β-Arrestins Regulate Interleukin-8-induced CXCR1 Internalization*

The functional role of neutrophils during acute inflammatory responses is regulated by two high affinity interleukin-8 receptors (CXCR1 and CXCR2) that are rapidly desensitized and internalized upon binding their cognate chemokine ligands. The efficient re-expression of CXCR1 on the surface of neutrophils following agonist-induced internalization suggests that CXCR1 surface receptor turnover may involve regulatory pathways and intracellular factors similar to those regulating β2-adrenergic receptor internalization and re-expression. To examine the internalization pathway utilized by ligand-activated CXCR1, a CXCR1-GFP construct was transiently expressed in two different cell lines, HEK 293 and RBL-2H3 cells. While interleukin-8 stimulation promoted CXCR1 sequestration in RBL-2H3 cells, receptor internalization in HEK 293 cells required co-expression of G protein-coupled receptor kinase 2 and β-arrestin proteins. The importance of β-arrestins in CXCR1 internalization was confirmed by the ability of a dominant negative β-arrestin 1-V53D mutant to block internalization of CXCR1 in RBL-2H3 cells. A role for dynamin was also demonstrated by the lack of CXCR1 internalization in dynamin I-K44A dominant negative mutant-transfected RBL-2H3 cells. Agonist-promoted co-localization of transferrin and CXCR1-GFP in endosomes of RBL-2H3 cells confirmed that receptor internalization occurs via clathrin-coated vesicles. Our data provides a direct link between agonist-induced internalization of CXCR1 and a requirement for G protein-coupled receptor kinase 2, β-arrestins, and dynamin during this process.

A general characteristic of inflammatory responses is the migration of leukocytes from the blood to sites of injury or infection. A number of chemoattractants have been shown to cause the directed migration of leukocytes in vitro and in vivo. These include complement fragment C5a, formylated bacterial peptides (fMLP), arachidonic acid metabolites (LTB4), and a group of low-molecular weight pro-inflammatory cytokines known as chemokines (1)(2)(3). The superfamily of chemokines is subdivided into different subsets based on the presence and positioning of highly conserved cysteine residues. The family, based on the configuration of the first two N-terminal located cysteines, is divided into CC, CXC, and CX 3 C subfamilies. All known neutrophil-targeted human chemokines belonging to the CXC subfamily (IL-8, 1 GRO␣, GRO␤, NAP-2, ENA78, and GCP-2) are defined by the presence of a glutamic acid-leucinearginine motif (ELR motif) in the portion of the molecule that lies N-terminal to the first highly conserved cysteine, thus representing the ELR-CXC chemokine subclass of CXC chemokines (3)(4)(5).
IL-8 and other neutrophil-directed chemokines stimulate neutrophils via specific seven-transmembrane guanine nucleotide-binding protein-coupled receptors (GPCRs) (5,7). The two human IL-8 receptors, CXCR1 and CXCR2, have 77% overall sequence homology. The two receptor subtypes differ notably in their N-terminal extracellular domains, as well as in their C-terminal intracellular domains, and possess differences in their ligand specificities. CXCR1 displays greater ligand specificity by binding to IL-8 and GCP-2 with high affinity, whereas CXCR2 binds with high affinity multiple CXC chemokines in addition to IL-8, including ENA 78, NAP-2, GRO␣, and GRO␤ (6 -9). Binding of the ligand to high affinity IL-8 receptors initiates a variety of cellular responses, including calcium translocation, chemotaxis, alterations in cytoskeletal architecture as well as cellular morphology, degranulation, and respiratory burst activation (3, 10 -12). ELR-CXC chemokines are produced by a variety of cell types including monocytes, T lymphocytes, fibroblasts, and endothelial cells (3,5,6).
It has been well documented that IL-8 receptors become rapidly desensitized and internalized upon agonist stimulation (13,14). The molecular mechanism(s) and cellular factors required for translocation of these agonist-occupied receptors from the membrane to cytosolic compartments are not well characterized. However, the rapid sequestration and re-expression of CXCR1 (14,16) is similar to the well described model of ␤ 2 -adrenergic receptor (␤ 2 -AR) regulation.
