- Email: [email protected]
A proposed role for ZO-1 in targeting connexin 43 gap junctions to the endocytic pathway
Biochimie 86 (2004) 241–244 www.elsevier.com/locate/biochi
A proposed role for ZO-1 in targeting connexin 43 gap junctions to the endocytic pathway D. Segretain a,b, C. Fiorini a,b, X. Decrouy a,b, N. Defamie a,b, J.R. Prat c, G. Pointis a,b,* a
Inserm EMI 00-09, IFR 50, Université Paris V, 45, rue des Saint-Pères, 75006 Paris, France b Faculté de Médecine, 28, avenue de Valombrose, 06107 Nice cedex 2, France c Institut Curie d’Orsay, Section de Recherche UMR 146, Plateforme d’Imagerie cellulaire, 91405 Orsay cedex, France Received 10 May 2004; accepted 13 May 2004 Available online 26 May 2004
Abstract Gap junctions are intercellular channels organized in plaque that directly link adjacent cells. Connexins (Cx), the constitutive proteins of gap junctions are associated with several partner proteins (cytoskeletal, anchoring) which could participate in plaque formation and degradation. Coimmunoprecipitation and indirect immunofluorescence analyses showed that ZO-1, a tight junction-associated protein, was linked to Cx43 in the testis. By using c-hexachlorocyclohexane (HCH), known to induce gap junction endocytosis, we demonstrated that endocytosis increased Cx43/ZO-1 association within the cytoplasm of treated Sertoli cells. In control cells, the two proteins were present, as expected, at the plasma membrane level, but poorly colocalized. The increased intracytoplasmic Cx43/ZO-1 complex was associated with a shift towards increased levels of Cx43 P1 and P2 isoforms. The HCH induced Cx43 hyperphosphorylation was abolished by the ERK inhibitor PD98059 suggesting that this effect could be mediated through activation of the ERK pathway. These data strongly support a novel role for ZO-1 in the turnover of Cx43 during gap junction plaque endocytosis. © 2004 Elsevier SAS. All rights reserved. Keywords: Z0-1; Gap junction plaque; Cx43; Endocytosis
1. Introduction Cell–cell communication is critical for tissue homeostasis and cell proliferation. Composed by channels which coupled adjacent cells, the functionality of gap junctions has been extensively examinated [1]. These channels are formed by an hexameric arrangement of connexins (Cx), among which Cx43 (43 kDa) is the most representative Cx in mammalians. Due to the rapid turnover of these proteins (2–5 h), the Cxs permanently cycle from synthesis to plaque clustering and formation toward their final internalisation and degradation [2]. These proteins have several partners, such as cytoskeletal components, anchoring proteins and adapters, which are possibly implicated in the traffic, plaque formation and functionality. Recently, it was established that Cx43 interacted with the tight junction-associated protein ZO-1, a member of the * Corresponding author. Tel.: +33-493-377-737; fax: +33-493-377-731. E-mail address: [email protected] (G. Pointis). 0300-9084/$ - see front matter © 2004 Elsevier SAS. All rights reserved. doi:10.1016/j.biochi.2004.05.003
MAGUK family, through the C terminal region of Cx43 and the second PDZ domain of ZO-1 [3,4]. Other Cxs have now been reported to bind to ZO-1 [5,6]. ZO-2 has been also demonstrated as another direct binding partner for Cx43 [7]. In contrast to Cx43, Cx45 interacted with the PDZ domains of ZO-3 but not with those of ZO-2 [8]. Since ZO-1 is tightly associated with a-spectrin, an actin linking protein, it has been suggested that ZO-1 might serve to link gap junctions to the actin cytoskeleton as reported for tight and adherent junctions [6,9]. It has been also proposed that ZO-1 could be involved in the targeting of connexons to specialized plasma membrane domains, regulation of interaction between different Cxs during the formation of heteromeric connexon, stabilization of Cx at the plasma membrane and control of channel function [10]. Using a carcinogenic toxicant (HCH), known to induce endocytosis of Cx43 gap junctions in cell types as hepatocytes [11] and Sertoli cells [12,13], we have examinated the close association of Cx43 and ZO-1 during gap junction endocytosis in a Sertoli cell line. Here, we suggest that an
242
D. Segretain et al. / Biochimie 86 (2004) 241–244
increased Cx43/ZO-1 interaction may be a key mechanism implicated in the gap junction plaque endocytosis.
