Morphologic and functional characterization of caveolae in rat liver hepatocytes

Morphologic and Functional Characterization of Caveolae inRat Liver Hepatocytes

MARIA CALVO,1 FRANCESC TEBAR,1 CARMEN LOPEZ-IGLESIAS,2 AND CARLOS ENRICH1

Caveolae are small pits on the plasma membrane found inseveral, if not all, differentiated cells. They are involved inpotocytosis, endocytosis, transcytosis, membrane traffick-ing, and signal transduction. Although caveolin has recentlybeen identified in subcellular fractions from rat liver there isno clear-cut morphologic evidence for the presence of pro-totypical caveolae on the surface of hepatocytes. In thisstudy the presence of caveolae at the cell surface of hepato-cytes was directly shown by rapid-freeze, deep-etching elec-tron microscopy. Moreover, combined deep-etching and im-munogold techniques revealed caveolin in caveolae of thedorsal membrane of primary culture hepatocytes. Using re-agents that perturb membrane cholesterol and interferewith endocytosis through the caveolae, a caveolae-depen-dent internalization of cholera toxin B and retinol-bindingprotein by hepatocytes in primary culture was shown. Fi-nally, immunocytochemical analysis of caveolin in non-parenchymal cells of the rat liver showed its presence inKupffer and stellate cells, however no caveolin could bedetected in endothelial cells. (HEPATOLOGY 2001;33:1259-1269.)

The hepatocyte is a highly polarized epithelial cell with aplasma membrane divided into three main functional do-mains: the sinusoidal domain, facing the blood and thehepatic endothelial cells; the lateral domain, containing thejunctional complexes (e.g., desmosomes and gap junc-tions); and the canalicular plasma membrane, involved inbile secretion. The sinusoidal domain of the hepatocyteplasma membrane (basolateral) in contact with the bloodcontains most of the receptors for hormones, growth fac-tors, and metabolites.1 A functional continuity of this ba-

solateral membrane with the endosomal compartment ismediated by clathrin-coated pits and vesicles, which even-tually fuse with the early/sorting endocytic compartment.Although most of receptors involved in endocytosis ortranscytosis are located in clathrin-coated pits, alternativeports of entry by nonclathrin-coated pits have been shownin various hepatocyte-derived cell lines.2,3

In biochemical analysis of isolated rat-liver plasma mem-branes, the presence of caveolin was not considered, mainlybecause of its low expression in the liver and the lack ofappropriate antibodies, but also because such cell-surfacestructures were believed to be restricted to endothelial cells.However, the expression of caveolin in liver homogenates wasshown by Northern4 and by Western blot,5 and the internal-ization of specific ligands through caveolae has been reported.5-7

Recently, we have shown the presence of caveolin in the en-docytic compartment and in a caveolae-enriched plasmamembrane fraction from rat liver,8 together with differentialdistribution of signaling molecules.9

The liver is composed of parenchymal, hepatocytes, andnonparenchymal cells, mainly Kupffer, stellate, and endothe-lial cells. Little is known about the cellular distribution ofcaveolae in liver cells. Smooth-surfaced caveolae have beendescribed on the perisinusoidal surface of stellate cells,10 andthey have also been observed in Kupffer cells.11 Unlike othertissues, liver endothelial cells do not seem to contain a signif-icant amount of caveolar structures, in the plasmamembrane,but they were enriched in clathrin-coated pits.12,13

Although clathrin-coated pits and vesicles transport recep-tors and ligands en route to lysosomes or for transcytosis,caveolae may be alternative endocytic pathways involved incholesterol transport and signal transduction. They are 50- to70-nm plasma membrane invaginations enriched in choles-terol and sphingolipids, and they contain caveolin (recentlyreviewed by Kurzchalia and Parton 1999,14 and Anderson199815).The biochemical finding that subcellular fractions from rat

liver, derived from plasma membrane and endosomes, con-tained caveolin prompted us to study the possibility that he-patic cells were endowed with caveolae, like other polarizedepithelial cells, e.g., MDCK cells.In the present study, rapid freeze deep-etching and im-

muno-electronmicroscopy were used to show the presence ofcaveolae in hepatocytes in primary culture and to characterizethem morphologically. We also found that internalization ofcholera toxin subunit B (CT) and retinol binding protein(RBP) was impaired by drugs or agents reported to bind oralter cholesterol-rich rafts, such as caveolae, in isolated hepa-tocytes.

Abbreviations: MDCK, Madin-Darby canine kidney cells; RBP, retinol-binding pro-tein; CT, cholera toxin subunit B; LDL, low density lipoprotein; ASF, asialofetuin; HPC,hepatocytes in primary culture; CEF, caveolin-enriched plasma membrane fraction;SR-BI, scavenger receptor type BI; GPI, glycosylphosphatidylinositol; EGF, epidermalgrowth factor; ERK, extracellular signal-regulated kinase;MAPK,mitogen-activated pro-tein kinase; CD, 2-hydroxypropyl�-cyclodextrin; NPC, nonparenchymal cells; KC,Kupffer cells; SC, stellate cells; ASGPR, asialoglycoprotein receptor.From the 1Departament de Biologia Cellular, Institut d’ Investigacions Biomediques

August Pi I Sunyer, Facultat de Medicina, Universitat de Barcelona; 2Serveis Cientıfi-cotecnics de la Universitat de Barcelona.Received July 10, 2000; accepted February 14, 2001.Supported by Ministerio de Educacion y Cultura grants: PM99-0166 and from Fun-

dacio Marato TV3 2000 (to C.E.).Address reprint requests to: Carlos Enrich, Ph.D., Departament de Biologia Cellular,

Facultat de Medicina, Universitat de Barcelona, Casanova 143, 08036-Barcelona, Spain.E-mail: [email protected]; fax: (34) 93-4021907.Copyright © 2001 by the American Association for the Study of Liver Diseases.0270-9139/01/3305-0031$35.00/0doi:10.1053/jhep.2001.23937

1259

MATERIALS AND METHODS

Antibodies

The rabbit and mouse anti-caveolin antibodies and the anti-�-adaptin were from Transduction Laboratories (Lexington, KY); amouse anti-caveolin, used for cryo-immunoelectron microscopy,was from Zymed Laboratories (South San Francisco, CA), the mouseanti-actin was from ICN (Costa Mesa, CA), and the anti-ASGP re-ceptor was prepared in our laboratory; the mouse anti-pIgR waskindly donated by Dr. K.E. Mostov (University of California, SanFrancisco, CA); the monoclonal anti-lgp120 antibody (clone GM10)was provided by K. Siddle and J. Hutton (University of Cambridge,UK); the monoclonal anti-clathrin X22 was donated by Dr. FrancesBrodsky (University of California, San Francisco, CA); the monoclo-nal mouse anti-MRC OX43 was used as an endothelial-cell marker(Harlan Bioproducts, Madison, WI); and a monoclonal mouse anti-CD11b (clone WT.5), FITC conjugated (Calbiochem-NovabiochemCorporation, Darmstadt, Germany), was used as a Kupffer cellmarker. Finally, fluorescence conjugated (FITC and Cy3) antibodieswere from Jackson ImmunoResearch (West Grove, PA) and gold-coupled secondary antibodies for immuno-electronmicroscopywerefrom British BioCell International Ltd. (Cardiff, UK).

Animals

Male Sprague-Dawley rats weighing 200 to 250 g were kept undera controlled lighting schedule with a 12-hour dark period. All ani-mals received humane care in compliance with institutional guide-lines. Food and water were available ad libitum.

