By continuing to browse this site you agree to us using cookies as described in About Cookies
Wiley Online Library is migrating to a new platform powered by Atypon, the leading provider of scholarly publishing platforms. The new Wiley Online Library will be migrated over the weekend of March 17 & 18. You should not experience any issues or loss of access during this time. For more information, please visit our migration page: http://www.wileyactual.com/WOLMigration/
P. G. Strange, School of Animal and Microbial Sciences, University of Reading, PO Box 228, Reading, RG6 6AJ, UK. Fax: + 44 118378 6537, Tel.: + 44 118378 8015, E-mail: P.G.Strange@rdg.ac.uk
CCR5 is a G protein-coupled receptor that binds several natural chemokines but it is also a coreceptor for the entry of M tropic strains of HIV-1 into cells. Levels of CCR5 on the cell surface are important for the rate of HIV-1 infection and are determined by a number of factors including the rates of CCR5 internalization and recycling. Here we investigated the involvement of the actin cytoskeleton in the control of ligand-induced internalization and recycling of CCR5. Cytochalasin D, an actin depolymerizing agent, inhibited chemokine-induced internalization of CCR5 and recycling of the receptor in stably transfected CHO cells and in the monocytic cell line, THP-1. CCR5 internalization and recycling were inhibited by Toxin B and C3 exoenzyme treatment in CHO and THP-1 cells, confirming activation of members of the RhoGTPase family by CCR5. The specific Rho kinase inhibitor Y27632, however, had no effect on CCR5 internalization or recycling. Ligand-induced activation of CCR5 leads to Rho kinase-dependent formation of focal adhesion complexes. These data indicate that CCR5 internalization and recycling are regulated by actin polymerization and activation of small G proteins in a Rho-dependent manner.
Chemokine-mediated responses are induced by the activation of G protein-coupled receptors (GPCR) on target cells . The chemokine receptor, CCR5, has been identified as a receptor for the chemokines MIP1-α, MIP1-β, RANTES, MCP-2, MCP-3 and MCP-4 . CCR5 also serves as a coreceptor for the entry of M-tropic HIV-1 strains into cells [3–6]. The number of coreceptors (CCR5) expressed on the cell surface is an important factor for virus infection of cells. Patients with a reduced number of CCR5 expressed on the cell surface – caused by a mutation in the CCR5 gene that leads to a truncated receptor that is not expressed on the cell surface (CCR5Δ32) – have a degree of resistance to HIV-1 infection [7–10].
The number of receptors on the cell surface depends on the rate of internalization and the rate of replacement (recycling and new synthesis). Previously, we and others showed that chemokine activation of CCR5 leads to its internalization, with the participation of clathrin-coated pits and caveolae, followed by recycling to the cell surface via recycling endosomes [11,12]. CCR5 recycling is independent of transport from early to late endosomes and cannot be inhibited with nocodazole , an inhibitor of microtubules. Other factors that are potential regulators of internalization and the recycling processes of GPCRs are yet to be defined.
The actin cytoskeleton and related proteins may be involved in ligand-induced intracellular trafficking of GPCRs [13,14]. Some of the evidence for this relates to CCR5. RANTES has been shown in T cells to induce the activation of two tyrosine kinases, ZAP70 and p125focal adhesion kinase (FAK), and stimulate both their association with paxillin and the tyrosine phosphorylation of the related Pyk2 kinase [15,16]. CCR5 is able to signal through activation of related adhesion focal tyrosine kinase (RAFTK) , Pyk2  and also through the protein tyrosine phosphatases SHP1 and SHP2 . It has been suggested that CCR5 directly interacts with FAK to activate this kinase [15,16,19]. These proteins are usually linked to activation of small G proteins (Rho GTPases). There is some evidence that GPCRs activate these small G proteins and affect actin polymerization [21,22]. Activated Rho GTPases interact with cellular target proteins or effectors to trigger a wide variety of cellular responses, including the reorganization of the actin cytoskeleton [23–26]. Cytochalasin D inhibits CXCR1 and CXCR2 internalization , suggesting an involvement of actin polymerization in chemokine receptor internalization and recycling. The present study therefore investigated the role of actin polymerization and small G protein activation in internalization and recycling of CCR5. The work has been performed using CHO cells stably transfected with CCR5 (CHO.CCR5 cells) or CCR5 and CD4 (CHO.CCR5.CD4 cells) and the monocytic cell line, THP-1.
