Identification and differential expression of multiple isoforms of mouse Coiled-coil-DIX1 (Ccd1), a positive regulator of Wnt signaling

Molecular Brain Research 135 (2005) 169 – 180 www.elsevier.com/locate/molbrainres

Research report

Identification and differential expression of multiple isoforms of mouse Coiled-coil-DIX1 (Ccd1), a positive regulator of Wnt signaling $ Kensuke Shiomi, Mizuki Kanemoto, Kazuko Keino-Masu, Sachine Yoshida, Katsunori Soma, Masayuki Masu* Department of Molecular Neurobiology, Institute of Basic Medical Sciences, Graduate School of Comprehensive Human Sciences, University of Tsukuba, Tsukuba, 1-1-1 Tennoudai, Tsukuba, Ibaraki 305-8577, Japan Accepted 13 December 2004 Available online 27 January 2005

Abstract The Wnt signaling plays important roles in cell growth, differentiation, polarity formation, and neural development. In the canonical pathway, two DIX domain-containing proteins, Dishevelled (Dvl) and Axin, regulate the degradation of h-catenin that activates Wnt target genes through TCF/LEF family transcription factors. Recently, we have isolated a third type of DIX domain-possessing protein, Coiled-coilDIX1 (Ccd1). Ccd1 forms homomeric and heteromeric complexes with Dvl and Axin, and regulates the neural patterning in zebrafish embryos through Wnt pathway activation. Here, we report the isolation and characterization of mouse Ccd1. Fourteen putative mRNA isoforms are generated by different promoter usage and alternative splicing, and each isoform shows different expression patterns in various tissues. The predicted Ccd1 proteins are classified into three subtypes, and a novel form, termed Ccd1A, possesses an N-terminal calponin homology domain, suggesting an additional interaction of the isoform with actin or other proteins. When Ccd1 proteins were singularly expressed in Hela cells, they showed almost no activation of TCF-dependent reporter transcription on their own. However, when Dvl protein, at the level that did not activate Wnt pathway by itself, was co-expressed with Ccd1, the reporter transcription was greatly potentiated in Ccd1-dose-dependent manner. In addition, Ccd1- and Wnt3a-dependent activation of Wnt pathway was inhibited by Axin or a dominant negative Ccd1. These results indicate that mouse Ccd1 functions as a positive regulator of the Wnt/h-catenin pathway. Furthermore, Ccd1 is highly expressed and co-localized with Wnt signaling molecules in the embryonic and adult brain, implicating the importance of Ccd1 in the Wnt-mediated neuronal development, plasticity, and remodeling. D 2004 Elsevier B.V. All rights reserved. Theme: Cellular and molecular biology Topic: Gene structure and function: general Keywords: Wnt signaling; DIX domain; Calponin homology domain; Coiled-coil domain; Isoform

1. Introduction The Wnt signaling pathway plays crucial roles in many developmental processes controlling embryonic induction, cell polarity, and cell growth [4,39]. In the nervous system,

$ The nucleotide sequences reported in this paper have been submitted to the GenBank database with accession numbers AY549883–AY549887. * Corresponding author. Fax: +81 29 853 3498. E-mail address: [email protected] (M. Masu).

0169-328X/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.molbrainres.2004.12.002

Wnt regulates development and maturation of the nervous system [3,9,25,26,30], hippocampal growth [20], proliferation and differentiation of neural progenitors [42], and neurite extension, synaptogenesis, and axonal remodeling in cerebellar neurons [14]. In the canonical Wnt pathway, when the Wnt stimulus is absent, the h-catenin degradation complex comprising Axin, APC (adenomatous polyposis coli), and GSK3h phosphorylates a key signal transducer, h-catenin, leading to its degradation by the ubiquitin proteasome system. Binding of Wnt ligands to their receptor, Frizzled, activates Dishevelled (Dvl), which then

170

K. Shiomi et al. / Molecular Brain Research 135 (2005) 169–180

blocks the activity of the h-catenin degradation complex, and accumulated h-catenin is translocated to the nucleus and interacts with TCF/LEF family transcription factors to activate Wnt target genes [23,27,31]. The key molecules in this canonical pathway are Dvl and Axin, both of which have a DIX domain, a conserved sequence of about 80 amino acids found at the N-terminus of Dvl and the Cterminus of Axin. A DIX domain is required for protein– protein interaction between Axin and Dvl themselves or between Axin, Dvl, and other proteins, and essential for canonical Wnt signal transduction [17,21,34]. Recently, we have cloned a novel zebrafish gene, coiled-coil-DIX1 (ccd1) encoding a third type of DIX domain-possessing protein [33]. Ccd1 has a coiled-coil domain at its N-terminus and a DIX domain at its Cterminus. Ccd1 interacts with Dvl and Axin through its DIX domain and also forms a homomeric complex using its coiled-coil domain and DIX domain. In vitro, Ccd1 activates TCF-dependent transcription, whereas a dominant-negative form of Ccd1 (DN-Ccd1) inhibits Ccd1- and Dvl-dependent activation of the Wnt pathway. In addition, over-expression of ccd1 in zebrafish embryos leads to a reduction of the size of the eyes and forebrain as a result of Wnt pathway activation, whereas DN-ccd1 overexpression causes the opposite phenotype. These results indicate that Ccd1 functions as a positive regulator of Wnt signaling and regulate neural patterning in zebrafish embryos [33]. Here, we report the molecular cloning and characterization of mouse Ccd1. We identify multiple isoforms including a novel subtype having a calponin homology domain. We show that Ccd1 is abundantly expressed in the embryonic and adult nervous system and activates Wnt signaling synergistically with Dvl. These data suggest that mouse Ccd1 plays a role in neural development and adult brain functions as a positive regulator of Wnt signaling.