In the case of ␤ 2 -AR, agonist binding induces a change in the receptor conformation, which is necessary for the interaction of the receptor with G protein-coupled receptor kinases (GRKs) (17,19). GRK-mediated phosphorylation of the ␤ 2 -AR C terminus promotes binding of arrestin proteins (␤-arrestins) which when bound, elicit uncoupling of the receptor from its G protein (18 -20). Recent data suggests that the synergistic action of cellular GRKs and ␤-arrestins determines the kinetics of ␤ 2 -AR internalization (21). Moreover, it was demonstrated that ␤-arrestins serve as adaptor proteins, specifically targeting agonistoccupied receptors to clathrin-coated vesicles (CCVs) (19,20,22). A critical step in receptor-mediated endocytosis of ␤ 2 -AR is the translocation of CCVs saturated with agonist-occupied receptor to the cytosol, which is a dynamin-regulated event (23). Desensitized ␤ 2 -ARs, internalized via CCVs, are thought to be resensitized in the acidified endosomal environment and recycled back to the cell surface to re-establish normal receptor signaling (24).
In the present work we examined the role of GRK2, ␤-arrestins, and dynamin in regulating CXCR1 internalization. For this purpose a CXCR1-green fluorescent protein (GFP) construct (CXCR1-GFP) was transiently expressed in human embryonic kidney 293 (HEK 293) and rat basophilic leukemia 2H3 (RBL-2H3) cell lines. We demonstrate that GRK2, ␤-arrestins, and dynamin are required for rapid agonist-induced internalization of CXCR1.

EXPERIMENTAL PROCEDURES
Materials-HEK 293 and RBL-2H3 cell lines were obtained from American Type Culture Collection (Manassas, VA). Dulbecco's modified essential medium and Eagle's modified essential medium (EMEM) were purchased from Biowhittaker (Walkerswille, MD). Chemiluminescent substrates and horseradish peroxidase-coupled donkey anti-rabbit antibody were purchased from Amersham (Amersham International). The plasmid containing a variant of a green fluorescence protein (pEGFP-N1) and a green fluorescence protein directed polyclonal antibody were purchased from CLONTECH (Palo Alto, CA). Human recombinant interleukin-8 (IL-8) was obtained from R&D Systems (Minneapolis, MN) and 125 I-IL-8 was obtained from Amersham. Texas Red-transferrin conjugates were purchased from Molecular Probes (Eugene, OR).
Construction of CXCR1-GFP-RNA from human neutrophils was isolated using TriPure TM Isolation Reagent (Roche Molecular Biochemicals) and CXCR1 cDNA synthesized using Superscript TM II reverse transcriptase (Life Technologies, Inc.) according to the manufacturer's instructions. The coding sequence of CXCR1 was amplified using forward (AAGAGGACATGTCAAATATTACAGAT) and reverse (TTCATC-GATGGTTTTCCGAGG) primers carrying EcoRI restriction enzyme recognition sequence. Polymerase chain reaction products were 5Ј and 3Ј terminal digested with EcoRI and cloned into pEGFP-N1 cloning vector. The final construct was sequenced through the region that was generated by polymerase chain reaction to confirm sequence fidelity.
Cell Cultures and Transfections-HEK 293 cells were grown in Dulbecco's modified essential medium whereas RBL-2H3 cells were grown in EMEM, both containing 10% fetal bovine serum and 1:100 dilution of penicillin/streptomycin (BioWhittaker) at 37°C in a humidified atmosphere of 95% air and 5% CO 2 . HEK 293 cells were seeded 12 h prior to transfection in 35-mm glass bottom plates (MatTek Corp., Ashland, MA) at a density of 2 ϫ 10 5 cells per dish and RBL-2H3 cells were seeded at a density of 1 ϫ 10 5 cells per dish 3 h prior to transfection. HEK 293 cells were transiently transfected with 5 g of CXCR1-GFP and 7.5 g of pcMV5 rat ␤-arrestin 1/pcMV5 rat ␤-arrestin 2, and/or 7.5 g of pcDNA1-Amp rat GRK2. For RBL-2H3 cells, 3 h prior to transfection the cells were washed and incubated in serum-free EMEM and then transiently transfected with 5 g of CXCR1-GFP alone or along with 5 g of pcDNA1-Amp rat ␤-arrestin 1-V53D or 5 g of pCB1 rat dynamin I-K44A. Cell lines were transfected with LipofectAMINE (Life Technologies, Inc.) following the manufacturer's instructions. Following transfection the cells were maintained in fresh complete medium for 12 h to recover. For stable transfections, RBL-2H3 cells were grown to 80% confluence in 100-mm dishes (Falcon) and transfected with 10 g of CXCR1-GFP in 30 l of LipofectAMINE. Three days after transfection, cells were harvested, diluted, and replated in media supplemented with 1 mg/ml of Geneticin (Life Technologies, Inc.). The media was replaced every 4 days and stable transformants were isolated approximately 3 weeks after transfection. Clonal selection was confirmed by observation of cells under confocal microscope.
