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

J. Biol. Chem., Vol. 276, Issue 49, 46251-46259, December 7, 2001

This Article Abstract Full Text (PDF) All Versions of this Article: 276/49/46251    most recent M104545200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Weixel, K. M. Articles by Bradbury, N. A. Search for Related Content PubMed PubMed Citation Articles by Weixel, K. M. Articles by Bradbury, N. A. Social Bookmarking                      What's this?

µ2 Binding Directs the Cystic Fibrosis Transmembrane Conductance Regulator to the Clathrin-mediated Endocytic Pathway^*

Kelly M. Weixel and Neil A. Bradbury

From the Cystic Fibrosis Research Center, Department of Cell Biology and Physiology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261

Received for publication, May 18, 2001, and in revised form, August 27, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The cystic fibrosis transmembrane conductance regulator (CFTR) contains a conserved tyrosine-based internalization motif, ¹⁴²⁴YDSI, which interacts with the endocytic clathrin adaptor complex, AP-2, and is required for its efficient endocytosis. Although direct interactions between several endocytic sequences and the medium chain and endocytic clathrin adaptor complexes have been shown by protein-protein interaction assays, whether all these interactions occur in vivo or are physiologically important has not always been addressed. Here we show, using both in vitro and in vivo assays, a physiologically relevant interaction between CFTR and the µ subunit of AP-2. Cross-linking experiments were performed using photoreactive peptides containing the YDSI motif and purified adaptor complexes. CFTR peptides cross-linked a 50-kDa subunit of purified AP-2 complexes, the apparent molecular mass of µ2. Furthermore, isolated µ2 bound to the sorting motif, YDSI, both in cross-linking experiments and glutathione S-transferase pull-down experiments, confirming that µ2 mediates the interaction between CFTR and AP-2 complexes. Inducible overexpression of dominant-negative µ2 in HeLa cells results in AP-2 complexes that fail to interact with CFTR. Moreover, internalization of CFTR in mutant cells is greatly reduced compared with wild type HeLa cells. These results indicate that the AP-2 endocytic complex selectively interacts with the conserved tyrosine-based internalization signal in the carboxyl terminus of CFTR, YDSI. Furthermore, this interaction is mediated by the µ2 subunit of AP-2 and mutations in µ2 that block its interaction with YDSI inhibit the incorporation of CFTR into the clathrin-mediated endocytic pathway.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cystic fibrosis, the most common lethal genetic disease of Caucasians, is caused by mutations in the gene encoding the protein for the cystic fibrosis transmembrane conductance regulator (CFTR)¹ (1). Alterations in the genetic sequence of CFTR result in the impairment of transepithelial chloride secretion in response to the activation of a cAMP-mediated signal transduction pathway in cystic fibrosis cells (2). Indeed, it is now established that CFTR functions at the apical plasma membrane of polarized epithelial cells to regulate chloride permeability in response to cAMP-dependent protein kinase (PKA)-mediated phosphorylation, ATP binding, and ATP hydrolysis (3-6).

Morphological, biochemical, and functional evidence indicates that, in addition to a cell surface localization, CFTR is also found in endosomal and recycling compartments. The presence of CFTR within endosomes was demonstrated functionally in several cell lines expressing endogenous and exogenous CFTR. Thus, PKA stimulated an anion conductance in isolated as well as in situ endosomes, a conductance that was susceptible to inhibition by monoclonal anti-CFTR antibodies (7). In addition, immunocytochemistry at the light microscopy level has revealed co-localization of CFTR with rab4, a member of the small GTP-binding protein family, and a component of recycling endosomes (8). Compelling evidence indicates that CFTR enters endosomal compartments through the clathrin-mediated endocytic pathway (9-12). Furthermore, perturbation of clathrin-coated vesicle formation inhibits the removal of CFTR from the plasma membrane (12, 13).

Clathrin-mediated internalization of integral membrane proteins relies on the presence of relatively short peptide sequences within their cytosolic tails. Although heterogeneous, most sorting signals fall into two main classes (14-17). The first class is characterized by an essential tyrosine residue, either as part of an NPXY motif (as initially identified in the low density lipoprotein receptor), or in the context of a YXX motif, where X is any amino acid and  is a bulky hydrophobic amino acid). The second class of internalization motifs typically contains a dileucine sequence, although in some cases one of the leucines may be replaced by an isoleucine, valine, alanine, or methionine. Such endocytic sorting signals have been most extensively studied in type I and type II membrane proteins. Depending upon the precise context of the sorting signal, such motifs can be also recognized in sorting events within the trans-Golgi network as well as endosomes (18, 19).

Clathrin adaptor complexes (APs) have been obvious candidates to recognize sorting signals and to act as adaptors between integral membrane proteins and the clathrin lattice. However, in only a very few cases have membrane proteins and adaptors co-immunoprecipitated (20). Several in vitro assays, including bead pull-down and surface plasmon resonance, had shown interaction between tyrosine-based sorting signals and clathrin adaptor complexes (21-25), yet surprisingly few data are available concerning whether such interactions occur in vivo or are physiologically important. Although a direct interaction between YXX internalization sequences and µ2 subunits has been demonstrated, the binding of receptors to AP-2 does not necessarily correlate with the internalization capacity of proteins bearing YXX motifs. For example, the epidermal growth factor (EGF) receptor strongly binds AP-2 through a YRAL sequence (26). However, mutations in the YRAL sequence that abolish the interaction of the EGF receptor with AP-2 do not significantly affect internalization of the receptor (26, 27). In contrast, transferrin receptors whose endocytic removal from the plasma membrane shows a strong dependence upon a YTRF motif (28), show very weak, if any, detectable interaction with the AP-2 endocytic adaptor complex (21, 29). In addition, although direct interaction between tyrosine-based sequences and the µ subunit of the endocytic adaptor complex AP-2 have been demonstrated by several types of protein-protein interaction assays (21, 23), a recent report suggests that the NPXY motif binds directly to the terminal domain of the clathrin heavy chain rather than directly to AP-2 (30).

In contrast to the many studies on endocytic signals in type I and II membrane proteins, relatively little is known about the endocytic signals in polytopic membrane proteins such as transporters and ion channels. The ₂-adrenergic receptor contains a highly conserved tyrosine residue (Tyr³²⁶) that is responsible for the ligand-induced internalization of the receptor (31). GLUT4, the insulin-responsive sodium-glucose co-transporter, is constitutively retrieved from the plasma membrane via clathrin-mediated pathways in the absence of insulin and contains a leucine endosomal targeting signal (32). Recently, a tyrosine (Tyr¹⁴²⁴)-based motif was identified as a potential endocytic targeting signal in the carboxyl cytoplasmic tail of CFTR (10, 33, 34). Mutation of this sequence, either in the context of a chimera consisting of transferrin receptor and the carboxyl-terminal tail of CFTR (10) or, in intact CFTR stably expressed in a heterologous cell line (35), inhibits the retrieval of CFTR from the plasma membrane.

