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

J. Biol. Chem., Vol. 275, Issue 33, 25562-25571, August 18, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/33/25562    most recent M000692200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mirzadegan, T. Articles by Jarnagin, K. Search for Related Content PubMed PubMed Citation Articles by Mirzadegan, T. Articles by Jarnagin, K. Social Bookmarking                      What's this?

Identification of the Binding Site for a Novel Class of CCR2b Chemokine Receptor Antagonists

BINDING TO A COMMON CHEMOKINE RECEPTOR MOTIF WITHIN THE HELICAL BUNDLE^*

Tara Mirzadegan, Frank Diehl, Bettina Ebi, Sunil Bhakta, Irene Polsky, Deborah McCarley, Mary Mulkins, Gabe S. Weatherhead, Jean-Marc Lapierre, John Dankwardt, David Morgans Jr., Robert Wilhelm, and Kurt Jarnagin^§

From Roche Bioscience, Palo Alto, California 94304

Received for publication, January 31, 2000, and in revised form, April 17, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Monocyte chemoattracant-1 (MCP-1) stimulates leukocyte chemotaxis to inflammatory sites, such as rheumatoid arthritis, atherosclerosis, and asthma, by use of the MCP-1 receptor, CCR2, a member of the G-protein-coupled seven-transmembrane receptor superfamily. These studies identified a family of antagonists, spiropiperidines. One of the more potent compounds blocks MCP-1 binding to CCR2 with a K_d of 60 nM, but it is unable to block binding to CXCR1, CCR1, or CCR3. These compounds were effective inhibitors of chemotaxis toward MCP-1 but were very poor inhibitors of CCR1-mediated chemotaxis. The compounds are effective blockers of MCP-1-driven inhibition of adenylate cyclase and MCP-1- and MCP-3-driven cytosolic calcium influx; the compounds are not agonists for these pathways. We showed that glutamate 291 (Glu²⁹¹) of CCR2 is a critical residue for high affinity binding and that this residue contributes little to MCP-1 binding to CCR2. The basic nitrogen present in the spiropiperidine compounds may be the interaction partner for Glu²⁹¹, because the basicity of this nitrogen was essential for affinity; furthermore, a different class of antagonists, a class that does not have a basic nitrogen (2-carboxypyrroles), were not affected by mutations of Glu²⁹¹. In addition to the CCR2 receptor, spiropiperidine compounds have affinity for several biogenic amine receptors. Receptor models indicate that the acidic residue, Glu²⁹¹, from transmembrane-7 of CCR2 is in a position similar to the acidic residue contributed from transmembrane-3 of biogenic amine receptors, which may account for the shared affinity of spiropiperidines for these two receptor classes. The models suggest that the acid-base pair, Glu²⁹¹ to piperidine nitrogen, anchors the spiropiperidine compound within the transmembrane ovoid bundle. This binding site may overlap with the space required by MCP-1 during binding and signaling; thus the small molecule ligands act as antagonists. An acidic residue in transmembrane region 7 is found in most chemokine receptors and is rare in other serpentine receptors. The model of the binding site may suggest ways to make new small molecule chemokine receptor antagonists, and it may rationalize the design of more potent and selective antagonists.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Chemokines are a large family of small proteins that mediate attraction of leukocytes to inflammatory sites (1-3). The chemokine family shares a common pattern of disulfide bonds and a common overall tertiary structure as shown in solution or crystallographically determined structures (4-6). The chemokine family is divided into four subfamilies based on the number of residues between the first and second cysteine. Among the chemokines, the CC chemokine monocyte chemoattracant-1 (MCP-1)¹ has received a great deal of attention because of its involvement in diseases. MCP-1 expression is elevated in the inflamed synovium of rheumatoid arthritis, and its expression is reduced by anti-arthritic drugs (7, 8). Other work has shown that MCP-1 is elevated in asthmatic patients; the amount of elevation correlates with symptoms. MCP-1 expression is suppressed by successful immunotheraphy (9-12). MCP-1 is elevated in human atherosclerosis and is elevated by consumption of high fat diets in monkeys and rabbits (13, 14). Treatment with MCP-1 neutralizing antibodies or other biological antagonists can reduce inflammation in a number of animal models; these include lung granuloma (15, 16), lipopolysaccharide-induced death (17), glomerulonephritis (17), delay type hypersensitivity in the skin (18), and adjuvant arthritis in MRL mice (19). Transgenic over expression of MCP-1 induces monocyte and T-cell migration to the site of expression in skin (20), lung (21), brain (22), and pancreas (23). Mice genetically deleted for the MCP-1 or its receptor CCR2 are protected from inflammation and atherosclerosis driven by bacterial products and high fat diets (24-26). These findings suggest that MCP-1 functions in vivo as a monocyte and T-cell attractant and that MCP-1 is present and elevated during disease. These findings also demonstrate that modulation of MCP-1 expression or activity will be beneficial in treating inflammatory diseases.

Because of the involvement of MCP-1 in pathophysiological inflammatory diseases, we initiated a program to discover small molecule antagonists of the MCP-1 receptor. These efforts lead to the identification of several classes of compounds that inhibited MCP-1 binding to its receptor with affinities between 1 and 15 µM; this report details the properties of one class of inhibitors. We examined several of these compounds in detail with regard to binding affinity and antagonist function at several receptor sites. One class, the spiropiperidines, has affinities that range from 60 nM to inactive and has distinct structure activity relationships; we wished to understand the mechanism of action for these compounds and model their interaction with CCR2 receptor. Using several types of function studies, we examined the effect of the compounds on several post-receptor signaling pathways simulated by MCP-1 and another CCR2 ligand, MCP-3. We also examined the effect of certain site-directed receptor mutations on the binding affinity of MCP-1 and several compounds. These studies revealed an unappreciated similarity between binding of compounds to biogenic-amine receptors and their binding to MCP-1 receptor. We have built models of the CCR2 receptor incorporating the mutagenesis data, the spiropiperidine structure activity relationships, and ligand binding and signaling models that we have previously presented (27, 28). These observations provide a model to understand the interaction of the small molecule chemokine antagonists with chemokine receptors and provide a path to the design of better antagonists.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Compound Synthesis-- Spiropiperidines were prepared by adding the appropriate ortho-lithiated tert-butoxycarbonyl-protected aniline (prepared from the tert-butoxycarbonyl-aniline and 2.2 eq of tert-butyl lithium in tetrahydrofuran at 78 °C) to the N-tert-butoxycarbonyl-4-piperidone followed by a spontaneous cyclization under (29). Yields averaged 60% after chromatography. Subsequent deprotection of the tert-butoxycarbonyl-protected spirobenzoxazinones using standard procedure (20 eq of trifluoroacetic acid in CH₂Cl₂) followed by alkylation with the appropriate alkyl bromide yielded the desired compounds (5 eq of Hunig's base and 1.2 eq of the bromide in acetonitrile at reflux, 5 h. The compounds were purified by column chromatography or by preparative thin layer chromatography using 5:4:1 hexanes/CH₂Cl₂/MeOH mixtures. Each sample was >98% pure.

