of hth: hth is expressed throughout the antennal primordium early in development, but in the second or early third instar is repressed in the central domain, which will produce the arista (Emmons, 2007).

The stable activation of B6.9 by Hth suggests that this fragment contains a 'cellular memory module'. The presence of a PRE within B6.9 is consistent with this idea. The ss locus binds Polycomb protein in salivary gland chromosomes and was recently shown to contain PREs by chromatin immunoprecipitation. In the latter work, ss PREs were localized to within the E1.6 subfragment of B6.9 as well as the EX8.2 fragment, both of which showed pairing dependent suppression in this work. PREs are generally thought of as functioning to stably repress genes. However, PREs can also be associated with activating elements to form memory modules that mediate stable activation. It seems likely that B6.9 contains such a module that responds to Hth. Like a memory module from the hedgehog gene, activity of the ss module is set sometime around the second instar. Surprisingly, it was found that activation of the 522 reporter by Hth can also be persistent, although not as stable as for B6.9. The 522 fragment does not appear to contain a PRE, suggesting that Hth may directly recruit factors to the 522 element that cause semi-stable transcriptional activation (Emmons, 2007).

ss is not a completely passive target of hth; ss partially represses hth in antennal discs, which causes hth to be expressed at a lower level in A3 than in A2. This repression appears to be important for normal development as ectopic expression of Hth can delete A3. Moreover, clones ectopically expressing Hth are largely blocked from entering A3 from the proximal (A2) side, suggesting that the different levels of Hth present in A2 and A3 cause a difference in cell affinities between these segments. Hth-expressing clones are similarly restricted to the two most proximal segments in leg discs, although here there is no endogenous expression of hth more distally (Emmons, 2007).

In contrast to hth, Dll is required continuously for expression of both B6.9 and 522 as Dll clones induced even very late in development lose expression of these reporters. This continuous requirement for Dll indicates that stable activation of the B6.9 memory module by Hth does not by itself commit the reporter to expression; rather, activation by Hth appears to render B6.9 open to interaction with Dll and perhaps other positive factors (Emmons, 2007).

Three lines of evidence suggest that Dll is the primary activator of the ss antennal enhancer. (1) It was found that expression of B6.9 and 522 is sensitive to the dosage of Dll+. Expression of both reporters is reduced in animals carrying only one dose of Dll+, and for 522, expression is enhanced in clones having extra doses of Dll+. This dose sensitivity suggests that ss is a direct target of Dll. (2) It was found that expression of both reporters is often induced within clones expressing ectopic Dll, even in the apparent absence of Hth expression. Such activation is seen in clones in the distal leg, wing and elsewhere. (3) It was found that the embryonic antennal enhancer carried by B6.9 is absolutely dependent upon Dll+, but independent of hth. Taken together, these observations suggest that Dll is a primary activator of the ss antennal enhancers. Hth may provide antennal specificity by boosting the level of activation by Dll in the antennal disc (Emmons, 2007).

Surprisingly, it was found that the regulatory relationship between ss and Dll is reversed in the maxillary palp. Here, ss is expressed prior to Dll and is required for the normal initiation of Dll expression. Although some Dll expression ultimately takes place in the palp primordium in ss animals, this expression is weak and occurs in only a few cells. It has not been worked out how ss is activated in the palp. However, it seems likely that dpp plays a role as the 531 subfragment of B6.9 drives expression in a stripe in the region of the palp that roughly coincides with a stripe of dpp expression. The positioning of ss upstream of Dll in the palp may explain why the region ventral to the antenna is so sensitive to ectopic expression of Ss. Strong activation of Dll here by ectopic Ss combined with endogenous expression of hth might be expected to cause frequent induction of ectopic antennae, as is observed. Since ss is normally expressed in the palp, why should earlier ectopic Ss cause the palp primordium to develop as antenna? It seems likely that timing is key, but level of Ss expression could also be important (Emmons, 2007).

The reciprocal regulatory roles of ss and Dll in the antenna and palp suggest a particularly close relationship between these genes. This relationship is reinforced by the finding that ss is required for the development of bracts in the femur, as is Dll (Emmons, 2007).

The finding that Dll and Hth are both activators of the ss antennal reporters is consistent with the proposal that antennal identity is defined by the combined activity of these regulators. However, the results indicate that this model is an oversimplification. Examination of clones expressing Dll, Hth, or both proteins together revealed little correlation between activation of the B6.9 and 522 antennal reporters and combined expression of Dll and Hth. Strikingly, Dll-expressing clones often activate the reporters ectopically without any apparent concomitant expression of Hth, and clones expressing both proteins usually do not activate the reporters. These experiments also reveal strong context dependence. Examples include the leg, where Dll-expressing clones can activate the reporters distally, but not proximally (where endogenous hth expression occurs) and the wing disc, where clones expressing Dll or both Dll and Hth activate the reporters in the wing pouch, but not at all in the notum. The level of expression of both proteins also appears to be key as high levels of Dll can activate the reporters in the leg in the absence of Hth and elevated levels of Hth can repress expression in the normal antennal domain. Previous results have shown that antennal structures can be induced by ectopic expression of Dll in the wing hinge region or proximal leg (which express hth endogenously) or by combined expression of Dll and Hth elsewhere. While this is true, the results indicate highly variable effects in such ectopic expression experiments and fail to detect the strongly synergistic activation of antennal identity by combined Hth and Dll implied by the model. The results indicate that Dll is the primary activator of the ss antennal reporters, that Hth serves to promote this activity and that activation by Dll and Hth is highly context-dependent (Emmons, 2007).

