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

J. Biol. Chem., Vol. 277, Issue 19, 17291-17299, May 10, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/19/17291    most recent M108232200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Brandt, S. M. Articles by Landau, N. R. Search for Related Content PubMed PubMed Citation Articles by Brandt, S. M. Articles by Landau, N. R. Social Bookmarking                      What's this?

Association of Chemokine-mediated Block to HIV Entry with Coreceptor Internalization^*

Stephanie M. Brandt, Roberto Mariani, Anne U. Holland^§, Thomas J. Hope^¶, and Nathaniel R. Landau^**

From the  Salk Institute for Biological Studies, Infectious Disease Laboratory, La Jolla, California 92037, the ^§ University of California San Diego, La Jolla, California 92037 and the ^¶ Department of Microbiology and Immunology, University of Illinois at Chicago College of Medicine, Chicago, Illinois 60612

Received for publication, August 27, 2001, and in revised form, December 9, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Chemokines inhibit entry of HIV into CD4⁺ T cells more effectively than into macrophages or transfected adherent cells. Here, we tested whether chemokine receptor internalization could account for cell type differences in the effectiveness of chemokines. Infection of CEM T cells expressing stably transduced wild-type CCR5 was much more readily inhibited by chemokine than were transduced HOS cells. This response correlated with the efficiency of CCR5 internalization. A mutated CCR5, termed M7-CCR5, in which the Ser/Thr phosphorylation sites in the cytoplasmic tail were changed to Ala, did not internalize in response to MIP-1. M7-CCR5 was expressed at slightly higher levels than wild-type on stably transduced cell lines and was somewhat more potent as an HIV-1 coreceptor. The mutated receptor mobilized intracellular Ca²⁺ in response to chemokine to a level 4-fold higher than did the wild type CCR5. Unexpectedly, the receptor was desensitized as efficiently as wild type, suggesting that desensitization does not require cytoplasmic tail phosphorylation. Entry of R5 HIV-1 reporter virus into cells stably expressing M7-CCR5 was largely resistant to blocking by MIP-1. As much as 80% of entry inhibition was attributed to receptor internalization. Aminooxypentane (AOP)-MIP-1 was able to induce a low level of M7-CCR5 internalization in HOS and to weakly inhibit HIV-1 entry. Introduction of dominant negative dynamin into HOS cells reduced the ability of chemokine to inhibit infection. The inefficiency of internalization of chemokine receptors in some cell types could allow virus to replicate in vivo in the presence of endogenous chemokine. Last, M7-CCR5 is a useful tool for discriminating coreceptor internalization from binding site masking in the evaluation of small molecule inhibitors of HIV-1 entry.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Chemokines are a family of chemotactic cytokines consisting of at least 40 members that bind seven transmembrane G-protein-coupled receptors (GPCRs)¹ on the surface of target cells. Chemokines are involved in the recruitment and activation of subsets of leukocytes in inflammatory responses (1, 2) and play a role in lymphocyte maturation (3). In addition to their role in the immune system, specific chemokine receptors are used by HIV as coreceptors for virus entry, and their chemokine ligands are potent inhibitors of HIV-1 replication (4, 5).

HIV-1 infection is initiated by the attachment of the virus envelope glycoprotein, gp120, to CD4 on the target cell. Binding to CD4 triggers a conformational change in gp120 that exposes a binding site for a chemokine receptor that acts as a coreceptor (6-9). Interaction with the coreceptor triggers a rearrangement of the transmembrane subunit of the envelope glycoprotein, gp41, that leads to fusion of the virus and cell membranes (10). The predominant chemokine receptors used as coreceptors for entry by primary isolates of HIV-1 are CCR5 and CXCR4 (11-13), although other chemokine receptors including CCR2 and CCR3 can be used by some virus isolates with much lower efficiency (14). The CCR5 ligands RANTES, MIP-1, and MIP-1 are potent inhibitors of HIV-1 isolates that use CCR5 for entry (15, 16). Stromal derived factor-1, the ligand for CXCR4, inhibits entry of isolates that use CXCR4 (17, 18).

Chemokine-mediated inhibition of HIV-1 entry appears to result from the combination of three mechanisms: (i) steric blocking of the interaction between gp120 and the coreceptor (19, 20); (ii) ligand-mediated internalization of the receptor, which reduces its availability for use by gp120 (21-23); and (iii) interference with receptor recycling (24). AOP derivatives of chemokines are particularly active at blocking virus entry (24). The potency of these derivatives results from their ability to induce CCR5 internalization and to prevent the recycling of the endocytosed chemokine receptor to the cell surface (24, 25). In addition to chemokines, small molecules such as TAK-779, which binds CCR5 (26), or AMD3100, which binds CXCR4 (27, 28), are potent inhibitors of CCR5 and CXCR4 HIV-1 entry, respectively.

For GPCRs in general, binding of ligand activates the coupled heterotrimeric G proteins, which mobilize intracellular second messengers such as inositol trisphosphates, cAMP, and intracellular [Ca²⁺] (29). Signaling is rapidly terminated as a result of receptor desensitization and internalization. These processes are mediated by cytoplasmic tail phosphorylation by a family of G protein-coupled receptor kinases (GRKs) and/or protein kinase C. The phosphorylated receptor then becomes a target for arrestin binding, which promotes dissociation of the G proteins and links the receptor to the endocytic machinery via adaptor proteins such as adaptor protein 2, leading to its association with clathrin-coated pits (29, 30). For CCR5 and CXCR4, chemokine binding activates coupled G_i or G_q trimeric G proteins resulting in a rapid, but transient increase in 1,4,5-trisphosphate and cytosolic [Ca²⁺] (31, 32) as well as phosphorylation of specific cytoplasmic tail Ser residues (33-35). The phosphorylated receptor then becomes associated with -arrestin (29, 36) and is internalized via clathrin-coated pits (37). The receptor may then traffic through low pH endosomes, resulting in ligand dissociation and recycling of the receptor back to the cell surface (25). HIV-1 entry does not require signaling by the coreceptor or its internalization (12, 33, 38-40); however, these events could be important for the ability of chemokines to inhibit infection.

