Structure/Function Relationships of CCR8 Agonists and Antagonists

We describe here the interactions of CCR8 with its ligands using both CCR8 transfectants and a T-cell line expressing the receptor endogenously. Of the CCR8 agonists reported previously, only CCL1 and vMIP-I exhibited potency in assays of intracellular calcium flux, chemotaxis, and receptor internalization, this latter mechanism being dependent upon the expression of β-arrestins 1 and 2 but independent of Gαi signaling. NH2-terminal extension of the mature CCL1 sequence by a serine residue (Ser-CCL1) resulted in a partial agonist with a reduced affinity for CCR8, suggesting that the NH2 terminus of the ligand plays a role in ligand binding to an intrahelical site. Attempts to identify key residues within this site revealed that the conserved glutamic acid residue in transmembrane helix 7, Glu-286, is crucial for trafficking of the receptor to the cell surface, while Asp-97 of transmembrane helix 2 is dispensable. CCL7 was found to inhibit both Ser-CCL1 and vMIP-I responses but not those of CCL1 itself. Similarly, vMIP-I responses were more than 2 orders of magnitude more sensitive to the specific CCR8 antagonist MC148 than those induced by CCL1, which is difficult to reconcile with the reported affinities for the receptor. Collectively, these data suggest that the CCR8 ligands are allotropic, binding to distinct sites within CCR8 and that the human immune system may have evolved to use CCL7 as a selective antagonist of viral chemokine activity at CCR8 but not those of the host ligand.

The mammalian immune system can be organized into innate and adaptive immunity with the directed movement of leukocytes being of paramount importance to the effectiveness of both systems. This is mediated largely by the chemokine family of proteins, whose activity is mediated by seven transmembrane spanning G protein-coupled receptors expressed on the leukocyte cell surface. Inadvertent or excessive production of chemokines has been implicated in the inflammatory components of many clinically important diseases (1) and the development of specific antagonists to inhibit chemokine receptor function remains a goal of the pharmaceutical industry (2).
The chemokine receptor CCR8 is expressed on lymphocytes of the Th2 lineage (3,4) and is thought to play a role in adaptive immunity. The level of infiltrating CCR8-expressing Th2 cells has been correlated with the severity of asthmatic responses in challenged individuals (5). These data have been confirmed in animal models of airways inflammation in mice deficient in CCR8 (6), albeit controversially (7,8). CCR8 has also been shown to be involved in the migration of dendritic cells from the lymph nodes (9). More recent data suggest that CCR8 also plays a role in the innate immune response, with CCR8 negatively correlated with the effective innate immune responses during models of septic peritonitis (10).
The ligand of CCR8 is CCL1 (11,12), which was identified originally as a polypeptide secreted by activated T-cells (13). In addition, CCL16 (14) and CCL4 and CCL17 (15) have also been proposed to act as CCR8 agonists. Moreover, viruses have effectively targeted this receptor. Genes encoding polypeptide agonists and antagonists of CCR8 are present in the genomes of human herpesvirus-8 (16,17) and Molluscum contagiosum virus, respectively (18). In this paper we re-address the repertoire of chemokines that activate CCR8 and dissect the molecular signature necessary for receptor activation by different ligands.
(HA) 4 antibody was purchased from Covance Research Products (Berkley, CA), and its corresponding IgG1 isotype control antibody was from Sigma. Fluorescein isothiocyanate-conjugated goat anti-mouse antibody was purchased from Dako cytomation (Ely, UK).
Generation of HA-CCR8 and Point Mutants-The cDNA encoding human CCR8 was tagged at the 5Ј region with a sequence encoding the HA-epitope as described previously for CCR4 (19). This results in an NH 2 -terminally HA-tagged construct when expressed in the pcDNA3 vector. Point mutants of human CCR8 were generated by PCR according to the manufacturer's instructions, using the QuikChange TM site-directed mutagenesis kit (Stratagene, Amsterdam, Netherlands). Verification of mutation was performed by DNA sequencing on both strands (MWG Biotech, London, UK).
Flow Cytometry-Approximately 5 ϫ 10 5 cells were harvested, washed once with fluorescence-activated cell sorting (FACS) buffer (0.25% bovine serum albumin and 0.01% NaN 3 in HEPES-modified phosphate-buffered saline), and incubated with 50 g/ml anti-HA antibody or the corresponding IgG1 isotype control in 100 l of ice-cold FACS buffer for 20 min. Cells were then washed with FACS buffer and incubated with 50 g/ml fluorescein isothiocyanate-conjugated secondary antibody for 20 min in 50 l of ice-cold FACS buffer. After this time, cells were washed twice and resuspended in 400 l of FACS buffer before being analyzed by flow cytometry, as described previously (20). Intracellular staining was assessed with the addition of 0.1% saponin to the FACS buffer.
Chemotaxis Assay-The chemotactic responsiveness of cells was ascertained using ChemoTx TM plates (Receptor Technologies Ltd., Oxon, UK), as described previously (11). Briefly, various concentrations of chemokine in duplicate were applied in a final volume of 31 l to the lower wells of a chemotaxis chamber. 2 ϫ 10 5 cells were applied to the filter, and following a 5 h incubation (3 h for CCR8 -4DE4 cells) in a humidified chamber at 37°C in the presence of 5% CO 2 , the number of migrating cells traversing a 5-m pore filter were counted using a hemocytometer. Data are shown as the percentage of migrating cells.
Intracellular Calcium Measurements-These were performed as described previously (23) using 4DE4-CCR8 cells loaded with the fluorescent dye Fura-2/AM (Molecular Probes, Inc., Sunnyvale, CA). Cells were stimulated with the appropriate chemokine and real-time data were recovered using a fluorimeter (LS-50B; PerkinElmer Life Sciences). Data are expressed as the relative ratio of fluorescence emitted at 510 nm after sequential stimulation at 340 and 380 nm.
Internalization Assay-Chemokine receptor internalization was assessed by the loss of cell surface receptor expression, as determined by flow cytometry using an antibody directed against the HA-tag introduced into the CCR8 protein sequence. 5 ϫ 10 6 cells were resuspended, in duplicate in 100 l of ice-cold culture medium. One duplicate also received the addition of 100 nM chemokine. All samples were then incubated at 37°C for the required duration, before washing, staining, and receptor expression analysis by flow cytometry. Alternatively, for the dose-response assays, cells in duplicate were resuspended in culture medium containing various concentrations of chemokine. One of these samples was incubated for 20 min at 37°C, while the other remained on ice throughout the assay. The remaining receptor expression was calculated using the following equation: 100 ϫ [(mean fluorescence of chemokine treated cells at 37°C Ϫ mean fluorescence of negative control cells)/(mean fluorescence of chemokine treated cells at 4°C Ϫ mean fluorescence of negative control cells)], where negative control cells were buffer-treated cells stained with an IgG1 isotype control antibody.
Competitive Radioligand Binding Assay-Whole cell binding assays on 4DE4 cells expressing human CCR8 stably were performed as described previously (23), using 0.1 nM 125 I-CCL1 purchased from Amersham Biosciences (Little Chalfont, UK) and various concentrations of unlabeled chemokine competitor. A Canberra Packard Cobra 5010 ␥-counter (Canberra Packard, Pangebourne, UK) was used to assess the level of competition for binding of 125 I-CCL1 with the receptor.
Purification and Analysis of CCL1-CCL1 was purified and analyzed from conditioned medium from megakaryocytic cells as described previously (24). Briefly, following fractionation by heparin-Sepharose chromatography and Mono S cation exchange fast protein liquid chromatography, CCL1 immunoreactivity was purified to homogeneity by reverse-phase high performance liquid chromatography and eluted from the C8 column by an acetonitrile gradient, while absorbance was measured at 220 nm. The amino-terminal sequence of mature CCL1 in fraction 43 was confirmed by automated Edman degradation on a 491 Procise cLC protein sequencer (Applied Biosystems, Foster City, CA). CCL1 immunoreactivity was measured by a sandwich ELISA using monoclonal anti-human CCL1 (R&D Systems) and polyclonal anti-human CCL1 (Peprotech) as capture and detection antibodies, respectively.
GTP␥S Assays-Membrane fractions were prepared from TK Ϫ 143 cells infected with v⌬B8R-7LHAN or v⌬B8R and assayed for their ability to bind [ 35 S]GTP␥S in response to stimulation by various concentrations of chemokine, as described previously (25).
Data and Statistical Analysis-Data are expressed as the mean Ϯ S.E. of the stated number of experiments and were analyzed with the application of a relevant statistical test, where appropriate, using PRISM software (GraphPad PRISM v4.03, San Diego, CA).

CCL1 and vMIP-I Are Potent CCR8
Agonists-Initial experiments were performed using the 4DE4 cell line that expresses CCR8 stably (11). We focused upon the ligand repertoire of CCR8, examining reported human CCR8 agonists in assays of chemotaxis and intracellular calcium release. CCL1 induced migration of CCR8 -4DE4 cells with a typical "bell-shaped" profile with 1 nM CCL1 inducing optimal migration (Fig. 1A). However, CCL4, CCL16, and CCL17 were unable to induce any significant migration of these cells. In contrast, the virally encoded chemokine vMIP-I induced CCR8 -4DE4 cell migration with a very similar profile to that of CCL1, which is surprising when one considers they share only 32% amino acid identity. These chemotaxis results were also replicated in the HUT-78 cell line that expresses CCR8 constitutively (data not shown). In keeping with these data, in assays measuring the change in [Ca 2ϩ ] i following stimulation with 25 nM of ligand, only CCL1 and vMIP-I were capable of provoking Ca 2ϩ release from 4DE4-CCR8 transfectants (Fig. 1, B and C), whereas CCL4, CCL16 and CCL17 were without effect ( Fig. 1

, D-F). CCL1 and vMIP-1 Induce the Internalization of CCR8 via a ␤-Arrestin-dependent, Pertussis Toxin-independent Pathway-
Following incubation with ligand, chemokine receptors typically undergo endocytosis. Generally this is assayed by flow cytometry using receptor specific antibodies. However, the lack of CCR8-specific antibodies that work efficiently in this application has hampered progress in this area. Therefore, we used an HA epitope-tagged CCR8 construct that was expressed transiently in the L1.2 cell line. Significant cell surface CCR8 expression was observed, as ascertained by flow cytometric analysis using an anti-HA monoclonal antibody and a relevant isotype control ( Fig. 2A). Cell surface expression of HA-CCR8 was reduced following treatment of the cells with 100 nM CCL1 for 20 min at 37°C ( Fig. 2A), suggesting that like other chemokine receptors, CCR8 also undergoes ligand-induced internalization. We subsequently extended these studies to assess both the time course and the dose-responsive nature of this process by treating cells with either CCL1 or vMIP-I (Fig. 2, B and C). The process of internalization was extremely rapid, reaching a maximum within 10 min of incubation with chemokine. In keeping with the previous chemotaxis data, both ligands were equally potent in inducing CCR8 internalization. In contrast, treatment with buffer alone had no significant effect on CCR8 cell surface levels. As reported for other chemokine receptors, the process by which HA-CCR8 is internalized is independent of G␣ i protein coupling, because pretreatment of the CCR8 expressing cells with pertussis toxin had no effect on the ability of CCL1 to induce receptor endocytosis (Fig. 2D). In contrast, chemotaxis of the same pertussis toxin-treated cells in response to 1 nM CCL1 was completely ablated (data not shown).
Next we addressed the intracellular mechanism by which CCR8 undergoes endocytosis. This was undertaken by the tran-sient expression of HA-CCR8 in MEFs derived from both WT and mice deficient in ␤-arrestins 1 and 2 (21). HA-CCR8 was expressed efficiently in both cell types and its internalization in response to CCL1 was observed in WT MEFs (Fig. 3A), with ϳ60% of cell surface CCR8 levels internalized over a 20 min time period. In contrast, the internalization of HA-CCR8 was negligible in MEFs from ␤-arrestin1/2-deficient mice, suggesting a critical role for ␤-arrestins in the internalization of CCR8 in these cells. This was supported further by dose-response experiments with little internalization of CCR8 observed in ␤-arrestin1/2-deficient MEFs at concentrations of CCL1 up to 100 nM (Fig. 3B).
Amino-terminal Modification of CCL1 Produces a Partial Agonist of Human CCR8-One of the original reports of CCL1 in the literature described the purification of a 73-amino acid protein following the transfection of COS cells with the corresponding cDNA (26). Data sheets accompanying two commercial sources of CCL1 (both recombinant proteins produced in Escherichia coli) stated that they differed by a single amino acid, with one form possessing an additional serine residue at the amino terminus when compared with the published primary sequence of the mature peptide. We named this extended form Ser-CCL1. In chemotaxis assays using the 4DE4-CCR8 line   DECEMBER 1, 2006 • VOLUME 281 • NUMBER 48 (Fig. 4A) and the HUT-78 cell line (Fig. 4B), Ser-CCL1 exhibited both a 10-fold lower potency and a reduced efficacy when compared with CCL1, suggesting that, in this assay at least, Ser-CCL1 is a partial agonist of CCR8. This is reminiscent of the activity of mutant forms of CCL5, AOP-RANTES (27), where additions to the amino terminus reduced function severely, presumably by hindering access of the chemokine to an activation site within the receptor. We subsequently determined the NH 2 -terminal sequence of natural CCL1. CCL1 was purified to homogeneity from the supernatants of megakaryocytic cells (Fig. 4C) and subsequent amino-terminal sequencing revealed that the major product was the 73 amino acid form of CCL1 described by Miller and Krangel (26), which lacks a serine residue at the amino terminus.

Structure/Function Analysis of CCR8
We consequently used Ser-CCL1 as a tool to examine the structural requirements for CCR8 activation. In keeping with the chemotaxis data, displacement of 125 I-CCL1 from 4DE4-CCR8 cells by Ser-CCL1 was reduced by an order of magnitude when compared with CCL1 (Fig. 5A), suggesting that in addition to activating the receptor, the amino-terminal portion of the chemokine makes a significant contribution to receptor binding. CCL7, which was reported to compete for CCL1 binding to both CCR8 (16) and to a poxviral orthologue 7L (25), displaced 125 I-CCL1 weakly with 50% displacement only apparent with a 1000-fold excess of cold competing CCL7.
To assess whether these partial agonist properties of Ser-CCL1 were specific for the human receptor, we used cells expressing 7L, an orthologue of CCR8 encoded by Yaba-like disease virus that shares 53% amino acid identity with CCR8 (25,28,29). G protein activation following stimulation with both CCL1 and Ser-CCL1 was assessed by GTP␥S assays using membranes from both 7L and mock-transfected cells (Fig. 5B). In contrast to the data obtained with human CCR8, Ser-CCL1 was a slightly more potent ligand than CCL1, with both forms of the chemokine inducing similar levels of maximal stimulation. . Data shown are the mean (ϮS.E.) percentage of migrating cells of three separate experiments, expressed as the percentage of migrating cells following subtraction of migration to buffer alone. C shows the purification of CCL1 from the conditioned medium of megakaryocytes by reverse-phase high performance liquid chromatography with elution by an acetonitrile gradient (dashed line). Proteins were detected by their UV absorbance at 220 nm (solid line). The amino-terminal sequence of mature CCL1 in fraction 43 was confirmed by automated Edman degradation to begin with a lysine residue (inset). Thus, it appears that the molecular mechanisms underlying the activation of 7L are distinct from those governing CCR8 activation.
E286 of TM7 Is Critical for CCR8 Trafficking to the Cell Surface-Previous structure/function analysis of chemokine receptors has suggested a two-step model of receptor activation in which the chemokine is initially bound by the amino terminus of the receptor, facilitating the subsequent delivery of the amino terminus of the chemokine to an intrahelical binding site. This latter step results in signaling (30 -33). Since the additional serine residue at the amino terminus of Ser-CCL1 reduced its affinity for CCR8 by an order of magnitude (Fig. 5A), we postulated that the serine residue might inhibit access to an intrahelical pocket within the receptor and that the amino terminus of the chemokine itself contributes significantly to receptor binding. Analysis of the primary sequence of CCL1 shows that it is unusually basic, with lysine and arginine residues at positions 1 and 9 of the mature peptide. We hypothesized that these residues might form ionic interactions with acidic residues within the transmembrane (TM) helices of CCR8. Only two acidic amino acids reside within this region, namely Asp-97 of TM2 and Glu-286 of TM7, which are highly conserved among chemokine receptors and class A GPCRs in general. Therefore, we mutated these residues of the HA-CCR8 cDNA to generate alleles encoding asparagine and glutamine residues at the respective positions. Transient expression of both constructs together with HA-CCR8 was carried out in L1.2 cells as before. While the D97N-CCR8 mutant was expressed on the surface of cells at levels similar to those of HA-CCR8, the E286Q construct was absent, with staining of these cells at levels similar to that of the mock transfection (Fig.  6A). Intracellular staining of the same cells indicated that despite poor cell surface expression, the E286Q protein was produced at levels comparable with HA-CCR8 (Fig. 6B). At a functional level, cells expressing the D97N-CCR8 mutant showed identical chemotactic profiles to WT-CCR8 when migrating in response to CCL1, Ser-CCL1, and vMIP-I (Fig. 7,  A-C). Unsurprisingly, a migratory response to all three chemokines was absent in cells transfected with the E286Q-CCR8 cDNA. These data indicate that the amino acid Glu-286 in TM domain 7 is critical for efficient cell surface expression of CCR8, while Asp-97 is dispensable for both effective receptor expression and CCR8 activation.
CCL7 Antagonizes Viral but Not Human Ligands of CCR8-Because CCL7 has been described as a natural antagonist of CCR5 (34), and both this and previous studies have indicated that CCL7 can bind to both human CCR8 (16) or its viral orthologue 7L (25), it was predicted that this chemokine might also   DECEMBER 1, 2006 • VOLUME 281 • NUMBER 48

Structure/Function Analysis of CCR8
be an antagonist at CCR8. To test this hypothesis, pretreatment with CCL7 was used in an attempt to inhibit ligand-induced [Ca 2ϩ ] i release. Fig. 8, A and B, show that CCL1 responses were unperturbed by pretreatment with an equimolar amount of CCL7 and that CCL7 alone was unable to induce a detectable calcium flux. In keeping with this latter finding, neither CCR8 -4DE4 cells nor HUT-78 cells migrated toward CCL7 in chemo-taxis assays (data not shown) nor was CCL7 able to induce CCR8 internalization in L1.2 transfectants (data not shown). In contrast to their effect upon CCL1 stimulation, the Ca 2ϩ responses to vMIP-I and Ser-CCL1 were reduced markedly by CCL7 pretreatment (Fig. 8, C-F). CCL7 pretreatment had no effect on the subsequent ability of CXCL12 to induce Ca 2ϩ release, suggesting that its inhibitory effects are specific for CCR8 (Fig. 8, G and H).
Next we compared the ability of CCL7 to inhibit the migration of CCR8-expressing cells with that of MC148, a CCR8 antagonist produced by the poxvirus M. contagiosum virus (18). Fig. 9A shows that the migration of HUT-78 cells to CCL1 was inhibited only by MC148; in contrast, responses to 10 nM Ser-CCL1 and 1 nM vMIP-I were inhibited significantly by both MC148 and CCL7 with little difference between the levels of inhibition observed. Dose-response curves were generated to further examine the inhibition of chemotaxis of CCR8 -4DE4 cells by increasing concentrations of CCL7 (Fig. 9B) and MC148 (Fig. 9C). Inhibition of Ser-CCL1 and vMIP-I  induced migration was of the same order of magnitude, with CCL7 being a more potent inhibitor of migration to Ser-CCL1 than vMIP-I (respective IC 50 values of 18.5 and 97.0 nM). In keeping with the earlier Ca 2ϩ data, CCL7 was ineffective at antagonising either CCL1-or CXCL12-induced migratory responses (Fig. 9B). In contrast, vMIP-I was much more sensitive to inhibition by MC148 than was CCL1 with respective IC 50 values of 0.4 pM and 5.7 nM. Collectively, these data suggest that CCL1 and vMIP-I interact with CCR8 in distinctly different ways although they are equally proficient in the activation of the receptor.

DISCUSSION
We report here the results of investigations into the structure/function relationships of CCR8 and its ligands. While CCL1 and vMIP-1 were potent activators of CCR8 in assays of chemotaxis and [Ca 2ϩ ] i release, we were unable to demonstrate any such activity for the previously described CCR8 ligands CCL4, CCL16, and CCL17. The described ability of CCL4 and CCL17 to induce migration of Jurkat cells expressing CCR8 (15) has been previously questioned by others (35) because these cells were reported to express CCR4 and CCR5, which may explain the observed activity of both ligands (36,37). We were unable to reproduce here in the 4DE4 murine pre-B cell line and also in the HUT-78 cell line (data not shown) the migration and adhesion of CCR8 HEK transfectants in response to CCL16 observed by others (14). Our findings are in agreement with those of Nomiyama et al. (38), who were also unable to observe activity of CCL16 on murine pre-B cell CCR8 transfectants and may reflect differential signaling pathways in HEK cells compared with pre-B cell lines. CCL1 also induced the internalization of HA-CCR8 in both L1.2 and MEFs and in accordance with the chemotaxis data, with similar efficacy and potency. The process of internalization appeared to be reliant upon ␤-arrestins as deduced using MEFs from arrestin-deficient mice. This is in agreement with reports for other chemokine receptors, including CXCR1 (39), CXCR4 (40,41), CCR5 (42), and the human decoy receptor D6 (22), and suggests that it follows the well characterized clathrin mediated endocytosis. The concentrations of chemokine required to induce significant down-regulation of CCR8 were also similar to those observed to induce chemotaxis, suggesting that these processes occur concurrently and could therefore be linked. For other chemokine receptors, such as CCR5, the levels of chemokine reported to be necessary for down-regulation are substantially greater than those required to induce chemotaxis (43). This may represent subtle differences in the mechanisms of induction of receptor internalization between different chemokine receptors. One such difference highlighted here is that CCR8 was shown to internalize independently of G␣ i signaling, whereas the internalization of CCR5, for example, was reported to be sensitive to pertussis toxin, i.e. dependent on G␣ i signaling (44).
The process by which chemokine receptors are activated has been postulated to involve two distinct stages. The first involves the tethering of the chemokine to the amino terminus of the receptor. This facilitates a second interaction whereby the amino terminus of the chemokine disrupts interactions between the transmembrane helices of the receptor, inducing a signaling conformation (30 -33). In this report, our data suggest that the amino terminus of CCL1 plays a significant role in both the binding to and activation of CCR8. Ser-CCL1, an extended form of CCL1, which we were unable to find in megakaryocytederived culture medium, was noticeably less potent and efficacious in assays of ligand binding and CCR8 activation. Therefore, we postulate that this additional serine residue sterically hinders access to a site within the receptor that contributes to both ligand binding and receptor activation. We attempted to identify corresponding residues within the transmembrane bundle of CCR8 that might interact with the amino terminus of CCL1. Mutation of Asp-97 (within TM2 of the receptor) to an asparagine was without effect on receptor expression or responses to any of the CCR8 ligands identified, suggesting that this acidic side chain plays little role in receptor function. In contrast, mutation of Glu-286 to glutamine resulted in a total loss of CCR8 trafficking to the cell surface despite production of the protein. Mutation of the analogous residues in CCR1 had little effect on receptor functionality (45). This suggests that subtle differences exist at the molecular level between conformations of different chemokine receptors and that the acidic side chain of Glu-286 most likely forms a stabilizing interaction with one or more residues within CCR8. While this interaction is presumably needed for the correct folding of CCR8 and its subsequent trafficking to the cell surface, it is plausible that the residue is also involved in CCR8 activation and that this interaction is broken by engagement with CCL1 or vMIP-1, resulting in a signaling receptor phenotype.
While CCL7 has never been considered as a CCR8 agonist, which we confirm here, it was reported to possess moderate to low affinity for the receptor (16). The binding data presented here confirm this earlier finding and show that while CCL7 is able to displace 125 I-CCL1 from CCR8 poorly, it has no observable antagonist activity against CCL1-mediated responses. This suggests that CCL7 is unlikely to be a physiologically relevant natural antagonist of host CCR8 responses, in contrast to its reported activity at CCR5 (34). CCL7 was, however, able to inhibit responses to both Ser-CCL1 and vMIP-I, suggesting that at the molecular level, the manner in which the three chemokine agonists interact with CCR8 are subtly different. At first glance it is tempting to put these differences solely down to their relative affinities for CCR8. While we have not assessed here the relative affinity of vMIP-I for CCR8, two previous competition binding studies have shown relatively high affinity for CCR8 with IC 50 values for CCL1 displacement of 1.2-4.7 nM (16,17). CCL7 was a poor competitor of CCL1 from CCR8, with displacement only observed using a 1000-fold excess of competitor (Fig. 5A), which roughly correlates with its ability to block the activity of Ser-CCL1 and vMIP-I chemotaxis at similar molar ratios (Fig.  9B). However, if differing affinities for CCR8 were the sole reason for the observed ability/inability of CCL7 to inhibit vMIP-I/CCL1 responses, one would not expect the strikingly similar dose-response profiles observed in chemotaxis of CCR8 expressing cells to CCL1 and vMIP-I (Fig. 1A). In addition, while MC148 inhibited chemotaxis to 10 nM CCL1 with an IC 50 of 7.1 nM, in excellent agreement with previous findings (18), its activity against responses to an equal concentration of vMIP-I were more than two orders of magnitude higher (IC 50 ϭ 0.2 pM). This suggests that CCL1, vMIP-I, and CCL7 may be allotropic ligands, binding to distinct sites within CCR8. This was also suggested for binding of CXCL10 and CXCL11 to CXCR3 (46). In the absence of commercially available radiolabeled vMIP-I and CCL7 we are unable to directly test this hypothesis.
In summary, we describe here the relative importance of the amino terminus of CCL1 in both receptor binding and CCR8 activation and show evidence for a complicated relationship between the receptor and its known ligands. Further identification of the molecular mechanisms involved in CCR8 activation may be aided by the recent identification of a small molecule agonist for the receptor, ZK756326 (47). The observation that CCL7 is an antagonist of vMIP-I but not CCL1-induced responses suggests that the human immune system may have evolved to use CCL7 as a selective antagonist of viral chemokine activities at CCR8 while sparing host responses.