Thus although AP-2 adaptors, and their medium chain subunits, have been implicated in the recruitment of several type of plasma membrane proteins into the clathrin-dependent endocytic pathway, the molecular details and the physiological significance of such interactions are poorly understood. Moreover, such information is completely lacking for ion channels. We have used both in vitro protein-protein interaction studies and in vivo studies with inducible dominant-negative µ2 to investigate the interaction of adaptor complexes with the tyrosine endocytic sequence of a clinically important ion channel, CFTR. The analyses of in vitro binding and endocytic trafficking reveal, for the first time in a polytopic ion channel, a unified model for CFTR endocytosis showing a strong correlation between in vitro binding of CFTR to µ2 and its endocytic capacity.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Monoclonal antibodies against -adaptin and 1/2-adaptin were obtained from BD Transduction Laboratories (Lexington, KY). Monoclonal antibody AC1-M11 against -adaptin was a generous gift from Dr. M. S. Robinson (University of Cambridge, Cambridge, United Kingdom). A rabbit polyclonal antiserum to an amino-terminal sequence of the µ2 chain was custom generated by Affinity Bioreagents (Golden, CO). Antibodies were obtained by immunization with a peptide corresponding to residues 11-29 of human µ2 (KGEVLISRVYRDDIGRNAV). This antibody was specific for µ2 and did not recognize the related µ1 protein (data not shown). Polyclonal anti-CFTR antibodies were from Affinity Bioreagents. Alexa-Fluor 488 goat anti-rabbit secondary antibodies were from Molecular Probes (Eugene, OR). Gelvatol was from Monsanto Co. (Augusta, GA). TNT^® T7 coupled reticulocyte lysate system was from Promega (Madison, WI). The QuickChange^TM site-directed mutagenesis kits were obtained from Stratagene (La Jolla, CA). SuperSignal^® West Pico chemiluminescent substrate and GelCode^® Blue stain reagent were from Pierce. Glutathione-Sepharose 4B and Redivue [³⁵S]methionine were from Amersham Pharmacia Biotech. Complete^TM EDTA-free protease inhibitor tablets were obtained from Roche Molecular Biochemicals. Peptides were synthesized by New England Peptide (Fitchburg, MA). Geneticin^®, LipofectAMINE 2000^® reagent, primers, and tissue culture media were from Life Technologies, Inc. All other antibiotics and materials were from Sigma and were of reagent grade quality.

Purification of AP Complexes-- Clathrin-coated vesicles were obtained from bovine calf brains (Pel-Freez) as described previously (36-38). Purified adaptor complexes were analyzed to establish that all four subunits were present in the preparation by protein staining with GelCode^® Blue reagent and immunoblot with subunit-specific antibodies.

Surface Plasmon Resonance (SPR)-- The interaction between CFTR carboxyl terminus peptides and AP-1 or AP-2 was analyzed in real time by SPR (39) using a Biacore X Biosensor (Biacore, Piscataway, NJ). Peptides (KVIEENKVRQYDSIQ) were coupled via their amino-terminal biotin moiety to an SA5 sensor chip (streptavidin surface) according the manufacturer's instructions. All binding studies were performed with buffer containing 20 mM HEPES-NaOH (pH 7.0), 150 mM NaCl, 10 mM KCl, 2 mM MgCl₂, 0.2 mM dithiothreitol at a flow rate of 20 µl/min. Purified adaptor and clathrin preparations were centrifuged at 250,000 × g for 30 min prior to the experiment to remove potential aggregates. Adaptors and clathrin were used at 100 nM unless otherwise noted. A short pulse injection (15 s) of 20 mM NaOH, 0.5% SDS was used to regenerate the sensor chip surface after each experiment. The peptide-derivatized sensor chip remained stable and retained its specific binding capacity throughout the experiments.

Peptide Synthesis-- Several peptides were synthesized by New England Peptide and are shown in Table I. The synthetic peptide *YQRL corresponds to the carboxyl terminus of the cytoplasmic tail of TGN38, and has been shown to bind to purified AP-2 (23). Photoreactive peptides *YDSI and *ADSA correspond to the peptides YDSI and ADSA and have a photoreactive probe, benzoylphenylalanine at position Tyr-3 and a biotin moiety added to the amino-terminal lysine to facilitate detection of cross-linked products using streptavidin-HRP. Peptides were stored at 4 °C until use when they were diluted in water to a final concentration of 2 µM.

Plasmid Constructs-- The GST-CT construct (kindly provided by Drs. R. Frizzell and F. Sun, University of Pittsburgh, Pittsburgh, PA) contains the carboxyl terminus of CFTR (amino acids 1404-1480) amplified from pBQ4.7 CFTR cDNA using the polymerase chain reaction and subcloned in to the pGEX 4T-1 vector. The cDNA for GST fusion proteins containing the human immunodeficiency virus type 1 p6 protein (GST-p6) and equine infectious anemia virus p9 protein (GST-p9) were kindly provided by Dr. R. Montellaro (University of Pittsburgh). GST constructs containing µ1 and µ2 were provided by Dr. J. Bonafacino (National Institutes of Health, Bethesda, MD). The cDNA of mouse µ2 was subcloned from a pACTµ2 construct kindly provided by J. Bonafacino into pCDNA3.1 for in vitro translation reactions. Epitope-tagged wild type µ2 in pcDNA 3.1 was obtained from Dr. A. Sorkin (University of Colorado, Denver, CO).

UV-induced Cross-linking Reactions-- Cross-linking experiments were performed as described (23); briefly, purified AP complexes were kept in Tris-buffer (250 mM Tris (pH 7.4), 1 mM EGTA, 0.5 mM MgCl₂, 0.5 mM dithiothreitol). AP concentrations ranging between 0.13 and 0.6 mg/ml were incubated with 0.2 µM photoreactive peptides. For competition experiments, peptides YQRL, ADSA, or YDSI were included at a final concentration of 200 µM. Cross-linking experiments were carried out in microtiter plates in a final volume of 20 µl. Plates were incubated on ice for 30 min in the dark. Samples were irradiated at 80,000 µJ/cm² to cross-link samples. Following irradiation, 5 µl of 5× Laemmli sample buffer was added to the samples and boiled for 5 min.

Detection of Cross-linked Products-- To identify the AP protein chains that were cross-linked to the biotinylated peptide, an aliquot of the cross-linking reaction was subjected to SDS-PAGE and transferred to nitrocellulose. Following blocking (10 mM Tris (pH 8.0), 150 mM NaCl, 0.05% Tween 20, 10% nonfat dry milk) for 1 h at room temperature, the membrane was rinsed in Tris-buffered saline plus Tween (10 mM Tris (pH 8.0), 150 mM NaCl, 0.05% Tween 20) and incubated for 30 min with streptavidin conjugated with horseradish peroxidase (HRP, 5 µl/ml; Zymed Laboratories Inc.) in blocking buffer. The membrane was then washed six times for 5 min with Tris-buffered saline plus Tween and processed for ECL using SuperSignal West Pico ECL reagent (Pierce). For recapturing experiments, AP complexes were photolabeled with *YSDI peptides and AP complexes were immunoprecipitated with AC1-M11 antibodies. Immunoprecipitates were washed five times in RIPA (1% Triton X-100, 0.3 M NaCl, 50 mM Tris-HCl (pH 7.0), 0.1% BSA), supplemented with Complete^TM EDTA-free protease inhibitors (Roche). Samples were resuspended in SDS buffer (0.1 M Tris-HCl (pH 7.4), 1% SDS, 10 mM dithiothreitol). Samples were shaken vigorously for 20 min at 4 °C and boiled for 5 min to release the immunoprecipitates from the antibodies and to denature the AP complex. The extract was diluted 20-fold with RIPA, clarified by centrifugation at 15,000 × g, 4 °C. Samples were then incubated with rotation at 4 °C for 1 h with one of the following antibodies (anti-, anti-1/2, or anti-µ). Immunoprecipitates were washed twice with RIPA and once with RIPA minus detergent. Samples were analyzed by SDS-PAGE and immunoblot using streptavidin-HRP as described above.