2. Materials and methods 2.1. Cell culture and treatments The 42GPA9 Sertoli cell line was maintained as previously described [14]. To better visualize the gap junction plaques, Cx43GFP cDNA was transfected in 42GPA9 Sertoli cells as previously described [2,15]. Induction of gap junction endocytosis was performed by exposure of the cells to 50 µM c-hexachlorocyclohexane (HCH, from Sigma) [12,13]. The Park Davis Inhibitor PD98059 (MEK1-specific inhibitor, New England Biolabs) was added at a concentration of (10 µM) for 90 min prior to the addition of HCH. 2.2. Immunoprecipitation and western blotting Rat testis lysates were immunoprecipitated with 2 µg of either anti-Cx43 (Transduction Laboratories, Lexington, KY, USA), or -ZO-1 (Zymed Laboratories, CA) antibodies for 16 h at 4 °C. Briefly, the immunocomplexes were precipitated by incubation with protein A-Sepharose (Pharmacia) for 1 h at 4 °C. Immunopellets were washed four times in PBS 0.5% NP 40, once in PBS, and then eluted with 40 µl of reduced Laemmli buffer at 100 °C for 5 min. Testis immunoprecipitates or lysates of cells cultured with or without HCH (50 µM) were separated on 5–15% SDS-PAGE, electroblotted onto a Polyvinylidene fluoride membrane (PVDF Immobilon-P, Millipore) and analyzed by western blotting with either anti-Cx43 (1:2000) or anti-ZO-1 (1:1500) antibodies as previously described [12]. 2.3. Immunocytochemical procedures Cells were fixed in cold methanol at –20 °C for 5 min and washed with 0.1% Tween 20-PBS for 5 min. Tubulin was revealed with a monoclonal anti-b-tubulin (Zymed Laboratories Inc., CA). For colocalization experiments fixed cells were incubated with a mixture containing the anti-Cx43 antibody and anti-ZO-1 antibody (1:100). Subsequently, the slides were incubated in a mixture containing FITC-conjugated goat anti-mouse IgG (1:50) (Dako) and rhodamine TRITC-conjugated anti-rat or anti-rabbit IgG (1:200) (Amersham) in PBS containing 3% bovine serum albumin. Slides were examined with a confocal microscope (Leica TCS SP) fitted with a 488 or 543 nm krypton/argon laser allowing simultaneous analysis of the fluorescein and rhodamine chromophores and/or a wide field deconvolution microscope as recently described [16].
3. Results The Sertoli cell line used in the present work strongly expressed Cx43 [17] and formed Cx43 gap junctional plaques, between adjacent cells, visualized when cells were transfected with Cx43-GFP cDNA (Fig. 1A). The dynamic of Cx is presented in an hypothetical model which represents gap junction assembly and turnover and the potential role of ZO-1 in these processes (Fig. 1B). To evaluate the possible
Fig. 1. Cx43 gap junction plaques in Cx43-GFP transfected Sertoli cells (A) Deconvolution microscopy of Cx43-GFP in transfected 42GPA9 Sertoli cells allows to focus on gap junction plaques between two adjacent cells. Cx43 is distributed within round plaque seen in face view (open arrow) and as small line for the side view (arrow). Endocytic Cx43-GFP particles are present in the cytoplasm (arrow head).Tubulin (red) fills the cell cytoplasm. The nucleus are DAPI stained (blue). X 800. (B) Schematic hypothetical model for the role of ZO-1 in Cx trafficking.
involvement of ZO-1 in endocytosis and degradation of the gap junction plaque, the molecular association of Cx43 and ZO-1 was first investigated in testis lysates. Cx43 and ZO-1 were immunoprecipitated from testis lysates and Cx43 or ZO-1-containing complexes were analyzed by western blotting with anti-Cx43 and -ZO-1 antibodies. As shown in Fig. 2 (upper panels) Cx43 antibodies coimmunoprecipitated ZO-1. The association of Cx43 and ZO-1 was further supported by the presence of Cx43 in ZO-1 immunoprecipitates. One band around 220 kDa for ZO-1 and three bands around 45 kDa for Cx43, corresponding to phosphorylated (P1 and P2) and unphosphorylated isoforms (P0) were detected. The indirect immunocolocalization experiments demonstrated that Cx43 and ZO-1 were present at the plasma membrane in cells cultured in the absence of HCH (Fig. 2, middle panels). However, at higher magnification the two proteins exhibited poor colocalization at the plasma membrane level. In contrast, when cells were exposed to HCH, known to induce the endocytic process, the plasma membrane signals for Cx43 and ZO-1 totally disappeared and was redistributed in the cytoplasm. High magnification revealed that the majority of the cytoplasmic dots exhibited a yellow region indicating a
D. Segretain et al. / Biochimie 86 (2004) 241–244
Fig. 2. Analysis of Cx43/ZO-1 association during induced endocytosis upper panels: Selective association of Cx43 with ZO-1. Testis lysates were immunoprecipitated (IP) with anti-Cx43 or anti-ZO-1 antibodies. Immunoprecipitates were analyzed by ECL western blotting immunodetection with anti-Cx43, and ZO-1 antibodies. Cx43 and ZO-1 are detected in testis lysates at the predicted size of 39–45, and 220 kDa respectively. P0, unphosphorylated; P1 and P2, phosphorylated isoforms of Cx43. Middle panels: Cx43 (green) and ZO-1 (red) colocalization in 42GPA9 cells. In control cells both proteins are located at the plasma membrane and underline the contour of the cells (X 600). Higher magnification reveals the moderate colocalization of the signals (arrows in inset). In HCH-treated cells, the two immunoreactive signals are cytoplasmic and exhibited a more pronounced colocalization at high magnification (inset, yellow signal). Lower panels: ECL western blot immunodetection of Cx43 isoforms in control and cells in which Cx43/ZO-1 endocytosis was induced by HCH. Note that HCH increased the levels of P1, P2 isoforms of Cx43 as compared to control cells. The use of a MAP kinase inhibitor (PD98059) restores the control level of the phosphorylated isoforms of the protein. Representative of three different experiments.
