Transcriptional Regulation

Bithorax complex genes repress exd in the posterior of the embryo (Rauskolb 1993).

Patterning function of homothorax/extradenticle in the thorax of Drosophila

In Drosophila, the morphological diversity is generated by the activation of different sets of active developmental regulatory genes in the different body subdomains. This study investigates the role of the homothorax/extradenticle (hth/exd) gene pair in the elaboration of the pattern of the anterior mesothorax (notum). These two genes are active in the same regions and behave as a single Hox independent functional unit. Their original uniform expression in the notum is downregulated during development and becomes restricted to two distinct, alpha and ß subdomains. This modulation appears to be important for the formation of distinct patterns in the two subdomains. The regulation of hth/exd expression is achieved by the combined repressing functions of the Pax gene eyegone (eyg) and of the Dpp pathway. hth/exd is repressed in the body regions where eyg is active and that also contain high levels of Dpp activity. Evidence is presented for a molecular interaction between the Hth and the Eyg proteins that may be important for the patterning of the alpha subdomain (Aldaz, 2005).

This study deals with a novel hth/exd function: its patterning role in the notum. It is not related to the specification of notum identity because notum identity is not affected by alterations of hth/exd activity. For example, in the absence of hth/exd, the cells still differentiate as notum, if an abnormal one. Conversely, high and uniform Hth levels also produce notum tissue but with abnormal pattern. This function is only required in part of the notum and is therefore linked to the modulation of hth expression during the development of the disc. The final result of this modulation is the appearance of the alpha and ß subdomains of hth that is reported in this study. These two subdomains differentiate distinct notum patterns, suggesting that Hth/Exd interact with other localised products to generate these patterns (Aldaz, 2005).

Thus, there are two principal aspects in the patterning function of hth/exd: (1) the spatial regulation, that eventually results in the restriction of its expression to the alpha and ß subdomains, and (2) the local interactions of Hth/Exd with other products in either of the subdomains (Aldaz, 2005).

Although hth and exd form a single functional unit, their mode of regulation is different: exd is expressed ubiquitously but is regulated at the subcellular level by hth, which promotes Exd nuclear transport. Therefore, the key element of hth/exd regulation is the transcriptional control of hth (Aldaz, 2005).

Originally, hth is expressed in all the notum cells and later becomes restricted to the alpha and ß subdomains. Consequently, the principal aspect of hth regulation is the mechanism(s) leading to its repression in the regions outside the alpha and ß subdomains. Two negative regulators have been identified, the eyg gene and the Dpp pathway, which probably acts through some unidentified downstream gene. In the notum hth behaves as a downstream target of both the Dpp pathway and eyg (Aldaz, 2005).

The role of Eyg as a negative regulator of hth is based on the following observations: (1) the beginning of the modulation of hth expression in the notum at the early third instar coincides with the initiation of eyg expression; (2) in eyg mutants the hth domain is expanded, extending to most of the notum; (3) mutant eyg clones show hth derepression in the inter-subdomains region, and conversely, ectopic eyg activity in the ß subdomain represses hth. The fact that this ectopic activity fails to affect hth in the alpha subdomain was expected since eyg and hth are normally co-expressed in this subdomain. In conclusion, eyg suppresses hth in the inter-subdomains region and also acts as a barrier for hth in the eyg/ß-hth border (Aldaz, 2005).

The role of the Dpp pathway as a negative regulator of hth is based on results showing that Mad - mutant clones in the inter-subdomains region show activation of hth. This is in contrast to the behaviour of those clones in the alpha subdomain, where they have no effect, or in the ß subdomain, where they show suppression of hth. It is believed that the reason for the latter effect is that eyg is up regulated in those clones, and in turn Eyg suppresses hth. The lack of effect of Mad - clones in the alpha subdomain is probably due to the low activity of Dpp in that region. In principle, the observation that the high activity levels generated in the TkvQD clones suppress hth in this subdomain supports this view. Expectedly, TkvQD clones do not affect hth expression in the ß subdomain, because it normally possesses high Dpp activity levels (Aldaz, 2005).

Taking all the results together, the following model of hth regulation is proposed. Since hth is originally expressed in all trunk embryonic cells and in all the notum cells in the early disc, the regulation of hth during wing disc development essentially reflects local repression in specific parts of the disc. The basic idea is that hth is repressed by the joint contribution of eyg and high/moderate levels of the Dpp pathway. Neither of these elements can repress hth individually. Although eyg appears to act uniformly in its domain, the repressing activity of Dpp is concentration dependent. Within the eyg domain, the hth alpha subdomain is located in the anterior region, in which the Dpp levels are too low to be effective and Eyg alone cannot repress hth/exd. In the inter-subdomains region the Dpp levels are high enough to repress hth, since here it acts together with Eyg. The ß subdomain is outside the eyg domain and therefore in the absence of Eyg even the high Dpp levels are not capable of repressing hth/exd. The model is also supported by the experiments of overexpressing eyg. The eyg-expressing clones in the ß subdomain suppress hth because the two repressors are active in the clones, while they have no effect in the alpha subdomain because it normally contains high eyg levels. In principle the experiments overexpressing the Dpp pathway (TkvQD clones) appear to support the model. These clones have no effect in the ß subdomain, which normally possesses high Dpp activity levels, but they suppress hth in the alpha subdomain. However, these clones are known to suppress eyg and therefore hth should not be repressed according to this model. It is possible that in certain circumstances the very high Dpp activity levels induced by these clones may be sufficient to down regulate hth, even in the absence of eyg (Aldaz, 2005).

The presence of two distinct repressors may suggest that the hth promoter region contains binding sites for Eyg and for Mad/Medea that would be responsible for the transcriptional repression. The ubiquitous expression in the absence of these two repressors may be due to a constitutive promoter (Aldaz, 2005).

The second aspect of the late patterning function of hth/exd arises from the observation that the alpha and ß subdomains form different patterns with similar levels of hth. This suggests the existence of interactions between Hth/Exd and products specifically localised to the different subdomains. In the case of the alpha subdomain, the obvious candidate for the interaction is Eyg. The joint activity of hth/exd and eyg specifies a notum pattern that is different from those specified by each of these genes alone (Aldaz, 2005).

The finding that the Eyg and Hth proteins associate to form a complex in vitro suggests a mechanism to achieve the pattern difference between the alpha and the ß subdomains. As it has been shown to be the case for the in vivo specificity of the Hox genes, the association of Hth/Exd with the different Hox products results in higher affinity and specificity for target sites. Here, the formation of an Eyg/Hth/Exd complex in the alpha subdomain may result in a constellation of gene activity different from that in the ß subdomain where Eyg is not present. In the latter subdomain hth/exd may act alone, for after all the two genes encode transcription factors. Alternatively, the Hth/Exd products may interact with some other yet unidentified co-factor (Aldaz, 2005).

An interesting aspect of the interaction of hth/exd and eyg is that it acts in two different ways. At the gene regulation level, eyg participates in the spatial control of hth/exd activity, but where the two genes are co-expressed their proteins interact, presumably to contribute to the in vivo affinity and specificity for target genes (Aldaz, 2005).

Functional dissection of the splice variants of the Drosophila gene homothorax (hth)

TALE-homeodomain family member Homothorax interacts with a second TALE-homeodomain protein Extradenticle to facilitate Exd entrance to the nucleus. The many different functions described for Hth rely on the complexity of the locus, from which six different isoforms arise. The isoforms can be grouped into full-length and short versions, which contain either one or the two conserved domains of the protein (homeodomain and Exd-interacting domain). This study used molecular and genetic tools to analyze the levels of expression, the distribution and the function of the isoforms during embryonic development. The results clearly show that the isoforms display distinct levels of expression and are differentially distributed in the embryo. This detailed study also shows that during normal embryonic development not all the Hth isoforms translocate Exd into the nucleus, suggesting that both the proteins can also function separately. The full-length Hth protein activates transcription of exd, augmenting the levels of exd mRNA in the cell. The higher levels of Exd protein in those cells facilitate its entrance to the nucleus. This work demonstrates that hth is a complex gene that should not be considered as a functional unit. The roles of the different isoforms probably rely on their distinct protein domains and conformations and, at the end, on interactions with particular partners (Corsetti, 2013).

Targets of Activity

The Extradenticle protein is ubiquitously expressed in early embryonic cells during the period when segmental identities are being determined. exd interacts genetically with partial loss-of-function Dfd mutant alleles, enhancing their lethality. Thus, exd function is required for the autoactivation phase of Deformed expression in the posterior head: Exd partners Dfd in Dfd-directed Dfd autoactivation. An epidermal autoregulatory enhancer (EAE) maps in a 2.7 kb fragment about 4 kb upstream of the Dfd start site. Mutations that change the affinity of the small autoactivation element (120 base pairs) for Exd protein result in corresponding changes in the element's embryonic activity. The Exd and Dfd proteins directly activate this element in maxillary cells without cooperative binding. Based on the types of homeotic transformations and changes in gene expression observed in exd mutant embryos, a new model for Exd/PBX action is proposed in which these proteins are required for HOX protein transcriptional activation functions, but dispensable for HOX transcriptional repression functions. Although the selection of a specific target gene activation site by one HOX protein versus another may be explained in some cases by Exd's selective modulation of HOX binding specificity, the favored idea in this instance is that Exd can interact with HOX proteins to switch them into a state where they are capable of transcriptional activation. The experiments provide no support for the idea that a specialized heterodimer-binding site that would bind a Dfd-Exd complex stably and tightly is required for the specific activation of the 120 bp element in embryos. In this case Exd functions as a coactivator and does not function to promote cooperative binding with Dfd. Binding of Exd without simultaneous binding of Dfd may mediate repression. At adjacent regions of the 120 bp element, a novel protein, DEAF-1 has been shown to bind. Mutant substitutions in these adjacent regions reduce their affinity for DEAF-1. It is possible the Exd and DEAF-1 collaborate in non-maxillary cells to prevent ectopic activation of this head-specific HOX element (Pinsonneault, 1997).

Two BX-C genes, Ultrabithorax and abdominal-A, require exd activity for their maintenance and function. Using an antibody directed against the Ubx protein, Ubx expression was examined in haltere imaginal discs containing exd clones induced at different times during larval development. In exd clones induced at 48-60 hr after egg laying (AEL), Ubx expression is abolished or greatly reduced in the region of the disc that will develop into metanotum. In exd clones induced later (72-84 hr AEL), Ubx expression is variable; some cells retain the Ubx antigen, whereas it is undetectable in others. As expected, exd clones in the region of the disc giving rise to the haltere pouch, where Exd is cytoplamic, have no effect on Ubx expression. This observation indicates that exd is involved in the maintenance of Ubx activity in the trunk region of the segment (metanotum), but not in the appendage. It is proposed that mutual interactions between Exd and BX-C proteins ensure the correct amounts of interacting molecules. Since the Hoxd10 gene has the same properties as Drosophila BX-C genes, it is suggested that the control mechanism of subcellular distribution of Exd found in Drosophila probably operates in other organisms as well (Azpiazu, 1998).

Homeotic gene transcription is not affected by EXD. EXD positively regulates wingless and teashirt, both target genes of homeotic genes. EXD is required for the correct expression of dpp in visceral mesoderm. EXD and UBX regulate dpp via the same visceral mesoderm enhancer. Thus EXD coordinates regulation of downsteam genes in cooperation with homeotic selector proteins (Rauskolb 1993).

In embryos lacking extradenticle, expression patterns for wingless, teashirt and decapentaplegic are altered in the embryonic visceral mesodermal midgut. EXD acts with UBX to activate dpp expression in parasegment 7 (PS7), via a minimal visceral mesoderm enhancer. Anterior to PS7, EXD represses dpp expression. Even when Ubx is ubiquitously expressed at high levels in exd embryos, UBX is incapable of activating dpp enhancer expression (Rauskolb, 1994). Thus EXD repression of dpp is independent of homeobox cluster genes.

Transcription of dpp in the Drosophila midgut is activated by UBX protein and repressed by ABD-A protein. A candidate cofactor, the EXD protein, binds to the dpp enhancer in close proximity to homeotic protein binding sites. Mutation at either this site or another conserved motif compromises enhancer function (Manek, 1994). Thus EXD is a cofactor both in the repression and activation of dpp.

The Drosophila Engrailed homeoprotein has been shown to directly activate a Polycomb-group gene, polyhomeotic, during embryogenesis. A study of the molecular mechanism involved in this activation detected two different types of Engrailed-binding fragments within the polyhomeotic locus. The P1 and D1 fragments contain several 'TTAATTGCAT' motifs, whereas the D2 fragment contains a long 'TAAT' stretch to which multiple copies of Engrailed bind cooperatively. Another homeodomain-containing protein, Extradenticle, establishes protein-protein interactions with Engrailed on the D2 fragment. Both types of Engrailed-binding sites (P1 or D1 and D2), as well as Extradenticle, are necessary to obtain activation by Engrailed. In vivo, normal polyhomeotic expression depends on extradenticle expression. Moreover, in the absence of Extradenticle, overexpression of Engrailed protein represses polyhomeotic expression (Serrano, 1998).

On the basis of these results, a model is proposed to explain ph regulation by En. In the absence of Exd, En protein is probably bound to P1 and D1, while D2 might be bound either by En or by other homeodomain-containing proteins. Under such conditions, no activation of ph is observed. This situation might represent what happens in the embryo prior to full germband extension. When Exd is present, activation of ph occurs. Although Exd is expressed prior to the time at which En activates ph, Exd might not be active at those earlier stages: Exd activity has been shown to be regulated by nuclear import and was described as exclusively cytoplasmic in the embryo before germband extension. Because the En-binding sites are so far apart within the ph locus and because D2 is required for activation, acting as an enhancer on both ph transcription units, it is further suggested that DNA bending might be involved in activation. In particular, En binding to P1 and D1 could be involved in bringing the proximal and distal ph promoters close to D2 in the presence of Exd, allowing ph activation from both transcription units. En binding to P1 and D1 prior to germband extension might thus prepare the chromatin for rapid activation. At least two lines of evidence support the idea that D2 interacts with P1 as well as D1, to mediate activation of both the proximal and the distal transcription units: (1) the P1-D2 and D1-D2 fusions behave similarly in CAT assays and (2) the in vivo activation was observed with the phlac enhancer trap, which corresponds to an insertion in the proximal transcription unit of ph (Serrano, 1998).

Homothorax is shown to limit Dpp and Wg expression in the leg disc. Expression of the Dpp and Wg targets omb and H15 is restricted to those cells that do not express Hth. To determine if hth inhibits target gene activation by Dpp and Wg, hth was either removed from its endogenous domain or either a GFP-Hth fusion protein or the murine hth homolog MEIS-1B was misexpressed in the distal portion of the leg disc. Removing hth function results in the expansion of wg and dpp target gene expression. Dorsally situated hth- clones result in the expansion of omb expression, as marked by the omb-lacZ reporter gene. Does hth repress Distal-less and dachshund? Similar to removing exd function, when hth loss-of-function clones were examined, dac was found to be only partially derepressed, and derepression was found to be more likely to occur in clones that arise near endogenous dac expression. hth- clones have no effect on Dll expression, regardless of where they are situated. However, when clones of GFP-Hth- or MYC-MEIS-expressing cells are generated, both Dll and dac can be repressed. These results suggest that the expression of Dll and dac requires two conditions: (1) the absence of Hth and (2) sufficient activity in the Dpp and Wg pathways. High levels of Wg and Dpp signaling are shown to repress the nuclear localization of Exd by repressing hth transcription. The direct action of both the Wg- and Dpp-signaling pathways is required to specify cell fates along the P/D axis. High levels of Wg and Dpp signaling are required to activate Dll, a determinant of distal cell fates, and to repress expression of dac, a determinant of intermediate fates along the P/D axis. At intermediate levels of Wg and Dpp signaling, dac, but not Dll, is activated. The distal edge of hth expression coincides with the proximal edge of dac expression, suggesting that the threshold of Dpp and Wg signaling required to activate dac is similar to that required to repress hth. To test this idea, either Wg or Dpp signaling was elevated in the hth expression domain by generating clones of cells that express either a membrane-tethered form of Wg or an activated Dpp receptor, Thickveins QD (TKV QD). When Wg-expressing clones were generated dorsally, where endogenous Wg levels are low but where Dpp is present at high concentrations, there was a loss of Hth protein and a shift of Exd protein to the cytoplasm. This suggests that sufficient levels of both Wg and Dpp signaling are required to repress Hth (Abu-Shaar, 1998).

High levels of Wg and Dpp signaling are shown to affect Hth and Exd, at least in part, by repressing hth transcription. The ability of Wg and Dpp to repress hth appears to be indirectly mediated by Dll and dac. Like Dll, Dac appears to have the capacity to repress hth. The expression patterns of Hth and Dll were examined in leg discs that were entirely devoid of dac function. In these discs, the hth domain appears expanded distally and the Dll domain appears to be expanded proximally, consistent with the idea that dac normally represses both hth and Dll (Abu-Shaar, 1998).

The domains of gene expression for Hth, Dac and Dll, as well the regulatory interactions between them, suggest that the leg is functionally divided into two major domains. The first is a proximal domain, which expresses hth, has nuclear Exd and does not express at least some of the potential target genes of the Wg- or Dpp-signaling pathways. The second is a distal domain, which does not express hth, has Exd localized to the cytoplasm, and expresses the targets of Wg, Dpp and Wg+Dpp signaling. These data suggest that the proximal domain is what has been referred to as the coxopodite, or an extension of the body wall, and is distinct from the distal domain, the telopodite. hth expression and nuclear Exd in the coxopodite would restrict the ability of the Wg and Dpp signals to activate their target genes. This idea is consistent with the observation that these two domains differ with respect to their requirement for Hh signaling: unlike the telopodite, which exhibits severe truncations upon the reduction of hh function, the coxopodite is less severely affected. These two domains also appear to have different cell surface properties; cells from one domain prefer not to mix with cells from the other domain. For example, Dll mutant clones almost always relocalize to the hth-expressing domain and hth mutant clones frequently sort into distal regions of the leg disc. This phenomenon is not observed in the wing disc, where hth and Dll are restricted to outside and within the wing pouch, respectively: hth or Dll mutant clones are positioned randomly in this tissue. The mutant phenotypes displayed by the loss of coxopodite gene function are qualitatively different from those displayed by the loss of telopodite gene function. Removal of coxopodite genes such as exd results in either ‘nonsense’ or proximal to distal cell fate transformations, whereas removal of telopodite gene functions such as Dll and dac results in deletions of the appendage. In summary, the data support the idea that the proximal and distal regions of the leg have independent origins and differ from each other primarily due to the expression of hth, which limits or alters the ability of proximal cells to respond to Wg and Dpp signaling (Abu-Shaar, 1998 and references).

dac and Dll are shown to mediate Wg and Dpp mediated repression of hth. The demonstration that Wg and Dpp signaling repressed hth transcription and Exd’s nuclear localization was surprising, because these two signaling molecules induce Exd’s nuclear localization in the endoderm of the embryonic midgut. An investigation was carried out into the possibility that the repression of hth by Wg and Dpp is indirect and perhaps mediated by dac and Dll, which are not expressed in the midgut. TKV QD-expressing clones were generated and Hth, Dll and Dac were examined. Loss of function clones of Dll and dac were generated. When Dll- clones were generated before ~72 hours of development, hth was found to be derepressed and Exd was nuclear. However, clones generated after ~72 hours have no effect on hth or Exd, suggesting that there is an alternative mechanism for maintaining hth repression. Like Dll, Dac appears to have the capacity to repress hth. The ability of Dac to repress hth expression was confirmed by generating dac- clones. These dac- clones suggest that there might be other regulators of hth in addition to dac and Dll. Completely removing dac function results in viable animals that have deletions along the P/D axes of their legs. The expression patterns of Hth and Dll were examined in leg discs that were entirely devoid of dac function. In these discs, the hth domain appears expanded distally and the Dll domain appears to be expanded proximally, consistent with the idea that dac normally represses both hth and Dll. It is an apparent paradox that Wg and Dpp repress hth in the leg disc while these same signals activate hth expression and nuclear Exd in the midgut endoderm. This may be explained because in the leg, Wg and Dpp repress hth indirectly, by activating the hth repressors Dll and dac. In the absence of Dll or dac, hth is derepressed in the leg disc, even in cells that receive high levels of the Wg and Dpp signals. In contrast, in the embryonic endoderm, dac and Dll are not activated by Wg and Dpp, nor are any other known hth repressors, allowing hth to be activated in these cells (Abu-Shaar, 1998).

