INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Bronchial asthma is characterized by chronic recruitment of eosinophils in the airways. It has been reported that bronchial eosinophil recruitment and activation even may occur in mild- moderate stable asthma and that bronchial epithelium damage and airway responsiveness may be partially associated with the eosinophilic inflammatory reaction (1). There is increasing evidence that the eosinophil-rich bronchial inflammation characteristic of asthma is orchestrated, at least partly, by cytokine products of activated T lymphocytes. Of particular interest is T-helper type 2 (Th₂) cell-derived interleukin-5 (IL-5) (2). This cytokine mediates the terminal differentiation of committed eosinophil precursors, activates mature eosinophils, and prolongs their survival in culture and, presumably, at sites of allergic inflammation (3). IL-5 also selectively enhances eosinophil degranulation, antibody-dependent cytotoxicity (7), and adhesion to vascular endothelium (8). By topical instillation of IL-5 into the lower airways, we recently indicated that IL-5 acts directly as a chemoattractant for eosinophils recruitment into human airway in asthmatics, and as an activator for the recruiting eosinophils (9). Because IL-5 contributes to the influx of inflammatory cells, we hypothesized that it may therefore be involved in the development of airway hyperresponsiveness in asthmatics. We tested part of this hypothesis by administering recombinant human IL-5 by nebulization to patients with allergic bronchial asthma, and measuring airway responsiveness by methacholine challenge and cell differential counts as well as concentrations of eosinophil cationic protein (ECP) in induced sputum in a blinded crossover study.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

The study protocol was approved by the Ethics Committee of Guangxi Medical University, People's Republic of China, and all subjects provided written consent. Eight nonsmoking patients (5 males, 3 females; 19 to 58 yr of age), who met the criteria for a diagnosis of asthma as defined by the American Thoracic Society (10) were enrolled in this study. All patients had mild atopic asthma, with baseline forced expiratory volume in one second (FEV₁) greater than 70% of predicted value (mean ± standard error of mean [SEM] = 98 ± 3% of predicted), requiring only intermittent use of inhaled ₂-agonists. All patients had a provocative concentration of methacholine producing a 20% fall in FEV₁ (PC₂₀ Mch) < 8 mg/ml. Each patient had one or more documented positive skin prick test responses to aeroallergens, but none was receiving immunotherapy including the use of inhaled corticosteroid therapy.

Assessment of Bronchial Responses

Baseline measurements of spirometry were followed by methacholine challenge. Spirometric measurements were FEV₁, forced vital capacity (FVC), and peak expiratory flow (PEF), using dry wedge spirometry (Autospiro AS-600; Minto Ltd., Japan). Methacholine inhalation tests were carried out by the method described by Hargreave and coworkers (11). Briefly, methacholine chloride (Sigma Chemical Company, St. Louis, MO) was dissolved in normal saline to make solutions of doubling concentrations from 0.04 to 20 mg/ml. Normal saline and methacholine solutions were inhaled from a nebulizer (Liuzhou Medical Instrument Ltd., Guangxi, China) operated by compressed air at 5 L/min. Normal saline was inhaled first for 2 min, and FEV₁ was measured. If the change in FEV₁ from the baseline after inhalation of normal saline was 10% or less, inhalation of methacholine was started. When inhaled normal saline caused a greater change in FEV₁, the test was stopped or postponed. Methacholine was inhaled for 2 min by mouth tidal breathing wearing a noseclip, and this was followed immediately by measurements of FEV₁. Increasing concentrations of methacholine were inhaled until a fall of 20% or more in FEV₁ was obtained. The measured values were plotted on semilogarithmic graph paper, and PC₂₀ Mch was determined. FEV₁ was measured three times, and the best FEV₁ value of three attempts was recorded each time. Following this, subjects underwent sputum induction.

Administration of IL-5 and Control Experiments

The following afternoon (24 h after initial methacholine challenge), 10 µg of recombinant human IL-5 (Genzyme Co., Boston, MA) in vehicle (0.1% bovine serum albumin in 0.9% saline) or vehicle only was inhaled as a 0.5 ml nebulized solution, the chamber was refilled twice with 0.5 ml vehicle, and the nebulization was repeated to scavenge any remaining IL-5. The dose of IL-5 was based upon a preliminary study involving two asthmatic patients.

Spirometry, methacholine challenge, and sputum induction were repeated 2, 24, and 48 h after the inhalation of IL-5 or vehicle. At least 4 wk were allowed to elapse between the two inhalations, and the order of inhalation of IL-5 or vehicle was randomized.

In addition, recombinant human IL-5 used for inhalation in this study was obtained commercially, which was contaminated with endotoxin (0.04 ng/µg, indicated by manufacturers), thus, total dose of endotoxin inhaled by each subject was 0.4 ng. To determine whether the endotoxin contaminated in IL-5 was responsible for airway hyperresponsiveness measured in this study, a separate control inhalation of 0.4 ng lipopolysaccharide (LPS, Escherichia coli serotype 026:B6 [Sigma]) was performed in five of eight asthmatics (the same subject group) more than 4 wk later in a similar manner.

Sputum Induction and Examination

Sputum induction was performed by the method of Pin and coworkers (12) with a slight modification. Briefly, subjects inhaled nebulized 3.5% saline and gargled orally with water prior to voluntary coughing every 2.5 min until 20 min had elapsed or until 5 ml of sputum had been expectorated. The induced sputum was added with an equal volume of 1 mmol/L dithiothreitol (Sigma) in Hanks' buffered salt solution, and then were mixed gently by vortex mixer and incubated at 37° C for 15 min to ensure complete homogenization. After incubation, aliquots of homogenized samples were removed for total cell counts by hemocytometer. The remainder of the homogenized sputum was centrifugated at 2,000 rpm for 5 min. the supernatants were aspirated and stored at 70° C for later detection of ECP. The cell pellets were prepared for differential cell counts by Diff-Quik staining of cytocentrifuge preparations. A total of 400 cells not including squamous cells were enumerated for differential cell counts, which identified macrophages, lymphocytes, neutrophils, and eosinophils. The absolute number of each type cell was calculated.

ECP Assay

Because mean percentages of squamous cell in sputum samples from all subjects on all four occasions were similar (data not shown) and because we believed that squamous cells in the sputum samples represented salivary contamination, we have estimated that the mean extent of salivary contamination of the sputum samples was similar. Thus, we have expressed the quantification of ECP in micrograms per liter of induced sputum. Induced sputum samples previously stored at 70° C were thawed. ECP concentrations were determined with commercially available ECP Fluoroimmunoassay kits (Pharmacia AB, Uppsala, Sweden) according to the procedures recommended by the manufacturers. Assay sensitivity for ECP was 2 µg/L.

Statistical Analysis

Results are expressed as arithmetic mean ± SEM for spirometric data, cell differential counts, and ECP concentrations. PC₂₀ Mch values were log₁₀-transformed for analysis and reported as geometric mean and geometric standard error of mean (GSEM). GSEM is the antilog of SEM of the log PC₂₀ Mch values and is a factor multiplication. Statistical analysis was done by repeated measures analysis of variance (ANOVA) for data conforming to a normal distribution, and by Friedman's test for those data with a nonparametric distribution (confirmed by the Shapiro-Wilk W test). p Values < 0.05 were considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effects of IL-5 on Spirometry and Methacholine Responsiveness

All subjects tolerated the procedures without complications such as fever, increased cough, or any other increased symptoms during or after the 48-h follow-up period.

Baseline spirometry showed no difference among the initial values of FEV₁, FVC, and PEF measured at the beginning of each arm of the experiment and at subsequent time points (data not shown). These results suggested that the inhalation of recombinant human IL-5 did not alter lung function in patients with allergic asthma.

Baseline measurements of PC₂₀ Mch before three challenges showed no significant difference (0.99 ± 1.57 mg/ml with vehicle, 0.90 ± 1.66 mg/ml with IL-5, and 1.14 ± 1.59 mg/ ml with LPS, respectively) (Figure 1). PC₂₀ Mch within both vehicle and LPS group did not appear to change from baseline at any time throughout the study (all p > 0.05). It meant neither vehicle nor LPS inhalation of asthmatics could lead to an increased airway responsiveness to methacholine. Two hours after IL-5 inhalation, PC₂₀ Mch (1.03 ± 1.70 mg/ml) was not different from baseline value (p > 0.05). PC₂₀ Mch fell from baseline to 0.32 ± 1.63 mg/ml (p < 0.01) at 24 h, and to 0.55 ± 1.49 mg/ml (p < 0.05) at 48 h after IL-5 inhalation; that is to say, IL-5-induced airway hyperresponsiveness reached a peak at 24 h and then slightly declined at 48 h.

View larger version (15K): [in this window] [in a new window]   Figure 1.   Changes in log PC20 Mch (mean ± SEM) after inhalation of vehicle control (n = 8), recombinant human IL-5 (n = 8), or endotoxin (n = 5) in patients with allergic bronchial asthma. *p < 0.05; **p < 0.01 compared with baseline measurement. The scale along the x-axis is nonlinear.

Effects of IL-5 on Sputum Cytology

Satisfactory sputum samples were obtained from all subjects on all occasions. There was an increase in the total cell numbers in induced sputum after IL-5 inhalation, which increased with time was significantly greater than in both control inhalations, reaching a maximum at 24 h (Table 1). Moreover, a similar increase in the percentage of eosinophils in induced sputum after IL-5 inhalation was observed compared with controls.

The absolute number of each type of cell was calculated; the results are presented in Figure 2. After allergic asthmatics were challenged with vehicle or 0.4 ng of LPS only, we did not observe increases of eosinophil counts in induced sputum obtained at three time points when compared with baseline measurement before inhalation (all p > 0.05). Compared with baseline value (0.8 ± 0.3 × 10⁸/L), no significant increase in the number of eosinophils could be seen at 2 h (1.2 ± 0.4 × 10⁸/L, p > 0.05) after IL-5 inhalation; The number of eosinophils increased with time, reaching a maximum at 24 h (5.0 ± 1.6 × 10⁸/L, p < 0.05); this significant sputum eosinophilia lasted at least 48 h (3.3 ± 1.0 × 10⁸/L, p < 0.05).

View larger version (44K): [in this window] [in a new window]   Figure 2.   Comparison of numbers (mean ± SEM) of each type of cell in induced sputum from allergic asthmatics challenged with vehicle (open bars, n = 8), recombinant human IL-5 (closed bars, n = 8), or endotoxin (hatched bars, n = 5). BL = baseline. *p < 0.05 compared with baseline measurement. (A) macrophage; (B) neutrophils; (C ) lymphocytes; (D) eosinophils.

We also noted that although percentages of macrophages and neutrophils decreased after IL-5 inhalation because of increase in percentage of eosinophils, the absolute numbers of macrophages and neutrophils were all not different from their baseline values, respectively.

Effects of Inhalation on ECP Levels in Sputum

Baseline measurements of ECP before three challenges showed no significant difference in induced sputum from allergic asthmatics (Figure 3). It was found that vehicle or 0.4 ng of LPS inhalation of asthmatics did not lead to elevations of ECP levels in sputum. Two hours after IL-5 inhalation, sputum ECP concentration (307.1 ± 52.3 µg/L) was still not different from baseline value (279.5 ± 45.5 µg/L); ECP concentration increased from baseline to a significantly higher extent at 24 h (773.5 ± 97.9 µg/L, p < 0.01), and at 48 h (738.5 ± 89.5 µg/L, p < 0.01) after IL-5 inhalation. IL-5-induced elevation of ECP reached a peak at 24 h and could last at least 48 h.

View larger version (15K): [in this window] [in a new window]   Figure 3.   Comparison of ECP concentrations (mean ± SEM) in induced sputum from allergic asthmatics challenged with vehicle (n = 8), recombinant human IL-5 (n = 8), or endotoxin (n = 5). *p < 0.01 compared with baseline measurement.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the present study, we provide the first direct evidence that IL-5 challenge of allergic asthmatics induced significant increases in airway reactivity, as well as increased numbers of eosinophils with an elevated ECP concentration in induced sputum. Our study confirms the hypothesis that IL-5 can mediate airway hyperreactivity and airway eosinophilia.

Previous studies have suggested that there might be some connection between the infiltration of eosinophils and IL-5 in the pathogenesis of bronchial asthma. In an animal study, Nakajima and coworkers have demonstrated that the in vivo depletion of CD4⁺ cells by pretreatment with anti-L3T4 monoclonal antibody (mAb) significantly decreased the eosinophilia induced by ovalbumin (OVA) inhalation in the trachea of sensitized mice; however, the in vivo depletion of CD8⁺ cells by pretreatment with anti-Lyt-2 mAb had no significant effect on OVA-induced eosinophil infiltration in the trachea; pretreatment with anti-murine IL-5 mAb also decreased OVA-induced eosinophil infiltration in the trachea (13). These materials provide direct evidence that CD4⁺ but not CD8⁺ cells, as well as CD4⁺ cells' product IL-5, mediated antigen-induced eosinophil recruitment in the airways. Other work also suggested a critical and possible independent role for IL-5 in the accumulation of eosinophils within the lung of sensitized guinea pigs (14). The importance of IL-5 is not only emphasized by studies in animals showing that eosinophil accumulation in the airway can be inhibited by anti-IL-5 mAb, but also further demonstrated by studies showing that in vivo administration of recombinant IL-5 increases the number of eosinophils in BAL fluid (15).

More recently, we have examined for the first time the effects of the instillation of IL-5 directly into the airways on the airway eosinophilic inflammation in patients with allergic asthma (9). Our results demonstrated that the total eosinophils (BMK-13⁺ cells) and the activated eosinophils (EG2⁺ cells) in bronchial mucosa, the eosinophil numbers in BAL fluid, as well as ECP in BAL fluid from saline-challenged segments were not different from those in unchallenged segments. However, a significant eosinophilia was observed in bronchial mucosa and bronchoalveolar lavage (BAL) fluid from IL-5-challenged sites. Eosinophil activation, as assessed by secretion of ECP, was also increased significantly in bronchial mucosa and BAL fluid. These results strongly suggested that IL-5 is capable of inducing eosinophil infiltration into the asthmatic airways, as well as the activation of infiltrating eosinophils. For practical and ethical reasons, it was not possible to perform a time-course study of changes of eosinophil numbers and ECP concentrations in bronchial mucosal and BAL fluid after IL-5 administration. We chose induced sputum as the study sample in the present study to observe the effect of inhaled IL-5 on time-course changes of airway inflammation because recent studies of sputum cell counts have demonstrated repeatability, responsiveness, and validity of the methods (12, 16), and sputum phase and biochemical analysis, such as detection of ECP, of induced sputum is also feasible and responsive in asthmatic and healthy subjects (17, 18).

The most important findings in this study were that IL-5 inhalation, not vehicle or LPS (0.4 ng) inhalation, of allergic asthmatics produced a marked progressive influx of eosinophils into the airways when studied by differentials of cells in induced sputum. The changes were time-course-related, in that the response was most marked at 24 h. Both the percentage and the absolute number of sputum eosinophils were significantly higher at 24 and at 48 h after IL-5 challenge than baseline measurement before challenge. We also found that sputum ECP levels were elevated in asthmatics by IL-5 inhalation in a similar manner. Our results demonstrated that inhalation of 10 µg IL-5 in patients with allergic asthma was not only able to chemoattract eosinophils recruited into airways, but also resulted in activation of increased eosinophils in sputum.

With respect to the relationship between IL-5 and airway reactivity, many studies have been undertaken. It has been reported that antigen challenge of A/J mice induced significant increases in airway reactivity, and that depletion of CD4⁺ cells prevented this antigen-induced airway hyperreactivity (20). In guinea pigs treated with anti-IL-5 mAb, the development of hyperreactivity to histamine and arecoline after OVA challenge was completely inhibited (15). Mauser and coworkers have investigated the effect of a neutralizing monoclonal antibody to murine IL-5 in a cynomolgus monkey model of allergic asthma and found that the inhibition of this pulmonary eosinophilia and bronchial hyperresponsiveness by TRFK-5 was seen for as long as 3 mo after treatment (21). From these data, it could be concluded that IL-5 was involved in the development of hyperreactivity in animal models.

In a double-blind placebo-controlled study, Robinson and colleagues (22) have shown that decreased bronchial responsiveness in asthma after prednisolone treatment was accompanied by a reduction in numbers of BAL cells expressing mRNA for IL-5. These results indicated that the improvement of airway reactivity is associated with the decrease of IL-5 in human asthma. Ackerman and coworkers (23) have found that within the group of atopic asthmatics, the level of bronchial responsiveness to inhaled methacholine best correlated with the total number of cells showing immunostaining for IL-5. In addition, there was a trend for the asthmatics who were IL-5 mRNA-positive to have a lower FEV₁ and PC₂₀ Mch when compared with the asthmatics who were IL-5 mRNA- negative (24).

Our present study showed increased airway responsiveness to methacholine after inhaled IL-5 in patients with allergic bronchial asthma, but no change in spirometric measures. The changes were time-course-related, in that the response was most marked at 24 h, and coincided with an eosinophil recruitment also observed in this study. An increased sensitivity to methacholine after the inhalation of nebulized IL-5, in association with eosinophil infiltration and its activation, suggests that IL-5 plays an important role in the development of airway hyperreactivity.

The exact mechanism by which IL-5 contributes to airway hyperresponsiveness is still unknown. It seems unlikely that IL-5 directly acts on airway and thereby increases its responsiveness, because it has been demonstrated that incubation of tracheal rings with IL-5 does not induce tracheal hyperreactivity to histamine (15). In asthmatics, sputum eosinophils, as well as eosinophil granule proteins have been reported to be inversely correlated with FEV₁ and PC₂₀ Mch, suggesting the direct contribution of eosinophils to the development of airway hyperresponsiveness (25). Because IL-5 is capable of attracting eosinophil recruitment into the asthmatic airways, as well as the activation of infiltrating eosinophils, it is possible that IL-5 induces airway hyperreactivity by means of a role for eosinophils in the asthmatic process. Eosinophils are known to secrete a number of basic proteins, including major basic protein (MBP), ECP, eosinophil-derived neurotoxin, and eosinophil peroxidase that have profound effects on airway cells (26, 27). Release of these products is associated with increases in vascular permeability, bronchoconstriction, and destruction of airway epithelial cells (26). In particular, MBP has been shown to cause bronchial hyperresponsiveness in animal models of asthma (28, 29). The importance of these proteins in human asthma has been illustrated by the demonstration of increased concentrations of MBP in BAL fluid of atopic asthmatics as compared with normal control subjects (27). In addition, eosinophil activation also results in the release of a number of important lipid mediators, including leukotriene C₄, which can contract airway smooth muscle, and platelet-activating factor, which can contract airway smooth muscle as well as increase bronchial responsiveness (30). Furthermore, platelet-activating factor is an eosinophil chemoattractant and functional primer. Thus, eosinophils possess properties that can directly or indirectly cause airway obstruction and promote bronchial hyperreactivity. Moreover, through the eosinophil's own ability to generate chemoattractants, or via stimulation of other cell secretion, additional eosinophil migration is accomplished and a self-perpetuating process established.

Recombinant human IL-5 used for inhalation in this study was obtained commercially, which was contaminated with endotoxin (0.04 ng/µg). The possibility that the airway effects observed were due to endotoxin should be evaluated because of the involvement of endotoxin in the pathogenesis of asthma. It has been reported that in normal subjects inhalation of 200 µg LPSthe major part of endotoxin present in the external membrane of gram-negative bacteriacauses bronchoconstriction (31); a dose of 20 µg LPS causes no response in normal subjects but causes bronchoconstriction in some asthmatic patients (31). Hunt and coworkers (34) have demonstrated that challenge with endotoxin-free allergen in asthmatics resulted in recruitment of significantly fewer neutrophils than challenge with endotoxin-contaminated allergen; and challenge with endotoxin-free allergen resulted in recruitment of significantly greater numbers of eosinophils than neutrophils. When asthmatics were challenged with 20 µg endotoxin, a significant circulating neutrophilia with a bronchial obstructive response could be found, but no eosinophilia was observed (35). These results suggest endotoxin is related to neutrophilia other than eosinophilia. In the present study, we noted no neutrophilia in both airway and circulation (the findings in circulation would be reported separately), therefore, the possibility that the observed eosinophilia in our study were due to endotoxin contamination could be excluded. On the other hand, we also observed that the inhalation of 0.4 ng LPS, which was equivalent to the dose contaminated in IL-5 and inhaled by each subject, was not able to cause any changes in airway reactivity at 2 h, 14 h, or 48 h after the inhalation (data not shown). Actually, even when the asthmatic patients were challenged with up to 2 µg LPS by inhalation, no airway obstruction and airway hyperresponsiveness could be observed (32). When the asthmatics were challenged with 20 µg inhaled LPS, increase in airway reactivity was shown at 4 to 6 h after LPS inhalation, and partially normalized at 24 and 48 h (31, 35). It would therefore be concluded that the observed airway hyperreactivity at 24 and 48 h after inhalation in this study was caused by IL-5, and was not endotoxin related.

In summary, we have shown that recombinant human IL-5 is functionally important in causing airway hyperresponsiveness to methacholine as well as airway eosinophilia and its activation. These data suggest that drugs that interfere with IL-5 synthesis or IL-5 receptor antagonists could be beneficial in the treatment of asthma.