Some cell types appear to be more susceptible to the blocking effects of chemokines than others. Infection of primary CD4⁺ T-cells is generally inhibited by low concentrations of chemokine (<1 ng/ml), whereas macrophages are much less susceptible and require as much as 1 µg/ml of chemokine to prevent virus entry (20, 41, 42). At 100-500 ng/ml, RANTES completely protected peripheral blood leukocytes from infection with non-syncytium-inducing virus, yet macrophage infection was not blocked and in some cases was slightly enhanced (20, 41, 42). The effectiveness of chemokines on macrophages is also influenced by their differentiation state (43). Monocyte-derived macrophages cultured without cytokines are 5 times less susceptible to HIV infection in the presence of RANTES than monocytes stimulated with macrophage colony-stimulating factor. This might be explained by the fact that monocytes are susceptible to modulation of CCR5 by chemokines (24), whereas differentiated macrophages are not (44).

In this study, we tested whether cell type differences in chemokine receptor internalization efficiency might account for the differential ability of chemokines to inhibit HIV infection. To investigate the mechanisms by which chemokines inhibit HIV-1 entry, we used an internalization-deficient mutant CCR5 termed M7-CCR5, in which seven cytoplasmic tail domain phosphorylation sites were changed to Ala. M7-CCR5 was deficient for chemokine-induced internalization, but its coreceptor activity was comparable with or slightly better than wild type. Importantly, entry was only slightly inhibited by chemokine. The ability of chemokines to block HIV entry could also be reduced by introducing a dominant negative dynamin into target cells, which interferes with receptor endocytosis. In addition, the susceptibility of two model cell lines to the inhibitory effects of chemokines correlated with the efficiency with which the receptor internalized in response to chemokine. Taken together, these findings suggest that internalization is a major mechanism by which chemokines prevent HIV infection and that the differential sensitivity of various cell types to the blocking effects of chemokines may be due to the efficiency with which they internalize CCR5 in response to ligand.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Plasmid Constructs-- M7-CCR5 (Ser³³⁶/Ser³³⁷/Tyr³³⁹/Thr³⁴⁰/Ser³⁴²/Thr³⁴³/Ser³⁴⁹)  Ala was generated by two rounds of PCR mutagenesis of a CCR5 cDNA (45). The first round used the oligonucleotide primers 5'-CTC GGA TCC GGT GGA ACA AGA TGG ATT AT (5' CCR5-BamHI) and 5'-TCA CAA GCC CAC AGC TAT TTC CTG CTC CCC AGC GGC TCG GGC GGC AAC TGC ACT GCT CG (C5M60). The second round used the oligonucleotides 5'-CCR5-BamHI and 5'-CCG TCG ACG AGT CCG TGT CAC AAG CCC ACA GCT AT (C5M34-SalI). Wild-type and M7-CCR5 cDNAs were fused in frame to EGFP in pEGFP-N1 vector (CLONTECH, Palo Alto, CA) by SalI and BamHI digestion after amplification with the oligonucleotides 5'-CTC GTC GAC GGT GGA ACA AGA TGG ATT AT (CCR5-1) and 5'-CTC GGA TCC GCC AAG CCC ACA GAT ATT TCC TG (CCR5-2). The cDNAs were excised by SalI and BamHI digestion, filled in with Klenow polymerase, and cloned into the SnaBI site in the retroviral vector pBABE-puro (46). Plasmid expression vectors for arrestin-2, arrestin-3, GRK2, GRK2-K220R, dynamin-K44A, and arrestin-3-(284-409) were provided by J. Benovic (Thomas Jefferson University) (37, 47).

Cell Lines-- The adherent human osteosarcoma cell line, HOS, and human embryonic kidney cell line, HEK-293T, were cultured in Dulbecco's modified Eagle's medium, 10% fetal bovine serum. Nonadherent cell lines derived from CEM.SS (CEM) were cultured in RPMI 1640, 10% fetal bovine serum. Cultures were maintained in 5% CO₂ at 37 °C. CCR5.EGFP and M7-CCR5.EGFP were stably transduced into HOS.CD4 (45) and CEM by retroviral vector infection using retroviral vector stocks produced by transfecting 293T cells (4.0 × 10⁶) with 10 µg of retroviral vector, 7 µg of pHT60 (48), and 3 µg of pCMV-VSV-G (49). Retrovirus-containing supernatants were harvested 48 h post-transfection, filtered, and stored at 80 °C. Cells were infected with 1.0 ml of virus and selected 2 days later in medium containing 1 µg/ml or 0.5 µg/ml puromycin for HOS.CD4 or CEM, respectively. HOS.CD4 cell clones were isolated with glass cloning cylinders and expanded. For CEM, single cells were deposited using a FACStar® (Becton Dickinson; Franklin Lakes, NJ) into 96-well plates and expanded. Individual cell clones were evaluated for CCR5 cell surface expression by staining with phycoerythrin-conjugated anti-CCR5 monoclonal antibody 2D7 (BD Pharmingen, San Diego, CA) and analyzed by flow cytometry. Single clones were chosen based upon CCR5 staining.