Materials and methods
Cells and materials
CHO cells stably expressing CCR5 (CHO.CCR5) or CCR5 and CD4 (CHO.CCR5.CD4) have been described previously [12,27]. THP-1 cells were from the MRC Aids Reagent Repository Programme (Potters Bar, UK) and grown in 10% (v/v) fetal bovine serum/RPMI/glutamine (2 mm). The MIP-1α isoform used is the MIP-1α D26A form and has been described in  and was a generous gift from British Biotech (Oxford, UK). Secondary antibodies were obtained from Sigma (Poole, UK). Anti-CCR5 IgGs, HEK/1/85a/7a and anti-CD4 IgGs have been described previously [12,27]. All other chemicals were from Sigma.
Cells were incubated for 1 h at 37 °C with cytochalasin D (1 µm), nocodazole (30 µm), Toxin B of Clostridium difficile (1 µg·mL−1), C3 exoenzyme of Clostridium botulinum (1 µg·mL−1), genistein (30 µg·mL−1), Y27632 (20 µm), staurosporine (2 ng·mL−1) or LY294002 (Sigma) (10 µm) before an internalization assay. Treatment with pertussis toxin (PTX; Sigma) was performed for 18 h at 100 ng·mL−1.
Internalization assay and flow cytometry analysis
THP-1, CHO.CCR5 and CHO.CCR5.CD4 cells each were incubated with serum-free medium for 1 h at 37 °C, harvested with 2 mm EDTA/NaCl/Pi and then resuspended in medium without serum at 5 × 106 cells·mL−1. Cells were then incubated with chemokines (50–100 nm) for 1 h at 37 °C, and washed in ice-cold NaCl/Pi containing 1% (v/v) fetal bovine serum and 1% (w/v) NaN3 for flow cytometry analysis. Cell surface expressed CCR5 was detected by flow cytometry using anti-CCR5 IgG HEK/1/85a/7a and fluorescein isothiocyanate (FITC) conjugated anti-rat IgG. Cells were incubated for 1 h at room temperature with HEK/1/85a/7a (saturating amounts of hybridoma supernatant), washed three times with NaCl/Pi buffer containing 1% (v/v) fetal bovine serum and 1% (w/v) NaN3 and incubated for 1 h with FITC-labelled anti-rat IgG. Samples were quantified on a FACScan and data analysed with cellquest software version 3.1 (Becton Dickinson, San Jose, CA, USA). For each experiment (e.g. when examining the effects of a chemokine on cell surface CCR5) the relative CCR5 surface expression (%) was calculated as:
where Fctc is the mean channel of fluorescence (chemokine-treated cells), Fneg is the mean channel of fluorescence (negative control cells) and Fmws is the mean channel of fluorescence (cells treated with medium without serum). CHO cells not expressing CCR5, as well as irrelevant monoclonal antibodies were used for negative controls with similar results. Where an inhibitor such as cytochalasin D was used in the experiments, it was present in the three sets of cells used, i.e. negative control, treated with medium without serum and chemokine-treated.
Recycling of receptor
Internalization was initiated as described. After 1 h incubation with chemokines, the cells were washed three times in medium without fetal bovine serum and resuspended in medium without fetal bovine serum at 37 °C. Samples were taken at different time points and cells were washed in NaCl/Pi buffer containing 1% (v/v) fetal bovine serum and 1% (w/v) NaN3. Cells were stained with antibodies as described.
Cells were grown on coverslips and incubated in medium without serum for 1 h before treatment with chemokines for 1 h. The cells were then washed with medium and incubated with CCR5 antibody for 1 h at room temperature. After washing, the cells were incubated with the corresponding secondary tetramethyl rhodamine isothiocyanate (TRITC) or FITC-labelled antibody for 1 h, washed and fixed in ice-cold methanol and mounted on glass slides. Images were taken using a Leica NT Confocal Imaging system or a Zeiss Axiovision Imaging system. For actin staining after a CCR5 stain, cells were fixed with 4% paraformaldehyde and 100% acetone according to the manufacturer's protocol and stained with Alexa Fluor 488 phalloidin (Molecular Probes, Eugene, OR). Focal adhesion complexes were stained with an anti-vinculin IgG (Sigma) and a corresponding secondary antibody.
