Alix (AIP1) is a vasopressin receptor (V2R)-interacting protein that increases lysosomal degradation of the V2R

Xianhua Yi , Richard Bouley, Herbert Y. Lin, Shaliha Bechoua, Tian-xiao Sun, Elisabetta del Re, Toshi Shioda, Malay K. Raychowdhury, Hua A. J. Lu, Abdul B. Abou-Samra, Dennis Brown, Dennis A. Ausiello


The vasopressin type 2 receptor (V2R) is a G protein-coupled receptor that plays a central role in renal water reabsorption. Termination of ligand (vasopressin) stimulation is an important physiological regulatory event, but few proteins that interact with the V2R during downregulation after vasopressin (VP) binding have been identified. Using yeast two-hybrid screening of a human kidney cDNA library, we show that a 100-kDa protein called ALG-2-interacting protein X (Alix) interacts with the last 29 amino acids of the V2R COOH terminus. This was confirmed by pull-down assays using a GST-V2R-COOH-tail fusion protein. Alix was immunolocalized in principal cells of the kidney, which also express the V2R. The function of the Alix-V2R interaction was studied by transfecting Alix into LLC-PK1 epithelial cells expressing V2R-green fluorescent protein (GFP). Under basal conditions, V2R-GFP localized mainly at the plasma membrane. On VP treatment, V2R-GFP was internalized into perinuclear vesicles in the nontransfected cells. In contrast, V2R-GFP fluorescence was virtually undetectable 2 h after exposure to VP in cells that coexpressed Alix. Western blotting using an anti-GFP antibody showed marked degradation of the V2R after 2 h in the presence of VP and Alix, a time point at which little or no degradation was detected in the absence of Alix. In contrast, little or no degradation of the parathyroid hormone receptor was detectable in the presence or absence of Alix and/or the PTH ligand. The VP-induced disappearance of V2R-GFP was abolished by chloroquine, a lysosomal degradation inhibitor, but not by MG132, a proteosome inhibitor. These data suggest that Alix increases the rate of lysosomal degradation of V2R and may play an important regulatory role in the VP response by modulating V2R downregulation.

  • LLC-PK1 renal epithelial cells
  • G protein-coupled receptors
  • receptor-mediated endocytosis
  • antidiuretic hormone
  • yeast two-hybrid screening
  • protein interaction

the termination of ligand-induced stimulation is an important phenomenon that regulates and limits cellular responses to hormonal action. The internalization of G protein-coupled receptor (GPCR) ligand complexes is complex and involves several interacting proteins at different points in the pathway. Binding of ligand triggers the phosphorylation of the receptor and coordinates the involvement of protein kinase-anchoring proteins, specific kinases, and in some cases small GTP-binding proteins (8, 24, 37). After phosphorylation has occurred, β-arrestin binds to the cytoplasmic domain of the receptors, which undergo internalization via clathrin-mediated endocytosis (25). The budding vesicle is pinched off from the membrane by a dynamin-dependent mechanism (51), and the newly formed vesicle is rapidly uncoated and fuses with early endosomes. Some receptors reappear at the cell surface rapidly, whereas others are sent for degradation in either the lysosomal compartment or the proteosome pathway.

Recent studies showed that the pathway taken by different receptors after internalization is dictated by specific protein interactions and specific sorting signals. The Na+/H+ exchanger regulatory factor (NHERF) that binds the PDZ-site at the COOH tail of the β-adrenergic receptor is important in conferring “rapid” recovery of functional membrane-binding sites after downregulation (12). Both the α-adrenergic and β-adrenergic receptors also interact with the α-subunit of the eukaryotic initiation factor 2B (22). ARAP1 is a protein that interacts with the COOH tail of the type 1 angiotensin II receptor to promote the reappearance of this receptor at the cell surface (14). In contrast, the α1b-adrenergic receptor is downregulated after interaction between its COOH-terminal cytoplasmic domain and the gC1q-R, a multifunctional protein (49). The COOH terminus of the delta opioid receptor binds the GTP-binding protein (G protein)-coupled receptor-associated sorting protein (GASP) that promotes lysosomal trafficking (38, 47), whereas interaction of the dopamine receptor COOH tail with the G-α-interacting protein known as GIPC prevents lysosomal degradation of the receptor (18). Recently, proteins from the endoplasmic reticulum (ER) such as DRIP 78 have been shown to interact with receptors including the dopamine D1 receptor and reduce surface ligand-binding sites. Receptor activity modifying protein (RAMPS) forms a stable dimer with several receptors in the ER and Golgi apparatus, and these complexes are maintained from translocation to the cell surface until their subsequent lysosomal degradation (31, 36).

The vasopressin type 2 receptor (V2R) that is involved in stimulating water reabsorption in the collecting duct shows a slow recycling pathway after stimulation with ligand (17). Like the β-adrenergic receptor, the V2R also interacts with β-arrestin, and is internalized in a clathrin-coated pit-dependent manner (4, 32). However, the V2R reappears at the cell surface very slowly after VP stimulation, reaching prestimulation levels only after several hours. Recently, we have shown that a significant fraction of intracellular V2R is targeted directly to lysosomes for degradation after ligand-induced internalization (3). This V2R degradation pathway was also reported independently by Robben et al. (35). This may partially explain the “slow” recycling phenotype of the V2R. However, the mechanism(s) by which V2R trafficking is regulated is not fully understood, and, in particular, the proteins involved in directing the V2R toward the lysosomal degradation compartment are still elusive.

