Journal of Cell Science 111, 2819-2830 (1998) 2819 Printed in Great Britain © The Company of Biologists Limited 1998 JCS4614

Differential regulation of CXCR4 and CCR5 endocytosis

Nathalie Signoret1, Mette M. Rosenkilde2, P. J. Klasse1, Thue W. Schwartz2, Michael H. Malim3, James A. Hoxie4 and Mark Marsh1,* 1Medical Research Council Laboratory for Molecular Cell Biology and Department of Biochemistry, University College London, Gower Street, London WC1E 6BT, UK 2Department of Pharmacology, The Panum Institute, Blegdamsvej 3, Copenhagen DK-2200, Denmark 3Howard Hughes Medical Institute and Departments of Microbiology and Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA 19104, USA 4Department of Medicine, University of Pennsylvania, 415 Curie Blvd, Philadelphia, PA 19104, USA *Author for correspondence (e-mail: [email protected])

Accepted 22 July; published on WWW 27 August 1998

SUMMARY

The chemokine receptors CCR5 and CXCR4 are major co- rich domain is required for ligand- but not phorbol ester- receptors/receptors for the CD4-dependent and CD4- induced CXCR4 internalisation. However, a Ser/IleLeu independent entry of human and simian immunodeficiency motif, similar to that required for the endocytosis of CD4 viruses. The chemokines that bind and activate these and the T cell receptor/CD3 complex, is required for receptors can inhibit the entry of viruses that use the phorbol ester-induced, but not ligand-induced, CXCR4 respective co-receptor molecules. Chemokine-induced co- endocytosis. By contrast, CCR5 internalisation is induced receptor internalisation is a significant component of the by the β-chemokine RANTES but not by phorbol esters. mechanism through which chemokines inhibit virus entry. CCR5 lacks the Ser/IleLeu sequence required for phorbol CXCR4 internalisation is induced by the CXCR4 ligand ester-induced uptake of CXCR4. Together these results stromal cell derived factor-1 (SDF-1), phorbol esters and, indicate that distinct mechanisms can regulate CXCR4 and in T cells, cellular activation. Here we show that CXCR4 CCR5 endocytosis and trafficking. endocytosis can be mediated through either one of two Key words: CXCR4, CCR5, SDF-1α, RANTES, Phorbol ester, distinct internalisation signals. A COOH-terminal serine Internalisation motif

INTRODUCTION been shown to inhibit HIV entry (Bleul et al., 1996; Cocchi et al., 1995; Oberlin et al., 1996; Simmons et al., 1997). For Chemokine receptors are members of the seven transmembrane CCR5, the β-chemokines RANTES, MIP-1α, MIP-1β and domain (7TM) G protein-coupled receptor superfamily several RANTES antagonists inhibit infection by M-tropic (GPCRs) that mediate chemotactic activity in leukocytes but HIV-1 strains (Arenzana-Seisdedos et al., 1996; Cocchi et al., are expressed on a wide range of cell types (Murphy, 1994; 1995; Simmons et al., 1997), while stromal cell-derived factor Premack and Schall, 1996). Several chemokine receptors have 1 (SDF-1α and SDF-1β) and eotaxin can inhibit the entry of been shown to act as co-receptors for the human and simian viruses that use CXCR4 and CCR3, respectively (Bleul et al., immunodeficiency viruses either independently of (Endres et 1996; Choe et al., 1996; Oberlin et al., 1996). The al., 1996), or together with, CD4 (Berger, 1997; Moore et al., mechanism(s) through which ligands and antagonists block 1997). CCR5 has been implicated as the major co-receptor for virus infection has been unclear. One model is that the ligands SIV and M-tropic strains of HIV-1 (Alkhatib et al., 1996; Deng block infection sterically by preventing interaction of the viral et al., 1996; Dragic et al., 1996) while CXCR4 permits entry envelope protein (Env) gp120 with critical sites on the co- of T-tropic and CD4-independent strains of HIV (Berson et al., receptor molecule. An alternative view is that ligand-induced 1996; Endres et al., 1996; Feng et al., 1996). Other chemokine chemokine receptor internalisation would effectively remove receptors, such as CCR2b (Doranz et al., 1996) and CCR3 the co-receptor from the cell surface (Wells et al., 1996). While (Choe et al., 1996; He et al., 1997) in addition to chemokine these two modes of action are not mutually exclusive, two receptor-like orphan proteins such as STRL33/Bonzo recent studies have suggested that chemokine-induced (Alkhatib et al., 1997; Deng et al., 1997), GRP-15/BOB (Deng internalisation of CCR5 and CXCR4 can contribute to the et al., 1997; Farzan et al., 1997) and GRP1 (Farzan et al., mechanism of chemokine-mediated inhibition of virus entry 1997), can also function as co-receptors for M and/or T tropic (Amara et al., 1997; Signoret et al., 1997). Furthermore, a HIV and SIV strains (Berger, 1997; Berger et al., 1998; Moore RANTES antagonist AOP-RANTES that is a particularly et al., 1997). potent inhibitor of M-tropic strains of HIV-1 (Simmons et al., In several cases chemokines or chemokine antagonists have 1997), induces very efficient CCR5 down-modulation by 2820 N. Signoret and others promoting receptor internalisation and blocking recycling stromal cell-derived factor 1α (SDF-1α) was expressed in and purified (Mack et al., 1998). from Escherichia coli (provided by Dr Mike Luther, Glaxo Wellcome The mechanisms regulating the endocytosis and trafficking Inc, RTP, NC, USA). This SDF-1α had an additional N-terminal of chemokine receptors are not well understood. For CXCR4, methionine residue that was not removed following synthesis and purification; however its properties were indistinguishable from those the SDF-1 receptor, ligand and phorbol ester treatment of cells α induce rapid internalisation of cell surface receptors (Amara et of a chemically synthesised SDF-1 (Crump et al., 1997; Signoret et al., 1997). al., 1997; Forster et al., 1998; Signoret et al., 1997). For T cells, stimulation through CD28, CD3 or CD2 can also down Antibodies modulate cell surface CXCR4 (Jourdan et al., 1998) and may Anti-human CXCR4 monoclonal antibodies 44702 and 44716 and be the physiological counterpart of the pathway activated by anti-human CCR5 mAbs 45517, 45523, 45529, 45531 and 45533 phorbol esters. Internalisation, at least in the case of phorbol were kindly provided by Dr Monica Tsang (R&D Systems, esters, occurs through coated pits and results in delivery of the Minneapolis Minnesota; USA). The anti-CD4 antibody, OKT4, was receptor and ligand to endosomes, from where the receptor is purchased from Ortho Diagnostics Systems Inc. (Raritan, New able to recycle to the cell surface (Amara et al., 1997; Signoret Jersey). The anti-CXCR4 mAb 12G5 (IgG2a) and the anti-CD4 mAb Q4120 (IgG1) were described previously (Endres et al., 1996; Healey et al., 1997). For CCR5, ligand also induces uptake into 125 endosomes (Amara et al., 1997; Mack et al., 1998), most likely et al., 1990). 12G5 was I-labelled using Bolton and Hunter reagent (Amersham International plc) as described (Signoret et al., 1997). through coated vesicles (Mack et al., 1998). For this receptor, Specific activities of 300-400 Ci/mmol were obtained for different G-protein coupled receptor kinases (GRKs) and members of iodinations. Radioiodinated proteins, diluted in PBS containing 0.25% β the family of -arrestins can facilitate internalisation (Aramori gelatin and 0.02% NaN3 and stored in small aliquots at −20¡C, were et al., 1997). In our studies of CXCR4 we have shown that stable for up to 4 months. phorbol esters and SDF-1α induce internalisation through distinct biochemical mechanisms. Phorbol ester-induced Cells uptake is inhibited by staurosporine and calphostin C Retrovirus mediated gene transfer was used to stably express the suggesting a role for PKC, while ligand-induced uptake is not CCR5 chemokine receptor in CEM-SS cells (CEMss/CCR5) such that (Signoret et al., 1997). As phorbol ester-induced uptake of CD4 they were rendered susceptible to HIV-1YU-2 infection. To achieve and the T cell antigen receptor (TcR) is known to involve this, a 100 mm-diameter culture of 293T cells was transiently transfected with 10 µg each of the vectors LP-M/CCR5, pHIT/G and phosphorylation-dependent endocytosis signals that are pHIT60 using calcium phosphate as described (Fouchier et al., 1997). recognised by the AP2 clathrin adaptor complex (Dietrich et Virus supernatant was then harvested at 24 hours and used to infect a al., 1997; C. Pitcher, and M. Marsh, unpublished), we parental CEM-SS culture. A stably transduced cell population suggested that CXCR4 might be capable of interacting with CEMss/CCR5 was selected and maintained in complete RPMI 1640 both the AP2 complex and with β-arrestin through distinct medium supplemented with 10% foetal calf serum (FCS) and 0.2 signalling motifs (Signoret et al., 1997). µg/ml puromycin. Previous studies have indicated that deletion of the carboxyl Mink Mv-1-Lu cells and Mv-1-Lu cells stably expressing human terminal domain of CXCR4 abrogates both SDF-1 and phorbol CD4 (Mv-1-Lu-CD4) were obtained from the MRC AIDS Reagents ester-induced receptor internalisation (Amara et al., 1997; Programme (NIBSC, Potters Bar, UK) and maintained in DMEM Haribabu et al., 1997; Signoret et al., 1997), suggesting that containing 10% FCS, 2 mM glutamine, 100 U/ml penicillin and 0.1 mg/ml streptomycin (PenStrep), with or without 1 mg/ml G418 as sequences required for endocytosis might reside within this required. Mv-1-Lu cells were transfected by electroporation with domain. Comparison of the COOH-terminal sequences from wild-type or mutated human CXCR4 cDNAs (see below) inserted in the established chemokine receptors indicates that although the expression vector pTEJ8 (Johansen et al., 1990). Stable these cytoplasmic domains have some features in common, transfectants were selected in the presence of 1 mg/ml G418 and there is also significant variability. In this paper we show: (1) colonies were screened for cell surface expression by indirect that in contrast to CXCR4, CCR5 is not down modulated by immunofluorescence using the anti-CXCR4 mAb 12G5 as previously treatment of cells with phorbol esters, (2) we identify a phorbol described (Signoret et al., 1997). Similarly, Mv-1-Lu-CD4 cells were ester-induced internalisation signal in CXCR4 similar to transfected with pTEJ8 vectors containing human CXCR4 or CCR5 signals previously identified in CD4 and the CD3γ subunit of cDNAs, together with the plasmid pBABE-hygro encoding the the TcR, and (3) we demonstrate that mutations in this signal hygromycin B resistance gene (Morgenstern and Land, 1990). Clones were selected in medium containing 1 mg/ml G418 and 500 µg/ml abrogate CXCR4 responses to phorbol ester but not to ligand. hygromycin B and screened for CXCR4 or CCR5 expression by Together these results demonstrate that the cell surface immunofluorescence using the 12G5 mAb or a mix of anti-CCR5 expression of CXCR4 and CCR5 can be modulated through mAbs (45517, 45523 and 45531), respectively. distinct biochemical mechanisms. Furthermore, the results COS-7 cells were cultured in DMEM F12 containing 5% FCS, 2 suggest that CXCR4 can interact with distinct sets of adaptor mM glutamine and PenStrep. Cells were transiently transfected using proteins that affect endocytosis. calcium phosphate as previously described (Kledal et al., 1997). Mutagenesis Truncated CXCR4 molecules were generated using human CXCR4 MATERIALS AND METHODS constructs in which stop codons were inserted at codon positions 322, 330 or 341 by site directed mutagenesis. Point mutations were made Reagents in the CXCR4 cDNA in order to replace specific amino acids (S324 Tissue culture reagents were from Gibco Ltd (Paisley, UK), and other and S325, I328 and L329) with alanine residues. The mutations were chemicals were from Sigma (Poole, UK), unless otherwise indicated. introduced by the polymerase chain reaction-mediated overlap Tissue culture plastic was from Nunc, and radioactive reagents were extension technique, as previously described (Rosenkilde et al., 1994). from Amersham International plc (Little Chalfont, UK). Recombinant All constructs were cloned into pTEJ8 vector and sequenced. Differential regulation of CXCR4 and CCR5 endocytosis 2821

