Substance P-induced Trafficking of b-Arrestins THE ROLE OF b-ARRESTINS IN ENDOCYTOSIS OF THE NEUROKININ-1 RECEPTOR*

Agonist-induced redistribution of G-protein-coupled receptors (GPCRs) and b-arrestins determines the subsequent cellular responsiveness to agonists and is important for signal transduction. We examined substance P (SP)-induced trafficking of b-arrestin1 and the neurokinin-1 receptor (NK1R) in KNRK cells in real time using green fluorescent protein. Green fluorescent protein did not alter function or localization of the NK1R or b-arrestin1. SP induced (a) striking and rapid (<1 min) translocation of b-arrestin1 from the cytosol to the plasma membrane, which preceded NK1R endocytosis; (b) redistribution of the NK1R and b-arrestin1 into the same endosomes containing SP and the transferrin receptor (2–10 min); (c) prolonged colocalization of the NK1R and b-arrestin1 in endosomes (>60 min); (d) gradual resumption of the steady state distribution of the NK1R at the plasma membrane and b-arrestin1 in the cytosol (4–6 h). SP stimulated a similar redistribution of immunoreactive b-arrestin1 and b-arrestin2. In contrast, SP did not affect Gaq/11 distribution, which remained at the plasma membrane. Expression of the dominant negative b-arrestin inhibited SP-induced endocytosis of the NK1R. Thus, SP induces rapid translocation of b-arrestins to the plasma membrane, where they participate in NK1R endocytosis. b-Arrestins colocalize with the NK1R in endosomes until the NK1R recycles and b-arrestins return to the cytosol.

Alterations in the subcellular distribution of GPCRs, GRKs, and ␤-arrestins determine cellular responsiveness to agonists and may be important for signal transduction. Receptor endocytosis contributes to desensitization by depleting the cell surface of receptors available to interact with agonists in the extracellular fluid (1), and endocytosis of ␤ 2 ARs is required for stimulation of mitogen-activated protein kinases (15). Agonists of several GPCRs induce translocation of GRK2/3 and ␤-arrestins from the cytosol to the plasma membrane where they interact with cell-surface receptors to participate in uncoupling and endocytosis (16 -18). However, the duration of these interactions and the fate of GRK2/3 and ␤-arrestins during endocytosis and recycling of GPCRs have not been examined. These are important issues because resensitization requires endocytosis and processing of receptors, which may entail dissociation of ligand and ␤-arrestin, receptor dephosphorylation, and recycling (19 -25).
The purpose of the present investigation was to examine substance P (SP)-induced trafficking of ␤-arrestins and to determine the role of ␤-arrestins in endocytosis of the SP or neurokinin-1 receptor (NK1R). Regulation of cellular responses to SP are of interest in view of the important role of the NK1R in pain transmission in the spinal cord, regulation of intestinal motility and secretion, mediation of neurogenic inflammation, and in depression in humans (26,27). ␤-Arrestins and GRKs may mediate NK1R uncoupling since GRK2/3 phosphorylate the NK1R in reconstituted systems and membrane assays (28,29), and disruption of ␤-arrestins abrogates NK1R desensitization (30). The role of ␤-arrestins in SP-induced endocytosis of the NK1R has not been examined. We and others (23,25,(31)(32)(33) have shown that SP induces endocytosis and recycling of the NK1R. By using fluorescent SP to detect the NK1R in neurons, we showed that SP causes translocation of ␤-arrestins to the plasma membrane and to endosomes containing the NK1R (34). However, limitations of using fluorescent SP include inability to detect receptors at later times when SP is degraded (23) and difficulty in colocalization with other proteins by immunofluorescence, since incompletely fixed peptides are washed out during staining (35).
In the present study, we used green fluorescent protein (GFP) and specific antibodies to localize NK1R, ␤-arrestin1/2, and G␣ q/11 , which couples to the NK1R (28,36). This approach requires transfected cells but enables direct observation of trafficking in real time for prolonged periods. We expressed dominant negative ␤-arrestin to determine the role of ␤-arrestins in NK1R endocytosis. Our aims were as follows: (a) generate cell lines expressing NK1R-GFP or wild type NK1R plus ␤-arrestin1-GFP (ARR-GFP) and investigate whether GFP affects the function of these proteins; (b) examine SPinduced trafficking of NK1R-GFP and ARR-GFP in real time and determine their precise subcellular distribution for 0 -6 h after stimulation with SP; (c) determine if SP induces redistribution of endogenous ␤-arrestin1, ␤-arrestin2, and G␣ q/11 in cells expressing wild type NK1R; and (d) examine the effect of a dominant negative mutant of ␤-arrestins on SP-induced endocytosis of the NK1R.
Antibodies-The sources of primary antibodies are shown in Table I. An antibody to GFP was raised in rabbits to a GST-GFP fusion protein.
A polymerase chain reaction fragment of cDNA encoding enhanced GFP (EGFP) was cloned into the HindIII/XbaI restriction sites of pGEX-2T. The resultant plasmid, pGEXGFP, was transfected into Escherichia coli strain BL21DE3. The GST-GFP protein was purified using glutathioneagarose. An emulsion was prepared from equal parts of GFP and complete Freund's adjuvant. Two New Zealand rabbits (female, 8 weeks) were immunized at intervals of 6 -8 weeks. At each immunization, both rabbits received 2.0 ml of emulsion containing approximately 1 mg of GFP, divided into 8 -10 intradermal sites. At the first immunization, 0.5 ml of Tri-immunol vaccine was injected intramuscularly. Antibody 9708 had a titer of 1:100,000 by enzyme-linked immunosorbent assay and was used in all experiments. Goat anti-mouse IgG coupled with (R)-phycoerythrin was from Caltag Laboratories (Burlingame, CA). The sources of other secondary antibodies have been described (37).