Mammalian Cell Culture and Transfection-- HeLa cell lines expressing hemagglutinin-tagged D176A/W421A mutant µ2 constructs under control of the tetracycline-off system were kindly provided by Dr. A. Sorkin. The HeLa cell line has been characterized previously and expresses an epitope-tagged µ2 containing mutations at Asp-176 and Trp-421 under control of the Tet-Off system. Studies in this cell line demonstrate that mutant hemagglutinin-µ2 incorporates into AP-2 and is targeted to coated pits. Inducible overexpression of the mutant µ2 resulted in the replacement of endogenous wild type µ2 in AP-2 complexes and the complete abrogation of AP-2 interactions with tyrosine-based internalization motifs (40). Cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 400 µg/ml G418, 200 ng/ml puromycin, and 2 ng/ml doxycycline, a tetracycline derivative. For experiments, cells were plated in the growth medium without selection markers with or without doxycycline. At 24 h after plating, medium was replaced with fresh medium with or without doxycycline; cells without doxycycline were supplemented with 2 mM sodium butyrate to ensure high levels of hemagglutinin-µ2 protein expression to replace the endogenous wild type µ2 in AP-2 complexes. Experiments were performed 3-4 days after plating. For transient transfections cells were plated at 50-60% confluence as outlined above. 48 h after plating, cells were transfected with pcDNA3.1 CFTR plasmids using the LipofectAMINE 2000^TM reagent according to the manufacturer's instructions (Life Technologies, Inc.). Transfected cells were used 48 h after transfection.

GST Pull-down Experiments-- HeLa cells cultured in the presence or absence of doxycycline were washed with PBS to remove media and to cool the cells. Cells were lysed in TGH buffer (50 mM HEPES (pH 7.4), 1% Triton X-100, 10% glycerol, 100 mM NaCl, 1 mM EDTA, 1 mM EGTA), supplemented with Complete^TM EDTA-free protease inhibitor tablets (Roche Molecular Biochemicals) for 20 min at 4 °C. Insoluble material and aggregates were removed by centrifugation for 45 min at 125,000 × g. Glutathione-Sepharose, loaded with 20 µg of GST or GST-CT, was incubated with HeLa lysates for 3 h to overnight at 4 °C. Beads were washed three times with TGH and once with HEPES buffer (20 mM HEPES, 150 mM KCl, 2 mM MgCl₂ (pH 7.2)). Beads were resuspended in Laemmli buffer, and bound proteins were resolved by SDS-PAGE, transferred to nitrocellulose, and processed for immunoblot according to standard protocols. Detection of bound primary monoclonal antibodies was performed using horseradish peroxidase-conjugated goat anti-mouse secondary antibodies and enhanced chemiluminescence (ECL) using SuperSignal^® West Pico reagents (Pierce).

pcDNA 3.1 constructs containing wild type or D176A/W421A µ2 were translated at 30 °C for 1.5 h in the presence of [³⁵S]methionine using the TNT coupled protein translation kit (Promega), according to the manufacturer's instructions. Prior to pull-down experiments, 10% of the translation was reserved for SDS-PAGE analysis. For each pull-down experiment, 45 µl of the translation was diluted into 1 ml of TGH buffer and cleared by centrifugation for 45 min at 125,000 × g. 20 µg of GST-CT, GST-P6, GST-P9, and GST preabsorbed onto glutathione-Sepharose were incubated with TGH diluted translation products for 3 h to overnight at 4 °C. Beads were washed three times with TGH buffer and resuspended in Laemmli buffer. Proteins were resolved by SDS-PAGE, and gels were dried and processed for autoradiography.

Expression of GST Fusion Proteins-- GST fusion proteins were expressed in Escherichia coli BL21 de3pLysS strain. Protein expression was induced by 0.4 mM isopropyl--D-thiogalactopyranoside for 3 h, and GST fusion proteins were purified on glutathione-Sepharose 4B (Amersham Pharmacia Biotech).

Immunofluorescence-- Cells grown on glass coverslips were cooled to 4 °C for 15 min and washed three times with ice-cold PBS. Cells transiently expressing CFTR were incubated with polyclonal anti-CFTR antibodies diluted 1:500 in PBS containing 1% BSA for 1 h at 4 °C. Cells were then rapidly warmed to 37 °C with pre-warmed media and incubated for 15 min in a 37 °C incubator at 5% CO₂. Cells were immediately cooled to 4 °C with PBS-containing 1% BSA and kept on ice. The cell surface was labeled with wheat germ agglutinin conjugated to rhodamine (Molecular Probes) for 30 min at 4 °C. Cells were then washed, fixed in 2% paraformaldehyde, and permeabilized in PBS containing 0.1% Triton X-100. The cells were incubated with Alexa-Fluor 488 goat anti-rabbit secondary antibodies for 1 h. The cells were washed and subsequently mounted onto glass slides using Gelvatol. Confocal microscopy was performed on a TCS confocal microscope equipped with krypton, argon, and helium-neon lasers (Leica, Deerfield, IL). Images were imported into Adobe Photoshop^® for final presentation. To examine colocalization of CFTR and cell surface markers, the images were imported into Metamorph^® imaging software (Universal Imaging Corp., Downingtown, PA). The stored images were converted into binary images, and a new series of images were generated by performing a Boolean operation "AND" in the pairs of images representing Alexa-Fluor 488 and rhodamine signals within the same optical section. The final presentation of colocalized proteins uses differential intensity spectral representation encompassing the black/white intensity range 0-255.