The 26 bp bx1 element from the regulatory region of Distal-less is capable of imposing control by the homeotic genes Ultrabithorax and abdominal-A on a general epidermal activator in Drosophila. This provides an assay to analyze the sequence requirements for specific repression by these Hox genes. Both the core Hox binding site, 5'-TAAT, and the adjacent Exd 5'-TGAT core site are required for repression by Ultrabithorax and abdominal-A. The Distal-less bx1 site thus fits with the model of Hox protein binding specificity based on the consensus PBX/HOX-family site TGATNNAT[g/t][g/a], where the key elements of binding specificity are proposed to lie in the two base pairs following the TGAT. A single base pair deletion in the bx1 sequence generates a site, bx1:A-mut, which on the consensus PBX/HOX model would be expected to be regulated by the Deformed Hox gene. It has been observed, however, that the bx1:A-mut site is regulated predominantly by Sex combs reduced, Ultrabithorax and abdominal-A. The analysis of this site indicates that the specificity of action of Hox proteins may depend not only on selective DNA binding but also on specific post-binding interactions (White, 2000).

A homeotic response element that mediates repression by the Hox genes Ubx and abd-A has been built. The element is based on two short sequence modules; the 21 bp binding site Grainyhead binding site element (Gbe) for the transcription factor GRH and the 26 bp UBX/ ABD-A bx1 footprint site from the Dll regulatory sequences. On its own the Gbe mediates uniform epidermal activation. However, the combination of the Gbe and the bx1 element produces a homeotically modulated response. Specifically in the domain of expression of the Hox genes Ubx and abd-A the epidermal expression is repressed. This homeotic element thus successfully recapitulates the features of endogenous regulatory elements from homeotic target genes giving tissue-specific regulation by multiple homeotic genes (e.g. connectin). However, while endogenous target gene regulatory elements tend to be large (typically covering several kb) and are correspondingly difficult to analyze, this constructed element is appealingly simple and provides a manageable system for analyzing the specificity of the homeotic response (White, 2000).

The basis of Hox target specificity has been proposed to depend on the nucleotides immediately following the Exd core site in the 5' -TGATNNAT[g/t][g/a] PBX/HOX consensus target site. In vivo studies have indicated that NN=GG produces a lab-specific response whereas NN=TA elicits a response to Dfd. Previous in vitro studies on the optimum target sequences for the whole range of PBX/HOX heterodimers have emphasized the importance of the nucleotide at position 7 in the PBX/HOX consensus sequence 5' - ATGATTNATGG. Preference for a particular nucleotide at N7 varies across the Hox complex. The most anteriorly-expressed proteins, HOXB1 and 2, prefer G at this position, the next more posteriorly-expressed Hox proteins allow N7=A, and from HOXB6-10 preference grows for N7=T. The optimum consensus for binding site for the PBX/HOXB7 heterodimer, equivalent to Ubx or Abd-A/Exd in Drosophila, is 5' -ATGATTTATGG. This is similar to the sequence ATGATTTAatt (differences in lower case) in the bx1 element. In particular, the nucleotide at N7 is T and thus fits with the predictions from the in vitro binding data. Taking the two elements previously studied in vivo into account, N7=G is in a lab (Hoxb1) element; N7=A in a Dfd (Hoxb4) element, and now N7=T is present in the bx1 element, which responds to Ubx and abd-A. Thus, the specificity of regulation of the bx1 element fits well with the expectations of the PBX/HOX model where the nucleotides immediately 3' of the TGAT PBX core provide the specificity for a particular subset of Hox proteins through interactions with the specific residues of the N-terminal arm of the homeodomain (White, 2000).

The bx1 element on its own does not drive reporter gene expression. It acts as a control module that together with the Grh activation module constitutes a homeotic response element. This may represent a general model for a target gene regulatory element, where proteins bound to a PBX/ HOX site interact with neighbouring regulatory complexes to modulate their activity. The PBX/HOX modules may act to overlay homeotic control on a whole variety of tissue-specific, signal transduction pathway-specific, temporal or other regulatory elements. It is worth noting that the regulatory regions that have been identified in homeotic target genes all drive highly tissue-restricted patterns, indicating the collaboration between tissue-specific and homeotic control. These observations give an insight into the modular construction of complex enhancers and also have important implications for the specificity of homeotic gene function. If the PBX/HOX complexes in target gene regulatory sequences act by interacting with other regulatory complexes, it seems very likely that different Hox proteins will vary in their ability to mediate specific interactions. Thus, there is clearly the potential for much of the specificity of responses to homeotic gene control to reside in post-binding interactions (White, 2000).

The cap'n'collar locus encodes three transcript and protein isoforms. The cncB transcript is expressed in an embryonic pattern that includes the labral, intercalary and mandibular segments, while cncA and cncC are expressed ubiquitously. CncB suppresses the segmental identity function of the Hox gene Deformed (Dfd) in the mandibular segment of Drosophila embryos. Evidence has been provided that the CncB-mediated suppression of Dfd requires the Drosophila homolog of the mammalian small Maf proteins, Maf-S, and that the suppression occurs even in the presence of high amounts of Dfd protein. Interestingly, the CncB/Maf-S suppressive effect can be partially reversed by overexpression of Homothorax (Hth), suggesting that Hth and Extradenticle proteins antagonize the effects of CncB/Maf-S on Dfd function in the mandibular segment (Veraksa, 2000).

Specificity of Distalless repression and limb primordia development by Abdominal Hox proteins

In Drosophila, differences between segments, such as the presence or absence of appendages, are controlled by Hox transcription factors. The Hox protein Ultrabithorax (Ubx) suppresses limb formation in the abdomen by repressing the leg selector gene Distalless, whereas Antennapedia (Antp), a thoracic Hox protein, does not repress Distalless. The Hox cofactors Extradenticle and Homothorax selectively enhance Ubx, but not Antp, binding to a Distalless regulatory sequence. A C-terminal peptide in Ubx stimulates binding to this site. However, DNA binding is not sufficient for Distalless repression. Instead, an additional alternatively spliced domain in Ubx is required for Distalless repression but not DNA binding. Thus, the functional specificities of Hox proteins depend on both DNA binding-dependent and -independent mechanisms (Gebelein, 2002).

This work begins with a characterization of a Ubx binding site in the Dll gene that is critical for Dll repression; both Exd and Hth play a role in Ubx binding and repression. The Dll304 enhancer is sufficient to recapitulate the early expression pattern of Dll in the embryonic leg primordia. In addition to activation functions, Dll304 contains two Hox binding sites, Bx1 and Bx2, that repress enhancer activity in the abdomen and thereby restrict Dll expression to the thorax. Most of the repression activity is conferred by Bx1, a sequence bound by Ubx and Abd-A. In agreement with this result, a Distalless minimal element (DME) that lacks the Bx2 site accurately recapitulates the expression of Dll304 in the embryonic thorax. The DME enhancer also shows no derepression within the abdomen, suggesting that Bx1 is sufficient to fully repress Dll (Gebelein, 2002).

To better understand how Bx1 represses Dll, the presence of Exd and Hth binding sites were sought near the previously characterized Hox binding site. A consensus Exd site and a near consensus Hth site are in close proximity to the Hox site of Bx1. The Hox/Exd site (5'-AAATTAAATCA-3'), however, is unlike other previously characterized Hox/Exd binding sites because it contains an additional base pair in between the Hox and Exd half-sites. The Bx1 region containing this Hox/Exd/Hth site is referred to as the Distalless repression element (DllR). To determine whether DllR is required to repress DME expression in the abdomen, it was deleted from the DME enhancer (DMEact) and its ability to activate a reporter gene was tested in vivo. DMEact drives gene expression in all abdominal segments as well as in the thoracic region. Because the thoracic expression driven by DMEact is similar to that of DME, the DllR region is not required for DME activation but solely functions in the repression of Dll in the abdomen (Gebelein, 2002).

To determine whether Exd and Hth stimulate Hox binding to DllR, electrophoretic mobility shift assays (EMSAs) were performed with purified Ubx, Exd, and Hth proteins. Unless stated otherwise, all of these experiments were performed with UbxIa, the most widely expressed of several Ubx isoforms. By themselves, Ubx or an Exd/Hth heterodimer are capable of weakly interacting with DllR. The combination of all three proteins results in a slower migrating band indicating the formation of a Ubx/Exd/Hth/DNA complex. The formation of this protein/DNA complex is highly cooperative when compared to the amount of binding observed with Ubx or Exd/Hth alone. To test the contribution of each binding site, point mutations were introduced within the individual Hox, Exd, and Hth sites. Mutation of any one of these sites results in a decrease in the formation of the trimeric protein/DNA complex, suggesting that all three are required for optimal binding to DllR (Gebelein, 2002).

To test whether the Hox, Exd, and Hth binding sites are also required for Dll repression in vivo, reporter constructs were created containing the lacZ gene under the control of mutant versions of the DME enhancer. Mutation of the Hox site (DMEHox) results in a similar level of derepression of reporter gene expression throughout the abdomen, as does the the complete deletion of DllR. Mutation of the Exd (DMEExd) and Hth (DMEHth) sites individually also results in derepression, albeit slightly weaker than mutation of the Hox site. However, if both the Exd and Hth sites are mutated together, full derepression is observed. Taken together, these results demonstrate that the efficient formation of a Hox/Exd/Hth trimeric complex on DllR is required for Dll repression within the abdomen (Gebelein, 2002).

These above data support a model in which a Ubx/Exd/Hth complex bound to DllR is necessary for Dll repression. Whether a single copy of DllR is sufficient to repress a heterologous enhancer element was tested. An artificial enhancer, called fkh(250con), is activated by Scr, Antp, and Ubx (with Exd and Hth), and thus provides a useful heterologous activator to test for DllR function. A reporter construct under the control of both fkh(250con) and DllR was created. Unlike fkh(250con), which is expressed in parasegments (PS) 2-6, the composite enhancer (fkh250con-DllR) is not expressed in PS 6, where Ubx is expressed. Ubx-mediated repression of fkh(250con)-DllR is more obvious in embryos mutant for abd-A, which derepress Ubx and, consequently, fkh(250con) throughout the abdomen. In this genetic background, fkh(250con)-DllR is still active only in PS 2-5. Furthermore, misexpression of Ubx throughout the embryo activates fkh(250con) but represses fkh(250con)-DllR. Taken together, these results indicate that DllR is sufficient to confer Ubx-mediated repression of a heterologous enhancer. In addition, these results also illustrate that Ubx/Exd/Hth complexes can mediate repression through DllR in the same cells as it mediates activation through fkh(250con) (Gebelein, 2002).

A general question for all transcription factors is how they achieve specificity in vivo. For the Hox proteins, a large number of studies have implicated sequences both within and outside the homeodomain as being important for their in vivo specificities. But how do these sequences function? Because DNA binding domains, including homeodomains, can also be protein interaction domains, studies that map the domains necessary for target gene regulation cannot answer this question by themselves. Instead, direct transcriptional targets must be identified and, once binding sites are characterized, DNA binding, in addition to target gene regulation, must be measured. The results allow two steps to be discriminated in the repression of Dll by Ubx. First, Exd and Hth stimulate Ubx, but not Antp, binding to DllR. In contrast, Ubx/Exd/Hth and Antp/Exd/Hth have similar affinities for a different "consensus" binding site (5'-CCATAAATCA-3'), suggesting that subtle differences in the DNA sequence, in addition to differences between Ubx and Antp, contribute to specificity. A C-terminal peptide in Ubx stimulates this cofactor-dependent binding to DllR. DNA binding, however, is not sufficient for Dll repression. Instead, an additional linker domain included in only a subset of Ubx isoforms is required for repression. Thus, a second step, the recruitment of additional factors to the Ubx/Exd/Hth complex bound to DllR, is implied by these data. In addition to the UbxIa linker, this step also requires the specific sequences and conformation imposed on the Ubx/Exd/Hth trimer by DllR (Gebelein, 2002).

Although the Ubx C terminus plays an important role in cofactor-dependent binding to DllR, additional domains contribute to optimal binding. In the presence of Exd and Hth, the AAUU chimera, but not heterologous AAUA or AAAU, binds DllR, suggesting that both the Ubx homeodomain and C terminus are important for optimal DNA binding to this site. The C terminus is not absolutely required for binding because a Ubx protein that lacks this domain (UUU*) is still able to bind well to DllR. Last, the finding that UUU*, but not AAU*, binds DllR suggests that a domain N terminal to the homeodomain also enhances DllR binding. Based on the crystal structures of Hox/Exd/DNA complexes, this difference could be due to the YPWM motif. Taken together, the data suggest that multiple regions of Ubx contribute to binding DllR and that no one domain is sufficient for full binding activity. This finding may be understood in light of the fact that the entire Ubx coding sequence has been constrained over millions of years of insect evolution to maintain leg (and Dll) repression in the abdomen (Gebelein, 2002).

How might the Ubx C terminus and YPWM motifs contribute to DNA binding? It is suggested that these regions could make additional protein-DNA contacts and/or protein-protein interactions that help stabilize the DllR-bound form of the trimeric complex. In support of this idea, the C termini of other homeodomain proteins also contribute to DNA binding. The Exd C terminus, for example, consists of an alpha helix that packs against its homeodomain and contributes to DNA binding. The C terminus of the MATalpha2 protein from yeast forms an alpha helix that contacts the MATa1 homeodomain to stabilize heterodimer formation on DNA. Interestingly, the two Hox proteins that repress Dll expression, Ubx and Abd-A, share sequence homology in their C termini, and are the only Drosophila Hox proteins predicted to form an alpha helix after their homeodomains (Gebelein, 2002).

The Ubx YPWM motif may also help stabilize complex formation on DllR. In the Hox/Exd/DNA crystal structures, this motif, together with flanking amino acids, directly contacts a hydrophobic pocket within the Exd homeodomain. These protein-protein contacts are thought to stabilize protein-DNA contacts made by the complex. The amino acids surrounding the YPWM motifs are different in Ubx and Antp and thus could contribute to DNA binding specificity by such an indirect mechanism (Gebelein, 2002).

The finding that UbxIa, but not UbxIVa, is able to repress Dll suggests that the linker region in UbxIa is required for repression. In addition, these results suggest that alternative splicing has the potential to modulate Ubx's control of gene expression. In support of this view, the expression of Ubx isoforms is temporally and spatially regulated. In addition, misexpression experiments using UbxIa and UbxIVa have shown that while both perform many of the same functions, only UbxIa efficiently transforms the peripheral nervous system. The finding that UbxIa and UbxIVa have different transcriptional regulatory properties provides a possible explanation for their distinct abilities to transform this tissue (Gebelein, 2002).

One argument against the idea that the different Ubx isoforms have distinct functions is that flies containing a genetic inversion that prevents the inclusion of the second microexon are, for the most part, normal. Although this mutation prevents the expression of UbxIa, it is unclear which other Ubx isoforms are expressed in this mutant because the inversion does not include both microexons. Furthermore, the effect that this mutation has on Dll expression has not been examined. A definitive test of the idea that Ubx isoforms have unique functions will require determining whether a Ubx allele in which both microexons are eliminated can provide all Ubx functions in vivo (Gebelein, 2002).

As in Drosophila, Dll expression is a marker for leg primordia in many animal phyla. Animals with appendages on their abdominal segments, such as crustaceans and onychophora, coexpress Ubx with Dll, demonstrating that Ubx is not a repressor of Dll in these species. The ability of Ubx to repress Dll probably arose in a subset of arthropods, the hexapods. Consistent with these findings, two recent studies suggest that one relevant difference between Ubx orthologs that repress Dll (for example, Drosophila Ubx) and Ubx orthologs that do not repress Dll (for example, onychophoran Ubx) maps to the C-terminal regions of these Hox proteins (Galant, 2002; Ronshaugen, 2002). These two groups, however, propose different mechanisms for how these sequences function. Galant suggests that the Drosophila Ubx C terminus actively represses transcription via a polyalanine motif that is present in the Ubx orthologs from all hexapods. Ronshaugen suggests that the Drosophila Ubx C terminus is only permissive for repression. Instead, they argue that crustaceans, which have abdominal legs, evolved a C-terminal sequence that inhibits Dll repression. However, neither study analyzed the binding of these proteins to the relevant binding sites in Dll, leaving open the possibility that the effects they observe could also be due to effects on DNA binding (Gebelein, 2002).

The data provide additional insights into how repression mechanisms may have evolved in these different species. It was found that the Drosophila Ubx C terminus contributes to DllR binding but is not sufficient for Dll repression in vivo. Thus, the positive role, observed by Galant, that the Drosophila sequence plays in Dll repression, could be due to an effect on DNA binding. These experiments also implicate the linker region of UbxIa as important for repression, but not DNA binding. Because some of the onychophora/ Drosophila and crustacean/ Drosophila chimeras lack this linker but are able to repress Dll, the crustacean and onychophoran Ubx orthologs must have repression domains that are different from the one identified in Drosophila Ubx (Gebelein, 2002).

Ronshaugen suggests that the phosphorylation of serine and threonine residues in the crustacean Ubx C terminus is necessary for it to prevent Dll repression (Ronshaugen, 2002). This is an intriguing possibility in light of the fact that phosphorylation of a Hox C terminus can inhibit cooperative DNA binding with Exd. Taken together with the current data that the C terminus of Ubx enhances DNA binding to DllR, it is suggested that the inhibition of Dll repression by the crustacean C terminus may be due to a reduced ability to bind DllR with Exd and Hth. This model accounts for why a Drosophila UbxIa protein containing the crustacean C terminus is unable to repress Dll (Ronshaugen, 2002) and for the inability of onychophora Ubx, which also contains a putative phosphorylation site in its C terminus, to repress Dll. Taken together, it is suggested that the evolution of limb suppression by Hox proteins, and probably many other Hox functions, depended upon the modification of both DNA binding-dependent and -independent mechanisms controlling Hox specificity (Gebelein, 2002).

Although these experiments focused on understanding why Antp is different from Ubx, the results provide some insights into the mechanism of transcriptional repression. The data strongly argue that a DNA-bound Ubx/Exd/Hth complex is necessary, but not sufficient, for repression. First, in addition to repressing Dll, Ubx/Exd/Hth activates fkh(250con). When both fkh(250con) and DllR simultaneously regulate the same reporter gene, DllR is able to repress gene expression in the same cells in which fkh(250con) normally activates gene expression. This result suggests that the repressor proteins required for DllR activity are not cell type specific and are widely expressed in the embryo. Further, these results suggest that differences between the fkh(250con) and DllR sequences determine whether transcription is activated or repressed. These sequences may recruit additional DNA binding factors that interact with the trimeric complex. These factors, which have not yet been identified, might provide or reveal a latent activation or repression domain within the Hox/Exd/Hth complex. Alternatively, another DNA binding factor may not be needed. Instead, the unique arrangement or spacing of the Hox, Exd, and Hth sites in these two elements may result in distinct conformations of the trimeric complex that recruit different coactivators or corepressors. Such a mechanism has been suggested for the nuclear receptor family of transcription factors and for the POU domain protein Pit-1, where a difference in spacing in a Pit-1 dimer binding site regulates the recruitment of a corepressor. Consistent with such a mechanism, it was found that the DllRcon binding site, which has one less base pair between the Hox and Exd half-sites than the DllR binding site, fails to repress transcription despite having a higher affinity for Ubx/Exd/Hth complexes. In addition, although repression activity for the UbxIa linker and C terminus in S2 cells can be measured, the experiments suggest that their activities are context dependent. The abdominal expression of DMEcon-lacZ suggests that the mere presence of these domains is not sufficient for repression. Thus, the data suggest that transcription factor domains have distinct properties when assayed by themselves versus when they are part of a multiprotein complex. Further, it is concluded that the unique architecture of the complex assembled on DllR is necessary for efficient repression (Gebelein, 2002).