Chemokines and Inhibitors-- MIP-1 (Chemicon International Inc., Temecula, CA) and AOP-MIP-1 (a gift from G. Graham (Beatson Institute for Cancer Research, Glasgow, United Kingdom)) were diluted in Dulbecco's modified Eagle's medium to 100 ng/µl stocks. TAK-779 was dissolved at 20 mM in dimethyl sulfoxide.

CCR5 Internalization Assay-- Cells (5 × 10⁵) in six-well plates were incubated for 4 h at 37 °C with various concentrations of chemokine. Cells were harvested and stained in PBS, 1% fetal calf serum, 0.1% sodium azide for 30 min at 4 °C with 2D7-phycoerythrin. Cells were washed twice with PBS, 1% fetal calf serum, 0.1% sodium azide; fixed in 1% paraformaldehyde; and analyzed on a FACScan (Becton-Dickinson, Franklin Falls, NJ).

Fluorescence Microscopy-- Cells were cultured overnight on 0.1% gelatin-treated glass coverslips in 24-well tissue culture dishes. Internalized chemokine receptor in cells was visualized by incubating them with chemokine for 4 h at 37 °C and then fixing with 4% paraformaldehyde, PBS. Cells were permeabilized with PBS, 0.1% Triton X-100, and nuclei were stained with 1 µg/ml Hoechst 33258 (Sigma) for 15 min at room temperature. Coverslips were washed twice with PBS and mounted with Gel/Mount (Biomeda Corp., Foster City, CA). Slides were photographed using SlideBook 2.1 software (Intelligent Imaging Innovations, Inc., Denver, CO). To visualize the process of internalization in real time, cells were plated on fibronectin-treated 40-mm number 1.5 glass coverslips (Bioptechs, Butler, PA) in CO₂-independent medium, 10% fetal bovine serum. The coverslips were mounted in a closed live cell micro-observation system (Bioptechs) kept at 37 °C connected to a microperfusion pump (Bioptechs) via Tygon high purity tubing (Bioptechs) used to administer agonist. The chamber was placed on an Olympus IX70 microscope and observed with a ×100 oil immersion objective equipped with an objective warmer. Images were deconvoluted using a Silicon Graphics O₂ computer and Deltavision software. Cells were monitored for 30 min under initial chamber conditions to establish base-line conditions. The chamber was then infused with medium containing 200 ng/ml AOP-MIP-1 for 1 h. Images were recorded at a rate of one/min. The coverslip was then scanned to ensure that the final images were representative of the cells on the entire coverslip.

[Ca²⁺] Flux Measurement-- Cells were incubated in Dulbecco's modified Eagle's medium, 10% fetal calf serum for 30 min at 37 °C with 2 µM Indo-1 AM, 0.01% pluronic F-127 and were then exposed to 100 ng/ml MIP-1. Fluorescence was monitored at 402 and 468 nm using a PTI bench top dual emission continuous spectrofluorimeter (Photon Technologies International, Inc., Lawrenceville, NJ). Intracellular [Ca²⁺] was calculated by the formula [Ca²⁺] = K_d (F  F_min)/(F_max  F), where K_d is the dissociation constant (250 nM) and F represents arbitrary fluorescent units. F_min was considered as base-line fluorescence; F_max was calculated as the fluorescence after stimulation with 1 µM ionomycin.

Transient Transfection and Luciferase Reporter Virus Assays-- HOS cells stably expressing CD4/CCR5.EGFP (2.0 × 10⁶) were transfected with 20 µg of plasmid DNA by calcium phosphate transfection (50). The next morning, the cells were washed with medium and then harvested by trypsinization and plated in the afternoon for infection on the following day. Transfection efficiency was 60-75% in each experiment as measured by fluorescence-activated cell sorting analysis of parallel transfections of HOS.CD4 with EGFP expression vector. Luciferase reporter viruses were produced by cotransfecting 293T cells with NL-Luc-ER (51, 52) and JR.FL Env expression vector (45). Culture supernatants were harvested 48 h postinfection, filtered, and frozen at 80 °C in aliquots. Viruses were quantitated by p24^gag enzyme-linked immunosorbent assay. Reporter assays were performed in triplicate with 1.0 × 10⁴ cells/well in 96-well plates. Cells were incubated with chemokine or TAK-779 for 30 min at 37 °C. JR.FL-pseudotyped NL-Luc-ER luciferase reporter virus (1 ng of p24) was added to a final volume of 100 µl and removed after a 4-h incubation at 37 °C. Luciferase activity was measured 3 days postinfection using the Luc-lite Plus reagent (Packard Bioscience Company). Light emission was quantitated using a Packard TopCount and expressed as counts per second (cps).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Virus entry into primary T cells and T cell lines is effectively blocked by 1 ng/ml concentrations of inhibitory chemokines, whereas entry into 293 or HeLa cells transiently transfected with expression vectors for CD4/CCR5 was not detectably diminished by concentrations of MIP-1, MIP-1, or RANTES as high as 1 µg/ml (20, 41, 42).² Because CCR5 conformation is not known to be affected by cell type and because chemokine receptor internalization has been shown to play a role in mediating the effects of chemokines on HIV entry (23), we speculated that the efficiency of CCR5 internalization in response to ligand might be the cause of the observed cell type differences in susceptibility to inhibition of virus entry. To test this hypothesis, we used two model cell lines, an adherent cell line derived from a human osteosarcoma, HOS, and the transformed CD4⁺ T cell line, CEM. The HOS cells were stably transfected with a CD4 expression vector to generate HOS.CD4, and both cell lines were then stably transduced with a retroviral vector expressing CCR5 fused at its carboxyl terminus to EGFP. The EGFP did not interfere with coreceptor or chemokine receptor function, as shown below.

