Cloning and Characterization of the Guinea Pig Eosinophil Eotaxin Receptor, C-C Chemokine Receptor-3: Blockade Using a Monoclonal Antibody In Vivo

Materials

The Dunkin-Hartley guinea pig liver genomic DNA library in EMBL3A was purchased from Stratagene Cloning Systems (La Jolla, CA). ¹²⁵I-labeled human monocyte chemotactic protein (hMCP)-3 and ¹²⁵I-labeled hRANTES (each 2200 Ci/mmol) were from New England Nuclear (Boston, MA). Cell culture media and reagents including RPMI 1640, HEPES, FCS, sodium bicarbonate, sodium pyruvate, 2-ME, and geneticin (G418) were obtained from Life Technologies (Paisley, U.K.). Percoll, BSA, and general laboratory reagents were obtained from Sigma (Poole, U.K.). Restriction enzymes were purchased from Boehringer Mannheim (Indianapolis, IN). Human IL-8 was produced in Escherichia coli as described (40). Human eotaxin was purified from Sf-9 insect cell culture supernatants infected with a recombinant eotaxin baculovirus obtained using the Bac-to-Bac baculovirus expression system (Life Technologies). All other chemokines other than those below were from Peprotech EC, (London, U.K.). The following were generous gifts. Synthetic guinea pig eotaxin was from B. Guarino, G. Andrews, and H. Showell (Pfizer, Groton, CT). Recombinant human C5a (rhC5a) was from Dr. J. van Oostrum (Ciba-Geigy, Summit, NJ). ENA-2 anti-human E-selectin F(ab′)₂ fragments were from Dr. M. Robinson (Celltech Therapeutics, Slough, U.K.). The murine pre-B cell line L1.2 was from Dr. E. Butcher (Stanford University, Stanford, CA).

Isolation of guinea pig CCR3 genomic bacteriophage clones

Guinea pig CCR3 DNA clones were isolated by screening 5 × 10⁵ plaque-forming units of the guinea pig genomic DNA library using ³²P-labeled human (27) and mouse (29) CCR3 cDNA sequences as probes. Plaques showing hybridization to both probes under high stringency wash conditions were purified, and restriction fragments of phage DNA with homology to the CCR3 probes were subcloned into pBluescript SK⁺ (Stratagene) and sequenced (Sequenase; Amersham Life Science, Cleveland, OH). A 1.1-kb single-exon open reading frame was identified, subcloned by PCR into the expression vector pcDNA3 (Invitrogen Corp., San Diego, CA) incorporating a 5′ BamHI and a 3′ XbaI linker, and resequenced to confirm PCR fidelity. The PCR N- and C-terminal primers for this reaction were as follows: GPCKR3/1, 5′-CGG GAT CCC GCC AGT GGC ATG GCG C-3′, and GPCKR3/2, 5′-GCT CTA GAG CCC TCG AGG TCG ACC-3′.

Generation of a guinea pig CCR3-expressing stable cell line

Twenty micrograms of the CCR3/pcDNA3 construct were linearized by digestion with Bsm1 (New England Biolabs, Beverly, MA) and used to transfect L1.2 cells as follows. Twenty five million cells were washed twice and resuspended in 0.8 ml of PBS. The cells were incubated for 10 min at room temperature with the linearized CCR3/pcDNA3 construct DNA and transferred to a 0.4-cm cuvette, and a single electroporation pulse was applied at 250 V, 960 μF. Electroporated cells were incubated for 10 min at room temperature and transferred to culture at 37°C in RPMI supplemented with 10% FCS. Geneticin (G418) was added to a final concentration of 800 μg/ml 48 h posttransfection, and the cells were plated in 96-well plates at 25,000 cells/well. After 2–3 wk under drug selection, geneticin-resistant cells were assessed for their ability to migrate to 100 nM guinea pig eotaxin in a Transwell (Costar, Cambridge MA) chemotaxis assay (8). Lines with robust migration were cloned by limiting dilution and rescreened for their ability to bind ¹²⁵I-labeled hMCP-3 (described below) and their ability to migrate to guinea pig eotaxin.

