Many proteins are targeted to proteasome degradation by a family of E3 ubiquitin ligases, termed SCF complexes, that link substrate proteins to an E2 ubiquitin-conjugating enzyme. SCFs are composed of three core proteins-Skp1, Cdc53/Cull, Rbx1/Hrt1-and a substrate specific F-box protein. The closest homologs to the human components of the SCF(betaTrCP) complex and the E2 ubiquitin-conjugating enzyme UbcH5 have been identified in Drosophila. Putative Drosophila SCF core subunits SkpA and Rbx1 both interact directly with Cu11 and the F-box protein Slimb. The direct interaction of UbcH5 related protein UbcD1 with Cul1 and Slimb is also reported. In addition, a functional complementation test performed on a Saccharomyces cerevisiae Hrt1p-deficient mutant shows that Drosophila Rbx1 is able to restore the yeast cells viability. These results suggest that Rbx1, SkpA, Cullin1, and Slimb proteins are components of a Drosophila SCF complex that functions in combination with the ubiquitin conjugating enzyme UbcD1 (Bocca, 2001).
Selective spatial regulation of gene expression lies at the core of pattern formation in the embryo. In Drosophila, localized transcriptional regulation accounts for much of the embryonic pattern. Properties of a newly identified gene, partner of paired (ppa), suggest that localized receptors for protein degradation are integrated into regulatory networks of transcription factors to ensure robust spatial regulation of gene expression. The Ppa protein interacts with the Pax transcription factor Paired (Prd) and contains an F-box, a motif found in receptors for ubiquitin-mediated protein degradation. In normal development, Prd functions only in cells in which ppa mRNA expression has been repressed by another segmentation protein, Even-skipped (Eve). When ppa is expressed ectopically in these cells, Prd protein, but not mRNA, levels diminish. When ppa function is removed from cells that express PRD mRNA, Prd protein levels increase. It is concluded that Ppa coordinates Prd degradation and is important for the correct localization of expressed Prd. In the presence of Ppa, Prd protein is targeted for degradation at sites where its mis-expression would disrupt development. In the absence of Ppa, Prd is longer-lived and regulates downstream target genes (Raj, 2000).
To gain further insight into the combinatorial regulation by Prd and Eve, a yeast two-hybrid screen was performed for a cDNA library derived from 0-12 hour old embryos, using as bait a 140 amino-acid fragment of Prd that included its homeodomain. From a total of 2.4 x 106 primary transformants, 22 classes of clones were identified by restriction analysis; mRNA in situ hybridization analysis of representatives from each class indicated that one of the cDNAs from the screen is expressed in a pair-rule pattern of stripes. This cDNA was named ppa. To test the specificity of the interaction of Ppa with Prd, the ppa cDNA was retransformed into yeast and tested using a mating assay against a panel of different baits, including Prd. Ppa interacts either with the original homeodomain-containing Prd fragment, or a fragment containing both the homeodomain and the Prd domain, but not with homeodomain-containing fragments of Ftz or Bicoid, or with several unrelated control baits. In addition, Ppa does not interact with Prd containing a Ftz homeodomain substitution, suggesting that the Prd homeodomain sequences are required for the protein interaction (Raj, 2000).
Sequence analysis of ppa provided insights into its possible functions. The Ppa open reading frame (ORF) contains 11 leucine-rich repeats (LRRs). These 20-29 amino-acid motifs have Leu residues at characteristic positions and have been implicated in protein-protein interactions. Indeed, the carboxy-terminal 92 amino acids of Ppa, encompassing three LRRs, is sufficient for the interaction with Prd in the two-hybrid screen. Sequence alignments indicate that the LRRs in Ppa are similar to those found in yeast glucose repression regulator 1 (Grr1), C. elegans CO2F5.7, an ORF of unknown function, and a human hypothetical ORF that has been named Ppa because it is 63% identical (78% similar) to the carboxy-terminal 392 amino acids of the Drosophila protein (Raj, 2000).