clear colocalization of the two immunoreactive signals (inset). It is well established that the phosphorylation status of Cx43 is essential not only for channel gating but also in the regulation of gap junction assembly and turnover. Fig. 2 (lower panels) shows that exposure of cells to HCH induced hyperphosphorylation of Cx43 with a shift towards increased levels of P1 and P2 isoforms of the protein concomitantly to the endocytosis of the Cx43/ZO-1 complex. This hyperphosphorylation was abolished by the addition of the ERK inhibitor PD98059 suggesting that this effect was mediated through an activation of the ERK pathway. 4. Discussion The present data clearly show the existence of a direct interaction between Cx43 and ZO-1 as seen by dual immunofluorescence and coimmunoprecipitation analyses in testis and in a Sertoli cell line. The present findings also demonstrate that this protein–protein association is not only maintained but also reinforced during the endocytosis of gap junction plaques.
243
Recently we demonstrated that stimulation of the endocytic process by HCH resulted in a intracytoplasmic delocalization of the membranous signal for both Cx43 and ZO-1, while localization of other tight junction proteins such as occludin and claudin was not affected [12]. The current immunofluorescence study associated with a fine analysis of the signals by the use of confocal microscopy confirms these data. Our results also show that the two signals for Cx43 and ZO-1 highly colocalized in treated cells while only a moderate plasma membrane colocalization of the Cx43 and ZO-1 was found in control cells. Such absence of complete colocalization of Cx43/ZO-1 signals was also observed in normal testis tissue analyzed by deconvolution microscopy (data not shown). There is now clear evidence that ZO-1 is required for Cx43 transport during the assembly of the gap junctions (see for review [9]). From the current data, we can now suggest that ZO-1 plays a previously uncovered role in the turnover of Cx43 specifically during gap junction plaque endocytosis. These results are in agreement with previous studies showing alteration of the gap junction turnover when ZO-1 interaction is abolished by mutation [18] or C-terminal tagging of Cx43 [15,19]. An increase in ZO-1 and Cx43 association was also reported after enzymatic dissociation of myocytes, a treatment known to induce gap junction plaques endocytosis [20]. In the latter study, it was proposed that the Cx43/ZO-1 complex appeared within the annular gap junction. By using deconvolution microscopy analysis of the Cx43 signal we recently reported that Cx43 was associated with early endosomes during testicular tumoral progression [16] or in HCHtreated Sertoli cells [13]. Since the Sertoli cell line used in the present study presented no annular gap junctions as demonstrated by electron microscopy studies [17], we speculate that the Cx43/ZO-1 association could be present in the Rab5 early endosomes. This proposal is under investigation. There is now strong evidence that the Cx43/ZO-1 complex linked to the actin filaments through an actin linking protein, a-spectrin [4]. Thus, ZO-1 could serve as an intermediate and/or a linker molecule for Cx43 positioning with actin. We recently demonstrated the essential role of actin during gap junction plaques internalization [2]. Altogether these findings suggest that the Cx43/ZO-1/actin complex could actively participate in the internalization process of gap junction plaques. Phosphorylation of Cx has been implicated in the regulation of gap junctional communication at different levels: trafficking, assembly–disassembly, gating of channels and degradation [21]. Endocytosis of Cx43 gap junctions has been previously reported to be associated with an increase in the phosphorylated isoforms of Cx43 in different models [11,20]. Interestingly, our data indicate that gap junction endocytosis induced by HCH exposure of the cells was mainly associated with increasing phosphorylation of P1 and P2 isoforms of Cx43. In addition, our findings suggest that the increased Cx43 phosphorylated isoforms resulted from activation of the ERK/mitogen-activated protein kinase pathway. This is supported by the observations that both the shift
244
D. Segretain et al. / Biochimie 86 (2004) 241–244
toward the P1 and P2 Cx43 isoforms and the delocalisation of the Cx43/ZO-1 association (data not shown) were abolished when the ERK pathway was inhibited by PD98059. Thus, the existence of a relationship between Cx43 phosphorylation status and the increased association between Cx43 and ZO-1 during gap junction plaques endocytosis could be suggested. In conclusion, the present data emphasize the implication of protein–protein interactions in Cx trafficking and focus on a specific role for ZO-1 in the rapid Cx turnover during gap junctional plaque endocytosis. Acknowledgments This work was supported by the Institut National de la Santé et de la Recherche Médicale (INSERM), Université Paris V and a fund from Région Ile de France. Authors are grateful to C. Avondet for technical assistance and to J. Gilleron, M. Tramoni and A. Valette for cell culture. C.F. was a recipient from Pfizer and X.D. from Aventis Pharma.
[9]
[10]
[11]
[12]
[13]
[14]
[15]
References [16] [1] [2]
[3]
[4]
[5]
[6] [7] [8]
F. Bukauskas, V. Verselis, Gap junction channel gating, Biochim. Biophys. Acta 1662 (2004) 42–60. D. Segretain, M. Falk, Regulation of connexin biosynthesis, assembly, gap junction formation and removal, Biochim. Biophys. Acta 1662 (2004) 3–21. B. Giepmans, W. Moolenaar, The gap junction protein connexin43 interacts with the second PDZ domain of the zona occludens-1 protein, Curr. Biol. 8 (1998) 931–934. T. Toyofuku, M. Yabuki, K. Otsu, T. Kuzuya, M. Hori, M. Tada, Direct association of the gap junction protein connexin-43 with ZO-1 in cardiac myocytes, J. Biol. Chem. 273 (1998) 12725–12731. P. Nielsen, A. Baruch, V. Shestopalov, B. Giepman, I. Dunia, L. Benedetti, N. Kumar, Lens connexins Cx46 and Cx50 interact with zonula occludens protein-1 (ZO-1), Mol. Biol. Cell 14 (2003) 2470–2481. B. Giepmans, Gap junctions and connexin-interacting proteins, Cardiovasc. Res. 62 (2004) 233–245. D. Singh, P. Lampe, Identification of connexin-43 interacting proteins, Cell Commun. Adhes 10 (2003) 215–220. P. Kausalya, M. Reichert, W. Hunziker, Connexin45 directly binds to ZO-1 and localizes to the tight junction region in epithelial MDCK cells, FEBS Lett. 505 (2001) 92–96.
[17]
[18]
[19]
[20]
[21]
J. Herve, N. Bourmeyster, D. Sarrouilhe, Diversity in protein–protein interactions of connexins: emerging roles, Biochim. Biophys. Acta 1662 (2004) 22–41. M. Thomas, S. Huang, A. Cokoja, O. Riccio, O. Staub, S. Suter, M. Chanson, Interaction of connexins with protein partners in the control of channel turnover and gating, Biol. Cell 94 (2002) 445–456. X. Guan, R. Ruch, Gap junction endocytosis and lysosomal degradation of connexin43-P2 in WB-F344 rat liver epithelial cells treated with DDT and lindane, Carcinogenesis 17 (1996) 1791–1798. N. Defamie, B. Mograbi, C. Roger, L. Cronier, A. Malassine, F. Brucker-Davis, P. Fenichel, D. Segretain, G. Pointis, Disruption of gap junctional intercellular communication by lindane is associated with aberrant localization of connexin43 and zonula occludens-1 in 42GPA9 Sertoli cells, Carcinogenesis 22 (2001) 1537–1542. B. Mograbi, E. Corcelle, N. Defamie, M. Samson, M. Nebout, D. Segretain, P. Fenichel, G. Pointis, Aberrant connexin 43 endocytosis by the carcinogen lindane involves activation of the ERK/mitogen-activated protein kinase pathway, Carcinogenesis 24 (2003) 1415–1423. V. Bourdon, A. Lablack, P. Abbe, D. Segretain, G. Pointis, Characterization of a clonal Sertoli cell line using adult PyLT transgenic mice, Biol. Reprod. 58 (1998) 591–599. U. Lauf, B. Giepmans, P. Lopez, S. Braconnot, S. Chen, M. Falk, Dynamic trafficking and delivery of connexons to the plasma membrane and accretion to gap junctions in living cells, Proc. Natl. Acad. Sci. USA 99 (2002) 10446–10451. D. Segretain, X. Decrouy, J. Dompierre, D. Escalier, N. Rahman, C. Fiorini, B. Mograbi, J.P. Siffroi, I. Huhtaniemi, P. Fenichel, G. Pointis, Sequestration of connexin43 in the early endosomes: an early event of Leydig cell tumor progression, Mol. Carcinog. 38 (2003) 179–187. A. Lablack, V. Bourdon, N. Defamie, C. Batias, M. Mesnil, P. Fenichel, G. Pointis, D. Segretain, Ultrastructural and biochemical evidence for gap junction and connexin 43 expression in a clonal Sertoli cell line: a potential model in the study of junctional complex formation, Cell Tissue Res. 294 (1998) 279–287. T. Toyofuku, H. Akamatsu, T. Zhang, T. Kuzuya, M. Tada, M. Hori, C-Src regulates the interaction between connexin-43 and ZO-1 in cardiac myocytes, J. Biol. Chem. 276 (2001) 1780–1788. A. Hunter, J. Jourdan, R. Gourdie, Fusion of GFP to the carboxyl terminus of connexin43 increases gap junction size in HeLa cells, Cell Commun. Adhes. 10 (2003) 211–214. R.J. Barker, R.L. Price, R.G. Gourdie, Increased association of ZO-1 with connexin43 during remodeling of cardiac gap junctions, Circ. Res. 90 (2002) 317–324. P. Lampe, A. Lau, The effects of connexin phosphorylation on gap junctional communication, Int. J. Biochem. Cell Biol. 36 (2004) 1171–1186.
A proposed role for ZO-1 in targeting connexin 43 gap junctions to the endocytic pathway D. Segretain a,b, C. Fiorini a,b, X. Decrouy a,b, N. Defamie a,b, J.R. Prat c, G. Pointis a,b,* a
Inserm EMI 00-09, IFR 50, Université Paris V, 45, rue des Saint-Pères, 75006 Paris, France b Faculté de Médecine, 28, avenue de Valombrose, 06107 Nice cedex 2, France c Institut Curie d’Orsay, Section de Recherche UMR 146, Plateforme d’Imagerie cellulaire, 91405 Orsay cedex, France Received 10 May 2004; accepted 13 May 2004 Available online 26 May 2004
Abstract Gap junctions are intercellular channels organized in plaque that directly link adjacent cells. Connexins (Cx), the constitutive proteins of gap junctions are associated with several partner proteins (cytoskeletal, anchoring) which could participate in plaque formation and degradation. Coimmunoprecipitation and indirect immunofluorescence analyses showed that ZO-1, a tight junction-associated protein, was linked to Cx43 in the testis. By using c-hexachlorocyclohexane (HCH), known to induce gap junction endocytosis, we demonstrated that endocytosis increased Cx43/ZO-1 association within the cytoplasm of treated Sertoli cells. In control cells, the two proteins were present, as expected, at the plasma membrane level, but poorly colocalized. The increased intracytoplasmic Cx43/ZO-1 complex was associated with a shift towards increased levels of Cx43 P1 and P2 isoforms. The HCH induced Cx43 hyperphosphorylation was abolished by the ERK inhibitor PD98059 suggesting that this effect could be mediated through activation of the ERK pathway. These data strongly support a novel role for ZO-1 in the turnover of Cx43 during gap junction plaque endocytosis. © 2004 Elsevier SAS. All rights reserved. Keywords: Z0-1; Gap junction plaque; Cx43; Endocytosis
1. Introduction Cell–cell communication is critical for tissue homeostasis and cell proliferation. Composed by channels which coupled adjacent cells, the functionality of gap junctions has been extensively examinated [1]. These channels are formed by an hexameric arrangement of connexins (Cx), among which Cx43 (43 kDa) is the most representative Cx in mammalians. Due to the rapid turnover of these proteins (2–5 h), the Cxs permanently cycle from synthesis to plaque clustering and formation toward their final internalisation and degradation [2]. These proteins have several partners, such as cytoskeletal components, anchoring proteins and adapters, which are possibly implicated in the traffic, plaque formation and functionality. Recently, it was established that Cx43 interacted with the tight junction-associated protein ZO-1, a member of the * Corresponding author. Tel.: +33-493-377-737; fax: +33-493-377-731. E-mail address: [email protected] (G. Pointis). 0300-9084/$ - see front matter © 2004 Elsevier SAS. All rights reserved. doi:10.1016/j.biochi.2004.05.003
MAGUK family, through the C terminal region of Cx43 and the second PDZ domain of ZO-1 [3,4]. Other Cxs have now been reported to bind to ZO-1 [5,6]. ZO-2 has been also demonstrated as another direct binding partner for Cx43 [7]. In contrast to Cx43, Cx45 interacted with the PDZ domains of ZO-3 but not with those of ZO-2 [8]. Since ZO-1 is tightly associated with a-spectrin, an actin linking protein, it has been suggested that ZO-1 might serve to link gap junctions to the actin cytoskeleton as reported for tight and adherent junctions [6,9]. It has been also proposed that ZO-1 could be involved in the targeting of connexons to specialized plasma membrane domains, regulation of interaction between different Cxs during the formation of heteromeric connexon, stabilization of Cx at the plasma membrane and control of channel function [10]. Using a carcinogenic toxicant (HCH), known to induce endocytosis of Cx43 gap junctions in cell types as hepatocytes [11] and Sertoli cells [12,13], we have examinated the close association of Cx43 and ZO-1 during gap junction endocytosis in a Sertoli cell line. Here, we suggest that an
242
D. Segretain et al. / Biochimie 86 (2004) 241–244
increased Cx43/ZO-1 interaction may be a key mechanism implicated in the gap junction plaque endocytosis.
