dorsal


PROTEIN INTERACTIONS (part 2/2)

Interaction with transcription factors

The simultaneous reduction in the levels of DL and any one of several helix-loop-helix (HLH) proteins results in severe disruptions in the formation of mesoderm and neuroectoderm. The area of twistand snail expression in the presumptive mesoderm is severely reduced in dl-daughterless double heterozygotes. The same is true in dorsal/achaete-scute double heterozygotes. Certain triple heterozygous combinations essentially lack mesoderm as a result of a block in ventral furrow formation during gastrulation. HLH proteins that have been implicated previously in sex determination and neurogenesis (daughterless, achaete, and scute) are required for the formation of these embryonic tissues. Evidence suggests that DL-HLH interactions involve the direct physical association of these proteins in solution mediated by the rel and HLH domains (Gonzalez-Crespo, 1993).

Transactional activation by low levels of DL in lateral regions (the presumptive neuroectoderm) depends on cooperative DNA binding interactions between DL and bHLH proteins. The snail repressor blocks this interaction and restricts expression to the neuroectoderm (Jiang, 1993).

DL-bHLH interactions mediating gene expression in the neuroectoderm and mesoderm are fundamentally distinct. Close proximity of DL and bHLH binding sites is essential for the synergistic activation of gene expression in the lateral neuroectoderm, where there are diminishing levels of the DL regulatory gradient. In contrast, sharp on/off patterns of gene expression in the presumptive mesoderm do not require linkage of these sites. These results are consistent with two distinct modes of DL-bHLH synergy: cooperative binding to DNA (requiring linkage of sites) and synergistic contact of basal transcription factors (no linkage required) (Szymanski, 1995).

Grainyhead/NTF-1 is a cofactor in the repression of decapentaplegic (dpp) and zerknüllt by Dorsal. Repression by Dorsal appears to require accessory proteins that bind to corepression elements in Dorsal-dependent regulatory modules called ventral repression regions (VRRs). A corepression element in dpp is located within a previously identified VRR and close to essential Dorsal-binding sites. In the dpp VRR, one of the dpp response elements overlaps the binding site for a potential activator protein, suggesting that one mechanism of ventral repression may be the mutually exclusive binding of repressor and activator proteins. The Dorsal response element binding protein is identical to NTF-1/ Grainyhead, a factor originally identified as an activator of the Ultrabithorax and Dopa decarboxylase promoters (Huang, 1995).

In the same cell, one protein can activate some genes and repress others. Dorsal activates the twist gene and represses the zen gene in the ventral region of early embryos. A Drosophila HMG1 protein, called DSP1 (dorsal switch protein), converts Dorsal and NF-kappa B from transcriptional activators to repressors. This effect requires a sequence termed a negative ventral regulatory element (VRE), found adjacent to Dorsal-binding sites in the zerknüllt promoter and adjacent to the NF-kappa B-binding site in the human interferon-beta (IFN-beta) enhancer. Previous studies have shown that another type of HMG protein, HMG I(Y), can stimulate NF-kappa B activity. Thus, the HMG-like proteins DSP1 and HMG I(Y) can determine whether a specific regulator functions as an activator or a repressor of transcription (Lehming, 1994 and Ip, 1995).

Krü, snail, and knirps do not appear to require DNA binding cofactors to mediate transcriptional repression. The VRE contains binding sites for a number of proteins suggesting that the repressor may function as a complex. Some of these components may be redundant. One putative corepressor, DSP1 (dorsal switch protein 1), binds to a sequence within the VRE and functions together with DL or other REL proteins to repress transcription. DSP-1 may be necessary for optimal silencing, but does not appear to be necessary for VRE activity, because minimal VRE sequences lacking DSP-1 binding sites can still mediate ventral repression. Another protein, NTF-1/Elf-1, was recently reported to bind to the ventral repression elements of the zen and dpp promoters. However, as with DSP1, it is still unclear whether the NTF-1/Elf-1 protein itself is involved in VRE activity. Interaction of DL and corepressor require fixed spacing between DL and corepressor binding sites, as separation of DL and corepressor sites with a 5-bp spacer, abolishes the silencer activity, while a 10-bp spacer restores silencer activity (Cai, 1996).

The Dorsal Switch Protein (DSP1) was identified in a genetic screen for activities which convert Drosophila Dorsal protein from an activator of transcription to a repressor. DSP1 shares homology with the HMG-1/2 family and inhibits activation by the rel transcription factors Dorsal and NF-kappaB in transfection studies. DSP1 protein can act as a potent transcriptional repressor for multiple activator proteins. DSP1 binds directly to the TATA binding protein (TBP) and forms a stable ternary complex with TBP bound to DNA. DSP1 preferentially disrupts the DNA binding of TBP complexes containing TFIIA and displaces TFIIA from binding to TBP. Consistent with inhibition of TFIIA-bound complexes, DSP1 inhibits activated (but not basal) transcription reactions. The ability of DSP1 to interact with TBP and to repress transcription has been found to map to the C-terminal domain, which contains two HMG boxes (Kirov, 1996).

The Drosophila HMG1-like protein DSP1 was identified in yeast by its ability to inhibit the transcriptional activating function of Dorsal in a promoter-specific fashion. DSP1 as well as its mammalian homolog hHMG2 bind to the mammalian protein SP100B; SP100B in turn binds to human homologs of HP1. "nuclear bodies" (NBs), found in nuclei of mammalian cells, are known to contain SP100A and PML. SP100A is an autoantigen recognized by antibodies from patients suffering from primary biliary cirrhosis, and PML has been implicated to play a role in acute promyelocytic leukemia. Both SP100A and PML are up-regulated by interferons: overexpression of PML results in slow growth. HP1 is a Drosophila protein involved in transcriptional silencing. Each of these proteins represses transcription when tethered to DNA in mammalian cells. These results suggest how heterochromatin proteins might be recruited to specific sites on DNA with resultant specific effects on gene expression (Lehming, 1998).

DSP1, fused to a GAL4 DNA-binding domain and bound to DNA at GAL4 sites, functions as a transcriptional repressor in HeLa cells. The reporter used in this experiment bore five GAL4 sites upstream of DNA sequences, taken from the TK promoter, that bind the activators C/EBP and SP1. A GAL4 fusion bearing the complete DSP1 represses the reporter some tenfold in mammalian cells. A fusion protein containing the carboxyl half of DSP1, which includes both HMG boxes, represses even more efficiently than does the full length fusion, whereas the amino half of DSP1 increases activation, a result of unknown significance. Intact human HMG2 fused to GAL4 also works as an efficient repressor in mammalian cells. In all cases, repression is abolished by deleting the GAL4-binding sites. In no case is repression found in yeast, suggesting that the repression seen in higher eukaryotes requires proteins absent from yeast (Lehming, 1998).

To identify such proteins from higher eukaryotes, a yeast two-hybrid screen was performed by using GAL4-DSP1(178-393) as bait and challenging with a human cDNA library derived from B cells. The cDNAs were fused to DNA encoding rII, an activating region derived from GAL4. One interacting candidate was found, a previously sequenced protein called SP100B, a splice variant of the NB protein SP100A. SP100A is 479 amino acids long; SP100B contains 688 amino acids. Both proteins are identical in their first 476 amino acids. The cDNA clone represents amino acids 5-528 of SP100B, referred to here as SP100B. SP100B, but not SP100A, interacts with the carboxyl half of DSP1, as well as with human HMG2. The B-specific domain of SP100B (amino acids 477-528) is sufficient for the interaction with both DSP1(178-393) and hHMG2 in vitro. The fusion protein GAL4-SP100B, like GAL4-DSP1 and GAL4-hHMG2, works as a repressor in mammalian cells but not in yeast. The repression depends on the presence of the GAL4 DNA-binding sites in the reporter (Lehming, 1998).

A second two-hybrid screen was performed, similar to the first, except that in this case the bait was GAL4-SPl00B. Two strongly interacting candidates were recovered, each of which contained both a chromodomain (CD) and a chromo shadow domain (CSD). Both proteins, hHPlalpha and hHP1gamma, are homologs of the Drosophila protein HPl (also called Su[var]205). In a yeast two-hybrid assay system, hHPlalpha and hHP1gamma interact with SP100B. The interaction requires the CSD of hHP1 but not the CD. For SP100B, the interaction domain is located between amino acids 286 and 333, a region identical in both SP100A and SP100B. SP100B also interacts with itself: this self-association determinant is probably located in the N terminus of the protein. SP100B interacts with both hHP1 proteins in vitro. Both GAL4-hHP1 fusion proteins work as repressors when bound to GAL4 sites in mammalian cells. In all cases, repression is abolished by deleting the GAL4-binding sites. Like the interaction with SP100B, repression requires the chromo shadow and not the CD. GAL4-hHPl represses approximately tenfold more efficiently than does GAL4-DSP1(178-393) and fivefold more efficiently than does GAL4-SP100B. Thus the CSD of hHP1alpha, comprising 53 amino acids, bears two functions: it confers the repression function when fused to GAL4, and it interacts with SP100B (Lehming, 1998).

