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

J. Biol. Chem., Vol. 277, Issue 25, 22710-22717, June 21, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/25/22710    most recent M202394200v1 Alert me when this article is cited 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 Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Huebert, R. C. Articles by LaRusso, N. F. Search for Related Content PubMed PubMed Citation Articles by Huebert, R. C. Articles by LaRusso, N. F. Social Bookmarking                      What's this?

Expression and Localization of Aquaporin Water Channels in Rat Hepatocytes

EVIDENCE FOR A ROLE IN CANALICULAR BILE SECRETION^*

Robert C. Huebert, Patrick L. Splinter, Fabiana Garcia^§, Raul A. Marinelli^§, and Nicholas F. LaRusso^¶

From the  Center for Basic Research in Digestive Diseases and the ^¶ Departments of Internal Medicine and Biochemistry and Molecular Biology, Mayo Medical School, Clinic and Foundation, Rochester, Minnesota 55905 and the ^§ Instituto de Fisiología Experimental, Consejo Nacional de Investigaciones Científicas y Técnicas, Facultad de Ciencias Bioquímicas y Farmacéuticas, Universidad Nacional de Rosario, 2000 Rosario, Santa Fe, Argentina

Received for publication, March 12, 2002, and in revised form, April 3, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Although bile formation requires that large volumes of water be rapidly transported across liver epithelia, including hepatocytes, the molecular mechanisms by which water is secreted into bile are obscure. The aquaporins are a family of 10 channel-forming, integral membrane proteins of ~28 kDa numbered 0-9 that allow water to rapidly traverse epithelial barriers in several organs including kidney, eye, and brain. We found transcripts of three of 10 aquaporins in hepatocytes (aquaporin 8  aquaporin 9 > aquaporin 0) by reverse transcription-polymerase chain reaction and quantitative ribonuclease protection assays; immunohistochemistry confirmed the presence of these three proteins in liver. Immunoblots of subcellular fractions of hepatocytes showed enrichment of aquaporins 0 and 8 in microsomes and canalicular plasma membranes; aquaporin 9 was enriched only in basolateral plasma membranes. Immunofluorescence of hepatocyte couplets confirmed the intracellular/canalicular localization of aquaporins 0 and 8 and the basolateral localization of aquaporin 9. Upon exposure of couplets to a choleretic stimulus (i.e. dibutyryl cAMP), aquaporin 8 redistributed to the canalicular plasma membrane; the subcellular distributions of aquaporins 0 and 9 were unaffected. In addition, exposure of couplets to dibutyryl cAMP caused an increase in canalicular water transport in the presence and absence of an osmotic gradient, an effect that was blocked by aquaporin inhibitors. These results provide evidence that aquaporins are present in hepatocytes and that aquaporins are involved in agonist-stimulated canalicular bile secretion.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Primary bile is secreted by hepatocytes at the bile canaliculus and is modified via absorption and secretion of water, ions, and solutes by cholangiocytes, the epithelial cells that line the intrahepatic bile ducts. Hepatocytes are polarized epithelial cells that possess well defined canalicular and basolateral plasma membranes and are capable of rapidly transporting large volumes of water (1). Bile consists of 99% water, and water transport by hepatocytes is thought to occur passively in response to local, transient, osmotic gradients generated by the active transport of osmotically active solutes, especially bile acids (2). Two pathways exist by which water could potentially move from blood to bile across the hepatocyte epithelial barrier: a paracellular pathway through the tight junctions between adjacent hepatocytes and a transcellular pathway across hepatocytes. Furthermore, transcellular water movement across individual hepatocytes could theoretically occur either by diffusion through the lipid portion of the sinusoidal and canalicular hepatocyte plasma membranes or through aquaporin water channels, proteins that span the plasma membrane and allow for bi-directional, passive flux of water in response to solute-induced osmotic gradients. The quantitative contributions of these potential pathways (i.e. paracellular versus transcellular) and mechanisms (i.e. diffusion versus channel-mediated) of water transport in hepatocytes are currently unclear.

The aquaporins (AQPs)¹ are a family of integral membrane proteins of ~28 kDa, most of which function as water-selective channels. To date, 10 AQPs, numbered 0-9, have been cloned in mammals. These proteins exist in cells as homotetramers, and water molecules move bidirectionally through the central pore of each monomer in response to osmotic gradients (3). Water channels have been shown to be expressed in many epithelial cells; typically, one or more specific AQP exists in a particular water-transporting epithelial cell and is either constitutively expressed or regulated, often by agonist-induced trafficking from an intracellular vesicular compartment to the plasma membrane allowing for rapid changes in membrane permeability, depending upon physiological needs (e.g. AQP2 in response to ADH in the tubule epithelia of the collecting duct of the kidney) (4-7).