To compare the efficiency of ligand-induced internalization by the HOS and CEM cell lines, the cells were incubated for 30 min at 37 °C with MIP-1 or the more potent AOP-MIP-1 at two different concentrations. The cells were analyzed for cell surface CCR5 by flow cytometry and for total cellular CCR5 by EGFP fluorescence. MIP-1 induced efficient internalization of CCR5 on CEM at 10 and 100 ng/ml. On the HOS cells, CCR5 was not affected by the lower concentration of chemokine, and at the higher concentration the receptor was reduced by only 25% (Fig. 1A). AOP-MIP-1, which is a potent inducer of CCR5 internalization (24), was active on the CEM cells and showed some activity on HOS. However, even at 100 ng/ml, AOP-MIP-1 removed only about 50% of the cell surface CCR5 on the HOS cells, while on CEM it was reduced to background. EGFP fluorescence, an indicator of total CCR5 abundance, was not detectably affected by any of the treatments (Fig. 1B), demonstrating that internalization does not target the receptor for degradation. Thus, internalization of CCR5 in CEM is triggered by lower concentrations of chemokine than HOS and is more efficient.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Internalization of wild-type and M7-CCR5 in response to MIP-1 and AOP MIP-1. Flow cytometry analysis of HOS and CEM stably expressing wild type (A) or M7-CCR5.EGFP (C). Cells were incubated for 4 h with 10 ng/ml (light solid line) or 100 ng/ml (dark solid line) MIP-1 or AOP-MIP-1 at 37 °C and stained with 2D7-PE. Shaded areas indicate receptor expression in untreated cells; dotted lines indicate the unstained control cells. Total CCR5.EGFP expression (B) as detected by EGFP fluorescence in the chemokine-treated and -untreated control cells. HOS and CEM cells stably expressing wild-type CCR5.EGFP are shown in the absence of chemokine (dark shaded area) overlaid by cells incubated with 100 ng/ml AOP-MIP-1 for 4 h (light solid line); the dotted line indicates the unstained control.

Cytoplasmic Tail Domain Phosphorylation Sites Are Required for Ligand-induced CCR5 Internalization-- Previous studies using transiently transfected cytoplasmic tail truncations of CCR5 (23) or CXCR4 (22) suggested that coreceptor internalization contributes to the blocking activity of chemokines. To evaluate the role of coreceptor internalization in chemokine blocking of HIV entry, we sought to generate an internalization-deficient CCR5 that could be stably expressed on the cell surface. GPCR internalization in response to ligand binding is initiated by phosphorylation of cytoplasmic tail Ser/Thr residues. The cytoplasmic tail of CCR5 contains nine Ser/Thr/Tyr residues (Tyr³⁰⁷/Ser³²⁵/Ser³³⁶/Ser³³⁷/Tyr³³⁹/Thr³⁴⁰/Ser³⁴²/Thr³⁴³/Ser³⁴⁹), four of which (Ser³³⁶/Ser³³⁷/Ser³⁴²/Ser³⁴⁹), are sites of GRK-mediated phosphorylation (34). We initially generated CCR5 tail truncations in which stop codons were introduced at positions 307, 313, 328, or 335. These proteins were not expressed (positions 307 and 313) or were poorly expressed (positions 328 and 335) on the cell surface in stably transduced cells (data not shown). We next tested a mutated CCR5 in which eight cytoplasmic tail Ser/Thr residues were changed to Ala; however, this was also poorly expressed on the cell surface (not shown). In contrast, a variant in which Tyr³⁰⁷ and Ser³²⁵ were left unchanged and the remaining seven residues were changed to Ala (termed M7-CCR5) was efficiently expressed on the cell surface (Fig. 2). Ser³²⁵ is not thought to be a phosphorylation site (34). This residue may be important for CCR5 cell surface stability, at least in the context of the other cytoplasmic tail substitutions.

View larger version (32K): [in this window] [in a new window]   Fig. 2.   Diagram of M7-CCR5. Ala substitutions are indicated above the sequence. The arrows indicate locations of CCR5 truncations that were tested and found not to be expressed on the cell surface by stable transduction (data not shown). The asterisk indicates the Ser left intact.

To test whether M7-CCR5 was impaired for ligand-induced internalization, the transduced HOS and CEM cells were incubated with chemokine for 30 min at 37 °C and analyzed by flow cytometry (Fig. 1C). Addition of increasing concentrations of MIP-1 did not significantly reduce M7 cell surface level on either HOS or CEM. AOP-MIP-1, was weakly active on both cell lines, inducing internalization of 20 and 30% on HOS and CEM, respectively. These results suggest a dependence of CCR5 internalization on cytoplasmic tail phosphorylation, consistent with what has been found for other GPCRs. They also suggest that the more potent AOP derivative can cause a low level of internalization by an alternative mechanism. Interestingly, in the absence of chemokine, M7-CCR5 surface expression was 2-fold higher than wild-type in both cell lines (Fig. 1, compare A with C). This may result from the decreased rate of constitutive internalization by the mutant as compared with the wild-type chemokine receptor. In keeping with this, entry of JR.FL-pseudotyped luciferase reporter viruses was reproducibly 20-40% higher for cells expressing M7-CCR5 than wild-type (60,500 ± 40 cps versus 48,000 ± 5800 cps for M7-CCR5 and wild-type on HOS.CD4, respectively, in a representative experiment). As expected, cytoplasmic tail phosphorylation was not important for virus entry, consistent with the findings of others using tail truncations (22, 23).

