| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
T-catenin is testis-restricted in expression and function
,3
* Department for Molecular Biomedical Research, VIB-Ghent University, Ghent, Belgium; and
Department of Molecular Biology, Ghent University, Ghent, Belgium
3Correspondence: Department for Molecular Biomedical Research, VIB-Ghent University, Technologiepark 927, B-9052 Ghent, Belgium. E-mail: f.vanroy{at}dmbr.ugent.be
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
T-Catenin is a recently identified member of the -catenin family of cell-cell adhesion molecules. For decades it was thought that -catenins mediate solid cell-cell adhesion by linking the cadherin-mediated cell-cell adhesion complex with the actin cytoskeleton. However, the roles of -catenins in this classical adhesion model have been questioned recently. T-Catenin has a restricted expression pattern, in contrast to the ubiquitously expressed E-catenin. High levels of T-catenin were detected in heart and testis. Northern and Western blot experiments indicated that besides the standard full-length T-catenin transcript, smaller alternative transcripts are expressed in testis. We report the cloning of two alternative transcripts of the mouse T-catenin gene (transcript-B and -X), both of which are expressed in a testis-restricted manner from two putative alternative promoters. Alternative transcript-X encodes a smaller protein, isoform-X, which lacks the amino-terminal ß-catenin binding domain of the standard mouse T-catenin protein, and is therefore unable to restore cell-cell adhesion in an -catenin-negative colon carcinoma cell line. Immunohistochemical analysis showed specific localization of the T-catenin isoform-X in the differentiating germ cells. In contrast to the standard full-length T-catenin protein, this shortened isoform-X can bind to l-afadin, an important component of the nectin/afadin/ponsin adhesion complex that reportedly is essential for spermatogenesis.—Goossens, S., Janssens, B., Vanpoucke, G., De Rycke, R., van Hengel, J., van Roy, F. Truncated isoform of mouse T-catenin is testis-restricted in expression and function.
Key Words: afadin • testis • spermatogenesis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-CATENINS ARE INDISPENSABLE for cadherin-mediated cell-cell adhesion (1)
. In the classical model for adhesion junctions, -catenins anchor the cadherin/catenin complex to the actin cytoskeleton by binding to ß-catenin and/or plakoglobin on one side and forming a direct or indirect interaction with actin filaments on the other side. However, most recently Drees et al. (2)
and Yamada et al. (3)
suggested on the basis of some compelling in vitro data that this model might not be correct. They postulated that -catenins favor intercellular adhesion by functioning as a molecular switch that regulates actin dynamics at the site of adherens junctions.
Three different -catenins are known (1)
. Though quite similar, they differ in patterns of expression. Besides the ubiquitously expressed E-catenin (4)
and the neural N-catenin (5
, 6)
, a third -catenin has been identified: T-catenin, which is predominantly expressed in heart and testis (7)
. In vitro, N- and T-catenin can substitute for the adhesive functions of E-catenin. However, the restricted expression patterns of N- and T-catenin indicate that they have other specific functions. E-catenin is a well-known invasion suppressor, and poor prognosis of various tumor types is often correlated with defects in E-catenin expression. In addition, E-catenin defects have been observed in several invasive cell lines, and introduction of an exogenous functional -catenin restored cell-cell adhesion and inhibited invasiveness (7
, 8)
. Besides its role in suppressing invasion, -catenin can also suppress tumor growth, as shown by conditional knockout of E-catenin in epidermis (9)
. Loss of E-catenin not only compromises intercellular adhesion but also leads in vivo and in vitro to sustained activation of the Ras-MAPK pathway, resulting in hyperproliferation of E-catenin-null keratinocytes. It was also shown that ablation of E-catenin is accompanied by activation of the NF-
B signaling pathway (10)
. These results indicate that in addition to playing a structural role in cell-cell adhesion, -catenins function in signal transduction.
Much controversy exists about the presence of the cadherin/catenin-based complex in testis (11)
. Recently, evidence has been emerging that cadherins and catenins do localize at the ectoplasmic specializations (ES), specialized junctions between Sertoli cells and differentiating germ cells, and that changes in protein-protein interactions of the N-cadherin/catenin complex regulate junction dynamics between Sertoli cells and germ cells (12)
. A well-organized system of breakage and reformation of these specialized junctions is crucial for migration of the germ cells across the seminiferous epithelium during differentiation.