Generation of NK1R-GFP, ARR-GFP Constructs-Constructs with EGFP at the C terminus of rat NK1R, ␤-arrestin1, and dominant negative rat ␤-arrestin1 319 -418 were generated by polymerase chain reaction using the following primers and Pfu DNA polymerase. For NK1R, the forward primer was 5Ј-GGCCGGGAATTCGCCACCATGG-ATAACGTCCTTCCTATG-3Ј (EcoRI site underlined, followed by the Kozak translation initiation site, and the N terminus of NK1R cDNA in bold), and the reverse primer was 5Ј-GGCCGGGGATCCCCGGCCAG-CATGTTAGAGTAGAA-3Ј (BamHI site underlined, C terminus of the NK1R cDNA in bold). A 1253-base pair fragment amplified from rat NK1R cDNA was separated on an agarose gel and purified using QiaEx extraction kit. For ␤-arrestin1, the forward primer was 5Ј-GGCCGGA-AGCTTGCCACCATGGGCGACAAAGGGACACGA-3Ј (HindIII restriction site underlined, N terminus of ␤-arrestin1 cDNA in bold), and the reverse primer was 5Ј-GGCCGGCCGCGGTCTGTTGTTGAGGT-GTGGAGA-3Ј (SacII site underlined, C terminus of ␤-arrestin1 in bold). A 1284-base pair fragment amplified from rat ␤-arrestin1 cDNA (Dr. R. Lefkowitz, Duke University) was purified. A fragment of bovine ␤-arrestin1 319 -418 acts as a dominant negative mutant (10). A construct of rat ␤-arrestin1 319 -418 was generated using the forward primer 5Ј-G-GCCGGAAGCTTGCCACCATGGTTTCCTACAAAGTCAAAGTG-3Ј (HindIII site underlined, sequence corresponding to rat ␤-arrest-in1 319 -325 in bold), and the same reverse primer as for wild type ␤-a-rrestin1. All constructs were ligated into pEGFP-N1 vector and used to transform JM109 E. coli in Luria broth containing 30 g/ml kanamycin. The sequences of the chimeras were verified using a reverse primer (5Ј-CGTCGCCGTCCAGCTCGACCAG-3Ј) corresponding to the 46 -67 region of the GFP. Transfection and Cell Culture-The generation and characterization of KNRK cells stably expressing rat NK1R (KNRK-NK1R cells) with N-terminal Flag (DYKDDDDK) epitope has been described (38). The Flag epitope does not affect binding, signaling, or trafficking of the NK1R. To obtain a cell line expressing NK1R-GFP, KNRK cells were transfected with cDNA encoding NK1-GFP, and colonies were screened for expression of GFP by fluorescence microscopy and flow cytometry (38). To generate a cell line expressing FlagNK1R plus ARR-GFP, KNRK-NK1R cells (expressing FlagNK1R) were transfected with cDNA encoding ARR-GFP. To verify coexpression of FlagNK1R plus ARR-GFP, clones were screened by immunofluorescence microscopy and flow cytometry to detect NK1R and ␤-arrestin (below). Highly expressing clones were further enriched by fluorescence-activated cell sorting. Cells were maintained and prepared for experiments as described (23,32,38). To express GFP constructs transiently, KNRK-NK1R cells (stably expressing FlagNK1R) were transfected by overnight incubation with 5 g/ml cDNA encoding control vector pEGFP-N1, ARR-GFP, or ARR 319 -418 -GFP and Lipofectin. Cells were plated on glass coverslips 48 -72 h before experiments.
Flow Cytometry-Flow cytometry was used to assess expression of constructs and to enrich populations of cells. The GFP signal was used to detect NK1R-GFP. In cells coexpressing FlagNK1R and ARR-GFP, NK1R was detected using an M2 antibody to the extracellular Flag epitope and a secondary antibody coupled with (R)-phycoerythrin, and the GFP signal was used to detect ARR-GFP. Cells were dissociated in enzyme-free buffer, and 1.5 ϫ 10 6 cells were resuspended in 200 l of Iscove's medium containing 1 mg/ml bovine serum albumin (BSA) and 1 mM CaCl 2 . Cells were incubated with the M2 antibody (3 g/ml) for 2 h at 4°C, washed, and incubated with goat anti-mouse antibody coupled to phycoerythrin (2 g/ml) for 2 h at 4°C. Cells were resuspended in 1 ml of DMEM with 5% enzyme-free cell dissociation buffer, 0.3% fetal bovine serum, and 2 g/ml propidium iodide and analyzed with a Facscan Flow Cytometer (Becton Dickinson). Fluorophores were excited at 488 nm, and emission was collected at 530/30 nm for GFP and 575/25 nm for phycoerythrin. Viability was assessed by exclusion of propidium iodide. Western Blotting-Western blotting was used to confirm expression of GFP-tagged proteins, determine whether ␤-arrestin1 and ␤-arrestin2 are expressed in KNRK cells, and to assess specificity of the antibodies (37). Cells (ϳ10 7 ) were pelleted and lysed in 1 ml of Laemmli or RIPA buffer containing a protease inhibitor mixture. Lysates were fractionated by SDS-PAGE on 4 -15 or 12% polyacrylamide gels and transferred to nitrocellulose. Membranes were incubated in 3% non-fat milk in PBS for 60 min or 3% BSA in PBS for 60 min and then with antibodies to GFP (1:10,000 overnight, 4°C), ␤-arrestin1 (1:250, 1 h, room temperature), ␤-arrestin2 (1:10,000, 1 h, room temperature), ␤-arrestin1 ϩ 2 (1:10,000, 1 h, room temperature), or G␣ q/11 (0.5 g/ml, 1 h, room temperature). Membranes were washed and incubated with goat antirabbit or mouse IgG conjugated to horseradish peroxidase (1:5,000) for 60 min at room temperature. Proteins were detected as described (37). In controls, primary antibodies were preincubated with 1-2 g/ml antigens overnight at 4°C.
Measurement of [Ca 2ϩ ] i -To examine signaling by cells expressing GFP-tagged proteins, [Ca 2ϩ ] i was measured with Fura-2/AM (24). Fluorescence was measured at 340 and 380 nm excitation and 510 nm emission, and results were expressed as the ratio of the fluorescence at the two excitation wavelengths, which is proportional to the [Ca 2ϩ ] i . To generate concentration-response curves, cells were exposed once to varying concentrations of SP. To examine desensitization, cells were exposed to SP, washed, and then re-exposed to SP. All observations were in n Ͼ 3 experiments.
Endocytosis of 125 I-SP-The rate of NK1R endocytosis was quantified with 125 I-SP as described (32). Cells were incubated in Hanks' balanced salt solution containing 50 pM 125 I-SP, 0.1% BSA, 0.2 mg/ml bacitracin, 20 g/ml leupeptin, and 20 g/ml chymostatin for 60 min at 4°C. They were washed and incubated at 37°C for 0 -30 min. Cells were washed with ice-cold PBS and incubated in 250 l of ice-cold 0.2 M acetic acid containing 50 mM NaCl, pH 2.5, on ice for 5 min to separate acid-labile (cell surface) from acid-resistant (internalized) label. Nonspecific binding was measured in the presence of 1 M SP and was subtracted to give specific binding. Observations were in triplicate in n Ͼ 3 experiments.
