inhibition of endocytosis by dominant-negative dynamin - AJP - Renal ...

6 downloads 56 Views 3MB Size Report
Dec 11, 2001 - oratories (West Grove, PA). cDNA constructs, adenovirus, and .... (Cressington Scientific Instruments, Watford, UK). Sample temperature was ...
Am J Physiol Renal Physiol 282: F998–F1011, 2002. First published December 11, 2001; 10.1152/ajprenal.00257.2001.

Aquaporin-2 localization in clathrin-coated pits: inhibition of endocytosis by dominant-negative dynamin TIAN-XIAO SUN, ALFRED VAN HOEK, YAN HUANG, RICHARD BOULEY, MARGARET MCLAUGHLIN, AND DENNIS BROWN Program in Membrane Biology and Renal Unit, Department of Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts 02114 Received 15 August 2001; accepted in final form 6 December 2001

Sun, Tian-Xiao, Alfred Van Hoek, Yan Huang, Richard Bouley, Margaret McLaughlin, and Dennis Brown. Aquaporin-2 localization in clathrin-coated pits: inhibition of endocytosis by dominant-negative dynamin. Am J Physiol Renal Physiol 282: F998–F1011, 2002. First published December 11, 2001; 10.1152/ajprenal.00257.2001.—Before the identification of aquaporin (AQP) proteins, vasopressin-regulated “water channels” were identified by freeze-fracture electron microscopy as aggregates or clusters of intramembraneous particles (IMPs) on hormonally stimulated target cell membranes. In the kidney collecting duct, these IMP clusters were subsequently identified as possible sites of clathrin-coated pit formation on the plasma membrane, and a clathrin-mediated mechanism for internalization of vasopressin-sensitive water channels was suggested. Using an antibody raised against the extracellular C loop of AQP2, we now provide direct evidence that AQP2 is concentrated in clathrin-coated pits on the apical surface of collecting duct principal cells. Furthermore, by using a fracture-label technique applied to LLC-PK1 cells expressing an AQP2-c-myc construct, we show that AQP2 is located in IMP aggregates and is concentrated in shallow membrane invaginations on the surface of forskolin-stimulated cells. We also studied the functional role of clathrin-coated pits in AQP2 trafficking by using a GTPase-deficient dynamin mutation (K44A) to inhibit clathrin-mediated endocytosis. Immunofluorescence labeling and freeze-fracture electron microscopy showed that dominant-negative dynamin 1 and dynamin 2 mutants prevent the release of clathrin-coated pits from the plasma membrane and induce an accumulation of AQP2 on the plasma membrane of AQP2-transfected cells. These data provide the first direct evidence that AQP2 is located in clathrin-coated pits and show that AQP2 recycles between the plasma membrane and intracellular vesicles via a dynamin-dependent endocytotic pathway. We propose that the IMP clusters previously associated with vasopressin action represent sites of dynamin-dependent, clathrin-mediated endocytosis in which AQP2 is concentrated before internalization.

are a family of transmembrane channel proteins that are present in diverse epithelial cell

membranes and facilitate transepithelial water reabsorption and body fluid homeostasis (1, 19, 39, 40). AQP2 is the vasopressin (VP)-regulated water channel that is expressed in collecting duct principal cells in the kidney (9, 19), although this protein has been located in the testis and epithelial cells lining the vas deferens (47, 56) as well as in the inner ear (45a). Mutations in AQP2 protein result in autosomal nephrogenic diabetes insipidus, confirming the importance of this protein in urinary concentration (15, 34). Many studies have shown that AQP2 is moved from intracellular vesicles to the plasma membrane by hormonally induced, regulated trafficking of AQP2 (9), but in transfected cells, AQP2 also appears to cycle constitutively between the cell surface and intracellular vesicles (23). Before the identification of AQPs, VP-sensitive “water channel” trafficking in the kidney was followed indirectly by freeze-fracture electron microscopy (EM) and by the use of endocytotic tracers such as horseradish peroxidase and FITC dextran (6). Freeze-fracture studies of VP target cells in amphibians and mammals revealed that aggregates or clusters of intramembraneous particles (IMPs) appeared on the plasma membranes of these cells only on hormonal stimulation of transepithelial water flow (13, 28, 33, 62). The area of cell surface occupied by these aggregates was directly correlated with the magnitude of the transepithelial water flow or urinary concentration in most instances (7, 32). Thus it was proposed that VP-sensitive water channels, now identified as AQP2 in mammals (20), may form clusters of IMPs when inserted into the plasma membrane. Subsequent studies revealed that these IMP clusters, which are probably sites at which water channels are aggregated in the plasma membrane, represent sites of clathrin-coated pit formation. Together with horseradish peroxidase tracer studies, this indirect evidence suggested that water channels were internalized by a clathrin-mediated pathway (10, 11, 58). However, since the identification of AQP2 and the production of antibodies against the cytoplasmic COOH terminus of this protein, it has been difficult to

Address for reprint requests and other correspondence: D. Brown, Program in Membrane Biology, Renal Unit, Massachusetts General Hospital East, 149 13th St., Charlestown, MA 02129 (E-mail: [email protected]).

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.

kidney epithelial cells; immunofluorescence; water channels; vasopressin stimulation; immunocytochemistry

AQUAPORINS (AQPs)

F998

0363-6127/02 $5.00 Copyright © 2002 the American Physiological Society

http://www.ajprenal.org

AQP2 IN CLATHRIN-COATED PITS

confirm the clathrin-coated pit localization of AQP2 by immunogold EM. In previously published studies from several groups (including our own) that used immunogold EM, AQP2 was shown to be randomly distributed on the plasma membrane of principal cells, with no obvious accumulation in clathrin-coated structures (27, 31, 43, 48, 49, 52, 63). In the present study, we have used an antibody raised against an extracellular C-loop domain of AQP2 (4, 24) to reexamine the cellsurface localization of AQP2 in collecting duct principal cells. Furthermore, we used the fracture-label technique (17, 18) to localize cell-surface AQP2 in LLC-PK1 cells transfected with an AQP2-c-myc construct (LLCAQP2 cells) (38). We also examined AQP2 localization in LLC-AQP2 cells expressing GTPase-deficient, dominant-negative dynamin. It is well known that dynamin, a 100-kDa GTPase, plays an essential role in the release of nascent clathrin-coated pits from the plasma membrane during endocytosis (30, 45, 53). The precise role of dynamin in the membrane scission process is unclear, but several observations support the idea that dynamin has a mechanochemical function in vesicle release from the membrane (42, 55, 57, 59). Expression of GTPase-deficient, dominant-negative mutants of dynamin inhibits clathrin-mediated endocytosis in mammalian cells (12, 16, 21). This approach has proven to be a useful diagnostic tool for discriminating between dynamin-dependent and dynamin-independent mechanisms of endocytosis and for determining the role of dynamin-dependent endocytosis in a variety of cellular processes. Among the membrane proteins that internalize via a dynamin-dependent pathway are ␤2-adrenergic receptors (22, 64), GLUT4 (35, 36, 51, 61), the Na⫹/H⫹ exchanger NHE3 (14), and epithelial sodium channels (54). In addition, ARF6-mediated endocytosis in Madin-Darby canine kidney cells is dynamin dependent (2). By using these approaches, we now provide direct evidence that AQP2 is located in clathrin-coated pits in collecting duct principal cells and that AQP2 internalization in LLC-AQP2 cells is a dynamin-dependent process. Materials and Methods Peptides, antibodies, and chemicals. Lysine VP was purchased from Sigma (St. Louis, MO). All cell culture reagents, including Geneticin, DMEM, and fetal bovine serum were purchased from GIBCO BRL (Grand Island, NY). The antic-myc monoclonal antibody (9E10) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-clathrin heavy chain monoclonal antibody was purchased from Transduction Laboratories (San Diego, CA). The anti-dynamin rabbit polyclonal antibody was purchased from Covance (Richmond, CA). The anti-dynamin 1 and anti-dynamin 2 rabbit polyclonal antibodies were from Affinity Bioreagents (Golden, CO). Secondary antibodies (FITC and CY3 conjugated) were purchased from Jackson Immunoresearch Laboratories (West Grove, PA). cDNA constructs, adenovirus, and cell lines. The recombinant adenovirus expressing the dynamin 1/K44A mutant was constructed by the method of Becker et al. (3) and was AJP-Renal Physiol • VOL