Flow cytometry/FACS analysis At the indicated times the cells were returned to 4¡C, the media Adherent cells were detached in PBS containing 10 mM EDTA and collected, and the cells washed with cold BM. For each time point six resuspended in culture medium. After centrifugation (2,000 rpm, 5 wells were used. For half of the wells, the cells were collected directly minutes), the cells were resuspended in PBS containing 0.1% bovine in 400 µl 0.2 M NaOH and transferred to tubes for γ-counting (total serum albumin (BSA) and 1% formaldehyde and left to fix overnight cell-associated activity). To determine the intracellular activity, the at 4¡C. After two washes in wash buffer (WB; PBS containing 0.1% remaining wells were rinsed twice with 0.5 ml 4¡C BM adjusted to BSA and 0.05% azide), cells were labelled with primary antibody in pH 2.0 and then incubated 2× 3 minutes with 1 ml of the same medium WB for 1 hour at room temperature. Subsequently, the cells were to remove cell surface antibody (acid-resistant activity). The cells stained in WB with a FITC-conjugated goat anti-mouse antibody were harvested in NaOH as above. The proportion of internalised (1/100; Pierce and Warriner [UK] Ltd, Chester, UK) for a further hour. activity for each time point was determined by dividing the acid- Cells were analysed using a Becton Dickinson FACSCalibur flow resistant activity by the total cell-associated activity. cytometer (Becton-Dickinson UK Ltd, Oxford, UK). SDF-1α mediated CXCR4 down modulation Phorbol ester induced receptor down modulation on Cells plated in 16 mm well 24-well plates were incubated in 37¡C BM CEMss/CCR5 cells with or without SDF-1α (125 nM) at 37¡C as indicated. After CEMss/CCR5 cells (2×107) were collected and washed twice in treatment, the cells were placed on ice and cooled by addition of 1 ml binding medium (BM: RPMI-1640 without bicarbonate, containing of ice-cold BM. After 4 washes with ice-cold BM and one 5 minutes 0.2% BSA, 10 mM Hepes, and adjusted to pH 7.4) at room wash in cold BM adjusted to pH 3.0, the cells were returned to pH temperature. Cells were resuspended in 7 ml of BM and one aliquot 7.4 BM. The cells were then labelled with 300 µl of 0.5 nM 125I-12G5 of 1 ml (3×106 cells) was put directly on ice in cold BM for time point for 2 hours at 4¡C under agitation. Subsequently, the cells were t=0. The rest of the cells were centrifuged (1,500 rpm, 5 minutes) and washed again in cold BM, harvested in 400 µl of 0.2 M NaOH and resuspended in 6 ml of prewarmed BM containing 100 ng/ml phorbol the bound radioactivity determined as above. 12-myristate 13-acetate (PMA). At each time point a 1 ml aliquot 125 α (3×106 cells) was transferred into a 15 ml tube containing 5 ml of cold I-SDF-1 binding BM and kept on ice. When the last sample was collected, cells were 125I-SDF-1α was prepared by oxidative iodination using Iodogen pelleted by centrifugation (1,500 rpm, 5 minutes), and washed twice (Pierce Warriner [UK] Ltd) followed by HPLC purification to separate in cold WB before fixation for FACS analysis. For each time point unlabelled from labelled compound. COS-7 cells transiently aliquots of 2.5×105 cells were placed in the well of a 96-well plate, expressing CXCR4 and CXCR4 mutants were used for SDF-1α labelled for either CD4, CXCR4 or CCR5 and analysed by flow competition binding assays as previously described (Rosenkilde et al., cytometry. 1994). Briefly, cells (at a density of 0.75 to 3×105 cells per well) were incubated for 3 hours at 4¡C in 0.5 ml binding buffer (BB: 25 mM Immunofluorescence microscopy Hepes buffer, pH 7.2, with 1 mM CaCl2, 5 mM MgCl2, and Mv-1-Lu-CD4 CXCR4 and CCR5 cells were grown on coverslips for supplemented with 0.5% BSA) containing 10 pM of 125I-SDF-1α and 48 hours before the experiment. The cells were washed in BM at room an increasing amount of unlabeled SDF-1α. The incubation was temperature and then incubated in BM with or without PMA (100 stopped by washing four times with 0.5 ml of ice-cold BB made 0.5 ng/ml), SDF-1α (125 nM) or RANTES (250 nM) at 37¡C. After 45 M with NaCl. Radioactivity was extracted from the cells with 8 M minutes the coverslips were placed on ice, washed with cold BM and urea in 3 M acetic acid and 1% NP-40. the cells incubated for 5 minutes in cold BM adjusted to pH 3.0 to remove the surface-bound ligand as previously described (Signoret et al., 1997). The cells were then returned to pH 7.4 by washing with RESULTS BM and fixed in PBS containing 3% PFA for 10 minutes. Free aldehyde groups were quenched using 50 mM NH4Cl in PBS for at CXCR4 but not CCR5 is internalised after PMA least 20 minutes, and the cells then treated with PBS containing 0.05% saponin (PBS-Sap) for 15 minutes, washed and stained with treatment antibodies diluted in PBS-Sap containing 0.2% gelatin. Cells were We previously demonstrated that the receptor for the CXC incubated with primary antibodies for 1 hour, either in a mix of anti- chemokine SDF-1α, CXCR4, is internalised from the cell CCR5 mAbs 45517, 45523 and 45531 (10 µg/ml each) or the anti- surface and down modulated when cells are treated with CXCR4 mAb 12G5 (5 µg/ml) or the anti-CD4 Q4120 (5 µg/ml). The phorbol esters (Signoret et al., 1997). To determine whether cells were washed in PBS-Sap and labelled for 1 hour with a biotin- other chemokine receptors involved in mediating HIV entry are conjugated anti-mouse antibody in PBS-Sap containing 0.2% gelatine similarly regulated we investigated the cell surface expression (1/2,000; Amersham International plc). After further washes in PBS- of CCR5 on cells treated with phorbol esters or ligands. We Sap, the coverslips were stained for 30 minutes with FITC-conjugated streptavidin (1/100; Amersham International plc). Subsequently the first used transfected mink lung epithelial cells (Mv-1-Lu) in coverslips were mounted in Mowiol and examined using a Nikon which we previously found CXCR4 to exhibit responses to Optiphot-2 microscope equipped with an MRC Bio-Rad 1024 laser phorbol esters and ligand similar to those seen in T cell lines scanner. The images were assembled in, and printed directly from, (Signoret et al., 1997). Mv-1-Lu-CD4 cells, which stably Adobe Photoshop. express human CD4, were transfected with cDNAs encoding either human CXCR4 or human CCR5 and clones selected for Endocytosis assays stable cell surface expression of the chemokine receptors. Endocytosis assays on adherent cells were performed essentially as Chemokine receptor expression was monitored by FACS described (Pelchen-Matthews et al., 1989, 1991; Signoret et al., 1997). analysis (not shown) and immunofluorescence (see below). Briefly, adherent cells were seeded in 16 mm wells in 24-well plates and grown for 2 days to a final density of 1-2×105 cells per well. The The distribution of CXCR4 and CCR5 was examined on cells were cooled on ice, washed with BM and incubated for 2 hours permeabilised cells by immunofluorescence, either before or at 4¡C with 300 µl of 0.5 nM 125I-12G5 antibody in BM. after treatment with phorbol ester or ligand (Fig. 1). We also Subsequently, the cells were washed in BM to remove free antibody, monitored the distribution of CD4 on these cells. On untreated and then warmed by addition of 1 ml 37¡C BM with or without PMA. cells, CXCR4 and CCR5 were present mainly on the plasma 2822 N. Signoret and others A B