Liver Cell Isolation and Culture

Isolation and Culture of Rat Hepatocytes. Hepatocytes were isolatedfrom a young male Sprague-Dawley rat using a sterile method.16 Theportal vein was cannulated, and the liver was perfused in a nonrecir-culating manner with buffered, warm Hanks’ balanced salt solution(HBSS) saturated with 95% O2 and 5% CO2 gas, with 0.5 mmol/LEGTA at a rate of 20 mL/min for 10 minutes. The liver was excisedfrom the animal while perfusion continued. Next, liver was dissoci-ated gently with a glass rod after recirculating perfusion with warmHBSS containing type IV collagenase (Sigma, Madrid, Spain)(0.035%) and 4 mmol/L CaCl2. Cell suspension was filtered throughgauze with cold Kreb’s Henseleit medium and allowed to settle for 10minutes. Cells were then decanted, resuspended in Kreb’s, andwashed 3 times. The obtained pellet containing hepatocytes wasresuspended in medium DMEM/HAMF12 (Sigma) 10% FCS andcounted. Trypan blue exclusion showed cell viability greater than95%. Cells were plated on 60-mm culture dishes at a density of 3 �106 cells per dish. Cultured cells were incubated at 37°C in a humid-ified 95% O2 and 5% CO2 atmosphere for 12 to 24 hours.

Isolation and Culture of NPCs. Isolation of rat stellate cells was per-formed as described in Geerts et al.17 After isolation of rat stellatecells, these were seeded in cell-culture dishes and glass coverslips inIscove’s medium supplemented with 0.1% nonessential amino acids,2 mmol/L glutamine, 1 mmol/L sodium pyruvate, 10% FCS, andantibiotics.Isolation of rat liver Kupffer and endothelial cells was performed

as described in Pertoft et al.18 The interface at 25% to 50% of thePercoll gradient, containing mainly Kupffer and endothelial cells,was seeded on cell-culture plates and incubated for 30 minutes at37°C. Purified Kupffer-cell preparations were obtained by the rapidadherence of these cells (30 minutes) to a glass or plastic surface.Kupffer cells were cultured for 1 day in RPMI 1640 medium supple-mented with 2 mmol/L glutamine 5% FCS and antibiotics. Nonat-tached cells (enriched in endothelial cells) after this period wereseeded on fibronectin covered cell-culture dishes or glass coverslipsin RPMI 1640 supplemented with heparin 8 U/mL, 15% Nu-serum,15% bovine calf serum and antibiotics, and cultured for 1 week.

Electron-Microscopy Procedures. Cell monolayers on petri disheswere fixed at room temperature in 2% glutaraldehyde in 0.1 mol/L

phosphate buffer for 1 hour. Cells were scraped and collected into0.1 mol/L phosphate buffer containing 2% of paraformaldehyde. Af-ter 3 rinses in 0.1 mol/L phosphate buffer, the pellets were postfixedin 1% OsO4 in 0.1 mol/L phosphate buffer. Finally, samples wereembedded in Epon.For cryo-immunoelectron microscopy, livers of young adult

Sprague-Dawley rats were fixed by perfusion with 3% paraformalde-hyde, 0.5% glutaraldehyde in 0.1mol/L phosphate buffer. After cryo-protection in 2.1mol/L sucrose, livers were frozen in liquid nitrogen,and ultrathin sections were cut using a Reichert Ultracut S cryo-ultramicrotome equipped with a cryochamber attachment (Leica,Heidelberg, Germany). Cryo-sections were immunolabeled as de-scribed elsewhere.19 After labeling with a monoclonal anti-caveo-lin-1 antibody (2.5 �g/mL) (Zymed) and a secondary anti-mouseconjugated to gold 15 nm, sections were rinsed in distilled water andembedded in 1.8% methyl cellulose 0.3% uranyl acetate.20 In someexperiments, cryo-sections were also labeled with anti-clathrin X22antibody (1:40).For cryo-substitution, hepatocytes in primary culture were grown

on transwell filters. Small pieces of the filters were cryofixed byprojection against a copper block cooled in liquid nitrogen(�196°C) using a Cryoblock (Leica) as described.21 Freeze-substi-tution was performed in a home-made cryosystem,22 using acetonecontaining 0.5% of uranyl acetate, for 3 days at�90°C. On the fourthday, the temperature was slowly increased, 5°C/h, to�50°C. At thistemperature samples were rinsed in acetone and then infiltrated andembedded in Lowicryl HM20. Ultrathin sections were picked up onFormvar-coated gold. For immunogold localization, samples wereblocked with 2% ovoalbumin for 30 minutes and incubated at roomtemperature for 1 hour withmonoclonal anti-caveolin antibody fromZymed (2.5 �g/mL). Washes were performed with PBS prior to add-ing goat anti-mouse conjugated to 15 nm colloidal gold for 45 min-utes at room temperature. Finally, samples were washed and con-trasted with 2% uranyl acetate for 30 minutes.

Freeze-Fixation, Freeze-Drying Electron Microscopy. Dorsal plasmamembranes of rat hepatocytes in primary culture were obtained by asandwich technique. Cell membranes on coverslips were preparedby overlaying a TESPA-coated coverslip onto the upper surface ofprimary cultured hepatocytes, applying gentle pressure, and remov-ing the coverslip with adherent membranes, all at 4°C.23,24 Ventralmembranes were obtained by a lysis-squirting technique. Briefly,cells grown on coverslips were chilled with cold PBS, lysed withhipotonic buffer (30 mmol/L HEPES, 70 mmol/L KCl, 5 mmol/LMgCl2, 3 mmol/L EGTA, pH 7.5) diluted 1:3 in distilled water, andsquirted by a stream of buffer (30 mmol/L HEPES, 70 mmol/L KCl, 5mmol/L MgCl2, 3 mmol/L EGTA, pH 7.5). For immunogold local-ization,membraneswere chilled to 4°C and fixed for 20minutes with3.7% paraformaldehyde in 0.1 mol/L phosphate buffer and rinsed inbuffer containing 20 mmol/L HEPES, 100 mmol/L KCl, 5 mmol/LMgCl2, 3 mmol/L EGTA, pH 6.8 before incubation with primaryantibody for 1 hour at 4°C. After washing, membranes were incu-bated with secondary antibody conjugated to gold. Then sampleswere washed and fixed with 2.5% glutaraldehyde before processingas for rapid-freezing and deep-etching electron microscopy. Briefly,coverslips were cryofixed by projection against a copper block asdescribed earlier. The frozen samples were stored at �196°C in liq-uid nitrogen until subsequent use. Samples were freeze-dried andcoated with platinum and carbon using a freeze-etching unit (modelBAF-060, BAL-TEC, Liechtenstein). A rotatory shadowing of theexposed surface was made by evaporating 1 nm platinum-carbon atan angle of 24° above the horizontal, followed by 10 nm of carbonevaporated at a 75° angle.24 The replica was separated from the cov-erslip by immersion in full-strength hydrofluoric acid, washed twicein distilled water, and digested with 5% sodium hypochlorite for 5 to10 minutes. Finally, the replicas were washed several times in dis-tilled water, broken into small pieces, and picked up on Formvar-coated copper grids for electron microscopy. When immunolabelingwas carried out, the replicas were only washed in distilled water. Allelectron micrographs were obtained on a Hitachi HU-600, operating

1260 CALVO ET AL. HEPATOLOGY May 2001

at 75 kV. The photographic negatives were digitalized without con-trast-reversing and treated by the IMAT program (Alejandro DiGior-gio, Serveis Cientıfico Tecnics, Universitat de Barcelona). As controlsfor single immunostaining, incubation with the second antibodyonly was accomplished. For double labeling, controls using only 1primary antibody, and the respective secondary antibodies were per-formed to establish that the colocalization was not the result of rec-ognition of the same primary antibody. In both cases, the labelingwas specific, as no signal was obtained (data not shown).