Based on existing data from other receptors, some of which are mentioned above, we hypothesized that the COOH-terminal domain of the V2R plays an important role in receptor targeting and fate determination. In this study, we screened for proteins binding to the COOH terminus of the human V2R by the yeast two-hybrid method. Five of the clones that were isolated encode proteins that are portions of the same gene: ALG-2-interacting protein X (Alix). Alix is a 100-kDa cytoplasmic protein containing two distinct domains, an NH2-terminal domain (amino acids 1–434) and a COOH-terminal domain (amino acid 468–869) (30, 45). The COOH tail of Alix contains a proline-rich domain (PRD) that has several possible SH3-binding sites that interact with SH3p4, SH3p8, and SH3p13, proteins collectively known as endophilins (33). When the same Alix COOH tail is coexpressed with endophilins, intracellular vacuoles/tubules are drastically enlarged and can even form very large spherical vacuoles with a diameter of up to several micrometers (7). It was also reported that an AIP1 homolog in yeast, Npi3/Bro1, plays a significant role in the downregulation, internalization, and degradation of a general amino acid permease Gap1 of Saccharomyces cerevisiae in a ubiquitin-dependent manner (40).

We show here that the V2R COOH tail interacts with Alix in yeast two-hybrid and GST-fusion pull-down experiments. Overexpression of Alix causes an increased rate of degradation of the V2R only after interaction of the receptor with ligand. Finally, localization of Alix in principal cells in the kidney, the site where V2R normally resides and functions to regulate vasopressin-mediated water transport, indicates that Alix is poised to affect V2R degradation also in vivo. We propose that Alix-mediated degradation is a novel mechanism for regulating the fate of the V2R, and perhaps other GPCRs after ligand induced internalization.


Cells, reagents, and antibodies.

Arginine vasopressin (VP), parathyroid hormone (134), BSA, chloroquine, alcohol, and inorganic salts were purchased from Sigma (St. Louis, MO). All cell culture reagents including DMEM, Lipofectamine 2000, PBS, and FBS were from GIBCO BRL (Grand Island, NY). Complete protease inhibitor cocktail was purchased from Roche Diagnostic (Indianapolis, IN). Carbobenzoxy-l-leucyl-l-leucyl-l-leucinal (MG132) was purchased from Calbiochem (San Diego, CA). Human embryonic kidney cells (HEK293 cells) were from ATCC (Manassas, VA). The LLC-PK1a cells were originally provided by Dr. Steven Krane [Arthritis Unit, Massachusetts General Hospital (MGH)]. The green fluorescent protein-tagged parathyroid hormone receptor COOH-terminal construct (PTHR-GFP) was provided by Dr. Abdul Abou-Samra (Endocrine Unit, MGH). Both GST, GST-agarose, and protein A-sepharose were obtained from Pierce (Rockford, IL). Secondary CY3-conjugated antibodies were from Jackson Immunoresearch Laboratories (West Grove, PA). Polyclonal antibodies against the vacuolar ATPase (H+V-ATPase) E-subunit and the aquaporin-2 (AQP2) COOH terminus have been used and described previously (5, 16). Polyclonal and monoclonal antibodies against FLAG and secondary horseradish peroxidase-conjugated antibodies were purchased from Sigma. Anti-GFP antibody and Lysotracker Red DND-99 were bought from Molecular Probes (Eugene, OR). Animal experiments were approved by the MGH Institutional Committee on Research Animal Care, in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

DNA constructs for two-hybrid screening and GST fusion proteins.

The last 29 amino acids from the COOH-terminal region of human V2R were cloned into GAL4:DBD vector pAS2–1 (Clontech, Palo Alto, CA) as a bait construct for yeast two-hybrid screening. The coding sequence of the entire V2R COOH-terminal cytoplasmic domain construct (amino acids 342–370) was also subcloned into the bacterial expression vector pET41a (Novagen, Madison, WI) as a fusion protein with a GST tag on its NH2 terminus. The DNA constructs encoding the GST V2R COOH-terminal fusion proteins (GST-V2R) were sequenced to confirm their reading frame and predicted composition. Expression of GST-V2R in Escherichia coli (Rosetta DE3 strain) was induced by isopropyl-1-thio-β-galactopyranoside (IPTG; 0.2 mM) at 37°C for 4–5 h. Bacteria were lysed by lysozyme (1 mg/ml) and 0.5% NP-40 in STE buffer (150 mM NaCl, 1 mM EDTA, 50 mM Tris·HCl, pH 7.4). The bacterial supernatant was allowed to bind to the glutathione-agarose column at 4°C for 40 min with gentle rocking. After being washed twice with washing buffer 1 (STE buffer with NaCl increased to 500 mM, 0.5% NP-40, and 0.1% Triton X-100) and twice with STE buffer, the GST fusion protein was allowed to either bind to the glutathione column or was eluted with elution buffer (5 mM reduced glutathione, 50 mM Tris·HCl, pH 8.0) following a standard protocol recommended by the manufacturer. Meanwhile, GST protein was also expressed and purified as described above. Both purified GST and GST-V2R were subsequently used for pull-down assays.

Yeast two-hybrid screening and yeast two-hybrid assays.

Yeast two-hybrid screening was performed in AH109 S. cerevisiae (MATa, trp1–901, leu2–3, 112, ura3–52, his3–200, gal4D, gal80D, LYS2::GAL1UAS-GAL1TATA-HIS3, GAL2UAS-GAL2TATA-ADE2, URA3: MEL1UAS-MEL1TATA-lacZ, Clontech) that contains three reporters, ADE2, HIS3, and MEL1, responding to GAL4 strongly and specifically. KC8 E. coli (hsdR, leuB600, trpC9830, pyrF:Tn5, hisB463, lacDX74, strA, galU,K) was employed to segregate the cDNA library plasmid from the bait plasmid by nutritional selection. XL1-Blue E. coli (Stratagene, La Jolla, CA) was used for routine plasmid propagation. Yeast media were purchased from Biogen. AH109 S. cerevisiae can be propagated on nonselective solid agar YPD plates or in liquid YPD medium. Selective media are used for selection of yeast transformants and/or the detection of BD-X/AD-Y interactions. In our system, the BD-expressing plasmid contains TRP1, while the AD-expressing plasmid contains LEU2. The three other prototrophic markers, HIS3, ADE2, and MEL1, used in our experiments serve as reporters. The medium-stringency selection medium lacks tryptophan, leucine, and histidine; the high-stringency selection medium lacks tryptophan, leucine, histidine, and adenine, but contains X-α-gal (20 μg/ml). 3-Aminotriazole (2.5 mM) was added to all media to inhibit HIS3 leakiness. M9/-Leu was used for nutritional selection of KC8 E. coli transformants to rescue the AD:cDNA plasmid.