Fig. 1. Cellular distribution of CXCR4 and CCR5 in MV-I-Lu-CD4 cells. Mv-1-Lu-CD4 expressing CXCR4 (A) or CCR5 (B) were incubated for 45 minutes at 37¡C in medium alone (I) or with medium containing 100 ng/ml PMA (II), 250 nM RANTES (III) or 125 nM SDF-1α (IV), respectively. The cells were subsequently fixed, permeabilised, stained for CXCR4 and CD4 (A) or CCR5 and CD4 (B) and viewed by confocal microscopy. Bar, 20 µm. membrane as indicated by the diffuse fluorescence illustrated in punctate intracellular structures on cells treated with phorbol Fig. 1A and B, panels I. Similar staining was seen on intact cells ester (Fig. 1A, CD4 panel II). Neither SDF-1α nor RANTES confirming that the antigens were located primarily on the cell appeared to affect the distribution of CD4 (not shown). When surface (not shown). When cells were stained for CD4, diffuse cells expressing CCR5 were examined, we found no detectable cell surface fluorescence and some intracellular punctate change in the cell surface staining on SDF-1α treated cells (Fig. fluorescence was seen (Fig. 1A and B, CD4 panels I). This 1B, CCR5 compare panels I and IV), whereas RANTES internal fluorescence likely corresponds to CD4 molecules induced internalisation of the receptor into intracellular located in the endocytic and exocytic pathways, as previously organelles (Fig. 1B, CCR5 panel III). Significantly, PMA had described (Marsh and Pelchen-Matthews, 1996). We then little effect on the distribution of CCR5 and did not appear to examined whether the distribution of the chemokine receptors induce redistribution of the cell surface receptor (Fig. 1B, was changed when cells were treated for 45 minutes at 37¡C CCR5 panel II). The CCR5 expressing cells were, however, with ligand (SDF-1α for CXCR4 and RANTES for CCR5) or able to respond to phorbol ester as demonstrated by their ability the phorbol ester PMA. As previously observed (Signoret et al., to down modulate CD4 when treated with PMA (Fig. 1B, CD4 1997), we found that SDF-1α and PMA induced the panel II). These observations indicated that although both redistribution of cell surface CXCR4 to intracellular structures CXCR4 and CCR5 were induced to internalise by their specific which appeared punctate by immunofluorescence and were ligands, CCR5 does not undergo the phorbol ester-induced often located in regions of the cell adjacent to the nucleus (Fig. internalisation seen for CXCR4. 1A, CXCR4 panels II and IV). By contrast, the CC chemokine RANTES had no effect on CXCR4 distribution (Fig. 1A, Chemokine receptor modulation on T cells CXCR4 panel III). Cell surface CD4 was redistributed into To determine whether CCR5 cell surface expression also fails Differential regulation of CXCR4 and CCR5 endocytosis 2823

to be modulated by phorbol esters in T cells we compared the A properties of this receptor expressed on CEMss/CCR5 cells. These cells were derived from the CEMss T cell line after transfection with cDNA for human CCR5. CEMss cells constitutively express both CD4 and CXCR4. Following transfection and selection, CCR5 could also be detected on the surface of these cells by FACS (Fig. 2A). This allowed us to investigate the regulation of surface expression for the three markers simultaneously. After incubation of CEMss/CCR5 cells at 37¡C in the presence of 100 ng/ml PMA for up to 2

Cell counts hours, surface expression of CD4, CXCR4 and CCR5 was determined using antibodies against the specific receptors and flow cytometry (Fig. 2B). On cells treated with PMA we found that cell surface expression of both CD4 and CXCR4 declined with time and was reduced by 60-80% after 2 hours of treatment (Fig. 2B panel a). The kinetics of down modulation Fluorescence for CXCR4 on these cells was slower than that previously seen with SupT1 cells (Signoret et al., 1997). Nevertheless, down modulation was reproducibly detected using several different B 150 anti-CXCR4 mAbs. When CCR5 expression was monitored, a CD4 CXCR4 we saw little change in cell surface expression over 2 hours 125 Q4120 12G5 44702 using 5 different anti-CCR5 mAbs (Fig. 2B, panel b). OKT4 44716 The observation that CCR5 and CD4 are internalised 100 independently supports the notion that there is little constitutive interaction of these two molecules on the surfaces 75 of cells in the absence of HIV (Signoret et al., 1997). Moreover, 50 the trimolecular complexes of Env, CD4 and co-receptor that have been proposed to form during viral entry (Lapham et al., 25 1996; Trkola et al., 1996; Ugolini et al., 1997) are likely to be induced directly by the viral Env protein. 0 CXCR4 contains an S(x)nLL type motif 150 B/Bo% b Chemokine receptors and related 7TM GPCRs have been 125 shown to undergo endocytosis on binding their specific ligands (Amara et al., 1997; Aramori et al., 1997; Ferguson et al., 100 1996a; Haribabu et al., 1997; Mueller et al., 1997; Prado et al., 1996; Signoret et al., 1997; Solari et al., 1997; von Zastrow 75 and Kolbika, 1992). Serine rich COOH-terminal domains of 45517 several 7TM GPCRs have been implicated in these events, 50 45523 though no distinct internalisation motifs has been identified. 45529 α 25 We previously demonstrated that SDF-1 and phorbol ester- 45531 induced internalisation of CXCR4 occurs through distinct 45533 0 biochemical mechanisms. Endocytosis induced through both 0 30 60 90 120 reagents requires elements in the COOH-terminal domain of CXCR4, as deletion of the bulk of this domain abrogates both time (min) ligand and phorbol ester-induced uptake. However, phorbol α Fig. 2. PMA treatment of CEMss/CCR5 cells. (A) Cell-surface ester, but not SDF-1 -induced internalisation is inhibited by expression of CXCR4, CCR5 and CD4 on CEMss/CCR5 cells was staurosporin and calphostin C, implicating a role for protein measured by FACScan analysis of fluorescently labelled cells. Anti- kinase C only in phorbol ester-induced internalisation CD4 monoclonal antibody Q4120 (...), anti-CXCR4 mAb 12G5 (Signoret et al., 1997). (—) and anti-CCR5 mAb 45517 (—). Cells labelled with the second We have previously shown that the cell surface expression layer, goat anti-mouse-FITC, were used as controls (solid histogram). of CD4 can be modulated by phorbol esters (Hoxie et al., 1988; Each histogram indicates the fluorescence intensity for 10,000 events Pelchen-Matthews et al., 1993). For CD4, endocytosis is acquired. (B) Cells were incubated in medium containing 100 ng/ml dependent on a signal in the 37 amino acid cytoplasmic domain PMA for up to 2 hours at 37¡C. At the indicated times an aliquot of of the molecule that involves a pair of leucine residues (L413 cell-suspension was taken and fixed. The cells were then labelled and L414 in human CD4) and the phosphorylation of an with a series of mAbs specific for CXCR4, CD4 (a) and CCR5 (b). Cells were then stained with a FITC-conjugated antibody and adjacent serine residue (S408) (Pelchen-Matthews et al., 1993; analysed by flow cytometry. The graph shows the mean fluorescence C. Pitcher and M. Marsh, unpublished; Shin et al., 1991). intensity (MFI) for PMA treated cells as a proportion of the MFI for Mutation of either L413, L414 or S408 blocks constitutive and untreated cells (B/Bo%) at the indicated times. The MFIs were phorbol ester-induced endocytosis of CD4. Phosphorylation of determined from 10,000 accumulated events for each sample. S408 activates the signal and leads to association of CD4 with 2824 N. Signoret and others

A TM 7 Serine rich region

CXCR1 L G FLH S CLNP IIYAF I G Q N F R H G FLK I L AMH ---G LVSK EFL ARH RVT SYT - S S SVNVS SNL CXCR2 L G ILH S CLNPLIYAF I G Q K F R H G LLK I L AIH ---G LISK DSL P K DSRPSRPG S S S G H T S TTL CXCR3 L G YMH CCLNPLLYAFV G V K F RERMWML L LRL---G CPNQ R G L Q R Q PSSSRRDSSW SETS EASYSG L CXCR4 LAFFH CCLNP ILYAFL G A K F K TSAQ H A L TSVSR-G -SSLK I L S K G K R G G H SSVS TESES SSFH SS CXCR5 L G LAH CCLNPML Y T F A G V K F R-SDLSLL T K L G CAG PASLCQ L FPNWR K SSLSES ENAT S LTTF

