| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* INSERM U773, Université Paris 7, Paris, France;
Institute of Theoretical and Experimental Biophysics, Russian Academy of Sciences, Moscow Region, Russia;
Laboratory of Experimental Cancerology, Ghent University Hospital, Ghent, Belgium,
Huntsman Cancer Institute, Departments of Biology and Oncological Sciences, University of Utah, Salt Lake City, Utah, USA; and
|| Departments of Molecular Biomedical Research and Molecular Biology, VIB-Ghent University, Ghent, Belgium
1Correspondence: INSERM U773, Centre de Recherche Biomedicale Bichat Beaujon CRB3, Faculté de Médecine Bichat, 16 rue Henri Huchard, 75018 Paris, France. E-mail: E.C., eric.chastre{at}inserm.fr; L.K., larissa.kotelevets{at}inserm.fr
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
B-dependent manner and impaired cell-cell aggregation at least in part by uncoupling adherens junctional complexes from the cytoskeleton. The TRIP6Stop473 mutant, which lacks the PDZ binding motif, was still able to increase NF-B and Akt activities but did not promote invasiveness or interfere with cell-cell aggregation. Intracellular delivery of competing peptides corresponding to TRIP6 or β-catenin C terminus restored invasive properties in MDCKts-src TRIP6Stop473 cells, highlighting the requirement of PDZ scaffolds in junctional complexes activity. TRIP6 overexpression in colon tumors suggest its critical role in cancer progression.—Chastre, E., Abdessamad, M., Kruglov, A., Bruyneel, E., Bracke, M., Di Gioia, Y., Beckerle, M. C., van Roy, F., Kotelevets, L. TRIP6, a novel molecular partner of the MAGI-1 scaffolding molecule, promotes invasiveness.
Key Words: PTEN • PDZ domain • NF-B • E-cadherins • PI3-kinase • Akt
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. We demonstrated that PTEN interacts indirectly with β-catenin by binding the scaffolding protein MAGI-1b. Ectopic expression of MAGI-1b potentiated the interaction of PTEN with junctional complexes, promoted E-cadherin-dependent cell-cell aggregation, and reverted the Src-induced invasiveness of kidney epithelial MDCKts-src cells (2)
. Thus, the recruitment of PTEN at adherens junctions by MAGI-1b and the local down-regulation of phosphatidylinositol-3,4,5-trisphosphate pools and downstream effector systems at the site of cell-cell contacts are focal points for restraining both the disruption of junctional complexes and the induction of tumor cell invasion.
To identify novel molecular partners of the MAGI-1/PTEN signalosome, we performed yeast 2-hybrid (Y2H) screening, using the fifth PDZ domain of MAGI-1 as a bait. We identified TRIP-6 (thyroid receptor interacting protein 6)/ZRP-1 (zyxin-related protein 1)/OIP1 (opa-interacting protein) (3
4
5)
, as a molecular partner of MAGI-1b. Human TRIP6 is a 476-amino acid (aa) protein containing a proline-rich N-terminal segment and 3 LIM domains (3)
. The C terminus of TRIP6 contains a PDZ-binding motif that reportedly interacts with the cytosolic protein tyrosine phosphatase hPTP1E/PTP-BL and with the Scrib adaptor molecule (4
, 6
, 7)
. TRIP6 belongs to a new family of LIM proteins, including Zyxin, LPP, Ajuba, and LIMD1 (8
9
10
11)
. Members of this protein family are located at the focal adhesion plaques, associated with membrane proteins and localized at the membrane, and in the nucleus, where they perform regulatory functions (11
12
13)
. These proteins are components of integrin-mediated adhesive complexes in fibroblasts (14)
and of cell-cell junction adhesive complexes in epithelial cells (15)
. They interact with filamentous actin either directly or indirectly (10
, 16)
. Studies on the involvement of TRIP6 in cell motility, F-actin organization, and focal-adhesion assembly yielded conflicting results (14
, 17
18
19)
. Overexpression of TRIP6 slows 10T1/2 cell migration (14)
and increases LPA-induced cell migration of ovarian carcinoma SKOV3 cells (19)
. Depletion of TRIP6 in lung carcinoma A549 cells or epidermoid carcinoma A431 cells resulted in an increase of the cell migration rate induced by wounding (17)
. Deletion of TRIP6 in human umbilical vein endothelial cells leads to reduction of F-actin fibers (20)
, and, in contrast, depletion from A549 or A431 cells leads to enhanced focal adhesion and stress fiber formation (17)
.
To delineate the involvement of TRIP6 in the scattering and invasiveness of epithelial cells, we transfected MDCK and their derivatives MDCKts-src cells expressing the thermosensitive v-Src with expression vectors encoding murine TRIP6. MDCKts-src cells allow investigation of permissive or restrictive activities of transgenes on invasiveness by changing culture conditions (21)
. The biological significance of the TRIP6 PDZ-binding motif was further assessed by using the TRIP6Stop473 mutant defective in the PDZ-binding motif and by intracellular delivery of competing peptides corresponding to the C terminus of TRIP6 or β-catenin. We evaluated actin cytoskeleton organization, morphological conversion, cell aggregation, and invasiveness into type I collagen in relation to the expression and subcellular localization of junctional complexes containing cadherin and the activity of effector systems known to regulate cell scattering, including Akt, Rac1, RhoA, and NFkB.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-Gal but lacking histidine, tryptophan, leucine, and adenine.