Kinetics of CCR5 Internalization in Live Cells-- The carboxyl-terminal EGFP allowed us to view the intracellular trafficking of CCR5 upon ligand binding in real time by fluorescence microscopy. For this, HOS.CD4 cells expressing wild-type or M7-CCR5 were plated on coverslips and exposed to 100 ng/ml AOP-MIP-1 at 37 °C. Images of the cells were collected at a rate of 1/min. Prior to chemokine addition, CCR5.EGFP was primarily localized to the plasma membrane (Fig. 3A). After 5 min, the receptor had redistributed forming a speckled pattern in the cytoplasm; however, redistribution was apparent as early as 1 min after chemokine addition when the first image was collected (not shown). After 15 min, the receptor had moved inwards in large hollow vesicles that accumulated at one side of the nucleus. After 30 min, the receptors were concentrated at a perinuclear region of the cell where they remained for the course of the experiment. Native MIP-1 was a less potent inducer of CCR5 internalization (Fig. 3B). Redistribution of the receptor was first apparent 15 min post-chemokine addition and the vesicles that formed remained distributed in the cytoplasm and did not show pronounced perinuclear accumulation.

View larger version (30K): [in this window] [in a new window]   Fig. 3.   CCR5 internalization in live cells. HOS cells stably expressing CD4 and wild-type CCR5.EGFP were exposed to (A) AOP-MIP-1 or (B) MIP-1 for indicated times and visualized under UV illumination in a 37 °C chamber. Chemokines were added with a perfusion pump to a final concentration of 200 ng/ml. Images were collected for 1 h at a rate of 1/min. Images shown were taken at 0, 5, 15, 30, and 60 min after chemokine addition. Cells shown are representative of those on the coverslip.

To compare the intracellular movement of the wild type and M7-CCR5 in response to chemokine, the cells were treated with chemokine and fixed after 4 h. The localization of wild-type or M7-CCR5 was unaffected by 10 ng/ml MIP-1 (Fig. 4), consistent with the insensitivity of HOS cells to chemokine that was found by flow cytometry. At 100 ng/ml, wild-type CCR5 redistributed to vesicles underlying the plasma membrane. In contrast, M7-CCR5 localization was unaffected. AOP-MIP-1 at high concentration caused redistribution of wild type CCR5 but did not affect the localization of M7-CCR5, confirming data gathered by flow cytometry.

View larger version (74K): [in this window] [in a new window]   Fig. 4.   Internalization of wild-type and M7-CCR5.EGFP in HOS cells. Cells were treated with medium or medium containing MIP-1 or AOP-MIP-1 (10 or 100 ng/ml) for 4 h at 37 °C. Cells were fixed, mounted, and visualized under UV light. Cells shown are representative of those on the coverslip.

M7-CCR5.EGFP Mediates Ligand-induced Signal Transduction-- Cytoplasmic tail phosphorylation is not required for signal transduction by GPCRs but is required for receptor desensitization (53). To test whether this was the case for CCR5, we compared [Ca²⁺] mobilization in response to MIP-1 in HOS.CD4 cells expressing wild-type or M7-CCR5.EGFP. Both receptors responded to the chemokine with a rapid increase in [Ca²⁺] (Fig. 5). M7-CCR5.EGFP reproducibly mobilized a higher concentration of intracellular [Ca²⁺] as compared with wild type. In addition, the [Ca²⁺] peak occurred later (60 s as compared with about 10 s post-ligand addition). This could have been due to less rapid shut-off of the mutant receptor due to the lack of cytoplasmic tail phosphorylation. Unexpectedly, the mutant receptor was as efficiently desensitized as the wild-type. GPCR desensitization is generally mediated by cytoplasmic tail domain phosphorylation. Thus, CCR5 appears to be able to desensitize via an alternative mechanism of desensitization that is independent of cytoplasmic tail phosphorylation.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   [Ca2+] flux mediated by wild-type and M7-CCR5 in response to MIP-1. HOS cells stably expressing wild-type or M7-CCR5.EGFP were loaded with Indo-1 AM. Cells were stimulated with 100 ng/ml MIP-1, and [Ca2+] was measured in a spectrofluorimeter. Chemokine was added at 60 and 180 s as indicated by the arrows. The parental cell line, HOS.CD4, did not signal in the presence of chemokine (data not shown). Note differences in the y axes.

Sensitivity to Chemokine Inhibition Is Associated with CCR5 Internalization-- The reduced ability of M7-CCR5 to internalize in response to chemokine allowed us to measure the component of chemokine-mediated inhibition of HIV-1 entry that was attributable to coreceptor internalization. If chemokine inhibits virus entry only by competing with gp120 for CCR5 binding, then M7-CCR5-mediated entry would be blocked as effectively as that mediated by wild-type CCR5. Alternatively, if chemokine inhibits virus entry only by ligand-induced internalization, then M7-CCR5-mediated entry would not be inhibited by chemokine. The contribution of the two mechanisms could thus be evaluated by comparing entry inhibition mediated by the two receptors. To evaluate the relative contributions of both mechanisms, HOS.CD4 cells expressing wild-type or M7-CCR5 were incubated with increasing concentrations of chemokine for 30 min and then infected with M-tropic, JR.FL-pseudotyped luciferase reporter virus (51, 52). MIP-1 inhibited 20% of wild type-mediated and M7-CCR5-mediated entry at 1 and 10 ng/ml, concentrations at which CCR5 internalization does not occur. At 100 ng/ml MIP-1, which induces internalization of 25% of the cell surface CCR5 on HOS, inhibition increased to 55%. MIP-1 did not fully inhibit entry into HOS, probably because not all of the CCR5 is cleared from the cell surface. In contrast, for M7-CCR5-mediated entry, inhibition remained constant at 20% for all chemokine concentrations. Correspondingly, M7-CCR5 is not internalized at any of these chemokine concentrations in HOS (Fig. 6). Thus, for M7-CCR5, only the binding site competition component of chemokine activity is active, and this accounts for 20% of entry inhibition.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   M7-CCR5-mediated HIV entry is resistant to chemokines. HOS and CEM cells stably expressing CD4 and wild-type or M7-CCR5.EGFP were incubated with MIP-1 or AOP-MIP-1 for 30 min at 37 °C and then infected with 1 ng/ml p24 JR.FL-pseudotyped NL-Luc-ER luciferase reporter virus (51, 52). Luciferase activity was measured 3 days postinfection. All infections were performed in triplicate, and a representative of three experiments is shown. Values are expressed as the percentage of inhibition as compared with the average luciferase activity in the absence of chemokine.