Microscopy and Immunofluorescence-To examine trafficking of GFP-labeled proteins in real time, cells were maintained at 37°C in DMEM containing 0.1% BSA (DMEM/BSA). Cells were observed before and after addition of 10 nM SP. To detect proteins by immunofluorescence, cells were incubated in DMEM/BSA containing 10 nM SP, 100 nM cyanine-3-conjugated SP (Cy3-SP), or no addition (control) for 60 min at 4°C (for equilibrium binding), washed at 4°C, and either fixed immediately or incubated in SP-free medium at 37°C for 1 min to 8 h (for trafficking to proceed). Cells were fixed with 4% paraformaldehyde in 100 mM PBS, pH 7.4, for 20 min at 4°C. NK1R, ␤-arrestin1, ␤-arrestin2, ␤-arrestin1 ϩ 2, G␣ q/11 , and markers of endosomes and lysosomes were localized by immunofluorescence (23,34). In some experiments, cells were incubated with Cell Tracker CM-DiI to identify the plasma membrane (37). All observations were in duplicate in n Ͼ 3 experiments. Cells were observed by confocal microscopy (25). For observation of live cells, care was taken to minimize laser exposure by limiting the number of optical sections (2-3 per time point) and the laser intensity (Ͻ3%, aperture of 3-4 mm), so as to maintain fluorescence intensity.

Generation of Cell Lines Stably Expressing NK1R-GFP and NK1R ϩ ARR-GFP
KNRK cells were stably transfected with cDNA encoding NK1R-GFP (KNRK-NK1R-GFP cells). Clone 3W was used for all experiments because a large proportion of cells expressed NK1R-GFP at a high and uniform level, assessed by flow cytometry ( Fig. 1A) and microscopy ( Fig. 1C). To generate cells coexpressing FlagNK1R and ARR-GFP (KNRK-NK1R ϩ ARR-GFP cells), KNRK-NK1R cells were transfected with cDNA encoding ARR-GFP. Clone 23 was selected because a large proportion of cells coexpressed both NK1R and ARR-GFP at high and uniform levels, assessed by flow cytometry (Fig. 1A)

FIG. 1. Expression of NK1R-GFP and ARR-GFP in KNRK cells.
To generate cells expressing NK1R-GFP, KNRK cells were transfected with NK1R-GFP (KNRK-NK1R-GFP cells). To generate cells expressing both the FlagNK1R plus ARR-GFP, KNRK-NK1R cells were transfected with cDNA encoding ARR-GFP (KNRK-NK1R ϩ ARR-GFP cells). A, analysis by flow cytometry with the GFP signal (vertical axis) to detect GFP-tagged proteins and the phycoerythrin (PE) signal (horizontal axis) to detect the FlagNK1R using a Flag M2 antibody and a phycoerythrin-conjugated secondary antibody. Note the low autofluorescence of untransfected KNRK cells and that a high proportion of cells express NK1R-GFP and FlagNK1R plus ARR-GFP. B, Western blots of untransfected KNRK cells, KNRK cells expressing the GFP vector without insert (KNRK-GFP), KNRK-NK1R-GFP cells, and KNRK-NK1R ϩ ARR-GFP cells (4 -15% SDS-PAGE, with 2 l of cell lysate/ lane). In the left panel the blot was probed with the GFP antibody. In the right panel the blot was probed with the GFP antibody preabsorbed with 1 g/ml GFP-GST that was used for immunization. Note the specific detection of proteins of the predicted masses in transfected cell lines. C, localization of GFP in KNRK-NK1R-GFP cells (upper panels) and KNRK-NK1R ϩ ARR-GFP cells (lower panels). The left panels show the GFP signal. The right panels show GFP immunoreactivity that was detected in the same cells using the GFP antibody and a secondary antibody coupled to Texas Red. The NK1R was present at the plasma membrane (arrowheads) and in some intracellular stores (arrows). In contrast, ␤-arrestin1 is in the cytosol (arrows). Scale bar ϭ 10 m. and microscopy (Fig. 5). Cells were studied at Ͻpassage 15 when they retained NK1R-GFP and NK1R plus ARR-GFP.
Western blotting verified expression of NK1R-GFP and ARR-GFP (Fig. 1B). In extracts of KNRK cells transfected with GFP vector without insert, the GFP antibody recognized a major protein with an apparent molecular mass of ϳ27 kDa, corresponding to the predicted size of GFP. A second minor protein was also detected, but its identity is unknown. In extracts of KNRK-NK1R-GFP cells, the GFP antibody recognized a broad band with an apparent molecular mass of ϳ90 kDa. The molecular mass of the rat NK1R is predicted to be ϳ46 kDa, but there are two potential N-linked glycosylation sites, and the glycosylated receptor has a larger mass (38). The broad protein band probably represents glycosylated NK1R-GFP. In extracts of KNRK-NK1R ϩ ARR-GFP cells, the GFP antibody recognized a major protein with an apparent molecular mass of ϳ75 kDa. This protein corresponds to the known mass of ␤-arres-tin1 (ϳ48 kDa) plus GFP (ϳ27 kDa). The GFP antibody did not interact with proteins in extracts of untransfected KNRK cells, and signals in the transfected cells were abolished by preabsorption of the antibody with 1 g/ml GFP-GST fusion protein.
These results confirm expression NK1R-GFP and ARR-GFP of the predicted sizes.
To determine the subcellular distribution of NK1R-GFP and ARR-GFP in unstimulated cells, we localized these proteins simultaneously using the GFP signal and immunofluorescence with the GFP antibody and a secondary antibody coupled to Texas Red. In KNRK-NK1R-GFP cells, NK1R-GFP was principally localized to the plasma membrane and also detected in an intracellular compartment (Fig. 1C, upper panels). In KNRK-NK1R ϩ ARR-GFP cells, ARR-GFP was uniformly distributed throughout the cytosol and was not found at the plasma membrane in vicinity of the NK1R (Fig. 1C, lower panels, arrows). Staining was similar by both methods, validating the specificity of the GFP antibody.

