C-C Chemokine Receptor 3 Antagonism by the β-Chemokine Macrophage Inflammatory Protein 4, a Property Strongly Enhanced by an Amino-Terminal Alanine-Methionine Swap

Met-Ckβ7 production and other chemokines

The coding sequence of Ckβ7 was amplified from an adult human lung cDNA library using primers to remove the signal peptide and replace the N-terminal alanine seen in the mature protein with a methionine. The nucleotides encoding the C terminus were altered to encode either LKLMPEA or LKLNA, and both cDNAs were cloned into the pQE7 expression vector. The resulting plasmids were transformed into Escherichia coli M15 Rep4 host cells, grown at 37°C in L-Broth containing ampicillin and kanamycin and protein induced by incubation with 0.2 mM isopropyl β-d-thiogalactoside (IPTG) for 3 h. Cells were harvested, resuspended in ET buffer (75 mM EDTA, 50 mM Tris (pH7.5)), and lysed by passing twice through a microfluidizer (Microfluidics, Newton, MA) at 6000–8000 psi. NaCl was added to 0.5 M, and the sample was centrifuged at 7000 × g for 15 min. The pellet was washed in ET plus 0.5 M NaCl and centrifuged at 7000 × g again for 15 min. These partially purified inclusion bodies were resuspended in 1.5 M guanidine hydrochloride and 50 mM Tris (pH 7.4) and were incubated overnight at 4°C and then centrifuged at 30,000 × g. The supernatant was mixed vigorously for 30 min at 4°C in 20 vol of 150 mM NaCl, 2 mM EDTA, and 50 mM sodium acetate (pH 4.5) and was left for 60 h at 4°C. This solution was clarified using a 0.16-μm sterile filter (Filtron; Pall, Port Washington, NY) and chromatographed over a strong cation exchange column (Poros HS-50; Perspective Biosystems, Framingham, MA) prewashed with 6 column vol of 250 mM NaCl and 40 mM sodium acetate (pH 5.5). Bound protein was eluted using 3 to 5 column vol of a stepwise gradient of 0.5 M, 1 M, and 1.5 M NaCl in 40 mM sodium acetate (pH 5.5). Positive fractions were pooled, diluted 3-fold with 40 mM sodium acetate (pH 5.5), and applied to a set of strong anion (Poros HQ-50) and weak cation (Poros CM-20) exchange columns in tandem mode prewashed with 150 mM NaCl and 40 mM sodium acetate (pH 5.5). The CM-20 column was eluted with a 10–20 column vol linear gradient of 0.15–1.25 M NaCl; fractions were analyzed through SDS-PAGE and positive fractions were combined. The proteins were greater than 95% pure by SDS-PAGE and reverse phase HPLC analysis. Peptide sequencing revealed the expected amino-termini of MQVGTNKEL.

Met-RANTES and AOP-RANTES were produced as previously described (37, 38). All other chemokines, including MIP4, were purchased from Peprotech (London, U.K.) or R&D Systems (Abingdon, U.K.).

Cell culture and preparation

HOS cells stably expressing human CCR3 were the generous gift of Dr. Nathaniel Landau (Salk Institute, La Jolla, CA) and were maintained in DMEM plus 10% FCS, antibiotics, and 1 μg/ml puromycin (Sigma, Poole, Dorset, U.K.). HEK293 cells stably expressing human CCR1, -2, -3, and -5 were generated by transfection with Transfectam (Promega, Southampton, U.K.), according to the manufacturer’s protocols, and selection in 0.8 mg/ml G418. Chinese hamster ovary (CHO) cells expressing human D6 are described elsewhere (40).

Eosinophils were purified from single donor leukopaks (American Red Cross, Baltimore, MD) as previously described (41), or by purifying granulocytes from peripheral blood and selecting those cells that were CD16 negative using MACS technology (Miltenyi Biotec, Bergisch, Germany). Eosinophils were greater than 90% pure as assessed on stained cytospun preparations.

Mononuclear cells were purified from samples of peripheral blood (Western Infirmary, Glasgow, U.K.). Blood containing at least 100 U/ml of preservative-free heparin was diluted 1:4 with PBS containing 0.6% acid citrate dextrose (Sigma). This was layered over Ficoll (1.077 density) in a ratio of 3:1 and centrifuged at 400 × g for 30 min at 22°C. Mononuclear cells were removed from the interface and washed three times in PBS/0.6% acid citrate dextrose.