During Drosophila embryogenesis, segments, each with an anterior and posterior compartment, are generated by the segmentation genes while the Hox genes provide each segment with a unique identity. These two processes have been thought to occur independently. This study shows tha abdominal Hox proteins work directly with two different segmentation proteins, Sloppy paired and Engrailed, to repress the Hox target gene Distalless in anterior and posterior compartments, respectively. These results suggest that segmentation proteins can function as Hox cofactors and reveal a previously unanticipated use of compartments for gene regulation by Hox proteins. The results suggest that these two classes of proteins may collaborate to directly control gene expression at many downstream target genes (Gebelein, 2004).

The segregation of groups of cells into compartments is fundamental to animal development. Originally defined in Drosophila, compartments are critical for providing cells with their unique positional address. The first compartments to form during Drosophila development are the anterior and posterior compartments and the key step to defining them is the activation of the gene engrailed (en). Expression of en, which encodes a homeodomain transcription factor, results in a posterior compartment fate, and the absence of en expression results in an anterior compartment fate. Once activated by gap and pair-rule genes, en expression and, consequently, the anterior-posterior compartment boundary later become dependent upon the protein Wingless (Wg), which is secreted from adjacent anterior compartment cells. Concurrently with anterior-posterior compartmentalization and segmentation, the expression of the eight Drosophila Hox genes is also initially established by the gap and pair-rule genes. The Hox genes, however, which also encode homeodomain transcription factors, do not contribute to the formation or number of segments but instead specify their unique identities along the anterior-posterior axis (Gebelein, 2004).

This flow of genetic information during Drosophila embryogenesis has led to the idea that anterior-posterior compartmentalization and segment identity specification are independent processes. In contrast to this view, this study shows that these two pathways are interconnected in previously unrecognized ways. Evidence is provided that Hox factors directly interact with segmentation proteins such as En to control gene expression. Moreover, Hox proteins collaborate with two different segmentation proteins in anterior and posterior cell types to regulate the same Hox target gene, revealing a previously unknown use of compartments to control gene expression by Hox proteins (Gebelein, 2004).

Distalless (Dll) is a Hox target gene that is required for leg development in Drosophila. In each thoracic hemisegment, wg, expressed by anterior cells adjacent to the anterior-posterior compartment boundary, activates Dll in a group of cells that straddle this boundary. A cis-regulatory element derived from Dll, called DMX, drives accurate Dll-like expression in the thorax. The abdominal Hox genes Ultrabithorax (Ubx) and abdominalA (abdA) directly repress Dll and DMX-lacZ in both compartments, thereby blocking leg development in the abdomen. DMX is composed of a large activator element (DMXact) and a 57-base-pair (bp) repressor element referred to here as DMX-R. Previous work demonstrated that Ubx and AbdA cooperatively bind to DMX-R with two homeodomain cofactors, Extradenticle (Exd) and Homothorax (Hth). In contrast, the thoracic Hox protein Antennapedia (Antp) does not repress Dll and does not bind DMX-R with high affinity in the presence or absence of Exd and Hth. Thus, repression of Dll in the abdomen depends in part on the ability of these cofactors to selectively enhance the binding of the abdominal Hox proteins to DMX-R (Gebelein, 2004).

Exd and Hth, as well as their vertebrate counterparts, are used as Hox cofactors at many target genes. Moreover, Hox/Exd/Hth complexes are used for both gene activation and repression, raising the question of how the decision to activate or repress is determined. One view posits that these complexes do not directly recruit co-activators or co-repressors, but instead are required for target gene selection. Accordingly, other DNA sequences present at Hox/Exd/Hth-targeted elements would determine whether a target gene is activated or repressed. Consistent with this notion, DMX-R sequences isolated from six Drosophila species show extensive conservation outside the previously identified Hox (referred to here as Hox1) Exd and Hth binding sites, suggesting that they also play a role in Dll regulation (Gebelein, 2004).

To test a role for these conserved sequences, a thorough mutagenesis of DMX-R was performed. Each mutant DMX-R was cloned into an otherwise wild-type, full-length DMX and tested for activity in a standard reporter gene assay in transgenic embryos. Thoracic expression was normal in all cases. However, surprisingly, many of the DMX-R mutations, such as X5, resulted in abdominal de-repression only in En-positive posterior compartment cells, whereas other mutations, such as X2, resulted in abdominal de-repression only in En-negative anterior compartment cells. Single mutations in the Hox1, Exd, or Hth sites also resulted in de-repression predominantly in posterior cells. In contrast, deletion of the entire DMX-R (DMXact-lacZ), or mutations in both the X2 and X5 sites (DMX[X2 + X5]-lacZ), resulted in de-repression in both compartments. These results suggest that distinct repression complexes bind to the DMX-R in the anterior and posterior compartments and that segmentation genes play a role in Dll repression (Gebelein, 2004).

One clue to the identity of the proteins in these repression complexes is that the sequence around the Hth site is nearly identical to a Hth/Hox binding site that had been identified previously by a systematic evolution of ligands by exponential enrichment (SELEX) approach using vertebrate Hox and Meis proteins. This similarity suggested the presence of a second, potentially redundant Hox binding site, Hox2. In agreement with this idea, mutations in both the Hox1 and Hox2 binding sites resulted in de-repression in both the anterior and posterior compartments of the abdominal segments. Similarly, although individual mutations in the Exd and Hth binding sites lead predominantly to de-repression in the posterior compartment, mutation of both sites resulted in de-repression in both compartments. These results suggest that a Hox/Exd/Hth/Hox complex may be used for repression in both compartments. Furthermore, they suggest that although single mutations in these binding sites are sufficient to disrupt the activity of this complex in the posterior compartment, double mutations are required to disrupt its activity in the anterior compartment (Gebelein, 2004).

To provide biochemical evidence for a Hox/Exd/Hth/Hox tetramer, DNA binding experiments were performed using DMX-R probes and proteins expressed and purified from E. coli. Previous experiments demonstrated that a Hox/Exd/Hth trimer cooperatively binds to the Hox1, Exd and Hth sites. The function of the Hox2 site was tested in two ways. First, binding was measured to a probe, DMX-R2, that includes the Exd, Hth and Hox2 sites, but not the Hox1 site. It was found that Exd/Hth/AbdA and Exd/Hth/Ubx trimers cooperatively bind to this probe and that mutations in the Hth, Exd or Hox2 binding sites reduced or eliminated complex formation (Gebelein, 2004).

Second, if both the Hox1 and Hox2 sites are functional, the full-length DMX-R may promote the assembly of Hox/Exd/Hth/Hox tetramers. Using a probe containing all four binding sites (DMX-R1 + 2), the formation of such complexes was observed. Mutation of any of the four binding sites reduced the amount of tetramer binding whereas mutation of both Hox sites or both the Exd and Hth sites eliminated tetramer binding. Furthermore, Antp, which does not repress Dll, formed tetramers with Exd and Hth that were approximately tenfold weaker than with Ubx or AbdA, but bound well to a consensus Hox/Exd/Hth trimer binding site. Because mutation of both Hox sites or both the Exd and Hth sites resulted in de-repression in both compartments, these experiments correlate the binding of a Hox/Exd/Hth/Hox complex on the DMX-R with the ability of this element to mediate repression in both compartments (Gebelein, 2004).

Although binding of a Hox/Exd/Hth/Hox tetramer is sufficient to account for the necessary abdominal Hox-input into Dll repression, it does not explain the compartment-specific de-repression exhibited by some DMX-R mutations. The X2 and X5 mutations, for example, result in abdominal de-repression but do not prevent the formation of the Hox/Exd/Hth/Hox tetramer. Sequence inspection of the DMX-R revealed that the X2 mutation, which resulted in de-repression specifically in the anterior compartment, disrupts two partially overlapping matches to a consensus binding site for Forkhead (Fkh) domain proteins. With this in mind, the expression pattern of Sloppy paired 1 (Slp1), a Fkh domain factor encoded by one of two partially redundant segmentation genes, slp1 and slp2, was examined. The two slp genes are expressed in anterior compartment cells adjacent and anterior to En-expressing posterior compartment cells. In the thorax, cells expressing Dll and DMX-lacZ co-express either Slp or En at the time Dll is initially expressed. In the abdomen, the homologous group of cells, which express DMXact-lacZ (a reporter lacking the DMX-R), co-express either Slp in the anterior compartment or En in the posterior compartment. The expression patterns of Slp and En were compared with Ubx and AbdA. Ubx levels are highest in anterior, Slp-expressing cells whereas AbdA levels are elevated in posterior, En-expressing cells. In contrast, both Exd and Hth are present at similar levels in both compartments throughout the abdomen (Gebelein, 2004).

On the basis of these data, a model is presented for Hox-mediated repression of Dll in both the anterior and posterior compartments of the abdominal segments. In the anterior compartment it is proposed that Slp binds to DMX-R directly with a Ubx/Exd/Hth/Ubx tetramer. In the posterior compartment it is suggested that En binds to DMX-R directly with an AbdA/Exd/Hth/AbdA tetramer. One important feature of this model is that Antp/Exd/Hth/Antp complexes fail to form on this DNA, thereby accounting for the lack of repression in the thorax. Furthermore, the model proposes that Slp and En should, on their own, have only weak affinity for DMX-R sequences because repression does not occur in the thorax, despite the presence of these factors. The Hox/Exd/Hth/Hox complex, perhaps in conjunction with additional factors, is required to recruit or stabilize Slp and En binding to the DMX-R. Both Slp and En are known repressor proteins that directly bind the co-repressor Groucho. Thus, the proposed complexes in both compartments provide a direct link to this co-repressor and, therefore, a mechanism for repression. DNA binding and genetic experiments are presented that test and support this model (Gebelein, 2004).

To test the idea that En is playing a direct role in Dll repression, the ability of En and Hox proteins to bind to DMX-R probes was examined. On its own, En binds to DMX-R very poorly. Surprisingly, it was found that En binds DMX-R with the abdominal Hox proteins Ubx or AbdA in a highly cooperative manner. The thoracic Hox protein Antp does not bind cooperatively with En to this probe. Mutations in the Hox1 or X5 binding sites block AbdA/En binding in vitro, consistent with these mutations showing posterior compartment de-repression in vivo. In contrast, the X6, X7 and Hth mutations do not affect AbdA/En complex formation (Gebelein, 2004).

On the basis of DMX-R's ability to assemble a Hox/Exd/Hth/Hox tetramer, whether En could bind together with an AbdA/Exd/Hth/AbdA complex was tested. Addition of En to reactions containing AbdA, Exd and Hth resulted in the formation of a putative En/AbdA/Exd/Hth/AbdA complex. This complex contains En because its formation is inhibited by an anti-En antibody. A weak antibody-induced supershift is also observed. Moreover, this complex fails to form on the X5 mutant, which causes posterior compartment-specific de-repression. It is noted that En/Exd/Hth complexes also bind to the DMX-R and that it cannot be excluded that an En/Exd/Hth/AbdA complex may be important for Dll repression. The model emphasizes a role for an En/AbdA/Exd/Hth/AbdA complex because it better accommodates the cooperative binding observed between En and AbdA on the DMX-R (Gebelein, 2004).

Repression in the anterior compartments of the abdominal segments requires the sequence defined by the X2 mutation, which is similar to a Fkh domain consensus binding site. The model predicts that this sequence is bound by Slp. Consistent with this view, Slp1 binds weakly to wild type, but not to X2 mutant DMX-R probes. However, in contrast to En, no cooperative binding was observed between Slp and Hox or Hox/Exd/Hth/Hox complexes, suggesting that additional factors may be required to mediate interactions between Slp and the abdominal Hox factors (Gebelein, 2004).

Together, these results suggest that En and Slp play a direct role in DMX-lacZ and Dll repression. However, these experiments do not unambiguously determine the stoichiometry of binding by these factors. Furthermore, in vivo, additional factors may enhance the interaction between these segmentation proteins and Hox complexes, thereby increasing the stability and/or activity of the repression complexes (Gebelein, 2004).

The model for Dll repression is supported by previous genetic experiments that examined the effect of Ubx and abdA mutants on Dll expression in the abdomen. Ubx abdA double mutants de-repress Dll in both compartments of all abdominal segments. In contrast, Ubx mutants de-repress Dll in the anterior compartment of only the first abdominal segment, which lacks AbdA. abdA mutant embryos de-repress Dll in the posterior compartments of all abdominal segments, where Ubx levels are low (Gebelein, 2004).

Several genetic experiments were performed to provide in vivo support for the idea that Slp and En work directly with Ubx and AbdA to repress Dll. The design of these experiments had to take into consideration that the activation of Dll in the thorax depends on wg, and that wg expression depends on both slp and en. Consequently, Dll expression is mostly absent in en or slp mutants, making it impossible to characterize the role that these genes play in Dll repression from examining en or slp loss-of-function mutants. However, some of the mutant DMX-Rs described here provide the opportunity to test the model in alternative ways (Gebelein, 2004).

According to the model, DMX[X5]-lacZ is de-repressed in the posterior compartments of the abdominal segments because it fails to assemble the posterior, En-containing complex. Repression of DMX[X5]-lacZ in the anterior compartments still occurs because it is able to assemble the anterior, Slp-containing complex. According to this model, DMX[X5]-lacZ should be fully repressed if Slp is provided in posterior cells. A negative control for this experiment is that ectopic Slp should be unable to repress DMX[X2]-lacZ because this reporter gene does not have a functional Slp binding site. To mis-express Slp, paired-Gal4 (prd-Gal4), which overlaps both the Slp and En stripes in the odd-numbered abdominal segments, was used. As predicted, ectopic Slp repressed DMX[X5]-lacZ but not DMX[X2]-lacZ, providing strong in vivo support for Slp's direct role in Dll repression in the anterior compartments (Gebelein, 2004).

Conversely, the model posits that DMX[X2]-lacZ is de-repressed in the anterior compartment because it cannot bind Slp, but remains repressed in the posterior compartment because it is able to assemble the En-containing posterior complex. Thus, providing En in the anterior compartment should repress DMX[X2]-lacZ. A complication with this experiment is that En is a repressor of Ubx, which is the predominant abdominal Hox protein in the anterior compartment. It was confirmed that prd-Gal4-driven expression of En represses Ubx and that AbdA levels remain low at the time Dll is activated in the thorax. Consequently, ectopic En expression is not sufficient to repress DMX[X2]-lacZ, consistent with the observation that low levels of abdominal Hox proteins are present. Therefore, to promote the assembly of the posterior complex in anterior cells, En was co-expressed with AbdA using prd-Gal4. As predicted, this combination of factors repressed DMX[X2]-lacZ but not DMX[X5]-lacZ, providing strong in vivo evidence for En playing an essential role in Dll repression in the posterior compartments (Gebelein, 2004).

Several observations provide additional support for the model. First, ectopic expression of AbdA or Ubx in the second thoracic segment (T2) represses DMX[X5]-lacZ in the anterior compartment, but not in the posterior compartment. Conversely, expression of AbdA or Ubx in T2 represses DMX[X2]-lacZ only in posterior compartment cells. Second, co-expression of Slp with Ubx completely represses DMX[X5]-lacZ in T2 but does not repress DMX[X2]-lacZ in T2. Third, in those cases where repression is incomplete (for example, En + AbdA repression of DMX[X2]-lacZ in the abdomen), cells that escape repression have low levels of either an abdominal Hox protein or Slp/En. Together, these data provide additional evidence that the abdominal Hox proteins work together with Slp and En to repress Dll (Gebelein, 2004).

The segregation of cells into anterior and posterior compartments during Drosophila embryogenesis is essential for many aspects of fly development. The results presented in this study reveal an unanticipated intersection between anterior-posterior compartmentalization by segmentation genes and segment identity specification by Hox genes. Specifically, it is suggested that the abdominal Hox proteins collaborate with two different segmentation proteins, Slp and En, to mediate repression of a Hox target gene (Dll) in the anterior and posterior compartments of the abdomen, respectively. This mechanism of transcriptional repression suggests a previously unknown use of compartments in Drosophila development. The mechanism proposed here contrasts with the alternative and simpler hypothesis in which the abdominal Hox proteins would have used the same set of cofactors to repress Dll in all abdominal cells, regardless of their compartmental origin (Gebelein, 2004).

These results provide further support for the view that Hox/Exd/Hth complexes do not directly bind co-activators or co-repressors but instead indirectly recruit them to regulatory elements. Consistent with previous analyses, it is suggested that Hox/Exd/Hth complexes are important for the Hox specificity of target gene selection. Additional factors, such as Slp or En in the case of Dll repression, are required to determine whether the target gene will be repressed or activated. In the future, it will be important to dissect in similar detail other Hox-regulated elements, to assess the generality of this mechanism (Gebelein, 2004).

These results also broaden the spectrum of cofactors used by Hox proteins to regulate gene expression. Although the analysis of Exd/Hth in Drosophila and Pbx/Meis in vertebrates has provided some insights into how Hox specificity is achieved, there are examples of tissues in which these proteins are not available to be Hox cofactors and of Hox targets in which Exd and Hth seem not to play a direct role. This study shows that En, a homeodomain segmentation protein, is used as a Hox cofactor to repress Dll in the abdomen. Although the complex defined at the DMX-R includes Exd and Hth, the DNA binding studies demonstrate that Hox and En proteins can bind cooperatively to DNA in the absence of Exd and Hth. These findings suggest that En may function with Ubx and/or AbdA to regulate target genes other than Dll, and perhaps independently of Exd and Hth. Consistent with this idea are genetic experiments showing that, in the absence of Exd, En can repress slp and this repression requires abdominal Hox activity. Although these experiments were unable to distinguish whether the Hox input was direct or indirect, the results suggest that En may bind directly with Ubx and AbdA to repress slp, and perhaps other target genes (Gebelein, 2004).

Finally, these results raise the question of why a compartment-specific mechanism is used by Hox factors to repress Dll. The activation of Dll at the compartment boundary by wg may be important for accurately positioning the leg primordia within each thoracic hemisegment, but this mode of activation requires that Dll is repressed in both compartments in each abdominal segment. The utilization of segmentation proteins such as En and Slp may be the simplest solution to this problem. Compartment-specific mechanisms may also provide additional flexibility in the regulation of target genes by Hox proteins by allowing them to turn genes on or off specifically in anterior or posterior cell types. For these reasons, compartment-dependent mechanisms of gene regulation may turn out to be the general rule instead of the exception (Gebelein, 2004).

Transcriptional activation by Extradenticle in the Drosophila visceral mesoderm

decapentaplegic is a direct target of Ultrabithorax (Ubx) in parasegment 7 (PS7) of the embryonic visceral mesoderm. This study demonstrates that extradenticle (exd) and homothorax (hth) are also required for dpp expression in this location, as well as in PS3, at the site of the developing gastric caecae. A 420 bp element from dpp contains Exd binding sites necessary for expressing a reporter gene in both these locations. Using a specificity swap, Exd was demonstrated to directly activate this element in vivo. Activation does not require Ubx, demonstrating that Exd can activate transcription independently of homeotic proteins. Restoration is restricted to the domains of endogenous dpp expression, despite ubiquitous expression of altered specificity Exd. Nuclear Exd is more extensively phosphorylated than the cytoplasmic form, suggesting that Exd is a target of signal transduction by protein kinases (Stultz, 2006).

Previous studies (Sun, 1995) demonstrated that Ubx directly regulates dpp in PS7 of the VM using a specificity swap strategy. Subsets of six Ubx binding sites were mutated in a 420 bp reporter construct (PX) from binding sites for Q50 homeodomains to binding sites for K50 homeodomains. For example, the wild-type UBX site 5/EXD site e2 was AGGCCTATCAATTAGCACC (with the EXD site underlined) and the mutant UBX site 5/EXD site e2 was AGGCCTAGGGATTAGCACC. It was then possible to restore the expression of these constructs by changing Q50 to K50 in the Ubx protein (called Ubx K50). However, it was not possible to restore expression of a reporter in which all six Ubx sites were altered. This suggested that an additional factor was required, and it was noted that the alterations in the fully substituted PX reporter also disrupted closely apposed Exd binding sites, suggesting that Ubx and EXD may co-regulate dpp (Stultz, 2006).