Positive 2-hybrid protein interactions were verified by transfection of the plasmids back into the reporter strain AH109 together with the original bait or with selected controls. The plasmids extracted from positive colonies were amplified in Escherichia coli DH5-. The inserts were amplified by PCR, and the corresponding cDNAs were sequenced.
Plasmid construction, cell transfection, and Ca2+ switch
The expression vector for green fluorescent protein (GFP)- tagged murine TRIP6 was kindly provided by Dr. W. Hendriks (University Nijmegen, Nijmegen, The Netherlands). The Flag-tagged murine TRIP6 expression vector had been described previously (14)
.
The Flag-TRIP6Stop473 and GFP-TRIP6Stop473 expression vectors were constructed by PCR amplification of a 1085 bp TRIP6-encoding sequence using oligonucleotides: sense 5'-ccagacagacaggcttttgag-3', and antisense 5'-agtctcgagagtggctcagagctcttggata-3' or 5'-agtgtcgacagagtggctcagagctcttggata-3' introducing a stop mutation at codon 473. The enzyme was Phusion High-Fidelity DNA Polymerase (Ozyme, St-Quentin-en-Yvelines, France). After restriction of the PCR products using, respectively, SstII and XhoI, or ScaI and SalI, the PCR products were subcloned in the Flag-TRIP6 and GFP-TRIP6 expression vectors in place of the corresponding wild-type (wt) sequence.
A Flag epitope-tagged MAGI-1b construct was kindly provided by Dr. I. Y. Dobrosotskaya (University of Texas, Dallas, TX, USA). The pRc/CMV-p65 and mp50 plasmids encoding p65 and subunits p50 of NF-kB, respectively, had been described previously (22)
.
The colonic cell line Caco-2, kidney epithelial HEK293 cells, HeLa human cervical cancer cells, MDCK cells, and the derivative MDCKts-src clone 2, which expresses the thermosensitive v-Src, were routinely grown in Dulbecco modified Eagle medium (DMEM) containing 4.5 g/L glucose, 10% fetal calf serum, 8 mM L-glutamine, and antibiotics. The MDCKts-src cells have been previously established after infection of MDCK cells with a murine leukemia retroviral construct expressing a thermosensitive v-src gene. The resulting MDCKts-src cells were then subcloned, yielding clone 2 (21)
. At the temperature restrictive for Src activity (40°C), MDCKts-src cells exhibit the feature and characteristics of the parental MDCK cells, including epithelial morphology and functional E-cadherin junctional complexes. At the temperature permissive for Src activity (35°C), MDCKts-src cells undergo an epithelial-mesenchymal transition characterized by a fibroblast-like morphology, the disruption of the E-cadherin junctional complexes, and the acquisition of the invasive phenotype (23)
. MDCKts-src and HEK293 cells were transfected with the above-mentioned constructs using lipofectamine, as described (2)
.
For extracellular Ca2+ depletion, trypsinized cells were plated to form monolayers on glass coverslips in 6-well plates and cultivated for 24 h. The cells were washed twice and incubated in Ca2+-free DMEM. The calcium switch was performed by substituting Ca2+-free medium for DMEM with Ca2+. At different time points after Ca2+ depletion/repletion, the cells were fixed and processed for immunofluorescence as described below.
Real-time PCR analysis
Total RNA was extracted from human colonic tumor samples paired with control mucosa (24)
. RNA was reverse transcribed using Moloney murine leukemia virus reverse transcriptase (Invitrogen Life Technologies, Cergy Pontoise, France), and the cDNAs were subjected to real-time PCR using a LightCycler 480 Roche qPCR (Roche Diagnostics, Meylan, France). Real-time PCR was conducted in duplicate, using 3 serial dilutions (1:3.16) of the cDNAs. The human TRIP6 and 18S primers and the corresponding TaqMan probes were purchased from Applied Biosystems (Applera France, Courtaboeuf, France). To determine the relative accumulation of TRIP6 transcripts in colonic tumors and paired control colonic mucosa, the threshold cycle (CT) values of TRIP6 were first normalized by substracting the corresponding CT values obtained from the 18S control rRNA used as the internal standard (
CT). The CT values were calculated by substracting the CT values obtained in tumors and in paired control mucosa, and the fold differences in TRIP6 mRNA accumulation in the tumors relative to the paired control mucosa was determined using the formula 2–CT.
Collagen type-I invasion and fast aggregation assays
Single-cell suspensions were seeded on top of type-I collagen gel (UBI, Lake Placid, NY, USA) and incubated for 24 h at 37, 35, or 40°C. Using a computer-controlled inverted microscope, we counted invasive and superficial cells in 12 fields of 0.157 mm2. The invasion index was expressed as the percentage of cells invading the gel (25)
.
