Human vascular smooth muscle cells possess functional CCR5.

CC chemokine receptors are important modulators of inflammation. Although CC chemokine receptors have been found predominantly on leukocytes, recent studies have suggested that vascular smooth muscle cells respond to CC chemokines. We now report that human smooth muscle cells express CCR5, a co-receptor for human immunodeficiency virus. CCR5 mRNA was detectable by RNA blot hybridization in human aortic and coronary artery smooth muscle cells. The cDNA generated by reverse transcription-polymerase chain reaction from aortic smooth muscle cells had 100% identity throughout the entire coding region with the CCR5 cloned from THP-1 cells. By immunohistochemistry, CCR5 and the CCR5 ligand, macrophage inflammatory protein-1beta (MIP-1beta), were detected in smooth muscle cells and macrophages of the atherosclerotic plaque. In smooth muscle cell culture, MIP-1beta induced a significant increase in intracellular calcium concentrations, which was blocked by an antibody to CCR5. In addition, MIP-1beta caused a calcium-dependent increase in tissue factor activity. Tissue factor is the initiator of coagulation and is thought to play a key role in arterial thrombosis. These data suggest that human arterial smooth muscle cells express functional CCR5 receptors and MIP-1beta is an agonist for these cells.

Smooth muscle cells (SMCs) 1 are the predominant cellular elements of the arterial media. In the normal arterial wall, SMCs are in a quiescent, contractile state and function primarily to regulate vascular tone. In atherosclerosis, SMCs modulate to a synthetic phenotype, allowing them to proliferate and migrate to form part of the intimal plaque (1). SMC proliferation and hypertrophy are also features of hypertension (2). A variety of growth factors and cytokines found in the atheroscle-rotic plaque and in the injured arterial wall have been shown to be agonists for SMCs and to modulate their growth and migratory properties (3)(4)(5).
CC chemokines are members of a family of highly related proteins characterized by two adjacent cysteine residues (6). These proteins play a crucial role in a variety of inflammatory processes by acting as leukocyte activators and chemoattractants (6,7). CC chemokines bind to a family of G proteincoupled receptors, CC chemokine receptors (CCRs), characterized by seven transmembrane spanning domains (8). CCRs have been identified predominantly on leukocytes (9,10). We recently reported that the CC chemokine, monocyte chemoattractant protein-1 (MCP-1) specifically binds to human aortic SMCs and induces tissue factor (TF), the initiator of the clotting cascade (11). In these cells, the mRNA encoding the known MCP-1 receptor, CCR2 (12) could not be identified by RNA blot hybridization or by reverse transcription-polymerase chain reaction (RT-PCR). In addition, these cells did not express mRNA for other CCRs, including CCR1 (13), CCR3 (14), CCR4 (15), and DARC (16), as determined by RT-PCR. However, the responsiveness of SMCs to MCP-1 suggested that they expressed a member(s) of the CCR family. Since our initial report, several additional CCRs have been identified (17). In particular, CCR5 has been shown to be the receptor for macrophage inflammatory protein-1 (MIP-1) ␣ and ␤ (18) and to be a co-receptor for macrophage-tropic or dual-tropic strains of human immunodeficiency virus (HIV) (19).
We now report that cultured human arterial SMCs possess CCR5 mRNA and protein. The SMC CCR5 was functionally coupled, responding to MIP-1␤ with increases in intracellular calcium concentration ([Ca 2ϩ ] i ) and in TF activity. Antibody to CCR5 blocked both the increases in [Ca 2ϩ ] i and in TF activity. CCR5 and MIP-1␤ antigens were also detected in SMCs of the atherosclerotic arterial wall. CCR5 and its ligand, MIP-1␤, may play a role in mediating the inflammatory and prothrombotic responses associated with atherosclerosis.

EXPERIMENTAL PROCEDURES
Growth Factors and other Reagents-Recombinant human anti-human CCR5 antibody (2D7) and its isotype control antibody were purchased from Pharmingen International (San Diego, CA). Polyclonal anti-human MIP-1␤, anti-human CCR5 (monoclonal antibody 182), and anti-human CCR2 antibodies were purchased from R & D Systems (Minneapolis, MN). An inhibitory monoclonal antibody against human tissue factor (20)  detection of activated Factor Xa was purchased from American Diagnostica (Greenwich, CT).