F999

generously provided by Dr. Jeffrey Pessin (University of Iowa). HEK-293 cells were used as the packaging cells for large-scale virus production. The resultant virus was purified by CsCl2 banding; the final yield was ⬃1012 viral particles, as estimated by measurements at an optical density of 260 nm. The adenovirus was stored at ⫺80°C in 10 mM Tris 䡠 HCl buffer, pH 8.0, containing 1 mM MgCl2 and 10% glycerol. The cDNA for the dynamin/K44A was generously provided by Dr. Sandra Schmid (Scripps Institute, La Jolla, CA). The constructs expressing dynamin 2/wt-green fluorescent protein (GFP), where wt is wild-type, and dynamin 2/K44A-GFP were provided by Dr. Mark McNiven (Mayo Clinic and Foundation, Rochester, MN) and prepared as previously described (36). The preparation of a stable LLC-PK1 cell line expressing AQP2 with a c-myc tag at its COOH terminus (referred to as LLC-AQP2 cells in the remainder of the text) has been described previously (38). Tissue fixation and immunogold detection of AQP2 in principal cells. Three adult Sprague-Dawley rats were injected intraperitoneally with arginine VP (16 ␮g/kg body wt in 1 ml of 0.9% NaCl) 30 min before fixation of the kidneys by intravascular perfusion with paraformaldehyde lysine periodate fixative as previously described (41, 56). After a 5-min perfusion in situ, kidneys were removed and cut into 1- to 2-mm slices that were fixed for a further 6 h in paraformaldehyde lysine periodate, washed, and stored in PBS (10 mM sodium phosphate buffer, 0.9% NaCl, pH 7.4). To examine AQP2 distribution in principal cells in the absence of VP stimulation, male Sprague-Dawley rats were anesthetized with pentobarbital sodium and the kidneys were briefly perfused via the abdominal aorta with Hanks’ balanced salt solution, pH 7.4, at 37°C equilibrated with 5% CO2-95% O2. When the kidneys were cleared of blood (⬃1 min), they were removed, and thin slices (⬃0.5 mm) were quickly cut with a Stadie-Riggs slicer (Thomas Scientific, Swedesboro, NJ). The slices were preincubated at 37°C for 15 min in equilibrated Hanks’ balanced salt solution as previously described. We have previously shown that this procedure results in VP washout and endocytosis of cell-surface AQP2 (4). For the detection of AQP2 in principal cells by EM, tissue slices were cut into thinner sections (50 ␮m) with a Vibratome (Ted Pella, Redding, CA). These final slices were incubated overnight with a previously characterized anti-AQP2 antibody (1:100 dilution) raised against an external epitope of AQP2 from the C-loop (C-GDLAVNALHNNATA) (24) without any detergent permeabilization step. This procedure ensured that only cell-surface (i.e., plasma membrane) AQP2 would be detected, as previously described (4). Some sections were incubated with preimmune rabbit serum or with antibody preabsorbed with the AQP2 peptide used for immunization as immunocytochemical controls. After being rinsed, the slices were incubated overnight in goat anti-rabbit IgG coupled to colloidal gold particles (10 nm diameter; Ted Pella). After a further rinsing, slices were fixed in 1% glutaraldehyde, postfixed in 2% OsO4 in distilled water for 2 h, dehydrated, and flat embedded in Epon 812. Thin sections were stained with uranyl acetate and lead citrate and examined with a Philips CM10 electron microscope (Philips Electronics, Mahwah, NJ). Quantification of AQP2 gold particle labeling was performed on 11 micrographs of cells taken from 1 animal. Images were taken at a final magnification of ⫻45,000. Total apical membrane length was measured in micrometers for each cell, and the length occupied by clathrin-coated pits was also measured. The number of gold particles in coated and noncoated membrane domains was then counted. All proce-

282 • JUNE 2002 •

www.ajprenal.org

F1000

AQP2 IN CLATHRIN-COATED PITS

dures were performed with a Wacom graphics tablet attached to a Macintosh computer running the National Institutes of Health’s Image 4.1 software. The number of gold particles per micrometer of membrane was calculated for each membrane region, and the ratio of gold particle density per micrometer of membrane was obtained for the coated and noncoated domains. Fracture labeling of LLC-AQP2 cells. LLC-AQP2 cells were incubated for 10 min with 10 ␮M forskolin followed by a 10-min incubation in medium alone to wash out forskolin in some cases. Control, forskolin-treated, and “washout” cells were cooled on ice while the medium was replaced with PBS containing 4% paraformaldehyde. Cells were kept on ice, and, after 15 min, the fixative solution was removed. Cells were incubated for 15 min with 50 mM NH4Cl to arrest the fixation process, followed by subsequent incubations of 1 h each with 10, 20, and 30% glycerol in 0.1 M cacodylate buffer (pH 7.4). Fresh 30% glycerol buffer was added for overnight infiltration. Sheets of cells were scraped off the plastic support with a rubber policeman, placed on flat copper specimen holders, and rapidly frozen in N2-cooled Freon 22. Frozen samples were transferred into a freeze-fracture apparatus (Cressington Scientific Instruments, Watford, UK). Sample temperature was raised to ⫺145°C, and samples were cut under vacuum (10⫺7 Torr) to a thickness of ⬃0.1 mm. The specimen temperature was then raised to ⫺120°C, and membranes were fractured with a final passage of the knife blade. Specimens were shadowed with 1.7-nm platinum at 45°, which was followed by a 6-nm carbon coat perpendicular to the fracture plane. The technique described here, SDS-digested freeze-fracture replica labeling, was adapted from the procedure of Fujimoto (17, 18). Postfracture labeling with antibodies was performed after SDS digestion of replicas from lightly fixed cells. Replicas were thawed and floated on 2.5% SDS in PBS or 2.5% SDS in 0.1 M Na⫹/HCO3⫺. Replicas that did not float were discarded. After shaking and prolonged incubation for 2 h, cleaned replicas were floated onto PBS containing 1% defatted and IgG-free bovine serum albumin (blocking solution) for 1 h. They were then floated for 2 h on anti-c-myc antibody (1:1,000) in blocking solution, washed three times for 15 min each with blocking solution, and incubated for 1 h on a drop of goat anti-mouse IgG-gold (15 nm diameter, 1:200 dilution). Replicas were washed in PBS, incubated with 2% glutaraldehyde in PBS for 2 min, and finally washed in distilled water for 2 min. The replicas, taken from two separate experiments, were examined on nickel grids with a Philips CM10 electron microscope. Transient transfection of LLC-AQP2 cells with electroporation. LLC-AQP2 (60% confluent) cells in five 15-cm plates were washed with PBS, pH 7.4, and trypsinized. The cells were pelleted at 1,000 rpm in a tabletop centrifuge, washed two times with DMEM, and resuspended in 0.4 ml of DMEM. The cell suspension was incubated with 25 ␮g of plasmid DNA at room temperature for 5 min before it was transferred to a 0.4-cm electroporation cuvette. Electroporation was performed on a Bio-Rad Gene Pulser II at 960 ␮F and 0.25 kV. Cells were allowed to rest for 10 min at room temperature before being plated onto one 15-cm plate. Sixteen hours after electroporation, cells were counted and replated onto 12 ⫻ 12-mm glass coverslips. The coverslips were incubated for 24 h to give a cell confluency of 80% before use in the AQP2 trafficking experiments. Adenovirus infection. LLC-AQP2 cells were grown on 12 ⫻ 12-mm glass coverslips for 48 h to 60% confluence. Purified adenovirus Ad-dynamin 1/K44A was added to the culture AJP-Renal Physiol • VOL