Yeast two-hybrid screening was initiated with the construction of a BD:V2R cytoplasmic domain construct as mentioned above. The BD:V2R plasmid was first tested for His leakiness and autoactivation of other reporters as recommended. The GAL4 AD:human kidney cDNA library constructed in the yeast plasmid pACT2 was purchased from Clontech. The AD:cDNA plasmid library in pACT2 was amplified according to the supplier's instructions. A total amount of 5 mg of library plasmid DNA was purified from amplified E. coli colonies using CsCl banding.

The AH109 containing the pAS2–1/ALCT plasmid was transformed with 0.2 mg of the AD:cDNA plasmid library using a modified lithium acetate method. The transformed cells were plated onto medium-stringency selection medium and incubated at 30°C for 14 days. Positive colonies were patched onto high-stringency selection medium plates and incubated at 30°C until sufficient growth was achieved and colonies turned blue. Positive colonies were maintained on high-stringency selection media until further characterized. Colonies that activated all of the reporter genes in the AH109 stain were further analyzed. The AD:cDNA plasmid encoding the interacting protein was isolated from yeast cells using the standard lycatase/alkaline method and was transformed to KC8 E. coli by electroporation. The colonies containing the AD:cDNA plasmid were rescued on M9/-Leu selection medium plates, and the plasmid DNA was isolated and sent for sequencing.

A yeast two-hybrid assay was performed to further confirm the positive interaction of pAS2–1/ALCT and AD:cDNA plasmids as follows. AD:cDNA plasmids isolated from the primary screen were used to cotransform AH109 with pAS2–1/ALCT to confirm activation of reporters; the potential interacting clones were also cotransformed to AH109 with BD vector alone to test for false positives via direct interaction with GAL4:DBD. Transformed cells were plated on high-stringency selection medium. AD:cDNA clones that were positive with pAS2–1/ALCT but negative with pAS2–1 were further characterized by DNA sequencing. Clones that failed to grow in the reconfirmation screen or showed a strong interaction with GAL4:DBD were not pursued further.

Antibody production and purification.

A peptide (DEIKKERESLENDL) corresponding to amino acids 561–574 of mouse Alix was synthesized by the Peptide Core Facility at MGH. A previous study reported using a similar sequence to raise a polyclonal antibody against Alix (45). An extra cysteine residue was added to the NH2 terminus for later construction of an affinity column for antibody purification. Part of the synthesized peptide was conjugated to keyhole limpet hemocyanin (KLH) for immunization. Affinity purification of antibodies from whole rabbit serum was carried out using a SulfoLink Kit from Pierce (Rockford, IL) according to the manufacturer's instructions. The specificity of the polyclonal antibody was verified by Western blotting.

Cell culture.

HEK293 cells were cultured in DMEM supplemented with 10% heat-inactivated FBS, penicillin (100 units/ml), and streptomycin (100 mg/ml). Stable transfection of V2R-GFP into LLC-PK1a cells, a cell line that expresses a low amount of endogenous receptor (3), was carried out by incubating cells (60% confluent) with a mixture of Lipofectamine (15 μl) and 4 μg of V2R-GFP plasmid DNA at 37°C for 4 h. After selection in medium, resistant individual colonies were isolated, and their [3H]-VP-binding abilities were analyzed as described previously (4). Several clones (LLC-V2R-GFP cells) were isolated and characterized. LLC-V2R-GFP cells and our existing LLC-PK1a cells stably expressing FLAG-V2R (LLC-FLAG-V2R cells) (4) were routinely maintained and passaged in DMEM supplemented with 10% heat-inactivated FBS and 1 mg/ml of Geneticin (G418).

Preparation of cell culture lysates.

Mammalian cell expression of Alix was carried out by transient transfection of HEK293 cells with FLAG-Alix. The empty vector (pcDNA3 vector) was used as a control. Cells were transfected when the cells were 60% confluent. For transfection, 15 μl of Lipofectamine were mixed with 4 μg of plasmid DNA incubated at room temperature for 30 min, before being applied to the cells. Cells were incubated with the transfection mixture for 4 h at 37°C. Transfected cells were harvested and solubilized for 20 min at 4°C in a cold RIPA buffer (50 mM Tris·HCl, pH 7.4, 150 mM NaCl, 0.5% sodium deoxycholate, 1% NP-40, 0.1% SDS, and protease inhibitor cocktail, Roche). Cell debris was spun down at 14,000 g, and the total protein concentration of the supernatant was determined using the BCA protein assay. Proteins from the supernatant were used to study protein expression and degradation levels in the presence of VP by Western blotting and were also used for the GST pull-down assay.

LLC-V2R-GFP cells were transfected as described above with pcDNA3 (as a control) or FLAG-Alix. Forty-eight hours after transfection, the cells were stimulated with 1 μM VP for 1 or 2 h. Cells were harvested and lysed in lysis buffer (50 mM HEPES, pH 7.5, 0.5% NP-40, 250 mM NaCl, 2 mM EDTA, 10% glycerol and protease inhibitor cocktail). Cells were spun down at 14,000 g for 20 min at 4°C after 30-min shaking at 4°C. Western blot analysis of the supernatants was performed using a rabbit polyclonal anti-GFP antibody. HEK cells were transiently transfected with V2R-GFP (6 μg) or with GFP-tagged parathyroid hormone receptor (PTHR-GFP; 6 μg). V2R-GFP and PTHR-GFP were cotransfected with or without FLAG-Alix (10 μg). After 48 h, transfected cells were stimulated with either 1 μM VP or 1 μM PTH (1–34) for 1–4 h before sampling as described above

Renal papilla preparation.

Kidney membranes were prepared as previously described (41). In brief, kidney papilla from six mice were minced and homogenized in iced-cold hypotonic buffer (5 mM Tris·HCl, pH 7.4, 3 mM MgCl2, 1 mM EDTA) containing protease inhibitors (Complete, Boehringer-Mannheim). After centrifugation, the homogenate was centrifuged at 35,000 g for 20 min, and the pellet was resuspended in RIPA buffer containing protease inhibitors before use in Western blot analysis.