We originally reported that AQP1, AQP2, and AQP4 were not expressed in hepatocytes and suggested that transmembrane water transport is mainly via a non-channel-mediated pathway in the basal state (i.e. in the absence of a choleretic stimulus) (8). Although our more recent studies, as well as those of others, have shown evidence for the presence and agonist-induced trafficking of AQP8 in hepatocytes, a comprehensive study of the expression, subcellular localization, and trafficking of all of the AQPs in hepatocytes is lacking (9-11). Furthermore, the functional involvement of AQPs in canalicular water movement has not been demonstrated. The expression of multiple, differentially localized and trafficked AQPs in hepatocytes would support the notion that channel-mediated, transcellular water movement is an important mechanism for canalicular bile secretion. Thus, the purpose of the current study was to determine the expression, subcellular localization, and trafficking of AQPs in hepatocytes and to assess their role in canalicular bile formation.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Isolation of Rat Hepatocytes-- Highly purified (>98%) populations of rat hepatocytes were isolated from the livers of male Fisher rats, as previously described (12). The livers were perfused with oxygenated Hepes buffer solution containing 0.02% EGTA (Sigma) to remove blood cells and then transferred to a temperature-controlled chamber at 37 °C and perfused with Hepes buffer solution containing 0.45 mg/ml collagenase D (Roche Molecular Biochemicals). Following perfusion, the hepatocytes were gently removed from the biliary tree by mechanical disruption and filtered twice through 40-µm nylon mesh. The hepatocytes were purified further by isopycnic centrifugation through a discontinuous Percoll gradient (Amersham Biosciences) and washed three times in Liebovitz's L-15 medium (Invitrogen). Viability was greater than 90% as assessed by trypan blue exclusion. Of the isolated hepatocytes, ~65% were isolated as single cells, 25% were isolated as couplets, and 10% were isolated as cell clusters.

RNA Isolation-- Total RNA was extracted from freshly isolated rat hepatocytes using Tri-Reagent (Sigma). Isolated hepatocytes were lysed in 1 ml of Tri-Reagent/10 × 10⁶ cells with 5 µl of Glyco-Blue (Ambion Inc., Austin, TX) added as a co-precipitant and stored at room temperature for 5 min. After addition of 0.1 ml of 1-bromo-3-chloro-propane/1 ml of Tri-Reagent, the samples were vortexed, incubated for 15 min at room temperature, and centrifuged at 12,000 × g for 15 min at 4 °C. The upper, aqueous phase was collected and transferred to a new tube; to this, 0.5 ml of isopropanol was added per 1 ml of Tri-Reagent used for the initial lysis. The samples were stored for 10 min and centrifuged at 12,000 × g for 15 min at 4 °C. After removing the supernatant, the RNA pellet was washed with 1 ml of 75% ethanol and repelleted by centrifugation at 12,000 × g for 15 min at 4 °C. RNA was resuspended in RNA Secure solution (Ambion), and the concentration and purity were determined by spectrophotometry.

Reverse Transcription-Polymerase Chain Reaction-- 5 µg of total RNA was reverse transcribed using an avian myeloblastosis virus reverse transcriptase system (Promega, Madison, WI). RNA was first incubated for 10 min at 70 °C. The reaction mixture included reverse transcription buffer, 25 mM MgCl₂, 10 mM deoxynucleotide triphosphates, avian myeloblastosis virus reverse transcriptase, RNasin® ribonuclease inhibitor, and random primers in a final volume of 95 µl. This mixture was added to the total RNA and incubated for 10 min at room temperature and then 1 h at 42 °C. Heating to 95 °C for 5 min stopped the reaction. The AQP cDNA was amplified using the polymerase chain reaction with primers designed toward nonconserved regions of each of the 10 AQP genes (Table I). The PCR products were electrophoresed in 1% agarose gels, and the bands were visualized by ethidium bromide staining. Sequencing was performed on all positive PCR products (Mayo Molecular Core Facility, Rochester, MN) to confirm the identity of the amplified genes.

Quantitative Ribonuclease Protection Assay (Q-RPA)-- AQP mRNA expression was confirmed and quantified using quantitative ribonuclease protection assays, as previously described (13). All positive PCR reactions were gel purified using a gel extraction kit (QIAquick; Qiagen, Inc., Valencia, CA) and cloned into the pCR II dual promoter vector (Invitrogen Corp., San Diego, CA). The ³²P-labeled antisense probes were prepared by transcribing with the SP6 polymerase and labeling with [-³²P]UTP. The sense products were transcribed with T7 polymerase and then diluted to 0.1, 0.2, 0.5, and 1.0 ng for the preparation of the standard curve. The RPAs were performed by using 40 µg of total RNA from pure populations of isolated hepatocytes. The RPA probe was purified by excision from a 5% acrylamide, 8 M urea denaturing gel and elution into 0.5 M ammonium acetate, 1 mM EDTA, 0.1% SDS at 37 °C. The antisense probe was hybridized with RNA for the standard curve or with total hepatocyte RNA for 12 h at 45 °C. The unhybridized RNA was digested using a mixture of RNase A and RNase T. The protected hybrid was electrophoresed in a 5% acrylamide, 8 M urea gel and detected by exposure to Biomax MS x-ray film (Eastman Kodak, Rochester, NY). Densitometric analysis was performed for each probe to create a standard curve using the in vitro transcribed sense strands as template.

Preparation of Mixed Plasma Membranes-- Mixed plasma membranes were prepared from total rat liver using the method of Cefaratti et al. (14) with modifications or from isolated hepatocytes, as previously described (11). Briefly, the livers were perfused with oxygenated Hepes buffer solution containing 1 mM EGTA and then minced and homogenized in 50 ml of isolation medium (250 mM sucrose, 50 mM Tris, pH 8.0) by 10 passes with a loose fitting (type A) Dounce homogenizer followed by three passes with a tight fitting (type B) Dounce homogenizer. The homogenate was diluted to 6% (w/v) in the same buffer and centrifuged at 1400 × g for 10 min. The pellets were resuspended again at 6% in isolation medium. 1.4 ml of Percoll (Amersham Biosciences) was added for every 10.4 ml of resuspension. After centrifugation at 34,500 × g for 30 min, the top, fluffy layer containing the mixed plasma membrane was removed, washed in incubation medium (250 mM sucrose, 50 mM Tris, pH 8.0), and resuspended in phosphate-buffered saline. In some experiments, isolated hepatocytes were preincubated with 100 µM d-cAMP and 100 µM IBMX at 37 °C for 10 min prior to membrane isolation. The aliquots of mixed plasma membrane from the total rat liver (but not from isolated hepatocytes) were used for subsequent preparation of canalicular and basolateral plasma membranes.