2. Materials and methods 2.1. Cell culture and treatments The 42GPA9 Sertoli cell line was maintained as previously described [14]. To better visualize the gap junction plaques, Cx43GFP cDNA was transfected in 42GPA9 Sertoli cells as previously described [2,15]. Induction of gap junction endocytosis was performed by exposure of the cells to 50 µM c-hexachlorocyclohexane (HCH, from Sigma) [12,13]. The Park Davis Inhibitor PD98059 (MEK1-specific inhibitor, New England Biolabs) was added at a concentration of (10 µM) for 90 min prior to the addition of HCH. 2.2. Immunoprecipitation and western blotting Rat testis lysates were immunoprecipitated with 2 µg of either anti-Cx43 (Transduction Laboratories, Lexington, KY, USA), or -ZO-1 (Zymed Laboratories, CA) antibodies for 16 h at 4 °C. Briefly, the immunocomplexes were precipitated by incubation with protein A-Sepharose (Pharmacia) for 1 h at 4 °C. Immunopellets were washed four times in PBS 0.5% NP 40, once in PBS, and then eluted with 40 µl of reduced Laemmli buffer at 100 °C for 5 min. Testis immunoprecipitates or lysates of cells cultured with or without HCH (50 µM) were separated on 5–15% SDS-PAGE, electroblotted onto a Polyvinylidene fluoride membrane (PVDF Immobilon-P, Millipore) and analyzed by western blotting with either anti-Cx43 (1:2000) or anti-ZO-1 (1:1500) antibodies as previously described [12]. 2.3. Immunocytochemical procedures Cells were fixed in cold methanol at –20 °C for 5 min and washed with 0.1% Tween 20-PBS for 5 min. Tubulin was revealed with a monoclonal anti-b-tubulin (Zymed Laboratories Inc., CA). For colocalization experiments fixed cells were incubated with a mixture containing the anti-Cx43 antibody and anti-ZO-1 antibody (1:100). Subsequently, the slides were incubated in a mixture containing FITC-conjugated goat anti-mouse IgG (1:50) (Dako) and rhodamine TRITC-conjugated anti-rat or anti-rabbit IgG (1:200) (Amersham) in PBS containing 3% bovine serum albumin. Slides were examined with a confocal microscope (Leica TCS SP) fitted with a 488 or 543 nm krypton/argon laser allowing simultaneous analysis of the fluorescein and rhodamine chromophores and/or a wide field deconvolution microscope as recently described [16].
3. Results The Sertoli cell line used in the present work strongly expressed Cx43 [17] and formed Cx43 gap junctional plaques, between adjacent cells, visualized when cells were transfected with Cx43-GFP cDNA (Fig. 1A). The dynamic of Cx is presented in an hypothetical model which represents gap junction assembly and turnover and the potential role of ZO-1 in these processes (Fig. 1B). To evaluate the possible
Fig. 1. Cx43 gap junction plaques in Cx43-GFP transfected Sertoli cells (A) Deconvolution microscopy of Cx43-GFP in transfected 42GPA9 Sertoli cells allows to focus on gap junction plaques between two adjacent cells. Cx43 is distributed within round plaque seen in face view (open arrow) and as small line for the side view (arrow). Endocytic Cx43-GFP particles are present in the cytoplasm (arrow head).Tubulin (red) fills the cell cytoplasm. The nucleus are DAPI stained (blue). X 800. (B) Schematic hypothetical model for the role of ZO-1 in Cx trafficking.
involvement of ZO-1 in endocytosis and degradation of the gap junction plaque, the molecular association of Cx43 and ZO-1 was first investigated in testis lysates. Cx43 and ZO-1 were immunoprecipitated from testis lysates and Cx43 or ZO-1-containing complexes were analyzed by western blotting with anti-Cx43 and -ZO-1 antibodies. As shown in Fig. 2 (upper panels) Cx43 antibodies coimmunoprecipitated ZO-1. The association of Cx43 and ZO-1 was further supported by the presence of Cx43 in ZO-1 immunoprecipitates. One band around 220 kDa for ZO-1 and three bands around 45 kDa for Cx43, corresponding to phosphorylated (P1 and P2) and unphosphorylated isoforms (P0) were detected. The indirect immunocolocalization experiments demonstrated that Cx43 and ZO-1 were present at the plasma membrane in cells cultured in the absence of HCH (Fig. 2, middle panels). However, at higher magnification the two proteins exhibited poor colocalization at the plasma membrane level. In contrast, when cells were exposed to HCH, known to induce the endocytic process, the plasma membrane signals for Cx43 and ZO-1 totally disappeared and was redistributed in the cytoplasm. High magnification revealed that the majority of the cytoplasmic dots exhibited a yellow region indicating a
D. Segretain et al. / Biochimie 86 (2004) 241–244
Fig. 2. Analysis of Cx43/ZO-1 association during induced endocytosis upper panels: Selective association of Cx43 with ZO-1. Testis lysates were immunoprecipitated (IP) with anti-Cx43 or anti-ZO-1 antibodies. Immunoprecipitates were analyzed by ECL western blotting immunodetection with anti-Cx43, and ZO-1 antibodies. Cx43 and ZO-1 are detected in testis lysates at the predicted size of 39–45, and 220 kDa respectively. P0, unphosphorylated; P1 and P2, phosphorylated isoforms of Cx43. Middle panels: Cx43 (green) and ZO-1 (red) colocalization in 42GPA9 cells. In control cells both proteins are located at the plasma membrane and underline the contour of the cells (X 600). Higher magnification reveals the moderate colocalization of the signals (arrows in inset). In HCH-treated cells, the two immunoreactive signals are cytoplasmic and exhibited a more pronounced colocalization at high magnification (inset, yellow signal). Lower panels: ECL western blot immunodetection of Cx43 isoforms in control and cells in which Cx43/ZO-1 endocytosis was induced by HCH. Note that HCH increased the levels of P1, P2 isoforms of Cx43 as compared to control cells. The use of a MAP kinase inhibitor (PD98059) restores the control level of the phosphorylated isoforms of the protein. Representative of three different experiments.