Radioactive Binding Assay-- The internalization efficiency of CFTR was measured by monitoring the cell surface density of CFTR with a rabbit polyclonal antibody against an extracellular epitope of CFTR and ¹²⁵I-labeled Protein A (PerkinElmer Life Sciences). Cell-surface CFTR was labeled for 1 h with anti-CFTR antibodies (3 µg/ml) in OptiMEM^® at 4 °C. Following incubation, the cells were washed with PBS and internalization was initiated by incubating the cells in pre-warmed 37 °C OptiMEM^® followed by incubation at 37 °C for 15 min. Cells were rinsed with ice-cold PBS, and anti-CFTR antibody remaining at the cell surface was measured by with binding of ¹²⁵I-Protein A (30 µCi total) in OptiMEM^® for 1 h at 4 °C. The cells were rinsed in ice-cold PBS and solubilized in 100 mM NaOH, 0.1% v/v Triton X-100. The radioactivity of solubilized cells was determined with a  counter. Nonspecific binding was measured in mock-transfected cells. The internalization efficiency of CFTR in wild type HeLa cells and HeLa cells expressing the dominant-negative mutant µ2 was expressed as a percentage of the decrease in the surface binding of ¹²³I-Protein A between the zero time point and 15 min. The results are a mean of three individual experiments ± S.E.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The Tyrosine-based Sorting Signal of CFTR Selectively Interacts with Plasma Membrane Adaptors, AP-2-- We employed surface plasmon resonance to monitor interaction of the carboxyl cytoplasmic tail of CFTR with purified clathrin, AP-1, and AP-2. This method has been used in several studies to analyze the interaction between purified adaptors and receptor tails. A typical binding experiment using adaptors from bovine brain is shown in Fig. 1. As the solution passed over a sensor surface immobilized with wild type CFTR peptide was changed from buffer alone to buffer containing AP-2, a rapid increase in signal was recorded reflecting AP-2 binding and an increase in mass at the sensor surface. Upon returning to buffer after 3 min of AP-2 exposure, a loss of signal was observed, consistent with a loss of mass from the sensor surface. The signal did not return entirely to the initial base line, indicating that some protein remained bound. After exposure to 20 mM NaOH, 0.5% SDS for 2 min, the surface was completely regenerated (data not shown). When AP-1-containing solution was passed over the CFTR-containing chip, a small binding signal as observed, which decreased upon AP-1 washout. Note, however, that the signal obtained with AP-1 was observed at a 10-fold higher adaptor concentration that that for AP-2 binding. As expected, clathrin showed no interaction with CFTR at all, consistent with the notion that CFTR/clathrin interactions are mediated via adaptor complexes.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Surface plasmon resonance analysis of CFTR peptide interactions with purified proteins of the clathrin-mediated endocytic pathway. CFTR peptides containing the 1424YDSI endocytic sorting motif were immobilized onto a sensor chip via avidin linkage as described under "Experimental Procedures." Samples containing purified AP-2, AP-1, or clathrin were injected under continuous flow conditions at the concentration indicated, and the bar indicates perfusion of protein-containing buffer. The resonance is given in arbitrary units (RU). Only solutions containing purified AP-2 complexes bound to the YDSI peptide of CFTR. Perfusion of solutions containing AP-1 complexes or clathrin did not result in a shift in resonance units much beyond base line.

The Tyrosine-based Sorting Signal in CFTR Binds to the µ2 Subunit of AP-2-- To characterize the subunit of AP-2 that interacts with the carboxyl terminus of CFTR, we performed in vitro cross-linking assays using synthetic photoactivable peptides that contained the tyrosine motif identified in CFTR, *YDSI (Table I). We initially established that *YDSI peptides specifically associate with plasma membrane adaptors, AP-2, (Fig. 2A, lane 4). The *YDSI peptide cross-links to a 50-kDa subunit of the AP-2 complex, the apparent molecular mass of µ2, but fails to recognize any of the subunits of the Golgi specific AP-1 complex (Fig. 2A, lane 2). Furthermore, peptides containing alanine substitution at tyrosine 1424 and isoleucine 1427, *ADSA, abolished the ability of the peptide to cross-link to AP-2 (Fig. 2A, lane 3). Specificity of binding was confirmed by performing cross-linking studies in the presence of competitor peptides. Cross-linking reactions were performed as described in Fig. 2 with or without the presence of 100-fold excess competitor peptides. Peptides containing mutations in the tyrosine based sorting motif failed to compete with wild type peptides for AP-2 cross-linking, (Fig. 3, lane 2), whereas peptides containing the tyrosine-based sorting motif from TGN38 successfully competed with CFTR peptide for AP-2 cross-linking, as did the wild type peptide YDSI, (Fig. 3, lanes 3 and 4, respectively). This is consistent with our hypothesis that the interaction between CFTR and AP-2 is dependent on the tyrosine-based motif YDSI.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   A, the *YDSI sorting motif of CFTR cross-links to the µ2 subunit of the AP-2 complex. Purified AP-2 and AP-1 complexes were incubated with photoactivable peptides and subject to UV cross-linking. Proteins were resolved by SDS-PAGE, transferred to nitrocellulose, and immunoblotted with streptavidin-HRP. *YDSI-containing peptides cross-link to a 50-kDa subunit of the AP-2 complex, the apparent molecular mass of µ2 (lane 4). Peptides with mutations in the sorting motif, *ADSA, fail to interact with AP-2 (lane 3). Neither peptide cross-links to any of the AP-1 subunits (lanes 1 and 2). B and C, immunoblot of AP-2 complexes (B) and Coomassie staining of cross-linking reactions (C) demonstrate that all four subunits of the AP-2 are present in the reaction.

View larger version (43K): [in this window] [in a new window]   Fig. 3.   Peptides containing tyrosine-based sorting motifs compete for AP-2 cross-linking. Cross-linking reactions were performed with the addition of 100-fold excess competitor peptides. Peptides containing mutations in the tyrosine sorting motif, ADSA, did not compete with WT peptides for AP-2 cross-linking (lane 2). Peptides containing the tyrosine-based sorting motif from TGN38 did successfully compete with WT peptides for AP-2 cross-linking (lane 3), as did an excess of nonbiotinylated WT peptide (lane 4).

Although the intact complex was present in the cross-linking reaction as detected by immunoblot and Coomassie staining (Fig. 2, B and C, respectively), a 50-kDa subunit was the only cross-linked species detected. To confirm that the cross-linked protein was µ2, recapturing experiments were performed. After photolabeling, AP-2 complexes were immunoprecipitated with the AP-2-specific antibody, anti--adaptin. Isolated AP-2 complexes were dissociated into individual subunits by denaturation and boiling, followed by immunoprecipitation of the resultant individual subunits. The recaptured subunits were then analyzed by SDS-PAGE and immunoblot with streptavidin-HRP conjugates to determine which subunit had been cross-linked by the biotinylated peptide. Confirming the results in Fig. 2A, the only subunit of the AP-2 complex that was cross-linked by the *YDSI peptide was µ2 (Fig. 4, lane 6). To confirm that the µ2 subunit of AP-2 is solely responsible for mediating the interaction between CFTR and AP-2, cross-linking experiments were performed with isolated µ2 subunits. A GST fusion protein containing the medium subunit of the AP-2 complex, µ2, was purified onto glutathione-Sepharose beads and incubated with the photoreactive peptides described above. After UV illumination, cross-linked proteins were resolved by SDS-PAGE and detected by immunoblot. *YDSI peptides cross-linked to GST-µ2, but failed to cross-link to GST alone (Fig. 5, lanes 9, 4, and 7, respectively). Furthermore, GST fusion proteins containing the medium subunit of AP-1, µ1, were not cross-linked by *YDSI (Fig. 5, lanes 5 and 8). As seen with the intact AP-2 complexes, cross-linking was specific for the tyrosine-based motif as peptide *ADSA failed to cross-link to µ2 (Fig. 5, lane 6). These results demonstrate that the interaction among plasma membrane adaptors, AP-2, and CFTR requires the intact sorting signal, YDSI. Furthermore, this interaction is mediated solely through the medium subunit of the AP-2 complex, µ2.