In previous work (Sun, 1995), it was not possible to restore expression of the fully substituted PX4–9 reporter using Ubx K50. This study shows that ubiquitous and simultaneous induction of Ubx K50 and Exd K50 restores expression of PX4–9 in a manner that is similar to wild-type PX. Induction of Exd K50 alone also restores PX4–9 in these domains but changes the balance of staining intensity between them, with PS7 expression appearing less prominent. This reflects a sufficiency of Exd K50 for activation of gene expression at both sites, but with an additional requirement for Ubx K50 to achieve wild-type levels in PS7. These experiments identify Exd as a direct activator of dpp's VM expression in both PS3 and PS7 (Stultz, 2006).

No HOX proteins are expressed in PS3. Thus, Exd K50 activates gene expression independently of HOX family proteins in this location. In cases where Ubx K50 restored partially substituted PX constructs, restoration was never seen in PS3 (Sun, 1995), further indicating that dpp expression here does not require HOX proteins. In addition, in PS7, where it is clearly established that Ubx contributes to activation of dpp expression, the results demonstrate that Ubx is not absolutely required for Exd K50 to activate transcription. Ubx increases the level of dpp expression, as demonstrated by the reduced PS7 expression in Exd K50-alone restorations, but is not required for Exd function. This point is further reinforced by the ability of Exd K50 to activate PX4–9 gene expression, even in Ubx homozygous mutants. On simple Exd binding sites, PBX proteins have not demonstrated transcriptional activation, but the data suggest that Exd can participate in gene activation without a HOX gene. Other unidentified factors in PS3 or PS7 could also be involved, and one candidate would be HTH, which is genetically required for dpp's VM expression and capable of binding to the PX element in concert with Exd. Genetic evidence for the ability of Exd/HTH to act in the absence of HOX proteins has been steadily accumulating, based on mutant phenotypes that cannot be attributed to HOX genes, and both genetic and in vitro data suggest that HTH/MEIS may have transcriptional activation capabilities (Stultz, 2006).

Two models for the role of Exd in regulating HOX targets have been proposed. The data indicate that the PX element is directly regulated by both Exd and Ubx, allowing evaluation of these two models based on the results. The 'co-selective binding' model proposes that Exd enhances the specificity and affinity of its HOX partner for a DNA binding site. This model requires that Exd and HOX proteins bind cooperatively as heterodimers to closely spaced Exd and HOX binding sites. This model predicts that the relative spacing and orientation of PBX/Exd and HOX binding sites must be tightly constrained, as has been shown by in vitro studies. Although the dpp cis-regulatory PX element contains multiple Ubx and Exd sites identified by DNA footprinting (Sun, 1995), only site e2 resembles the optimal site for binding by a PBX1/HOXB7 heterodimer. Even this site is not a perfect match, and data indicate that this site may be more likely to bind Exd/HTH in vivo. The electrophoretic mobility shift data demonstrate that Exd K50 can bind TAATCCC sites (the optimal site for HOX K50) that replace Ubx sites, as well as unaltered Exd sites. Thus, Exd K50 restores PX4–9 by binding to some or all of these sites. This demonstrates that an Exd protein altered only in its binding specificity can act in vivo through sites of altered spacing and orientation and is not necessarily constrained to act in close proximity to a HOX protein (Stultz, 2006).

The second model, 'widespread binding', proposes that Exd determines the outcome of HOX protein action. According to this model, either Exd or HOX proteins in isolation can bind DNA and act as transcriptional repressors. When both proteins are present, a complex that activates transcription is formed. For Deformed, Exd activates an otherwise silent transcriptional activation domain within the Deformed protein. The physical association between the proteins stabilizes their binding to DNA, but they do not have to bind as heterodimers. This model is more consistent with both the spacing of Exd and HOX sites in the dpp PX element and the apparent flexibility in the location of Exd-responsive sites observed in the experiments. However, this model predicts that independent Exd action is repressive, based on a Deformed-responsive target. In contrast, the data indicate that Exd can also activate reporter gene expression without a HOX partner, suggesting that repression is not the default action of Exd in the absence of HOX proteins (Stultz, 2006).

This study has shown that Exd is a direct activator of dpp expression in the VM. In PS3, the normal action of Exd does not require input from any homeotic protein. In PS7, input from Ubx is critical to achieving the correct level of gene expression, but the data do not support a model where Ubx is absolutely necessary for transcriptional activation. The data suggest that Exd can activate transcription in the absence of HOX proteins but that, in many cases, it also collaborates with HOX proteins, allowing the complex to achieve a more robust level of transcriptional activation. The current notion is that Exd is an essential cofactor for homeotic proteins. An equally tenable model for gene activation is that HOX proteins are the cofactors of Exd, imparting additional spatial regulation, site specificity, and activity to this transcriptional regulator (Stultz, 2006).

The striking restriction of reporter restoration to domains influenced by kinase-mediated signaling pathways led to an examination Exd protein for evidence of phosphorylation. The primary sequence of Exd contains more than 15 potential sites for various protein kinases, including Protein Kinase A (PKA) and Casein Kinase II upstream of its NLS. Protein kinase action is required for gene activation by PBX proteins in tissue culture cells, PKA converts HOX/PBX complexes from repressors to activators on the Hoxb1 autoregulatory element, and phosphorylation by PKA induces nuclear import of PBX1 independently of the PBX/MEIS nuclear localization mechanism. While it was not possible to establish a connection between DPP signaling and Exd phosphorylation, nonetheless, Exd clearly exists in multiple phosphoprotein forms, and the increased phosphorylation is clearly correlated to subcellular localization in Drosophila as well. Thus, Exd must be a target of kinase action, although whether this activity is solely required for nuclear translocation or for activity once in the nucleus is unresolved (Stultz, 2006).

dpp requires both its own expression and that of wg to achieve normal gene expression in the VM. These data led to a hypothesis that the spatial restriction observed in the restoration must be connected to DPP or WG signaling. However, the data do not support this hypothesis, and it is more likely that the major inputs generating dpp's localization in the VM are repressive in nature. In previous work, it was postulated that dpp's spatial regulation in the VM was the result of dual modes of regulation involving both general activation and spatially specific repression and spatially restricted activation (Sun, 1995). The general activator has been identified as biniou (bin), a member of the FoxF/forkhead family of transcription factors. This factor is capable of inducing dpp expression throughout the posterior half the VM, including PS7, when its action is not specifically repressed. This repression comes from multiple inputs. dpp is a direct target of posterior repression via Abd-A. dpp is also repressed outside of PS3 and PS7 via the action of Drosophila T Cell Factor (dTCF) in the absence of WG signaling. The ectopic PS4–6 expression of longer dpp constructs in exd or hth null embryos identifies exd and hth or a downstream target of these genes as another repressor of dpp in PS4–6. Such a downstream target could be teashirt, a known repressor whose VM expression is lost in exd null embryos and is expressed in PS4–6 (Stultz, 2006).

To this model of multiple general activators and spatially specific repressors is added the spatially localized strong activator Ubx. Ubx directly regulates dpp and may also have indirect inputs to dpp's PS7 gene expression, as the reduced restoration in Ubx9.22 null embryos indicates. Ubx is itself repressed via chromatin factors such as Polycomb and osa in the anterior midgut and Abd-A posterior to PS7. dpp autoregulation provides additional weak activation via inputs from SMAD proteins and through DPP-mediated schnurri repression of the repressor brinker. Thus, dpp expression is the cumulative result of general activation constrained by spatially specific repression and augmented by spatially specific activation. Clearly, evolution has deemed the formation of the embryonic midgut of sufficient importance to create a highly buffered, reinforced system of gene expression (Stultz, 2006).

Sequence conservation and combinatorial complexity of Drosophila neural precursor cell enhancers

The presence of highly conserved sequences within cis-regulatory regions can serve as a valuable starting point for elucidating the basis of enhancer function. This study focuses on regulation of gene expression during the early events of Drosophila neural development. EvoPrinter and cis-Decoder, a suite of interrelated phylogenetic footprinting and alignment programs, were used to characterize highly conserved sequences that are shared among co-regulating enhancers. Analysis of in vivo characterized enhancers that drive neural precursor gene expression has revealed that they contain clusters of highly conserved sequence blocks (CSBs) made up of shorter shared sequence elements which are present in different combinations and orientations within the different co-regulating enhancers; these elements contain either known consensus transcription factor binding sites or consist of novel sequences that have not been functionally characterized. The CSBs of co-regulated enhancers share a large number of sequence elements, suggesting that a diverse repertoire of transcription factors may interact in a highly combinatorial fashion to coordinately regulate gene expression. Information gained from the comparative analysis was used to discover an enhancer that directs expression of the nervy gene in neural precursor cells of the CNS and PNS. The combined use EvoPrinter and cis-Decoder has yielded important insights into the combinatorial appearance of fundamental sequence elements required for neural enhancer function. Each of the 30 enhancers examined conformed to a pattern of highly conserved blocks of sequences containing shared constituent elements. These data establish a basis for further analysis and understanding of neural enhancer function (Brody, 2008).

To determine the extent to which neural precursor cell enhancers share highly conserved sequence elements, cis-Decoder analysis was performed of in vivo characterized enhancers. This analysis revealed the presence of both novel elements and sequences that contained consensus DNA-binding sites for known regulators of early neurogenesis. None of the illustrated conserved neural specific sequence elements within two or more neural precursor cell enhancers were present in a collection of 819 CSBs from in vivo characterized mesodermal enhancers, thus ensuring their enrichment in neural enhancers. Consensus binding sites for known TFs were represented: basic Helix-Loop Helix (bHLH) factors and Suppressor of Hairless [Su(H)], respectively acting in proneural and neurogenic pathways; Antennapedia class homeodomain proteins, identified by their core ATTA binding sequence, and the ubiquitously expressed Pbx- (Pre-B Cell Leukemia TF) class homeodomain protein Extradenticle, a cofactor of many TFs, identified by the core binding sequence of ATCA. More than half the conserved elements, termed cis-Decoder tags or cDTs were novel, without identified interacting proteins. Many of the CSBs consisted of 8 or more bp, and often contained core sequences identical to binding sites for known factors as well as other core sequences that aligned with shorter novel cDTs, suggesting that the longer cDTs may contain core recognition sequences for two or more TFs (Brody, 2008).

Most cDTs discovered in this analysis represent elements that are shared pairwise, i.e., by only two of the NB enhancers examined (see the website for a list of cDTs that are shared by only two of the enhancers examined). The fact that the majority of cDTs are shared two ways, with only a small subset of sequences being shared three or more ways, suggests that the cis-regulation of early neural precursor genes is carried out by a large number of factors acting combinatorially and/or that many of the identified cDTs may in fact represent interlocking sites for multiple factors, and the exact orientation and spacing of these sites may differ among enhancers (Brody, 2008).

During Drosophila neurogenesis, bHLH proteins function as proneural TFs to initiate neurogenesis in both the central and peripheral nervous system. TFs encoded by the achaete-scute complex function in both systems, while the related Atonal bHLH protein functions exclusively in the PNS. Different proneural bHLH TFs, acting together with the ubiquitous dimerization partner Daughterless, bind to distinct E-boxes that contain different core sequences. In addition to the core recognition sequence, flanking bases are important to the DNA binding specificity of bHLH factors (Brody, 2008).

One of the principle observations of this study was that the core central two bases of the hexameric E-box DNA-binding site (CANNTG; core bases are bold throughout) were conserved in all the species used to generate the EvoPrint. All of the enhancers included in this study contained one or more conserved bHLH-binding sites, with NB and PNS enhancers averaging 3.9 and 4.1 binding sites respectively. More than a third of the core bases in NB bHLH sites contained a core GC sequence, and more than a third of the core bases in PNS bHLH sites contained either a core GC or a GG sequence. The most common E-box among the NB CSBs was CAGCTG with 14 sites in four of the six enhancers. The CAGCTG and CAGGTG E-boxes are high-affinity sites for Achaete/Scute bHLH proteins. However the CAGCTG site itself is not specific to NB enhancers, as evidenced by its presence in four of the mesodermal enhancer CSBs . The most common bHLH-binding site among PNS enhancers was also the CAGCTG E-box with 11 occurrences in six of the 13 enhancers. In contrast, the most common bHLH motif in enhancers of the E(spl)-complex was CAAGTG, with 16 occurrences in 8 of the 11 enhancers. CAGGTG, previously shown to be an Atonal DNA-binding site, was also common in E(spl) enhancers, with 9 occurrences in 8 of the 13 enhancers, but was less prevalent among NB enhancers. The CAGGTG box was also overrepresented in PNS and E(spl) enhancers relative to its appearance in NB enhancers, and it was also present in four of the characterized mesodermal enhancer CSBs. The CAGATG box was present six times among PNS enhancers but not at all among NB enhancers. Thus there appears to be some specificity of E-boxes in the different enhancer types. The fact that each of these E-boxes is conserved in all the species in the analysis, suggests that there is a high degree of specificity conferred by the E-box core sequence (Brody, 2008).

The analysis also revealed that not only are the core bases of E-boxes shared between similarly regulated enhancers, but bases flanking the E-box were also found to be highly conserved and are also frequently shared by these enhancers. Among the E-boxes found in CSBs of NB enhancers (many are illustrated in the accompanying Table aaCAGCTG (core bases of E-box are bold, flanking bases lower case) is repeated three times in nerfin-1 and once in scrt; gCACTTG is repeated three times in scrt; CAGCTGCA is repeated twice in wor, and CAGCTGctg is repeated twice in scrt . In the dpn CNS NB enhancer, the E-box CAGCTG is found twice, separated by a single base (CAGCTGaCAGCTG). None of these sequences were present in mesodermal enhancers examined, but each is found in PNS enhancers; CAGCTGCA is repeated multiple times among PNS enhancers. Among the conserved PNS enhancer E-boxes (CAAATGca, gcCAAATG, cacCAAATGg, CACATGttg, gCACGTGtgc, ttgCACGTG, agCACGTGcc, aCAGATG, ggCAGATGt, CAGCTGccg, CAGCTGcaattt, gCAGGTGta and cCAGGTGa) each, including flanking bases, is found in two or three PNS enhancers, and these are distributed among all 13 enhancers. Of these, only agCACGTGcc, CAGCTGccg, cCAGGTGa were found once in the sample of neuroblast enhancers and none were found in the sample of mesodermal enhancers. The sequence aaCAAGTG is found in 4 E(spl) complex enhancers, those for E(spl)m8, mγ, HLHmδ and m6, and the sequence aCAGCTGc is found twice in E(spl)m8 and once in m4 and m6; neither sequence was found in the mesodermal enhancers. Therefore, although a given hexameric sequence may often be shared by all three types of enhancers, NB, PNS and E(spl), when flanking bases are taken into account there appears to be enhancer type-specific enrichment for different E-boxes (Brody, 2008).

Antennapedia class homeodomain proteins play essential roles in multiple aspects of neural development including cell proliferation and cell identity. The segmental identity of Drosophila NBs is conferred by input from TFs encoded by homeotic loci of the Antennapedia and bithorax complexes. For example, ectopic expression of abd-A, which specifies the NB6-4a lineage, down-regulates levels of the G1 cyclin, CycE. Loss of Polycomb group factors has been shown to lead to aberrant derepression of posterior Hox gene expression in postembryonic NBs, which causes NB death and termination of proliferation in the mutant clones (Brody, 2008).

This study examined the enhancer-type specificity of sequences flanking the Antennapedia class core DNA-binding sequence, ATTA. Nearly 25% of the NB and PNS CSBs examined in this study contain this core recognition sequence. ATTA-containing sites were found multiple times in selected NB and PNS enhancers. The cis-Decoder analysis identified 18 different neural specific ATTA containing cDTs that were exclusively shared by two or more PNS enhancers or CNS enhancers and 10 were found to be shared between PNS and CNS. The most common cDT, ATTAgca, was shared by two CNS and two PNS enhancers; consensus homeodomain-binding sites are bold, flanking sequence lower case). In addition, 6 homeodomain-binding site cDTs were found twice in wor CSBs, aATTAccg, tttgaATTA, aatcaATTA, ATTAATctt and aaacaaATTAg, but not in other CNS or PNS enhancer CSBs. In some cases these cDTs were found repeated in given enhancer CSBs. Only one of these cDTs aligned with CSBs of enhancers of the E(spl) complex. Given that 2/3 of the occurrences of HOX sites in these promoters can be accounted for by cDTs whose flanking sequences are shared between enhancers, it is unlikely that the appearance of these shared sequences occurs by chance (Brody, 2008).

In summary, the appearance of Hox sites in the context of conserved sequences shared by functionally related enhancers suggests that the specificity of consensus homeodomain-binding sites is conferred by adjacent bases, either through recognition of adjacent bases by the TF itself or in conjunction with one or more co-factors (Brody, 2008).

Examination of the cDTs from Drosophila NB and PNS enhancers revealed that many contained the core Pbx/Extradenticle docking site ATGA. In Drosophila , Extradenticle has been shown to have Hox-dependent and independent functions. Studies have also shown that Pbx factors provide DNA-binding specificity for homeodomain TFs, facilitating specification of distinct structures along the body axis. In the CNS enhancers of Drosophila , most predicted Pbx/Extradenticle sites are not, however, found adjacent to Hox sites (Brody, 2008).

Cytoscape analysis of Pbx motifs revealed that 8 were shared between CNS and PNS enhancer types, and 16 were shared between similarly expressed enhancers, thus indicating that there appears to be some degree of specificity to Pbx site function when flanking bases are taken into account. Three of the Pbx binding-site containing elements also exhibit ATTA Hox sites: 1) the dodecamer GATGATTAATCT (Pbx site is ATGA, Hox sites in bold) shared by the PNS enhancers edl and amos , contains a homeodomain ATTA site that overlaps the Pbx site by a single base, and 2) the smaller heptamer ATGATTA, shared by pfe and ato, likewise contains a homeodomain ATTA site (bold) that overlaps ATGA Pbx site by a single base. Adjacent Hox and Pbx sites have been documented to facilitate synergy between the two factors. Taken together these findings suggest that, as with homeodomain-binding sites, the conserved bases flanking putative Pbx sites are functionally important. These flanking bases are likely to confer different DNA-binding affinities for Pbx factors or are required for binding of other TFs (Brody, 2008).

Also indicating a degree of biological specificity of enhancer types is the distribution of Suppressor of Hairless Su(H) binding sites among neural enhancers. Su(H) is the Notch pathway effector TF of Drosophila . The members of the E(spl) complex, both the multiple basic helix-loop-helix (bHLH) repressor genes and the Bearded family members, have been shown to be Su(H) . The consensus in vitro DNA binding site for Su(H) is RTGRGAR (where R = A or G). Notch signaling via Su(H) occurs through conserved single or paired sites and the presence of conserved sites for other transcription regulators associated with CSBs containing Su(H) binding sites has been documented (Brody, 2008).

Within the CSBs of the six NB enhancers examined, only two, dpn and wor, contained conserved putative Su(H)-binding sites; two dpn sites matched one of the Su(H) consensus sites (GTGGGAA) and two wor sites match the sequence ATGGGAA. Only one of the two dpn sites contained flanking bases conforming to the widely distributed CGTGGGAA site of E(spl) Su(H) binding sites and none of the NB enhancers contained paired Su(H) sites typical of the E(spl) enhancers. Of the 13 PNS cis-regulatory regions examined, only four enhancers contained putative Su(H)-binding sites [sna and ato (ATGGGAA), brd (GTGGGAG)] and dpn (GTGGGAA). dpn also contained a pair of sites that conforms to the SPS configuration frequently found in Su(H) enhancers (CSB sequence: AATGTGAGAAAAAAACTTTCTCACGATCACCTT, Su(H) sites in bold, Pbx site is ATCA). The lack of Su(H) sites in PNS enhancers has been noted in a previous study, and it was suggested that these enhancers are directly regulated by the proneural proteins but not activated in response to Notch-mediated lateral inhibitory signaling. Among the conserved sequences of E(spl) gene enhancers there is an average of 3.4 consensus Su(H) binding sites per enhancer, with most enhancers containing both types of sites, i.e., those with either A or G in the central position (Brody, 2008).