The [35S]GTP[γS]-binding assays were carried out essentially as described in . Cell membranes (30 µg) were incubated in [35S]GTP[γS]-binding buffer containing 20 mm Hepes, 100 mm NaCl, 10 mm MgCl2 and 0.1% (w/v) BSA, pH 7.4 and 10 µm GDP in a final volume of 0.9 mL in the absence or presence of an agonist (50 nm) to measure the basal and stimulated nucleotide exchange, respectively. The reaction was initiated by the addition of cell membranes and the tubes were incubated at 30 °C for 30 min. This preincubation ensured that ligand binding was at equilibrium before addition of 100 µL of [35S]GTP[γS] to give a final [35S]GTP[γS] concentration of 100 pm. The assay was incubated for a further 30 min before termination by rapid filtration through GF/C filters with four washes of 3 mL of ice-cold NaCl/Pi (140 mm NaCl, 10 mm KCl, 1.5 mm KH2PO4, 8 mm Na2HPO4) using a Brandel cell harvester (Gaithersburg, MD, USA). The filters were soaked for at least 6 h in 2 mL of LKB OptiPhase Hisafe 3 scintillation fluid after which bound radioactivity was determined by liquid scintillation counting.
Analysis of intracellular calcium ion concentration
Cells were harvested with 2 mm EDTA/NaCl/Pi and washed twice in buffer (148 mm NaCl, 5 mm KCl, 2.5 mm CaCl2, 10 mm Hepes, 1 mm glucose, 2.5 mm probenecid, 0.1% (w/v) BSA, pH 7.4) and incubated with 4 µm Fura-2AM (Molecular Probes, Eugene, OR, USA) at 37 °C. After washing cells with buffer, cells were resuspended at 2 × 106 cells·mL−1 of buffer. Chemokine-induced intracellular calcium mobilization was determined by spectrofluorometry using a Perkin Elmer LS50 fluorometer and the peak calcium ion concentration was determined as in .
Data were analysed using graphpad prism (GraphPad Software, San Diego, CA, USA). Statistical analysis was performed using Student's t-test. Internalization data represent the means and SEM of at least three independent experiments.
Cytochalasin D treatment inhibits CCR5 internalization and recycling
CHO.CCR5 and CHO.CCR5.CD4 cells were treated with cytochalasin D (1 µm) for 1 h before inducing internalization with MIP-1α (50 nm) or LD78β (100 nm), respectively (Table 1). Treatment with cytochalasin D inhibited ligand-induced internalization of the receptor significantly in both cell lines and for both chemokines used. The effect of cytochalasin D on the binding of chemokines to the receptor and the function of the receptor was assessed using a [35S]GTP[γS] binding assay (Table 2). MIP-1α stimulation of [35S]GTP[γS] binding to membranes of CHO.CCR5.CD4 cells was not affected by this inhibitor showing that cytochalasin D does not affect chemokine binding or activation of CCR5 after acute treatment. Pretreatment of CHO.CCR5.CD4 cells with cytochalasin D (1 h, 1 µm) did not affect the ability of MIP-1α to mobilize intracellular calcium in the cells  (Table 2) showing that longer term treatment with this inhibitor does not reduce the affinity of MIP-1α to bind to CCR5 or its ability to activate the receptor.
Table 1. Effects of cytochalasin D on CCR5 internalization. Effects of cytochalasin D on CCR5 internalization induced by MIP-1α and LD78β. CHO.CCR5 or CHO.CCR5.CD4 cells, pretreated with cytochalasin D (1 µm) for 1 h where indicated, were treated with 50 nm MIP-1α or 100 nm LD78β and cell surface CCR5 was determined using flow cytometry as described in the Materials and methods section. Cell surface expression in chemokine-treated cells was calculated as a percentage of control cells that were not treated with chemokine but treated with inhibitor in the same way as chemokine-treated cells. Data represent means ± SEM (n experiments), * P < 0.05, ** P < 0.001 for effect of cytochalasin D relative to chemokine control.