Purification of guinea pig leukocytes

Guinea pigs (400–450 g) were humanely killed by the administration of an overdose of sodium pentobarbitone i.p. (May and Baker, Oxford, U.K.), and blood was taken by cardiac puncture into 10 mM EDTA. For some experiments, guinea pigs were pretreated with 1 μg/kg recombinant human IL-5 i.v. 90 min before euthanasia. The whole blood was spun at 260 × g for 20 min and the plasma discarded. Dextran (5 ml; 9% in normal saline) and 20 ml of normal saline were added per 15 ml of the buffy coat/erythrocyte cell pellet and gently mixed, and the erythrocytes were allowed to sediment over 30 min. The leukocyte-rich suspension was harvested, and the leukocytes were pelleted by centrifugation. The mononuclear cells and polymorphonuclear leukocytes (PMNLs, comprising eosinophils and neutrophils) were separated over a two-layer 70%/80% discontinuous PBS/Percoll gradient centrifuged at 400 × g for 25 min. Contaminating erythrocytes were removed by hypotonic shock lysis (41). The cells were washed in the appropriate assay buffer and counted using Kimura’s stain.

Chemotaxis assays

Chemotaxis experiments with L1.2 cells or L1.2 receptor transfectant cell lines were performed using Transwell tissue culture inserts (Costar) as described (8).

For guinea pig leukocyte chemotaxis, PMNLs were prepared from IL-5-treated animals as above, and the chemotaxis assay was performed as described (18). Briefly, aliquots of 5 × 10⁵ PMNLs (containing 18% ± 3.2% (SD) eosinophils) were pretreated in 100 μl of buffer (RPMI 1640 containing 20 mM HEPES and 0.25% BSA) with or without Ab for 10 min at room temperature. The cells were transferred to Transwell inserts within 24-well tissue culture plates containing 400 μl of buffer, guinea pig eotaxin, or rhC5a per well. The plates were incubated for 1 h at 37°C, and the chemotactic responses were determined using FACS analysis to measure the eosinophil recruitment to the lower chamber, as described previously (18).

Radiolabeled ligand binding assays

Chemokine binding studies of transfected cells using 0.1 nM ¹²⁵I-labeled hMCP-3 or 0.1 nM ¹²⁵I-labeled hRANTES as the ligand were performed using a modified method previously reported (8). Briefly, cells were washed once in PBS and resuspended in buffer (1 × 10⁷ cells/ml in PBS with 50 mM HEPES, pH 7.5; 1 mM CaCl₂; 5 mM MgCl₂; 0.5% BSA; and 0.05% azide). Aliquots of 50 μl (5 × 10⁵ cells) were dispensed into 1.5-ml Eppendorf tubes, followed by the addition of cold competitor and radiolabeled chemokines as indicated. The final reaction volume was 200 μl. Nonspecific binding was determined in the presence of 500 nM unlabeled MCP-3. After a 60-min incubation at 37°C, the cells were washed twice with 200 μl of buffer containing 0.5 M NaCl. Cell pellets were then counted. The data are presented as the percentage of specific binding calculated according to the formula 100 × (S − B)/(T − B), where S is the radioactivity of the sample, B is background nonspecific binding, and T is total binding without competitors. Duplicates were used throughout the experiments, and the SDs were always less than 10% of the mean. All experiments were repeated at least three times. Curve fit was calculated by KaleidaGraph software (Synergy Software, Reading, PA). In some experiments, binding of ¹²⁵I-labeled hMCP-3 and ¹²⁵I-labeled hRANTES was performed in the RPMI-based buffer as used for chemotaxis assays (see above).