View larger version (50K): [in this window] [in a new window]   Fig. 4.   Recapturing experiments confirm that *YDSI only cross-links µ2. AP-2 complexes purified from bovine brain were incubated with the cross-linking peptide *YDSI and UV-irradiated. AP-2 complexes were immunoprecipitated with AP-2-specific antibodies against the -subunit. AP-2 complexes were fractionated by SDS-PAGE, and cross-linked species were detected by immunoblot with streptavidin-HRP. *YDSI-containing peptides cross-link to a 50-kDa subunit of the AP-2 complex, the apparent molecular mass of µ2 (lane 2). In parallel photolabeling reactions, isolated AP-2 complexes were dissociated into individual subunits by denaturation and boiling followed by immunoprecipitation with antibodies against the individual AP-2 subunits,  (lane 4), 1/2 (lane 5), or µ2 (lane 6). Recaptured subunits were analyzed by SDS-PAGE and immunoblot with streptavidin-HRP. µ2 was the only subunit of the AP-2 complex that was cross-linked by the *YDSI peptide (lane 6).

View larger version (25K): [in this window] [in a new window]   Fig. 5.   YDSI specifically cross-links to isolated µ2. GST, GST-µ2, and GST-µ1 were incubated with the photoreactive peptides *YDSI and *ADSA and subjected to UV irradiation. Proteins were resolved by SDS-PAGE, transferred to nitrocellulose, and incubated with streptavidin-HRP. Coomassie Blue staining of the GST fusion proteins is shown in lanes 1-3. Lanes 4-6 are the ECL observed with the mutant cross-linking peptide, *ADSA. Lanes 7-9 demonstrate the cross-linking results obtained with the WT peptide, *YDSI.

Dominant-negative Mutant µ2 Fails to Interact with CFTR-- Binding assays and crystal structure data have identified that amino acid residues Asp-176 and Trp-421 of µ2 as essential for the binding of YXX motifs. To characterize the binding site of µ2 for YDSI, we performed pull-down assays using in vitro translated µ2 and GST fusion proteins containing the YDSI motif of CFTR, GST-CT. Wild type µ2 or µ2 with mutations D176A/W421A were translated in vitro in the presence of [³⁵S]methionine and incubated with either GST-CT or GST alone. GST-CT efficiently bound to wild type µ2, but failed to interact with mutant µ2, (Fig. 6, lanes 2 and 1, respectively). Similar experiments were performed as controls using the p9 protein from equine infectious anemia virus that has been shown previously to interact with µ2, and the human immunodeficiency virus type 1 p6 protein that does not interact with µ2 (Fig. 5, lanes 4 and 6, respectively) (41). Thus, mutagenesis and pull-down assays demonstrate that the two regions on µ2 containing Asp-176 and Trp-421 are critical in µ2 interaction with the YDSI sorting signal of CFTR.

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Dominant-negative mutant µ2 fails to interact with CFTR. Wild type µ2 or dominant-negative mutant µ2 (D176A/W421A) was translated in vitro in the presence of [35S]methionine. GSH-Sepharose-immobilized GST fusion proteins were incubated with 4% of translation reaction diluted in binding buffer as described under "Experimental Procedures." Bound proteins were resolved by SDS-PAGE and processed for autoradiography. GST fusion proteins containing the carboxyl terminus of CFTR (GST-CT) efficiently pulled-down WT µ2 but not mutant µ2 (lanes 8 and 9, respectively). GST alone did not bind either µ2 protein (lanes 2 and 3). A GST fusion protein containing the equine immunodeficiency virus protein p9 was used as a positive control for wild type µ2 binding (lane 5), and a GST fusion protein containing the human immunodeficiency virus type 1 p6 protein, which does not interact with µ2, was used as a negative control (lane 4). An aliquot of the µ2 translation reaction (2%) is shown in lane 1.

Inducible Overexpression of Dominant-negative µ2 Inhibits the Interaction of AP-2 and CFTR-- A previously characterized HeLa cell line that expresses a dominant-negative mutant µ2 (D176A/W421A) under control of the Tet-Off system (40) was utilized to investigate how the replacement of endogenous µ2 by mutant µ2 affects the interaction between AP-2 and CFTR. Lysates from HeLa cells expressing either the wild type µ2 or dominant-negative µ2 were incubated with GST-CT or GST alone. The amount of AP-2 that bound to GST-CT was determined by immunoblot of GST precipitates using an antibody specific for the AP-2 complex, -adaptin. Wild type AP-2 complexes efficiently bound to GST-CT, but not GST alone (Fig. 6, lanes 3 and 4, respectively). In contrast, when the expression of mutant µ2 was induced by the removal of doxycycline, binding of AP-2 and GST-CT was abolished (Fig. 7, lane 2). The amount of AP-2 in the input faction from cells expressing WT or dominant-negative µ2 was the same (data not shown). These results demonstrate that the µ2 subunit of the AP-2 complex alone mediates the interaction between coated pit adaptor and the internalization motif YDSI in CFTR.

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Mutant AP-2 complexes fail to interact with the COOH terminus of CFTR. GST fusion proteins containing the COOH terminus of CFTR were incubated with cell lysates from HeLa cells expressing wild type AP-2 (+Tet) and HeLa cells expressing the dominant-negative AP-2 complexes (Tet). Bound proteins were resolved by SDS-PAGE, transferred to nitrocellulose, and immunoblotted with the AP-2-specific antibody, anti--adaptin. GST-CT efficiently interacted with WT AP-2 complexes (lane 3); however, mutant AP-2 complexes failed to interact with GST-CT (lane 1). GST alone (lanes 2 and 4) failed to interact with either AP complex.

To examine the consequence of blocking AP-2 and CFTR interactions on the internalization of CFTR, full-length CFTR was transiently expressed in wild type and dominant-negative µ2-expressing HeLa cells. The internalization of CFTR was followed by immunofluorescence using a polyclonal antibody against an extracellular epitope of CFTR. HeLa cells were labeled for 1 h at 4 °C with anti-CFTR antibodies, cells were then warmed to 37 °C for 15 min to allow for internalization. Following internalization, cells were returned rapidly to 4 °C and the cell surface was labeled with rhodamine-conjugated wheat germ agglutinin (WGA). The internalized pool of CFTR was then detected by fixing and permeabilizing the cells and incubating with Alexa-Fluor 488-conjugated secondary antibodies. In wild type cells following the internalization period, CFTR was found primarily as discrete punctate vesicles within the cytoplasm (Fig. 8, panel B). CFTR signal (green) did not co-localize with the WGA signal (red) at the cell surface in merged images, demonstrating that CFTR was rapidly internalized from the plasma membrane in cells expressing wild type µ2 (Fig. 8, panels B and C). However, in cells expressing the dominant-negative mutant µ2, the immunolocalization pattern of CFTR was different, with the majority of the signal remaining at the cell surface (Fig. 8, panels E and F). The signal for CFTR in the mutant cells co-localizes with that of the cell surface marker WGA, yielding a yellow cell surface staining pattern upon merge of the separate signals (Fig. 8, panel F). These observations suggest that CFTR internalization was inhibited in cells expressing the dominant-negative mutant µ2. These images were imported into Metamorph^® imaging software to compare the regions of colocalization in HeLa cells expressing endogenous wild type µ2 or dominant-negative mutant µ2. The degree of colocalization of green (CFTR) and red (WGA) signal is depicted as a spectral plot encompassing the black/white intensity range 0-255, with white being the most intense. Cells expressing the dominant-negative mutant µ2 show significantly greater co-localization of CFTR with WGA (Fig. 9 panel A) compared with cells expressing the endogenous wild type µ2 (Fig. 9, panel B). In addition, the regions where CFTR and WGA overlap in HeLa cells expressing the dominant-negative mutant µ2 show a greater intensity of colocalization compared with cells expressing endogenous wild type µ2 in differential intensity maps (compare panels A and B in Fig. 9), suggesting that more CFTR remains at the cell surface in cells expressing the dominant-negative mutant µ2 compared with cells expressing wild type endogenous µ2.