Consistent with direct control of the antennal reporters by Dll and Hth, two highly conserved regions within the 522 fragment contain apparent binding sites for Dll, Hth, and the Hth dimerization partner Extradenticle. The functional importance of these binding sites is currently being tested (Emmons, 2007).

This study has show that the proximal boundary of B6.9 and 522 expression is defined by repression by cut. This repression likely explains why ectopic Cut causes a transformation of arista to tarsus. cut has been shown to define the proximal expression limit of distal antenna (dan) and distal antenna related (danr); since ss lies upstream of these genes, it seems very likely that their regulation by cut is indirect. The mechanism of action of Cut is not well understood, since only one direct target has been characterized in Drosophila (Emmons, 2007).

Ectopic expression of Antp in the antenna represses the B6.9 and 522 reporters. This finding was expected, since it is well known that expression of Antp or other Hox genes in the antenna causes a transformation to leg. The conventional view is that this transformation results from the repression of hth by ectopic Hox proteins. Repression of hth early in development would be expected to lead secondarily to loss of ss expression and loss of distal antennal identity. However, it was found that clones expressing Antp repress the B6.9 and 522 reporters even when these clones are induced very late in development, long after the requirement for activation by hth has passed. Late repression of the antennal reporters by Antp must therefore occur independently of hth and could be direct. One possibility, currently being tested, is that Antp might compete with Dll for binding to the 522 enhancer. Late repression of the ss antennal enhancer by Antp is consistent with the effects of Antp-expressing clones on antennal identity: such clones induced in the mid to late third instar cause transformations of distal antenna to leg (Emmons, 2007).

Clones induced late that ectopically express Antp in a sustained fashion were examined. In contrast, previous work studied the effects of pulses of Antp expression induced by one-hour heat shocks in a heat shock/Antp line. It had been found that transformations of arista to tarsus were induced by such pulses only when they were administered at the end of the second instar. Why do pulses of Antp at this time cause a stable, heritable transformation of the distal antenna? The current results suggest an explanation. The period sensitive to Antp pulses coincides roughly with when the ss antennal enhancer becomes independent of hth. This correlation suggests that pulses of Antp in the second instar cause heritable transformations by interfering with the stable activation of ss by Hth. Recently, it has been reported that ectopic Antp does not repress hth in the antenna early in larval development. This observation suggests that Antp might act directly on the ss antennal enhancer to prevent its stable activation by Hth (Emmons, 2007).

The regulation of ss by ectopic Antp suggests that Antp may normally play a significant role in repressing ss antennal enhancer activity in the legs. Although this idea has not been tested directly, it seems unlikely that Antp is primarily responsible for keeping the ss antennal enhancers inactive in the leg. Antp null clones do cause activation of the ss target gene dan in leg discs, implying ectopic activation of ss. However, this activation occurs only proximally, with the distal leg appearing to develop independently of Antp. Expression of Antp in the proximal leg may account for why Dll-expressing clones fail to activate B6.9 or 522 in this location. Ectopic activation of the ss antennal enhancers in the leg primordia of the embryo is not seen in an Antp null mutant (Emmons, 2007).

These studies suggest that antennal structures are specified in a combinatorial fashion by Hth, Dll, Ss and probably other factors. In A3, all three proteins are required for normal antennal identity. In ss antennae, hth continues to be expressed in A3 (although at elevated levels), as does Dll. Despite this continued expression of hth and Dll, A3 develops without antennal characteristics and produces only naked cuticle. Thus, Hth and Dll are unable to specify A3 characters in the absence of Ss. Conversely, assuming that ss is stably activated in the antenna by Hth, as is B6.9, then hth clones induced late would show persistent expression of both ss and Dll in A3. Such clones are transformed to leg, implying that Ss and Dll have no ability to direct A3 identity in the absence of Hth. Taken together, these observations suggest that Hth, Dll and Ss must act together to specify A3 identity. This requirement for combined action accounts for why ectopic expression of Ss does not induce A3 tissue in the medial leg, since hth is not normally expressed here. The view of combinatorial control suggests that many A3-specific target enhancers might be identifiable in genome searches as regions that contain clustered binding sites for Hth, Dll and Ss; tests of this prediction will be presented elsewhere (Emmons, 2007).

In contrast to A3, the aristal primordium appears to be specified by ss and Dll acting together in the absence of hth expression. hth is expressed in the aristal region early in development, where it functions to establish ss expression, but it is soon repressed here. Therefore, for most of development, the arista is specified by Ss and Dll acting without input from Hth. Consistent with this picture, the arista adopts leg identity in ss null mutants, and ectopic expression of ss causes the distal tip of the leg to develop as arista (Emmons, 2007).

In ss mutants, the distal antenna is terminated by a single tarsal segment (the fifth). In contrast, in ss mutants that lack only antennal enhancer activity (e.g. the breakpoint mutations ssD114.3 and ssD114.7, the distal antenna develops with a near complete set of tarsal segments. This difference likely reflects the activity of the tarsal enhancer in the antenna. In support of this view, the ss tarsal enhancer drives expression in the segmented base of the arista, a region known as the basal cylinder. This region transforms to tarsal segments 2-4 in Antp-induced transformations of antenna to leg. However, the question arises as to why normal antennal expression of ss causes the proximal arista to develop as basal cylinder, whereas ss expression driven by the tarsal enhancer alone causes this same region to develop as tarsal segments. Likely, the key difference is that expression driven by the tarsal enhancer is transient, whereas expression driven by the antennal enhancer is sustained. Perhaps transient expression of ss allows growth and subsegmentation to produce a full set of tarsal segments, whereas sustained expression inhibits growth, producing the basal cylinder. Consistent with this idea, sustained expression of ss driven by the GAL4 method can cause deletion of tarsi in the legs. The levels of expression driven by the tarsal and antennal enhancers may also be important as flies having only one dose of ss show a partial transformation of the basal cylinder to tarsus. The ss tarsal enhancer drives weak expression in A3 as well as in the basal cylinder, likely accounting for the presence of some specialization of A3 in ss mutants lacking the antennal enhancers (Emmons, 2007).

The view that antennal identity is specified by the combined action of Hth, Dll and Ss contradicts the now prevalent view that antennal identity is determined solely by hth. The major evidence supporting the latter view is that early hth clones transform the entire antenna to leg, and ectopic expression of Hth can induce ectopic antennal structures in the anal plates. Moreover, Dll shows little antennal specificity, being expressed in the distal portions of all of the ventral appendages, and ss expression in the antenna is dependent upon hth+. Should hth be viewed as the antennal 'selector' gene? hth does not seem to be a selector in the same sense as the Hox genes; it is expressed very broadly in the embryo and in other imaginal discs and plays no role in activating ss in the antennal segment of the embryo. Moreover, the ability of ectopic Hth to induce antennal structures is very limited: transformations of anal plate to distal antenna have been reported following ectopic expression of Hth or Meis1, a mammalian homolog. However, others have been unable to reproduce this effect by ectopic expression of Hth, matching the results of this study. That anal plates are susceptible to transformation at all is likely due to the fact that Dll and ss are coexpressed here in normal development. A further dissimilarity is that hth acts only as an establishment regulator of ss in the antennal disc, unlike the continuous requirements usually seen for the Hox genes. Ultimately, assessment of the importance of hth will depend on whether its function in the antenna is conserved. The expression pattern of hth in the antenna does appear to be conserved in the milkweed bug Oncopeltus. However, localization of nuclear Exd (a proxy for Hth expression) indicates that Hth is not differentially expressed in the antenna and leg of the cricket. Expression of hth in the crustacean Porcellio also appears to be identical in the second antenna and the legs. Characterization of hth, Dll and ss expression and function in additional arthropods will be required to assess properly the importance of these genes in antennal specification (Emmons, 2007).

Control of the spineless antennal enhancer: direct repression of antennal target genes by Antennapedia

It is currently thought that antennal target genes are activated in Drosophila by the combined action of Distal-less, homothorax, and extradenticle, and that the Hox gene Antennapedia prevents activation of antennal genes in the leg by repressing homothorax. To test these ideas, a 62bp enhancer was isolated from the antennal gene spineless that is specific for the third antennal segment. This enhancer is activated by a tripartite complex of Distal-less, Homothorax, and Extradenticle. Surprisingly, Antennapedia represses the enhancer directly, at least in part by competing with Distal-less for binding. Antennapedia is required in the leg only within a proximal ring that coexpresses Distal-less, Homothorax and Extradenticle. It is concluded that the function of Antennapedia in the leg is not to repress homothorax, as has been suggested, but to directly repress spineless and other antennal genes that would otherwise be activated within this ring (Duncan, 2010).

This report examines the regulation of an enhancer from the antennal gene ss that drives expression specifically in the third antennal segment (A3). The work provides the first look at how the homeodomain proteins Dll, Hth, and Exd function in the antenna to activate antennal target genes. These proteins form a trimeric Dll/Hth/Exd complex on the enhancer, suggesting that Dll acts much like a Hox protein in antennal specification. This work also reveals how the Hox protein Antp functions in the leg to repress antennal development. The conventional view has been that the primary function of Antp is to repress hth in the distal leg, which then prevents the activation of all downstream antennal genes. However, this study found that Antp represses the ss A3 enhancer directly. This repression is essential within a proximal ring in the leg that coexpresses the antennal gene activators Dll, Hth, and Exd. Antp competes with Dll for binding to the enhancer, and this competition is part of a molecular switch that allows the ss A3 element to be activated in the antenna, but represses its activation in the leg. The results suggest that repression of antenna-specific genes in the proximal ring is the sole function of Antp in the leg imaginal disc (Duncan, 2010).

At 62 bp, the ss A3 enhancer (called D4) is one of the smallest enhancers to be identified in Drosophila, and yet it is quite strong; only a single copy is required to drive robust expression of lacZ reporters. The enhancer is also very specific, driving expression in A3 and nowhere else in imaginal discs. It has been proposed that antennal identity in Drosophila is determined by the combined action of Dll, Hth, and Exd. Consistent with this proposal, all three of these factors were found to be required for D4 expression. Although these activators are coexpressed in both A2 and A3, D4/lacZ expression is restricted to A3 by Cut, which represses the enhancer in A2. Like ss itself, D4/lacZ is also repressed by ectopically expressed Antp (Duncan, 2010).

A previous report (Emmons, 2007) showed that the antennal expression pattern of ss is reproduced by lacZ reporters containing a 522 bp fragment from the ss 5' region. This fragment contains five conserved (41%-90% identity) domains, each of which was deleted and tested for effect on expression in vivo. Expression in the arista and the third antennal segment (A3) prove to be under separate control; expression in the arista requires domains 1, 3 and 5, whereas expression in A3 is lost only when domain 4 is deleted. Moreover, reporters containing domain 4 alone show expression in A3 and nowhere else in imaginal discs. Thus, domain 4 is both necessary and sufficient for A3-specific expression. Domain 4 (D4) is 62 bp in length and is highly conserved, being invariant at 50/62 base pairs in the 12 Drosophila species sequenced (Duncan, 2010).

Surprisingly, Dll, Hth, Exd, Cut, and Antp all act directly upon D4. The activators Hth and Exd bind with strong cooperativity to directly adjacent sites. Their joint binding site matches the optimum site for in vitro binding of the mammalian homologs of Hth and Exd (Meis and Prep), consistent with the robust activity of the enhancer in vivo. Mutation of either of these sites abolishes activity of the enhancer. The coactivator Dll binds three sites in D4; one of these sites (Dlla) is required for almost all activity of the enhancer. Dll shows strong cooperativity with Hth and Exd for binding to D4, indicating that Dll interacts physically with these proteins. This interaction requires DNA binding, as Dll protein containing a missense change that blocks DNA binding (a change of asn51 to ala in the homeodomain) shows no ability to associate with D4-bound Hth and Exd. A curious feature of the cooperativity seen in the binding studies is that although Hth and Exd increase the affinity of Dll for D4, Dll appears to have little effect on the affinity of Hth and Exd for the enhancer. Since Hth and Exd already bind cooperatively with one another, it may be that additional cooperative interactions with Dll have little effect. Alternatively, it may be that Hth and Exd interact with Dll only after binding DNA. If so, Hth and Exd would be expected to increase Dll binding to D4, but Dll would have little effect on the binding of Hth and Exd, as observed. Interactions between Dll and Hth in the absence of DNA have been reported in immunoprecipitation experiments. However, this study was unable to repeat these observations. Moreover, the finding that the asn51 mutant of Dll fails to associate with D4-bound Hth and Exd argues strongly against such interactions (Duncan, 2010).

The repressor Cut also acts directly upon D4. Binding of Cut requires two sites, one overlapping Dlla and the other overlapping the joint Hth/Exd site. These binding sites suggest that D4 is controlled by Cut in much the same way that a structurally similar Abdominal-A (Abd-A) regulated enhancer from the rhomboid gene is controlled by the repressor Senseless (Sens). In the rhomboid enhancer, adjacent Hth and Exd sites are also present, and these create a binding site for Sens. Activity of the rhomboid enhancer is controlled by a competition between binding of the Sens repressor and binding of the activators Abd-A, Hth, and Exd. It seems likely that D4 is controlled similarly, with the repressor Cut competing for binding with the activators Dll, Hth, and Exd. It will be of interest to determine whether enhancers similar to D4 are used more widely to control Cut targets involved in its role as an external sense organ determinant (Duncan, 2010).

A key finding in this work is that Antp represses D4 by direct interaction. Antp binds a single site in D4, which overlaps or is identical to the Dlla binding site. Like Dll, Antp binds cooperatively with Hth and Exd. Using purified proteins, it was showm that binding of Dll and Antp to the Dlla site is mutually exclusive. This indicates that Antp represses the enhancer at least in part by competing with Dll for binding. Similar competition may occur at other enhancers; when Antp expression is driven artificially in the distal leg, variable deletions of the tarsal segments occur. These defects might arise because Antp competes with Dll for binding to its target genes in the distal leg. In most other contexts examined, Antp is an activator of transcription; why it fails to activate D4 is not clear. The similar behavior of Dll and Antp in binding to D4 supports the idea that Dll behaves like a Hox protein in activating D4 (Duncan, 2010).

Although the initial focus of this study was on the antenna, the finding that Antp interacts directly with D4 led to an examination of D4 regulation in the leg, where Antp is normally expressed. In second leg imaginal discs, Antp is required only in a proximal ring of cells that coexpresses Dll and Hth. This ring appears in the early third instar, and is of uncertain function. Large Antp clones in T2 leg discs that do not enter this ring appear to develop completely normally, regardless of whether they are located distal or proximal to the ring. However, clones that overlap the ring show activation of D4/lacZ within the ring cells. Importantly, such clones have no effect on the expression of Dll or Hth within the ring. By examining Antp clones of increasing age the following sequence of events is inferred. First, D4/lacZ is activated in cells of the ring that are included within Antp clones. Second, many such clones begin expressing the antennal markers Ss and Cut, indicating a transformation to antenna, and round up as if they have lost affinity for neighboring cells. Third, such clones appear to extend and move distally in the disc (Duncan, 2010).

The events described for Antp clones in the leg make sense of several previously enigmatic observations. It has been noted that many Antp clones in the leg do not transform to antenna and appear to develop normally. The finding that only clones that overlap the proximal ring undergo transformation accounts for this observation. Antp clones that do contain transformations usually show apparent nonautonomy in that not all cells in the clone are transformed to antenna. The current results account for this observation as well, since within an Antp leg clone only those cells located in the proximal ring undergo transformation to antenna; cells located elsewhere in the clone retain normal leg identity. Most importantly, these observations provide an explanation for why ss is controlled directly by Antp. Antp clones have no effect on hth or Dll expression in the proximal ring. Therefore, Antp must function in the ring at the target gene level to repress antennal genes that would otherwise be activated by combined Hth and Dll (and Exd). Since several such targets are known, it seems likely that several, perhaps many, antennal genes in addition to ss are repressed directly by Antp (Duncan, 2010).

Transformed Antp clones in the leg often show ectopic hth expression in distal locations. If hth is not directly controlled by Antp in the leg, as this study suggests, then why is hth ectopically expressed within such clones? A likely explanation is that downstream antennal genes that have become activated in such clones feed back to activate hth. This interpretation is strongly supported by the finding that ectopic expression of the antennal genes ss, dan, or danr in the distal leg causes ectopic activation of hth. Thus, the distal expression of hth seen in Antp leg clones is likely a consequence rather than a cause of the transformation to antenna. Whether repression of hth in the antenna by ectopic Antp is also indirect is not clear. Dll is also expressed ectopically in transformed Antp leg clones, suggesting that it is also subject to feedback activation by downstream antennal genes (Duncan, 2010).

The function of the proximal Dll- and Hth-expressing ring in the proximal leg is not well understood. The ring is highly conserved among the insects, and may serve as a boundary between the proximal and distal portions of the legs. In the context of this work, a striking feature of the ring is that it contains a microcosm of gene expression domains corresponding to the three major antennal segments. Thus, proceeding from proximal to distal through the ring, cells express hth alone, hth + Dll, and hth + Dll + strong dachshund. These expression combinations are characteristic of the A1, A2, and A3 antennal segments, respectively. Looked at in this way, the ring would appear to resemble a repressed antennal primordium within the leg (Duncan, 2010).

It has been known for almost thirty years that Antp is required in the leg to repress antennal identity. However, an understanding of how this repression occurs has been lacking. The current results indicate that Antp functions within the proximal ring to directly repress antennal genes that would otherwise be activated by combined expression of Dll, Hth, and Exd. This appears to be the only function of Antp in the leg, at least during the third instar larval stage. The results are entirely consistent with the idea that second leg is the 'ground state' ventral appendage (the limb type that develops in the absence of identity specification) and that the role of Antp in the leg is to preserve this ground state by repressing the activation of 'head-determining' genes (Duncan, 2010).

Turtle functions downstream of Cut in differentially regulating class specific dendrite morphogenesis in Drosophila

Dendritic morphology largely determines patterns of synaptic connectivity and electrochemical properties of a neuron. Neurons display a myriad diversity of dendritic geometries which serve as a basis for functional classification. Several types of molecules have recently been identified which regulate dendrite morphology by acting at the levels of transcriptional regulation, direct interactions with the cytoskeleton and organelles, and cell surface interactions. Although there has been substantial progress in understanding the molecular mechanisms of dendrite morphogenesis, the specification of class-specific dendritic arbors remains largely unexplained. Furthermore, the presence of numerous regulators suggests that they must work in concert. However, presently, few genetic pathways regulating dendrite development have been defined. The Drosophila gene turtle belongs to an evolutionarily conserved class of immunoglobulin superfamily members found in the nervous systems of diverse organisms. Turtle is differentially expressed in Drosophila da neurons. Moreover, MARCM analyses reveal Turtle acts cell autonomously to exert class specific effects on dendritic growth and/or branching in da neuron subclasses. Using transgenic overexpression of different Turtle isoforms, context-dependent, isoform-specific effects were found on dendritic branching in class II, III and IV da neurons. Chromatin immunoprecipitation, qPCR, and immunohistochemistry analyses demonstrated that Turtle expression is positively regulated by the Cut homeodomain transcription factor, and genetic interaction studies demonastrated that Turtle is downstream effector of Cut-mediated regulation of da neuron dendrite morphology. These findings reveal that Turtle proteins differentially regulate the acquisition of class-specific dendrite morphologies. In addition, a transcriptional regulatory interaction between Cut and Turtle was established, representing a novel pathway for mediating class specific dendrite development (Sulkowski, 2011).

Analyses of tutl-GAL4 and Tutl protein expression reveal differential expression levels in da neuron subclasses. The highest levels of tutl expression were observed in the more complex class III and IV da neurons, whereas moderate levels of Tutl expression were observed in class II da neurons, followed by weaker levels in class I da neurons. This differential expression pattern correlates with certain aspects of morphological complexity by which these neurons have been subdivided into classes. Specifically, the highest levels of Tutl expression correlate with the appearance of numerous, short processes extending from longer primary branches characteristic of class III and IV da neurons. Interestingly, the transcription factor Cut, a known regulator of dendritic growth, shows a strikingly similar pattern of expression in da neurons In addition to differential Tutl expression in da neuron cell bodies, Tutl localization to dendrites and axons was observed (Sulkowski, 2011).

Consistent with differential Tutl expression levels, novel, cell autonomous and class specific defects were observed in da neuron dendrite morphogenesis in tutl mutants. In class III da neurons, tutl mutation had objectively the most severe phenotype. A significant reduction in both branching and length was observed. Overall, class III da neurons mutant for tutl appeared smaller than their wild-type counterparts, with reduced branching and a reduction in number and length of their characteristic 'spine-like' processes. In fact, by directly analyzing the length of terminal branches, it was possible to show that tutl is required for full extension of these processes. MARCM analyses further revealed that class IV and class II da neurons each require Tutl for normal dendrite development. However, Tutl appears to regulate different aspects of morphogenesis in each class. For the complex, space-filling dendrites of class IV neurons, Tutl is required for dendrites to reach the appropriate length and establish full dendritic field coverage. In contrast, class II neurons require Tutl primarily to promote an appropriate number of branches (Sulkowski, 2011).

Many of the findings reported herein sharply contrast with those of a previous study (Long, 2009) in which a very similar set of experimental analyses were performed to investigate tutl function in da neurons. Among the differences in findings between studies is the question of tutl function in class I da neurons. A previous study implicated Tutl in restricting class I neuron dendritic branching complexity (Long, 2009), however, in general, this study found that Tutl is largely dispensable for class I branching. In light of this striking difference, the previous studies were repeated using the same tutl23 mutant allele, and the analyses were extended to incorporate additional novel allelic combinations and to include all three class I da neuron subtypes, however the results of the previous study in any class I neuron could not be reproduced. These results demonstrate that, contrary to previous findings, Tutl function is largely dispensable for class I dendrite development and moreover, indicate that allelic variation is not a contributing factor to the observed differences between studies at least for class I neurons (Sulkowski, 2011).

In addition, this study found that tutl exerts differential effects on class II and III dendritic branching and growth, whereas a previous study (Long, 2009), using the tutl23 mutant allele, failed to reveal any significant effects on class II and III dendrite development. The basis for the differences between the two studies are not entirely clear, however, distinctions between studies include the use of independent tutl alleles, suggesting the potential of allelic variation with respect to phenotypic effects, as well as the fact that the previous study focused exclusively on analyses of the class II ddaB and class III ddaA neurons, whereas the present study examined effects on class II or III neurons as a whole incorporating data from all subtypes within an individual class. Finally, while previous reports, using the tutl23 allele, suggested that Tutl was required to mediate dendritic self-avoidance in class IV da neurons, this study could not detect any significant effect on self-avoidance. Ultimately, the basis for the observed phenotypic differences with respect to self-avoidance remain unclear, however may again potentially be attributable to allelic variation as independent alleles were used in each study (Sulkowski, 2011).

Isoform-specific Tutl overexpression was found to differentially regulate dendrite branching in class II, III, and IV da neurons, however produced no significant defects in class I da neuron dendrite development. Consistent with these findings, the Long (2009) study reported an inhibition of class IV dendrite branching upon overexpression of the membrane bound Tutl RE40452 protein isoform and observed no defects in class I dendrite development upon Tutl RE40452 overexpression. Interestingly, the current results demonstrate that overexpression of one of the previously characterized non-membrane bound Tutl isoforms, AT02763, inhibited terminal dendritic branching in class IV ddaC neurons, whereas overexpression of two membrane-bound Tutl isoforms, HL01565 and LD28224, promoted terminal dendritic branching in ddaC neurons. Given that both membrane bound and soluble isoform overexpression inhibit class IV dendrite branching, whereas two other membrane bound isoforms promote branching there is no simple relationship in terms of soluble vs. trans-membrane Tutl isoforms in regulating class IV dendritic branching. Moreover, analyses of isoform-specific overexpression in class II and III da neurons further illustrate the complexity of Tutl function in mediating class specific dendritic branching. Overexpression of the soluble AT02763 and GH15753 isoforms in class II neurons promotes dendritic branching, whereas the membrane bound isoforms had no effect. In contrast, in class III neurons, overexpression of AT02763 and GH15753 isoforms in class III neurons had opposing effects on dendritic branching and again overexpression of the membrane bound forms had no effect. Consistent with these isoform-specific dendritic effects, a recent study demonstrated Tutl is required for normal CNS axon development and found that specific Tutl protein isoforms function as attractive axon guidance cues and promote axon branching and invasiveness (Al-Anzi, 2009). The Al-Anzi study demonstrated that overexpression of the membrane bound Tutl isoforms, HL01565 and LD28224, produced increased branching of ISNd motor neurons and sprouting of extra axonal processes in retinal neurons similar to the increased dendritic branching observed with overexpression of these Tutl isoforms in class IV da neurons. Moreover, this study demonstrated that soluble Tutl isoforms (e.g., AT02763) can act as attractive axon guidance cues suggesting Tutl may also exhibit non-cell autonomous functions. However, overexpression of AT02763 in class IV da neurons led to inhibition of dendrite branching, suggesting a potential repulsive function for dendrite branching. Multiple functioning of a gene in both dendrite and axon growth is not without precedent. Another conserved IgSF member, the receptor Roundabout (Robo), first discovered for its regulation of axonal midline crossing has recently been shown as necessary for proper dendrite growth in the CNS and PNS. In da neurons, Robo serves to restrict dendrite outgrowth, consistent with its role as a receptor of repulsive signals. Collectively, these isoform-specific overexpression studies suggest that distinct Tutl isoforms may function in a complex context-dependent, cell-type specific manner to regulate dendritic branching morphology. It is possible that these various Tutl isoforms function in concert to 'fine tune' class specific dendritic branching complexity (Sulkowski, 2011).

The Cut homeodomain transcription factor was among the first genes shown to regulate class specific da neuron dendrite morphogenesis; however, to date the pathway through which it functions has been unknown. Although Cut was shown to influence the level of the Knot transcription factor expression, a direct transcriptional relationship was not established. The current studies reveal that Tutl expression is positively regulated by the Cut. Cut specifically binds to the tutl promoter sequence and acts as a transcriptional activator of Tutl expression in da neurons; this constitutes a novel transcriptional regulatory mechanism by which Tutl may mediate class-specific dendrite morphogenesis. Moreover, genetic interaction studies reveal that Tutl functions as a downstream effector of Cut-mediated regulation of a da neuron dendrite morphogenesis (Sulkowski, 2011).

Expression in class I da neurons, albeit at low levels, established that Cut is not the sole transcriptional regulator of Tutl since Cut is not normally detectable in class I neurons. Consistent with this conclusion, this study and that of Long (2009) were able to detect Tutl protein expression in cut mutant MARCM clones. Collectively, these results indicate that Cut is not absolutely or solely required for Tutl expression in da neurons, however this does not preclude the potential that Cut contributes to transcriptional regulation of Tutl in da neurons as the data indicate. Interestingly, ectopic expression of Cut in class I da neurons leads to a significant increase in dendritic branching complexity, whereas overexpression of Tutl in class I da neurons has no demonstrable effect on dendritic complexity, although the current genetic interaction data reveal that Tutl is required for the observed Cut overexpression effects on class I dendritic complexity. These data suggest that Cut regulation of Tutl expression in class I neurons alone is insufficient to modulate dendritic morphology. This raises the possibility that Cut may positively regulate transcription of both Tutl and a putative partner molecule and thus the gain of function dendrite phenotype observed with ectopic Cut expression is the result of co-upregulation of both Tutl and an as yet unidentified Tutl-interacting protein. In support of this possibility are recent findings in which the deletion of the Tutl carboxy-terminal domain (CTD) of the longest known membrane bound Tutl isoform, RE40452, failed to produce specific loss of function defects in dendrite morphogenesis (Long, 2009), whereas tutl mutations in general produce defects suggesting that a potential partner, possibly a co-receptor, may be required to mediate downstream signaling essential for normal class specific dendrite development. Other members of the IgSF require cis or trans interactions to perform cellular functions. For example, Robo-1 interacts with DCC via conserved cytoplasmic regions to neutralize the attractive effect of Netrin-1. Furthermore, axonal repulsion following Slit binding appears to rely on interactions between Robo-1 and the intracellular phosphoprotein Enabled, which modulates cytoskeleton dynamics. As Cut has been shown to modulate dendritic complexity through the actin cytoskeleton, it would be interesting to determine if potential cytoskeletal changes accompany tutl mutation. The differential changes based on neuronal class that tutl mutants display may manifest via different changes in the underlying dendritic cytoskeletal architecture (Sulkowski, 2011).

The Drosophila tutl gene shares evolutionary homology with both humans (KIAA1355) and mice (Dasm1). The data indicate that Tutl is required for the differential regulation of class specific dendrite morphogenesis including dendritic extension and branching in class II-IV da neurons. These results are consistent with initial studies of Dasm1 which was found to be required in promoting dendrite outgrowth in cultured rat hippocampal neurons (Shi, 2004). However, later studies using a Dasm1 genetic knockout did not observe the reported morphological defects, nor any overt behavioral phenotype. A possible explanation for this discrepancy is the presence of the co-expressed and closely related ortholog IgSF9b which may contribute to functional redundancy. As such, the contribution of mammalian homologs to mediating dendrite morphogenesis remains somewhat unclear. Future studies examining double knockout animals will likely be needed to fully elucidate the contribution of mammalian Tutl homologs to the regulation dendrite development. Due to its apparent lack of functional redundancy, further studies of the Tutl/Dasm1/KIAA1355 family of proteins using Drosophila may be a more direct route to further study this evolutionarily conserved gene family (Sulkowski, 2011).

Similar to Tutl, evolutionarily conserved Cut-related proteins in vertebrates have likewise been implicated in regulating class-specific dendrite morphogenesis. Previous studies have demonstrated that Cut-related protein, CDP, is functionally conserved in regulating dendrite morphogenesis as expression of CDP can rescue cut mutant dendritic defects in Drosophila da neurons and ectopic expression of CDP can phenocopy dendritic defects observed with ectopic Cut expression. Moreover, a recent study has implicated the Cut-related, Cux1 and Cux2 genes, in regulating dendritic branching, spine morphology, and synaptogenesis in upper layer neurons of the cortex. Moreover, these analyses implicate Cux1 and Cux2 in regulating subclass-specific mechanisms of synapse regulation providing further evidence for the evolutionarily conserved role of Cut-related proteins in the specification of class-specific dendritic morphologies. The fact that both Tutl- and Cut-related proteins display evolutionarily conserved roles in directing dendrite morphogenesis, coupled with the findings of a transcriptional regulatory interaction between Tutl and Cut, together with genetic interaction data indicating Tutl functions as a downstream of effector of Cut in da neuron dendrite morphogenesis suggests the potential that murine Cut homologs may also function in transcriptionally regulating the murine Tutl homolog, DASM1 and/or the highly orthologous IgSF9b. Ultimately, elucidating the molecular mechanisms by which Tutl and Cut mediate differential effects on dendrite development will require additional studies aimed at the identification of Tutl-interacting proteins which may be distinct in individual neuron subclasses and identification of other downstream effectors of Cut transcriptional regulation in da neurons (Sulkowski, 2011).

A transcriptional network controlling glial development in the Drosophila visual system

In the nervous system, glial cells need to be specified from a set of progenitor cells. In the developing Drosophila eye, perineurial glia proliferate and differentiate as wrapping glia in response to a neuronal signal conveyed by the FGF receptor pathway. To unravel the underlying transcriptional network, this study silenced all genes encoding predicted DNA-binding proteins in glial cells using RNAi. Dref and other factors of the TATA box-binding protein-related factor 2 (TRF2) complex were previously predicted to be involved in cellular metabolism and cell growth. Silencing of these genes impaired early glia proliferation and subsequent differentiation. Dref was found to control proliferation via activation of the Pdm3 transcription factor, whereas glial differentiation was regulated via Dref and the homeodomain protein Cut. Cut expression was controlled independently of Dref by FGF receptor activity. Loss- and gain-of-function studies showed that Cut was required for glial differentiation and was sufficient to instruct the formation of membrane protrusions, a hallmark of wrapping glial morphology. This work discloses a network of transcriptional regulators controlling the progression of a naïve perineurial glia towards the fully differentiated wrapping glia (Bauke, 2015).

Using a genome-wide RNAi-based screen this study has unravelled the transcriptional machinery responsible for such a switch during gliogenesis in the Drosophila eye. During embryonic development the anlage of the eye imaginal disc is formed. It is attached to the forming brain through the so-called Bolwig's nerve. A few glial cells reside along this nerve, presumably generated in the segmental nerves, as are most of the glia. These glial cells proliferate extensively during larval stages to form ~300 glial cells within each eye imaginal disc. During the third larval stage ~50 of these cells differentiate into wrapping glia in an FGFR-dependent manner. This study shows that the proliferation of the glial progenitor pool requires the activity of Pdm3 and the DNA replication-related element-binding factor (Dref), which are both strongly expressed by proliferating perineurial glia. Dref was first identified as an important factor required for efficient transcription of the proliferating cell nuclear antigen (PCNA), a key regulator of replication. Dref protein associates with the TATA box-binding protein related factor 2 (TRF2), which functions as a core promoter selectivity factor that governs a restricted subset of genes co-ordinately regulated. Interestingly, pan-glial knockdown of TRF2 also results in lethality, suggesting that the Dref/TRF2 complex is active in glia. Knockdown of CG30020, encoding a member of the Dref/TRF2 complex, or osa and moira, which had been shown previously to interact with Dref, caused similar glial phenotypes in the visual system (Bauke, 2015).

TRF2 targets several classes of TATA-less promoters present in more than 1000 genes, including a cluster of ribosomal protein genes. Likewise, Dref was found to associate with many genes involved in protein synthesis and cell growth, and loss of Dref results in reduced organismal growth rates. Most likely, dividing glial cells as well as differentiating wrapping glia have an increased protein synthesis demand, which might explain the observed defects in proliferation and differentiation. This study shows that expression of the transcription factor Pdm3 depends on Dref. Previously, Pdm3 has been associated with axonal pathfinding. The current results indicate that Pdm3 also regulates cell number. In contrast to Dref, Pdm3 expression is repressed by FGFR signalling, ensuring that perineurial glia routed to differentiation do not express Pdm3 anymore (Bauke, 2015).

Previous work suggested that in the Drosophila eye imaginal disc perineurial glial cells at the anterior margin of the eye field are competent to react to a neuronal signal inducing their glial differentiation. During this phase glial cells have reduced Dref expression but increased FGFR activity. Whereas in the absence of FGFR signalling no glial differentiation can be observed, high levels of FGF signalling trigger the expression of Cut specifically in wrapping glia. In addition to Cut, Dref is essential for proper glial differentiation. Dref is required for normal Cut expression levels but gain of Dref function is unable to activate Cut ectopically in perineurial glial cells. This requires additional FGFR signalling, indicating that two parallel molecular pathways converge on the activation of the transcription factor Cut to orchestrate wrapping glial differentiation (Bauke, 2015).

In the Drosophila PNS Cut controls the ES/ChO lineage decision. By contrast, during glial cell development this work defined Cut as a master regulator organizing elaborated membrane growth, which is required during the wrapping of axons. Similarly, Cut instructs the morphogenesis of multi-dendritic neurons. In mammals, the Cut homologues Cux1/2 also control dendritic branching, the number of dendritic spines and synapses. The number of filopodial extensions correlates to the level of Cut expression, corresponding to these findings. It was recently shown that Cut-dependent filopodia formation depends on the function of CrebA, which activates components of the secretory pathway. Cut might not only orchestrate membrane organization through the modulation of the secretory pathway, it also directly controls cytoskeletal dynamics. In larval sensory da neurons, the actin bundling protein Fascin is necessary for a Cut-dependent induction of spiked cell protrusions. However, eye disc glial cells still form long cell processes when fascin expression is suppressed by RNAi. Further understanding of wrapping glial cell differentiation will require the identification of the transcriptional targets of Cut (Bauke, 2015).

In conclusion, this study demonstrates that the specification of wrapping glial cells in the developing visual system does not require a single lineage switch gene but rather appears as a gradual process. The specification of wrapping glia is orchestrated by a transcriptional network comprising Pdm3, Dref and Cut that is modulated by the activity of the FGFR (Bauke, 2015).

cut: Biological Overview | Evolutionary Homologs | Regulation | Developmental Biology | Effects of Mutation | References

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

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