AOP-MIP-1 was significantly more active than MIP-1 at blocking entry through wild-type CCR5 on HOS. At concentrations as low as 1 ng/ml, AOP-MIP-1 inhibited 50% of virus entry. At 100 ng/ml AOP-MIP-1, inhibition was boosted to 70%, reflecting the inefficient receptor internalization in these cells. For M7-CCR5-mediated entry, AOP-MIP-1 was even less effective, increasing inhibition only slightly over the entire concentration range. Thus, the inefficiency with which this receptor is internalized, even in response to a potent ligand, demonstrates the importance of internalization in the blocking effects of chemokines.

In CEM cells, CCR5 internalization is induced by lower chemokine concentrations, resulting in more effective inhibition of virus entry. At 1 ng/ml MIP-1, a concentration that does not induce receptor internalization, virus entry mediated by wild-type CCR5 was inhibited by 25%. At 10 and 100 ng/ml MIP-1, concentrations at which internalization is induced, entry was inhibited by 80 and 100%, respectively. Entry mediated by M7-CCR5 was less effectively blocked by MIP-1, increasing to only 40% at a concentration as high as 100 ng/ml. This finding again demonstrated the importance of coreceptor internalization for chemokine blocking of HIV entry. AOP-MIP-1 blocked wild-type CCR5-mediated entry very effectively in CEM, with 95% of entry blocked. In contrast, the M7-CCR5-mediated entry was largely resistant to the effects of the more potent AOP derivative, consistent with the inefficiency with which the receptor is internalized in response to ligand.

CCR5 Antagonist TAK-779 Inhibits M7-CCR5-mediated Entry-- To rule out the possibility that M7-CCR5 was inherently less susceptible to entry inhibitors, we tested the efficacy of the CCR5 antagonist TAK-779 (26, 54, 55) on wild type and M7-CCR5. TAK-779 was similarly effective at inhibiting virus entry mediated by wild type or M7-CCR5. The drug does not induce internalization of wild type (Fig. 7) or mutant receptor (not shown). Thus, M7-CCR5-mediated entry is fully susceptible to inhibition by interference with its gp120 interaction site. This difference between the action of chemokine and the small molecule inhibitor highlights the difference in the mechanisms by which the two compounds block virus entry. The finding also demonstrates that the chemokine inhibition results described above for M7-CCR5 were not due to the trivial explanation of differences in receptor cell surface abundance.

View larger version (48K): [in this window] [in a new window]   Fig. 7.   Inhibition of HIV by TAK-779 is independent of internalization. A, HOS stably expressing CD4 and wild-type or M7-CCR5.EGFP were incubated with TAK-779 for 30 min at 37 °C and then infected with JR.FL-pseudotyped NL43 RE luciferase reporter virus (1 ng of p24) (51, 52). Cells were washed 4 h postinfection and cultured at 37 °C for an additional 3 days. Luciferase activity was measured for triplicate infections, and a representative of two experiments is shown. Values are expressed as the percentage of inhibition as compared with the average luciferase activity in the absence of inhibitor. B, intracellular localization of CCR5.EGFP after treatment with 1 × 106 M TAK-779 for 4 h at 37 °C. Cells were fixed and viewed under UV illumination.

Overexpression of Internalization Accessory Proteins Enhances Chemokine Protection-- GRK phosphorylation of the cytoplasmic tail of GPCRs results in association with arrestin and internalization via clathrin-coated pits. If internalization efficiency of the coreceptor limits the blocking activity of chemokines in HOS cells as compared with CEM, then increasing the intracellular abundance of endocytic pathway components in HOS might lead to more efficient protection by chemokines. We tested this prediction by transiently expressing GRK2, Arr-2, and Arr-3 in HOS.CD4.R5 cells and measuring chemokine blocking activity in the luciferase reporter virus entry assay (Fig. 8A). At 10 ng/ml MIP-1, Arr-2, Arr-3, and GRK2 all modestly increased chemokine inhibition. Coexpression of the factors did not significantly further increase their activity. At 100 ng/ml MIP-1, the internalization cofactors did not increase blocking efficiency, suggesting that these internalization-related cofactors could not fully restore sensitivity to chemokine.

View larger version (27K): [in this window] [in a new window]   Fig. 8.   Expression of transfected internalization components and dominant negative derivatives alters the ability of chemokines to inhibit HIV infection. HOS (2 × 106) stably expressing CD4 and CCR5.EGFP were transiently transfected with vectors expressing various internalization accessory factors (A) or dominant-negative variants (B). One day posttransfection, the cells were replated, and the next day they were incubated with chemokine for 30 min at 37 °C and then infected with JR.FL-pseudotyped NL43 RE luciferase reporter virus (51, 52). Values are expressed as the percentage of inhibition as compared with the average luciferase activity in the absence of chemokine. The amount of entry in the absence of chemokine was defined as 0% inhibition (1-5 × 104 cps). 100% inhibition was defined as cps in cultures with no reporter virus (50-200 cps). Infections were performed in triplicate, and a representative of two experiments is shown.