Functional Characterization of Cells Expressing NK1R-GFP and ARR-GFP
SP-induced Ca 2ϩ Mobilization-GFP is a compact protein, but its molecular mass is almost half that of the NK1R and ␤-arrestin1. To determine whether GFP interferes with signal transduction of the NK1R, we measured SP-induced Ca 2ϩ mobilization. SP stimulated a prompt but transient increase in [Ca 2ϩ ] i in both KNRK-NK1R-GFP cells and KNRK-NK1R ϩ ARR-GFP cells with a similar potency and efficacy (EC 50 ϳ0.5 nM) ( Fig. 2A). This potency is similar to that we have previously reported in KNRK cells expressing wild type NK1R and ␤-arrestins (ϳ0.6 nM) (38). Thus, GFP does not interfere with signal transduction of the NK1R. To determine if GFP interferes with either the ability of the NK1R to undergo desensitization or with the role of ␤-arrestin1 in desensitization, we compared desensitization of SP-induced Ca 2ϩ mobilization in cells expressing NK1R-GFP or FlagNK1R plus ARR-GFP. Exposure of KNRK-NK1R-GFP cells to 1 nM SP for 2 min caused a prompt increase in [Ca 2ϩ ] i , which rapidly declined to basal values even in the continued presence of SP (Fig. 2B). When cells were washed and exposed again to 1 nM SP 5 min after the first challenge, the response was strongly desensitized. Identical results were obtained in KNRK-NK1R ϩ ARR-GFP cells (Fig.  2C). This desensitization is similar to that we have reported in cells expressing wild type NK1R and ␤-arrestin (24). Thus, GFP does not interfere with NK1R desensitization or with the capacity of ␤-arrestin1 to participate in this process.
Internalization of 125 I-SP-To determine if endocytosis of the NK1R was affected by GFP, and whether GFP altered the ability of ␤-arrestin1 to participate in endocytosis, we quanti-fied internalization of 125 I-SP in cells expressing NK1R-GFP or FlagNK1R plus ARR-GFP. In KNRK-NK1R-GFP cells, 93.6 Ϯ 0.4% of specifically bound SP was at the surface and 5.0 Ϯ 0.1% was internalized after 60 min at 4°C (Fig. 2D). Warming to 37°C resulted in a rapid decline in surface label and a concomitant increase in internalized label, which was almost maximal after 20 min (18.9 Ϯ 5.9% surface, 79.1 Ϯ 5.3% internalized). Similar results were obtained using KNRK-NK1R ϩ ARR-GFP cells. At 4°C, 96.7 Ϯ 0.6% of specifically bound SP was at the surface and 5.3 Ϯ 0.8% was internalized (Fig. 2D). After 20 min at 37°C, 19.0 Ϯ 4.5% of specifically bound SP was at the surface and 79.4 Ϯ 5.0% was internalized. These rates of internalization are similar to those observed in KNRK cells expressing wild type NK1R and ␤-arrestins (32). Thus, GFP does not interfere with endocytosis of the NK1R or with the capacity of ␤-arrestin1 to participate in this process. Total specific binding at 4°C was similar in both cell lines (KNRK-NK1R-GFP, 16.7 Ϯ 5.4% total counts; KNRK-NK1R ϩ ARR-GFP, 16.7 Ϯ 5.6% total counts), which suggests that these cell lines expressed similar numbers of NK1Rs.

SP-induced Trafficking of NK1R-GFP and ARR-GFP in Real Time
The effects of SP on the subcellular distribution of the NK1R and ␤-arrestins have not been examined in real time. There- fore, we directly observed SP-induced trafficking of NK1R-GFP in KNRK-NK1R-GFP cells and ARR-GFP in KNRK-NK1R ϩ ARR-GFP cells over time.
NK1R-GFP-Before addition of SP, NK1R-GFP was principally localized at the plasma membrane (Fig. 3A). After 30 s incubation with 10 nM SP, NK1R-GFP was still mostly detected at the plasma membrane (Fig. 3B). After 10 min, NK1R-GFP was present in numerous endosomes localized beneath the plasma membrane and in a perinuclear location (Fig. 3C, arrows), and the intensity of the signal at the cell surface was diminished (Fig. 3C, arrowheads). After 60 min, NK1R-GFP was localized in a prominent perinuclear pool of endosomes, and there was diminished labeling of the plasma membrane (Fig. 3D, arrows). These results show that SP binds to NK1R-GFP in live cells and rapidly induces receptor internalization to superficial and then perinuclear endosomes.
ARR-GFP-Before addition of SP, ARR-GFP was uniformly distributed throughout the cytosol (Fig. 3E). Within 30 s of adding 10 nM SP, ARR-GFP was prominently detected at the plasma membrane, and the intensity of the cytosolic signal declined (Fig. 3F, arrowheads). After 10 min, ARR-GFP was detected in endosomes that were located immediately beneath the plasma membrane and in a perinuclear location (Fig. 3G,  arrows), and the intensity of the signal at the plasma membrane was diminished. After 60 min, ARR-GFP was localized in a prominent perinuclear pool of endosomes (Fig. 3H). Thus, when SP binds to the NK1R at the plasma membrane, a large proportion of ARR-GFP rapidly translocates to the plasma membrane. Membrane translocation of ARR-GFP precedes endocytosis of the NK1R but is quickly followed by redistribution of ARR-GFP into endosomes.

Colocalization of Cy3-SP with ARR-GFP
SP induced translocation of NK1R-GFP and ARR-GFP into vesicles that appeared to be similar in size, shape, and location. To determine if these proteins were colocalized, we used Cy3-SP to localize the NK1R. The advantage of using Cy3-SP to detect the NK1R is that it interacts initially only with functional receptors at the plasma membrane and, unlike a receptor antibody or a GFP tag, will not detect the NK1R in the biosynthetic pathway. Therefore, signals are especially clear. We have previously reported that Cy3-SP colocalizes with the wild type NK1R in KNRK cells (23). To determine if Cy3-SP similarly colocalizes with NK1R-GFP and whether GFP alters trafficking of the NK1R, we examined SP-induced trafficking in KNRK-NK1R-GFP cells. When incubated with cells at 4°C, Cy3-SP colocalized with NK1R-GFP at the cell surface (not shown). After 1-10 min at 37°C, Cy3-SP and NK1R colocalized in endosomes (not shown). We have previously shown that after internalization the NK1R recycles and SP is degraded in KNRK cells (23). Thus, Cy3-SP can be used to localize the NK1R during early stages of endocytosis. To determine if the NK1R colocalizes with ARR-GFP, we examined SP-induced trafficking in KNRK-NK1R ϩ ARR-GFP cells. After incubation with 100 nM Cy3-SP at 4°C, ARR-GFP (Fig. 4A) and Cy3-SP (Fig. 4B) was colocalized at the plasma membrane (Fig. 4C). After 1-10 min at 37°C, ARR-GFP and Cy3-SP were colocalized in the same superficial and perinuclear vesicles (Fig. 4, D-F). Because Cy3-SP colocalizes with the NK1R at the plasma membrane and in endosomes, these results indicate that the NK1R colocalizes with ␤-arrestins during the early stages of endocytosis.

