HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 1, 566-574, January 4, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/1/566    most recent M106968200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mohammad-Panah, R. Articles by Bear, C. E. Search for Related Content PubMed PubMed Citation Articles by Mohammad-Panah, R. Articles by Bear, C. E. Social Bookmarking                      What's this?

The Chloride Channel ClC-4 Co-localizes with Cystic Fibrosis Transmembrane Conductance Regulator and May Mediate Chloride Flux across the Apical Membrane of Intestinal Epithelia^*

Raha Mohammad-Panah, Cameron Ackerley, Johanna Rommens, Monideepa Choudhury, Yanchun Wang, and Christine E. Bear^§

From the Program in Structural Biology, Research Institute, Hospital for Sick Children, Toronto M5S 1G8, Canada

Received for publication, July 23, 2001, and in revised form, October 23, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cystic fibrosis (CF) causing mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) lead to mislocalization of CFTR protein from the brush border membrane of epithelial tissues and/or its dysfunction as a chloride channel. In initial reports, it was proposed that certain channels from the ClC family of chloride channels may provide compensatory or alternative pathways for epithelial chloride secretion in tissues from cystic fibrosis patients. In the present work, we provide the first evidence that ClC-4 protein is functionally expressed on the surface of the intestinal epithelium and hence, is appropriately localized to act as a therapeutic target in this CF-affected tissue. We show using confocal and electron microscopy that ClC-4 co-localizes with CFTR in the brush border membrane of the epithelium lining intestinal crypts in mouse and human tissues. In Caco-2 cells, a cell line thought to model human enterocytes, ClC-4 protein is expressed on the cell surface and also partially co-localizes with EEA1 and transferrin, marker molecules of early and recycling endosomes, respectively. Hence, like CFTR, ClC-4 may cycle between the plasma membrane and endosomal compartment. Furthermore, we show that ClC-4 functions as a chloride channel on the surface of these epithelial cells as antisense ClC-4 cDNA expression reduced the amplitude of endogenous chloride currents by 50%. These studies provide the first evidence that ClC-4 is endogenously expressed and may be functional in the brush border membrane of enterocytes and hence should be considered as a candidate channel to provide an alternative pathway for chloride secretion in the gastrointestinal tract of CF patients.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The disease cystic fibrosis (CF)¹ affects the epithelium lining multiple organs including the respiratory tract, the gastrointestinal tract, sweat ducts, and the reproductive organs (1). Normally, the protein product of the CF gene, the cystic fibrosis transmembrane conductance regulator (CFTR) resides on the apical surface of these epithelia. However, the most common disease causing mutation in CFTR (i.e. CFTRF508) promotes misfolding, leading to its mislocalization and degradation in intracellular compartments (2). Alternatively, other mutations lead to alterations in CFTR function as a phosphorylation and nucleotide-regulated anion channel (3). The anion channel function of CFTR is thought to provide the primary driving force for fluid transport and clearance of mucus and bacteria and lack of this function is the major cause for mucus obstruction in CF-affected organs. As proposed initially by Clarke et al. (4), non-CFTR chloride channels could compensate for lack of CFTR in certain tissues if they were appropriately localized on the apical, brush border membrane of the epithelium and could be opened under physiological conditions. In fact, we recently reported that increased basal chloride secretion correlates with amelioration of disease severity in intestinal tissues of a subpopulation of Cftr-deficient mice (5). Therefore, there is a compelling rationale for studying the expression of other chloride channels that may mediate chloride secretion across the gastrointestinal epithelium as these proteins may provide a strategic therapeutic target for treatment of cystic fibrosis.

There are nine mammalian members of the superfamily of voltage-gated ClC channels (6). ClC-1, ClC-2, ClCKa, and ClCKb are closely related, ClC-3, ClC-4, and ClC-5 form another arm of the family and finally, ClC-6 and ClC-7 comprise a distinct branch. Some of these family members exhibit quite a restricted tissue expression and hence are unlikely to contribute to chloride transport across the epithelium of the gastrointestinal tract, i.e. ClC-1 is only expressed in muscle tissue and ClCKa and ClCKb are expressed exclusively in the kidney (7-9). On the other hand, ClC-2, ClC-3, ClC-4, ClC-5, ClC-6, and ClC-7 are relatively widely distributed. However, most of these family members, with the exception of ClC-2 and possibly ClC-3, are thought to be primarily expressed on intracellular membranes (6, 10, 11).

We have recently reported that ClC-2 is endogenously expressed at a unique location in intestinal epithelia, in proximity to the tight junctions at the apical boundary between interacting differentiated enterocytes (12, 13). Furthermore, we showed using an antisense strategy that endogenous ClC-2 channels in intestinal epithelial cells can mediate chloride flux. Hence, we proposed that ClC-2 contributes to chloride secretion across certain epithelia. These findings were substantiated in a recent report describing Clc-2 knockout mice, in which the chloride current across the retinal pigment epithelium was decreased in the Clc-2 knockout animals (14). However, as the authors suggest, the impact of disrupting ClC-2 in intestinal tissue may only be apparent in Cftr-null animals.

At present, the localization and physiological role of ClC-3 remains controversial as some studies suggest that ClC-3 functions in intracellular vesicles (6, 15) and others support a role for ClC-3 channels on the plasma membrane (16, 17, 18). However, in our own immunolocalization studies of ClC-3 expression in the human intestinal cell line, Caco-2, we found that ClC-3 protein was predominantly expressed in intracellular vesicles (13).

Mutations in ClCN5 are associated with Dent's disease in humans, a kidney disease characterized by proteinuria and hypercalcuria (6, 7, 19, 20). Recently, examination of the native expression of ClC-5 in differentiated epithelial tissues revealed that this channel resides predominantly in intracellular membranes. In fact Clc-5 knockout mice predominantly show a defect in endocytosis of protein from lumen of the renal proximal tubule (11). Similarily, Vandewalle et al. (10) showed that ClC-5 protein localizes in intracellular compartments, i.e. endosomes and Golgi in the rat intestinal mucosa. Hence, in light of these immunohistochemical studies of native tissue, it seems unlikely that ClC-5 will contribute significantly to chloride currents across the apical membrane of the intestinal epithelium.

ClC-4 message is known to be expressed in brain, skeletal muscle, heart, kidney (21-23), and intestine (24), however, the subcellular distribution and function of ClC-4 protein in these tissues has yet to be determined. In our present work, we provide the first evidence that ClC-4 is endogenously expressed in the apical plasma membrane of murine and human enterocytes. Furthermore, we show using an antisense strategy, that endogenous ClC-4 protein mediates chloride currents across the plasma membrane of Caco-2 cells, cells that model human enterocytes. From the prospective of our current knowledge about the ClC family of chloride channels, these findings were unexpected. It was predicted that the subcellular distribution of ClC-4 would be similar to that of the related channel proteins, ClC-5 and ClC-3, and reside primarily in the membranes of intracellular organelles. To our knowledge, ClC-4 is the first chloride channel protein shown to co-localize with CFTR on the brush border membrane of intestinal crypt epithelial cells endogenously where it could mediate chloride currents into the gut lumen. Hence, these data strongly support a role for ClC-4 in intestinal chloride secretion and suggest that it may be capable of functionally complementing CFTR in vivo.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Culture and Transfection of Caco-2 Cells-- Caco-2 cells were obtained from the American Type Culture Collection (Manassas, VA). They were grown in Earl's -minimum essential medium (Wisent Inc., Montreal, Canada) containing 10% fetal calf serum, with 2 mM glutamine, 100 units of penicillin G, and 100 µg/ml streptomycin sulfate at 37 °C in an atmosphere of 5% CO₂. For patch clamp studies, cells were used 1-2 days after plating onto 35-mm coverslips (Fisher). For analysis of ClC-4 localization in fully differentiated epithelia, Caco-2 cells were seeded on semipermeable filters (Corning Costar) and grown for 4 days past confluency, conditions previously reported to induced a differentiated monolayer (25). Caco-2 cell cultures at 40% confluency (undifferentiated cells) were transfected with antisense ClC-4 cDNA or vector control using Lipofectin (Invitrogen) and the recommended transfection protocol was followed as previously described (13).

Tissues-- Tissues were obtained from adult male rats (Wistar) fed a standard diet. Fragments (0.5 cm long) of ileum were removed, rapidly frozen in liquid nitrogen, and kept at 80 °C until used. Portions of a small human bowel biopsy deemed normal by a pathologist were used for electron microscopic studies.

Immunoblotting-- Expression of ClC-4 protein in mouse tissues and Caco-2 cells was determined by immunoblotting as described in our previous papers (12, 13). In brief, tissues from mouse were homogenized and then centrifuged at 2,000 × g to pellet the nuclei. The supernatant was further centrifuged at 100,000 × g to yield a membrane pellet. Caco-2 cells were homogenized and centrifuged at 80,000 × g for 10 min at 4 °C to isolate a crude membrane preparation. 50 µg of each preparation was analyzed by SDS-polyacrylamide gel electrophoresis (8% gel) using anti-ClC-4 antibody at a concentration of 11.8 µg/ml. This polyclonal antibody was generated against a GST fusion peptide containing amino acids (1-52) of mouse ClC-4 (ClC-4 cDNA obtained from E. Rugarli, Milano, Italy). The antiserum was pre-adsorbed to a GST-coupled matrix to remove anti-GST antibodies. For the competition studies, the anti-ClC-4 antibody was preincubated with a 4.8-fold excess of the antigenic fusion peptide containing peptide from ClC-4, or a GST fusion protein containing residues 1-52 of hClC-5 overnight at 4 °C before incubation (hClC-5 cDNA kindly provided by T. Jentsch, Hamburg, Germany). Immunoreactive protein was detected using the ECL system (Amersham Bioscience, Inc.).

Northern Blot Analysis and RT-PCR Analysis-- Total RNA was isolated from Caco-2 cell monolayers by dissolution in guanidinium isothiocyanate and centrifugation through a CsCl cushion (26). Total RNA (5 µg) was analyzed on agarose gels (1%) containing 0.6 M formaldehyde and transferred to Hybond-N membranes (Amersham Bioscience Inc.). Blots were cross-linked with UV radiation and hybridized with mouse-specific ClC-4 cDNA fragments radiolabeled by random priming (27). Final conditions of washing included 0.2 × SSC (sodium chloride/sodium citrate) with 0.1% SDS at 60 °C. The blots were exposed to X-Omat film (Kodak) for 24-72 h at 0.1% SDS at 70 °C with on intensifying screen. Expression of ClC-4 was analyzed by reverse transcription-PCR, using the primers with following sequences: human ClC-4, sense (5'-TCCTCGATGAGCCGTTCCCTGATGT-3') and antisense (5'-AGGATGTACATTAAGTAATTCAGA-3'), producing a PCR product of 402 bp. The sequence identity of the PCR products was confirmed by BLAST sequence data base search.

Antisense ClC-4 Constructs-- The antisense murine ClC-4 was generated by cloning the ClC-4 open reading frame with BamHI (5') and EcoRI (3') into the eukaryotic vector pCDNA 3.1() (Promega, Madison, WI) such that the reversed restriction sites on this vector would reverse the orientation of the open reading frame to create the antisense plasmid. The murine ClC-4 sequence shares 87% sequence identity with the human sequence.

Immunofluorescence Staining-- Caco-2 cells were washed with PBS (phosphate-buffered saline), fixed with paraformaldehyde AM (4% in PBS), and permeabilized with 0.05% Triton X-100 in PBS. Blocking was done by using 5% normal goat serum in PBS for 1 h prior to primary antibody incubation. Primary antibodies were dissolved in blocking solution as follows. Mouse anti-endosomal early antigen 1 (1:200, Transduction Laboratories), mouse anti-giantin (1/1000, kind gift of H. P. Hauri, Basel, Switzerland), rabbit anti-ClC-3 (1/10, Alomone Labs Ltd., Jerusalem, Israel), rabbit anti-ClC-5 (1:150, kind gift of T. Jentsch, Hamburg, Germany), and rabbit anti-ClC-4 antibody (1/200). The cells were then incubated for 2.5 h at room temperature in primary antibody and then rinsed in PBS and incubated with Cy3-conjugated or fluorescein isothiocyanate-conjugated anti-rabbit or anti-mouse secondary antibodies (1:1000, Molecular Probes) and washed again before mounting. For double labeling experiments, two primary antibodies developed in different species were applied together, followed by the simultaneous detection using Cy3 and fluorescein isothiocyanate-coupled secondary antibodies. In some experiments we loaded Caco-2 cells for 1 h at 37 °C with 50 µg/ml transferrin conjugated to iron and tetramethylrhodamine (transferrin-Fe²⁺-Rhd, Molecular Probes) diluted in Earl's -minimum essential medium without serum. Following fixation, cells we labeled with the ClC-4 antibody. For immunofluorescence labeling of rat tissues, cryosections (5 µm thick) of intestine were fixed in ice-cold methanol for 10 min prior to blocking. Samples were then processed for fluorescence as above. Slides were viewed with a ×63 objective on a Carl Zeiss LSM 510 equipped with an Axiovert 100 confocal microscopy.

Electron Microscopy-- Portions of small bowel from a normal human biopsy, ileum from mice, and Caco-2 cells grown on semi-permeable filters were collected and prepared for freeze substitution and Lowicryl HM20 embedding (28). The sections were incubated with the ClC-4 antibody diluted 1:100 in PBS, 0.5% bovine serum albumin. Sections were then labeled as previously described (12). Controls included the omission of either the primary or secondary antibody or a peptide competition. In double labeling experiments, sections of human small bowel were first labeled with the ClC-4 antibody and then with a monoclonal antibody against the R-domain of CFTR (1/100, R&D, Minneapolis, MN). Following ClC-4 labeling the sections were incubated for an hour in the CFTR antibody diluted 1:10 with PBS/bovine serum albumin. The sections were then labeled with goat anti-mouse IgG 5 nm complex, washed thoroughly with distilled water, and stained with uranyl acetate and lead citrate prior to examination in a JEOL JEM 1230 transmission electron microscope (JEOL USA Inc., Peabody, MA). Images were acquired with an AMT CCD digital camera (AMT Corp., Danvers, MA).

Intranuclear Injection of Plasmid-- Caco-2 cells were microinjected with plasmids at day 1 after plating on glass coverslips for patch clamp experiments as described (13). Plasmids were diluted to a final concentration of 300 µg/ml for antisense ClC-4 and antisense ClC-2. Fluorescein isothiocyanate-labeled dextran (0.5%, Sigma) was also added to the injection medium to identify successfully microinjected cells.

Patch Clamp Studies-- Caco-2 cell membrane currents were measured using conventional whole cell patch clamp technique as described (13). Data were collected and analyzed with an Axopatch-200A amplifier and pCLAMP software (Axon Instruments, Foster City, CA). The bath solution contained (in mM): 140 N-methyl-D-glutamine chloride, 2 MgCl₂, 2 CaCl₂, 5 HEPES, while the pipette solution contained (in mM) 140 N-methyl-D-glutamine chloride, 2 MgCl₂, 2 EGTA, and 5 HEPES. Both pipette and bath solutions were adjusted to pH 7.4. The tip resistance was 3-5 M when filled with the pipette solution.

Statistics-- Patch clamp measurements are presented as the mean ± S.E. Statistical analyses were performed using the Student's t test. Probabilities (p) of 0.05 or less were considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ClC-4 Message and Protein Expression in Intestinal Tissue and the Human Intestinal Cell Line, Caco-2-- To assess ClC-4 protein expression, we raised a polyclonal antibody against a GST fusion protein containing residues 1-52 of the amino terminus of mouse ClC-4. As expected, given the abundant ClC-4 message expression in the rodent brain (22), our antibody recognized a broad, prominent signal in immunoblots of mouse brain lysate (Fig. 1i, A). The predominant band corresponds to a protein of ~90 kDa in molecular mass, close to the ClC-4 mass predicted from the primary sequence of 83 kDa. This signal is specific for ClC-4 as it is competed by the fusion peptide of the amino terminus of ClC-4 (residues 1-52) but not a fusion protein containing the amino terminus (residues 1-52) of the closely related channel protein, ClC-5 (Fig. 1i, B). Using this polyclonal antibody, we detected a prominent 97-kDa band in immunoblots of mouse ileum suggesting that ClC-4 protein is also endogenously expressed in intestinal tissue (Fig. 1ii).

View larger version (29K): [in this window] [in a new window]   Fig. 1.   ClC-4 message and protein expression in rodent brain and small intestine as well as the human intestinal epithelial cell line, Caco-2. i: A, anti-ClC-4 immunoblot of murine brain tissue. B, competition of ClC-4 band with a 4.8-fold excess of ClC-4 fusion protein (middle lane), but not with ClC-5 fusion protein (left lane). ii, anti-ClC-4 immunoblot of murine small intestine tissue. iii: A, Northern blot analysis showing that the murine ClC-4 cDNA probe recognizes a band of ~4.7 kb in Caco-2 cells. B, RT-PCR, with human-specific ClC-4 primers spanning 402 bp, confirms ClC-4 mRNA expression in the Caco-2 cells (+). No ClC-4 was detected in non-reverse transcribed Caco-2 cells RNAs () or by omitting cDNA (H2O). C, anti-ClC-4 immunoblot in Caco-2 cells.

ClC-4 mRNA and protein could also be detected in the human intestinal epithelial cell line, Caco-2. ClC-4 mRNA in Caco-2 cells was detected as a 4.7-kb transcript by Northern analysis, using a mouse ClC-4 specific probe (Fig. 1iii, A). A smaller, as yet unidentified, transcript of less than 1.9 kb was also detected in Caco-2 cells. ClC-4 mRNA expression in Caco-2 cells was confirmed using RT-PCR analysis with sequence specific primers to human ClC-4 (Fig. 1iii, B). ClC-4 protein expression in Caco-2 cells was detected by immunoblotting using the polyclonal antibody described above. As in the immunoblots of mouse intestine, ClC-4 protein was detected as a predominant 97-kDa protein in Caco-2 cells, confirming that ClC-4 is also expressed in human intestinal epithelial cells (Fig. 1iii, C).

Comparison of ClC-4 and ClC-5 Protein Distribution in the Intestinal Mucosa-- The distribution of ClC-5 protein in rat intestinal tissue has been previously described by Vandewalle et al. (10) using a ClC-5-specific polyclonal antibody originally characterized by Gunther et al. (29). We also observed that ClC-5 specific staining was primarily perinuclear in cross-sections of intestinal crypts, showing a Golgi-like pattern of distribution (Fig. 2i, A). ClC-5 specific signal also appeared to be localized close to the luminal membrane of the intestinal crypt. It was previously reported that this juxtaluminal membrane distribution of ClC-5 may be localized in early endosomes (10). Therefore, we processed cryosections of rat ileum for double immunofluorescence labeling using the polyclonal ClC-5 antibody and a monoclonal antibody against EEA1, one of the most specific early endosomal markers (30) (Fig. 2i, A, middle, green). The immunolabeled ClC-5 protein appeared to co-localize with immunolabeled EEA1 protein in a distinct ring around the apical membrane (Fig. 2i, A, right, yellow). These results provide a benchmark for our comparative studies of ClC-4 distribution in this tissue.

View larger version (88K): [in this window] [in a new window]   Fig. 2.   i, confocal images of ClC-4 distribution in the intestinal mucosa. A, cryosections of rat ileum, double-labeled with anti-ClC-5 and anti-EEA1 antibodies. Merge, overlay of the two images. The circle identifies the boundary of a single intestinal crypt. B, cryosections of rat ileum, labeled with anti-ClC-4 antibody. Differential interference contrast (DIC) is used to define structure of the crypt lumen. C, cryosections of rat ileum, double-labeled with anti-ClC-4 and anti-EEA1 antibodies. D, double labeling of rat ileum cryosections with anti-ClC-4 and anti-giantin antibodies. Immunostaining was absent when the primary antibody was omitted (left). Bar, 20 µm. ii, electron micrographs depicting localization of immunogold labeled ClC-4 in brush border of intestinal epithelia (A) ClC-4 immunogold labeling of an absorptive cell from the crypt of the mouse ileum. B, ClC-4 labeling (arrows) in an adjacent goblet cell. C, double immunogold labeling of a section of human small bowel with gold-labeled ClC-4 antibody (large gold grains, arrows) and CFTR antibody (small grains). D, ClC-4 labeling in the apical cytoplasm and membrane of goblet cells in human tissue (arrows). Bar, 100 nm for A and C and 500 nm for B and D. iii, immunogold labeling of ClC-4 is specific. All four images show effective competition of ClC-4 gold labeling with a 4.8-fold excess of of antigenic ClC-4 fusion protein. Bar, 500 nm for A-D.

Overall, the pattern observed for ClC-4-specific staining of the intestinal mucosa is different from that observed for ClC-5. Cross-sections of intestinal crypts were most intensely labeled around the luminal membrane of the gland suggesting the presence of ClC-4 in the apical membrane and/or subapical membrane compartment (Fig. 2i, B, left, red). The apical or luminal membrane of the crypt can be readily discerned by differential interference contrast microscopy (Fig. 2i, B, middle and right). To determine if ClC-4 protein is located in early endosomes, we performed a double labeling experiment, wherein, intestinal sections were labeled with both our ClC-4 antibody and the EEA1 monoclonal antibody described above (Fig. 2i, C, middle, green). These images show that the two patterns partially overlap (Fig. 2i, C, right, yellow), suggesting that ClC-4 protein, like ClC-5 is (at least transiently) localized in early endosomes. Furthermore, double labeling of intestinal crypts with ClC-4 antibody (red) and a Golgi marker (giantin, green) clearly shows the lack of overlap and the definite apical polarization of ClC-4 (Fig. 2i, D).

Our immunofluorescence studies on rat intestinal tissue suggest that ClC-4 is localized in proximity to the apical membrane of intestinal epithelial cells. The confocal images do not have the resolution necessary to determine whether ClC-4 protein is inserted in the apical plasma membrane. Therefore, we examined ultra-thin sections of freeze-substituted mouse and human intestinal tissues labeled with ClC-4 antibody (described above) by electron microscopy. Fig. 2ii, A, shows that in mouse tissue, immunogold-labeled ClC-4 is localized primarily along the brush border membrane and the apical cytoplasm of an absorptive cell from the crypt of the ileum. In adjacent goblet cells, ClC-4 labeling was found predominantly in clusters in the apical cytoplasm (Fig. 2ii, B). Double labeling of a section of human small bowel with gold-labeled ClC-4 antibody (large grains, arrows) and CFTR antibody (small gold grains) shows that both of these proteins co-localize primarily on the brush border membrane of human enterocytes (Fig. 2ii, C). ClC-4 labeling intensity was also detected in the apical cytoplasm and membrane of goblet cells in human tissue (Fig. 2ii, D). The ClC-4 labeling pattern in mouse enterocytes (Fig. 2iii, A) and goblet cells (Fig. 2iii, B) as well as in human enterocytes (Figs. 2iii, C) and goblet cells (Fig. 2iii, D) was effectively competed with the ClC-4 fusion protein against which the ClC-4 antibody was raised, supporting its specificity. Clearly, as ClC-4 was found in the apical membrane of the absorptive cells in close proximity to CFTR, it was next necessary to determine whether ClC-4 is functional as a chloride channel in the plasma membrane of enterocytes.

ClC-4 Is Functionally Expressed on the Plasma Membrane of Caco-2 Cells-- We suggest that the Caco-2 cell line is an effective model for analysis of ClC-4 function in intestinal epithelia, as ClC-4 localization in confluent Caco-2 monolayers is identical to that observed in the native epithelia. Fig. 3i, A, shows an electron micrograph of an immunogold-labeled section of Caco-2 cells grown for 4 days after confluency on a semi-permeable filter labeled using ClC-4 antibody. Consistent with our previous electron micrographs obtained from native intestinal tissues, we found electron-dense grains corresponding to ClC-4 localized primarily to intact and sloughed brush border membranes. This labeling pattern was effectively competed with the ClC-4 fusion protein against which the ClC-4 antibody was raised testifying to its specificity (Fig. 3i, B). Fig. 3ii, A, shows representative confocal microscopy image in the XZ plane of ClC-4 localization in confluent Caco-2 monolayers grown on a semipermeable filter labeled with ClC-4 antibody. This image indicates that the ClC-4 signal is limited to the apical membranes of the polarized cells.

View larger version (52K): [in this window] [in a new window]   Fig. 3.   Cellular distribution of ClC-4 in Caco-2 cells. i, A, the apical membrane of differentiated Caco-2 cells was labeled intensely by immunogold. B, competition of ClC-4 gold labeling with a 4.8-fold excess of ClC-4 fusion protein. Bar, 500 nm for both A and B. ii: A, immunolocalization of ClC-4 on apical membrane of Caco-2 cells grown 4 days post-confluency on semipermeable filter. Confocal image in the XZ plane of cells shows predominant apical staining. The scale bar represents 10 µm. B, immunofluorescence labeling of ClC-4 in Caco-2 cells by confocal microscopy (left). Immunostaining was absent when the anti-ClC-4 antibody was omitted (right). C, partial co-localization of the intracellular ClC-4 signal with the EEA1 distribution in Caco-2 cells. D, partial overlapping of transferrin-Fe2+-Rhd and ClC-4 distributions in transferrin-Fe2+-Rhd-loaded Caco-2 cells. Bar, 10 µm.

In subconfluent, non-polarized Caco-2 cells, our ClC-4 polyclonal antibody labeled both the cell surface and intracellular vesicles (Fig. 3ii, B). The intracellular ClC-4-specific signal appeared punctate and perinuclear. Intracellular ClC-4 specific label partially overlapped with the pattern associated with EEA1 staining (Fig. 3ii, C). These observations suggest that ClC-4 may be cycled between the plasma membrane and the endosomal compartment. To test this hypothesis, we incubated Caco-2 cells with transferrin conjugated to iron and tetramethylrhodamine (transferrin-Fe²⁺-rhodamine). This marker has been previously shown to recycle from the plasma membrane to the endosomal compartment following its interaction with the transferrin receptor (31, 32). Fig. 3ii, D, shows that the pattern of ClC-4 protein distribution partially overlaps (Fig. 3ii, D, right, yellow) with the intracellular distribution of rhodamine transferrin (Fig. 3ii, D, left, red), as it recycles through the endocytic recycling compartments.

To further assess the specificity of our ClC-4 antibody and the subcellular distribution pattern detected using this antibody, we examined the effects of antisense ClC-4 cDNA (aClC-4) expression on the pattern detected by immunofluorescence and confocal microscopy of Caco-2 cells. DNA coding for green fluorescence protein (GFP) was co-transfected with aClC-4 or empty vector as a control into Caco-2 cells to identify transfected cells (Fig. 4i, A). We used the Scion imaging program (Scion Corp., Frederick, MD) to compare the ClC-4 immunofluorescence intensity in aClC-4 and in vector-transfected Caco-2 cells (Fig. 4i, B). We found that the fluorescence intensity of the signal (red) corresponding to the expression of ClC-4 was reduced by ~65.2% in aClC-4 transfected Caco-2 cells (47.42 units ± 2.6, n = 25, p < 0.0001) relative to the intensity of the ClC-4 signal in mock (vector alone) transfected cells (136.4 units ± 3.2, n = 33). We then evaluated the specificity of the effects of aClC-4 by assessing its effect on ClC-5 and ClC-3 proteins, as these channel-forming proteins share 77 and 79% sequence identity with ClC-4, respectively (Fig. 4ii). In cells transfected with empty vector alone, immunoreactive ClC-5 was primarily expressed in intracellular membranes (Fig. 4ii, A), consistent with our previous studies in native intestinal tissue and Caco-2 cells and those reported by Vandewalle et al. (10). In contrast to the effect of aClC-4 expression on ClC-4 protein (65%), we found that there was only a minor effect on ClC-5 protein expression (6.8%, 123 units ± 4.14, n = 32, versus 132 units ± 1.48, n = 37 in vector-transfected cells, p = 0.03). Immunofluorescence corresponding to expression of ClC-3, a related family member, was not affected by aClC-4 or vector transfection (Fig. 4ii, B). Immunoreactive ClC-3 was primarily expressed in intracellular membranes although there was a signal detected at the cell surface in a subpopulation of cells (Fig. 4ii, B), consistent with our previous studies in Caco-2 cells (13) and those reported by Shimada et al. (17). Fig. 4ii, B, shows that the ClC-3 immunofluorescence (red) in aClC-4 transfected Caco-2 cells (129.3 units ± 6.4, n = 22) were similar to that in vector and GFP co-transfected cells (128.7 units ± 6.08, n = 23, p = 0.95). Therefore, these data show that our antibody recognizes ClC-4 protein specifically in immunofluorescence studies of these intestinal epithelial cells.

View larger version (45K): [in this window] [in a new window]   Fig. 4.   ClC-4 antisense reduces ClC-4 protein expression in Caco-2 cells. i: A, ClC-4 immunostaining in Caco-2 cells co-transfected with vector (Vtr) and GFP (upper panel) or with antisense ClC-4 (aClC-4) and GFP (lower panel). B, bar graph showing the reduction of ClC-4 signal in aClC-4-transfected cells. ii: A, ClC-5 signal in Caco-2 cells co-transfected with vector and GFP (upper panel) or aClC-4 and GFP (lower panel). Bar graph showing a minor but significant decrease of ClC-5 protein aClC-4 transfected cells. B, ClC-3 signal in vector and GFP (upper panel) or aClC-4 and GFP (lower panel) co-transfected cells. Bar graph showing that transfection of aClC-4 does not significantly alter ClC-3 protein expression. All bar graphs show the mean ± S.E. of fluorescence intensities. Bar, 10 µm. Fluorescence intensities were quantified using two methods and both techniques revealed similar antisense ClC-4 effects. Pixel intensity of the grayscale image (0 units = white, 255 units = black) was averaged for the whole cell or for the perinuclear region (by superpositioning a box at 8 locations around the nucleus).

Next, we used the above ClC-4 cDNA antisense strategy to determine if ClC-4 functions endogenously as a chloride channel on the surface membrane of Caco-2 cells. Previous studies performed in either Xenopus oocytes or in HEK 293 fibroblasts (15) showed that overexpression of ClC-4 conferred chloride currents which were activated by depolarization. Therefore, we functionally isolated anion-dependent currents by using intracellular (pipette) and extracellular (bath) solutions which contained N-methyl-D-glutamine-Cl as the predominant salt and chose a voltage step protocol which would include depolarizing voltage steps (15). Briefly, from a holding potential of 30 mV, the membrane potential was stepped by 20 mV increments from 100 to + 100 mV. As shown in Fig. 5, A and B, currents typical of those previously associated with ClC-4 expression, i.e. showing activation with depolarizing voltage steps and an outward-rectifying current-voltage relationship, were detected in Caco-2 cells. At the depolarized membrane potential of +100 mV, these currents had a magnitude of 14 pA/pF ± 1.5 (n = 6). These depolarization activated currents reversed close to the estimated equilibrium potential of chloride (E_Cl = 0 in symmetrical N-methyl-D-glutamine-Cl solutions (Fig. 5B).

View larger version (39K): [in this window] [in a new window]   Fig. 5.   Functional expression of ClC-4 in Caco-2 cells. i: A, depolarization activated chloride currents are significantly reduced in Caco-2 cells after nuclear injection of our aClC-4 construct (300 µg/ml). B, mean current-voltage (I-V) curves of chloride currents obtained from control Caco-2 cells (n = 6, Ctl, squares) and from cells injected with 300 µg/ml aClC-4 (n = 11, circle). Currents were normalized to cell capacitance (pF). C, mean I/V curves of depolarization-activated chloride currents in antisense ClC-2 injected (n = 8, aClC-2, circle) and in non-injected (n = 6, Ctl, squares) Caco-2 cells. ii: A, ClC-4 immunostaining is reduced in Caco-2 cells after nuclear injection with antisense ClC-4 (aClC-4) and GFP cDNA relative to non-injected cells. The bar graph shows the reduction of ClC-4 signal in aClC-4 injected cells (black) versus non-injected cells (gray) (p < 0.0001). B, ClC-5 signal in Caco-2 cells after nuclear injection with aClC-4 and GFP cDNA. Bar graph showing a minor but significant decrease of ClC-5 protein in aClC-4-injected cells (black) versus non-injected cells (gray) (p = 0.04). C, ClC-3 signal in aClC-4 and GFP co-injected cells. Bar graph shows that transfection of aClC-4 does not significantly alter ClC-3 protein expression (p = 0.8). All bar graphs show the mean ± S.E. of fluorescence intensities. Bar, 10 µm. Fluorescence intensities were quantitated as described in the legend to Fig. 4.

We used an antisense strategy to confirm that the above currents were mediated by ClC-4 since this channel is not inhibited by any of the typical blockers of chloride channels (15). For the relatively labor intensive patch clamp studies, we manipulated ClC-4 expression using the intranuclear plasmid injection technique (13) as this method permits control of plasmid copy number and hence limits cell to cell variation in antisense expression. Fluorescein isothiocyanate-dextran was co-injected with the plasmid to permit identification of manipulated cells. We found that microinjection of aClC-4 into Caco-2 cells decreased the endogenous outward rectifier currents (Fig. 5i, A-B). The positive whole cell current measured at + 100 mV decreased from 14 pA/pF ± 1.5 (n = 6) in uninjected cells to 7 ± 0.8 pA/pF (n = 11, p = 0.0004) in 300 µg/ml aClC-4-injected cells, respectively.

We then evaluated the effectiveness and specificity of the nuclear injection technique of antisense ClC-4 expression by assessing ClC-4, ClC-5, and ClC-3 protein quantity by immunofluorescence. Intranuclear injection of antisense ClC-4 effectively decreased ClC-4 protein expression. The fluorescence intensity of the signal (red) corresponding to the expression of ClC-4 (Fig. 5ii, A) was reduced by ~73% in aClC-4 transfected Caco-2 cells (36.2 units ± 1.9, n = 27, p < 0.0001) relative to the intensity of the ClC-4 signal in non-injected cells (133.6 units ± 2.4, n = 27). We then evaluated the specificity of the effects of aClC-4 by assessing its effect on ClC-5 (Fig. 5ii, B) and ClC-3 proteins (Fig. 5ii, C). As in lipid-based transfection method, we found only a minor effect of aClC-4 on ClC-5 protein expression (7.7%, 119 units ± 3.4, n = 34, versus 128.9 units ± 3.2, n = 34 in non-injected cells, p = 0.04, Fig. 5ii, B) and no effect of aClC-4 on ClC-3 protein expression (122.8 units ± 8.7, n = 14, versus 125.9 units ± 9.2, n = 14 in non-injected cells, p = 0.8, Fig. 5ii, C). Therefore, these data show the specificity of the antisense strategy using the microinjection method. In addition, inhibition of outward-rectifying chloride currents by ClC-4 antisense expression was a specific response, as nuclear injection of antisense ClC-2, a distinct member of the ClC chloride channel family (6), did not affect these endogenous currents (Fig. 5i, C). The currents measured in antisense ClC-2-injected Caco-2 cells (14.6 ± 0.6 pA/pF at + 100 mV, n = 8) were not significantly different from that measured in non-injected cells (14 ± 1.5 pA/pF, p = 0.8). Together, these results indicate that ClC-4 natively expressed in Caco-2 cells contributes to the endogenous outwardly rectifying currents recorded in these cells.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Jentsch and colleagues (6) suggested in a recent research article that the mammalian ClC channel forming proteins ClC-3 through ClC-7 function predominantly in intracellular compartments. His own elegant studies of ClC-3, ClC-5 and ClC-7 knockout mice support this statement. The primary defect associated with disruption of murine ClC-3 relates to synaptic vesicle function and defective internalization of neurotransmitters by these vesicles in Clc3(/) mice (33). The primary defect in Clc-5 knockout mice targets endocytic vesicles and leads to deficient uptake by endosomes of low molecular weight proteins from the lumen of the proximal tubule of the kidney (11, 34). This phenotype in mice mirrors the clinical phenotype of proteinuria in patients bearing mutations in ClCN5 (20). Finally, disruption of murine Clc-7 primarily affects acid secretion across the membrane of the osteoclast which interfaces with bone, i.e. the "ruffled" membrane, and consequently impairs bone resorption (35). The ruffled membrane of the osteoclasts is similar to lysosomal membranes with respect to its biochemical properties. Overall, these studies show convincingly, at least in the tissues studied, that these members of the ClC family function in intracellular organelles. Given its sequence similarity with ClC-3 and ClC-5, it was predicted that ClC-4 would also function in intracellular compartments.

The present report provides the first description of the subcellular distribution and function of the ClC-4 channel forming protein. In contrast to the above prediction, our evidence suggests that at least in the intestinal epithelium of rats, mice, and humans, ClC-4 is expressed primarily on the apical brush border membrane where it functions to modify extracellular chloride concentrations. We showed by immunofluorescence labeling and confocal microscopy that ClC-4 is primarily localized in proximity to the apical, brush border membrane of intestinal enterocytes and goblet cells. This ClC-4 specific staining pattern partially overlaps with that of EEA1, suggesting that ClC-4 also exists in apical early endosomes. Significantly, while immunogold-labeled ClC-4 can be detected in the subapical cytoplasm (presumably in endocytic vesicles), the label is primarily localized in the apical plasma membrane of these cells. Furthermore, we confirmed that ClC-4 is functionally expressed in the plasma membrane of intestinal epithelial cells using an antisense strategy combined with patch clamp electrophysiology. As ClC-4 antisense expression specifically reduced endogenous ClC-4 protein expression in Caco-2 intestinal cells, and this decrease in protein correlated with a significant decrease in an endogenous voltage-activated chloride conductance, we argue that ClC-4 mediates this native current.

As yet, we have not yet identified the signals that target ClC-4 to the brush border of the intestinal epithelium. Although ClC-4 shares ~80% identity with ClC-5 at the sequence level, the subcellular distributions of ClC-4 and ClC-5 are quite distinct in intestinal epithelial tissue, as apparent in Fig. 2i. As previously mentioned, ClC-4 protein exhibits apical polarization and concentration in apical endosomes and brush border membrane. Immunolabeled ClC-5 is diffusely distributed throughout the entire cell interior and concentrated in Golgi and endosomal (EEA1-stained) vesicles. Schwake et al. (36) identified a PY motif on the carboxyl terminus of ClC-5 (PPLPPY) which is absent in ClC-4 (PELPAN). These authors showed that a PY motif in ClC-5 may be involved in internalization of ClC-5 channels from the cell surface via an interaction with the ubiquitin ligase, NEDD4 (36). NEDD4 was previously shown to regulate the surface expression of the epithelial sodium channel, ENaC through this mechanism (37, 38). Hence, this PY motif may account for the relative absence of ClC-5 from the cell surface of intestinal epithelial cells. However, there are likely to be other, as yet unidentified sequences within ClC-4 that promote a specific apical polarization of this protein in intestinal epithelia.

We consider it likely that ClC-4 participates in chloride secretion by the intestine, as it shares the same tissue distribution pattern as CFTR, the chloride channel thought to be primarily responsible for this function in intestine. Interestingly, Clc4 was previously identified as a candidate modifier gene for intestinal disease severity in Cftr(/) mice (39). In the laboratory mouse strain C57BL/6, Clc4 is located on chromosome 7 in close proximity to a "modifier" locus shown to correlate with a "milder" intestinal phenotype in Cftr(/) mice. However, in another mouse strain; Mus spretus, Clc4 is located on the X chromosome, providing an exception to Ohno's hypothesis regarding conservation of linkage to the X chromosome (22). Interestingly, ClCN4 is also located on human chromosome X, hence, ClC-4 is no longer considered a candidate for intestinal disease modification for cystic fibrosis (23). However, it does remain formally possible that activation of ClC-4 function as a chloride channel on the brush border membrane may partially compensate for the absence of CFTR function in the intestine.

Presently, our understanding of the regulation of ClC-4 channel activity is quite rudimentary, so it is difficult to predict its relative contribution to net transepithelial chloride secretion under resting and/or hormone stimulated conditions. Clc4-null males, generated from backcrossing the progeny of C57BL/6 and M. spretus matings with congenic M. spretus mice exhibit no gross phenotype, arguing that other chloride channels, probably CFTR, may perform a similar function in the intestine. Although the primary sequence of ClC-4 possesses consensus sites for phosphorylation by PKC and PKA, as yet, there is no in vivo evidence for a role for regulation of channel function by phosphorylation. Extracellular acidification has been shown to exert a minor inhibitory effect on both ClC-5 and ClC-4 channel function and whereas extracellular alkalinization exerts a minor stimulatory effect on ClC-5 activity, there is no such stimulatory effect of alkaline pH on ClC-4. The only known agonist of ClC-4 is membrane depolarization. However, as for the related protein ClC-5, depolarization-dependent activation of ClC-4 at +20 to +40 mV seems improbable as a physiological trigger. Jentsch and his colleagues (15) have postulated for ClC-5 that there might be neighboring proteins which could interact with these depolarization-activated chloride channels (i.e. ClC-4 and ClC-5) and modify the regulation of ClC-4 and ClC-5 channel function. Hypothetically, CFTR could provide such a regulatory partner as it is localized on the brush border of enterocyte close to ClC-4 and has been implicated in the modification of several other chloride channels (40-42). Future studies will focus on understanding the regulatory properties of ClC-4 and identifying those accessory proteins which are important for its function in situ.