Pull-down assay using GST-V2R.

Purified GST-V2R and GST alone were used for pull-down studies using HEK293 cell lysates expressing FLAG-Alix. Before the final pull-down assay, the cell lysates were precleared on a glutathione column at 4°C for 40 min. Approximately 25 mg of the GST-V2R fusion protein were incubated with the glutathione column, and 1 mg of precleared cell lysate was applied at 4°C with gentle rocking for 2 h. The glutathione column was washed four times with PBS containing 1 M NaCl, Triton X-100 (0.1%), and protease inhibitors to remove nonspecifically bound proteins. Purified GST protein was used as a control in side-by-side experiments. The pulled down complexes were subjected to SDS-PAGE and immunoblot analysis using a monoclonal antibody against FLAG.

Protein analysis with SDS-PAGE and Western blotting.

SDS-PAGE was conducted as described by the manufacturer (Invitrogen, Carlsbad, CA). Reduced samples from cell lysates (10 μg) or kidney membrane preparations (10 μg) were loaded onto each lane of a standard NUPAGE 4–12% Bis-Tris Gel and then electroblotted on polyvinylidene difluoride (PVDF) filters (Bio-Rad, Hercules, CA). The filters were blocked with 5% nonfat milk in PBS-T buffer (PBS buffer; 10 mM sodium phosphate, 2 mM potassium phosphate, 2.7 mM KCl, 137 mM NaCl, pH 7.4, with 0.1% Tween 20) by shaking for 2 h at room temperature (or overnight at 4°C) and were further incubated with either a polyclonal anti-GFP (0.4 μg/ml) or anti-FLAG antibody (3.5 μg/ml) in PBS-T buffer containing 1% nonfat milk, respectively, for 1 h. After being washed three times in PBS-T, the PVDF sheets were incubated with the appropriate secondary antibody coupled to peroxidase (0.16 μg/ml) in TBS-T buffer for 1 h. The membranes were washed several times with PBS-T, and peroxidase activity was detected with the ECL enhanced chemiluminescence system (PerkinElmer, Boston, MA).

Single and double immunofluorescence staining of kidney sections.

Mice weighing ∼20 g were anesthetized by intraperitoneal injection of pentobarbital sodium (2.5 mg/kg body wt) 20 min before perfusion in accord with a protocol approved by the MGH Institutional Subcommittee on Research Animal Care. The kidneys were perfused briefly through the left ventricle with PBS (pH 7.4) at 37°C. When kidneys were cleared of blood (∼1 min of perfusion), they were fixed by perfusion with paraformaldehyde-lysine-periodate fixative (5). After further incubation in paraformaldehyde-lysine-periodate overnight at 4°C, tissues were rinsed 3 × 5 min in PBS, cryoprotected in 30% sucrose in PBS, placed in OCT compound 4583 (Tissue-Tek, Elkhart, IN), frozen in the microtome chamber, and sectioned at 5 μm with a Reichert Frigocut microtome (Leica Microsystems, Deerfield, IL). Sections were picked up on Fisher Superfrost Plus slides (Fisher Scientific, Philadelphia, PA) for immunohistochemistry.

After rehydration in PBS buffer, the kidney tissue sections were treated with 1% SDS in PBS pH 7.4 for 3 min as an antigen retrieval step (6). Nonspecific binding interactions were reduced by incubating the kidney tissues sections for 20 min in PBS-BSA 1% solution before the addition of the primary antibody. Then kidney sections were incubated 1 h in the presence of immunopurified anti-Alix antibody (10 ng/ml final concentration) in the absence or presence of the corresponding antigenic peptide (100 ng/ml) or in the presence of preimmune serum (10 ng/ml). After being washed with PBS, tissues were then incubated with a CY3-conjugated secondary antibody (2 μg/ml) for 60 min at room temperature. After final washes with PBS, sections were mounted in Vectashield (Vector Labs, Burlingame, CA) and examined using a Zeiss/Bio-Rad Radiance 2000 laser confocal microscope (Carl Zeiss microImaging, Thornwood, NY).

To perform double labeling, the anti-Alix antibody was first detected using a tyramide signal amplification procedure to avoid signal cross-reactivity that would otherwise result from double labeling using primary antibodies raised in the same species. This procedure has been described previously in work from our laboratory (5). The kidney tissue sections were rehydrated in PBS buffer, then treated with 1% SDS in PBS, pH 7.4, for 3 min, and washed three times before incubation in the presence of 3% hydrogen peroxide solution for 10 min. The tissues were blocked with a biotin-blocking reagent (Molecular Probes) and then blocked again following the instructions of the TSA amplification kit (PerkinElmer Lifesciences). Kidney sections were incubated with the affinity-purified rabbit primary antibody against Alix diluted to 0.5 ng/ml for 90 min. At this antibody concentration, interaction between the Alix antibody and the anti-rabbit IgG-CY3 secondary antibody used in the second step (see below) was not detectable. To visualize Alix fluorescence, the antibody signal was amplified using the TSA-FITC fluorescence system, following the instructions of the manufacturer. After washing with PBS, tissues were then reincubated with either polyclonal anti-AQP2 antibody (diluted 1/50) or an anti-V-ATPase antibody (diluted 1/20) for 1 h. A CY3-conjugated secondary antibody (2 μg/ml) was applied for 60 min at room temperature. The interaction of this reagent with the anti-Alix primary antibody was below the threshold of detection due to the low concentration of anti-Alix IgG applied in the first step. Thus no signal overlap was seen despite the use of two antibodies raised in the same species. After final washes with PBS, sections were mounted and examined as describe above.

Immunofluorescence of LLC-V2R-GFP cells transfected with FLAG-Alix.