Radioligand Sequestration Assay-RBL-2H3 cells were subcultured overnight in 6-well plates (1 ϫ 10 6 cells/well) in complete EMEM. Cells were washed twice with serum-free EMEM containing 1% bovine serum albumin and 25 mM HEPES (pH 7.2) and preincubated in the same media 1 h prior to 125 I-IL-8 treatment. Nontransfected RBL-2H3 and RBL-2H3 cells stably expressing CXCR1-GFP were stimulated with 50 nM 125 I-IL-8 (3000 Ci/mmol) at 4 and 37°C for 45 min. The reaction was stopped with 1 ml of ice-cold PBS (pH 7.4) supplemented with 1% bovine serum albumin, the cells were washed with the same buffer three times and lysed in the lysis buffer (0.5% Nonidet P-40 and 0.5% Triton X-100 in PBS) on ice for 30 min. After incubation, the total volume of the well was transferred onto 10% sucrose, PBS cushion and pelleted at 10,000 rpm for 20 min. Equal volumes of the supernatant (100 l) were aliquoted and the amount of 125 I-IL-8 in supernatants was measured using a ␥-counter (LKB/Wallac, Turku, Finland). Incorporation of nonspecific radioactivity was determined in supernatants of nontransfected RBL-2H3 cells.
Secretion of ␤-Hexosaminidase-Cells were seeded as described for the "Radioligand Sequestration Assay," washed, and preincubated in serum-free EMEM for 30 min. Nontransfected and RBL-2H3 cells stably expressing CXCR1-GFP were then stimulated with 50 nM IL-8 for 60 min at 37°C. The reaction was terminated by placing 6-well plates on ice for 15 min. The amount of the secreted ␤-hexosaminidase was determined by incubating 50 l of the overlaying medium with 50 l of 1 mM p-(nitrophenyl)-N-acetyl-␤-D-glucosamide in 0.1 M sodium citrate buffer (pH 4.5) at 37°C for 1 h. At the end of the incubation 500 l of a 0.1 M Na 2 CO 3 , NaHCO 3 buffer (pH 4.5) was added and the absorbance was measured at 400 nm.
Confocal Microscopy of Single Cell Time Courses and Colocalization Studies-Confocal microscopy was performed on a Bio-Rad MRC-600 confocal microscope under ϫ 60 oil immersion objective, using a fluorescein isothiocyanate filter with the emission wavelength of 488 nm. Transiently transfected HEK 293 and RBL-2H3 cells were maintained in fresh complete media. For time course studies, the cells were treated with increasing concentrations of IL-8 (10, 25, 40, 50, and 75 nM) and events following agonist stimulation were observed in 5-min time intervals up to 90 min post-agonist stimulation. To determine the effect of de novo protein synthesis that occurs during the time of observation, RBL-2H3 cells transiently expressing CXCR1-GFP were first pretreated with cyclohexamide (10 ng/ml) for 45 min at 37°C and then stimulated with 50 nM IL-8. Events following agonist treatment were observed under confocal microscope.
For colocalization studies, RBL-2H3 cells transiently expressing CXCR1-GFP were stimulated with 50 nM IL-8 and labeled with Texas Red-transferrin conjugates (15 ng/ml) for 45 min at 37°C. The reaction was terminated by washing the cells twice with ice-cold PBS (pH 7.4). The cells were fixed in 3.6% paraformaldehyde solution and confocal microscopy was performed as described above.
Subcellular Cell Fractionation-HEK 293 cells transiently expressing CXCR1-GFP, CXCR1-GFP, and GRK2, and CXCR1-GFP, GRK2, and ␤-arrestin 1 were stimulated with 50 nM IL-8 for 45 min at 37°C. The cells were washed twice with ice-cold PBS (pH 7.4), removed from plates by gentle washing, and pelleted at 100 rpm for 10 min. The cell pellet was resuspended in 3 ml of buffer A (10 mM Tris-HCl, pH 7.4, 2 mM EDTA), incubated on ice for 30 min and homogenized using a Dounce homogenizer. Nuclei were removed by centrifugation at 200 rpm for 10 min. The supernatant was loaded on a stepwise sucrose cushion (35 and 5% sucrose in PBS) and centrifuged at 35,000 rpm for 90 min at 4°C. The supernatant was removed and the 35% sucrose interface fraction containing endosomes (the light vesicular fraction) was collected, diluted in buffer A, and re-centrifuged at 35,000 rpm for 45 min at 4°C. The pellets were resuspended in 100 l of buffer A containing 2 ϫ SDS sample buffer and 100 g of each protein sample was loaded onto SDS-polyacrylamide gel electrophoresis.
Protein Determination-Protein levels in the whole cell lysates of HEK 293 and RBL-2H3 cells were determined using Bio-Rad protein assay (Richmond, CA) with bovine serum albumin as a standard.
Western Blotting-Expression levels of ␤-arrestin 1, ␤-arrestin 2, and GRK 2 in HEK 293 and RBL-2H3 cells were examined using specific polyclonal antisera as described previously (20). Equivalent amounts (100 g) of total cell protein were separated on a 10% polyacrylamide gel and transferred onto nitrocellulose membrane (Bio-Rad). The endogenous amounts of ␤-arrestin 1, ␤-arrestin 2, and GRK 2 were determined using anti-␤-arrestin 2 and anti-GRK2 rabbit polyclonal sera at a dilution of 1:2500 and horseradish peroxidase-conjugated anti-rabbit secondary antibodies using the ECL system (Amersham) according to manufacturer's instructions. The amount of total ␤-arrestin 1, ␤-arrestin 2, and GRK 2 in RBL-2H3 cells were determined relative to their respective endogenous expression in HEK 293 cells. Amounts of CXCR1-GFP in the light vesicular subcellular fraction were determined using GFP-directed polyclonal antibody at dilution 1:1000 (CLONTECH).
Statistical Analysis of the Sequestration Data-The relative membrane and cytosol luminosity was measured using SigmaScan Pro software. Data was statistically analyzed and plotted using Microsoft Excel software. Results are the average Ϯ S.D. from three separate identical experiments.

RESULTS
Agonist-promoted Internalization of CXCR1-GFP in HEK 293 and RBL-2H3 Cells-Cells transfected with CXCR1-GFP were positive for fusion protein expression within 24 -36 h post-transfection as evidenced by robust membrane fluorescence in 15% of the RBL-2H3 and 70% of the HEK 293 cells visualized by confocal microscopy. In some transfected cells under nonstimulated conditions the CXCR1-GFP conjugate accumulated in the Golgi apparatus. Dose-response experiments indicated that maximal internalization of CXCR1-GFP occurred in a range of 40 -75 nM IL-8. Over a 45-min time period, agonist-occupied CXCR1-GFP conjugates carried in specific membrane-associated vesicles, gradually translocated from the plasma membrane to the cytosol (Fig. 1A, i). Sequestration data acquired in RBL-2H3 cells showed a mean decrease of 61% in membrane luminosity and a 4.2-fold increase in cytosol fluorescence intensity in response to agonist stimulation over the 45-min time period (Fig. 1B). Although there was a significant decrease of membrane luminosity in response to IL-8 treatment in RBL-2H3 cells transiently expressing the CXCR1-GFP construct not all of the expressed receptor was internalized 45 min post-stimulation. Pretreatment of RBL-2H3 cells transiently expressing the CXCR1-GFP fusion construct with cyclohexamide resulted in rapid CXCR1-GFP sequestration with no membrane fluorescence after 45 min post-stimulation, indicating that residual membrane fluorescence was due to de novo synthesis of CXCR1-GFP conjugates (data not shown). Unstimulated cells showed very little redistribution of CXCR1- GFP over the same time frame (Fig. 1A, ii). Stably transfected CXCR1-GFP cells internalized 125 I-IL-8 at 37°C whereas at 4°C, CXCR1-GFP transfected cells failed to internalize 125 I-IL-8 (Fig. 1C), indicating that internalization of CXCR1-GFP is an agonist and temperature-dependent process, which is similar to IL-8 receptor internalization observed in neutrophils (14). To assess whether CXCR1-GFP transduced functional responses, we performed ␤-hexosaminidase assays on IL-8 stimulated and unstimulated CXCR1-GFP stably transfected cells.