The establishment of mesoderm and neuroectoderm in the early Drosophila embryo relies on interactions between the Dorsal morphogen and basic-helix-loop-helix (bHLH) activators. Dorsal and the bHLH activator Twist synergistically activate transcription in cell culture and in vitro from a promoter containing binding sites for both factors. Somewhat surprisingly, a region of Twist outside the conserved bHLH domain is required for the synergy. The two N-terminal Gln-rich regions of Twist appear to mediate synergistic activation by Dorsal and Twist and are also required for binding to Dorsal in vitro. In Dorsal, the rel homology domain appears to be sufficient for synergy. The interaction between Dorsal and Twist does not appear to be of sufficient strength to yield cooperative binding to DNA. It is suggested that the interaction betweein Dorsal and Twist induces a conformational change in one of the factors that enables it to efficiently activate transcription (Shirokawa, 1997).

The primary developmental effect of Drosophila Creb-binding protein (CBP: Nejire) mutation involves the function of Dorsal. nejire mutants exhibit, at germband elongation, the clockwise or counter-clockwise twisting of the embryo, just behind the cephalic furrow, often with the posterior side down. The ventral and cephalic furrows appear normal, but the mesodermally derived internal tissues and a block of ectodermal cells are often missing. On the basis of this phenotype, the Drosophila CBP mutant was named nejire, which means 'twist' in Japanese. Mesoderm formation is crucial event that takes place during early embryogenesis. To initiate the differentiation of the mesoderm in Drosophila, multiple zygotic genes such as twist (twi) and snail (sna), which encode a basic-helix-loop-helix and a zinc finger transcription factor, respectively, are required. The transcription of these genes is induced by maternal Dorsal protein, a transcription factor that is homologous to the NF-kappa B family of proteins. Drosophila CBP mutants, devoid of both maternal and zygotic nejire expression, fail to express twi and therefore generate twisted embryos. This is explained by results showing that dCBP is necessary for Dorsal-mediated activation of the twi promoter. In vitro Dorsal has been shown to bind the N-proximal portion of Drosophila CBP in a phosphorylation independent manner. A region of Dorsal lying between amino acids 186 and 356 (a part of the Rel homology domain) is required for interaction with CBP (Akimaru, 1997). Further studies indicated that a CBP-Dorsal complex is formed on the twist promoter (Perkins, 1996).

Preferential genome targeting of the CBP co-activator by Rel and Smad proteins in early Drosophila melanogaster embryos

CBP and the related p300 protein are widely used transcriptional co-activators in metazoans that interact with multiple transcription factors. Whether CBP/p300 occupies the genome equally with all factors or preferentially binds together with some factors is not known. Therefore Drosophila CBP (nejire) ChIP-seq peaks were compared with regions bound by 40 different transcription factors in early embryos, and high co-occupancy was found with the Rel-family protein Dorsal. Dorsal is required for CBP occupancy in the embryo, but only at regions where few other factors are present. CBP peaks in mutant embryos lacking nuclear Dorsal are best correlated with TGF-β/Dpp-signaling and Smad-protein binding. Differences in CBP occupancy in mutant embryos reflect gene expression changes genome-wide, but CBP also occupies some non-expressed genes. The presence of CBP at silent genes does not result in histone acetylation. Polycomb-repressed H3K27me3 chromatin does not preclude CBP binding, but restricts histone acetylation at CBP-bound genomic sites. It is concluded that CBP occupancy in Drosophila embryos preferentially overlaps factors controlling dorso-ventral patterning and that CBP binds silent genes without causing histone hyperacetylation (Holmqvist, 2012).

By comparison of CBP-bound regions in 2-4 hour old Drosophila embryos to previously mapped transcription factors, an extensive overlap of CBP peaks was found with the key activator of dorsal-ventral patterning, the Rel-family transcription factor Dorsal. The genome-wide distribution of CBP was determined in embryos where Dorsal cannot enter the nucleus (gd7 mutants), and it was found that CBP peaks that overlap regions where Dorsal, but few other factors bind in wild-type are selectively reduced in gd7 mutant embryos. Instead, strong CBP-bound regions in gd7 mutants overlap best with regions bound by the Smad protein Medea, a mediator of Dpp-signaling. Signaling by the TGF-β molecule Dpp is exceptionally sensitive to a small decline in the level of CBP in Drosophila embryos. The current results are consistent with a function for CBP in the genomic response to Dpp-signaling (Holmqvist, 2012).

Less overlap of the CBP peaks is found with mapped activators of anterior-posterior patterning such as Stat92E, Fushi-tarazu (Ftz), Paired, Caudal, and Bicoid. Previous work has indicated that CBP may function as a Bicoid co-activator. When Bicoid and CBP are expressed in S2 cells, they can interact, and Bicoid-mediated activation of reporter genes in these cells is influenced by CBP levels. 43% of the 300 strongest Bicoid-binding regions overlap a CBP peak in wild-type embryos, indicating that CBP may participate in Bicoid-mediated activation in vivo. However, many of the Bicoid peaks are found in HOT regions that bind several transcription factors. Therefore, it may not be Bicoid that targets CBP to these sites. Furthermore, although the shape of the Bicoid gradient is slightly changed in embryos from the CBP hypomorph nej1, activation of Bicoid-target genes is not compromised by the decrease in CBP levels in nej1 embryos. Consistent with a non-essential function for CBP in Bicoid-mediated activation, there is no co-occupancy of CBP and Bicoid at the known target genes hb, otd, kni, and eve. Thus, although CBP may contribute to Bicoid-mediated activation of some target genes, it seems to make a more widespread contribution to Dorsal-mediated activation. In conclusion, both genetic and genomic evidence points to a particularly important function for CBP in controlling the two key events in dorsal-ventral patterning of Drosophila embryos, the Dorsal gene regulatory network and Dpp-signaling. Perhaps CBP serves to coordinate the Dorsal and Dpp pathways in dorsal-ventral patterning (Holmqvist, 2012).

In embryos where Dorsal cannot enter the nucleus, places were found where CBP occupancy is increased, unchanged, decreased or lost. Regions that are unchanged bind several transcription factors, evident in their high HOTness, indicating that in the absence of Dorsal, other factors maintain CBP binding at these sites. Surprisingly, regions where CBP binding is increased are even HOTer, and therefore associated with even more factors in wild-type embryos. Although many CBP peaks in the genome are found where also GAF binds, the regions where CBP occupancy increases in gd7 embryos are lacking strong GAF binding, despite their high HOTness. Perhaps binding of GAF to these sites is not compatible with proper regulation of the corresponding genes. Instead, many of these regions bind Medea and Dichaete, especially the places where CBP binding is strong already in wild-type. In gd7 embryos, Dpp/Medea-regulated genes are expressed in more cells, resulting in increased CBP signal. These data indicate that also Dichaete-regulated genes are more highly expressed in gd7 mutants, and that CBP-binding therefore increases at these regions (Holmqvist, 2012).

Unexpectedly, median gene expression level of genes associated with gd7 Up regions is high in wild-type embryos. Most genes associated with these regions increase in expression even further in the absence of Dorsal, in most cases probably due to an expansion in the number of cells expressing the gene. It is therefore expected that these CBP-binding sites would be situated in promoter regions, and the increase in CBP binding a consequence of increased gene activity. However, it was found that these sites are mainly found in intronic and intergenic regions associated with H3K4me1, a mark of transcriptional enhancers. This indicates that CBP becomes recruited to these enhancers to mediate gene activation, rather than passively associating with active gene regions (Holmqvist, 2012).