medium directly at a ratio of five multiplicities of infection. The cells were incubated with adenovirus for 16 h before use. AQP2 trafficking in LLC-AQP2 cells and immunofluorescence detection. LLC-AQP2 cells or the transfectants on glass coverslips were incubated in DMEM without serum for 2 h before each experiment. To stimulate AQP2 movement to the plasma membrane, cells were incubated in 10 nM VP at 37°C for 10 min followed by fixation at room temperature for 20 min with 4% paraformaldehyde in PBS containing 5% sucrose. In the VP-washout experiments, the stimulated cells were washed with warm DMEM six times and incubated at 37°C for 90 min before fixation. After fixation, the cells were permeabilized at room temperature for 5 min with Triton X-100 (0.1%) for staining with anti-c-myc antibody (to detect epitope-tagged AQP2) or with SDS (1%), an antigen retrieval technique (8), for staining with the anti-clathrin antibody. Nonspecific antibody binding sites were blocked by a 10-min incubation with 1% BSA in PBS. The anti-c-myc monoclonal antibody was used to label AQP2 protein in cells, whereas the inherent fluorescence of the GFP tag was used to detect dynamin 2 in double-labeling experiments on LLC-AQP2 cells expressing the dynamin 2-GFP construct. Preliminary data that used an anti-dynamin 1 antibody showed that after adenoviral infection most of the cells in each culture expressed dynamin 1 (an isoform not normally detectable in nonneuronal cells). All primary and secondary antibodies were applied for 1 h at room temperature. Coverslips were mounted on slides with Vectashield (Vector Labs, Burlingham, CA). The cells were examined with a Nikon Eclipse 800 microscope or a Bio-Rad Radiance 2000 confocal microscope. For final printing, the positive staining was pseudocolored in red or green, and images were printed with a Tektronix Phaser 440 dye sublimation printer. Transient transfections and adenovirus infections were both performed on three separate occasions, and data presented are representative of all three sets of experiments in each condition. Thin-section EM of LLC-AQP2 cells expressing dynamin 1/K44A. LLC-AQP2 cells infected 16 h earlier with the dynamin 1/K44A adenovirus were fixed in 2% glutaraldehyde containing 0.1% tannic acid and 0.1% saponin. This procedure allows better visualization of cytoskeletal elements and clathrin-coated pits than fixation in glutaraldehyde alone (44). Cells from two separate infections were postfixed in 2% OsO4, dehydrated in graded ethanols, and flat embedded in Epon 812. Thin sections were examined with a Philips CM10 electron microscope. Conventional freeze fracture of LLC-AQP2 cells expressing dynamin 1/K44A. Conventional freeze fracture was performed on control cells and on cells expressing dominantnegative dynamin 1/K44A after adenoviral infection for 16 h. This method provides an en face view of large areas of plasma membrane on which images of membrane fusion and fission events can be visualized more readily than by thin-section EM. Cells were fixed by immersion for 1 h in 2% glutaraldehyde in 0.1 M sodium cacodylate buffer. After cryoprotection for at least 1 h in 30% glycerol, cells were scraped from the culture dish, and clumps of cells were placed on a copper freeze-fracture support and frozen in Freon 22 cooled by liquid nitrogen. Freeze-fracture replicas were produced as described in Fracture labeling of LLC-AQP2 cells. After removal from the freeze-fracture device, the replicas were cleaned by immersion for 2 h in concentrated sodium hypochlorite bleach, washed three times for 5 min each with distilled water, picked up on copper EM grids, and examined with a Philips CM10 electron microscope. Random areas of plasma membrane from 20 control cells and 20 dynamintransfected cells were photographed at a final magnification

282 • JUNE 2002 •

www.ajprenal.org

AQP2 IN CLATHRIN-COATED PITS

F1001

of ⫻21,000. The number of small circular membrane fusion sites, interpreted to reflect images of the necks of clathrincoated pits, was counted and expressed as the number per square micrometer of membrane surface. RESULTS

Coated pit localization of AQP2 in thin sections of collecting duct principal cells. In ultrathin sections of kidney collecting ducts from VP-treated rats, the apical plasma membranes of principal cells were heavily labeled with gold particles, representing sites of AQP2 antigenicity. As illustrated in Fig. 1, the majority of gold particles were positioned on the outermost surface of the plasma membrane of principal cells, which is consistent with the detection by this antibody of an externally oriented epitope of AQP2 as previously described (4). Although gold labeling in the form of clusters as well as individual gold particles was detected over microvilli and flat intermicrovillar regions of the membrane, many invaginated areas of the membrane were also heavily labeled. Most of these invaginated regions could be readily identified as clathrin-coated pits on the basis of the thick, electron-dense coat of clathrin that characterizes these structures in electron micrographs. The association of clathrin antigenicity with these coated domains on the plasma membrane of collecting duct principal cells has been described previously (11). The appearance of the gold-labeled clathrin-coated membrane domains varied from relatively shallow invaginations to much more pronounced invaginations or so-called “⍀-figures” (Fig. 1, inset). Because no permeabilization step was included in the labeling procedure, intracellular vesicles containing AQP2 were not labeled in these sections, and some clathrin-coated vesicles that may have already detached from the cell surface were also unlabeled. Tissue slices that had been incubated in VP-free Hanks’ buffer for 15 min were also incubated with anti-AQP2 antibodies. As shown in Fig. 1B, very few gold particles were detectable on the apical plasma membrane of principal cells from these animals, indicating that under these VPwashout conditions, very little AQP2 is detectable at the cell surface. Furthermore, antibody labeling was not detectable on collecting duct intercalated cells that do not express AQP2 (4). Finally, no staining was seen in sections of tissue incubated in the absence of primary antibody or with antibody that was preincubated with the immunizing peptide. Quantification of gold particle labeling in VP-treated rats revealed that, on average, the ratio of gold particle labeling per micrometer of membrane length in coated pits vs. noncoated apical membrane was 3.15 ⫾ 0.73, showing a clear (⬎3-fold) enrichment of AQP2 in coated membrane domains. However, there was variability in the enrichment factor among different cells, ranging from no enrichment to an almost eightfold enrichment of AQP2 in coated pits, perhaps reflecting different activity states of the epithelial cell population. The total number of coated pits examined was 45, AJP-Renal Physiol • VOL

Fig. 1. Immunogold localization of aquaporin-2 (AQP2) in clathrincoated pits on the apical plasma membrane of kidney collecting duct principal cells from the inner stripe of a vasopressin-treated animal (A and inset) and on a principal cell from a kidney slice after vasopressin washout in vitro (B). Anti-AQP2 antibody directed against an extracellular epitope of the water channel was used to detect AQP2 antigenic sites on nonpermeabilized tissue slices by using a preembedding labeling procedure. A: after vasopressin treatment, gold particles are scattered over the external leaflet of the plasma membrane but are concentrated in an invaginating clathrincoated pit (arrow); another large cluster of gold particles to the left of the arrow is probably also associated with a coated pit that has been tangentially sectioned and is not clearly visible in the plane of this section. Inset: examples of more deeply invaginated clathrin-coated pits that also contain numerous gold particles, indicating the presence of AQP2 in these structures. B: in contrast, very few gold particles are present on the apical plasma membrane of principal cells after vasopressin washout. Bars ⫽ 0.2 ␮m (A and inset); 0.5 ␮m (B).