View larger version (25K): [in this window] [in a new window]   Fig. 8.   Dominant-negative AP-2 complexes inhibit the internalization of CFTR. CFTR was transiently expressed in HeLa cells expressing WT AP-2 complexes or in HeLa cells expressing mutant AP-2 complexes. Cell surface CFTR was labeled with anti-CFTR antibodies at 4 °C followed by incubation at 37 °C for 15 min to monitor internalization of CFTR from the cell surface (panels B and E). Prior to fixation rhodamine-conjugated wheat germ agglutinin was used to label the cell surface (panels A and B). Panel C shows a merge of panels A and B; intracellular green staining demonstrates that CFTR endocytosis proceeds in wild type cells. Panel F, a merge of D and E, shows that CFTR remains at the cells surface, indicated by the overlap of red (cell surface WGA) and green (CFTR) signals to yield a yellow cell surface staining pattern.

View larger version (9K): [in this window] [in a new window]   Fig. 9.   CFTR remains at the cell surface in dominant-negative mutant HeLa cells. The co-localized pixel densities of WGA and CFTR are shown by differential intensity plot using black/white intensity representation derived from confocal image acquisition of HeLa cells expressing dominant-negative mutant µ2 (panel A) and wild type cells (panel B). CFTR and cell surface WGA showed a higher degree of co-localization in mutant cells (panel A) compared with wild type cells (panel B), indicated by the appearance of significantly more pixels in panel A compared with panel B. The differential intensity maps (lower panels) of the co-localized regions in mutant cells (panel A) and wild type cells (panel B) show that CFTR and cell surface WGA share greater spectral intensity, indicative of a higher degree of co-localization.

A radioactive antibody-binding assay was implemented to compare quantitatively the rates of endocytosis of CFTR in cells expressing wild type and dominant-negative µ2. The extracellularly exposed epitope of CFTR was saturated with anti-CFTR antibodies at 4 °C. Endocytosis was subsequently initiated by shifting the temperature to 37 °C. After 15 min, anti-CFTR antibody remaining at the cell surface was measured by the specific binding of ¹²⁵I-labeled protein-A at 4 °C. The rapid disappearance of cell surface anti-CFTR antibodies in HeLa cells expressing wild type µ2 indicated that CFTR is internalized with high efficiency in HeLa cells (Fig. 10), as has been shown for many other cell types. Approximately 60% of cell surface CFTR was endocytosed during a 15-min incubation. This implies an internalization rate of 4.2 ± 0.3%/min (n = 3) for CFTR in cells expressing wild type µ2. In contrast, the internalization rate of CFTR in cells expressing dominant-negative µ2 was significantly slower (1.8 ± 0.2%/min; n = 3) (Fig. 10).

View larger version (17K): [in this window] [in a new window]   Fig. 10.   Dominant-negative mutant µ2 inhibits the internalization of CFTR. Full-length CFTR was transiently expressed in HeLa cells expressing wild type µ2 (+Tet; dark shading) or HeLa cells expressing dominant-negative mutant µ2 (Tet; light shading). The surface density of CFTR after 15 min of internalization was determined with the radioactive antibody binding assay as described under "Experimental Procedures." Cell surface CFTR was labeled with anti-CFTR antibodies at 4 °C, followed by 15 min of internalization at 37 °C. The remaining CFTR antibody at the cell surface was measured by 125I-Protein A. Results are expressed as a percentage of CFTR internalized from the cell surface at 15 min relative to the amount of total cell surface CFTR at time 0. Results are mean ± S.E. for three separate experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Both in vitro (including co-immunoprecipitation, pull-down, SPR) and in vivo (yeast two-hybrid and dominant-negative cell lines) assays have been utilized to investigate the interaction between internalized integral membrane proteins and subunits of the endocytic AP-2 clathrin adaptor complex. However, although a direct interaction between YXX internalization sequences and µ2 subunits has been demonstrated, the binding of receptors to AP-2 does not necessarily correlate with the internalization capacity of proteins bearing YXX motifs. For example, the EGF receptor strongly binds AP-2 through a YRAL sequence (26). However, mutations in the YRAL sequence that abolish the interaction of the EGF receptor with AP-2 do not significantly affect internalization of the receptor (26). In contrast, transferrin receptors whose endocytic removal from the plasma membrane shows a strong dependence upon a YTRF motif (28), show very weak, if any, detectable interaction with the AP-2 endocytic adaptor complex (21, 29). Moreover, although such studies as those described above have been performed on monotopic abundantly expressed receptors, there is a paucity of information regarding the interaction of clathrin adaptor proteins and polytopic low abundance proteins, such as ion channels. Our data provide the first evidence linking in vitro binding data with in vivo internalization data for a clinically important ion channel, the CFTR.

Although the primary function of CFTR is to regulate transepithelial chloride permeability by functioning as an apical chloride channel, immunolocalization and functional studies demonstrate that CFTR also resides in the endosomal and recycling compartments (8, 42). Cell surface biotinylation indicates that CFTR undergoes rapid and efficient internalization in both polarized epithelial and nonpolarized cells (9, 10). In addition, several studies indicate that the distribution of CFTR may be regulated by PKA by promoting the translocation of CFTR from an intracellular pool to the plasma membrane and by inhibiting the internalization of CFTR from the plasma membrane (43-47). These results suggest that the amount of CFTR in the plasma membrane may be regulated by insertion and retrieval mechanisms and that these processes may provide a mechanism to augment PKA activation in regulating CFTR in the plasma membrane.

Studies investigating the traffic of CFTR argue that it is internalized exclusively through the clathrin-mediated endocytic pathway (12, 13). Although inhibition of clathrin-coated vesicle formation inhibits the removal of CFTR from the cell surface of native epithelial cells, disruption of alternative pathways, such as caveolae, does not affect the internalization of CFTR (12). Furthermore, both immunological and electrophysiological techniques show that CFTR can be found in clathrin-coated vesicles in both native epithelial cells and in cultured cells (11). Proteins that target to clathrin-coated vesicles are recruited to these structures by interacting with clathrin adaptor complexes. Association of AP complexes with membrane proteins is mediated by the presence of sorting motifs in the cytoplasmic tails of such proteins (for review see Ref. 20). We have demonstrated previously that CFTR interacts with the endocytic adaptor complex, AP-2, and that this interaction is mediated through the carboxyl tail of CFTR (33). Analysis of the carboxyl tails of several species of CFTR shows the presence of a highly conserved tyrosine-based sorting motif at amino acids ¹⁴²⁴YDSI (10, 33). This conforms to the paradigm of a YXX sorting motif, where X is any amino acid and  is a hydrophobic amino acid. Indeed, mutations in this sorting motif inhibit the internalization of full-length CFTR and block the interaction between the carboxyl terminus of CFTR and AP-2 complexes (10, 33, 35).