LLC-V2R-GFP cells were plated on 12 × 12-mm glass coverslips (Fisher Scientific, Pittsburgh, PA). Cells were transfected with FLAG-Alix in the pcDNA3 vector as described above. Transfected cells were incubated with or without VP (1 μM) at 37°C in either the presence or absence of diverse agents as detailed below. Cells were pretreated 20 min with the proteasome inhibitor (MG132, 10 μM) or the lysosome degradation inhibitor (chloroquine, 0.1 μM). Those concentrations have been shown to inhibit V2R degradation efficiently (3, 35). Both drugs were present during the incubation with VP (1 μM) for 90 min. After treatment, cells were fixed with 4% paraformaldehyde (Electron Microscopy Sciences, Washington, PA)-5% sucrose in PBS for 20 min at room temperature. The cells were washed three times in PBS and then used for immunocytochemistry. Cells were permeabilized with 0.1% Triton in PBS buffer for 3 min at room temperature before incubation with primary antibodies as above for tissue sections. Cells were incubated with monoclonal anti-FLAG antibody (11 μg/ml) for 1 h at room temperature. Cy3-conjugated goat-anti-mouse IgG (1.5 μg/ml) was applied for 1 h at room temperature. Coverslips were mounted on slides with Vectashield medium (Vector Labs) and examined using a Zeiss/Bio-Rad Radiance 2000 confocal microscope.

In addition to these antibodies, we also used a fluorescent marker to study the intracellular localization of FLAG-Alix. Cells were preincubated for 30 min with Lysotracker (500 nM, Molecular Probes), a marker of acidic compartments including lysosomes and late endosomes/multivesicular bodies, before the VP treatment. Vesicles containing Alix were then compared with the distribution of Lysotracker-labeled vesicles. After incubation, the cells were fixed and visualized as described aboved.


Yeast two-hybrid screening for V2R COOH tail-interacting proteins.

A peptide with the sequence of the last 29 amino acid of the V2R COOH terminus was used as bait to screen a human kidney cDNA library. BLAST analysis revealed that five of the clones obtained were from the same protein, Alix. To characterize the specificity of the interaction, the yeast two-hybrid screen was performed in the presence of positive and negative controls as shown in Table 1. The data suggested a possible interaction between these two proteins in vivo. Alix did not interact with an AQP2 COOH-terminal probe in a two-hybrid screening that was performed in parallel (data not shown). Alix is an 869-amino acid cytoplasmic protein with two distinct domains, the COOH domain from amino acids 1–434 and the NH2 domain from amino acids 468–869 (7). Sequence analysis showed that all five clones encoded for the COOH domain of Alix, but all clones also contained at least a part of the NH2 domain (Fig. 1). Clone 1 starts at amino acid (aa) 206, clone 2 at aa 211, clone 3 at aa 305, clone 4 at aa 307, and clone 5 from aa 382. This observation suggested that the COOH-terminal domain of Alix might be involved in mediating the interaction between Alix and V2R. Thus we investigated further the interactions in vitro by binding assays.

Fig. 1.

Schematic representation of the alignment of ALG-2-interacting protein X (Alix) clones obtained from yeast 2-hybrid screening and comparison with full-length Alix. Clone 1 starts from amino acid (aa) 206, clone 2 starts from aa 211, clone 3 starts from aa 305, clone 4 starts from aa 307, and clone 5 starts from aa 382. All 5 clones have an intact COOH-terminal domain of Alix, indicating that this region might be involved in mediating the interaction between Alix and the vasopressin type 2 receptor (V2R).

View this table:
Table 1.

Yeast two-hybrid results of V2R-COOH tail-Alix interaction

Pull-down assays with GST fusion proteins.

Purified GST-V2R and GST alone were used in pull-down assays using cell lysates from HEK293 cells transfected with FLAG-Alix. The pull-down complex was subjected to SDS-PAGE and Western blotting with antibody against FLAG peptide. FLAG-Alix, a protein of ∼100 kDa, was very strongly expressed in HEK293 cells (Fig. 2, lane 1), but this band was not detected in pull-down assays using nontransfected cells (Fig. 2, lane 2). The GST-V2R COOH tail (Fig. 2, lane 4) but not GST alone (Fig. 2, lane 3) was able to pull down FLAG-Alix from overexpressing cells. These data provide further evidence that the V2R COOH tail interacts with Alix, although pull-down assays do not necessarily imply that the interaction is direct.

Fig. 2.

GST pull-down assay using the GST-V2R COOH tail fusion protein shows the ability of this portion of the V2R to interact (directly or indirectly) with Alix. Western blot analysis of the proteins pulled-down from FLAG-Alix-transfected HEK293 cell lysates was performed using an anti-FLAG antibody as a probe. A 100-kDa band was detected in the FLAG-Alix-transfected cell homogenate using anti-FLAG IgG (lane 1). No band was observed when the pull-down was performed with the the GST-V2R COOH tail and nontransfected cell lysate (lane 2) or when the pull-down assay was done with GST alone and FLAG-V2R-transfected cells (lane 3). In contrast, a 100-kDa band was detected using anti-FLAG IgG when the pull-down assay was performed from lysates of HEK293 cells transfected with FLAG-Alix using the complete GST-V2R COOH-tail fusion protein (lane 4). This result is representative of 3 independent experiments.

Characterization of the anti-Alix antibody and immunolocalization of Alix in kidney sections.

Characterization of the anti-Alix antibody is shown in Fig. 3. The polyclonal antibody detects a 100-kDa band in HEK293 cells transfected with Alix (Fig. 3A, lane 4) but not in untransfected cells (Fig. 3A, lane 3), reflecting a low expression of endogenous Alix in these cells. A monoclonal anti-FLAG antibody shows a similar pattern in Fig. 3A, lanes 1 and 2. Furthermore, the anti-Alix antibody detected a major band at ∼100 kDa in mouse kidney and a lower minor band at ∼60 kDa (Fig. 3B, lane 1) that is displaced by peptide competition (Fig. 3B, lane 2). Those bands are not observed using preimmune serum from the same rabbit before immunization (Fig. 3B, lane 3).

Fig. 3.