Time Course of Colocalization of ARR-GFP with
Immunoreactive NK1R SP is degraded in lysosomes after endocytosis, and therefore Cy3-SP cannot be used to localize the NK1R during later stages of receptor trafficking. To permit simultaneous localization of the NK1R and ␤-arrestin1 in KNRK-NK1R ϩ ARR-GFP cells for prolonged periods, we localized the NK1R using a primary antibody to the C-tail and a Texas Red-conjugated secondary antibody, and we localized ␤-arrestin1 using GFP. In the absence of SP, NK1R was mainly localized to the plasma membrane (Fig. 5A, arrowheads), and ARR-GFP was distributed throughout the cytosol (Fig. 5B, arrows), with minimal colocalization (Fig. 5C). When cells were incubated with 10 nM SP for 60 min at 4°C, NK1R was principally localized to the cell surface (Fig. 5D), and ARR-GFP was redistributed from the cytosol to the plasma membrane (Fig. 5E), where it colocalized with the NK1R (Fig. 5F). After 5 min at 37°C, NK1R was detected in numerous superficial vesicles, and there was diminished NK1R immunoreactivity at the cell surface (Fig. 5G). ARR-GFP was no longer detected at the plasma membrane but was observed in vesicles of the same size, shape, and location as those containing NK1R, as indicated by superimposition of confocal images (Fig. 5, H and I). After 30 min, NK1R was

FIG. 3. SP-induced trafficking of NK1R-GFP in KNRK-NK1R-GFP cells (upper panels) and of ARR-GFP in KNRK-NK1R ؉ ARR-GFP cells (lower panels) in real time.
The same cells are shown in both rows. Cells were maintained at 37°C and observed before (A and E) and at 30 s (B and F), 10 min (C and G), and 60 min (D and H) after adding SP (10 nM). Note the redistribution of NK1R-GFP from the plasma membrane (A and B, arrowheads) to vesicles (C and D, arrows), and the translocation of ARR-GFP from the cytosol (E, arrows) to the plasma membrane (F, arrowheads), and to vesicles (G and H, arrows). Translocation of ARR-GFP preceded endocytosis of NK1R-GFP. These results are representative of at least 8 experiments. Scale bar ϭ 10 m.
localized to a prominent perinuclear pool of large vesicles (Fig.  5J). ARR-GFP was colocalized with the NK1R in these centrally located vesicles (Fig. 5, K and L). After 240 min, NK1R was prominently detected at the cell surface (Fig. 5M, arrowhead) with diminished localization in vesicles (Fig. 5M, arrow). ARR-GFP was still present in some centrally located vesicles that also contained the NK1R (Fig. 5, N and O, arrows). NK1R-GFP and ARR-GFP resumed their steady state distribution at the cell surface and in the cytosol, respectively, 6 h after exposure to SP. Thus, SP induces translocation of ␤-arrestin1 from the cytosol to the plasma membrane, where it colocalizes with the NK1R, followed by endocytosis of the NK1R and ␤-arres-tin1 to the same vesicles. The NK1R and ␤-arrestin1 remain colocalized until the NK1R returns to the plasma membrane and ␤-arrestin1 resumes its cytosolic distribution.

Identification of Organelles Containing NK1R-GFP and ARR-GFP
To identify organelles containing NK1R-GFP and ARR-GFP, we stained KNRK-NK1R-GFP cells and KNRK-NK1R ϩ ARR-GFP cells with antibodies to the transferrin receptor and to lysosomal-associated membrane protein-1 (LAMP-1). To prevent degradation of proteins in lysosomes and loss of signal, cells were incubated with 10 mM NH 4 Cl during the experiment (23). The transferrin receptor and LAMP-1 were detected by immunofluorescence using a Texas Red-conjugated secondary antibody. When cells were incubated with 10 nM SP for 60 min at 4°C, washed, incubated at 37°C for 30 -60 min, ARR-GFP (Fig. 6, A-C) and NK1R-GFP (not shown) were detected in vesicles in a central and peripheral location that also contained the transferrin receptor. Thus, these vesicles are early endosomes. In contrast, vesicles containing ARR-GFP (Fig. 6, D-F) and NK1R-GFP (not shown) were distinct from lysosomes. We have previously shown that SP induces redistribution of immunoreactive NK1R into early endosomes, which also contain the transferrin receptor, but not to lysosomes (23). Even when cells were incubated with 10 nM for 4 or 6 h at 37°C, NK1R-GFP and ARR-GFP were principally detected in vesicles that were distinct from lysosomes (not shown). Thus, GFP does not alter trafficking of the NK1R to endosomes, and SP induces translocation of both ARR-GFP and NK1R-GFP to early endosomes but not lysosomes.

SP-induced Trafficking of Immunoreactive NK1R, ␤-Arrestins, and G␣ q/11 in KNRK-NK1R Cells
SP-induced trafficking of the NK1R and ␤-arrestin1 could be affected by GFP or by overexpression of ␤-arrestin1. To determine whether SP causes a similar redistribution of endogenous ␤-arrestin1, we simultaneously localized ␤-arrestin1 with a monoclonal antibody and the NK1R with a polyclonal antibody in KNRK cells expressing FlagNK1R. To examine whether SP caused a similar redistribution of endogenous ␤-arrestin2, we localized ␤-arrestin2 using a polyclonal antibody and the NK1R using a monoclonal M5 Flag antibody. We similarly colocalized G␣ q/11 with a polyclonal antibody and NK1R with a monoclonal antibody to determine whether G␣ q/11 remains at the plasma membrane with the NK1R or is sorted to endosomes. Incubation of cells in SP-free medium at 4°C and warming to 37°C had no effect on the subcellular distributions of these proteins.
In cells that were incubated in SP-free medium at 4°C and immediately fixed, the NK1R was prominently localized to the plasma membrane, and ␤-arrestin1 was cytosolic (Fig. 7A-C). After incubation with 10 nM SP at 4°C, the NK1R was localized at the plasma membrane, and ␤-arrestin1 was detected at the plasma membrane and diffusely in the cytosol (Fig. 7, D-F). After 2-60 min at 37°C, ␤-arrestin1 colocalized with the NK1R in superficial and perinuclear endosomes (Fig. 7, G-I). SP had a similar effect on the subcellular localization of ␤-arrestin2. In the absence of SP, ␤-arrestin2 was principally cytoplasmic. SP induced translocation of ␤-arrestin2 to the plasma membrane, where it colocalized with the NK1R at 4°C and then to endosomes containing the NK1R at 37°C (Fig. 7, J-L). Similar results were obtained using a well characterized antibody that interacts with both ␤-arrestin1 ϩ 2 (Fig. 7, M-O). The NK1R and ␤-arrestins resumed their steady state distribution 6 h after exposure to SP. Thus, SP induces a similar redistribution of endogenous ␤-arrestin1 and ␤-arrestin2 and GFP-tagged ␤-arrestin1.
In cells that were incubated with 10 nM SP at 4°C, G␣ q/11 was detected at the plasma membrane where it colocalized with the NK1R (not shown). After 10 min at 37°C, G␣ q/11 remained at the plasma membrane, and the NK1R was detected in endosomes (Fig. 8, A and B). Thus, whereas SP induces membrane translocation of ␤-arrestins followed by endocytosis of ␤-arrestins and the NK1R into the same vesicles, SP has no effect on the subcellular distribution of G␣ q/11 . Specificity of Antibodies-We have previously shown that the Flag antibody and the C-terminal NK1R antibody specifically interact with the NK1R in KNRK-NK1R cells (38). Preincubation of antibodies to Flag, NK1R, and G␣ q/11 with the peptides used for immunization abolished the staining of cells (not shown), confirming specificity. The fusion proteins used to generate antibodies to ␤-arrestins were not available for preabsorption controls. However, these antibodies were affinity purified before use and have been characterized (5). To investigate further antibody specificity, we examined expression of ␤-arrestins and G␣ q/11 by Western blotting. Antibodies to ␤-ar-

FIG. 4. Localization of ARR-GFP (A and D) and Cy3-SP (B and E). KNRK-NK1R ϩ ARR-GFP cells were incubated
with 100 nM Cy3-SP for 60 min at 4°C, washed, and incubated at 37°C for 0 min (A-C) or 5 min (D-F). Images in the right panels are formed by superimposition of images from the two other panels in the same row. At 4°C, ARR-GFP and Cy3-SP colocalized at the plasma membrane (A-C, arrowheads). Thus, even at 4°C SP induces translocation of ARR-GFP to the plasma membrane. After 5 min at 37°C, ARR-GFP and Cy3-SP were in the same endosomes (D-F, arrows), and there was diminished surface labeling. Scale bar ϭ 10 m. restin1 and ␤-arrestin2 detected single proteins of ϳ50 and 45 kDa, respectively (Fig. 9). The antibody to ␤-arrestin1 ϩ 2 detected a broad band that appears to comprise two proteins of ϳ45-50 kDa. The antibody to G␣ q/11 detected a single prominent band of ϳ42 kDa. Preincubation of G␣ q/11 antibody overnight at 4°C with the peptide antigen diminished the signal (not shown). Thus, the antibodies specifically interact with proteins of the predicted sizes in KNRK cells. These results also confirm expression of ␤-arrestin1 and ␤-arrestin2 by KNRK cells.
The Role of ␤-Arrestin1 in Endocytosis of the NK1R-To determine the role of ␤-arrestin1 in endocytosis of the NK1R, KNRK-NK1R cells were transiently transfected with GFP vector (control), ARR-GFP, or dominant negative ARR 319 -418 -GFP  O). NK1R was localized using a C-terminal primary antibody and a secondary antibody coupled to Texas Red (left panels). ␤-Arrestin1 was detected using GFP (center panels). Images in the right panels are formed by superimposition of images from the two other panels in the same row. In untreated cells, NK1R was at the plasma membrane (A, arrowheads) and ARR-GFP was cytosolic (B, arrows), with minimal colocalization (C). After incubation with SP at 4°C, NK1R remained at the cell surface (D, arrowheads), and ARR-GFP was also detected at the plasma membrane (E, arrowheads), where it colocalized with the NK1R (F). After 5-30 min at 37°C, the NK1R and ARR-GFP colocalized in superficial and the centrally located endosomes (G-arrows). After 240 min, NK1R was detected at the cell surface (M, arrowhead) and in some perinuclear pools where it colocalized with ARR-GFP (N and O, arrow). Scale bar ϭ 10 m. In cells expressing vector without insert, GFP was uniformly distributed throughout cells. Binding and endocytosis of Cy3-SP were unaffected by expression of GFP (Fig. 10A). In cells transfected with wild type ␤-arrestin1-GFP, ARR-GFP was uniformly distributed throughout the cytosol before exposure to Cy3-SP (not shown). After 60 min at 4°C, Cy3-SP was detected at the plasma membrane of cells expressing ARR-GFP and in non-transfected cells (Fig. 10B, upper panel, arrowheads). This binding induced redistribution of ARR-GFP from the cytosol to the plasma membrane. After 5-30 min at 37°C, Cy3-SP was detected in superficial and then perinuclear endosomes in cells expressing ARR-GFP and in non-transfected cells (Fig. 10B, center and lower panels, arrows). At all time points, Cy3-SP colocalized with ARR-GFP in endosomes. Thus, expression of ARR-GFP does not affect binding or endocytosis of Cy3-SP.
In cells transfected with dominant negative ␤-arrestin 319 -418 -GFP, ARR 319 -418 -GFP was detected in superficial and perinuclear vesicles before exposure to Cy3-SP. This distribution was unaffected by Cy3-SP. At 4°C, Cy3-SP was detected at the plasma membrane of cells expressing ARR 319 -418 -GFP and in non-transfected cells (Fig. 10C, upper panels, arrowheads). After 5-30 min at 37°C, Cy3-SP was prominently detected at the plasma membrane in cells expressing ARR 319 -418 -GFP at high levels (Fig. 10C, center and lower panels, arrowheads). In marked contrast, Cy3-SP was detected in endosomes in a superficial and then perinuclear location in non-transfected cells. Expression of ARR 319 -418 -GFP strongly inhibits SP-induced endocytosis of the NK1R. Thus, ␤-arrestins play an important role in SP-induced endocytosis of the NK1R. DISCUSSION In unstimulated cells, The NK1R and G␣ q/11 colocalize at the plasma membrane. SP caused the following: (a) rapid (1 min) and striking translocation of ␤-arrestins from the cytosol to the plasma membrane, where they are required for NK1R endocy-tosis; (b) marked redistribution of ␤-arrestins and NK1R from the plasma membrane to the same endosomes (2-10 min), whereas G␣ q/11 remains at the plasma membrane; (c) prolonged (Ͼ60 min) association of ␤-arrestins and the NK1R in endosomes; and (d) gradual (4 -6 h) redistribution of ␤-arrestins to the cytosol and NK1R to the plasma membrane. To our knowledge, this is the first detailed examination of SPinduced trafficking of ␤-arrestins, and the first direct demonstration that ␤-arrestins participate in SP-induced endocytosis of the NK1R.