Like GRR1, C. elegans CO2F5.7, and the human ORF, Drosophila Ppa also contains an F-box motif amino-terminal to the LRRs. Previously characterized F-box proteins, including Grr1 and yeast Cdc4, have been shown to be receptors that target their substrates for ubiquitin-mediated protein degradation. These proteins interact through their F-boxes with Skp1, which associates with Cdc53/Cullin, forming an SCF complex (Skp1/Cullin/F-box). The SCF complex functions as a ubiquitin ligase enzyme (E3), which facilitates transfer of ubiquitin from a ubiquitin conjugation enzyme (E2) to the substrate. The F-box proteins provide a vital link between this machinery and specific substrates to be degraded, the substrate interaction typically being mediated through WD40 or LRR protein-interaction motifs within the F-box protein. Thus, the F-box proteins provide for specificity of substrate choice. Unlike Grr1, CO2F5.7 or other described F-box proteins, Ppa also contains a region rich in Ala, His and Pro, which is similar to Ala-rich domains observed in previously identified transcriptional repressor proteins, including Kruppel, Knirps, Eve and En. The presence of the F-box and Ala/His/Pro motifs suggests that Ppa might function as a receptor for protein degradation, or as a transcriptional co-repressor, or both (Raj, 2000).
The ppa mRNA is not detected in unfertilized embryos, suggesting that ppa is not expressed maternally. Uniform expression throughout the embryo is first detected at nuclear cycle 10, and gradually increases in intensity during cycles 11-14. Ppa expression diminishes in the pole regions during cycle 13. During cycle 14 and early gastrulation, the expression of ppa transformed into a pair-rule striped pattern with the formation of interbands within which ppa expression is lost. This is followed during germ-band elongation by splitting of the ppa stripes to generate a one-segment-repeated pattern of reiterated interbands. The ppa stripes do not have sharp borders. Expression of ppa is lost throughout the ventral region of the embryo, which contributes to the ventral furrow during gastrulation, presumably as a result of dorsoventral regulators. The ppa mRNA is localized in the basal regions of cells, in contrast to the apical localization of most pair-rule gene mRNAs (Raj, 2000).
To assess the possible functional relationships with Prd and Eve, embryo fillets were double-stained for ppa mRNA and Prd or Eve protein. During the early stages of cycle 14, when ppa expression is being restricted to stripes, there are significant levels of ppa expression overlapping the stripes of Prd protein. As cycle 14 proceeds, the posterior regions of the forming ppa stripes transiently overlap the anterior regions of the primary Prd stripes but, by early gastrulation, the Prd and ppa stripes are almost distinct. This transient but limited overlap in the expression of ppa and Prd is consistent with the model that Ppa negatively regulates Prd protein function (Raj, 2000).
Comparison of ppa mRNA with Eve protein shows almost reciprocal expression of the two genes (ppa interbands coincide with Eve stripes), raising the possibility that Eve might repress ppa expression, thereby giving rise to the ppa interbands. This interpretation is supported by examination of eve mutant embryos, which have uniform instead of striped ppa expression during cycle 14 and germ-band elongation. Moreover, adding back a transgene (P[eve.2,3,7] that expresses eve stripes 2, 3 and 7 in an otherwise eve mutant background, results in ppa interbands corresponding to these three Eve stripes (Raj, 2000).
The spatial expression and sequence of ppa suggest that Ppa might negatively regulate Prd, either by transcriptional co-repression or degradation of the Prd protein. To test these possibilities, ppa was ectopically expressed in the Prd-expressing cells to determine whether activation of en transcription or levels of Prd protein are affected. A transgene with the full ORF of ppa driven by an hsp70 promoter (hs-ppa) was introduced into embryos. Heat treatment of hs-ppa embryos during cycle 14 has pronounced effects. The odd-numbered, Prd-dependent, en stripes are weakened or completely absent, suggesting that Prd activation of these stripes is repressed. This is not observed in heat-treated wild-type embryos processed in parallel (Raj, 2000).