Preparation of Canalicular and Basolateral Plasma Membranes-- Canalicular and basolateral plasma membrane fractions were isolated from aliquots of total rat liver mixed plasma membrane using the method of Cefaratti et al. (14) with modifications. Mixed plasma membranes were washed with five volumes of washing buffer (250 mM sucrose, 25 mM K-Hepes, pH 7.4) and centrifuged at 34,500 × g for 30 min. The pellet was resuspended to a concentration of 5 mg protein/ml, homogenized by 75 passes with a tight fitting (type B) Dounce homogenizer, and layered onto a discontinuous sucrose gradient (2.6 ml each of 43%, 46%, and 52% sucrose). After centrifugation at 93,000 × g for 1 h, the pellet (basolateral plasma membrane) and the bands at the top of the 43% layer and at the 43%/46% interface (canalicular plasma membrane) were collected. The bands were diluted with five volumes of washing buffer and spun at 35,000 × g for 30 min.

Preparation of Microsomal Membranes-- A microsomal fraction consisting of intracellular vesicles was isolated from total rat liver, as previously described (15). Liver homogenate was centrifuged at 200,000 × g for 60 min on a discontinuous 1.3 M sucrose gradient. The plasma membrane band and pellet were removed. The 1.3 M sucrose gradient was sonicated, diluted to 0.3 mol/liter, and centrifuged at 17,000 × g for 30 min. The resulting supernatant was centrifuged at 200,000 × g for 60 min to yield the microsomal membrane fraction.

Confirmation of Canalicular and Basolateral Membrane Purity-- The purity of the canalicular and basolateral membrane subfractions was confirmed using a combination of enzymatic assays and Western blots for specific canalicular and basolateral protein markers (Table II). Alkaline phosphatase (a canalicular marker) activity was assessed using a commercially available enzyme assay kit (Sigma). Na⁺/K⁺ ATPase (a basolateral marker) activity was assessed using the method of Scharschmidt et al. (16). Immunoblots were performed for aminopeptidase N (a canalicular marker) and the Na⁺/taurocholate co-transporting protein (a basolateral marker) (17, 18). Enzymatic assays for alkaline phosphatase (n = 3) showed a 7.5-fold enrichment in canalicular versus basolateral plasma membrane. Assays for the Na⁺/K⁺ ATPase (n = 3) showed a 12.5-fold enrichment in basolateral versus canalicular plasma membranes. Western blots for aminopeptidase N (15 µg of protein/lane) and Na⁺/taurocholate co-transporting protein (25 µg of protein/lane) showed canalicular and basolateral localization, respectively, as expected.

Immunoblotting-- MPM, CPM, BPM, and microsomal membranes from total rat liver or MPM from isolated rat hepatocytes were heated to 60 °C for 10 min in sample buffer containing 0.8 M dithiothreitol (Sigma) and 10% SDS (Fisher) for protein denaturation and solubilization. The samples were then subjected to electrophoresis through 7.5% or 12% SDS-polyacrylamide gels and transferred overnight to nitrocellulose sheets. The blots were blocked with 5% (w/v) nonfat dry milk and 0.2% (v/v) Tween 20. After blocking, the blots were incubated with affinity-purified rabbit anti-rat antibodies to the various AQPs (Alpha Diagnostic International, San Antonio, TX) at a dilution of 1:2000 to 1:5000 for 1 h at room temperature or overnight at 4 °C. Goat anti-aminopeptidase N (Santa Cruz Biotechnology, Santa Cruz, CA) was used at a dilution of 1:100. Rabbit anti-Na⁺/taurocholate co-transporting protein fusion protein (kindly provided by Dr. Peter Meier) was used at a dilution of 1:500. The blots were washed and incubated for 1 h at room temperature with a horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin (Tago, Inc., Burlingame, CA) or horseradish peroxidase-conjugated donkey anti-goat immunoglobulin (Santa Cruz Biotechnology). Protein bands were detected using an enhanced chemiluminescence detection system (ECL Plus; Amersham Biosciences). After exposing the nitrocellulose sheets to Kodak XAR film, the autoradiographs were scanned and quantified by densitometry using Molecular Analyst software (Bio-Rad).

Immunohistochemistry-- Whole liver was perfused with Hepes buffer solution, pH 7.4, containing 0.02% EGTA and then harvested, sliced, and fixed by immersion with 4% paraformaldehyde. The tissue was embedded in paraffin or Optimal Cutting Temperature embedding medium. Paraffin and frozen sections were cut to a thickness of 4-8 µm. The sections were placed in 10 mM citrate buffer (pH 6.0) and microwaved to improve staining by antigen unmasking. After washing and quenching of endogenous peroxidase, the sections were blocked and incubated with rabbit affinity-purified antibodies against the various rat AQPs (2-10 µg/ml; Alpha Diagnostics International) overnight at 4 °C. The remaining steps were carried out using an immunoperoxidase detection procedure (Vectastain Elite ABC Kit; Vector Laboratories, Burlingame, CA). The peroxidase was visualized by reaction with diaminobenzidine and hydrogen peroxide (diaminobenzidine peroxidase substrate kit; Vector Laboratories) and counterstaining with hematoxylin.

