Cloning, in Vitro Expression, and Functional Characterization of a Novel Human CC Chemokine of the Monocyte Chemotactic Protein (MCP) Family (MCP-4) That Binds and Signals through the CC Chemokine Receptor 2B*

Here we describe the characterization of a novel human CC chemokine, tentatively named monocyte chemotactic protein (MCP-4). This chemokine was detected by random sequencing of expressed sequence tags in cDNA libraries. The full-length cDNA revealed an open reading frame for a 98-amino acid residue protein, and a sequence alignment with known CC chemokines showed high levels of similarity (59–62%) with MCP-1, MCP-3, and eotaxin. MCP-4 cDNA was cloned into Drosophila S2 cells, and the mature protein (residues 24–98) was purified from the conditioned medium. Recombinant MCP-4 induced a potent chemotactic response (EC50 = 2.88 ± 0.15 nm) and a transient rise in cytosolic calcium concentration in fresh human peripheral blood monocytes but not in neutrophils. Binding studies in monocytes showed that MCP-4 and MCP-3 were very potent in displacing high affinity binding of125I-MCP-1 (IC50 for MCP-4, MCP-3, and unlabeled MCP-1 of 2.1 ± 1.4, 0.85–1.6, and 0.7 ± 0.2 nm respectively), suggesting that all three chemokines interact with the CC chemokine receptor-2 (MCP-1 receptor). This was confirmed in binding studies with Chinese hamster ovary cells, stably transfected with the CC chemokine 2B receptor. Northern blot analysis in extracts of normal human tissues showed expression of mRNA for MCP-4 in small intestine, thymus, and colon, but the level of protein expression was too low to be detected in Western blot analysis. However, expression of MCP-4 protein was demonstrated by immunohistochemistry in human atherosclerotic lesion and found to be associated with endothelial cells and macrophages.

Dextran T500 was obtained from Pharmacia Biotech Inc.; Nyco-Prep TM1.068 was from Nycomed UK; PBS and RPMI 1640 were from Life Technologies, Inc.; HEPES (ultrapure) and ionomycin were from Calbiochem; Fura-2/AM was from Molecular Probes, Inc.; bovine serum albumin (fatty acid-free) and naphthyl-butyrate esterase kit were purchased from Sigma; Diff-Quik stain was from B. M. Browne Ltd.; multiple tissue blots for Northern blotting and a tissue-specific protein solution of normal human tissues for Western blotting were purchased from CLONTECH Inc. (Palo Alto, CA). Ready made human multiple tissue blots from DNA Technologies Inc. (Gaithersburg, MD) were also used for Western blotting. All other reagents were analytical grade (AnalaR).

Cloning and Expression of Recombinant Human MCP-4
MCP-4 was initially identified at Human Genome Sciences (Rockville, MD) by random sequencing of expressed sequence tags in a cDNA library from an 8-week-old human fetus, and a full-length clone was identified and sequenced subsequently.
MCP-4 was expressed in a Drosophila cell culture system (17,18). The MCP-4 coding region was amplified by polymerase chain reaction (PCR) using the cDNA clone as substrate and the following pair of primers: 5Ј-TCCCCGCGGCCACCATGAAAGTTTCTGCAGTGCTT-3Ј (SacII site and initiator methionine codon underlined) and 5Ј-GCTCTA-GACTATTAAGTCTTCAGGGTGTGAGCT-3Ј (XbaI site underlined). The PCR product was digested with SacII and XbaI, cloned into SacII/ XbaI-digested vector pMTAL, and sequenced to ensure the absence of PCR-generated mutations. This placed the open reading frame between a copper sulfate-inducible metallothionein promoter and an SV40 late polyadenylation site. This vector was introduced into Drosophila Schneider 2 (S2) cells by co-transfection with a vector carrying a hygromycin B resistance gene. Hygromycin B-resistant cells were selected as a stable polyclonal population, induced with copper sulfate, and examined for expression by Western blot analysis using an antiserum raised in rabbits immunized with a MCP-4 peptide. Cells were adapted to serum-free medium (19) and scaled to 10 liters to generate conditioned medium for purification.

