![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,
Maria Fink,1
Antonio Simeone,2,3,4 and
Thomas Czerny1*
Institute of Animal Breeding and Genetics, University of Veterinary Medicine, Veterinärplatz 1, A-1210 Vienna, Austria,1 CEINGE Biotecnologie Avanzate, Via Comunale Margherita 482, 80145 Naples, Italy,2 MRC Centre for Develeopmental Neurobiology, New Hunt's House, King's College London, Guy's Campus, London Bridge, London SE1 1UL, United Kingdom,3 Institute of Genetics and Biophysics A. Buzzati-Traverso, CNR, Via P. Castellino 111, 80131 Naples, Italy4
Received 8 May 2006/ Returned for modification 18 July 2006/ Accepted 27 September 2006
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Tle4 is one of the four full-length Groucho proteins in mammals (39, 63). The founding member of this conserved corepressor family is the Groucho gene of Drosophila. Groucho is known to play important roles in various developmental processes, including sex determination, segmentation, neurogenesis and dorsoventral patterning (22, 50). Groucho family members are characterized by a conserved N-terminal glutamine-rich region (Q domain) and a conserved C-terminal WD40 repeat domain (reviewed in references 15 and 26). The WD40 repeat domain binds to a variety of proteins (21, 32, 65). The Q domain is required for the homotetramerization of Groucho (17, 51, 62) but is also involved in several protein-protein interactions (9, 23, 43, 55). The central region of Groucho contains three less-well-conserved domains, the GP domain, the CcN domain, and the SP domain. The GP region is important for interactions with histone deacetylase HDAC1 (9, 16, 19), which results in a repressive chromatin structure, thus explaining the corepressor function of full-length Groucho proteins. In contrast, a truncated member of the family, Aes (amino-terminal enhancer of split) contains Q and GP domains but lacks WD40 repeats and does not interact with HDAC1 (9). When expressed in the same cell, inactive heteromultimers involving Aes proteins and full-length Groucho family members form. This dominant-negative function of Aes on Tle-mediated repression has been demonstrated in both in vitro and in vivo systems (6, 43, 57).
A number of developmentally important transcription factors contain a small peptide motif that was first identified in Engrailed proteins. This eh1 (engrailed homology region 1) motif is highly conserved through evolution and was initially characterized as a transcriptional repressor element (61). Groucho corepressor proteins were then determined to interact with eh1 motifs via WD40 domains located at their C termini (32). eh1-mediated interactions control various developmental processes, such as the dorsoventral polarization of the neural tube. Here, repression by Nkx proteins requires the association of Tle proteins with eh1-like motifs of these homeodomain factors (19, 48). Pax proteins also exert transcriptional repression by recruiting Tle family members via their so-called octapeptide, which shows extensive sequence similarity to the eh1 motif (23). On the other hand, Tcf/Lef transcription factors, the nuclear targets of canonical Wnt signaling, were shown to interact with the N-terminally located Q domain of Groucho proteins (57).
Contiguous expression and mutual repressive interactions of Otx2 and Gbx2 in the early embryo define the position of the MHB. During further development of this organizing center, the Groucho interaction partners Engrailed, Pax2/5/8, and Tcf/Lef play key roles. In this report, we show that also Otx2 and Gbx2 are associated with Groucho proteins. Using pull-down, coimmunoprecipitation, and colocalization assays, we demonstrate that Gbx2 is able to interact with Tle4 in vitro as well as in vivo. Similar to Otx2, Gbx2 contains an eh1-like motif, and we show here that both elements physically interact with the WD40 repeat domain of Tle4. Using a heat-inducible promoter for misexpression we found that these eh1 motifs function as Groucho-dependent repression domains in the MHB of medaka embryos. Together, these data indicate that Gbx2 as well as Otx2 can function as active repressors by recruiting corepressors of the Groucho/Tle family, thus offering a molecular explanation of how these homeodomain proteins repress the transcription of target genes in an essential organizer during vertebrate brain development.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
149-182) (52), which was subcloned from existing Otx2 constructs. For coimmunoprecipitation experiments, an N-terminal Flag tag and 18 copies of a myc tag were added. Green fluorescent protein (GFP) and DsRed were also fused N terminally. The glutathione S-transferase (GST)-Tle4 construct was described previously (23). The F65E mutation (61) in En2 was introduced by PCR. A medaka Gbx2 fragment was amplified by PCR from cDNA of 24- to 48-h medaka embryos with the primers 5'-GGAATTCGCAAGAAGTACCTGTCCTTA-3' and 5'-GCGAGAAGCTTCAGAGCTGCAACAGTTTAA-3'. For heat shock-induced expression in the embryos, a heat shock element (HSE)-driven promoter was used (7).
Cell lines and cell transfection assays. Monkey COS-7 kidney cells, human U2-OS osteosarcoma cells, and mouse NIH 3T3 fibroblasts were transfected in 24-well plates using polyethylenimine (8). For luciferase reporter assays, 300 ng of the Gal4-responsive construct lGC-luc (10) and 50 ng of the Renilla luciferase expression plasmid phRG-TK (Promega) were transfected. Luciferase activities were measured 48 h after transfection with a Dual-Luciferase reporter assay (Promega). The activity of firefly luciferase was standardized relative to the reference level of Renilla luciferase to normalize for transfection efficiencies. The expression levels of Gal4 fusion constructs were analyzed by Flag tag detection in Western blotting. The cellular extracts were obtained from upscaled transient transfection experiments in six-well plates. For colocalization assays, U2-OS cells were transfected with Lipofectamine 2000 (Invitrogen) according to the supplier's instructions. Cells were analyzed 48 h after transfection.