For the fast aggregation assay, single-cell suspensions were incubated in an isotonic buffer containing 1.25 mM Ca2+ at 35°C or 40°C for 30 min with gyratory shaking. Particle diameters were measured in a Coulter particle size counter (LS 200; Coulter, Lake Placid, NY, USA) at the start (t0) and after a 30 min incubation (t30) and plotted against percentage volume distribution (26)
.
Synthesis of peptides and their use in cell culture assays
A cell-permeable peptide of the Drosophila homeodomain protein antennapedia and N-terminally biotinylated peptides corresponding to either the C terminus of Trip6 or the C terminus of β-catenin were synthesized commercially to our specifications (Eurogentec, Seraing, Belgium) and purified to >85% purity by reverse-phase high-performance liquid chromatography (HPLC). The sequence of peptide TRIP6 C-term (ELSATVTTDC) corresponds to the C terminus of murine TRIP6 (aa 471-480), and that of β-ctn C-term (CNQLAWF DTDL) corresponds to the C terminus of β-catenin (aa 772-781). The sequence of the control peptide (CGDSNQLAWFD) was shifted 3 aa toward the N terminus to delete the C-terminal PDZ binding motif of β-catenin (aa 769-778). The extra Cys residue in each β-catenin and control peptide sequence was used for coupling the peptide to the 16-aa cell-permeable sequence of antennapedia: RQIKIWFQNRRMKWKK. After coupling, the peptides were again purified, and purity was analyzed by mass spectrography.
Cells were incubated with peptides at a concentration of 10 µM for 3 h in serum-free medium. They were harvested, and single-cell suspensions were assayed for invasion of type-I collagen gel or cell aggregation (see above).
Cell viability was assessed by trypan blue staining. Peptide internalization was analyzed with goat anti-biotin pAb (Pierce, Rockford, IL, USA).
NF-kB p65 inhibitor peptide sets were purchased from Imgenex (CliniSciences, Montrouge, France). Inhibitor peptide PTD-p65 and control peptide were used at a concentration of 150 µM as described previously (27)
.
Immunoprecipitation, immunoblotting, and detergent fractionation
Cells were lysed in 10 mM Tris-HCl, 150 mM NaCl, 1% Triton X-100, and a cocktail of protease and phosphatase inhibitors (1)
. E-cadherin and associated proteins in the lysates were immunoprecipitated with specific antibodies and protein-G Sepharose (Amersham Biosciences, Orsay, France). Immunoprecipitates and whole-cell extracts were separated on 8% polyacrylamide gels and blotted onto polyvinylidene difluoride membranes (Millipore Corp., Bedford, MA, USA). Mouse mAbs directed against β-catenin or ZRP1/TRIP6 were purchased from Transduction Laboratories; mouse mAb and rabbit pAb anti-PTEN and the rabbit polyclonal antibodies anti-p65 and anti-p50 were from obtained Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Human mouse mAb HECD-1 against E-cadherin was from TaKaRa Biochemicals (Tokyo, Japan); rat mAb DECMA-1 against E-cadherin, the mouse anti-Flag mAb, and rabbit pAb against MAGI-1 were obtained from Sigma-Aldrich (St-Quentin Fallavier, France). To separate the Triton-soluble fraction, cell lysates were centrifuged at 4°C for 10 min at 20,000 g. Total protein concentration in Triton-soluble supernatant was determined. The Triton-insoluble pellet was washed with PBS and then solubilized in sodium dodecyl sulfate sample buffer (using the half of the volume of the lysis buffer) for 20 min at room temperature and boiled for 5 min.
Determination of Akt, RhoA, and Rac1 activities
Akt activity was detected by an in vitro phosphorylation assay of GSK-3 (Cell Signaling Technology, Ozyme, St-Quentin-en-Yvelines, France). Signals were visualized with Ig coupled to horseradish peroxidase (HRP), using the enhanced chemiluminescence system (Amersham Biosciences).
Activation assays for the small GTPases were performed using the classical CRIB-peptide pull-down assay (28)
or the G-Lisa RhoA or Rac1 absorbance-based Biochem kits (Cytoskeleton, Denver CO, USA). Briefly, cell lysates corrected for total protein concentration were applied on ELISA plates and treated by specific anti-GTPase antibodies, and the signals were developed with HRP detection reagents and read by measuring absorbance at 490 nm using a microplate spectrophotometer.
In parallel, an aliquot from each extract was analyzed by SDS-PAGE and Western blot for endogenous RhoA and Rac1 levels to confirm comparable total protein content between samples.
