INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Asthma is a chronic disease characterized by episodic chest tightness and wheezing associated with variable airway obstruction (1). At the level of the bronchi, chronic inflammation and epithelial shedding are regarded to be characteristic to airway pathology in asthma (2). The inflammatory process is considered to be mediated by the selective release of cytokines and other mediators by infiltrating leukocytes, but also by resident cells within the airways (3), bringing about the characteristic accumulation and activation of eosinophilic granulocytes and other accompanying features such as airway wall edema and luminal exudation (4).

Asthma exacerbations appear to be associated in time with respiratory virus infections (5). The detection rate of such viruses during wheezing episodes has been reported to be as high as 83% in 9- to 11-yr-old children, rhinovirus accounting for approximately 50% of the identified viruses (5). Furthermore, experimental rhinovirus infections in asthma and atopic rhinitis have been shown to transiently enhance airway hyperresponsiveness to histamine (6, 7) and to increase the maximal bronchoconstrictive response to methacholine in mildly asthmatic subjects (8).

These functional changes in asthma suggest a rhinovirus- induced increase in airway inflammation (9). Indeed, Fraenkel and coworkers (6) recently described an increase in numbers of submucosal T-lymphocytes, without a change in the CD4/ CD8 ratio in bronchial biopsies after experimental RV16 infection in a group of 11 normal subjects and six atopic asthmatics. Moreover, the numbers of activated eosinophils in the epithelium increased in the acute phase of infection in both normal subjects and asthmatics, whereas, interestingly, in the latter group the eosinophils counts remained elevated even after recovery. These findings are suggestive of rhinovirus-induced eosinophilic airway inflammation in asthma.

One of the potential explanations for enhanced cellular infiltration into the airways after infection may be virus-induced production of chemoattractant mediators such as chemokines or other cytokines within the airway mucosa. In vitro studies have shown that bronchial epithelial cell lines and primary epithelial cell subcultures produce several cytokines in response to incubation with rhinovirus (10). This is in keeping with results in human in vivo, showing an increase in the levels of IL-8 in the nasal lavage after experimental rhinovirus infections (7). Hence, it can be postulated that epithelial-derived chemokines are involved in the cellular and physiologic responses of the airways to a rhinovirus infection.

In the present study we hypothesized that experimental RV16 infection in asthmatic subjects induces changes in the cellular constituents and the production of cytokines within the airways in atopic, mildly asthmatic subjects. To that end, we applied the recently validated, noninvasive technique of sputum induction by means of inhalation nebulized hypertonic saline. We performed sputum inductions and measured lung function and airway responsiveness to histamine before and twice after experimental RV16 infection in atopic, mildly asthmatic subjects in vivo. Sputum was examined for cell differentials and for albumin, ECP, IL-8, and IL-6 in sputum supernatant. The cellular origin of one of the cytokines was investigated by intracellular IL-8 protein staining.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Twenty-seven nonsmoking, atopic subjects with mild persistent asthma participated in this study (1). The subjects' characteristics are listed in Table 1. Their FEV₁ was greater than 70% of the predicted value, and the provocative concentration of histamine causing a 20% fall in FEV₁ (PC₂₀) was smaller than 8 mg/ml (1). Atopy was examined by skin prick test, using a panel of 10 common aeroallergens (Vivodiagnost, ALK, The Netherlands). The subjects did not report symptoms of a common cold in the month preceding the study, and they had not used inhaled or oral corticosteroids for at least 3 mo, nor had they used theophyllines, antihistamines, sodium cromoglycate, or neodocromil sodium for at least 6 wk preceding the study. Symptoms of asthma were stable and controlled by on-demand usage of inhaled salbutamol alone, which was withheld for at least 8 h before the measurements. Before entry into the study, four of eight placebo-treated subjects did not have detectable levels of circulating neutralizing antibodies against RV16 (titer  1:1) against 25 times the 50% tissue culture infective dose (TCID₅₀), whereas four subjects had titers in the range of 1:2 to 1:128. In the RV16 group 14 of 19 subjects had no detectable neutralizing antibodies, whereas five subjects had low titers, between 1:2 and 1:16. The study was conducted during the months of July to December. The study was approved by the Medical Ethics Committee, and informed consent was obtained from all participants.

Design

The study had a double-blind parallel placebo-controlled design. Each subject was screened for the inclusion and exclusion criteria before entry into the study. Before inoculation with RV16 or placebo a sputum induction (Day 5) and a histamine challenge were performed (Day 3). Subsequently, RV16 or placebo was inoculated on two consecutive days (Days 0 and 1). Thereafter, sputum induction was repeated at Days 2 and 9 after the first inoculation, and the histamine challenge was repeated at Days 4 and 11. The sputum inductions were preceded by nasal lavage.

RV16 Inoculation

The RV16 strain and stock was the same as used in previous experiments in humans in vivo by others (13) and by ourselves (7, 8). The virus was cultured according to standards of good laboratory practice, and the inoculum was tested to be safe for human in vivo usage (14). Nasal inoculation of the RV16 was performed following a previously described protocol (8). Briefly, a dose of 0.5 to 2.9 × 10⁴ was nasally administered by inhalation from a nebulizer (DeVilbiss 646; DeVilbiss Co., Somerset, PA), spraying by atomizer (DeVilbiss 286), and instillation by pipette into both nostrils. This procedure was repeated after 24 h.

Confirmation of Infection

A fourfold or greater increase in RV16-specific neutralizing antibody in serum, and/or recovery of RV16 from the nasal lavages was considered to be confirmative of infection. Serum titers of neutralizing antibodies were determined by a neutralization assay, using 25 TCID₅₀ RV16. Nasal lavages were inoculated onto human embryonic lung fibroblasts (HEL) cultures and incubated at 32° C for 14 d. In the positive cultures RV16 was identified by neutralization assay, using RV16 specific guinea pig immune serum (1126AS /GP-VR; American Type Tissue Culture Collection, Rockville, MD). All nasal lavages were also inoculated onto LLC-MK2 cell cultures, Hep2 cell cultures, and HEL cell cultures and cultured at 37  ° C in order to exclude other intercurrent respiratory virus infections. In addition, the subjects recorded their cold symptoms, including sneezing, sore throat, nasal discharge, stuffy nose, headache, cough, malaise, and chills on a 4-point scale, ranging from none (0) to severe (3). These scores were added up to a maximum of 24, and they are referred to as cold score. Cold scores were recorded three times daily throughout the study period (7 ). The highest cold score recorded after inoculation is presented in Table 1.

Histamine Challenge

The histamine challenge was performed using histamine-biphosphate (Bufa, Uitgeest, The Netherlands) according to a standardized procedure (15). Serial doubling concentrations of nebulized histamine (0.03 to 8 mg/ml; DeVilbiss 646 nebulizer, output 0.13 ml/min) were inhaled by tidal breathing for 2 min at 5-min intervals with the nose clipped. The response was measured by FEV₁, using a dry rolling-seal spirometer (Morgan spiroflow; P. K. Morgan, Rainham, UK) and an analogue recorder (X-Y recorder BD 90; Kipp, Delft, The Netherlands). Baseline FEV₁ was determined as the mean of three reproducible measurements. Subsequently, FEV₁ was measured after each dose. The test was discontinued when FEV₁ decreased by more than 20% from baseline. The provocative concentration causing 20% fall in FEV₁ (PC₂₀) was calculated by log-linear interpolation of the last two data points.

Sputum Induction and Processing

Sputum was induced and processed according to the method described by Fahy and colleagues (16), which was slightly modified (17). Sodium chloride (4.5% wt/vol) was aerosolized using an ultrasonic nebulizer (DeVilbiss Ultraneb 2000) with a calibrated mass median aerodynamic diameter of 4.5 µm, and the output set at 2.5 ml/min. The aerosol was inhaled through a tube 100 cm long with an internal diameter of 22 mm, connected to a two-way valve (Hans-Rudolph, Kansas City, MO) equipped with a mouthpiece while the subject's nose was clipped. The aerosol was inhaled for serial doubling time periods (0.5 to 8 min), and subsequently for repeated 5-min periods, for a maximum of 30 min. After each inhalation period FEV₁ was measured. Then, the subjects were instructed to rinse their mouths and throat and gargle thoroughly with water, and, if necessary to blow their noses. Subsequently, the subjects were encouraged to cough up and expectorate sputum into a plastic container. A sputum weight of at least 2 g was considered sufficient. If FEV₁ dropped by more than 20% from baseline value, 200 µg of salbutamol were administered.

Sputum Processing and Cell Differentials

In order to homogenize the sputum, the whole sputum samples were gently mixed with an equal weight of dithiothreitol 0.1% (wt/vol) (Sputolysin; Calbiochem, La Jolla, CA), using a wide bore plastic pipette, and placed in a shaking water bath at 37  ° C for 15 min. The samples were then centrifuged at 350 × g for 10 min. The supernatants were collected and stored at 70 ° C until further analysis. The pellets were resuspended in phosphate-buffered saline and filtered through gauze, and cytospins were made (200 × g for 3 min) (Shandon 3 cytocentrifuge; Shandon Southern Instruments, Sewickley, PA). Of each sputum sample two cytospins were stained with Diff-quik and coded. All cell differentials were made by one observer (H.H.S.), counting at least 500 cells per cytospin. In addition, at least four cytospins were dried and stored at 70° C pending intracellular IL-8 staining.

Biochemical Analysis

In sputum, supernatant albumin levels were assessed by nephelometric assay (Beckman, Brea, CA), IL-8 levels by enzyme-linked immunosorbent assay (ELISA) using a monoclonal antibody (mAb) against IL-8 (CLB, Amsterdam, The Netherlands), IL-6 levels by mAb ELISA (CLB), and ECP by fluoroenzyme immunoassay (FEIA) (Pharmacia, Uppsala, Sweden) according to the manufacturers' instructions. The detection limits of the ELISA for IL-8 and IL-6 were 4 and 1.5 pg/ml, respectively, and 4 pg/ml for the ECP FEIA. Addition of DTT to the recombinant cytokines resulted in > 90% recovery of the proteins in the ELISA.

Intracellular IL-8 Staining

The intracellular IL-8 staining procedure (18) was performed on thawed, paraformaldehyde-fixed cytospins, according to Dolhain and colleagues (19). In order to obtain and maintain permeabilization of the cells continuously, all washing and staining solutions contained saponin 0.1% (wt/vol) (Riedel-de Häen, Seelze, Germany). First, endogenous peroxidase (H₂O₂ 0.3% [vol/vol] azide 0.3% [wt/vol]) and nonspecific binding sites were blocked (fetal calf serum 5%; Gibco, Breda, The Netherlands). The actual staining procedure consisted of a monoclonal IgG₁ antibody against IL-8 (CLB), followed by horseradish peroxidase-linked rabbit antimouse and swine antirabbit antibody amplifying steps (DAKO, Glostrup, Denmark). Then, the cytospins were incubated in the dark with 0.1% H₂O₂ and diaminobenzidine (DAB) (0.5 mg/ml) for 30 min. Finally, the cytospins were counterstained with Mayer's Lösung and eosin. In pilot experiments it was shown that stimulation with LPS of peripheral blood polymorphonuclear cells and macrophages isolated from bronchoalveolar lavage greatly increased the intracellular staining for IL-8. Likewise, cells of the pulmonary epithelial cell line A549 cells stimulated with TNF- stained positively for IL-8, whereas unstimulated cells were always negative. TNF- stimulated A549 cells were therefore used as a positive control for the staining procedure. As a negative control, a second cytospin of all samples was stained with control mouse IgG₁ (CLB) instead of mAb anti-IL8. All cytospins were counted twice, counting at least 500 cells. Cells were characterized by cell type and presence or absence of the typical perinuclear IL-8 staining pattern (18, 19).

Statistical Analysis

The squamous cells in sputum cell differentials were expressed as a percentage of all cells, whereas all other cell types were expressed as a percentage of all nonsquamous cells. IL-8-positive staining cells were expressed as a percentage of all nonsquamous cells, or as a percentage of particular cell type. Eosinophils and lymphocytes only occasionally stained positively for IL-8, and they were therefore also lumped as one cell type in the statistical analysis. The mean data of the two differential cell counts were used in the analysis. Regrettably, data on sputum volume were lost. Therefore, cell concentrations could not be calculated. PC₂₀ and levels of ECP, IL-8, IL-6, and albumin were log-transformed before analysis in order to obtain normally distributed data. Changes in PC₂₀ were expressed as doubling dose (DD), and changes in levels of soluble markers were expressed in doubling concentrations (DC).

Multivariate analysis of variance (MANOVA) was applied, either with factor time for analysis of the separate groups, or with RV16 or placebo as a between-group factor, and time as a within-group factor for analysis of all subjects (20). Significant effects were explored using Student's paired t tests for within-group effects, and Student's unpaired t tests for between-group effects. Values of p < 0.05 were considered statistically significant (20). The summary statistics were expressed as mean ± SEM, or geometric mean ± SEM in DC or DD for the log-transformed data.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

One of the RV16-treated subjects (Subject 1 in Table 1) dropped out at 7 d after the first RV16 inoculation because of an asthma exacerbation that required treatment with oral prednisone. Two subjects (Subjects 6 and 15) did not produce sputum at all three time points and one subject (Subject 22) did not produce sputum at the baseline visit (Day 5).

In all RV16-treated subjects the infection was confirmed by a rise in RV16 neutralizing antibody titer in serum and/or at least one rhinovirus-positive culture of the nasal lavage, whereas in none of the placebo-treated subjects could a viral infection be demonstrated (Table 1).

Lung Function and Histamine Challenge

Baseline lung function was slightly higher in the RV16 group than in the placebo group (p = 0.04). During the study, there were no significant changes in FEV₁ either within the group (MANOVA, p  0.34) or between the groups (MANOVA, p = 0.61).

PC₂₀ to histamine was not significantly different between the groups at baseline (p = 0.93). In the placebo group there were no significant changes in PC₂₀ during the study (MANOVA, p = 0.67). In the RV16 group PC₂₀ decreased significantly at Day 4 as compared with a baseline (mean difference, 0.65 ± 0.25 DD; p = 0.2), but not at Day 11 (0.40 ± 0.30 DD; p = 0.19). These changes were not significantly different from the changes in the placebo group (p  0.10).

Sputum Cell Differentials

At baseline sputum cell differentials were not significantly different between the groups (p  0.32). There were no significant changes in cell differentials within the placebo group (MANOVA, p  0.68). In the RV16 group there was only a significant decrease in the percentage squamous cells at Days 2 and 9 after inoculation as compared with the baseline (mean change, 13.2 ± 5.2; p = 0.02 and 17.3 ± 5.5; p = 0.006, respectively) (Table 2). Changes in cell differentials were not significantly different between the placebo and RV16 groups (MANOVA, p  0.25) (Table 2). However, in the RV16 group the change in the percentage eosinophils between Days 5 and 9 correlated significantly with the change in PC₂₀ (r = 0.58, p = 0.02) (Figure 1A).

View larger version (14K): [in this window] [in a new window]   Figure 1.   (A) Relationship between the change in the percentage eosinophils in sputum and the change in PC20 in the RV16 group in the second week after infection as compared with baseline. (B) Relationship between the change in ECP in sputum supernatant and the change in PC20 in the RV16 group in the first week after infection as compared with baseline.

Sputum-soluble Mediators

At baseline, albumin levels were higher in the RV16 group than in the placebo group (mean difference, 2.35 DC; p = 0.045). However, within each group there were no significant changes in albumin levels during the study (MANOVA, p  0.10) (Table 3).

The groups were not statistically different at baseline with respect to the levels of IL-8, IL-6, and ECP (p  0.34). In the RV16 group the levels of IL-8 increased significantly at Days 2 and 9 after inoculation as compared with baseline (mean change, 1.18 ± 0.38 DC; p = 0.007 and 1.01 ± 0.48 DC; p = 0.049, respectively). In addition, IL-6 levels increased significantly during the study (mean change at Day 2, 2.89 ± 0.74 DC; p = 0.001; Day 9, 2.87 ± 0.78 DC; p = 0.002), whereas ECP levels increased only significantly at Day 2 (mean change, 0.68 ± 0.32 DC; p = 0.047). There were no significant effects on the biochemical markers in the placebo group (MANOVA, p  0.17), and the effects were not significantly different between the placebo and the RV16 groups (MANOVA, p 0.11) (Table 3). In the RV16 group, only the increase in ECP between Days 5 and 2 correlated significantly with the decrease in PC₂₀ between Days 3 and 4 (r = 0.60, p = 0.01) (Figure 1B).

Intracellular IL-8 Staining

Three samples were rejected because the quality of the cytospins was not adequate for interpreting the intracellular staining (Subject 14, Visit 1; Subject 22, Visit 3; Subject 16, Visit 3). A representative example of a cytospin stained for IL-8 and its negative control are shown in Figure 2. At baseline, the groups were not significantly different with respect to the percentage IL-8 positive nonsquamous cells (mean ± SEM; RV16, 7.0 ± 1.6%; placebo, 8.7 ± 3.6%; p = 0.64) or the percentages of the positively staining individual cell types (p  0.24). The IL-8 positive cells appeared to be predominantly neutrophils (Figure 3), which at baseline accounted for 77.9% of all positively staining nonsquamous cells. Within the individual cell types, 22.5 ± 4.1% of all neutrophils stained positively, and 2.6 ± 0.7% of the epithelial cells and macrophages stained positively. Lymphocytes and eosinophils only occasionally stained positively, whereas squamous cells were never observed to stain positively for IL-8.

View larger version (99K): [in this window] [in a new window]   View larger version (98K): [in this window] [in a new window]   Figure 2.   Example of typical intracellular staining pattern for IL-8 (A ) and the negative control (B) in sputum cells. Arrows indicate positively staining neutrophils.

View larger version (28K): [in this window] [in a new window]   Figure 3.   Intracellular IL-8 staining: negatively or positively staining cells, expressed as a percentage of nonsquamous cells in the RV16 group (upper panel ) and the placebo group (lower panel ). White: negatively staining nonsquamous cells. Black: IL-8 positive eosinophils and lymphocytes. Vertically hatched: IL-8 positive epithelial cells and macrophages. Horizontally hatched: IL-8 positive neutrophils. The upward and downward error bars represent the SEM for the negatively staining cells and the positively staining neutrophils, respectively. In the RV16 group the percentage negatively staining cells decreased significantly at Day 2 after inoculation (mean change ± SEM, 6.1 ± 2.4; p = 0.02), whereas the percentage positively staining neutrophils increased at this time point (mean change ± SEM; 5.4 ± 2.1; p = 0.02).

In the RV16 group the percentage IL-8 positive nonsquamous cells increased at Day 2 (mean change, 6.1 ± 2.4; p = 0.02), but this increase was no longer significant at Day 9 (mean change, 6.3 ± 4.0; p = 0.13) (Figure 3). The increase at Day 2 was due to an increase in the percentage IL-8 positive neutrophils (mean change, 5.4 ± 2.1; p = 0.02), with a trend towards an increase in the percentage positively staining epithelial cells and macrophages (mean change, 0.8 ± 0.4%; p = 0.07) (Figure 3). There was no significant change in the percentage IL-8 positive nonsquamous cells in the placebo group (MANOVA, p = 0.23), and the changes were not significantly different between the groups (MANOVA, p = 0.89). In the RV16 group, the change in the percentage IL-8 positive neutrophils between Days 2 and 9 correlated significantly with the change in IL-8 levels in sputum supernatant (r = 0.64, p = 0.008), whereas there was no significant correlation between changes in IL-8 positive staining cells and changes in airway hyperresponsiveness (p 0.19).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study we have demonstrated that RV16 infection in atopic asthmatic subjects does not change the cellular composition of the sputum or levels of albumin in sputum supernatant. However, an increase in ECP levels was shown in the first week after infection. The increase in airway hyperresponsiveness to histamine correlated positively and significantly with the increase in ECP during the first week, and with the change in the percentage eosinophils during the second week after infection. Furthermore, RV16 infection induced an increase in the levels of IL-8 and IL-6 for as long as 9 d after infection. When staining IL-8 in permeabilized sputum cells, it appeared that IL-8 is predominantly present in neutrophils and that the percentage IL-8 positive neutrophils increased at Day 2 after infection. These results suggest that experimental rhinovirus infection increases lower airway inflammation, involving the release of proinflammatory cytokines into the airway lumen.

This is the first study showing changes in inflammatory mediators in the airways after rhinovirus infection in asthmatic subjects. The relationship between RV16-induced changes in eosinophil markers and changes in airway hyperresponsiveness are in keeping with cross-sectional studies, demonstrating a correlation of sputum eosinophilia with lung function (21, 22) during exacerbations of asthma of unknown cause. The present study extends the observation by Fraenkel and coworkers (6), showing an increase in activated (EG₂-positive) eosinophils in the bronchial epithelium after experimental RV16 infection in a group of normal and asthmatic subjects. Hence, the correlation between the increase in ECP, the change in the percentage eosinophils, and the change in airway hyperresponsiveness support a role for the eosinophil in rhinovirus-induced exacerbations of asthma.

The presently described increase in IL-8 levels after RV16 infection, together with intracellular staining for IL-8 predominantly in neutrophils, is in keeping with a relatively high percentage of neutrophils and high levels of IL-8 in sputum found during acute exacerbations of asthma (23). On the basis of the chemotactic properties of IL-8 for neutrophils, T-lymphocytes, basophils, and primed eosinophils (24), and the pleiotrophic effects of IL-6, including B-cell activation and proliferation of cytotoxic T-cells (25), the involvement of these cytokines in the immune response against a rhinovirus cold could be postulated. This is further supported by an increase of neutrophil numbers in the nasal lavage (26) and peripheral blood (7), and the increase of T-lymphocytes in the bronchial submucosa during a rhinovirus cold (6). However, the role of IL-8 and IL-6 in the development of rhinovirus-induced changes in airway physiology remains to be elucidated since we did not find a correlation between changes in the percentages neutrophils and lymphocytes, IL-8, IL-6 levels, or IL-8 positive staining cells and changes in airway hyperresponsiveness.

In the present study, we applied a placebo-controlled design in order to differentiate the responses to rhinovirus inoculation from fluctuations in airway pathophysiology that are characteristic of asthma. Despite evident changes in the soluble mediators in the RV16 group, there were no significant differences between the RV16 and placebo groups. One could speculate that this was due to the smaller sample size of the placebo group, resulting in low statistical power to detect such between-group differences (17).

Validated methods were used to inoculate the rhinovirus and measure the physiologic responses (7, 8, 13, 15). The natural way of transmission was mimicked by using a combination of three methods of virus administration, including nasal inhalation. In this way, virus particles may even have reached the lungs. However, we have not attempted to assess lower airway infection by viral culture of the sputum since oropharyngeal contamination would have biased a positive result.

Sputum induction was performed according to a validated technique (17), which included careful avoidance of oropharyngeal contamination during the sputum induction, thereby reducing, (but not excluding) possible bias by mixing the sputum sample with upper airway secretions and saliva (27). In cross-sectional studies, the sputum eosinophil numbers and ECP levels have been shown to correlate significantly with those in bronchoalveolar lavage and bronchial wash (28), whereas a trend towards a significant correlation has been shown between eosinophils in sputum and the submucosa of bronchial biopsies (29, 30). Moreover, sputum analysis has revealed a decrease in eosinophil numbers and ECP levels after glucocorticoid treatment in asthma (31), in accordance with findings in bronchial biopsies (32). Therefore, the cellular and biochemical constituents of hypertonic saline-induced sputum seem to be an adequate reflection of rhinovirus-induced airway inflammation in asthma.

The decrease in the percentage of squamous cells in the RV16-treated subjects might have been caused by an increase in sputum volume and/or sputum cell concentration after infection. Unfortunately, the data on sputum volume were lost in the present study, which precluded calculations of cell concentrations. However, the cell differentials by themselves seem to provide the most useful information since these appear to be rather stable under conditions in which absolute cell counts are variable because of volume dilution effects (27).

How do we interpret the results? The increase in airway hyperresponsiveness to histamine, in the absence of a significant decrease in lung function, may be explained by inflammatory changes such as airway wall swelling or thickening, potentiating the airway narrowing effect of smooth muscle shortening (9). Such inflammation may be characterized by an infiltrate of inflammatory cells, typically eosinophils and lymphocytes, and the release of proinflammatory mediators, which may bring about features such as vasodilation and vascular hyperpermeability, oxidative stress, cellular activation, and tissue damage. The current association between changes in PC₂₀ and changes in sputum ECP and the percentage eosinophils suggest that eosinophilic inflammation is involved in rhinovirus-induced enhancement of airway hyperresponsiveness.

There are several mechanisms potentially driving the cell infiltration and activation after RV16 infection. First, airway inflammation may be driven by T-lymphocytes infiltrating the airways (6, 33), and producing several cytokines such as IL-2, IFN-, IL-4, and IL-5 (3), thereby orchestrating the augmentation of the immune response. Second, there is increasing evidence that cultured epithelial cells and fibroblasts produce proinflammatory cytokines such as IL-8, IL-6, GM-CSF, and RANTES in response to infection with several rhinovirus serotypes, including RV16 (10). Other airway cells such as macrophages release TNF- and IL-1 upon uptake of rhinovirus (34), which may in turn induce and potentiate cytokine production in, for example, epithelial cells and neutrophils (10). It is presently uncertain as to whether rhinovirus actually infects lower airway tissues in humans in vivo, although there is evidence to suggest that rhinovirus is present in the lower airways during a cold (35). Thus, one could envision the interaction between airway resident cells in the induction of an immune response involving the release of chemotactic mediators such as chemokines, which may lead to infiltration of inflammatory cells and subsequent pathophysiologic changes.

What are the clinical implications of this study? The present results suggest that a common cold increases airway hyperresponsiveness, together with an increase in lower airway inflammation. This fits in with the epidemiologic association between respiratory virus infections and exacerbations of asthma (5). Glucocorticoids are the treatment of choice for exacerbations of asthma (1). In view of their effectiveness against eosinophilic inflammation (31, 32) it could be postulated that our data support a rationale for such therapeutic use of glucocorticoids during rhinovirus-induced exacerbation of asthma. However, it remains to be established whether steroids can actually prevent the resulting pathologic and functional changes in the airways.

We conclude that experimental RV16 infection in atopic asthmatic subjects enhances airway hyperresponsiveness to histamine in conjunction with a rise in sputum levels of IL-8, IL-6, and ECP, which suggests an increase in lower airway inflammation. The relationship between changes in ECP, the percentage eosinophils, and changes in airway responsiveness suggests that eosinophilic inflammation is involved in rhinovirus-induced enhancement of airway hyperresponsiveness. In order to further investigate the mechanisms of rhinovirus- induced exacerbations of asthma, the state of activation and cytokine production of resident cells and infiltrating cells within the airway wall/mucosa need to be examined.