clear colocalization of the two immunoreactive signals (inset). It is well established that the phosphorylation status of Cx43 is essential not only for channel gating but also in the regulation of gap junction assembly and turnover. Fig. 2 (lower panels) shows that exposure of cells to HCH induced hyperphosphorylation of Cx43 with a shift towards increased levels of P1 and P2 isoforms of the protein concomitantly to the endocytosis of the Cx43/ZO-1 complex. This hyperphosphorylation was abolished by the addition of the ERK inhibitor PD98059 suggesting that this effect was mediated through an activation of the ERK pathway. 4. Discussion The present data clearly show the existence of a direct interaction between Cx43 and ZO-1 as seen by dual immunofluorescence and coimmunoprecipitation analyses in testis and in a Sertoli cell line. The present findings also demonstrate that this protein–protein association is not only maintained but also reinforced during the endocytosis of gap junction plaques.
243
Recently we demonstrated that stimulation of the endocytic process by HCH resulted in a intracytoplasmic delocalization of the membranous signal for both Cx43 and ZO-1, while localization of other tight junction proteins such as occludin and claudin was not affected [12]. The current immunofluorescence study associated with a fine analysis of the signals by the use of confocal microscopy confirms these data. Our results also show that the two signals for Cx43 and ZO-1 highly colocalized in treated cells while only a moderate plasma membrane colocalization of the Cx43 and ZO-1 was found in control cells. Such absence of complete colocalization of Cx43/ZO-1 signals was also observed in normal testis tissue analyzed by deconvolution microscopy (data not shown). There is now clear evidence that ZO-1 is required for Cx43 transport during the assembly of the gap junctions (see for review [9]). From the current data, we can now suggest that ZO-1 plays a previously uncovered role in the turnover of Cx43 specifically during gap junction plaque endocytosis. These results are in agreement with previous studies showing alteration of the gap junction turnover when ZO-1 interaction is abolished by mutation [18] or C-terminal tagging of Cx43 [15,19]. An increase in ZO-1 and Cx43 association was also reported after enzymatic dissociation of myocytes, a treatment known to induce gap junction plaques endocytosis [20]. In the latter study, it was proposed that the Cx43/ZO-1 complex appeared within the annular gap junction. By using deconvolution microscopy analysis of the Cx43 signal we recently reported that Cx43 was associated with early endosomes during testicular tumoral progression [16] or in HCHtreated Sertoli cells [13]. Since the Sertoli cell line used in the present study presented no annular gap junctions as demonstrated by electron microscopy studies [17], we speculate that the Cx43/ZO-1 association could be present in the Rab5 early endosomes. This proposal is under investigation. There is now strong evidence that the Cx43/ZO-1 complex linked to the actin filaments through an actin linking protein, a-spectrin [4]. Thus, ZO-1 could serve as an intermediate and/or a linker molecule for Cx43 positioning with actin. We recently demonstrated the essential role of actin during gap junction plaques internalization [2]. Altogether these findings suggest that the Cx43/ZO-1/actin complex could actively participate in the internalization process of gap junction plaques. Phosphorylation of Cx has been implicated in the regulation of gap junctional communication at different levels: trafficking, assembly–disassembly, gating of channels and degradation [21]. Endocytosis of Cx43 gap junctions has been previously reported to be associated with an increase in the phosphorylated isoforms of Cx43 in different models [11,20]. Interestingly, our data indicate that gap junction endocytosis induced by HCH exposure of the cells was mainly associated with increasing phosphorylation of P1 and P2 isoforms of Cx43. In addition, our findings suggest that the increased Cx43 phosphorylated isoforms resulted from activation of the ERK/mitogen-activated protein kinase pathway. This is supported by the observations that both the shift
244
D. Segretain et al. / Biochimie 86 (2004) 241–244
toward the P1 and P2 Cx43 isoforms and the delocalisation of the Cx43/ZO-1 association (data not shown) were abolished when the ERK pathway was inhibited by PD98059. Thus, the existence of a relationship between Cx43 phosphorylation status and the increased association between Cx43 and ZO-1 during gap junction plaques endocytosis could be suggested. In conclusion, the present data emphasize the implication of protein–protein interactions in Cx trafficking and focus on a specific role for ZO-1 in the rapid Cx turnover during gap junctional plaque endocytosis. Acknowledgments This work was supported by the Institut National de la Santé et de la Recherche Médicale (INSERM), Université Paris V and a fund from Région Ile de France. Authors are grateful to C. Avondet for technical assistance and to J. Gilleron, M. Tramoni and A. Valette for cell culture. C.F. was a recipient from Pfizer and X.D. from Aventis Pharma.
