- Email: [email protected]
The effect of an experimental rhinovirus 16 infection on bronchial lavage neutrophils
The effect of an experimental rhinovirus 16 infection on bronchial lavage neutrophils Nizar N. Jarjour, MD, James E. Gern, MD, Elizabeth A. B. Kelly, PhD, Cheri A. Swenson, BS, Claire R. Dick, BS, and William W. Busse, MD Madison, Wis Background: Viral respiratory tract infections are the most frequent cause of asthma exacerbations. Of the respiratory viruses associated with these exacerbations, rhinovirus (RV) is the most common. It is proposed that these RV infections may enhance airway inflammation and thus provoke asthma. Objective: It is our hypothesis that RV infections generate nasal proinflammatory mediators that are associated with an initial increase in circulating leukocytes and may contribute to later development of neutrophilic airway inflammation. Methods: To evaluate this hypothesis, subjects with a history of allergic asthma were experimentally inoculated with strain 16 RV (RV16). The effect of this experimental infection was evaluated on circulating leukocytes, nasal-derived mediators, and markers of bronchial inflammation that were obtained by bronchoscopy and lavage. Results: RV16 inoculation was associated with an initial increase in circulating neutrophils. Paralleling these acute changes in circulating neutrophils was an increase in nasal concentrations of IL-8 and granulocyte–colony-stimulating factor (G-CSF). The RV16-associated changes in circulating and nasal G-CSF correlated with increases in peripheral blood neutrophils (rs = 0.874, P < .001 and rs = 0.898, P < .001, respectively). Bronchial lavage samples showed no increase in neutrophils 48 hours after RV16 inoculation; however, 96 hours after RV inoculation there was a significant increase in bronchial neutrophils compared with preinoculation values. Conclusions: These results suggest that the production of nasal mediators associated with the RV infection, particularly GCSF, may be important to the eventual development of neutrophilic bronchial inflammation and thus contribute to asthma exacerbations. (J Allergy Clin Immunol 2000;105:1169-77.) Key words: Virus-induced asthma, rhinovirus, asthma, neutrophils
Viral respiratory tract infections are the most frequent cause of asthma exacerbations.1-4 Of the respiratory viruses associated with these asthma exacerbations, the common cold virus, rhinovirus (RV), has been the most frequently identified illness in children and adults.2,3 To
From the Departments of Medicine and Pediatrics, University of Wisconsin Medical School, University of Wisconsin Hospital, Madison, Wis. Supported by National Institutes of Health grants No. AI 34891 and AI 26609. Received for publication Dec 7, 1999; revised Feb 10, 2000; accepted for publication Feb 10, 2000. Reprint requests: William W. Busse, MD, University of Wisconsin Medical School, Departments of Medicine and Pediatrics, University of Wisconsin Hospital, H6/367 Clinical Science Center, 600 Highland Ave, Madison, WI 53792-3244. Copyright © 2000 by Mosby, Inc. 0091-6749/2000 $12.00 + 0 1/1/106376 doi:10.1067/mai.2000.106376
Abbreviations used BL: Bronchial lavage EDN: Eosinophil-derived neurotoxin G-CSF: Granulocyte–colony-stimulating factor LTB4: Leukotriene B4 MPO: Myeloperoxidase RV16: Rhinovirus strain 16 TCID50: Tissue culture infective dose 50% WBC: White blood cell
evaluate the mechanisms of an RV infection on asthma exacerbations, we have experimentally inoculated subjects with RV, including strain 16 (RV16).5-7 This experimental respiratory infection has caused a symptomatic illness, an increase in bronchial responsiveness,5,6 and an enhanced allergic inflammatory response to allergen.5-7 Because RV infections are, at least initially, confined to the upper airway, it is likely that the nasal immune response to this virus plays a key role in initiating an airway inflammatory response in patients with asthma.1 Thus we have hypothesized that RV infections generate proinflammatory cytokines in the nose and that these cytokines initiate neutrophilic responses in the blood and the upper and lower airways. To evaluate this hypothesis, we experimentally inoculated 8 asthmatic subjects with RV16 and evaluated timedependent effects of the respiratory infection on circulating leukocytes, nasal-derived mediators, and bronchial (by bronchoscopy) markers of inflammation.
Methods Subject selection Eight allergic asthmatic subjects were recruited for participation in this study (Table I). Each subject had a physician’s diagnosis of asthma, a history of occasional seasonal wheezing (presumably in relationship to environmental allergen exposure), and a prescription for an inhaled β-agonist. None of the subjects, however, required regular use of asthma medication or had a history of wheezing with a respiratory infection. Each subject had a positive prick skin test to one or more aeroallergens (including house dust mite, ragweed, grass, trees, cat, dog, and Alternaria antigen extract) and then underwent a medical history, physical examination, and pulmonary function testing by spirometry. These subjects did use any medications during the study, including bronchodilators, antihistamines, or topical nasal corticosteroids. We have previously found the effects of the RV16 experimental infection on airway responsiveness to be primarily limited to individuals with allergic disease.8 In addition, the effects of RV16 on markers of lower airway inflammation, when evaluated by bron1169
1170 Jarjour et al
J ALLERGY CLIN IMMUNOL JUNE 2000
TABLE I. The effect of the experimental RV16 inoculation on FEV1 values and RV16 antibody titers, virus titers, and cold symptoms Baseline FEV1 (L [% predicted]) Subject
Sex
Age (y)
1 2 3 4 5 6 7 8
M M F F F M M F
26 28 29 24 20 29 37 22
Precold
4.13 (91) 4.33 (99) 3.38 (115) 3.53 (129) 3.12 (106) 4.82 (103) 4.48 (102) 3.03 (81)
Serum RV16 antibody titers
Acute infection
Acute
4.17 (92) 4.33 (99) 3.38 (115) 3.11 (114) 3.07 (104) 4.72 (101) 4.49 (102) 2.81 (77)
<1 1 <1 <1 1.4 <1 <1 2.8
Convalescent
2.8 45 45 90.5 2 5.6 1.4 45
Peak nasal virus titers* TCID50/mL
101 103 103 105.5 104 105 103 101
Peak symptoms†
7 13 17 12.5 12 13 4 15
TCID50, Tissue culture infective dose 50%; M, male; F, female. *Peak TCID50/100 µL of nasal fluid during acute infection. †Peak subject-related symptom score is tabulation of 13 cold symptoms with highest possible score of 39.
choscopy and lavage, were principally confined to subjects with allergic disease.7 Because the goal of this study was to determine the effects of an experimental RV infection on upper and lower airway inflammatory events in asthma, a study population was selected in which such changes are likely to occur (ie, allergic individuals with a history of asthma). The study was approved by the University of Wisconsin–Madison Center for Health Sciences Human Subjects Committee. Informed consent was obtained from each subject before participation.
Study design The recruited subjects were evaluated during 2 separate phases: (1) precold and (2) cold (see Table II for study design). On study day 1 of the precold phase, a blood draw, nasal wash, and bronchoscopy with lavage were performed. Two days later (study day 3), a blood sample was drawn and bronchoscopy and lavage were repeated. These studies provided the baseline or precold data for comparison with similar temporal values during the cold phase. Approximately 1 month later the subjects entered the cold phase of the study (Table II). On day 1 of the cold phase each subject had a nasal wash performed to determine whether a respiratory infection was present at the time of inoculation. The subjects were then inoculated (see below for methods) with live RV16 on study days 1 and 2. Two days after the initial RV16 inoculation (study day 3) and at a time when all participants had symptoms of a cold, bronchoscopy, and airway lavage plus a blood draw were performed. The bronchoscopy, airway lavage, and blood draw were repeated 48 hours later (study day 5). Nasal washes were obtained daily during the acute cold phase.