Immunofluorescence-- Collagen-coated coverslips were seeded with freshly isolated hepatocytes containing ~100,000 couplets and incubated for 4 h at 37 °C in Liebovitz's L-15 medium (Invitrogen) to re-establish cell polarity. After washing with medium, the cells were fixed with 2% paraformaldehyde and 0.01% glutaraldehyde for 10 min at 37 °C. Fixative was removed with six 1-min washes in PBS. The cells were then permeablized for 5 min in 0.2% Triton. The cells were again washed with six 1-min washes in PBS and then blocked with 5% goat serum and 5% glycerol for 1 h at room temperature. The cells were incubated with rabbit affinity-purified antibodies against the various rat AQPs (1:500 dilution; Alpha Diagnostic International) for 1 h at room temperature. After three 10-min washes in PBS, the cells were incubated with a Texas Red-conjugated goat anti-rabbit secondary antibody (Molecular Probes, Eugene, OR). The coverslips were again washed three times for 10 min in PBS followed by one 10-min wash in deionized water and then mounted using Prolong® Antifade mounting medium (Molecular Probes). The cells were examined using scanning laser confocal microscopy. Fluorescence intensity was quantified and displayed graphically using LSM 510 Image Examiner software (Carl Zeiss, Thornwood, NY). In some experiments, the couplets were preincubated with 100 µM d-cAMP and 100 µM IBMX at 37 °C for 10 min prior to fixation.

Osmotic Water Transport Studies-- 35-mm glass-bottomed culture dishes (MatTek, Ashland, MA) were coated with type I collagen from calf skin (Sigma) and seeded with ~100,000 freshly isolated hepatocyte couplets. The couplets were cultured for 6 h in Liebovitz's L-15 medium. The couplets were then exposed to hypertonic (350 mosmol) Krebs-Ringers buffer to shrink the canalicular space to ~40% of its original size. Then the hypertonic Krebs-Ringers buffer was removed and replaced with hypotonic (200 mosmol) Krebs-Ringers buffer to stimulate secretion into and re-expansion of the canalicular space. The canalicular changes were recorded using time lapse microscopy with differential interference contrast optics. Major (a) and minor (b) axes of the canaliculi were measured every 20 s for a total of 10 min. Canalicular volume was calculated assuming an ellipsoid shape of the lumen according to the following equation.

(Eq. 1)

In some experiments, the couplets were preincubated with 100 µM d-cAMP and 100 µM IBMX (Sigma) at 37 °C for 10 min and/or with 500 mM Me₂SO (Sigma) or 500 µM phloretin (Sigma), two AQP inhibitors that did not affect hepatocyte viability as determined by trypan blue staining (data not shown) (11, 19-22).

d-cAMP-induced Secretion Experiments-- 35-mm glass-bottomed culture dishes (MatTek) were coated with type I collagen from calf skin (Sigma) and seeded with ~100,000 freshly isolated hepatocyte couplets. The couplets were cultured for 6 h in Liebovitz's L-15 medium. The couplets were incubated for 60 min with 100 µM d-cAMP and 100 µM IBMX (Sigma) and/or with 500 mM Me₂SO (Sigma) or 500 µM phloretin (Sigma). The canalicular volume changes over 60 min were measured using differential interference contrast microscopy, as above. Because focal drift becomes a significant problem during this time frame, 4-µm beads were added to each culture dish as a focal reference. In addition, hepatocyte cellular volume was assessed by measuring the cellular diameter and calculating cellular volume according to the following equation.

(Eq. 2)

Protein Assay-- Protein concentrations were determined using the fluorescamine method with bovine serum albumin as a standard.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

AQP Expression by RT-PCR-- RT-PCR was run for each of the 10 AQPs on total RNA derived from pure populations of isolated rat hepatocytes (Fig. 1). Fig. 1 shows 1% agarose gels of the PCR products with ethidium bromide staining for AQP0, AQP8, and AQP9. The PCR products were sequenced, and the identities of the amplicons were verified by data base homology searches (BLAST; NCBI, National Institutes of Health). When using primers for AQPs 1, 2, 4, 5, 6, and 7, no amplification products were observed. These data are consistent with data previously published from our laboratory showing that AQP1, AQP2, and AQP4 are not expressed in hepatocytes. A positive band was also seen for AQP3 (data not shown), but AQP3 was shown to be negative by Q-RPA. This band for AQP3 may represent extremely low levels of AQP3 expression in hepatocytes (below the detection limits of the RPA) or a false positive because of small amounts of contaminating mRNA from endothelial or blood cells. The extreme sensitivity of RT-PCR leaves it vulnerable to false positives resulting from small numbers of contaminating cells carried over during hepatocyte isolation. Thus, this method was used only as an initial screening technique to rule out those AQPs that are not expressed in hepatocytes. Each AQP cDNA was amplified in a separate PCR reaction alongside appropriate positive and negative controls.

View larger version (48K): [in this window] [in a new window]   Fig. 1.   RT-PCR showing expression of AQPs in hepatocytes. Total RNA from rat hepatocytes was reverse transcribed using random primers and then PCR-amplified with primers designed to amplify a nonconserved region of each AQP cDNA. Positive bands are shown for AQP0, AQP8, and AQP9.

Quantitation of AQP mRNAs by Q-RPA-- Q-RPAs were used to confirm and quantify the mRNAs in pg of AQP mRNA/µg of total RNA (Table III). The data are expressed as the means ± S.E. of three separate experiments. The presence of three AQP mRNAs (AQP0, AQP8, and AQP9) was confirmed and quantified by Q-RPA. The AQP8 message was expressed at the highest level followed by AQP9 > AQP0. The relative abundance of the three mRNA transcripts suggests that the three AQP homologues exist in varying amounts. However, caution is necessary when extrapolating conclusions about protein abundance from nucleic acid data because of varying rates of mRNA degradation. AQP3 was not detected by Q-RPA (data not shown). Each AQP mRNA was labeled in a separate RPA alongside appropriate positive and negative controls (data not shown).

Immunohistochemistry-- Immunohistochemical staining of whole liver sections was used to confirm the presence of the AQP proteins (Fig. 2). Hepatocytes were positively stained for AQP0, AQP8, and AQP9. AQP0 (Fig. 2, A and D) showed strong intracellular staining, mainly in pericentral hepatocytes; any canalicular staining was most likely obscured by the strong intracellular signal. AQP8 (Fig. 2, B and E) showed a punctate, intracellular staining pattern and some staining at the canaliculi with increased intensity in the pericentral hepatocytes. AQP9 (Fig. 2, C and F) showed very strong staining on the basolateral surfaces of all hepatocytes. These data confirm the presence of AQP0, AQP8, and AQP9 in hepatocytes and suggest differential subcellular localization of the three proteins as well as differences in their lobular distribution. Controls using nonimmune IgG (Vector Laboratories) or omission of primary or secondary antibody revealed no labeling (data not shown).

View larger version (113K): [in this window] [in a new window]   Fig. 2.   Immunohistochemistry in whole liver sections. 4-8-µm sections were stained via a diaminobenzidine/peroxidase reaction, counterstained with hematoxylin, and visualized by light microscopy at magnifications of 4× (A-C) and 60× (D-F). A and D, staining for AQP0 was mainly intracellular and was most intense in hepatocytes surrounding the central vein. B and E, staining for AQP8 was mainly intracellular and was also most intense in pericentral hepatocytes. C and F, staining for AQP9 was strong at the basolateral membrane surfaces of all hepatocytes.

Immunoblotting for AQPs-- Immunoblots for AQP0, AQP8, and AQP9 in the basal state were run on subcellular fractions of hepatocytes enriched in MPM, CPM, BPM, or intracellular vesicles (i.e. microsomal membranes) (Fig. 3). The data are expressed as the means densitometric values ± S.E. of three separate experiments with a representative immunoblot pictured below. AQP0 and AQP9 showed bands near 28 kDa, as expected. AQP8 showed a band at 34 kDa, consistent with its known glycosylation (11). Immunoblots for AQP0 (Fig. 3A) showed enrichment in microsomes versus MPM and in CPM versus BPM, respectively. Similarly, immunoblots for AQP8 (Fig. 3B) showed enrichment in both microsomes and in CPM. Immunoblots for AQP9 (Fig. 3C) showed enrichment in MPM versus microsomes and in BPM versus CPM, respectively. These results are consistent with our immunohistochemical data.

View larger version (41K): [in this window] [in a new window]   Fig. 3.   Immunoblots for AQP0, AQP8, and AQP9 in the basal state. MPM, CPM, BPM, and microsomes from total rat liver were loaded onto 12% SDS-polyacrylamide gels using 25 µg of total protein/lane as starting material. Immunoblots were performed using affinity-purified rabbit anti-AQP primary antibodies and goat anti-rabbit secondary antibodies, visualized via chemiluminescence, and quantified using densitometry. The data are expressed in arbitrary densitometry units (n = 3; means ± S.E.) with representative immunoblots pictured below. A, blots for AQP0 showed enrichment in microsomes versus MPM and in CPM versus BPM. *, p < 0.05 versus MPM; #, p < 0.05 versus BPM. B, blots for AQP8 also AQP0 showed enrichment in microsomes versus MPM and in CPM versus BPM. *, p < 0.05 versus MPM; #, p < 0.05 versus BPM. C, blots for AQP9 showed enrichment in MPM versus microsomes and in BPM versus CPM. % p < 0.05 versus microsomes; +, p < 0.05 versus CPM.

Immunofluorescence-- Immunofluorescent staining of isolated hepatocyte couplets and subsequent quantitative analysis further confirmed the subcellular localization of AQP0, AQP8, and AQP9 in hepatocytes as well as demonstrating a d-cAMP-stimulated redistribution of AQP8 (Fig. 4). It has been shown that hepatocyte couplets redevelop their polarity upon isolation and short term culture (23). The interface area between the two coupled cells contains proteins derived from canaliculi, whereas the outer perimeter of the couplets contains proteins derived from basolateral surfaces. Representative couplets are displayed as raw confocal images, three-dimensional fluorescence intensity plots, and fluorescence profile graphs. In the basal state, AQP0 (Fig. 4A) and AQP8 (Fig. 4B) show strong staining in intracellular vesicles and some staining at the canaliculus. AQP8 staining was clearly redistributed from the interior of the cell to the CPM upon stimulation with d-cAMP (Fig. 4E), whereas AQP0 localization was unaffected (Fig. 4D). AQP9 stained mainly the BPM in both the basal (Fig. 4C) and d-cAMP-stimulated (Fig. 4F) states, suggesting that this AQP is constitutively expressed on the basolateral plasma membrane and is not trafficked in response to d-cAMP. These experiments provide further evidence of the intracellular and canalicular localization of AQP0 and AQP8 and the constitutively basolateral localization of AQP9 in the basal state. In addition, the data show that AQP8 (but not AQP0 or AQP9) can be redistributed to the canalicular plasma membrane by the choleretic agonist, d-cAMP. Controls using nonimmune IgG (Vector Laboratories) or omission of primary or secondary antibody revealed no specific (punctate) labeling, but a small amount of autofluorescence was evident when using the same confocal settings as the AQP experiments (data not shown).

View larger version (66K): [in this window] [in a new window]   Fig. 4.   Immunofluorescence in isolated hepatocyte couplets. The couplets were fixed and stained with affinity-purified rabbit anti-AQP primary antibodies and Texas Red-conjugated goat anti-rabbit secondary antibodies. The couplets were then visualized by laser scanning confocal microscopy, and the fluorescence was quantified using Zeiss 510 Image Examiner software. The results are displayed as raw confocal images, fluorescence profile graphs, and three-dimensional fluorescence intensity plots. A, staining for AQP0 in the basal state was observed within the cells and at the canaliculus. B, staining for AQP8 in the basal state was observed primarily within the cells but also at the canaliculus. C, staining for AQP9 was localized mainly at the BPM in the basal state. D, AQP0 staining was unaffected by treatment with d-cAMP. E, AQP8 staining redistributed from an intracellular vesicular compartment to the CPM upon treatment with d-cAMP. F, AQP9 staining remained strong at the BPM after d-cAMP treatment.

Immunoblotting with or without d-cAMP-- Immunoblotting of plasma membrane fractions of isolated rat hepatocytes treated with or without d-cAMP confirmed that AQP8 (but not AQP0 or AQP9) is trafficked to the plasma membrane in response to a choleretic stimulus (Fig. 5). Immunoblots for AQP0 and AQP9 (Fig. 5, A and C) show no change after treatment with d-cAMP, whereas there is a 4-fold increase in plasma membrane AQP8 (Fig. 5B). These data further support the concept that AQP8 undergoes regulated exocytic insertion into the canalicular plasma membrane upon stimulation with a choleretic agonist.

View larger version (45K): [in this window] [in a new window]   Fig. 5.   Immunoblots with or without d-cAMP. MPM from isolated hepatocytes treated with or without 100 µM d-cAMP were loaded onto 12% SDS-polyacrylamide gels using 20 µg of total protein/lane as starting material. Immunoblots were performed using affinity-purified rabbit anti-AQP primary antibodies and goat anti-rabbit secondary antibodies, visualized via chemiluminescence, and quantified using densitometry. The data are expressed in arbitrary densitometry units (n = 3; mean ± S.E.) with representative immunoblots pictured below. A, blots for AQP0 show no significant change after d-cAMP treatment. B, blots for AQP8 show a significant increase after d-cAMP treatment. *, p < 0.05 versus control. C, blots for AQP9 showed no significant change after d-cAMP treatment.

Osmotic Water Transport Studies-- Functional studies in isolated hepatocyte couplets were used to assess the importance of AQPs in osmotic water movement across hepatocytes (Fig. 6). Osmotic water transport in the basal state was unaffected by the AQP inhibitors, Me₂SO or phloretin (Fig. 6A). After initial shrinkage of the canalicular space to ~40% of original size, control cells showed a recovery of canalicular volume (to ~65% of original size) during 10 min of hypoosmotic stress. This recovery was unaffected by AQP inhibitors, suggesting that hepatocyte water movement is not channel-mediated in the basal or unstimulated state, consistent with previous studies from our laboratory (8). After incubation with d-cAMP, the rate of canalicular water transport increased during 10 min of hypoosmotic stress, resulting in a larger total increase in canalicular volume (to ~120% of original size). This effect was completely abolished when the couplets were also exposed to Me₂SO or phloretin (Fig. 6B), suggesting that the d-cAMP-induced increase in osmotic water movement into the canaliculus is AQP-dependent.

View larger version (34K): [in this window] [in a new window]   Fig. 6.   Osmotic water transport studies. Freshly isolated hepatocyte couplets were cultured for 6 h and then pretreated with or without 100 µM d-cAMP and 100 µM IBMX in the presence and absence of Me2SO or phloretin. After initial shrinkage of the canalicular space using a hypertonic buffer solution, secretion was stimulated by exposing the cells to a hypotonic stress. The images were obtained during a period of 10 min using time lapse differential interference contrast imaging, and then canalicular volume was calculated assuming an ellipsoid shape of the lumen. The data are expressed in relative canalicular volume as percentages of initial versus time. A, in the basal state, treatment of the couplets with the AQP inhibitors Me2SO or phloretin did not significantly affect osmotic water transport into the canaliculus (n = 4-6; mean ± S.E.). B, treatment with d-cAMP significantly increased osmotic water transport into the canaliculus, an effect that could be completely abrogated by inhibitors of AQP water channels (n = 4-6; mean ± S.E.). *, p < 0.05.

d-cAMP-induced Secretion-- d-cAMP-induced secretion in isolated hepatocyte couplets in the absence of an external osmotic gradient was also shown to be AQP-dependent (Fig. 7). The driving force for this secretion is of physiological size and is generated by the action of the choleretic agonist, d-cAMP. Fig. 7 shows the mean ± S.E. of four to six separate experiments for each treatment group. A representative couplet after 60 min of treatment is also shown for each group. Canalicular volume in control cells increased only slightly during 60 min (< 20%). In contrast, couplets treated with d-cAMP showed canalicular volume increases of ~300% during the 60-min experiment. This secretory effect was reduced to control levels when cells were also treated with the AQP inhibitors, Me₂SO or phloretin. Treatment with Me₂SO alone or phloretin alone had no significant effect on canalicular volume for 60 min (data not shown). These results suggest an AQP dependence of the d-cAMP-induced increase in canalicular secretion in a physiologically relevant model system. Hepatocyte cellular volume was unaffected by all the experimental perturbations (d-cAMP, Me₂SO, and phloretin, alone or in combination; data not shown).