For preparation of 3-chlorobenzyl 2-pyrrole carboxylic acid, 2-trifluoroacetylpyrrole (25 g, 0.12 mol) was combined with sodium hydride (6.2 g, 015 mol) in dimethylformamide (300 ml) at 0 °C. 3-Chlorobenzylbromide was added, and the resulting mixture was stirred for about 15 h at room temperature. The reaction was quenched with water followed by dilute sodium hydroxide (3 M, 80 ml) and stirred for 2 h. Additional sodium hydroxide was added, and the water layer was washed with ether. Addition of 6 M HCl to the aqueous phase until the pH was 2 resulted in the precipitation of the desired carboxylic acid (RS-136270). The compound was filtered and washed with cold water. The crude acid was dissolved in CH₂Cl₂, and decolorizing charcoal was added. The mixture was stirred on a steam bath and concentrated to 800 ml. The solution was filtered and placed in a refrigerator for 10 days. Filtration and washing with cold ether then afforded 18 g of the pure carboxylic acid (>95%). All compounds were characterized by NMR spectroscopy (¹H and ¹³C), IR spectroscopy, mass spectroscopy, melting point, and elemental analysis; the data were consistent with the desired structure.

Binding Assay-- A detailed description of the binding assay is described in a previous manuscript (30). Briefly, binding was measured using membranes prepared from two cell lines, THP-1 and CCR2-CHL cells. Each competition assay (Table I) was composed of cell membranes, 50 pM ¹²⁵I-MCP, MCP buffer, protease inhibitors, and test compound. Equilibrium was achieved by incubation at 28 °C for 90 min. Membrane-bound ¹²⁵I-MCP was collected by filtration through GF/B filters presoaked in polyethyleneimine and bovine serum albumin, followed by four rapid washes with approximately 0.5 ml of ice-cold buffer containing 0.5 M NaCl and 10 mM HEPES, pH 7.4. MCP buffer consists of 50 mM HEPES, pH 7.2, 1 mM CaCl₂, 5 mM MgCl₂, and 0.1% bovine serum albumin. Protease inhibitors include 0.1 mM phenylmethylsulfonyl fluoride, 1 µM leupeptin, and 0.35 mg/ml pepstatin. THP-1 cells are a human monocyte cell line (ATCC TIP-202) that express both CCR1 and CCR2. CCR2-CHL cells are Chinese hamster lung cells (ATCC CRL-1657) that have been stably transformed with an expression vector bearing the human CCR2b receptor (30).

The other binding assays reported in Table I were performed as described above with appropriate changes in radioligand, buffer, temperature, and cell or tissue source. Thus, CCR1 and CXCR1 binding assays used 50 pM ¹²⁵I-MIP1a or ¹²⁵I-interleukin-8, a buffer composed of 50 mM HEPES, 5 mM MgCl₂, 1 mM CaCl₂, 0.1% bovine serum albumin and a pH of 7.4, a temperature of 28 °C and 100,000 CHO-CCR1 cells or CHO-CXCR1 cells. These are transformed CHO lines containing the pSW104 vector containing the human CCR1 cDNA or the human CXCR1 cDNA. The _1a or _1b binding assay used 300 pM [³H]prazosin a buffer composed of 50 mM Tris-Cl, 0.5 mM EDTA, pH 7.4, a temperature of 28 °C, and 50,000 CHO-_1a or CHO-_1d membranes from transformed CHO cell lines containing the pSW104 vector containing the human _1a or _1d adrenergic receptor. The 5HT_1a binding assay used 300 pM [³H]8-hydroxyaminopropylaminotetralin, a buffer composed of 50 mM Tris-Cl, 0.5 mM EDTA, pH 7.4, a temperature of 32 °C, and rat cortex membranes containing 0.25 mg/ml protein. All the binding buffers also contained 100 µM phenylmethylsulfonyl fluoride and 1 µM leupeptin. The biogenic amine binding assays used 0.1 M NaCl to wash the glass fiber filters, and the chemokine assay filters were washed as for the CCR2 assay. The cell lines bearing biogenic amine receptors were a kind gift from David Chang and Rick Salazar (Roche Bioscience).

Freeze-thaw lysed L1.2 cell clones bearing the wild type receptor and mutant receptors were used for saturation binding experiments with MCP-1 (Table I) and for competition experiments with MCP-1 and various compounds (Table II). Freeze-thaw lysed cells were prepared from washed centrifuge-packed cell pellets, which had been stored at -80 °C prior to binding analysis. These cells were thawed at room temperature and refrozen once. Between 1 × 10⁶ and 1 × 10⁵ freeze thawed cells were used for saturation binding experiments. These experiments used ¹²⁵I-MCP-1 at 440 Ci/mmol and varied the MCP-1 concentration between 3 and 0.023 nM. Data analysis for these saturation experiments was performed as described previously (31). Competition experiments using the L1.2 cell bearing mutant receptors also used freeze-thaw lysed cells; 1 × 10⁵ wild type bearing cells and 5 × 10⁵ mutation bearing cells were used for the competition experiments. These experiments used ¹²⁵I-MCP-1 at 440 Ci/mmol and MCP-1 concentrations of 50 pM for wild type bearing cells and 100 pM for cells bearing mutations. Data analysis of these type of competition was performed as described by Jarnagin et al. (31). The buffers used for these experiments as well as the equilibration times and temperatures were identical to those used with THP-1 cells.

Intracellular Signaling and Chemotaxis-- The methods for measurement of intracellular signaling and chemotaxis have been described in detail in our previous paper (32). Briefly, cytosolic calcium influx was measured in CCR2-CHL cells loaded with the fluorescent dye Fura-2-AM. Quantitation of signal intensity used the integrated signal intensity for 82 s after the addition of chemokine and thus has units of M·s. Antagonism by various compounds of calcium influx was measured using an approximate ED₅₀ dose of MCP-1 (3 nM) and an approximate ED₂₅ dose for MCP-3 (5 nM).

Chemotaxis was measured over 1 h using THP-1-5X cells in a 96-well Boyden chamber apparatus. Cell migration through the polycarbonate filter was quantified by fluorescent staining using propidium iodide in 0.1% Triton X-100. These assays typically gave stimulated to unstimulated migration of 6-fold, range 4-10-fold, using a maximally effective concentration of MCP-1. Chemotaxis antagonist measurements used 3 nM MCP-1 or RANTES; these concentrations are near the ED₉₅ attractant concentration for MCP-1 and for RANTES as agonists. The data are expressed by normalization to the uninhibited migration caused by the agonist chemokine. The antagonist was present in both chambers of the Boyden apparatus.

Adenylate cyclase inhibition was measured using a reporter linked gene transcription assay; in which firefly luciferase was attached 3' to a cyclic-AMP response enhancer, CREB; these elements were contained within the cell clone, CHO-K1-CCR2b-cAMP-Luc-neo-22. The assay measured adenylate cyclase inhibition after 6 h of exposure to the appropriate compounds in the presence of 2.0 µM forskolin and 5 nM MCP-1. This assay reflects biochemical measurements of cAMP production with a correlation coefficient of 0.96 (32).

Preparation, Isolation, and Characterization of Mutant Receptor Encoding cDNA and Verification of Transfectant Receptor Sequence-- Mutations were introduced into a pcDNA3.1 vector containing an amino-terminal FLAG-tagged hCCR2 receptor (33) using a polymerase chain reaction method similar to that of Jarnagin et al. (31). This method amplifies the region of receptor sequence immediately surrounding the target mutation site; thus it avoids amplification of the whole receptor. The region of the plasmid DNA amplified during the mutagenesis procedure was completely sequenced to verify the presence of the mutation and confirm the absence of unintended mutations.

Clones of L1.2 cells bearing a single mutant receptor were further checked after cloning to confirm that each isolated cell line contained DNA encoding the desired mutation. Genomic DNA from the clone was isolated using a QIAamp blood DNA isolation kit (Qiagen GmbH, Hilden Germany). The genomic DNA was used as the template in polymerase chain reaction reactions using primers encoded by the T7 region of the pCDNA3.1 vector and a 3' primer within pCDNA3.1 vector, 5'-TAGAAGGCACAGTCGAGG. The polymerase chain reaction product was further purified using Geneclean columns (Bio101, La Jolla, CA) and then sequenced. The mutations in the third extracellular loop and at the top of the seventh transmembrane domain were sequenced using a primer that included the receptor stop codon, 5'-CTA CGC GTC GAC TTA TAA ACC AGC CGA G. These sequences confirmed that all of the clonal lines reported contained the desired mutant receptor.

Preparation and Isolation of L1.2 Transfectants-- L1.2 cells were a gift from E. Butcher (Stanford University, CA). Untransfected L1.2 cells were grown at 37 °C in humidified 5% CO₂ and in RPMI 1640 medium (Life Technologies, Inc.) containing 10% (v/v) heat-inactivated fetal calf serum (Hyclone, Logan UT), 2 mM minimal essential medium sodium-pyruvate (Life Technologies, Inc.), 55 µM -mercaptoethanol, 50 units/ml penicillin G, and 50 µg/ml streptomycin (Life Technologies, Inc.). Transfected stable L1.2 cells were grown under the same conditions with additional 400 µg/ml Geneticin® (G-418, Life Technologies, Inc). L1.2 cells were transfected for 4 h at 37 °C using 30 µg of LipofectAMINE and 5 µg of plasmid DNA/2 × 10⁶ cells in Opti-MEM serum-free medium. The cells were allowed to recover for 24 h in RPMI medium containing 15% fetal calf serum. After a washing, the cells were cultured for 6 days in RPMI medium containing 1000 µg/ml G-418. Cells were maintained for two passages in 400 µg/ml G-418 prior to fluorescence-activated cell sorter-based cell cloning.

L1.2 cells express chemokine receptors more efficiently following incubation of the cells for 5.5 h with 2 mM sodium butyrate in RPMI medium at a density of 1 × 10⁵ cells/ml. The treatment was performed prior to all receptor quantitation, affinity determinations, and cell cloning steps. This treatment increased the cell surface receptor expression by 6.9-fold as measured by epitope tag-specific fluorescent antibody staining; the treatment increased MCP-1 binding to total cell membranes by 4.1-fold, from 5100 receptors/cell to 21,000 receptors/cell.

Immunostaining of Cells-- Before every study the expression of the MCP-1 receptor was monitored by flow cytometry. For the expression analysis, an aliquot of the induced cells (5 × 10⁵) was centrifuged for 5 min at 200 × g. The supernatant was removed, and the pellet was resuspended in 250 µl of RPMI 1640 medium with 10% fetal calf serum and 1 µg of Anti-FLAG® M1 monoclonal antibody (Eastman Kodak Co.). The cells and the primary antibody were incubated for 1 h at room temperature with shaking. The cells were than centrifuged for 5 min at 500 × g. After the supernatant was removed the pellet was resuspended in 1 ml of the RPMI 1640 medium and washed. The pellet was then resuspended in 250 µl of RPMI 1640 (+10% fetal calf serum) and 5.6 µg of fluorescein isothiocyanate-conjugated goat F(ab')₂, anti-mouse IgG was added (CALTAG Laboratories, Burlingame, CA). The cell antibody mixture was incubated in the dark for 30 min with shaking. After washing once, 5 µg/ml propidium iodide was added. To reduce clogging of the cytometer nozzle, the cell suspension was passed through a 70-µm nylon cell strainer (Becton Dickinson, San Jose, CA). During individual live cell cloning steps, cells with high fluorescence were put in single wells of 96-well tissue culture plates containing complete growth medium and conditioned growth medium in a ratio of 5:1. Conditioned medium is L1.2 cell culture medium in which tissue L1.2 cells had previously been grown; the medium was prepared by filtration through a sterile 0.2-µm filter after removal of cells.

Statistical Analysis of Data-- The data were analyzed for statistical significance by the use of pK_d or pIC₅₀; significance determined using analysis of variance is explained in the footnotes to the tables. The use of pK_d and pIC₅₀ is justified by our examination of the distribution of K_d and IC₅₀ values determined for the experiments of these types (32); the data are log normally distributed, as expected for the ratio definition of K_d.

Molecular Modeling-- The primary sequence of more than 100 G-protien-coupled receptors including CCR2 were aligned using the Needleman-Wunsch multi-sequence alignment as implemented in the PILEUP command in GCG (receptors that bind peptides and other lignads were included). The sequences were further aligned manually with special attention to transmembrane regions. The recently published bacteriorhodopsin structure and low resolution structures of rhodopsin was used as a template for the modeling of human CCR2 receptor (27, 28, 34). The methods used for building these first models are described elsewhere (35). Briefly, the appropriate residues of the helices of the rhodopsin template were changed to the corresponding amino acid of the CCR2. The amino acid side chains were energy minimized and placed in a reasonable conformation. For this study the loops were discarded, and only the transmembrane bundle was used in the small molecule antagonist binding site analysis. The final structure was minimized with a limited cycle using the Discover software (Biosym) and CVFF force field. Only a few cycles are used to improve the stereochemistry of the model and to remove the unfavorable clashes.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Identification and Modification of Small Molecule MCP-1 Receptor Antagonists-- Because of the involvement of MCP-1 in several inflammatory pathologies, we began a search to find small molecule antagonists of the MCP-1 receptor. A large chemical library was screened using a high throughput radioligand binding and a secondary chemotaxis inhibition assay. This screen identified about 10 structurally diverse classes of small molecule inhibitor with affinities between 5 and 30 µM for receptor binding and chemotaxis inhibition.

Inhibitor Structure Activity Relationships-- Two of these classes were judged to be drug-like and amenable to medicinal chemistry optimization. The spiropiperidine (SP) class, represented by RS-21825, RS-29634, RS-102895, and RS-504393 (Fig. 1), had a distinct structure activity relationship and affinities that spanned IC₅₀ values from 90 nM to completely inactive compounds. The central features of this class includes a pharmacophore defined by a basic nitrogen in a piperidine ring; another important feature is the orthogonal relationship between a phenyl urethane system and a piperidine, imposed by a spiro-carbon atom. The hydrogen bonding potential provided by a urethane moiety is very important. Another important feature is the restriction to small moieties for substituents on the phenyl-urethane heterocycle. Finally, a large range of hydrophobicities and sizes are acceptable for substituents in the phenethyl portion of the inhibitors, RS-504393 for example.

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Structure and affinities of several small molecule inhibitors of chemokine binding; their affinities for biogenic amine receptors and their potencies as chemotaxis inhibitors driven by MCP-1 or RANTES. MCP-1 binding was measured in THP-1 cells using 125I-MCP-1. MIP1a binding was measured in a transfected cell line CHO-CCR1 using 125I-MIP1a. The biogenic amine binding receptors were measured using transfected cell lines bearing the human 1a and 1d receptors or a tissue preparation for the 5HT1a receptor, rat brain cortex. Characteristic radioligands appropriate for each receptor were used: prazosin for 1a and 1d receptors or 8-hydroxyaminopropylaminotetralin for 5HT1a receptors. Chemotaxis inhibition was measured using THP-1-5X strain of THP-1 cells and 3 nM MCP-1 or RANTES as the chemoattractant. All measurements are reported in nM, and the error statistic is the standard deviation for the number of replicate experiments shown in parentheses. ND, not determined.

The 2-carboxy-pyrrole (2CP) class, represented by RS-136270 in Fig. 1, had a less distinct and weak structure activity relationship. The best compounds had affinities near 5 µM. Two-carboxypyrroles and 2-carboxyindoles were acceptable. Substituents at position three or four (pyrroles) were acceptable if small; however, large substituents of any kind were not well tolerated. Benzyl substitution of the nitrogen was required for activity, and halogen substituents on the ring were well tolerated. Large benzyl substituents were not well tolerated. The requirement for a carboxylic acid at position 2 was absolute.

Inhibition of Chemotaxis by SP and 2CP Class-- All of the compounds shown in Fig. 1 inhibited chemotaxis driven by MCP-1. The chemotaxis inhibitor concentrations of the spiropiperidine class are well correlated with the binding affinities and thus with the binding-derived structure activity relationship. Similar close correlation between binding affinities, structure and chemotaxis were observed for the carboxypyrrole and carboxyindole class (data not shown). Fig. 2A shows representative dose response relationships for chemotaxis inhibition. These curves are single phase curves with Hill slopes close to 1, suggesting that the compounds interact with a single population of receptors. Figs. 1 and 2A also illustrate that although these compounds are potent inhibitors of MCP-1-driven chemotaxis mediated by CCR2, none of the compounds are potent inhibitors of RANTES-driven chemotaxis mediated by CCR1. The best spiropiperidine, RS-505393, was 8000-fold better as a MCP-1-driven chemotaxis inhibitor than as a RANTES-driven chemotaxis inhibitor (Fig. 1). The highest affinity CCR2 binder, RS-504393, was a 700-fold better binder on CCR2 than CCR1. Neither class of compound, SP or 2CP, displayed any affinity for CXCR1 or CCR3 (not shown). Thus, both classes of compounds are highly selective among chemokine receptors.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   The spiropiperidine class and the carboxypyrrole class of inhibitors block MCP-1-driven chemotaxis and MCP-1-driven adenylate cyclase suppression. A, inhibition of chemotaxis of THP-1 cells to the attractants MCP-1 (3 nM) or RANTES (3 nM) by several spiropiperidine drugs. Each data point is the mean of quadruplicate data points; the standard deviations for the two curves shown are representative of all the curves. The data shown are representative of experiments repeated on from 2-6 separate days. B, the antagonistic effect of spiropiperidine and carboxypyrrole compounds on MCP-1 caused inhibition of cAMP-dependent luciferase expression. The left-most curve shows the inhibition of luciferase expression (cAMP-dependent reporter) caused by MCP-1. The three right-most curves use RS-102895, RS-136270, and RS-29634 as antagonists of MCP-1 (5 nM). In both the antagonism experiments and the MCP-1 control, 2.0 µM forskolin was used to stimulate cAMP production. The data are presented as a percentage of the luciferase expression caused by 2.0 µM forskolin (100%) and the background luciferase expression in untreated cells (0%). The data points shown are the means of triplicate determinations. The standard deviations are about 10% and are omitted for clarity. The curves shown are representative of dose titrations repeated from two to four times.

Inhibition of MCP-1-stimulated cAMP Inhibition-- Fig. 2B illustrates that both SP compounds and 2CP compounds inhibit MCP-1-stimulated cAMP inhibition, measured here using a pentameric cAMP responsive promoter element, CREB binding element, linked to a reporter cDNA, luciferase. Previous studies (30, 32) have demonstrated that this reporter assay accurately reflects cAMP changes in these cells. The rank order of potencies observed for these two SP compounds, RS-29634 and RS-102895, and one 2CP compound are very similar to their rank order receptor binding and chemotaxis inhibition potency.

Inhibition of MCP-1- and MCP-3-stimulated Calcium Influx-- Fig. 3 demonstrates that both SP compounds and 2CP compounds inhibit MCP-1- and MCP-3-stimulated calcium influx into CCR2-CHL cells. The IC₅₀ values for RS-504393, RS102895, and RS-136270 as MCP-1 inhibitors are, respectively, 35 ± 9, 32 ± 7, and 260 ± 45 nM, and as MCP-3 inhibitors they are 160 ± 26, 130 ± 27, and 500 ± 170 nM. We note that as antagonists of calcium flux, RS-504393 and RS-102895 do not show the 3-fold separation in potency that is exhibited by these two compounds in CCR2 binding assays and chemotaxis antagonist assays. This finding may reflect an equal ability of these two inhibitors to block formation of a conformation of the receptor that is able to stimulate calcium influx. Thus both classes of compounds inhibit calcium influx caused by either MCP-1 or MCP-3. None of these compounds were able to stimulate calcium influx at up to 50 µM for RS-504393 and RS-102895 or 20 µM for RS-136270 (data not shown). They are not agonists for calcium influx, adenylate cyclase inhibition, or chemotaxis. These small molecular weight ligands block all post receptor events and do not distinguish between calcium stimulatory receptor conformations caused by either CCR2 agonist, MCP-1, or MCP-3.

View larger version (28K): [in this window] [in a new window]   Fig. 3.   The spiropiperidine class and the carboxypyrrole class of inhibitors block cytosolic calcium influx caused by MCP-1 or MCP-3. A, inhibition of MCP-1 (3 nM) stimulated calcium influx by two spiropiperidines, RS-102895 and RS-504393, and by a carboxypyrrole, RS-136270. B, inhibition of MCP-3 (5 nM) stimulated calcium influx by RS-102895, RS-504393, and RS-136270. The influx was measured in the transformed cell line CCR2-CHL cells. The antagonist data (right axis) in both panels are plotted as the percentages calcium influx caused by MCP-1 (3 nM) or MCP-3 (5 nM). The data shown are single determinations and are representative of dose titrations repeated two or three times. In both panels the left-most curve shows the integrated calcium influx (left axis) dose response for the chemokine (MCP-1 in A and MCP-3 in B).

Biogenic Amine Binding Properties of the SP and 2CP Class-- Despite the high chemokine receptor specificity exhibited by the spiropiperidine compounds, these compounds are potent ₁-adrenergic receptor blockers (Fig. 1); lower affinities were observed for other types of adrenergic receptors, ₂ and  (not shown). In addition several other GPC-7TM receptors, the serotonin 5HT_1a and µ-opioid receptors for example, also bound SP compounds. In the case of the serotonergic receptors, structures with improved CCR2 affinity had decreased 5HT_1a (Fig. 1). A remarkable property of the SP class is its high affinity for -adrenergic receptors, particularly _1a receptors and to a lesser extent _1d receptors. Affinity at _1b receptors was much lower (not shown). Affinity at _1d sites became less because the compounds had improved affinity on CCR2 (Fig. 1). Affinity at _1a receptors also decreased as CCR2 affinity increased; however, because the _1a affinity of the initial lead compound, RS-21825, was 5.6 nM, the decrease in _1a affinity was insufficient to eliminate _1a affinity of the SP compounds. The 2CP class of compounds had no significant affinity for biogenic amine receptors (Fig. 1).

Selection of Residues Glu²⁹¹ and Asp²⁸⁴ for Mutagenesis Studies-- The findings of a strong requirement for a basic nitrogen in the SP class of blockers suggests that an acidic residue on CCR2 might provide part of the interaction site. The alignment of several chemokine receptors (Fig. 4) highlights an unusual feature of chemokine receptors; chemokine receptors contain an acidic glutamic acid, Glu²⁹¹, 1.5-2 turns within TM-7. Acidic residues are rare in this position in GPC-7TM receptors (36, 37). We also noted in our sequence alignment Asp²⁸⁴. This residue is located two helical turns above and on the same helical face as Glu²⁹¹ and is also conserved between chemokine receptors. Thus, we chose to make mutations of CCR2 at Glu²⁹¹ and Asp²⁸⁴. We made mutations to alanine to eliminate the residues side chain or made changes to asparagine or glutamine to remove the charge but maintain the hydrogen bonding potential of the side chain.

View larger version (73K): [in this window] [in a new window]   Fig. 4.   Alignment of 14 human chemokine receptors from the middle of transmembrane-6 through transmembrane-7. A, the transmembrane region-7 acidic residue unique to chemokine receptors is shown in bold type (residue 291 in CCR2). The residue two helical turns above the TM-7 acidic residue is also shown in bold type (residue 284 in CCR2). B, schematic illustration showing the location of all the extracellular and transmembrane localized acidic residues (black background with white letters). The probable topology and disulfide bonding pattern of CCR2 is shown, as are potential helical transmembrane and juxtamembrane regions (shaded background).

Creation and Characterization of Glu²⁹¹ and Asp²⁸⁴ Mutants-- The mutant receptors were made in an amino-terminal FLAG epitope-tagged receptor (38) and then transfected into L1.2 cells. Stable clones expressing moderate receptor densities (12-36 thousand receptors/cell) were isolated using flow cytometry. Fig. 5 shows representative cell population distributions of three important cell lines, epitope-tagged wild type receptors and E291A and E291Q mutant receptors. This figure demonstrates that the various stable mutations have similar population distributions and surface staining with anti-epitope antibody to mean levels 50-250 times the staining observed in untransfected L1.2 cells. However, we note with interest (Table I) that the only rough correlation between the surface receptor staining measured by flow cytometry and the binding assay measured receptor density, particularly for the D284A receptor mutant (Table I). The crudeness of the correlation may reflect a difference in intracellular membrane system distribution for the various mutants and the difference in sampling inherent in the two assay methods.

View larger version (35K): [in this window] [in a new window]   Fig. 5.   Transfected cell lines bearing mutant and wild type MCP-1 receptors have substantial expression of the epitope-tagged MCP-1 receptor on their surface. Shaded, number of cells (count) versus fluorescence intensity (FL-1 height) for untransfected L1.2 cells; closely spaced dots, wild type CCR2 bearing cell line; widely spaced dots, E291A bearing cell line; solid line, E291Q bearing cell line. Three thousand cells from each line were analyzed.

The use of amino-terminal FLAG epitope-tagged MCP-1 receptor does not affect the MCP-1 binding affinity of wild type receptors. The wild type receptor, naturally expressed in the human monocyte cell line THP-1, binds MCP-1 with a K_d of 0.035 ± 0.02 nM (32), whereas the epitope-tagged receptors bind MCP-1 with a K_d of 0.06 ± 0.01 (Table I and Ref. 38).

Binding Affinities of Receptor Mutants Measured by Saturation-- To fully characterize the mutant receptors, we preformed saturation binding assays for wild type receptor bearing cell lines as well as lines bearing D284N, E291Q, and D284A/E291A (Table I). For these mutant receptors affinities ranged between 0.06 nM, for wild type to 1.53 nM for D284N. The double receptor mutations, such as D284A/E291A, have higher affinities than are measurable in saturation binding experiments using available radioactive ligands.

However, four mutant receptors have substantial surface expression (Table I) but bind with an affinity of less than 3 nM, the lowest affinity detectable in these saturation binding assays. These mutants, D284K, D284A/E291A, D284N/E291Q, D284K/E291K, and DOM4A, are well expressed on the cell surface with staining of 53-177-fold over untransfected control cells (Table I). Their population distributions were all quite similar to wild type receptors and the other mutants shown in Fig. 5. These findings indicate that these mutant receptors are cell surface localized but have very low MCP-1 affinities.

Binding Affinities of Receptor Mutants Measured Using Competition-- To more fully characterize the mutants and to measure the effect of the mutations on drug binding, competition binding experiments were performed. Table II and the example graphs of Fig. 6 show that mutations at positions 284 and 291 significantly reduced MCP-1 binding affinity; however, none of these changes completely eliminated the ability of MCP-1 to bind to the receptor. The most detrimental combination of mutations, the double mutant D284N/E291Q, reduced the binding affinity by 190-fold to 11.3 nM. Other single amino acid mutants had reduced binding affinity by 4.8-8.7-fold. This observation, along with the ligand mutagenesis data showing the critical importance of certain basic residues on the MCP-1 surface (30), suggests that Asp²⁸⁴ or Glu²⁹¹ may be a site for receptor ligand interaction with some of the MCP-1 basic residues. A model of this interaction has been presented (30).

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Displacement of [125I]MCP-1 from wild type and several mutant MCP-1 receptors by MCP-1, and by the spiropiperidines RS-102895 or RS-504393 or by RS-136270, a carboxypyrrole. The data points shown are the means of triplicate determinations; the standard deviations for each point are 10-15% and are omitted for clarity. Each curve representative of dose titrations that were repeated three or four times.

Binding Affinities of Mutants for Small Molecular Weight Antagonists-- The ability of these mutants to bind MCP-1 allowed competition measurements of binding affinity of several of the small molecule antagonists. Fig. 6 and the summary data presented in Table II show that the spiropiperidine compounds, RS-102895 and RS-504393, bind extremely poorly when Glu²⁹¹ is changed to alanine (affinity reduced by 610-fold) or to glutamine (affinity reduced by 250-fold). These findings suggest that the negative charge of glutamic acid is important because glutamine has hydrogen bonding potential similar to that of glutamic acid, yet E291Q has significantly reduced binding affinity with RS-504393, 250-fold. In contrast to the spiropiperidine compounds, the 2-carboxypyrroles, represented by RS-136270, bind to all of the single acidic amino acid mutant receptors with affinities similar to wild type receptors (Fig. 6). Together these observations suggest that Glu²⁹¹ interacts with the basic nitrogen of SP molecules.

Residue Asp²⁸⁴ mutants bind SP compounds with affinities that are only very modestly changed from wild type affinities, which is in contrast to the large changes in affinity caused by mutations at residue Glu²⁹¹ (Table II). Of the single Asp²⁸⁴ mutations only the affinity of RS-504393 was significantly affected by the change to alanine; however, the magnitude of reduction in affinity of this mutant (3.9-fold) is small when compared with the magnitude of Glu²⁹¹ changes (610- and 250-fold). Thus it would seem that Asp²⁸⁴ contributes only very slightly to a SP compound binding pocket.

The 2CP compound, RS-136270 was unaffected by any of the single mutations at residues 284 or 291 (Table II). Thus these acidic residues are not required for 2CP binding, and the contrast in binding contact requirements compared with SP compounds suggests very different binding sites for these two classes of antagonists.