and the average membrane length of these pits was 0.27 ⫾ 0.02 ␮m. AQP2 localization in LLC-AQP2 cells by fracture labeling. Because the technique of fracture labeling is not easily adapted to whole tissues such as the kidney, we sought to examine the relationship between cellsurface AQP2 distribution and morphological features

282 • JUNE 2002 •

www.ajprenal.org

F1002

AQP2 IN CLATHRIN-COATED PITS

previously associated with water channel clusters (see the beginning of this study) in AQP2-transfected epithelial cells. Cells were first treated with forskolin (10 ␮M for 10 min) to induce the appearance of AQP2 on the basolateral cell surface, as previously described (38). Cells were then washed in buffer alone to stimulate the reinternalization of AQP2 by endocytosis, as previously described (37). Little or no gold label was found on the plasma membranes of nonstimulated cells (not shown). After forskolin stimulation, the amount of cell-surface labeling was increased, but the gold label was randomly dispersed on the basolateral membrane surface (4.8 gold particles/␮m2; Fig. 2). After forskolin washout for 10 min, the AQP2 gold labeling became concentrated over membrane invaginations and IMP clusters (200 gold particles/␮m2), although randomly dispersed gold particles were still present on other areas of the membrane (4.1 gold particles/␮m2; Fig. 3, A–C). These IMP clusters resembled the loose clusters of IMPs previously described on apical membranes of VP-stimulated principal cells (10, 25, 26). In addition to the gold-labeled clusters that were present in flat membrane domains (Fig. 3B), others were associated with deeper invaginations that are suggestive of endocytotic domains of the plasma membrane (81.3 gold particles/␮m2; Fig. 3C). In this image, only 2.7 gold particles/␮m2 were present on noninvaginated membrane domains. Expression of dynamin 1 and 2 in LLC-AQP2 cells. To show the functional role of dynamin in clathrinmediated endocytosis, we overexpressed dynamin 1 and 2 mutants in LLC-AQP2 cells. The expression was verified by immunocytochemistry, as shown in Fig. 4. Dynamin 1, a neuronal specific isoform, is not endog-

enously expressed in LLC-PK1 cells. Staining of LLCAQP2 cells with anti-dynamin 1 antibody showed no background (data not shown), whereas dynamin 1 virus-infected cells showed a strong positive staining (Fig. 4A). Dynamin 2, a ubiquitous isoform, is expressed endogenously in LLC-PK1 cells but at levels too low to be detected by immunocytochemistry (not shown). When cells were electroporated with a dynamin 2-GFP construct, ⬃50% of the cells overexpressed dynamin 2-GFP (Fig. 4B). This overexpressed dynamin 2-GFP was also detectable in an identical pattern by using an antibody against the dynamin 2 COOH terminus (data not shown). Expression of dynamin 1/K44A inhibits clathrincoated pit detachment from the plasma membrane. LLC-AQP2 cells were infected with the dynamin 1/K44A virus. They were treated with VP for 10 min and subsequent washout of VP for up to 90 min. Immunofluorescence labeling with anti-clathrin antibodies showed that clathrin was distributed mainly in a perinuclear and cytoplasmic location under baseline conditions, with little clear definition of the plasma membrane labeling (Fig. 5A). Expression of dynamin 1/K44A resulted in the appearance of a well-defined peripheral membrane pattern of staining in addition to intracellular staining (Fig. 5B). Membrane staining was also present after VP treatment (Fig. 5D) and VP washout (Fig. 5F). In the control cells, clathrin distribution was not clearly modified by VP stimulation and washout (Fig. 5, C and E). The punctate, linear clathrin staining pattern in dynamin 1/K44A-expressing cells is suggestive of the accumulation of clathrin-coated pits that could not be released from the plasma membrane after the process

Fig. 2. Fracture labeling of the basolateral plasma membrane of an LLCAQP2 cell that had been stimulated for 10 min with forskolin. After freezefracturing and platinum-carbon replication, membranes were incubated with an anti-c-myc antibody (to detect epitope-tagged AQP2 in the transfected LLC-AQP2 cells) followed by goat anti-mouse IgG coupled to 15-nm colloidal gold. Electron-dense gold particles (some are indicated with arrows) are scattered randomly over the surface of the plasma membrane. Small globular structures are intramembraneous particles (IMPs) that represent integral membrane proteins in freezefracture replicas. Smooth, flat islandlike structures represent patches of membrane from an adjacent cell that detached during the fracture process. Bar ⫽ 0.5 ␮m.

AJP-Renal Physiol • VOL

282 • JUNE 2002 •

www.ajprenal.org

AQP2 IN CLATHRIN-COATED PITS

F1003

Fig. 3. Label-fracture detection of AQP2 on the plasma membrane of LLCAQP2 cells after forskolin stimulation for 10 min followed by 10-min washout. During this time, AQP2 begins to be internalized. A: gold particles are clearly concentrated over distinct IMP clusters on the plasma membrane (arrows), although a few particles remain outside the clusters. B: at higher magnification, the presence of gold particles associated with IMP clusters can be readily appreciated (arrows). C: in some membranes, gold particles were concentrated over small pits or depressions in the membrane, with the freeze-fracture appearance of clathrincoated pits (arrows). Bars ⫽ 1 ␮m (A); 0.2 ␮m (B and C).

of invagination. This was confirmed by conventional and freeze-fracture EM. In thin sections of dynamin 1/K44A-expressing cells, many images of invaginated coated pits with long narrow necks ⬃30–40 nm in diameter were found at the plasma membrane both by conventional EM (Fig. 6) and by freeze-fracture EM (Fig. 6, insets). Such images were rarely if ever found in noninfected cells. Freeze-fracture EM was performed to quantify this increase in cell-surface invaginations. Many small, circular depressions corresponding to the fractured necks of these “frozen” clathrin-coated pits were found in the dynamin 1/K44A-expressing cells AJP-Renal Physiol • VOL

(Fig. 7). The average diameter of these fractured necks (Fig. 7, inset) was 30.7 ⫾ 6.8 nm (n ⫽ 28), and the number per square micrometer of membrane was 2.1 ⫾ 0.4 (n ⫽ 11 cells) in the dynamin 1/K44Aexpressing cells. Invaginations with similar morphology were not found in any of the 12 control cells that were examined. In cross fractures, vesicles attached to the plasma membrane by necks of various lengths could be detected (Fig. 6, inset). Similar structures were found on both the apical and the basolateral plasma membranes of the dynamin 1/K44A-expressing LLC-AQP2 cells.

282 • JUNE 2002 •

www.ajprenal.org

F1004

AQP2 IN CLATHRIN-COATED PITS

Fig. 4. Expression of dynamin 1 (Dyn1) and dynamin 2 in LLCAQP2 cells. A: LLC-AQP2 cells were incubated with dynamin 1 adenovirus for 16 h and were fixed and stained with a dynamin 1-specific antibody. More than 95% of the cells expressed dynamin 1, as seen by their red (CY3 fluorophore) staining pattern. B: LLCAQP2 cells were transfected with a dynamin 2-green fluorescent protein (Dyn2-GFP) construct by electroporation. Forty hours after electroporation, the cells were fixed and subjected to confocal immunofluorescence microscopy to examine the intrinsic green fluorescence of dynamin 2-GFP. Thirty to fifty percent of the cells expressed dynamin 2. Bars ⫽ 10 ␮m.

Expression of dynamin 1/K44A inhibits AQP2 endocytosis. We previously demonstrated that AQP2 recycles between intracellular vesicles and the plasma membrane in response to hormonal stimulation and withdrawal (37, 38). As previously described, AQP2 translocated from a perinuclear vesicle pool (Fig. 8A) to the plasma membrane of LLC-AQP2 cells (Fig. 8C) on VP treatment (10 nM for 10 min), and it returned to intracellular vesicles 90 min after washout of VP (Fig. 8E). In contrast, AQP2 in dynamin 1/K44A-expressing cells showed a constitutive membrane location under baseline conditions (Fig. 8B), after VP treatment (Fig. 8D), and after VP washout (Fig. 8F). Expression of dynamin 2/K44A also induces an accumulation of clathrin and AQP2 at the cell surface. The data obtained from dominant-negative mutant dynamin 1/K44A infection support the idea that dynamin AJP-Renal Physiol • VOL

plays a key role in AQP2 endocytosis and recycling from plasma membrane. However, dynamin 1 is an isoform of the protein that is restricted to neural tissue. Only dynamin 2, the ubiquitous isoform, is endogenously expressed in kidney epithelial cells. Thus we characterized the effect of dynamin 2 on clathrin localization and AQP2 endocytosis. We found that dynamin 2/K44A also inhibits AQP2 internalization, similar to the effect of the dynamin 1/K44A mutant on this process. The constructs for dynamin 2 expression were tagged with GFP, a 30-kDa peptide, that potentially could affect the function of dynamin 2 in the endocytotic process. Therefore, we transfected LLC-AQP2 cells with dynamin 2/wt-GFP to determine any potential effect of the GFP tag on dynamin 2 activity. In all of the figures showing dynamin 2 transfections, the effect of dynamin 2 on clathrin and AQP2 distribution can be easily seen by comparing their localization (red) in adjacent cells that either do or do not express the dynamin 2/wt-GFP construct (identified by its green fluorescence). The wild-type construct had no effect on clathrin localization (Fig. 9, A–C), whereas dynamin 2/K44A transfection resulted in an accumulation of clathrin on the cell surface (Fig. 9, D–F). Furthermore, wild-type dynamin transfection did not affect the intracellular localization of AQP2 (Fig. 10, A–C), and a normal VP-induced translocation of AQP2 to the cell surface was also seen in these transfected cells (Fig. 10, D–F). In contrast, dynamin 2/K44A transfection resulted in a cell-surface accumulation of AQP2 without VP stimulation (Fig. 11, A–C). Nontransfected cells retained a perinuclear distribution of AQP2. After VP stimulation, all cells showed the expected plasma membrane localization of AQP2 (Fig. 11, D–F). DISCUSSION

The present data demonstrate that AQP2 can be associated with clathrin-coated pits in the apical plasma membrane of collecting duct principal cells, which supports earlier indirect evidence that AQP2 is internalized by a clathrin-mediated mechanism (10, 11, 58). The fracture-label data obtained by using LLCAQP2 cells are also in agreement with this hypothesis, which is strongly supported by our finding that expression of dominant-negative, GTPase-deficient dynamin 1 and dynamin 2 in these cells causes accumulation of AQP2 on the cell surface secondary to an inhibition of endocytosis. In this model system, AQP2 was associated with membrane invaginations with the freezefracture characteristics of clathrin-coated pits (10, 46) after forskolin stimulation and washout. Furthermore, IMP clusters that resemble those induced by VP in principal cells, and which are proposed to be sites of AQP aggregation at the cell surface (6, 10, 11, 25, 26), were also labeled with gold particles in stimulated LLC-AQP2 cells. Thus our data support the hypothesis that VP-sensitive water channels are internalized by clathrin-mediated endocytosis and that IMP clusters

282 • JUNE 2002 •

www.ajprenal.org

AQP2 IN CLATHRIN-COATED PITS

F1005

Fig. 5. Expression of dynamin 1/K44A (K44A/Dyn1) causes accumulation of clathrin at the plasma membrane. Nontransfected LLC-AQP2 cells and cells expressing dynamin 1/K44A were stimulated with vasopressin (VP; C and D) followed by VP washout (W/O; E and F). Cells were then fixed and labeled with anti-clathrin heavy chain monoclonal antibody. Compared with the mainly perinuclear localization of clathrin in control (CON) cells not expressing the dynamin mutant (A, C, and E), dynamin 1/K44A-expressing cells show a marked staining at the cell periphery under all conditions examined (B, D, and F). VP treatment did not result in a detectable increase in cell-surface expression of clathrin in control cells (C). Bar ⫽ 10 ␮m.

are hallmarks of the membrane domains in which these water channels are concentrated before internalization. However, some IMP clusters were also detectable in nonstimulated LLC-AQP2 cells, where they were not labeled with gold particles (not shown). This indicates that although IMP clusters can contain AQP2, their appearance does not depend absolutely on the presence of AQP2 at the cell surface. This is consistent with the fact that clathrin-mediated endocytosis of many other membrane proteins occurs in these and other cells and that the generalized clustering of

proteins destined for internalization in these regions results in IMP clusters that are detectable by freezefracture microscopy (46). This may indicate that not all of the individual IMPs that make up any given IMP cluster represent the AQP2 protein, although more direct evidence will be required to confirm this idea. The label-fracture technique used in the present study was initially described by Fujimoto (17, 18), who made use of a rapid-freezing protocol for replicating and gold-labeling nonfixed cells. The use of fixed material was believed to preclude adequate cleaning of Fig. 6. Electron microscopy showing the morphology of clathrin-coated pits in the plasma membrane of LLC-AQP2 cells expressing K44A dominant-negative dynamin 1. Use of tannic acidsaponin fixation (see MATERIALS AND METHODS) provides greater contrast and staining to the coated pits than conventional fixation and staining. Main panel: thin section of two invaginating coated pits (arrows) on the apical plasma membrane of the LLCAQP2 cell. The pits have a rounded appearance, and in 1 case, the neck connecting the pit to the cell surface is clearly visible. Insets: 2 images of clathrin-coated pits seen by freezefracture electron microscopy. Right inset: invagination with a long, narrow neck. Bars ⫽ 0.5 ␮m (0.2 ␮m, insets).

AJP-Renal Physiol • VOL

282 • JUNE 2002 •

www.ajprenal.org

F1006

AQP2 IN CLATHRIN-COATED PITS

Fig. 7. Freeze-fracture electron microscopy of dynamin 1/K44A-infected LLC-AQP2 cells. P-face of an apical membrane of a dynamin 1/K44A-expressing cell shows cross-fractured membrane projections (large oval structures) and many small IMPs. In addition, many small pits and depressions are present throughout the membrane P-face (arrows). These round, 30- to 40-nm-diameter depressions are shown at higher magnification in the inset and represent cross fractures through the necks of the many clathrin-coated pits that are “frozen” at the cell surface because of dynamin 1/K44A overexpression. These “neck” structures were virtually absent from control, nontransfected cells (see text for quantification). Bars ⫽ 0.3 ␮m (0.1 ␮m, inset).

Fig. 8. Expression of dynamin 1/K44A inhibits AQP2 endocytosis. LLC-AQP2 cells expressing dynamin 1/K44A (B, D, and F) as well as noninfected control cells (A, C, and E) were subjected to vasopressin (VP) stimulation and subsequent washout before they were fixed and stained with an anti-cmyc antibody. A: in the control cells, AQP2 is located in perinuclear vesicles under baseline conditions. It relocates predominantly to the plasma membrane after VP stimulation (C) and is retrieved to an intracellular compartment after VP washout (E). In dynamin 1/K44A-expressing cells, AQP2 is located on the plasma membrane under all three conditions (B, D, and F). Bar ⫽ 10 ␮m.

