INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Eotaxin, a C-C chemokine which was discovered as a selective chemoattractant for eosinophils in bronchoalveolar lavage fluid (BALF) obtained from an experimental model of allergen exposure of sensitized guinea pigs (1), induces selective pulmonary and intradermal eosinophil recruitment, indicating a role for eotaxin for eosinophil homing and tissue recruitment (1). When administered to the airways of mice, eotaxin causes no significant increase in lung eosinophils, but in interleukin-5 (IL-5) transgenic mice, eotaxin induces a large eosinophilic response in the airways (5, 6). This observation pointed to the important cooperative effects between eotaxin and IL-5 which is an important mediator of eosinophil growth, differentiation, and activation on eosinophil mobilization (7- 10). Eotaxin administered to IL-5 transgenic mice also induced activation of lung eosinophils, together with bronchial hyperresponsiveness (BHR) (6). Other mediators of allergic inflammation have also been implicated, namely the cysteinyl-leukotrienes (11). These lipid mediators derived from the action of the 5-lipoxygenase enzyme on arachidonic acid are also active in recruiting eosinophils to the lungs (12, 13), and have been shown to be involved in the BHR induced by allergen exposure in sensitized mice and asthmatics (14, 15). In addition, leukotriene (LT) receptor antagonists such as pranlukast have been shown to reduce eosinophil numbers in bronchial biopsies of patient with asthma (16). Eotaxin has recently been shown to increase cytokine-mediated release of leukotriene C₄ (LTC₄) from eosinophils in vitro (17). We speculated that eotaxin could induce eosinophil recruitment and BHR through the elaboration of cysteinyl-leukotrienes. Because this has not been ascertained before, we therefore determined whether these effects observed in IL-5 transgenic mice could be prevented by an inhibitor of 5-lipoxygenase, SB210661, or by an antagonist of the cysteinyl-leukotriene receptor, pranlukast.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals

Eight- to 10-wk-old CBA/Ca mice were purchased from Harlan-Olac (Bicester, UK). Age-matched CBA/Ca mice overexpressing the murine IL-5 gene (Tg1 mice) (18) were obtained from Glaxo Wellcome (Stevenage, UK), were bred and housed in air-conditioned animal facilities (temperature, 20 ± 2° C; relative humidity, 55 ± 10%) and fed a standard laboratory diet and tap water ad libitum.

Protocol

Mice were anesthetized with an intraperitoneal injection of 150 to 200 µl of anesthetic solution, containing 0.5 ml midazolam (Roche Products Ltd., Welwyn Garden City, UK) and 0.5 ml Hypnorm (0.315 mg/ ml of fentanyl citrate and 10 mg/ml of fluanisone; Janssen Pharmaceuticals Ltd., Wantage, UK). Mice were intubated with an angled 18-gauge blunt needle, through which recombinant murine eotaxin (5 µg dissolved in 10 ml vehicle) or 10 ml control vehicle (10 mM phosphate-buffered saline [PBS]/0.1% bovine serum albumin [BSA], pH 7.4) was instilled. In order to study the role of leukotrienes, we used a 5-lipoxygenase inhibitor, SB210661 (19), and the cysteinyl-leukotriene receptor antagonist, pranlukast (20). SB210661 (40 mg/kg dissolved in 1% polyethylene glycol PEG200), pranlukast (40 mg/kg dissolved in 0.25M NaOH in 1% PEG200), or respective vehicles were administered by gavage (in 50-ml delivery dose) 30 min before eotaxin instillation. Airway responsiveness to acetylcholine chloride (Ach) was measured 24 h after instillation of eotaxin (R&D Systems Ltd, Oxford, UK).

We first studied the effect of eotaxin in IL-5 transgenic mice by comparing the effect of intratracheal administration of diluent for eotaxin (Group A, n = 8) to that of intratracheal administration of eotaxin alone (Group B, n = 6).

For the effect of SB210661, the corresponding groups (n = 6 in each group) were examined: Group C: pretreatment with SB210661 followed by intratracheal administration of diluent; Group D: pretreatment with vehicle for SB210661 followed by intratracheal administration of eotaxin; Group E: pretreatment with SB210661 followed by intratracheal administration of eotaxin.

In order to examine the effect of pranlukast on eotaxin-induced effects, we studied the following groups (n = 6 in each group): Group F: pretreatment with pranlukast followed by intratracheal administration of diluent; Group G: pretreatment with vehicle for pranlukast followed by intratracheal administration of eotaxin; Group H: pretreatment with pranlukast followed by intratracheal administration of eotaxin.

Measurement of Airway Responsiveness to Ach

Following anesthesia, a tracheal cannula (18-gauge) was inserted into the cervical trachea through a tracheotomy. The cannula was connected to a mouse ventilator (Model 687; Harvard Apparatus, Kent, UK) with an in-line pressure transducer. Animals were ventilated at 120 breaths/min. A 27-gauge needle was used to establish intravenous access via the tail vein. A paralytic agent (suxamethonium, 0.5 mg/kg; Antigen Pharmaceuticals Ltd, Roscrea, Ireland) was administered to eliminate spontaneous respiration. After recording of a stable baseline airway pressure, Ach (0.1, 0.32, 1.0, 3.3, 10.8 µg/g; Sigma, St. Louis, MO) was infused intravenously over 1 s via the tail vein and changes in airway pressure were recorded for 5 min on a standard polygraph (Multitrace 2 Recorder 5022; Lectromed Ltd., Jersey, Channel Islands, UK). Increasing half-log concentrations were administered at 5-min intervals, with one hyperinflation of the tidal volume applied 3 min after Ach administration, which was accomplished by manually blocking the outflow of the ventilator. Final airway pressure changes were recorded as the maximum percentage of change in airway pressure from baseline (maximum airway pressure  baseline airway pressure/baseline airway pressure). PC₅₀Ach, the provocative concentration of Ach in µg/g causing a 50% increase in airway pressure, was calculated.

BAL and Cell Counting

After measurement of airway responsiveness, mice were given a lethal dose of pentobarbital (60 mg/kg intraperitoneally, Sagatal; May & Baker Ltd, Dagenham, UK), and the lungs were lavaged 6 times with 5-ml aliquots of 0.9% NaCl solution through the tracheotomy. The lavage fluid was centrifuged (300 × g for 10 min at 4° C), and the BAL fluid supernatant was stored at 20° C prior to eicosanoid assay. The cell pellet was resuspended in 1 ml of Hanks' balanced salt solution. Total cell counts were performed by adding 10 µl of the cell suspension to 90 µl of Kimura stain, and cells were counted in a Neubauer chamber (American Optical Corp., Southbridge, MA) under light microscopy. Differential cell counts were made on cytospin preparations, prepared by centrifuging at 300 rpm for 6 min and staining with May-Grunwald stain. Cells were identified as macrophages, neutrophils, eosinophils, lymphocytes, and epithelial cells according to standard morphology. Five hundred cells were counted under ×400 magnification, and the percentage and absolute number of each cell type were calculated.

Histology

Lungs were inflated by tracheal instillation of 1 ml of optimal cutter temperature compound (OCT; Sakura, Torrance, CA) embedding medium, diluted 1:1 with PBS. The lobes were dissected and mounted over cork disks, covered with OCT compound and snap-frozen in isopentane (BDH, Poole, UK) cooled by liquid nitrogen. The frozen blocks were kept at 20° C before use. Sections alongside the main intrapulmonary bronchus (6 µm) were cut in a cryostat at 30° C and collected on glass slides previously coated with poly-L-lysine (Sigma), fixed in acetone (BDH) for 10 min, wrapped in aluminum foil, and kept at 20° C before staining. Hematoxylin-eosin was used for the staining of the tissue, and cyanide-resistant eosinophil peroxidase (EPO) activity, using potassium cyanide, diaminobenzidine, and hydrogen peroxide (BDH), was used to stain the eosinophils.

Sections were coded and read in a blind fashion. Positive cells were enumerated around the bronchi (mucosal and submucosal areas) at a magnification of ×400. Counts were performed on a minimum of five randomly selected intrapulmonary bronchi and expressed as per millimeter of basal lamina using computer-assisted image analysis.

Eicosanoid Assays

BAL fluid was assayed for LTB₄ and LTC₄ in duplicate using a commercially available ELISA kit (Cayman Chemical Co., Ann Arbor, MI). The limit of sensitivity of the assay was 7 pg/ml for LTB₄ and 10.3 pg/ml for LTC₄.

Data Analysis

All values are expressed as means ± SEM. Nonparametric analysis of variance (ANOVA, Kruskal-Wallis method) was used to determine significance among the groups in the pranlukast and SB210661 protocols. We used the Mann-Whitney U test to analyze for significant differences between individual groups, and a value of p < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Eotaxin-induced Eosinophilic Inflammation

Eotaxin increased the number of eosinophils in the IL-5 transgenic mice both in lung tissue and in BALF. Thus, after vehicle administration, eosinophil count in BALF was 1.2 ± 0.4 × 10⁵ eosinophils/ml, increasing to 16.8 ± 1.8 × 10⁵ eosinophils/ ml after eotaxin administration. Pranlukast alone had no effect on the number of eosinophils recovered in BALF or measured in lung tissue, but significantly inhibited eosinophil influx from 18.4 ± 2.1 × 10⁵ eosinophils/ml to 6.1 ± 1.7 × 10⁵ eosinophils/ml (p < 0.01) in BALF and from 115.4 ± 21.8 eosinophils/mm basal lamina to 41.1 ± 8.8 eosinophils/mm basal lamina (p < 0.05) in lung tissues (Figure 1). Similar results were obtained with SB210661 with a significant suppression of eosinophil numbers in tissues or BALF.

View larger version (14K): [in this window] [in a new window]   Figure 1.   Effect of pranlukast and SB210661 on BHR to Ach induced by intratracheal instillation of murine eotaxin in IL-5 transgenic mice. Bronchial responsiveness is expressed as the concentration of Ach needed to cause an increase in baseline insufflation pressure by 50% (PC50). In the first two groups, mice treated with eotaxin intratracheally (B) had a significantly lower log PC50 than diluent-treated mice (A). Groups C, D, and E are mice treated with SB210661 followed by intratracheal administration of diluent, with vehicle for SB210661 followed by eotaxin, and with SB210661 followed by eotaxin, respectively. Pranlukast partially inhibited the reduction in log PC50 induced by eotaxin. Groups F, G, and H are the corresponding groups for the study of the effect of pranlukast which also inhibited partially the reduction in logPC50 induced by eotaxin. *p < 0.01 compared with effect of diluent alone; #p < 0.05 compared with vehicle-treated eotaxin-challenged mice.

Eotaxin-induced BHR

Intratracheal instillation of eotaxin induced a significant increase in bronchial responsiveness to Ach, with a reduction in logPC₅₀ from 0.87 ± 0.08 to 0.34 ± 0.22 (p < 0.01; Figure 2). Both pranlukast and SB210661 had no effect on the logPC₅₀ of IL-5 transgenic mice, but inhibited eotaxin-induced BHR. After pranlukast, logPC₅₀ was 0.39 ± 0.10 after eotaxin, whereas after diluent treatment, logPC₅₀ was significantly lower at 0.43 ± 0.16 (p < 0.05). Similarly, after SB210661, the corresponding values were 0.53 ± 0.10 and 0.22 ± 0.10 (p < 0.05).

View larger version (19K): [in this window] [in a new window]   Figure 2.   Effect of SB210661 and pranlukast on eosinophil counts in BALF (A) and in the airways (B) induced by intratracheal instillation of murine eotaxin in IL-5 transgenic mice. Legend similar to Figure 1. Eosinophils in the airways were detected as cells staining for major basic protein per millimeter of basement membrane. Eotaxin induced a significant increase in eosinophil counts which were inhibited by both SB210661 and pranlukast. *p < 0.01 compared with effect of diluent alone; #p < 0.05 compared with vehicle-treated eotaxin-challenged mice; +p < 0.01 compared with vehicle-treated eotaxin-challenged mice.

Eotaxin-induced LT Release in BALF

Eotaxin administration induced a significant 19-fold increase in the concentrations of LTC₄ in BALF compared with vehicle administration (Figure 3). We also measured LTC₄ concentrations in BALF of wild type CBA/ca mice and could not detect any levels within the limits of our assay. Both pranlukast- and SB210661-treated mice inhibited eotaxin-induced LTC₄ release, but these compounds had no effect on the baseline levels of LTC₄. Concentrations of LTB₄ were detected in BALF of IL-5 transgenic mice but were not increased by eotaxin administration. To ensure that other proteins in the BALF were not interfering with the assay of LTC₄, we removed proteins of molecular weights (MW) > 3,000 daltons by using a cellulose membrane (Microcon Centrifugal Filter Devices; Millipore Corp., Watford, Herts, UK). We found that similar levels of LTC₄ were measurable in the BALF with no significant loss, indicating that there was no interference of proteins of MW > 3,000 with the LTC₄ assay.

View larger version (15K): [in this window] [in a new window]   Figure 3.   Effect of SB210661 and pranlukast on the levels of leukotriene B4 and C4 in BALF. The groups are similar to those described in Figure 1. There were no significant changes in LTB4 levels. However, after eotaxin instillation, the concentrations of LTC4 were significantly increased; these concentrations were inhibited by pretreatment with either pranlukast or SB210661. p < 0.05 compared with effect of diluent alone; #p < 0.05 compared with vehicle-treated eotaxin-challenged mice.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have shown that the eosinophilia and BHR induced by eotaxin in IL-5 transgenic mice were inhibited by a receptor antagonist of the cysteinyl-leukotriene receptor, pranlukast, and by an inhibitor of the 5-lipoxygenase, SB210661. In addition, we found evidence of production of the cysteinyl leukotriene, LTC₄, in BALF, but not of the other LT of this pathway, notably LTB₄, indicating a specific activation of the cysteinyl-leukotrienes. Concentrations of LTC₄ declined with the 5-lipoxygenase inhibitor, and surprisingly with the cysteinyl-leukotriene receptor antagonist. Our data indicate that eotaxin may stimulate certain lung cells in the IL-5 transgenic mouse to produce cysteinyl-leukotrienes, which in turn mediates eosinophil recruitment and BHR.

An interaction between IL-5 and eotaxin has already been shown (3, 5). Eotaxin alone does not induce BHR and eosinophilia in wild-type mice but is potent in these respects in IL-5 transgenic mice (6). Although IL-5 transgenic mice demonstrate high levels of circulating eosinophil blood counts, there is only a small increase in eosinophils in lung and airway tissues. Eotaxin induces chemotaxis of circulating eosinophils into the tissues. It is unlikely that there is involvement of cysteinyl-leukotrienes in the mild tissue eosinophilia observed in IL-5 transgenic mice because the 5-lipoxygenase inhibitor and the cysteinyl-leukotriene receptor antagonist had no effect on the number of eosinophils recovered by BAL or in airway tissue. In addition, the IL-5 transgenic mouse does not demonstrate BHR, and bronchial responsiveness is not affected by these compounds.

Concentrations of LTC₄ in BALF from IL-5 transgenic mice were not significantly different from those of wild-type mice. However, it is possible that IL-5 may promote the release of cysteinyl-leukotrienes from the eosinophil when mice were exposed to eotaxin as has been shown in vitro (23). In the guinea pig, there is evidence to indicate that cysteinyl-leukotrienes may induce the release of IL-5 (13), but there is no such evidence in the mouse.

The role of eotaxin in BHR and tissue eosinophilia has been elucidated in studies of the ovalbumin-exposed sensitized mouse. Using blocking eotaxin antibodies to nullify the effect of endogenously released eotaxin, eotaxin has been shown to be an important chemokine involved in the eosinophil influx and in BHR (24, 25). Similarly cysteinyl-leukotrienes are also involved in similar effects in this particular model (15). Other studies have demonstrated that cysteinyl-leukotrienes can induce lung eosinophilia and also BHR (12, 26, 27). Our findings support the view that eotaxin may induce the 5-lipoxygenase enzyme to produce cysteinyl-leukotrienes to cause BHR and eosinophil influx. Preliminary data indicate that the release of cytokine-mediated eosinophil LTC₄ can be increased by eotaxin in vitro (17), similar to the effect of IL-5 (23). Therefore, the eosinophil may be an important site of production of cysteinyl-leukotrienes in response to eotaxin, which also increases the adhesion molecules very late antigen 4 (VLA₄) and CD18 and is known to cause activation of eosinophils (28). A potential sequence of events is that eotaxin induces recruitment of eosinophils which have been primed in IL-5 transgenic mice to the airways, and activates eosinophils recruited to the airways with the release of cysteinyl-leukotrienes, which can further enhance eosinophil recruitment.

The reduction in the concentrations of eotaxin-induced LTC₄ by pranlukast in BALF is rather surprising because such an effect would not be expected from a cysteinyl-leukotriene receptor antagonist. Possible explanations include nonspecific effects of pranlukast as a 5-lipoxygenase inhibitor, or as antagonizing the effects of eotaxin either at the eotaxin receptor level or acting as an antibody to eotaxin. Indeed, pranlukast has recently been shown to inhibit blood eosinophil activation as measured by a reduction in the release of LTC₄ and eosinophil cationic protein (ECP) (31). Pranlukast, on the other hand, has been shown to be a specific and potent inhibitor of the effects of cysteinyl-leukotrienes (20). An alternative explanation would be the presence of a feedback mechanism leading to a reduction in 5-lipoxygenase activity after blockade of the cysteinyl receptor by pranlukast. SB210661 (19) is a potent inhibitor of 5-lipoxygenase and, as expected, inhibited eotaxin-induced increase in BAL concentrations of LTC₄.

Eotaxin induced a predominant airway eosinophilia, and, in our previous study, we have shown that inhibition of eosinophilia using an antibody to the ₄-integrin led to a significant reduction in BHR (6). This indicates that eosinophils alone may be important for the appearance of BHR, which is supported by previous work (32, 33). In the present study, a reduction in eotaxin-induced eosinophil influx through inhibition of cysteinyl-leukotrienes was accompanied by a reduction in BHR, further supporting the close relationship between eosinophils and BHR.

In summary, eotaxin-induced eosinophil recruitment to the airways and lungs and BHR are partly dependent on the generation of cysteinyl-leukotrienes in IL-5 transgenic mice. This effect may be another example of the cooperative effects between IL-5 and eotaxin in increasing the generation of cysteinyl-leukotrienes from the eosinophil.