View larger version (65K): [in this window] [in a new window]   Fig. 7.   d-cAMP-induced bile secretion. Freshly isolated hepatocyte couplets were cultured for 6 h and then pretreated with and without 100 µM d-cAMP and 100 µM IBMX in the presence and absence of Me2SO or phloretin. Secretion into the canalicular space was recorded for a period of 60 min using differential interference contrast imaging, and then canalicular volume was calculated assuming an ellipsoid shape of the lumen. 4-µm beads were added to each culture dish as a focal reference. The data are expressed in relative canalicular volume as percentages of initial with a representative couplet from each treatment group after 60 min pictured above. d-cAMP caused a significant increase in canalicular secretion, and this effect could be completely abrogated by both Me2SO and phloretin. *, p < 0.05.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The major findings reported here relate to the expression and subcellular localization of aquaporin water channels in rat hepatocytes and their possible role in canalicular bile secretion. Using molecular, biochemical, and morphological approaches in both highly purified rat hepatocytes and whole rat liver, we showed that: (i) the message and protein of three (AQP0, AQP8, and AQP9) of the 10 known AQPs are expressed in hepatocytes in variable amounts (AQP8 AQP9 > AQP0); (ii) AQP9 is found principally on the basolateral hepatocyte membrane, whereas AQP0 and AQP8 are principally localized intracellularly and to a lesser degree on the canalicular plasma membrane; (iii) AQP0 and AQP8 are expressed to a larger degree in hepatocytes surrounding the central vein, whereas AQP9 is uniformly distributed within the hepatic lobule; (iv) the amount of AQP8 in the canalicular plasma membrane markedly increases after exposure of hepatocytes to a choleretic agonist; and (v) two inhibitors of AQPs block agonist-induced (but not basal) canalicular secretion in isolated hepatocyte couplets. Our results are consistent with the notion that AQPs play a role in agonist-induced canalicular bile secretion by hepatocytes.

AQPs have a wide range of tissue distribution in mammals, but their expression is most notably associated with tissues involved in the rapid transport of large volumes of water such as kidney, eye, brain, secretory glands, and more recently, colon (4-7, 9). Some epithelial cells express more than one AQP with the different AQP homologues existing at different subcellular locations. Also, some AQPs undergo agonist-induced exocytic insertion and endocytic retrieval into and out of the plasma membrane as a mechanism for regulating AQP-mediated water transport. For example, the principle cells of the collecting duct in the kidney contain AQP2 in intracellular vesicles and AQP3 and AQP4 on their basolateral plasma membrane in the basal state; after exposure to vasopressin, the amount of AQP2 (but not AQP3 or AQP4) in the apical plasma membrane is increased via insertional exocytosis promoting reabsorption of water within this segment of the nephron (6). After withdrawal of vasopressin, AQP2 molecules are removed from the apical plasma membrane by endocytosis. Similarly, the cells of the salivary and lacrimal glands contain AQP5 at the apical domain and AQP3 or AQP4 at the basolateral domain (4, 5). AQP5 in the salivary gland, which is normally sequestered in intracellular vesicles, redistributes to the apical plasma membrane upon stimulation by the neurotransmitter, epinephrine (24). Finally, we have demonstrated that AQP4 is constitutively expressed on the basolateral membrane of cholangiocytes, whereas AQP1, normally residing in intracellular vesicles, undergoes secretin-induced exocytic insertion into the apical cholangiocyte membrane (15, 25, 26). Thus, a generalizable model accounting for the physiological role of AQPs in water-transporting epithelia is developing in which multiple, differentially localized, constitutively expressed or trafficked AQPs function together to facilitate regulated transcellular water movement in water-transporting epithelia in both the basal and stimulated states.

d-cAMP has been shown to stimulate canalicular bile secretion in the perfused rat liver (27) as well as in isolated rat hepatocyte couplets (28). Moreover, recent evidence suggests that cAMP stimulates the canalicular targeting of vesicles containing the Cl/HCO (29), the bile salt export pump (28), and the organic anion transporter, Mrp2 (30, 31), processes that may underlie the choleretic effect of cAMP. In this study, we showed that d-cAMP stimulated the insertion of AQP8-containing vesicles into the canalicular membrane of hepatocytes. We propose that this process facilitates the movement of water coupled to the secretion of osmotically active solutes and, in turn, the elaboration of canalicular bile.

We found that AQP0 was present in hepatocytes and was localized principally to intracellular vesicles and that, unlike AQP8, this localization was unaffected by d-cAMP. This protein had previously been thought to be expressed exclusively in the lens fiber cells of the eye where it functions as a water channel of low permeability but also as a structural protein important in cell-cell adhesion (32). The significance of AQP0 expression in hepatocytes is currently unclear, but we speculate that its role may be unrelated to canalicular bile secretion.

Recently, we showed that AQP8 is localized largely to intracellular vesicles within isolated hepatocytes in the unstimulated state and was redistributed to the hepatocyte plasma membrane via stimulation with d-cAMP (11). Our results here extend these findings by demonstrating that the canalicular membrane is the specific domain of the plasma membrane into which the insertion of AQP8 occurs. This mechanism of regulation may allow hepatocytes to quickly and selectively increase the osmotic permeability of their canalicular plasma membrane in response to a choleretic stimulus. AQP8 expression has also been demonstrated on the luminal side of seminiferous tubules, the apical membrane of pancreatic acinar cells, and in both absorptive epithelia of the colon and acinar cells of the salivary gland; whether AQP8 in these tissues is also regulated by a trafficking mechanism has not been explored (10, 33-36).