2. Materials and methods 2.1. Animal experiments All the experimental procedures using animals were approved by the Animal Care and Use Committee of University of Tsukuba, and performed in accordance with its guidelines. Time-pregnant female C57BL/6J mice or 6week old male C57BL/6 mice were purchased from Japan SLC Inc. (Hamamatsu, Japan). The day when the presence of a vaginal plug was observed was taken as embryonic day 0 (E0). Tissues or embryos were taken after the animals were sacrificed under deep anesthesia induced by intraperitoneal injection of an excess of sodium pentobarbital. 2.2. Cloning of the mouse Ccd1 gene A 981-bp DNA fragment of mouse Ccd1 was obtained by RT–PCR using primers (Mccd1-F1 and Mccd1-R1 in Table 1) designed based on mouse EST sequences (GenBank accession numbers: BB359336, BB467918, AI480555, and AW455507) that show high homology with zebrafish ccd1. The fragment contained the sequence between exon 2 and exon 12 and is homologous to the 5V- two-thirds of the open reading frame of zebrafish ccd1. The digoxigenin (DIG)-labeled RNA probe made by in vitro transcription from the above fragment was used as a pan-Ccd1 probe to screen a Large-Insert cDNA Library of adult mouse brain (Clontech). Fourteen positive clones were isolated from 2.5  105 phage plaques and characterized. Their nucleotide sequences were determined using a Thermo Sequenase Cycle Sequencing Kit (Amersham), sequencer model 4200L (Li-Cor), and CEQ2000 (Beckman Coulter). 5V-RACE was performed using a BD SMART RACE cDNA Amplification kit (Clontech). Database searches were carried out using nucleotide and protein sequence databases (http://www.ncbi.nlm.nih.gov/BLAST/) and

Table 1 The oligonucleotides used in this study Name

Sequence

Direction

Location

Combination

Mccd1-F1

5V-ccacagtgctaagagcgaatccat-3V

Sense

(EST, exon 2)

Mccd1-R1

Mccd1-R1 Mccd1A-F1

5V-agacgctccatgaggctagagatt-3V 5V-ccaaacgggccgatagtttatgag-3V

Antisense Sense

(EST, exon 12) Exon A1

Mccd1-7R

Mccd1A-F2

5V-gtcctgcaagagggcttcaatgag-3V

Sense

Exon A3

Mccd1-7R

Mccd1B-F1 Mccd1C-F1 Mccd1-7R Mccd1-9F

5V-catgggagggacacaagtcaaatg-3V 5V-aggagcagctgacaaaggctttgt-3V 5V-atccatatttgccctcagcagctc-3V 5V-agttggaagaggcgcttcggaaac-3V

Sense Sense Antisense Sense

Exon Exon Exon Exon

Mccd1-7R Mccd1-7R

Mccd1-14R

5V-gcggtcggtgaaatagagcacctt-3V

Antisense

Exon 14

Mccd1-3R G3PDH-F G3PDH-R

5V-ggatgatggagatgctggcgca-3V 5V-ccatcaacgaccccttcattgacc-3V 5V-ctgttgaagtcgcaggagacaacc-3V

Antisense Sense Antisense

Exon 3 (cDNA) (cDNA)

B1 C1 7 9

Mccd1-14R

G3PDH-R

Size (bp) 981

1328 (Ccd1Aa1) 1249 (Ccd1Aa2) 1236 (Ccd1Ab1) 1158 (Ccd1Ab2) 561 582 515 (Ccd1L) 425 (Ccd1S)

770

Experiment Cloning of the mouse Ccd1 fragment RT–PCR (Ccd1Aa-specific) RT–PCR (Ccd1Ab-specific) RT–PCR RT–PCR RT–PCR RT–PCR

(Ccd1B-specific) (Ccd1C-specific) (Ccd1 common) (Ccd1 common)

RT–PCR (Ccd1 common), 5V-RACE 5V-RACE RT–PCR (control) RT–PCR (control)

K. Shiomi et al. / Molecular Brain Research 135 (2005) 169–180

NCBI mouse genome assembly (http://www.ensembl.org/ Mus_musculus/). The CH domain, coiled-coil domain, and DIX domain were determined using the SMART program (http://smart.embl-heiderberg.de). 2.3. Expression analysis by RT–PCR and Northern blotting For RT–PCR analyses, the cDNAs synthesized from total RNAs of mouse tissues using Superscript II Reverse Transcriptase (Invitrogen) and oligo(dT)12–18 were used. PCR primers designed to amplify the sequences common to all isoforms or specific to individual isoforms were used (summarized in Table 1). PCR was performed using AmpliTaqGold (Applied Biosystems) by incubation at 95 8C for 10 min; 26 cycles of 94 8C for 30 s, 65 8C for 30 s, 72 8C for 1 min (2 min for the Ccd1Aa and Ccd1Ab reactions); and incubation at 72 8C for 10 min. The amplified cDNA fragments were subjected to nucleotide sequence determination. Glyceraldehyde-3-phosphate dehydrogenase (G3PDH) was used as a positive control. For Northern blot analysis, 2 Ag of total RNA was fractionated by electrophoresis on a formaldehyde-agarose gel, transferred to a Biodyne A membrane (Pall), and hybridized with a DIG-labeled pan-Ccd1 probe (nucleotides 1323 to 2341 of the Ccd1Aa1L) at 65 8C for 16 h. After washing with 2 SSC, 0.1% SDS at room temperature for 5 min twice, and with 0.1 SSC, 0.1% SDS at 68 8C for 15 min twice, the membrane was incubated with an alkaline phosphatase (AP)-conjugated anti-DIG antibody (Roche Diagnostics). The chemiluminescence reaction was performed using CSPD reagent (Roche Diagnostics). 2.4. In situ hybridization In situ hybridization was performed as previously described [24]. E9.5 and E13.5 mouse embryos were fixed with 4% paraformaldehyde (PFA) in phosphate buffered saline (PBS) at 4 8C overnight. Adult mice were transcardially fixed with 4% PFA/PBS under deep anesthesia, and dissected brains were immersed in the same fixative at 48 C overnight. After cryoprotection with 30% sucrose in PBS, 10-Am sections were made using a cryostat. The sections were treated with 1 Ag/ml proteinase K in PBS with 0.1% Tween-20 (PBT) at 37 8C for 5 min, washed and fixed with 4% PFA/PBS, and then hybridized with 100 ng/ml of a DIG-labeled pan-Ccd1 RNA probe (nucleotides 1323 to 2341 of the Ccd1Aa1L) in hybridization solution (50% formaldehyde, 5 SSC-pH 4.5, 1% SDS, 50 Ag/ml heparin, 50 Ag/ml yeast RNA) at 65 8C for 16 h. They were washed with 50% formaldehyde, 5 SSC, 1% SDS at 65 8C for 30 min, and 50% formaldehyde, 2 SSC at 65 8C for 30 min three times, and then incubated with an AP-conjugated antiDIG antibody (Roche) at 4 8C overnight. After the sections were washed with Tris-buffered saline containing 0.1% Tween-20 (TBST), signals were detected using a precipitating AP substrate, BM purple (Roche Diagnostics), in the