Internalization Assays With Cholera Toxin-Gold

To study the binding and the internalization of cholera toxin-goldby caveolae, a pre-embedding procedure was performed. Briefly,hepatocytes cultured on 12-well cell-culture plates for 20 hours werewashed once with DMEM/25mmol/L HEPES containing 0.1% BSA at4°C and then incubated for 2 hours with cholera toxin-gold (15�g/mL) at 4°C. After several washes, the medium was replaced byprewarmed fresh medium to induce the internalization for 10 min-utes. Finally, cells were fixed with 2.5% glutaraldehyde – 2% para-formaldehyde in phosphate buffer for 1 hour at room temperature.Fixed cells were scrapped and prepared for electron microscopy asdescribed earlier.

Other Procedures

SDS-PAGE and Western Blotting. SDS/PAGE of proteins was per-formed in 10% polyacrylamide, as described by Laemmli.25 ForWestern blotting, polypeptides were transferred electrophoreticallyat 60 V for 90 minutes at 4°C to immobilon-P transfer membranes(Millipore Iberica, Madrid, Spain), and antigens were identified us-ing specific antibodies diluted in TBS (Tris-buffered saline) contain-ing 0.5% powdered skimmed milk, and finally the reaction productwas detected using the ECL system (Amersham Iberica, Madrid,Spain). Image analysis of Western-blots and band quantification wasperformed with a Bio-Image system (Millipore). The protein contentof the samples was measured by the method of Bradford26 using BSAas standard.

Conjugation of Asialofetuin and RBP to FITC. Asialofetuin (Sigma,Ma-drid, Spain) was diluted (10mg/mL) in 50mmol/L sodium carbonatebuffer pH 9.2, and RBP (Sigma) (2.5 mg/mL) in 0.2 mol/L boratebuffer pH 9.0 and then mixed end-over-end with fluorescein isothio-cyanate (Sigma) dissolved (10 mg/mL) in ethanol for 2 hours in thedark at room temperature. The conjugates were transferred to a PBSpre-equilibrated PD-10 column to separate FITC-conjugated asia-lofetuin from free FITC.

Conjugation of Low-Density Lipoprotein to DiI. Human low-densitylipoprotein (LDL) was isolated and dialyzed27 in 0.15 mol/L NaCl at4°C, was brought to a concentration of 2 mg/mL in PBS 0.5% BSA at37°C, and mixed with 1,1�-dioctadecyl-3,3,3�,3�-tetramethylindo-carbocyane (DiI) (Molecular Probes, Eugene, OR) (30 mg/mL inDMSO) at a conjugation ratio of 0.3 mg DiI/2 mg LDL for 8 to 10hours at 37°C. The solution was brought to a density of 1,063mg/mLwith KBr and centrifuged at 36,000 rpm for 20 hours in a Beckman(Palo Alto, CA) SW 50.1. The top supernatant was collected anddialyzed in 0.15 mol/L NaCl at 4°C for 1 to 2 days.28

Preparation of Cholera Toxin-B-Gold. Colloidal gold (17 nm) wasprepared as described by Slot and Geuze.29 The pH of the solutionwas adjusted to 6.9. The cholera toxin-gold was stabilized with BSAto a concentration of 0.1% and then washed twice by centrifugation.Immediately before use, the gold was washed once more and resus-pended in medium or PBS containing 0.1% BSA.

Ligands Internalization and Immunofluorescence. Hepatocytes on glasscoverslips were incubated for 30 minutes with HEPES-modifiedDMEM 1% BSA and treated with 5 �g/mL filipin (Sigma) for 1 houror with 1.5% cyclodextrin (Sigma) for 30 minutes. Control andtreated cells were allowed to uptake cholera toxin subunit B-FITCconjugated (8�g/mL) (Sigma), asialofetuin-FITC or DiI-LDL (20-50�g/mL) for 10 minutes with the presence of the drugs. Cells werequickly rinsed and incubated in the absence of labeled ligands, and

with or without the presence of drugs in the medium for differentperiods. Cells were fixed with 3.7% paraformaldehyde, washed inPBS, and mounted in Mowiol (Calbiochem, La Jolla, CA).Cells grown on glass coverslips were fixed in 3.7% paraformalde-

hyde for 10minutes at room temperature and then permeabilized for15 minutes in PBS, 1% BSA, 0.1% saponin. After 3 washes in PBS,cells were incubated for 1 hour at 37°C in a humidified atmospherewith primary antibody in PBS 0.1% BSA, 0.1% saponin. Coverslipswere then washed 3 times in PBS and incubated for 1 hour at 37°Cwith corresponding secondary antibodies in PBS 0.1% BSA, 0.1%saponin. After 3 washes in PBS, samples were mounted on glassslides with Mowiol and examined under a confocal microscope. Forthe detection of CD11b in Kupffer cells, fixation was carried out withacetone (2 minutes at �20°C), and no secondary antibody wasneeded as mouse anti-CD11b was conjugated to FITC.The measure of internalization of fluorescence-ligands in control

and treated hepatocytes were digitalized using an MC80 camera. Forall these experiments, the same settings of the camera and systemwere used. Surface fluorescence and internalization were measuredwith KS100 Kontron Imaging System software. For this purpose, theperiphery of cells was manually defined, and the mean cell-associ-ated fluorescence intensity and the area of the cells were determined.The integrated optical density of the cellular staining was then de-termined, and themeanwas calculated. The results are the average ofat least 3 separate experiments, and the values are expressed as apercentage of the maximum signal-detected: control cells. For eachexperiment a minimum of 100 cells of control and treated groupswere analyzed.

Measurement of Endocytosis of 125I-CT and 125I-ASF. Hepatocytes werewashed briefly in DMEM-HEPES incubated with and without CD(1.5%) (2-hydroxypropyl�-cyclodextrin) for 15 minutes at 37°C orpotassium depleted for the inhibition of clathrin-dependent internal-ization.30,31 125I-CT and 125I-asialofetuin (125I-ASF) were then addedto the cells and incubated for 30minutes at 0°C inDMEM-HEPES 1%BSA or simplified medium (140 mmol/L NaCl, 1 mmol/L CaCl2, 1mmol/L MgCl2, 5.5 mmol/L glucose, 0.1% BSA, 20 mmol/L HEPES,pH7.3) for potassium-depleted cells. After washes in cold medium,cells were incubated at 37°C for 30 minutes with and without CD orsimplified medium. Control for the K� depletion was made by incu-bating the cells in 10mmol/L KCl. Endocytosed 125I-CT and 125I-ASFwere measured and calculated as the percentage of total cell-associ-ated CT and ASF. Surface-bound CT was measured as the amount ofCT that could be released by low pH after 3 rapid washes in mediumat 0°C. The cells were then incubated at 0°C for 5 minutes with lowpH buffer (0.5 mol/L NaCl and 0.2 mol/L acetic acid, pH 2.5), fol-lowed by 1 rapid wash inmedium. Surface-bound ASFwasmeasuredas the amount of ASF that could be released after 3 rapid washes inmedium at 0°C and then incubated at 0°C for 5 minutes with 5mmol/L EGTA in medium. After a quick wash with medium, endo-cytosed ligands were measured as the amount of CT and ASF thatcould not be removed by these treatments. The results are the aver-age of at least 3 separate experiments and the values are expressed asa percentage of the maximum detected signal: control cells.