Characterization of the antibody against Alix. Western blotting on either cell lysates from HEK293 cells transfected with either FLAG-Alix or an empty vector (A) or a kidney membrane preparation (B) using an antibody directed against 14 amino acids of mouse Alix, raised in rabbit and affinity purified is shown. A: anti-Alix antibody (lanes 3 and 4) was compared with an anti-FLAG monoclonal antibody, a positive control (lanes 1 and 2). Both antibodies detected a broad band at ∼100 kDa in HEK293 cell lysates transfected with FLAG-Alix (lanes 2 and 4, respectively). No bands were observed in lysates of cells transfected with an empty vector pcDNA3. B: anti-Alix antibody (lane 1) detected a major band at ∼100 kDa in kidney preparations as well as a weaker band at a lower molecular mass. Both bands were undetectable when anti-Alix antibody was preincubated with the immunizing peptide (lane 2). Furthermore, these bands were not detectable using preimmune serum from the same rabbit that was used to raise the anti-Alix antibody (lane 3).

Kidney tissue slices were immunostained with a polyclonal antibody against human Alix and with either an anti-AQP2 antibody that labeled principal cells or anti-V-ATPase antibodies to identify intercalated cells. The anti-Alix antibody showed bright staining throughout the cytoplasm in kidney tissue (Fig. 4A). Staining is abolished when kidney tissues are incubated with antibody pretreated with the immunizing peptide or with preimmune serum (Fig. 4, B and C, respectively). The immunolocalization data show that Alix is expressed in kidney where it is associated with different types of cells. Figure 5, AC, shows that Alix is present on large intracellular vesicles in AQP2-positive principal cells in the collecting duct. Figure 5, DF, shows that Alix is also expressed in V-ATPase-positive collecting duct intercalated cells. Thus while Alix expression is not specific for one cell type, the localization of Alix in principal cells that express AQP2 and the V2R supports the notion that Alix might play a role in protein and receptor trafficking in these cells.

Fig. 4.

Characterization of anti-Alix antibody specificity. Kidney sections were incubated with polyclonal anti-Alix antibody (A). Kidney sections incubated with either the anti-Alix antibody in the presence of peptide inhibitor (B) or with preimmune serum (C) showed no staining. These results are representative of 3 independent immunostaining experiments in different animals.

Fig. 5.

Immunostaining of kidney sections using the anti-Alix antibody (A and D) was followed by signal amplification using a commercial TSA amplification kit. After amplification, tissues were costained with either an anti-AQP2 antibody (B, red) or an anti-V-ATPase antibody (E, red). Merged images from the double stainings are shown in C and F. In these collecting ducts from the border between the outer medulla and inner medulla, Alix is present in both aquaporin-2 (AQP2)-positive principal cells (PC) and V-ATPase-positive intercalated cells (IC). These results are representative of 3 independent immunostaining experiments in different animals.

Immunofluorescence of Alix and V2R in transfected LLC-PK1 cells.

FLAG-Alix was transfected into LLC-PK1 cells stably expressing V2R-GFP (Fig. 6A), and it was found to have a cytoplasmic distribution (Fig. 6A). Overexpression of FLAG-Alix does not affect either the level of expression or the localization of V2R-GFP (Fig. 6B). V2R-GFP is localized mainly in the plasma membrane, but some is also present in the cytoplasm: the pattern of fluorescence is mostly distinct from that of Alix (Fig. 6C). In the presence of VP, however, the membrane GFP-V2R fluorescence is lost, whereas intracellular GFP fluorescence increases. The V2R-GFP appears to be internalized into juxtanuclear vesicles (Fig. 6, E and H) in most cells. However, in cells expressing Alix (red), most of the GFP fluorescence disappeared not only from the plasma membrane but also from the cytoplasm (Fig. 6, E and H, arrows). FLAG-Alix staining was consistently associated with an almost complete loss of GFP fluorescence, but only after VP treatment (Fig. 6, F and I). This result suggested that FLAG-Alix is involved in the loss of V2R-GFP fluorescence. In control cells, Alix was not colocalized with Lysotracker, which had a distinct cellular distribution (Fig. 7, AD). However, in LLC-V2R-GFP cells treated with VP for 2 h, Alix was clearly colocalized with Lysotracker (Fig. 7, E and F), as indicated by the purple/pink color in the merged image (Fig. 7H). In contrast, under these conditions, V2R-GFP was not detectable in the Alix-transfected cells (Fig. 7G), as also shown in Fig. 6. These results suggest that the rate of delivery of V2R-GFP to lysosomes for degradation is increased in the presence of FLAG-Alix, which also becomes localized in Lysotracker-positive structures after VP treatment.

Fig. 6.

Expression of Alix affects the ligand-induced degradation of V2R-green fluorescent protein (GFP). FLAG-Alix was transiently transfected into the LLC-V2R-GFP cell line. FLAG-Alix was detected using an anti-FLAG antibody and a CY3-conjugated secondary antibody (red). In the basal condition (AC), FLAG-Alix is located throughout the cytoplasm (A, red) and V2R-GFP is mainly localized at the membrane (B, arrows). Little colocalization between FLAG-Alix and V2R-GFP was observed [as shown in the overlap picture (C, yellow)]. The effects of vasopressin (VP) incubation for 45 min in 2 different experiments are presented individually. Thus DF show a similar treatment to GI. In the presence of VP for 45 min, V2R-GFP is internalized into the cytoplasm (E and H), except in the case where the cells overexpressed FLAG-Alix (see arrows in E and H). Internalized V2R-GFP green fluorescence is not detectable in cells expressing FLAG-Alix in the overlap pictures (F and I). This result is representative of 8 independent experiments. Bar = 10 μm.

Fig. 7.