Cell Culture-SMCs were isolated from human thoracic aortas and coronary arteries harvested from explanted hearts at the time of cardiac transplantation. Coronary artery SMCs were prepared by enzyme digestion as described previously for rat aortic SMCs (21). Aortic SMCs were prepared by explant (22). Human internal mammary arteries were obtained from patients undergoing coronary artery bypass surgery. SMCs from these specimens were also prepared by explant (23). SMCs were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% FBS, 100 units/ml penicillin, and 100 g/ml streptomycin and were serially passaged before reaching confluence. Cells were identified as smooth muscle by their typical appearance on light microscopy and by immunostaining with antibody to human smooth muscle ␣-actin (0.1 g/ml) (IA4, DAKO). The lack of endothelial cell contamination of SMCs in culture was established by the absence of staining with antibody (0.3 g/ml) to human von Willebrand factor (F8186, DAKO). Experiments were performed on cells from passages 3-7, except for internal mammary arteries, in which only passages 1-3 were used.
RNA Preparation and Blot Hybridization-Extraction of total RNA, agarose gel electrophoresis, transfer to nitrocellulose, and hybridization to 32 P-labeled DNA were as described previously (21). Prehybridization and hybridization were performed at 42°C. Final washes for all blots were in 0.1ϫ SSC (1ϫ ϭ 0.15 M NaCl/0.015 M sodium citrate, pH 7.0) and 0.1% SDS at 50°C for 1 h. Hybridization was performed using a 1116-bp segment of the human CCR5 cDNA (bases 584 -1699), labeled by random oligomer priming to a specific activity of greater than 10 8 cpm/mg, and used at 2 ϫ 10 6 cpm/ml.
RT-PCR-RT-PCR was done using the following cycling conditions: for reverse transcription, 5 g of total RNA was incubated with 0.5 g of oligo(dT) [12][13][14][15][16][17][18] (Life Technologies, Inc., Gaithersburg, MD), at 68°C for 10 min and quick chilled. RNA was then added to reverse transcriptase mix consisting of two hundred units of Superscript II reverse transcriptase (Life Technologies, Inc.), 2.5 mM dNTPs, 1ϫ reverse transcriptase buffer, (Life Technologies, Inc.), 0.1 N dithiothreitol (Sigma). This mixture was incubated for 1 h at 42°C and then 5 min at 99°C. Subsequent PCRs were performed under the following cycling conditions: initial incubation at 94°C for 3 min, followed by 35 cycles of 30 s at 94°C, 30 s at 55°C, and a final incubation for 10 min at 72°C. RT-PCR products were ligated into the TA3 vector using the TA cloning kit (Invitrogen, San Diego, CA) and sequenced commercially (Biotechnology Center, Utah State University). One set of primer pairs corresponded to nucleotides 453-471 (sense) and 745-764 (antisense) of CCR5. The second set corresponded to nucleotides 57-77 (sense) and 1137-1157 (antisense), encompassing the entire CCR5 coding region and 50 bases into the 5Ј untranslated region. The PCR for the latter primer pair was performed as above except that the temperature of the annealing step was 52°C.
Immunohistochemistry-SMCs were incubated in single-well Permanox slides (Nunc Inc., Naperville, IL). Slides were fixed in 4% paraformaldehyde (in phosphate-buffered saline (PBS), pH 7.4), blocked with normal goat serum and 3% H 2 O 2 in water, washed in PBS, and incubated with monoclonal anti-human CCR5 antibody (5 g/ml) for 2 h at 37°C. Human anti-CCR5 primary antibody was detected with a rabbit anti-mouse IgG antibody conjugated with biotin (Biogenex, San Ramon, CA). Secondary antibodies were detected with streptavidinperoxidase and developed with diaminobenzidine. Slides were counterstained with hematoxylin, coverslipped, and examined.
Primary antibody was detected as described above. Controls for each experiment included processing the specimens using the nonimmune IgG isotype as the primary antibody and omission of primary antibody. THP-1 cells were used as a positive for CCR5.

Measurement of [Ca 2ϩ ] i -Human
SMCs were grown to 60 -70% confluence on sterile 25-mm coverslips. Cells were loaded with 10 M fura-2/AM in 0.1% dimethyl sulfoxide in HEPES-buffered solution con-taining145 NaCl mM, 5 KCl mM, 2.5 mM CaCl 2, 1.2 mM MgSO 4 , 11 mM glucose, and 10 mM HEPES for 1 h at 25°C. Measurements of [Ca 2ϩ ] i were performed as described previously (23,(25)(26)(27). Loaded cells were washed three times and placed in a Leiden chamber with 1 ml of HEPES-buffered solution on the stage of a Nikon Dapshot microscope (Nikon Corp., Melville, NY). All experiments were carried out at 25°C in HEPES-buffered solution. MIP-1␤ was diluted in 0.2 M filtered PBS with 0.1% bovine serum albumin (cell-culture grade) and stored at Ϫ80°C as per the manufacturer's guidelines. Prior to the addition of agonists, an equal volume of diluent alone was added to cells and recorded for 10 min as a control for possible contamination. Cells that did not respond to MIP-1␤ were treated with endothelin-I (10 ng/ml) to establish that they were capable of agonist-mediated [Ca 2ϩ ] i mobilization. Cells that did not respond to either agonist were excluded from analysis. Cells that responded to endothelin-I but not MIP-1␤ were classified as nonresponders. For the blocking experiments, cells were pretreated antibodies or isotype-matched nonimmune IgG for 15 min prior to addition of MIP-1␤ (100 nM).