Binding assays using ¹²⁵I-labeled guinea pig eotaxin were performed using a modified previously described method (41). Briefly, guinea pig eotaxin was iodinated using the Iodogen (Pierce & Warriner, Chester, U.K.) method as described (6), with a specific activity of 195 Ci/mmol. Transfectants were incubated with 0.3 nM ¹²⁵I-labeled guinea pig eotaxin in RPMI 1640 + 20 mM HEPES + 0.5% BSA + 0.05% sodium azide (pH 7.4) and varying concentrations of unlabeled chemokines (in a final volume of 50 μl) at room temperature for 60 min. Fifty microliters of assay buffer containing NaCl was added to a final concentration of 0.5 M NaCl, and the samples were mixed and layered onto tubes containing 150 μl of silicone oil. The cells were pelleted through the oil by centrifugation (13,000 × g for 5 min at 15°C). The cell pellet and supernatant were counted separately in a Canberra Packard Cobra 5010 gamma counter (Canberra Packard, Pangebourne, U.K.). Data are presented without the subtraction of nonspecific binding. Curve fits and calculation of ligand K_d were performed using MacLigand software (NIH version 4.92).

Measurement of cytosolic calcium in purified guinea pig eosinophils

Guinea pig peritoneal eosinophils were elicited in response to horse serum and purified over discontinuous Percoll gradients as previously described (5). Eosinophils (1 × 10⁷/ml) were loaded with 1 μM fura 2 as previously described (41), washed in PBS + 0.1% BSA, and resuspended at a final concentration of 1 × 10⁶ cells/ml in PBS + 0.25% BSA + 10 mM HEPES + 10 mM glucose. Aliquots of cells (2 ml) were dispensed into quartz cuvettes and equilibrated with 1 mM calcium at 37°C for 10 min before use. Changes in fluorescence were measured in a Perkin-Elmer (Norwalk, CT) LS-50 fluorescence spectrophotometer as previously described (41). Calculation of intracellular free calcium was derived from the fluorescence spectra (excitation wavelengths, 340 and 380 nm; emission wavelength, 510 nm) in accordance with established methods (41).

Generation of anti-guinea pig CCR3 receptor mAbs and flow cytometry

mAbs reactive with guinea pig CCR3 were generated by immunizing C57BL6 mice with 1–2 × 10⁷ receptor transfectants i.p. six to eight times at 2-wk intervals. The final immunization was given i.v. Three days later, the spleen was removed and the splenic lymphocytes were fused with the SP2/0 cell line as described (27). Candidate anti-guinea pig CCR3 mAbs were initially identified by flow cytometry. Hybridoma supernatants were incubated with guinea pig CCR3 receptor transfectants and untransfected L1.2 cells, and bound Ab were detected using FITC-conjugated F(ab′)₂ goat anti-mouse IgG (Jackson ImmunoResearch Laboratories, West Grove, PA). mAbs showing selectivity for guinea pig CCR3 transfectants were cloned by limiting dilution and rescreened against a panel of L1.2 cell chemokine receptor transfectants and guinea pig leukocytes. For flow cytometry of guinea pig leukocytes, 5 × 10⁵ cells were incubated with mAbs in 100 μl of buffer (PBS containing 10 mM HEPES, 10 mM glucose, and 0.1% BSA) for 60 min on ice, washed twice (220 × g for 7 min), and incubated for 30 min on ice with FITC-conjugated F(ab′)₂ goat anti-mouse IgG. The cells were washed once and resuspended in PBS before FACS analysis.

Generation of F(ab′)₂ fragments of mAb 2A8

For use in vivo, F(ab′)₂ fragments of mAb 2A8 were generated by TSD BioServices (Germantown, NY) as follows. The 2A8 (IgG2a) Ab was first purified from cell culture supernatant by passage over a protein A-Sepharose column (Pharmacia) and elution of Ab with 20 mM sodium acetate, pH 4.0. F(ab′)₂ fragments were prepared from the Ab by digestion with immobilized pepsin (Pierce and Warriner, Chester, U.K.) and purified from Fc by protein A chromatography. The final product contained <3.0 endotoxin units (EU) of endotoxin/mg.

Inhibition of ¹¹¹In-labeled eosinophil accumulation by mAb pretreatment in vivo

Eosinophils from four donor animals were purified over discontinuous Percoll gradients, pooled, and labeled with ¹¹¹InCl₃ as previously described (5). The radiolabeled eosinophils were divided into two aliquots and pretreated with either a control F(ab′)₂ Ab (ENA2, anti-human E selectin; 100 μg/ml) or 100 μg/ml 2A8 F(ab′)₂ in buffer (HBSS + 0.25% BSA + 30 mM HEPES) for 10 min at room temperature. The cells were washed and injected i.v. via an ear vein into sedated recipient animals (2.5 × 10⁶ cells/animal, with four animals receiving control F(ab′)₂ Ab-treated cells and four animals receiving anti-guinea pig CCR3 2A8 F(ab′)₂ Ab-treated cells). Five minutes later, test agents (in 100 μl) were injected intradermally into the shaved dorsal skin of the recipients in a randomized pattern. After 2 h, the recipients were humanely killed. Blood was taken by cardiac puncture for calculation of circulating eosinophil numbers, the back skin of the animals was removed, and the skin sites were punched out and counted in a Cobra 5010 gamma counter. Results are expressed as the calculated number of ¹¹¹In-labeled eosinophils accumulating per site according to established protocols (5, 6).

Cloning of the guinea pig CCR3 receptor

Two clones from the guinea pig genomic library showing homology to both human and murine CCR3 probe sequences were selected for further investigation and were found to contain the same sequence; one clone alone contained the complete open reading frame. Analysis of the open reading frame showed it to encode a G protein-coupled seven-transmembrane receptor with 67 and 69% homologies to human and murine CCR3, respectively (Fig. 1⇓). The predicted protein sequence contained the familiar amino acid motif, DRYLAIV, a highly conserved chemokine receptor amino acid sequence across both receptor types and animal species. Similarly to human CCR3, and in contrast to many other human chemokine receptors (including CCR1, CCR2, CCR4, CXCR1, and CXCR2), the N-terminus of the guinea pig receptor lacks sites for N-linked glycosylation. In common with other C-C and C-X-C chemokine receptors, the sequence contains a serine- and a threonine-rich C terminus, sites that are likely to be important in receptor regulation through phosphorylation by G protein receptor kinases (41, 42, 43). RT-PCR using specific primers for the guinea pig CCR3 sequence amplified a band of predicted size from mRNA purified from guinea pig peritoneal eosinophils (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Sequence of the guinea pig eotaxin receptor, guinea pig CCR3. A 1.1-kb sequence was isolated from a guinea pig genomic DNA library encoding a 358-amino acid protein with 67 and 69% homology to human and murine CCR3, respectively. The predicted protein sequence of the open reading frame is shown above, together with the sequences of human and murine CCR3, with human CCR1 and CCR2b for comparison. Residues showing homology with these comparative sequences are shown in boldface type. The sequence has been deposited in the GenBank database and assigned accession number AF060698.

Construction of guinea pig CCR3 stable transfectants and characterization of receptor ligand binding

Stable guinea pig CCR3 transfectants were generated and assessed for their ability to migrate in response to guinea pig eotaxin. Two cell lines showing robust chemotaxis to guinea pig eotaxin (2A1 and 4B4) were further characterized for their ability to bind radiolabeled chemokines. Cell lines 2A1 and 4B4 showed identical ligand binding profiles, and therefore only data from the 4B4 line are shown here.

The 4B4 cells bound ¹²⁵I-labeled guinea pig eotaxin (Fig. 2⇓a), which was displaced by unlabeled guinea pig eotaxin in a concentration-dependent fashion. In previous studies using guinea pig eosinophils, binding of ¹²⁵I-labeled guinea pig eotaxin was displaced both by hMCP-3 and hRANTES as well as guinea pig eotaxin (44). In keeping with these data, 4B4 cells, but not their untransfected parent cell line, also bound ¹²⁵I-labeled hMCP-3 (Fig. 2⇓, a and b) and ¹²⁵I-labeled hRANTES (Fig. 2⇓b). Scatchard analysis of the binding data indicated dissociation constants (K_d) of 3.0 nM for hMCP-3 (Fig. 2⇓a) and 4.5 nM for guinea pig eotaxin (mean of three experiments; data not shown), with a mean receptor density of 47,000 receptors per cell (Fig. 2⇓a).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

a, The ligand binding of the guinea pig CCR3 stable transfectant cell line 4B4. i, Transfectants were incubated with 0.3 nM ¹²⁵I-labeled guinea pig eotaxin and varying concentrations of unlabeled guinea pig eotaxin. After washing, the transfectant cell pellets were counted in a gamma counter, and the results are presented as a percentage of maximum binding. All experiments were performed in triplicate, and data shown are the means ± SEM of three or four experiments. ** signifies p < 0.01, and *** signifies p < 0.001 (both significant competitions; ANOVA and Bonferroni’s post test). ii, Competition of 0.1 nM ¹²⁵I-labeled hMCP-3 binding by increasing concentrations of unlabeled ligand. Inset, Scatchard transformation of the data, calculation of which showed a receptor density of 47,000 receptors/cell with a K_d of 3.0 nM. Data are from a representative experiment (duplicates ± SD) of two to four experiments. b, Competition of binding of ¹²⁵I-labeled guinea pig eotaxin, ¹²⁵I-labeled hMCP-3, and ¹²⁵I-labeled hRANTES to 4B4 cell transfectants. Transfectants were incubated with 0.3 nM ¹²⁵I-labeled guinea pig eotaxin (i) or 0.1-nM concentrations of ¹²⁵I-labeled hMCP-3 (ii) and ¹²⁵I-labeled hRANTES (iii) and 100-nM concentrations of the unlabeled chemokines indicated. After washing, the transfectant cell pellets were counted in a gamma counter; ligand binding is displayed as a percentage of maximum specific binding. i, Data are the means from three or four experiments ± SEM. ii and iii, Data are from a representative experiment (duplicates ± SD) of two to four experiments. c, Chemotactic responses of the guinea pig CCR3 transfectant cell line. 4B4 cell transfectants were placed in Transwell cell culture inserts and incubated at 37°C in 24-well plates containing buffer or varying concentrations of guinea pig, human, and murine eotaxin; hMCP-3; hMCP-4; and human IL-8 as indicated. The chemotactic responses of the transfectants in response to these chemokines were measured by flow cytometry. Guinea pig eotaxin alone induced a chemotactic response that was maximal at 100 nM. No other tested agonist induced a chemotactic response, and these symbols overlap on the baseline. Data shown are from a representative experiment ± SD, which was repeated three times.

Binding of ¹²⁵I-labeled guinea pig eotaxin, ¹²⁵I-labeled hMCP-3, and ¹²⁵I-labeled hRANTES was displaced by an excess of unlabeled guinea pig eotaxin or hMCP-3 (Fig. 2⇑b). hRANTES failed to compete for ¹²⁵I-labeled guinea pig eotaxin or ¹²⁵I-labeled hMCP-3 binding in these experiments using the buffers described. However, when ¹²⁵I-labeled hMCP-3 binding studies were performed in the RPMI buffer used in transfectant cell chemotaxis assays, 100 nM hRANTES competed for 53% of 0.1 nM ¹²⁵I-labeled hMCP-3 binding to 4B4 cells (data not shown).

To evaluate ligand specificity further, we tested the chemotactic responses of the 4B4 cell line to the chemokines used in the radiolabeled ligand binding assays. Human IL-8 was included as a negative control. Untransfected L1.2 cells were unresponsive to all chemokines tested (Ref. 27 and data not shown). Fig. 2⇑c shows that only guinea pig eotaxin was able to induce a chemotactic response in the receptor transfectants, producing a characteristic bell-shaped dose-response profile with a peak response at 100 nM. This profile is similar to that found for human eotaxin acting on human CCR3 transfectants (27). Despite the similar K_ds of hMCP-3 and guinea pig eotaxin for guinea pig CCR3, hMCP-3 was unable to induce a chemotactic signal at concentrations up to 500 nM.

Development of mAbs to guinea pig CCR3

Previous studies with a blocking mAb to human CCR3 have demonstrated the importance of this receptor in chemokine-induced chemotactic responses of eosinophils in vitro (12). This mAb does not cross-react with guinea pig CCR3 on either transfectants or guinea pig eosinophils, and therefore a panel of anti-guinea pig CCR3 mAbs was generated to explore the in vivo relevance of CCR3 in well characterized guinea pig models of allergic inflammation (6, 7, 10). Following the first fusion, 1000 hybridomas were generated, and two Abs, 1C3 and 2A8, were isolated. Both mAbs bound to guinea pig CCR3 transfectants and showed minor cross-reactivity with the human CCR3 transfectant cell line (data not shown). Neither Ab stained a murine CCR3 transfectant or any other human CCR (1, 2, 4, 5) or CXCR (1, 2, 3) transfectant tested (data not shown). The specificity of Abs 1C3 and 2A8 for guinea pig leukocytes was therefore investigated, and Fig. 3⇓ shows that these Abs bound selectively to guinea pig eosinophils in PMNL populations purified from blood. There was no detectable binding of either 1C3 or 2A8 to guinea pig blood neutrophils (Fig. 3⇓), monocytes, or lymphocytes as determined by flow cytometry (data not shown), in keeping with the previously described expression patterns of human CCR3 (27). A second fusion and round of hybridoma formation resulted in the identification of four further anti-guinea pig CCR3 mAbs with binding characteristics identical to those of 1C3 and 2A8 (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Binding of anti-guinea pig CCR3 mAbs 1C3 and 2A8 to guinea pig leukocytes. Guinea pig PMNLs (comprising eosinophils and neutrophils) were incubated with purified 1C3 and 2A8 IgG (10 μg/ml). The cells were washed and incubated with FITC-conjugated goat anti-mouse secondary Abs. Following one further wash, the presence of bound Ab was detected by flow cytometry. A, Separation of guinea pig eosinophils and neutrophils in PMNL populations according to their light-scatter characteristics on FACS analysis. Both Abs bound to eosinophils (B), and neither Ab bound to neutrophils (C).

Blockade of ligand binding and chemotactic responses of guinea pig CCR3 transfectants by anti-guinea pig CCR3 mAbs

To investigate the potential of the anti-guinea pig CCR3 mAbs to block ligand binding, the 4B4 cell line stable guinea pig CCR3 transfectants were pretreated with mAb 1C3 or 2A8, and the subsequent binding of ¹²⁵I-labeled hMCP-3 was determined. Pretreatment of transfectants with mAb 2A8 blocked ¹²⁵I-labeled hMCP-3 binding to transfectants in a concentration-dependent manner; in contrast, mAb 1C3 did not inhibit ligand binding at any concentration tested (Fig. 4⇓). The binding characteristics of guinea pig CCR3 deduced from Fig. 2⇑, a and b, and Ref. 44 predict that guinea pig eotaxin and hMCP-3 share a common binding site on guinea pig CCR3. Accordingly, pretreatment with 2A8 also blocked chemotaxis of the guinea pig CCR3 transfectants to guinea pig eotaxin in a concentration-dependent manner, at similar concentrations to those needed to inhibit ¹²⁵I-labeled hMCP-3 binding (Fig. 4⇓). Pretreatment of cells with the mAb 1C3 did not inhibit chemotaxis.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Inhibition of 4B4 cell transfectant ligand binding and chemotaxis by pretreatment with mAb 2A8. In binding studies, Ab-pretreated cells were incubated with 0.1 nM ¹²⁵I-labeled hMCP-3, and the cell pellets were counted in a gamma counter. In chemotaxis assays, Ab-pretreated cells were placed in Transwell cell culture inserts and incubated at 37°C in 24-well plates containing guinea pig eotaxin. The chemotactic responses of the transfectants were measured by flow cytometry (cells/30 s). The Ab 2A8, but not mAb 1C3, inhibited both ligand binding and chemotaxis at similar concentrations. Data shown are representative of at least three (inhibition of chemotaxis) and two (inhibition of radioligand binding) experiments.