Coimmunoprecipitation and GST pull-down assays. Transiently transfected COS-7 cells were lysed after 48 h in buffer A (20 mM Tris-HCl, pH 7.9, 120 mM KCl, 5 mM MgCl2, 0.2 mM EDTA, 0.2% NP-40, 1 mM dithiothreitol, and 10% glycerol) and supplemented with a protease inhibitor cocktail (Sigma) and 1 mM phenylmethylsulfonyl fluoride. Lysates were incubated for 30 min on ice, cleared of cellular debris by centrifugation, and subsequently mixed with 20 µl of anti-Flag M2 affinity beads (Sigma) for 2 h at 4°C under constant rotation. Beads were extensively washed with buffer A. For coimmunoprecipitation of Tle4 with Gbx2 and En2, the concentration of KCl was increased to 500 mM. The precipitated proteins were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blotting using monoclonal anti-Myc (9E10) and anti-Flag M2 affinity beads (Sigma). GST pull-down assays were performed as previously described (23).
Fish strains and DNA injection. Medaka embryos and adults of the Cab inbred strain were used for all experiments (42). Embryonic stages were determined according to Iwamatsu (29). DNA was coinjected with the I-SceI meganuclease enzyme (64) into single blastomeres at the one- to two-cell stage. After injection, the embryos were incubated at 28°C. Heat treatments were performed for 2 h at 39°C (7), followed by an incubation at 28°C for 6 to 24 h.
Whole-mount in situ hybridization. Embryos were fixed in 4% paraformaldehyde-2PTW (2x concentrated phosphate-buffered saline at pH 7.5 and 0.1% Tween 20). Whole-mount in situ hybridization was performed at 65°C (53).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
To investigate this protein complex in vivo, we expressed a myc epitope-tagged Tle4 alone and with Flag-tagged Gbx2 in COS-7 cells. Gbx2 was then precipitated from cell lysates with a monoclonal anti-Flag antibody and analyzed for the presence of Tle4 by Western blotting. A low, but significant amount of myc-tagged Tle4 was detected (Fig. 2D). Therefore, Tle4 interacts with Gbx2 in vitro and in vivo.
The interaction of Gbx2 with full-length Tle4 depends on an eh1-like motif. Gbx proteins display extensive similarity throughout their amino-terminal region including a proline-rich domain of unknown function and a highly conserved motif of 10 amino acids located N-terminally to this domain. We identified this region as an eh1-like motif, present also in other classes of homeoproteins such as engrailed, goosecoid, Nkx/Nk, Msx/Msh, and Hesx/Anf (61). The eh1 motif mediates interactions with Groucho corepressors (21, 32, 48). This motif consists of a conserved seven-amino-acid core with divergent flanking residues. The core sequence, starting with a phenylalanine residue (Fig. 3A), is identical in mouse Gbx1 and Gbx2 proteins. The Drosophila homologue of Gbx, Unplugged, also displays low divergence. Within the core sequence, a single conservative amino acid exchange has occurred during evolution (position 5 in the alignment). Figure 3A shows a comparison of eh1-like core sequences present in selected proteins of different organisms. The eh1-like motif of Otx proteins is similarly conserved among vertebrates but also occurs in the invertebrate species Tribolium castaneum.
|
29N). In addition, we deleted the C-terminal region starting from amino acid 227 (227C) that includes the homeodomain (Fig. 1, Gbx2 227C and Gbx2 29N 227C), which we previously found to reduce the cytotoxic effects of overexpression. Coimmunoprecipitation assays resulted in a Tle4 signal for Gbx2 227C but showed a strong reduction for Gbx2 29N 227C which lacks the eh1-like sequence (Fig. 3B). As a control, we performed a coimmunoprecipitation experiment for mouse En2 proteins under the same conditions (data not shown). A strong interaction was detected for the wild-type protein but not for an F65E mutation within the eh1 motif (61). We therefore conclude that the eh1-like motif of Gbx2 is able to interact with Tle4 but at a lower level than that of En2.
Since physical interaction between Otx2 and Tle4 was recently demonstrated (52), we repeated the experiment using a homeodomain-truncated Otx2 construct lacking 100 N-terminal residues, similar to our strategy used for Gbx2 (Fig. 1, Otx2 100N). As expected, we detected Tle4 protein in the coimmunoprecipitate, which was significantly reduced for the Otx2 mutant protein (Otx2 100N 149-182) (Fig. 3C). A comparison of the precipitated bands (Fig. 3B and C) indicates that the affinity of Tle4 for Otx2 was below that for Gbx2. Taken together, these results demonstrate the importance of the eh1-like motif of Gbx2 and Otx2 for Tle4 interaction.
Development of a colocalization assay for sensitive detection of Tle4 interactions.
To independently verify the results of the coimmunoprecipitation experiments, we established a cell culture colocalization assay. We initially examined the subcellular distribution of a Gfp-Tle4 fusion protein and found it localized to the nucleus of U2-OS cells (Fig. 3D), as previously reported for Groucho proteins (63). Gbx2 and Otx2 were also localized to the nucleus when fused with DsRed protein (Fig. 3E) but, in contrast to Tle4, displayed a "speckled" pattern. When coexpressed with the homeodomain proteins, Tle4 reproduced the nuclear signal pattern of Otx2 and Gbx2, suggesting a physical interaction between each pair of expressed proteins (Fig. 3F to F" and H to H"). To test for the specificity of these interactions, we performed the same experiments with the eh1 mutant versions. Both Gbx2 29N and Otx2 149-182 failed to alter the GFP-Tle4 pattern (Fig. 3G to G" and I to I"). As references, we tested En2 and Pax5, which were previously identified as Groucho interaction partners. In both cases we observed a dramatic relocation of Gfp-Tle4 to produce the speckled nuclear patterns characteristic of both En2 and Pax5, alone. Likewise, mutations within the eh1-like motifs of both references failed to do this. Therefore, this assay detected eh1-specific interactions of Otx2 and Gbx2 with Tle4 in the nucleus.
Tle4 represses the transcriptional activity of Gbx2 and Otx2 in an eh1-dependent manner.
Gbx2 has been shown to operate both as a transcriptional repressor (28, 66) and activator (40). To investigate the functional consequences of its physical interaction with Tle4, we tested whether the corepressor can modify the repression of Gbx2 in cell culture. The basal activity of the Gal4-responsive reporter was increased after cotransfection with small amounts of a Gal4-VP16 fusion construct. Gbx2 weakly reduced the transcriptional activity of the reporter (partly due to positive effects on the expression of the Renilla luciferase reference construct [data not shown]). But when Gal4-Gbx2 was coexpressed with Tle4, the Gal4 promoter activity dropped considerably (sevenfold below basal expression), even though the Gal4-Gbx2 expression level was reduced due to the coexpression of Tle4 (Fig. 4A). The expression of Tle4 alone did not significantly alter reporter activity, showing that the repressive effect on the promoter is exerted specifically by Gal4-Gbx2. Furthermore, the combined effect of Gbx2 and Tle4 was not seen when the eh1 mutant of Gbx2 (Gal4-Gbx2 29N) was used. The effect of Gbx2 and Tle4 together became more pronounced when higher amounts of Gal4-Gbx2/Gal4-Gbx2 29N expression vectors (30 ng) were used (eightfold difference in reporter activity) (data not shown). In the absence of Tle4, both Gbx2 versions repressed Gal4 reporter gene activity equally (Fig. 4A). These data show that in cell culture, Tle4 mediates transcriptional repression of Gbx2 depending on its interaction with the eh1-like motif of Gbx2.
|
149-182) significantly increased the activity of the Gal4-responsive reporter (Fig. 4B). However, coexpression of Tle4 completely abolished the effect of wild-type Otx2, whereas the Otx2 protein lacking the eh1 motif was unaffected by the corepressor. This reduction may in part be caused by a reduced Gal4-Otx2 level (due to Tle4 coexpression, but note an even lower level upon coexpression of Tle4WD40, resulting in strong activation). Our cell culture experiments are therefore consistent with a Groucho-mediated repression conferred by the eh1-like domains of these two homeobox proteins.
The WD40 repeat domain of Tle4 is sufficient for the interaction with Gbx2 and Otx2.
After identifying the eh1-like sequences as Groucho interaction motifs, we next asked which domain of Tle4 mediates this contact. The eh1 region of Engrailed has been shown to interact with the WD40 repeat domain of Groucho (32). We therefore tested a Tle4 mutant lacking the carboxy-terminal region including the WD40 repeat domain in our reporter assays (Fig. 4A, Tle4 WD40). Whereas full-length Tle4 in the reporter assays reached only low levels of expression, the WD40 version was highly expressed in these experiments (data not shown). Nevertheless, the mutant Tle4 protein was not able to confer repressive activity on Otx2 and Gbx2 (Fig. 4A and B). These results suggest that the WD40 repeat domain is essential for full repression exerted by Gbx2 and Otx2 and the Tle4 corepressor, possibly by a direct interaction involving the eh1 motif.
We next investigated the potential physical interaction with the WD40 repeat domain using colocalization assays. In contrast to results with the full-length protein, overexpression of Gfp-Tle4WD40 resulted in the formation of small aggregates within the nucleus or closely associated with it (Fig. 5A). Such distributions are often observed for truncated proteins fused with GFP following overexpression (27). However, coexpressing Gal4 fused with the Tle4 Q domain (Gal4-Tle4 Q) distributed the GFP fusion protein again into a speckled pattern within the nucleus (Fig. 5B), suggesting that the two proteins interact at the Q domain. On the other hand, neither Otx2 nor Gbx2 coexpression had such an effect (Fig. 5C and D), indicating that the Tle4 protein lacks the corresponding interaction domain.
|
|
Expression of Gbx2 in medaka embryos. So far, no data are available on Gbx2 function in medaka; we therefore performed a BLAST search using the zebra fish Gbx2 protein sequence in a medaka draft genome database (UT Genome Browser; http://medaka.utgenome.org). A homologous gene appeared that was consistent with the results of a previous PCR-based screen for homeobox genes in medaka (41). We then isolated a cDNA fragment, including the homeodomain region, by reverse transcription-PCR and analyzed the expression pattern of the Gbx2 gene during medaka embryogenesis.
The results of whole-mount in situ hybridization experiments with wild-type embryos are shown in Fig. 6. In medaka, Gbx2 transcripts are first detectable at 70% epiboly as a patchy expression domain of two stripes at the level of the prospective hindbrain (Fig. 6A). At late neurula stages, Gbx2 is expressed in the neural tube within the hindbrain region (Fig. 6B). The most anterior part of this domain forms a ring within the neural tube (Fig. 6B to G, arrowheads). A stripe with a more posterior positioning also appears transiently during early somitogenesis (Fig. 6B and D, arrows) and is lost around the seven-somite stage (Fig. 6E and F). The anterior Gbx2-positive ring initially appears diffuse (Fig. 6C, arrowhead) but sharpens during early somitogenesis at its anterior boundary (Fig. 6D, arrowhead). This domain marks the anterior-most hindbrain in other species. Two-color in situ hybridization experiments show that at four somites, this boundary coincides with the posterior limit of the Otx2 territory, thus identifying the region as the MHB (Fig. 6C and D). In addition, Gbx2 is expressed in neural crest cells at the beginning of the neurula stage. Migration of Gbx2-positive neural crest cells during later stages generates a rapidly changing, diffuse pattern within the hindbrain (Fig. 6C to F). At the seven-somite stage, Gbx2 transcripts appear in the dorsal telencephalon and otic primordia (Fig. 6E and F) and are later detected in neurons of the spinal cord and the medial side of otic vesicles (Fig. 6H). The medaka expression domains of Gbx2 largely coincide with those documented in zebra fish (36, 56).
|
Injection of Gbx2 mRNA into zebra fish blastomeres was reported to result in reduced forebrains and eyes (36). Activation of the Gbx2 HSE expression construct at gastrulation strongly reduced the expression of this phenotype. Additionally, we observed an anterior shift of the caudal Otx2 limit on the dorsal side of the embryos (in 19% of the GFP-positive embryos) (Fig. 7B and Table 2). Endogenous Gbx2 expression domains followed the new Otx2 boundaries, indicating a complete shift of the MHB (Fig. 7B). These data are in agreement with reports demonstrating that Gbx2 represses Otx2 expression in the posterior midbrain (46, 66). The repressive effect of Gbx2 on Otx2 was lost when the Gbx2 mutant lacking the eh1-like domain (Gbx2 29N) was used for misexpression (Table 2), indicating that Tle is required for this repression. In 50% of the embryos, we observed ectopic expression of Otx2 in both nonneural (Fig. 7H) and neural tissue (data not shown). Otx2 ectopic expression became more pronounced when embryos were incubated to late stages. Taken together, these data indicate that Gbx2 represses Otx2 in a Tle-dependent manner.
|
|
In addition, we analyzed the repressive effect of Otx2 on fibroblast growth factor 8 (Fgf8). Secreted Fgf8 protein plays a pivotal role in isthmic organizer activity at early somite stages. In this region, Otx gene products are required to repress Fgf8 in nonexpressing tissues, thereby restricting the spatial expression of Fgf8 gene products (2). As expected, the overexpression of full-length Otx2 clearly reduced the expression of Fgf8 (Fig. 7F and Table 2). The Otx2 mutant lacking the eh1-like motif failed to down-regulate Fgf8 expression and even produced small, ectopic patches of Fgf8 transcripts (Fig. 7G and Table 2). Therefore, Otx2 represses Fgf8 in a Tle-dependent manner. This contrasts with its effects on Gbx2, which in our experiments appear to be independent of the corepressor.
Our coimmunoprecipitation and colocalization experiments with Otx2/Gbx2 and Tle4 suggested a medium affinity interaction of these proteins. Covalent bonding of the transcription factors in a fusion protein with the corepressor should, therefore, enhance Groucho-dependent repression effects. To test this hypothesis, we fused Tle4 with the DNA binding domains of Gbx2 (Gbx2HD-Tle4) and Otx2 (Otx2HD-Tle4) and misexpressed them in medaka embryos. Both fusion constructs exhibited repressive functions on the same target genes as the full-length proteins (Table 2), thereby further supporting a corepressor-mediated function. Furthermore, the repressive effects of Otx2 on Fgf8 and of Gbx2 on Otx2 were strongly enhanced by using Tle4 fusion constructs (Table 2 and see Fig. S1 in the supplemental material).
The function of Groucho/Tle proteins can be blocked efficiently by misexpression of the short family member Aes, which interferes with the function of the Q-domain (43, 55, 57). Similarly, the interaction of proteins with the WD40 repeats should specifically be blocked by overexpression of the WD40 domain. For the experiments, we selected a DsRed-WD40 fusion protein, which gave better results in colocalization experiments compared to the isolated WD40 domain. Misexpression of this protein indeed resulted in phenotypes reminiscent of Aes injection experiments in medaka (6; B. Bajoghli, N. Aghaallaei, D. Soroldoni, and T. Czerny. submitted for publication). The eyes and the midbrain were smaller (see Fig. S1 in the supplemental material) and heart looping was affected (data not shown). Injection of an eh1-deleted Gbx2 had resulted in ectopic activation of Otx2 (see above). We therefore tested whether dominant negative interference at the level of the WD40 interaction similarly resulted in a derepression of the endogenous Gbx2. Indeed, we observed ectopic activation of Otx2 in 12% of the embryos (n = 25) (see Fig. S1 in the supplemental material). Taken together, the in vivo experiments demonstrate the involvement of Tle proteins in the genetic interactions that form the isthmic organizer.
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Tle-dependent transcriptional repression of Gbx2 and Otx2. Groucho binding transcription factors can be classified according to their transcriptional properties. One group comprises strong repressors like Hairy-related basic helix-loop-helix proteins (50) and the homeodomain proteins En and Gsc (32, 33). Other DNA-binding partners for Groucho proteins are known to act as both transcriptional repressors and activators, such as Tcf/Lef, the nuclear targets of canonical Wnt signaling (14, 57). We found that Gbx2 acts as a repressor when fused to Gal4. A major part of this repression is mediated by Groucho/Tle proteins. On the other hand, induction of reporter gene activity has been reported for Gbx2 at the myelomonocytic growth factor minimal promoter of chick, which contains additional AP-1 binding sites (40).