Images were sequentially obtained from a manual gain charge-coupled device camera at excitation wavelengths of 350 and 380 nm by means of a filter wheel (Shutter Instruments, Novato, CA) with emission set at 510 nm by a dicroic mirror. Data were analyzed for [Ca 2ϩ ] i levels using the Image 1AT/FL software package (Media, PA). Pixel by pixel ratio imaging was used to obtain spatial maps. Background light levels were subtracted, and the ratios of the intensities of the individual pixels at the two wavelengths were calculated (operationally logarithms were subtracted). Ratio maps were used to obtain spatial maps of [Ca 2ϩ ] i in individual SMCs (25,28). As such, these dual wavelength measurements permitted for calculation of [Ca 2ϩ ] i in nM, independent of indicator concentration. For these studies, the peak amplitude of the agonist-induced [Ca 2ϩ ] i transient was calculated from whole cell averages of cytosolic Ca 2ϩ levels. All images from which such measurements were obtained represented the ratio (350:380) of the average of eight images obtained at each wavelength.
Measurement of Tissue Factor Activity-Aortic SMCs (passages 3-7) were grown in six-well plates and grown in DMEM with 10% FBS. Prior to treatment, cells were placed DMEM with 0.3% FBS for 24 h. MIP-1␤ in DMEM (40 l) or DMEM alone was added directly to the cultures and TF activity was measured aft 4 h. For blocking experiments, antibodies at a final concentration of 20 g/ml, or BAPTA/AM (10 g/ml) were added 1 h prior to treatment with MIP-1␤ (100 ng/ml). TF activity was measured as described previously (22). Cells were washed two times with 10 mM HEPES (pH 7.6), 140 mM NaCl, and 4 mM CaCl 2 ; trypsinized; and then resuspended in 0.5 ml of the same solution. An 80-l aliquot of the cell suspension was diluted 1:1 in octyl-␤-D-glycopyranoside (15 mM) and transferred to a 96-well microtiter Dyne plate. Factors VIIa and X, at 1 and 150 nM (final concentration) respectively, were added sequentially to cell lysates. Aliquots (40 l) were then taken every 2 min for 6 min and placed in a 96-well plate containing 100 l of Bicine buffer (pH 8.5, 1 g/liter bovine serum albumin and 25 mM EDTA) to stop the reaction. Factor Xa was assayed by adding 25 l of Spec-trozyme® (2 mM) to each well and measuring the changes in absorption at 405 nm at 35°C in an enzyme-linked immunosorbent assay plate reader (Tmax Molecular Devices, Menlo Park, CA). The concentration of FXa was calculated from the slope of the absorption over time. TF activity, in fmol/cm 2 , was calculated assuming a k cat of the TF-VIIa complex of 300 min Ϫ1 at 20°C.

Identification of CCR5 in Human
SMCs-To determine whether CCR5 is expressed in human SMCs, RNA blot analysis was performed using a 1.1-kb fragment of the human CCR5 cDNA. The CCR5 probe hybridized to a band of a size identical to that found in THP-1 cells (Fig. 1A). No band was seen with RNA from human endothelial cells. To provide further evidence that this band represents CCR5 mRNA, RT-PCR was performed with two different primer pairs. One pair encoded a 300-bp product within the coding region (Fig. 1B); the second encompassed the entire coding region (not shown). Bands of identical size were seen in lanes containing RNA from SMCs and THP-1 cells. No bands were detected when reverse transcriptase was omitted from the reactions. The SMC bands from both sets of primers were subcloned and sequenced. The sequences had 100% identity with that of the previously published CCR5 cDNA (GenBank TM accession number U54994) (18).
To establish that human SMCs express CCR5 protein, immunohistochemistry was performed with a monoclonal antibody to human CCR5. As shown in Fig. 2, human coronary SMCs stained intensely for CCR5 in the cytoplasm and on the membrane.