To purify and activate T cells, heparinized blood was collected from healthy donors, separated over ficoll-hypaque, and washed four times in PBS. These cells were then adhered to plastic for 2 h at 37°C in RPMI/10% FCS, and the nonadherent cells were then stimulated for 5 days in 4 μg/ml Con A (Sigma). Cells were then repurified over ficoll-hypaque as before and stimulated at 10⁶ cells/ml in 20 U/ml IL-2 (Peprotech) for 7–14 days. CD3⁺ and CD45RA⁺ cells were purified by positive selection from PBMCs (isolated as described above) using anti-CD3 and anti-RA Abs, respectively, and MACS technology. Purity was >95% as determined by FACS analysis.

Ca²⁺ flux assays

HOS-hCCR3 cells were harvested by trypsinization, washed in SR buffer (136 mM NaCl, 4.8 mM KCl, 5 mM glucose, 1 mM CaCl₂, 0.025% BSA, and 25 mM HEPES (pH 7.6)), and then incubated in SR with 10 μM fura-2-AM (Sigma) for 1 h at 37°C. Cells were then washed in SR and resuspended in SR to ∼2 × 10⁶ cells/ml, and 2 ml was incubated at 37°C in a continuously stirred cuvette in a Perkin-Elmer LS50 Spectrometer. After 2 min, fluorescence emission was recorded every 100 ms (340 nm (λ_ex); 500 nm (λ_em)) for 20 to 40 s, agonist or antagonist was added to a defined concentration, and fluorescence was recorded every 100 ms for a further 80 to 120 s. For agonist dose-response experiments, all ligands were compared with a full dose response performed with human eotaxin to avoid day-to-day experimental variation. PBMCs and lymphocytes were also loaded and analyzed using this procedure.

Eosinophils were loaded for 30 min at room temperature with 2 μM fura-2-AM in a modified SR buffer (125 mM NaCl, 5 mM KCl, 0.5 mM glucose, 1 mM MgCl₂, 1 mM CaCl₂, 0.025% BSA, and 20 mM HEPES (pH 7.4)), washed, and resuspended at 10⁶ cells/ml. Intracellular Ca²⁺ changes from a 2-ml sample were measured in an F2000 spectrometer (Hitachi Instruments, San Jose, CA) by monitoring fluorescence emission at 37°C over time (340 nm and 380 nm (λ_ex); 510 nm (λ_em)).

In all experiments, the distance from the baseline emission to the highest point of the flux was calculated and converted to a percentage of the maximal flux induced in each experiment.

Eosinophil chemotaxis assays

Purified eosinophils were washed with HBSS/BSA (HBSS with 0.1% BSA) and resuspended in this medium at 5 × 10⁶ cells/ml with 1 μM calcein-AM (Molecular Probes, Eugene, OR). After 30 min at 37°C, cells were washed in HBSS/BSA and resuspended to 5 × 10⁶ cells/ml; 20 μl of this suspension was dispensed into each upper chamber of a 96-well chemotaxis plate filter (Neuro Probe, Cabin John, MD). Different concentrations of agonist were added to the bottom chamber, and antagonist was added to either the bottom chamber, the top and bottom chamber, or neither chamber. Cells were allowed to migrate for 3 h through the polycarbonate filter (8-μM pores; polyvinylpyrolidone-free) between the two chambers, and the number of migrated cells in the bottom chamber was quantitated using a fluorescence plate reader (Cytofluor PerSeptive Biosystems). The ratio between the number of cells migrated in the presence of agonist and the number of cells migrated in buffer alone is defined as the chemotaxis index.

Radioiodination of Met-Ckβ7

Five micrograms of Met-Ckβ7 was incubated in 50 μl of PBS containing 100 μg of IODO-GEN (Pierce, Rockford, IL) and 1 mCi Na¹²⁵I (DuPont- NEN, Hounslow, U.K.) for 15 min on ice. The reaction was then run down a D-Salt Excellulose Desalting Column (40–100 micron), and 0.5-ml fractions were taken with PBS. Two-microliter aliquots of the fractions were counted in a Beckman γ 5500B counter (Beckman, High Wycombe, U.K.), and positive fractions were combined.

Eosinophil ¹²⁵I-chemokine binding assay

Purified eosinophils (2 × 10⁵) were placed in each well of a 96-well plate in binding buffer (1 mM CaCl₂, 5 mM MgCl₂, 0.5% BSA, 0.05% sodium azide, and 50 mM HEPES (pH 7.5)). Iodinated chemokine (final concentration of 0.1 nM [¹²⁵I]eotaxin or [¹²⁵I]MCP4, spec. act. 2200 Ci/mmol (NEN, Boston, MA)) was added in the absence or presence of unlabeled chemokines to a final volume of 100 μl. The binding reaction was incubated for 60 min at room temperature, and the cells were then transferred to filter plates (Silent Screen with Ioprodyne membrane (Nalge Nunc, Rochester, NY)) pretreated with 0.1% polyethylenimine and washed three times with binding buffer containing 0.5 M NaCl. The plates were dried and counted after addition of 50 μl of liquid scintillant in each well. Each point was done in triplicate and is presented as the mean of these results with SE. For experiments with [¹²⁵I]Met-Ckβ7, 2 × 10⁵ eosinophils were incubated in DMEM plus 10% FCS, 0.4% sodium azide, and 20 mM HEPES (pH 7.6) containing 45 nM [¹²⁵I]Met-Ckβ7 and with or without 500 nM unlabeled chemokine as competitor for 2 h at room temperature. Cells were then washed twice with ice cold PBS and counted in a Beckman 5500B gamma counter. Each point was done in triplicate and is presented as the mean of these results with SE.

Met-Ckβ7 production

A cDNA clone, termed Ckβ7, was amplified from an adult human lung cDNA library. Predictive algorithms suggest that the putative signal peptide will be cleaved from this protein to leave an amino terminus beginning AQVGT-, although QVGT- may also be produced. Production in insect cells generated a protein with AQVGTNKEL- at the N terminus, although variable amounts of truncated variants starting TNKEL- or NKEL- were also found in our studies. To produce this protein in vitro with a homogeneous N terminus, we generated a cDNA construct in which the region coding the signal sequence was removed and the nucleotides encoding the alanine residue found at the putative N terminus of the mature protein were replaced with an ATG encoding a methionine residue. This cDNA was transformed into bacteria, and the protein was purified (see Materials and Methods). Sequencing revealed the expected N termini of MQVGT-. During the production and analysis of this protein, Ckβ7 was described by others and named PARC, MIP4, DCCK1, or AMAC1 (33, 34, 35, 36). These sequences were identical to Ckβ7, with the exception that the predicted C terminus of Ckβ7 ends LKLMPEA rather than LKLNA, a difference that may be due to allelic variation within this gene and is currently under investigation. Corroborating our observations, in two of these publications, production in COS or insect cells also demonstrated that the N terminus of the mature protein begins AQVGT- (33, 34). To assess the impact of the C-terminal sequence variation, we made a protein in bacteria that as before was engineered to contain a methionine in place of the first N-terminal alanine, but had a C terminus of LKLNA rather than LKLMPEA. Importantly, in all assays tested, these two proteins exhibited identical activity showing that the C termini differences have no affect on the properties described (data not shown). The data shown use the protein with the C terminus of LKLMPEA (since this protein has been more extensively used), and it is referred to hereafter as Met-Ckβ7, with the “Ckβ7” name retained to indicate that it contains the C-terminal sequence derived from our cDNA clone.

Activity of Met-Ckβ7 on chemokine receptors

This protein was tested for its ability to elicit a Ca²⁺ flux through chemokine receptors CCR1, -2, -3, and -5 expressed in heterologous cells. At concentrations up to 1 μM, no signaling was detectable through these receptors whereas known ligands signaled efficiently (data not shown). Also, 700-nM Met-Ckβ7 was unable to displace any [¹²⁵I]MIP1α from the promiscuous D6 chemokine receptor (40) in binding assays on CHO cells expressing this receptor (not shown). However, with HOS cells stably transfected with human CCR3 (HOS-CCR3 cells), pretreatment with 500 nM Met-Ckβ7 prevented subsequent Ca²⁺ fluxes induced with known CCR3 agonists (Fig. 1⇓, A–D). This activity was not seen on CCR1, -2, or -5, even when CCR3 ligands MCP4 (that also uses CCR2) or RANTES (that also signals through CCR1 and -5) were used as agonists (not shown). To further assess the specificity of this antagonist, we isolated mononuclear cells and CD3⁺ T lymphocytes from human peripheral blood and also prepared Con A/IL-2-activated T cells (see Materials and Methods), and tested whether Met-Ckβ7 could inhibit Ca²⁺ fluxes induced with a range of chemokines. None of the human chemokines that gave a Ca²⁺ flux in these various cell types (melanoma growth-stimulating activity (MGSA), IL-8, stromal cell-derived factor (SDF) 1, RANTES, MIP1α, MIP1β, MCP1, fractalkine, MIP3α, MIP3β, secondary lymphoid-tissue chemokine (SLC), and IFN-γ-inducible 10-kDa protein (IP-10)) were antagonized by 500 nM Met-Ckβ7 (not shown). Further, no agonist activity was detected with Met-Ckβ7 in these assays. One example of these experiments using PBMCs, and human MIP1α as the agonist, is shown in Fig. 1⇓E. It is of note that unmodified MIP4 gave no detectable Ca²⁺ flux in any of the cell types tested (not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Met-Ckβ7 prevents signaling through CCR3 induced by eotaxin, MCP4, RANTES, and eotaxin2. Fura-2-loaded HOS-CCR3 cells (A–D) or PBMCs (E) stimulated at 37°C with agonist in the absence (left panels) or presence (right panels) of 500 nM Met-Ckβ7. Fluorescence emission is recorded every 0.1 s for 100 s (340 nm (λ_ex); 500 nm (λ_em)).

Met-Ckβ7 is a potent CCR3 antagonist

Next, we wanted to define the potency of this CCR3 antagonist. Several β-chemokines are known to act as CCR3 agonists, and Met-Ckβ7 may exert differential effects on these ligands. Thus, we first compared the potency of these known CCR3 agonists by performing dose-response experiments examining Ca²⁺ flux into HOS-CCR3 cells at different concentrations of ligand. As shown in Fig. 2⇓, eotaxin, eotaxin-2, and MCP4 induced strong signals through CCR3 in the low nanomolar range and above, with slight Ca²⁺ fluxes still detectable at 1 nM. RANTES was much less potent, and MCP3 gave a barely detectable signal, even at 100 nM. The number of CCR3 receptors on the HOS-CCR3 cells is low (not shown) and may account for the relative ineffectiveness of RANTES and MCP3 in these assays. We then performed dose-response experiments using a range of concentrations of Met-Ckβ7 to examine its effect on a subsequent Ca²⁺ flux induced with a concentration of CCR3 agonist known to give a strong signal. As shown in Fig. 3⇓A, half maximal inhibition of a Ca²⁺ flux induced with 25 nM eotaxin or MCP4, or 50 nM eotaxin-2, was seen at ∼25 nM Met-Ckβ7. With 100 nM RANTES, which only produces a weak signal into HOS-CCR3 cells, 50% inhibition was observed at a slightly lower concentration of Met-Ckβ7 (∼10 nM) (not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

CCR3 ligands signal with different potencies into HOS-CCR3 cells. Dose-response curves for ligand-induced calcium ion fluxes into fura-2-loaded HOS cells expressing human CCR3. Each flux was compared with a maximal defined as that induced with 100 nM MCP4.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Met-Ckβ7 is a potent antagonist of signaling through CCR3 into HOS-CCR3 cells or eosinophils. The peak of calcium ion flux (detected by fura-2 fluorescence) induced by a set amount of agonist in the presence of a range of Met-Ckβ7 concentrations is represented as a percentage of the flux induced in the absence Met-Ckβ7. A, HOS-CCR3 cells; (B) purified human eosinophils. Arrows indicate where no Ca²⁺ flux was detectable.

We also used these HOS-CCR3 signaling assays to examine the activity of two known RANTES receptor antagonists, Met-RANTES and AOP-RANTES (37, 38) in comparison with Met-Ckβ7. Met-RANTES used at 100 nM only slightly reduced a 25-nM eotaxin-induced flux, whereas 100 nM AOP-RANTES was more potent, reducing this flux by ∼35% (Fig. 4⇓). In contrast, 100 nM Met-Ckβ7 reduced a 25-nM eotaxin-induced flux by ∼85% (Fig. 4⇓D). It is of note that Met-RANTES and AOP-RANTES are able to induce a moderate Ca²⁺ flux through CCR1 and CCR5 and that AOP-RANTES is in fact fully active on CCR5 in these assays⁵ (39). It has been shown that AOP-RANTES (100 nM), and to a lesser degree Met-RANTES, have weak Ca²⁺ mobilizing activity on CHO-CCR3 and L1.2-CCR3 transfectants, although this activity was never superior to 50% of that induced by RANTES.⁶ Neither Met- nor AOP-RANTES (at 100 nM) mobilize calcium in the HOS-CCR3 transfectants (Fig. 4⇓), perhaps due to the low receptor level on these cells, but AOP-RANTES at higher concentrations (250 nM) does induce very weak, but detectable, Ca²⁺ fluxes into these cells (data not shown). On the contrary, Met-Ckβ7 shows no CCR3 signaling potential in Ca²⁺ flux assays with CCR3 transfectants or eosinophils, even at 1 μM.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Met-Ckβ7 more effectively inhibits CCR3-mediated Ca²⁺ flux than Met- or AOP-RANTES. Fura-2-loaded HOS-CCR3 cells stimulated at 37°C with 25 nM eotaxin (A), or with 25 nM eotaxin in the presence of 100 nM Met-RANTES (B), 100 nM AOP-RANTES (C), or 100 nM Met-Ckβ7 (D). Fluorescence emission is recorded every 0.1 s for 100 s (340 nm (λ_ex); 500 nm (λ_em)).

Met-Ckβ7 can prevent eosinophil Ca²⁺ flux induced by CCR3 agonists

We extended our studies to examine whether Met-Ckβ7 could antagonize the function of CCR3 agonists on eosinophils. The concentration of CCR3 agonist required to induce detectable Ca²⁺ fluxes into these cells was considerably lower than that required with the CCR3-transfected HOS cells used above, most likely due to higher receptor levels on eosinophils. Met-Ckβ7 completely blocked Ca²⁺ fluxes into eosinophils induced by 1 nM eotaxin or MCP4, at 50 nM or 10 nM of Met-Ckβ7, respectively (Fig. 3⇑B). When 10 nM RANTES or MCP3 were used as agonists, whereas low concentrations of Met-Ckβ7 (1 nM) caused a profound reduction in the magnitude of the Ca²⁺ flux, complete antagonism was not achieved even with 100 nM Met-Ckβ7. This is explained by the presence of CCR1 on these cells acting as a receptor for RANTES and MCP3, further demonstrating that Met-Ckβ7 does not abrogate signaling through this receptor.

Analysis using radio-iodinated ligands suggest a direct interaction between Met-Ckβ7 and CCR3

Next, we examined the ability of unlabeled Met-Ckβ7 to displace radio-iodinated MCP4 or eotaxin from CCR3, in comparison with unlabeled eotaxin and MCP4. The low level expression of CCR3 on the transfected HOS-CCR3 cells necessitated the use of eosinophils in these assays. Fig. 5⇓A shows that Met-Ckβ7 more efficiently displaced [¹²⁵I]eotaxin from these cells (IC₅₀ ∼ 6 nM) than unlabeled eotaxin (IC₅₀ ∼ 10 nM) or MCP4 (IC₅₀ ∼ 60 nM), whereas with the displacement of [¹²⁵I]MCP4, eotaxin and Met-Ckβ7 behaved similarly (IC₅₀ ∼ 5 nM) and were more effective than unlabeled MCP4 (IC₅₀ ∼ 25 nM) (Fig. 5⇓B). Thus, Met-Ckβ7 is able to successfully compete with known CCR3 agonists for binding to CCR3. This is most likely due to a direct interaction between the antagonist and the receptor. We have been able to demonstrate this directly by radiolabeling Met-Ckβ7 and showing binding to eosinophils that could be competed by unlabeled Met-Ckβ7 and eotaxin, but not by unlabeled MIP-1α (Fig. 5⇓C). Interestingly, unlabeled MIP4 was also able to displace some of the labeled Met-Ckβ7 from the surface of eosinophils, suggesting that it too interacts weakly with this receptor (Fig. 5⇓C; and see below). Subsequent experiments with radiolabeled eotaxin also demonstrated heterologous displacement by MIP4, although this ligand was considerably less effective than Met-Ckβ7 (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Displacement of ¹²⁵I-labeled eotaxin, MCP4, or Met-Ckβ7 from eosinophils. A and B, Purified eosinophils (2 × 10⁵) were incubated for 60 min at room temperature in azide-containing binding buffer with 0.1 nM of [¹²⁵I]eotaxin (A) or 0.1 nM of [¹²⁵I]MCP4 (B) plus a range of concentrations of unlabeled chemokine, either Met-Ckβ7 (▪), MCP4 (X), or eotaxin (○). After washing with binding buffer containing 0.5 M NaCl, the percentage of radio-iodinated ligand remaining bound was calculated compared with assays in which no unlabeled chemokine was added. Each point is the mean of three identical incubations and SE is included. C, Purified eosinophils (2 × 10⁵) were incubated for 2 h at room temperature in azide-containing binding buffer with 45 nM [¹²⁵I]Met-Ckβ7 plus 500 nM of the various unlabeled chemokines indicated at the bottom of the graph. Cells were washed twice with PBS and remaining [¹²⁵I]Met-Ckβ7 bound was counted. Results are the mean of three identical incubations represented as a percentage of the binding seen in the absence of competitor, and SE is included.

Met-Ckβ7 prevents eosinophil chemotaxis induced by CCR3 ligands

To further test the potency of the antagonist on eosinophil function, we performed chemotaxis assays with purified human eosinophils. Results from three donors are shown in Figs. 6⇓ and 7. Eotaxin and MCP4 induced chemotaxis of eosinophils most effectively when 1 nM or 10 nM of the chemokine was present in the lower well, although the maximal chemotaxis index achieved varied considerably between donors. These concentrations are considerably lower than those required to induce maximal Ca²⁺ fluxes through CCR3 into HOS-CCR3 cells or eosinophils (Fig. 3⇑A and not shown), indicating that low receptor occupancy is optimal for chemotaxis induction. Met-Ckβ7 was unable to stimulate chemotaxis over the concentrations tested but was able to efficiently block eotaxin- and MCP4-mediated chemotaxis. The potency of the antagonist in these assays varied slightly between donors: thus, whereas 1 nM was sufficient to inhibit the eotaxin- and MCP4-induced chemotaxis of eosinophils from donor A, this was not sufficient to prevent eotaxin-induced chemotaxis of donor B’s eosinophils unless it was added to the top and the bottom wells of the assay chamber. Curiously, with donor C, addition of Met-Ckβ7 to the bottom compartment of the well only, reduced eosinophil chemotaxis in response to eotaxin and MCP4, but addition of the antagonist to both upper and lower chambers reduced chemotaxis to baseline levels. The reason for this donor-to-donor variability is currently uncertain (see Discussion). These data demonstrate, however, that Met-Ckβ7 is a potent inhibitor of eosinophil chemotaxis induced with CCR3 ligands, with low nanomolar concentrations being sufficient to completely abrogate cell migration.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Met-Ckβ7 inhibits eotaxin-induced chemotaxis of human eosinophils from three donors. Chemotaxis assays were performed for 3 h with a range of eotaxin (or Met-Ckβ7) concentrations. In some experiments, Met-Ckβ7 was added at 1 or 10 nM to the bottom chamber, or the top and bottom chambers, of the chemotaxis well as indicated in the key. Chemotaxis index is calculated as the ratio of cells migrated in the test sample compared with cells migrated in buffer alone. A, Donor A; (B) donor B; (C) donor C.