CBP occupancy in gd7 embryos is reduced at regions where only few factors bind. The bigger the reduction in CBP occupancy compared to wild-type, the fewer the factors that are associated with such a region in wild-type, i.e. the lower the HOTness of the region. CBP peaks that are reduced in gd7 embryos are much more common at regions where Dorsal binds in wild-type compared to other factors, consistent with a requirement for Dorsal in targeting CBP to chromatin. Although not all of the gd7 Down CBP peaks overlap the top 300 Dorsal-binding regions, 92% overlap Dorsal when all Dorsal-binding regions are considered. Peaks where CBP is reduced in gd7 embryos are found in several known Dorsal target genes, such as twi, brk, htl, and Mef2. Furthermore, 10 of the 20 strongest Dorsal peaks overlap a region where CBP binding is reduced in gd7 embryos. Together, these data show that in early embryos, chromatin binding of CBP to many sites in the genome is dependent on Dorsal (Holmqvist, 2012).

A number of genomic regions were found where CBP occupancy in gd7 embryos is reduced to a level approaching background, the gd7 Lost regions. These regions are mostly devoid of histone modifications and occupied by very few or none of the 40 transcription factors. The factors found at these regions bind at very low levels, indicating that they may not contribute to regulation of the corresponding genes at this stage of development. Further, most genes associated with the gd7 Lost regions are expressed at very low levels or completely silent. These CBP-binding regions may therefore represent regulatory sequences that are poised for subsequent activation. Consistent with this interpretation, mean expression of the corresponding genes increases at later stages of development. Why is CBP occupancy lost from these regions in gd7 embryos? Perhaps these genes are not, and will not be expressed in the dorsal ectoderm, and are therefore not associated with CBP in gd7 mutants that convert the entire embryo into dorsal ectoderm. Alternatively, CBP binding to these regions is dependent on Dorsal. Although binding is weak, Dorsal occupies many of these regions in wild-type. It is possible that even small amounts of Dorsal is sufficient and necessary for CBP recruitment to these sites, and that CBP binding is consequently lost in the absence of Dorsal (Holmqvist, 2012).

Although CBP occupancy is reduced predominantly at Dorsal-binding regions in gd7 mutant embryos, expression of Dorsal target genes is also altered. The decrease in CBP occupancy in mutant embryos may therefore be a consequence of transcriptional inactivity, rather than a lack of recruitment by Dorsal. Indeed, CBP occupancy is on average reduced at down-regulated genes and increased at up-regulated genes. Therefore, although Dorsal and CBP occupancy often coincide, Dorsal may not directly recruit CBP to regulatory DNA sequences. However, there are also places where CBP occupancy is reduced without a corresponding change in gene expression. One such example is at the promoter of the caudal (cad) gene, which is co-occupied by Dorsal and CBP but where CBP binding is reduced more than two-fold in gd7 embryos, although the gene continues to be expressed. Furthermore, Dorsal and CBP associate in vivo. It is believed, therefore, that Dorsal may directly recruit CBP to many sites in the genome (Holmqvist, 2012).

There are also genomic sites where CBP occupancy is not dependent on either Dorsal or gene expression. Several known Dorsal target genes, including sna, neur, ind and ths, continue to associate with CBP in gd7 embryos. Although in general, HOTness is major determinant of CBP occupancy, there is no big difference in HOTness of the Dorsal target gene regions where CBP-binding is reduced (e.g., twi, htl, brk) compared to Dorsal target gene regions where CBP binding is not changed (e.g., sna, ind, ths). What maintains CBP binding on these genes in the absence of Dorsal is not clear. Presumably, other factors recruit CBP to these sites in the absence of Dorsal, but no common factor was found for the regions where CBP binding is unchanged. It is noted, however, that GAGA-factor (GAF) associates with many of the CBP-binding regions in wild-type embryos, but much less with CBP-binding regions in gd7 embryos. It is possible that GAF contributes to the recruitment of CBP to chromatin (Holmqvist, 2012).

Dorsal is converted to a repressor when it binds in proximity to AT-rich sequences, and thereby prevents expression of dorsal ectoderm target genes in the neuroectoderm and mesoderm. Consequently, these target genes, e.g. dpp, zen, and tld, are activated in all cells of gd7 mutant embryos. As expected, CBP occupancy increases at these target genes in gd7 embryos, since more cells express the genes. The Zelda protein is a maternally contributed activator of these genes. It has been previously shown that in nej1 embryos containing reduced amounts of CBP, tld expression is diminished, whereas dpp and zen expression remains unaffected. It is possible, therefore, that more activators than Zelda contribute to activation of tld, zen, and dpp in the dorsal ectoderm. Until these factors are identified, it may not be possible to explain why tld expression is particularly sensitive to a reduction in CBP amount in early embryos (Holmqvist, 2012).

When Dorsal functions as a repressor, it recruits the Groucho co-repressor. The yeast Tup1 protein, which is related to Groucho, was recently shown to block recruitment of co-activators to target genes. By contrast, it was found that CBP continues to associate with the tld and zen genes in the neuroectoderm although they are being repressed by Dorsal/Groucho. Groucho binds the histone deacetylase Rpd3 (HDAC1), which may be important for repression. Indeed, it was found that when tld and zen are repressed by Dorsal in the neuroectoderm and mesoderm, the genes are hypoacetylated despite the presence of CBP (Holmqvist, 2012).

Contrary to the general trend, some genes recruit CBP even though they are silent. Why are these genes not activated? In the cases examined, histone acetylation is low despite the presence of CBP when the genes are not expressed. Since lysine methylation and acetylation are mutually exclusive, histone methylation was measured at CBP-bound regions and it was found that Polycomb-repressed H3K27me3 chromatin is present at Dorsal-target genes in some tissues where these genes are not expressed. Although H3K27me3-decorated chromatin restricts DNA accessibility, it was found that H3K27me3-chromatin does not preclude CBP binding, but restrains histone acetylation at these CBP-bound genomic sites. Interestingly, all histone acetylations that were measured are blocked by H3K27me3-chromatin, not only the mutually exclusive H3K27ac. This indicates that despite the ability of CBP to bind to genes enclosed in H3K27me3-chromatin, the histones are not accessible for acetylation by CBP and other HATs. The data are consistent with a model for Polycomb silencing that allows access of proteins and pol II to DNA, but that restrains pol II elongation. Perhaps high levels of histone acetylation are necessary for release of pol II from the promoter, for example by recruiting the bromodomain protein Brd4 that brings in the P-TEFb kinase to phosphorylate pol II (Holmqvist, 2012).

In cells depleted of CBP and p300, global levels of H3K18ac and H3K27ac are greatly diminished whereas other histone acetylations remain unaffected, suggesting that these are in vivo targets of CBP acetylation. CBP can also acetylate H3K56, which occurs in response to DNA damage. It was found that H3K18ac and H3K27ac levels do not always correlate with changes in CBP occupancy at Dorsal target genes, although H3K18ac levels are most similar to CBP abundance. In part, this can be explained by the presence of H3K27me3-chromatin, that precludes histone acetylation. However, in the neuroectoderm (Tollrm9/rm10 embryos), the twi promoter contains less histone acetylation than in the dorsal ectoderm (gd7 embryos) although H3K27me3 levels are reduced and CBP binding not decreased compared to dorsal ectoderm. Together, these results show that CBP's HAT activity is regulated by substrate availability, but that it may also be regulated by genomic context or signaling (Holmqvist, 2012).

Genome occupancy of CBP/p300 and H3K4me1 can be used to predict cis-regulatory DNA sequences. However, what fraction of regulatory sequences that can be identified in this way is not known. It was found that CBP binding to many known enhancer sequences that are active in early embryos is below the cut-off for high-confidence peaks, although average CBP occupancy was determined to be 1.73 times the genomic background at 97 previously described early embryonic enhancers. The results also show that CBP binding differs greatly between wild-type and mutant embryos, and that some gene regulatory networks rely on CBP to a much larger extent than others. Together, these results suggest that although CBP/p300 binding can be used to successfully identify transcriptional regulatory sequences, many enhancer sequences will be missed because they are not bound by CBP/p300 or bound at levels below criteria for high-confidence peaks. Even though mapping CBP/p300 binding in different cell-types will increase the number of putative regulatory sequences, it is anticipated that a substantial number of enhancers will require alternative strategies for their identification, e.g. genome occupancy of other HATs (Holmqvist, 2012).

In conclusion, this study shows that association of CBP with the genome is dependent on the number and types of transcription factors that bind the DNA sequence, that CBP preferentially associates with some gene regulatory networks, that CBP binding correlates with gene activity, but that CBP also binds silent genes without causing histone hyperacetylation (Holmqvist, 2012).