Finally, we have provided evidence to suggest that ClC-4 may provide a pathway, operating in parallel with CFTR, to mediate the flux of chloride ion across the brush border membrane of the epithelium lining intestinal crypts in rodents and humans. Therefore, we suggest that ClC-4 protein may provide a therapeutic target for treatment of intestinal disease in cystic fibrosis.

	ACKNOWLEDGEMENTS

We are grateful to Professor T. Jentsch (Hamburg, Germany) for his kind gift of ClC-5 polyclonal antibody, Professor Elena Rugarli for her generous gift of Clcn4 cDNA, and Dr. H. P. Hauri for his kind gift of giantin monoclonal antibody (Switzerland). We thank Afshandoht Amini for excellent technical assistance.

	FOOTNOTES

^* This work was supported in part by operating grants awarded by the Canadian Institutes of Health and the National Institutes of Health (to C. E. B.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Supported by a fellowship from The Canadian Cystic Fibrosis Foundation.

^§ To whom correspondence should be addressed: Program in Structural Biology, Research Institute, Hospital for Sick Children, 555 University Ave., Toronto, Canada. Tel.: 416-813-5981; Fax: 416-813-5028; E-mail: bear@sickkids.on.ca.

Published, JBC Papers in Press, October 23, 2001, DOI 10.1074/jbc.M106968200

	ABBREVIATIONS

The abbreviations used are: CF, cystric fibrosis; CFTR, cystic fibrosis transmembrane conductance regulator; GST, glutathione S-transferase; RT, reverse transcriptase; PBS, phosphate-buffered saline; GFP, green fluorescent protein.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				N. S. Joo, J. J. Wine, and A. W. Cuthbert Lubiprostone stimulates secretion from tracheal submucosal glands of sheep, pigs, and humans Am J Physiol Lung Cell Mol Physiol, May 1, 2009; 296(5): L811 - L824. [Abstract] [Full Text] [PDF]

				R. Mohammad-Panah, L. Wellhauser, B. E. Steinberg, Y. Wang, L. J. Huan, X.-D. Liu, and C. E. Bear An essential role for ClC-4 in transferrin receptor function revealed in studies of fibroblasts derived from Clcn4-null mice J. Cell Sci., April 15, 2009; 122(8): 1229 - 1237. [Abstract] [Full Text] [PDF]

				H. Okkenhaug, K.-H. Weylandt, D. Carmena, D. J. Wells, C. F. Higgins, and A. Sardini The human ClC-4 protein, a member of the CLC chloride channel/transporter family, is localized to the endoplasmic reticulum by its N-terminus FASEB J, November 1, 2006; 20(13): 2390 - 2392. [Abstract] [Full Text] [PDF]

				L. Huang, J. Cao, H. Wang, L. A. Vo, and J. G. Brand Identification and Functional Characterization of a Voltage-gated Chloride Channel and Its Novel Splice Variant in Taste Bud Cells J. Biol. Chem., October 28, 2005; 280(43): 36150 - 36157. [Abstract] [Full Text] [PDF]

				S. Spiegel, M. Phillipper, H. Rossmann, B. Riederer, M. Gregor, and U. Seidler Independence of apical Cl-/HCO3- exchange and anion conductance in duodenal HCO3- secretion Am J Physiol Gastrointest Liver Physiol, November 1, 2003; 285(5): G887 - G897. [Abstract] [Full Text] [PDF]

				R. Mohammad-Panah, R. Harrison, S. Dhani, C. Ackerley, L.-J. Huan, Y. Wang, and C. E. Bear The Chloride Channel ClC-4 Contributes to Endosomal Acidification and Trafficking J. Biol. Chem., August 1, 2003; 278(31): 29267 - 29277. [Abstract] [Full Text] [PDF]

				Y. Li, S. T. Halm, and D. R. Halm Secretory activation of basolateral membrane Cl- channels in guinea pig distal colonic crypts Am J Physiol Cell Physiol, April 1, 2003; 284(4): C918 - C933. [Abstract] [Full Text] [PDF]

				M. Gentzsch, L. Cui, A. Mengos, X.-b. Chang, J.-H. Chen, and J. R. Riordan The PDZ-binding Chloride Channel ClC-3B Localizes to the Golgi and Associates with Cystic Fibrosis Transmembrane Conductance Regulator-interacting PDZ Proteins J. Biol. Chem., February 14, 2003; 278(8): 6440 - 6449. [Abstract] [Full Text] [PDF]

				C. Isnard-Bagnis, N. Da Silva, V. Beaulieu, A. S. L. Yu, D. Brown, and S. Breton Detection of ClC-3 and ClC-5 in epididymal epithelium: immunofluorescence and RT-PCR after LCM Am J Physiol Cell Physiol, January 1, 2003; 284(1): C220 - C232. [Abstract] [Full Text] [PDF]

				O. Devuyst and W. B. Guggino Chloride channels in the kidney: lessons learned from knockout animals Am J Physiol Renal Physiol, December 1, 2002; 283(6): F1176 - F1191. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/1/566    most recent M106968200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mohammad-Panah, R. Articles by Bear, C. E. Search for Related Content PubMed PubMed Citation Articles by Mohammad-Panah, R. Articles by Bear, C. E. Social Bookmarking                      What's this?