Others recently suggested that AQP9 resides principally on the sinusoidal (basolateral) plasma membranes of hepatocytes (9). Our results support this topography using multiple experimental techniques, but also show that AQP9 localization does not change in response to a choleretic stimulus. This constitutively basolateral localization of AQP9 is thought to facilitate trans-basolateral water movement, a critical component of transcellular movement of water across hepatocytes. Other studies have shown AQP9 to be expressed in epididymus, testis, spleen, and brain, but the subcellular location of AQP9 in these cell types is unclear.

Historically, water transport in hepatocytes has been assumed by many to occur via a paracellular pathway (i.e. between cells). This assumption was based on evidence involving tight junction penetration of nonaqueous substances and on in vitro electrical resistance measurements suggesting that hepatocytes are "leaky epithelia" (37-39). More recently, we suggested that transcellular water movement might account for the high osmotic permeability coefficient of isolated hepatocytes (i.e. P_f = 66.4 × µm/s) we previously observed from careful measurements. This value is actually higher, for example, than that of cholangiocytes (i.e. P_f = 50.0 × µm/s), which we have shown to transport water via a transcellular, channel-mediated pathway (20, 40). We previously proposed that the transcellular mechanism by which hepatocytes transported water was mainly via a non-channel-mediated pathway (i.e. via the lipid phase) in the basal state (8). This conclusion was based on careful measurements demonstrating a temperature dependence of hepatocyte membrane permeability and on an apparent lack of water channel expression. The calculated activation energy (E_a) from these experiments was 12.8 kcal/mol (outside of the typical range for channel-mediated transport) and was decreased to 3.6 kcal/mol in the presence of amphotericin B, a substance that forms artificial water channels in plasma membranes. Our functional studies here using AQP inhibitors also support the concept that transcellular water transport by hepatocytes in the basal state occurs mainly via a non-channel-mediated pathway. Our discovery of the expression of three water channels with distinct subcellular locations within hepatocytes and our data demonstrating that the insertion of AQP8 into the canalicular plasma membrane can be stimulated by a choleretic agonist suggest that water transport across the hepatocyte epithelial barrier in response to a choleretic stimulus (as opposed to the basal, unstimulated state) is due to a transcellular pathway facilitated by the regulated and constitutive expression of AQP8 and AQP9, respectively. The role of AQP0, which appears to be localized primarily in intracellular vesicles, remains obscure. Indeed, we have now demonstrated that in addition to the diffusional and paracellular routes, a regulated, channel-mediated component for water transport by hepatocytes also exists (Fig. 8). The quantitative contributions of these various pathways to total canalicular bile secretion is under active investigation.

View larger version (85K): [in this window] [in a new window]   Fig. 8.   Proposed model for water transport during canalicular bile secretion. AQP0 and AQP8 are localized in intracellular vesicles and on the canalicular plasma membrane, whereas AQP9 is localized on the basolateral plasma membrane. AQP8 is inserted into the canalicular plasma membrane in response to d-cAMP, but the localization of AQP0 and AQP9 is unaffected by d-cAMP. The regulated insertion of AQP8-containing vesicles into the secretory pole (i.e. canalicular membrane) of hepatocytes would facilitate the movement of water coupled to the secretion of osmotically active solutes and, in turn, the elaboration of canalicular bile.

In summary, our findings show that hepatocytes express multiple differentially localized and differentially trafficked AQPs, a pattern increasingly being recognized in water transporting epithelial cells. Although hepatocyte water movement appears to be non-channel-mediated in the basal state, d-cAMP-induced water secretion can be abrogated by inhibitors of AQPs. The data suggest that water channels play an important role in the transcellular transport of water during primary bile secretion by the hepatocyte. Further studies will be necessary to determine the quantitative contribution of channel-mediated water movement relative to diffusion across and movement between hepatocytes, as well as the exact role and mechanism of regulation for each AQP homologue.

	ACKNOWLEDGEMENTS

We acknowledge Deb Hintz for secretarial support, Dr. Bing Huang for assistance with confocal microscopy, and Pamela Tietz, Dr. Anatoly Masyuk, and Dr. Giuseppe Calamita for valuable discussions.

	FOOTNOTES

^* This work was supported by Grant DK24031 from the National Institutes of Health (to N. F. L.), by funds from the Mayo Foundation, by Grant PICT 03589 from Agencia Nacional de Promoción Científica y Tecnológica (R. A. M.), and by a Grant from Subsecretaría de Investigación y Tecnología, Ministerio de Salud.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.

To whom correspondence should be addressed: Center for Basic Research in Digestive Diseases, Mayo Medical School, Clinic and Foundation, 200 First St. SW, Rochester, MN 55905. Tel.: 507-284-1006; Fax: 507-284-0762; E-mail: larusso.nicholas@mayo.edu.

Published, JBC Papers in Press, April 3, 2002, DOI 10.1074/jbc.M202394200

	ABBREVIATIONS

The abbreviations used are: AQP, aquaporin; d-cAMP, dibutyryl cAMP; IBMX, isobutyl methyl xanthine; RT, reverse transcription; Q-RPA, quantitative ribonuclease protection assay; PBS, phosphate-buffered saline; MPM, mixed plasma membrane; CPM, canalicular plasma membrane; BPM, basolateral plasma membrane.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

G. Calamita, D. Ferri, P. Gena, F. I. Carreras, G. E. Liquori, P. Portincasa, R. A. Marinelli, and M. Svelto Altered expression and distribution of aquaporin-9 in the liver of rat with obstructive extrahepatic cholestasis Am J Physiol Gastrointest Liver Physiol, October 1, 2008; 295(4): G682 - G690. [Abstract] [Full Text] [PDF]