The tyrosine-based sorting motif YDSI has been shown to be necessary for efficient endocytosis of CFTR. To evaluate the selectivity of the interaction between this sorting motifs and clathrin adaptor complexes, we performed surface plasmon resonance analysis using an immobilized peptide corresponding to the YDSI motif of CFTR and recorded interactions with purified adaptor complexes AP-2 and Golgi-specific adaptor complexes, AP-1. No interaction was detected with AP-1 complexes or clathrin. However, CFTR peptides did interact with AP-2 adaptors, indicating that the sorting motif in the carboxyl terminus of CFTR specifically interacts with plasma membrane adaptors, consistent with our previous studies (33). AP-2 and AP-1 complexes are responsible for selecting cargo that traffics through clathrin-coated vesicles that originate at the plasma membrane and TGN, respectively, and recognize similar sorting signals in the cytoplasmic domains of such proteins. The observation that the tyrosine signal, YDSI, selectively interacts with AP-2 complexes but not AP-1 complexes indicates that this signal in the carboxyl terminus of CFTR is important for directing CFTR to endocytic clathrin-coated vesicles at the plasma membrane.

Despite many protein-protein interaction assays, including bead pull-down, surface plasmon resonance, phage display, interaction overlay, and yeast two-hybrid analysis, demonstrating the association between endocytic sorting motifs, from a variety of proteins, and AP-2 clathrin adaptor complexes, whether such interactions occur in vivo or are physiologically relevant has not always been established. For example, one study has shown in vitro interactions between the asialoglycoprotein receptor tail and the  subunit of AP-2 complexes (25). However, most evidence, including yeast two-hybrid, peptide cross-linking, and crystal structure data, suggests that the µ subunit mediates the interaction between adaptor complexes and tyrosine-based sorting motifs (22, 48-50). Having established that CFTR contains a tyrosine-based sorting motif that interacts with AP-2 complexes, we sought to define the amino acid requirements for the interaction as well as identify the subunit of AP-2 that recognizes this motif. Photoreactive peptides containing the YDSI sorting motif of CFTR cross-linked a 50-kDa subunit of the AP-2 complex, which is the apparent molecular mass of µ2. Recapturing experiments confirmed that this cross-linked species was indeed µ2. This interaction between CFTR peptides and µ2 was dependent upon the tyrosine motif, as mutations at Tyr-1424 and Ile-1427 abolished cross-linking. Furthermore, peptides containing the tyrosine-based sorting motif of TGN38 (SDYQRL) inhibited the cross-linking of CFTR peptides to µ2, indicating that the sorting motif of CFTR and TGN38 likely interact at the same or similar sites on µ2.

Mutagenesis studies on µ2 have revealed at least two regions on µ2, Asp-176 and Trp-421, that are important for interacting with tyrosine-based sorting motifs (40). The crystal structure of µ2 (residues 158-435) complexed with tyrosine-containing peptides has been resolved and corroborates mutagenesis studies, demonstrating that Asp-176 and Trp-421 are directly involved in binding the tyrosine motif of TGN38 (50). GST fusion proteins containing the carboxyl tail of CFTR (GST-CT) efficiently bound isolated wild type µ2, but mutations at D176A/W421A in µ2 abolished the interaction. These results are consistent with previous data that show Asp-176 and Trp-421 constitute the internalization signal-binding interface of µ2. Furthermore, although other subunits have been proposed to mediate the interaction between tyrosine-sorting motifs and AP-2 complexes, these results demonstrate that µ2 exclusively mediates the interaction between YDSI and AP-2. AP-2 complexes containing mutant µ2 failed to bind to CFTR in pull-down experiments, confirming that the tyrosine-based sorting motif of CFTR solely interacts with the µ2 subunit of AP-2. Another tyrosine-based sorting motif binding site has been proposed at the amino-terminal domain of µ2 (residues 102-125); however, as the D176A/W421A mutations blocked the interaction between µ2 and CFTR, it does not appear that these residues contribute significantly to the interaction with the tyrosine sorting motif of CFTR (51).

Recent studies examining the interaction between the EGF receptor and µ2 have shown that, although these two proteins interact in vitro, mutations that affect their interaction do not result in the inhibition of EGF internalization, suggesting the possibility of µ2-dependent and µ2-independent pathways of clathrin-mediated endocytosis (40). These results also highlight the importance of evaluating the functional relevance of protein-protein interactions observed in vitro. Our results show that CFTR was efficiently internalized from the plasma membrane when transiently expressed in cells expressing the wild type endogenous µ2, but that the overexpression of the dominant-negative mutant µ2 significantly reduced the efficiency with which CFTR was retrieved from the cell surface. Furthermore, quantitative analysis of CFTR internalization in the dominant-negative mutant µ2-expressing cells supports the phenomenon observed by immunofluorescence, that the D176A/W421A mutation inhibits the endocytosis of CFTR. Although over 60% of CFTR is internalized in 15 min in wild type HeLa cells, only 25% of CFTR is removed from the cell surface in cells expressing the dominant-negative mutant µ2. These results strongly correlate with the reduction in endocytosis observed in CFTR Y1424A mutants (10, 35), suggesting that the interaction specifically affected is that between µ2 and the ¹⁴²⁴YDSI endocytosis signal. Thus, the interactions between µ2 and CFTR that were initially demonstrated in vitro correspond to a functional interaction that is necessary for the selective and efficient mechanism of CFTR endocytosis. Furthermore, this provides the first characterized example of an endocytosis signal identified for an ion channel.

Although the above results clearly demonstrate the functional relevance of AP-2 and CFTR interactions for CFTR internalization, the data also indicate that a portion of CFTR is still able to undergo internalization in HeLa cells expressing the dominant-negative mutant µ2. This could argue for additional endocytosis signals in CFTR. Indeed, very recent studies by Hu et al. (34) indicate that the efficient endocytosis of CFTR requires multiple internalization signals located in the carboxyl terminus. These authors identified that a phenylalanine, F¹⁴¹³ and a dileucine signal, ¹⁴³⁰LL, co-operate with the ¹⁴²⁴YDSI signal in the carboxyl terminus of CFTR to efficiently drive its internalization from the plasma membrane. These results appear to be at odds, however, with previous studies by Prince et al. (10) examining the dileucine signal in the carboxyl terminus of CFTR, as similar studies performed with CFTR/transferrin receptor chimeras demonstrate that mutagenesis of the dileucine motif ¹⁴³⁰LL to a dialanine has no effect on the rate of CFTR internalization. Moreover, the only mutation identified that affected CFTR internalization was the YXX motif in the carboxyl terminus of CFTR, Y1424A. Presumably, additional signals could interact with clathrin adaptor proteins much in the same manner described here for the YXX motif ¹⁴²⁴YDSI. For example, AP complexes have been shown to interact with dileucine signals in vitro much the same as tyrosine-based signals. However, co-immunoprecipitation studies have yet to show the interaction between AP-2 complexes and dileucine-containing proteins in vivo, prompting some groups to conclude that dileucine signals constitute a low affinity site for AP-2 interactions (51, 52). Although acknowledging the potential for multiple signals in CFTR that could bind to AP-2, the results presented here clearly demonstrate that the YXX sorting motif identified in the carboxyl terminus of CFTR follows the paradigm of other well described YXX motifs, such that it interacts with the defined YXX binding site on µ2, and is important in the efficient endocytosis of CFTR.