The double mutations to alanine and the amides, D284A/E291A and D284N/E291Q, respectively, have undetectable affinities for the SP class of compounds and are the most reduced in binding affinity for MCP-1. Thus, this combination of residues is important for both the SP and the MCP-1 binding interaction. In contrast, the double mutations to alanine and to the amides, D284A/E291A and D284N/E291Q, affect the 2CP compounds binding only modestly, 28- and 2.7-fold, respectively (Table II); these findings highlight the clear distinction between the binding mode of SP compounds and 2CP compounds.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Structure-Activity Relationship of SP Compounds-- After identification of the spiropiperidines as CCR2 ligands we were able to define several central features of the structure-activity relationship of these compounds. The spiro arrangement between a phenyl urethane moiety and a piperidine is one of the striking features. The piperidine ring was di-substituted at the 4 position with a spiro-phenyl urethane system. Ring systems other than piperidine and straight chain substitutions of the piperidine system were inactive or quite reduced in affinity. In all the more potent analogs, the nitrogen was connected via an aliphatic chain to an aromatic system. For the SP class, the aromatic system-binding pocket was tolerant of a wide range of substituents of a variety of sizes, provided that they were nonpolar. Polar substituents were not well tolerated; this implied a binding pocket with substantial hydrophobicity. This phenyl urethane system was mostly intolerant of substitution, particularly large moieties. The urethane moiety could not be replaced or altered without significant losses in activity. The strength of this structure activity relationship and its defined preference for particular sizes and polarities in different parts of the molecule implies several different steric constraints around the phenyl-urethane portion. The absolute requirement for the urethane system, with the nitrogen bearing a hydrogen, implies the presence of a hydrogen bond partner for the urethane nitrogen in the binding pocket. Possibly, the most significant feature of this structure activity relationship is the absolute requirement for a tertiary nitrogen, preferably contained within a piperidine ring. Alterations of the nitrogen pK_a lead to complete loss of binding affinity. These features suggest that the register of the compound within its binding site was controlled by an acid-base interaction.

Our starting point of the optimization of these compounds was RS-21825 (IC₅₀, 11.4 µM), and we were able to make steady improvements by increasing the affinity to RS-504393, IC50 90 nM. The discovery of small molecular weight CCR2 antagonists of several different classes was gratifying because, when we started this effort, no drug-like chemokine receptor antagonist had been revealed. In addition, we and others had suggested that the ligand was too big (molecular weight of 8,600) to allow small molecular weight compounds to inhibit its binding. This view held that the receptor ligand complex involved several thousand square angstroms of surface and was spread over a large flat interaction interface similar in nature to the surfaces defined for growth hormone and its receptor (39) or antibody-protein interactions surfaces (40). Clearly these results demonstrate the feasibility of development of MCP-1 antagonists, and they suggest that some part of the essential interaction space from MCP-1 contains a crevice or pocket with features necessary to bind small molecular weight inhibitors.

Structure-Activity Relationship of 2-CP Compounds-- In contrast to the substantial improvements in affinity we obtained in the SP inhibitor class, affinities of the 2-CP inhibitor class improved by less than 4-fold from the initial screening hit (20 µM). This class of compounds had no requirement for a basic nitrogen, nor did it possess distinct geometric requirements imposed by a spiro-center or the molecular length and size requirements displayed by the SP inhibitor class. The absolute requirement for an acidic residue further distinguished the 2CP compounds from the SP compounds. These contrasting features clearly hint that the binding sites on CCR2 for SP compounds and 2CP compounds are distinct. Given the extraordinary contrast, we speculate that the sites were spacially distinct and certainly controlled by opposite acid-base interactions.

Binding of SP and 2-CP Compounds to Other Chemokine Receptors-- These groups of compounds are uniquely able to interact with a binding pocket on CCR2 and not other chemokine receptors. Because both classes of inhibitors are extremely weak binders to other chemokine receptors, including CXCR1, CCR1, and CCR3. Small molecular weight ligands for CCR1 and CXCR4 have been described (41-43); these compounds have no affinity for CCR2, findings that we have corroborated.² None of the CCR1 or CXCR4 antagonists similar structurally to either the SP or CP classes. Thus, inspite of the sequence similarities between the chemokines (30-90%), their receptors (30-70%), and their conserved overall three-dimensional structures, small molecular weight inhibitors that distinguish the various family members can be found.

Blockade of Receptor Signaling by SP Compounds-- In addition to inhibiting MCP-1 binding and MCP-1-driven chemotaxis, the SP compounds are potent inhibitors of MCP-1-driven cAMP-mediated gene transcription as well as MCP-1- or MCP-3-driven calcium influx. Furthermore, these compounds are not agonists of MCP-1-mediated events, even at concentrations 500 times their apparent affinities. These findings differentiate small molecule antagonists from MCP-1 mutants that are chemotaxis antagonists (32). These ligand-derived antagonists were of two types; one type ( (1+9-76)hMCP) stimulates cAMP signaling through CCR2 but is not able to stimulate chemotaxis or calcium signaling. Another chemotaxis antagonistic ligand mutant (Y13A) was able to stimulate cAMP signaling and calcium signaling through CCR2 but was not able to stimulate chemotaxis. These findings suggest that ligand-derived antagonists can promote some receptor conformations that are necessary to drive signaling of post receptor pathways but are not able to promote the conformation necessary to drive chemotaxis. Unlike the ligand-derived antagonists, the small molecular weight inhibitors are not able to separate these functions; they block all measured post receptor events and are not agonists. Thus they have no ability to promote CCR2 conformations that can couple to post-receptor pathways. These findings indicate that the small molecular weight antagonists interact with the receptor quite differently than ligand-derived antagonists.

Receptor Distribution within Cells-- The point mutations we made in CCR2 were designed to investigate the role of specific residues in MCP-1 and small molecule antagonist binding; they also revealed interesting findings regarding GPC-7TM receptor intracellular sorting. As part of our efforts to characterize each mutant thoroughly, we measured receptor density using two techniques. The correlation between the two measurement methods was weak. We believe that the correlation is affected by the difference in receptor populations measured by the two techniques; the receptor binding assay is performed in crude cell membranes and thus measures binding competent receptors on the cell surface or on intracellular membranes. In contrast, the cell staining estimation of receptor density measures only surface receptors because it is performed on unpermeablized cells. These differences could account for the weak correlation between the methods. Several other workers have noted that various mutant GPC-7TM receptors do not transport well to the cell membrane (44-47). CCR2a, an alternative splice variant of CCR2b studied here, does not efficiently localize to the plasma membrane (48). These observations indicate that for D284A and possibly other receptors, some part of the receptor population is localized on intracellular membranes. These findings and the aberrant transport of several other GPC-7TM receptors to the cell surface are interesting and could indicate the existence of cellular membrane sorting systems that detects individual proteins and their conformations. Further investigation into the effect of these and similar point mutations on the distribution of GPC-7TM receptors may reveal novel details of the protein sorting process. Nevertheless, even in the most dramatic example of intracellular retention, D284A, adequate surface receptor was present to allow high quality saturation binding experiments.

Binding of SP and 2-CP Compounds to Biogenic Amine Receptors-- The first identified member of the SP family, RS21825, had been previously synthesized as an -adrenergic receptor antagonist (29) to be used as a treatment for hypertension. In addition to ₁-adrenergic receptor affinity, some members of this compound class also had significant affinity to 5HT₁, ₂-adrenergic, and opioid receptors (not shown). Thus we sought to remove the -adrenergic and biogenic amine receptor binding properties from this class while preserving CCR2 affinity. We were successful at eliminating or significantly reducing the affinities of these compounds for _1d, 5HT_1a, ₂, and opioid receptors while improving the affinities for CCR2. These results show that differences exist among the binding pockets on these G-protein coupled-seven transmembrane receptors; these differences can be exploited to convert molecules that bind to one class into molecules that bind to another class. However, the fact that these receptors were all from families that bind ligands that are basic amines was very provocative. This finding of a shared basic nitrogen motif and other results (below) formed the basis of our hypothesis of similar ligand binding modes between SP compounds and biogenic amines.

Significance of Acidic Residues Glu²⁹¹ and Asp²⁸⁴-- The results of our mutagenesis experiments establish that the SP compounds interact with Glu²⁹¹ and to a lesser extent with Asp²⁸⁴. The SP class of antagonists requires a basic nitrogen for high affinity. In addition the SP class has significant affinity for biogenic-amine receptors. These findings suggest that the way in which SP molecules bind to CCR2 may be similar to well characterized binding modes of biogenic amine receptor antagonists (49-51). The distribution of acidic amino acids in CCR2 further stimulated our interest (Fig. 4); the extracellular loops and transmembrane regions contain only 13 acidic residues. Five of the 13 residues are in an acidic cluster located in the amino terminus of the receptor. (This cluster was examined in a previous paper (30).) Parts of this cluster are important for MCP binding, but none of it is important for small molecular weight ligand binding.³ The alignment of several chemokine receptors (Fig. 4) highlights an unusual feature of chemokine receptors; chemokine receptors contain an acidic glutamic acid 1.5-2 turns within TM-7. Acidic residues are very rare in this position in GPC-7TM receptors (36, 37).

Model of SP Compounds Binding to Receptor-- Models of GPC-7TM receptors were constructed based on the low resolution electron cryomicroscopic data available for rhodopsin (28, 34, 52) and the higher resolution structures of bacteriorhodopsin (27). These models predict that Glu²⁹¹ of TM-7 is inside the ovoid-helical bundle in a mirror image position to the critical binding residue of the biogenic amine receptor, an aspartic acid from TM-3.

This binding interaction is illustrated in Fig. 7. The model depicts Glu²⁹¹ interacting with the basic nitrogen of piperidine ring and the phenylurethane moiety extending the length of the ovoid bundle to interact with TM-3 and TM-6. The phenethyl moiety extends in the opposite direction toward TM-1 and TM-2, forming a hydrophobic pocket. The binding pocket is 1.5 to 2 helical turns within the transmembrane helices; and thus, no more than one-third of the extracellular portion of the transmembrane width is occupied by the SP compound.

View larger version (86K): [in this window] [in a new window]   Fig. 7.   Model of RS-102895 bound to the MCP-1 receptor. RS-102895 is the space-filling molecule in the center of the bundle of helices, indicated by the ribbons. Receptor residue Glu291 is shown as a ball and stick structure; it is shown interacting with the basic nitrogen of the spiropiperidine structure.

How Does the Binding of a SP Compound Prevent MCP-1 Interactions with CCR2?-- Comparison of the model presented in Fig. 7 with our model of MCP-1 binding to CCR2 (27) indicates that SP compounds prevent MCP-1 binding by occupying the same space, the inter-helical bundle region on the extracellular side of the receptor. Our models of the MCP-1/CCR2 complex place the amino terminus of MCP-1 in close proximity to Asp²⁸⁴ and Glu²⁹¹, crossing through the same volume of space as would be occupied by SP compounds.

In conclusion, we have discovered and optimized a class of small organic molecule antagonists of the CCR2 receptor, spiropiperidines. These compounds are not chemotaxis agonists and do not stimulate post receptor signaling of any kind. The compounds block MCP-1 and MCP-3 signaling through CCR2; however, they are not antagonists of CXCR1, CCR1, or CCR3. These antagonists block the receptor by occupation of a binding site that includes acidic residue Glu²⁹¹. We believe that the acidic moiety interacts with the piperidine basic nitrogen. The spiropiperidine blockade of MCP binding occurs by occupation of some of the space that CCR2-bound MCP-1 occupies. These results provide an excellent starting point for the design of new experiments directed toward refining theses models and may provide insights on how to improve the affinity and specificity of small organic ligands. Most of the chemokine receptors contain acidic residues in TM-7 in an analogous location to CCR2. Many of the early lead antagonists disclosed in the patent literature contain basic sites similar to the piperidine nitrogen; thus our models may provide a general framework to explain small molecule binding to chemokine receptors.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence may be addressed. Tel.: 650-855-5814; Fax: 650-354-7554; E-mail: tara.mirzadegan@roche.com.

^§ To whom correspondence may be addressed. Present address: Iconix Pharmaceuticals Inc., 850 Maude Ave., Mountain View, CA 94043. Tel.: 650-567-5503; Fax: 650-526-3034; E-mail: kjarnagin@iconixpharm.com.

Published, JBC Papers in Press, April 17, 2000, DOI 10.1074/jbc.M000692200

² T. Mirzadegan, F. Diehl, B. Ebi, S. Bhakta, I. Polsky, D. McCarley, M. Mulkins, G. S. Weatherhead, J.-M. Lapierre, J. Dankwardt, D. Morgans, Jr., R. Wilhelm, and K. Jarnagin, unpublished data.

³ I. Polsky and K. Jarangin, unpublished data.

	ABBREVIATIONS

The abbreviations used are: MCP-1, monocyte chemoattracant-1; GPC, G-protein-coupled; TM, transmembrane region; CHO, Chinese hamster ovary; RANTES, regulated on activation normal T cell expressed; SP, spiropiperidine; 2CP, 2-carboxy-pyrrole.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES