Abstract
The common cold is a highly prevalent, uncomplicated upper airway disease. However, rhinovirus (RV) infection can lead to exacerbation of asthma, with worsening in airway hyperresponsiveness and bronchial inflammation. The current authors questioned whether such involvement of the intrapulmonary airways is disease specific.
Twelve nonatopic, healthy subjects (forced expiratory volume in one second (FEV1) >80% predicted, provocation concentration causing a 20% fall in FEV1 (PC20) >8 mg·mL−1) were experimentally infected with RV16. Next to PC20 and the maximal response to methacholine (MFEV1 and MV′40p), the numbers of mucosal inflammatory cells and epithelial intercellular adhesion molecule (ICAM)‐1 expression in bronchial biopsies were assessed before and 6 days after RV16 inoculation.
RV16 infection induced a small but consistent increase in maximal airway narrowing, without a change in PC20. There was a significant increase in bronchial epithelial ICAM‐1 expression after RV16, whereas inflammatory cell counts did not change. Nevertheless, the change in the number of submucosal CD3+ cells was correlated with the change in MV′40p.
In conclusion, rhinovirus infection in normal subjects induces a limited, but significant increase in maximal airway narrowing, which is associated with changes in bronchial T‐cell numbers. Together with the upregulation of bronchial epithelial intercellular adhesion molecule‐1, these findings indicate that, even in healthy subjects, rhinovirus infection affects the intrapulmonary airways.
This study was funded by “Stichting Astma Bestrijding”, the Netherlands.
Acute upper respiratory tract infection is the most prevalent illness in humans 1. It not only causes school and work absence, but also results in a high expenditure for treatment each year 1. The viruses most frequently involved are adenoviruses, influenza viruses, parainfluenza viruses, respiratory syncytial viruses, coronaviruses, and rhinoviruses 2, 3. The latter are the major cause of the common cold and have >100 serotypes 1–3. Fortunately, in healthy humans the common cold is usually a self-limiting upper airway disease with a short duration.
However, common colds in individuals with pre-existing airway diseases, such as asthma, are often associated with transient worsening of the disease, sometimes even leading to life-threatening exacerbations 4. Such clinical and epidemiological data are supported by human studies using experimental rhinovirus (RV) infection. Asthmatic subjects who were experimentally infected with RV16 showed worsening of symptoms, variable airway obstruction, and airway hyperresponsiveness, associated with both increased sensitivity 5 and augmented maximal airway narrowing to bronchoconstrictor agents 6. These physiological findings strongly suggest lower (intrapulmonary) airway involvement. Indeed, an increase in bronchial mucosal lymphocytes and eosinophils was found during an experimental RV16 cold in a study using a combined group of asthmatic and nonasthmatic subjects 7. However, it remains to be established whether such lower airway inflammation is disease specific or whether it equally occurs in nonatopic nonasthmatic subjects.
In this study, the null-hypothesis, that RV infection in healthy subjects does not result in worsening of airway responsiveness or demonstrable lower airway inflammation, was tested. Therefore, the dose/response curve to inhaled methacholine and the degree of mucosal inflammation in bronchial biopsy specimens before and after experimental RV16 infection in a carefully selected group of nonatopic nonasthmatic subjects was investigated.
Materials and methods
Subjects
Twelve nonsmoking, nonatopic, nonasthmatic healthy subjects were recruited (table 1⇓). They had a baseline lung function (spirometry) within the normal range (forced expiratory volume in one second (FEV1) >80% predicted), and normal responsiveness to inhaled methacholine (provocative concentration causing 20% fall in FEV1 (PC20) >8 mg·mL−1) 8. Their nonatopic status was determined by negative skin-prick tests to 16 common aero-allergen extracts (ALK Abelló, Nieuwegein, the Netherlands). The subjects had low titres of circulating antibodies specific to RV16 using a RV16 serum neutralisation assay (≤1:2 serum dilution against 20–25×50% tissue culture infective dose (TCID50)). The study was approved by the Medical Ethics Committee of the Leiden University Medical Center, Leiden, the Netherlands, and the subjects gave their written, informed consent.
Design
Experimental RV infection was performed by RV16 inoculation on 2 consecutive days (days 0 and 1). Methacholine challenges were performed on days −3, 5 and 14, to obtain the degree of airway sensitivity and maximal airway narrowing. Blood samples were drawn on days −2, 3, 6 and 27, and nasal washes were carried out on days −2, 3 and 6. Bronchial biopsies were taken by fibreoptic bronchoscopy on days −2 and 6.
Rhinovirus-16 inoculation
The RV16 inoculum was obtained from the same strain and stock as used in previous experiments in humans in vivo 5, 6. A total dose of 0.6–24×104 TCID50 diluted in 3 mL Hanks' balanced salt solution (HBSS) with 0.5% (weight/volume (w/v)) gelatin was administered to each subject according to a previously described procedure 5, 6. Confirmation of RV16 infection was established by an at least four-fold increase in virus-specific neutralising antibody titre in serum and/or by recovery of the virus from the nasal washes. Possible intercurrent respiratory infections were excluded 5, 6. In addition, the subjects scored their cold symptoms three-times daily on a four-point scale 5. The highest cold score after infection is presented in table 1⇑.
Leucocyte counts in peripheral blood
Before and on days 3 and 6 after virus inoculation, differential leucocyte counts were assessed by automated blood count analysis (Technicon H1; Technicon, Tarrytown, NY, USA).
Airway responsiveness
Standardised methacholine challenge tests were performed according to the 2‐min tidal breathing method 8. The response was measured by FEV1 and the more sensitive partial flow/volume curves (flow when 40% of the forced vital capacity remains to be expired (V′40p)) 5, 8. Airway sensitivity was determined by the provocative concentration of methacholine causing a 10% fall in FEV1 (PC10FEV1) and 40% fall in V′40p (PC40V′40p). Maximal responses (MFEV1 and MV′40p) were calculated by averaging the consecutive points on the dose-response plateau or, in absence of a plateau, by taking the highest data point 6.
Bronchoscopy and biopsy processing
Fibreoptic bronchoscopy was performed according to a standardised and validated protocol 9. Six biopsy specimens were taken at (sub)segmental level. Three biopsies were immediately fixed in phosphate buffered saline (PBS) buffered formalin 10% (v/v) and three were embedded in ornithyl carbamyl-transferase (OCT) medium (Miles Inc. Diagnostics Division, Elkhart, USA) and snap-frozen in isopentane cooled by iced carbon dioxide 9. The formaline-fixed biopsies were embedded in paraffin and stored until further processing. The snap-frozen samples were stored in airtight containers at −70°C.
Immunohistochemical staining of bronchial biopsy specimens
Cell markers
Immunohistochemical staining on paraffin embedded tissue was performed on 3 µm thick biopsy sections. Cell-type specific antibodies against CD3, CD4, CD8, CD68, tryptase (AA1) and neutrophil elastase were obtained from DAKO (Glostrup, Denmark), and the antibody against eosinophils (EG2) was purchased from Pharmacia (Uppsala, Sweden). Antigen expression was demonstrated with appropriate dilutions of the primary antibodies, followed by a secondary biotinylated antibody and a tertiary complex of streptavidin-biotin conjugated to horseradish peroxidase (SABC/HRP). 3‐amino‐9‐ethyl-carbazole (AEC) was used as a chromogen. The horseradish peroxidase conjugated antimouse EnVision system (DAKO) was used for the detection of CD4, with NovaRED (Vector, Burlingame, CA, USA) as the chromogen. The sections were counterstained with Mayer's haematoxyline (Klinipath, Duiven, the Netherlands). For negative controls, the primary antibody was omitted from this procedure 9.
Intercellular adhesion molecule‐1
Since studies in asthma have demonstrated that bronchial mucosal ICAM‐1 is upregulated by RV infection 10, staining for ICAM‐1 was performed on 4 µm sections of the snap-frozen tissue using monoclonal anti-CD54 as primary antibody (clone MEM-111; Monosan, Uden, the Netherlands) 10.
Analysis of bronchial biopsies
Cell markers
Digital images from the stained sections were obtained using a three-chip colour camera (1.732×106 pixels; 1320×992 µm2; 3×256 grey values) (Zeiss Vision KS-400 system; Kontron/Zeiss, Weesp, the Netherlands). Fully-automated or point-interactive (CD3+ and CD4+ cells in epithelium) cell counts were performed in the epithelium and in the lamina propria by a validated method 9. Positively stained cells were expressed as the number of cells·0.1 mm−2.
Intercellular adhesion molecule‐1
Staining of ICAM‐1 in the epithelium was analysed semiquantitatively, using a four-point scale for intensity of the staining (0=absent, 1=weak, 2=medium, 3=intense) 10. Variability of the intra-observer scoring of the semiquantitative ICAM‐1 staining, expressed as weighted kappa (κw), was satisfactory (κw>0.7).
Statistical analysis
PC10FEV1, PC40V′40p, and cell counts were log-transformed before statistical analysis. Paired t‐tests were applied for comparing cell counts and functional data before and after infection. The Wilcoxon signed-ranks test was applied to test for differences in ICAM‐1 staining. Relationships between outcome parameters were investigated using Spearman correlation tests. A p‐value of ≤0.05 was considered statistically significant.
Results
RV16 infection was confirmed in all subjects, except patient 10 who was excluded from analysis. Patient 1 dropped out on day 6 after the virus inoculation, because of lignocaine-associated adverse effects during the second bronchoscopy. No other respiratory viruses were detected in any of the nasal washings (table 1⇑). The common cold symptom score increased significantly after infection compared with before (mean±sem) (day −3: 0.91±0.34; postinfection: 3.73±0.73; p=0.003). The differential leucocyte counts in peripheral blood showed a significant decrease in lymphocyte numbers on day 3, as compared with day −2 (day −2: 35.5±1.4; day 3: 29.9±1.3%; p=0.006), with a subsequent significant increase back to baseline value on day 6 (35.9±1.6%; p=0.0008). The neutrophil counts in peripheral blood, however, did not change significantly during RV16 infection (days −2, 3 and 6: 51.6±1.7, 55.3±1.6, 51.3±1.5%; p>0.08).
Airway responsiveness
During the study, there were no significant changes in baseline FEV1 as compared to the values at entry of the study (p>0.1). After infection, there were no significant changes in PC10FEV1 and PC40V′40p (p>0.1). However, the maximal response to methacholine, as measured by both MFEV1 and MV′40p, increased significantly 5 days after virus infection as compared with before. At day 14, both MFEV1 and MV′40p were not significantly different from baseline values (p>0.1) (fig. 1⇓).
Bronchial biopsies
Cell markers
There were no significant changes in the numbers of EG2-, elastase-, AA1-, CD68-, CD3-, CD4-, and CD8-positive cells after RV infection as compared to before, both in epithelium (p>0.4) and in the lamina propria (p>0.2) (tables 2⇓ and 3⇓). However, there was a significant and positive correlation between the change in CD3+ cell numbers in the lamina propria and the change in maximal airway narrowing as measured by MV′40p (between days −3 and 5: rs=0.65, p=0.03) (fig. 2⇓).
Intercellular adhesion molecule‐1
RV16 infection was associated with a significant increase in ICAM‐1 expression of the bronchial epithelium (scores 0, 1, 2 and 3: preRV16: 0, 37.5, 62.5 and 0%; postRV16: 0, 0, 62.5 and 37.5%; p=0.048).
Discussion
The results of this study show a small, but consistent, effect of experimental RV16 infection on maximal airway narrowing to methacholine in nonatopic, nonasthmatic volunteers. There was a significant increase in ICAM‐1 expression in the bronchial epithelium. Even though there were no significant changes in bronchial inflammatory cell counts as such, it appeared that the changes in CD3+ cells in the bronchial lamina propria occurring after RV16 infection were positively correlated with the changes in maximal airway narrowing. These data indicate that a common cold leads to physiological and cellular changes in the intrapulmonary airways, even in normal subjects.
There are very few comparative data on the relationship between the physiological and inflammatory outcomes of RV infection in nonasthmatic healthy controls. This is of fundamental importance in order to assess the disease-specific mechanisms of virus-induced asthma exacerbations 11. Firstly, when considering the physiological changes in normal subjects, the observation of an increase in maximal airway narrowing extends previous findings as obtained by experimental RV infection in asthmatics 6. Secondly, systemic effects, such as the fall in peripheral lymphocytes, do occur in RV-infected normal subjects, as has been observed in asthmatics 6, 7. However, at the level of histopathology in the intrapulmonary airways, normal subjects seem to behave differently from asthmatics 7, 12, since significant increases in airway infiltrative cells were not observed.
All the study procedures were carried out according to validated protocols. The sample size of 12 subjects was carefully chosen, especially concerning the analysis of intrapulmonary cell counts. With the current analysis of 11 subjects, the statistical power to detect a two-fold change in, for example, mucosal eosinophils or PC20, was 0.77 and 0.85, respectively 8, 9. Objective confirmation of infection with RV16 was established by a rise in antibody titre and/or positive virus culture. In this study, in nonatopic normal subjects, the cold scores were relatively low, especially when compared with the studies by Grünberg and coworkers 10, 12, in which the same virus batch was used. Apparently, nonatopic healthy subjects have less severe cold symptoms than asthmatics, even though this is still a controversial issue 13, 14.
How can these data be interpreted in normal subjects? In this study, both the physiological and cellular findings are indicative of intrapulmonary effects of RV infection in nonasthmatic subjects. The increase in maximal airway narrowing per se is likely to reflect small airway involvement 15. This might be associated with effects of RV on microvascular leakage leading to airway wall swelling 16, or direct effects on airway smooth muscle 17, 18. It has been suggested that infection of the bronchial epithelial cells in the lower airways by RV 19, 20 may be enhanced by the virus-induced upregulation of ICAM‐1 21. Indeed, the present findings confirm such increased epithelial ICAM‐1 expression during RV infection in healthy subjects similar to the observations in asthmatic subjects 10. Release of various mediators by infected epithelial and other cells 22 can attract inflammatory cells, such as eosinophils, neutrophils and T‐lymphocytes, thereby enhancing intrapulmonary inflammation. An increase in CD3+ T‐cells in the epithelium or the lamina propria in normal subjects, as has been observed in asthma 7, 12, was not observed. However, the correlation between the changes in maximal response and those in T‐cell counts supports a possible inflammatory origin of physiological changes. Bronchial eosinophilic infiltration after RV has been controversial in asthma 7, 12. The absence of an increase in eosinophil counts in normal subjects is in agreement with the current authors' recent observations in asthma 12. However, a virus-induced increase in the activity of eosinophils in normal subjects, as has been demonstrated in asthmatics by Grünberg et al. 23, who showed elevated sputum ECP levels after RV infection in asthmatics, cannot be excluded.
In conclusion, rhinovirus infection leads to systemic and intrapulmonary effects even in nonatopic, nonasthmatic subjects. However, changes in airway responsiveness and bronchial biopsies are certainly limited. The authors speculate that any differences in response to a common cold between normal and asthmatic subjects might be explained by host factors, such as pre-existing allergic inflammation and/or airway hyperresponsiveness 24, thereby potentially facilitating the infection, intrapulmonary inflammation, and clinical worsening in pre-existing disease.
Acknowledgments
The authors would like to sincerely thank all of the volunteers in the study. They also thank J.J. Brahim, A. ten Brinke, J. Stolk, and L.N.A. Willems of the Dept of Pulmonology, Leiden University Medical Center, Leiden, the Netherlands for their skilled performance of the bronchoscopies, and A.C.J.A. Tiré for technical assistance during those bronchoscopies. The authors also acknowledge the technical assistance of C.J.G. van Zeijl-van der Ham of the Virology Dept (Leiden University Medical Center).
- Received May 23, 2001.
- Accepted February 20, 2002.
- © ERS Journals Ltd