The most clinically important mutation of CFTR, F508 CFTR, results in insufficient quantities of CFTR at the plasma membrane. A great deal of effort has gone toward rescuing this mutation in the biosynthetic pathway by chemical chaperones or manipulation of endogenous chaperones with the goal of increasing the delivery of F508 to the plasma membrane. However, recent reports suggest that F508 CFTR is removed from the plasma membrane at a rate faster than that of wild type CFTR (53). Therefore, to increase the amount of CFTR at the plasma membrane, it is necessary to address the issue of residence time of CFTR in the plasma membrane as well as CFTR delivery to the cell surface. Thus, gaining a clearer understanding of the mechanisms by which CFTR is removed from the plasma membrane and the molecular interactions involved in this process will aid in identifying strategies to enhance CFTR expression at the plasma membrane. The results presented here demonstrate that the interaction between the tyrosine-based sorting motif in the carboxyl terminus of CFTR (YDSI) and the µ2 subunit of plasma membrane clathrin adaptors AP-2 provide a mechanism for the selective internalization of CFTR into clathrin-coated vesicles.

	ACKNOWLEDGEMENTS

We gratefully acknowledge the technical assistance of Mark R. Silvis. We appreciate the generosity of Dr. Juan Bonafacino for providing the GST-µ constructs and Dr. Alexander Sorkin for providing the HeLa Tet-off cell line and the cDNA for µ2. We also thank Drs. Michael Marks, Alexander Sorkin, and Juan Bonafacino for helpful discussion.

	FOOTNOTES

^* This work was supported by National Institutes of Health NIDDK Grant 1P50DK56490 and the North American Cystic Fibrosis Foundation Grant BRADBU00G0.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Dept. of Cell Biology and Physiology, University of Pittsburgh School of Medicine, 3500 Terrace St., Pittsburgh, PA 15261. Tel.: 412-648-2845; Fax: 412-648-8330; E-mail: nabrad+@pitt.edu.

Published, JBC Papers in Press, September 17, 2001, DOI 10.1074/jbc.M104545200

	ABBREVIATIONS

The abbreviations used are: CFTR, cystic fibrosis transmembrane conductance regulator; GST, glutathione S-transferase, PAGE, polyacrylamide gel electrophoresis; SPR, surface plasmon resonance; PKA, cAMP-dependent protein kinase; EGF, epidermal growth factor; RIPA, radioimmune precipitation buffer; WGA, wheat germ agglutinin; BSA, bovine serum albumin; PBS, phosphate-buffered saline; AP, adaptor protein; HRP, horseradish peroxidase; WT, wild type.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				A. Jurkuvenaite, K. Varga, K. Nowotarski, K. L. Kirk, E. J. Sorscher, Y. Li, J. P. Clancy, Z. Bebok, and J. F. Collawn Mutations in the Amino Terminus of the Cystic Fibrosis Transmembrane Conductance Regulator Enhance Endocytosis J. Biol. Chem., February 10, 2006; 281(6): 3329 - 3334. [Abstract] [Full Text] [PDF]

				W. B. Guggino and S. P. Banks-Schlegel Macromolecular Interactions and Ion Transport in Cystic Fibrosis Am. J. Respir. Crit. Care Med., October 1, 2004; 170(7): 815 - 820. [Abstract] [Full Text] [PDF]

				M. Gentzsch, X.-B. Chang, L. Cui, Y. Wu, V. V. Ozols, A. Choudhury, R. E. Pagano, and J. R. Riordan Endocytic Trafficking Routes of Wild Type and {Delta}F508 Cystic Fibrosis Transmembrane Conductance Regulator Mol. Biol. Cell, June 1, 2004; 15(6): 2684 - 2696. [Abstract] [Full Text] [PDF]

				J. Walker, H. B. Jijon, T. Churchill, M. Kulka, and K. L. Madsen Activation of AMP-activated protein kinase reduces cAMP-mediated epithelial chloride secretion Am J Physiol Gastrointest Liver Physiol, November 1, 2003; 285(5): G850 - G860. [Abstract] [Full Text] [PDF]

				J. A. Picciano, N. Ameen, B. D. Grant, and N. A. Bradbury Rme-1 regulates the recycling of the cystic fibrosis transmembrane conductance regulator Am J Physiol Cell Physiol, November 1, 2003; 285(5): C1009 - C1018. [Abstract] [Full Text] [PDF]

				C. A. Bertrand and R. A. Frizzell The role of regulated CFTR trafficking in epithelial secretion Am J Physiol Cell Physiol, July 1, 2003; 285(1): C1 - C18. [Abstract] [Full Text] [PDF]

				D. Diviani, A.-L. Lattion, L. Abuin, O. Staub, and S. Cotecchia The Adaptor Complex 2 Directly Interacts with the {alpha}1b-Adrenergic Receptor and Plays a Role in Receptor Endocytosis J. Biol. Chem., May 23, 2003; 278(21): 19331 - 19340. [Abstract] [Full Text] [PDF]

				K. R. Hallows, G. P. Kobinger, J. M. Wilson, L. A. Witters, and J. K. Foskett Physiological modulation of CFTR activity by AMP-activated protein kinase in polarized T84 cells Am J Physiol Cell Physiol, May 1, 2003; 284(5): C1297 - C1308. [Abstract] [Full Text] [PDF]

				M. R. Silvis, J. A. Picciano, C. Bertrand, K. Weixel, R. J. Bridges, and N. A. Bradbury A Mutation in the Cystic Fibrosis Transmembrane Conductance Regulator Generates a Novel Internalization Sequence and Enhances Endocytic Rates J. Biol. Chem., March 21, 2003; 278(13): 11554 - 11560. [Abstract] [Full Text] [PDF]

				Y. Xu, J. C. Clark, B. J. Aronow, C. R. Dey, C. Liu, J. L. Wooldridge, and J. A. Whitsett Transcriptional Adaptation to Cystic Fibrosis Transmembrane Conductance Regulator Deficiency J. Biol. Chem., February 21, 2003; 278(9): 7674 - 7682. [Abstract] [Full Text] [PDF]

				K. Peter, K. Varga, Z. Bebok, C. M. McNicholas-Bevensee, L. Schwiebert, E. J. Sorscher, E. M. Schwiebert, and J. F. Collawn Ablation of Internalization Signals in the Carboxyl-terminal Tail of the Cystic Fibrosis Transmembrane Conductance Regulator Enhances Cell Surface Expression J. Biol. Chem., December 13, 2002; 277(51): 49952 - 49957. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 276/49/46251    most recent M104545200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Weixel, K. M. Articles by Bradbury, N. A. Search for Related Content PubMed PubMed Citation Articles by Weixel, K. M. Articles by Bradbury, N. A. Social Bookmarking                      What's this?