Another actin-based cell-cell adhesion complex reported to be localized at the ES is the nectin/afadin/ponsin complex (NAP complex) (13)
. Nectin is an immunoglobulin-like cell adhesion molecule (14)
that has been shown to be crucial for spermatogenesis. Adhesion in nectin-2–/– mice is impaired, and morphogenesis and positioning of spermatids are aberrant, leading to male infertility (15)
. Afadin is an actin binding protein that links nectin to the actin cytoskeleton, thereby strengthening adhesion. This NAP complex interacts with the cadherin/catenin complex (16)
, possibly via a direct link between l-afadin and -catenin (17)
or an indirect link via ADIP and -actinin (18)
.
Here, we report the identification of two testis-restricted alternative transcripts (AT-B and AT-X) of the mouse T-catenin gene, Ctnna3. AT-X encodes a smaller protein, isoform-X, that cannot bind ß-catenin, but in contrast to the standard full-length T-catenin protein, can strongly bind l-afadin. Based on these data, we suggest that mouse T-catenin isoform-X may play a key role in spermatogenesis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
T-catenin alternative transcriptsT-catenin gene (Ctnna3). Primers were designed on the basis of the full-length mouse T-catenin cDNA sequence we determined earlier (GenBank accession #AF344871) (19)
. The 5' RACE reaction was performed using the GeneRacer Kit, version B (Invitrogen, San Diego, CA, USA), and mouse testis cDNA as template. Capped RNA was isolated from mouse testis by the RNeasy method (Qiagen, Hilden, Germany) and used to synthesize cDNA employing a gene-specific primer, MCBU#2821 (5'-ACCCCCAATGTTTTATGTTATTTG-3'). The 5'-RACE products were obtained with primer MCBU#2481 (5'-CTTGGTGGAGGCAATGTATGAC-3'). They were cloned in pGEMTeasy (Promega, Madison, WI, USA) and sequenced. Two alternative transcripts were identified, transcript-B and –X; the consensus cDNA sequences were deposited with GenBank (accession #DQ380429 and #DQ380430).
In vitro transcription/translation analysis of mouse T-catenin alternative transcript-X
On the basis of the consensus cDNA sequence of alternative transcript-X, primers MCBU#3214 (5'-TGCCAGCTCTTCCTCTGA-3') and MCBU#2918 (5'-ACAGCCAGAAAGTCGTCAATG-3') were designed to amplify the 5' region (810 bp). The amplified fragment was cloned in pGEMTeasy, yielding pGEMTe PCR3214 + 2918. An NcoI fragment of this plasmid was ligated into the NcoI-digested pGEMTe-mTctn (1–2979) (19)
to produce pGEMTe-mTctn AT-X(33–2137), containing the complete ORF of alternative transcript-X. The latter was expressed in vitro in a coupled transcription/translation reaction, using the TNT Coupled Reticulocyte Lysate System (Promega). pGEMTe-mTctn FL(1–2979) (19)
, containing the standard full-length mouse T-catenin cDNA, was used as a positive control and pGBKT7-hKaiso as a negative control. Full-length cDNA of hKaiso was amplified from genomic DNA of the human colon carcinoma cell line HCT8/E8 with primers MCBU#2210 (5'-GCGGAATTCAAAGGCATGGAGAGTAGA-3') and MCBU#2211 (5'-CCGCTCGAGAAAATCCCACTACACTTC-3') and cloned in the EcoRI-SalI restricted pGBKT7 to obtain pGBKT7-hKaiso.
Proteins synthesized in vitro were analyzed by standard SDS-PAGE, followed by Western blot.
Expression analysis by RT-PCR
For reverse transcription-polymerase chain reaction (PCR) (RT-PCR), RNA was prepared from different mouse tissues with the RNeasy method (Qiagen) and cDNA was synthesized using a commercial kit (Invitrogen). For alternative transcript-B, exon 1b-specific primer MCBU#3243 (5'-CCGCCCGGAGGCTGAG-3') and exon 4-specific primer MCBU#2914 (5'-TGCAAGAGGCACATGACATC-3') amplified a 522 bp fragment. For alternative transcript-X, primers MCBU#3214 (5'-TGCCAGCTCTTCCTCTGA-3') and MCBU#2566 (5'-GCAGCATAGTCTTTGATTTCCTT-3') amplified a 532 bp fragment. The standard mouse T-catenin transcript was amplified using exon 1a-specific primer MCBU#2820 (5'-CCCCTTTCTCTCTTATCCTGAG-3') and exon 3-specific primer MCBU#2839 (5'-GCTGCCAGCTCTTCCTTTAAA-3'), yielding a fragment of 426 bp. As a control, a 452 bp fragment of mouse GAPDH was amplified with primers MCBU#2219 (5'-ACCACAGTCCATGCCATCAC-3') and MCBU#2220 (5'-TCCACCACCCTGTTGCTG TA-3').
Expression plasmids
Standard full-length mouse T-catenin cDNA and alternative transcript-X cDNA were cloned in pENTR (Invitrogen) for Gateway cloning. Cloning of pENTR2B-mT-ctn FL(120–2979) has been described (20)
. By performing an LR Gateway reaction between pENTR2B-mTctn FL(120–2979) and destination vector pdGBKT7, we obtained yeast expression plasmid pdGBKT7-mTctn FL(119–2979). pdCS-mTctn FL(120–2979) was made similarly (20)
. Myc-mTctn FL(120–2979) was shuttled as a HindIII-XbaI fragment to pcDNA4/TO (Invitrogen) to obtain a doxycycline-inducible expression vector, pcDNA4/TO-MT+mTctn FL.
For alternative transcript-X, first the 5' end was amplified with primers MCBU#3463 (5'-GGGATCCGGCTTATGTCTTAAAGAAAT-3') and MCBU#2918 (5'-ACAGCCAGAAAGTCGTCAATG-3') using Pfu polymerase (Stratagene, La Jolla, CA, USA), then cloned in pGEMTe. The 5' end of alternative transcript-X was subsequently shuttled, by a three-point ligation, as a BamHI-NotI fragment together with the 3' end of the mouse T-catenin (NotI-NcoI) to BamHI-NcoI restricted pENTR3C. This produced pENTR3C-mTctn AT-X(101–2137), containing the full-length ORF of the mouse T-catenin alternative transcript-X. By performing an LR Gateway reaction between this entry clone and the destination vectors pdGBKT7 and pdCS3, we obtained expression plasmids pdGBKT7-mTctn AT-X(101–2137) and pdCS3-mTctn AT-X(101–2137). The myc-mTctn AT-X was shuttled as a HindIII-XbaI fragment to plasmid construct DNA (pcDNA)/TO to obtain a doxycycline inducible expression vector, pcDNA4/TO-MT+mTctn AT-X.
The full-length ß-catenin cDNA was kindly provided as plasmid pBAT-ßCAT by Dr. J. Behrens (Erlangen, Germany). The amino-terminal fragment 239–717 of the ß-catenin cDNA was obtained as an NcoI-PstI restriction fragment. The NcoI end was filled in with Pfu polymerase, and the fragment was cloned in SmaI-PstI opened pGAD424, to produce pGAD424-ATßctn. The pGAD424HA-l-afadin construct was kindly provided by Dr. Y. Takai (Osaka, Japan).
Yeast two-hybrid transformation
The yeast strain AH109 (Matchmaker, Clontech, CA, USA) was used for cotransformation of pGBKT7 bait and pGAD424 or pGADT7 prey plasmids harboring the cloned inserts of interest. The yeast cells were grown in YPD medium to log phase (OD600
0.8) and transformed by the lithium acetate procedure (21)
. Cotransformants were selected by plating on synthetic dropout (SD) minimal medium lacking leucine and tryptophan. After 3 days, colonies were picked and grown overnight in SD without leucine and tryptophan, but containing 0.07 M potassium phosphate. Replica plates selecting for molecular interactions were made on SD lacking leucine, tryptophan, histidine, and adenine but containing 0.07 M potassium phosphate and 80 mg/ml 5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside (X-ß-gal) (Duchefa, Haarlem, The Netherlands).
Generation of T-catenin isoform-X-specific antibodies
To generate monoclonal antibodies, peptide P1158 (MLAPKEDRLNANKNl-C), corresponding to the amino terminus of mouse T-catenin isoform-X, and peptide P1159 (C-KIHPVQVMSEFRGRQVY), corresponding to the carboxyl terminus of standard T-catenin and its isoform-X, were synthesized and coupled via the additional cysteine residue to keyhole-limpet hemocyanin (Pierce, Rockford, IL, USA). The peptides (50 µg) were combined with Titermax adjuvant (Sigma) and injected in Wistar rats. Boosts were given every 2 wk, and sera were tested by peptide ELISA until a titer of 1:5000 without any loss of reactivity was obtained after 8 wk. Hybridomas were generated by fusing spleen cells of the immunized rats with Sp20_Ag14 myeloma cells. Supernatants of hybridoma cells were tested by ELISA. Strongly reacting clones were tested by Western blot for recognition of the myc-tagged T-catenin isoform-X in lysates of HEK-293 cells transfected with pdCS3-mTctn AT-X (101–2137) and for recognition of the endogenous T-catenin isoform-X in protein lysates of mouse testis.
Generation of E-catenin negative DLD1-TR21 cells and stable transfection
Dr. H. Clevers (Utrecht, The Netherlands) kindly provided us with the tetracycline-inducible DLD1-TR21 cell line. These cells, like their parental human colon carcinoma cells, acquire spontaneous mutations in the CTNNA1 gene, which results in a dramatic change from an epithelial to a round phenotype (22
, 23)
. Different clones of these round cells were picked and analyzed by Western blot and immunofluorescence staining for loss of E-catenin expression. This yielded the E-catenin-negative DLD1-TR21/R6 cells.
These cells (5x106) were stably transfected with 9 µg of either pcDNA/TO MT+mT-ctn FL, pcDNA4/TO MT+mTctn AT-X, or the empty vector, using the nucleofector AMAXA Biosystem (Amaxa Biosystems, Cologne, Germany). Stable transfectants were selected in 500 µg/ml zeocin (Invitrogen) and 10 µg/ml blasticidin S (Invitrogen). Expression of MT+mT-catenin was induced by adding doxycycline hydrochloride (DOX; 1 µg/ml; Duchefa).
Immunocytochemistry of cell cultures
Cells were grown to confluency on glass coverslips, rinsed briefly with PBS, and fixed with ice-cold methanol for 10 min. They were then incubated for 1 h with primary antibody diluted in PBS + 0.4% gelatin, washed with PBS, and incubated for 30 min with secondary antibodies diluted in PBS + 0.4% gelatin. Secondary anti-mouse IgG or anti-rabbit IgG antibodies were coupled to either Alexa 594 or Alexa 488 (Molecular Probes, Eugene, OR, USA). Finally, cells were treated with 4'-6-diamidine-2-phenylidole-dihydrochloide solution (4',6'-diam idino-2-phenylidole (DAPI); Roche Diagnostics, Mannheim, Germany) to mark nuclear DNA and mounted in Vectashield (Vector Laboratories, Burlingame, CA, USA) to prevent photobleaching. Samples were examined with a Zeiss Axiophot microscope, and images were recorded with a microMAX camera (Princeton, Trenton, NJ, USA) and MetaMorph software (Image Universal Corporation, NY, NY, USA).
Immunohistochemical analysis
A paraformaldehyde-fixed adult mouse testis was embedded in paraffin, and 4 µm-thick sections were immunostained by standard procedure using rat monoclonal antibodies recognizing the mouse T-catenin isoform-X: 1158_AT-X (dilution 1:250) and 1159_AT-FL (1:250). After deparaffinization, an antigen retrieval step was performed by microwaving the slides twice for 5 min at 95–100°C in an EDTA-buffered antigen retrieval solution, pH 9.0 (DakoCytomation, Glostrup, Denmark).
For immunofluorescence staining, sections were incubated for 1 h at room temperature with primary antibodies diluted in PBS + 0.4% gelatin. After washing with PBS, the sections were incubated for 1 h at room temperature with secondary goat anti-rabbit IgG and goat anti-rat IgG antibodies labeled with Alexa 594 (Molecular Probes). Finally, sections were mounted using Vectashield with DAPI (Vector Laboratories).
For chromogen end point staining, the LSAB2 System-HRP (ready-to-use 3-amino-9-ethyl carbazole) was used according to the manufacturer’s instructions (DakoCytomations). Counterstaining was done with hematoxylin (Biogenex, San Ramon, CA, USA) and mounting with Aquatex (Merck, Whitehouse Station, NJ, USA). Samples were examined with an Olympus BX51 microscope and images were captured using a Coolsnap camera (Photometrix, Tucson, AZ, USA) and RSI image software (Roper Scientific, Trenton, NJ, USA).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
T-catenin gene (Ctnna3)T-catenin transcript in heart and testis, there is strong expression of a smaller alternative transcript in testis (data not shown). Using 5'-RACE experiments on mouse testis RNA, we were able to identify two alternative mouse T-catenin cDNA sequences (Fig. 1
): alternative transcript-B (AT-B: GenBank accession #DQ380429) and alternative transcript-X (AT-X: GenBank accession #DQ380430).
|
AT-B (2907 bp) lacks the first, noncoding exon of the standard T-catenin transcript, which is replaced by an alternative exon B of 82 bp. By BLAT search (http://genome.ucsc.edu) of mouse genomic sequences, we located this exon B more than 27.7 kb in front of Ctnna3, suggesting that the alternative transcript-B is transcribed from a putative alternative promoter located upstream of Ctnna3. The complete open reading frame of the standard mouse T-catenin transcript is preserved in this alternative transcript, which means that both transcripts encode the same 100 kDa protein.
The second alternative transcript, AT-X, lacks the first six exons of the standard mouse T-catenin transcript, which are replaced by an alternative exon X. This exon is located in intron 6 of Ctnna3, indicating that the alternative transcript-X is transcribed from a putative internal promoter. The alternative 160 bp exon X contains a start-codon at position 119, preceded by an in-frame stop codon at position 98. The ORF encodes a 70 kDa protein, isoform-X, composed of 628 amino acid residues (AA). The first stretch of 14 AA is isoform-X-specific; the remainder is identical to the central and carboxyl-terminal part of the standard T-catenin protein. Using an exon-X-specific probe, we showed that the alternative transcript, which we had detected by Northern blot, is identical to the alternative transcript-X that we identified by 5' RACE experiments (data not shown). The alternative transcript-X was expressed in an in vitro transcription/translation assay. Protein products were analyzed by Western blot and could be detected by a rabbit polyclonal anti-T-catenin antibody (#952), which recognizes a peptide corresponding to the C terminus of both T-catenin proteins (7)
. The protein synthesized in vitro comigrates with a 70 kDa protein detected by a rabbit polyclonal anti-T-catenin antibody in mouse testis lysates (Fig. 2
).
|
Both alternative transcripts show a testis-restricted expression pattern
We analyzed the tissue distribution of the newly identified transcripts by RT-PCR using primer combinations specific for each of the mouse T-catenin alternative transcripts. This revealed that both alternative transcripts are expressed exclusively in testis (Fig. 3
), whereas the full-length T-catenin was highly expressed in heart as well. Northern and Western blot analysis of RNA and protein lysates from various mouse tissues confirmed these findings (data not shown). These results also agree with the mouse and rat expressed sequence tag (EST) sequences that we identified by basic local alignment search tool (BLAST) search analysis using exon B and exon X sequences as a query (data not shown).
|
Isoform-X is expressed only during spermiogenesis
To investigate whether isoform-X might play a role during spermatogenesis, we compared its expression in premature and mature male mice. Western blot analysis of lysates from testes showed no expression of isoform-X in premature males (Fig. 4
). Strong expression of isoform-X could only be seen after wk 4, at the onset of spermatogenesis. No differences in expression of the standard full-length T-catenin could be observed in lysates of premature and mature testes.
|
A rat monoclonal antibody was generated against a mouse T-catenin isoform-X-specific peptide. This antibody, named 1158_AT-X, strongly and specifically detects endogenous isoform-X in Western blots (data not shown). Another rat monoclonal antibody, named 1159_AT-FL and raised against a peptide shared by both the standard full-length T-catenin protein and the truncated isoform-X, detects both T-catenin forms.
We then used these antibodies to localize T-catenin on paraffin sections of adult murine testis. Strong staining of the isoform-X protein was detected in the inner mass of the seminiferous tubules (Fig. 5
), at the location of the elongating spermatids. Expression was seen exclusively in the cytoplasm of differentiating germ cells of tubules that were in stage II-VIII of the spermatogenic cycle (Fig. 5B
) (24)
.
|
Isoform-X is unable to restore cadherin-mediated cell-cell adhesion in an -catenin-negative colon carcinoma cell line
Isoform-X lacks the N-terminal domain of the standard T-catenin, which contains the ß-catenin binding domain. To assess whether the absence of this N-terminal domain has functional implications for localization of this isoform or for formation of stable cell-cell contacts, we carried out rescue experiments by overexpressing isoform-X in round -catenin-negative DLD1-TR21 cells. This cell line was subcloned from DLD1-TR21 cells and analyzed by immunofluorescence staining and Western blot. Most of the selected DLD1-TR21/R clones, including DLD1-TR21/R6, showed no expression of -catenins (Fig. 6
A, B). These cells where subsequently used for stable transfections of plasmids encoding either isoform-X or the standard full-length T-catenin protein. Immunostaining of cells transfected with cDNA encoding the full-length T-catenin and induced with doxycycline showed that the protein localized at the cell-cell contacts and that the round cells adopted a more epithelial-like morphology (Fig. 6C, D
). In cells transfected with isoform-X cDNA, however, induction did not lead to morphological changes. The cells retained their round morphology, and only cytoplasmic staining for the Myc-tagged isoform-X could be observed (Fig. 6C, D
). These data demonstrate that isoform-X cannot function as a genuine -catenin, since it is unable to restore cell-cell adhesion in an -catenin-negative colon carcinoma cell line.
|
Isoform-X is unable to bind ß-catenin, but interacts strongly with another cell-cell adhesion protein, l-afadin
Yeast two-hybrid experiments further confirmed that isoform-X, in contrast to the full-length T-catenin, cannot bind ß-catenin (Table 1
). On the other hand, we could show that isoform-X can strongly bind to l-afadin, a crucial component of the nectin/afadin cell-cell adhesion complex. In the same assay, the full-length T-catenin could bind l-afadin only weakly, most likely because the l-afadin binding site is cryptic in this full-length isoform.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
T-catenin indicated the expression of at least one smaller alternative transcript. Here we report on the identification of two alternative transcripts whose expression is restricted to the testis.
Alternative transcript-B (AT-B) is transcribed from a putative alternative promoter-B located upstream of the standard promoter-A of Ctnna3 (24)
. This transcript differs from the standard T-catenin transcript only in its 5'-untranslated region, and both transcripts encode the same 100 kDa protein. The second alternative transcript, AT-X, transcribed from a putative internal promoter-X, lacks the first six exons of the standard T-catenin transcript and encodes a smaller 70 kDa protein, isoform-X. We showed by in vitro transcription/translation that AT-X and isoform-X correspond to the same smaller transcript and protein we previously detected in RNA and protein extracts from mouse testis.
Although exon X and putative promoter-X are highly conserved between mouse and rat, until now we could not find a conserved exon X in T-catenin genes of other animals, including humans. However, EST sequences derived from the chicken reproductive tract (accession #CD218203) and a human EST sequence (accession #BX494002) indicate the existence of alternative internal promoters used to transcribe similar (though not identical) alternative transcripts that lack the 5' part of the standard T-catenin transcript.
This newly identified isoform-X of T-catenin lacks the ß-catenin domain typically found at the amino terminus of -catenins. On that basis, we hypothesized that isoform-X cannot bind ß-catenin, and subsequently confirmed this by yeast two-hybrid analysis. In line with this, isoform-X cannot restore cell-cell adhesion in an -catenin-negative colon carcinoma cell line, in contrast to the standard full-length T-, E- and N-catenins.
In the many tissues we tested, we detected expression of the truncated T-catenin isoform only in testis. Whereas full-length T-catenin is expressed throughout postnatal testis development, isoform-X begins to be expressed only at the onset of puberty. This indicates that isoform-X plays a role in spermatogenesis. Such a function is also supported by the cellular localization of the isoform-X protein. Immunofluorescence with isoform-X-specific monoclonal antibodies revealed this isoform only in elongating spermatids. The expression of T-catenin isoform-X in spermatids is in apparent contrast to the peritubular T-catenin expression we had observed earlier (7)
. This can be explained to some extent. The human section we analyzed before seems not to contain the more mature forms of sperm cells. In addition, weak T-catenin expression could often be detected at the periphery of mouse seminiferous tubules, but only when using an antibody recognizing both mouse T-catenin isofoms, and not with an isoform-X-specific antibody, suggesting that only the full-length T-catenin might be expressed there. Notably, aspecific staining of Leydig cells by the currently available antibodies hinders correct interpretation. Alternatively, because there is no strict orthologue of exon X in the human genome, it is possible that the functional counterpart of isoform-X, as evidenced by a human EST sequence, is either very weakly expressed or poorly recognized by the antibodies used. More work is needed to prove or disprove these possibilities.
Although isoform-X cannot bind the cadherin/catenin complex via ß-catenin, it may still play a role in its dynamic breakage and reformation. It may, for instance, compete with the full-length E- and T-catenin for binding to F-actin or other interaction partners. In this way it will weaken cell-cell adhesion and allow release of fully differentiated spermatozoa into the lumen of the seminiferous tubules. Spermatozoa will then leave the testis and pass into the epididymis for storage and further maturation.
On the other hand, isoform-X might be involved in the formation of another cell-cell adhesion complex, the nectin/afadin/ponsin complex (NAP complex), which is reportedly localized at the apical ectoplasmic specialization. Close interaction between the cadherin/catenin complex and the NAP complex has been reported (16
, 25)
. Direct interaction between l-afadin and E-catenin is controversial, since full-length E-catenin seems to have a cryptic central binding domain for l-afadin (17)
. Indeed, full-length E-catenin was found to bind significantly less l-afadin than did a fragment of E-catenin (residues 385–651). In a previous study (16)
that included yeast two-hybrid analysis and assays for in vitro binding with recombinant proteins, only the C-terminal half of E-catenin (residues 508–906), but not the full-length protein, bound to l-afadin. Collectively, these data indicate that there is a cryptic binding site for l-afadin in full-length E-catenin. In contrast, an E-catenin protein lacking the amino-terminal domain, and strongly resembling the T-catenin alternative isoform-X, seems to bind l-afadin (16
, 17)
. Our yeast two-hybrid experiments showed that isoform-X can indeed bind l-afadin strongly.
It has been reported that -catenins are involved in actin dynamics (26)
, which is essential for many marked morphological changes during spermiogenesis (27)
. Recently a mechanism was suggested in which -catenins might play a key role as a molecular switch that regulates actin assembly (2
, 3)
. A significant increase in the local concentration of -catenin would favor homodimerization, resulting in reorganization of the actin filaments on the one-hand, by suppression of the Arp2/3-mediated formation of a branched actin network, and on the other by recruitment of formins, which promote actin bundling and linear actin cable formation. The expression of isoform-X of T-catenin might be another way to regulate this molecular switch and coordinate the actin cytoskeleton during spermiogenesis.
In conclusion, we report here the identification of two alternative transcripts of Ctnna3 transcribed from two putative alternative promoters and expressed in a testis-restricted manner. One of the alternative transcripts encodes a smaller protein, isoform-X, which can bind l-afadin strongly but lacks the amino-terminal ß-catenin binding domain, and therefore is unable to restore cell-cell adhesion in -catenin-negative human colon carcinoma. It will be challenging to discover the exact role of the T-catenin isoforms during spermatogenesis by analyzing conditional T-catenin-deficient mice.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Current address: Wiley-VCH, Boschstrasse 12, D-69469 Weinheim, Germany. 
2 Current address: Peakadilly N.V., Bio-incubator, Technologiepark 4/8, B-9052 Ghent, Belgium.
Received for publication March 7, 2006. Accepted for publication October 11, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-catenin, NF-B, and squamous cell carcinoma in skin. Proc. Natl. Acad. Sci. U. S. A. 103,2322-2327E-catenin gene (CTNNA1) acts as an invasion-suppressor gene in human colon cancer cells. Oncogene 18,905-915[CrossRef][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