Dominant Negative Internalization Cofactors Prevent Chemokine-mediated Protection-- As shown above, the effectiveness of chemokine at inhibiting virus entry is modestly improved by increasing the abundance of various cellular endocytic components in HOS cells. To further address the role of internalization in chemokine function, we asked whether interfering with endocytosis would prevent chemokine blocking activity. To do this, dominant negative forms of individual endocytic pathway components were introduced into cells using transient transfection of appropriate expression vectors. Three dominant negatives were used: dynamin K44A, Arr-3-(284-409), and GRK2-K220R. Dynamin mediates the scission of clathrin-coated vesicles from the cell membrane. Dynamin K44A is deficient in GTP binding, and thus interferes with dynamin-mediated scission of the clathrin-coated vesicles (56). Arr-3- (284-409) lacks the receptor binding region of arrestin but competes for clathrin binding (47). GRK2-K220R is mutated in the catalytic domain but retains the ability to bind receptor (57). Dynamin-K44A and arrestin-3-(284-409) have been shown to block CXCR4 internalization (36). The three dominant negative components were transiently expressed in HOS.CD4.R5.EGFP, and the cells were challenged 48 h posttransfection with JR.FL pseudotyped luciferase reporter viruses. Dominant-negative arrestin-3 and GRK2 modestly relieved inhibition, while dynamin K44A eliminated the inhibition at 10 ng/ml MIP-1 and reduced it to 20% at 100 ng/ml (Fig. 8B). In some experiments, a slight enhancing activity on virus entry was noted in dynamin K44A-expressing cells incubated with 10 ng/ml MIP-1. This may be due to increased steady-state cell surface CCR5 that results from the impairment of internalization.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We present evidence here that the efficiency of receptor internalization can cause cell type differences in susceptibility to the blocking effects of chemokines on HIV entry. A mutated CCR5 that lacked the cytoplasmic tail phosphorylation sites failed to internalize in response to chemokine and was resistant to the blocking effects of chemokine on HIV entry. This demonstrated that chemokine receptor internalization is a major component of the inhibitory mechanism of chemokines, consistent with findings of others (21-23). The effectiveness of chemokine-mediated virus entry inhibition correlated with receptor internalization in two cell lines, HOS and CEM. 40-Fold more MIP-1 was required to comparably inhibit entry into HOS cells than CEM. Correspondingly, HOS cells inefficiently internalized CCR5 upon chemokine binding in comparison with CEM. These findings could explain the relative resistance of cell types such as macrophages to the blocking effects of chemokines. The mechanistic basis for cell type differences in receptor internalization rates is not clear. It is possible that cell type differences in endogenous levels of internalization-related factors such as GRKs, dynamin, or arrestins play a role. The kinetics of GPCR internalization have been shown to correlate with endogenous arrestin expression (58), and transient overexpression of GRKs and arrestins synergistically enhanced the internalization of GPCRs (36, 59). In support of this, we found that increasing the abundance of arrestin and GRK by transient transfection could modestly increase the susceptibility of HOS cells to chemokine blocking of HIV entry.

The relative resistance of M7-CCR5 to chemokine-mediated inhibition of HIV entry highlights the importance of receptor internalization, a finding first made using transiently transfected cytoplasmic tail truncations of CCR5 (23) and CXCR4 (22). In initial studies, we found that CCR5s with cytoplasmic tail truncations were poorly expressed on the cell surface, suggesting that this domain is important for trafficking of the receptor to the cell surface. Ser³²⁵ was also important. A CCR5 mutated at seven of the nine cytoplasmic tail Ser/Thr/Tyr residues was expressed on the cell surface at least as well as wild type. This molecule served as a useful tool for studying internalization. The use of stably expressed full-length CCR5 reduced the possibility of conformational alteration to the mutated receptor and further ensured that resistance to chemokine inhibition was not the result of overexpression that can occur in transient transfection. The finding that TAK-779 was equally effective at preventing wild type- or M7-CCR5-mediated infection further demonstrated that the mutant was correctly folded and was equally inhibitable by steric blocking and that the results were not due to the 2-fold increased cell surface abundance of M7-CCR5.

Because the M7-CCR5 was not internalized in HOS cells in response to native chemokine, we were able to distinguish the relative contributions of binding site interference and receptor internalization to virus entry inhibition. M7-CCR5-mediated virus entry into HOS cells was inhibited by about 20% at low chemokine concentration and did not increase with increasing chemokine concentration. In contrast, wild-type CCR5-mediated entry was inhibited by 17% at low chemokine concentration and increased steadily with increasing chemokine concentration. Thus, 20% of the chemokine effect can be attributed to steric blocking of the interaction between gp120 and CCR5, and the remainder can be attributed to receptor internalization. As the chemokine concentration was increased, the wild-type but not the mutant receptor was increasingly internalized, reducing virus entry. The AOP-modified chemokine inhibited entry through M7-CCR5 by about 20% throughout the range of chemokine concentrations. It was significantly more active on wild type CCR5, inhibiting entry more efficiently by 50% at low concentration and increasing to 75% with greater chemokine concentration. These findings are consistent with the increased potency of the AOP-modified chemokine to induce internalization of the wild type receptor and to prevent its recycling to the cell surface (24). Taken together, these findings suggest that as much as 80% of the chemokine blocking of HIV entry is due to receptor internalization. The finding that high concentrations of chemokine do not block entry through M7-CCR5 leads to the conclusion that a chemokine receptor when bound by ligand can maintain its ability to interact with gp120. This conclusion is unexpected given the partial overlap between the chemokine and gp120 binding sites (7, 9). It is possible that chemokine does not fully obscure the gp120 interaction site on CCR5 or that occasional dissociation of the chemokine from its receptor allows interaction with gp120.