IL-8 stimulation of CXCR1-GFP transfected cells resulted in a 13.4% release of hexosaminidase compared with a 4.8% release from untransfected RBL-2H3 cells stimulated with IL-8 (Fig. 1D). Stimulation with MCP-1, a CC chemokine that does not bind CXCR1, did not induce hexosaminidase release. These results show that CXCR1-GFP expressed in RBL-2H3 cells retains several features of the wild type receptor expressed in neutrophils; the receptor can transduce signals that result in granule release, undergo agonist-induced internalization, and sequester IL-8.
In contrast to RBL-2H3 cells, HEK 293 cells transiently expressing the fusion protein construct did not internalize CXCR1-GFP when stimulated with IL-8 ( Fig. 2A, i). Since previous studies (30) have demonstrated that HEK 293 cells require increased expression of ␤-arrestins and GRKs for internalization of some GPCRs we explored whether co-expression of these two classes of molecules with CXCR1-GFP could restore agonist-induced receptor internalization in HEK 293 cells.

Inhibition of CXCR1-GFP Sequestration in RBL-2H3 Cells by Overexpression of ␤-Arrestin 1-V53D and Dynamin I-K44A
Mutants-To explore the role of ␤-arrestins in CXCR1-GFP internalization in RBL-2H3 cells we co-expressed CXCR1-GFP along with the ␤-arrestin 1-V53D dominant negative mutant in RBL-2H3 cells and stimulated with IL-8. In the presence of ␤-arrestin 1-V53D there was no redistribution of membrane fluorescence that followed IL-8 stimulation (Fig. 4, A, ii, and B). This was in sharp contrast to cells expressing CXCR1-GFP alone (Fig. 4A, i) or cells expressing CXCR1-GFP and wild type ␤-arrestin 1 (data not shown), which showed marked receptor internalization following IL-8 stimulation. These observations complement the results obtained with HEK 293 cells and clearly demonstrate a role of ␤-arrestins in agonist-induced CXCR1 internalization.
␤-Arrestins are thought to act as scaffolding proteins in coupling GPCRs to CCVs (22, 24 -26). Agonist stimulation promotes the formation of CXCR1-GFP containing vesicles, which are pinched off from the plasma membrane and translocated into post-endocytic compartments (19,24,26). The pinching or sealing off of the vesicles from the plasma membrane is dependent upon dynamin, a GTPase containing molecule (27)(28)(29). The dominant negative dynamin I-K44A mutant has been utilized in determining whether GPCRs are internalized via a dynamin-dependent pathway involving CCVs. We explored whether CXCR1 required dynamin for agonist-induced receptor internalization by co-expressing CXCR1-GFP with the dy-namin I-K44A dominant negative mutant. The expression of dynamin I-K44A successfully blocked redistribution of CXCR1-GFP from the membrane to the cytosol. Vesicles formed in cells expressing the dynamin I-K44A mutant simply did not pinch off from the inner surface of the plasma membrane (Fig. 4A, iii,  45 min). These results indicate that agonist-induced internalization of CXCR1 occurs via CCVs and requires functional ␤-arrestins and dynamin molecules.
