INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Allergic rhinitis is characterized by a marked increase in eosinophils in the nasal submucosa and epithelium. Eosinophils have been implicated as primary effector cells in the pathogenesis of allergic rhinitis. It has been considered that the recruitment of eosinophils to allergic mucosa arises from the combined action of a number of cellular and molecular signals. Yet, the mechanism responsible for their selective accumulation has not been fully understood.

Eotaxin (CCL11), a potent eosinophil chemoattractant belonging to the class of C-C chemokine, has been recently described (1). Eotaxin signals only via one specific receptor, CCR-3, which is a G-protein-coupled, 7-transmembrane-domain receptor and is expressed on eosinophils but not on neutrophils (4). In addition, eotaxin may help eosinophil accumulation to the nasal mucosa through its effect on the expression of adhesion molecules on microvascular endothelial cells (5). Thus, eotaxin has multifaceted effects on eosinophils and is likely to be a key mediator of tissue eosinophilia. There is increasing evidence that patients with atopic asthma have high concentrations of eotaxin in bronchoalveolar lavage fluid or sputum when compared with normal control subjects (6, 7). In addition, increased expression of eotaxin mRNA and protein as well as accumulation of eotaxin-positive cells have been demonstrated in the tissue or bronchoalveolar lavage after allergen challenge in these patients (8, 9). In nasal polyposis, allergic rhinitis and both allergic and nonallergic sinusitis, upregulation of eotaxin mRNA and protein has been detected (3, 10, 11).

In the present study we investigated the levels of various kinds of eosinophil chemoattractants in nasal lavage fluids of patients with allergic rhinitis sequentially obtained after antigen challenge and compared them with eosinophil counts and eosinophil protein X (EPX). In addition, we evaluated the effect of eotaxin on eosinophil transendothelial migration (TEM) through the nasal microvascular endothelial cells. Furthermore, to examine the relative role of eotaxin among various eosinophil chemoattractants in nasal mucosa, we investigated the effect of anti-CCR-3 monoclonal antibody (mAb) on eosinophil TEM induced by homogenate of nasal mucosa.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects and Study Design

A total of 10 Japanese cedar pollinosis patients participated in this study. The study was performed in midwinter, out of the pollen season to avoid natural provocation. Before the study, written informed consent was obtained after adequately explaining the objective and procedures of the present study. The protocols were reviewed and approved by the local ethics committee for human studies. An antigen challenge was performed using two pollen disks for each nasal cavity (total of four disks; each disc was saturated with 300 µg protein of Japanese cedar). Nasal lavage was performed 15 min and immediately before antigen challenge, as well as 30 min, and 2, 4, 6, 8, and 10 h after antigen challenge. An A1 acoustic rhinometer ("GM" Instruments, Ltd., London, UK) was used to determine nasal airway volume (NAV). The nasal cavity was washed with 10 ml of physiological saline warmed at 37° C by reciprocating the piston of the syringe 10 times on each side. Lavage fluid was filtered through a 52-µm nylon filter to remove mucin, and the filtrate was centrifuged for 20 min at 400 × g. The supernatant was stored frozen at 80° C until assayed for mediator levels. The sediment was subjected to cell count measurement.

Measurement of Mediators

The levels of the following mediators were measured by using commercially available kits (IL-5: Amersham International, Buckinghamshire, UK; IL-16: BioSource International, Inc., Camarillo, CA; RANTES: R&D Systems Inc., Mineapolis, MN; PAF: PAF scintillation proximity assay system, Amersham; LTB4: LTB4 enzyme immunoassay system, Amersham; EPX: EPX RIA kit, Pharmacia, Uppsala, Sweden). The method for the measurement of human eotaxin was described previously (12).

mAbs and Cytokines

Eotaxin was kindly donated by the Shionogi Pharmaceutical (Osaka, Japan). mAbs to RANTES, MCP-3, and MCP-4 were purchased from Genzyme (Cambridge, MA). Preparation of anti-eotaxin mAb was described previously (12). Anti-human interleukin (IL)-5R (YT01) was purchased from R&D. PAF receptor antagonist WEB 2086 and LTB4 receptor antagonist Y-24180 were generous gifts from Boehringer Ingelheim (Germany) and Yoshitomi Pharmaceutical Industries Inc. (Osaka, Japan), respectively. Preparation of mAbs to CCR-3 has been described elsewhere (13). RANTES, IL-5, and MCP-4 were purchased from Genzyme. LTB4 and PAF were purchased from Cascade Biochem, Berkshire, UK.

Transendothelial Migration Assay

Isolation and preparation of HMMECs from nasal mucosa were described previously (14). TEM was investigated by the method reported previously (14, 15). To examine the role of eotaxin on eosinophil migration, we performed an inhibition assay. Inferior nasal turbinate mucosas obtained at the time of surgery were homogenized in 5 ml of ice cold phosphate-buffered saline (PBS) by Polytron (Kinematica, Sweden) and briefly sonicated. The eosinophils were preincubated for 30 min at 37° C in the presence or absence of a saturating concentration of antibodies prior to TEM.

Statistical Assessment

The significances of differences between within-group and between intergroup comparison were determined using Wilcoxon signed rank tests and Mann-Whitney U test, respectively. Correlations were evaluated by the Spearman rank correlation test. Differences were considered significant when the p value was less than 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Clinical Responses and Changes in NAV

When the patients received the allergen challenge, NAV decreased immediately after challenge NAV returned to baseline values by 4 h after challenge but then began to decrease again by 8 h after challenge (Figure 1).

View larger version (13K): [in this window] [in a new window]   Figure 1.   The NAV immediately before a nasal challenge, 30 min after, and at 2-h intervals during the following 10 h after the challenge. A range of 8.5-10.5 cm from the wave tube of the acoustic rhinometer was used to determine NAV. Closed triangles denote changes in NAV after antigen challenge in the allergy group and open triangles denote those after antigen challenge in the control group. The results are value ± SEM (n = 10) at each time point. *p < 0.05 compared with prechallenge levels. **p < 0.01 compared with prechallenge levels.

Analysis of Nasal Lavage Fluid

In all patients with nasal allergy, the number of eosinophils in nasal lavage fluids increased in the 30-min samples (p < 0.01, Figure 2A) and then returned to the prechallenge values 2 h after the nasal antigen challenge. During the late observation hours, the number of eosinophils gradually increased to a maximum 10 h after the provocation. Similarly, the levels of EPX and eotaxin increased in the early and the late phases (Figures 2B and 2C). As the recovery rate of nasal lavage fluids could vary from subject to subject and from lavage to lavage, levels of eotaxin were also analyzed after normalization by concentrations of albumin. The results obtained did not differ significantly from those obtained when eotaxin was expressed as a concentration in nasal lavage fluids (data not shown). For this reason, all lavage data are presented as absolute concentrations and were not normalized to any other parameters.

View larger version (23K): [in this window] [in a new window]   View larger version (22K): [in this window] [in a new window]   Figure 2.   (A-H ) Changes in eosinophil cell counts, EPX, and various eosinophil chemoattractant levels in nasal lavage fluids. Nasal lavage was performed 15 min and immediately before antigen challenge, as well as 30 min, and 2, 4, 6, 8, and 10 h after antigen challenge. The levels immediately before antigen challenge are defined as prechallenge levels. Closed circles denote allergy group values and open circles denote control group values. The results are value ± SEM (n = 10) at each time point. *p < 0.05 compared with prechallenge levels. **p < 0.01 compared with prechallenge levels.

Allergic rhinitis patients exposed to allergen challenge release IL-5 and IL-16 into nasal lavage fluid during the early- and late-phase allergic responses (Figures 2D and 2E). On the other hand, LTB4 and PAF were released into nasal lavage fluid mostly during the early phase (Figures 2G and 2H). Normal subjects had no changes in any of the measured parameters observed before and after the antigen challenge (Figures 2A through 2H).

The lavage levels of eosinophil counts and EPX showed a strong statistical correlation as expected (Table 1). The lavage levels of eosinophil counts and EPX (r = 0.304, p < 0.01) weakly but significantly correlated with the percentage change in NAV: r = 0.782 (p < 0.001) and r = 0.312 (p < 0.01), respectively. The levels of eotaxin and, to a lesser extent, those of IL-5 and IL-16 correlated significantly with lavage levels of eosinophil counts and EPX levels (Table 1). The correlation coefficient between eotaxin and eosinophil cell counts was 0.726 (p < 0.001) and between eotaxin and EPX was 0.673 (p < 0.001), as shown in Figures 3A and 3B and Table 1. The correlation coefficient between IL-5 and eosinophil cell counts was 0.500 (p < 0.001) and between IL-16 and eosinophil cell counts was 0.46 (p < 0.001) (Table 1). No significant correlation was observed between PAF and eosinophil counts or EPX, between LTB4 and eosinophil counts or EPX, or between RANTES and eosinophil counts or EPX.

View larger version (25K): [in this window] [in a new window]   Figure 3.   Correlation among eosinophil cell counts, EPX and eotaxin levels. (A) Correlation between eosinophil cell counts and eotaxin levels. (B) Correlation between EPX levels and eotaxin levels.

Transendothelial Migration Assay

Eotaxin, MCP-4, and RANTES induced eosinophil TEM in a concentration-dependent manner with the following rank order of potency as shown in Figure 4: eotaxin > MCP-4, RANTES > IL-5 > PAF > LTB4, IL-16.

View larger version (19K): [in this window] [in a new window]   Figure 4.   Concentration dependence of eosinophil chemoattractant-induced eosinophil TEM. Eosinophils labeled with 35S were added to each of the upper compartments. Then, eosinophil chemoattractant was added into the lower compartments. After the incubation period, the inserts were removed, and the contents in the lower chambers were collected to count the eosinophil number, which had transmigrated from the upper compartments through the HMMECs monolayers and pored membrane filter. Data shown represent mean ± SEM from at least five experiments. Control TEM (PBS) was 5.1 ± 0.5%. *p < 0.05 compared with control levels. #p < 0.05 compared with levels of other eosinophil chemoattractants-induced eosinophil TEM.

Finally, to examine the relative role of eotaxin among various eosinophil chemoattractants in nasal mucosa, we investigated the effects of mAb to CCR-3 on eosinophil TEM induced by homogenate of nasal mucosa. In the preliminary experiment, we confirmed that 30 µg/ml of CCR-3 mAb, completely inhibited eosinophil TEM induced by eotaxin, RANTES, MCP-3, and MCP-4, appeared in the homogenates of nasal mucosa. When eosinophils were pretreated with anti-CCR-3 mAb or anti-IL-5R mAb, eosinophil TEM induced by the homogenate of nasal mucosa was reduced in a dose-dependent manner. At optimal concentrations, CCR-3 mAb (30 µg/ ml) and IL-5R (30 µg/ml) mAb were inhibited by 35.2 ± 9.2% and 12.4 ± 8.1%, respectively (Figure 5). The treatment of eosinophils with CCR-3 mAb and IL-5R mAb together further blocked eosinophil TEM by 62.8 ± 15.1%. On the other hand, a PAF receptor antagonist and an LTB4 receptor antagonist only weekly inhibited the eosinophil TEM. In this experiment, when eotaxin in the homogenates of nasal mucosa was neutralized by excessive anti-eotaxin mAb, eosinophil TEM induced by the homogenate of nasal mucosa was reduced to 27.2 ± 4.4%. On the other hand, anti-RANTES mAb, anti-MCP-3 mAb, and anti-MCP-4 mAb were only inhibited by 9.2 ± 3.1%, 6.4 ± 2.9%, and 5.4 ± 1.9%, respectively.

View larger version (25K): [in this window] [in a new window]   Figure 5.   Effects of Abs against receptors of eosinophil chemoattractants on the nasal homogenate-induced TEM. Eosinophils were pretreated with CCR-3 mAb (30 µg/ml), IL-5R mAb (30 µg/ml), WEB 2086 (10 µM), or Y-24180 (10 µM) prior to eosinophil TEM. Eosinophils labeled with 35S were added to each of the upper compartments. Then, homogenate of nasal mucosa was added into the lower compartments. Data shown represent mean ± SEM from at least five experiments. *p < 0.05 compared with control levels. **p < 0.01 compared with control levels.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Of particular significance in the present study is the finding that allergen challenge leads to parallel increases of numbers of eosinophils and of eotaxin concentration in nasal lavage fluid. In our experiments of TEM assay using HMMECs, eotaxin showed the most potent effect, which was consistent with the findings in a similar experiment using HUVEC (16). Taken together, these observations suggest that eotaxin strongly contributes to the pathogenesis of allergic rhinitis by the specific recruitment of eosinophils into the airways. Most recently, Hanazawa and colleagues observed that intranasal administration of eotaxin actually increases nasal eosinophils and nitric oxide in patients with allergic rhinitis (17). It is possible that eosinophils induced by eotaxin increase nitric oxide and contribute to oxidative stress.

Patients with allergic rhinitis exposed to allergen challenge release IL-5 into nasal lavage fluid during the early- and late-phase responses, as with eotaxin. The amount of IL-5 significantly correlated with eosinophil counts and EPX levels. It was reported that eotaxin had an important local role in eosinophil recruitment from blood microvessels, whereas IL-5 facilitates this process by acting remotely as a hormone to stimulate the release of a rapidly mobilizable pool of bone marrow eosinophils into the circulation (18). In addition, preincubation of eosinophils with IL-5 significantly enhanced eotaxin-induced TEM across resting HUVEC (16). We also confirmed this synergy in eosinophil TEM across HMMECs (data not shown). Taken together, these results suggest that both eotaxin and IL-5 play crucial roles in eosinophilic inflammation in nasal mucosa.

IL-16 has recently been characterized as an eosinophil chemoattractant (19). In this study, we observed that the change in the concentration of IL-16 showed biphasic increase after allergen challenge as eotaxin and correlated significantly with lavage levels of eosinophil counts and EPX. However, the correlation coefficients between IL-16 levels and eosinophil counts or EPX levels were 0.46 and 0.48, respectively, which were less than those obtained with eotaxin (Table 1). In addition, the magnitude of the chemotactic response to IL-16 by eosinophils in a transmigration assay was much lower than eotaxin. These results suggest that IL-16 contributes less in the chemotactic activity of eosinophils than eotaxin.

RANTES is a candidate for selective recruitment of both eosinophils and macrophages. Sim and coworkers demonstrated that RANTES were released in nasal fluid after allergen challenge in allergic subjects and the levels correlated with corresponding total symptom scores during late-phase response (20). However, we detected RANTES in only 19 of 70 samples of nasal lavage fluids obtained before and after allergen challenge in allergic subjects. Other samples contained RANTES below the sensitivity of the assay we used. Moreover, no significant correlation was observed between RANTES and eosinophil counts or EPX. The low number of detectable samples certainly contributed to this low level of correlation. The discrepancy between the study of Sim and coworkers and our study might come from the procedure of collecting samples. We performed nasal lavage instead of using the matrix technique because we wished to analyze many kinds of cytokines and chemokines to evaluate their relative roles in eosinophilic inflammation in nasal mucosa. In addition, the matrix technique does not allow analysis of the number of inflammatory cells. The relative role of RANTES in eosinophilic inflammation in nasal mucosa remains to be established.

One of the most important roles of PAF is the activation and recruitment of eosinophils. Leukotriene B4 is known to have a potent chemotactic activity on neutrophils, eosinophils, and monocytes. However, in our study using eosinophils separated by MACS system, eosinophils showed a weak response to PAF or LTB4. Very recently, it was demonstrated that Percoll-isolated eosinophils migrated to the lipid mediators, LTB4 and PAF, in a dose-responsive fashion. Although MACS isolation provided a greater number and higher purity of eosinophils, these eosinophils migrated less to LTB4 and PAF (21). Thus, we could not deny the possibility that we underestimated the ability of PAF and LTB4 in TEM. However, considering that both PAF and LTB4 did not show significant correlation with eosinophil counts and EPX levels, these mediators may not be important chemoattranctants in nasal mucosa.

Finally, to examine the relative role of eotaxin among various eosinophil chemoattractants in nasal mucosa, we investigated the effect of blocking CCR-3 on eosinophil TEM induced by homogenates of nasal mucosa. When eosinophils were pretreated with anti-CCR-3 mAb or anti-IL-5R mAb, eosinophil TEM induced by nasal mucosal homogenate was partially but significantly reduced. The treatment of eosinophils with anti-CCR-3 mAb and anti-IL-5R mAb together further blocked eosinophil TEM. On the other hand, the PAF receptor antagonist UK-74505, at either optimal or higher concentrations, inhibited eosinophil TEM to almost the same levels as when no antireceptor mAb was used. Because there is no specific receptor for eotaxin, we cannot evaluate the exact contribution of eotaxin to the activity in extracts in this transendothelial eosinophil migration assay. We cannot completely exclude the possibility that the other CCR-3-activating chemokines could be more important than eotaxin itself. However, when eotaxin in nasal extracts was neutralized by excessive anti-eotaxin mAb, eosinophil TEM induced by the homogenate of nasal mucosa was reduced by 27.2 ± 4.4%, which is similar to the percentage inhibition by anti-CCR-3 mAb. On the other hand, the level of inhibition by anti-RANTES mAb, anti-MCP-3 mAb, and anti-MCP-4 mAb was minimal. Thus, it is very possible that eotaxin is the most important eosinophil chemoattractant among CCR-3-activating chemokines. Recently, Grimaldi and coworkers showed that repeated anti-CCR-3 mAb treatment in Nippostrongyus brasiliensis-infected mice significantly reduced tissue eosinophilia in the lung tissue and bronchoalveolar lavage fluid (22). In addition, an antagonist of CCR-3 was derived by a modified cloning strategy for macrophage inflammatory protein-4, by use of primers to remove the signal peptide and to replace the N-terminal alanine with a methionine residue. This antagonist was termed Met-Ck7, was specific for CCR-3, and, at concentrations as low as 1 nmol/L, was able to block eosinophil chemotaxis to eotaxin (23). In the near future, it is possible that small-molecule inhibitors for CCR-3 might be developed. The lack of complete inhibition of eosinophil TEM, even after pretreatment with anti-CCR-3 mAb and anti-IL-5R mAb, suggests that some other eosinophil chemoattractants contribute to eosinophil TEM.