Human SMCs Contain Functioning CCR5-MIP-1␤ is the most potent known agonist for CCR5 (18). To determine whether the SMC CCR5 is functionally coupled, the ability of MIP-1␤ to mobilize [Ca 2ϩ ] i was examined. As shown in Table I and Fig. 3A, MIP-1␤ induced increases in [Ca 2ϩ ] i in internal mammary artery (3.7 Ϯ 0.7-fold), coronary artery (3.3 Ϯ 0.2fold), and aortic SMCs (10.8 Ϯ 2.6-fold). The average time to a response was 1-4 min and appeared to vary by the anatomic source of the SMCs. As shown in Table I and Fig. 3B, the MIP-1␤-induced increase in [Ca 2ϩ ] i was blocked by two different antibodies to CCR5, whereas it was not blocked by antibody to the CC chemokine receptor, CCR2. An antibody to TF, an abundant SMC surface molecule not known to mobilize [Ca 2ϩ ] i did not block the induction of [Ca 2ϩ ] i by MIP-1␤. Nonimmune isotype-matched IgG for CCR5 also had no effect on the induction of [Ca 2ϩ ] i by MIP-1␤ (not shown).
MIP-1␤ Induces TF Activity in SMCs-Agonist-induced mobilization of [Ca 2ϩ ] i has been found to be coupled to gene induction in a variety of cells, including SMCs. TF synthesis, in particular, is mediated via a Ca 2ϩ -dependent mechanism (21,   29). To assess the biologic relevance of MIP-1␤-induced [Ca 2ϩ ] i mobilization, we examined the induction of TF activity in aortic SMCs. MIP-1␤ increased TF activity in a concentration dependent manner (Fig. 4A). To determine whether the response to MIP-1␤ (100 ng/ml) was specific for CCR5, the cells were pretreated with a monoclonal antibody to CCR5 (20 g/ml) (2D7) (Fig. 4B). This antibody significantly reduced the induction of TF activity by MIP-1␤, whereas antibody to CCR2 (20 g/ml) had no effect (Fig. 4B). The nonimmune IgG isotype (20 g/ml) also had no effect on the induction of TF activity by the MIP-1␤ (Fig. 4B). To verify that the induction of TF activity by MIP-1␤ was a consequence of the increase in [Ca 2ϩ ] i , cells were pretreated with BAPTA/AM (10 g/ml) to chelate [Ca 2ϩ ] i prior to addition of MIP-1. This treatment has been shown to block growth factor-mediated increases in [Ca 2ϩ ] i and the induction of early genes, including TF, in SMCs (21, 29 -31). BAPTA/AM completely blocked the induction of TF activity (Fig. 4C).
CCR5 Expression in the Arterial Wall-Abundant CCR5 antigen was detected in the fibrous cap, necrotic core, and media of human coronary artery atheroma specimens and localized with ␣-actin positive SMCs and CD-68 positive macrophages (Figs. 5, A-C). CCR5 was not detected in normal human coronary arteries (Fig. 5C, inset). MIP-1␤ antigen had a similar distribution in the plaque to CCR5, with antigen detected in the SMCs and macrophages (Fig. 5D). In contrast, CCR2 antigen was detected in the macrophage-rich areas of the plaque (Figs. 5F, green arrows) but not in the ␣-actin positive regions (Figs. 5F, red arrows). CCR3 antigen had a similar macrophage-restricted pattern of expression (not shown).

DISCUSSION
This report describes the presence of functional CCR5 in human vascular SMCs. CCR5 is a member of the family of CCRs and the only known receptor, to date, that binds MIP-1␤ at physiologic concentrations (7). CCRs have been found largely on leukocytes and have been shown to mediate a variety of biologic processes, including inflammation, chemotaxis, proliferation, and cell activation (6,32,33). CCR5 has received particular attention because of its role as a co-receptor for macrophage-tropic strains of HIV (34).
There has been little information about the presence of CCR5 or other CCRs on cells of the arterial wall. CCR5 antigen was identified in blood vessels from a variety of tissues using mono- clonal antibodies (35). Staining was largely confined to the endothelium but some staining was reported in SMCs underlying positive endothelium. Luo et al. (36) examined the binding and activity of the CC chemokine, TCA3, in rat aortic SMCs. Mouse TCA3, MIP-1␣, and MCP-1 stimulated chemotaxis of SMCs and increased SMC adhesiveness to type III collagen. Rat SMCs exhibited high affinity (3 nM) binding sites for TCA3. Mouse MIP-1␤ had no binding to or biologic effect on rat SMCs.