This study offers three insights with respect to Su(H) binding sites. First, although in vitro DNA-binding studies suggest there is a flexibility in the Su(H) binding site, like the bHLH E-box, comparative analysis shows that within any one the Su(H) sites there is no sequence flexibility. Except for the pair of Su(H) sites in the dpn PNS enhancer, none of the CNS or PNS sites contained a central A; less that a quarter of the E(spl) sites consisted of a central A, and all these were conserved across all species examined. In light of the high conservation in these regions the invariant core and flanking sequences are important for the unique Su(H) function at any particular site (Brody, 2008).

A second finding was the extensive conservation of bases flanking the consensus Su(H) sequence in the E(spl) complex genes. For example, the cDT GTGGGAAACACACGAC [Su(H) site bold] was present in HLHm3 and HLHm5 enhancer CSBs, and ACCGTGGGAAAC was conserved in HLHm3 and HLHmβ enhancers. The conservation of bases flanking the consensus Su(H) binding site suggests that the Su(H) site may be flanked by additional binding sites for co-operative or competitive factors, or else, that Su(H) contacts additional bases besides the consensus heptamer (Brody, 2008).

A third observation is that in most cases Su(H) binding sites are imbedded in larger CSBs, suggesting that CSB function is regulated by the integrated function of multiple TFs. For example the dpn NB enhancer Su(H) site is imbedded in a CSB of 24 bases, and the atonal PNS enhancer Su(H) site is imbedded in a CSB of 45 bases. In the E(spl) complex, CSB #6 of HLHmγ, consisting of 30 bases and CSB#13 of m8, consisting of 31 bases (each contains a GTGGGAA Su(H) site, a CACGAG element, conforming to a Hairy N-box consensus CACNAG, and an AGGA Tramtrack (Ttk) DNA-binding core recognition sequence, but the order and context of these three sites is different for each enhancer). Although Su(H) binding sites were present in only a minority of NB and PNS enhancers, the conservation of core bases, as well as the complexity of their flanking conserved sequences points to a diversity of Su(H) function and interaction with other factors (Brody, 2008).

Neural specific cDTs contain core DNA-binding sites for other known TFs. Two of these elements, one exclusively present in NB enhancers (CAGGATA) and a second exclusively present in PNS enhancers (GTAGGA), contained consensus core AGGA DNA-binding sites for Ttk, a BTB domain TF that has been shown to regulate pair rule genes during segmentation and to repress neural cell fates. Another site (CACCCCA), shared by both NB and PNS enhancers, conforms to the consensus binding site of IA-1 (ACCCCA), the vertebrate homolog of nerfin-1 . Most of the neural specific sequence elements illustrated in the paper do not contain sequences corresponding to consensus binding-sites of known regulators of NB expression. The fact that they are represented multiple times in NB CSB sequences suggests that they contain binding sites for unknown regulators of neurogenesis in Drosophila (Brody, 2008).

Neural enriched cDTs that are shared between multiple NB enhancers and also exhibit a low frequency in the sample of mesodermal enhancers examined in this study serve as a resource for understanding enhancer elements that may not have an exclusive neural function [see cis-Decoder tags with multiple hits on two or more NB enhancers]. Notable here is the presence of CAGCTG bHLH DNA binding sites (all with flanking A, CC and TC) and Antennapedia class homeobox (Hox) core DNA binding site ATTA, as well as additional Ttk and Pbx/Extradenticle sites. Present in this list are portions of sequences conforming to Su(H) binding sites. Of particular interest are sequences that are also enriched in the PNS; these sites may bind factors that play similar developmental roles in different tissues. For example, the presumptive Ttk site, AAAGGA (core sequence in bold) is highly enriched in segmental enhancers. Thus, some of these sites can be identified as targets of known TFs, but the identity of most are as yet unknown. These elements shared by multiple enhancers may be useful in identifying other enhancers driving expression in NBs (Brody, 2008).

EvoPrint analysis revealed that all of the enhancer regions examined in this study contained multiple CSBs that were greater that 15 to 20 bases in length. The occurrence of overlapping DNA-binding sites for different TFs is currently the best explanation for the maintenance of intact CSB sequences across ~160 millions of years of collective species divergence. This analysis has revealed that the sequence context, order and orientation of shared cDTs can differ between co-regulating enhancers (Brody, 2008).

Two examples are given here of the complex contextual appearance of cDTs. Each of the eight illustrated CSBs shown was nearly fully 'covered' by cDTs of the NB library, suggesting that each contains multiple overlapping binding sites for a number of TFs. In these two examples, there is no consistent spatial constraints to the association of known TF-binding sites (i.e., bHLH-binding E-box sites) with novel cDTs; a picture that emerges is one of combinatorial complexity, in which known or novel cDTs are associated with each other in different contexts on different CSBs (Brody, 2008).

The information derived from cis-Decoder analysis of neural precursor cell enhancers was used to search for other genomic sequences with similar cis-regulatory properties. Having identified cDTs found multiple times among NB enhancers, the genomic search tool FlyEnhancer was used to identify Drosophila melanogaster genomic sequences that contained clusters of the following cDTs (number in parenthesis is the total number of each cDT in the sample of six NB enhancers): GGCACG (6), GGAATC (4), TGACAG (6), TGGGGT (4), CAGCTG (14), TGATTT (9) CAAGTG (7), CATATTT (5), TGATCC (7) and CTAAGC (6). As a lower limit, a minimum of three CAGCTG bHLH sites was set for this search, because of the prevalence of this site in nerfin-1 and deadpan NB enhancers. Each sequence detected by this search was subjected to EvoPrinter analysis to determine the extent of its sequence conservation. Among the cDT clusters identified, the search identified a 5' region adjacent to the nervy gene that contained three conserved CAGCTG sites as well five other sites identical to TGACAG, GGAATC, TGGGGT, GGCACG and CATATTT. nervy, originally identified as a target of homeotic gene regulation, is expressed in a subset of early CNS NBs, as well as in PNS SOP cells. Later studies have implicated nervy, along with cyclic adenosine monophosphate (cAMP)-dependent protein kinase (PKA) in antagonizing Sema-1a-PlexA-mediated axonal repulsion, and nervy has been shown to promote mechanosensory organ development by enhancing Notch signaling (Brody, 2008).

EvoPrinter analysis revealed that the cluster of neural precursor cell enhancer cDTs positioned 90 bp upstream from the nervy transcribed sequence contains highly conserved sequences. This region contains 10 CSBs that include six conserved E-boxes, three of which conform to the CAGCTG sequence that was prominent in nerfin-1 and deadpan promoters. To determine if this region functions as a neural precursor cell enhancer, transformant lines were generated containing the nervy CSB cluster linked to a minimal promoter/GFP reporter transgene. This analysis of the reporter expression driven by the nervy upstream fragment revealed a pattern indistinguishable from early nervy mRNA expression. Specifically, expression was detected in a large subset of early delaminating NBs and in SOPs and secondary precursor cells of the PNS. Significantly, the nervy enhancer, unlike nerfin-1 and deadpan NB enhancers, activates reporter expression in then PNS and not just in early NBs (Brody, 2008).

The major finding of this study is that enhancers of co-regulated genes in neural precursor cells possess complex combinatorial arrangements of highly conserved cDT elements. Comparisons between NB and PNS enhancers identified CNS and PNS type-specific cDTs and cDTs that were enriched in one or another enhancer type. cis-Decoder analysis also revealed that many of the conserved sequences contain DNA-binding sites for classical regulators of neurogenesis, including bHLH, Hox, Pbx, and Su(H) factors. Although in vitro DNA-binding studies have shown that many of these factors have a certain degree of flexibility in the sequences to which they bind, defined in terms of a position weight matrix, the studies described in this paper show that for any given appearance these sites are actually highly conserved across all species of the Drosophila genus. The genus invariant conservation in many of these characterized binding sites indicates that there are distinct constraints to that sequence in terms of its function (Brody, 2008).

The high degree of conservation displayed in the enhancer CSBs could derive from unique sequence requirements of individual TFs, or the intertwined nature of multiple DNA-binding sites for different TFs. Thus there is a higher degree of biological specificity to these sites than the flexibility that is detected using in vitro DNA-binding studies. As an example, the requirement for a specific core for the bHLH binding site, i.e., for a CAGCTG E-box for nerfin-1, deadpan and nervy, suggests that it is the TF itself that demands sequence conservation; however, the requirement for conserved flanking sequences suggests that additional specific factors may be involved. Although the inter-species conservation of core and flanking sites has been noted by others, the extent of this conservation is rather surprising. To what extent and how evolutionary changes in enhancer function take place, given the conservation of core enhancer sequences, remains a question for future investigation (Brody, 2008).

In addition to classic regulators of neurogenesis, cis-Decoder reveals additional conserved novel elements that are widely distributed or only detected in pairs of enhancers. Many of these novel elements flank known transcription binding motifs in one CSB, but appear independent of known motifs in another. The appearance of novel elements in multiple contexts suggests that they may represent DNA-binding sites for additional factors that are essential for enhancer function. Only through discovery of the factors binding these sequences will it become clear what role they play in enhancer function (Brody, 2008).

Preliminary functional analysis of CSBs within the nerfin-1 neuroblast enhancer reveals that CSBs carry out different regulatory roles. Altering cDT sequences within the nerfin-1 CSBs reveals that most are required for cell-specific activation or repression or for normal enhancer expression levels. CSB swapping studies reveals that, for the most part, the order and arrangement of a number of tested CSBs was not important for enhancer function in reporter studies. The discovery of the nervy neural enhancer by searching the genome with commonly occurring NB cDTs underscores the potential use of EvoPrinter and cis-Decoder analysis for the identification of additional neural enhancers. By starting with known enhancers and building cDT libraries from their CSBs, one now has the ability to search for other genes expressed during any biological event (Brody, 2008).

Hox and senseless antagonism functions as a molecular switch to regulate EGF secretion in the Drosophila PNS

Hox factors are key regulators of distinct cells, tissues, and organs along the body plan. However, little is known about how Hox factors regulate cell-specific gene expression to pattern diverse tissues. This study shows an unexpected Hox transcriptional mechanism: the permissive regulation of EGF secretion, and thereby cell specification, by antagonizing the Senseless transcription factor in the peripheral nervous system. rhomboid expression in a subset of sensory cells stimulates EGF secretion to induce hepatocyte-like cell development. A rhomboid enhancer was identified that is active in these cells; an abdominal Hox complex directly competes with Senseless for enhancer binding, with the transcriptional outcome dependent upon their relative binding activities. Thus, Hox-Senseless antagonism forms a molecular switch that integrates neural and anterior-posterior positional information. As the vertebrate Senseless homolog is essential for neural development as well as hematopoiesis, it is proposed Hox-Senseless antagonism will broadly control cell fate decisions (Li-Kroeger, 2008).

Hox genes have long been known to specify distinct cell types along the body axes of both vertebrates and invertebrates. However, it has remained elusive how Hox factors regulate transcription in a tissue- or cell-specific manner. In this study, a Hox-regulated enhancer (Rho654) active within a subset of PNS cells was identified. Rho654 drives gene expression in abdominal C1-SOP cells to induce oenocytes, and an Exd/Hth/Abd-A complex stimulates gene expression by directly competing with Sens for this enhancer. These findings have three main implications: (1) They demonstrate how a Hox selector gene integrates A-P positional information with a PNS factor to differentially regulate gene expression along the body plan. (2) They uncover a permissive rather than instructive role for Hox factors in regulating transcription. (3) As Hox and Sens binding sites share a common core sequence, they suggest that additional target genes will be regulated through this mechanism. Moreover, genetic studies in mice have linked Gfi1 and Hox factors to both neural and blood cell development, and this study found that vertebrate Hox and Gfi1 factors compete for binding sites in blood cells (Li-Kroeger, 2008).

Sensory organs within the fly head, thorax, and abdomen require sens for their development. However, the type, location, and number of sensory organs that form in different body regions are regulated, at least in part, by Hox factors. The results provide new insight into how Hox factors provide positional information to modify gene expression in sensory cells. A series of point mutations was used to demonstrate that Hox-Sens competition forms a molecular switch whose outcome correlates with the binding activity of each factor. Intrinsic to this model is the following prediction: If Hox factors differ in their ability to interact with composite sites, then A-P differences in Hox-Sens target expression will be observed. Previous biochemical studies revealed that posterior Hox factors have higher affinity for DNA when bound with Pbx (Exd) than anterior Hox proteins (LaRonde-LeBlanc, 2003). Consistent with these results, this study found that a posterior Hox complex (Abd-A/Hth/Exd) that stimulates Rho654 binds 5-fold more RhoA than an anterior Hox complex (Antp/Hth/Exd) that fails to stimulate Rho654. Thus, differences in binding activities between Hox factors for Hox-Sens composite sites result in the differential regulation of gene expression along the A-P axis of the sensory system (Li-Kroeger, 2008).

Hox proteins instructively regulate gene expression by either activating and/or repressing transcription. In fact, the same Hox factor can perform both functions. Abd-A directly binds regulatory elements to activate wingless (wg) and repress decapentaplegic (dpp) in the same cells of the visceral mesoderm. So what determines if a Hox factor activates or represses transcription? Two recent studies revealed that the transcriptional outcome depends upon the binding of additional transcription factors (Gebelein, 2004; Walsh, 2007). The repression of Distal-less (Dll) by the Abd-A and Ultrabithorax (Ubx) Hox factors requires the binding of two transcription factors in addition to Exd and Hth. In posterior compartment cells, the Engrailed (En) protein collaborates with Abd-A/Exd/Hth to bind DNA and repress Dll. In anterior compartment cells, the Sloppy-paired (Slp) protein binds DNA near the Hox complex to repress Dll (Gebelein, 2004). As both En and Slp interact with the Groucho (Gro) corepressor, their recruitment by Hox factors suggests a mechanism to repress transcription. Similarly, Walsh and Carroll found that Ubx and Smad binding are required to repress spalt-major (salm) in the wing. In this case, the Smad proteins recruit the Schnurri corepressor to inhibit transcription. Thus, Hox factors collaborate with additional factors to determine the transcriptional outcome (Li-Kroeger, 2008).

Studies on Abd-A stimulation of a rho enhancer reveal an unexpected mechanism by which Hox factors control gene expression: through competition with the Sens repressor for DNA binding sites. Sens binds RhoA to repress thoracic gene expression, whereas in the abdomen Exd/Hth/Abd-A is permissive for activation by out-competing Sens. Importantly, mutations that disrupt both Sens and Hox binding to RhoA (SensM/HoxM) are expressed in the thorax and abdomen, revealing that Exd/Hth/Abd-A binding is not required to activate gene expression. In addition, coexpression of Exd, Hth, and Abd-A in cultured cells failed to stimulate Rho654- or RhoAAA-luciferase unless Abd-A is fused to a potent activation domain. Thus, unlike other Hox target genes, Hox complexes on RhoA are permissive rather than instructive and stimulate Rho654 by interfering with the binding of a transcriptional repressor (Li-Kroeger, 2008).

A comparison of consensus Sens, Hox/Exd, and Exd/Hth sites reveal a shared core sequence, suggesting that additional target genes will be regulated through Hox-Sens antagonism. In fact, bioinformatics reveals many Hox-Sens composite sites throughout the Drosophila and mammalian genomes. However, both the Sens and Hox sites extend beyond this core sequence, indicating that only a subset of target genes will comprise composite sites. Thus, three types of target genes for those factors are proposed: (1) those regulated by only Hox factors, (2) those regulated by only Sens/Gfi1, and (3) those regulated by both Hox and Sens/Gfi1. For example, many of the previously characterized Hox target genes in the Drosophila embryo are controlled in tissues that do not express Sens, suggesting they are only regulated by Hox genes. However, the Hox and Sens/Gfi1 factors are coexpressed in many neural cells of the developing PNS in both flies and vertebrates, indicating that similarly to rho regulation in abdominal SOP cells, additional targets will be coregulated by Hox and Sens (Li-Kroeger, 2008).

Like Hox genes, the Sens gene family is conserved in C. elegans (Pag-3), Drosophila, and vertebrates (Gfi1 and Gfi1b). These zinc finger transcription factors are essential for nervous system development in all three organisms. In addition, Gfi1 plays a critical role in hematopoiesis, where it participates in regulating stem cell renewal as well as specific blood cell lineages. Interestingly, Hox factors also regulate blood cell differentiation, proliferation, and stem cell renewal. HoxA9, for example, is required for normal hematopoiesis in mice, and alterations in HoxA9 expression have been implicated in acute myeloid leukemia (AML). In fact, a study analyzing the expression profile of 6817 genes in AML patients who either responded or did not respond to treatment found the highest correlated gene associated with poor prognosis is HoxA9. To determine if the Hox-Sens mechanism uncovered in Drosophila is conserved in mammals, in vitro DNA binding assays were used to show that HoxA9 forms a complex with Pbx and Meis that competes with Gfi1 for common binding sites. Moreover, mouse genetic studies support the hypothesis that Hox-Gfi1 factors antagonize each other to regulate gene expression and blood cell development. Thus, Hox-Sens/Gfi1 competition for composite binding sites is likely a conserved mechanism for the regulation of gene expression in organisms from flies to humans (Li-Kroeger, 2008).

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).

Abdominal-A mediated repression of Cyclin E expression during cell-fate specification in the Drosophila central nervous system

Homeotic/Hox genes are known to specify a given developmental pathway by regulating the expression of downstream effector genes. During embryonic CNS development of Drosophila, the Hox protein Abdominal-A (AbdA) is required for the specification of the abdominal NB6-4 lineage. It does so by down regulating the expression of the cell cycle regulator gene cyclin E (CycE). CycE is normally expressed in the thoracic NB6-4 lineage to give rise to mixed lineage of neurons and glia, while only glial cells are produced from the abdominal NB6-4 lineage due to the repression of CycE by AbdA. This study investigated how AbdA represses the expression of CycE to define the abdominal fate of a single NB6-4 precursor cell. Both in vitro and in vivo, the regulation was examined of a 1.9 kb CNS-specific CycE enhancer element in the abdominal NB6-4 lineage. CycE was shown to be a direct target of AbdA and it binds to the CNS specific enhancer of CycE to specifically repress the enhancer activity in vivo. These results suggest preferential involvement of a series of multiple AbdA binding sites to selectively enhance the repression of CycE transcription. Furthermore, the data suggest a complex network to regulate CycE expression where AbdA functions as a key regulator (Kannan, 2010).

All progenies of both thoracic and abdominal NB6-4 can be traced using Eagle (Eg) as a lineage marker, and Reversed polarity (Repo) for differentiating glial cells. Thus, Eg-only expression marks neuronal fate. The thoracic variant of NB6-4 (NB6-4t) gives rise to both neuronal and glial cells, whereas the abdominal variant (NB6-4a) gives rise to only glial cells. The Hox gene Antennapedia (Antp) is expressed in the NB6-4t lineage of thoracic segments (T1-T3) whereas abdominal A (abdA) and Abdominal B (AbdB) are expressed in the NB6-4a lineage of abdominal segments A1-A6 and A7-A8, respectively. However, loss of Antp function does not affect the lineage development in contrast to loss of abdA or AbdB, which results in NB6-4a to NB6-4t homeotic transformations. Thus, thoracic identity of NB6-4 lineage acts as a default state without the requirement of any Hox gene input, while abdominal identity of the lineage is imposed by the function of abdA and AbdB. AbdA and AbdB function by suppressing the expression of CycE, a cell cycle molecule necessary for G1-S phase transition. This study has focused on the mechanism by which AbdA regulates CycE expression (Kannan, 2010).

Exd and AbdA cooperatively bind as a heterodimer to a consensus DNA sequence. During development, nuclear localization of Exd is regulated by interaction with another homeodomain protein Homothorax. The expression pattern of the cofactors Exd and Hth was examined in NB6-4 lineages of wild-type embryos. Exd expression was detected in glial precursors of the NB6-4a lineage, Interestingly, it was not found in the NB6-4t lineage, although in the ectoderm expression levels of nuclear Exd are higher in thoracic segments than in the abdominal segments. In the case of Hth, the protein was detected in NB6-4a glial cells only after late stage 11, and also weakly in NB6-4t derived glia. Thus, consistent with their requirement to modulate the function of Hox proteins, Exd was found expressed in the NB6-4a lineage and not in NB6-4t lineage, although weak expression of Hth was detected in glial cells of NB6-4t lineage. Assuming that both are required together to modulate Hox function, modulation of Hox function is predicted only in NB6-4a lineage (where abdA and AbdB are expressed) and not in NB6-4t (where Antp is expressed). Indeed, loss of abdA and AbdB show NB6-4a to NB6-4t transformations, while loss of Antp has no phenotypic consequence (Kannan, 2010).