Transcriptional coactivator activity
Both plasmids, pcDNA3-βgal (Invitrogen) together with the reporter vector pIgB-luc (D-121/+232 HIV-LTR-luciferase) (22)
, were introduced into MDCKts-src cells, and their derivatives expressing either TRIP6wt or mutant TRIP6Stop473 by lypofectamine transfection protocol. Twenty-four hours after transfection, cells were lysed, and the amount of luciferase and β-galactosidase were quantified using the Luciferase Assay System (Promega, France, Charbonnières Les Bains, France) and Galacto-Light/Plus Assay (Applera France), respectively.
Immunofluorescence microscopy
Monolayers prepared for fluorescent staining were grown on glass coverslips. Cells were fixed with 4% paraformaldehyde in PBS supplemented with Ca2+ and Mg2+ and then permeabilized with 0.2% Triton X-100. Immunofluorescent staining was performed with primary antibody in 0.4% gelatin/PBS for 1 h. This was followed by incubation with Alexa-594 or Alexa-488 conjugated secondary antibody (Invitrogen Life Technologies). For visualization of actin filaments, TRITC-conjugated phalloidin was added during the incubation with secondary antibodies. After 20 min, the cells were washed twice in PBS, mounted with Vectashield, and examined with a Zeiss Axiovert 200 inverted microscope and LSM 510 software version 3.2 (Carl Zeiss, Oberkochen, Germany).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). The TRIP6 sequence is remarkably conserved across species, with human and mouse aa sequences sharing more than 86% identity, including the PDZ binding motif. The main difference relates to the insertion of 4 more aa residues in the murine TRIP6 sequence (480 aa) between residues 139 and 140 of the human sequence.
|
TRIP6 interacts specifically with the fifth PDZ domain of MAGI-1b
The presence of a C-terminal fragment in all Y2H clones of TRIP6 indicates that an intact C terminus might be required for TRIP6/MAGI-1 interaction. To determine the contribution of the various domains of TRIP6 including its C-terminal residues to its interaction with the PDZ5 domain of MAGI-1, several cDNA constructs were subcloned in plasmid pGADT7 (Fig. 1B
). Mutation of Thr to Ala at position 478 did not abolish the binding of TRIP6 to the fifth PDZ of MAGI-1b but greatly reduced it, indicating that this aa residue is important but not required. On the other hand, the deletion of the 8 C-terminal aa residues (TRIP6Stop473 and TRIP6 386–473) impaired the interaction with MAGI-1, demonstrating the involvement of the TRIP6 PDZ-binding motif in the interaction with the fifth PDZ domain of MAGI-1b. To confirm the specificity of these interactions, we investigated TRIP6 interaction with the second and fourth PDZ domains of MAGI-1. We found that TRIP6 (Fig. 1B
) and β-catenin (Fig. 1C
) interacted with the fifth PDZ domain, but were unable to interact with the second or fourth PDZ domain of MAGI-1b. In contrast, as previously demonstrated (2)
, PTEN interacts selectively with the second PDZ domain of MAGI-1b (Fig. 1C
).
It had been reported that TRIP6 interacts with the second PDZ domain of the tyrosine-phosphatase hPTP1E/PTP-BL. We show here that, although both TRIP6 and β-catenin interact with the fifth PDZ domain of MAGI-1b, only TRIP6 can interact with the second PDZ domain of PTP-BL (Fig. 1C
). These results demonstrate the specificity of these interactions via PDZ domains.
TRIP6/MAGI-1b complexes in human epithelial cell lines
To confirm the observed interaction between TRIP6 and MAGI-1b in a more physiological context, coimmunoprecipitation studies were carried out using the HeLa human cervical cancer cells and the human embryonic kidney cell line HEK293. These cells were transiently transfected with expression vectors encoding Flag-labeled MAGI-1b alone or combined with plasmids expressing GFP-labeled TRIP6. Proteins in cellular lysates were immunoprecipitated with anti-MAGI-1, anti-TRIP6, or anti-β-catenin antibodies, and the presence of TRIP6, β-catenin, and MAGI-1b in the immunoprecipitates was detected by Western blot.
β-catenin interacted with MAGI-1 at endogenous protein levels in HeLa cells, in line with our previous report (2)
. Ectopic expression of MAGI-1b, markedly enhanced the amount of associated endogenous β-catenin protein (Fig. 2A
). We demonstrated the interaction of endogenous TRIP6 with Flag-MAGI-1b and potentiation of these interactions in HEK293 cell line overexpressing GFP-TRIP6 (Fig. 2B
). Furthermore, GFP-TRIP6 expression results in decrease of β-catenin coimmunoprecipitated with endogenous or overexpressed Flag-MAGI-1b in both HeLa (Fig. 2A
) and HEK293 (Fig. 2B
) cell lines. These results suggest the competition of TRIP6 and β-catenin for binding to the fifth PDZ domain of MAGI-1b.