RESULTS

Biochemical Detection of Caveolin in Rat Liver, in Isolated Hepa-tocytes, and in NPCs. Homogenates from rat liver and fromprimary cultured hepatocytes were used to demonstrate thepresence of caveolin byWestern blotting using a specific anti-caveolin antibody. Figure 1A shows the comparison and dif-ferential expression of caveolin, �-adaptin and asialoglycop-rotein receptor (ASGP-R) in lung and liver homogenates aswell as in hepatocytes in primary culture (HPC). Anti-actinwas used as control of protein loading. Figure 1B shows arepresentative Western blotting comparing the amount ofcaveolin in hepatocytes, NPC crude cellular fraction, and inisolated Kupffer and stellate cells.

HEPATOLOGY Vol. 33, No. 5, 2001 CALVO ET AL. 1261

Isolated cells were also analyzed by immunocytochemistrywith a confocal microscopy. Figure 2 shows the pattern ofcaveolin by confocal microscopy. Note that hepatocytes (Fig.2A) and Kupffer cells (Fig. 2C) have caveolin located at theplasma membrane; in stellate cells caveolin was mainly intra-cellular, punctate, or concentrated in the Golgi region (Fig.2E). No caveolin was observed in endothelial cells (Fig. 2G,arrows). Figure 2G and H show a double labeling of endothe-lial enriched fraction, which also contained some stellate cells,with anti-caveolin (Fig. 2G) and anti-MRC-OX43 (Fig. 2H)antibodies; note that no caveolin can be observed in endothe-lial cells (arrows), labeled with anti-MRC-OX43 antibody (en-dothelial cell marker). Contaminant stellate cells were posi-tively labeled with anti-caveolin (arrowheads).

Electron Microscopy Analysis of Hepatocyte Cell Surface. Wefirst followed a morphologic approach to define the existenceand the location of caveolae in these two systems: intact ratliver and primary cultured hepatocytes. Figure 3 shows ultra-thin cryosections of intact rat liver with caveolae-like struc-tures or vesicles beneath the sinusoidal plasma membrane,amongmicrovilli (Fig. 3A, arrows). When these sections wereused for immunodetection with a polyclonal anti-caveolin an-tibody, followed by an anti-rabbit IgG-conjugated to 10 nmgold, it can be observed that these caveolae were specificallylabeled with gold (Fig. 3B).Ultrathin cryosections from intact livers, immunolabeled

with anti-clathrin or with anti-caveolin antibodies, were usedfor amorphometric analysis to assess the distribution of clath-rin (and clathrin-coated pits/vesicles) and caveolin (andcaveolae or intracellular caveolin) in the sinusoidal (basolat-eral) and the canalicular (apical) plasma membrane domains.Data shown in Table 1 clearly indicated that, whereas clathrinand/or clathrin-coated structures were observed in similarproportion in the basolateral as well as in the canalicularplasma membranes, no caveolae could be found in the cana-licular plasma membrane, and the amount of caveolin de-tected in the basolateral domain was 86% compared with the14% found in the canalicular membranes.

Isolated hepatocytes were cryofixed and prepared for im-muno-electron detection by freeze-substitution and embed-ding in Lowicryl HM20. Ultrathin sections were labeled witha monoclonal anti-caveolin antibody followed by an anti-mouse conjugated to gold (15 nm). Figure 4A shows a caveo-lae decorated with the anti-caveolin on its protoplasmic sur-face. As an additional direct evidence for the presence ofcaveolae in the hepatocyte we used cholera toxin B-gold as ahighly specific probe shown to bind the GM1 in caveolae.32-34Cholera toxin-gold was incubated at 4°C for 2 hours, and thenhepatocytes were warmed and internalization was proceededfor 10 minutes at 37°C. Figure 4B and C show the specific

FIG. 2. Distribution of caveolin in hepatocytes and in NPCs from rat liver.Isolated hepatocytes (A, B), Kupffer cells (C, D), stellate cells (E, F), andendothelial cells (G, H) were prepared for immunocytochemistry and exam-ined in the confocal microscopy. The staining with anti-caveolin shows thathepatocytes and Kupffer cells contained caveolin at the cell surface (A, C); instellate cells, caveolin was mainly located in the Golgi region (E); no caveolincan be observed in endothelial cells (arrows in [G]). Controls of isolationinclude anti-ASGPR, for hepatocytes (B); anti-CD11b, for Kupffer cells (D);autofluorescence, for stellate cells (F); and anti–MRC-OX43, for endothelialcells (H). As detailed in the text, a double immunolabeling was performed incoverslips with endothelial cells. Arrowheads show the caveolin staining incontaminating stellate cells (G) (slightly different from [C] because thesecells are from 1 week); however, these cells were not labeled with anti–MRC-OX43, endothelial marker (H). Bar � 10 �m.

FIG. 1. Expression of caveolin in rat liver. (A) The expression of caveolinwas analyzed by Western blotting in homogenates of lung, liver, and fromHPC. Lung with a high expression of caveolin was used for comparison.Other proteins studied as controls include the ASGPR, specific for hepato-cytes, �-adaptin, and actin as a control of load. (B) The comparison of caveo-lin levels in lysates of hepatocytes (HPC) and NPC; separately, Kupffer cells(KC) and stellate cells (SC) were also analyzed. The same amount (20 �g) ofprotein was loaded in each lane.


location of cholera toxin-gold in caveolae at the cell surface orin caveolae-derived membrane structures.Interestingly, a survey of the hepatocyte cell surface in pri-

mary culture showed significant differences between dorsaland ventral plasma membranes, and a detailed analysis of anumber of sections indicated a polarized distribution of clath-rin-coated pits at the ventral surface and caveolae-like struc-tures at the dorsal membranes. These topological featureswere not resolved in a previous immunofluorescence analysisof isolated hepatocytes, where staining with anti-clathrin,35 orwith anti-caveolin antibodies8 revealed an almost homoge-neous staining around the cell surface.To examine the fine morphology of these specific microdo-

mains involved in the endocytosis, we used the rapid-freeze,deep-etching technique. Figure 5 shows the comparison ofthin sections and rapid-freeze, deep-etch views of the dorsal(Fig. 5A and C) and ventral (Fig. 5B and D) inner surfaces ofa hepatocyte. Themost striking observation was that, whereasthe ventral membrane contained many clathrin-coated pits,sheets, and almost-formed vesicles, the dorsal cell surface hadsmooth invaginations, some with a characteristic striatedcoat, flask-shaped morphology, size, and uniform curvature(Fig. 5C, arrows) of caveolae.

Confirmation of thin-section labeling was obtained by ap-plying the immuno-gold technique to ventral and dorsalmembranes obtained by lysis-squirting and sandwich tech-nique, respectively, from hepatocytes in primary culture at 12hours after plating. Figure 6A shows the freeze-dried replicaof the protoplasmic surface of a hepatocyte dorsal membraneimmunogold labeled for caveolin (10 nm). Caveolin immu-nogold was almost restricted to areas rich in smooth vesicles.In Fig. 6B, some caveolae show a striated coat and gold labelon the cytoplasmic face.Because the polymeric immunoglobulin receptor (pIgR) is

exclusively located in the clathrin-coated pits of the hepato-cyte plasma membrane, we double-labeled the ventral mem-branes with anti-caveolin (15 nm) and anti-pIgR (10 nm) andthen processed them for rapid-freeze, deep etching (Fig. 6Cand D). Arrows indicate labeling with anti-caveolin (15 nm)scattered on irregular structures that lack clathrin lattices;arrowheads indicate clathrin-coated pits and sheets labeledwith anti-pIgR (10 nm).In this study, rat hepatocytes were obtained by liver perfu-

sion and plated for 12 to 16 hours. Such primary culturehepatocytes show two plasma membrane surfaces with differ-ent functions. The substratum adherent plasma membrane

FIG. 3. Identification of caveolae in theplasma membrane of hepatocytes in intact liver.Ultrathin cryo-sections of rat liver show a regionof the sinusoidal plasma membrane with severalvesicle-like structures close or attached to theplasma membrane (arrows) (A). When thesecryo-sections were immunolabeled with anti-caveolin and a secondary antibody coupled withgold (10 nm), it can be observed that gold parti-cles precisely decorate a caveolae-like structure ofapproximately 70 nm in diameter in the sinusoi-dal plasma membrane (B). Bar � 100 nm.


contains different types of adhesion structures (e.g., focal con-tacts), and it participates in secretion or uptake of ligands,etc.36 This cell surface is called the ventral membrane. The topsurface is referred to as the dorsal membrane. The analysis ofthese hepatocytes, by biochemical or morphologic means,showed that no polarization occurred.37

Caveolin-Dependent Endocytosis in Hepatic Cells. To analyzethe functionality of caveolae in hepatocytes, we tested theinternalization of ligands considered specific to caveolae. CTand RBP have been described such that liver- or hepatic-de-rived cell lines can be used as specific ligands of the entrythrough the caveolae.2,5,6,8 Two different approaches were un-

dertaken: (1) ligands were coupled with fluorescent tags tovisualize their internalization by confocal microscopy and (2)selective disruption of caveolae or inhibition of clathrin-de-pendent entry and subsequent quantification of the internal-ized radiolabeled or fluorescence-labeled ligands.Figure 7A shows the internalization of CT coupled to FITC

in isolated hepatocytes at different times. After incubation for10 minutes at 37°C (0 minute chase), labeling was mainlyassociated with the cell surface. However, after 20 or 30 min-utes chase (37°C) CT-FITC was concentrated underneath theplasma membrane and in the perinuclear (Golgi) region,though some labeling remained at the cell surface. The inter-nalization of RBP-FITC was also analyzed in hepatocytes; af-ter 30 minutes RBP-FITC was detected in intracellular punc-tate structures, most probably early and late endosomes (Fig.7D); these structures did not colocalize with the prelysosomal(lysosomal) marker lgp120 (Fig. 7E).When the internalization of RBP was compared with other

ligands that enter via clathrin-coated pits, such as asialofetuin(ASF-FITC), it can be observed that, though both ligandsshared peripheral vesicular structures, ASF appeared to bemore efficiently internalized than RBP (see the high degree ofcolocalization with prelysosomes/lysosomes, immunolabeledwith anti-lgp120) (a couplet of hepatocytes in Fig. 7B and C).These results are in agreement with work published byMalaba et al.,5 and our own studies in isolated rat-liver endo-somal fractions.8Hepatocytes were treated with reagents that interfere with

cholesterol, filipin and cyclodextrin, and therefore modify thestructure and the function of caveolae.38,39 Figure 8 shows theeffect of cyclodextrin on the internalization of RBP, ASF, andLDL. Internalization of RBP was significantly inhibited by cy-clodextrin (88%) (Fig. 8A) (filipin only inhibited the internal-ization of RBP by 26%, data not shown). On the other hand,treatment of hepatocytes with CD hardly affected the internal-ization of ASF (Fig. 8B) or LDL (Fig. 8C). Figure 8D, E, and Fshow representative images of the internalization of RBP-FITC, ASF-FITC, and DiI-LDL, respectively, in hepatocytesafter CD treatment.Finally, we also compared and quantified the entry of

125I-CT and 125I-ASF by the hepatocytes after K� depletion, atreatment known to markedly reduce the rate of endocytosis

FIG. 4. Electron microscopy of isolated rat hepatocytes. (A) Immunoelectron microscopy of cryo-fixed hepatocytes prepared by the freeze-substitutiontechnique. Ultrathin Lowicryl HM20 section shows an isolated hepatocyte immunolabeled with anti-caveolin (15 nm). A single caveolae on the cell surfacespecifically decoratedwith gold particles can be observed (arrow). (B andC) Location of the cholera toxin-gold (17 nm) by pre-embedding procedures. Choleratoxin binds specifically to ganglioside GM1, located in caveolae. Gold labeling was observed in caveolar structures at the cell surface or in intracellularmembrane structures (10 minutes at 37°C). Arrowhead in (B) shows a glycogen granule. Bar � 100 nm.

TABLE 1. Morphometric Analysis of Clathrin and Caveolin Labeling inLiver Parenchymal Cells

Clathrin Membrane PM CCPCC

Vesicles ICM

Au, % of total Basolateral 65 55 68 49Apical 35 45 32 51

Caveolin Membrane PMPM

Caveolae Vesicles ICM

Au, % of total Basolateral 86 100 80 56Apical 14 0 20 44

NOTE. Clathrin and caveolin distribution in hepatocyte membranes of ratliver was analyzed using ultrathin cryosections immunolabeled with anti-caveolin or anti-clathrin X22 antibodies and examined and photographedusing a JEOL JEM 1010 electron microscope. A total of 384 and 230 goldparticles associated withmembranes were counted after clathrin and caveolinlabeling, respectively. A total of 57 (clathrin) and 72 (caveolin) micrographslabeled cryosections (at � 50,000 magnification) were quantified. Gold par-ticles associated with hepatocytes were counted; labeling at the plasma mem-brane (PM) was distinguished from labeling associated with intracellularmembranes (ICM). Linear membranes in the cytoplasm of unknown or-ganelle type were scored as intracellular membranes. On close examination ofthe plasma membranes, gold particles associated with plasma membraneswere distinguished from those associated with caveolae (PM caveolae) orclathrin-coated pits (CCP) attached to the plasma membranes and the caveo-lae or clathrin-coated vesicles that were apparently detached (free) (vesicles).Abbreviations: PM, plasma membrane; CCP, clathrin-coated pits; CC ve-

hicles, clathrin-coated vesicles; ICM, intracellular membrane; PM, caveolae,plasma membrane caveolae.


FIG. 5. Comparison of ultrathinsections and rapid-freeze, deep-etch-ing replicas of the dorsal (A, C) andventral (B, D) hepatocyte plasmamembrane. (A) A detail of the dorsalsurface of an isolated hepatocytewith a fully formed caveolae can beobserved (arrow); (B) a flat region ofthe ventral membrane with someclathrin-coated pits and vesicles (ar-rowheads); (C) the dorsal membranewith several caveolae (arrows); (D)freeze-dry replicas of this ventralsurface show the enrichment ofclathrin-lattices (arrowheads). Bar �100 nm.

FIG. 6. Combination of deep-etching andimmuno-gold techniques with a polyclonalanti-caveolin (15 nm) in dorsal membranes(A) and (B) and a double labelingwith a rabbitanti-caveolin (15 nm) and a mouse anti-pIgR(10 nm) in ventral hepatocyte plasma mem-branes (C, D). Detail of the ventral surfacewith a clathrin-coated pit labeled with anti-pIgR (D). Arrowheads indicate the labeling ofthe anti-pIgR in clathrin-coated pits; arrowsindicate the labeling with anti-caveolin. Bar �100 nm.


of ligands that enter by clathrin-coated pits.40 Internalizationof 125I-ASF was significantly reduced (39%) compared withthe entry of 125I-CT that was almost unaffected (9%) (Fig. 9A).However, endocytosis of 125I-CT in hepatocytes treated withcyclodextrin was reduced by 45% (Fig. 9B).

DISCUSSION

The presence of caveolae in the hepatocyte plasma mem-brane was directly shown by rapid-freeze, deep-etching elec-tron microscopy. Moreover, combined deep-etching and im-munogold techniques revealed the presence of caveolin inprototypic caveolae of the dorsal membrane of primary cul-tured hepatocytes. Interestingly, immunofluorescence analy-sis of isolated nonparenchymal cells also revealed that Kupfferand stellate cells contained caveolin but endothelial cellslacked it. In this study, we achieved a morphologic and func-tional characterization of caveolae in hepatocytes.Hepatocytes are extremely active in endocytic processes,

and its major intracellular trafficking pathways have been ex-amined in detail. The “default” pathway for the delivery of thecontent of endosomes to lysosomes, the transcytotic route

involved in apical secretion but also in the transport of apicalplasma membrane proteins to the bile canalicular plasmamembrane (e.g., GPIs) and the recycling pathway, exempli-fied by the transferrin receptor, have been investigated.41-57

These endocytic pathways begin in clathrin-coated pits at thesinusoidal plasma membrane. However, different ports of en-try have now been shown to be operative in different cells:clathrin-independent and through caveolae.We have described a subcellular fraction from rat liver

highly enriched in caveolin and morphologically identical tothe caveolar-membrane fraction isolated from endothelialcells.58 A comprehensive characterization of this caveolin-en-riched plasma membrane fraction (CEF) showed that besidescaveolin it contained other molecules present in the caveolaeof endothelial, smooth muscle, adipocytes, or fibroblast cells(e.g., PKC, Ras, Raf-1, Mek or SR-BI).9,59 Interestingly, al-though several publications have shown that the caveolin de-tected in liver may originate from caveolae of hepatocytes, noclear-cut morphologic or immunocytochemical evidence hasbeen presented.The squirting technique for exposing the protoplasmic sur-

face of the plasma membrane reveals the topography of largeareas of plasma membrane and allows direct identification ofthose structural elements that are morphologically distinct,such as clathrin sheets, coated vesicles, caveolae or actin-mi-crofilaments. A striking finding in hepatocytes was the pres-ence of multiple caveolae clustered in restricted areas of thedorsal membrane, whereas this face of hepatocyte was poor inclathrin complexes, compared with their ventral face. Thepresence of numerous clathrin complexes on the protoplas-mic surface of the ventral plasma membrane of hepatocyteshas been reported elsewhere.60,61 These authors suggestedthat the first stages of cell adhesion to the substratum aremediated by specialized regions of the plasmamembrane, richin receptors, with clathrin-coated structures, that strongly in-teract with the extracellular matrix components (e.g., colla-gen). On the other hand, the presence of scattered caveolincontaining structures (other than caveolae) in the ventralmembrane of hepatocytes might be related with its function oflinking integrins with signal transduction. Indeed, a recentarticle62 showed that caveolin-1 functions as a membraneadaptor to link the �-subunit of integrin to the tyrosine kinaseFyn. Once activated, Fyn binds to Src, which on phosphory-lation recruits Grb and eventually regulates the Ras-ERK cas-cade. The assembly of signaling molecules surrounding theintegrin family of adhesion receptors remains poorly under-stood. Caveolin binds cholesterol and several signaling mole-cules potentially linked to integrin function, e.g., the Src fam-ily of kinases, although caveolin has not been directlyimplicated in integrin-dependent cell adhesion.63

Finally, although hepatocytes are quiescent cells, it wasshown that there is basal activity of Raf-1 and Mek, restrictedto the early/sorting endocytic compartment,9 and EGF trig-gers a recruitment of caveolin from the plasma membrane tothe early endocytic compartment.64 Thus, the machinery forsignal transduction pre-organized in the caveolae at the cellsurface could be recruited into the early/sorting endosomes,where phosphorylation of Raf-1, Mek, and eventually MAPKtakes place. Binding of EGF could induce the dissociation ofthe caveolin/Ras complex and allow access of Ras to GTP, thusactivating the MAPK cascade.

FIG. 7. Endocytosis in hepatocytes. Binding and internalization of CT inisolated hepatocytes (A). CT-FITCwas incubated for 10minutes (pulse), andthen cells were fixed and prepared for confocal microscopy; labeling wasmainly associated with the cell surface (0 minutes), but after 20 and 30minutes (chase), an increasing amount of labeling can be observed under-neath the plasma membrane and in the Golgi region of the cells. Internaliza-tion of ASF-FITC (B) and RBP-FITC (D) was performed for 30 minutes, andhepatocytes were fixed and immunolabeled with anti-lgp120 antibody (C andE, respectively). It can be observed that ASF-FITC colocalizes with lgp120 atthe prelysosomal compartment (arrow in [B] and [C]); however, little colo-calization can be detectedwith RBP-FITC and lgp120 (arrows in [D] and [E]).Bar � 10 �m.


Is There a Caveolae-Dependent Endocytic Pathway in Hepato-cytes? Evidence for a caveolae-dependent endocytic pathwayhas been reported in various systems.65-70 In hepatocytes, atleast the following possibilities should be considered: (1) theuptake of retinol-binding protein for subsequent transport tostellate cells, (2) “fluid phase” transcytosis (e.g., albumin)71 orthe specific targeting of GPI-anchored proteins from the ba-solateral to the apical plasma membrane, and (3) receptor-mediated endocytosis of HDL by the scavenger receptor SR-BIspecifically detected in caveolae72 or in caveolin-enriched ratliver membrane fractions.59 Furthermore, because the hepa-tocyte is the only site of catabolism of free cholesterol to bileacids, which are secreted to the bile canaliculi along with freecholesterol and phospholipids, one may question the possibleinvolvement of the caveolae in cholesterol transport.Internalization assays were performed using 2 ligands that

internalize via caveolae-mediated processes such as retinol-binding protein and cholera toxin subunit B. Treatment ofcells with drugs that bind or alter cholesterol-rich rafts such ascaveolae significantly inhibit the internalization of these li-gands, indicating that, in hepatocytes, their internalization isdependent on cholesterol-rich rafts. Thus, in this study, usingthe squirting technique for exposing the protoplasmic surfaceof plasma membranes and by biochemical means, we showthat hepatocytes from rat liver have prototypic caveolae andthat these cell surface microdomains are functional.Finally, caveolin was also detected in nonparenchymal rat

liver Kupffer and stellate cells. Its precise cellular or subcellu-lar distribution, compared with hepatocytes, may be differentand is currently under investigation.

FIG. 9. Effects of cyclodextrin and K� depletion on 125I-CT and 125I-ASFinternalization in isolated hepatocytes. Cells were cultured for 12 hours in10% FCS and then incubated for 1 hour in serum-free without (control) andwith cyclodextrin (CD) for 15 minutes. 125I-CT or 125I-ASF were incubated at0°C for 60 minutes and then internalized for 30 minutes at 37°C (A). In parallelexperiments, hepatocytes were K�-depleted and incubated with 125I-CT and 125I-ASF (B). Controls in (B) were incubated with 10 mmol/L KCl.

FIG. 8. Effect of cyclodextrin on the in-ternalization and distribution of RBP-FITC,ASF-FITC, and DiI-LDL in hepatocytes.Histograms represent the quantification ofinternalized ligands: (A), RBP-FITC; (B),ASF-FITC; (C), DiI-LDL (see Materials andMethods for details), and the correspondingpanels on the right are representative fieldsvisualized by confocal microscopy at thesame experimental conditions: (D), RBP;(E), ASF; (F), LDL. Control, untreated; CD,cyclodextrin.


Acknowledgment: The authors are grateful to Esther Titos(Liver Unit, Hospital Clınic, Barcelona) and to Dr. Sofia Perezdel Pulgar (Immunology Unit, Fundacio Clınic, Barcelona)for the advice in the preparation of Kupffer cells and the gift ofstellate cells respectively. They also thank the technical staff,Anna Bosch and Marta Taules, from Serveis Cientıficotecnicsde la Universitat de Barcelona, for skillful assistance in confo-cal and electron microscopy.

REFERENCES

1. Enrich C, EvansWH. The endocytic compartments of normal and regen-erating liver. Sub-Cell Biochem 1993;19:195-222.

2. Montesano R, Roth A, Orci L. Non-coated membrane invaginations areinvolved in binding and internalization of cholera and tetanus toxins.Nature 1982;296:651-653.

3. Henley JR, Krueger EWA, Oswald BJ, McNiven MA. Dynamin-mediatedinternalization of caveolae. J Cell Biol 1998;141:85-99.

4. Glenney JR, Jr. The sequence of the human caveolin reveals identity withVIP21, a component of transport vesicles. FEBS Lett 1992;314:45-48.

5. Malaba L, Smeland S, Senoo H, NorumKR, Berg T, Blomhoff R, KindbergGM. Retinol-binding protein and asialo-orosomucoid are taken up bydifferent pathways in liver cells. J Biol Chem 1995;270:15686-15692.

6. Janicot M, Clot JP, Desbuquois B. Interactions of cholera toxin withisolated hepatocytes. Effects of low pH, Chloroquine and monensin ontoxin internalization, processing and action. Biochem J 1988;253:735-743.

7. Janicot M, Fouquet F, Desbuquois B. Activation of rat liver adenylatecyclase by cholera toxin requires toxin internalization and processing inendosomes. J Biol Chem 1991;266:12858-12865.

8. Pol A, Calvo M, Lu A, Enrich C. The “early-sorting” endocytic compart-ment of rat hepatocytes is involved in the intracellular pathway of caveo-lin-1 (VIP-21). HEPATOLOGY 1999;29:1848-1857.

9. Pol A, Calvo M, Enrich C. Isolated endosomes from quiescent rat livercontain the signal transduction machinery. Differential distribution ofactivated Raf-1 and Mek in the endocytic compartment. FEBS Lett 1998;441:34-38.

10. Tanuma Y, Ohata M, Ito T. Electron microscopic studies on the sinusoi-dal cells in the monkey liver. Arch Histol Jpn 1983;46:401-426.

11. Ohata M, Ito T. Experimental study on the fine structure of chicken liverparenchyme with special references to extrasinusoidal macrophages andsinusoidal blood cells. Arch Histol Jpn 1986;49:83-103.

12. Pitas RE, Boyles J, Mahley RW, Bissell DM. Uptake of chemically modi-fied low density lipoproteins in vivo is mediated by specific endothelialcells. J Cell Biol 1985;100:103-117.

13. Oh P, McIntosh DP, Schnitzer JE. Dynamin at the neck of caveolae me-diates their budding to form transport vesicles by GTP-driven fissionfrom the plasma membrane of endothelium. J Cell Biol 1998;141:101-114.

14. Kurzchalia TV, Parton RG. Membrane microdomain in caveolae. CurrOpin Cell Biol 1999;11:424-431.

15. Anderson RGW. The caveolae membrane system. Annu Rev Biochem1998;67:199-225.

16. Bissell DM, Hammaker LE, Meyer UA. Parenchymal cells from adult ratliver in nonproliferating monolayer culture. I. Functional studies. J CellBiol 1973;59:722-734.

17. Geerts A, Niki T, Hellemans K, De Craemer D, Van Den Berg K, LazouJM, Stange G, et al. Purification of rat hepatic stellate cells by side scatter-activated cell sorting. HEPATOLOGY 1998;27:590-598.

18. Pertoft H, Smedsrod B. Separation and characterization of liver cells. In:Pretlow TG, Pretlow TP, eds. Cell separation, methods and selected ap-plications. Vol 4 New York; Academic, 1987:1-24.

19. Griffiths G, McDowall A, Back R, Dubochet J. On the preparation ofcryosections for immunocytochemistry. J Ultrastruct Res 1984;89:65-78.

20. Tokuyasu KT. A study of positive staining of ultrathin frozen sections. JUltrastruct Res 1978;63:287-307.

21. Escaig J. New instruments which facilitate freezing at 83 and 6K. J Mi-croscopy 1982;126:221-229.

22. Quintana C. Cryofixation, cryosubstitution, cryoembedding for ultra-structural, immunocytochemical and microanalytical studies. Micron1994;25:63-99.

23. Nicol A, Nermut MV. A new type of substratum adhesion structure inNRK cells revealed by correlated interference reflection and electronmicroscopy. Eur J Cell Biol 1987;43:348-357.

24. Sanan DA, Anderson RG. Simultaneous visualization of LDL receptordistribution and clathrin lattices on membranes torn from the uppersurface of cultured cells. J Histochem Cytochem 1991;39:1017-1024.

25. Laemmli UK. Cleavage of structural proteins during the assembly of thehead of bacteriophage T4. Nature 1970;227:680-685.

26. Bradford MM. A rapid and sensitive method for the quantitation of mi-crogram quantities of protein utilizing the principle of protein-dye bind-ing. Anal Biochem 1976;72:248-254.

27. Havel RJ, Eder HA, Bragdon JH. The distribution and chemical compo-sition of ultracentrifugally separated lipoproteins in human serum. J ClinInvest 1955;34:1345-1353.

28. Pitas RE, Inneraty TL, Weinstein JN, Mahley RW. Acetoacetylated li-poproteins used to distinguish fibroblasts from macrophages in vitro byfluorescence microscopy. Arteriosclerosis 1981;1:177-185.

29. Slot JW, Geuze HJ. A new method of preparing gold probes for multiple-labeling cytochemistry. Eur J Cell Biol 1985;38:87-93.

30. Larkin JM, Donzell WC, Anderson RG. Potassium-dependent assemblyof coated pits: new coated pits form as planar clathrin lattices. J Cell Biol1986;103:2619-2627.

31. Heuser JE, Anderson RG. Hypertonic media inhibit receptor-mediatedendocytosis by blocking clathrin-coated pit formation. J Cell Biol 1989;108:389-400.

32. Parton RG. Ultrastructural localization of gangliosides; GM1 is concen-trated in caveolae. J Histochem Cytochem 1994;42:155-166.

33. Critchley DR, Streuli CH, Kellie S, Patel B. Characterization of the chol-era toxin receptor on BALB/c 3T3 cells as a ganglioside similar to, oridentical with, ganglioside GM1. Biochem J 1982;204:209-219.

34. Fishman PH. Role of membrane gangliosides in the binding and action ofbacterial toxins. J Membr Biol 1982;69:85-97.

35. Enrich C, Jones GR, Bordas J, EvansWH. A 220 kDa polypeptide, immu-nolocalized to epithelial tight junctions, is associated with brain clathrinpreparations. Eur J Cell Biol 1989;50:304-312.

36. Nermut MV. Strategy and tactics in electron microscopy of cell surfaces.Electron Microsc Rev 1989;2:171-196.

37. Talami MA, Kappus B, Hubbard A. Repolarization of hepatocytes in cul-tured. HEPATOLOGY 1997;25:167-172.

38. Rodal SK, Skretting G, Garred Ø, Vilhardt F, van Deurs B, Sandvig K.Extraction of cholesterol with methyl-(�-cyclodextrin perturbs forma-tion of clathrin-coated endocytic vesicles. Mol Biol Cell 1999;10:961-974.

39. Orlandi PA, Fishman PH. Filipin-dependent inhibition of cholera toxin:evidence for toxin internalization and activation through caveolae-likedomains. J Cell Biol 1998;141:905-915.

40. Larkin JM, Brown MS, Golstein JL, Anderson RGW. Depletion of intra-cellular potassium arrest coated pit formation and receptor-mediatedendocytosis in fibroblasts. Cell 1983;33:273-285.

41. Geuze HJ, Slot JW, Strous GJAM, Peppard J, von Figura K, Hasilik A,Schwartz AL. Intracellular receptor sorting during endocytosis: compar-ative immunoelectron microscopy of multiple receptors in rat liver. Cell1984;37:195-204.

42. Wall DA, Hubbard AL. Receptor-mediated endocytosis of asialoglyco-proteins by rat liver hepatocytes: biochemical characterization of theendosomal compartments. J Cell Biol 1985;101:2104-2112.

43. Hope CA, Connolly TP, Hubbard AL. Transcellular transport of poly-meric IgA in the rat hepatocyte: biochemical and morphological charac-terization of the transport pathway. J Cell Biol 1985;101:2113-2123.

44. Courtoy PJ. Dissection of endosomes. In: “Intracellular Trafficking ofProteins” (Steer C, Hannover J, eds). Cambrige University Press; Cam-bridge, 1991;103-156.

45. Jackle S, Runquist EA, Brady S, Hamilton RL, Havel RJ. Isolation andcharacterization of three endosomal fractions from the liver of normalrats after lipoproteins loading. J Lipid Res 1991;32:485-498.

46. Jackle S, Runquist E, Miranda-Brady S, Havel RJ. Trafficking of the epi-dermal growth factor receptor and transferrin in three hepatocytic endo-somal fractions. J Biol Chem 1991;266:1396-1402.

47. Sztul E, Kaplin A, Saucan L, Palade G. Protein traffic between distinctplasma membrane domains: isolation and characterization of vesicularcarriers involved in transcytosis. Cell 1991;64:81-89.

48. Barr VA, Hubbard AL. Newly synthesized hepatocyte plasma membraneproteins are transported in transcytotic vesicles in the bile duct-ligatedrat. Gastroenterology 1993;105:554-571.

49. Bu G, Schwartz AL. Receptor-mediated endocytosis. In: Arias IM, BoyerLL, Fausto N, Jakoby WB, Schachter DA, Schfritz DA, eds. The Liver:Biology and Pathobiology. Raven Press, Ltd., New York, 1994;259-274.


50. Oda H, Stockert RJ, Collins C, Wang H, Novikoff PM, Satir P, WolkoffAW. Interaction of the microtubule cytoskeleton with endocytic vesiclesand cytosplasmic dynein in cultured rat hepatocytes. J Biol Chem 1995;270:15242-15249.

51. Baas PC, Di Guglielmo GM, Authier F, Posner BI, Bergeron JJM. Com-partmentalized signal transduction by receptor tyrosine kinases. TrendsCell Biol 1995;5:465-470.

52. Enrich C, Jackle S, Havel RJ. The polymeric immunoglobulin receptor isthe major calmodulin-binding protein in an endosome fraction from ratliver enriched in recycling receptors. HEPATOLOGY 1996;24:226-232.

53. Crawford JM. Role of vesicle-mediated transport pathways in hepatocel-lular bile secretion. Sem Liver Dis 1996;16:169-189.

54. Pol A, Ortega D, Enrich C. Identification and distribution of proteins inisolated endosomal fractions of rat liver: involvement in endocytosis,recycling and transcytosis. Biochem. J 1997;323:435-443.

55. Pol A, Enrich C. Membrane transport in rat liver endocytic pathways:preparation, biochemical properties and functional roles of hepatic en-dosomes. Electrophoresis 1997;18:2548-2557.

56. Enrich C, Pol A, Calvo M, Pons M, Jackle S. Dissection of the multifunc-tional “receptor-recycling” compartment of hepatocytes. HEPATOLOGY1999;30:1115-1120.

57. Verges M, Havel RJ, Mostov KE. A tubular endosomal fraction from ratliver: biochemical evidence of receptor sorting by default. Proc Natl AcadSci U S A 1999;96:10146-10151.

58. Schnitzer JE, McIntosh DP, Dvorak AM, Liu J, Oh P. Separation of caveo-lae from associated microdomains of GPI-anchored proteins. Science1995;269:1435-1439.

59. Calvo M, Enrich C. Biochemical analysis of a caveolae-enriched plasmamembrane fraction from rat liver. Electrophoresis 2000;21:3386-3395.

60. Aggeler J, Heuser JE, Werb Z. Presence of clathrin at adhesion sites inphagocytosing macrophages. 40th Ann Proc Electron Microsc Soc Amer1982.

61. Nermut MV, Williams L, Stamatoglou SC, Bissell DM. Ultrastructure of

ventral membranes of rat hepatocytes spread on type IV collagen. EurJ Cell Biol 1986;42:35-44.

62. Wary KK, Mariotti A, Zurzolo C, Giancotti FG. A requirement for caveo-lin-1 and associated kinase Fyn in integrin signaling and anchorage-dependent cell growth. Cell 1998;94:625-634.

63. Wei Y, Yang X, Liu Q, Wilkins JA, Chapman HA. A role for caveolin andthe urokinase receptor in integrin-mediated adhesion and signaling.J Cell Biol 1999;144:1285-1294.

64. Pol A, Calvo M, Lu A, Enrich C. EGF triggers caveolin redistributionfrom the plasma membrane to the early/sorting endocytic compartmentof hepatocytes. Cell Signalling 2000;12:537-540.

65. Rothberg KG, Heuser JE, Donzell WC, Ying YS, Glenney JR, AndersonRG. Caveolin, a protein component of caveolae membrane coats. Cell1990;68:673-682.

66. Parton RG, Joggerst B, Simons K. Regulated internalization of caveolae.J Cell Biol 1994;127:1199-1215.

67. Parton RG. Caveolae and caveolins. Curr Opin Cell Biol 1996;8:542-548.

68. Smaby JM, Kulkarni VS, Momsen M, Brown RE. The interfacial elasticpacking interactions of galactosylceramides, sphingomyelins, and phos-phatidylcholines. Biophys J 1996;70:868-877.

69. Chang JY, Chavis JA, Liu LZ, Drew PD. Cholesterol oxides inducedprogrammed cell death in microglial cells. Biochem Biophys Res Com-mun 1998;249:817-821.

70. Hailstones D, Sleer LS, Parton RG, Stanley KK. Regulation of caveolin andcaveolae by cholesterol in MDCK cells. J Lipid Res 1998;39:369-379.

71. Wilton JC, Matthews GM. Polarised membrane traffic in hepatocytes.BioEssays 1996;18:229-236.

72. Babitt J, Trigatti B, Rigotti A, Smart EJ, Anderson RGW, Xu S, KriegerM. Murine SR-BI, a high density lipoprotein receptor that mediatesselective lipid uptake, is N-glycosylated and fatty acylated and colo-calizes with plasma membrane caveolae. J Biol Chem 1997; 272:13242-13249.


Morphologic and functional characterization of caveolae in rat liver hepatocytes

Documents