Because the Drosophila embryo develops very quickly, the segmentation gene products are expected to be short lived. This is indeed the case for those products examined and is also likely to be true for the Prd protein, perhaps even in the absence of ppa function. Indeed, it is difficult to assess whether Prd protein levels are reduced in hs-ppa embryos because of the fairly broad range of immuno-staining signals observed between different embryos, a problem inherent to the detection technique. To overcome this problem, ppa was ectopically expressed over only part of the embryo, so that the effects of ectopic ppa could be assessed relative to regions of the same embryo where ppa expression is normal. eve mutant embryos with a transgene P[eve.2,3,7] that expresses only eve stripes 2, 3 and 7 have well-formed ppa interbands at these locations. Thus, it is possible to compare Prd protein expression at stripe 2, which overlaps the ppa interband at eve stripe 2, with Prd expression at stripe 4, where ppa is expressed ectopically. Examination of prd mRNA signals in eve-;P[eve.2,3,7] embryos reveals strong expression of stripe 4 when compared with stripe 2, consistent with previous observations that eve represses prd transcription, thereby contributing to refinement of Prd stripes. In contrast, Prd protein signal at stripe 4 is significantly lower than at stripe 2, correlating with the ectopic ppa expression at stripe 4, and suggesting that Ppa regulates Prd protein levels. Even though there is 50% more mRNA signal at stripe 4 than stripe 2 after ppa upregulation, there is 25% less protein. Note that it is formally possible that the reduced Prd protein levels result from changes in genes other than ppa that are regulated by eve. Nevertheless, these analyses of hs-ppa and ppa mutant embryos suggest that regulation by ppa is responsible (Raj, 2000).
The equivalent analysis of wild-type embryos shows similar mRNA signals at stripes 4 and 2, whereas the Prd protein signal at stripe 4 is somewhat reduced compared with stripe 2, correlating with the residual ppa expression normally still present at gastrulation at the ppa interband corresponding to Prd stripe 4. This decrease in Prd protein is less pronounced than in eve-;P[eve.2,3,7] embryos, presumably because the difference in ppa expression at stripes 2 and 4 is smaller. To confirm the interpretation that Ppa regulates Prd protein levels, the stripe 4 to stripe 2 ratios for mRNA and protein would be expected to be similar in heat-treated hs-ppa embryos, in which ppa is expressed ectopically at both stripes 2 and 4. This was indeed observed, supporting the conclusion that ppa regulates Prd protein levels. Because Ppa has an F-box, this regulation is most likely through targeted protein degradation rather than translational repression. Consistent with these data, Western analysis of embryo extracts indicates that Prd protein levels are reduced by approximately 50% in hs-ppa embryos, as compared with the wild type (Raj, 2000).
To confirm the role of Ppa in Prd degradation, a small chromosomal deletion was generated that removes the Ppa ORF, starting 345bp upstream and ending 304bp downstream of the ORF. Consistent with the observation that high levels of ppa are normally only observed in regions where Prd function is not required, the homozygous ppa mutants survive to adulthood but with reduced viability and abnormal nuclear cycling. To analyze the mutants, advantage was taken of the normal anterior-posterior progression of ppa stripe development in wild-type embryos: at the gastrulation stage, the more anterior Prd stripes (for example, stripe 2) has little overlap with ppa expression, whereas the more posterior stripes (for example, stripes 4 and 6) still have significant overlap. When prd mRNA and protein signals at stripes 4 and 6 are measured relative to stripe 2 in the same embryos, the protein signals at stripes 4 and 6 are found to be significantly lower than the corresponding mRNA levels in wild-type embryos. In ppa mutant embryos, however, the mRNA and protein signals are similar, indicating that ppa normally reduces Prd protein expression (Raj, 2000).
Supporting the conclusion that Ppa mediates Prd degradation, Ppa was found to interact with Drosophila Skp1, the component of the protein degradation machinery that is predicted to link Ppa to the ubiquitin-mediated degradation pathway. Expressed-sequence-tag (EST) cDNAs for Drosophila Skp1 were identified from the Berkeley Drosophila Genome Project; yeast two-hybrid assays show that the Skp1 protein interacts with Ppa (amino acids 131-538, which lacks the Ala/His/Pro-rich region). As expected, the smaller fragment of Ppa (amino acids 447-538) originally identified in the two-hybrid screen does not interact with Skp1, presumably because it has no F-box. The interaction between Ppa and Skp1 was confirmed by co-immunoprecipitation analysis of yeast cell extracts. Immunoprecipitation of a hemagglutinin (HA) epitope-tagged Ppa fragment (amino acids 131-538) using anti-HA antibody also brings down LexA-tagged Skp1, which was detected with anti-LexA antibody. The Ppa-Skp1 interaction was also observed by immunoprecipitation with anti-LexA antibody, and the interaction of Ppa with Prd was also verified in these experiments (Raj, 2000).
This analysis of ppa function indicates that, when it is expressed ectopically in Prd-expressing cells, the levels of Prd protein diminish about twofold. A similar change in substrate stability (2-4-fold) is observed when GRR1, the yeast gene most similar to ppa, is mutated. It is also possible that, in addition to reducing Prd protein levels, Ppa might function as a transcriptional co-repressor, interacting with Prd to reduce its activation of en. Together, these two repression functions would ensure robust negative regulation of Prd in the Ppa-expressing cells (Raj, 2000).
Ppa is the first example of an F-box receptor localized in stripes. Loss- and gain-of-function analyses show that, while the presence of ppa expression in stripes is important for embryo development, it is the absence of ppa expression in the interbands that is crucial. Homozygous ppa minus mutants survive to adulthood, but show 50%-70% lethality, consistent with the fact that Ppa works in conjunction with the transcriptional repressors slp and run to localize Prd function. The partial lethality may be due to altered Prd expression or abnormal nuclear cycling. In contrast, in embryos with a functional ppa gene, it is absolutely essential that its expression be spatially regulated (by eve). Even basal expression of one of the hs-ppa transgenes (two copies without heat treatment) causes complete lethality, whereas the same transformant is not lethal before removal of an FRT cassette that blocks transcription. This predicts that cis-regulatory mutations in ppa causing loss of spatial regulation will have profound detrimental effects on embryogenesis, and this could also apply to the vertebrate homologs of ppa, which have such striking sequence similarity (Raj, 2000).
With the recent cloning of F-box proteins and the realization that they provide specific links between substrates and the protein degradation machinery, it has been predicted that F-box proteins would play important roles in development. Because F-box-regulated degradation normally depends on phosphorylation of substrates, localized action of signal transduction systems can, in principle, lead to localized protein degradation. This is likely to be the case for the signal-dependent localized degradation of Drosophila Cactus, a homolog of vertebrate IkappaB, whose degradation is a prerequisite for nuclear import of the Dorsal transcription factor (a homolog of NFkappaB) in the ventral portion of the embryo. Degradation of Cactus is mediated by the F-box protein Slimb (a homolog of ß-TrCP), which is also implicated in Wingless and Hedgehog pathways. In contrast to these signal transduction systems, the localized protein degradation in the Ppa system depends on spatially regulated expression of the Ppa F-box protein itself. By having its transcription regulated by a segmentation protein (Eve), and by targeting other segmentation proteins for degradation (Prd), the Ppa F-box protein forms an integrated link in the segmentation protein regulatory cascade that serves to strengthen the spatial refinement required for pattern formation. It is predicted that integration into transcriptional cascades may be a property of an important subfamily of F-box proteins, which, as suggested above, may also have recruited transcriptional repression functions to optimize their negative regulation of targeted transcription factors (Raj, 2000).
Ppa is the first example of a localized F-box receptor for protein degradation that works alongside transcription factors to ensure localized gene expression in the Drosophila segmentation cascade. These analyses suggest that Ppa targets the Prd transcription factor for degradation in cell rows in which Prd function is inappropriate, and that it is crucial that ppa expression is removed, through repression by eve, from cell rows in which Prd function is required for normal embryonic development (Raj, 2000).
Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.