171

presence of 2 mM Levamisole (Sigma) at room temperature for 1–5 days. Slides were embedded with Aqua Poly/Mount (Polysciences), and observed and photographed with an Axioplan2 microscope (Carl Zeiss). Two non-overlapping antisense probes (nucleotides 1323 to 2341 and 2342 to 3707 of the Ccd1Aa1L) gave rise to essentially identical signals, whereas sense probes, used as negative controls, resulted in no signal. 2.5. Western blotting For heterologous expression of Ccd1 protein in Hela cells, Ccd1 cDNAs tagged with FLAG sequence at the Ctermini of their ORFs were subcloned into pCS2+ vector. Hela cells were transfected individually with the Ccd1Aa1L, Ccd1BaL, and Ccd1CL expression constructs for 27 h using LipofectAMINE PLUS (Invitrogen) according to the manufacturer’s protocol. Mouse tissues and transfected Hela cells were homogenized and sonicated in an SDS sample buffer (50 mM Tris–HCl pH 6.8, 2% SDS, Complete Mini Protease Inhibitor Cocktail (Roche)). The soluble proteins collected by centrifugation were resolved by SDS polyacrylamide gel electrophoresis and transferred to an Immobilon PVDF membrane (Millipore). The membrane was serially treated with anti-Ccd1(N2) or anti-Ccd1(C1) antibodies diluted 1:500, an HRP-conjugated anti-rabbit IgG (Bio-Rad) diluted 1:2000, and an ECL kit (Amersham) for detection of immunopositive bands. The anti-Ccd1(N2) and anti-Ccd1(C1) antibodies were generated by immunizing rabbits with the KLH-conjugated peptides HGSLPEDEQTSTLSC and CWEGKIVAWVEEDH, respectively (Transgenic, Inc.). Anti-Ccd1(N2) and anti-Ccd1(C1) react with the sequences in the coiled-coil domain (encoded by exon 4) and the C-terminus (encoded by exon 16), which are present in all the Ccd1 isoforms. 2.6. Luciferase reporter assay Transfection of Hela cells was performed in duplicate in a 24-well plate using FuGENE 6 transfection reagent (Roche Diagnostics). For each well, 200 ng of the firefly luciferase reporter plasmid (TOPtkLuciferase or FOPtkLuciferase [19]), 10 ng of Renilla luciferase as an internal control, and 200 ng of one of the following plasmids in the absence or presence of 20 ng or 200 ng of Xenopus Dvl were used. The constructs used were mouse Ccd1Aa1L, mouse Ccd1BaL, mouse Ccd1CL, zebrafish ccd1, dominant-negative Ccd1 (DN-Ccd1) [33], zebrafish Axin1, and nuclear h-galactosidase cDNAs. The total amount of the plasmids in each well was adjusted to 600 ng using a nuclear h-galactosidase cDNA. For stimulation by a Wnt ligand, Wnt3a conditioned medium was added to the cells 6 h after transfection. Luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega) 27 h after transfection. Firefly luciferase activity was normalized by Renilla luciferase activity.

172

K. Shiomi et al. / Molecular Brain Research 135 (2005) 169–180

3. Results 3.1. Multiple isoforms of mouse Ccd1 cDNAs are generated by different promoter usage and alternative splicing To isolate mouse Ccd1, we first amplified a mouse cDNA fragment that is homologous to the zebrafish ccd1 by RT–PCR using mouse EST sequences in the databases. We then screened an adult mouse brain cDNA library and isolated 14 cDNA clones. Sequence determination revealed that they were identical except for some sequence variations in their 5V regions (described below). We determined the 5V ends of the mouse Ccd1 cDNAs by sequencing cDNA and 5V-RACE clones. We next examined exon–intron organization of Ccd1 gene by comparing our cDNA sequences with the genomic sequences obtained from the public database (NCBI build 30 mouse assembly). This analysis revealed that the mouse Ccd1 gene consists of 25 exons spanning 77 kb in chromosome 9 (Fig. 1A and Table 2). Exons A1–A7, B1–B2, and C1 are specifically used in the Ccd1A, Ccd1B, and Ccd1C isoforms, respectively, while exons 2–16 are commonly used in all Ccd1 isoforms. Fig. 1B shows 14 putative mRNA isoforms, which are generated by usage of multiple transcription start sites and alternative splicing.

3.2. A novel Ccd1 subtype has an N-terminal calponin homology domain Fig. 2A shows the structure of Ccd1 proteins predicted from the putative mRNA isoforms. They are classified into three subtypes, Ccd1A, Ccd1B, and Ccd1C, on the basis of their N-terminal sequences. Ccd1B corresponds to the original form of zebrafish Ccd1, and Ccd1C partially lacks the N-terminal portion. Ccd1A possesses an N-terminal calponin homology (CH) domain in addition to the coiled-coil and DIX domains that are present in all isoforms. In the Ccd1A group, Ccd1Aa and Ccd1Ah have specific N-terminal sequences of 19 amino acids (encoded by exon A2) and 20 amino acids (encoded by exon A3), respectively. Ccd1Aa2 and Ccd1Ah2 have 26-amino-acid deletions in the CH domain compared to Ccd1Aa1 and Ccd1Ah1, respectively. In the Ccd1B group, Ccd1Ba and Ccd1Bh have specific N-terminal sequences of 8 amino acids (encoded by exon B1) and 11 amino acids (encoded by exon B2), respectively. In all subtypes, there are short (S) and long (L) forms: the S forms have 30-amino-acid deletions compared with the corresponding L forms in the region between the coiled-coil domain and the DIX domain.

Fig. 1. Structure of mouse Ccd1 gene and its multiple transcripts. (A) Genomic organization of mouse Ccd1 gene. Vertical lines and boxes indicate exons. A1– A7, B1–B2, and C1 denote Ccd1A, Ccd1B, and Ccd1C-specific exons, respectively, while 2–16 indicate the exons common to all isoforms. Arrows indicate four transcription start sites. (B) Putative mRNA isoforms generated by usage of different transcription start sites and alternative splicing. Dashed lines illustrate splicing, while bold and thin lines indicate protein-coding regions and non-coding regions, respectively. Alternative splicing in exon A4 generates a1/a2 and h1/h2 isoforms of Ccd1A, whereas the presence or absence of exon 11 generates long (L) and short (S) forms.