Unmodified MIP4 also antagonizes signaling through CCR3

Data shown above using radio-iodinated ligands suggested that unmodified MIP4 also demonstrates some ability to displace ligands from human CCR3. Therefore, we next tested whether the commercially available MIP4 protein, with an N-terminal alanine residue and LKLNA at the C terminus, was able to affect CCR3 signaling. No CCR3-mediated Ca²⁺ flux was detectable with up to 250 nM MIP4, but, surprisingly, this protein exhibited CCR3 antagonistic activity in Ca²⁺ flux assays with fura-2 loaded HOS-CCR3 cells (Fig. 8⇓) or eosinophils (Fig. 9⇓). However, it is not as potent as Met-Ckβ7 (∼5- to 10-fold less active), suggesting that the amino acid sequence differences in Met-Ckβ7 conspire to amplify a property present in the natural protein. Thus, whereas 10 nM Met-Ckβ7 completely abrogates Ca²⁺ fluxes induced into eosinophils by 1 nM MCP4, only 50% inhibition is seen with MIP4 at this concentration (Fig. 9⇓B). However, at higher concentrations (100 nM), MIP4 is able to completely abrogate signaling induced with 1 nM MCP4. Similar results were obtained with abrogation of 1 nM eotaxin signaling (Fig. 9⇓A), with complete inhibition observed with 50 nM Met-Ckβ7 but ∼10% of the signal remaining in the presence of 100 nM MIP4.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Met-Ckβ7 inhibits MCP4-induced chemotaxis of human eosinophils from three donors. Chemotaxis assays were performed for 3 h with a range of MCP4 (or Met-Ckβ7) concentrations. In some experiments, Met-Ckβ7 was added at 1 or 10 nM to the bottom chamber, or the top and bottom chambers, of the chemotaxis well as indicated in the key. Chemotaxis index is calculated as the ratio of cells migrated in the test sample compared with cells migrated in buffer alone. A, Donor A; (B) donor B; (C) donor C.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8.

Unmodified MIP4 inhibits eotaxin-induced signaling through CCR3. Fura-2-loaded HOS-hCCR3 cells stimulated at 37°C with 25 nM eotaxin in the absence (A) or presence (B) of 250 nM MIP4. Fluorescence emission is recorded every 0.1 s for 100 s (340 nm (λ_ex); 500 nm (λ_em)).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 9.

MIP4 is a less potent antagonist of CCR3 signaling into eosinophils than Met-Ckβ7. The peak of calcium ion flux into human eosinophils (detected by fura-2 fluorescence), induced by a set amount of agonist in the presence of a range of MIP4 or Met-Ckβ7 concentrations, is represented as a percentage of the flux induced in the absence of antagonist. Arrows indicate where no Ca²⁺ flux was detectable. Agonist used is (A) 1 nM eotaxin, or (B) 1 nM MCP4, as indicated in the key.

Chemotaxis inhibition by MIP4

To extend these observations, we sought to determine the potency of MIP4 in inhibiting eosinophil chemotaxis using MCP4 or eotaxin as agonists, the two most potent CCR3 ligands in our assays. As above, maximal chemotaxis with these agonists was observed between 1 and 10 nM for both donors tested (Fig. 10⇓). When 10 nM MIP4 was added to the top and bottom chambers of the chemotaxis assay plate, then eosinophil chemotaxis induced by 1 nM MCP4 or eotaxin is reduced to near baseline levels, and chemotaxis induced by 10 nM of the agonists was significantly reduced. MIP4 was unable to induce chemotaxis when present in the lower well at concentrations ranging from 0.1 to 100 nM (data not shown). These results corroborate the observations obtained in the Ca²⁺ flux signaling assays and show that MIP4 is able to inhibit CCR3 function, but that it is less potent than the Met-Ckβ7. The physiological relevance of this observation, with respect to the biological function of MIP4, is discussed below.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 10.

MIP4 inhibits eosinophil chemotaxis induced by MCP4 or eotaxin. Chemotaxis assays were performed with eosinophils from two donors (A and B) for 3 h with a range of MCP4 or eotaxin concentrations. In some experiments, MIP4 was added at 10 nM to the top and bottom chambers of the chemotaxis well as indicated in the key. Chemotaxis index is calculated as the ratio of cells migrated in the test sample compared with cells migrated in buffer alone.