Colocalization of FLAG-Alix and intracellular compartment markers in LLC-V2R-GFP cells (arrow) after VP stimulation. In the absence of VP, V2R-GFP is localized mainly at the plasma membrane (C, green), whereas FLAG-Alix and Lysotracker are scattered throughout the cytoplasm (A, blue and B, red, respectively). The merged image does not show colocalization (D). After 2 h with 1 μM VP, FLAG-Alix (E, blue), Lysotracker (F, red), and V2R-GFP (C, green) were localized in the cytoplasm. Only a limited amount of overlap was seen between the lysosomal marker and V2R-GFP (D, yellow), whereas a partial overlap was observed between Lysotracker and FLAG-Alix (D, purple). The presence of Alix in the lysosomal compartment suggests that V2R-GFP was degraded in lysosomes in the presence of FLAG-Alix. This result is representative of 3 independent experiments. Bar = 10 μM.

To characterize the role of FLAG-Alix interactions, the degradation of V2R-GFP was examined by Western blot analysis using an anti-GFP antibody. Transfection of FLAG-Alix did not affect the staining pattern of V2R-GFP in LLC-PK1a cells where the major band detected is a doublet at ∼57–68 kDa (Fig. 8, lane 1 vs. lane 4). In the presence of VP, a time- and VP-dependent increase in the intensity of a lower molecular weight band at ∼46 kDa was observed in cells transfected with FLAG-Alix (Fig. 8, lane 3 vs. lane 6). The intensity of the 46-kDa band increased sevenfold between the first and the second hour of incubation in the presence of VP in cells transfected with FLAG-V2R (Fig. 8, lanes 13), whereas no significant modification in intensity was observed in nontransfected cells (Fig. 8, lanes 46).

Fig. 8.

Western blot using anti-GFP antibody showing that V2R-GFP degradation in the presence of FLAG-Alix is time dependent. The V2R-GFP band pattern is similar between nontransfected cells and cells transfected with FLAG-Alix (lanes 1 and 4, respectively). Both conditions showed a predominant band of 57–68 kDa. In the presence of FLAG-Alix and VP, there is an increase of the 46-kDa band intensity between 0 and 2 h of incubation (lanes 1 and 3, respectively). No significant increase in the 46-kDa band was observed in the presence of VP alone at this time point (lanes 46). This result is representative of 3 independent experiments.

We have recently shown that this band is a degradation product and is not due to changes in glycosylation of V2R-GFP (3). The smaller, higher molecular weight band that appears after 2 h of VP treatment in cells not overexpressing Alix (Fig. 8, lane 6) may also represent a partial degradation product of the V2R-GFP chimera. This result confirms our previous data that VP causes degradation of V2R-GFP, but suggests that the rate or the extent of degradation is increased by Alix transfection, since the major 46-kDa degradation product that we reported previously (3) is barely detectable at the 2-h time point in nontransfected cells. To confirm the specificity of FLAG-Alix in V2R degradation after VP-stimulation, we examined the effect of overexpressing FLAG-Alix in HEK cells transiently transfected with either V2R-GFP or PTHR-GFP (Fig. 9, A and B, respectively). In the presence of VP, V2R-GFP is degraded as described above (Fig. 9A, lanes 15). Quantification of the intensity of the higher molecular weight band representing mature V2R-GFP after normalization with a β-actin loading control shows a reduction of 30% in the intensity of the 70-kDa band (Fig. 9A, lane 1 vs. lane 5). In contrast, the PTHR-GFP band intensity is not affected in FLAG-Alix-transfected cells, even after 4-h incubation with the PTH ligand (Fig. 9B, lanes 15). The effects of a lysosomal degradation inhibitor, chloroquine, were next tested along with the effect of MG132, an inhibitor of proteasomal degradation. When LLC-V2R-GFP cells were pretreated with MG132 during stimulation with VP, V2R-GFP was degraded in the presence FLAG-Alix as indicated by the loss of intracellular fluorescence in Alix-transfected cells (Fig. 10, AC). In contrast, when cells were pretreated with chloroquine, V2R-GFP was internalized after VP stimulation, and no subsequent loss of intracellular fluorescence was detectable even in cells overexpressing Alix (Fig. 10, DF). Interestingly, Alix did not appear to colocalize with the internalized V2R-GFP under these conditions (Fig. 10F). This suggests that chloroquine treatment might have an effect on the cellular distribution of Alix that could also contribute to the inhibitory effect of chloroquine on V2R degradation that we have reported previously (3). This result supports the notion that the loss of V2R-GFP fluorescence after 2-h VP treatment in Alix-overexpressing cells is probably attributed to lysosomal degradation, whereas the proteosome is not involved in this process.

Fig. 9.

Western blot using anti-GFP antibody showing that V2R-GFP but not PTHR-GFP degradation occurs in the presence of FLAG-Alix. After VP treatment, the 70-kDa “mature” V2R-GFP band intensity is reduced with time (A, lane 1 vs. lane 5). In contrast, the PTHR-GFP band pattern is not affected in the presence of PTH (B, lane 1 vs. lane 5). This result is representative of 3 independent experiments.

Fig. 10.

Effects of proteasomal and lysosomal degradation inhibitors on V2R-GFP in the presence of both VP and FLAG-Alix. LLC-V2R-GFP cells were transfected with FLAG-Alix. Cells were pretreated for 20 min with a proteasome inhibitor (MG132) or a lysosomal degradation inhibitor (chloroquine). When cells were stimulated with 1 μM VP for 90 min in the presence of MG132, V2R-GFP (B, green) was internalized in the cytoplasm, and GFP fluorescence was lost in cells overexpressing FLAG-Alix (A, arrow, red). After VP stimulation, V2R-GFP was internalized in cells treated with chloroquine but green fluorescence was still detectable (E, green) even though FLAG-Alix (D, red) was similarly overexpressed. This result shows the role of lysosomes and FLAG-Alix in V2R-GFP degradation. The figure is representative of 3 independent experiments. Bar = 10 μM.