Expression of Functional NK1R-GFP and ARR-GFP-We placed GFP at the intracellular C terminus of the NK1R. Despite the importance of the NK1R C-tail for desensitization (39,40) and trafficking (41), comparisons of SP-induced Ca 2ϩ mobilization, uncoupling and endocytosis in cells expressing ARR-NK1R, with our previous reports in KNRK cells expressing wild type NK1R and in neurons that naturally express the NK1R, suggest that GFP does not affect the function of the NK1R. First, SP stimulated Ca 2ϩ mobilization in cells expressing NK1R-GFP with a similar potency to that which we have previously reported in KNRK cells expressing wild type NK1R (38). Second, the Ca 2ϩ response in cells expressing NK1R-GFP was transient in the continued presence of SP and desensitized to repeated stimulation by SP. We have previously shown that wild type NK1R similarly desensitizes in transfected cells and neurons (24,34). Finally, SP induced endocytosis and trafficking of NK1R-GFP in a similar manner to that reported for wild type NK1R in KNRK cells and neurons (23,25,32). Thus, NK1R-GFP behaves similarly to wild type NK1R in transfected cells and in cells that naturally express this receptor (23-25, 34, 38). In support of our results, others have reported that GFP does not affect signaling or trafficking of cholecystokinin A and ␤ 2 -adrenergic receptors (42)(43)(44).
Because GFP has been attached to the C terminus of ␤-ar-restin2 without affecting its function (18), we placed GFP at the C terminus of ␤-arrestin1. The C terminus of ␤-arrestins interacts with clathrin (45) to mediate endocytosis, and N-terminal domains interact with GRK-phosphorylated GPCRs to uncou-FIG. 6. Identification of organelles containing ARR-GFP. KNRK-NK1R ϩ ARR-GFP cells were incubated with 10 nM SP for 60 min at 4°C, washed, and incubated at 37°C for 60 min. Endosomes were localized with an antibody to the transferrin receptor and a Texas Red-conjugated secondary antibody. Lysosomes were localized with the GM10 antibody LAMP-1 and a Texas Red-conjugated secondary antibody. Images in the right panels are formed by superimposition of images from the two other panels in the same row. A-C, ARR-GFP (A, arrows) colocalized with endosomes (B, arrows). D-F, ARR-GFP (D, white arrows) did not colocalize with lysosomes (E, yellow arrows). Scale bar ϭ 10 m.
ple them from G-proteins (46). Comparisons of cells expressing ARR-GFP, with our previous observations in KNRK cells and neurons expressing wild type ␤-arrestins, suggest that GFP does not affect the function of ␤-arrestin1. SP-induced endocytosis of the NK1R and desensitization of SP-induced Ca 2ϩ mobilization in cells expressing ARR-GFP in a similar manner to its effects on endocytosis and desensitization of the NK1R in cell lines and neurons that express wild type ␤-arrestins (23-25, 34, 38). Furthermore, SP stimulated a similar redistribution of ARR-GFP and immunoreactive ␤-arrestins in KNRK KNRK-NK1R cells were incubated in the absence (control) or presence of 10 nM SP for 60 min at 4°C, washed, and incubated at 37°C for various times. A-I, NK1R was detected using a polyclonal antibody, and ␤-arrestin1 was detected with a monoclonal antibody. J-O, NK1R was detected using a Flag monoclonal antibody, and ␤-arrestin2 or ␤-arrestin1 was detected with a polyclonal antibody. The NK1R was visualized with fluorescein isothiocyanate-conjugated secondary antibodies (left panels); ␤-arrestins were detected using Texas Red-conjugated secondary antibodies (center panels). Images in the right panels are formed by superimposition of images from the two other panels in the same row. A-I, localization of NK1R and ␤-arrestin1. In the absence of SP (A-C), the NK1R was at the plasma membrane (A, arrowheads); ␤-arrestin1 was cytosolic and in vesicles (B, arrows), and there was no colocalization (C). After incubation with SP at 4°C, the NK1R (D, white arrowheads) and ␤-arrestin1 (E, yellow arrowheads) colocalized at the plasma membrane (F). After 15 min at 37°C, the NK1R (G) and ␤-arrestin1 cells and neurons (34). Thus, ARR-GFP participates as expected in endocytosis and uncoupling of the NK1R, and ARR-GFP redistributes similarly to wild type ␤-arrestins in cell lines and neurons. These observations suggest that GFP does not alter the function or trafficking of ␤-arrestin1.
SP-induced Trafficking of the NK1R, G␣ q/11 , and ␤-Arres-tin1/2-SP caused rapid endocytosis of the NK1R, whereas G␣ q/11 remained at the plasma membrane. Thus, the NK1R may couple to G␣ q/11 at the plasma membrane and rapidly dissociate. Studies in reconstituted systems and in membrane assays also suggest that the NK1R couples to G␣ q/11 (28,36). The ␤ 2 AR and G s ␣ similarly colocalize at the plasma membrane of unstimulated cells, but agonists cause redistribution of G s ␣ to the cytosol and endocytosis of the ␤ 2 -AR (47).
SP stimulated a striking, rapid, and transient redistribution of ␤-arrestin1 and ␤-arrestin2 from the cytosol to the plasma membrane, followed by internalization of the NK1R and ␤-arrestins into the same endosomes. SP similarly stimulates translocation of ␤-arrestins to the plasma membrane and endosomes of neurons (34). Remarkably, membrane translocation of ␤-arrestins occurred even at 4°C, whereas endocytosis of the NK1R is temperature-dependent. This rapid membrane translocation may be due to the cytosolic localization of ␤-arrestins. In support of our results, agonists of several GPCRs induce rapid translocation of ␤-arrestins to the plasma membrane and endosomes (7,18). Although the mechanism by which SP induces membrane translocation of ␤-arrestins is unknown, receptor phosphorylation by GRKs increases the affinity with which ␤-arrestins interact with GPCRs (48). It is likely that SP also induces translocation of GRK2/3 from the cytosol to the plasma membrane of KNRK cells. In support of this suggestion, SP stimulates a minor and transient translocation of GRK2/3 to the plasma membrane of neurons expressing the NK1R (34). Membrane targeting of GRK2/3 entails their interaction with ␤␥ subunits of heterotrimeric G-proteins, a precise mechanism because free ␤␥ subunits are found in the plasma membrane at sites of receptor activation (16,17). GRK2/3 probably phosphorylate the NK1R at the plasma membrane and thereby facilitate interaction of the NK1R and ␤-arrestins, which mediate uncoupling and endocytosis. GRK2/3 phosphorylate the NK1R in a reconstituted system and in membranes (28,29). The NK1R possess numerous potential phosphorylation sites in the C-tail, and truncation of the C-tail impairs desensitization (39,40) and endocytosis (41). GRK2/3 also phosphorylate the ␤ 2adrenergic and m2 muscarinic cholinergic receptors to promote uncoupling and endocytosis (2,6,(11)(12)(13)49). Phosphorylation may also regulate the activity of ␤-arrestins (50). ␤-Arrestins are constitutively phosphorylated in the cytosol, and dephosphorylation, which occurs at the plasma membrane, is required for their participation in receptor endocytosis.