Blockade of guinea pig eotaxin signaling on guinea pig eosinophils by mAb 2A8

mAb 2A8 bound specifically to both guinea pig CCR3 transfectants and guinea pig eosinophils and blocked both ligand binding and chemotactic responses of guinea pig CCR3 transfectants. Therefore, the ability of 2A8 to block the in vitro responses of guinea pig eosinophils to guinea pig eotaxin was investigated using assays of chemotaxis and intracellular calcium flux. Pretreatment with 2A8 (10 μg/ml) of PMNLs purified from the blood of IL-5-treated guinea pigs caused a potent inhibition of eosinophil chemotaxis in response to guinea pig eotaxin but had no effect on the eosinophil chemotactic response to rhC5a (Fig. 5⇓). In the same cell populations, guinea pig neutrophils showed a chemotactic response to rhC5a, but not to guinea pig eotaxin, and this was not inhibited by 2A8 pretreatment (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Inhibition of guinea pig eosinophil chemotactic and cytosolic calcium responses to eotaxin by pretreatment with mAb 2A8. Guinea pig PMNLs purified from the blood of IL-5-treated donor animals were pretreated for 10 min with 2A8 (10 μg/ml) or buffer and were placed in Transwell cell culture inserts, which were incubated for 1 h at 37°C in 24-well plates containing guinea pig eotaxin (30 nM) or rhC5a (3 nM). The chemotactic responses of the leukocytes in response to these agonists were measured by flow cytometry (top). Pretreatment of the cells with 2A8 inhibited the chemotactic response to guinea pig eotaxin (*p < 0.05, ANOVA and Bonferroni’s post test) but not to rhC5a. Data shown are the means of four experiments ± SEM. In the lower panels, purified eosinophils were loaded with fura 2, and their intracellular calcium responses to guinea pig eotaxin were determined in the presence of buffer, 2A8 (12 μg/ml), and 1C3 (12 μg/ml). mAb 2A8 completely inhibited the guinea pig eosinophil calcium response to eotaxin.

To investigate further the blocking efficacy of 2A8 on guinea pig CCR3-mediated signaling in guinea pig eosinophils, elicited eosinophils were loaded with fura 2, and their intracellular calcium response to guinea pig eotaxin was tested. Fig. 5⇑ shows that mAb 2A8 (12 μg/ml) completely blocked the calcium signal in response to 3 nM guinea pig eotaxin. In contrast, mAb 1C3 had no effect on the calcium flux induced by guinea pig eotaxin at the same concentration.

Blockade of guinea pig eosinophil CCR3 signaling by F(ab′)₂ fragments of mAb 2A8

In pilot experiments, i.v. administration of mAb 2A8 (1 mg/kg) to guinea pigs caused a rapid clearance of eosinophils from the circulation (data not shown), possibly mediated through nonspecific interactions of the Fc portion of the mAb. Therefore, for the use of 2A8 as a blocking anti-guinea pig CCR3 Ab in vivo, F(ab′)₂ fragments of 2A8 were prepared. Fig. 6⇓ shows that 2A8 F(ab′)₂ fragments (30 μg/ml) bound to eosinophils similarly to whole Ab (10 μg/ml). 2A8 F(ab′)₂ fragments (30 μg/ml) retained their ability to inhibit the eosinophil cytosolic calcium response to 3 nM guinea pig eotaxin but did not inhibit responses to rhC5a, used as a control to show that the eosinophil responses to other ligands were not blocked (Fig. 6⇓). Surprisingly, the results shown in Fig. 6⇓ also suggest that guinea pig eotaxin partially desensitized the calcium response to rhC5a and that the blockade of eotaxin-induced responses by 2A8 pretreatment prevented this heterologous desensitization of the rhC5a response. Previous studies have found little evidence of heterologous desensitization of chemoattractant signaling in human eosinophils (45). This is in contrast to the patterns of heterologous desensitization observed in human neutrophils, in which C5a is a potent desensitizer of responses to chemokines, but C5a responses are not readily desensitized by chemokines (41, 45, 46).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

2A8 F(ab′)₂ binds to guinea pig eosinophils and inhibits cytosolic calcium responses induced by guinea pig eotaxin. 2A8 IgG (10 μg/ml) and F(ab′)₂ 2A8 (10 and 30 μg/ml) were incubated with guinea pig PMNLs for 60 min on ice, the cells were washed, and the presence of bound Ab was detected by flow cytometry after incubation of the cells with FITC-conjugated goat anti-mouse secondary Abs. Eosinophils and neutrophils were separated according to their light scatter characteristics (Fig. 3⇑). 2A8 bound, as whole molecules and F(ab′)₂ fragments, to guinea pig eosinophils but not to neutrophils. Binding of F(ab′)₂ fragments at a concentration of 30 μg/ml was similar to that obtained for whole IgG binding at a concentration of 10 μg/ml. Purified eosinophils were loaded with fura 2, and their calcium response to guinea pig eotaxin was determined in the presence of F(ab′)₂ fragments of 2A8 (0, 10, and 30 μg/ml). There was a concentration-dependent inhibition of the intracellular calcium response to eotaxin by pretreatment of cells with F(ab′)₂ fragments of 2A8 that was associated with an apparent enhancement of the subsequent calcium response to rhC5a used as a positive control for cell responses.

Blockade of ¹¹¹In-labeled eosinophil accumulation in vivo by mAb 2A8

In contrast to unfragmented 2A8 IgG, F(ab′)₂ 2A8 did not result in any clearance of circulating eosinophils when administered i.v. to guinea pigs at a dose of 1 mg/kg (data not shown). Eosinophils were labeled with ¹¹¹In and pretreated with 100 μg/ml concentrations of either a control F(ab′)₂ mAb (ENA-2, anti-human E selectin) or F(ab′)₂ 2A8 as described in Materials and Methods. The cells were washed to remove unbound Ab and injected i.v. into recipient animals. Buffer, guinea pig eotaxin, or leukotriene B₄ (LTB₄) was then injected intradermally into the recipients, and the local accumulation of circulating radiolabeled eosinophils was allowed to proceed for 2 h. Fig. 7⇓ shows that pretreatment of eosinophils with F(ab′)₂ fragments of 2A8 significantly inhibited the dermal accumulation of eosinophils in response to eotaxin, but not in response to LTB₄. There was no significant difference in circulating ¹¹¹In-labeled eosinophil numbers between control and 2A8-treated groups (14.7% ± 2.35% (SD) for ENA-2-pretreated cell recipients vs 10.1% ± 4.28% for 2A8-pretreated cell recipients).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Inhibition of eosinophil accumulation in vivo by pretreatment with F(ab′)₂ fragments of 2A8. Purified eosinophils were labeled with ¹¹¹In and pretreated with 100 μg/ml F(ab′)₂ fragments of 2A8 or control Ab. After washing to remove unbound Ab, the cells were injected i.v. into eight recipients (four per Ab). Duplicate injections of buffer, guinea pig eotaxin, or LTB₄ were performed intradermally in the recipients in a randomized pattern. Two hours later, the animals were humanely killed, and the skin sites were punched out and counted in a gamma counter to determine the number of ¹¹¹In-labeled eosinophils accumulating per skin site. * signifies p < 0.05 (significant difference between control F(ab′)₂ fragment-treated cells and 2A8 F(ab′)₂ fragment-treated cells; ANOVA and Bonferroni’s post test).