|
The interaction of TRIP6 with MAGI-1b involved the PDZ-binding motif of TRIP6, because GFP-TRIP6 Stop473 could not be coimmunoprecipitated with Flag-MAGI-1 (Fig. 3
). Furthermore, we determined that GFP-TRIP6 and PTEN coimmunoprecipitated with MAGI-1b, whereas GFP-TRIP6 coimmunoprecipitated with PTEN. This suggests the existence of a ternary complex involving MAGI-1b, TRIP6, and PTEN (Fig. 3
). The formation of this complex was associated with a decreased association of PTEN with E-cadherins and β-catenins, as evaluated by coimmunoprecipitation experiments (Fig. 3
). In this connection, in the human colonic adenocarcinoma Caco-2 cells, the intracellular transfer of the TRIP6 C-term peptide (PDZ-binding motif) impaired the recruitment of PTEN to the plasma membrane, without affecting the accumulation of PTEN or of E-cadherins (Supplemental Fig. 1). In this cell line we previously reported the interaction of endogenous PTEN with MAGI-1 and identified MAGI-1 in E-cadherin junctional complexes (2)
.
|
Ectopic expression of TRIP6 induced cell invasiveness and decreased cell aggregation of MDCK and MDCKts-src cells
To gain further insight into the biological significance of TRIP6/MAGI-1b complexes, we investigated the effect of TRIP6 on cell invasiveness after ectopic expression in the canine kidney epithelial MDCK cells and their Src-expressing derivatives, MDCKts-src cells. At the temperature restrictive for Src activity (40°C), MDCKts-src cells exhibited the features and characteristics of their parental MDCK cells, including epithelial morphology and functional E-cadherin junctional complexes. At the temperature permissive for Src activity (35°C), MDCKts-src cells underwent an epithelial-mesenchymal transition characterized by a fibroblast-like morphology, disruption of the E-cadherin junctional complexes, and acquisition of the invasive phenotype (1)
. Invasion assays in type-I collagen gels demonstrated that ectopic expression of TRIP6wt induced invasiveness of MDCK cells (data not shown) and MDCKts-src at the temperature restrictive for Src activity (Fig. 4A
). TRIP6 did not affect the invasive phenotype of MDCKts-src cells at the temperature permissive for Src activity.
|
Whether TRIP6 effects on invasion involved junctional complexes was addressed by fast aggregation assays. As shown in Fig. 4B
, MDCKts-src cells formed aggregates after incubation for 30 min at the temperature restrictive for Src activity (Fig. 4B
). This aggregation was impaired in MDCKts-src derivatives overexpressing TRIP6. These results demonstrate that the promotion of invasiveness by TRIP6 relates to junctional complexes activity.
To delineate the involvement of the TRIP6 PDZ binding motif in invasion, we investigated whether the Stop473 mutation affects TRIP6-induced invasiveness. TRIP6 Stop473 was unable to promote invasiveness (Fig. 4A
) or to impair aggregation (Fig. 4B
) of MDCKts-src cells at the temperature restrictive for Src activity, demonstrating that the C terminus of TRIP6 is required to decrease the strengh of cell-cell adhesion. Accordingly, we postulated that the C terminus of TRIP6 is involved in invasiveness either by recruiting of TRIP6 to invasion-promoting molecular complexes, or by destabilizing complexes involved in the cell-cell junction. Previously we had demonstrated that PTEN is recruited to specific subcellular microenvironments, such as adherens junctions, by the binding of its PDZ-binding motif to scaffolding molecules, such as MAGI-1b, which in turn bind to the cadherin/catenin complex via β-catenin (2)
.
According to the hypothesis that TRIP6 C terminus destabilizes complexes involved in cell-cell adhesion, the intracellular targeting of a 10-aa peptide, corresponding to the C terminus of either TRIP6 or β-catenin and including the PDZ-binding motifs, should compete with endogenous β-catenin for binding to MAGI-1b. For the intracellular delivery of these peptides, we used a cell-permeable 16-aa motif of the Drosophila homeodomain protein antennapedia. This motif is an internalization sequence that facilitates the efficient transport of peptides across cell membranes without any cytotoxicity or plasma membrane damage (2)
. The dose effect was determined in order to choose an appropriate working concentration. We showed that at a concentration of 10 µM the peptides TRIP6 C-term (aa 471-480) and β-ctn C-term (aa 772–781) did not induce invasiveness and did not impair aggregation of the parental MDCKts-src cells at the temperature restrictive for Src activity (Fig. 4A, B
). However, both peptides could restore invasiveness and decrease cell-cell adhesion in MDCKts-src TRIP6Stop473 derivatives. This effect was clearly dependent on the competitive inhibition of the PDZ binding domain because the control peptide lacking the PDZ binding motif was ineffective. In this connection, the suppressive effects of the ectopic expression of MAGI-1b on Src-induced invasiveness (at the permissive temperature) required its interaction with β-catenin, because it was counteracted by the TRIP6 and β-catenin competing peptides but not by the control peptide.
Our findings suggest that both the core protein of TRIP6 and the PDZ-binding motif cooperate to promote invasiveness.
TRIP6 ectopic expression decreased cell-cell adhesion and uncoupled cadherin from actin cytoskeleton
To elucidate whether TRIP6-induced reduction of cell-cell adhesion capacity in MDCK and MDCKts-src cell lines was related to changes in the accumulation and/or the subcellular localization of E-cadherin complexes, we performed confocal microscopy and Western blot analysis. As shown in Fig. 5A
, MDCKts-src derivatives did not display any change in membrane staining of the cadherin/catenin complex upon TRIP6 overexpression. Accumulation of E-cadherin in MDCK or MDCKts-src cells was not markedly affected by ectopic TRIP6 expression (Fig. 5B
). The interaction of E-cadherin junctional complexes with the actin cytoskeleton is required for full adhesive activity, and biochemically characterized as detergent-insoluble fraction. A marked decrease in E-cadherin in Triton-insoluble fractions was observed in MDCK and MDCKts-src cells overexpressing TRIP6wt (Fig. 5B
).
|
The colocalization of E-cadherin and actin was further investigated by the analyses of consecutive optical sections by confocal microscopy. As shown in Fig. 5C
, the colocalization of actin and E-cadherin was markedly decreased in MDCKts-src TRIP6wt cells as compared to the parental MDCKts-src and MDCKts-src TRIP6Stop473 cells. These results suggested that TRIP6 expression weakened cell-cell adhesion, and that this process required TRIP6 PDZ-binding motif.
The actin cytoskeleton is implicated in the maintenance of cell shape as well as in cell motility. We analyzed any changes in the actin cytoskeleton caused by the overexpression of TRIP6 (Fig. 5D
). Confocal microscopy observations of parental cells grown under semiconfluent conditions revealed that parallel lines of actin filaments were assembled continuously between neighboring cells. By contrast, actin filaments between neighboring TRIP6wt tranfected cells were disordered and blurred (Fig. 5Da, b
). Actin/phalloidin staining revealed enhanced stress-fiber formation in cells transfected with TRIP6wt, but these actin bundles were almost absent in the parental cell line (Fig. 5Dc, d
).
Effect of TRIP6 on E-cadherin response to changes in extracellular calcium
The generation of cell-cell contacts requires E-cadherin with its capacity to form cell surface homophilic complexes stabilized by extracellular calcium. Accordingly, depletion of extracellular calcium in medium leads to disruption of contacts between neighboring MDCK cells and concomitant disappearance of E-cadherin from cell-cell adhesion sites. Immunofluorescence analysis revealed that a 1 h calcium depletion affected more markedly E-cadherin junctional complexes in MDCK TRIP6wt3 cells as compared to the parental MDCK cell line or the derivatives expressing TRIP6Stop473 (Fig. 6A
).
|
To examine further the effects of TRIP6 overexpression on the formation of cell-cell contacts, we induced formation of cell-cell adhesions in parental and TRIP6 derivative cell lines by changing the Ca2+ concentration in the culture medium from low to high. In parental MDCK cells, junctional complexes had actively reformed 1 h after calcium switch (Fig. 6Bg, j
). By contrast, such organized E-cadherin localization in neighboring cells was almost absent in TRIP6-transfected cells (Fig. 6Bh, k
). Moreover, we often observed that, instead of forming adhesions with or departing from neighboring attached cells, TRIP6-overexpressing cells migrated under or over neighboring cells (Fig. 6Bn, q
). These results suggest that TRIP6 overexpression perturbs formation of cell-cell adhesion.
Effect of TRIP6 on Rho GTPases, Akt, and NFkB activities
Rho GTPases are critical for establishment of cell-cell junctions, remodeling of the cytoskeleton, production of lamellipodia, and formation of new focal complexes at the leading edge. We proceeded to monitor the activation profiles of Rac1 in MDCKts-src cell derivatives. Strikingly, TRIP6wt, but not mutant TRIP6Stop473, expression was associated with a decrease in Rac1 activity. Similar results were obtained after transient transfection of HEK293 cells (Fig. 7A
).
|
Because TRIP6-overexpressing cells showed enhanced formation of actin stress fibers and membrane protrusion, we examined the levels of active GTP-bound RhoA. Surprisingly we could not detect an obvious change in total RhoA activity after TRIP6 overexpression in MDCK, MDCKts-src, and HEK293 cells (data not shown).
We have previously demonstrated the critical role of the PI3K/PTEN pathway in the control of invasiveness in MDCK and MDCKts-src cells (1
, 2
, 23)
, and more recently we reported that constitutively active Akt, a downstream effector of PI3K, was sufficient to trigger invasiveness (2)
. Here we show that TRIP6-induced invasiveness was dependent on PI3K activity, because this effect was reverted by the PI3K selective inhibitor LY294002 (Fig. 7A
). As TRIP6-induced invasiveness required PI3K activity, we investigated whether TRIP6 might modulate Akt activity. Ectopic expression of TRIP6 stimulated Akt activity in MDCKts-src and HEK293 cells (Fig. 7C
and data not shown). This effect was independent of the interaction of TRIP6 with MAGI-1b, since it was mimicked by TRIP6Stop473.
It has been recently demonstrated that TRIP6 may function as a coactivator of NF-B-regulated promoter (12)
. In line with this report, we found a strong activation (3-fold) of a NF-B-regulated reporter in MDCKts-src cells derivatives expressing either TRIP6wt or TRIP6 Stop473 cotransfected with plasmids pIgkB-luc and pcDNA3-βgal (Fig. 7D
). Since TRIP6 enhanced NF-B transcriptional activity, we evaluated the implication of NF-B in TRIP6-induced cell invasion. The treatment of invasive MDCKts-src TRIP6wt cells (clone 15) with NF-B inhibitory peptide PTD-p65, but not with the control peptide, results in inhibition of invasiveness (Fig. 7E
).
Furthermore, to investigate the direct involvement of NF-B in invasiveness, MDCKts-src cells were stably transfected with constructs expressing either p50 or p65 subunits of NF-B and then seeded on type-I collagen at the temperature restrictive for Src activity (40°C). Overexpression of either NF-B p50 or p65 induced invasiveness of MDCKts-src cells at the temperature restrictive for Src activity (Fig. 7E
), indicating the involvement of this pathway in invasiveness.
These data suggest that NF-B is involved in invasiveness mediated by TRIP6, pointing to a possible mechanism responsible for the permissive role of TRIP6 core protein in this process.
TRIP-6 expression in human colorectal tumors
Because TRIP6 overexpression promotes invasiveness, we evaluated the accumulation of the corresponding transcripts in colorectal tumors. We analyzed by qRT-PCR 12 human colonic tumor samples paired with control mucosa, that is, 5 adenomas and 7 adenocarcinomas, and 2 preparations of purified colonic epithelial crypts (Fig. 8
). TRIP6 transcripts were identified in colonic mucosa, as well as in highly purified colonic epithelial crypts, demonstrating the expression of TRIP6 in colonic epithelial cells. The accumulation of TRIP6 transcripts was 2.1 ± 0.60-fold higher in adenoma (median value 1.72), and 4.89 ± 1.10-fold higher in adenocarcinomas (P<0.02; median value 4.23) as compared to the paired control colonic mucosa. In these samples, histological analysis revealed that stromal compartment accounted for less than 20%. Furthermore, we did not determine any change in TRIP6 expression in an inflammatory sample of colonic mucosa (diverticulitis) as compared to control mucosa (0.96 vs. 1, not shown), suggesting that TRIP6 overexpression in colonic tumors is not related to inflammatory process. Thus, the overexpression of TRIP6 in human colonic adenocarcinomas might exert a critical role in the progression and the invasive properties of colorectal cancers.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. In the present study, we aimed at identifying, by Y2H screening, molecular partners for MAGI-1b that might antagonize β-catenin binding to its fifth PDZ domain. In this way we identified TRIP6. Our present report provides the first evidence for the association of TRIP6 with the MAGI-1b scaffolding molecule in epithelial cells and demonstrates the importance of this multimolecular complex in cell signaling and invasiveness. So far, no study has evaluated the involvement of the PDZ-binding domain of TRIP6 and related molecules in cell scattering. We demonstrated that TRIP6 ectopic expression decreases cell-cell aggregation and promotes invasiveness of MDCK kidney epithelial cells and their MDCKts-src derivatives. This effect is clearly dependent on the TRIP6 PDZ-binding motif, because the depletion of TRIP6 C terminus (mutant TRIP6 Stop473) impairs the induction of the invasive phenotype. A peptide corresponding to this TRIP6 C terminus was by itself unable to trigger invasiveness but complemented TRIP6Stop473 in achieving this effect. This transcomplementation indicates that the core protein of TRIP6 exerts a permissive role that requires its C terminus to weaken adherens junctions through the competitive inhibition of PDZ-dependent scaffolds.
TRIP6 was recently reported to interact with the second PDZ domain of the Tyr phosphatase PTP-BL/hPTP1E and with the third PDZ domain of the tumor suppressor Scrib (4
, 6
, 7)
. Using 2-hybrid assays, we confirmed the interaction of PTP-BL with TRIP6, but we did not observe an interaction of PTP-BL with β-catenin. Because the peptides corresponding to the C terminus of either TRIP6 or β-catenin complemented TRIP6Stop473, it is unlikely that PTP-BL is the TRIP6 target involved in invasiveness. Therefore, MAGI-1b, which we identified here as a novel molecular partner of TRIP6, is a good candidate as a downstream target of TRIP6.
We previously reported that the ectopic expression of MAGI-1b stabilizes junctional complexes and reverts invasiveness by recruiting PTEN to cadherin/β-catenin complexes (2)
. We show here that both TRIP6 and β-catenin interact with the fifth domain of MAGI-1b, and that the intracellular delivery of competing C-terminal β-catenin and C-terminal TRIP6 peptides, but not the control peptide, reverts the invasion suppression activity of MAGI-1b in MDCKts-src cells grown at the temperature permissive for Src activity. These results are a strong indication that the decrease of cell-cell adhesion by TRIP6 is partly dependent on competition with β-catenin for binding to MAGI-1b, and related to the inability to recruit the tumor suppressor PTEN to cell-cell contacts.