Purification of MCP-4 from Drosophila-conditioned Media
Ten ml of 0.5 M EDTA (pH 7.4) was added per liter of conditioned medium, which was then diluted 1:1 (v/v) with deionized/distilled water. The diluted medium was filtered through a 0.45-m membrane filter to remove particulates and subsequently loaded, via a peristaltic pump, at 14 -15 ml/min onto a 10-ml bed volume Sepharose-S Fast Flow column previously equilibrated with 10 mM HEPES, pH 7.0. After sample loading, the column was washed with 10 mM HEPES, pH 7.0, until the absorbance at 280 nm of the effluent returned to base line. Bound material was then eluted from the column with 1 M NaCl in 10 mM HEPES, pH 7.0. The eluate was concentrated using an Amicon Stir Cell concentrator using a YM3 3,000-Da cut-off membrane. The concentrated eluate was applied to a Superdex 75 size exclusion column equilibrated in 20 mM sodium phosphate, 150 mM NaCl, pH 7.0, and developed using the same buffer. Fractions were analyzed by 17.5% SDS-PAGE, and those containing MCP-4 were pooled, concentrated in an Amicon Stir Cell concentrator using a YM3 membrane, filter-sterilized through a 0.22-m membrane filter, aliquoted, and stored at Ϫ20°C. Samples of the purified MCP-4 were subjected to N-terminal sequence analysis, quantitative amino acid analysis, and matrix-assisted laser desorption time-of-flight mass spectrometry to confirm the integrity and concentration of the product.

Amino Acid and N-terminal Sequence Analysis of MCP-4
MCP-4 was hydrolyzed in vacuo under 6 N HCl for 20 h at 110°C. The resulting hydrolysate was analyzed by ion exchange amino acid analysis using postcolumn ninhydrin detection on a Beckman 6300 analyzer equipped with a System Gold data acquisition system (Beckman Instruments, Fullerton, CA). N-terminal sequence analysis was performed on an Applied Biosystems 470A gas phase protein sequencer (Applied Biosystems, Inc., Foster City, CA) equipped with a Beckman 126/166 system (Beckman Instruments, Fullerton, CA) for on-line phenylthiohydantoin analysis. Data were acquired using System Gold chromatography software (Beckman Instruments, Palo Alto, CA). Samples were spotted directly onto Polybrene-coated GF/C filters (Applied Biosystems), and standard ABI sequencing cycles were used.

Mass Spectrometry
Matrix-assisted laser desorption mass spectrometry data for MCP-4 were obtained on a Vestec LaserTec Research laser desorption time-offlight mass spectrometer (Vestec Corp., Houston, TX). Samples were prepared for analysis by mixing MCP-4, in PBS, pH 7.0, with sinapinic acid (saturated solution in 33% CH 3 CN, 0.1% trifluoroacetic acid (v/v)) for a final concentration of 1 M. Apomyoglobin (Sigma) was included as an internal calibrant (MH 2ϩ ϭ 8476.8 Da). Desorption/ionization was accomplished using a 337-nm nitrogen laser. Spectra were averaged over ϳ50 laser scans. Calibrations were carried out using a customized version of IGOR on a Macintosh personal computer.

Monocyte Isolation
Human peripheral blood monocytes were prepared from the blood of normal healthy volunteers, essentially as described by Boyum (20). Blood was collected into anticoagulant (1 part 50 mM EDTA, pH 7.4, to 9 parts blood) and then centrifuged for 5 min at 600 ϫ g. The upper layer of platelet-rich plasma was removed and centrifuged for 15 min at 900 ϫ g to pellet the platelets. The upper layer of platelet-poor plasma was removed and added back to the packed red cells; the pelleted platelets were discarded. Dextran T500 was added (10 volumes of EDTA blood to 1 volume of 6% (w/v) dextran in 0.9% (w/v) NaCl), and the erythrocytes were allowed to sediment at unit gravity for 30 min. The resultant leukocyte-rich plasma was removed and centrifuged for 5 min at 400 ϫ g. The cell pellet was resuspended in 5 ml of the supernatant, and the suspension was underlayered with 3 ml of Nyco-Prep TM1.068 and then centrifuged for 15 min at 600 ϫ g. The mononuclear layer at the interface between the plasma and the NycoPrep TM1.068 was removed and washed through PBS by centrifugation for 5 min at 400 ϫ g. The mononuclear layer typically contained Ն80% monocytes, determined by staining cytocentrifuge preparations for nonspecific esterase using ␣-naphthyl butyrate. Cell viability (typically Ͼ95%) was assessed as the ability to exclude trypan blue.

Preparation of Chinese Hamster Ovary Cells, Expressing the Human CC-CKR2B Receptor
Oligonucleotide primers 5Ј-TGGGGTACCCCACAACATGCTGTCC-ACATCTCGTTCTCG-3Ј (sense) and 5Ј-CAAACGGATCCTCGTTTT-ATAAACCAGCCGAG-3Ј (antisense) containing KpnI and BamHI restriction sites, respectively (underlined), were synthesized based on the published sequence of the human CC-CKR-2B receptor (10). These primers were used to produce a full-length cDNA for this receptor by PCR with human placenta cDNA as the template. The expected product of 1.1 kilobase pairs was digested with KpnI and BamHI and inserted into the equivalent sites of the mammalian expression vector pCDN. The insert of the resulting construct was completely sequenced to confirm its identity and orientation. To obtain a stable cell line expressing the CC-CKR-2B receptor, the pCDN:CC-CKR-2B expression plasmid was electroporated into Chinese hamster ovary cells. Stable cell lines were isolated by growing cells in the absence of nucleosides. Those cell lines that grew out of selection were then analyzed for CC-CKR-2B expression using a radiolabeled MCP-1 binding assay. The cell line used here, C11, was found to express approximately 200,000 receptors/cell.

Chemokine Binding and Binding Displacement Studies
Monocytes were suspended in RPMI 1640 medium without bicarbonate containing 0.2% bovine serum albumin and 0.1% azide. 125 I-Labeled chemokines (MCP-1 or RANTES, about 2 ϫ 10 4 cpm) were incubated with 1-2 ϫ 10 6 cells in the presence or absence of increasing concentrations of unlabeled chemokines (MCP-1, MCP-3, MCP-4, RANTES, or MIP-1␣) for 15 min at 37°C in a 96-well plate in a final volume of 0.2 ml. After the incubation, 0.5 ml of ice-cold wash buffer (20 mM Tris, 0.5 M NaCl, pH 7.4) was added, and cells were collected onto a polyethyleneimine-treated Whatman GF/C filter using a Brandell cell harvester. Filters were washed with 4 ml of cold wash buffer, and the radioactivity bound to the filters was counted in a ␥-counter. For competition studies, the IC 50 was calculated with a curve fitting program (GraFit, Erithacus Software, London), using a four-parameter logistic, cpm bound ϭ cpmmax /(1 ϩ ([L]/IC 50 ) s )ϩ cpm ns , where cpm max represents the binding without competitor, [L] is the competitor concentration, cpm ns is the nonspecific binding, and s is the slope factor. The cpm bound was corrected for "no cell" controls. To obtain the K d and capacity of binding for individual chemokines, specific binding data from homologous displacement experiments were fitted into a single site ligand binding equation using the GraFit best fit program. Chemokine binding to the CHO cells, expressing the human CC-CKR-2B receptor, were performed essentially as described above except that 125 I-labeled chemokines were di-luted with unlabeled chemokine to a specific activity of 400 Ci/mmol and used in a final concentration of 0.6 nM. Cells were incubated with 125 I-labeled chemokines plus or minus unlabeled chemokines for 30 min at 37°C.

Measurement of [Ca 2ϩ ] i
For measurements of cytosolic Ca 2ϩ concentrations, monocytes were incubated with 0.5 M Fura-2/AM for 30 min at 37°C in HEPESbuffered saline (145 mM NaCl, 5 mM KCl, 1 mM MgCl 2 , 10 mM HEPES, and 10 mM glucose), pH 7.4, at 37°C, supplemented with 1% albumin (w/v) and 1 mM CaCl 2 . After loading with Fura-2, the cells were centrifuged for 5 min at 300 ϫ g and then resuspended in buffer containing no added albumin, to a cell density of 1.5 ϫ 10 6 cells/ml, and kept at room temperature until use. This protocol resulted in a cytosolic Fura-2 concentration of ϳ100 M. Serial dilutions of chemokines in PBS plus 0.1% albumin (w/v) (sterile filtered) were added to aliquots (0.7 ml) of cell suspension. The Fura-2 fluorescence of monocyte suspension was measured at 37°C in a single excitation, single emission (500 nm) wavelength Perkin-Elmer LS5 fluorometer. [Ca 2ϩ ] i was calculated from changes in fluorescence measured at a single excitation wavelength of 340 nm, as described by Grynkiewicz et al. (21). [Ca 2ϩ ] i measurements in CHO cells, expressing the human CC-CKR-2B receptor were performed essentially as described above for monocytes. After loading with Fura-2/AM, cells (1 ϫ 10 6 /ml) were kept in ice-cold medium (118 mM NaCl, 4.6 mM KCl, 25 mM NaHCO 3 , 1 mM KH 2 PO 4 , 11 mM glucose, 50 mM HEPES, 1 mM MgCl 2 , 1 mM CaCl 2 , 0.1% gelatin (pH 7.4). Aliquots (2 ml) of cell suspension were prewarmed at 37°C for 5 min in 3-ml plastic cuvettes, and fluorescence was measured in a fluorometer (Johnson Foundation Biomedical Group) with magnetic stirring and temperature controlled at 37°C. Excitation was set at 340 nm, and emission was set at 510 nm. [Ca 2ϩ ] i was calculated as described for monocytes.
For studies in monocytes on cross-desensitization of calcium responses, chemokines were added sequentially with a 2-min interval, and [Ca 2ϩ ] i transients were recorded. The concentrations used in these studies varied for each chemokine and were set at levels known to induce the maximal response for [Ca 2ϩ ] i mobilization (see above).

Monocyte Chemotaxis Assay
Cell migration was evaluated using a 48-well modified Boyden microchemotaxis chamber (Neuroprobe, Cabin John, MD). 25-l aliquots of chemokines, diluted in RPMI 1640 containing 0.1% bovine serum albumin (w/v) were placed in the wells of the lower compartment, and 50-l aliquots of monocyte suspension in RPMI 1640 (1.5 ϫ 10 6 cells/ml) were placed in the upper wells of the chamber; the two components were separated by a polycarbonate filter with a 5-m pore size. The chamber was incubated at 37°C in humidified air containing 5% CO 2 , for 60 min. After incubation, the filter sheet was removed, and the nonmigrated cells were scraped from its upper surface. The filter was stained with Diff-Quik, and the number of migrated cells was counted (5 fields/well), using a ϫ 40 objective and ϫ 10 ocular containing a 10-mm 2 counting grid.

MCP-4 mRNA Expression
Multiple Tissue Blot Analysis-Multiple tissue blots were commercially purchased and prepared as described in the accompanying literature from the manufacturer. Denatured MCP-4 cDNA (50 ng) was labeled with [ 32 P]CTP using Klenow reagent. Multiple tissue blots were probed using ExpressHyb solution and after washing were exposed on X-Omat (Kodak) x-ray development film for between 48 and 96 h.
Expression in Human Pulmonary Smooth Muscle Cells-Total RNA (3 g/sample) isolated from quiescent human pulmonary smooth muscle cells and cells treated with serum for the indicated times were immobilized on nitrocellulose filters and hybridized to either 32 P-labeled MCP-4 cDNA or glyceraldehyde-3-phosphate dehydrogenase cDNA, as described previously (22).
Antibodies to MCP-4 -Initial antibodies used for the monitoring of the expression of recombinant MCP-4 were raised in rabbits after injection of a synthetic MCP-4 peptide (cysteine 58 to isoleucine 73). Subsequent polyclonal antibodies against MCP-4 were raised in rabbits using purified E. coli-expressed His-tagged MCP-4. The antiserum CS09, used for immunodetection of MCP-4 in the present study, was highly specific for MCP-4 (E. coli-as well as Drosophila-expressed MCP-4) and showed no cross-reactivity whatsoever with related chemokines (MCP-1, MCP-2, MCP-3, eotaxin, MIP-1␣, or RANTES) in Western blot analysis.
Western Analysis of Human Tissue Extracts-MCP-4 protein expression in human tissues was studied by Western blot analysis. Tissue-specific protein solutions of normal human tissues were obtained from CLONTECH, fractionated (75 g of protein/lane) on a 16% polyacrylamide gel and blot on nitrocellulose membranes. Commercial, ready made human multiple tissue blots (75 g of protein extracts of normal human tissues) were also used and obtained from DNA Technologies, Inc. (Gaithersburg, MD). Immunoblots were probed with antiserum CS09 (1:2000 dilution) using the ECS (Amersham) detection system. The lower limit of detection was determined using dilutions of pure MCP-4 and found to be ϳ1 ng of MCP-4.
Immunohistochemistry-Tissues used in this study were coronary arteries (n ϭ 6) from heart transplant recipients and carotid endarterectomy specimens (n ϭ 2) from patients undergoing vascular reconstructive surgery for arterial occlusive disease. Tissue specimens were snap frozen in liquid nitrogen and stored at Ϫ70°C. Specimens were mounted in Tissue-Tek OCT compound embedding medium (Agar Scientific Ltd, Essex, UK) for cryostat sectioning. Serial transverse tissue sections (10 m) were thaw mounted onto Superfrostϩ slides (BDH, Leicestershire, UK). Sections were fixed in acetone for 10 min and stored at Ϫ70°C until use. Immunohistochemistry was performed using the avidin-biotin-peroxidase complex method and the rabbit polyclonal antiserum for MCP-4 CS09 (see above). All incubations were done at room temperature unless otherwise stated. Briefly, endogenous peroxidase activity was blocked by incubating slides in 0.6% H 2 O 2 in methanol for 15 min. Nonspecific binding was blocked by incubating sections in normal swine serum before applying the primary antibody. Optimal primary antibody concentrations were predetermined by titration, and slides were incubated overnight at 4°C in a 1:1000 dilution. Slides were then incubated in biotinylated swine anti-rabbit antibody (Dako, Buckinghamshire, UK) for 30 min, washed, and incubated with avidin/biotinylated horseradish peroxidase complex (Vector Laboratories, Peterborough, UK) for 30 min. Slides were then stained with diaminobenzidine tetrahydrochloride substrate (Vector Laboratories). The sections were counterstained with hematoxylin, dehydrated through a graded series of alcohols, and mounted in DPX. Negative controls were omission of primary antibody and substitution of primary antibody with normal rabbit serum.
The cellular composition of the plaques was determined using the following mouse monoclonal antibodies anti-CD68, (a specific human monocyte/macrophage marker), anti-␣-actin (smooth muscle), and anti-Von Willebrand factor (endothelial cells), all from Dako. The method was the same as above except that primary antibody incubations were for 30 min at room temperature. These were followed by incubation with a biotinylated rabbit anti-mouse secondary antibody (Dako).

RESULTS
Sequence Comparison with Other Chemokines-The derived amino acid sequence of the full-length open reading frame for the novel chemokine is given in Fig. 1. Full-length MCP-4 consists of 98 amino acid residues, and an amino acid sequence alignment with known human chemokines showed a high homology to MCP-1 (62% similarity), MCP-3 (61%), and eotaxin (59%) with lower homologies to MCP-2, MIP-1␣, MIP-1␤, and RANTES (52, 37, 39, and 29%, respectively). Based upon these sequence homologies and functional properties (see below), we propose to adopt the name MCP-4 as used throughout this manuscript.
Expression and Purification of Recombinant MCP-4 -To allow functional studies with this novel chemokine, MCP-4 was expressed in Drosophila S2 cells, a system previously shown to correctly recognize and remove human cytokine signal sequences (18,23). Growth medium from a stably transfected Drosophila S2 cell line carrying the expression vector for MCP-4 was analyzed by immunoblot, using an antiserum raised against a synthetic MCP-4 peptide. A 9-kDa protein band was detected that was absent from medium from control cells ( Fig. 2A). After purification by ion exchange and size exclusion chromatography, the final preparation contained a 9-kDa species as visualized on a stained polyacrylamide gel (Fig. 2B). Mass spectrometry showed a predominant peak with a mass of 8583 Da, corresponding to the size expected (8580 Da) for a protein extending from glutamine 24 (cyclized to pyroglutamatic acid) to the end of the open reading frame (Fig. 2C). A minor peak at 8782 Da was also observed, which is close to the mass expected (8780 Da) for a molecule extending from leucine 22 to the end of the open reading frame. N-terminal sequence analysis indicated that most of the material in the preparation was blocked, as expected for a protein with an N-terminal pyroglutamatic acid. However, minor sequences starting at unblocked glutamine 24 (QPDALNVPST) and at leucine 22 (LAQPDALNVP) were detected. These results show that over 90% of the material in this preparation has glutamine 24 (cyclized or free) at the N terminus. This N-terminal sequence aligns very well with the consensus sequence QPDAV for mature human CC chemokines (compare Fig. 1) and is likely to represent the product of MCP-4 maturation in humans. The nature of the minor sequence in Drosophila cell expression (MCP-4 starting at leucine 22) is presently unknown; however, MCP-4 maturation by signal sequence cleavage at the glycine 21-leucine 22 bond and subsequent removal of the leucine 22-alanine 23 dipeptide seems unlikely, since proline at a position 3 residues upstream of a signal peptide cleavage site (as proline 19 would be) is not generally found (24).
Effects of MCP-4 on Monocyte Chemotaxis-The effect of mature (residues 24 -98) MCP-4 (hereafter referred to as MCP-4) on chemotaxis of freshly isolated human peripheral blood monocytes was tested in a 48-well Boyden chamber and compared with chemotactic responses toward known human chemokines. Dose-response curves of these studies are given in Fig. 3. All chemokines gave the expected bell-shaped response curves with inhibition of chemotaxis at high chemokine concentrations due to receptor desensitization. MCP-4 was found to be a potent stimulator of monocyte chemotaxis (EC 50 ϭ 2.88 Ϯ 0. cytosolic calcium (EC 50 ϭ 3.96 Ϯ 0.28 nM, data not shown) as reported by others previously (27).
Binding Competition Studies-To investigate which monocyte chemokine receptors are involved in functional responses of MCP-4, chemokine binding studies were performed. Binding was measured at 37°C in the presence of sodium azide to arrest chemokine uptake and processing. As demonstrated by others earlier (28), freshly isolated peripheral blood monocytes possess specific high affinity binding sites for 125 I-radiolabeled MCP-1 (3384 Ϯ 1223 binding sites/cell, K d ϭ 0.51 Ϯ 0.06 nM). Competition analysis showed that the binding of 125 I-MCP-1 to these cells could be competed for completely by a surplus of unlabeled MCP-1 as well as unlabeled MCP-3 (Fig. 5A). Interestingly, MCP-4 was also found to be a potent and complete competitor of 125  are probably too high to be of physiological significance. Therefore, we expect that receptors with high affinity for RANTES receptors are unlikely to be involved in MCP-4 binding and signaling.