To gain additional evidence on the relevance of exd and hth expression in the NB6-4a lineage, their loss of function mutations were analyzed. exd mutant embryos showed an increase in the number of NB6-4a progeny. Some of these cells migrated medially in a pattern similar to glial cells of NB6-4t, while others migrated to the dorso-lateral cortex, suggesting neuronal identity. Abdominal hemisegments of hth mutant embryos did not show an increase in glial progeny, but generated ectopic neurons in the dorsal lateral cortex suggesting homeotic transformation of NB6-4a to NB6-4t. The fact that mutations in both the cofactors of AbdA independently induced homeotic transformations, although at much lesser degree (11% in exd mutants and 7% in hth mutants, compared to 100% in abdA mutants), suggests that this observation is a phenocopy of the abdA loss of function phenotype. The mildness of the phenotypes could be an indication of their role as cofactors to enhance the effect of AbdA rather than essential factors to regulate cell-fate specification (Kannan, 2010).

Next the expression pattern of CycE transcripts was investigated in the transformed abdominal NB6-4 lineage in exd mutant embryos. Consistent with the phenotype at the cellular level, CycE mRNA was observed exclusively in neuronal cells of transformed NB6-4a (Eg expressing cells) in exd mutant background as is the case for the thoracic NB6-4. In hemisegments that show no transformation and thus represent the wild-type NB6-4a, absence of CycE mRNA was found (Kannan, 2010).

The complex cis-regulatory region of zygotic CycE comprises of tissue and stage specific activator and repressor elements within an at least 10 kb genomic region including upstream and downstream elements. Based on the expression pattern of a 1.9 kb lacZ reporter gene (CycE-lacZ) in transgenic assays, it is evident that this region includes cis-acting sequences that drive zygotic CycE transcription both in epidermis (mitotic cycles 14-16) and CNS and regulatory elements responsible for CycE down regulation at the end of st11. Similar to CycE transcripts and CycE protein, CycE-lacZ is not expressed in NB6-4a and is activated in abdA, AbdB double mutant embryos. Thus, this 1.9 kb lacZ reporter reliably reflects the CycE expression in the abdominal NB6-4 lineage (Kannan, 2010).

To elucidate the mechanisms that AbdA specifies in collaboration with Exd/Hth, the 1.9 kb CNS-specific CycE regulatory element for known AbdA, Exd and Hth-binding sites. The element harbours at least 3 binding sites for AbdA, and one each for Exd and Hth in close association. Since strong repression of β-Gal expression from the 1.9 kb CycE-lacZ regulatory element was observed in the NB6-4a lineage, it was wondered whether this regulation could be due to the presence of AbdA and Exd/Hth-binding sites. Therefore, the 1.9 kb CycE-lacZ fragment in more detail both in vitro and in vivo (Kannan, 2010).

Binding of AbdA, Exd and Hth to the regulatory sequences of CycE was tested by electro-mobility shift assays (EMSAs) on three spatially separated AbdA binding sites named AbdA-1, -2 and -3. AbdA-2 is in close association with binding sites of cofactors Exd and Hth, named as 68 bp fragment. EMSA suggested physical association of AbdA protein with all the three putative binding sites (AbdA-1, -2 and -3) when tested independently with corresponding oligosequences. In addition, the association of AbdA, Exd and Hth complex was observed in the 68 bp fragment. To check whether the putative AbdA, Exd and Hth sites identified within the CycE enhancer are responsible for assembling the complex, the core sequences that make critical contact to each of these factors was mutated. Mutations in Hox binding sites resulted in the loss of association of AbdA, Exd and Hth to the core sequence. These results suggest cooperative binding of AbdA and its cofactors Exd and Hth to the CycE enhancer element (Kannan, 2010).

To test the functional relevance of binding of AbdA, Exd and Hth in vivo, reporter gene constructs were constructed, with presence or absence of either of three AbdA binding sites, upstream of a minimal promoter driving β-Gal expression. Transgenic flies were generated by P-element mediated transformation (Kannan, 2010).

Embryos homozygous for lacZ transgenes (two independent insertion lines for each transgene to rule out position variation effects) were stained for Eg, Repo and β-Gal to visualize the regulatory behaviour of the CNS-specific CycE element in the abdominal NB6-4 lineage. The enhancer elements deleted independently for putative binding sites AbdA-1 and AbdA-2 drive β-Gal expression in abdominal NB6-4 cells, suggesting the preferential requirement of both sites for transcriptional repression of CycE. In contrast, the transgene deleted for AbdA-3 showed the wild-type 1.9 kb CycE-lacZ expression pattern i.e. no expression in NB6-4a, suggesting that AbdA-3 may not be a preferred binding site for repressive activity. As expected, regulatory elements deleted for both AbdA-1 and AbdA-3 or AbdA-2 and AbdA-3 drive lacZ expression in abdominal NB6-4 progenies. However, deletion of both the repressor elements abdA1 and abdA2 (CycE-lacZAbdA-1&2) did not result in de-repression of lacZ in NB6-4a. Interestingly, deletion of all the three elements resulted in the activation of lacZ in NB6-4a. This suggests that AbdA-3 may act as a cryptic repressor, functional only in the absence of both AbdA-1 and AbdA-2. In addition, while deletion of AbdA-2 alone resulted in the activation of β-Gal expression in NB6-4a, CycE-lacZ68bp element deleted for Exd/Hth and AbdA-2 mimicked wild-type β-Gal expression pattern suggesting that in its absence, AbdA-1 and AbdA-3 may maintain repression of CycE in NB6-4a (Kannan, 2010).

These results do not rule out the possibility of other sequences in the regulatory region of CycE that contribute to the AbdA-mediated repression. The fact that lacZ is strongly repressed in CycE-lacZAbdA-1&2 and CycE-lacZ68bp embryos, but de-repressed in CycE-lacZAbdA-1,2&3, CycE-lacZAbdA-1&3 and CycE-lacZAbdA-2&3 embryos, suggest that other regulators may function together with this enhancer in vivo. There is a possibility that this regulatory region is between AbdA-1 and AbdA-2, which is required to assemble a repressor complex. Computational screening of the 1.9 kb enhancer fragments revealed the existence of at least 3 Engrailed (En)-binding sites. Two of the En sites are between AbdA-1 and AbdA-2. It is likely that a minimum of two AbdA binding sites along with this regulatory region is required to assemble a repressor complex that also involves Exd/Hth and probably En. When either AbdA-1 or AbdA-2 is deleted, this repressor complex fails to assemble and hence leads to activation of CycE-lacZ in NB6-4a. In the absence of both AbdA-1 and AbdA-2, the putative En-binding region may come closer to AbdA-3 and still be able to assemble a repressor complex. Interestingly, repression of lacZ was observed when the whole 68 bp region is deleted. This could also be due to the fact that AbdA-3 is now much closer to the putative En-binding sites. Unfortunately, this could not be tested in the background of loss of function of en since NB6-4 itself is not born in those embryos. Nevertheless, the above mentioned model appears to be identical to the way Dll expression is repressed in the epithelial cells, which is mediated by Ubx and En. Further investigation in this direction involves Chromatin immunoprecipitation for AbdA or En followed by Western blot analyses for the other protein under different conditions (Kannan, 2010).

To conclude, these results suggest the preferential involvement of a series of multiple AbdA binding sites for enhanced repression of CycE transcription. These data suggests a complex network to regulate CycE expression where AbdA functions as a key regulator. This may have evolved to ensure tight repression of CycE as it is a potent regulator of cell fate in NB6-4 and possibly other CNS lineages (Kannan, 2010).

Alternative splicing modulates Ubx protein function in Drosophila melanogaster

The Drosophila Hox gene Ultrabithorax (Ubx) produces a family of protein isoforms through alternative splicing. Isoforms differ from one another by the presence of optional segments-encoded by individual exons-that modify the distance between the homeodomain and a cofactor-interaction module termed the 'YPWM' motif. To investigate the functional implications of Ubx alternative splicing, this study analyzed the in vivo effects of the individual Ubx isoforms on the activation of a natural Ubx molecular target, the decapentaplegic (dpp) gene, within the embryonic mesoderm. These experiments show that the Ubx isoforms differ in their abilities to activate dpp in mesodermal tissues during embryogenesis. Furthermore, using a Ubx mutant that reduces the full Ubx protein repertoire to just one single isoform, specific anomalies were obtained affecting the patterning of anterior abdominal muscles, demonstrating that Ubx isoforms are not functionally interchangeable during embryonic mesoderm development. Finally, a series of experiments in vitro reveals that Ubx isoforms also vary in their capacity to bind DNA in presence of the cofactor Extradenticle (Exd). Altogether, results indicate that the structural changes produced by alternative splicing have functional implications for Ubx protein function in vivo and in vitro. Since other Hox genes also produce splicing isoforms affecting similar protein domains, it is suggested that alternative splicing may represent an underestimated regulatory system modulating Hox gene specificity during fly development (Reed, 2010).

The experiments described in this study indicate that the generation of structural differences among Ubx proteins by alternative splicing is relevant for the functional specificity of Ubx in vivo. These structural features modulate essential biochemical properties of Ubx proteins such as their DNA-binding profiles in the presence of a cofactor (Reed, 2010).

The differential effects of Ubx in vivo are apparent in the posterior visceral mesoderm, but not in the anterior. To understand this it must first be noted that the mechanism of dpp674 visceral mesoderm enhancer repression is different anterior and posterior to PS7. In the anterior visceral mesoderm, repression requires Exd, since in Exd null mutants, dpp674 is ectopically expressed anterior to PS7. But Hox genes appear to play no role in the normal repression of dpp in this region. The same Exd-dependent mechanism may also be acting in the posterior, but it is not necessary, for no posterior ectopic expression is observed in Exd null mutants. The posterior repression depends instead on the Hox protein Abd-A, which is presumably able to repress dpp674 in the absence of Exd as a cofactor (Reed, 2010).

All Ubx protein isoforms are able to induce dpp ectopically in the anterior, suggesting that they can all override the normal repression mediated by Exd. However, they exert differential effects in the posterior, where Abd-A is the controlling repressor. Abd-A has a very similar DNA-binding specificity to that of Ubx. It binds to multiple sites in the dpp674 enhancer, including those to which Ubx binds. Through some of these sites it serves as an activator, but through others it acts as a repressor. With Ubx form Ia, the repressing effect by Abd-A is dominant. The results suggest that Ubx form IVa is either able to compete more effectively as an activator with the repressing action of Abd-A bound at other sites or able to displace the binding of Abd-A at the sites where it mediates repression (Reed, 2010).

Ubx form IVa also overrides the normal specificity of the dpp-lacZ enhancer for the visceral mesoderm, activating it ectopically in the somatic mesoderm of most trunk segments. It is not known whether the activity of this enhancer is normally restricted to the visceral mesoderm by a required cofactor that is present only in the visceral mesoderm, by repressors present in other tissues, or by both mechanisms. However, UbxIVa at high levels seems able to override this normal tissue specificity. The more efficient DNA binding of this isoform (in the presence of Exd) may bypass the requirement for a cofactor or displace a repressor more effectively (Reed, 2010).

A quantitatively controlled study showed that the levels of Ubx protein are very important to determine the functional outcome of Ubx in vivo. In this context, the results show that in spite of significant variation across expression, levels of UbxIVa protein in different transgenic lines, this isoform is consistently able to produce a similar output in terms of dpp target activation in posterior regions. This suggests that for UbxIVa dpp target activation and protein concentration may relate to one another in the form of a sigmoidal function with a narrow protein concentration interval acting as a threshold that is crossed by all UbxIVa lines tested in this study. Given that UbxIa lines achieving comparable protein expression levels to UbxIVa lines, it was possible to activate dpp in posterior regions of the embryo, it is concluded that qualitative differences in Ubx protein structure as determined by alternative splicing are causal to the observed differential behavior in target activation (Reed, 2010).

The tissue-specific effects of Ubx isoform ectopic expression emphasize the likely role that the splice isoforms play in mediating specific Ubx functions in different tissues. Indeed, the isoforms have different tissue distributions in embryogenesis with UbxIa expressed predominantly in epidermis, mesoderm, and peripheral nervous system whereas UbxIVa appears to be exclusively expressed in the central nervous system. The analysis of the UbxMX17 mutation, which retains the full Ubx expression pattern but generates only isoform UbxIVa, supports the endogenous relevance of splice isoforms for tissue-specific Hox function. Although in UbxMX17 embryos UbxIVa can replace the function of Ubx isoforms I and II in the epidermis with the generation of a normal cuticle pattern, the peripheral nervous system is affected, and in this study clear defects were found in the segmental specification of somatic muscles. Tissue-specific isoform functions may be mediated by effects on cofactor interaction or may also involve effects on collaborative regulatory interactions between Hox proteins and tissue-specific regulators (Reed, 2010).

The results of in vitro binding studies of the Ubx/Exd element show that Ubx isoforms differ in their capacity to interact with DNA in the presence of the cofactor Exd. A possibility that emerges from these results is that different Ubx isoforms display differential levels of interaction with Exd (in the presence of target DNA) and, accordingly, could perform specific functions in vivo as a consequence to the distinct levels of nuclear Exd available in different regions of the embryo (Reed, 2010).

in vitro studies also show that ablation of the YPWM motif has little effect on the ability of Ubx form Ia to form a complex with Exd, but it significantly reduces complex formation by Ubx form IVa (Reed, 2010).

The observation that mutated forms of the Ubx protein lacking the hexapeptide interact with Exd at all is at first sight surprising, especially in view of the structural studies showing that the YPWM motif provides the major contact between Hox and Exd proteins bound to DNA. However, this finding is not inconsistent with earlier work. The article that originally described cooperative interactions between Ubx and Exd used Ubx proteins deleted of all sequences located amino-terminal to the homeodomain. These proteins therefore lacked the hexapeptide motif entirely. Interactions with Exd that do not require the hexapeptide have also been reported in a study focused on Ubx functions independent from Exd and Hth proteins. In addition, alanine replacement of the hexapeptide does not affect the way another Hox protein (i.e., Abd-A) interacts with Exd. Furthermore, a recent study suggests that Ubx-Exd recruitment may rely not on a single, but on several different mechanisms, some of which require a short evolutionarily conserved motif originally termed UbdA. Thus, the hexapeptide is unlikely to be the only Hox protein motif that interacts with Exd (Reed, 2010).

Crystallographic studies show that the Ubx and Exd homeodomains are closely adjacent, almost touching each other, the extent of this proximity being revealed by a significant reduction in solvent-accessible surface area within the area of putative contact. In addition, the distance between the Ubx recognition helix C terminus and the Exd recognition helix N terminus is just 9 Å, with one particular residue of Ubx (Lys58) well positioned to form hydrogen bonds with a residue of Exd (Ser48). Thus, regions within the Ubx homeodomain may be responsible for additional interactions with Exd. Sequences elsewhere in the proteins may also contribute to these interactions. A fine combination of protein mapping and crystallographic studies may be required to reveal the structural details of such interactions (Reed, 2010).

These experiments would be consistent with the possibility that the Ubx linker region itself could be a previously uncharacterized region of Ubx that makes contacts with Exd. This would be supported by experiments of protein-protein interaction in yeast, which also suggest that the Ubx linker region could affect the interaction between Ubx and Exd, and by the high evolutionary conservation of these sequences across phylogenetically distant species of flies. The results could also be accommodated in a model in which the Ubx linker region induces a conformational change elsewhere in the Ubx protein, such that the degree of interaction with Exd is affected. Alternatively, the regulatory potential of the Ubx protein could be affected. In particular, linker-dependent conformational changes may affect the behavior of critical regulatory motifs of Ubx, such as the recently studied QA motif involved in the modulation of Ubx activities in a tissue-specific manner (Hittinger, 2005) and the SSYF motif (Tour, 2005), an evolutionarily conserved motif close to the N terminus of the Ubx protein that is involved in transcriptional activation (Reed, 2010).

Other reports have also emphasized that the Hox linker regions are not just passive spacers. One such study shows that mutation of the short linker region in the Hox protein Abd-A specifically disables its capacity to activate wingless, a natural target gene, without affecting its ability to repress dpp. When comparing cofactor and DNA-binding properties of this Abd-A mutant protein with those of the wild-type Abd-A on a repressor element of the Distalless gene (DllR), no differences were seen in the interaction between the mutant form of Abd-A and Exd protein. (It is perhaps worth noting that these in vitro experiments required the presence of Homothorax protein; on this particular DNA target (DllR), the formation of complexes with Abd-A and Exd alone was below the limit of detection of the assay. Another report described Ubx isoform-specific functions for the repression of the Dll gene. These studies suggest that sequences within the Hox linker regions may modify the trans-regulatory potential of Hox proteins without necessarily affecting the DNA-binding properties of these proteins. The results extend these studies, showing that linker regions may at times affect target gene regulation and the DNA binding/cofactor interaction abilities of a Hox protein. The binding results are consistent with a recent study that integrates DNA-binding tests with computational predictions of ordered and disordered segments of the Ubx protein (Liu, 2008), proposing that sequences outside the homeodomain can reduce the DNA-binding activities of the Ubx protein by twofold (Reed, 2010).

Beyond the effects that alternative splicing may have on modulating Hox interactions with cofactors, other possibilities must also be considered. Given that in these experiments a very unstable binding of the Ubx proteins to target DNA was observed in the absence of Exd, it is difficult to advance arguments regarding the affinities of the individual Ubx isoforms for this molecular target on firm grounds. In spite of this, given that in various physiological contexts the binding of Hox proteins to target sites in the absence of cofactors has been demonstrated, it is pertinent to suggest that alternative splicing might also be influencing Hox DNA binding independently from cofactor interactions, at the level of changing DNA-binding affinities of the individual isoforms for their target DNAs. This possibility would be suitable for experimental investigation in the physiological context of Drosophila adult appendage development, as Ubx is known to act independently from cofactors during the development of these structures (Reed, 2010).

Examples from other Drosophila Hox genes further support a more general role of differential splicing in the diversification of Hox function during development, affecting protein modules outside the Ubx linker region discussed above. For instance, genetic dissection of the Abd-B gene demonstrated the existence of two distinct gene functions, originally termed morphogenetic (m) and regulatory (r). The spectrum of mRNAs derived from the Abd-B locus have been cloned, and a family of Abd-B transcripts generated by differential promoter use has been revealed , that, in turn, leads to different splicing variants affecting 5' sequences of the gene. Two proteins products, called m and r, are produced from the Abd-B transcripts; m differs from r in that it encodes an additional large glutamine-rich amino-terminal domain. Furthermore, ectopic expression of each Abd-B protein class leads to specific effects on the larval cuticle, suggesting that the isoforms have specific developmental functions (Reed, 2010).

Given that in Drosophila several Hox proteins possess alternatively spliced modules modifying the distance between the hexapeptide and the homeodomain, while others produce functionally different isoforms via differential promoter use coupled to alternative splicing, alternative splicing may truly represent a very important, yet underexplored regulatory mechanism modulating the functional specificity of Hox proteins during development (Reed, 2010).

Segment-specific regulation of the Drosophila AP-2 gene during leg and antennal development

Segmentation involves subdivision of a developing body part into multiple repetitive units during embryogenesis. In Drosophila and other insects, embryonic segmentation is regulated by genes expressed in the same domain of every segment. Less is known about the molecular basis for segmentation of individual body parts occurring at later developmental stages. The Drosophila transcription factor AP-2 gene, dAP-2, is required for outgrowth of leg and antennal segments and is expressed in every segment boundary within the larval imaginal discs. To investigate the molecular mechanisms generating the segmentally repetitive pattern of dAP-2 expression, transgenic reporter analyses was performed and multiple cis-regulatory elements were isolated that can individually or cooperatively recapitulate endogenous dAP-2 expression in different segments of the appendages. An enhancer specific for the proximal femur region, which corresponds to the distal-most expression domain of homothorax (hth), was analyzed in the leg imaginal discs. Hth is known to be responsible for the nuclear localization and, hence, function of the Hox cofactor, Extradenticle (Exd). Both Hth and Exd were shown to be required for dAP-2 expression in the femur, and a conserved Exd/Hox binding site was found to be essential for enhancer activity. These loss- and gain-of-function studies further support direct regulation of dAP-2 by Hox proteins and suggest that Hox proteins function redundantly in dAP-2 regulation. This study reveals that discrete segment-specific enhancers underlie the seemingly simple repetitive expression of dAP-2 and provides evidence for direct regulation of leg segmentation by regional combinations of the proximodistal patterning genes (Ahn, 2011).

The segmentally repeated expression of dAP-2 in the developing leg and antennal discs may suggest that its expression in each segment is regulated in a similar manner by upstream segmentation genes. Alternatively, each domain (ring) of dAP-2 expression could result from the combinatorial activities of multiple transcription factors, which themselves are not expressed in a repeated pattern, but instead occupy distinct and broader domains along the PD axis of the appendages. Current data provide strong evidence that the latter strategy is utilized to establish dAP-2 expression in all but the tarsal segments. It seems that the tarsus has adopted a strategy different from that of other leg segments to regulate dAP-2 expression (Ahn, 2011).

In an effort to understand molecular mechanisms controlling dAP-2 expression during leg development, the regulatory potential of dAP-2 genomic fragments was tested using transgenic reporter analyses. Multiple enhancers were successfully isolated which can independently direct reporter expression in specific leg segments and together recapitulate, almost completely, the endogenous expression pattern. It is intriguing that the relative positions of these enhancers on the chromosome are well correlated with the position of their activity along the PD axis of the leg. Importantly, the presence of segment-specific enhancers suggests that dAP-2 expression is differentially regulated in each leg segment. It is likely that each domain of dAP-2 expression in the true joints is regulated by a combination of upstream regulators involved in PD patterning using segment-specific enhancers similar to the distinct enhancers used to regulate expression of the pair-rule gene, even-skipped, in every other parasegment during embryonic segmentation. Interestingly, dAP-2 expression in the coxa is differentially regulated along the DV axis and depends on two region-specific enhancers. In addition, the EB fragment displayed relatively weaker activity in the ventral region compared to the larger E6 fragment. These data raise the possibility that DV patterning genes are also involved in dAP-2 regulation in the proximal segments. It is possible that the use of multiple region-specific enhancers is a general mechanism establishing expression of segmentation genes during leg development (Ahn, 2011).

Current data indicate that dAP-2 expression in antennal discs also requires multiple region-specific enhancers. Some of the leg enhancers showed an antennal expression pattern similar to their leg patterns with respect to the PD axis. However, there are also enhancers specific for either antennal or leg discs implying that the genes required for normal identity of the two homologous appendages might be involved in regulation of dAP-2 expression in some segments. One of the features that distinguish antennae from legs is that in antennal discs, hth expression is expanded to the intermediate region where dac expression is missing. In contrast, the expression patterns of Dll, dac and hth are very similar in the proximal and distal regions of the two appendages. It is interesting to note that the dAP-2 enhancers for the most proximal and distal regions are shared between the two appendages while the intermediate region utilizes distinct enhancers. This implies that dAP-2 expression in the intermediate region is more likely to be regulated by antennal- or leg-specific regulatory pathways. For example, although the femur and the AIII are homologous structures, the Hox-dependent proximal femur enhancer is active in the leg, but not in the antenna. Likewise, the BE enhancer is active in the proximal AIII of the Hox-free antenna, but not in the leg (Ahn, 2011).

The Hox gene Antp has been considered to be a key factor in determining leg identity since Antp mutant clones in the T2 leg cause a leg-to-antenna transformation, mainly outside of the Hth domain. Previous studies suggested that Antp performs its selector function by acting as a repressor of hth and other antennal genes in the intermediate leg. In contrast, both Antp and hth are expressed in the proximal leg, and are required for growth and segmentation of this region. Therefore, it has been proposed that the role of Antp as a repressor of hth is limited to the intermediate leg, and that both Antp and Hth contribute to proper development of the proximal leg (Ahn, 2011).

The similar loss-of-function phenotypes of hth and exd suggest that Hth and Exd act on common target genes during development of the proximal leg. In certain developmental contexts, Hth can directly bind to DNA through its homeodomain in a ternary complex including Exd and Hox proteins to regulate expression of target genes. However, it has been shown that a Hth isoform lacking the homeodomain can execute the function of Hth in PD patterning of Drosophila leg discs indicating that direct DNA binding is not necessary for its function in proximal leg discs. Since no conserved, consensus Hth binding site were found in the proximal femur enhancer of dAP-2, Hth is likely functioning in dAP-2 expression through a mechanism independent of its direct binding to DNA through its homeodomain. Instead, Hth may regulate dAP-2 expression in the proximal femur by facilitating the nuclear localization of Exd or by interacting with other transcription factors which bind DNA (Ahn, 2011).

As a cofactor of Hox proteins, Exd, and its mammalian homolog Pbx, cooperatively bind DNA with Hox proteins and regulate expression of their target genes which are involved in a variety of developmental processes in both vertebrates and Drosophila. Although previous genetic analyses have revealed essential functions of Exd and Hox proteins in leg development, it has been unclear whether these factors act together on common target genes during this process. This study has identified a conserved Exd/Hox binding site which is required for activity of the proximal femur enhancer of dAP-2. Through clonal analyses, it was demonstrated that hth, exd and Antp are necessary for dAP-2 expression in the presumptive proximal femur of leg imaginal discs. This is the first example of a direct target gene of an Exd/Hox complex in Drosophila limb development. This study also provides insight into the molecular mechanism integrating the combinatorial actions of PD patterning genes in the regulation of region-specific expression of leg segmentation genes (Ahn, 2011).

Although Antp is expressed in all three pairs of legs, most of the prothoracic (T1) and metathoracic (T3) legs with Antp mutant clones appeared to be normal, except for a rare fusion between the femur and tibia. However, Scr/Antp double mutant clones in T1 legs and Antp/Ubx double mutant clones in T3 legs generated leg defects indistinguishable from those generated by Antp mutant clones in T2 legs. It was proposed that the low penetrance of the Antp mutant phenotypes in T1 and T3 legs is due to redundancy with Scr and Ubx, which are expressed in T1 and T3 leg discs, respectively. This idea is consistent with the previous observations that Scr and Ubx both can induce antenna-to-leg transformations when ectopically expressed in antennal discs. It is proposed that Antp, Scr and Ubx can redundantly activate dAP-2 expression in the proximal femur as Exd/Hox heterodimers based on the following observations. First, dAP-2 expression in T2 leg discs, but not in T1 and T3 leg discs, requires Antp. Secondly, EMSA results demonstrate that all three Hox proteins bind strongly to the binding site in the proximal femur enhancer as Exd/Hox heterodimers. Thirdly, all three Hox proteins can activate PrF enhancer function when ectopically expressed in antennal discs (Ahn, 2011).

Competition for cofactor-dependent DNA binding underlies Hox phenotypic suppression

Hox transcription factors exhibit an evolutionarily conserved functional hierarchy, termed phenotypic suppression, in which the activity of posterior Hox proteins dominates over more anterior Hox proteins. Using directly regulated Hox targeted reporter genes in Drosophila, this study shows that posterior Hox proteins suppress the activities of anterior ones by competing for cofactor-dependent DNA binding. Furthermore, a motif in the posterior Hox protein Abdominal-A (AbdA) was identified that is required for phenotypic suppression and facilitates cooperative DNA binding with the Hox cofactor Extradenticle (Exd). Together, these results suggest that Hox-specific motifs endow posterior Hox proteins with the ability to dominate over more anterior ones via a cofactor-dependent DNA-binding mechanism (Noro, 2011).

fkh250 is a 37-base-pair (bp) element from the forkhead (fkh) gene, which is directly regulated by Scr; it contains a single Hox-Exd-binding site that, compared with other Hox-Exd heterodimers, is preferentially bound by Scr-Exd in vitro (Ryoo, 1999b). When lacZ is placed under the control of fkh250, fkh250-lacZ is specifically expressed in PS2 in an exd- and Scr-dependent manner. Indeed, misexpression of Scr throughout the Drosophila embryo can ectopically activate fkh250-lacZ. Notably, ectopic activation of fkh250-lacZ occurs even in the abdomen, in the presence of endogenous, more posterior Hox (Noro, 2011).

In contrast to fkh250, fkh250CON (for 'consensus') is an artificial variant of fkh250 with two base pair substitutions that enable fkh250CON-lacZ to be directly regulated by four Hox genes in an exd-dependent manner: Scr, Antp, and Ubx activate this reporter in PS2-PS6, while AbdA represses it in abdominal segments (Ryoo, 1999b). Consistent with its relaxed specificity in vivo, fkh250CON binds well to Scr-Exd, Antp-Exd, Ubx-Exd, and AbdA-Exd heterodimers in vitro. The promiscuous binding and regulation by multiple Hox proteins classifies fkh250CON as a shared Hox target gene, while the Scr-specific regulation of and binding to fkh250 suggests that it is a specific Hox target gene (Mann, 2009; Noro, 2011).

Because of their distinct specificities, fkh250-lacZ and fkh250CON-lacZ provide an ideal system to examine the molecular mechanism of phenotypic suppression. In accordance with the premise of posterior dominance, coexpression of Scr and AbdA throughout the fly embryo leads to repression of fkh250CON-lacZ by AbdA. Note that both fkh250 and fkh250CON-lacZ require direct binding by the Hox cofactor Exd. The primary distinction between these two readouts is that AbdA-Exd binds well to fkh250CON but not to fkh250. Accordingly, it is concluded that AbdA cannot suppress the activities of Scr if it cannot bind to the target element. Furthermore, in this system, posterior dominance cannot be mediated by miR activity or competition for factors, such as Exd, off DNA. Rather, these data support a model in which competition for cofactor-dependent DNA binding underlies phenotypic suppression for shared Hox target genes (Noro, 2011).

If AbdA is outcompeting Scr for binding to fkh250CON, AbdA would be expected to have a higher affinity for this sequence compared with Scr. To test this prediction, the affinities of AbdA-Exd and Scr-Exd heterodimers for fkh250CON were measured in vitro. AbdA-Exd heterodimers bound more than twofold more tightly to fkh250CON compared with Scr-Exd. Thus, at the same concentration, AbdA-Exd is more likely than Scr-Exd to be bound to fkh250CON, consistent with the idea that competition depends on cofactor-dependent DNA binding (Noro, 2011).

Binding to fkh250CON is Exd-dependent for both AbdA and Scr, implying that AbdA has domains that allow higher binding affinity with Exd to this target site. In general, Hox interactions with Exd are mediated by the highly conserved, four-amino-acid motif YPWM, which directly binds to a hydrophobic pocket established by the three-amino-acid loop extension (TALE) in the Exd homeodomain (Mann, 2009). For some Hox proteins, the YPWM-TALE interaction is necessary and sufficient for cooperative DNA binding with Exd and target gene regulation in vivo (Joshi, 2010). In addition to the YPWM motif, AbdA, but not Scr, has a second well-conserved tryptophan-containing motif, TDWM, which could play a role in mediating AbdA-Exd interactions. However, when a mutant form of AbdA in which both the YPWM and TDWM motifs are mutated to alanines (2WAla) was coexpressed with Scr in the phenotypic suppression assay, fkh250CON-lacZ was repressed to the same extent as by wild-type AbdA. Thus, although the YPWM and TDWM may contribute to interactions with Exd, these motifs are not necessary for AbdA to dominate over Scr (Noro, 2011).

Immediately C-terminal to its homeodomain, AbdA contains a so-called UbdA motif, a nine-amino-acid sequence also present in Ubx, which has been suggested to mediate cooperative binding with Exd to some DNA sequences (Merabet, 2007). In fact, UbdA is part of a larger 23-residue conserved region adjacent to the AbdA homeodomain, which is referred to as the UR motif (for UbdA-RRDR). To determine whether this or other regions in the C-tail of AbdA are involved in mediating phenotypic suppression, a series of C-terminal truncations were tested for their ability to compete with Scr for the repression of fkh250CON-lacZ in vivo. All AbdA variants were epitope-tagged, allowing use of transgenes that express at similar levels (Noro, 2011).

AbdA's ability to compete with Scr for fkh250CON regulation is eliminated when the entire C terminus is removed (ΔC197). Adding back only the UR motif partially restores AbdA's ability to dominate over Scr (ΔC220). Consistently, an internal deletion that removes most of the UR motif (Δ200-220) exhibits a reduced ability to repress fkh250CON-lacZ. No additional loss of repressive activity is displayed by an AbdA variant in which both the YPWM and TDWM motifs are mutated in combination with this internal deletion (2WAlaΔ200-220). Additional sequences in the C-tail of AbdA may account for the residual activity of variants lacking the UR motif (Δ200-220 and 2WAlaΔ200-220). All AbdA variants used in this study are capable of repressing the exd-independent target gene spalt in the wing imaginal disc, confirming that these mutants are still functional transcription factors. Furthermore, these mutants retain the ability to repress gene expression in vivo, arguing that AbdA's repressive activity is not sufficient to account for its ability to dominate Scr. Together, these data highlight the importance of the UR motif for phenotypic suppression (Noro, 2011).

The above data show that the UR motif is required for AbdA to compete with Scr in vivo. To test the hypothesis that UR carries out this function by facilitating cooperative DNA binding with Exd, the ability of the truncated AbdA variants to bind fkh250CON in complex with Exd was analyzed. In general, the results correlate with the in vivo phenotypic suppression assay: Those mutants that failed to suppress Scr's ability to activate fkh250CON-lacZ (ΔC197, Δ200-220, and 2WAlaΔ200-220) were severely compromised in binding fkh250CON with Exd in vitro. Together, these data strongly suggest that cooperative DNA binding with Exd is required for phenotypic suppression and that domains unique to AbdA are critical for its ability to dominate over Scr. More specifically, they argue that AbdA's UR motif is necessary for cooperative binding of AbdA and Exd to fkh250CON and that the YPWM and TDWM motifs are not sufficient to mediate this interaction on this binding site. The insufficiency of the YPWM motifs to mediate cooperative binding with Exd has been observed for other Hox proteins, suggesting that the use of paralog-specific motifs such as UR may be a general phenomenon (Noro, 2011).

To test the generality of AbdA's dependency on its UR motif for posterior dominance, the same AbdA variants were analyzed for their ability to suppress the activity of the thoracic Hox protein Antp in the patterning of the larval epidermis. When ectopically expressed, Antp transforms the head and first thoracic segment (T1) toward the identity of the second thoracic segment (T2), where Antp is normally expressed). In contrast, when AbdA is ectopically expressed, the head and thorax acquire abdominal segmental identities. Consistent with the rules of phenotypic suppression, wild-type AbdA is able to produce this transformation even in the presence of exogenous Antp. However, similar to the results with fkh250CON-lacZ, AbdA mutants that are compromised in their ability to cooperatively bind DNA with Exd (e.g., ΔC197, Δ200-220, and 2WAlaΔ200-220) fail to suppress the activity of Antp (Noro, 2011).

Taken together, these data support a model in which phenotypic suppression depends on a competition for cofactor-dependent DNA binding. It follows that this mechanism would only apply to readouts that depend on regulatory elements that are targeted by multiple Hox proteins. For example, ectopic Scr can activate fkh and other target genes required for salivary gland development in more posterior segments, illustrating that this Hox-specific function does not obey phenotypic suppression. Furthermore, it is particularly noteworthy that, compared with the anterior Hox protein Scr, AbdA has additional motifs that facilitate complex formation with Exd on DNA. These data suggest that when phenotypic suppression is observed, the more posterior Hox proteins may have a higher affinity for shared binding sites; this higher affinity is a consequence of the quantity and quality of motifs that mediate cooperative DNA binding with Exd. It is speculated that these motifs may be used differently at different target genes and binding sites. It is suggested that the YPWM motif provides a common, basal level of interaction between Hox proteins and Exd. In the context of Hox-specific regulatory elements, this motif may be sufficient to enable Hox-Exd regulation of some target genes. In contrast, in the context of shared enhancers and when multiple Hox proteins are coexpressed, additional, paralog-specific motifs present in the more posterior Hox proteins enable tighter binding of Hox-Exd dimers to DNA, leading to more posterior phenotypes. This was shown to be the case for a single shared Hox-Exd enhancer and suggest that the generality of this mechanism for phenotypic suppression will become apparent as more shared and specific targets for Hox proteins are characterized at high resolution (Noro, 2011).

Engrailed cooperates directly with Extradenticle and Homothorax on a distinct class of homeodomain binding sites to repress sloppy paired

Even skipped (Eve) and Engrailed (En) are homeodomain-containing transcriptional repressors with similar DNA binding specificities that are sequentially expressed in Drosophila embryos. The sloppy-paired (slp) locus is a target of repression by both Eve and En. At blastoderm, Eve is expressed in 7 stripes that restrict the posterior border of slp stripes, allowing engrailed (en) gene expression to be initiated in odd-numbered parasegments. En, in turn, prevents expansion of slp stripes after Eve is turned off. Prior studies showed that the two tandem slp transcription units are regulated by cis-regulatory modules (CRMs) with activities that overlap in space and time. An array of CRMs that generate 7 stripes at blastoderm, and later 14 stripes, surround slp1. Surprisingly given their similarity in DNA binding specificity and function, responsiveness to ectopic Eve and En indicates that most of their direct target sites are either in distinct CRMs, or in different parts of coregulated CRMs. Cooperative binding sites for En, with the homeodomain-containing Hox cofactors Extradenticle (Exd) and Homothorax (Hth), were located within two CRMs that drive similar expression patterns. Functional analysis revealed two distinct, redundant sites within one CRM. The other CRM contains a single cooperative site that is both necessary and sufficient for repression in the en domain. Correlating in vivo and in vitro analysis suggests that cooperativity with Exd and Hth is a key ingredient in the mechanism of En-dependent repression, and that apparent affinity in vitro is an unreliable predictor of in vivo function (Fujioka, 2012).

Consistent with the fact that Eve is expressed earlier than En, with some overlap at embryonic stages 8–9, slp CRMs tended to respond to ectopically expressed Eve at earlier stages than to En. Transgenic dissections further showed that they have distinct responsive regions within CRMs, suggesting that many of their binding sites are distinct. This is somewhat surprising because they are both homeodomain-containing repressors that set the posterior borders of slp stripes, and they have been seen to have similar in vitro binding specificities. A possible explanation is that they cooperate in DNA binding with different cofactors, making their functional sites distinct. Despite detailed analyses of Eve function in segmentation, no candidate co-factors for specifying target genes have emerged (Fujioka, 2012).

A recent study showed that CRM u8172 drives ectopic expression within odd-numbered parasegments in cells that normally do not express detectable levels of slp RNA. However, when combined with the promoter-proximal CRM u3125, which drives properly restricted expression within even-numbered parasegments, ectopic expression is repressed, suggesting that an Eve-responsive element resides within this region. Consistent with these findings, transgenes containing this region responded to ectopically expressed Eve, and rescue-type transgenes carrying u8172 without this region drove ectopic Slp, causing embryonic defects (Fujioka, 2012).