Another interesting point concerns the effector systems of TRIP6 scaffolds involved in the induction of invasiveness. We show here that TRIP6 modulates Rac1, Akt, and NFB activities. Rac1 activation in epithelia is associated with opposite outcomes: first, promotion of cell-cell junctions and enhancement of the epithelial phenotype and, second, formation of lamellipodia during epithelial cell migration (29
30
31)
. Our data point to the fact that overexpression of TRIP6wt in epithelial cells results in the induction of invasiveness and the down-regulation of total Rac1 activity in MDCKts-src derivatives. In this connection, it was found that Rac1 activity is elevated in TRIP6-depleted HeLa cells, resulting in abnormal burst of actin polymerization and dynamic membrane protrusions (18)
. In the present study, we determined that the inactivation of Rac1 required TRIP6 PDZ-binding motif. Because TRIP6wt uncoupled E-cadherin junctions from actin cytoskeleton, down-regulation of Rac1 activity could serve to promote destabilization of the actin cytoskeleton at the subapical adhesion belt of epithelial monolayers. Accordingly, activation of the ARF6 GTPase promotes the decrease in Rac1-GTP levels in addition to facilitating the endocytosis of E-cadherin (32)
. The induction of invasiveness by TRIP6 requires PI3K activity, because it is reverted by the selective inhibitor LY294002. Downstream of PI3K, we determined that TRIP6 induces Akt activation. This effect of TRIP6 might be partly involved in the induction of invasiveness because we previously reported that constitutively active Akt is sufficient to destabilize junctional complexes and to trigger invasiveness. The activation of Akt by TRIP6 does not seem to be related to changes in PTEN accumulation, nor to recruitment of TRIP6 in PDZ molecular scaffolds, as it was phenocopied by TRIP6Stop473. In this respect it was recently reported that in cardiomyocytes zyxin, which does not contain a PDZ binding motif, interacts with Akt and promotes its nuclear accumulation in response to cGMP (33)
. It should be stressed that TRIP6 was also reported to modulate the transcriptional activity of AP-1 and NF-B (12)
. We have demonstrated here that the ectopic expression of TRIP6wt or TRIP6Stop473 enhanced the transactivation of NF-B responsive elements. TRIP6-dependent invasiveness required NF-B activity, because it was reverted by the PTD-p65 inhibitory peptide. Conversely, the overexpression of the NF-B p65 or p50 subunits promoted invasiveness of MDCKts-src cells at the temperature restrictive for Src activity. In this connection, it was recently reported that activation of NF-B induces the epithelial-mesenchymal transition (34)
.
Thus, TRIP6 might act at several molecular levels to weaken adherens junctions, to promote actin cytoskeleton reorganization and cell invasiveness. Our results concerning a role for TRIP6 in cell scattering is in agreement with recent studies demonstrating that TRIP6 enhances lysophosphatidic acid (LPA) -induced cell migration by interacting with the LPA2 receptor. Indeed, TRIP6 directly binds to the C-terminal tail of the LPA2 receptor through its LIM domains, and LPA-dependent recruitment of TRIP6 to the plasma membrane promotes its targeting to focal adhesions and the migration of SKOV3 ovarian carcinoma and 3T3 fibroblast cell lines (19
, 35)
.
Taken together, our results indicate that TRIP6 functions at a point of convergence of several signaling pathways, including activity of junctional complexes, actin cytoskeleton remodeling, and gene expression. The differential biological effects of TRIP6 are likely to be dependent on subtle balances in accumulation of TRIP6 itself and of its molecular partners, their subcellular targeting (i.e., plasma membrane, cytoplasm, focal adhesion or nucleus), competition with other zyxin-related proteins with LIM domain (e.g., LPP, zyxin or ajuba), and existence of a PDZ-binding motif present in both TRIP6 and LPP but missing in ajuba, zyxin, and LIMD1. We propose that, according to the cellular context, TRIP6 might compete with β-catenin for the binding with the MAGI-1b/PTEN signalosome to destabilize E-cadherin junctional complexes and to promote cell motility through the regulation of Akt/NF-B targets and/or effectors of focal adhesion (Fig. 9
). These findings are in line with an important role of PDZ-mediated interactions in the shaping and organization of submembranous microenvironments of cell adhesion in epithelial cells and in the control of cell scattering. In this connection, the TRIP6 overexpression we documented here in human colorectal adenocarcinomas sustains a critical role of this adaptor in the neoplastic progression.
|
|
ACKNOWLEDGMENTS |
|---|
Received for publication April 17, 2008. Accepted for publication October 16, 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. K. McMichael, S. M. Meyer, and B. S. Lee c-Src-mediated Phosphorylation of Thyroid Hormone Receptor-interacting Protein 6 (TRIP6) Promotes Osteoclast Sealing Zone Formation J. Biol. Chem., August 20, 2010; 285(34): 26641 - 26651. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