Using a CCR5.EGFP fusion protein, we were able to visualize the process of chemokine-induced CCR5 internalization in living cells by fluorescence microscopy. The carboxyl-terminal EGFP did not alter CCR5 function as the receptor signaled and internalized efficiently in response to ligand. Chemokine receptor internalization was remarkably rapid. As early as 1 min after the addition of AOP-MIP-1, in the first image that could be collected, chemokine receptor clustering had already initiated. After 15 min, the receptors localized to one side of the nucleus in vesicles in a region close to the Golgi apparatus. Although this accumulation could have been in the Golgi, we do not believe that this was the case, since two-color analysis of CCR5.EGFP and the Golgi marker, mannose 6-phosphate receptor, showed only partially overlapping fluorescence.³ In cells treated with native MIP-1, the CCR5 was clustered in cytoplasmic vesicles. These did not show a pronounced perinuclear localization but were more closely associated with the plasma membrane. This difference in localization is consistent with the ability of the AOP derivatives, but not the native chemokine, to prevent the chemokine receptors from recycling to the cell surface (24).

GPCR desensitization is in general mediated by phosphorylation of cytoplasmic tail residues followed by association with -arrestin and internalization. Thus, the finding that M7-CCR5 was shut off and desensitized as well as wild type was unexpected. M7-CCR5 mobilized a rapid increase in intracellular [Ca²⁺] that reached a peak concentration about 4-fold higher than that of the wild type receptor. [Ca²⁺] returned to base line slightly more rapidly than for the wild-type, indicating a rapid shut-off of the receptor and efficient receptor desensitization. The increased [Ca²⁺] peak may have been due to the 2-fold higher cell surface abundance of M7-CCR5, but clearly, desensitization of CCR5 by chemokine does not require cytoplasmic tail phosphorylation. It may be that CCR5 differs from canonical GPCRs in that arrestin binding to the ligand-bound receptor does not require cytoplasmic tail phosphorylation. Alternatively, desensitization may proceed by a novel mechanism that is independent of phosphorylation or internalization.

The differences we detected in the ability of cells to internalize CCR5 in response to chemokine has implications both in HIV pathogenesis and for therapeutic intervention. Monocytes and lymphocytes are susceptible to modulation of CCR5 by chemokines (24); however, differentiated macrophages lose the capacity to down-regulate CCR5 in response to chemokines (44). Donor differences in the propensity of T cells or macrophages to internalize CCR5 in response to physiologically produced chemokines could influence cell surface coreceptor density and thus affect virus replication rates. Sabbe et al. (60) found that CCR5 promoter polymorphisms in the human genome that are linked to changes in disease progression rates are associated with differences in CCR5 recycling rates. Given the ability of various internalization accessory factors, such as GRKs or arrestins, to influence chemokine blocking of HIV entry, it is conceivable that the polymorphisms are linked to genes that alter these cellular cofactors. With regard to development of entry inhibitors, it is important to distinguish whether these act by blocking the gp120 interaction site on the coreceptor or by inducing coreceptor internalization. Inducing internalization has the advantage of receptor removal from the cell surface, plus the virus cannot readily become resistant to the effects of the inhibitor. Optimal inhibitors may be ones that act through both mechanisms. M7-CCR5 may be a useful tool for distinguishing these two mechanisms of inhibition.

	ACKNOWLEDGEMENTS

We thank Ge Wei, Beth Rasala, Don Kaiser, David Chambers, and Kelly Hardwicke for technical assistance; Gerard Graham, Robb Nibbs, and Jeffery Benovic for reagents; Brian Egan and Sohela DeRozieres for critical reading of the manuscript; and Lynn Artale for graphics.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants AI43252, CA7214, and AI42397; Elisabeth Glaser Pediatric AIDS Foundation Grant 77328 (to R. M.); University of California San Diego Center for AIDS Research development grant AI36214 (to R. M.); American Foundation for AIDS Research Grant 02758-30-RGT (to R. M.); and the Pendleton Foundation.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

A predoctoral fellow of the Howard Hughes Medical Institute.

^** An Elizabeth Glaser Scientist of the Pediatric AIDS Foundation. To whom correspondence should be addressed: The Salk Institute for Biological Studies, Infectious Disease Laboratory, 10010 N. Torrey Pines Rd., La Jolla, CA 92037. Tel.: 858-453-4100; Fax: 858-554-0341; E-mail: landau@salk.edu.

Published, JBC Papers in Press, January 8, 2002, DOI 10.1074/jbc.M108232200

² S. M. Brandt, R. Mariani, A. U. Holland, T. J. Hope, and N. R. Landau, unpublished observations.

³ T. Hope, unpublished observation.

	ABBREVIATIONS

The abbreviations used are: GPCR, G-protein-coupled receptor; CD4, cluster of differentiation 4; CCR5, CC chemokine receptor 5; MIP, macrophage inflammatory protein; HIV, human immunodeficiency virus; gp, glycoprotein; CXCR4, CXC chemokine receptor 4; RANTES, regulated on activation normal T cell expressed and secreted; EGFP, enhanced green fluorescent protein; GRK, G-protein coupled receptor kinase; HOS, human osteosarcoma; CEM, CEM.SS; AOP, aminooxypentane; cps, counts/s; Arr, arrestin.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES