JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M605261200 on February 5, 2007

J. Biol. Chem., Vol. 282, Issue 13, 9789-9796, March 30, 2007
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
282/13/9789    most recent
M605261200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Duffy, H. S.
Right arrow Articles by Spray, D. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Duffy, H. S.
Right arrow Articles by Spray, D. C.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

The Gap Junction Protein Connexin32 Interacts with the Src Homology 3/Hook Domain of Discs Large Homolog 1*

Heather S. Duffy{ddagger}1, Ionela Iacobas§, Kylie Hotchkiss{ddagger}, Bethany J. Hirst-Jensen, Alejandra Bosco§, Nadine Dandachi§, Rolf Dermietzel||, Paul L. Sorgen, and David C. Spray§

From the {ddagger}Department of Pharmacology, Columbia University, New York, New York 10032, §Department of Neuroscience, Albert Einstein College of Medicine, Bronx, New York 10461, Department of Biochemistry and Molecular Biology, University of Nebraska, Omaha, Nebraska 68918, and ||Department of Neuroscience and Molecular Brain Research, University of Bochum, D-44780 Bochum, Germany

Received for publication, June 1, 2006 , and in revised form, February 5, 2007.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Scaffolding of membrane proteins is a common strategy for forming complexes of proteins, including some connexins, within membrane microdomains. Here we describe studies indicating that Cx32 interacts with a PDZ-containing scaffolding protein, Dlgh1 (Discs Large homolog 1). Initial screens of liver lysates using antibody arrays indicated an interaction between Cx32 and Dlgh1 that was confirmed using coimmunoprecipitation studies. Yeast two-hybrid complementation determined that the Cx32 bound via interaction with the SH3/Hook domain of Dlgh1. Confocal microscopy of liver sections revealed that Cx32 and Dlgh1 could colocalize in hepatocyte membranes in wild type mice. Examination of levels and localization of Dlgh1 in livers from Cx32 null mice indicate that, in the absence of Cx32, Dlgh1 was decreased, and the remainder was translocated from the hepatocyte membrane to the nucleus with some remaining in cytoplasmic compartments. This translocation was confirmed by Western blots comparing Dlgh1 levels in nuclear extracts from wild type and Cx32 null murine livers. Using SKHep cells stably transfected with Cx32 under the control of a tet-off promoter, we found that acute removal of Cx32 led to a decrease of membrane-localized Dlgh1 and an increase in the nuclear localization of this tumor suppressor protein. Together, these results suggest that loss of Cx32 alters the levels, localization, and interactions of the tumor suppressor protein Dlgh1, events known in other systems to alter cell cycle and increase tumorigenicity.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Connexins, of which there are more than 20 isoforms in mice and humans, are tetraspan proteins that oligomerize to form hexameric connexons. Connexons contributed by each cell interact head-to-head across the extracellular space, forming channels providing direct cytoplasmic continuity between cells. Aggregates of these channels, gap junctions, are found in almost all tissues. Connexins have intracellular amino-terminal cytoplasmic loops and carboxyl-terminal domains. Although much of the connexin sequence is highly conserved throughout the protein family, the amino acid sequences of intracellularly localized cytoplasmic loop and carboxyl-terminal domains vary markedly between connexin isoforms, and these regions likely confer isoform specificity. Carboxyl-terminal connexin domains, in particular, contain multiple sites for protein-protein interactions. For example, the major gap junction protein of heart and astrocytes, Connexin43 (Cx43),2 possesses binding sites for Src homology 3 (SH3), SH2, WW, MAPK (mitogen-activated protein kinase), and PDZ within its carboxyl-terminal domain, and other connexins, although less well mapped, also display potential binding sites. The function of these protein-protein interactions is not well understood, but many of them regulate the function of the gap junction channel (1). Interestingly, analysis of the primary sequence of the {alpha} subdivision of connexins (which includes Cx43) for consensus binding sites indicates that many of these connexins have potential PDZ interaction domains, whereas the primary sequences of the beta subdivision of connexins (including the major gap junction protein of liver and myelinating glia, Cx32) do not contain these potential interaction sites. This leads to the question of whether the {alpha} connexins are scaffolded while the beta connexins are not or whether there are alternative binding domains within the PDZ-containing proteins that interact with beta connexins.

The PDZ domain-containing proteins, including Discs Large and its mammalian homolog (Dlgh1), belong to a family of proteins known as membrane-associated guanylate kinases (MAGUKs). These proteins have multiple functions, including scaffolding of ion channels, heterodimerization to form protein microdomains, and regulation of junctional complexes. Dlgh1, in particular, has been shown to be involved in cellular growth control via a nuclear translocation mechanism (2). The PDZ domains are the primary sites of protein-protein interactions with many macromolecular components, but the SH3 region, with its "hook" region also participates in the formation of protein-protein interactions (reviewed in Funke et al. (3)).

We here report our studies examining the interaction of a beta connexin, Cx32, with the MAGUK protein Dlgh1. In normal murine livers, Cx32 and Dlgh1 were found to colocalize at sites of cell-cell contact, whereas in livers of Cx32 null mice and in hepatoma cells, where Cx32 was acutely down-regulated, expression of Dlgh1 was reduced and the remaining Dlgh1 protein was largely localized within the nucleus. As revealed by yeast two-hybrid complementation, the interaction of Cx32 with Dlgh1 is non-canonical, involving the SH3/Hook/GUK region rather than the PDZ domains. When Dlgh1 is localized to cell membranes, it regulates cell growth, blocking cell cycle in the G0/G1 phase. Translocation of hDlg to the nucleus has been shown to increase cell proliferation (3). Loss of Cx32 has been shown to promote the formation of induced hepatic tumors (4, 5). Thus, we propose that one mechanism by which Cx32 may act to control cell proliferation in liver is through maintenance of membrane-associated Dlgh1 in addition to the intracellular communication that gap junctions provide.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Liver Lysate Preparation—For antibody arrays (whole lysates from 5 mg of liver tissue), Western (50 µg) and coimmunoprecipitation studies (whole lysates from 5 mg of tissue protein matched across samples). Adult C56B mice were anesthetized using isoflurane and decapitated prior to removal of the liver. Whole liver was extracted, rinsed in phosphate-buffered saline (PBS), pH 7.4, and lysed in complete lysis buffer (50 mM Tris-HCl, pH 7.4, 0.25 mM sodium-deoxycholate, 150 mM NaCl, 2 mM EGTA, 0.1 mM Na3VO4, 10 mM NaF, 1 mM phenylmethylsulfonyl fluoride, 1% Triton X-100, 1/2 tablet of Complete protease inhibitor (Roche Applied Science). Lysates were sonicated for 30 s, maintained on ice for 30 min, and then triturated and spun at 10,000 revolutions/min for 10 min. Total protein was assessed using the BCA protein assay kit (Pierce). Lysates were then equalized to 50 µg by dilution in lysis buffer. The supernatant was retained for use in arrays and coimmunoprecipitation studies.

Antibody Arrays—Hypromatrix (Worcester, MA) antibody arrays were reacted with liver lysate and probed with HRP-tagged Cx32 antibody according to the manufacturer's recommendations. In-house antibody arrays were produced on polyvinylidene difluoride membranes and spotted with 32 different antibodies (6). For this, membranes were wet in 100% methanol and rinsed three times in PBS. Antibodies (0.05 µg/spot) were individually spotted onto the membrane using a BioDot array spotter (Bio-Rad). All arrays were made in triplicate for statistical analysis. Following drying of the array, the membrane was blocked for 2 h in 5% skim milk with 0.05% Tween 20, incubated with lysates from whole liver (see above) for 2 h, rinsed three times for 10 min in PBS with 0.05% Tween 20 (PBST), exposed to HRP-conjugated Cx32 antibody (antibody conjugated to HRP using the EZ Link Biotech kit (Pierce catalog number 31489) according to the manufacturer's instructions for 2 h in 5% skim milk and then rinsed, and HRP was detected using the ECL protein detection kit (Amersham Biosciences). Membranes were exposed to x-ray film and then developed on an Exomat developing system.

Coimmunoprecipitation Studies—Coimmunoprecipitations were performed as previously described (3). Briefly, whole liver lysates were treated as noted above to produce supernatant (50 µg of total protein). Supernatant was precleared by spinning with 50 µl of Sepharose A beads (Amersham Biosciences) in buffer (250 mg of Sepharose A), incubated in 2 ml of lysis buffer on ice for 15 min. The beads were spun at 3000 revolutions/min for 3 min and washed 3 times in lysis buffer. Following the wash, the beads were diluted 1:1 beads:lysis buffer and kept at 4 °C. Following preclear, 1.5 µg of Cx32 antibody was added (Cx32, polyclonal Zymed Laboratories Inc., Temecula, CA), and Dlgh1, monoclonal antibody (Affinity Bioreagents), and supernatant were rotated for 4 h at 4°C. Thereafter, 100 µl of Sepharose beads was added, and the sample was rotated overnight at 4 °C. The following day, the sample was spun at 5000 revolutions/min for 10 min and supernatant discarded. The beads were rinsed six times in lysis buffer, incubated in 60 µl of loading buffer (2x Laemmli buffer plus dithiothreitol), incubated in 100 °C water for 2 min, and then placed immediately on ice for 2 min. The sample was then spun and the loading buffer removed from the beads for running on Western blots using 10% SDS-polyacrylamide gels. Proteins were transferred to nitrocellulose membrane, and Cx32 or Dlgh1 were detected with the appropriate antibodies.

Yeast Two-Hybrid Complementation Assays—The Saccharomyces cerevisiae yeast strain AH109 was cotransformed with the following GAL4ad fusion constructs: GAL4ad-NT (hDlg-(1–233)), GAL4ad-PDZ1–2 (hDlg-(200–463)), GAL4ad-PDZ3 (hDlg-(451–587)), GAL4ad-SH3/Hook (hDlg-(569–764)), and GAL4ad-SV40 Large T antigen (control) or with the following GAL4bd fusion constructs: GAL4bd-Cx32CT (Cx32-(209–283)), GAL4bd-Cx32CL (Cx32-(96–127)), GAL4bd-Cx32NT (Cx32-(1–22)), and GAL4bd-pVA3 murine p53 (control). Cotransformants were assayed for growth on nonselective (+His) and selective (–His) media.

Cell Culture—Parental and Cx32-tet-off transfected SkHep cells were plated directly on coverslips and grown until 80–90% confluent. The cells were treated with 30 µM tetracycline (13) for 0, 24, 48, or 96 h and used for immunocytochemistry.

Immunofluorescence Microscopy—Rapidly frozen liver samples were sectioned (15 microns) using a Shandon electronic cryotome. Sections or cultured SkHep1 cells were fixed in 4% formaldehyde for 30 min at room temperature and then incubated in 50 mM NH4Cl for 3 h to quench autofluorescence. Following quench, sections were blocked (PBS plus 10% goat serum plus 0.4% Triton X-100) for 1 h at room temperature and then incubated with primary monoclonal antibodies directed against Cx32 (Sigma) and polyclonal Dlgh1/SAP97 (Affinity Bioreagents) at 4 °C overnight. Following 30 min of rinsing (3x for 10 min in PBS plus 0.4% Triton X-100), the slices were incubated with secondary antibodies (Alexa Fluor anti-mouse 488 and anti-rabbit 595) for 1 h at room temperature. The slices were rinsed for 50 min (5x for 10 min) and mounted on glass microscope slides with Vectrashield anti-fade agent (Vector Laboratories, Burlingame, CA) and examined on a confocal microscope (Olympus BX61WQ Fluoview confocal system). Images were processed using Imaris Image Pro software.

Nuclear Extraction—For SKHep1 cells, at 0, 24, 49, and 96 h time points, 100-mm dishes of confluent SkHep cells were washed with PBS and prepared for nuclear extraction (Bio-Rad) according to their instructions. Samples were subjected to protein level determination using the BCA protein assay kit (Pierce), run on either 10% (Cx32) or 7.5% (Dlgh1) SDS-polyacrylamide gels, and probed using antibodies against Cx32 and Dlgh1, respectively, as described above.


Figure 1
View larger version (32K):
[in this window]
[in a new window]

  		FIGURE 1.
Antibody array detection of Cx32 interaction with Dlgh1. A, antibody arrays were spotted with different antibodies (Spot a = ANT; Spot b = Claudin-1; Spot c = Dlgh1; Spot d = VDAC), incubated with lysates from murine liver, and then probed with HRP-tagged anti-Cx32 antibodies (Sigma). Although there was slight reaction above background with Spots a, b, and d, a strong HRP signal was seen at Spot c on all arrays. This indicates the presence of Cx32 in the complex captured by the Dlgh1 antibody. B, intensity of interactions was determined by densitometry. Note the high intensity of the interaction of Cx32 with the complex captured by the Dlgh1 antibody, suggesting an interaction between Cx32 and Dlgh1 (n = 3). Confirmation of the interaction between Cx32 and Dlgh1 was done using coimmunoprecipitation studies. C, lysates of murine liver were incubated with antibodies directed against either Cx32 (Cx32IP) or Dlgh1 (Dlgh1IP) and probed for Dlgh1. Note that both the Cx32 antibody as well as the Dlgh1 antibody pulled down a band of 140 kDa labeled with anti-Dlgh1. The reverse coimmunoprecipitation experiment shows that both the Cx32 antibody as well as the Dlgh1 antibody pull a protein that labels with the antibody directed against Cx32. These studies confirm the antibody array data and indicate that Cx32 and Dlgh1 are within the same protein complex in murine liver. Western blot analysis of Dlgh1 was done from wild type and Cx32 null murine livers (C, lanes 1 and 2). The protein levels were normalized against beta-actin and GAPDH, and the change in Dlgh1 was quantified. Quantification of total levels of Dlgh1 shows an ~75% decrease in cytoplasmic Dlgh1 protein in the liver from Cx32 null mice (D).

 
Segregation of Nuclear and Cytoplasmic Fractions—Isolation and preparation of cells harvested from whole mouse liver. Wild type and Cx32 null mice were anesthetized with isoflurane (inhaled). The left ventricle was located and perfused with an EGTA-containing solution (0.5 mM EGTA, 15 mM HEPES, 130 mM NaCl, 0.5 mM Na2HPO4, 10 mM glucose, 4 mM KCl, 1 mM MgCl2, 5 mM taurine) to remove blood and flush the organs; the right atrium was cut to facilitate solution outflow. The system was perfused until the liver was completely blanched, and the liver was dissected and washed with PBS. To obtain a cell suspension, the liver was ground with a pestle and successively strained through two mesh cell separators of 230 and 80 µm each. The cells were collected by centrifugation, washed twice with PBS, and placed on ice.

Analysis of Cytoplasmic and Nuclear Proteins for Cx32 and Dlgh1—The cytoplasmic and nuclear samples were assayed for protein concentration (Bio-Rad) and their volumes determined. Cytoplasmic (50 µg) and a proportional amount of nuclear protein were loaded in Laemmli buffer for separation by 10% SDS-PAGE. Proteins were transferred to polyvinylidene difluoride membrane (Millipore) for Western blotting. The membranes were probed with antibodies to Cx32 and Dlgh1 (Santa Cruz Biotechnology) and visualized on x-ray film using chemiluminescence (Pierce). The purity of the cytoplasmic and nuclear fractions was confirmed by probing the membranes with antibodies to markers of the cell cytoplasm (Rho-GDI, Santa Cruz Biotechnology) and nucleus (Histone H1, NeoMarkers; or Lamin A/C, Santa Cruz Biotechnology). Western blotting was analyzed by densitometry.

Measurements of Junctional Conductance—Cell pairs in cultures seeded at low density were voltage-clamped at a holding potential of 0 mV using patch pipettes connected to an Axo 1C patch clamp amplifier; junctional conductance was calculated as the current recorded in one cell divided by a 10 mV step applied to the other cell.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
In targeted immunoprecipitation experiments with likely partners, Cx32 has previously been shown to interact with Caveolin1 (7), Cx26 (8), occludin (9), and calmodulin (10). To identify additional binding partners, using an unbiased approach, we initially screened for Cx32 interactions within liver lysates using Hypromatrix antibody arrays, finding a number of novel potential interactions including Dlgh1 (results not shown). These were followed up with in-house antibody arrays in which we spotted membranes with antibodies that we had screened for specificity for the target protein. When incubated with whole liver lysates and probed with HRP-tagged anti-Cx32, the in-house antibody arrays indicated that, within liver, Cx32 interacts with several complexes of proteins, including a complex containing the scaffolding protein Dlgh1 (Fig. 1A). In this illustration, Spot c contained antibodies directed against Dlgh1, and Cx32-HRP binding determined that Cx32 was within the complex of proteins captured at this spot. Densitometric analysis of the spot indicates that the signal was much stronger for the Cx32 interaction with Dlgh1 than with ANT (adenine nucleotide translocator) protein (spot a), Claudin-1 (spot b), or VDAC (voltage-dependent anion-selective channel) (spot d), although for each of these spots the signal was higher than background, presumably indicating weak or higher order interactions with Cx32 (Fig. 1B, n = 3). The Dlgh1-Cx32 interaction was confirmed using coimmunoprecipitation studies (Fig. 1C). The reverse coimmunoprecipitation also indicated an interaction between these two proteins with the Dlgh1 antibody immunoprecipitates containing a 32-kDa protein that was immunoreactive to the Cx32 antibodies. Immunoprecipitation of Dlgh1 from Cx32 null livers indicated that there is no signal for Dlgh1 when Cx32 antibody was used for pulldowns in the Cx32 null background (data not shown). As a control for immunoprecipitation experiments, we performed Western blots of liver homogenate from Cx32 null mice. As shown in Fig. 1C, second lane, Dlgh1 remained in the liver lysate from Cx32 null mice. However, in quantitative experiments in which Dlgh1 levels were normalized against beta-actin levels, Dlgh1 abundance was substantially reduced (~4-fold) in livers from Cx32 null mice (1D).

It was of interest to determine whether the interaction between Cx32 and hDlg was via a PDZ domain, particularly because the Cx32 carboxyl-terminal domain does not contain a canonical PDZ interaction site. Thus, we performed a yeast two-hybrid assay to identify which of the Cx32 and Dlgh1 domains physically associate in vivo (Fig. 2). The CT, CL, and NT domains of Cx32 and the NT, PDZ1–2, PDZ3, and SH3/Hook domains of Dlgh1 (see "Experimental Procedures") were coexpressed in AH109 yeast cells as GAL4 DNA-binding domain (bd) fusion proteins and as GAL4 transactivation domain (ad) fusion proteins, respectively. Interestingly, there was no interaction between any Cx32 cytoplasmic domain and any PDZ domain of Dlgh1. Instead, a direct interaction was observed between the hDlg SH3/Hook domain and both the Cx32CT and Cx32CL domains. To determine whether this interaction was seen in native tissues, we performed immunocytochemistry on wild type and Cx32 null liver. For these studies, 15-µm sections of liver were stained for localization of Cx32 and Dlgh1 (Fig. 3, A–H). The cell nuclei were stained with DAPI (A and E). In wild type liver, Cx32 localizes around cell membranes (Fig. 3B) as does Dlgh1 (Fig. 3C). An overlay of these images shows apparent colocalization of Cx32 and Dlgh1 at sites of cell membranes (Fig. 3D). In contrast, in the Cx32 null livers, as expected, there is no staining for Cx32 (Fig. 3F), and Dlgh1 is no longer primarily localized to the membranes (Fig. 3G). Overlay of these images indicates that the remaining Dlgh1 is found within the cell nuclei rather than at the membranes. Higher magnification images of wild type (Fig. 3I) and Cx32 null hepatocytes (Fig. 3J) show how dramatic the changes in Dlgh1 localization are in the Cx32 null livers. To verify translocation of Dlgh1 protein into the cell nuclei, where it might play a functional role, we extracted nuclei from wild type and Cx32 null livers and probed the nuclear extracts for both Cx32 and Dlgh1 (Fig. 3K). Using Rho-GD1 to determine the purity of the cytoplasmic fraction and histone to determine the purity of the nuclear fraction, we found that Cx32 was primarily present in the cytoplasmic fraction of cells from wild type livers, with little signal seen in the nuclear component, whereas Cx32 was absent from both liver fractions from Cx32 null mice. The small amount of Cx32 seen in the nucleus in wild type liver likely corresponds to a small cytoplasmic contaminant evidenced by a low but visible Rho-GD1 signal in the wild type liver nuclear fraction. Dlgh1 was found in the cytoplasm and at low levels in the nucleus of the hepatocytes from wild type livers. However, there was a large increase in the levels of Dlgh1 within the nuclear fraction of the Cx32 null livers (Fig. 3, K and L).


Figure 2
View larger version (30K):
[in this window]
[in a new window]

  		FIGURE 2.
Characterization of the hDlg and Cx32 interaction by the yeast two-hybrid assay. The S. cerevisiae yeast strain AH109 was cotransformed with GAL4ad fusion constructs GAL4ad-NT (hDlg-(1–233)), GAL4ad-PDZ1–2 (hDlg-(200–463)), GAL4ad-PDZ3 (hDlg-(451–587)), GAL4ad-SH3/Hook (hDlg-(569–764)), and GAL4ad-SV40 Large T antigen (control) or with GAL4bd fusion constructs GAL4bd-Cx32CT (Cx32-(209–283)), GAL4bd-Cx32CL (Cx32-(96–127)), GAL4bd-Cx32NT (Cx32-(1–22)), and GAL4bd-pVA3 murine p53 (control). Cotransformants were assayed for growth on nonselective (+His) and selective (–His) media. ad, activation domain; bd, DNA-binding domain. B, schematic representation of the interaction between Cx32 and Dlgh1. The exact amino acid sequences of the Cx32 are as yet unknown.

 


Figure 3
View larger version (88K):
[in this window]
[in a new window]

  		FIGURE 3.
Confocal immunolocalization of Cx32 and Dlgh1 in wild type and Cx32 null murine liver. A, the number and location of cells was marked by DAPI staining of the hepatocyte nuclei. B, Cx32 was localized in a punctate fashion on the plasma membrane of hepatocyte around the entire periphery of most cells. C, Dlgh1 localized to the plasma membrane in a solid line of protein. D, confocal colocalization of these two proteins indicates that these two proteins overlap at most sites of Cx32 immunoreactivity, suggesting that Cx32 is in close contact with Dlgh1 in these cells. E, nuclear staining by DAPI. F, as expected, there was no reactivity to the anti-Cx32 antibody, confirming the knock-out of Cx32 in these animals. G, staining for Dlgh1 showed an altered localization of this protein in the livers of knock-out Cx32 mice. In these cells, Dlgh1 is no longer primarily localized to cell membranes in these hepatocytes; instead it appears to localize either around or within the nuclei of these hepatocytes (H). Higher magnification image of Cx32 and Dlgh1 at a section between two hepatocytes shows punctate Cx32 staining around Dlgh1 regions on the membrane (I) but primarily nuclear localization in the Cx32 null livers (J). To determine whether the Dlgh1 protein was translocating into the nucleus, Western blot analysis of nuclear Dlgh1 levels in Cx32 knock-out liver was done. Following nuclear extraction and purification, liver nuclear extracts from both wild type and Cx32 knock-out mice were blotted for Cx32 and Dlgh1 (K). Histone was used to determine the purity of the nuclear fraction, whereas Rho-GD1 was used to determine the purity of the cytoplasmic fraction. The nuclei of wild type animals (n = 3) showed little Cx32 but did contain a small amount of Dlgh1, whereas nuclei from Cx32 knock-out mice (n = 3) showed no Cx32 but had a large increase in the levels of Dlgh1 (K) within these purified nuclear extracts (L). There were no changes in overall cadherin levels (M), and localization of cadherin in the Cx32 null hepatocytes (O) was identical to that seen in the wild type hepatocytes (N).

 


Figure 4
View larger version (52K):
[in this window]
[in a new window]

  		FIGURE 4.
Colocalization profiles Cx32, Dlgh1, and nuclei in confocal stacks imaged from sections of wild type (left) and Cx32 null (right) liver. Colocalization profiles of the boxed regions show that, in wild type livers, Dlgh1 and Cx32 colocalize at plasmallemal regions of the cells, and Dlgh1 shows no nuclear localization (left). In contrast, Dlgh1 and the cell nuclei show strong colocalization profiles in the Cx32 null liver sections (right).

 
Cadherin, the functional component of adherens junctions, functions to maintain normal cell-cell contact. To determine whether there were changes in cadherin expression that might alter the ability of cells to adhere and form proper cell-cell contacts in the Cx32 null livers, we evaluated cadherin levels using Western blot analysis and immunostaining. Using a pancadherin antibody, which identifies both the 135-kDa isoform as well as a 120-kDa isoforms, we found no significant difference between the levels of cadherin in hepatocytes from Cx32 null livers (Fig. 3M). In conjunction with the Western blot analysis, immunostaining studies of cadherin revealed no difference in the staining pattern for cadherin in wild type and Cx32 null liver sections (Fig. 3, N and O). Colocalization analysis of Cx32, Dlgh1, and nuclei were done on confocal stacks of images from both wild type and Cx32 null liver sections using the Imaris colocalization algorithms (11). Regions of the wild type liver were selected in which colocalization profiles of Cx32 with Dlgh1 were characterized (Fig. 4, high magnification image in upper right panel), and regions of Cx32 null liver sections were selected for calculating colocalization of Dlgh1 with DAPI-stained nuclei (Fig. 4, high magnification image in upper right panel). Wild type liver exhibited regions along cell membranes where Cx32 and Dlgh1 were colocalized within the entire confocal stack (Fig. 4, lower right panel). In contrast, the localization profile of Dlgh1 showed strong colocalization with the nuclei in Cx32 null livers throughout the confocal stack (Fig. 4, lower right panel). Chronic long term loss of Cx32, as is seen in the Cx32 null mouse, might be expected to trigger alterations in cell function that subsequently lead to a loss of Dlgh1 at cell membranes. Thus, the loss of Cx32 might not be a primary step in the Dlgh1 alterations seen in the Cx32 null livers. To determine whether similar changes occurred with acute loss of Cx32, we used the normally connexin-deficient SkHep1 cell human hepatoma cell line transfected with Cx32 driven by a tet-off promoter (12). In the absence of tetracycline, Cx32 is robustly expressed in these cells (Fig. 5A), and Dlgh1 is prominent at cell membranes (Fig. 4). Overlay shows regions of overlap of these two proteins. Dual whole cell voltage clamp studies on cell pairs indicate that these cells are highly coupled (Fig. 5M). After 24 h of tetracycline treatment, Cx32 was no longer seen at cell membranes (Fig. 5D); Dlgh1 was still present, although its localization at the membrane was patchy, generally not fully encircling the cells (Fig. 5E). Overlay of these images shows substantially less colocalization of these two proteins than before tetracycline treatment (Fig. 5F). Junctional conductance was reduced by >10-fold in cells treated for 24 h with tetracycline (Fig. 5M). Extending the tetracycline treatment to 48 h further decreased Cx32 levels (Fig. 5G), and Dlgh1 was primarily located in the cytoplasm rather than at the membrane (Fig. 5H). Junctional conductance remained very low (Fig. 5M). Overlay of the images indicates no colocalization of Cx32 and Dlgh1. By 96 h of treatment, complete loss of Cx32 protein occurred (Fig. 5J), and Dlgh1 was then seen only localizing with cell nuclei (Fig. 5K, overlay 5L).


Figure 5
View larger version (64K):
[in this window]
[in a new window]

  		FIGURE 5.
Confocal studies of SkHep cells transfected with tet-off Cx32. SkHep cells were immunostained for either Cx32 (green) or Dlgh1 (red) and examined for the amount of overlay of the two proteins. A shows images of non-tetracycline-treated SkHep cells, where Cx32 can be seen at sites of cell-cell contact, and as is common with connection-transfected cells, perinuclearly. Dlgh1 can be seen around the periphery of the cells outlining the cell membranes (B). Overlay images show that these two proteins show areas of colocalization in these untreated SkHep cells (C). Following 24 h of tetracycline treatment, Cx32 was reduced to the perinuclear fraction (D), whereas the Dlgh1 was still seen on parts of the membrane, although slightly decreased as compared with the untreated cells (E). Overlay images show less colocalization of Cx32 with Dlgh1 following the 24 h treatment period (F). G shows that 48 h of tetracycline treatment obliterates almost all Cx32, and Dlgh1 is then found to be almost exclusively nuclear or perinuclear (H). Overlay images show no colocalization between the two proteins (I). By 96 h of tetracycline treatment, there was no Cx32 expression (J), the Dlgh1 was localized within the nuclei (K), and again, no colocalization of the two proteins was seen (L). Electrophysiological studies indicate that the Cx32 forms functional channels in these cells when tetracycline is absent, but coupling is lost as early as only 24 h after treatment and remained low for the entire treatment period (M).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
Initial studies showing interaction of connexins with scaffolding proteins have begun to find multiple interacting proteins with a small subset of connexins. The studies reported here are the first to show interaction of Cx32 with such binding partners. Although connexins turn over in the plasma membrane rather rapidly (13), scaffolding proteins remain. Thus, although the connexins may aid in the organization of the microdomain, the scaffolding protein is likely to be the more stable link. This would set up a "target" for cycling connexins that may underlie how connexins are added to the plaque along the outer edges of the plaques (14). Cross-talk between these proteins could then allow for a symbiotic relationship where, at least in this case, Cx32 may "signal" where the microdomain forms, and once recruited, Dlgh1 is needed to form a stable platform. Thus, we hypothesize that, when Cx32 is initially removed from the cells, the Dlg1 remains for a while, but when newly translated Cx32 does not insert into the plasma membrane, the organizational signal is lost and the Dlgh1 is no longer directed to the plasma membrane but rather to the cell nucleus as a portion of the signal that indicates loss of cell-cell coupling. The remaining Dlgh1 is degraded over time, causing a delay between the loss of Cx32 and translocation of Dlgh1.

Previous reports suggest that, in addition to providing direct intracellular communication between coupled cells, connexins (either alone or as components of the Nexus (15) complex) might be involved in other processes, including growth control (16–18), migration (19), intercellular and cell-matrix adhesion (20), and signaling through intracellular space (20–22). There are also numerous studies reporting that manipulation of gap junction gene expression modifies the expression of other genes (23). There are at least two distinct mechanisms by which expression of other genes may be modified by expression of connexins, channel-dependent mechanisms and channel-independent mechanisms. Channel-dependent mechanisms that have been proposed include the direct transfer of genetic information, a concept that was attractive in the early days of gap junction research but turned out to be artifactual (24). However, this possibility has recently reappeared with the demonstration that RNA interference-sized oligonucleotides may pass from cell to cell (25) and by a report that gap junctions may permit passage of large peptides, permitting "cross-presentation" of antigens (26). Moreover, it has been proposed that intracellular communication provided by Cx43 can modulate gene transcription in osteocytes by differential recruitment of Sp1/Sp3 transcription factors to the osteocalcin promoter (27, 28). In this scenario, the effectors are believed to be gap junction-permeant second messenger molecules that may act through increasing or decreasing the phosphorylation state of transcription factors. Connexin-dependent but channel-independent mechanisms for the control of gene expression include the possibility that the carboxyl-terminal domain of Cx43 can translocate to the nucleus, acting there as a transcription factor (29), or that connexins might bind to molecules with both membrane or cytosolic localization and also with transcriptional activity; in this scenario, varying connexin abundance would lead to changes in the expression of other genes by shifting the equilibrium between the connexin-bound pool and a nonbound pool that could translocate to the nucleus. The first evidence for this mechanism was the study of Wnt-1 regulation of Cx43 expression and function in cardiac myocytes (30); beta-catenin was shown to interact directly with Cx43 and also to translocate to the nucleus when activated by Wnt-1 signaling. beta-catenin transactivated Cx43 transcription, providing a mechanism whereby Cx43 binding to another molecule, independent of its channel function, could regulate its own transcription. A similar mechanism has been recently proposed involving the growth regulator CCN3 (NOV) (31, 32), in which NOV is localized to Cx43 at the membrane. When Cx43 is down-regulated, NOV translocates to the nucleus producing a transformed phenotype with increased growth rate. Similarly, Penes et al. (33) have reported that the transcription factor ZO-1-associated nucleic acid-binding protein (ZONAB) is colocalized and coimmunoprecipitates with ZO-1 and connexins in astrocytes and oligodendrocytes, providing an additional substrate for the regulation of transcription factor availability by the abundance of connexins. Nuclear localization of members of the MAGUK family of proteins has been reported for both ZO-1 and CASK (calcium/calmodulin-dependent serine kinase) with nuclear translocation causing alterations of cell proliferation, activation of the TCF/LEF family of transcription factors and changes in the levels of cell cycle regulators such as Cyclin D and c-Myc (34). Our finding that, in response to cellular perturbation, Dlgh1 also translocates to the nucleus expands the repertoire of regulatory roles of MAGUK proteins. Regulation of this translocation may be one mechanism by which Cx32 regulates cell proliferation.

In summary, we describe here the novel interaction of Cx32 and the scaffolding protein Dlgh1 by a noncanonical interaction of the SH3/Hook domain of Dlgh1 and the Cx32 cytoplasmic loop and carboxyl tail. Further research is needed to determine the functions of this interaction and will likely yield interesting information on the role of connexins in macromolecular complexes.


   FOOTNOTES
 
* This work was funded, in part, by grants from National Institutes of Health (DK41918, to D. C. S.), the Marion Bessin Liver Center at the Albert Einstein College of Medicine (to H. S. D.), the United States Public Health Service Grant M072631, as well as the American Heart Association Grant 0560050Z (to P. L. S.) The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

1 To whom correspondence should be addressed: Dept. of Pharmacology, College of Physicians and Surgeons, Columbia University, 630 W. 168th St., New York, NY 10032. Tel.: 212-305-3411; Fax: 212-305-8780; E-mail: hsd2102{at}columbia.edu.

2 The abbreviations used are: Cx43, Connexin43; SH, Src homology; MAGUK, membrane-associated guanylate kinase; PBS, phosphate-buffered saline; HRP, horseradish peroxidase; DAPI, 4',6-diamidino-2-phenylindole; CT, carboxyl-terminal; NT, amino-terminal; CL, cytoplasmic loop. Back



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Ahmad, S., Martin, P. E., and Evans, W. H. (2001) Eur. J. Biochem. 268, 4544–4552[Medline] [Order article via Infotrieve]
  2. Ishidate, T., Matsumine, A., Toyoshima, K., and Akiyama, T. (2000) Oncogene 19, 365–372[CrossRef][Medline] [Order article via Infotrieve]
  3. Funke, L., Dakoji, S., and Bredt, D. S. (2005) Annu. Rev. Biochem. 74, 219–245[CrossRef][Medline] [Order article via Infotrieve]
  4. Dagli, M. L., Yamasaki, H., Kurtovskikh, V., and Omori, Y. (2004) Carcinogenesis 25, 483–492[Abstract/Free Full Text]
  5. Temme, A., Buchmann, A., Gabriel, H. D., Nelles, E., Schwarz, M., and Willecke, K. (1997) Curr. Biol. 7, 713–716[CrossRef][Medline] [Order article via Infotrieve]
  6. Duffy, H. S., Iacobas, I., Spray, D. C., and Ashton, A. W. (2005) in Methods in Cardiovascular Res., pp. 916–935, Springer Press, Berlin
  7. Schubert, A. L., Schubert, W., Spray, D. C., and Lisanti, M. P. (2002) Biochemistry 41, 5754–5764[CrossRef][Medline] [Order article via Infotrieve]
  8. Kojima, T., Sawada, N., Chiba, H., Kokai, Y., Yamamoto, M., Urban, M., Lee, G. H., Hertzberg, E. L., Mochizuki, Y., and Spray, D. C. (1999) Biochem. Biophys. Res. Commun. 266, 222–229[CrossRef][Medline] [Order article via Infotrieve]
  9. Nusrat, A., Chen, J. A., Foley, C. S., Liang, T. W., Tom, J., Cromwell, M., Quan, C., and Mrsny, R. J. (2000) J. Biol. Chem. 275, 29816–29822[Abstract/Free Full Text]
  10. Peracchia, C., Sotkis, A., Wang, X. G., Peracchia, L. L., and Persechini, A. (2000) J. Biol. Chem. 275, 26220–26224[Abstract/Free Full Text]
  11. Costes, S. V., Daelemans, D., Cho, E. H., Dobbin, Z., Pavlakis, G., and Lockett, S. (2004) Biophys. J. 86, 3993–4003
  12. Fishman, G. I., Gao, Y., Hertzberg, E. L., and Spray, D. C. (1995) Cell Adhes. Commun. 3, 353–365[Medline] [Order article via Infotrieve]
  13. Laird, D. W. (2006) Biochem. J. 394, 527–543[CrossRef][Medline] [Order article via Infotrieve]
  14. Lauf, U., Giepmans, B. N., Braconnot, S., Chen, S. C., and Falk, M. M. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 10446–10451[Abstract/Free Full Text]
  15. Duffy, H. S., Delmar, M., and Spray, D. C. (2002) J. Physiol. (Paris) 96, 243–249[CrossRef][Medline] [Order article via Infotrieve]
  16. Hirschi, K. K., Xu, C. E., Tsukamoto, T., and Sager, R. (1996) Cell Growth & Differ. 7, 861–870[Abstract]
  17. Naus, C. C. (2002) Can. J. Physiol. Pharmacol. 80, 136–141[CrossRef][Medline] [Order article via Infotrieve]
  18. Naus, C. C., Bond, S. L., Bechberger, J. F., and Rushlow, W. (2000) Brain Res. Brain. Res. Rev. 32, 259–266[CrossRef][Medline] [Order article via Infotrieve]
  19. Xu, X., Li, W. E., Huang, G. Y., Meyer, R., Chen, T., Luo, Y., Thomas, M. P., Radice, G. L., and Lo, C. W. (2001) Cell Commun. Adhes. 8, 321–324[Medline] [Order article via Infotrieve]
  20. Lin, J. H., Takano, T., Cotrina, M. L., Arcuino, G., Kang, J., Liu, S., Gao, Q., Jiang, L., Li, F., Lichtenberg-Frate, H., Haubrich, S., Willecke, K., Goldman, S. A., and Nedergaard, M. (2002) J. Neurosci. 22, 4302–4311[Abstract/Free Full Text]
  21. Suadicani, S. O., De Pina-Benabou, M. H., Urban-Maldonado, M., Spray, D. C., and Scemes, E. (2003) Glia 42, 160–171[CrossRef][Medline] [Order article via Infotrieve]
  22. Scemes, E., Suadicani, S. O., and Spray, D. C. (2000) J. Neurosci. 20, 1435–1445[Abstract/Free Full Text]
  23. Stains, J. P., and Civitelli, R. (2005) Birth Defects Res. C Embryo Today 75, 72–80[CrossRef][Medline] [Order article via Infotrieve]
  24. Pitts, J. D., Finbow, M. E., and Kam, E. (1988) Br. J. Cancer Suppl. 9, 52–57[Medline] [Order article via Infotrieve]
  25. Valiunas, V., Polosina, Y. Y., Miller, H., Potapova, I. A., Valiuniene, L., Doronin, S., Mathias, R. T., Robinson, R. B., Rosen, M. R., Cohen, I. S., and Brink, P. R. (2005) J. Physiol. 568, 459–468[Abstract/Free Full Text]
  26. Neijssen, J., Herberts, C., Drijfhout, J. W., Reits, E., Janssen, L., and Neefjes, J. (2005) Nature 434, 83–88[CrossRef][Medline] [Order article via Infotrieve]
  27. Stains, J. P., Lecanda, F., Screen, J., Towler, D. A., and Civitelli, R. (2003) J. Biol. Chem. 278, 24377–24387[Abstract/Free Full Text]
  28. Stains, J. P., and Civitelli, R. (2005) Mol. Biol. Cell. 16, 64–72[Abstract/Free Full Text]
  29. Dang, X., Doble, B. W., and Kardami, E. (2003) Mol. Cell. Biochem. 242, 35–38[CrossRef][Medline] [Order article via Infotrieve]
  30. Ai, Z., Fischer, A., Spray, D. C., Brown, A. M., and Fishman, G. I. (2000) J. Clin. Investig., 105, 161–171[Medline] [Order article via Infotrieve]
  31. Fu, C. T., Bechberger, J. F., Ozog, M. A., Perbal, B., and Naus, C. C. (2004) J. Biol. Chem. 279, 36943–36950[Abstract/Free Full Text]
  32. Gellhaus, A., Dong, X., Propson, S., Maass, K., Klein-Hitpass, L., Kibschull, M., Traub, O., Willecke, K., Perbal, B., Lye, S. J., and Winterhager, E. (2004) J. Biol. Chem. 279, 36931–36942[Abstract/Free Full Text]
  33. Penes, M. C., Li, X., and Nagy, J. I. (2005) Eur. J. Neurosci. 22, 404–418[CrossRef][Medline] [Order article via Infotrieve]
  34. Hsueh, Y. P. (2006) Curr. Med. Chem. 13, 1915–1927[CrossRef][Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 Am. J. Physiol. Endocrinol. Metab.Home page
G. Prost, F. Bernier-Valentin, Y. Munari-Silem, S. Selmi-Ruby, and B. Rousset
Connexin-32 acts as a downregulator of growth of thyroid gland
Am J Physiol Endocrinol Metab, February 1, 2008; 294(2): E291 - E299.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
282/13/9789    most recent
M605261200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Duffy, H. S.
Right arrow Articles by Spray, D. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Duffy, H. S.
Right arrow Articles by Spray, D. C.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2007 by the American Society for Biochemistry and Molecular Biology.