K. Shiomi et al. / Molecular Brain Research 135 (2005) 169–180

173

Table 2 Genomic organization of the mouse Ccd1 gene Exon a

Intron

Amino acid

Number

Position (nucleotide)

Size (bp)

Size (bp)

Ccd1Aa1L (aa)

Ccd1BaL (aa)

Ccd1CL (aa)

A1 A2 A3 A4 A5 A6 A7 B1 C1 B2 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

50,862,453–50,862,374 50,860,843–50,860,752 50,851,083–50,850,894 50,834,078–50,833,871 50,829,310–50,829,185 50,826,403–50,826,172 50,825,077–50,824,970 50,822,073–50,821,608 50,820,828–50,820,640 50,819,878–50,819,777 50,818,633–50,818,521 50,816,741–50,816,593 50,812,959–50,812,870 50,811,666–50,811,613 50,810,421–50,810,371 50,807,942–50,807,838 50,807,417–50,807,364 50,806,976–50,806,884 50,806,806–50,806,735 50,805,636–50,805,547 50,805,210–50,805,043 50,804,563–50,804,497 50,792,564–50,792,457 50,790,777–50,790,669 50,789,284–50,785,761

79b 92 190b 208 126 232 108 466b 89b 102 113 149 90 54 51 105 54 93 72 90 168 67 106 109 3524

1530 26,673c 16,815d 4560e 2781 1094 6336f 2974f 2006f 1143f 1779 3633e 1203 1191 2428 422 387 77 1098 336 479 11,932 1681 1384

5V-UTR 5V-UTR, 1–19 – 20–88 89–130 131–207 208–243 – – – 244–281 282–331 332–361 362–379 380–396 397–431 432–449 450–480 481–504 505–534 535–590 591–612 613–647 648–684 685–710, 3V-UTR

– – – – – – – 5V-UTR, 1–8 – – 9–45 46–95 96–125 126–143 144–160 161–195 196–213 214–244 245–268 269–298 299–354 355–376 377–411 412–448 449–474, 3V-UTR

– – – – – – – – 5V-UTR – 5V-UTR 5V-UTR, 1–12 13–42 43–60 61–77 78–112 113–130 131–161 162–185 186–215 216–271 272–293 294–328 329–365 366–391, 3V-UTR

a b c d e f

Indicates the region in the mouse chromosome 9 (in the contig NT 039473 registered as gi:28490892). Indicates the largest sizes of the first exons determined by cDNA cloning, 5V-RACE, or EST search. Indicates the sizes of the introns between exon A2 and exon A4. Indicates the sizes of the introns between exon A3 and exon A4. Indicates the sizes of the intron that contain undetermined sequences. Indicates the sizes of the introns between exons A7/B1/C1/B2 and exon 2.

Fig. 2B shows the amino acid sequence alignment of the CH domains of mouse Ccd1A, h-spectrin, dystrophin, and a-actinin3. Amino acid identity between Ccd1 and the others is 25–27%. The CH domain in Ccd1A shows highest homology to the type 1 CH domain among the five types of CH domains (data not shown). Fig. 2C shows the amino acid sequence alignment of the DIX domains. Amino acid identity between mouse Ccd1 and zebrafish Ccd1, mouse Dvl1, and mouse Axin1 is 83%, 35% and 35%, respectively. 3.3. Ccd1 isoform mRNAs are differentially expressed in mouse tissues To assess the possible involvement of Ccd1 in development and tissue-specific functions, we studied the temporal and spatial expression of mouse Ccd1 mRNA. First, RT– PCR using a primer set that amplified all isoforms revealed that total Ccd1 expression was abundant in the brain and testis, and moderate in the lung, kidney, colon, ovary, and urinary bladder in adult tissues (Fig. 3A). Ccd1 was also detected in the whole embryos from embryonic day 9.5 (E9.5) to E17.5, with relatively high expression at E9.5 and decreased expression along with development (Fig. 3B). The L form predominated over the S form in all the tissues

examined, while the ratio of S/L was relatively high in early embryos. Next, the expression of individual isoforms was examined by RT–PCR using primer sets that selectively amplified Ccd1Aa, Ccd1Ab, Ccd1B, and Ccd1C. Ccd1Aa was robustly expressed in the testis, whereas Ccd1Ab was expressed in the brain, testis, ovary, and whole embryos (Fig. 3). In Ccd1Aa and Ccd1Ab, the shorter forms, Ccd1Aa2 and Ccd1Ab2, were dominant. Ccd1B was abundantly expressed in the adult brain, lung, colon, testis, and urinary bladder, while Ccd1C was hardly detected in the brain, testis, and urinary bladder (Fig. 3). These results indicate that the predominant forms of Ccd1 in the 14 putative isoforms are Ccd1Aa2L, Ccd1Ab2L, and Ccd1BaL. Next, Northern blot with a pan-Ccd1 probe was performed: it detected major bands of 6.4 kb and/or 6.0 kb and a minor band of 4.3 kb. Strong signals were detected in the adult brain and testis, and in early embryos, while weak signals were seen in the adult lung, colon, ovary, and urinary bladder, and in late embryos (Fig. 3). The 6.4 kb- and 6.0 kbbands appear to correspond to Ccd1A and Ccd1B mRNA, respectively, because the tissue distributions of the 6.4 kbband (the adult brain, testis, and ovary, and whole embryos) and of the 6.0 kb-band (the adult brain, lung, colon, and urinary bladder) are in good agreement with those of Ccd1A

174

K. Shiomi et al. / Molecular Brain Research 135 (2005) 169–180

Fig. 2. Structure of mouse Ccd1 protein. (A) Structures of the mouse Ccd1 proteins predicted from putative cDNA sequences. CH: calponin homology domain, CC: coiled-coil domain, DIX: DIX domain. The size (amino acid residues) and calculated molecular weight (kDa) of each protein are shown on the right. (B) Amino acid sequence alignment of the CH domains of mouse Ccd1Aa2 (Ccd1), h-spectrin, dystrophin, and a-actinin3. (C) Amino acid sequence alignment of the DIX domains of mouse Ccd1Aa1L (Ccd1), zebrafish Ccd1 (ZCcd1), mouse Dvl1, and mouse Axin1. Asterisks indicate stop codons. Shading indicates the amino acids conserved among more than two sequences. Numbers indicate the location of the residues in each amino acid sequence.

