Monocyte chemotactic protein-2 activates CCR5 and blocks CD4/CCR5-mediated HIV-1 entry/replication.

Human immunodeficiency virus, type I (HIV-1) cell-type tropism is dictated by chemokine receptor usage: T-cell line tropic viruses use CXCR4, whereas monocyte tropic viruses primarily use CCR5 as fusion coreceptors. CC chemokines macrophage inflammatory protein (MIP)-1alpha, MIP-1beta, and RANTES (regulated on activation normal T cell expressed and secreted) inhibit CD4/CCR5-mediated HIV-1 cell fusion. MCP-2 is also a member of the CC chemokine subfamily and has the capacity to interact with at least two receptors including CCR-1 and CCR2B. In an effort to further characterize the binding properties of MCP-2 on leukocytes, we observed that MCP-2, but not MCP-1, effectively competed with MIP-1beta for binding to monocytes, suggesting that MCP-2 may interact with CCR5. As predicted, MCP-2 competitively inhibited MIP-1beta binding to HEK293 cells stably transfected with CCR5 (CCR5/293 cells). MCP-2 also bound to and induced chemotaxis of CCR5/293 cells with a potency comparable with that of MIP-1beta. Confocal microscopy indicates that MCP-2 caused remarkable and dose-dependent internalization of CCR5 in CCR5/293 cells. Furthermore, MCP-2 inhibited the entry/replication of HIV-1ADA in CCR5/293 cells coexpressing CD4. These results indicated that MCP-2 uses CCR5 as one of its functional receptors and is an additional potent natural inhibitor of HIV-1.

on chemokine receptor usage, and T-cell line tropic viruses use CXCR4, whereas monocyte tropic viruses primarily use CCR5 as fusion coreceptors. A minority of HIV-1 strains may also use other CC chemokine receptors as their fusion co-factors. Dual tropic HIV-1 strains presumably interact with more than one type of chemokine receptor (5,8). The CC chemokines MIP-1␣, MIP-1␤, and RANTES were able to inhibit the entry of monocyte tropic viruses (9), whereas the CXC chemokine SDF-1 abrogates CXCR4-mediated fusion by T lymphotropic HIV-1 strains (6,7).
Monocyte chemotactic protein (MCP)-2 is a CC chemokine co-purified with MCP-1 and MCP-3 from human osteosarcoma cells (10 -12). It shares over 60% amino acid identity with MCP-1 and MCP-3 and has about 30% identity with the CC chemokines MIP-1␣, RANTES, and MIP-1␤ (10 -12). MCP-2, similar to MCP-3, is chemotactic for and activates a wide variety of inflammatory cells, including monocytes, T lymphocytes, NK cells, basophils, mast cells, and eosinophils (12), but differs from MCP-1, which is not active on eosinophils (13). We recently reported that MCP-2 uses CCR1 and CCR2B as its functional receptors (14), which may account for its action on a greater variety of target cells. In the course of studies on leukocyte activation by MCP-2, we observed that MCP-2 could competitively inhibit the binding to monocytes of 125 I-MIP-1␤, a CC chemokine that is believed thus far to exclusively use CCR5 as a functional receptor on human leukocytes (15)(16)(17). This prompted us to further investigate the effect of MCP-2 on cloned CCR5. We report that MCP-2 is also an efficient ligand for CCR5 and a potent inhibitor of CD4/CCR5-mediated HIV-1 entry/replication.

EXPERIMENTAL PROCEDURES
Chemokines-Recombinant human (rh) MCP-2 and other chemokines were purchased from PeproTech Inc. (Rocky Hill, NJ). Radioiodinated MCP-2 was a kind gift from Dr. G. Brown (NEN Life Science Products). Other radioiodinated chemokines were purchased from NEN. All radioiodinated chemokines have a specific activity of 2200 Ci/mmol. Cells-Human peripheral blood monocytes were isolated from normal donors (National Institutes of Health Clinical Center, Transfusion Department, Bethesda, MD) with an iso-osmotic Percoll (Pharmacia Biotech Inc.) gradient as described elsewhere (18). The monocyte preparations were Ͼ90% pure. The 293 human embryonic kidney epithelial cell line stably transfected with FLAG-tagged CCR5 (CCR5/293) was generated and grown in monolayers as described (17).
Binding Assays with Radiolabeled Chemokines-Binding assays were performed by using a single concentration of radiolabeled chemokines in the presence of increasing concentrations of unlabeled ligands as described previously (14,18). Cells (2 ϫ 10 6 /sample for monocytes and 1 ϫ 10 6 /sample for CCR5/293 cells) were suspended in 200 l of modified binding medium composed of RPMI 1640, 1 mg/ml bovine serum albumin, 25 mM HEPES, and 0.05% sodium azide and incubated in duplicates at room temperature for 40 min. After incubation, the cells were pelleted through a 10% sucrose/phosphate-buffered saline cushion, and the radioactivity associated with cell pellets was determined in a ␥-counter (Clinigamma-Pharmacia, Gaithersburg, MD). Experiments were also performed at 4°C in the absence of sodium azide and yielded similar binding and competition curves as obtained at room temperature. The binding data were then analyzed with a Macintosh computer program LIGAND (P. Munson, Division of Computer Research and Technology, NIH, Bethesda, MD). The degree of competition for binding by unlabeled chemokines was calculated as follows: % competition for * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
ʈ binding ϭ 1 Ϫ (cpm obtained in the presence of unlabeled ligand/cpm obtained in the absence of unlabeled ligand) ϫ 100.
Chemotaxis Assay-The migration of HEK 293 cells transfected with cDNA clones was assessed by a 48-well microchamber technique (14,19). Different concentrations of chemokines were placed in the lower wells of the chamber. The CCR5/293 cells (50 l, 0.5-1 ϫ 10 6 /ml) were loaded in the upper wells. The lower and upper wells were separated by a polycarbonate filter (polyvinylpyrrolidone-free, 10-m pore size; Poretics, CA) precoated with 20 g/ml mouse collagen type IV or 50 g/ml collagen type 1 for 2 h at 37°C. The chamber was incubated at 37°C for 5 h in humidified air with 5% CO 2 . At the end of the incubation, after removal of nonmigrating cells, the filter was fixed and stained with Diff-Quik (Biochemical Sciences, NJ). The cells migrating across the filter were counted in three high power fields under light microscopy in triplicates with all samples coded. The chemotaxis index was calculated as the number of cells migrating to chemokines/number of cells migrating to medium. The significance of the difference between test and control groups was analyzed by paired Student's t test.
Confocal Microscopy-CCR5/293 cells were pretreated for 3 h at 37°C with PMA (100 nM), RANTES, IL-8 (120 nM), or different concentrations of MCP-2. The cells were centrifuged on to slides and permeabilized using 0.15% saponin in phosphate-buffered saline. The slides were then stained with an anti-FLAG monoclonal antibody (M1, Kodak, New Heaven, CT) followed by incubation with FITC-labeled goat antimouse IgG F(abЈ) 2 fragments. Slides were examined using a Zeiss 310 Confocal Laser Scanning Microscope (Carl Zeiss). Nomarski and FITC (488 nm, green) images were prepared for each specimen, and fluorescent images were superimposed on Nomarski images.
HIV-1 Inhibition Assay-CCR5/293 cells were co-transfected using electroporation with a CD4 expression vector (fragment excised from pTB4, NIH AIDS Research and Reference Reagent Program, Bethesda, MD) and cloned into a plasmid vector, pSVZeo (Invitrogen, CA). 24 h after electroporation, the cells (CCR5/CD4/293) were plated at a density of 10 5 /well in 24-well plate for 24 h. The culture media were removed, and the cells were preincubated for 30 min with 800 ng/ml recombinant human MCP-2 or RANTES. The monocyte tropic HIV-1ADA was added at a final multiplicity of infection between 1 and 0.1. Cultures were continued for 24 h and washed, and total genomic DNA was isolated and purified (20). HIV-1 entry/reverse transcription was monitored by polymerase chain reaction (PCR) using primers specific for HIV-1gag region to detect late reverse transcription products (21). Two primers are as follows: M667 corresponds to nucleotide positions 496 -516 in the HIV-1 JR-CSF sequence; M661 (antisense) corresponds to the positions 695-672. PCR products were resolved on 2% agarose gels and photodocumented after ethidium bromide staining. No PCR products were detected in CCR5/CD4/293 cells infected by the lymphocyte tropic virus HIV-1RF nor in CCR5/293 cells without CD4 coexpression incubated with HIV-1ADA.

RESULTS AND DISCUSSION
Recombinant human MCP-2 is a potent monocyte chemoattractant and has been shown to activate at least two promiscuous CC chemokines receptors, CCR1 and CCR2B (14). The binding of 125 I-labeled MCP-2 to human monocytes was of high affinity and was efficiently competed by MCP-1, which is a ligand for CCR2B. MCP-3, which uses both CCR1 and CCR2B, also efficiently competed with MCP-2 for binding to monocytes. MIP-1␣ and RANTES, two ligands for both CCR1 and CCR5, partially competed with MCP-2 for binding to monocytes, whereas MIP-1␤, which only uses CCR5, had no effect (14). We further observed that although unlabeled MIP-1␤ did not competitively inhibit MCP-2 binding to monocytes, the binding of 125 I-MIP-1␤ was completely inhibitable by unlabeled MCP-2 (IC 50 ϭ 2.9 nM versus 1.7 nM for unlabeled MIP-1␤ itself, Fig. 1). In contrast, the binding of 125 I-MIP-1␤ to monocytes was only partially but significantly displaced by unlabeled MCP-1 at high concentrations (IC 50 ϭ 425 nM), in agreement with our earlier observations (22). There was no significant difference between the competition curves yielded at room temperature (Fig. 1A) and at 4°C (Fig. 1B). The uni-directional inhibition of MIP-1␤ binding to monocytes by MCP-2 prompted us to investigate whether MCP-2, in addition to using CCR1 and CCR2B as functional receptors, also uses CCR5. HEK293 cells stably expressing CCR5 (CCR5/293) showed a high level of specific
Further support for the ability of MCP-2 to interact with CCR5 was obtained with binding studies using 125 I-MCP-2. As shown in Fig. 3, 125 I-MCP-2 specifically bound to CCR5/293 cells with an estimated K d in the 5 nM range (Fig. 3A). The binding of 125 I-MCP-2 to CCR5/293 cells was efficiently inhibited by unlabeled MIP-1␤ and RANTES (Fig. 3B) but less effectively by MCP-1 and MCP-3, which in contrast were able to completely displace MCP-2 binding to CCR1 and CCR2B (14). Thus, MCP-2 possesses binding domains for interaction with CCR5 in addition to CCR1 and CCR2B. This was supported by the observations that although MCP-1, MCP-2 and MCP-3 are all functional ligands for CCR2B, the binding of MCP-1 on CCR2B was poorly competed for by MCP-2 or MCP-3 (14, 23) suggesting differential utilization of certain binding domains on a receptor by multiple ligands. Because promiscuity is a common feature of chemokines and their receptors (24,25) and receptor activation appears to be dependent on the relative affinity to interact with a ligand, structure analyses and mutagenesis studies are required to more precisely determine the functional epitopes on both ligands and receptors.
We next determined the functional role of MCP-2 on CCR5. Both rhMCP-2 and rhMIP-1␤ are poor Ca 2ϩ mobilizers in monocytes (18,26), and we were not able to observe significant Ca 2ϩ flux in CCR5/293 cells with these two ligands, although a RANTES (120 nM)-induced signal was obtained (data not shown). We therefore utilized chemotaxis assays, which in our previous studies were demonstrated to be very sensitive and reproducible in assessing the activity of a chemokine on a given receptor (14,19). CCR5/293 cells showed a significant chemotactic response to MCP-2 and MIP-1␤ (Fig. 4), yielding a typical bell-shaped dose response. MCP-2 was similar in potency to MIP-1␤ with an almost 6-fold increase over medium control at 1.2-6 nM concentration range, but it was less efficacious (EC 50 : MCP-2, 0.12 nM; MIP-1␤, 0.04 nM, respectively) in inducing CCR5/HEK293 cell migration. Other chemokines known to activate CCR5, such as MIP-1␣ and RANTES also bound to and induced considerable chemotactic migration of CCR5/293 cells. Although MCP-1 partially displaced MCP-2 binding to CCR5/ 293 cells, it only bound to and induced migration of CCR2B/293 but not CCR5/293 cells (data not shown). Likewise, MCP-3 also partially displaced MCP-2 binding to CCR5/293 cells; we did not detect significant binding of 125 I-MCP-3 to the same cells, nor did it induce migration of CCR5/293 cells at a wide range of concentrations (data not shown). Further proof of the activation of CCR5 by MCP-2 is provided by confocal microscopy (Fig.  5) in which MCP-2 markedly and dose-dependently induced CCR5 internalization in CCR5/293 cells. Another CCR5 ligand RANTES and PMA, a potent protein kinase C activator, similarly caused internalization of CCR5. In contrast, CXC chemokine IL-8, which uses CXCR1 and CXCR2 as functional receptors, did not induce CCR5 internalization (Fig. 5).
The identification of CCR5 as the principal co-receptor for cellular entry by monocyte tropic HIV-1 strains (2)(3)(4)(5)8)  MCP-2 Activates CCR5 and Blocks HIV-1 Entry/Replication 4291 ability to interrupt virus entry, with an antiviral activity equivalent to RANTES. MCP-2 also appears to be a natural inhibitor of CD4/CCR5-mediated HIV-1 entry/replication in host cells in addition to MIP-1␣, MIP-1␤, and RANTES (6,7,9). MCP-2 is constitutively expressed in tumor cells and is inducible by pro-inflammatory cytokines in mononuclear cells and fibroblasts (10,11). MCP-2 exhibits a broader spectrum of targeted cells, including cells of dendritic phenotype (27). Therefore, MCP-2 may play an important role in recruiting/ activating immune cells at inflammatory and neoplastic foci. In the present study, we demonstrated for the first time using cloned CCR5 that MCP-2 is a highly efficacious ligand for this receptor in addition to CCR1 and CCR2B (14). In our preliminary study, HEK293 cells transfected with CCR3, which is a receptor for CC chemokines eotaxin and MCP-4 (28,29), also were induced to migrate significantly and reproducibly in response to MCP-2 in addition to eotaxin and MCP-4 (data not shown). Thus, MCP-2 appears to use CCR1, CCR2B, CCR5, and possibly CCR3, suggesting it has a more promiscuous functional pattern than other known CC chemokines.
Despite an apparent redundancy in chemokines and their receptor family (24,25,30,31), CC chemokines have been implicated as important mediators of many pathological conditions such as chronic inflammation, immune diseases, neoplasia, and atherosclerosis (30,31). CCR5 ligands MIP-1␣, MIP-1␤, and RANTES are also major HIV-1 inhibitors produced by activated mononuclear cells (9). Our current observations extend the functional scope of MCP-2 as a potent inhibitor of monocyte tropic HIV-1 infection in CD4 ϩ /CCR5 ϩ cells. Investigation into the shared and unique functional domains on MCP-2 in comparison with other chemokine ligands will be important in the development of therapeutic approaches to chemokine-and chemokine receptor-mediated pathological states.