CCR1 I AYTH CCV NPVIYAFV G ERF R K Y LRQ L F H RRVAVH LVKWLPFL SVDRDLERVT S STSPS T G E H ELSAG F CCR2b L GMT H CC I NP IIYAFV G E K F RRY LSVFF R K H ITK RFCK - Q CPVFYRETVEG VT S SNTPS T G E Q EVSAG L CCR3 IAYSH CCMNPVIYAFV G ERF R K Y LRH F F H R H LLMH L G RY I PFLP SEK -LDR-TS SVSPS TAEPELS IVF CCR4 LAFVH CCL NP IIYFFL G E K F R K Y ILQ L F K TCRG LFVLCQ YCG LLQ IYSAETPSS VYTQ S TMD H DLH DLAL CCR5 L GMT H CC I NP IIYAFV G E K F RNY LLVFF Q K H IAK RFCK -CCSIFQ Q EAPARAS S -YTRS T G E Q EISVG L CCR6 LAFLH CCL NPVLYAF I G Q K F RNY ----F L K ILK DLWCVRRK Y K SSG FSCRG RYS VEN I S R Q T SETADNDNAS SFTM CCR7 LACVRCCV NPFLYAF I G V K F RNDLFK L F K DLG CL SQ E Q LRQW SSCRH IRESSM S CEAETTTTFSP CCR8 ISFTH CCV NPVLYAFV G E K F K K H LSEIF Q K SCSQ I FNYLG R QM P-RESCSK SSS L Q Q H S SRSSSVDYIL CCR9 IAFLH CCFSP ILYAF SS H R F R Q Y L-K A F LAAVLGWH LAPG TAQ ASLSSCE - - SS ITAQ EEMT GMNDLG ERQ SENY PNK EDVG N K SA

307 321 329 340 B CXCR4:

WT AKFKTSAQHALTSVSRGSSLKILSKGKRGGHSSVSTESESSSFHSS

31 AKFKTSAQHALTSVS

23 AKFKTSAQHALTSVSRGSSLKIL

12 AKFKTSAQHALTSVSRGSSLKILSKGKRGGHSSV

S324A;S325A AKFKTSAQHALTSVSRGAALKILSKGKRGGHSSVSTESESSSFHSS

I328A;L329A AKFKTSAQHALTSVSRGSSLKAASKGKRGGHSSVSTESESSSFHSS

S324A;S325A I328A;L329A AKFKTSAQHALTSVSRGAALKAASKGKRGGHSSVSTESESSSFHSS

Fig. 3. (A) Alignment of amino acids sequences for the COOH-terminal domains of chemokine receptors. The amino acids sequences of COOH-terminal portion of the seventh transmembrane domain and the COOH-terminal cytoplasmic domains of CXC and CC chemokine receptors were aligned using the Clustal Method algorithm in the DNAStar Megalign package. In addition to defined CXC and CCR chemokine receptors, we included the receptor for the CXC chemokine B-lymphocyte chemoattractant (BLC), Burkitt’s lymphoma receptor 1 (BLR-1) recently defined as CXCR5 (Gunn et al., 1998). The black boxes highlight identical amino acids and the grey boxes conservative changes of amino acid residues for one or both subclasses of chemokine receptors. The di-leucine based motif specific of CXCR4 is highlighted by a double line clear box. (B) Mutations in the cytoplasmic tail of CXCR4. Six mutations were introduced into the COOH-terminal domain of the human CXCR4 cDNA by site directed mutagenesis. Three mutations introduced premature stop codons that truncate the cytoplasmic domain. The other three replace the indicated amino acids with alanine residues. clathrin coated pits, most likely through interaction with the CXCR4 COOH-terminal domain mutants clathrin AP2 adaptor complex (C. Pitcher and M. Marsh, To examine the role of the SSLKIL sequence in both ligand unpublished). A very similar S(X)nLL type signal in the γ and phorbol ester-induced internalisation of CXCR4 we subunit of the CD3 complex is also involved in the endocytosis generated truncation and site specific mutations in human of the T cell receptor (TcR)/CD3 complex and mediates CXCR4 (Fig. 3B). Stop codons were introduced in place of the association with AP2 (Dietrich et al., 1994, 1997). Inspection codons for amino acids R322, S330 and S341 to generate of the cytoplasmic sequences (loops 1-3 and the COOH- molecules with COOH-terminal domains truncated by 31 terminal domain) of the five human CXC and nine distinct CC (∆31), 23 (∆23) and 12 (∆12) amino acids, respectively. The chemokine receptors characterised to date, indicates that ∆31 truncation removes the serine rich domain and the CXCR4 contains a S(X)nLL-like sequence SSLKIL in its SSLKIL sequence, ∆23 retains the SSLKIL sequence but COOH-terminal domain (Fig. 3A, NB, isoleucine can replace removes the serine rich domain, and ∆12 removes the bulk of leucine in di-leucine sorting signals; Sandoval and Bakke, the serine rich domain. In addition we changed the S324 and 1994). None of the other human chemokine receptors contains S325 in the full length CXCR4 sequence to alanines a similar sequence in this location, though CCR8 does have the (S324A;S325A), and I328 and L329 to alanines (I328A;L329). sequence SSSVDYIL at its extreme COOH terminus. A third mutant contained all four of these alanine substitutions Moreover, the SSLKIL sequence is completely conserved in (S324A;S325A/I328A;L329). The mutants and wt CXCR4 the eight species for which full length CXCR4 sequences have cDNAs were transfected into Mv-1-Lu cells and stable clones been determined to date. We therefore examined the role of this derived by selection with G418. Resistant colonies were tested motif in phorbol ester-induced internalisation of CXCR4. for CXCR4 surface expression by flow cytometry with the Differential regulation of CXCR4 and CCR5 endocytosis 2825

to 37¡C for periods of up to 60 minutes. At the indicated times 31 23 duplicate sets of cells were placed on ice and the intracellular radioactivity determined using an acid stripping assay that removes only the cell surface counts. As observed previously (Signoret et al., 1997), we measured only low levels of constitutive endocytosis for the wt and all six CXCR4 mutants, with maximally 12% of the initial cell surface pool of receptor internal after 60 minutes at 37¡C (Fig. 5A). When similar experiments were performed with 100 ng/ml PMA included in the medium during the 37¡C incubation we 12 S324A; S325A found that the rate and extent of endocytosis of ∆31 was similar to that of untreated cells (Fig. 5B). By comparison, the rate of uptake of the wt molecule was increased 7-8 fold (as measured over the first 5 minutes of uptake) and 70% of the initial cell surface pool was internal after 60 minutes of incubation at 37¡C. The ∆23 and ∆12 mutants showed intermediate effects. Both showed an increase in the rate and extent of endocytosis in the presence of PMA, though neither construct was I328A; L329A S324A; S325A internalised as rapidly nor to the same extent as wt CXCR4. Cell counts I328A; L329A Of these two mutants, ∆12 was less impaired than the ∆23 mutant. Thus CXCR4 molecules containing the COOH- terminal SSLKIL sequence exhibit phorbol ester-induced increases in endocytosis, although the magnitude of the response may be regulated by other elements in the COOH- terminal domain. Analysis of the alanine mutations indicated that all three constructs had lost the ability to undergo phorbol ester-induced WT endocytosis (Fig. 5C). Taken together, the data support the notion that the SSLKIL sequence is crucial for the phorbol ester-induced internalisation of CXCR4 and that the serine residues and the isoleucine/leucine pair are both essential components of the motif. The SSLKIL sequence is not required for SDF-1α- induced endocytosis of CXCR4 To determine whether the mutations described above influence Fluorescence the ability of CXCR4 to undergo chemokine-induced down modulation, cells were incubated in medium containing 125 Fig. 4. Cell surface expression of CXCR4 and CXCR4 mutants on α Mv-1-Lu transfectants. Mv-1-Lu cells transfected with wt or mutant nM SDF-1 at 37¡C for periods up to 60 minutes. At the end CXCR4 were labelled in suspension with the 12G5 mAb (5 µg/ml) of the incubation the cells were placed on ice, cell surface and then with a goat anti-mouse FITC-conjugated antibody. Cells bound SDF-1α was eluted by washing the cells in low pH labelled only with the FITC-conjugated antibody were used as medium and the level of cell surface CXCR4 measured using negative control. The histograms represent the fluorescence intensity 125I-12G5. In these experiments we found that truncation of the for 10,000 acquired events for each sample. cytoplasmic domain of CXCR4 progressively abrogated the ability of the molecule to undergo ligand-induced down modulation. Removal of the COOH-terminal 12 and 23 amino 12G5 mAb. All the mutants were expressed at levels acids (∆12 and ∆23, respectively) reduced the response by 75 comparable to that of the wt protein as indicated by FACS and 50%, respectively. As previously shown for a CXCR4∆Cyt analysis (Fig. 4), and 125I-SDF-1α binding experiments construct lacking the bulk of the cytoplasmic domain (Signoret indicated that as with other COOH-terminally truncated et al., 1997), the ∆31 construct showed no significant SDF-1α- chemokine receptors (Arai et al., 1997; Ben-Baruch et al., induced down modulation (Fig. 6A). By contrast, all three of 1995) the CXCR4 mutants bind SDF-1α with similar affinities the alanine mutants showed essentially wt responses to SDF- (see below). 1α (Fig. 6B). Together the data show that the SSLKIL sequence in The SSLKIL sequence is required for phorbol ester- CXCR4 is not required for SDF-1α-induced CXCR4 down induced endocytosis of CXCR4 modulation and is likely to explain the ability of this receptor We first verified that the mutations did not affect the to undergo phorbol ester-induced internalisation. constitutive endocytosis of the receptor using the anti-CXCR4 antibody 12G5, as previously described (Signoret et al., 1997). SDF-1α binding to wild-type and mutated CXCR4 After cell surface labelling with 125I-12G5 mAb (0.5 nM) for molecules 2 hours at 4¡C, Mv-1-Lu CXCR4 transfectants were warmed SDF-1α-induced down modulation of the CXCR4 COOH- 2826 N. Signoret and others

100 100 100 A B C 80 80 80

60 60 60

40 40 40 Internal (%) 20 20 20

0 0 0 0 10203040506070 0 10203040506070 0 10203040506070 Mins. Fig. 5. The SSLKIL sequence is required for PMA induced-CXCR4 endocytosis. Confluent cultures of Mv-1-Lu cells expressing CXCR4 or CXCR4 COOH-terminal mutations were labelled for 2 hours at 4¡C with 0.5 nM 125I-12G5 mAb in BM, washed and warmed to 37¡C in the absence (A) or presence (B and C) of 100 ng/ml PMA. The amount of internalised antibody was determined by acid washing as described in Materials and Methods. Each time point indicates the acid-resistant radioactivity (internal) as a proportion of the total cell-associated activity. WT (ᮀ), ∆31 (᭜), ∆23 (᭹) and ∆12 (᭡) CXCR4 deletion mutants, S324A;S325A (᭛), I328A;L329A (᭺) and S324A;S325A/I328A;L329A (᭝) CXCR4 mutants. All data points show means and standard deviations for triplicate samples from a representative experiment. terminal domain truncation mutants was reduced compared to shown). This may reflect the presence of mink CXCR4 wt CXCR4 (Fig. 6A). This most likely reflects the loss of molecules (or another chemokine receptor able to bind SDF- crucial elements in the COOH terminus required for 1α) on these cells that cannot be detected with our anti-human endocytosis. However, the mutations might also affect the CXCR4 antibodies. COS-7 cells have previously been used to ability of the molecule to bind SDF-1α. We therefore compared assay CXCR4/SDF-1α interaction (Kledal et al., 1997). We the binding properties of SDF-1α to the wt CXCR4 and all 6 therefore transiently expressed the wt and COOH-terminal COOH-terminal mutants. In initial experiments with Mv-1-Lu CXCR4 mutants in these cells for competition binding cells, we observed high affinity SDF-1α binding to these cells analysis. in the absence of transfected human CXCR4 molecules (not Fig. 7 shows the titration curves for 125I-SDF1α binding in the presence of increasing concentrations of unlabelled SDF- 1α. The wt and mutant CXCR4 molecules exhibited very similar SDF-1α binding activity. IC50, Hill coefficient and 100 A Bmax values were calculated from four independent experiments and did not show significant differences between the wild type and mutant CXCR4 molecules (Table 1). These α 75 data indicate that SDF-1 binding affinity was not affected by the COOH-terminal mutations. B/Bo% 50 DISCUSSION

The role of chemokine receptors as the principal or co- 25 receptors for HIV and SIV entry is now well established, with 0 10203040506070 CCR5 and CXCR4 implicated as major chemokine receptors involved in virus infection in vivo (Berger, 1997; Berger et al., 1998; Moore et al., 1997). Cellular entry of HIV and SIV is believed to occur by membrane fusion, induced through the 100 B Fig. 6. The SSLKIL sequence is not required for SDF-1α-induced down modulation of CXCR4. Mv-1-Lu cells expressing wt or 75 mutated CXCR4 molecules were incubated in BM containing 125 nM SDF-1α at 37¡C for up to 60 minutes. At the indicated time the cells were cooled on ice, washed with cold BM and the bound SDF- B/Bo% 1α was removed by incubation in acid medium. The cell surface 50 CXCR4 molecules were then detected by labelling with 125I-12G5 antibody for 2 hours at 4¡C. The graphs show the cell-associated 12G5 binding for SDF-1α treated cells as a proportion of 12G5 25 binding on untreated cells (B/Bo%) for the indicated time points. 0 10203040506070 (A) WT (ᮀ), ∆31 (᭜), ∆23 (᭹) and ∆12 (᭡) CXCR4 deletion mutants. (B) S324A;S325A (᭛), I328A;L329A (᭺) and ᭝ timetime (min.)(min) S324A;S325A/I328A;L329A ( ) CXCR4 mutants. Differential regulation of CXCR4 and CCR5 endocytosis 2827

Table 1. Binding constants for SDF-1α on wild-type and has been proposed to play a role in both desensitisation and mutant CXCR4 constructs resensitisation (Ferguson et al., 1996a; Krueger et al., 1997). The CC chemokines MIP-1α, MIP-1β, RANTES, eotaxin, the CXCR4 IC50 (nM) Hill Bmax (nM) CMV-encoded vMIP-II, and the CXC chemokine SDF-1, have WT 3.1±0.4 −1.0±0.2 53.8±7.7 ∆31 5.3±2.0 −1.3±0.1 78.8±22.1 been shown to inhibit CCR5, CCR3 or CXCR4 dependent ∆23 5.5±2.0 −1.2±0.1 98.3±22.4 HIV-1 infection (Bleul et al., 1996; Cocchi et al., 1995; Oberlin ∆12 4.5±1.5 −1.2±0.2 74.2±7.3 et al., 1996; Simmons et al., 1997). However, the mechanism(s) S324A;S325A 4.5±1.2 −1.2±0.1 63.2±12.2 through which this inhibition occurs has not been established. − I328A;L329A 5.2±2.0 1.2±0.1 78.8±23.2 Several recent studies have indicated that agonist-induced S324A;S325A/I328A;L329A 5.0±1.0 −1.2±0.2 71.4±12.3 chemokine receptor internalisation may be a component of the IC50, Bmax and Hill coefficient values were calculated from the mechanism of chemokine inhibition (Amara et al., 1997; Mack competition binding data illustrated in Fig. 7. The data represent the mean ± et al., 1998; Signoret et al., 1997). s.e. of four independent experiments. The mechanisms through which chemokines induce endocytosis of their receptors are not well understood. interaction of the viral envelope protein with CD4 and/or a Following on from studies of the β2-adrenergic receptor (β2- chemokine receptor, at the surface of target cells. Endocytosis AR) (Ferguson et al., 1996b; Godiska et al., 1997; Goodman of virus is not known to play a role in this process (Maddon et et al., 1996, 1997; Menard et al., 1997; Zhang et al., 1996), it al., 1988; Orloff et al., 1991; Pelchen-Matthews et al., 1995). has been shown that agonist-induced internalisation of CCR5 Indeed, cytoplasmic domain deleted forms of CD4, CCR5 and is enhanced by over-expression of G-protein receptor kinases CXCR4, which show significantly reduced endocytic (GRKs) and β-arrestins (Aramori et al., 1997), and that properties, will function efficiently as virus receptors (Amara internalisation is likely to proceed through clathrin coated et al., 1997; Edinger et al., 1997; Maddon et al., 1988; Orloff vesicles (Mack et al., 1998). Thus, ligand-binding is believed et al., 1991; Pelchen-Matthews et al., 1995; Signoret et al., to initiate GRK-induced phosphorylation of cytoplasmic 1997). Although receptor internalisation appears not to be sequences of the receptor and increase the association of the required for virus entry, receptor down modulation from the receptor with β-arrestin. β-arrestin couples the receptor to cell surface may still influence virus infection. Ligand-induced clathrin and facilitates endocytosis via clathrin coated vesicles. internalisation has been described for a number of 7TM For CXCR4, phorbol esters can induce receptor internalisation GPCRs including both CC and CXC chemokine receptors, the independently of ligand association (Amara et al., 1997; related C5a and fMLP receptors, the β2-adrenergic receptor, Forster et al., 1998; Haribabu et al., 1997; Lapham et al., 1996; thrombin receptor and many others (Amara et al., 1997; Signoret et al., 1997). In T cell lines that constitutively express Aramori et al., 1997; Ferguson et al., 1996a; Forster et al., CXCR4, and in transfected Mv-1-Lu cells, phorbol ester 1998; Haribabu et al., 1997; Hoxie et al., 1993; Jourdan et al., treatment increases the rate of CXCR4 endocytosis from a 1998; Mueller et al., 1997; Prado et al., 1996; Signoret et al., basal rate of about 1% of the cell surface pool per minute to 1997; Solari et al., 1997; von Zastrow and Kolbika, 1992) and about 6-8% per minute (Signoret et al., 1997). Phorbol ester- induced CXCR4 internalisation occurs through clathrin coated pits (Signoret et al., 1997) and, in contrast to ligand-induced uptake, involves activation of PKC (Signoret et al., 1997). Although the physiological relevance of the phorbol ester- induced internalisation of CXCR4 is still to be established it likely reflects signalling events that activate PKC (Acres et al., 1986). Indeed recent studies have shown that stimulation of T cells and T cell clones by cross-linking cell surface CD28, CD2 and CD3 can induce CXCR4 down modulation (Jourdan et al., 1998). Unlike ligand-induced down modulation, the rapid phorbol ester-induced increase in endocytosis seen with CXCR4 is not observed for all chemokine receptors. Here we show that the cell surface expression of CCR5 is not altered following exposure of cells to phorbol esters, indicating that there is no increase in endocytic activity and no consequent down modulation. The experiments were performed in transfected T cell lines expressing CCR5 together with CD4 and CXCR4, Fig. 7. Competition SDF-1α binding on CXCR4 wild-type and and in transfected Mv-1-Lu cells with and without CD4. In 125 COOH-terminal mutants. I-SDF-1α binding on COS-7 cells both cellular backgrounds CD4 and CXCR4 could be transiently expressing the CXCR4 constructs was competed with α ᭹ ∆ ᮀ demonstrated to undergo PMA-induced down modulation but increasing concentrations of unlabeled SDF-1 . WT ( ), 31 ( ), the CCR5 molecule was unaffected. We have previously ∆23 (᭺), ∆12 (᭝), S324A;S325A (᭡), I328A;L329A (᭿), S324A;S325A/I328A;L329A (᭜). The graph shows the mean of four studied the effects of phorbol esters on CD4 and found that for independent experiments. The results are expressed as the percentage this molecule, phosphorylation of a specific serine residue of maximum binding measured in absence of unlabelled ligand. Non- (S408 in human CD4) in the cytoplasmic domain of the specific binding, determined in presence of 1 µM unlabeled SDF-1α, molecule, activates an endocytosis signal that leads to CD4 was subtracted. recruitment into clathrin coated pits and endocytosis (Marsh 2828 N. Signoret and others and Pelchen-Matthews, 1996). The endocytosis signal involves Our results indicate that the cell surface expression and a pair of leucine residues four amino acids COOH-terminal to endocytosis of CXCR4 and CCR5 can be differentially S408. We noted the presence of a similar sequence (SSLKIL) regulated. Why CXCR4 alone of the chemokine receptors at position 324-329 in the COOH-terminal domain of human contains this S(X)nLL type signal is unclear. CCR5 is a non- CXCR4 but not the other well characterised chemokine essential protein that appears to function primarily in leukocyte receptors (Fig. 3a, with the possible exception of CCR8; see chemotaxis (Liu et al., 1996; Samson et al., 1996). By contrast, above). Here we have shown that this motif is required for in mice at least, CXCR4 is essential. It is widely expressed not phorbol ester-induced endocytosis of CXCR4, but not for only on leukocytes (Bleul et al., 1997; Forster et al., 1998) but ligand-induced internalisation. Mutation of either the also on neurones (Hesselgesser et al., 1997) and endothelial isoleucine/leucine pair and/or the serine pair to alanine cells (Gupta et al., 1998; Volin et al., 1998). Moreover, SDF completely abrogated the phorbol ester response without and CXCR4 have been implicated in vascular development, affecting SDF-1 induced uptake. In contrast, limited truncation neuronal patterning, haematopoiesis as well as T cell migration (∆12) of the cytoplasmic domain of CXCR4 inhibited SDF-1- and activation (Nagasawa et al., 1996; Tachibana et al., 1998; induced down modulation without significantly affecting the Zou et al., 1998). The additional endocytosis signals may be phorbol ester response. required for the wider functional activities of CXCR4. The endocytosis signal in the cytoplasmic domain of CD4 Alternatively, the presence of a S(X)nLL motif on CXCR4, as is dependent on phosphorylation of serine 408 for activity. In well as CD4 and the TcR/CD3 complex (Dietrich et al., 1994, the absence of phosphorylation CD4 has similar properties to 1997; Marsh and Pelchen-Matthews, 1996; Shin et al., 1991), a cytoplasmic domain-deleted form of the protein (Marsh and may allow the cell surface expression of these molecules to be Pelchen-Matthews, 1996). For CXCR4, both phorbol ester and co-ordinately regulated on T cells. Significantly, primary T ligand have been found to induce phosphorylation of the cells and dendritic cells appear to have intracellular stores of receptor cytoplasmic COOH-terminal domain (Haribabu et al., CXCR4 that can be relocated to the cell surface under certain 1997), though the precise residues have not been identified. conditions (Jourdan et al., 1998; Zaitseva et al., 1997). Thus Ligand-induced receptor phosphorylation has also been the trafficking of CXCR4 and perhaps other chemokine observed for CCR5 (Aramori et al., 1997). As with a number receptors is tightly controlled and is likely to depend on the of other 7TM GPCRs the extreme COOH-terminal domain of regulation of multiple trafficking or sorting signals. the protein is rich in serine residues, many of which are potential targets for phosphorylation. Our mutational analysis, We are grateful to colleagues who have contributed reagents, ideas together with the previous observation that inhibitors of PKC and discussions to this work, in particular to Dr Mike Luther and α activity block phorbol ester but not SDF-1α induced CXCR4 colleagues at GlaxoWellcome Inc. for providing SDF-1 , Marie-José internalisation (Signoret et al., 1997), suggests that ligand and Biljmakers and Annegret Pelchen-Matthews for critically reading the manuscript. M.M. is supported by grants from the UK Medical phorbol esters are likely to induce phosphorylation on distinct Research Council and GlaxoWellcome Inc. N.S. is supported by a sites in the COOH-terminal domain of CXCR4. Crucially, European Union TMR Marie Curie Research training grant phorbol esters are likely to induce phosphorylation of S324 ERBFMBICT961751. P.J.K. is supported by a grant from the UK and/or S325, with or without activity on other COOH-terminal Medical Research Council. J.A.H. is supported by grants from the residues. Phosphorylation of the COOH-terminus following National Institutes of Health AI40880 and AI38225. M.M. and J.A.H. SDF-1(α) treatment is likely to exclude S324 and S325. are supported by a NATO collaboration grant. The recruitment of β2-AR into coated pits has been shown to be enhanced through association with a β-arrestin, β-arrestin REFERENCES being able to couple the receptor to clathrin directly without Acres, B. R., Conlon, P. J., Mochizuki, D. Y. and Gallis, B. (1986). Rapid the apparent involvement of clathrin adaptors complexes phosphorylation and modulation of the T4 antigen on cloned helper T cells (Goodman et al., 1996, 1997). Data from studies with CCR5 induced by phorbol myristate acetate or antigen. J. Biol. Chem. 34, 16210- suggest that chemokine receptors may undergo similar 16214. recruitment into clathrin coated pits (Aramori et al., 1997). The Alkhatib, G., Combadiere, C., Broder, C. C., Feng, Y., Kennedy, P. E., Murphy, P. M. and Berger, E. A. (1996). CC-CKR5: A RANTES, MIP- phosphorylated S(X)nLL endocytosis signal in CD4 is believed 1, MIP-1 receptor as a fusion cofactor for macrophage-tropic HIV-1. Science to promote interaction with the AP2 clathrin adaptor complex. 272, 1955-1958. A related sequence in the γ subunit of the TcR/CD3 complex Alkhatib, G., Liao, F., Berger, E. A., Farber, J. M. and Peden, K. W. C. has been found to interact with AP2 complexes and we (1997). A new SIV co-receptor STRL33. Nature 388, 238. similarly found that peptides encoding the CD4 internalisation Amara, A., Gall, S. L., Schwartz, O., Salamero, J., Montes, M., Loetscher, P., Baggiolini, M., Virelizier, J. L. and Arenzana-Seisdedos, F. (1997). signal associate with AP2 complexes (C. Pitcher and M. HIV Coreceptor Downregulation as Antiviral Principle: SDF-1alpha- Marsh, unpublished). Thus it might be expected that the dependent Internalization of the Chemokine Receptor CXCR4 Contributes SSLKIL sequence in CXCR4 also provides a binding site for to Inhibition of HIV Replication. J. Exp. Med. 186, 139-146. AP2 and would explain the observation that minor sequence Arai, H., Monteclaro, F. S., Tsou, C.-L., Franci, C. and Charo, I. F. (1997). Dissociation of chemotaxis from agonist-induced receptor internalisation in changes that completely abolish the phorbol ester response (i.e. a lymphocyte cell line transformed with CCR2B. J. Biol. Chem. 272, 25037- I328A/L329A) have no effect on ligand-induced 25042. internalisation. Together the observations suggest that CXCR4 Aramori, I., Zhang, J., Ferguson, S. S., Bieniasz, P. D., Cullen, B. R. and has the capacity to interact with two distinct sets of endocytic Caron, M. G. (1997). Molecular mechanism of desensitization of the adaptors, the AP2 complex and a β-arrestin, respectively, and chemokine receptor CCR-5: receptor signaling and internalization are dissociable from its role as an HIV-1 co-receptor. EMBO J. 16, 4606-4616. that these interactions are likely to be regulated through Arenzana-Seisdedos, F., Virelizier, J.-L., Rousset, D., Clark-Lewis, I., phosphorylation of distinct sets of one or more serine residues Loetscher, P., Moser, B. and Baggiolini, M. (1996). HIV blocked by HIV in the COOH-terminal domain of the molecule. chemokine antagonists. Nature 383, 400. Differential regulation of CXCR4 and CCR5 endocytosis 2829

Ben-Baruch, A., Bengali, K. M., Biragyn, A., Johnston, J. J., Wang, J. M., Menard, L. and Caron, M. G. (1996b). Role of beta-arrestin in mediating Kim, J., Chuntharapai, A., Michiel, D. F., Oppenheim, J. J., et al. (1995). agonist-promoted G protein-coupled receptor internalization. Science 271, Interleukin-8 receptor beta. The role of the carboxyl terminus in signal 363-366. transduction. J. Biol. Chem. 270, 9121-9128. Forster, R., Kremmer, E., Schubel, A., Breitfeld, D., Kleinschmidt, A., Berger, E. A. (1997). HIV entry and tropism: the chemokine receptor Nerl, C., Bernhardt, G. and Lipp, M. (1998). Intracellular and surface connection. AIDS 11 (suppl. A), S3-S16. expression of the HIV-1 coreceptor CXCR4/fusin on various leukocyte Berger, E. A., Doms, R. W., Fenyö, E.-M., Korber, B. T. M., Littman, D. subsets: rapid internalization and recycling upon activation. J. Immunol. R., Moore, J. P., Sattentau, Q. J., Schuitemaker, H., Sodroski, J., et al. 160, 1522-1531. (1998). A new classification for HIV-1. Nature 391, 240. Fouchier, R. A., Meyer, B. E., Simon, J. H., Fischer, U. and Malim, M. H. Berson, J. F., Long, D., Doranz, B. J., Rucker, J., Jirik, F. R. and Doms, (1997). HIV-1 infection of non-dividing cells: evidence that the amino- R. W. (1996). A seven-transmembrane domain receptor involved in fusion terminal basic region of the viral matrix protein is important for Gag and entry of T-cell-tropic human immunodeficiency virus type 1 strains. J. processing but not for post-entry nuclear import. EMBO J. 16, 4531-4539. Virol. 70, 6288-6295. Godiska, R., Chantry, D., Raport, C. J., Schweickart, V. L., Trong, H. L. Bleul, C. C., Farzan, M., Choe, H., Parolin, C., Clark-Lewis, I., Sodroski, and Gray, P. W. (1997). Monocyte chemotactic protein-4: tissue-specific J. and Springer, T. A. (1996). The lymphocyte chemoattractant SDF-1 is expression and signaling through CC chemokine receptor-2. J. Leukoc. Biol. a ligand for Lestr/fusin and blocks HIV-1 entry. Nature 382, 829-832. 61, 353-60. Bleul, C. C., Wu, L. J., Hoxie, J. A., Springer, T. A. and Mackay, C. R. Goodman, O. B., Krupnick, J. G., Santinin, F., Gurevich, V. V., Penn, R. (1997). The HIV coreceptors CXCR4 and CCR5 are differentially expressed B., Gagnon, A. W., Keen, J. H. and Benovic, J. L. (1996). β-Arrestin acts and regulated on human T lymphocytes. Proc. Nat. Acad. Sci. USA 94, 1925- as a clathrin adaptor in endocytosis of the β2-adrenergic receptor. Nature 1930. 383, 447-450. Choe, H., Farzan, M., Sun, Y., Sullivan, N., Rollins, B., Ponath, P. D., Wu, Goodman, O. B., Krupnick, J. G., Gurevich, V. V., Benovic, J. L. and Keen, L., C.R., M., LaRosa, G., et al. (1996). The β-Chemokine receptors CCR3 J. H. (1997). Arrestin/clathrin interaction. Localization of the arrestin and CCR5 facilitate infection by primary HIV-1 isolates. Cell 85, 1135- binding locus to the clathrin terminal domain. J. Biol. Chem. 272, 15017- 1148. 15022. Cocchi, F., DeVico, A. L., Garzino-Demo, A., Arya, S. K., Gallo, R. C. and Gunn, M. D., Ngo, V. N., Ansel, K. M., Ekland, E. H., Cyster, J. G. and Lusso, P. (1995). Identification of RANTES, MIP-1 alpha, and MIP-1 beta Williams, L. T. (1998). A B-cell-homing chemokine made in lymphoid as the major HIV-suppressive factors produced by CD8+ T cells. Science follicles activates Burkitt’s lymphoma receptor-1. Nature 391, 799-803. 270, 1811-1815. Gupta, K., Lysko, P. G., Pillarisetti, K., Ohlstein, E. and Stadel, J. M. Crump, M. P., Gong, J. H., Loetscher, P., Rajarathnam, K., Amara, A., (1998). Chemokine receptors in human endothelial cells. J. Biol. Chem. 273, Arenzana-Seisdedos, F., Virelizier, J. L., Baggiolini, M., Sykes, B. D., et 4282-4287. al. (1997). Solution structure and basis for functional activity of stromal cell- Haribabu, B., Richardson, R. M., Fisher, I., Sozzani, S., Peiper, S. C., derived factor-1; dissociation of CXCR4 activation from binding and Horuk, R., Ali, H. and Snyderman, R. (1997). Regulation of human inhibition of HIV-1. EMBO J. 16, 6996-7007. chemokine receptors CXCR4: Role of phosphorylation in desensitization Deng, H., Liu, R., Ellmeier, W., Choe, S., Unutmaz, D., Burkhart, M., Di and internalization. J. Biol. Chem. 272, 28726-28731. Marzio, P., Marmon, S., Sutton, R. E., et al. (1996). Identification of a He, J., Chen, Y., Farzan, M., Choe, H., Ohagen, A., Gartner, S., Busciglio, major co-receptor for primary isolates of HIV-1. Nature 381, 661-666. J., Yang, X., Hofmann, W., et al. (1997). CCR3 and CCR5 are co-receptors Deng, H., Unutmaz, D., KewalRamani, V. N. and Littman, D. R. (1997). for HIV-1 infection of microglia. Nature 385, 645-649. Expression cloning of new receptors used by simian and human Healey, D., Dianda, L., Moore, J. P., McDougal, J. S., Moore, M. J., Estess, immunodeficiency viruses. Nature 388, 296-300. P., Buck, D., Kwong., P. D., Beverley, P. C. L., et al. (1990). Novel anti- Dietrich, J., Hou, X., Wegener, A.-M. and Geisler, C. (1994). CD3γ contains CD4 monoclonal antibodies separate human immunodeficiency virus a phosphoserine-dependent di-leucine motif involved in down-regulation of infection and fusion of CD4+ cells from virus binding. J. Exp. Med. 172, the T cell receptor. EMBO J. 13, 2156-2166. 1233-1242. Dietrich, J., Kastrup, J., Nielsen, B. L., Odum, N. and Geisler, C. (1997). Hesselgesser, J., HalksMiller, M., DelVecchio, V., Peiper, S. C., Hoxie, J., Regulation and function of the CD3gamma DxxxLL motif: a binding site Kolson, D. L., Taub, D. and Horuk, R. (1997). CD4-independent for adaptor protein-1 and adaptor protein-2 in vitro. J. Cell Biol. 138, 271- association between HIV-1 gp120 and CXCR4: Functional chemokine 281. receptors are expressed in human neurons. Curr. Biol. 7, 112-121. Doranz, B. J., Rucker, J., Yi, Y., Smyth, R. J., Samson, M., Peiper, S., Hoxie, J. A., Rackowski, J. L., Haggarty, B. S. and Gaulton, G. N. (1988). Parmentier, M., Collman, R. J. and Doms, R. W. (1996). A dual-tropic, T4 Endocytosis and phosphorylation induced by phorbol esters but not by primary HIV-1 isolate that uses Fusin and the §-chemokine receptors CKR- mitogen or HIV infection. J. Immunol. 140, 786-795. 5, CKR-3, and CKR-2b as fusion cofactors. Cell 85, 1149-1158. Hoxie, J. A., Ahuja, M., Belmonte, E., Pizarro, S., Parton, R. and Brass, Dragic, T., Litwin, V., Allaway, G. P., Martin, S. R., Huang, Y., Nagashima, L. F. (1993). Internalization and recycling of activated thrombin receptors. K. A., Cayanan, C., Maddon, P. J., Koup, R. A., et al. (1996). HIV-1 entry J. Biol. Chem. 268, 13756-13763. into CD4+ cells is mediated by the chemokine receptor CC-CKR-5. Nature Johansen, T. E., Scholler, M. S., Tolstoy, S. and Schwartz, T. W. (1990). 381, 667-673. Biosynthesis of peptide precursors and protease inhibitors using new Edinger, A. L., Amedee, A., Miller, K., Doranz, B. J., Endres, M., Sharron, constitutive and inducible eukaryotic expression vectors. FEBS Lett. 267, M., Samson, M., Lu, Z. H., Clements, J. E., et al. (1997). Differential 289-294. utilization of CCR5 by macrophage and T cell tropic simian Jourdan, P., Abbal, C., Nora, N., Hori, T., Uchiyama, T., Vendrell, J. P., immunodeficiency virus strains. Proc. Nat. Acad. Sci. USA 94, 4005-4010. Bousquet, J., Taylor, N., Pene, J., et al. (1998). IL-4 induces functional Endres, M. J., Clapham, P. R., Marsh, M., Ahuja, M., Davis-Turner, J., cell-surface expression of CXCR4 on human T cells. J. Immunol. 160, 4153- McKnight, A., Thomas, J., Stoebenau-Haggarty, B., Choe, S., et al. 4157. (1996). CD4-independent infection by HIV-2 is mediated by fusin. Cell 87, Kledal, T. N., Rosenkilde, M. M., Coulin, F., Simmons, G., Johnsen, A. H., 745-756. Alouani, S., Power, C. A., Luttichau, H. R., Gerstoft, J., et al. (1997). A Farzan, M., Choe, H., Martin, K., Marcon, L., Hofmann, W., Karlsson, broad-spectrum chemokine antagonist encoded by Kaposi’s sarcoma- G., Sun, Y., Barrett, P., Marchand, N., et al. (1997). Two orphan seven- associated herpesvirus. Science 277, 1656-1659. transmembrane segment receptors which are expressed in CD4-positive cells Krueger, K. M., Daaka, Y., Pitcher, J. A. and Lefkowitz, R. J. (1997). The support simian immunodeficiency virus infection. J. Exp. Med. 186, 405- role of sequestration in G protein-coupled receptor resensitization. 411. Regulation of beta2-adrenergic receptor dephosphorylation by vesicular Feng, Y., Broder, C. C., Kennedy, P. E. and Berger, E. A. (1996). HIV-1 acidification. J. Biol. Chem. 272, 5-8. entry cofactor: Functional cDNA cloning of a seven-transmembrane, G Lapham, C. K., Quyang, J., Chandrasekhar, B., Nguyen, N. Y., Dimitrov, protein-coupled receptor. Science 272, 872-877. D. S. and Golding, H. (1996). Evidence for cell-surface association Ferguson, S. S., Barak, L. S., Zhang, J. and Caron, M. G. (1996a). G- between fusin and the CD4-gp120 complex in human cell lines. Science 274, protein-coupled receptor regulation: role of G-protein-coupled receptor 602-605. kinases and arrestins. Can. J. Physiol. Pharmacol. 74, 1095-1110. Liu, R., Paxton, W. A., Choe, S., Ceradini, D., Martin, S. R., Horuk, R., Ferguson, S. S., Downey, W. E. r., Colapietro, A. M., Barak, L. S., MacDonald, S. R., Stuhlmann, H., Koup, R. A., et al. (1996). 2830 N. Signoret and others

Homozygous defect in HIV-1 co-receptor accounts for resistance of some Rosenkilde, M. M., Cahir, M., Gether, U., Hjorth, S. A. and Schwartz, T. multiply-exposed individuals to HIV-1 infection. Cell 86, 367-377. W. (1994). Mutations along transmembrane segment II of the NK-1 receptor Mack, M., Luckow, B., Nelson, P. J., Cihak, J., Simmons, G., Clapham, P. affect substance P competition with non peptide antagonists but not R., Signoret, N., Marsh, M., Stangassinger, F., et al. (1998). substance P binding. J. Biol. Chem. 269, 28160-28164. Aminooxypentane-RANTES induces CCR5 internalization but inhibits Samson, M., Libert, F., Doranz, B. J., Rucker, J., Liesnard, C., Farber, C.- recycling: a novel inhibitory mechanism of HIV infectivity. J. Exp. Med. M., Saragosti, S., Lapouméroulie, C., Cognaux, J., et al. (1996). 187, 1215-1224. Resistance to HIV-1 infection in caucasian individuals bearing mutant Maddon, P. J., McDougal, J. S., Clapham, P. R., Dalgleish, A. G., Jamal, alleles of the CCR-5 chemokine receptor gene. Nature 382, 722-725. S., Weiss, R. A. and Axel, R. (1988). HIV infection does not require Sandoval, I. and Bakke, O. (1994). Targetting of membrane proteins to endocytosis of its receptor, CD4. Cell 54, 865-874. endosomes and lysosomes. Trends Cell Biol. 4, 292-297. Marsh, M. and Pelchen-Matthews, A. (1996). Endocytic and exocytic Shin, J., Dunbrack, R. L., Lee, S. and Strominger, J. L. (1991). regulation of CD4 expression and function. Curr. Top. Microbiol. Immunol. Phosphorylation-dependent down-modulation of CD4 requires a specific 205, 107-135. structure within the cytoplasmic domain of CD4. J. Biol. Chem. 266, 10658- Menard, L., Ferguson, S. S., Zhang, J., Lin, F. T., Lefkowitz, R. J., Caron, 10665. M. G. and Barak, L. S. (1997). Synergistic regulation of beta2-adrenergic Signoret, N., Oldridge, J., Pelchen-Matthews, A., Klasse, P. J., Tran, T., receptor sequestration: intracellular complement of beta-adrenergic receptor Brass, L. F., Rosenkilde, M. M., Schwartz, T. W., Holmes, W., et al. kinase and beta-arrestin determine kinetics of internalization. Mol. (1997). Phorbol esters and SDF-1 induce rapid endocytosis and down Pharmacol. 51, 800-808. modulation of the chemokine receptor CXCR4. J. Cell Biol. 139, 651- Moore, J. P., Trkola, A. and Dragic, T. (1997). Co-receptors for HIV-1 entry. 664. Curr. Opin. Immunol. 9, 551-562. Simmons, G., Clapham, P. R., Picard, L., Offord, R. E., Rosenkilde, M. Morgenstern, J. P. and Land, H. (1990). Advanced mammalian gene M., Schwartz, T. W., Buser, R., Wells, T. N. C. and Proudfoot, A. E. transfer: high titre retroviral vectors with multiple drug selection markers (1997). Potent inhibition of HIV-1 infectivity in macrophages and and complementary helper-free packaging cell line. Nucl. Acids Res. 18, lymphocytes by a novel CCR5 antagonist. Science 276, 276-279. 3587-3596. Solari, R., Offord, R. E., Remy, S., Aubry, J. P., Wells, T. N. C., Whitehorn, Mueller, S. G., White, J. R., Schraw, W. P., Lam, V. and Richmond, A. E., Oung, T. and Proudfoot, A. E. I. (1997). Receptor-mediated (1997). Ligand-induced desensitization of the human CXC chemokine Endocytosis of CC-chemokines. J. Biol. Chem. 272, 9617-9620. receptor-2 is modulated by multiple serine residues in the carboxyl-terminal Tachibana, K., Hirota, S., Iizasa, H., Yoshida, H., Kawabata, K., Kataoka, domain of the receptor. J. Biol. Chem. 272, 8207-8214. Y., Kitamura, Y., Matsushima, K., Yoshida, N., et al. (1998). The Murphy, P. M. (1994). The molecular biology of leukocyte chemoattractant chemokine receptor CXCR4 is essential for vascularization of the factors. Annu. Rev. Immunol. 12, 593-633. gastrointestinal tract. Nature 393, 591-594. Nagasawa, T., Hirota, S., Tachibana, K., Takakura, N., Nishikawa, S.-I., Trkola, A., Dragic, T., Arthos, J., Binley, J. M., Olson, W. C., Allaway, G. Kitamura, Y., Yoshida, N., Kikutani, H. and Kishimoto, T. (1996). P., Chenh-Mayer, C., Robinson, J., Maddon, P. J., et al. (1996). CD4- Defects of B-cell lymphopoiesis and bone-marrow myelopoiesis in mice dependent, antibody sensitive interactions between HIV-1 and its co- lacking the CXC chemokine PBSF/SDF-1. Nature 382, 635-638. receptor CCR-5. Nature 384, 184-187. Oberlin, E., Amara, A., Bacherlerie, F., Bessia, C., Virelizier, J.-L., Ugolini, S., Moulard, M., Mondor, I., Barois, N., Demandolx, D., Hoxie, Arenzana-Seisdedos, F., Schwartz, O., Heard, J.-M., Clark-Lewis, I., et J., Brelot, A., Alizon, M., Davoust, J., et al. (1997). HIV-1 gp120 induces al. (1996). The CXC chemokine SDF-1 is the ligand for Lestr/Fusin and an association between CD4 and the chemokine receptor CXCR4. J. prevents infection by T-cell-line-adapted HIV-1. Nature 382, 833-835. Immunol. 159, 3000-3008. Orloff, G. M., Orloff, S. L., Kennedy, M. S., Maddon, P. J. and McDougal, Volin, M. V., Joseph, L., Shockley, M. S. and Davies, P. F. (1998). J. S. (1991). Penetration of CD4 T cells by HIV-1. The CD4 receptor does Chemokine receptor CXCR4 expression in endothelium. Biochem. Biophys. not internalise with HIV and CD4-related signal transduction events are not Res. Commun. 242, 46-53. required for entry. J. Immunol. 146, 2578-2587. von Zastrow, M. and Kolbika, B. K. (1992). Ligand-regulated internalization Pelchen-Matthews, A., Armes, J. E. and Marsh, M. (1989). Internalization and recycling of human β2-adrenergic receptors between the plasma and recycling of CD4 transfected into HeLa and NIH-3T3 cells. EMBO J. membrane and endosomes containing the transferrin receptor. J. Biol. Chem. 8, 3641-3649. 267, 3530-3538. Pelchen-Matthews, A., Armes, J. E., Griffiths, G. and Marsh, M. (1991). Wells, T. N. C., Proudfoot, A. E. I., Power, C. A. and Marsh, M. (1996). Differencial endocytosis of CD4 in lymphocytic and non-lymphocytic cells. Chemokine receptors Ð the new frontier for AIDS research. Chem. Biol. 3, J. Exp. Med. 173, 575-587. 603-609. Pelchen-Matthews, A., Parsons, I. J. and Marsh, M. (1993). Phorbol ester- Zaitseva, M., Blauvelt, A., Lee, S., Lapham, C. K., Klaus-Kovtun, V., induced down-regulation of CD4 is a multi-step process involving Mostowski, H., Manischewitz, J. and Golding, H. (1997). Expression and dissociation from p56lck, increased association with clathrin-coated pits and function of CCR5 and CXCR4 on human Langerhans cells and altered endosomal sorting. J. Exp. Med. 178 , 1209-1222. macrophages: implications for HIV primary infection. Nature Med. 3, 1369- Pelchen-Matthews, A., Clapham, P. and Marsh, M. (1995). The role of CD4 1375. endocytosis in HIV infection. J. Virol. 69, 8164-8168. Zhang, J., Ferguson, S. S. G., Barak, L. S., Menard, L. and Caron, M. G. Prado, G. N., Suzuki, H., Wilkinson, N., Cousins, B. and Navarro, J. (1996). Dynamin and beta-arrestin reveal distinct mechanisms for G protein- (1996). Role of the C terminus of the interleukin 8 receptor in signal coupled receptor internalization. J. Biol. Chem. 271, 18302-18305. transduction and internalization. J. Biol. Chem. 271, 19186-19190. Zou, Y.-R., Kottmann, A. H., Kuroda, M., Taniuchi, I. and Littman, D. R. Premack, B. A. and Schall, T. J. (1996). Chemokine receptors: gateways to (1998). Function of the chemokine receptor CXCR4 in haematopoiesis and inflammation and infection. Nature Med. 2, 1174-1178. in cerebellar development. Nature 393, 595-599.