and Ccd1B mRNA determined by RT–PCR, and Ccd1A cDNA is approximately 400 bp longer than Ccd1B cDNA. 3.4. Expression of Ccd1 protein We next examined the expression of Ccd1 protein using Western blot with polyclonal antibodies that were generated

against Ccd1 sequences (anti-Ccd1(N2) and anti-Ccd1(C1)). Fig. 4A shows the results obtained with anti-Ccd1(C1), while similar results were observed with anti-Ccd1(N2), indicating their specificity (data not shown). When the crude extracts of Hela cells transfected with Ccd1Aa1L expression constructs were analyzed, two bands of 84 kDa and 52 kDa were detected. The size of the larger band is close to the

K. Shiomi et al. / Molecular Brain Research 135 (2005) 169–180

175

Fig. 3. Expression profile of mouse Ccd1 mRNA. (A–B) RT–PCR and Northern blot analyses of mouse Ccd1 mRNA in adult tissues (A) and whole embryos (B). The bands obtained by a primer set that amplified all isoforms (uppermost column) and primers that were specific to Ccd1Aa, Ccd1Ab, Ccd1B, and Ccd1C (lower columns) are shown. Upper and lower bands in the Ccd1, Ccd1Aa, and Ccd1Ah reactions correspond to the L form and the S form, Ccd1Aa1 and Ccd1Aa2, and Ccd1Ab1 and Ccd1Ab2, respectively. (–) indicates PCR with no cDNA. Glyceraldehyde-3-phosphate dehydrogenase (G3PDH) is a positive control. In Northern blots with a pan-Ccd1 probe, multiple bands of about 6.4-kb, 6.0-kb, and 4.3-kb were detected.

calculated molecular weight of Ccd1Aa1L protein (80 kDa), while the smaller bands may correspond to the proteins produced from a cryptic translation start site or by proteolysis from a longer form, although the actual origin of this band remains unknown. When Ccd1BaL and Ccd1CL were analyzed, bands of about 51 kDa and 43 kDa were detected, respectively. These sizes were very close to the calculated molecular weights of Ccd1BaL (54 kDa) and Ccd1CL proteins (45 kDa), respectively. When mouse tissue extracts were analyzed, a major band of about 52 kDa was strongly detected in the adult brain, adult testis, and embryonic brain (Fig. 4A). The band appears to correspond to Ccd1B protein and/or the smaller band observed with the recombinant Ccd1A expression.

Weak bands of about 100 kDa and 70 kDa, which could have been generated from Ccd1A mRNA, were observed in embryonic tissues. The band of about 33 kDa detected with anti-Ccd1(C1) antibody appears to be a degraded fragment of Ccd1A and/or Ccd1B protein or cross-reactivity with an uncharacterized protein, because the band was not detected with anti-Ccd1(N2) antibody (data not shown). 3.5. Mouse Ccd1 potentiates Dvl-dependent Wnt signaling in vitro We then tested whether mouse Ccd1 can activate the Wnt signaling pathway in vitro by using a luciferase reporter assay. The effects on TCF-dependent transcription were

176

K. Shiomi et al. / Molecular Brain Research 135 (2005) 169–180

determined in HeLa cells transfected with mouse Ccd1A, Ccd1B, Ccd1C, and zebrafish ccd1 expression constructs (200 ng for each) in the absence or presence of Dvl (20 ng

or 200 ng) together with a luciferase reporter plasmid. Figs. 4B–D show the fold activation of the TCF-dependent transcription, which was calculated by dividing the luciferase activity obtained with a TOPtkLuciferase reporter plasmid (containing multimeric TCF binding sites) by that obtained with FOPtkLuciferase (containing mutated TCF binding sites) [19]. Transfection of mouse Ccd1A or Ccd1C alone did not activate TCF-dependent transcription significantly compared with a negative control, while transfection of mouse Ccd1B alone led to weak but significant activation (Fig. 4B). This is in contrast with robust activation (~6.6fold activation) obtained with the expression zebrafish ccd1 alone (Fig. 4B). However, when co-transfected with 20 ng of a Dvl expression plasmid that did not significantly activate the reporter activity by itself at this low dose, mouse Ccd1A, Ccd1B, or Ccd1C increased the luciferase activity 4.6-, 8.5-, and 5.4-fold, respectively (Fig. 4B). In addition, co-transfection of mouse Ccd1A, Ccd1B, or Ccd1C together with 200 ng of Dvl resulted in an increase of the luciferase activity that was comparable to that obtained with transfection of 200 ng of Dvl alone (Fig. 4B), indicating that Ccd1 and Dvl function in the same signaling pathway. Next, to examine the differences of Ccd1 isoforms in the extent of the Dvl-dependent Wnt activation, titration experiments (cotransfection of 20–200 ng of Ccd1 cDNAs with 20 ng of Dvl) were performed. These results revealed that Wnt pathway activation in the presence of a low dose of Dvl is dependent on Ccd1 protein expression levels and that potency of the reporter activation is different between Ccd1 isoforms (Fig. 4C). Furthermore, Wnt pathway activation by Ccd1 transfection with 20 ng of Dvl or by the addition of Wnt3a conditioned medium was inhibited by co-expression of Axin or DN-Ccd1 (Fig. 4D), indicating

Fig. 4. Ccd1 protein expression and activation of TCF-dependent transcription. (A) Western blot analysis of the Ccd1 protein. Crude lysates prepared from Hela cells transfected with Ccd1Aa1L (Ccd1A), Ccd1Ba1L (Ccd1B), or Ccd1CL (Ccd1C) (left), and lysates from the brain (B), liver (L), spleen (S), and testis (T) of adult mice, and the embryonic brain (EB), liver (EL), and spleen (ES) of E15.5 mice (right) were analyzed. Sizes (kDa) of the markers are shown on the left. (B–D) Activation of TCFdependent transcription determined by a luciferase reporter assay. Values shown are means F SEM from three independent experiments performed in duplicate. Nuclear h-galactosidase (hgal) was used as a negative control. (B) The effects of Ccd1A, Ccd1B, Ccd1C, and zebrafish ccd1 (ZCcd1) in the absence and presence of 20 ng or 200 ng of the Dishevelled (Dvl) expression construct are shown. * and ** indicate statistical significance ( P b 0.05 and P b 0.01 compared with a negative control in the absence of Dvl, resepectively) by Student’s t test. (C) The Ccd1-mediated potentiation in the presence of 20 ng Dvl is dependent on the Ccd1 protein levels. The lower panel shows the Western blot of the Ccd1 proteins in the cell lysates used in the luciferase reporter assay for the upper panel. ** indicates statistical significance ( P b 0.01 compared with a negative control) by Student’s t test. (D) Axin and a dominant-negative Ccd1 (DN-Ccd1) inhibit the Wnt pathway activation induced by the addition of Wnt3a conditioned medium or transfection of Ccd1 in the presence of 20 ng of Dvl. * and ** indicate statistical significance ( P b 0.05 and P b 0.01, respectively) by Student’s t test.

