Jerky/Earthbound facilitates cell-specific Wnt/Wingless signalling by modulating β-catenin-TCF activity
2011; Springer Nature; Volume: 30; Issue: 8 Linguagem: Inglês
10.1038/emboj.2011.67
ISSN1460-2075
AutoresHassina Benchabane, Nan Xin, Ai Tian, Brian P. Hafler, Kerrie Nguyen, Ayah Ahmed, Yashi Ahmed,
Tópico(s)Kruppel-like factors research
ResumoArticle11 March 2011free access Jerky/Earthbound facilitates cell-specific Wnt/Wingless signalling by modulating β-catenin–TCF activity Hassina Benchabane Hassina Benchabane Department of Genetics and the Norris Cotton Cancer Center, Dartmouth Medical School, Hanover, NH, USA Search for more papers by this author Nan Xin Nan Xin Department of Genetics and the Norris Cotton Cancer Center, Dartmouth Medical School, Hanover, NH, USA Search for more papers by this author Ai Tian Ai Tian Department of Genetics and the Norris Cotton Cancer Center, Dartmouth Medical School, Hanover, NH, USA Search for more papers by this author Brian P Hafler Brian P Hafler Department of Molecular Biology, Princeton University, Princeton, NJ, USA Search for more papers by this author Kerrie Nguyen Kerrie Nguyen Department of Genetics and the Norris Cotton Cancer Center, Dartmouth Medical School, Hanover, NH, USA Search for more papers by this author Ayah Ahmed Ayah Ahmed Department of Genetics and the Norris Cotton Cancer Center, Dartmouth Medical School, Hanover, NH, USA Search for more papers by this author Yashi Ahmed Corresponding Author Yashi Ahmed Department of Genetics and the Norris Cotton Cancer Center, Dartmouth Medical School, Hanover, NH, USA Search for more papers by this author Hassina Benchabane Hassina Benchabane Department of Genetics and the Norris Cotton Cancer Center, Dartmouth Medical School, Hanover, NH, USA Search for more papers by this author Nan Xin Nan Xin Department of Genetics and the Norris Cotton Cancer Center, Dartmouth Medical School, Hanover, NH, USA Search for more papers by this author Ai Tian Ai Tian Department of Genetics and the Norris Cotton Cancer Center, Dartmouth Medical School, Hanover, NH, USA Search for more papers by this author Brian P Hafler Brian P Hafler Department of Molecular Biology, Princeton University, Princeton, NJ, USA Search for more papers by this author Kerrie Nguyen Kerrie Nguyen Department of Genetics and the Norris Cotton Cancer Center, Dartmouth Medical School, Hanover, NH, USA Search for more papers by this author Ayah Ahmed Ayah Ahmed Department of Genetics and the Norris Cotton Cancer Center, Dartmouth Medical School, Hanover, NH, USA Search for more papers by this author Yashi Ahmed Corresponding Author Yashi Ahmed Department of Genetics and the Norris Cotton Cancer Center, Dartmouth Medical School, Hanover, NH, USA Search for more papers by this author Author Information Hassina Benchabane1, Nan Xin1, Ai Tian1, Brian P Hafler2, Kerrie Nguyen1, Ayah Ahmed1 and Yashi Ahmed 1 1Department of Genetics and the Norris Cotton Cancer Center, Dartmouth Medical School, Hanover, NH, USA 2Department of Molecular Biology, Princeton University, Princeton, NJ, USA *Corresponding author. Department of Genetics, Dartmouth Medical School, Remsen 7400, Hanover NH 03755, USA. Tel.: +1 603 650 1027; Fax: +1 603 650 1188; E-mail: [email protected] The EMBO Journal (2011)30:1444-1458https://doi.org/10.1038/emboj.2011.67 PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Wnt/Wingless signal transduction directs fundamental developmental processes, and upon hyperactivation triggers colorectal adenoma/carcinoma formation. Responses to Wnt stimulation are cell specific and diverse; yet, how cell context modulates Wnt signalling outcome remains obscure. In a Drosophila genetic screen for components that promote Wingless signalling, we identified Earthbound 1 (Ebd1), a novel member in a protein family containing Centromere Binding Protein B (CENPB)-type DNA binding domains. Ebd1 is expressed in only a subset of Wingless responsive cell types, and is required for only a limited number of Wingless-dependent processes. In addition, Ebd1 shares sequence similarity and can be functionally replaced with the human CENPB domain protein Jerky, previously implicated in juvenile myoclonic epilepsy development. Both Jerky and Ebd1 interact directly with the Wnt/Wingless pathway transcriptional co-activators β-catenin/Armadillo and T-cell factor (TCF). In colon carcinoma cells, Jerky facilitates Wnt signalling by promoting association of β-catenin with TCF and recruitment of β-catenin to chromatin. These findings indicate that tissue-restricted transcriptional co-activators facilitate cell-specific Wnt/Wingless signalling responses by modulating β-catenin–TCF activity. Introduction The evolutionarily conserved secreted ligands Wnt/Wingless activate a signal transduction pathway that governs fundamental aspects of metazoan development and is misregulated in several human diseases, including the majority of colorectal carcinomas (Clevers, 2006; MacDonald et al, 2009). The Wnt signalling cascade elicits distinct responses that control cell proliferation, fate specification, differentiation, and apoptosis. The majority of Wnt target genes are cell specific, and this specificity depends on numerous contextual factors, including developmental history, as well as strength, timing, and duration of Wnt stimulation (Logan and Nusse, 2004). For example, during mammalian skeletal muscle regeneration, early exposure of myogenic precursor cells to Wnt induces their specification as fibroblasts, whereas later exposure signals a switch from myoblast proliferation to myocyte differentiation (Brack et al, 2007, 2008). The observation that the same signal and signalling pathway can elicit many different outcomes underlies two long-standing questions: how is specificity achieved in Wnt signalling response and how does cell identity influence this choice? Despite the intense effort that has informed our current understanding of Wnt signal transduction, the molecular mechanisms that determine how context-specific responses are elicited upon Wnt stimulation remain largely unknown. The half-life of β-catenin/Armadillo (Arm), an essential transcriptional co-activator, is the primary determinant of Wnt pathway activity (Huang and He, 2008). β-Catenin is targeted for proteasomal degradation by a complex that includes Adenomatous polyposis coli (APC), Axin, and glycogen synthase kinase-3 (GSK3). Wnt stimulation inhibits the activity of this complex, resulting in β-catenin stabilization and nuclear translocation. Nuclear β-catenin interacts with the sequence-specific DNA-binding proteins TCF/LEF (T-cell factor/lymphoid enhancer factor) to activate Wnt target gene transcription (Arce et al, 2006). Widely utilized factors that modify chromatin or function in the basal transcription machinery are recruited by β-catenin–TCF to Wnt target gene enhancers, as are the Wnt pathway transcriptional co-factors Pygopus and BCL9/Legless (Mosimann et al, 2009). As the β-catenin–TCF complex is expressed ubiquitously, cell-specific Wnt target gene activation is likely facilitated by other factors; yet, genetic evidence for such factors has been scarce. Here, we describe a genetic screen in Drosophila for components that promote Wingless signalling, and identify a novel nuclear factor, Earthbound 1 (Ebd1), which belongs to a protein family containing Centromere Binding Protein B (CENPB)-type DNA binding domains. Ebd1 is expressed in a restricted subset of Wingless responsive cells and is required for only a limited number of signalling responses. Ebd1 associates with Armadillo/β-catenin and TCF. Two other Drosophila CENPB domain proteins have partially redundant roles with Ebd1. In addition, the human CENPB domain protein Jerky, previously implicated in development of certain forms of juvenile epilepsy, functions as an Ebd1 homologue. Jerky replaces Ebd activity when expressed in flies, promotes Wnt signalling in colon carcinoma cells, stabilizes the β-catenin–TCF complex, and facilitates recruitment of β-catenin to chromatin. These studies provide evidence that tissue-restricted transcriptional co-activators facilitate cell-specific Wnt signalling responses by modulating β-catenin–TCF activity. Results Identification of Ebd1 in a genetic screen for factors that promote Wingless signalling A gradient of Wingless activity patterns the Drosophila pupal retina, with the highest Wingless levels inducing apoptosis of all photoreceptors at the retinal periphery (Tomlinson, 2003; Lin et al, 2004). Our previous work revealed that loss of Adenomatous polyposis coli 1 (Apc1) results in aberrant activation of Wingless signalling in all retinal photoreceptors, thereby inducing their ectopic apoptosis (Figure 1A and B; Ahmed et al, 1998; Benchabane et al, 2008). This Apc1 mutant eye phenotype is a sensitive indicator of Wingless pathway activity, as reduction of only two-fold in gene dosage of arm or TCF is sufficient for suppression of the apoptosis (Ahmed et al, 1998; Benchabane et al, 2008). Therefore, to identify components that promote Wingless signalling, we performed a genetic modifier screen in which we searched for suppressors of Apc1 mutant retinal apoptosis. Among the suppressors we identified is a lethal allele of legless, which encodes a co-activator in the Armadillo–TCF transcription complex (Kramps et al, 2002), validating our approach. We also isolated an allele of a locus we named earthbound 1 (ebd1). Partial suppression of Apc1 mutant apoptosis is observed in ebd1QF1 heterozygotes, whereas nearly complete suppression is present in ebd1QF1 homozygotes (Figure 1C). We mapped ebd1 to the novel gene CG3371 at cytological position 61C (Supplementary Figure S1A–D). Two large deletions in this region, Df(3L)4136 and Df(3L)27-3, which when transheterozygous eliminate 11 genes including ebd1, prevent apoptosis in the Apc1 mutant (Figure 1D; Supplementary Figure S1A), as does the P element EY01876, which is inserted in the ebd1 5′ untranslated region (Supplementary Figure S1C–E). Two smaller deletions, Df(3L)5 and ebd1240, which when transheterozygous eliminate only ebd1, confirmed the suppressed apoptosis (Supplementary Figure S1E and F). Finally, a transgene encoding only Ebd1, expressed in the retina using UAS/Gal4 (Fischer et al, 1988; Brand and Perrimon, 1993), restores apoptosis, confirming identification of the correct gene (Figure 1E). Figure 1.Ebd1 promotes Wingless signalling. (A–E) Cross-sections through adult retinas. (A) As in wild-type, eight photoreceptors, seven of which are seen in the plane of focus, are present in Apc1Q8 heterozygous flies. Each group of photoreceptor cells is surrounded by a pigment cell lattice, identified by small, dark pigment granules. In Apc1 homozygous mutants, photoreceptors undergo apoptosis (B), which is suppressed in ebd1QF1 Apc1Q8 double mutants (C), and in flies transheterozygous for deficiencies Df(3L)4136 and Df(3L)27-3 (D). (E) Expression of UAS-ebd1 under the control of the eye-specific enhancer GMR-Gal4 restores apoptosis. (F) Schematic representation of Ebd1 with its two putative CENPB-type DNA binding domains, CENPB-N and CENPB (blue). The amino-acid sequence of the second CENPB domain and isoleucine to phenylalanine substitution found in Ebd1QF1 are shown. (G, H) Adult wings in which dominant-negative Lgs17E is expressed under control of the spalt major enhancer (salE) are shown. Wings are oriented with proximal left and anterior up. Expression of Lgs17E attenuates Wingless signalling, causing wing notches (G), which are rescued by co-expression of Ebd1 (H). Download figure Download PowerPoint Ebd1 is a member of a family of proteins containing CENPB-type DNA binding domains (Schultz et al, 1998; Letunic et al, 2006), named after the sequence-specific DNA-binding protein CENPB (Masumoto et al, 1989). Proteins containing CENPB domains are found from yeast to humans and function in diverse processes, including centromere assembly and transcriptional activation (Mojzita and Hohmann, 2006; Okada et al, 2007). Many CENPB domain proteins, including Ebd1, contain two tandem homeodomain-like helix-turn-helix domains, termed CENPB-N and CENPB, that are important for DNA binding (Pluta et al, 1992; Murakami et al, 1996). The QF1 mutation is located in the second CENPB domain of Ebd1 (Figure 1F), and results in an isoleucine to phenylalanine substitution at position 188, a highly conserved residue. Ebd1 promotes Wingless signalling To determine whether Ebd1 promotes Wingless signal transduction or solely promotes photoreceptor apoptosis, we examined the effects of Ebd1 on several other Wingless-dependent processes. A dominant-negative Legless protein, Lgs17E, which is impaired in interaction with Armadillo, attenuates Wingless signalling when expressed in the wing primordium, causing a fully penetrant notched wing phenotype (Figure 1G; Kramps et al, 2002; Mosimann et al, 2006). Co-expression of Ebd1 rescues this phenotype in all wings analysed, indicating restoration of Wingless transduction (Figure 1H), whereas co-expression of GFP or β-galactosidase has no effect (Figure 1G). We also examined two other phenotypes resulting from ectopic Wingless signalling in the Apc1 mutant retina: aberrant photoreceptor fate specification, as indicated by ectopic homothorax expression (Supplementary Figure S2A–C; Benchabane et al, 2008), and aberrant shortening in photoreceptor length (Supplementary Figure S2H and I; Benchabane et al, 2008). Inactivation of ebd1 partially prevents ectopic homothorax expression (Supplementary Figure S2D–G) and completely eliminates shortened photoreceptors (Supplementary Figure S2J and K). In addition, ectopic Wingless signalling resulting from Apc2 inactivation in retinal neurons (Benchabane et al, 2008) is prevented by elimination of Ebd1 (Supplementary Figure S2L–N). These data indicate that Ebd1 facilitates Wingless signal transduction. Ebd1 is expressed in only a subset of Wingless-dependent cell types To identify cells and developmental stages in which Ebd1 functions, we generated an Ebd1 polyclonal antiserum and immunostained embryonic, larval, and pupal tissues. During embryogenesis, Ebd1 is expressed in neurons, myoblasts and myofibres, glial cells, and nephrocytes (garland cells) (Figure 2A–H; Supplementary Figure S3). In larvae and pupae, Ebd1 expression persists in these same cell types, and is also present in salivary glands (Figure 2I–L; Supplementary Figure S4 and data not shown). We observed a similar tissue-specific pattern with a reporter gene activated by two independently derived ebd1-GAL4 lines (Figure 2M–O and data not shown). Together, these data reveal cell type-specific Ebd1 expression that is restricted primarily to myocytes, neurons, and glial cells. In contrast with the ubiquitous expression of core Wingless pathway components, Ebd1 is notably absent in ectodermally derived epithelial cells. Figure 2.Ebd1 is expressed in neurons and muscles. (A–H) Confocal images of late embryo stained for Ebd1 (green; A, E). Myoblasts and myofibres are marked with α-Mef2 (blue; B, F), and neurons with α-Elav (red; C, G). Higher magnification is shown in (E–H) and merged images in (D) and (H). (I–K) Confocal images of immunostained pupal thorax, 24 h after puparium formation (APF). Pupa is stained for Ebd1 (green; I) and Mef2 (magenta; J). The dashed line indicates the midline (I). A merged image is shown in (K). Ebd1 and Mef2 are expressed in dorsal longitudinal muscle (DLM) fibres and myoblasts. (L) Schematic representation of pupa indicating the position of DLMs (green). (M–O) Confocal images of immunostained DLMs, 22 h APF in which β-galactosidase is expressed under control of the ebd1 enhancer. Pupa is stained for β-gal (green; M) and Mef2 (magenta; N). Merged image is shown in (O). β-galactosidase is expressed in all six DLM fibres and some myoblasts (arrowhead). Download figure Download PowerPoint Ebd1 is required for indirect flight muscle development, but not for the majority of Wingless-dependent processes Inactivation of core Wingless pathway components causes widespread aberrant cell fate specification throughout development and characteristic external morphological defects in many adult tissues (Nusslein-Volhard and Wieschaus, 1980; Wieschaus and Riggleman, 1987; Baker, 1988a, 1988b). By contrast, ebd1 null mutants are viable and display a largely normal exterior morphology, suggesting that defects resulting from ebd1 loss are restricted to specific tissues or developmental stages. In addition, ebd1 null mutants have no observable defects in the specification or migration of embryonic RP2 motor neurons (Supplementary Figure S5; Supplementary Table 1), two developmental processes dependent on Wingless signalling (Chu-LaGraff and Doe, 1993; Bhat, 2007). However, we observed that some ebd1 mutant adults have abnormal wing posture and are unable to fly, and examined their flight musculature to identify the cause. The bulk of adult thoracic muscles are made up of two groups of indirect flight muscles: dorsal longitudinal muscles (DLMs) and dorso-ventral muscles (DVMs) (Figure 3A). Each thoracic hemisegment contains six DLMs and three sets of DVMs. All females and a third of ebd1 mutant males lack one or more DLMs, with an increased size in remaining DLMs (Figure 3B; Supplementary Table 2). In addition, most ebd1 mutants have reduced number of DVMs (data not shown). Figure 3.ebd1, ebd2, and Wingless signalling mutants lack a subset of indirect flight muscles. (A–G, I, J) Transverse section of adult thoraces. (A–E) Dorso-ventral muscles (DVMs, ▴) and dorsal longitudinal muscles (DLMs, *) are indicated. Wild-type (A), homozygous ebd1240 (B), ebd2136 (C), and ebd1240 ebd2136 double mutants (D). (E) Expressing an ebd1 transgene rescues muscle loss in ebd1240ebd2136 double mutants. (F–K) Attenuation of Wingless signalling in ebd1-expressing cells reduces the size and number of indirect flight muscles, resulting in loss of flight activity. This is rescued by expression of ArmS10 or Ebd1. (F–H) Adult females expressing dominant-negative dTCFΔN protein, under control of the ebd1 enhancer were raised at 25°C until larval third instar and then 22°C until eclosion. Co-expression of GFP and β-galactosidase (F) or ArmS10 (G) is shown. DLMs (*) are indicated in (G). (H) Flight assay of ebd1>dTCFΔN; UAS-GFP-lacZ and ebd1>dTCFΔN; UAS-armS10 adult flies. Number of flies examined (n) is indicated. (I–K) Adult ebd1>dTCFΔN males raised at 25°C; co-expression of GFP and β-galactosidase (I) or Ebd1 (J) is shown. (K) Flight assay of ebd1-Gal4>UAS-dTCFΔN; UAS-GFP-lacZ and ebd1-Gal4>UAS-dTCFΔN; UAS-ebd1 adult flies. Download figure Download PowerPoint In addition to Ebd1, the Drosophila melanogaster genome encodes four other proteins with CENPB domains (Supplementary Figure S6A). We isolated deletions in each of these genes; even upon simultaneous inactivation of all five CENPB domain proteins, viable flies with no external patterning defects are observed (data not shown). However, we examined ectopic expression of two of these CENPB proteins, encoded by CG12972 and CG13895, and found that they prevent attenuated Wingless signalling resulting from dominant-negative Legless, indicating that they may also promote Wingless transduction in a limited subset of cells (Supplementary Figure S6B and C). To determine whether there exists functional redundancy between Ebd1 and other CENPB family members, we examined flight muscles in mutants with deletions in these genes. A DLM loss phenotype is also present in 83% of CG12972/earthbound 2 (ebd2) mutant females, but in no males (Figure 3C; Supplementary Table 2). Further, the severity of the ebd1 mutant muscle loss phenotype, with respect to both penetrance and expressivity, increases upon simultaneous inactivation of ebd2. All ebd1 ebd2 double mutants display DLM loss (Figure 3D; Supplementary Table 2) and loss of flight activity (data not shown). The muscle loss is rescued by expressing an ebd1 transgene under control of the ebd1-Gal4 driver (Figure 3E). Functional redundancy also exists between Ebd1, Ebd2, and the CENPB domain protein encoded by CG13895 in promoting ectopic Wingless signalling in Apc1 mutant retinal neurons (Supplementary Figure S6E and F). Together, these data reveal that some redundancy exists in Ebd function, and also confirm the tissue- and stage-specific requirement for Ebd proteins. Attenuated Wingless signalling and Ebd1 loss result in similar flight muscle defects that are reversed upon Ebd1 overexpression Previous studies revealed that Wingless expression in a subset of larval wing ectodermal cells promotes proliferation and maintains cell fate in overlying myoblasts, which contribute to adult flight muscle (Sudarsan et al, 2001). Indeed, we find that disruption of Wingless transduction in indirect flight muscles, by expression of dominant-negative TCF (dTCFΔN; van de Wetering et al, 1997) in ebd1-expressing cells, results in ‘held-up’ wings, severe reduction in indirect flight muscle size (Figure 3F and I), and loss of flight activity (Figure 3H and K). Similarly, overexpression of the negative regulator Axin in wing disc-associated myoblasts or ebd1-expressing cells results in partial to complete loss of indirect flight muscles (Supplementary Figure S7A–C). Most convincingly, reduction in Wingless levels, as found in hypomorphic wingless mutants, also results in indirect flight muscle defects (Supplementary Figure S7D and E). Loss of flight muscles and flight activity caused by dTCFΔN is rescued by co-expression of a constitutively active Armadillo protein, ArmS10 (Pai et al, 1997) or Pygopus, whereas expression of GFP or β-galactosidase has no effect (Figure 3F–H and data not shown). Thus, although dominant-negative TCF is markedly reduced in ability to associate with Armadillo (van de Wetering et al, 1997), high levels of Armadillo generated by ArmS10 expression appear sufficient to promote an ArmS10–dTCFΔN interaction, either directly or indirectly, or to promote formation of an ArmS10–endogenous TCF complex that competes effectively for chromatin association. Together, these results indicate that Wingless transduction is required in ebd1-expressing myofibres for flight muscle development. In addition, the muscle loss phenotypes observed upon Wingless pathway attenuation are more severe than those observed upon loss of Ebd1 and Ebd2, or even upon simultaneous loss of all five CENPB proteins, indicating that some Wingless-dependent flight muscle development proceeds normally in the absence of Ebd family proteins. To examine whether Ebd1 can promote Wingless signalling during flight muscle development, we co-expressed Ebd1 and dTCFΔN with the ebd1-GAL4 driver. Upon co-expression of Ebd1, flight muscles appear wild type, and flight activity is restored in all flies examined (Figure 3I–K). We ruled out the possibility that this rescue results from transcriptional repression of the ebd1-GAL4 driver by Ebd1 (Supplementary Figure S8A). The similarity between some muscle defects in ebd1 mutants and by inhibition of Wingless signalling in ebd1-expressing cells, as well as the ability of Ebd1 to rescue this attenuated Wingless signalling, suggest that ebd1 mutant muscle loss results from disruption of Wingless transduction. Ebd1 acts downstream of the Apc/Axin/GSK3 destruction complex To determine at what level in the Wingless pathway Ebd1 functions in vivo, we used the Apc/Axin/GSK3 destruction complex as a reference point in genetic epistasis. Ectopic Wingless signalling in retinal neurons resulting from inactivation of either Apc1 or Apc2 is prevented by Ebd1 loss (Figure 1; Supplementary Figures S1 and S2), indicating that Ebd1 acts downstream of the destruction complex. This conclusion is supported by the observation that Ebd1 loss prevents ectopic Wingless signalling resulting from expression of constitutively active Armadillo (ArmS10), which cannot be targeted for degradation by the destruction complex. Wild-type photoreceptors extend the entire length of the retina, from lens to base (Wolff and Ready, 1993). By contrast, ectopic Wingless signalling resulting from ArmS10 shortens photoreceptor length; photoreceptors are detected apically, but absent basally (Figure 4A and B; Benchabane et al, 2008). Ebd1 inactivation prevents this morphological defect (Figure 4C and D). These results indicate that Ebd1 promotes Wingless signalling downstream of Apc/Axin/GSK3. Figure 4.ebd1 functions downstream of Axin/Apc. (A–D) Cross-section through adult retinas expressing ArmS10 under control of the eye-specific long GMR enhancer is shown. Schematic representations of transverse sections through retina with photoreceptors (black) and pigment cells (yellow) are shown on right. Locations of cross-sections presented in (A–D) are indicated. (A, B) Photoreceptors (arrowheads) are present at apical (A), but not basal (B) levels of armS10 retina with two wild-type copies of ebd1 (data not shown), and in armS10 flies heterozygous for ebd1240. (C, D) Complete loss of ebd1 prevents the shortening of photoreceptors, which are now present at apical (C) and basal (D) levels. Download figure Download PowerPoint To rule out the possibility that promotion of Wingless signalling by Ebd1 is due to a general effect on transcription, or a specific effect on the transcription of Wingless pathway components, we analysed expression of Glycerol-3-phosphate dehydrogenase, dTCF, Arm, and Pygo in flies lacking ebd1 and ebd2. The expression level of these genes is not reduced in ebd1 ebd2 double mutants, indicating that Ebd does not promote Wingless signalling through an indirect effect on transcription (Supplementary Figure S8B–E). Human Jerky shares sequence similarity with Ebd1 and can replace Ebd1 function We next sought to determine whether there exist any human homologues of Ebd1. The human genome encodes 17 proteins containing at least one CENPB domain, and 13 proteins that contain two tandem CENPB domains, including CENPB, TIGGER transposable element derived 1 (TIGD1) through TIGD7, and Jerky (JRK/JH8) (SMART analysis; Letunic et al, 2006; Schultz et al, 1998). We examined several human CENPB domain proteins to determine whether they are able to activate the Wnt signalling transcriptional reporter TOPFLASH (Korinek et al, 1997) in the Wnt responsive cell line HEK293T. Human Jerky increases TOPFLASH reporter activity by two- to three-fold in response to Wnt3a addition, whereas CENPB, TIGD1, TIGD2, and TIGD4 do not (Figure 5A and B and data not shown). By contrast, Jerky does not activate FOPFLASH, a control reporter containing mutated TCF-binding sites, indicating specificity of the Jerky-mediated TCF reporter activation (Figure 5B). Figure 5.The human CENPB domain protein Jerky promotes Wnt signalling. (A, B) Jerky overexpression enhances Wnt3A-dependent transcriptional activation in HEK293T cells, as measured using the TOPFLASH luciferase reporter. (A) Results are expressed as ratio of Wnt3A dependent to basal induction of TOPFLASH. Human Jerky enhances Wnt3A-dependent activation of TOPFLASH, whereas TIGD1, TIGD2, and CENPB do not. (B) Basal and Wnt3A-dependent induction of TOPFLASH by Jerky are shown. Jerky has no effect on the control reporter FOPFLASH. (C) Jerky expression rescues muscle defects in ebd1 ebd2 double mutants. Transverse adult thoracic section showing 12 DLMs (*) with wild-type morphology when jerky is expressed in ebd1 ebd2 double mutants. (D) jerky siRNAs reduce Jerky levels in DLD-1 cells. Lysates of cells transfected with jerky siRNA or control siRNA were immunoprecipitated with Jerky or control IgG antibodies and immunoblotted with α-Jerky. (E) jerky siRNA reduces expression of ectopic Jerky in DLD-1 cells. Cells were transfected with siRNA and FLAG-jerky cDNA containing endogenous sequence or silent point mutations in the siRNA targeted region. Lysates were immunoblotted with α-Flag or anti-α-tubulin. The expression of Jerky encoded by unmodified cDNA is reduced upon siRNA addition, whereas expression of Jerky encoded by mutant cDNA, containing silent point mutations in the region targeted by siRNA, is not affected. (F) siRNA-mediated knockdown of jerky reduces SuperTOPFlash luciferase activity in DLD-1 human colonic carcinoma cells, whereas a control siRNA does not. SuperTOPFlash activity in cells transfected with jerky siRNA is rescued with jerky cDNA containing silent point mutations in the region targeted by siRNA. Jerky encoded by an unmodified cDNA does not rescue SuperTOPFlash. siRNAs have no effect on the control reporter SuperFOPflash. Download figure Download PowerPoint To determine whether human Jerky is functionally homologous to Ebd1, we expressed Jerky in flies under control of the ebd1 promoter, and examined the ability of Jerky to rescue flight muscle defects in ebd1 ebd2 double mutants. Whereas ebd1 ebd2 mutants display muscle loss defects with 100% penetrance (Figure 3D; Supplementary Table 2), expression of Jerky rescues this phenotype, but expression of GFP or β-galactosidase does not (Figure 5C and data not shown). Thus, human Jerky can compensate for Ebd1 and Ebd2 loss, suggesting that Jerky is a human homologue of Ebd. Endogenous human Jerky promotes Wnt transduction in colon carcinoma cells In support of our genetic studies, human Jerky was identified as a positive modulator of β-catenin–TCF-mediated transcription in a high-throughput RNA interference (RNAi) screen in colon carcinoma cells (Major et al, 2008); this work revealed that Jerky facilitates activation of a TCF reporter, as well as several endogenous β-catenin–TCF target genes. We independently confirmed these conclusions and performed a numb
Referência(s)