Table 2. Effects of cytochalasin D on MIP-1α binding and signalling via CCR5. MIP-1α stimulation of [35S]GTP[γS] binding to membranes of CHO.CCR5.CD4 cells was determined as described in  under control conditions or in the presence of 1 µm cytochalasin D. Concentration–response curves were determined and the percentage stimulation over basal and the concentration of MIP-1α giving a half maximal response (EC50) were recorded. pEC50≡−log EC50. MIP-1α stimulation of intracellular calcium ion release in CHO.CCR5.CD4 cells treated for 1 h with 1 µm cytochalasin D or buffer (control) was determined as described in the Materials and methods section and the response to 100 nm MIP-1α is given. Data are mean ± SEM for three experiments ([35S]GTP[γS] binding) or four experiments (Ca2+ release).
MIP-1α stimulated [35S]GTP[γS] binding
MIP-1α stimulated Ca2+ release
% Stimulation over basal
9.03 ± 0.10 (0.93)
227 ± 20
26.1 ± 2.2
8.98 ± 0.01 (1.05)
209 ± 14
24.0 ± 2.2
The effects of cytochalasin D on recycling of CCR5 were then characterized. Receptor internalization was induced for 1 h with MIP-1α and then cells were incubated for the times indicated with cytochalasin D or medium as a control. Cytochalasin D completely inhibited receptor recycling in CHO.CCR5.CD4 cells as well as in CHO.CCR5 cells (Table 3). These results suggest an involvement of actin polymerization in CCR5 internalization and recovery.
Table 3. Effects of inhibitors on CCR5 recycling following internalization induced by MIP-1α. CHO.CCR5, CHO.CCR5.CD4 or THP-1 cells were treated with 50 nm MIP-1α for 1 h and recycling of CCR5 was determined after washing and incubating in medium at 37 °C, as described in the Materials and methods section. Where indicated, inhibitors were present during the recycling phase of the receptor. For these experiments, data should be compared to the MIP-1α control data for each cell line. Data represent means ± SEM (n, experiments). *** P < 0.0001, ** P < 0.01, * P < 0.05, relative to control (MIP-1α). ND, not determined.
The results above were obtained in transfected CHO cells. In order to investigate the internalization and recycling of CCR5 in a nontransfected system, we used the monocytic cell line THP-1 that expresses CCR5 as determined using flow cytometry analysis. The expression level in THP-1 cells is approximately half of that in CHO.CCR5 cells (Fig. 1). Ligand-induced internalization of CCR5 could be detected in THP-1 cells to a comparable extent to that seen in CHO cells (Table 3). We next investigated the recycling of the receptor in THP-1 cells. Internalization of CCR5 was induced by treatment with MIP-1α (50 nm). During the recovery phase, cells were incubated either in medium or in medium with cytochalasin D or in medium with nocodazole. In congruence with the results obtained in CHO cells , nocodazole (30 µm) did not inhibit CCR5 recycling, whereas cytochalasin D was able to block receptor recycling completely (Table 3). The recycling mechanisms in THP-1 cells therefore seem to be similar to those observed in CHO cells, whereby inhibition of actin polymerization affects CCR5 internalization and recycling, whilst microtubules do not seem to be involved in these processes.
The role of RhoGTPases in CCR5 internalization and recycling, and actin polymerization
To investigate the importance of small G protein activation for CCR5-induced actin polymerization, cells were treated with Toxin B of C. difficile (1 µg·mL−1), which has been shown to inactivate all members of the RhoGTPase family at that concentration . Pre-treatment of cells resulted in an inhibition of internalization (Table 4) in all cell lines tested. Toxin B also inhibited LD78β (100 nm) induced internalization in CHO.CCR5 and CHO.CCR5.CD4 cells (data not shown). These results were similar to those obtained with cytochalasin D. Internalization was then induced in untreated cells and Toxin B was added only during the recycling phase of the receptor. Toxin B completely inhibited receptor recycling (Table 4). To specify which member of the RhoGTPase family is activated by CCR5, C3 exoenzyme of C. botulinum was used to inhibit specifically Rho activation in cells . As shown in Table 4, C3 exoenzyme (1 µg·mL−1) inhibits internalization to a similar extent as seen with Toxin B, however, the extent of CCR5 recycling is higher after C3 exoenzyme treatment than after Toxin B treatment. This could suggest that inhibition of Rho is not completely sufficient to inhibit all receptor recycling. C3 exoenzyme also inhibits LD78β (100 nm) induced internalization in CHO.CCR5.CD4 and CHO.CCR5 cells (data not shown).