Agonist-induced Colocalization of Transferrin and CXCR1-GFP in Endosomes and the Presence of CXCR1-GFP Conjugates in the Light Vesicular Subcellular Fraction-To further investigate and confirm the identity of membrane-derived vesicles that translocate agonist-occupied CXCR1-GFP from the membrane to post-endocytic compartments, we labeled RBL-2H3 cells transiently expressing the receptor-GFP construct with a Texas Red-transferrin conjugate. Transferrin has been shown to undergo receptor-mediated endocytosis through CCVs upon binding to its cognate transferrin receptor and it has been described as a significant endosomal marker (37,38). Agonist stimulation promoted colocalization of CXCR1-GFP and dye-labeled transferrin conjugate within CCVs (Fig. 5, ii) whereas unstimulated RBL-2H3 cells transiently expressing CXCR1-GFP did not display any colocalization (Fig. 5, i). These results were further supported by isolation of the light vesicular (endosomal) subcellular fraction from transiently transfected HEK 293 cells expressing CXCR1-GFP and CXCR1-GFP, GRK2, and ␤-arrestin 1 that were stimulated or unstimulated with IL-8. A 9.5-fold increase in CXCR1-GFP was found in the light vesicular subcellular fraction isolated from IL-8 stimulated HEK 293 cells expressing CXCR1-GFP, GRK2, and ␤-arrestin 1 (Fig. 6, lane 4 versus lane 2). However, only a modest increase (3-fold) in CXCR1-GFP was found in the light vesicular subcellular fraction isolated from HEK 293 cells in the absence of transfected GRK2 and ␤-arrestin 1 (Fig. 6, lane 3 versus lane 1) indicating that GRK2 and ␤-arrestin 1 substantially enhance the efficiency of CXCR1 internalization. Together these results support a model for CXCR1 sequestration via clathrin-coated pits that are regulated by GRKs, ␤-arrestins, and dynamin. DISCUSSION Agonist-dependent regulation of chemokine receptor desensitization, internalization, and sequestration is an important mechanism for regulating leukocyte responsiveness to chemokine stimulation. Several studies have demonstrated that exposure of neutrophils to high concentrations of ELR-CXC chemokines renders the exposed neutrophils unresponsive to additional homologous chemokine stimulation. The refractory state of neutrophils following stimulation with high chemokine concentrations appears to be dependent upon desensitization, internalization, and sequestration of CXCR1, CXCR2, or both IL-8 receptor subtypes. Initial studies using radiolabeled IL-8 showed that IL-8-binding sites were rapidly lost from the neutrophil surface following stimulation of high concentrations of IL-8 (13). These initial studies were later confirmed by additional work utilizing monoclonal antibodies directed at the external domains of CXCR1 and CXCR2, that demonstrated a rapid loss of CXCR1 and CXCR2 following stimulation of neutrophils with high concentrations of ELR-CXC chemokines (14,33).
Recent studies by Richardson et al. (32) have demonstrated that phosphorylation of critical serine residues in the C-terminal region of CXCR1 is important for internalization of the receptor following agonist stimulation. Our work here compliments these observations by demonstrating that GRK2 is necessary for internalization of CXCR1 in HEK 293 cells (Fig. 2A,  iv and v). GRK2 is a serine-threonine kinase and a member of a multigene family whose members regulate GPCR function and internalization by phosphorylating serine/threonine residues located within the cytoplasmic regions of various receptors (17). GRK2 is abundantly expressed in human peripheral leukocytes (Ref. 31 and data not shown) and may represent the endogenous kinase responsible for CXCR1 phosphorylation in neutrophils. Alternatively one of the other members of the GRK family (GRK1 and 3-5) may serve the same function in regulating phosphorylation of CXCR1 in neutrophils. CCR5, a member of the CC chemokine family of receptors (4), is preferentially phosphorylated by GRK2 and GRK3 indicating that GRK phosphorylation likely represents a common feature of both CC and CXC chemokine receptor regulation (30).
While GRK phosphorylation represents a critical step in regulating the desensitization and internalization of a subset of GPCRs, it is the ␤-arrestin proteins which facilitate the translocation of GPCRs from the plasma membrane to CCVs. Our data in the present study places ␤-arrestins as central regulators of CXCR1 internalization in response to agonist stimulation. This has been substantiated using two separate cells lines displaying two different phenotypes. HEK 293 cells which have low expression of ␤-arrestins require expression of ␤-arrestin 1 or ␤-arrestin 2 for agonist induced internalization. In sharp contrast to HEK 293 cells, RBL-2H3 cells express higher levels of ␤-arrestins and do not require additional expression of ␤-arrestins for CXCR1 internalization (Fig. 3A). However, agonistinduced internalization of CXCR1 could be blocked by co-expressing the dominant negative ␤-arrestin 1-V53D mutant in RBL-2H3 cells (Fig. 4, A, ii, and B). These experiments provide strong evidence for ␤-arrestin regulation of agonist-induced CXCR1 internalization. Additionally, it is also clear from our studies that cellular factors other than GRK2 and ␤-arrestin are involved in the CXCR1 internalization machinery. Dynamins have been previously described as key proteins involved in the pinching off or sealing of CCVs from the membrane by stimulating GTP/GDP exchange which facilitates endocytic vesicle release (27,28). In contrast to the angiotensin II type 1A receptor, which is able to undergo dynamin-independent endocytosis (23), our studies indicate that in the presence of dynamin I-K44A mutant, CCVs saturated with agonist-occupied receptor are not released from the membrane into the cytosolic compartment (Fig. 4A, iii). Thus CXCR1 appears to undergo internalization and sequestration through a dynamin-driven and clathrin-dependent internalization pathway similar to several other GPCRs (20,26,30). This is supported by two additional pieces of data, agonistpromoted colocalization of transferrin and CXCR1-GFP in endosomal vesicles of RBL-2H3 cells (Fig. 5) and redistribution of CXCR1-GFP to the light vesicular fraction following IL-8 stimulation in HEK 293 cells transiently expressing CXCR1-GFP, GRK2, and ␤-arrestin 1 (Fig. 6). Even though CXCR1 and CXCR2 display 77% amino acid identity and elicit several similar functional responses they appear to have divergent pathways for internalization and recycling. Chuntharapai et al. (14) showed that CXCR1 but not CXCR2 was recycled back to the plasma membrane following agonist-induced internalization. The mechanism for the divergence of CXCR1 and CXCR2 recycling is presently unknown, although differences in the C-terminal region may play a role in how the two proteins undergo intracellular trafficking (15). In this context CXCR1 and CXCR2 may have differential requirements for ␤-arrestin proteins, which target CXCR1 and CXCR2 to different intracellular compartments. Preliminary data in our laboratory suggests that CXCR2 internalization is regulated differently from CXCR1 in HEK293 cells. The precise role of ␤-arrestins in regulating CXCR1 and CXCR2 recycling awaits further investigations. It is interesting to note that ␤-arrestins can function as signaling molecules since studies have implicated a role for ␤-arrestins in the activation of tyrosine kinases (36). Thus, it is conceivable that some of the functional responses elicited by CXCR1 are due to signals transduced by ␤-arrestins coupling to the CXCR1 receptor.
While there are at least five independent mechanisms for endocytotic internalization including the clathrin-and nonclathrin-coated pits, micropinocytosis, caveolae, and phagocytosis, GPCRs appear to utilize only two: clathrin-dependent and dynamin-independent endocytotic pathways (23,34,35). We present here a previously undescribed model for CXCR1 chemokine receptor internalization whereby we have demonstrated that GRK2, ␤-arrestin, and dynamin are necessary molecules for the entry of CXCR1 into the cell. Upon IL-8 binding, GRK2 phosphorylates C-terminal serine-threonine residues on CXCR1 allowing ␤-arrestins to couple phosphorylated receptor to cytoplasmic complexes containing clathrin. Furthermore, our data demonstrates that dynamin is required to pinch off CCVs containing CXCR1 and allow vesicular entry of the activated receptor into the cell. The importance of chemokine receptor internalization may be to serve as a mechanism of reducing the chemotactic activity of leukocytes under conditions of high exposure to inflammatory stimuli thereby preventing their continued migration and departure from the site of inflammation. These studies provide insight into the biochemical factors involved in chemokine receptor entry into the cell and thus may further our understanding of the inflammatory process.