␤-Arrestins participate in endocytosis of other GPCRs that internalize by clathrin-mediated mechanisms. Overexpression . KNRK-NK1R cells were incubated with 10 nM SP for 60 min at 4°C, washed, and incubated at 37°C for 10 min. The NK1R was detected with the monoclonal Flag M5 antibody and a fluorescein isothiocyanate-conjugated secondary antibody (left panel). G␣ q/11 was detected using a polyclonal antibody and Texas Red-conjugated secondary antibodies (center panel). The images in the right panel are formed by superimposition of images from the two other panels. The NK1R was present in early endosomes (A, arrow), whereas G␣ q/11 remained at the plasma membrane (B, arrowheads). The NK1R and G␣ q/11 did not colocalize in endosomes (C). Scale bar ϭ 5 m.
FIG. 9. Characterization of antibodies by Western blotting. Extracts of KNRK cells (50 g protein/lane) were separated on 12% SDS-PAGE gels and probed with antibodies to ␤-arrestin1, ␤-arrestin2, ␤-arrestin1 ϩ 2, and G␣ q/ll . of dominant negative mutants ␤-arrestin1-V53D and ␤-arrestin 319 -418 inhibits endocytosis of the ␤ 2 AR (8 -10), whereas overexpression of wild type ␤-arrestins promotes agonist-induced endocytosis of ␤ 2 -adrenergic and m2 muscarinic acetylcholine receptors (8,52). In contrast, agonist-induced endocytosis of the angiotensin II type 1A receptor and the m1, m3, and m4 muscarinic cholinergic receptors does not depend on ␤-arrestins (9,14). Some GPCRs, such as the endothelin 1 and cholecystokinin A receptors, internalize in part by caveolindependent pathways (53,54), but the potential role of ␤-arrestins in this pathway is unknown. Since the mechanism of endocytosis is distinct for different GPCRs and may also depend on the cellular environment, it is important to examine receptor regulation in cells that naturally express receptors at physiological levels. Thus, our report that SP induces translocation of ␤-arrestins to the plasma membrane of neurons suggests that ␤-arrestins play a physiological role in regulating the NK1R (34).
An unexpected observation was the prolonged colocalization of the NK1R and ␤-arrestins in the same endosomes. We do not know whether the NK1R is physically associated with ␤-arrestins in endosomes, but the prolonged colocalization suggests that ␤-arrestins regulate internalized receptors. Resensitization of responses to SP in KNRK cells and neurons is blocked by endocytic inhibitors, phosphatase inhibitors, and acidotropic agents that prevent receptor recycling, suggesting that the NK1R must be internalized, dephosphorylated, and recycled for resensitization to occur (23)(24)(25). Similarly, endocytosis and recycling are important for resensitization of the ␤ 2 AR (19 -22). Whether ␤-arrestins participate in intracellular trafficking or signaling of GPCRs is unknown, although endocytosis is required for the ␤ 2 AR to stimulate mitogen-activated protein kinases (15). The mechanism by which the NK1R and ␤-arrestins redistribute to different compartments is also unknown. Dephosphorylation of the NK1R in acidified endosomes may be required for dissociation of ␤-arrestins. In support of this possibility, endosomes are enriched in phosphatases that dephosphorylate GPCRs (55), and endosomal acidification is also necessary for dephosphorylation of the ␤ 2 -AR (56).
␤-Arrestins may also interact with the GRK-phosphorylated . Cells were transfected as described under "Experimental Procedures." After 72 h, cells were incubated with 100 nM Cy3-SP for 60 min at 4°C, and were either immediately fixed (0 min, upper panels) or washed and incubated for 10 (center panels) or 30 min (lower panels) at 37°C and then fixed. The same cells are shown in each row. The GFP signal is shown in the left panels, and the Cy3-SP signal is shown in the right panels. * denotes the transfected cells. A, GFP was distributed throughout the cells, and the distribution was unaffected by exposure to Cy3-SP. At 0 min, Cy3-SP was present at the plasma membrane (arrowheads). After 10 and 30 min, Cy3-SP was detected in endosomes (arrows). Binding and internalization of Cy3-SP was unaffected by expression of GFP. B, At 0 min, ARR-GFP was present at the plasma membrane where it colocalized with Cy3-SP (arrowheads). After 10 and 30 min, ARR-GFP and Cy3-SP colocalized in prominent endosomes (arrows) and were not detected at the plasma membrane. Binding and internalization of Cy3-SP was unaffected by expression of ARR-GFP. C, ARR 319 -418 -GFP was present in vesicles in a superficial and perinuclear location that was unaffected by exposure to Cy3-SP. At 0 min, ARR 319 -418 -GFP was present at the plasma membrane (arrowheads). After 10 and 30 min, Cy3-SP was detected in prominent endosomes in non-transfected cells (arrows). In contrast, at these times Cy3-SP was retained at the plasma membrane of cells expressing ARR 319 -418 -GFP (arrowheads). Comparison of cells expressing ARR 319 -418 -GFP and either non-transfected cells or cells expressing ARR-GFP indicated that expression of this dominant negative mutant inhibited endocytosis of Cy3-SP. These results are representative of Ͼ5 experiments. Scale bar ϭ 10 m.
NK1R at the plasma membrane to mediate uncoupling of the NK1R from G␣ q/11 and thereby terminate signal transduction, since inositol pentakisphosphate, which disrupts the interactions of ␤-arrestins with receptors, attenuates desensitization of the NK1R (30). This possibility could be confirmed by overexpression of the dominant negative mutant ␤-arrestin1-V53D, which inhibits interaction of endogenous ␤-arrestins with GPCRs.
In summary, SP induces a marked redistribution of ␤-arrestins to the plasma membrane where they participate in clathrin-mediated endocytosis of the NK1R.