To strengthen the evidence for MCP-4 being a ligand for the CC-CKR-2B receptor, recombinant CHO cells expressing the human CC-CKR-2B receptor were constructed, and displacement of binding of 125 I-MCP-1 by MCP-4 and other chemokines was tested. As in freshly isolated human peripheral blood monocytes, the binding of 125 I-MCP-1 to CHO cells expressing CKR-2B was completely displaced by unlabeled MCP-1, MCP-3, and MCP-4 with the following relative potencies: MCP-1 Ͼ MCP-4 Ͼ MCP-3 (IC 50 values of 3.6 Ϯ 0.52, 7.1 Ϯ 5.19, and 12.7 Ϯ 3.46 nM, respectively, n ϭ 3) (Fig. 6). MIP-1␣ and RANTES did not displace any binding of radiolabeled MCP-1 Functional response studies added further evidence. Like monocytes, CHO cells expressing the CC-CKR-2B receptor were found to generate potent transient increases in cytosolic Ca 2ϩ concentration following exposure to MCP-1. Since coupling of the receptor to the internal signaling system was evidently present, calcium transients following exposure to different concentrations of chemokines were measured (Fig. 7). MCP-1, MCP-3, and MCP-4 were all found to provoke an increase in cytosolic calcium (mean values Ϯ S.D. for EC 50 of 4.2 Ϯ 1.21, 35.3 Ϯ 15.4, and 51.7 Ϯ 18.9 nM, respectively, n ϭ 3-4), while RANTES and MIP-1␣ were all ineffective up to 100 nM (data not shown). Not unexpectedly (see above), MCP-2 was also found to be ineffective. No 125 I-MCP-1 binding or calcium transients were seen in untransfected CHO cells exposed to any of these chemokines. From these results it is evident that MCP-4 can bind and signal through interaction with the CC-CKR-2B receptor.
Desensitization Studies in Monocytes-Rapid successive exposure to their ligands is known to desensitize the signaling capacity of chemokine receptors. We have used heterologous desensitization of cytosolic Ca 2ϩ transients in monocytes, following chemokine exposure to investigate whether ligands are signaling through the same or different types of chemokine receptors. We used chemokine concentrations shown earlier to result in a maximum [Ca 2ϩ ] i response (compare Fig. 4). The results of these cross-desensitization studies are shown in Fig. 8.
Preexposure of monocytes to MCP-1 completely blocked a response to MCP-4, while initial exposure to MCP-4 greatly diminished a subsequent MCP-1 signal. These results suggest that MCP-4 and MCP-1 signal at least partially through a shared receptor on monocytes, probably the CC-CKR-2B receptor. MCP-3 and MCP-4 also showed mutual heterologous desensitization, although the signal of the second chemokine was not completely blocked by the administration of the first, suggesting partial sharing of receptors. Since MCP-3 interacts with the CC-CKR-2B receptor, cross-desensitization between MCP-3 and MCP-4 is to be expected if this receptor is involved in MCP-4 signaling. Interesting results were also obtained in cross-desensitization studies with RANTES and MIP-1␣. MCP-4 was able to desensitize RANTES signaling, but RAN-TES was found to be nearly inactive in desensitizing an MCP-4 signal. A partial mutual desensitization was found with the MCP-4 mRNA Distribution-Northern blot analysis of mRNA for MCP-4 in various tissues revealed an approximately 1-kilobase pair transcript in small intestine, thymus, and colon (Fig. 9A). No transcripts were detected in RNA from spleen, prostate, ovaries, and testes. In addition (not shown) a 1-kilobase pair transcript was found in lung, trachea, stomach, lymph node, and colonic mucosa, but not in a number of other tissues including spinal cord, adrenal gland, bone marrow, heart, brain, placenta, liver, skeletal muscle, kidney, and pancreas. Human pulmonary artery smooth muscle cells also showed a low level of MCP-4 mRNA expression, and the level of this transcript could be increased transiently by short exposure of these cells to 10% human serum (Fig. 9B), a treatment known to induce differentiation of smooth muscle cells toward a synthetic phenotype (29).
Immunological Studies: Tissue Localization by Western Blot Analysis-The expression of MCP-4 in normal human tissues was investigated by Western blot analysis using a highly specific polyclonal rabbit antiserum against MCP-4. Using commercially available protein extracts for Western blot analysis (CLONTECH) from small intestine, thymus, lung, and stomach (tissues for which MCP-4 mRNA expression was seen) as well an extract from testes, we were unable to detect any MCP-4 protein expression in blots loaded with 75 g of tissue protein.
In another study using a ready made multiple normal human tissue Western blot of extracts from brain, heart, small intestine, pancreas, liver, colon, thymus, and lung (75 g protein/ lane) again we were unable to detect any expression of MCP-4. Based on the detection limit for MCP-4 in our analysis (around 1 ng), it is concluded that the protein expression levels for MCP-4 in these extracts are below 1 ng/75 g of tissue protein.

Localization of MCP-4 in Human Vascular
Tissues-MCP-1 is known to be expressed in human atherosclerotic lesions, mainly by macrophage-foam cells but also in intimal smooth muscle cells and endothelial cells (30 -32). We studied the expression of MCP-4 in atherosclerosis using coronary artery specimens from heart transplant recipients and carotid endarterectomy specimens from patients undergoing vascular reconstructive surgery for arterial occlusive disease. MCP-4 expression was found to be associated with the luminal endothelium of human atherosclerotic coronary vessels (Fig. 10). It was also found on the endothelium of small adventitial blood vessels in all of the coronary vessels examined regardless of the degree of atherosclerosis. Four of the coronaries examined did not contain severe lesions (noncalcified, slightly thickened intima) but had a discrete macrophage infiltrate. We did not find MCP-4 to be strongly associated with these macrophages, although some areas of very weak staining could be seen. However, two of the coronary vessels were almost totally occluded and contained large areas of calcification. In these specimens, MCP-4 was found to be expressed by a subset of macrophages deep in the intimal lesion near areas of calcification. We also found MCP-4 strongly associated with macrophages in the two carotid endarterectomy specimens examined (Fig. 11). In these specimens, the lack of staining for Von Willebrand factor revealed loss of the luminal endothelium. Numerous intimal microvessels could be seen, but MCP-4 did not appear to be expressed by the endothelial cells lining these vessels.
Using primers designed to flank the open reading frame of MCP-4 cDNA, we carried out reverse transcriptase-PCR on RNA extracted from eight carotid endarterectomy specimens. mRNA expression was found in seven of eight specimens (results not shown).

DISCUSSION
In the present work we describe the identification and characterization of a novel human CC chemokine of the MCP family, MCP-4. This 75-amino acid residue chemokine was shown to induce chemotaxis in monocytes with nanomolar potency and to provoke a transient rise in cytosolic Ca 2ϩ concentrations in these cells. Similar to other members of the CC chemokine family, MCP-4 was found to be inactive toward human peripheral blood neutrophils, but in a concurrent study by Uguccioni and co-workers (33), MCP-4 was shown to induce functional responses in T cells, eosinophils, and monocytes.
Using chemokine binding studies as well as functional response data, we have investigated which receptors on monocytes are implicated in MCP-4 binding and signaling. In monocytes MCP-4 as well as MCP-3 completely displaced binding of 125 I-labeled MCP-1. Since MCP-1 and MCP-3 are known ligands of the CC-CKR-2B receptor, while only MCP-1 seems to interact with the CC-CKR-2A splicing variant (10,34,35), we concluded that at least part of the high affinity binding of MCP-4 to monocytes is due to the CC-CKR-2B receptor. This conclusion was further supported by binding studies to CHO cells, stably transfected with the human CC-CKR-2B receptor. Displacement of 125 I-MCP-1 binding in these cells by MCP-4, MCP-3, and other chemokines was very similar to that seen in monocytes, although affinities of all members of the MCP family were somewhat lower for the recombinant CC-CKR-2B receptor in CHO cells as compared with monocytes. It is possible that the latter effect is due to differences in the microenvironment for the receptor between monocytes and the recombinant CHO cells.
Functional response studies and desensitization studies in monocytes with MCP-4 and other ␤-chemokines are largely in agreement with the binding studies. MCP-1, MCP-3, and Chemokine [Ca 2ϩ ] i response desensitization studies also indicate that MCP-4, MCP-1, and MCP-3 share a common chemokine receptor. MCP-4 also showed a partial desensitization of RANTES and MIP-1␣ calcium transients, although our binding displacement studies did not show clear competition between MCP-4 and these chemokines (only at very high concentrations of MCP-4 were we able to displace 125 I-labeled RANTES binding; see Fig. 5B). In addition to the CC-CKR-2A and -B receptors, monocytes are known to express mRNAs for the CC-CKR-1 (MIP-1␣/RANTES receptor) (9, 36) and the pro- miscuous basophil receptor CC-CKR-4 (15,37). The CC-CKR-3 (eotaxin) receptor is reported not to be present on these cells (14). No data are as yet available on the presence or absence of the CC-CKR-5 receptor. All of these receptors (CC-CKR-1, -3, -4, and -5) have nanomolar affinity for RANTES (9,13,15,16), a ligand whose binding in monocytes is not very well displaced by MCP-4. Based on this finding, we would suggest that this set of receptors is unlikely to possess a high affinity for MCP-4. Whether any of these receptors may accept MCP-4 with low affinity needs to be investigated further.
At present, MCP-4 has not yet been purified from an authen-tic human source. Therefore, the primary structure of the biologically relevant (mature) form has not yet been established. However, the alignment of its sequence with those of well characterized members of the ␤-chemokine family strongly suggests that the recombinant form synthesized by Drosophila cells (which starts with residue 24 at QPDAL) is the correct mature protein.
The mature forms of all other members of the MCP family are known to start at this position. Preliminary data on mRNA expression for MCP-4 shows a wide tissue distribution. Small intestine, thymus, colon, and lung were all positive, while no transcript was seen in peripheral blood leu- kocytes, liver, heart, brain, skeletal muscle, and some other tissues. However, we have as yet been unable to demonstrate MCP-4 protein expression in any of the tissues tested. Using Western blot analysis, a procedure that could detect levels of MCP-4 as low as 1 ng/lane, no MCP-4 expression could be detected in blots containing 75 g of tissue protein extract, indicating that the constitutive expression of MCP-4 protein in these tissues is very low.
MCP-1 has been found in human atherosclerotic lesions (30 -32) and is thought to play a central role in the recruitment of monocytes into the subendothelial space during atherogenesis (38). Given the fact that MCP-4 is also a potent chemotactic agent for monocytes, we investigated the expression of MCP-4 in human atherosclerotic carotid and coronary arteries. Using reverse transcriptase-PCR, we have established that human atherosclerotic lesions express MCP-4 mRNA. Furthermore, we have determined that the MCP-4 protein expression showed many similarities to previously published mRNA/protein expression data for MCP-1 in human atherosclerosis. We found variable levels of expression of MCP-4 on the luminal endothelium of atherosclerotic coronary arteries. Similar to our data for MCP-4, Takeya et al. (32) found MCP-1 to be expressed by luminal endothelial cells of human aortic atherosclerotic plaques. We also found MCP-4 protein expression on endothelial cells lining the vasa vasorum of coronary arteries regardless of the severity of the lesion, while Nelken et al. (31) failed to detect MCP-1 mRNA expression in these vessels. MCP-4 expression in adventitial vessels may be important in lesion development. It has been found that newly formed intimal vessels, thought to be important in plaque rupture, largely originate from the adventitial vasa vasorum (39). Of the six coronary vessels examined, four were found to contain only a small amount of intimal thickening. We did not find MCP-4 expression in the intima of these vessels, similar to results obtained by Nelken et al. (31) for MCP-1. In all reports on MCP-1 expression in advanced human atherosclerosis, macrophages have been identified as the main cell type expressing MCP-1 mRNA/protein (30 -32). Our data indicate that MCP-4 is expressed by macrophages in advanced lesions as found in both of the carotid plaques examined and also in two of the six coronary arteries. In the specimens studied, no clear indication for MCP-4 expression in (intimal) smooth muscle cells was found. Our in vitro studies with (pulmonary) smooth muscle cells have indicated that these cells express mRNA for MCP-4 when exposed to serum. Smooth muscle cell-associated MCP-1 has been seen in some studies on atherosclerotic lesions (31,40) but not in others (30) and may relate to the type of lesion and its severity. More detailed studies on MCP-4 expression in atherosclerotic lesions are required to settle this point. Given the functional data for MCP-4 and its expression patterns in advanced atherosclerosis, we postulated that MCP-4, like MCP-1, may be involved in the recruitment of monocytes into the arterial wall during the disease process. The constitutive expression of MCP-4 mRNA in tissues such as lung, colon, and small intestine suggests that this chemokine may also play a role in the monocyte attraction in tissues chronically exposed to exogenous pathogens.
In addition to a role in atherogenesis, MCP-4 may be implicated in other inflammatory conditions involving monocytes T-cells and/or eosinophils and associated pathophysiological conditions. Therefore, MCP-4 and/or its receptors may represent targets for pharmacological intervention. However, more studies are required before the origin and specific function of MCP-4 in relation to other CC chemokines can be established. Studies along this line are currently in progress.