In contrast, the Drosophila Dorsal protein is known to activate the transcription of specific target genes in vitro (20) but has also been shown to exhibit Groucho-dependent repression in vivo (22). The Dorsal protein therefore appears to intrinsically activate, but can repress transcription with the binding of specific cofactors (26). When we fused Otx2 to the heterologous Gal4 DNA-binding domain, it exerted transcriptional activation. This is in agreement with cell culture studies in which Otx protein could activate a promoter containing Bicoid target sites (52, 60). Interestingly, in vivo studies in Xenopus suggest that transcriptional repression by Otx2 is critical for positioning of the MHB (28). Thus, Otx2 appears to belong to a class of transcription factors that are converted from activators to repressors by the recruitment of Groucho proteins.
The interaction between Gbx2/Otx2 and Groucho proteins is mediated by eh1-like motifs and the WD40 repeat domain. The physical interaction between Gbx2 and Tle4 appears to be direct, since it is detected in assays using bacterially produced proteins and in vitro translated Gbx2 preparations. This association depends on an eh1-like motif of Gbx2, which is located within a conserved N-terminal domain. eh1-related motifs have been found in various classes of homeodomain-containing proteins, including En, Gsc, Nkx/Nk, Msx/Msh, Hesx/Anf, and UNC-4 and have been shown to be critical for both repression and Groucho recruitment (21, 32, 33, 48, 61, 68). Deletion of this motif in Gbx2 and Otx2 strongly reduced signals in coimmunoprecipitation assays and prevented redistribution of the proteins in colocalization studies. Moreover, it is essential for Groucho-dependent transcriptional repression of both Gbx2 and Otx2 (Fig. 4).
The WD40 repeat domain is thought to represent an ancient eukaryotic protein-protein interaction surface resulting from tandem repeats of a 40-amino-acid motif that form a ß-propeller structure (38, 49). The WD40 repeat domain of Groucho has been shown to participate in interactions with the eh1 motif of Engrailed proteins (31, 32) and is required for interactions involving Hesx1 (21) and Cbfa1/Runx2 (45). In yeast two-hybrid assays, the Six domain of Six3 interacts with the WD40 domain of Tle1 and Tle3 (37, 43). In our experiments, a Tle4 mutant lacking this domain was unable to mediate transcriptional repression by Gbx2 and Otx2 in cell culture and failed to associate with these transcription factors in colocalization assays. Similar arguments based on the absence of activity in mutants have been previously used to infer essential functions of the WD40 domain in other interactions (21, 32). Although coimmunoprecipitation experiments using the isolated domain so far failed (21), we detected a WD40-mediated interaction for both Gbx2 and Otx2. The success of these experiments might be attributed to an increased detection sensitivity using the 18-copy myc tag.
Since our cell culture colocalization assay turned out to be highly sensitive for the detection of full-length Groucho protein interactions in vivo, we used it for the isolated WD40 domain as well. Initial experiments using GFP fusion proteins were negative (data not shown), but the application of the DsRed protein was successful. Red-WD40 protein relocated from the cytoplasm to the nucleus when cotransfected with En2 in almost 100% of the cells, whereas a single point mutation in the eh1 region of En2(F65E) completely abolished this effect (Fig. 5 G and H). Interestingly, DsRed is an obligate tetramer, in contrast to Gfp (71), and multimerization has previously been shown to be important for the repressive function of Groucho proteins (17). The assay gave positive results also for Gbx2 and Otx2 although at a reduced efficiency (Fig. 5I and J and Table 1). Altogether, we found that the WD40 repeat domain is not only essential but also sufficient for interactions with the eh1 motifs of Gbx2 and Otx2.
Development of a sensitive colocalization assay for the detection of protein-protein interactions in the nucleus. A number of methods are used to demonstrate protein-protein interactions, such as coimmunoprecipitation and pull-down assays; however, the presence of "sticky" domains within the interacting proteins may limit the sensitivity of these assays. While Engrailed proteins reliably coimmunoprecipitated Tle4 in our experiments (data not shown), considerable optimization of conditions was necessary to detect specific interactions with Otx2 and Gbx2 above background. This might reflect weaker interactions between Groucho corepressors and Gbx2 and Otx2 compared to those involving Engrailed. In addition, we observed considerable nonspecific binding of individual Tle4 domains (data not shown), suggesting that we reached the sensitivity limits of the assay. We therefore searched for an alternative method to analyze protein-protein interactions. We looked for an in vivo method to avoid artifacts of in vitro conditions. The relocation of binding partners to different intracellular compartments has been used before to demonstrate Groucho protein interactions (57). We observed that Tle4 proteins localize to different intranuclear positions compared to DNA-binding transcription factors, whereas coexpression leads to a redistribution of Tle4. In a previous experiment, AML proteins also targeted Tle to a speckled pattern within the nucleus (30). Using fluorescently labeled proteins, this approach provides an exceptionally detailed analysis of such interactions. After refining the experimental conditions, we detected interactions in nearly 100% of the cells with the strong binding partners, Engrailed, and Hes1. For Otx2 and Gbx2, we detected interactions in a large number of the cells. In addition, specific mutations within the interaction motifs completely abolished these effects (Fig. 3G and I). Our in vivo assay therefore provides highly specific detection of protein-protein interactions for experiments using transient transfection.
Interactions between Otx2/Gbx2 and Groucho/Tle proteins regulate the genetic network of the mid-hindbrain organizer in fish embryos. The midbrain-hindbrain organizer is a signaling center that orchestrates development in the midbrain and the anterior hindbrain primordia. One of the earliest events in the development of the anterior central nervous system is its subdivision into anterior Otx2-positive and posterior Gbx2-positive domains. These homeobox genes antagonize each other (11, 35, 46, 66), and this reciprocal relationship is critical for the positioning of the MHB. The molecular basis of this interaction, however, is not yet known. Fusion protein experiments in Xenopus suggest that transcriptional repression rather than activation underlies the mutual antagonism of Otx2 and Gbx2 (28). A physical interaction between Otx proteins and Tle4 was also recently demonstrated (52), but the consequences of this association on the positioning of the MHB have not been investigated.
mRNA injection of Gbx2 has previously been shown to produce truncations in anterior neural structures, such as the forebrain and eyes (36, 66). This early phenotype is not dependent on Groucho corepressors in medaka, as N-terminal deletions of Gbx2 including the eh1-like interaction domain do not affect the results of mRNA injection experiments (data not shown). However, by shifting the onset of Gbx2 expression to a later developmental stage using a heat shock-inducible system (7), we observed specific effects on MHB positioning. The Otx2 and Gbx2 expression domains, as well as their common boundary, shifted to a more anterior position (particularly on the dorsal side of the embryo) (Fig. 7B). A similar anterior shift of the MHB region was previously observed in mouse mutants with reduced levels of Otx gene products (1). This suggests that Otx2 gene function is down-regulated by Gbx2 activity.
Our inducible misexpression system was also instrumental for the Otx2 experiments since overexpression of this gene is known to block the convergent extension movements of gastrulation (47). In these experiments, heat shock activation of Otx2 resulted in significantly reduced Gbx2 expression. These effects occurred primarily in the anterior-most expression domain of Gbx2, directly at the MHB and abutting the Otx2-positive territory. Interestingly, the position of the MHB was not affected in these experiments. These misexpression experiments therefore reproduce the reciprocal repression of the two homeodomain transcription factors in MHB specification, as seen in other experimental systems.
The effect of Gbx2 on Otx2 was dependent on Groucho protein interactions since misexpression of Gbx2-Tle4 fusion proteins extended the phenotype of wild-type Gbx2, whereas an expression construct lacking the eh1 motif did not show repression (Table 2). On the contrary, misexpression of the Gbx2 mutant resulted in ectopic expression of Otx2 in a region posterior to the Otx2-positive territory (Fig. 7H), suggesting a switch of Gbx2 repressing functions to activation of its target gene, Otx2. These data are in agreement with our misexpression experiments using the WD40 domain to block interaction of the endogenous Gbx2 protein with the corepressor. Furthermore, overexpression of the dominant negative Aes in medaka embryos caused ectopic expression of Otx2 in the hindbrain (43).
The reciprocal Otx2 misexpression experiments did not, however, indicate corepressor dependence (Table 2). Interestingly, Otx2 and Otx2-Tle4 misexpression resulted in the repression of Fgf8 in the MHB, which was not the case for the Otx2 deletion construct (Fig. 7F and Table 2). Fgf8 plays a pivotal role in isthmic organizer activity (reviewed in reference 58), and Otx gene products have been shown to regulate Fgf8 expression. Thus, in Otx1–/– Otx2+/– mouse embryos, the Fgf8 expression domain was abnormally enlarged in the presumptive diencephalon (1). Therefore, a critical level of Otx gene activity is required to repress Fgf8 expression in the midbrain. Our experiments suggest that Groucho proteins play an important role in this process.
Taken together, our misexpression studies in medaka embryos fully support the hypothesis that specific interactions between Gbx2 and Otx2 are crucial for normal MHB development. Groucho proteins regulate the transcriptional repression exerted by Gbx2 on Otx2 in the MHB, as well as the repression of Fgf8 by Otx2 in this critical region.
|
ACKNOWLEDGMENTS |
|---|
This work was supported by the Austrian Science Fund (grant P15185), the Italian Association for Cancer Research, and the MRC (grant G0400410).
|
FOOTNOTES |
|---|
Published ahead of print on 23 October 2006.
Supplemental material for this article may be found at http://mcb.asm.org/.
Present address: Cellzome AG, Meyerhofstrasse 1 69117 Heidelberg, Germany.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2. Acampora, D., M. Gulisano, V. Broccoli, and A. Simeone. 2001. Otx genes in brain morphogenesis. Prog. Neurobiol. 64:69-95.[CrossRef][Medline]
3. Acampora, D., S. Mazan, Y. Lallemand, V. Avantaggiato, M. Maury, A. Simeone, and P. Brulet. 1995. Forebrain and midbrain regions are deleted in Otx2–/–mutants due to a defective anterior neuroectoderm specification during gastrulation. Development 121:3279-3290.[Abstract]
4. Adams, B., P. Dorfler, A. Aguzzi, Z. Kozmik, P. Urbanek, I. Maurer Fogy, and M. Busslinger. 1992. Pax-5 encodes the transcription factor BSAP and is expressed in B lymphocytes, the developing CNS, and adult testis. Genes Dev. 6:1589-1607.
5. Aghaallaei, N., B. Bajoghli, I. Walter, and T. Czerny. 2005. Duplicated members of the Groucho/Tle gene family in fish. Dev. Dyn. 234:143-150.[CrossRef][Medline]
6. Bajoghli, B., N. Aghaallaei, and T. Czerny. 2005. Groucho corepressor proteins regulate otic vesicle outgrowth. Dev. Dyn. 233:760-771.[CrossRef][Medline]
7. Bajoghli, B., N. Aghaallaei, T. Heimbucher, and T. Czerny. 2004. An artificial promoter construct for heat-inducible misexpression during fish embryogenesis. Dev. Biol. 271:416-430.[CrossRef][Medline]
8. Baker, A., M. Saltik, H. Lehrmann, I. Killisch, V. Mautner, G. Lamm, G. Christofori, and M. Cotten. 1997. Polyethylenimine (PEI) is a simple, inexpensive and effective reagent for condensing and linking plasmid DNA to adenovirus for gene delivery. Gene Ther. 4:773-782.[CrossRef][Medline]
9. Brantjes, H., J. Roose, M. van De Weteringqq, and H. Clevers. 2001. All Tcf HMG box transcription factors interact with Groucho-related co-repressors. Nucleic Acids Res. 29:1410-1419.
10. Braselmann, S., P. Graninger, and M. Busslinger. 1993. A selective transcriptional induction system for mammalian cells based on Gal4-estrogen receptor fusion proteins. Proc. Natl. Acad. Sci. USA 90:1657-1661.
11. Broccoli, V., E. Boncinelli, and W. Wurst. 1999. The caudal limit of Otx2 expression positions the isthmic organizer. Nature 401:164-168.[CrossRef][Medline]
12. Bulfone, A., L. Puelles, M. H. Porteus, M. A. Frohman, G. R. Martin, and J. L. Rubenstein. 1993. Spatially restricted expression of Dlx-1, Dlx-2 (Tes-1), Gbx-2, and Wnt-3 in the embryonic day 12.5 mouse forebrain defines potential transverse and longitudinal segmental boundaries. J. Neurosci. 13:3155-3172.[Abstract]
13. Cai, Y., P. D. Brophy, I. Levitan, S. Stifani, and G. R. Dressler. 2003. Groucho suppresses Pax2 transactivation by inhibition of JNK-mediated phosphorylation. EMBO J. 22:5522-5529.[CrossRef][Medline]
14. Cavallo, R. A., R. T. Cox, M. M. Moline, J. Roose, G. A. Polevoy, H. Clevers, M. Peifer, and A. Bejsovec. 1998. Drosophila Tcf and Groucho interact to repress Wingless signalling activity. Nature 395:604-608.[CrossRef][Medline]
15. Chen, G., and A. J. Courey. 2000. Groucho/TLE family proteins and transcriptional repression. Gene 249:1-16.[CrossRef][Medline]
16. Chen, G., J. Fernandez, S. Mische, and A. J. Courey. 1999. A functional interaction between the histone deacetylase Rpd3 and the corepressor groucho in Drosophila development. Genes Dev. 13:2218-2230.
17. Chen, G., P. H. Nguyen, and A. J. Courey. 1998. A role for Groucho tetramerization in transcriptional repression. Mol. Cell. Biol. 18:7259-7268.
18. Chiang, C., K. E. Young, and P. A. Beachy. 1995. Control of Drosophila tracheal branching by the novel homeodomain gene unplugged, a regulatory target for genes of the bithorax complex. Development 121:3901-3912.[Abstract]
19. Choi, C. Y., Y. H. Kim, H. J. Kwon, and Y. Kim. 1999. The homeodomain protein NK-3 recruits Groucho and a histone deacetylase complex to repress transcription. J. Biol. Chem. 274:33194-33197.
20. Courey, A. J., and J. D. Huang. 1995. The establishment and interpretation of transcription factor gradients in the Drosophila embryo. Biochim. Biophys. Acta 1261:1-18.[Medline]
21. Dasen, J. S., J. P. Barbera, T. S. Herman, S. O. Connell, L. Olson, B. Ju, J. Tollkuhn, S. H. Baek, D. W. Rose, and M. G. Rosenfeld. 2001. Temporal regulation of a paired-like homeodomain repressor/TLE corepressor complex and a related activator is required for pituitary organogenesis. Genes Dev. 15:3193-3207.
22. Dubnicoff, T., S. A. Valentine, G. Chen, T. Shi, J. A. Lengyel, Z. Paroush, and A. J. Courey. 1997. Conversion of dorsal from an activator to a repressor by the global corepressor Groucho. Genes Dev. 11:2952-2957.
23. Eberhard, D., G. Jimenez, B. Heavey, and M. Busslinger. 2000. Transcriptional repression by Pax5 (BSAP) through interaction with corepressors of the Groucho family. EMBO J. 19:2292-2303.[CrossRef][Medline]
24. Fainsod, A., and Y. Greunbaum. 1989. A chicken homeo box gene with developmentally regulated expression. FEBS Lett. 250:381-385.[CrossRef][Medline]
25. Fink, M., G. Flekna, A. Ludwig, T. Heimbucher, and T. Czerny. 2006. Improved translation efficiency of injected mRNA during early embryonic development. Dev. Dyn.
26. Fisher, A. L., and M. Caudy. 1998. Groucho proteins: transcriptional corepressors for specific subsets of DNA-binding transcription factors in vertebrates and invertebrates. Genes Dev. 12:1931-1940.
27. Garcia-Mata, R., Z. Bebok, E. J. Sorscher, and E. S. Sztul. 1999. Characterization and dynamics of aggresome formation by a cytosolic GFP-chimera. J. Cell Biol. 146:1239-1254.
28. Glavic, A., J. L. Gomez-Skarmeta, and R. Mayor. 2002. The homeoprotein Xiro1 is required for midbrain-hindbrain boundary formation. Development 129:1609-1621.
29. Iwamatsu, T. 2004. Stages of normal development in the medaka Oryzias latipes. Mech. Dev. 121:605-618.[CrossRef][Medline]
30. Javed, A., B. Guo, S. Hiebert, J. Y. Choi, J. Green, S. C. Zhao, M. A. Osborne, S. Stifani, J. L. Stein, J. B. Lian, A. J. van Wijnen, and G. S. Stein. 2000. Groucho/TLE/R-esp proteins associate with the nuclear matrix and repress RUNX (CBF(alpha)/AML/PEBP2(alpha))-dependent activation of tissue-specific gene transcription. J. Cell Sci. 113:2221-2231.[Abstract]
31. Jennings, B. H., L. M. Pickles, S. M. Wainwright, S. M. Roe, L. H. Pearl, and D. Ish-Horowicz. 2006. Molecular recognition of transcriptional repressor motifs by the WD domain of the Groucho/TLE corepressor. Mol. Cell 22:645-655.[CrossRef][Medline]
32. Jimenez, G., Z. Paroush, and D. Ish-Horowicz. 1997. Groucho acts as a corepressor for a subset of negative regulators, including hairy and engrailed. Genes Dev. 11:3072-3082.
33. Jimenez, G., C. P. Verrijzer, and D. Ish-Horowicz. 1999. A conserved motif in goosecoid mediates Groucho-dependent repression in Drosophila embryos. Mol. Cell. Biol. 19:2080-2087.
34. Joyner, A. L., A. Liu, and S. Millet. 2000. Otx2, gbx2 and fgf8 interact to position and maintain a mid-hindbrain organizer. Curr. Opin. Cell Biol. 12:736-741.[CrossRef][Medline]
35. Katahira, T., T. Sato, S. Sugiyama, T. Okafuji, I. Araki, J. Funahashi, and H. Nakamura. 2000. Interaction between Otx2 and Gbx2 defines the organizing center for the optic tectum. Mech. Dev. 91:43-52.[CrossRef][Medline]
36. Kikuta, H., M. Kanai, Y. Ito, and K. Yamasu. 2003. gbx2 homeobox gene is required for the maintenance of the isthmic region in the zebrafish embryonic brain. Dev. Dyn. 228:433-450.[CrossRef][Medline]
37. Kobayashi, M., K. Nishikawa, T. Suzuki, and M. Yamamoto. 2001. The homeobox protein Six3 interacts with the Groucho corepressor and acts as a transcriptional repressor in eye and forebrain formation. Dev. Biol. 232:315-326.[CrossRef][Medline]
38. Komachi, K., M. J. Redd, and A. D. Johnson. 1994. The WD repeats of Tup1 interact with the homeo domain protein alpha 2. Genes Dev. 8:2857-2867.
39. Koop, K. E., L. M. MacDonald, and C. G. Lobe. 1996. Transcripts of Grg4, a murine Groucho-related gene, are detected in adjacent tissues to other murine neurogenic gene homologues during embryonic development. Mech. Dev. 59:73-87.[CrossRef][Medline]
40. Kowenz-Leutz, E., P. Herr, K. Niss, and A. Leutz. 1997. The homeobox gene GBX2, a target of the myb oncogene, mediates autocrine growth and monocyte differentiation. Cell 91:185-195.[CrossRef][Medline]
41. Kurosawa, G., K. Yamada, H. Ishiguro, and H. Hori. 1999. Hox gene complexity in medaka fish may be similar to that in pufferfish rather than zebrafish. Biochem. Biophys. Res. Commun. 260:66-70.[CrossRef][Medline]
42. Loosli, F., R. W. Koster, M. Carl, R. Kuhnlein, T. Henrich, M. Mucke, A. Krone, and J. Wittbrodt. 2000. A genetic screen for mutations affecting embryonic development in medaka fish (Oryzias latipes). Mech. Dev. 97:133-139.[CrossRef][Medline]
43. Lopez-Rios, J., K. Tessmar, F. Loosli, J. Wittbrodt, and P. Bovolenta. 2003. Six3 and Six6 activity is modulated by members of the Groucho family. Development 130:185-195.
44. Matsuo, I., S. Kuratani, C. Kimura, N. Takeda, and S. Aizawa. 1995. Mouse Otx2 functions in the formation and patterning of rostral head. Genes Dev. 9:2646-2658.
45. McLarren, K. W., R. Lo, D. Grbavec, K. Thirunavukkarasu, G. Karsenty, and S. Stifani. 2000. The mammalian basic helix loop helix protein HES-1 binds to and modulates the transactivating function of the runt-related factor Cbfa1. J. Biol. Chem. 275:530-538.
46. Millet, S., K. Campbell, D. J. Epstein, K. Losos, E. Harris, and A. L. Joyner. 1999. A role for Gbx2 in repression of Otx2 and positioning the mid/hindbrain organizer. Nature 401:161-164.[CrossRef][Medline]
47. Morgan, R., M. H. Hooiveld, M. Pannese, G. Dati, F. Broders, M. Delarue, J. P. Thiery, E. Boncinelli, and A. J. Durston. 1999. Calponin modulates the exclusion of Otx-expressing cells from convergence extension movements. Nat. Cell Biol. 1:404-408.[CrossRef][Medline]
48. Muhr, J., E. Andersson, M. Persson, T. M. Jessell, and J. Ericson. 2001. Groucho-mediated transcriptional repression establishes progenitor cell pattern and neuronal fate in the ventral neural tube. Cell 104:861-873.[CrossRef][Medline]
49. Neer, E. J., C. J. Schmidt, R. Nambudripad, and T. F. Smith. 1994. The ancient regulatory-protein family of WD-repeat proteins. Nature 371:297-300.[CrossRef][Medline]
50. Paroush, Z., R. L. Finley, T. Kidd, S. M. Wainwright, P. W. Ingham, R. Brent, and D. Ish-Horowicz. 1994. Groucho is required for Drosophila neurogenesis, segmentation, and sex dete