AJP-Renal Physiol • VOL

282 • JUNE 2002 •

www.ajprenal.org

AQP2 IN CLATHRIN-COATED PITS

F1007

Fig. 9. Expression of dynamin 2/K44A-GFP, but not dynamin 2/wt-GFP, where wt is wild-type, induces cell-surface accumulation of clathrin. LLC-AQP2 cells were transiently transfected with dynamin 2/wt-GFP (A–C) or dynamin 2/K44A-GFP (D–F) and stained for clathrin (red). A and D: green cells are dynamin 2-transfected cells, whereas those cells that appear dark are nontransfected. A–C: clathrin is located predominantly in a perinuclear and cytoplasmic pattern in both nontransfected cells and in the cells expressing dynamin 2/wt-GFP. D–F: in contrast, clathrin shifts to a punctate distribution along the plasma membrane of cells expressing dynamin 2/K44A-GFP (arrows, E). E and F: nontransfected cells surrounding the green transfected cells retain the normal perinuclear pattern of clathrin distribution.

platinum-carbon replicas, which is necessary to allow labeling reagents to access the bottom surface of the replica, while preserving the antigenicity of attached proteins. The SDS treatment was used to digest away all biological material from the replica, with the exception of a very thin layer of protein that is in direct contact with the platinum. This thin, transparent layer of protein contains the antigenic sites that are detected by the subsequent antibodygold-labeling procedure. We found, however, that brief fixation of LLC-AQP2 cells with 4% paraformaldehyde allowed preservation of antigenicity while not preventing the digestion of underlying cellular material. However, attempts to reproduce this procedure on the kidney have so far proven unsuccessful in our laboratory because of the difficulty in removing the thicker layer of biological material that remains attached to the replica after the fracturing procedure has been performed. Thus antibodies and gold reagents cannot gain access to the underside of the replica and cannot label the proteins that are present in the replicated membrane. AJP-Renal Physiol • VOL

The reason that previous immunogold labeling procedures have failed to reveal a concentration of AQP2 in clathrin-coated pits at the cell surface is uncertain, but it could have two possible explanations. First, clathrin-coated domains are more difficult to positively identify in tissues that have been prepared for immunogold EM by embedding in hydrophilic resins such as Lowicryl K4M (11). Second, it is possible that antigenic sites on the COOH-terminal cytoplasmic tail of AQP2 are masked by interaction with clathrin and the accessory proteins present in these cell membrane domains. Because previously applied antibodies were raised against the extreme COOH terminus of the AQP2 protein, masking of this epitope would result in the negative findings. The use of an anti-external domain antibody in the present studies has now allowed the unequivocal localization of AQP2 in clathrin-coated pits in collecting duct principal cells. However, attempts to use this antibody to detect cell-surface AQP2 in LLC-AQP2 cells have so far proven unsuccessful for unknown reasons. The antibody was raised against a part of the external sequence of AQP2 that contains the

282 • JUNE 2002 •

www.ajprenal.org

F1008

AQP2 IN CLATHRIN-COATED PITS

Fig. 10. Expression of dynamin 2/wt-GFP does not interfere with VP-induced AQP2 trafficking. Dynamin 2/wt-GFP transfected LLC-AQP2 cells were treated with VP for 10 min before fixation. AQP2 was labeled with a red fluorophore (CY3), and the cells expressing dynamin 2/wt-GFP were detected by the intrinsic green fluorescence of GFP. In cells expressing dynamin 2/wt-GFP (green cells; A), AQP2 is located in intracellular vesicles under baseline conditions in both transfected and nontransfected cells (B). C: merged image. D–F: after VP treatment, AQP2 shifts from intracellular vesicles to a membrane location after VP treatment in the cells overexpressing dynamin 2/wt -GFP. Bar ⫽ 10 ␮m.

N-glycosylation site, and it is possible that in LLCAQP2 cells a cell-specific pattern of AQP2 glycosylation restricts antibody access to its binding site on the protein. Interestingly, the same antibody does recognize AQP2 in many intracellular vesicles in LLC-AQP2 cells (24). These data imply that the cell-surface form of this water channel is antigenically distinct from at least some of the intracellular protein in these transfected cells. In support of our immunocytochemical data, the use of dominant-negative dynamin constructs demonstrated that AQP2 is internalized from the cell surface in a dynamin-sensitive pathway. Both dynamin 1/K44A (in adenovirus-infected cells) and dynamin 2/K44A (in transiently transfected cells) caused an accumulation of clathrin and AQP2 on the plasma membrane in the absence of VP or forskolin stimulation. VP induced no further cell-surface accumulation of AQP2 in K44A dynamin-overexpressing cells, and, after agonist washout, AQP2 remained on the cell surface. In control LLC-AQP2 cells, cellsurface AQP2 labeling was weak or absent under baseline conditions, greatly increased 10 min after AJP-Renal Physiol • VOL

VP stimulation, and was restored to a mainly intracellular location 90 min after agonist washout, as expected (37). However, no apparent increase in cellsurface staining for clathrin was obvious in VPstimulated cells that were not expressing dominantnegative dynamin isoforms. However, we have previously shown that an increase in the number of cell-surface clathrin-coated pits can be quantified after VP stimulation of collecting duct principal cells (10). It is possible that the lower resolution afforded by immunofluorescence staining was not able to detect any VP-induced increase in plasma membranecoated pits in the presence of a large amount of cytoplasmic staining. The more marked increase in plasma membrane-associated coated pits resulting from dynamin/K44A overexpression was, in contrast, readily detectable by immunofluorescence. Taken together, our present data provide compelling evidence that AQP2 recycles constitutively between an intracellular pool and the cell surface, as previously proposed (23). This recycling pathway is blocked at the level of endocytosis from the cell surface on expression of dominant-negative dy-

282 • JUNE 2002 •

www.ajprenal.org

AQP2 IN CLATHRIN-COATED PITS

F1009

Fig. 11. Expression of dynamin 2/K44A-GFP induces an accumulation of AQP2 on the plasma membrane. LLC-AQP2 cells transiently expressing dynamin 2/K44A-GFP were stimulated with VP for 10 min and then stained for AQP2 with an anti-c-myc antibody. AQP2 was labeled in red, and dynamin 2/K44A-GFP is green. A–C: in the transfected cells, AQP2 shifts to a membrane location even under baseline conditions (arrows; B), whereas the surrounding nontransfected cells show a perinuclear staining for AQP2. D–F: after VP stimulation, AQP2 relocates to the plasma membrane of nontransfected cells and the staining in the transfected cells remains on the membrane. Bar ⫽ 10 ␮m.

namin, resulting in the observed accumulation even in the absence of agonist stimulation. Although dynamin has also been implicated in caveolin-mediated endocytosis (29, 50), we have found no morphological evidence that AQP2 is located in caveolae either in LLC-AQP2 cells or in principal cells (not shown). Indeed, we have previously reported that caveolin is a basolateral protein in principal cells (5), whereas AQP2 is predominantly apical in most regions of the kidney collecting duct. Thus our present data provide strong evidence to support the involvement of clathrin-mediated endocytosis in AQP2 recycling. These data also indicate that AQP2 accumulation at the cell surface can be a result of decreased endocytosis, at least under the relatively long (16 h) time scale of the present experiments. The relative importance of the rates of exo- and endocytosis in the acute cell-surface accumulation of AQP2 that occurs in response to VP stimulation remains to be determined. In summary, a role for clathrin-coated pits in water channel recycling was proposed in earlier studies on the basis of circumstantial evidence. We now provide AJP-Renal Physiol • VOL

direct immunocytochemical evidence in collecting duct principal cells in situ and freeze-fracture evidence in AQP2-expressing cells in culture that AQP2 is located in clathrin-coated pits and IMP aggregates, respectively. Furthermore, we show that AQP2 accumulates in the plasma membrane of LLC-AQP2 cells overexpressing dominant-negative dynamin 1 and dynamin 2. These data indicate that AQP2 internalization from the cell surface occurs by means of a clathrin-mediated endocytotic process. The data also demonstrate that AQP2 accumulation at the cell surface occurs after endocytosis is blocked by overexpression of GTPasedeficient dynamin isoforms in LLC-AQP2 cells, indicating that AQP2 constitutively recycles between an intracellular compartment and the plasma membrane in these cells. We are grateful to Dr. Jeffrey Pessin, University of Iowa, for providing the dynamin wild-type and mutant constructs that were used in this study. This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-38452 (to D. Brown) and DK-55864 (to A. V. Hoek). R. Bouley and T.-X. Sun were partially