K. Shiomi et al. / Molecular Brain Research 135 (2005) 169–180

that Ccd1 functions downstream of Wnt receptor activation and upstream of Axin. 3.6. Expression of Ccd1 mRNA in the mouse brain Finally, we examined Ccd1 mRNA expression in the mouse brain by in situ hybridization using the pan-Ccd1 probes that detected all the isoforms. Two non-overlapping antisense probes gave rise to identical patterns (data not shown), whereas sense probes, used as negative controls, resulted in no signal (Fig. 5B). In E9.5 embryos, Ccd1 mRNA was abundantly expressed in the mesenchymes surrounding the spinal cord and brain vesicles (Figs. 5A, C). In the spinal cord, Ccd1 was detected in a few cells in the ventral marginal zone, which appeared to be motor neurons (Fig. 5A). In the brain region of E13.5 embryos, Ccd1

177

mRNA was abundantly expressed in the cortical plate of the cerebral cortex (Figs. 5D–F). Ccd1 was also strongly expressed in the dorsal thalamus (Figs. 5D, F), and also detected in the cerebellum, medulla, spinal cord, choroid plexus, superior colliculus, and the inner layer of the retina (Figs. 5D–G, data not shown). In the adult mouse, Ccd1 expression was observed throughout the brain. Strong signals were detected in the main and accessory olfactory bulbs, cerebral cortex, piriform cortex, hippocampus, habenular nucleus, dorsal thalamus (including medial geniculate nucleus), superior and inferior colliculi, and cerebellum (Fig. 6). In the cerebral cortex, a strong signal was detected in layers II–IV and VI (Fig. 6G). In the hippocampus, Ccd1 was expressed in the pyramidal neurons in the CA1–4 fields and the granule cells in the dentate gyrus (Figs. 6H, I). In the

Fig. 5. Ccd1 mRNA expression in mouse embryos. Expression of Ccd1 mRNA at E9.5 (A–C) and E13.5 (D–G) embryos revealed by in situ hybridization. The transverse sections of the spinal cord (A–B), the horizontal sections of the forebrain region (C, E), the diencephalic region (F), and the eye (G), and a sagittal section of the head region (D) are shown. Panel (B) is a negative control hybridized with a sense probe. Abbreviations: Cb: cerebellum, Chp: choroid plexus, Cx: cerebral cortex, Di: diencephalon, DT: dorsal thalamus, IR: infundibular recess, MGE: medial ganglionic eminence, L: lens, LGE: lateral ganglionic eminence, LV: lateral ventricle, MN: motor neuron, OD: optic disc, OP: olfactory placode, OV: optic vesicle, P: pigment epithelium, SC: superior colliculus, TV: telencephalic vesicle, V: ventricle. Scale bar: 280 Am in (A–B), 560 Am in panel (C), 1600 Am in panel (D), 400 Am in panels (E–F), and 200 Am in panel (G).

178

K. Shiomi et al. / Molecular Brain Research 135 (2005) 169–180

Fig. 6. Ccd1 mRNA expression in the adult mouse brain. Expression of Ccd1 mRNA in coronal sections of the olfactory bulb (A), cerebrum (B–E, G–I), and cerebellum (F, J) is revealed by in situ hybridization. (G–J) Higher magnification views of the cerebral cortex (G), hippocampus (H), hippocampal CA1 field (I), and cerebellar cortex (J). Abbreviations: I–VI: layers I–VI of the cerebral cortex, 3V: third ventricle, 4V: fourth ventricle, AOB: accessory olfactory bulb, Aq: aqueduct, Cb: cerebellum, CPu: caudate putamen, Cx: cerebral cortex, DG: dentate gyrus, GCL: granule cell layer, Hb: habenular nucleus, Hip: hippocampus, IC: inferior colliculus, LV: lateral ventricle, MG: medial geniculate nucleus, ML: molecular layer, OB: olfactory bulb, Or: stratum oriens, P: pons, PCL: Purkinje cell layer, Pir: piriform cortex, Py: stratum pyramidale, Rad: stratum radiatum, SC: superior colliculus. Scale bar: 600 Am in panel (A), 1600 Am in panels (B–F), 300 Am in panel (G), 400 Am in panel (H), 50 Am in panels (I–J).

cerebellum, Ccd1 was highly expressed in the granule cells (Fig. 6J).

4. Discussion We have cloned and characterized mouse Ccd1. The orthologues are found in human, rat, and zebrafish but not in invertebrates, like some Wnt signaling molecules, such as GBP, Frodo, and Dpr [6,13,15,16,29,41], suggesting that Wnt signaling in vertebrates is regulated, in part, in a manner distinct from invertebrates and that Ccd1 plays a role in vertebrate specific signaling. A novel isoform, Ccd1A, possesses a CH domain. The CH domain is found in a variety of actin binding proteins, cytoskeletal proteins, and signaling molecules, and classified into five types on the basis of sequence homology, location in the protein, and function [12,18,35]. The type 1 CH domain, which is found in the N-termini of actinbinding proteins such as a-actinin and h-spectrin, is usually

followed by a type 2 CH domain. The combination of a type 1 and a type 2 CH domains forms a strong actin-binding region, while a single type 1 CH domain alone can bind to actin very weakly [37]. The CH domain in Ccd1A shows highest homology to the type 1 CH domain and lacks the following type 2 CH domain, suggesting that it mediates weak actin binding. Mouse Ccd1 proteins show almost no activation of TCF-dependent transcription on their own, which is in contrast to zebrafish Ccd1 that activates Wnt signaling by itself. Mouse Ccd1 thus functions in a similar fashion to Xenopus Frodo, a positive regulator of Wnt signaling that binds to and act synergistically with Dvl in secondary axis induction and neural development [13]. Axin and DNCcd1 inhibited the Ccd1- or Wnt3a-mediated Wnt signal activation, suggesting that Ccd1 functions downstream of Wnt receptor stimulation and upstream of Axin. Therefore, Ccd1 is thought to regulate Wnt/h-catenin pathway by activating Dvl and inhibiting Axin through somehow regulating their subcellular localization, phosphorylation,

K. Shiomi et al. / Molecular Brain Research 135 (2005) 169–180

stability, or binding with other proteins in the canonical Wnt pathway [1,5,8,10,28,38]. These results, however, do not exclude the possibility that Ccd1 plays a role in a noncanonical Wnt pathway or other signaling pathways. Indeed, recently Wong et al. reported that Ccd1 inhibits c-Jun N-terminal kinase activation by Axin and Dvl through distinct mechanisms [40]. Further studies will be necessary to elucidate the molecular mechanism of Ccd1 function. Ccd1 mRNA was expressed at high levels in the adult brain, testis, and lung, and embryos, where Wnt genes are highly expressed [11]. Because Ccd1 was abundantly expressed in the embryonic brain, it might be related to Wnt signaling in neural development and maturation [3,9,14,20,26,30,42]. Ccd1 expression in the early-born neurons suggests a role in the differentiation and maturation of postmitotic neurons. Co-expression of Ccd1 and Tcf4 in the dorsal thalamus of the embryos [7,36] and an evidence that a Wnt signal is necessary for specifying dorsal thalamus identity [2] suggest that Ccd1 may be involved in thalamic specification. In addition, persistent Ccd1 expression in the adult brain is intriguing. Ccd1 was highly expressed in the olfactory bulb, hippocampus, neocortex, and dorsal thalamus, where several Wnt, Frizzled, and secreted Frizzled-related proteins are strongly and specifically expressed [32]. In addition, a mapping study of Wnt/h-catenin signaling revealed that the brain is the most active tissue for Wnt signaling in the adult [22]. The Wnt-responsive brain regions are the hippocampus, sensory cortex, several thalamic nuclei, colliculi, and cerebellar cortex [22], which are overlapping with the Ccd1-expressing regions. These results implicate the possible role of Ccd1 in the Wnt signaling in neuronal plasticity, remodeling of axons and dendrites, new synapse formation, and neurogenesis. Further studies will be required to elucidate the physiological function of Ccd1 gene in the nervous system.

[2]

[3] [4] [5]

[6]

[7]

[8]

[9] [10]

[11]

[12] [13] [14]

[15]

[16]

[17]

Acknowledgments We thank T. Akiyama for TOPtkLuciferase and FOPtk Luciferase, M. Tada for Xenopus Dishevelled, M. Hibi for zebrafish Axin1, S. Takada for Wnt3a conditioned medium, and Y. Hatanaka and M. Yamamoto for critical reading of the manuscript. This work was supported by Grants-in-Aid for Scientific Research on Priority Areas and the 21st century COE program from the Ministry of Education, Culture, Sports, Science and Technology, Japan.

References [1] J.D. Axelrod, J.R. Miller, J.M. Shulman, R.T. Moon, N. Perrimon, Differential recruitment of Dishevelled provides signaling specificity

[18] [19]

[20]

[21]

[22]

179

in the planar cell polarity and Wingless signaling pathways, Genes Dev. 12 (1998) 2610 – 2622. M.M. Braun, A. Etheridge, A. Bernard, C.P. Robertson, H. Roelink, Wnt signaling is required at distinct stages of development for the induction of the posterior forebrain, Development 130 (2003) 5579 – 5587. S.J. Burden, Wnts as retrograde signals for axon and growth cone differentiation, Cell 100 (2000) 495 – 497. K.M. Cadigan, R. Nusse, Wnt signaling: a common theme in animal development, Genes Dev. 11 (1997) 3286 – 3305. D.G. Capelluto, T.G. Kutateladze, R. Habas, C.V. Finkielstein, X. He, M. Overduin, The DIX domain targets dishevelled to actin stress fibres and vesicular membranes, Nature 419 (2002) 726 – 729. B.N. Cheyette, J.S. Waxman, J.R. Miller, K. Takemaru, L.C. Sheldahl, N. Khlebtsova, E.P. Fox, T. Earnest, R.T. Moon, Dapper, a dishevelled-associated antagonist of h-catenin and JNK signaling, is required for notochord formation, Dev. Cell 2 (2002) 449 – 461. E.A. Cho, G.R. Dressler, TCF-4 binds h-catenin and is expressed in distinct regions of the embryonic brain and limbs, Mech. Dev. 77 (1998) 9 – 18. A. Cliffe, F. Hamada, M. Bienz, A role of dishevelled in relocating axin to the plasma membrane during wingless signaling, Curr. Biol. 13 (2003) 960 – 966. G.W. Davis, The making of a synapse: target-derived signals and presynaptic differentiation, Neuron 26 (2000) 551 – 554. F. Fagotto, E. Jho, L. Zeng, T. Kurth, T. Joos, C. Kaufmann, F. Costantini, Domains of axin involved in protein–protein interactions, Wnt pathway inhibition, and intracellular localization, J. Cell Biol. 145 (1999) 741 – 756. B.J. Gavin, J.A. McMahon, A.P. McMahon, Expression of multiple novel Wnt-1/int-1-related genes during fetal and adult mouse development, Genes Dev. 4 (1990) 2319 – 2332. M. Gimona, K. Djinovic-Carugo, W.J. Kranewitter, S.J. Winder, Functional plasticity of CH domains, FEBS Lett. 513 (2002) 98 – 106. J. Gloy, H. Hikasa, S.Y. Sokol, Frodo interacts with Dishevelled to transduce Wnt signals, Nat. Cell Biol. 4 (2002) 351 – 357. A.C. Hall, F.R. Lucas, P.C. Salinas, Axonal remodeling and synaptic differentiation in the cerebellum is regulated by WNT-7a signaling, Cell 100 (2000) 525 – 535. J. Jonkers, H.C. Korswagen, D. Acton, M. Breuer, A. Berns, Activation of a novel proto-oncogene, Frat1, contributes to progression of mouse T-cell lymphomas, EMBO J. 16 (1997) 441 – 450. J. Jonkers, R. van Amerongen, M. van der Valk, E. RobanusMaandag, M. Molenaar, O. Destree, A. Berns, In vivo analysis of Frat1 deficiency suggests compensatory activity of Frat3, Mech. Dev. 88 (1999) 183 – 194. S. Kishida, H. Yamamoto, S. Hino, S. Ikeda, M. Kishida, A. Kikuchi, DIX domains of Dvl and axin are necessary for protein interactions and their ability to regulate h-catenin stability, Mol. Cell. Biol. 19 (1999) 4414 – 4422. E. Korenbaum, F. Rivero, Calponin homology domains at a glance, J. Cell Sci. 115 (2002) 3543 – 3545. V. Korinek, N. Barker, P.J. Morin, D. van Wichen, R. de Weger, K.W. Kinzler, B. Vogelstein, H. Clevers, Constitutive transcriptional activation by a h-catenin–Tcf complex in APC / colon carcinoma, Science 275 (1997) 1784 – 1787. S.M. Lee, S. Tole, E. Grove, A.P. McMahon, A local Wnt-3a signal is required for development of the mammalian hippocampus, Development 127 (2000) 457 – 467. L. Li, H. Yuan, C.D. Weaver, J. Mao, G.H.r. Farr, D.J. Sussman, J. Jonkers, D. Kimelman, D. Wu, Axin and Frat1 interact with Dvl and GSK, bridging Dvl to GSK in Wnt-mediated regulation of LEF-1, EMBO J. 18 (1999) 4233 – 4240. S. Maretto, M. Cordenonsi, S. Dupont, P. Braghetta, V. Broccoli, A.B. Hassan, D. Volpin, G.M. Bressan, S. Piccolo, Mapping Wnt/betacatenin signaling during mouse development and in colorectal tumors, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 3299 – 3304.

180

K. Shiomi et al. / Molecular Brain Research 135 (2005) 169–180

[23] R.T. Moon, B. Bowerman, M. Boutros, N. Perrimon, The promise and perils of Wnt signaling through h-catenin, Science 296 (2002) 1644 – 1646. [24] T. Ohto, H. Uchida, H. Yamazaki, K. Keino-Masu, A. Matsui, M. Masu, Identification of a novel nonlysosomal sulphatase expressed in the floor plate, choroid plexus and cartilage, Genes Cells 7 (2002) 173 – 185. [25] M. Packard, E.S. Koo, M. Gorczyca, J. Sharpe, S. Cumberledge, V. Budnik, The Drosophila Wnt, wingless, provides an essential signal for pre- and postsynaptic differentiation, Cell 111 (2002) 319 – 330. [26] A. Patapoutian, L.F. Reichardt, Roles of Wnt proteins in neural development and maintenance, Curr. Opin. Neurobiol. 10 (2000) 392 – 399. [27] M. Peifer, P. Polakis, Wnt signaling in oncogenesis and embryogenesis—A look outside the nucleus, Science 287 (2000) 1606 – 1609. [28] U. Rothbacher, M.N. Laurent, M.A. Deardorff, P.S. Klein, K.W. Cho, S.E. Fraser, Dishevelled phosphorylation, subcellular localization and multimerization regulate its role in early embryogenesis, EMBO J. 19 (2000) 1010 – 1022. [29] T. Saitoh, J. Moriwaki, J. Koike, A. Takagi, T. Miwa, K. Shiokawa, M. Katoh, Molecular cloning and characterization of FRAT2, encoding a positive regulator of the WNT signaling pathway, Biochem. Biophys. Res. Commun. 281 (2001) 815 – 820. [30] P.C. Salinas, Synaptogenesis: Wnt and TGF-h take centre stage, Curr. Biol. 13 (2003) R60 – R62. [31] M.J. Seidensticker, J. Behrens, Biochemical interactions in the wnt pathway, Biochim. Biophys. Acta 1495 (2000) 168 – 182. [32] T. Shimogori, J. VanSant, E. Paik, E.A. Grove, Members of the Wnt, Fz, and Frp gene families expressed in postnatal mouse cerebral cortex, J. Comp. Neurol. 473 (2004) 496 – 510. [33] K. Shiomi, H. Uchida, K. Keino-Masu, M. Masu, Ccd1, a novel protein with a DIX domain, is a positive regulator in the Wnt

[34]

[35] [36]

[37]

[38]

[39] [40]

[41]

[42]

signaling during zebrafish neural patterning, Curr. Biol. 13 (2003) 73 – 77. M.J. Smalley, E. Sara, H. Paterson, S. Naylor, D. Cook, H. Jayatilake, L.G. Fryer, L. Hutchinson, M.J. Fry, T.C. Dale, Interaction of Axin and Dvl-2 proteins regulates Dvl-2-stimulated TCF-dependent transcription, EMBO J. 18 (1999) 2823 – 2835. T. Stradal, W. Kranewitter, S.J. Winder, M. Gimona, CH domains revisited, FEBS Lett. 431 (1998) 134 – 137. Y. Suda, Z.M. Hossain, C. Kobayashi, O. Hatano, M. Yoshida, I. Matsuo, S. Aizawa, Emx2 directs the development of diencephalon in cooperation with Otx2, Development 128 (2001) 2433 – 2450. M. Way, B. Pope, A.G. Weeds, Evidence for functional homology in the F-actin binding domains of gelsolin and a-actinin: implications for the requirements of severing and capping, J. Cell Biol. 119 (1992) 835 – 842. K.A. Wharton Jr., Runnin’ with the Dvl: proteins that associate with Dsh/Dvl and their significance to Wnt signal transduction, Dev. Biol. 253 (2003) 1 – 17. A. Wodarz, R. Nusse, Mechanisms of Wnt signaling in development, Annu. Rev. Cell Dev. Biol. 14 (1998) 59 – 88. C.K. Wong, W. Luo, Y. Deng, H. Zou, Z. Ye, S.-C. Lin, The DIX domain protein coiled-coil-DIX1 inhibits c-Jun N-terminal kinase activation by axin and dishevelled through distinct mechanisms, J. Biol. Chem. 279 (2004) 39366 – 39373. C. Yost, G.H. Farr III, S.B. Pierce, D.M. Ferkey, M.M. Chen, D. Kimelman, GBP, an inhibitor of GSK-3, is implicated in Xenopus development and oncogenesis, Cell 93 (1998) 1031 – 1041. D. Zechner, Y. Fujita, J. Hulsken, T. Muller, I. Walther, M.M. Taketo, E.B. Crenshaw III, W. Birchmeier, C. Birchmeier, h-catenin signals regulate cell growth and the balance between progenitor cell expansion and differentiation in the nervous system, Dev. Biol. 258 (2003) 406 – 418.