Recently, a striking number of distinct CRMs surrounding the slp1 transcription unit were found to drive expression that overlaps in both space and time. Extensive dissection of this regulatory region and rescue of slp mutants with various transgenes suggested that apparent redundancy may be necessary to provide fully functional levels of expression across the various stages of slp expression. This study shows that there are functionally redundant En/Exd/Hth binding sites within CRM u1523. In vitro binding analysis identified a strongly cooperative binding site and a weaker, but still highly cooperative site. Despite the apparent difference in in vitro binding affinity, either site is sufficient to confer repression in the En domain, and both sites must be mutated to cause significant derepression. Thus, apparent redundancy exists at multiple levels in slp regulation. Whether apparent redundancy at this level has a function in increasing the robustness of functional gene expression within the organism, as does apparent redundancy among multiple enhancers regulating the same gene, remains to be determined. Furthermore, cooperativity with cofactors in vitro seems to be a significant indicator of function in vivo, in addition to affinity. It was found that while the B1b site has the same apparent affinity as A2a, A2a confers considerably stronger repression activity, and shows greater cooperativity in binding by En with Exd/Hth. The discrepancy between relative affinity and functionality may be attributed to the challenge of reproducing functional binding conditions in vitro, where protein–protein interactions leading to cooperativity may be less sensitive to the differences in conditions than are protein–DNA interactions. Relatedly, competition with a variety of DNA binding proteins in vivo for sites on the DNA may lead to a greater reliance on cooperativity in vivo for occupancy of functional sites (Fujioka, 2012).

Previous studies indicated that En requires the Hox co-factors Exd and Hth to efficiently repressslp, especially in the anterior half of the embryo, and En was found to act cooperatively on target sites in the distalless gene with both Exd/Hth and posteriorly-expressed Hox gene products. Although it remains possible that the relatively weak, yet functional binding site (A2a) within i1523 might bind En with other cofactors in addition to Exd/Hth, dissection and construction experiments with this and other sites have not revealed any clear anterior–posterior differences in their activity that might suggest a functional interaction with cofactors such as Hox proteins that are restricted in expression along the anterior–posterior axis. Nonetheless, previous studies suggested that regulation of slp by En might utilize posterior-specific factors. Further analysis will be required to more fully explore this possibility (Fujioka, 2012).

The relative arrangement of consensus En and Exd sites that facilitate cooperative binding appears to be quite flexible. For example, the A2a site contains no canonical consensus core for En binding (ATTA), while for the other two functional sites, the distance between the centers of the En and Exd sites is 10–12 bp for A1a and only 2 bp for B1a. The latter is reminiscent of En–Exd/Hth binding in distalless, where simultaneous Hox binding occurs, although the position of the En site is on the opposite side of the Exd core consensus ATCA. This relative arrangement of En and Exd sites (En binding 5′ of the Exd core ATCA) is seen for all of the functional sites analyzed in this study. This arrangement is similar to the relative positions of Hox and Exd binding to sites where there is no En involvement. The flexibility overall is consistent with that seen for Exd/Hth binding in conjunction with the Hox gene products, and suggests that while homeodomain family transcription factors are able to function combinatorially in vivo on a wide variety of binding sites, there are significant constraints on the positions of contact by the individual homeodomains. A full understanding of the similarities and differences between En binding in conjunction with Exd and Hth, and Hox binding with these cofactors, will require further investigation (Fujioka, 2012).

The highly cooperative, strong En/Exd/Hth binding site B1a was both necessary and sufficient for repression of u4734 in the En domain. However, it did not fully substitute for the entire repression element that contains it, located between −3.9 and −3.4 kb from the slp1 TSS. This finding suggests that there may be other functional En binding sites in this region. Consistent with this, in vitro binding suggested that other subregions (B2 and/or B3) harbor some binding activity. Thus, like i1523, there may be partial redundancy in En complex binding within u4734, despite the existence of a single essential binding site (Fujioka, 2012).

This study has established the functional significance of three cooperative En/Exd/Hth binding sites within slp. Interestingly, two of them are well conserved among the 12 species of Drosophila whose genomes have been sequenced, and the other site is conserved within the more closely related species. The duplication that generated the twin slp transcription units apparently took place before the divergence of these 12 species, as all drosophilids (but not mosquitoes) contain two tandem slp-related protein coding regions. This might suggest that the two conserved En/Exd/Hth sites were duplicated along with the locus as a whole. It has been shown that Drosophila enhancers contain clusters of conserved sequences blocks, and the two CRMs analyzed in this study contain such conserved sequence clusters. However, the patterns of conservation in the regions surrounding the conserved En/Exd/Hth sites do not suggest that they are directly related to each other. Furthermore, both CRMs are more closely linked to slp1 than to slp2. Clearly, there have been other chromosomal rearrangements in the history of the slp locus, precluding a simple description of its evolution (Fujioka, 2012).

A recent study investigating the genome-wide distribution of En binding showed a peak on i1523, but not on u4734. The data were derived from 7–24 h-old embryos, which were mostly at later stages than those at which these CRMs are active. In addition, the data show peaks where our analysis has not identified functional CRMs. Such sites may function to assist those within the core enhancer regions, or they might be functional during larval or adult stages to keep slp in the off state. Alternatively, they might not be functionally important. Further study will be required to address these issues (Fujioka, 2012).

Hox proteins coordinate peripodial decapentaplegic expression to direct adult head morphogenesis in Drosophila

The Drosophila BMP, decapentaplegic (dpp), controls morphogenesis of the ventral adult head through expression limited to the lateral peripodial epithelium (P e) of the eye-antennal disc by a 3.5 kb enhancer in the 5' end of the gene. A 15 bp deletion mutation within this enhancer was recovered that identified a homeotic (Hox) response element that is a direct target of labial and the homeotic cofactors homothorax and extradenticle. Expression of labial and homothorax are required for dpp expression in the peripodial epithelium, while the Hox gene Deformed represses labial in this location, thus limiting its expression and indirectly that of dpp to the lateral side of the disc. The expression of these homeodomain genes is in turn regulated by the dpp pathway, as dpp signalling is required for labial expression but represses homothorax. This Hox-BMP regulatory network is limited to the peripodial epithelium of the eye-antennal disc, yet is crucial to the morphogenesis of the head, which fate maps suggest arises primarily from the disc proper, not the peripodial epithelium. Thus Hox/BMP interactions in the peripodial epithelium of the eye-antennal disc contribute inductively to the shape of the external form of the adult Drosophila head (Stultz, 2012).

dpp expression in the lateral PE of the eye-antennal disc is necessary for correct morphogenesis of the adult Drosophila head. This study shows that dpp expression related to ventral head formation is part of a Hox/BMP genetic network restricted to the PE of the eye-antennal disc. The homeotic gene lab, and its cofactors hth and exd positively regulate PE dpp expression. This is supported by the observation that Lab, Exd and Hth bind in vitro to the dpphc enhancer and the consensus sites for these factors are required in vivo for expression. In addition, individually, lab and hth are both genetically necessary and together demonstrate sufficiency for expression from dpphc enhancer, as shown from both LOF and GOF clonal analyses. lab exerts positive control over Dfd expression, as indicated by loss of Dfd expression in lab LOF clones. In contrast, Lab is ectopically expressed in Dfd LOF clones, demonstrating that Dfd represses lab in domains of its expression. Dpp signalling is genetically required for the transcription of lab, as expression from a lab reporter construct is reduced in tkv LOF clones. Expression of a hth enhancer trap increases in tkv LOF clones, and is reduced when activated Tkv is ectopically expressed, indicating that hth transcription is negatively regulated by Dpp signalling. Finally, dpp directly autoregulates its own expression (Stultz, 2006), and may be spatially limited to domains of signalling by repression by brk, as demonstrated by the ability of ectopically expressed Brk to repress expression from the dpphc enhancer. Lab and Hth (acting with Exd) activate the expression of dpp. Lab also contributes to the activation of Dfd, which when expressed, represses lab, acting as a switch to limit the extent of lab expression. It is envisioned that during disc development, lab initiates both dpp and Dfd, and when Dfd reaches a certain threshold level, it turns off lab, establishing the boundary between the two Hox proteins. However, while loss of Dfd is capable of derepressing lab throughout the disc, it does not do so to dpp, so further negative regulation must exist. brk may provide this repression to further ensure the lateral boundary of PE dpp through a potential AE element in the enhancer. These inputs collaborate to define the sharp boundary of PE dpp expression. The level of dpp transcription is positively modulated by feedback between lab and dpp and autoregulation of dpp, presumably through Mad/Med binding to the AE element. Negative feedback between dpp and hth provides a brake on expression; others may exist. For example, the inhibitory Smad protein, daughters against dpp is a target of peripodial Dpp expression (Stultz, 2006). It is presumed these interactions activate dpp expression rapidly but shut it down when a certain expression level is reached (Stultz, 2012).

The Hox response region represents one of what will likely be many inputs into the expression of this 3.5 kb enhancer. Another input, opa, is homologous to the Zinc Finger Protein of the Cerebellum or Zic family of transcription factors, and was identified due to its genetic interaction with dpps-hc mutations. Other transcription factors and signalling pathways display genetic interactions, and their contribution to PE dpp expression is being actively investigated, although it is noteworthy that lab, Dfd, hth, and exd are not among them. It is expected that many transcription factors and signalling pathways impinge on the dpphc enhancer. In this regard, the dpphc enhancer may resemble the dpp visceral mesoderm enhancer, another identified Hox target, where direct Ubx, Abdominal A, Exd, and Hth homeodomain inputs collaborate with the Fox-F-related factor binou, as well as Dpp and Wingless signalling to control gene expression. Enhancers that respond to signalling pathways often demonstrate characteristic behaviours: 'activator insufficiency', 'cooperative activation', and 'default repression', and the dpphc enhancer conforms to this model. No single activator is able to induce expression over the entire disc, as shown by GOF experiments. Ectopically expressed Lab produced activation only in close proximity to the domain of endogenous dpp, while Hth activated only in the PE of the posterior eye disc. Addition of two inputs together (Lab and Hth or Lab and Dpp signalling, activated over a much broader area. Only Opa has broad ability to activate on its own over the PE but only in concert with Lab was it able to activate outside the PE. Thus each activator is insufficient individually; the enhancer requires simultaneous cooperative inputs of multiple factors to produce correct spatial expression. Brk would provide the default repression, preventing Lab and Hth individually from successfully activating in the middle of the disc, away from domains of dpp activity (Stultz, 2012).

Based on the transcriptional inputs so far identified, it is proposed that activation is controlled on the lateral side at a minimum by Lab, Hth, Exd, Mad, Med, and Opa. In the middle of the disc, the presence of only Hth and Exd is insufficient to activate the enhancer, particularly over resident default repression provided by Brk. On the medial (future dorsal) side of the disc, Dpp and phosphorylated Mad expression are observed, controlled by an unknown area of the dpp gene. Lab, Hth, Exd, and Opa are expressed there as well, so an additional repressor was hypothesised to be needed that limits expression driven by the dpphc enhancer to the lateral side. In this model, Lab is the activity required for peripodial specificity, with its cofactors Hth and Exd, while Mad/Med and Opa act as necessary collaborative activators of the enhancer (Stultz, 2012).

At the nucleotide level, the Hox response element in and adjacent to the dpps-hc1 deficiency bears sequence homology to previously identified Lab response elements: the mouse Hoxb1 autoregulatory enhancer (b1-ARE), which also generates a lab-like pattern, dependent on lab and exd activity, in Drosophila, and the lab autoregulatory enhancer. Both these enhancers have binding sites for Hox (Hoxb1, Lab), PBC (Pbx, Exd) and MEIS (Prep, Hth) proteins. The orientation of the bipartite Exd/Lab site relative to the MEIS site is the same in these elements as seen in the dpphc Hox response element, and the relative spacing between the PBC/Hox and MEIS components is very similar. However, the dpphc Hox response element has a cluster of three overlapping Hth sites, two residing on the opposite strand, and an additional functional Exd site downstream of the Hth sites, as determined by its requirement for expression in vivo). The expression of mutated reporter constructs in vivo, as well as LOF analyses of lab and hth, indicate that Hth/Exd plays a more critical role in enhancer activity than does Lab, as neither mutations in the Lab binding site nor Lab loss-of-function within somatic clones completely extinguished expression. This suggests that there may be multiple ways that homeodomain transcription factors activate the enhancer, depending on the cellular context. It is noted that the expression driven by the dpphc enhancer actually manifests as two separate lines {see also Stultz, 2006b). The level of Lab associated with each of these lines is not equivalent, therefore the control of expression may be specific to each line. This would be reminiscent of a situation seen within dpp itself, where the Ubx responsive visceral mesoderm enhancer is activated by Ubx/Exd/Hth in parasegment seven, but only requires Hth/Exd for activation in parasegment three. The in vitro EMSA data further support this, as Hth and Exd bind synergistically to more locations within the enhancer than Lab. The TALE family homeodomain proteins function independently of Hox proteins in many contexts. An additional explanation for the apparent primacy of hth may be because it plays both direct and indirect roles on enhancer expression. Hth acts with the transcription factor Yorkie (Yki) as part of the Hippo signalling pathway, and the nuclear activity of Yki and Hth are required to specify the PE of the eye-antennal disc. In the absence of hth, the PE is incorrectly fated. This may effect early gene expression upstream of the Hox/BMP interactions described in this study, magnifying the genetic affect of hth (Stultz, 2012).

The Hox/BMP network described in this study plays a prominent role in the external appearance of the adult head, yet is restricted completely to the PE of the eye-antennal disc. The terminal mutant phenotypes of dpps-hc, Dfd, and lab have similarities, but are sufficiently distinct that additional targets for each must exist, and for the cell autonomous Dfd and lab, these targets must reside in the PE. Other signalling proteins such as Wingless and Hedgehog, and the Notch pathway ligands Serrate and Delta, are expressed in the PE of the eye-antennal disc. While some adult structures derive from the PE, and PE cells likely contribute to other adult structures, it is likely that much of the effect of the PE on head morphogenesis is via inductive interactions with the DP, either through secreted signalling molecules, or targeted cell protrusions. Based on the cuticular alterations seen in dpps-hc, Dfd, and lab mutations, such interactions are capable of exerting structural modifications on the final head shape. Dipterans demonstrate great variety in the external morphology of their heads often with sexually dimorphic alterations within a species. Much of this variety involves changes in the relative proportions of eye and head capsule tissue. BMP expression has been implicated in shaping the jaws of cichlid fish and the beak shape of finches, while dpp expression itself is correlated with the growth of beetle horns, a specialized cuticular structure of the head. It is speculated that the PE specific Hox/BMP network described in this study could be a motor for such types of shape change in the Drosophila species (Stultz, 2012).

Low affinity binding site clusters confer hox specificity and regulatory robustness

In animals, Hox transcription factors define regional identity in distinct anatomical domains. How Hox genes encode this specificity is a paradox, because different Hox proteins bind with high affinity in vitro to similar DNA sequences.This study demonstrates that the Hox protein Ultrabithorax (Ubx) in complex with its cofactor Extradenticle (Exd) binds specifically to clusters of very low affinity sites in enhancers of the shavenbaby gene of Drosophila. These low affinity sites conferred specificity for Ubx binding in vivo, but multiple clustered sites were required for robust expression when embryos developed in variable environments. Although most individual Ubx binding sites are not evolutionarily conserved, the overall enhancer architecture-clusters of low affinity binding sites-is maintained and required for enhancer function. Natural selection therefore works at the level of the enhancer, requiring a particular density of low affinity Ubx sites to confer both specific and robust expression (Crocker, 2015).

This study has demonstrated that the Hox protein Ubx regulates separate enhancers of the svb gene by binding, with its cofactors Exd and Hth, to clusters of low affinity binding sites. Combining in vitro and in vivo assays, experimental demonstration is provided of an affinity-specificity tradeoff for Hox proteins, such that enhancers that integrate Hox inputs to drive regionalized expression are unlikely to utilize high affinity Hox binding sites. Forced to utilize low affinity sites, enhancers have evolved to contain multiple binding sites to ensure regulatory robustness to genetic and environmental variations. Most individual Ubx-Exd sites have evolved rapidly, but evolution has conserved overall enhancer architecture, with clusters of low affinity sites (Crocker, 2015).

Homotypic clusters of transcription factor binding sites are pervasive in animal genomes and several models have been proposed to explain their existence. The current results provide experimental evidence that homotypic clusters of Hox binding sites can confer robustness to enhancers. This may reflect a more widespread phenomenon. Although many enhancers contain homotypic clusters with low affinity sites, previous studies have rarely detected changes in expression by deleting individual binding sites. However, these mutated enhancers have not been tested in variable environments. It is possible that many of these clustered sites confer regulatory robustness (Crocker, 2015).

It is useful to compare these results with previous studies that have demonstrated specific regulatory functions for homotypic clusters. For example, clustered binding sites in an enhancer of the Drosophila hunchback gene mediate cooperative DNA binding by Bicoid, which provides threshold-dependent enhancer activity. In other cases, clusters of homotypic binding sites act in a noncooperative manner to allow enhancers to respond in a graded fashion, for example to determine expression levels in response to transcription factor concentrations. It is worth noting that in these cases, where homotypic clusters mediate specific linear or nonlinear outputs, enhancers are bound by transcription factors that belong to small paralogous families: e.g., two paralogs for Msn2; three for p53; two for Dorsal; and five for NFκB. In contrast, there are 84 homeodomain-containing proteins encoded in the Drosophila genome, many with overlapping specificities. Therefore, in previously described examples of homotypic clusters, binding affinity may not be a strong constraint on specificity (Crocker, 2015).

For the Hox regulated svb enhancers, low affinity Ubx/AbdA-Exd binding sites enable specificity, while the clustering of low affinity sites confers phenotypic robustness. This is a fundamentally different constraint on clustered binding sites than observed in all previous examples. The affinity-specificity tradeoff, initially supported by computational analysis of in vitro data, was confirmed in vivo by progressively increasing the affinity of the Ubx-Exd binding sites. While replacement of low affinity sites with higher affinity sites always quantitatively altered enhancer activity, either positively or negatively, most higher affinity sites generated strong ectopic expression. This ectopic expression is driven, at least in part, by gaining the binding of additional Hox proteins, which are normally not involved in the regulation of these enhancers. Other studies have performed replacement of low affinity sites with higher affinity sites and, in some cases, they have also observed ectopic expression. These altered patterns of expression may reflect increased sensitivity of enhancers to the same transcription factor that binds to the wild-type enhancer. This study observed a similar effect for Ubx and AbdA-dependent upregulation of svb enhancers in the cells in which they are normally expressed. In addition, however, it was found that sites with higher affinity resulted in a reduced specificity, due to the binding of additional homeodomain proteins, such as Scr, to svb enhancers. Computational analyses suggest that this affinity-specificity tradeoff is a fundamental property of Hox proteins and would therefore influence the architecture of enhancers that must generate specific outputs in response to Hox factors. It is suggested that transcription factors that belong to other large paralog groups may exhibit a similar affinity-specificity tradeoff and that enhancers regulated by these factors may also exploit clusters of low affinity sites (Crocker, 2015).

The results help to explain previous difficulties with bioinformatic prediction of functional Hox binding sites, because low affinity sites are difficult to detect reliably. Indeed, the low affinity sites that implement Hox regulation within svb enhancers share little similarity with canonical Hox or Hox-Exd binding sites. Consequently, a very large number of seemingly disparate DNA sequences can confer low affinity binding for Hox proteins. If Hox-Exd sites are often clustered in the genome, then signals from genome-wide chromatin immunoprecipitation sequencing (ChIP-seq) will reflect binding to the entire cluster (as was observed) and the signals associated with individual low affinity sites may be difficult to discern from noise. Identification of important low affinity sites will require a change in computational approaches to analyzing genome-wide data. Currently, it is de rigueur to apply an arbitrary threshold to genome-wide data and then to analyze only signals above this threshold. This approach is likely to bias detection toward high affinity sites, whose functions may be distinct from those of clusters of low affinity sites (Crocker, 2015).

These findings provide insight into how different Hox proteins regulate specific target genes to generate phenotypic diversity across the anterior-posterior axis. One unanswered question is how the many low affinity DNA sequences, which appear to share little in common, are bound by the same Hox-Exd complex with apparently similar affinity. It is possible that variations in DNA shape (deviations from the structure of canonical B-DNA) influence Hox-Exd binding to low affinity sites. It remains unclear if very different sequences can adopt similar shapes, or whether instead the Hox-Exd complex can recognize a range of shapes. Resolution of this question will require structural studies of Hox-Exd complexes bound to a range of low affinity DNA sequences and quantitative analysis of their binding dynamics in vivo (Crocker, 2015).

extradenticle: Biological Overview | Evolutionary Homologs | Protein Interactions | 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.