We recently reported (29) that human aortic SMCs express TF in response to MCP-1. Although high affinity MCP-1 binding sites were identified, RT-PCR failed to demonstrate mRNA for the MCP-1 receptor, CCR2, or for CCR1, CCR3, CCR4, and DARC in cells grown in 10% serum. In a study also employing RT-PCR (37), mRNAs for CCR1 and CCR2 were expressed by human saphenous vein SMCs incubated in serum-free medium for 24 h. CCR3, CCR4, CCR5, CXCR1, and CXCR2 mRNA were not identified in these cells. The detection of CCR5 in our study may be due to differences in study conditions. The former study (37) utilized saphenous vein SMCs, whereas the present study used only arterial SMCs. There were also differences in the culture conditions. Our mRNA was isolated from cells grown in 10% serum, whereas theirs was harvested from cells grown in serum-free medium. Our data demonstrate that in addition to possessing CCR5 mRNA, human SMCs contain significant amounts of CCR5 antigen. Most importantly, CCR5 appears to be functionally active, in that these cells respond to MIP-1␤, the ligand specific for CCR5. Although it is possible that MIP-1␤ could be acting through a different CC chemokine receptor, the ability of antibodies to CCR5 to completely block the calcium response and TF induction by MIP-1␤ strongly argues that CCR5 is responsible.
MIP-1␣ and ␤ are small heparin-binding proteins that are produced by a variety of cells, including macrophages, neutrophils, fibroblasts, and epithelial cells (13). MIP-1␣ and ␤ are chemotactic agents and activators of monocytes and lymphocytes and, as such, exhibit potent proinflammatory properties. To our knowledge, this is the first report showing that SMCs respond to MIP-1␤. MIP-1␤ caused increases in [Ca 2ϩ ] i in SMCs derived from aorta, coronary, and internal mammary arteries. The magnitude and time to induction (1-4 min) are similar to those shown for the response of SMCs to plateletderived growth factor and epidermal growth factor (38 -41) 2 but distinct from the more rapid (15 s) and pronounced response to the potent vasoconstrictor, angiotensin II (42). These growth factor-mediated changes in [Ca 2ϩ ] i have been associated with a contractile response in isolated aortic rings (38 -41) and with the induction of early response genes in SMC culture (21,30,31).
MIP-1␤ treatment resulted in a marked increase in TF activity in human SMCs. This increase was blocked by antibody to CCR5 and by BAPTA/AM, suggesting that its induction was due to CCR5-mediated [Ca 2ϩ ] i mobilization. TF is considered a key mediator of thrombosis in the setting of atherosclerotic plaque rupture (43)(44)(45) and acute arterial injury (46,47). Of note, the 6-fold induction of TF by MIP-1␤ is more potent than the induction previously reported in these cells in response to MCP-1 (29) or platelet-derived growth factor (21).
CCR5 acts as a co-receptor for HIV-1 (48). Several experimental animal models of HIV are associated with vascular abnormalities. Transgenic mice carrying a replication-defective HIV-1 provirus with expression restricted to SMCs were shown to develop extensive vasculopathy, characterized by intimal thickening, significant luminal narrowing, and thrombotic occlusion (49). A severe arteriopathy, also characterized by intimal thickening, luminal narrowing, and thrombosis, was seen in macaques infected with simian immunodeficiency virus (SIV) (50).
To our knowledge, there have been no reports of direct SMC infection by HIV. The expression of functional CCR5 on SMCs suggests that the question of their infectability should be reexamined. SMCs do not express CD4; however, data using SIV indicate that many primary strains use CCR5 in the absence of CD4. For example, simian microvascular brain endothelial cells that express CCR5 but lack CD4 can be infected with neurotropic SIV strains. This infection can be blocked using CCR5 ligands (51). There is also evidence that soluble CD4, through its binding to GP120, induces conformational changes that facilitate SIV binding and infection via CCR5 (52-55). Thus, it is possible that CD4-independent, CCR5-dependent strains of HIV can be generated that would have the potential to infect SMCs. SMCs could play a role in HIV infection without actually incorporating virus. Infected cells within the vessel wall may elaborate MIP-1␤ that would then activate SMCs. This could contribute to the arteriopathies reported in patients with AIDS and Kaposi's sarcoma (56).
MIP-1 antigens and/or mRNAs have been identified at various sites of inflammation (57)(58)(59). However, there is limited information regarding their presence in atherosclerotic plaques. By in situ hybridization, low levels of MIP-1 mRNA were found in human carotid endartarectomy specimens (3). Our study suggests that MIP-1␤ antigen is abundant in SMCrich areas of the plaque. Significantly, intimal and medial SMCs also possess the CCR5 receptor and therefore appear to be competent to respond in vivo to MIP-1␤. This report thus raises the possibility that MIP-1␤ plays a role in the inflammatory and prothrombotic response of SMCs in atherosclerosis.