Dorsal and Groucho

The Dorsal morphogen acts as both an activator and a repressor of transcription in the Drosophila embryo in order to regulate the expression of dorsal/ventral patterning genes. Circumstantial evidence has suggested that Dorsal is an intrinsic activator and that additional factors (corepressors) convert it into a repressor. These corepressors, however, have previously eluded definitive identification. Via the analysis of embryos lacking the maternally encoded Groucho corepressor and via protein-binding assays it has been shown that recruitment of Groucho to the template by protein:protein interactions is required for the conversion of Dorsal from an activator to a repressor. Specifically, Groucho is required for the Dorsal, mediated repression of Zerknullt and Decapentaplegic. Groucho is not required for the spatially regulated expression of genes that are activated by Dorsal such as twist and snail. Groucho is therefore a critical component of the dorsal/ventral patterning system (Dubnicoff, 1997).

In the Drosophila embryo, Dorsal, a maternally expressed Rel family transcription factor, regulates dorsoventral pattern formation by activating and repressing zygotically active fate-determining genes. Dorsal is distributed in a ventral-to-dorsal nuclear concentration gradient in the embryo, the formation of which depends upon the spatially regulated inhibition of Dorsal nuclear uptake by Cactus. Using maternally expressed Gal4/Dorsal fusion proteins, the mechanism of activation and repression by Dorsal has been explored. A fusion protein containing the Gal4 DNA-binding domain fused to full-length Dorsal is distributed in a nuclear concentration gradient that is similar to that of endogenous Dorsal, despite the presence of a constitutively active nuclear localization signal in the Gal4 domain. Whether this fusion protein activates or represses reporter genes depends upon the context of the Gal4-binding sites in the reporter. A Gal4/Dorsal fusion protein lacking the conserved Rel homology domain of Dorsal, but containing the non-conserved C-terminal domain also mediates both activation and repression, depending upon Gal4-binding site context. A region close to the C-terminal end of the C-terminal domain has homology to a repression motif in Engrailed -- the eh1 motif. Deletion analysis indicates that this region mediates transcriptional repression and binding to Groucho, a co-repressor known to be required for Dorsal-mediated repression. As has previously been shown for repression by Dorsal, activation by Dorsal, in particular by the C-terminal domain, is modulated by the maternal terminal pattern-forming system (Flores-Saaib, 2001).

The results presented here show that just as Dorsal sites function in a context-dependent manner in the presence of endogenous Dorsal, so too do Gal4 sites function in a context-dependent manner in the presence of a Gal4/Dorsal fusion protein. When Gal4/Dorsal*/nt1 binds to multiple tandemly repeated Gal4 sites upstream of a core promoter, the result is activation. In contrast, when Gal4/Dorsal*/nt1 binds a modified dpp VRR in which two critical Dorsal-binding sites have been replaced by Gal4-binding sites, the result is repression. Thus, bringing Dorsal to its target sites is sufficient for both activation and repression -- the rel homology domain (RHD) itself need not be directly engaged with the DNA. Perhaps Dorsal, other DNA-bound repressors (the assistant repressors) and co-repressors such as Gro cooperatively assemble at the ventral silencer to form a 'silencesome'. As might be expected if silencer function required the assembly of such a complex, silencing by the zen VRR is crucially dependent upon the spacing between the sites for the DNA-binding proteins. Changing the spacing (by a non-integral multiple of the DNA helical repeat distance) severely abrogates silencing, presumably by rotating DNA-bound proteins onto opposite faces of the helix. Very similar spacing effects have been observed for enhancesomes (Flores-Saaib, 2001).

The co-repressor Gro, which is required for Dorsal-mediated repression, interacts with the Dorsal RHD. This finding is consistent with the observation that truncated forms of Dorsal consisting of little more than the RHD are able to mediate partial repression of target genes such as zen and dpp. However, the repression directed by the RHD alone is weak relative to that directed by full-length Dorsal and it is therefore not surprising to discover an additional Gro-interaction domain in Dorsal, this one in the CTD. Although the CTD is not conserved between Rel family proteins, the Dorsal-related immunity factor (Dif) can partially substitute for Dorsal during embryogenesis. In addition, patterning of the chick limb may involve the regulation by NF-kappaB of the vertebrate orthologs of Dorsal-target genes. Given these similarities in function, how is it possible to explain the apparent absence of the eh1-like repression domain from Dorsal-homologs such as Dif and NF-kappaB? One possibility is that Rel family protein-mediated transcriptional repression is of relatively minor importance to pattern formation. This is possible because other redundant mechanisms involving Short gastrulation (Sog)-family inhibitors exist to ensure that Dpp-orthologs will not be active at inappropriate positions along the dorsal/ventral axis of the metazoan embryo. The additional Gro-interacting repression domain in the Dorsal CTD may have arisen relatively recently, perhaps as an evolutionary adaptation to allow more complete or more reliable repression of dpp and other genes that interact with dpp to pattern the dorsal ectoderm (Flores-Saaib, 2001).

The studies presented here suggest that potent repression by Dorsal does require a region with homology to the eh1 motif. Thus, Engrailed and Dorsal may use a similar interface to recruit Gro. In this respect, it is interesting to note that Engrailed and Dorsal actually have a ~150 amino acid region of similarity, with the eh1 motif at the C-terminal end of this region. The similar region contains polyalanine stretches, which is a characteristic associated with other repression domains. Perhaps this extended region of similarity plays some role in repression beyond that played by the eh1 motif (e.g. the recruitment of another co-repressor). While Dorsal can function as either an activator or repressor, Engrailed and all other previously characterized repressors containing eh1 motifs appear to be dedicated repressors. The conserved phenylalanine in the eh1 domain is required for efficient Gro recruitment and transcriptional repression. The absence of this phenylalanine in the Dorsal motif could explain the ability of Dorsal to act as either an activator or a repressor depending upon binding site context. Perhaps this 'defect' in the Dorsal eh1 motif prevents Dorsal from recruiting Gro without help from other nearby DNA-bound repressor proteins (assistant repressors). In this respect, it is very interesting to note that Hairy family proteins, which are dedicated repressors, use a C-terminal WRPW motif to recruit Gro, while Runt family proteins, which can function as both activators and repressors, recruit Gro, at least in part, via a C-terminal WRPY motif. Perhaps the conversion of the C-terminal tryptophan to a tyrosine weakens Gro recruitment thereby allowing Runt family proteins to function as either activators or repressors depending upon binding site context (Flores-Saaib, 2001).

Consistent with previous experiments showing that the CTD contributes to transcriptional activation in Drosophila S2 cells and in vitro, this domain mediates activation in embryos. Transcriptional activation by the CTD may be mediated by the previously described interactions of this domain with TAFII110 and TAFII60. Interestingly, the deletion that removes the eh1-like motif and prevents repression by the CTD also results in reduced transcriptional activation. There are multiple possible explanations for this observation. Perhaps Gro has some role in activation in addition to repression. This is reminiscent of studies suggesting that Tup1, a possible yeast ortholog of Gro, functions in both activation and repression. Alternatively, it is possible that the activation and repression domains overlap in the CTD, but that they function via completely different co-regulators. If this is true, then one might expect the binding of the co-repressor and the co-activator to be mutually exclusive, thus ensuring that Dorsal cannot function at cross-purposes by simultaneously recruiting a co-activator and a co-repressor (Flores-Saaib, 2001).

When Gal4/CTD is targeted to the anterior end of the embryo, the resulting zone of repression does not include the anterior pole of the embryo. A key finding in the study of this phenomenon came with the discovery and analysis of capicua (cic), a gene that encodes an HMG-box family transcription factor. In addition to being required for terminal pattern formation, Cic is also required for efficient Dorsal-mediated repression. The finding that Cic appears to be degraded in response to Tor activation suggests that Cic may be a direct target of the terminal pattern forming system. Previous evidence also hinted at a role for the terminal system in modulating Dorsal-mediated activation. When an artificial anterior-to-posterior gradient of Dorsal is established in the embryo, activation of a reporter gene under the control of the proximal twi VAR does not extend to the anterior pole of the embryo. This effect has attributed to the possible presence of Tor response elements in the twi VAR. However, as reported here, even when activation is mediated by nothing but tandem Dorsal sites, this activation is still inhibited at the termini of the embryo by Tor. Likewise, Tor also blocks activation by Gal4/CTD through multiple Gal4 sites. Since these artificial reporters are unlikely to contain Tor response elements distinct from the Dorsal or Gal4 sites, it is likely that the Tor pathway interferes directly with Dorsal-mediated activation, either by modifying Dorsal itself or by modifying a co-activator required for Dorsal activity. Consistent with the possibility that Dorsal itself is the direct target of the terminal system, it has been found that elimination of Tor activity results in an increase in the lower SDS-PAGE mobility form of Dorsal. Because phosphorylation usually decreases SDS-PAGE mobility, this finding suggests that Tor activation might result in the dephosphorylation of Dorsal, either by inactivating a Dorsal kinase or by activating a Dorsal phosphatase. In addition to blocking the activation of Dorsal target genes directly, the terminal system also blocks their activation indirectly, since huckebein, a zygotic target of the terminal system, clearly directs sna repression at the poles. Thus, there appear to be multiple perhaps partially redundant mechanisms to ensure that mesodermal determinants such as twi and sna will not be inappropriately expressed at the poles (Flores-Saaib, 2001 and references therein).

The effect of a tor gain-of-function mutation on activation by Dorsal and the Gal4/Dorsal fusion is not what would be predicted based upon the simple idea that Tor inhibits Dorsal-mediated activation. Instead of resulting in a further retraction of expression from the pole of the embryo, the gain-of-function mutation causes no obvious change in the size of the anterior gap. In addition, this mutation results in an expansion towards the posterior of Gal4/CTD-driven activation and a broadening in the D4/lacZ expression domain. These findings appear to be consistent with a model in which Tor has two completely different effects on Dorsal-mediated activation, inhibiting it at the poles and strengthening it away from the poles. This is precisely what has been observed for the interaction between Bcd and the terminal system. Thus, the effects of Tor on activation may be very general. How Tor is able to function in these two opposite ways depending upon position in the embryo is not clear (Flores-Saaib, 2001 and references therein).

Regulation of Dorsal by calcium levels and PKA

Increased cytoplasmic calcium concentration and the expression of constitutively active Toll receptors can induce the relocalization of Dorsal. Activation of endogenous Protein kinase A and expression of wild-type Toll receptors, have only a marginal effect on the cellular distribution of the Dorsal protein. Treatment of cells with activators of Protein kinase C and radical oxygen intermediates, both of which activate Nuclear factor kappa B, also has little effect on Dorsal protein localization. Thus different threshold levels of Dorsal activation can be established by distinctly regulated signal transduction pathways (Kubota, 1993).

Control over the nuclear import of transcription factors represents a level of gene regulation integral to cellular processes such as differentiation and transformation. The Drosophila transcription factor Dorsal shares with other family members of rel oncogene a phosphorylation site for the cAMP-dependent protein kinase (PKA) located 22 amino acids N-terminal to the nuclear localization signal (NLS) at amino acids 335-340. This study examines the nuclear import kinetics of Dorsal fusion proteins in rat hepatoma cells in vivo and in vitro. Nuclear uptake is not only NLS-dependent, but also strongly dependent on the PKA site, where by using site-directed mutagenesis, the substitution of Ser312 by either Ala or Glu severely reduces nuclear accumulation. Either exogenous cAMP or PKA catalytic subunit significantly enhance the nuclear import of wild-type proteins both in vivo and in vitro. Using a direct binding assay, the molecular basis of PKA site enhancement of Dorsal fusion protein nuclear import was determined to be PKA site-mediated modulation of NLS recognition by the importin 58/97 complex. The physiological relevance of these results is supported by the observation that Drosophila embryos expressing PKA site Dorsal mutant variants are impaired in development. It is concluded that the Dorsal NLS and PKA site constitute a phosphorylation-regulated NLS essential to Dorsal function and able to function in heterologous mammalian cell systems, where phosphorylation modulates the affinity of NLS recognition by importin (Briggs, 1998).

In Drosophila, dorsal-ventral polarity is determined by a maternally encoded signal transduction pathway that culminates in the graded nuclear localization of the Rel protein, Dorsal. Dorsal is retained in the cytoplasm by the IkappaB protein, Cactus. Signal-dependent phosphorylation of Cactus results in the degradation of Cactus and the nuclear targeting of Dorsal. An in-depth study of the functional importance of Dorsal phosphorylation is presented. Dorsal is phosphorylated by the ventral signal while associated with Cactus, and Dorsal phosphorylation is essential for its nuclear import. In vivo phospholabeling of Dorsal is limited to serine residues in both ovaries and early embryos. A protein bearing mutations in six conserved serines abolishes Dorsal activity; it is constitutively cytoplasmic, and appears to eliminate Dorsal phosphorylation, but still interacts with Cactus. Two individual serine-to-alanine mutations have produced unexpected results. In a wild-type signaling background, a mutation in the highly conserved PKA site (S312) produces only a weak loss-of-function; however, it completely destabilizes the protein in a cactus mutant background. Significantly, the phosphorylation of another completely conserved serine (S317) regulates the high level of nuclear import found in ventral cells. It is concluded that the formation of a wild-type Dorsal nuclear gradient requires the phosphorylation of both Cactus and Dorsal. The strong conservation of the serines suggests that phosphorylation of other Rel proteins is essential for their proper nuclear targeting (Drier, 1999a).

Two of the serines that were changed to alanines are predicted to be phosphorylated by specific kinases. S79 is part of a predicted casein kinase II (CK II) site, and interestingly, this is the only mutant that retains wild-type function, suggesting that if the serine is phosphorylated, the modification is not essential. S312 is part of a PKA recognition site. This site may or may not be phosphorylated, but it is not the target of the signal-dependent phosphorylation that occurs on Dorsal. Rather, if S312 is phosphorylated, the modification may occur already in the ovary, and stabilize the protein. A serine (S317) that is not part of any known kinase recognition site is the likely target of the signal-dependent phosphorylation. This raises the question as to what kinase is responsible for the modification. Tube and Pelle are components of the dorsal-ventral signal transduction pathway that function downstream from the Toll receptor. They interact directly with Dorsal, albeit in the amino-terminal domain 1 of the Rel homology region (RHR). This interaction represents an essential step in the transduction of the ventral signal. Pelle is a serine-threonine kinase and may be directly responsible for the phosphorylation of Dorsal. However, as S317 is in domain 2 of the RHR, it is possible that additional kinases, such as the IB kinases, phosphorylate Dorsal and that Pelle controls their activity (Drier, 1999a).

Drosophila mediator complex is used by Dorsal

To decipher the mechanistic roles of Mediator proteins in regulating developmental specific gene expression and compare them to those of TATA-binding protein (TBP)-associated factors (TAFs), a multiprotein complex containing Drosophila Mediator (dMediator) homologs was isolated and analyzed. dMediator interacts with several sequence-specific transcription factors and basal transcription machinery and is critical for activated transcription in response to diverse transcriptional activators. The requirement for dMediator does not depend on a specific core promoter organization. By contrast, TAFs are preferentially utilized by promoters having a specific core element organization. Therefore, Mediator proteins are suggested to act as a pivotal coactivator that integrates promoter-specific activation signals to the basal transcription machinery (Park, 2001).

Previous studies in yeast and human cells have suggested that transcriptional activator proteins interact with Mediator complexes. The requirement of dMediator for the activated transcription in response to Gal4-VP16 indicates that dMediator may also serve as a binding target of transcriptional activators. Because several coactivators, such as TAFs and the GCN5 histone acetyltransferase (HAT) complex, have been suggested to interact directly with transcriptional activators, the relative binding affinities of these coactivator complexes with the VP16 protein were examined. After incubation of nuclear extracts with an excess of GST fusion protein beads containing either wild-type or mutant (Delta456FP442) VP16 activation domain, the supernatants were analyzed by immunoblotting with Abs against the components of the coactivator complexes. Almost all of the dMediator proteins in the nuclear extract (TRAP80, MED6, and Trfp) were removed by incubating with GST-VP16 but not with GST-VP16Delta456FP442. However, the amounts of dGCN5, dTAFII40, dTAFII250, and dTBP in the extract were not reduced at all by the incubation. When the proteins bound to the beads were analyzed, a large amount of dMediator was retained only in the GST beads containing the functional VP16 activation domain. The TFIID and dGCN5 HAT complexes did bind to the wild-type VP16 beads, but the amounts were less than 2% of the total amounts present in the extract. These data indicate that, among known transcriptional coactivator complexes, Mediator is most strongly bound to and most readily recruited to the activation domain (Park, 2001).

In addition to the model VP16 activator derived from herpesvirus, dMediator interacts with Drosophila transcriptional activators Dorsal and heat shock factor (dHSF). When dMediator complex was incubated with FLAG-Dorsal or GST-dHSF fusion protein beads, more than 20% of the dMediator input was retained specifically on the beads even after extensive washing. To extend this study to other sequence-specific transcription factors important for Drosophila development, dMediator was immobilized on protein G-agarose beads through anti-dSOH1 Ab and the binding of diverse 35S-labeled Drosophila transcription factors was examined. Bicoid, Krüppel, and Fushi-tarazu are retained specifically on the dMediator beads; Twist and Hunchback are not. Therefore, dMediator functions as a binding target for many, but not all, developmental specific transcription factors (Park, 2001).

To evaluate the requirement of dMediator for activated transcription in response to the Drosophila activator proteins that interact with dMediator, the ability of dMediator-deficient nuclear extracts to support transcriptional activation by the Dorsal and Gal4-dHSF proteins was examined. The addition of Dorsal or Gal4-dHSF to mock-depleted extract causes 20- and 25-fold increases, respectively, in transcription levels from the adenovirus early region 4 (E4) promoter linked to the appropriate DNA binding site. However, the level of transcriptional activation is reduced significantly (five- and three-fold activations, respectively) in nuclear extract that has been depleted by anti-dSOH1 Ab. Therefore, dMediator is absolutely required for transcriptional activation by all the activators tested. Addition of purified dMediator back to depleted extracts partially recovers activation by Dorsal and Gal4-dHSF in much the same way as it does in the case of Gal4-VP16. dMediator is not required for transcriptional repression by the sequence-specific transcription factor Even-skipped (Park, 2001).

dMediator is generally required for transcriptional activation from both TATA-containing and TATA-less promoters through direct communication with transcriptional activators. The function of dMediator seems to be exclusively related to sequence-specific transcription factors placed at upstream enhancer elements. However, the requirement of TAFs, or at least dTAFII250, in activated transcription appears to be redundant in the in vitro transcription system used and affected by such factors as the core promoter organization or nucleosomal structure of transcriptional templates. Several TAF components in the TFIID complex indeed have biochemical activities and structural motifs adequate for the recognition of specialized settings of transcription templates. For example, certain TAFs recognize the Inr and DPE sequences located in many Drosophila core promoters and increase the stability of TFIID-promoter interactions. In addition, TFIID contains dTAFII250, which has a HAT catalytic activity and also possesses a histone octamer-like module comprising the histone H2B-, H3-, and H4-like TAFs. Although not experimentally demonstrated, these TAFs may have some roles in the transcriptional regulation of nucleosomal templates (Park, 2001).

The sequence-specific transcription factors which interact physically with dMediator include VP16, Dorsal, dHSF, Bicoid, Krüppel, and Fushi-tarazu. These factors contain different types of activation domains (acidic and glutamine-rich domains). Most of these transcription factors have been shown to activate transcription either constitutively or inducibly. It is noteworthy that dHSF interacts with and requires dMediator for transcriptional activation because previous reports have shown that transcriptional activation by HSF in yeast does not require the function of the Mediator protein Srb4. However, the recent finding that activation by HSF depends on another Mediator protein, Rgr1 (Trap170), suggests that some function of Mediator is required for HSF-mediated transcriptional activation in yeast, as well. Since Rgr1, but not Srb4, is conserved between yeast and Drosophila, transcriptional activation by HSF might utilize the conserved Rgr1 components of the Mediator complexes (Park, 2001).

Although some human Mediator complexes appear to have a negative effect on activated transcription, dMediator does not exhibit such an activity in an in vitro transcription system reconstituted with Drosophila transcription factors. In addition, Even-skipped, a well-known Drosophila transcriptional repressor, does not interact with, or depend for its transcriptional repression on dMediator. Previous reports have shown that the repression domain of Even-skipped directly targets TBP. It has also been confirmed that the TFIID complex in the nuclear extract specifically interacts with Even-skipped under the same conditions in which Even-skipped fails to interact with dMediator. Although Krüppel has a well-characterized repressor function in Drosophila development, it can also act as a transcriptional activator under certain conditions. Therefore, it is more plausible that the dMediator-Krüppel interaction observed is a part of the mechanism for transcriptional activation rather than transcriptional repression. Taken together with the fact that dMediator is dispensable for basal transcription, the lack of defect of the dMediator-depleted nuclear extracts on transcriptional repression by Even-skipped protein suggests that dMediator is required mainly for the mediation of transcriptional activation signals to the basal transcription machinery. Very recently, developmental roles of certain dMediator proteins found in the Drosophila genome database have begun to be also identified in genetic studies. Genetic interactions between dMediator proteins and a homeotic regulator Sex combs reduced implicate dMediator proteins as a transcriptional activator-specific target critical for Drosophila development (Park, 2001).

Like yeast Mediator, dMediator bind with the CTD repeats of Drosophila Pol II. This implies that though dMediator was purified separately from Pol II, these two complexes indeed interact with each other and act together during transcriptional initiation. Besides the physical interaction with Pol II, dMediator also has some binding affinity for TBP, TFIIB, TFIIE, TFIIF, and TFIIS. Such interactions may be involved in the regulation of Pol II preinitiation complex assembly. Related with this idea, it has been reported that in yeast, recruitment of general transcription factors such as TBP, TFIIB, and TFIIH to active promoters requires the function of Mediator. Also, TFIIE interacts with the Mediator protein Gal11. Further analyses will be required to clarify whether these interactions, observed both in yeast and Drosophila, participate in the control of the stepwise preinitiation complex assembly in the course of transcription activation or simply reflect the affinities between the components of preassembled Pol II holoenzymeG (Park, 2001).

dMediator contains the protein kinase component Cdk8, which can phosphorylate serine residues in the CTD. This catalytic kinase subunit seems responsible, at least in part, for the Pol II phosphorylation by dMediator. In particular, dMediator and TFIIH synergistically phosphorylate the serine 5 residue of the carboxy-terminal Pol II repeats, suggesting the presence of a functional interaction between these complexes. Given that Pol II phosphorylation at serine 5 by TFIIH has been correlated with transcriptional activation processes, the synergy in the serine 5 phosphorylation by TFIIH and dMediator may be intimately linked with the regulatory effects that the Mediator complex exerts on Pol II transcription (Park, 2001).

The MADF-BESS domain factor Dip3 potentiates synergistic activation by Dorsal and Twist

The transcription factors Dorsal and Twist regulate dorsoventral axis formation during Drosophila embryogenesis. Dorsal and Twist bind to closely linked DNA elements in a number of promoters and synergistically activate transcription. A novel protein named Dorsal-interacting protein 3 (Dip3) has been identified that may play a role in this synergy. Dip3 functions as a coactivator to stimulate synergistic activation by Dorsal and Twist, but does not stimulate simple activation of promoters containing only Dorsal or only Twist binding sites. In addition, Dip3 is able to bind DNA in a sequence specific manner and activate transcription directly. Dip3 possesses an N-terminal MADF domain and a C-terminal BESS domain, an architecture that is conserved in at least 14 Drosophila proteins, including Adf-1 and Stonewall. The MADF domain directs sequence specific DNA binding to a site consisting of multiple trinucleotide repeats, while the BESS domain directs a variety of protein-protein interactions, including interactions with itself, with Dorsal, and with a TBP-associated factor. The possibility is assessed that the MADF and BESS domains are related to the SANT domain, a well-characterized motif found in many transcriptional regulators and coregulators (Bhaskar, 2002).

NF-kappaB, IkappaB, and IRAK control glutamate receptor density at the Drosophila NMJ

NF-κB signaling has been implicated in neurodegenerative disease, epilepsy, and neuronal plasticity. However, the cellular and molecular activity of NF-κB signaling within the nervous system remains to be clearly defined. This study shows that the NF-κB and IκB homologs Dorsal and Cactus surround postsynaptic glutamate receptor (GluR) clusters at the Drosophila NMJ. Mutations in dorsal, cactus, and IRAK/pelle kinase specifically impair GluR levels, assayed immunohistochemically and electrophysiologically, without affecting NMJ growth, the size of the postsynaptic density, or homeostatic plasticity. Additional genetic experiments support the conclusion that cactus functions in concert with, rather than in opposition to, dorsal and pelle in this process. Finally, evidence is provided that Dorsal and Cactus act posttranscriptionally, outside the nucleus, to control GluR density. Based upon these data it is speculated that Dorsal, Cactus, and Pelle function together, locally at the postsynaptic density, to specify GluR levels (Heckscher, 2007).

NF-κB signaling has been implicated in the mechanisms of neural plasticity, learning, epilepsy, neurodegeneration, and the adaptive response to neuronal injury. The data presented in this study advance the understanding of neuronal NF-κB signaling in two ways. First, multiple lines of evidence are presented that NF-κB/Dorsal signaling is required for the control of GluR density at the NMJ. These data provide a synaptic function for NF-κB signaling that may be directly relevant to the diverse activities ascribed to NF-κB in the nervous system. Second, molecular and genetic evidence is provided that Dorsal, Cactus, and Pelle may function posttranscriptionally, at the postsynaptic side of the NMJ, to specify GluR density during postembryonic development (Heckscher, 2007).

Several independent lines of experimentation suggest that Cactus, Dorsal, and Pelle function together at the PSD to specify GluR density. Evidence is provided that Cactus and Dorsal localize to a similar postsynaptic domain. In addition, overexpression of a GFP-tagged Pelle protein that is sufficient to rescue a pelle mutation, can traffic to the PSD where Cactus and Dorsal reside. Next, genetic evidence is presented that cactus, dorsal, and pelle function together, in the same genetic pathway, to control GluR density. It is particularly surprising that mutations in cactus behave similarly to dorsal and pelle. In other systems (embryonic patterning and immunity), Cactus inhibits Dorsal-mediated transcription by binding and sequestering cytoplasmic Dorsal protein. As a result, in these other systems, cactus mutations cause phenotypes that are opposite to those observed in dorsal mutations. This study used the same cactus and dorsal mutations that previously have been observed to generate the predicted opposing phenotypes during embryonic patterning, and yet it was observed that cactus phenocopies the dorsal mutations. In addition, genetic epistasis experiments indicate that these genes function together to facilitate GluR density. Thus, at the NMJ, Cactus functions in concert with, rather than in opposition to, Dorsal (Heckscher, 2007).

One explanation for this observation could be that Dorsal does not function as a nuclear transcription factor during the control of GluR levels. In support of this idea it has been demonstrated that (1) Dorsal protein is not detected in the nucleus, (2) reporters of Dorsal-dependent transcription fail to show activity in muscle nuclei, and (3) mutation of the Dorsal transactivation domain, dlU5 does not impair GluR abundance even though this same mutation has been shown to impair transcription-dependent patterning during embryogenesis. An alternative explanation for the observation that dorsal and cactus have similar phenotypes at the NMJ could be that Cactus and Dorsal act synergistically to control the transcription of GluRs at the NMJ. Indeed, there is evidence in other systems that IκB can shuttle with NF-κB to the nucleus. A previous study shows Cactus accumulation in Drosophila larval muscle nuclei in a dorsal mutant background (Cantera, 1999). However, this result could not be repeated despite examination of Cactus localization in five allelic combinations of dorsal. Furthermore, the data from vertebrate systems suggest that IκB should shuttle into the nucleus with NF-κB, not in its absence. Thus, a model is favored in which Dorsal and Cactus function together at the postsynaptic membrane to facilitate GluR abundance during development (Heckscher, 2007).

If this model is correct, then it is predicted that NF-κB does not control GluR density through transcriptional regulation. This prediction is supported by two experimental observations: (1) GluR transcript levels (assessed by QT PCR) are not statistically different from wild-type in dorsal and cactus mutations that cause an ~50% decrease in GluR abundance; (2) it was demonstrated that overexpression of a myc-tagged GluRIIA cDNA using a heterologous, muscle-specific promoter is not able to restore synaptic GluRIIA levels in either the cactus or dorsal mutant backgrounds. These data are consistent with Dorsal and Cactus acting posttranscriptionally to control GluR density at the NMJ. There are two general mechanisms by which GluR levels could be controlled posttranscriptionally: (1) altered receptor delivery to the NMJ or (2) altered receptor internalization/degradation. If receptor internalization/degradation were enhanced in the cactus, dorsal, or pelle mutant backgrounds, one might expect GluRIIA-myc overexpression to overcome this change and restore normal receptor levels. In addition, less myc-tagged protein might be seen in the mutants in comparison to wild-type. This is not what was observed. Therefore, the hypothesis is favored that Cactus, Dorsal, and Pelle function together to promote the delivery of glutamate receptors to the NMJ during development (Heckscher, 2007).

The possibility that Cactus, Dorsal, and Pelle act posttranscriptionally to control GluR density raises many questions. For example, do Dorsal and Cactus exist as a protein complex at the PSD? If so, is this complex regulated and how might such a complex influence GluR density? Since pelle kinase-dead mutants impair GluR density, it is possible that Dorsal and Cactus recruit Pelle to the PSD. If so, what are the targets of Pelle kinase that are relevant to establishing or maintaining the proper density of glutamate receptors at the PSD? Finally, the demonstration that cytoplasmic NF-κB/Dorsal can influence GluR density does not rule out the possibility that NF-κB/Dorsal may also translocate to the muscle nucleus at the Drosophila NMJ under certain stimulus conditions. Indeed, in both the vertebrate central and peripheral nervous systems NF-κB is found within neuronal and muscle nuclei, and nuclear translocation can be stimulated by neuronal activity, glutamate, injury, and disease. For nuclear entry of Dorsal, two events must occur: (1) Cactus must be degraded and (2) Dorsal must be phosphorylated. It remains possible that one or both of these criteria are not met during the normal development of the Drosophila NMJ but could be met under as-yet-to-be-identified stimulus conditions. The possibility that NF-κB acts both locally at the synapse and globally via the nucleus is not unique to this signaling pathway. A similar organization has been documented for wingless/wnt signaling where noncanonical cytoplasmic signaling can impact cytoskeletal organization while canonical signaling involves the nuclear translocation of downstream beta-catenin and TCF-dependent gene transcription (Heckscher, 2007).

It remains unknown how NF-κB signaling is activated at the Drosophila NMJ. In Drosophila embryonic patterning and innate immunity, NF-κB signaling is initiated through activation of Toll or Toll-like receptors. There are nine Toll and Toll-like receptors encoded in the Drosophila genome. However, none of these receptors appear to be present in Drosophila larval muscle. The Toll receptor is expressed in a subset of embryonic muscle fibers, but is absent from larval muscle. None of the Toll-like receptors are expressed in Drosophila embryonic muscle and none appear to be expressed in larval muscle. An alternative possibility is that TNF-α receptors activate NF-κB in Drosophila muscle as has been observed in vertebrate skeletal muscle. Indeed, a TNF-α receptor homolog (Wengen) has been identified, and it is expressed in Drosophila skeletal muscle. The possibility that TNF-α signaling is mediated via NF-κB is intriguing given the recent demonstration that TNF-α regulates GluR abundance in the vertebrate central nervous system. In both cultured neurons and hippocampal slices glial-derived TNF-α signaling is required for the increase in postsynaptic AMPA receptor abundance observed following chronic activity blockade. Thus, the current data in combination with work in the vertebrate CNS raise the possibility that a conserved TNFα/NF-κB signaling system controls GluR abundance at both neuromuscular and central synapses during development and in response to chronic activity blockade (Heckscher, 2007).

The nucleoporin Nup214 sequesters CRM1 at the nuclear rim and modulates NFkappaB activation in Drosophila

CRM1-mediated protein export is an important determinant of the nuclear accumulation of many gene regulators. This study shows that the NFkappaB transcription factor Dorsal is a substrate of the exportin CRM1 and requires the nucleoporin Nup214 for its nuclear translocation upon signaling. Nup214 binds to CRM1 directly and anchors it to the nuclear envelope. In nup214 mutants CRM1 accumulates in the nucleus and NES-protein export is enhanced. Nup214 forms complexes with Nup88 and CRM1 in vivo and Nup214 protects Nup88 from degradation at the nuclear rim. In turn, Nup88 was sufficient for targeting the complex to the nuclear pores. Overexpression experiments indicated that Nup214 alone attracts a fraction of CRM1 to the nuclear envelope but does not interfere with NES-GFP export. By contrast, overexpression of the Nup214-Nup88 complex traps CRM1 and Dorsal to cytoplasmic foci and inhibits protein export and immune response activation. It is hypothesized that variation in levels of the Nup214-Nup88 complex at the pore changes the amount of NPC-bound CRM1 and influences the relative strength and duration of NFkappaB signaling responses (Xylourgidis, 2006).

The cytoplasmic filaments of the NPC extend ~40 nm from the nuclear envelope. Functional analysis of their constituents suggests a role in the cytoplasmic release of nuclear export substrates and the assembly and docking of nuclear import complexes. Nup214 and Nup88 homologues form complexes at the cytoplasmic side of the pores in yeast, amphibians and mammals but their roles in animal development and physiology remain elusive. This paper investigates the role of Nup214 in CRM1-mediated nuclear export in vivo and its functional relationship with its binding partner Nup88 in Drosophila larvae (Xylourgidis, 2006).

Nup214 does not play a key role in maintaining the NPC architecture. Staining with an antibody against several FG-repeat-containing nucleoporins did not reveal detectable abnormalities in their abundance or localization in nup214 mutants. The major structural component of the cytoplasmic filaments is RanBP2-Nup358. Owing to the lack of a specific detection reagent for RanBP2-Nup358, antibodies against RanGAP were used to indirectly assess the integrity of the cytoplasmic filaments in nup214 larvae. RanGAP binds directly to RanBP2(Nup358) and this binding is required for its accumulation at the nuclear rim. RanGAP was still localized along the nuclear envelope, but the staining appeared ~35% weaker and punctuated, suggesting that the Nup214-Nup88 complex does not play a major role in maintaining RanBP2(Nup358) on the cytoplasmic filaments (Xylourgidis, 2006).

The interdependence of Nup88 and Nup214 at the NPC however, has been consistently observed in all species in which it has been analyzed. Nup88 is undetectable in cells derived from Nup214-deficient embryos and RNAi inhibition of each of the genes in tissue culture results in reduced protein levels for the other. In addition, yeast cells with a temperature-sensitive mutation of Nup82p, show a reduction of Nup159p from the nuclear rim. The molecular mechanisms underlying this interplay remain unknown. Nup88 is undetectable in nup214 mutants whereas the levels of the nup88 transcript remain constant. In tissue culture RNAi experiments, and in nup214 heterozygotes and homozygous mutants, the reduction of Nup88 is proportional to the amount of Nup214. In addition, epoxomycin can inhibit the Nup88 reduction caused by the inactivation of Nup214. Since Nup214 and Nup88 bind to each other directly, the results suggest that Nup214 binding of Nup88 at the NPC protects it from proteasome degradation. This protection mechanism may involve interference with the degradation of Nup88 selectively at the pore, because Nup88 appears to be a stable protein in the cytoplasm when overexpressed (Xylourgidis, 2006).

In mbo/nup88 mutants Nup214 detaches from the nuclear rim and localizes within the nucleus. Overexpression of Nup88 in nup214 mutants, lacking endogenous Nup214 and Nup88, results in accumulation of the overexpressed protein at the nuclear rim indicating that Nup88 alone is sufficient to target the complex to the pore. The high levels of overexpressed Nup88 in nup214 mutants might be explained by a protective function of minimal residual amounts of Nup214 or by the inability of the protein degradation machinery to cope with the overproduced Nup88. The analysis suggests an intriguing posttranslational mechanism for the interdependence of the two nucleoporins. Nup88 alone is sufficient to associate with the NPC and this association is a prerequisite for the localization of Nup214 to the nuclear membrane. In turn, the binding of Nup214 increases the stability of Nup88 proteins at the nuclear envelope and may thereby increase the potential of additional Nup88 molecules to associate with the NPC. In sporadic cells of nup214 mutants that expressed high concentrations of Nup88 and low amounts of Nup214 after heat shock, the complex was localized along the nuclear envelope. By contrast, in cells expressing relatively low Nup88 and high Nup214 levels, localization of the complex became nuclear. This distinct distribution of the proteins correlating with the relative expression levels of the two Nups in different cells, suggests that localization of the complex is dynamic and depends on the relative concentrations of the two nucleoporins (Xylourgidis, 2006).

Nup214 binds to CRM1 and the cytoplasmic accumulation of the GFP-NES reporter is increased in nup214 mutants, whereas CRM1 is mislocalized from the NPC. These phenotypes suggest that Nup214 acts as an inhibitor of NES-mediated export. The physiological significance of the inhibitory function of Nup214 on CRM1 export is further emphasized by the phenotype caused by the reduction of Nup214 in crm1 mutants. Removal of one chromosomal copy of nup214 in emb2 mutants, which die as second instar larvae, allows these animals to proceed into pupariation and to develop adult structures. This extension of the life span of emb2 mutants suggests that anchoring of CRM1 to the NPC is a general mechanism that limits CRM1 activity. The endogenous substrates of CRM1 required for progression through the larval stages and pupariation remain to be identified (Xylourgidis, 2006).

The phenotypes of Nup214 mutants are identical to CRM1-export defects of mbo/nup88 larvae. However, overexpression of Nup88 alone in wild-type animals does not affect CRM1 localization and overexpression of Nup214 only causes a minor enrichment of CRM1 concentration on the nuclear rim and does not interfere with NES export. The co-expression of both nucleoporins under the control of the heat-shock promoter results in the gross mislocalization of CRM1 from the nuclear envelope and disrupted GFP-NES export. Thus, the Nup214-Nup88 complex is necessary and sufficient to tether a fraction of CRM1 and attenuate protein export (Xylourgidis, 2006).

Why would nuclear pore components negatively regulate CRM1 function? One possible explanation is that export complex formation depends on the levels of CRM1 in the nucleoplasm. The binding affinity of CRM1 to natural NESs varies, and cargoes with low affinity NESs may be exported less efficiently. The introduction of an artificial, high-affinity NES disrupts CRM1 export indicating that natural NESs are selected for their weaker affinity for the export factor. Removal of the Nup214-Nup88 sub-complex from the pore increases the nuclear concentration of CRM1 and it would also increase nuclear export of cargoes with low-affinity NESs. Tethering or releasing the NPC-bound fraction of the export factor may provide the means for controlling the nuclear concentration of proteins carrying low-affinity export signals (Xylourgidis, 2006).

Dorsal contains a functional NLS embedded in its Rel-homology domain. This sequence is sufficient to target a ß-galactosidase reporter into the nucleus and is required for the nuclear accumulation of Dorsal during embryogenesis. Owing to the phenotypes of nup214 larvae in the nuclear accumulation of import reporters, the fact that the defects of nup214 mutants in the activation of immune responses may be partly due to a reduction in the nuclear import levels of Dorsal and Dif cannot be excluded. The hypothesis is favored that the nup214 phenotype in NFkappaB translocation is primarily due to increased levels of protein export. Mutations inactivating Nup88, the partner of Nup214, disrupt NFkappaB translocation and show concurrent enhanced levels of NES-mediated protein export but do not exhibit any detectable effects on the nuclear import of the same reporters. In addition, this study identified a functional, CRM1-dependent NES, required for the cytoplasmic accumulation of Dorsal. This NES4 motif is deleted in hypomorphic dorsal alleles expressing truncated forms of the protein and causing an extended nuclear gradient in the embryo. These mutants still retain their Cactus-binding domain and their phenotypes become further enhanced by reduction of cactus activity, suggesting that CRM1 export is an additional novel determinant of Dorsal localization and activation in S2 cells and the embryo. The requirement of the Nup214-Nup88 complex for the full activation of the immune response and its inhibitory function on CRM1 export in larvae suggest that the amplitude and duration of Toll signaling may be influenced by the export rates of Dorsal and Dif. The interference of the complex with NFkappaB localization and activity upon overexpression, further suggests that changes in the relative amounts of the nucleoporin complex and the fraction of CRM1 bound to it, may provide a regulatory node for the nuclear concentration of Dorsal and Dif. Variations in the NPC-bound CRM1 pool could be accomplished in two ways: First by modifications of Nup88 or/and Nup214, which could influence their binding capacity to CRM1. Variations in the affinity of the nucleoporin complex for CRM1 could explain the changes in the amounts of co-precipitated CRM1, whereas the amounts of Nup88, Nup214 and CRM1 remain constant in extracts from 5-10 and 10-15 hour Drosophila embryos. Nup88 phosphorylation has been detected in Xenopus oocytes, and such modifications might influence the affinity of CRM1 for the complex. Alternatively, transcriptional control of Nup88 during fly development might also influence the levels of Nup214 and CRM1 in turn, at the NPC. The steady-state ratios of Nup88 to Nup214 have been determined by proteomic studies of yeast and rat liver NPCs and revealed a 2:1 and an 8:1 excess of Nup88, respectively. Changes of the wild-type stoichiometry by overexpression of the two nucleoporins leads to lethality in Drosophila larvae and apoptosis and G0 arrest in human cells. The zygotic expression pattern of Nup88 in Drosophila is tissue- and stage-specific. The interdependence of Nup214 and Nup88 at the NPC may provide an elegant titration mechanism that continuously monitors the structure and function of the Nup214-Nup88 complex and the amount of CRM1 bound to it (Xylourgidis, 2006).

back to Dorsal: Protein Interactions part 1/2


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

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

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