Ligand-induced endocytosis contributes to the physiological regulation of a wide variety of signaling receptors, and the endocytosis of GPCRs has, therefore, been a widely studied phenomenon. It is now well established that an array of proteins such as β-arrestin, dynamin, amphiphysin, synaptojanin and endophilin are closely related to the formation, fission, and uncoating of clathrin-coated pits and have a strong tie to GPCR downregulation, recycling, and resensitization (10, 23, 50). However, the proteins and signaling pathways that are involved in human V2R trafficking are still relatively unknown. In the present study, we used the V2R COOH tail as bait in the yeast two-hybrid assay. Five positive clones were isolated, which all coded for overlapping sequences from the same protein, Alix. This 100-kDa cytosolic protein is also known as AIP1 and Hp95 and is involved in apoptosis, cell proliferation (30, 45, 48), and inducing vacuolization when overexpressed in some cell types (7).

Interaction of Alix with the V2R was confirmed using a GST-V2R COOH-terminal fusion protein to pull down Alix from cell supernatants (30, 45). While such pull-down assays can also reveal indirect protein interactions, our combined data strongly suggest that Alix binds directly to the V2R. Immunofluorescence analysis confirmed data from other labs that Alix is present in the kidney (30, 45) and revealed that it is expressed in cells of the collecting duct. The presence of Alix in VP-sensitive principal cells supports the hypothesis that Alix might be involved in V2R regulation, but its localization in adjacent intercalated cells that do not express the V2R indicates that this interaction is not exclusive. Indeed, Alix has several protein-protein interaction motifs that might allow multiple distinct regulatory activities. For example, Alix and its yeast homolog (BRO1) were found to interact with other membrane proteins such as the EGF receptor or GAP1P, respectively (15, 39, 44). Based on the yeast two-hybrid data, the COOH terminus of Alix is able to bind to the last 29 amino acids of hV2R, but the exact motifs that are involved in this process are not yet known. The sequence of this COOH-terminal region is 70% conserved among V2Rs from different species, but it has no homology with any other members of the VP/oxytocin receptor family (2, 13). Analysis of the hV2R COOH tail shows a class IV WW domain interaction motif (TP) that interacts with short proline-rich regions in other proteins (19). However, the sequence is not conserved in all species, and more work will be necessary to identify the amino acid domains involved in the Alix-V2R interaction.

While the presence of some Alix at the plasma membrane cannot be ruled out by our immunofluorescence studies, it is also localized throughout the cytoplasm as previously observed in other cell lines (45). Using Lysotracker as a marker that accumulates in highly acidic organelles, the vesicles containing Alix are suggested to be lysosomes or late endosomes/multivesicular bodies (MVBs), cellular compartments in which several other Alix-interacting proteins including the V2R have been localized (3, 20, 21, 33, 42, 46). It is still unclear when and where Alix first interacts with the V2R. Alix may play a role at the plasma membrane by interacting with proteins involved in endocytosis. One of these is endophilin, an enzyme that regulates membrane shape during endocytosis (26). Endophilin is an SH3 domain-containing protein and it functions as a lysophosphatidic acid acyltransferase that appears to be important for the invagination of clathrin-coated vesicles. Its major binding partner is synaptojanin, which also binds to dynamin (28, 29). This observation suggests that Alix may influence the fate of the receptor when it interacts with V2R at the membrane; the complex is sent to lysosomes and/or late endosomes for degradation rather than being recycled (3, 35). Another protein, β-arrestin, is also involved in the lysosomal trafficking of the V2R (17), and it is unknown how (or if) these two distinct proteins act together to coordinate intracellular V2R trafficking. It is also possible that Alix, which is mainly localized in MVBs in some cell types, interacts only with internalized V2R after β-arrestin binding has occurred. Alix was located in vesicles that originated from the MVBs that are intermediates between early and late endosomes (43). Bro1, the yeast homolog of Alix, is a key protein that is involved in the formation of luminal vesicles by inward budding of the limiting membrane of MVBs. The role of Alix in the control of vesicle budding in MVBs is also supported by several studies performed in eukaryotic cells (26, 33, 34, 42, 46). The V2R/Alix complex might, therefore, regulate V2R delivery to a degradative compartment by facilitating passage through the MVB and/or increasing the extent or efficiency of the production of luminal vesicles inside MVBs. Formation of these intraluminal vesicles is under the control of a number of other proteins that form the so-called ESCRT complex (1, 9), with which Alix interacts (42), and is critical for the subsequent lysosomal degradation of membrane proteins, including the EGF receptor (11). Furthermore, overexpression of Alix affects also the number and/or the shape MVBs without modifying other intracellular compartments; such an increase in MVB “activity” might influence the rate of degradation of membrane proteins (7, 27) and explain the increased rate of V2R degradation found in the present study. In previous work, we were able to detect a significant increase in protein bands representing V2R degradation only 4 h after ligand addition to LLC-PK1 cells (3). After transfection of the same cells with Alix, a lower molecular weight degradation band was already detectable 1 h after ligand stimulation and was greatly increased 2 h after hormone addition. It should be pointed out that our ability to detect the V2R and any degradation products depends critically on the retention of GFP antigenicity (by Western blotting) or fluorescence (by microscopy), since no useful antibodies against the native V2R are available at this time. Therefore, it is possible that other smaller or distinct degradation products are present under the different conditions examined, but they would be invisible to our fluorescence and Western blotting assays.

In summary, we show here that the 100-kDa soluble protein Alix interacts in eukaryotic cells with the V2R, a member of the GPCR family. Transfection of renal epithelial cells with Alix increases the rate and/or the efficiency of V2R trafficking to lysosomes after ligand-induced internalization.


This work was supported by National Institutes of Health (NIH) Grants DK-19406 (D. A. Ausiello), DK-38452 (D. Brown), CA-082230 (T. Shioda), and DK-11794 (A. B. Abou-Samra). H. Y. Lin was supported by NIH KO8 award DK-02716. R. Bouley received fellowship support from the kidney foundation of Canada. H. A. J. Lu is supported by NIH KO8 Grant K08-DK-075940–01. The Microscopy Core facility of the Massachusetts General Hospital Program in Membrane Biology receives additional support from the Boston Area Diabetes and Endocrinology Research Center (DK-57521) and the Center for the Study of Inflammatory Bowel Disease (DK-43341).


  • * X. Yi and R. Bouley contributed equally to this study.

  • 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.


View Abstract