Table 4. Effects of inhibitors on CCR5 internalization and recycling induced by MIP-1α. CHO.CCR5, CHO.CCR5.CD4 or THP-1 cells were treated with 50 nm MIP-1α and internalization and recycling of CCR5 was determined as described in the Materials and methods section. In one set of experiments, cells were pretreated with C3 exoenzyme, Toxin B and Y276322 were indicated for 1 h before internalization was induced with MIP-1α. After 1 h cells were washed and incubated in medium with inhibitor at 37 °C. At given time points samples were taken and subjected to flow cytometry stain. In a second set of experiments, CCR5 internalization was induced by treatment with MIP-1α as above and inhibitors were present only during the recycling phase of the receptor. For this set of experiments, data should be compared to the MIP-1α control data for each cell line. Data represent means ± SEM (n, experiments). *** P < 0.0001, ** P < 0.01, * = P < 0.05, relative to control (MIP-1α). ND, not determined.
Cell surface CCR5 (% untreated control)
Recycling 60 min
Recycling 120 min
Recycling 60 min
Recycling 120 min
Recycling 60 min
Recycling 120 min
52.7 ± 4.5 (7)
79.8 ± 4.8 (5)**
90.3 ± 7.5 (11)**
68.6 ± 1.6 (12)
93.6 ± 7.1 (10)**
91.4 ± 6.5 (7)**
69.1 ± 3.8 (9)
88.3 ± 5.8 (10)*
92.5 ± 7.4 (10)*
Cells pretreated with inhibitors
73.5 ± 6.2 (7)*
72.6 ± 6.7 (6)
68.4 ± 4.6 (7)
79.3 ± 6.9 (5)*
77.6 ± 8.3 (4)
78.6 ± 9.1 (4)
97.9 ± 9.6 (6)*
96.0 ± 17.5 (6)
105.9 ± 20.9 (6)
78.1 ± 8.5 (3)*
77.0 ± 7.3 (3)
81.8 ± 8.9 (4)
79.8 ± 5.6 (4)**
87.2 ± 6.9 (3)
71.8 ± 1.8 (3)
91.7 ± 7.0 (4)*
97.9 ± 5.2 (4)
87.7 ± 6.5 (4)
51.4 ± 5.5 (3)
79.8 ± 4.8 (5)**
90.3 ± 7.5 (11)*
57.6 ± 2.8 (8)
92.3 ± 7.4 (6)***
107.1 ± 11.9 (6)**
63.3 ± 5.2 (6)
87.9 ± 6.9 (6)*
94.6 ± 7.5 (6)
Inhibitors present only during recycling
67.2 ± 4.2 (6)
67.3 ± 4.1 (6)
63.5 ± 7.9 (4)
59.1 ± 4.4 (7)
79.2 ± 5.2 (5)
69.0 ± 7.1 (5)
43.9 ± 9.2
54.3 ± 7.0 (5)
49.1 ± 12.3 (4)
66.4 ± 11.0 (4)
66.9 ± 5.6 (5)
In further experiments, CCR5 was activated with MIP-1α (50 nm) for 1 h and both CCR5 and actin were visualized by immunofluorescence. Activation of CCR5 leads to loss of fluorescence and therefore to receptor internalization and also to induction of actin polymerization as shown by formation of stress fibres (Fig. 2). Cytochalasin D (1 µm) treatment inhibited CCR5 internalization as well as stress fibre formation (Fig. 2). We also used immunofluorescence to monitor the effects of Toxin B treatment on actin polymerization. Toxin B completely blocked any formation of actin stress fibre formation induced by CCR5 (Fig. 2). Immunofluorescence following C3 exoenzyme treatment showed similar results to those obtained with Toxin B treatment (Fig. 2). These data, however, clearly show an involvement of Rho activation in CCR5 internalization and recycling.
Internalization of CCR5 is Rho kinase-independent
The data presented above showed that activation of Rho, as well as stress fibre formation, were important for CCR5 internalization and recycling. We therefore used an inhibitor for Rho activated kinase (ROCK) to investigate downstream pathways of Rho in the internalization system. Cells were treated with Y27632 (20 µm) , a specific inhibitor for ROCK, before internalization and recycling assays were performed. Y27632 treatment did not significantly inhibit internalization or recycling of CCR5 in CHO cells (Table 4), although Y27632 completely inhibited the formation of actin stress fibres in the cells (Fig. 3). Even though our previous data suggest activation of Rho by CCR5, we are able to rule out ROCK as a regulatory protein for CCR5 internalization or recycling. Similar results for CCR5 internalization and recycling were also observed after treatment of CHO cells with genistein (30 µg·mL−1) and staurosporine (2 ng·mL−1) (data not shown). Although these two compounds do inhibit ROCK, they are not specific for this kinase and also inhibit other protein tyrosine kinases in cells [31,32]. However these data, taken together, clearly demonstrate that ROCK activation is not essential for CCR5 internalization and recycling. Similarly the formation of stress fibres in cells is not essential for receptor trafficking.
Activation of stress fibres and focal adhesion complex formation are ROCK-dependent
Pretreatment of cells with Y27632 completely inhibited chemokine induced formation of stress fibres through CCR5 but did not prevent CCR5 internalization (Fig. 3). Neither LY294002 (10 µm), a phosphatidyl inositol 3-kinase inhibitor, nor pertussis toxin (100 ng·mL−1, 18 h) inhibited CCR5 induced stress fibre formation. (Fig. 3). We have shown previously that pertussis toxin treatment inhibits CCR5 internalization using flow cytometry  and these results can be verified in confocal microscopy (Fig. 3). LY294002 however, does not inhibit CCR5 internalization nor is there an effect on stress fibre formation (Fig. 3).
CCR5 activation also leads to focal adhesion complex formation that can be visualized with a specific antibody against vinculin (Fig. 4). Y27632, Toxin B and C3 exoenzyme treatment prevented the formation of the complex (Fig. 4). These data show that activation of Rho and ROCK is necessary for the formation of focal adhesion complexes.
As the experiments above were performed in stably transfected cell lines, we used the nontransfected, monocytic cell line, THP-1 to examine the generality of the results described above. Experiments with Y27632 were repeated in the monocytic cell line THP-1 with comparable results to those in CHO cells (Table 4), suggesting that the signalling pathways described are not cell type specific and can also be detected in a nontransfected cell system.
Expression of CCR5 on the cell surface is important for the ability of M-tropic HIV-1 strains to infect cells. Low expression of CCR5 or complete absence of CCR5 from the cell surface is associated with immunity against HIV-1 infection [7–10], as well as a slower progression of the disease . It is important, therefore, to understand how CCR5 expression on the cell surface is regulated. CCR5 is internalized following agonist-activation through clathrin-coated pits after β-arrestin-binding and caveolae may also be involved . Recycling of the receptor back to the cell surface is independent of protein synthesis. CCR5 seems therefore to be transported into recycling endosomes [11,12] and then back to the cell surface. In the present study, we investigated in detail the mechanisms involved in CCR5 internalization and recycling and the role of the actin cytoskeleton in these processes.
Cytochalasin D has been shown to inhibit actin polymerization in cells. Pre-treatment of stably transfected CHO cells or the monocytic cell line THP-1 with cytochalasin D inhibits MIP-1α-induced internalization of CCR5. As well as the effects of cytochalasin D on CCR5 internalization, the inhibitor also affected recycling of CCR5 in all cell lines tested. These data suggest that stimulation of actin polymerization and rearrangement of the cytoskeleton is necessary for movement of CCR5 from the cell surface to the cytoplasm and back again. As actin polymerization is induced by activation of members of the RhoGTPase family, we then investigated whether activation of these small G proteins is also essential for the transport of CCR5. Both Toxin B and C3 exoenzyme inhibited ligand-induced CCR5 internalization and recycling with the two MIP-1α isoforms, suggesting a role for Rho GTPases, particularly Rho, in controlling transport of the receptor.
We then investigated the downstream pathways of Rho. Rho can activate several effectors, such as mDia or ROCK . Using a ROCK-specific inhibitor (Y27632) it was shown that CCR5 internalization and recycling are independent of ROCK activation, but stress fibre and focal adhesion complex formation are dependent on activation of ROCK. The data obtained with C3 exoenzyme treatment suggest activation of Rho by CCR5, but ROCK activation does not seem to be involved in regulating CCR5 internalization or recycling. It therefore seems likely that CCR5 is able to activate Rho and effectors of Rho including ROCK. Whereas ROCK activation leads to stress fibre activation, other Rho effectors are involved in receptor internalization and regulation of recycling. Differences between two Rho effectors have been described previously. ROCK causes the contraction of pre-existing actin filaments whereas mDia induces de novo actin polymerization [23,34]. It has also been described that mDia and ROCK have opposing effects on adherens junctions . Furthermore mDia colocalizes with stable microtubules when over expressed and associates with microtubules in vitro. ROCK, however, is not necessary for the formation of stable microtubules . Whether CCR5 is able to activate mDia or another Rho effector other than ROCK remains to be clarified. The data obtained with cytochalasin D, toxin B and C3 exoenzyme suggest that the presence of stress fibres is important for CCR5 internalization and recycling. The effects of ROCK inhibition and subsequent inhibition of stress fibre formation but not internalization and recycling, demonstrate that parallel stress fibres are induced by ligand-induced CCR5 activation but are not essential for internalization or recycling. It seems likely that other cytoskeletal rearrangements, like membrane ruffling or diffuse localization of actin filaments  induced by mDia, which are not as easily visible as stress fibres using confocal microscopy are involved in trafficking CCR5 into the cells and back to the cell surface (Fig. 5).
Stress fibre activation in CHO cells is therefore induced by CCR5 after challenge with an agonist and can be prevented by inhibition of ROCK. This activation however, is independent of Gi/o protein activation as shown by studies with pertussis toxin and LY294002. These inhibitors do not affect formation of stress fibres in the cells tested. Activation of CCR5 also leads to the formation of focal adhesion complexes in CHO cells. The formation of these complexes was prevented by the inhibitors Y27632, Toxin B as well as C3 exoenzyme, suggesting a dependence on Rho activation.
There seem to be two distinct regulatory systems for CCR5 internalization and recycling (Fig. 5). Whereas heterotrimeric G proteins (Gi/o) may be involved in internalization, as shown by the effects of pertussis toxin [27, Fig. 3], these G proteins are not involved in actin stress fibre formation following CCR5 activation. Nevertheless Rho-dependent actin rearrangement is important for CCR5 internalization and recycling.
This is the first report to show the importance of cytoskeletal rearrangement and Rho activation for CCR5 internalization and recycling. Recently it has been shown that internalization of CXCR1 and CXCR2 is inhibited by cytochalasin D  and also that IL-8 activates Rho and actin stress fibre formation in endothelial cells due to activation of CXCR1 . These data together with the present study suggest that rearrangement of the cytoskeleton and activation of Rho GTPases are a common feature of the mechanisms involved in regulating receptor activation, internalization and recycling for chemokine receptors. Activation of changes in the cytoskeleton control the capacity of the receptor to internalize and recycle, showing that a complex system is responsible for controlling receptor movement. Rho seems to be a central regulator, controlling receptor internalization and recycling and the induction of stress fibres and focal adhesion complex formation. CCR5 activation subsequently leads to activation of ROCK and induction of stress fibres and focal adhesion complex formation, but ROCK is not a regulatory protein for receptor internalization or recycling. mDia has been identified among others as another effector that is regulated by Rho activation. It has been shown that a dominant active mutant of mDia induced only the diffuse localization of actin filaments in contrast to parallel stress fibres that are induced by active ROCK . The data presented here suggest that this change in the diffuse localization of actin filaments or activation of other Rho effectors could be important for controlling CCR5 internalization and recycling.
The present report has provided some more details of the pathways involved in the internalization and recycling of CCR5. This has wider relevance given that CCR5 is important for the entry of HIV-1 in to cells and that the levels of CCR5 on cells, dependent on the balance of internalization and recycling, are important for the rate of entry. The present report may therefore provide new targets for the design of compounds that could modulate levels of CCR5 on cells and hence the rate of HIV-1 entry.
We thank Dr Jane McKeating for her involvement in the inception of the project and for generating cell lines and antibodies. We thank Dr Christine Shotton for monoclonal antibodies to CCR5 and Dr Lloyd Czaplewski (British Biotech) for various chemokine reagents. We thank the Centralized Facility for AIDS Reagents, supported by EU Programme EVA (contract BMH4 97/2515) and the UK Medical Research Council. This work was supported by the Biotechnology and Biological Sciences Research Council.