282 • JUNE 2002 •

www.ajprenal.org

F1010

AQP2 IN CLATHRIN-COATED PITS

supported by fellowship grants from the National Kidney Foundation. REFERENCES 1. Agre P, Brown D, and Nielsen S. Aquaporin water channels: unanswered questions and unresolved controversies. Curr Opin Cell Biol 7: 472–483, 1995. 2. Altschuler Y, Liu S, Katz L, Tang K, Hardy S, Brodsky F, Apodaca G, and Mostov K. ADP-ribosylation factor 6 and endocytosis at the apical surface of Madin-Darby canine kidney cells. J Cell Biol 147: 7–12, 1999. 3. Becker TC, Noel RJ, Coats WS, Gomez-Foix AM, Alam T, Gerard RD, and Newgard CB. Use of recombinant adenovirus for metabolic engineering of mammalian cells. Methods Cell Biol 43: 161–189, 1994. 4. Bouley R, Breton S, Sun T-X, McLaughlin M, Nsumu NN, Lin HY, Ausiello DA, and Brown D. Nitric oxide and atrial natriuretic factor stimulate cGMP-dependent membrane insertion of aquaporin 2 in renal epithelial cells. J Clin Invest 106: 1115–1126, 2000. 5. Breton S, Lisanti MP, Tyszkowski R, McLaughlin M, and Brown D. Basolateral distribution of caveolin-1 in the kidney. Absence from H⫹-ATPase-coated endocytic vesicles in intercalated cells. J Histochem Cytochem 46: 205–214, 1998. 6. Brown D. Structural-functional features of antidiuretic hormone-induced water transport in the collecting duct. Semin Nephrol 11: 478–501, 1991. 7. Brown D, Grosso A, and DeSousa RC. Correlation between water flow and intramembrane particle aggregates in toad epidermis. Am J Physiol Cell Physiol 245: C334–C342, 1983. 8. Brown D, Lydon J, McLaughlin M, Stuart-Tilley A, Tyszkowski R, and Alper S. Antigen retrieval in cryostat tissue sections and cultured cells by treatment with sodium dodecyl sulfate (SDS). Histochem Cell Biol 105: 261–267, 1996. 9. Brown D and Nielsen S. The cell biology of vasopressin action. In: The Kidney (6th ed.), edited by Brenner BW. Orlando, FL: Saunders, 2000, p. 575–594. 10. Brown D and Orci L. Vasopressin stimulates the formation of coated pits in rat kidney collecting ducts. Nature 302: 253–255, 1983. 11. Brown D, Weyer P, and Orci L. Vasopressin stimulates endocytosis in kidney collecting duct principal cells. Eur J Cell Biol 46: 336–341, 1988. 12. Ceresa BP, Kao AW, Santeler SR, and Pessin JE. Inhibition of clathrin-mediated endocytosis selectively attenuates specific insulin receptor signal transduction pathways. Mol Cell Biol 18: 3862–3870, 1998. 13. Chevalier J, Bourguet J, and Hugon JS. Membrane-associated particles: distribution in frog urinary bladder epithelium at rest and after oxytocin treatment. Cell Tissue Res 152: 129–140, 1974. 14. Chow CW, Khurana S, Woodside M, Grinstein S, and Orlowski J. The epithelial Na⫹/H⫹ exchanger, NHE3, is internalized through a clathrin-mediated pathway. J Biol Chem 274: 37551–37558, 1999. 15. Deen PM and Knoers NV. Vasopressin type-2 receptor and aquaporin-2 water channel mutants in nephrogenic diabetes insipidus. Am J Med Sci 316: 300–309, 1998. 16. Duan D, Li Q, Kao AW, Yue Y, Pessin JE, and Engelhardt JF. Dynamin is required for recombinant adeno-associated virus type 2 infection. J Virol 73: 10371–10376, 1999. 17. Fujimoto K. Freeze-fracture replica electron microscopy combined with SDS digestion for cytochemical labeling of integral membrane proteins. Application to the immunogold labeling of intercellular junctional complexes. J Cell Sci 108: 3443–3449, 1995. 18. Fujimoto K. SDS-digested freeze-fracture replica labeling electron microscopy to study the two-dimensional distribution of integral membrane proteins and phospholipids in biomembranes: practical procedure, interpretation and application. Histochem Cell Biol 107: 87–96, 1997. 19. Fushimi K and Marumo F. Water channels. Curr Opin Nephrol Hypertens 4: 392–397, 1995. AJP-Renal Physiol • VOL

20. Fushimi K, Uchida S, Hara Y, Hirata Y, Marumo F, and Sasaki S. Cloning and expression of apical membrane water channel of rat kidney collecting tubule. Nature 361: 549–552, 1993. 21. Gaborik Z, Szaszak M, Szidonya L, Balla B, Paku S, Catt KJ, Clark AJ, and Hunyady L. ␤-arrestin- and dynamindependent endocytosis of the AT1 angiotensin receptor. Mol Pharmacol 59: 239–247, 2001. 22. Gagnon AW, Kallal L, and Benovic JL. Role of clathrinmediated endocytosis in agonist-induced down-regulation of the ␤-2-adrenergic receptor. J Biol Chem 273: 6976–6981, 1998. 23. Gustafson CE, Katsura T, McKee M, Bouley R, Casanova JE, and Brown D. Recycling of AQP2 occurs through a temperature- and bafilomycin-sensitive trans-Golgi-associated compartment. Am J Physiol Renal Physiol 278: F317–F326, 2000. 24. Gustafson CE, Levine S, Katsura T, McLaughlin M, Aleixo MD, Tamarappoo BK, Verkman AS, and Brown D. Vasopressin regulated trafficking of a green fluorescent protein-aquaporin 2 chimera in LLC-PK1 cells. Histochem Cell Biol 110: 377–386, 1998. 25. Harmanci MC, Kachadorian WA, Valtin H, and DiScala VA. Antidiuretic hormone-induced intramembranous alterations in mammalian collecting ducts. Am J Physiol Renal Fluid Electrolyte Physiol 235: F440–F443, 1978. 26. Harmanci MC, Stern P, Kachadorian WA, Valtin H, and DiScala VA. Vasopressin and collecting duct intramembranous particle clusters: a dose-response relationship. Am J Physiol Renal Fluid Electrolyte Physiol 239: F560–F564, 1980. 27. Hayashi M, Sasaki S, Tsuganezawa H, Monkawa T, Kitajima W, Konishi K, Fushimi K, Marumo F, and Saruta T. Expression and distribution of aquaporin of collecting duct are regulated by vasopressin V2 receptor in rat kidney. J Clin Invest 94: 1778–1783, 1994. 28. Hays RM, Franki N, and Ding G. Effects of antidiuretic hormone on the collecting duct. Kidney Int 31: 530–537, 1987. 29. Henley JR, Krueger EW, Oswald BJ, and McNiven MA. Dynamin-mediated internalization of caveolae. J Cell Biol 141: 85–99, 1998. 30. Hinshaw JE. Dynamin and its role in membrane fission. Annu Rev Cell Dev Biol 16: 483–519, 2000. 31. Ishikawa S, Saito T, Fujita N, Sasaki S, and Marumo F. Expression and subcellular localization of water channel aquaporin-2 in conscious rats. Kidney Int 61 Suppl: S6–S9, 1997. 32. Kachadorian WA, Coleman RA, and Wade JB. Water permeability and particle aggregates in ADH-, cAMP-, and forskolin-treated toad urinary bladder. Am J Physiol Renal Fluid Electrolyte Physiol 253: F120–F125, 1987. 33. Kachadorian WA, Wade JB, and DiScala VA. Vasopressin: induced structural changes in toad bladder luminal membrane. Science 190: 67–69, 1975. 34. Kamsteeg EJ, Deen PM, and van Os CH. Defective processing and trafficking of water channels in nephrogenic diabetes insipidus. Exp Nephrol 8: 326–331, 2000. 35. Kao AW, Ceresa BP, Santeler SR, and Pessin JE. Expression of a dominant interfering dynamin mutant in 3T3L1 adipocytes inhibits GLUT4 endocytosis without affecting insulin signaling. J Biol Chem 273: 25450–25457, 1998. 36. Kao AW, Yang C, and Pessin JE. Functional comparison of the role of dynamin 2 splice variants on GLUT-4 endocytosis in 3T3L1 adipocytes. Am J Physiol Endocrinol Metab 278: E825– E831, 2000. 37. Katsura T, Ausiello DA, and Brown D. Direct demonstration of aquaporin-2 water channel recycling in stably transfected LLC-PK1 epithelial cells. Am J Physiol Renal Fluid Electrolyte Physiol 270: F548–F553, 1996. 38. Katsura T, Verbavatz JM, Farinas J, Ma T, Ausiello DA, Verkman AS, and Brown D. Constitutive and regulated membrane expression of aquaporin 1 and aquaporin 2 water channels in stably transfected LLC-PK1 epithelial cells. Proc Natl Acad Sci USA 92: 7212–7216, 1995. 39. King LS and Agre P. Pathophysiology of the aquaporin water channels. Annu Rev Physiol 58: 619–648, 1996.

282 • JUNE 2002 •

www.ajprenal.org

AQP2 IN CLATHRIN-COATED PITS 40. Knepper MA, Verbalis JG, and Nielsen S. Role of aquaporins in water balance disorders. Curr Opin Nephrol Hypertens 6: 367–371, 1997. 41. Lencer WI, Weyer P, Verkman AS, Ausiello DA, and Brown D. FITC-dextran as a probe for endosome function and localization in kidney. Am J Physiol Cell Physiol 258: C309–C317, 1990. 42. Marks B, Stowell MH, Vallis Y, Mills IG, Gibson A, Hopkins CR, and McMahon HT. GTPase activity of dynamin and resulting conformation change are essential for endocytosis. Nature 410: 231–235, 2001. 43. Marples D, Knepper MA, Christensen EI, and Nielsen S. Redistribution of aquaporin-2 water channels induced by vasopressin in rat kidney inner medullary collecting duct. Am J Physiol Cell Physiol 269: C655–C664, 1995. 44. Maupin P and Pollard TD. Improved preservation and staining of HeLa cell actin filaments, clathrin-coated membranes, and other cytoplasmic structures by tannic acid-glutaraldehyde saponin fixation. J Cell Biol 96: 51–62, 1983. 45. McNiven MA, Cao H, Pitts KR, and Yoon Y. The dynamin family of mechanoenzymes: pinching in new places. Trends Biochem Sci 25: 115–120, 2000. 45a.Merves M, Bobbitt B, Parker K, Kishore BK, and Choo D. Developmental expression of aquaporin 2 in the mouse inner ear. Laryngoscope 110: 1925–1930, 2000. 46. Montesano R, Perrelet A, Vassalli P, and Orci L. Absence of filipin-sterol complexes from large coated pits on the surface of culture cells. Proc Natl Acad Sci USA 76: 6391–6395, 1979. 47. Nelson RD, Stricklett P, Gustafson C, Stevens A, Ausiello D, Brown D, and Kohan DE. Expression of an AQP2 Cre recombinase transgene in kidney and male reproductive system of transgenic mice. Am J Physiol Cell Physiol 275: C216–C226, 1998. 48. Nielsen S, Chou CL, Marples D, Christensen EI, Kishore BK, and Knepper MA. Vasopressin increases water permeability of kidney collecting duct by inducing translocation of aquaporin-CD water channels to plasma membrane. Proc Natl Acad Sci USA 92: 1013–1017, 1995. 49. Nielsen S, DiGiovanni SR, Christensen EI, Knepper MA, and Harris HW. Cellular and subcellular immunolocalization of vasopressin-regulated water channel in rat kidney. Proc Natl Acad Sci USA 90: 11663–11667, 1993. 50. Oh P, McIntosh DP, and Schnitzer JE. Dynamin at the neck of caveolae mediates their budding to form transport vesicles by GTP-driven fission from the plasma membrane of endothelium. J Cell Biol 141: 101–114, 1998. 51. Omata W, Shibata H, Suzuki Y, Tanaka S, Suzuki T, Takata K, and Kojima I. Subcellular distribution of GLUT4 in

AJP-Renal Physiol • VOL

52.

53. 54. 55.

56.

57.

58.

59. 61.

62. 63.

64.

F1011

Chinese hamster ovary cells overexpressing mutant dynamin: evidence that dynamin is a regulatory GTPase in GLUT4 endocytosis. Biochem Biophys Res Commun 241: 401–406, 1997. Sabolic I, Katsura T, Verbavatz JM, and Brown D. The AQP2 water channel: effect of vasopressin treatment, microtubule disruption, and distribution in neonatal rats. J Membr Biol 143: 165–175, 1995. Sever S, Damke H, and Schmid SL. Garrotes, springs, ratchets, and whips: putting dynamin models to the test. Traffic 1: 385–392, 2000. Shimkets RA, Lifton RP, and Canessa CM. The activity of the epithelial sodium channel is regulated by clathrin-mediated endocytosis. J Biol Chem 272: 25537–25541, 1997. Smirnova E, Shurland DL, Newman-Smith ED, Pishvaee B, and van der Bliek AM. A model for dynamin self-assembly based on binding between three different protein domains. J Biol Chem 274: 14942–14947, 1999. Stevens AL, Breton S, Gustafson CE, Bouley R, Nelson RD, Kohan DE, and Brown D. Aquaporin-2 is a vasopressin-independent, constitutive apical membrane protein in rat vas deferens. Am J Physiol Cell Physiol 278: C791–C802, 2000. Stowell MH, Marks B, Wigge P, and McMahon HT. Nucleotide-dependent conformational changes in dynamin: evidence for a mechanochemical molecular spring. Nat Cell Biol 1: 27–32, 1999. Strange K, Willingham MC, Handler JS, and Harris HW Jr. Apical membrane endocytosis via coated pits is stimulated by removal of antidiuretic hormone from isolated, perfused rabbit cortical collecting tubule. J Membr Biol 103: 17–28, 1988. Sweitzer SM and Hinshaw JE. Dynamin undergoes a GTPdependent conformational change causing vesiculation. Cell 93: 1021–1029, 1998. Volchuk A, Narine S, Foster LJ, Grabs D, De Camilli P, and Klip A. Perturbation of dynamin II with an amphiphysin SH3 domain increases GLUT4 glucose transporters at the plasma membrane in 3T3–L1 adipocytes. Dynamin II participates in GLUT4 endocytosis. J Biol Chem 273: 8169–8176, 1998. Wade JB. Membrane structural studies of the action of vasopressin. Federation Proc 44: 2687–2692, 1985. Yamamoto T, Sasaki S, Fushimi K, Kawasaki K, Yaoita E, Oota K, Hirata Y, Marumo F, and Kihara I. Localization and expression of a collecting duct water channel, aquaporin, in hydrated and dehydrated rats. Exp Nephrol 3: 193–201, 1995. Zhang J, Ferguson SS, Barak LS, Menard L, and Caron MG. Dynamin and ␤-arrestin reveal distinct mechanisms for G protein-coupled receptor internalization. J Biol Chem 271: 18302–18305, 1996.

282 • JUNE 2002 •

www.ajprenal.org