[9]
[10]
[11]
[12]
[13]
[14]
[15]
References [16] [1] [2]
[3]
[4]
[5]
[6] [7] [8]
F. Bukauskas, V. Verselis, Gap junction channel gating, Biochim. Biophys. Acta 1662 (2004) 42–60. D. Segretain, M. Falk, Regulation of connexin biosynthesis, assembly, gap junction formation and removal, Biochim. Biophys. Acta 1662 (2004) 3–21. B. Giepmans, W. Moolenaar, The gap junction protein connexin43 interacts with the second PDZ domain of the zona occludens-1 protein, Curr. Biol. 8 (1998) 931–934. T. Toyofuku, M. Yabuki, K. Otsu, T. Kuzuya, M. Hori, M. Tada, Direct association of the gap junction protein connexin-43 with ZO-1 in cardiac myocytes, J. Biol. Chem. 273 (1998) 12725–12731. P. Nielsen, A. Baruch, V. Shestopalov, B. Giepman, I. Dunia, L. Benedetti, N. Kumar, Lens connexins Cx46 and Cx50 interact with zonula occludens protein-1 (ZO-1), Mol. Biol. Cell 14 (2003) 2470–2481. B. Giepmans, Gap junctions and connexin-interacting proteins, Cardiovasc. Res. 62 (2004) 233–245. D. Singh, P. Lampe, Identification of connexin-43 interacting proteins, Cell Commun. Adhes 10 (2003) 215–220. P. Kausalya, M. Reichert, W. Hunziker, Connexin45 directly binds to ZO-1 and localizes to the tight junction region in epithelial MDCK cells, FEBS Lett. 505 (2001) 92–96.
[17]
[18]
[19]
[20]
[21]
J. Herve, N. Bourmeyster, D. Sarrouilhe, Diversity in protein–protein interactions of connexins: emerging roles, Biochim. Biophys. Acta 1662 (2004) 22–41. M. Thomas, S. Huang, A. Cokoja, O. Riccio, O. Staub, S. Suter, M. Chanson, Interaction of connexins with protein partners in the control of channel turnover and gating, Biol. Cell 94 (2002) 445–456. X. Guan, R. Ruch, Gap junction endocytosis and lysosomal degradation of connexin43-P2 in WB-F344 rat liver epithelial cells treated with DDT and lindane, Carcinogenesis 17 (1996) 1791–1798. N. Defamie, B. Mograbi, C. Roger, L. Cronier, A. Malassine, F. Brucker-Davis, P. Fenichel, D. Segretain, G. Pointis, Disruption of gap junctional intercellular communication by lindane is associated with aberrant localization of connexin43 and zonula occludens-1 in 42GPA9 Sertoli cells, Carcinogenesis 22 (2001) 1537–1542. B. Mograbi, E. Corcelle, N. Defamie, M. Samson, M. Nebout, D. Segretain, P. Fenichel, G. Pointis, Aberrant connexin 43 endocytosis by the carcinogen lindane involves activation of the ERK/mitogen-activated protein kinase pathway, Carcinogenesis 24 (2003) 1415–1423. V. Bourdon, A. Lablack, P. Abbe, D. Segretain, G. Pointis, Characterization of a clonal Sertoli cell line using adult PyLT transgenic mice, Biol. Reprod. 58 (1998) 591–599. U. Lauf, B. Giepmans, P. Lopez, S. Braconnot, S. Chen, M. Falk, Dynamic trafficking and delivery of connexons to the plasma membrane and accretion to gap junctions in living cells, Proc. Natl. Acad. Sci. USA 99 (2002) 10446–10451. D. Segretain, X. Decrouy, J. Dompierre, D. Escalier, N. Rahman, C. Fiorini, B. Mograbi, J.P. Siffroi, I. Huhtaniemi, P. Fenichel, G. Pointis, Sequestration of connexin43 in the early endosomes: an early event of Leydig cell tumor progression, Mol. Carcinog. 38 (2003) 179–187. A. Lablack, V. Bourdon, N. Defamie, C. Batias, M. Mesnil, P. Fenichel, G. Pointis, D. Segretain, Ultrastructural and biochemical evidence for gap junction and connexin 43 expression in a clonal Sertoli cell line: a potential model in the study of junctional complex formation, Cell Tissue Res. 294 (1998) 279–287. T. Toyofuku, H. Akamatsu, T. Zhang, T. Kuzuya, M. Tada, M. Hori, C-Src regulates the interaction between connexin-43 and ZO-1 in cardiac myocytes, J. Biol. Chem. 276 (2001) 1780–1788. A. Hunter, J. Jourdan, R. Gourdie, Fusion of GFP to the carboxyl terminus of connexin43 increases gap junction size in HeLa cells, Cell Commun. Adhes. 10 (2003) 211–214. R.J. Barker, R.L. Price, R.G. Gourdie, Increased association of ZO-1 with connexin43 during remodeling of cardiac gap junctions, Circ. Res. 90 (2002) 317–324. P. Lampe, A. Lau, The effects of connexin phosphorylation on gap junctional communication, Int. J. Biochem. Cell Biol. 36 (2004) 1171–1186.