RV16 inoculation, confirmation of a viral infection, and assessment of cold severity The technique for RV inoculation and subsequent symptom monitoring followed previously described methods.9,10 Briefly, no patients had neutralizing antibody in their sera (≤1:1) at the initial screening for entry into the study. (Note: subject 8 was found to have a titer of 2.8 at the time of inoculation although the screening neutralizing antibody titer was <1:1.) RV colds were induced in the subjects by instilling, on 2 successive days, 0.25 mL of RV16 suspension (320-3200 TCID50) into each nostril by pipette and then spraying approximately the same amount of virus inocula into each nostril with an atomizer (DeVilbiss, No. 286, Somerset, Pa) powered by a DeVilbiss 561 series compressor. Colds were evaluated with a questionnaire that was completed by the participant hourly during waking hours as previously described.9,11 Briefly, the symptoms evaluated included, but were
not limited to, cough, nasal discharge, sneezing, stuffy nose, headache, malaise, chills, or fever. Symptoms were graded on the following scale: 0 = absent, 1 = mild, 2 = moderate, or 3 = severe. Total symptom scores of ≥12 have been used to define a severe cold, 7 to 11 a moderate cold, and <7 a mild cold. Shed virus was assayed by titration of the nasal wash into diploid (WI-38) cell cultures. Nasal washes were performed daily during the acute infection. Prewarmed HBSS with 0.5% gelatin was instilled (5 mL into each nostril) into the nose, retained for 5 seconds, and then forcefully expelled into a sterile petri dish. The nasal wash fluid was then transferred to a test tube to assess volume and was vortexed and an aliquot centrifuged, creating a cell-free supernatant that was stored at –80°C for later analysis. The remainder of the nasal wash was inoculated into WI-38, primary rhesus monkey, and HEp-2 cell cultures; washings obtained after RV16 inoculation were placed into WI-38 cells only.9,10 Confirmation that the respiratory illness was caused by rhinovirus included at least one identified isolate for RV16 or by a 4-fold or greater increase in neutralizing antibody in the convalescent serum specimens to this strain of virus. Antibody titers were expressed as initial serum dilutions in which 1:1 = undiluted serum, 1:2 = one part serum and one part diluent, etc.
Bronchoscopy and bronchial lavage Bronchoscopy and bronchial lavage (BL) were conducted as previously described.7 Briefly, one bronchopulmonary segment was identified, and the fiberoptic bronchoscope was wedged into that segment. The bronchoscope was held in a wedge position and lavage (60 mL) was performed. This volume was selected to obtain a wash of fluids in the airway and to be a likely representative of air space cells and other mediators. A second bronchoscopy was done 48 hours later. At that time, BL was performed in a similar fashion. The use of paired bronchoscopies over 48 hours provided an opportunity to evaluate the effect of RV infection on the lower airway at two time points.
BL fluid analysis The BL fluid return, which was similar in volume at precold and during the acute RV16 illness, was centrifuged (400g for 10 minutes) to sediment cells. The supernatant was removed and stored at –80°C for later analysis. BL cells were washed twice with HBSS containing 2% newborn calf serum. A cell count was made with a hemocytometer. Cytocentrifuge slides were prepared, air-dried, fixed in methanol, and stained (Diff-Quick Scientific Products, Chicago, Ill). For the differential cell count, 300 leukocytes were enumerated and identified as lymphocytes, neutrophils, eosinophils, macrophages, or epithelial cells on the basis of staining and morphologic characteristics.
Jarjour et al 1171
J ALLERGY CLIN IMMUNOL VOLUME 105, NUMBER 6, PART 1
TABLE II. Study design Precold evaluations (study day) Procedure
1
Blood draw Nasal wash Bronchoscopy RV16 inoculation
X X X
2
Cold evaluations (study day)
3
1
2
3
4
5
6
7
X
X X
X
X X X
X
X X X
X
X
X
X
X
X, Procedure performed.
Determination of cytokines and other mediators BL fluid was concentrated 10-fold by a centrifugal concentration device containing a low protein-binding membrane YM filter with a molecular weight cutoff value of 3000 (Amicon, Beverly, Mass). Nasal washes and serum were analyzed without concentration. IL-8 and granulocyte-colony stimulating factor (G-CSF) were measured with a commercial kit according to the manufacturer’s directions (Bioscience, Camerillo, Calif, and R&D, Minneapolis, Minn, respectively). Concentrated BL fluid was also analyzed for IL-5, IFN-γ, IL1β, and TNF-α by a 2-step sandwich ELISA, as previously published.12 The concentrations of TNF-α, IL-5, and IFN-γ were calculated by comparison with a standard curve consisting of known amounts of recombinant human TNF-α, IL-5, or IFN-γ. The sensitivity of this assay is 5 pg/mL (equivalent to 0.5 pg/mL in neat BL fluid). Myeloperoxidase (MPO) concentrations were determined in the samples (nasal washes or bronchial lavage fluid) by a commercial RIA according to the directions of the manufacturer (Pharmacia Diagnostics, Columbus, Ohio). Eosinophil-derived neurotoxin (EDN) was determined in lavage fluid by methods previously described.13 Leukotriene B4 (LTB4) was also measured in the lavage fluid.14
Data analysis The data are presented as medians with 25% and 75% interquartiles. Comparisons between the results from the 2 separate lavage samples were done with the Wilcoxon test. Correlation between cytokine levels, viral titers, and proportions of various BL cells were evaluated with the Spearman rank correlation test.
RESULTS Patient characteristics and confirmation of RV16 infection (Table I) There was no effect of the RV16 infection on FEV1 values. All subjects, except number 8, had serum RV16 antibody titers that were ≤1:1 at the time of inoculation. Subject 8 had an antibody titer <1:1 at the time of initial screening; on the day of RV16 inoculation, however, the serum antibody titer was ≥2.8. This subject had symptoms, shedding of RV16 in the nasal cultures, and an increase in antibody titer in the convalescence serum sample (≥45). Consequently, data from subject 8 were included in the evaluations. All subjects had a symptomatic upper respiratory infection as indicated by evaluation of peak symptoms, which ranged from mild (7) to severe (>12).
Effect of RV16 infection on peripheral white blood cell counts Peripheral white blood cell (WBC) counts and cell dif-
ferential counts were determined precold and then after inoculation with RV16 (see Study design in Tables II and Table III). There was a nonsignificant increase in the percentage of blood neutrophils in the precold phase at study day 3; this may reflect the effect of the previous bronchoscopy. The RV16 infection caused a small but significant increase in the peripheral WBC count, which was detected in the blood sample taken 48 hours after day 1 virus inoculation compared with a similar evaluation time point precold (6.4 ± 1.8 × 104 cells/mL [precold] vs 7.8 ± 2.5 × 106 cells/mL [acute cold], P < .05, mean ± SD). There was also a significant (P < .02) increase in the percentage of circulating neutrophils (55.1% ± 5.1% [precold] vs 68.3% ± 11.3% [acute cold]). In contrast, and in keeping with published observations,15,16 the percentage of circulating lymphocytes significantly decreased during the acute phase of the RV16 infection, precold to acute cold (31.4% ± 9.3% vs 16.3% ± 11.2%, P < .02). Ninety-six hours after RV inoculation (study day 5, acute cold), both neutrophil and lymphocyte peripheral blood counts had returned to preinoculation values; at this time, however, there was an increase in eosinophils over values at the initial determination (acute cold study day 3) (1.9% ± 2.1% vs 5.1% ± 3.2%, P < .05).
Nasal and serum concentrations of cytokines Nasal fluids from the precold and cold phases of study were analyzed for G-CSF (Fig 1, A) and IL-8 (Fig 1, B); beginning 1 to 2 days after RV16 inoculation, there were increases in G-CSF and IL-8 in the nasal lavage fluid. The increase in these cytokine values persisted for up to 4 days after RV16 inoculation. Serum G-CSF values increased significantly after RV16 inoculation (5.5 pg/mL [precold] vs 62.0 pg/mL [cold], study day 3 median values, P < .05). A strong correlation was found between nasal and serum G-CSF values on study day 3 of the cold phase (48 hours post-RV16 inoculation, rs = 0.95, P ≤ .001). Correlations between nasal concentrations either of IL-8 or G-CSF and RV16 virus titers from the nasal lavage fluids were determined. Forty-eight hours after RV16 inoculation, both nasal IL-8 and G-CSF values showed striking correlations with quantitative measurements of RV16 shedding in nasal secretions (rs = 0.77, P = .02 and rs = 0.85, P < .005, respectively). Furthermore, the serum G-CSF value and nasal RV16 virus titer had a strong correlation (rs = 0.78, P = .015).
1172 Jarjour et al
J ALLERGY CLIN IMMUNOL JUNE 2000
A
B FIG 1. The effect of RV16 inoculation on nasal lavage fluid G-CSF (A) and IL-8 (B). Box blots represent median with 25% and 75% interquartiles; the error bars represent the 10th and 90th percentiles. n = 8. Asterisk, P < .05 compared with preinoculation (pre-inoc) values.
Relationships between serum and nasal G-CSF and changes in peripheral blood neutrophils Serum (Fig 2, A) and nasal G-CSF (Fig 2, B) values were evaluated for a possible relationship to changes in peripheral blood neutrophils from the precold (study day 1) to acute cold phase of study (study day 3, 48 hours after RV16 inoculation). There were strong correlations between nasal and serum G-CSF values and changes in peripheral blood neu-
trophils (rs = 0.898, P < .001 and rs = 0.874, P < .001, respectively). When comparisons were made between nasal concentrations of IL-8 and changes in peripheral blood neutrophils, the correlation between these 2 variables showed a trend but did not achieve significance (rs = 0.663, P = .072).
The effect of RV16 infection on the cellular composition of BL fluid The percent recovery of instilled saline solution (60 mL) was similar in the paired lavage samples (BL 1, pre-
J ALLERGY CLIN IMMUNOL VOLUME 105, NUMBER 6, PART 1
Jarjour et al 1173
A
B
FIG 2. Correlation between serum (A) and nasal lavage fluid (B) G-CSF and change in peripheral blood neutrophils. The serum and nasal lavage G-CSF values were obtained 2 days after RV16 inoculation. The change in blood neutrophils represents the difference between preinoculation (precold) and acute symptoms (study day 3) values and are represented as 106cells per milliliter.
cold = 54.4% ± 7.9% vs BL 1, cold = 56.3% ± 14.5%; BL 2, precold = 41.9% ± 11.0% vs BL 2, cold = 43.8% ± 6.8% [mean ± SD, n = 8]). There was no increase in the total WBCs in bronchial lavage related to the cold either at 48 (BL1) or 96 (BL2) hours after RV16 inoculation (Fig 3, A). There was, however, an increase in neutrophils in samples from the second lavage (precold BL1 vs pre-
cold BL2), median values 0.02 versus 0.11 × 106 cells, P < .05 (Fig 3, B), an effect previously described.17 To assess, in a limited fashion, the kinetics of the lower airway inflammatory response to nasal inoculation with RV16, bronchoscopy lavage was performed on 2 occasions after the RV16 inoculation (see above and Table II). Forty-eight hours after the RV16 inoculation, there was
1174 Jarjour et al
J ALLERGY CLIN IMMUNOL JUNE 2000
A
B FIG 3. The effect of RV16 respiratory infection on BL total WBC (A) and neutrophil (B) values. Bronchoscopy and lavage were performed on 2 occasions before cold ([BL-1] study day 1 and [BL-2] 48 hours later, precold study day 3); similar assessments were made during the acute experimental RV16 infection (2 [BL-1] and 4 days [BL-2] after RV16 inoculation). Data are expressed as the median values with 25% and 75% interquartiles, n = 8. Total cell values are expressed as cells (WBCs and neutrophils × 106 cells per BL fluid). Asterisk, P < .05 compared with BL values (precold and cold); dagger, P < .05 compared with BL2 precold.
no difference between BL neutrophils (cold BL1) and precold values (precold BL1). Two days later and approximately 96 hours after RV16 inoculation, there was a significant increase in lavage neutrophils over the initial values (cold BL1) and both the precold values
(precold BL1 and BL2) (Fig 3, B). Although bronchoscopy by itself may increase the numbers of neutrophils (see precold BL1 vs BL2), the increase in neutrophils in the second lavage fluid was significantly greater during the experimental cold.
Jarjour et al 1175
J ALLERGY CLIN IMMUNOL VOLUME 105, NUMBER 6, PART 1
TABLE III. The effect of the acute RV16 infection on peripheral blood WBC counts and differential (mean ± SD, n = 8) WBC counts (mean ± SD) Precold
(×106/mL)
WBC Neutrophils (%) Lymphocytes (%) Monocytes (%) Eosinophils (%)
Acute cold
Study day 1
Study day 3
Study day 3
Study day 5
6.4 ± 1.3 55.1 ± 5.1 31.4 ± 9.3 11.3 ± 5.8 2.3 ± 1.3
6.3 ± 2.0 61.9 ± 8.7 23.4 ± 5.6 11.9 ± 6.5 2.8 ± 2.3
7.8 ± 1.6* 68.3 ± 11.3* 16.3 ± 11.2* 13.6 ± 5.3 1.9 ± 2.1
6.8 ± 1.6 57.8 ± 6.4† 24.8 ± 8.9 12.4 ± 6.2 5.1 ± 3.2*†
Precold values are from study day 1. Acute cold values are 48 hours after initial RV16 inoculation (cold study day 3) and 4 days after initial RV16 inoculation (cold study day 5). *P < .05 compared with precold values at study day 1. †P < .03 compared with study day 3 of the acute cold.
Serum and nasal G-CSF concentrations during the acute RV16 illness were compared with bronchial lavage neutrophils during the acute cold (BL2). Serum (rs = 0.762, P = .02) and nasal (rs = 0.69, P = .05) G-CSF concentrations correlated with the bronchial lavage neutrophil values. Moreover, there was a trend for bronchial lavage concentrations of G-CSF to increase during the RV16 infection (3 vs 119 pg/mL [median values], P = .06) and for the bronchial lavage G-CSF values (BL2) to correlate with BL neutrophils, rs = 0.685, P = .07. The concentrations of IL-8, TNF-α, IL-5, IL-1β, and IFn-γ in the lavage fluid, however, did not change after inoculation with RV16 (data not shown). There was no effect of the RV16 infection on lavage fluid concentrations of LTB4 or EDN. MPO was measured in the nasal and BL fluids. During the acute cold phase of study, there was a significant increase in MPO in the second (BL2) lavage sample (0.9 ± 1.2 vs 8.8 ± 7.8 mg/mL, P = .02). When MPO concentrations were compared with neutrophil numbers in the lavage fluid, a significant correlation (r = 0.833, P = .005) was found but only during the acute RV16 infection. MPO concentrations in the nasal lavage fluid were compared with levels of nasal lavage G-CSF during the acute RV16 cold; there was also a strong correlation between these 2 values, rs = 0.78, P = .01.
DISCUSSION Nasal inoculation with RV16 increased peripheral blood neutrophils and decreased lymphocytes within 48hours of virus administration. Changes in nasal lavage fluid G-CSF and IL-8 concentrations during the acute cold also paralleled changes in circulating neutrophils. The novel components of our study were the finding that the intensity of the nasal G-CSF response to RV16 inoculation correlated with circulating levels of G-CSF, both of which related to changes in circulating neutrophils. Second, in contrast to the rapid changes in blood neutrophils, we found an increase in BL fluid neutrophils 4 days after RV inoculation. Finally, the changes in lavage neutrophils correlated both with nasal and serum G-CSF, and there was a positive trend with G-CSF in lavage fluid, thus suggesting a potential importance and rele-
vance of rhinovirus-induced generation of G-CSF to an eventual development of neutrophilic bronchial inflammation. The importance of neutrophils in viral respiratory illnesses and asthma have been noted by others.15,16 Fahy et al,18 for example, detected sputum neutrophilia in patients during an acute asthma exacerbation and noted that sputum neutrophil counts were greater (85% vs 57%, P = .05) in those reporting that a respiratory infection had precipitated their asthma. Teran et al19 detected an increase in neutrophil MPO in nasal lavage samples from children with documented viral respiratory infections that provoked asthma. Finally, Grünberg et al20 found increased nasal lavage neutrophils in patients experimentally inoculated with the same RV16 strain we used. These studies have parallels to our observations, and overall current observations suggest that increases in airway neutrophils may be a principal inflammatory cellular response during RV respiratory infections and may contribute to an asthma exacerbation. Both G-CSF and IL-8 were increased in nasal secretions during the acute RV16 illness. IL-8 is a potent chemoattractant for neutrophils and has been linked to neutrophilic inflammation during viral infections. Teran et al21 used segmental antigen challenge by bronchoscopy to show a correlation between increases in airway neutrophils and IL-8 concentrations. In RV16infected subjects Grünberg et al22 found an increase in sputum IL-8 that correlated with sputum neutrophils and changes in airway responsiveness. These latter findings suggest that RV-induced IL-8 is an important airway chemoattractant for neutrophils during the acute infection. Although we found significant increases in nasal IL-8 levels during the RV16, there were no changes in bronchial lavage fluid IL-8 levels. The differences between our findings and those of Grünberg et al,20,22 in particular, may reflect sample analyzed (sputum vs lavage), methods of inoculation (nasal plus lower airway vs nasal alone), or timing of specimen collection. Our study provides new evidence that RV infection induces high concentrations of G-CSF in the nasopharynx and that this effect on G-CSF is associated with increases in the systemic circulation. Moreover, RV infection-induced changes in G-CSF, both in the nasal
1176 Jarjour et al
J ALLERGY CLIN IMMUNOL JUNE 2000
FIG 4. Proposed mechanisms of RV-induced neutrophilic inflammation. An initial nasal response (day 1) to RV16 inoculation includes the generation of IL-8 and G-CSF. The generation of G-CSF is associated with increases in peripheral blood neutrophils. Changes in lower airway inflammatory markers are found later in the cold (ie, at 4 days) and are characterized by increases in neutrophils.
lavage fluid and serum, correlated with increases in circulating neutrophils (Fig 2). This finding, along with the well-documented biologic effects of G-CSF on neutrophil precursors in the bone marrow,23,24 supports the hypothesis that G-CSF is an important RV-induced signal from the upper airway to cause the temporary and temporal increase in circulating neutrophils during the acute phase of viral respiratory infections. The source for the increase in circulating G-CSF was not determined. GCSF was also detected in BL fluid and tended to increase approximately 4 days after RV16 inoculation. Interestingly, G-CSF levels in BL correlated with concentrations of MPO, raising the possibility that G-CSF can participate in the activation of lower airway neutrophils during RV infection. Alternately, the correlation between G-CSF and MPO could be a reflection of the secretion of G-CSF by activated neutrophils, although we are unaware of any data to test this hypothesis. The mechanisms by which neutrophils are recruited to the lower airway were not determined. No changes in LTB4 were noted in relationship to the RV16 infection. Furthermore, no change in BL cytokines, which either might increase neutrophil recruitment (GM-CSF) or adhesion molecules on endothelial cells (IL-1β, IFN-γ), were found. Thus the signal for neutrophil recruitment to the lower airway remains to be fully identified and is under further study. The design of our study did not allow us to address a number of important questions. First, the effect of the RV16 infection on airway responsiveness was not measured. This particular rhinovirus strain, however, has
been shown to increase airway responsiveness in previous studies by us5,6 and others.20,25 Because the major focus of our study was to evaluate the cellular and mediator responses in the lower airway in relationship to nasal inoculation with RV16, bronchial challenges either with histamine or methacholine were not performed to avoid any influence this intervention may have on the lower airway independent of the viral infections. Second, we did not use a placebo inoculation; rather, we performed a cross-over study with baseline evaluations and waited at least 4 weeks between the initial evaluation and virus inoculation. The precold and cold initial lavage values (BL1, precold and BL2, cold) were similar, supporting our contention that lower airway effects of the RV16 inoculation were related to the cold and not previous instrumentation before the cold. There was also an increase in peripheral blood eosinophils 96 hours after RV16 inoculation (Table III). Although we did not find increases in lavage fluid eosinophils or eosinophil mediators (EDN), it is possible that RV16-associated changes in this cell occur later or require a concomitant interaction with antigen to be detected in the lower airway.7 Finally, our study was not designed to evaluate the important question as to whether healthy subjects and those with allergic disease respond differently to a rhinovirus infection.26 In previous studies we have found that nonallergic subjects do not show changes in airway responsiveness8 or markers of inflammation in lavage fluid after RV16 inoculation, whereas allergic and asthmatic subjects have increased airway inflammation and responsiveness.7 To enhance the likelihood that we could
J ALLERGY CLIN IMMUNOL VOLUME 105, NUMBER 6, PART 1
determine effects of RV16 on lower airway inflammation and to elucidate the kinetics of viral infection on upper and lower airway responses, this study was limited to individuals with allergic airway disease. Whether a similar profile of response is seen in healthy subjects is the focus of additional studies. In summary, there is increasing evidence that neutrophils are an important component of RV-induced inflammation and may contribute to both upper and lower airway responses to the viral infection. Our findings suggest a temporal progression of responses to an RV infection that eventually increase airway neutrophilic inflammation and could, in the susceptible host, contribute to exacerbations of asthma (Fig 4). First, there is an acute upper respiratory response to the RV infection, which generates G-CSF and IL-8, and these cytokines are likely to contribute to increases in circulating neutrophils and their recruitment, respectively. The eventual development of a bronchial neutrophilic inflammatory response occurs later in the cold (ie, 4 days after inoculation). The mechanisms that regulate neutrophil recruitment to the lung are not established but may involve the presence of RV-infected cells in the lower airway and generation of chemoattractants, as suggested by previous work,27 or the migration of activated leukocytes from the upper airway to the lung. Additional studies will be needed to determine the significance of changes in lower airway neutrophils and modifications in lung function that can lead to an asthma exacerbation. We thank Dr Sally Wenzel for analysis of lavage fluids for leukotrienes, Dr Gerald Gleich for EDN assays, Ms Mary Jo Jackson and Ms Ann Dodge for research-nursing assistance in the conduct of this study, and Ms Kathryn Purcell for assistance in preparation of the manuscript.
REFERENCES 1. Folkerts G, Busse WW, Nijkamp FP, Sorkness R, Gern JE. Virus-induced airway hyperresponsiveness and asthma. Am J Respir Crit Care Med 1998;157:1708-20. 2. Johnston SL, Pattemore PK, Sanderson G, Smith S, Lampe F, Josephs L,et al. Community study of role of viral infections in exacerbations of asthma in 9-11 year old children. BMJ 1995;310:1225-9. 3. Nicholson KG, Kent J, Ireland DC. Respiratory viruses and exacerbations of asthma in adults. BMJ 1993;307:982-6. 4. Pattemore PK, Johnston SL, Bardin PG. Viruses as precipitants of asthma symptoms, I:epidemiology. Clin Exp Allergy 1992;22:324-36. 5. Lemanske RF Jr, Dick EC, Swenson CA, Vrtis RF, Busse WW. Rhinovirus upper respiratory infection increases airway hyperreactivity and late asthmatic reactions. J Clin Invest 1989;83:1-10. 6. Calhoun WJ, Swenson CA, Dick EC, Schwartz LB, Lemanske RF Jr, Busse WW. Experimental rhinovirus 16 infection potentiates histamine release after antigen bronchoprovocation in allergic subjects. Am Rev Respir Dis 1991;144:1267-73. 7. Calhoun WJ, Dick EC, Schwartz LB, Busse WW. A common cold virus, rhinovirus 16, potentiates airway inflammation after segmental antigen bronchoprovocation in allergic subjects. J Clin Invest 1994;94:2200-8.
Jarjour et al 1177
8. Gern JE, Calhoun WJ, Swenson C, Shen G, Busse WW. Rhinovirus infection preferentially increases lower airway responsiveness in allergic subjects. Am J Respir Crit Care Med 1997;155:1872-6. 9. D’Alessio DJ, Peterson JA, Dick EC. Transmission of experimental rhinovirus colds in volunteer married couples. J Infect Dis 1976;133:28-36. 10. D’Alessio DJ, Meschievitz CK, Schultz SB, Dick EC. A model for obtaining predictable natural transmission of rhinoviruses in human volunteers. J Infect Dis 1984;150:195-201. 11. Jackson GG, Dowling HF, Spresman IG, Boand AV. Transmission of the common cold to volunteers under controlled conditions. Arch Intern Med 1958;101:267-73. 12. Kelly EAB, Rodriguez RR, Busse WW, Jarjour NN. The effect of segmental bronchoprovocation with antigen on airway lymphocyte function. Am J Respir Crit Care Med 1997;156:1421-8. 13. Sedgwick JB, Calhoun WJ, Gleich GJ, Kita H, Abrams JS, Schwartz LB, et al. Immediate and late airway response of allergic rhinitis patients to segmental antigen challenge: characterization of eosinophil and mast cell mediators. Am Rev Respir Dis 1991;144:1274-81. 14. Wenzel SE, Trudeau JB, Kaminsky DA, Cohn J, Martin RJ, Westcott JY. Effect of 5-lipoxygenase inhibition on bronchoconstruction and airway inflammation in nocturnal asthma. Am J Respir Crit Care Med 1995; 52:897-905. 15. Turner RB. The role of neutrophils in the pathogenesis of rhinovirus infections. Pediatr Infect Dis J 1990; 9:832-35. 16. Levandowski RA, Weaver CW, Jackson GG. Nasal secretion leukocyte populations determined by flow cytometry during acute rhinovirus infection. J Med Virol 1988; 25:423-32. 17. Jarjour NN, Calhoun WJ, Kelly EAB, Gleich GJ, Schwartz LB, Busse WW. The immediate and late allergic responses to segmental bronchopulmonary provocation in asthma. Am J Respir Crit Care Med 1997;155:1515-21. 18. Fahy JV, Kim KW, Liu J, Boushey HA. Respiratory pathophysiologic responses: prominent neutrophilic inflammation in sputum from subjects with asthma exacerbation. J Allergy Clin Immunol 1995;95:843-52. 19. Teran LM, Johnston SL, Schröder JM, Church MK, Holgate ST. Role of nasal interleukin-8 in neutrophil recruitment and activation in children with virus-induced asthma. Am J Respir Crit Care Med 1997;155:1362-6. 20. Grünberg K, Timmers MC, Smits HH, De Klerk EPA, Dick EC, Spaan WJM, et al. Effect of experimental rhinovirus 16 colds on airway hyperresponsiveness to histamine and interleukin-8 in nasal lavage in asthmatic subjects in vivo. Clin Exp Allergy 1997;27:36-45. 21. Teran LM, Carroll MP, Frew AJ, Redington AE, Davies DE, Lindley I, et al. Leukocyte recruitment after local endobronchial allergen challenge in asthma: relationship to procedure and to airway interleukin-8 release. Am J Respir Crit Care Med 1996;154:469-76. 22. Grünberg K, Smits HD, Timmers MC, de Klerk MPA, Dolhain RJEM, Dick EC, et al. Experimental rhinovirus 16 infection effects on cell differentials and soluble markers in sputum in asthmatic subjects. Am J Respir Crit Care Med 1997;156:609-16. 23. Khwaja A, Linch DC. Effects of granulocyte colony-stimulating factor and granulocyte-macrophage colony-stimulating factor on neutrophil formation and function. Curr Opin Hematol 1994;1:216-20. 24. Schmitz S, Franke H, Brusis J, Wichmann HE. Quantification of the cell kinetic effects of G-CSF using a model of human granulopoiesis. Exp Hematol 1993;21:755-60. 25. Cheung D, Dick EC, Timmers MC, De Klerk EPA, Spaan WJM,Sterk PJ. Rhinovirus inhalation causes long-lasting excessive airway narrowing in response to methacholine in asthmatic subjects in vivo. Am J Respir Crit Care Med 1995;152:1490-6. 26. Fleming HE, Little FF, Schurr D, Avila PC, Wong H, Liu J, et al. Rhinovirus-16 colds in healthy and in asthmatic subjects. Am J Respir Crit Care Med 1999;160:100-8. 27. Gern JE, Galagan DM, Jarjour NN, Dick EC, Busse WW. Detection of rhinovirus RNA in lower airway cells during experimentally induced infection. Am J Respir Crit Care Med 1997; 155:1159-61.
Viral respiratory tract infections are the most frequent cause of asthma exacerbations.1-4 Of the respiratory viruses associated with these asthma exacerbations, the common cold virus, rhinovirus (RV), has been the most frequently identified illness in children and adults.2,3 To
From the Departments of Medicine and Pediatrics, University of Wisconsin Medical School, University of Wisconsin Hospital, Madison, Wis. Supported by National Institutes of Health grants No. AI 34891 and AI 26609. Received for publication Dec 7, 1999; revised Feb 10, 2000; accepted for publication Feb 10, 2000. Reprint requests: William W. Busse, MD, University of Wisconsin Medical School, Departments of Medicine and Pediatrics, University of Wisconsin Hospital, H6/367 Clinical Science Center, 600 Highland Ave, Madison, WI 53792-3244. Copyright © 2000 by Mosby, Inc. 0091-6749/2000 $12.00 + 0 1/1/106376 doi:10.1067/mai.2000.106376
Abbreviations used BL: Bronchial lavage EDN: Eosinophil-derived neurotoxin G-CSF: Granulocyte–colony-stimulating factor LTB4: Leukotriene B4 MPO: Myeloperoxidase RV16: Rhinovirus strain 16 TCID50: Tissue culture infective dose 50% WBC: White blood cell
evaluate the mechanisms of an RV infection on asthma exacerbations, we have experimentally inoculated subjects with RV, including strain 16 (RV16).5-7 This experimental respiratory infection has caused a symptomatic illness, an increase in bronchial responsiveness,5,6 and an enhanced allergic inflammatory response to allergen.5-7 Because RV infections are, at least initially, confined to the upper airway, it is likely that the nasal immune response to this virus plays a key role in initiating an airway inflammatory response in patients with asthma.1 Thus we have hypothesized that RV infections generate proinflammatory cytokines in the nose and that these cytokines initiate neutrophilic responses in the blood and the upper and lower airways. To evaluate this hypothesis, we experimentally inoculated 8 asthmatic subjects with RV16 and evaluated timedependent effects of the respiratory infection on circulating leukocytes, nasal-derived mediators, and bronchial (by bronchoscopy) markers of inflammation.
Methods Subject selection Eight allergic asthmatic subjects were recruited for participation in this study (Table I). Each subject had a physician’s diagnosis of asthma, a history of occasional seasonal wheezing (presumably in relationship to environmental allergen exposure), and a prescription for an inhaled β-agonist. None of the subjects, however, required regular use of asthma medication or had a history of wheezing with a respiratory infection. Each subject had a positive prick skin test to one or more aeroallergens (including house dust mite, ragweed, grass, trees, cat, dog, and Alternaria antigen extract) and then underwent a medical history, physical examination, and pulmonary function testing by spirometry. These subjects did use any medications during the study, including bronchodilators, antihistamines, or topical nasal corticosteroids. We have previously found the effects of the RV16 experimental infection on airway responsiveness to be primarily limited to individuals with allergic disease.8 In addition, the effects of RV16 on markers of lower airway inflammation, when evaluated by bron1169
1170 Jarjour et al
J ALLERGY CLIN IMMUNOL JUNE 2000
TABLE I. The effect of the experimental RV16 inoculation on FEV1 values and RV16 antibody titers, virus titers, and cold symptoms Baseline FEV1 (L [% predicted]) Subject
Sex
Age (y)
1 2 3 4 5 6 7 8
M M F F F M M F
26 28 29 24 20 29 37 22
Precold
4.13 (91) 4.33 (99) 3.38 (115) 3.53 (129) 3.12 (106) 4.82 (103) 4.48 (102) 3.03 (81)
Serum RV16 antibody titers
Acute infection
Acute
4.17 (92) 4.33 (99) 3.38 (115) 3.11 (114) 3.07 (104) 4.72 (101) 4.49 (102) 2.81 (77)
<1 1 <1 <1 1.4 <1 <1 2.8
Convalescent
2.8 45 45 90.5 2 5.6 1.4 45
Peak nasal virus titers* TCID50/mL
101 103 103 105.5 104 105 103 101
Peak symptoms†
7 13 17 12.5 12 13 4 15
TCID50, Tissue culture infective dose 50%; M, male; F, female. *Peak TCID50/100 µL of nasal fluid during acute infection. †Peak subject-related symptom score is tabulation of 13 cold symptoms with highest possible score of 39.
choscopy and lavage, were principally confined to subjects with allergic disease.7 Because the goal of this study was to determine the effects of an experimental RV infection on upper and lower airway inflammatory events in asthma, a study population was selected in which such changes are likely to occur (ie, allergic individuals with a history of asthma). The study was approved by the University of Wisconsin–Madison Center for Health Sciences Human Subjects Committee. Informed consent was obtained from each subject before participation.
Study design The recruited subjects were evaluated during 2 separate phases: (1) precold and (2) cold (see Table II for study design). On study day 1 of the precold phase, a blood draw, nasal wash, and bronchoscopy with lavage were performed. Two days later (study day 3), a blood sample was drawn and bronchoscopy and lavage were repeated. These studies provided the baseline or precold data for comparison with similar temporal values during the cold phase. Approximately 1 month later the subjects entered the cold phase of the study (Table II). On day 1 of the cold phase each subject had a nasal wash performed to determine whether a respiratory infection was present at the time of inoculation. The subjects were then inoculated (see below for methods) with live RV16 on study days 1 and 2. Two days after the initial RV16 inoculation (study day 3) and at a time when all participants had symptoms of a cold, bronchoscopy, and airway lavage plus a blood draw were performed. The bronchoscopy, airway lavage, and blood draw were repeated 48 hours later (study day 5). Nasal washes were obtained daily during the acute cold phase.
RV16 inoculation, confirmation of a viral infection, and assessment of cold severity The technique for RV inoculation and subsequent symptom monitoring followed previously described methods.9,10 Briefly, no patients had neutralizing antibody in their sera (≤1:1) at the initial screening for entry into the study. (Note: subject 8 was found to have a titer of 2.8 at the time of inoculation although the screening neutralizing antibody titer was <1:1.) RV colds were induced in the subjects by instilling, on 2 successive days, 0.25 mL of RV16 suspension (320-3200 TCID50) into each nostril by pipette and then spraying approximately the same amount of virus inocula into each nostril with an atomizer (DeVilbiss, No. 286, Somerset, Pa) powered by a DeVilbiss 561 series compressor. Colds were evaluated with a questionnaire that was completed by the participant hourly during waking hours as previously described.9,11 Briefly, the symptoms evaluated included, but were
not limited to, cough, nasal discharge, sneezing, stuffy nose, headache, malaise, chills, or fever. Symptoms were graded on the following scale: 0 = absent, 1 = mild, 2 = moderate, or 3 = severe. Total symptom scores of ≥12 have been used to define a severe cold, 7 to 11 a moderate cold, and <7 a mild cold. Shed virus was assayed by titration of the nasal wash into diploid (WI-38) cell cultures. Nasal washes were performed daily during the acute infection. Prewarmed HBSS with 0.5% gelatin was instilled (5 mL into each nostril) into the nose, retained for 5 seconds, and then forcefully expelled into a sterile petri dish. The nasal wash fluid was then transferred to a test tube to assess volume and was vortexed and an aliquot centrifuged, creating a cell-free supernatant that was stored at –80°C for later analysis. The remainder of the nasal wash was inoculated into WI-38, primary rhesus monkey, and HEp-2 cell cultures; washings obtained after RV16 inoculation were placed into WI-38 cells only.9,10 Confirmation that the respiratory illness was caused by rhinovirus included at least one identified isolate for RV16 or by a 4-fold or greater increase in neutralizing antibody in the convalescent serum specimens to this strain of virus. Antibody titers were expressed as initial serum dilutions in which 1:1 = undiluted serum, 1:2 = one part serum and one part diluent, etc.
Bronchoscopy and bronchial lavage Bronchoscopy and bronchial lavage (BL) were conducted as previously described.7 Briefly, one bronchopulmonary segment was identified, and the fiberoptic bronchoscope was wedged into that segment. The bronchoscope was held in a wedge position and lavage (60 mL) was performed. This volume was selected to obtain a wash of fluids in the airway and to be a likely representative of air space cells and other mediators. A second bronchoscopy was done 48 hours later. At that time, BL was performed in a similar fashion. The use of paired bronchoscopies over 48 hours provided an opportunity to evaluate the effect of RV infection on the lower airway at two time points.
BL fluid analysis The BL fluid return, which was similar in volume at precold and during the acute RV16 illness, was centrifuged (400g for 10 minutes) to sediment cells. The supernatant was removed and stored at –80°C for later analysis. BL cells were washed twice with HBSS containing 2% newborn calf serum. A cell count was made with a hemocytometer. Cytocentrifuge slides were prepared, air-dried, fixed in methanol, and stained (Diff-Quick Scientific Products, Chicago, Ill). For the differential cell count, 300 leukocytes were enumerated and identified as lymphocytes, neutrophils, eosinophils, macrophages, or epithelial cells on the basis of staining and morphologic characteristics.
Jarjour et al 1171
J ALLERGY CLIN IMMUNOL VOLUME 105, NUMBER 6, PART 1
TABLE II. Study design Precold evaluations (study day) Procedure
1
Blood draw Nasal wash Bronchoscopy RV16 inoculation
X X X
2
Cold evaluations (study day)
3
1
2
3
4
5
6
7
X
X X
X
X X X
X
X X X
X
X
X
X
X
X, Procedure performed.
Determination of cytokines and other mediators BL fluid was concentrated 10-fold by a centrifugal concentration device containing a low protein-binding membrane YM filter with a molecular weight cutoff value of 3000 (Amicon, Beverly, Mass). Nasal washes and serum were analyzed without concentration. IL-8 and granulocyte-colony stimulating factor (G-CSF) were measured with a commercial kit according to the manufacturer’s directions (Bioscience, Camerillo, Calif, and R&D, Minneapolis, Minn, respectively). Concentrated BL fluid was also analyzed for IL-5, IFN-γ, IL1β, and TNF-α by a 2-step sandwich ELISA, as previously published.12 The concentrations of TNF-α, IL-5, and IFN-γ were calculated by comparison with a standard curve consisting of known amounts of recombinant human TNF-α, IL-5, or IFN-γ. The sensitivity of this assay is 5 pg/mL (equivalent to 0.5 pg/mL in neat BL fluid). Myeloperoxidase (MPO) concentrations were determined in the samples (nasal washes or bronchial lavage fluid) by a commercial RIA according to the directions of the manufacturer (Pharmacia Diagnostics, Columbus, Ohio). Eosinophil-derived neurotoxin (EDN) was determined in lavage fluid by methods previously described.13 Leukotriene B4 (LTB4) was also measured in the lavage fluid.14
Data analysis The data are presented as medians with 25% and 75% interquartiles. Comparisons between the results from the 2 separate lavage samples were done with the Wilcoxon test. Correlation between cytokine levels, viral titers, and proportions of various BL cells were evaluated with the Spearman rank correlation test.
RESULTS Patient characteristics and confirmation of RV16 infection (Table I) There was no effect of the RV16 infection on FEV1 values. All subjects, except number 8, had serum RV16 antibody titers that were ≤1:1 at the time of inoculation. Subject 8 had an antibody titer <1:1 at the time of initial screening; on the day of RV16 inoculation, however, the serum antibody titer was ≥2.8. This subject had symptoms, shedding of RV16 in the nasal cultures, and an increase in antibody titer in the convalescence serum sample (≥45). Consequently, data from subject 8 were included in the evaluations. All subjects had a symptomatic upper respiratory infection as indicated by evaluation of peak symptoms, which ranged from mild (7) to severe (>12).
Effect of RV16 infection on peripheral white blood cell counts Peripheral white blood cell (WBC) counts and cell dif-
ferential counts were determined precold and then after inoculation with RV16 (see Study design in Tables II and Table III). There was a nonsignificant increase in the percentage of blood neutrophils in the precold phase at study day 3; this may reflect the effect of the previous bronchoscopy. The RV16 infection caused a small but significant increase in the peripheral WBC count, which was detected in the blood sample taken 48 hours after day 1 virus inoculation compared with a similar evaluation time point precold (6.4 ± 1.8 × 104 cells/mL [precold] vs 7.8 ± 2.5 × 106 cells/mL [acute cold], P < .05, mean ± SD). There was also a significant (P < .02) increase in the percentage of circulating neutrophils (55.1% ± 5.1% [precold] vs 68.3% ± 11.3% [acute cold]). In contrast, and in keeping with published observations,15,16 the percentage of circulating lymphocytes significantly decreased during the acute phase of the RV16 infection, precold to acute cold (31.4% ± 9.3% vs 16.3% ± 11.2%, P < .02). Ninety-six hours after RV inoculation (study day 5, acute cold), both neutrophil and lymphocyte peripheral blood counts had returned to preinoculation values; at this time, however, there was an increase in eosinophils over values at the initial determination (acute cold study day 3) (1.9% ± 2.1% vs 5.1% ± 3.2%, P < .05).
Nasal and serum concentrations of cytokines Nasal fluids from the precold and cold phases of study were analyzed for G-CSF (Fig 1, A) and IL-8 (Fig 1, B); beginning 1 to 2 days after RV16 inoculation, there were increases in G-CSF and IL-8 in the nasal lavage fluid. The increase in these cytokine values persisted for up to 4 days after RV16 inoculation. Serum G-CSF values increased significantly after RV16 inoculation (5.5 pg/mL [precold] vs 62.0 pg/mL [cold], study day 3 median values, P < .05). A strong correlation was found between nasal and serum G-CSF values on study day 3 of the cold phase (48 hours post-RV16 inoculation, rs = 0.95, P ≤ .001). Correlations between nasal concentrations either of IL-8 or G-CSF and RV16 virus titers from the nasal lavage fluids were determined. Forty-eight hours after RV16 inoculation, both nasal IL-8 and G-CSF values showed striking correlations with quantitative measurements of RV16 shedding in nasal secretions (rs = 0.77, P = .02 and rs = 0.85, P < .005, respectively). Furthermore, the serum G-CSF value and nasal RV16 virus titer had a strong correlation (rs = 0.78, P = .015).
1172 Jarjour et al
J ALLERGY CLIN IMMUNOL JUNE 2000
A
B FIG 1. The effect of RV16 inoculation on nasal lavage fluid G-CSF (A) and IL-8 (B). Box blots represent median with 25% and 75% interquartiles; the error bars represent the 10th and 90th percentiles. n = 8. Asterisk, P < .05 compared with preinoculation (pre-inoc) values.
Relationships between serum and nasal G-CSF and changes in peripheral blood neutrophils Serum (Fig 2, A) and nasal G-CSF (Fig 2, B) values were evaluated for a possible relationship to changes in peripheral blood neutrophils from the precold (study day 1) to acute cold phase of study (study day 3, 48 hours after RV16 inoculation). There were strong correlations between nasal and serum G-CSF values and changes in peripheral blood neu-
trophils (rs = 0.898, P < .001 and rs = 0.874, P < .001, respectively). When comparisons were made between nasal concentrations of IL-8 and changes in peripheral blood neutrophils, the correlation between these 2 variables showed a trend but did not achieve significance (rs = 0.663, P = .072).
The effect of RV16 infection on the cellular composition of BL fluid The percent recovery of instilled saline solution (60 mL) was similar in the paired lavage samples (BL 1, pre-
J ALLERGY CLIN IMMUNOL VOLUME 105, NUMBER 6, PART 1
Jarjour et al 1173
A
B
FIG 2. Correlation between serum (A) and nasal lavage fluid (B) G-CSF and change in peripheral blood neutrophils. The serum and nasal lavage G-CSF values were obtained 2 days after RV16 inoculation. The change in blood neutrophils represents the difference between preinoculation (precold) and acute symptoms (study day 3) values and are represented as 106cells per milliliter.
cold = 54.4% ± 7.9% vs BL 1, cold = 56.3% ± 14.5%; BL 2, precold = 41.9% ± 11.0% vs BL 2, cold = 43.8% ± 6.8% [mean ± SD, n = 8]). There was no increase in the total WBCs in bronchial lavage related to the cold either at 48 (BL1) or 96 (BL2) hours after RV16 inoculation (Fig 3, A). There was, however, an increase in neutrophils in samples from the second lavage (precold BL1 vs pre-
cold BL2), median values 0.02 versus 0.11 × 106 cells, P < .05 (Fig 3, B), an effect previously described.17 To assess, in a limited fashion, the kinetics of the lower airway inflammatory response to nasal inoculation with RV16, bronchoscopy lavage was performed on 2 occasions after the RV16 inoculation (see above and Table II). Forty-eight hours after the RV16 inoculation, there was
1174 Jarjour et al
J ALLERGY CLIN IMMUNOL JUNE 2000
A
B FIG 3. The effect of RV16 respiratory infection on BL total WBC (A) and neutrophil (B) values. Bronchoscopy and lavage were performed on 2 occasions before cold ([BL-1] study day 1 and [BL-2] 48 hours later, precold study day 3); similar assessments were made during the acute experimental RV16 infection (2 [BL-1] and 4 days [BL-2] after RV16 inoculation). Data are expressed as the median values with 25% and 75% interquartiles, n = 8. Total cell values are expressed as cells (WBCs and neutrophils × 106 cells per BL fluid). Asterisk, P < .05 compared with BL values (precold and cold); dagger, P < .05 compared with BL2 precold.
no difference between BL neutrophils (cold BL1) and precold values (precold BL1). Two days later and approximately 96 hours after RV16 inoculation, there was a significant increase in lavage neutrophils over the initial values (cold BL1) and both the precold values
(precold BL1 and BL2) (Fig 3, B). Although bronchoscopy by itself may increase the numbers of neutrophils (see precold BL1 vs BL2), the increase in neutrophils in the second lavage fluid was significantly greater during the experimental cold.
Jarjour et al 1175
J ALLERGY CLIN IMMUNOL VOLUME 105, NUMBER 6, PART 1
TABLE III. The effect of the acute RV16 infection on peripheral blood WBC counts and differential (mean ± SD, n = 8) WBC counts (mean ± SD) Precold
(×106/mL)
WBC Neutrophils (%) Lymphocytes (%) Monocytes (%) Eosinophils (%)
Acute cold
Study day 1
Study day 3
Study day 3
Study day 5
6.4 ± 1.3 55.1 ± 5.1 31.4 ± 9.3 11.3 ± 5.8 2.3 ± 1.3
6.3 ± 2.0 61.9 ± 8.7 23.4 ± 5.6 11.9 ± 6.5 2.8 ± 2.3
7.8 ± 1.6* 68.3 ± 11.3* 16.3 ± 11.2* 13.6 ± 5.3 1.9 ± 2.1
6.8 ± 1.6 57.8 ± 6.4† 24.8 ± 8.9 12.4 ± 6.2 5.1 ± 3.2*†
Precold values are from study day 1. Acute cold values are 48 hours after initial RV16 inoculation (cold study day 3) and 4 days after initial RV16 inoculation (cold study day 5). *P < .05 compared with precold values at study day 1. †P < .03 compared with study day 3 of the acute cold.
Serum and nasal G-CSF concentrations during the acute RV16 illness were compared with bronchial lavage neutrophils during the acute cold (BL2). Serum (rs = 0.762, P = .02) and nasal (rs = 0.69, P = .05) G-CSF concentrations correlated with the bronchial lavage neutrophil values. Moreover, there was a trend for bronchial lavage concentrations of G-CSF to increase during the RV16 infection (3 vs 119 pg/mL [median values], P = .06) and for the bronchial lavage G-CSF values (BL2) to correlate with BL neutrophils, rs = 0.685, P = .07. The concentrations of IL-8, TNF-α, IL-5, IL-1β, and IFn-γ in the lavage fluid, however, did not change after inoculation with RV16 (data not shown). There was no effect of the RV16 infection on lavage fluid concentrations of LTB4 or EDN. MPO was measured in the nasal and BL fluids. During the acute cold phase of study, there was a significant increase in MPO in the second (BL2) lavage sample (0.9 ± 1.2 vs 8.8 ± 7.8 mg/mL, P = .02). When MPO concentrations were compared with neutrophil numbers in the lavage fluid, a significant correlation (r = 0.833, P = .005) was found but only during the acute RV16 infection. MPO concentrations in the nasal lavage fluid were compared with levels of nasal lavage G-CSF during the acute RV16 cold; there was also a strong correlation between these 2 values, rs = 0.78, P = .01.
DISCUSSION Nasal inoculation with RV16 increased peripheral blood neutrophils and decreased lymphocytes within 48hours of virus administration. Changes in nasal lavage fluid G-CSF and IL-8 concentrations during the acute cold also paralleled changes in circulating neutrophils. The novel components of our study were the finding that the intensity of the nasal G-CSF response to RV16 inoculation correlated with circulating levels of G-CSF, both of which related to changes in circulating neutrophils. Second, in contrast to the rapid changes in blood neutrophils, we found an increase in BL fluid neutrophils 4 days after RV inoculation. Finally, the changes in lavage neutrophils correlated both with nasal and serum G-CSF, and there was a positive trend with G-CSF in lavage fluid, thus suggesting a potential importance and rele-
vance of rhinovirus-induced generation of G-CSF to an eventual development of neutrophilic bronchial inflammation. The importance of neutrophils in viral respiratory illnesses and asthma have been noted by others.15,16 Fahy et al,18 for example, detected sputum neutrophilia in patients during an acute asthma exacerbation and noted that sputum neutrophil counts were greater (85% vs 57%, P = .05) in those reporting that a respiratory infection had precipitated their asthma. Teran et al19 detected an increase in neutrophil MPO in nasal lavage samples from children with documented viral respiratory infections that provoked asthma. Finally, Grünberg et al20 found increased nasal lavage neutrophils in patients experimentally inoculated with the same RV16 strain we used. These studies have parallels to our observations, and overall current observations suggest that increases in airway neutrophils may be a principal inflammatory cellular response during RV respiratory infections and may contribute to an asthma exacerbation. Both G-CSF and IL-8 were increased in nasal secretions during the acute RV16 illness. IL-8 is a potent chemoattractant for neutrophils and has been linked to neutrophilic inflammation during viral infections. Teran et al21 used segmental antigen challenge by bronchoscopy to show a correlation between increases in airway neutrophils and IL-8 concentrations. In RV16infected subjects Grünberg et al22 found an increase in sputum IL-8 that correlated with sputum neutrophils and changes in airway responsiveness. These latter findings suggest that RV-induced IL-8 is an important airway chemoattractant for neutrophils during the acute infection. Although we found significant increases in nasal IL-8 levels during the RV16, there were no changes in bronchial lavage fluid IL-8 levels. The differences between our findings and those of Grünberg et al,20,22 in particular, may reflect sample analyzed (sputum vs lavage), methods of inoculation (nasal plus lower airway vs nasal alone), or timing of specimen collection. Our study provides new evidence that RV infection induces high concentrations of G-CSF in the nasopharynx and that this effect on G-CSF is associated with increases in the systemic circulation. Moreover, RV infection-induced changes in G-CSF, both in the nasal
1176 Jarjour et al
J ALLERGY CLIN IMMUNOL JUNE 2000
FIG 4. Proposed mechanisms of RV-induced neutrophilic inflammation. An initial nasal response (day 1) to RV16 inoculation includes the generation of IL-8 and G-CSF. The generation of G-CSF is associated with increases in peripheral blood neutrophils. Changes in lower airway inflammatory markers are found later in the cold (ie, at 4 days) and are characterized by increases in neutrophils.
lavage fluid and serum, correlated with increases in circulating neutrophils (Fig 2). This finding, along with the well-documented biologic effects of G-CSF on neutrophil precursors in the bone marrow,23,24 supports the hypothesis that G-CSF is an important RV-induced signal from the upper airway to cause the temporary and temporal increase in circulating neutrophils during the acute phase of viral respiratory infections. The source for the increase in circulating G-CSF was not determined. GCSF was also detected in BL fluid and tended to increase approximately 4 days after RV16 inoculation. Interestingly, G-CSF levels in BL correlated with concentrations of MPO, raising the possibility that G-CSF can participate in the activation of lower airway neutrophils during RV infection. Alternately, the correlation between G-CSF and MPO could be a reflection of the secretion of G-CSF by activated neutrophils, although we are unaware of any data to test this hypothesis. The mechanisms by which neutrophils are recruited to the lower airway were not determined. No changes in LTB4 were noted in relationship to the RV16 infection. Furthermore, no change in BL cytokines, which either might increase neutrophil recruitment (GM-CSF) or adhesion molecules on endothelial cells (IL-1β, IFN-γ), were found. Thus the signal for neutrophil recruitment to the lower airway remains to be fully identified and is under further study. The design of our study did not allow us to address a number of important questions. First, the effect of the RV16 infection on airway responsiveness was not measured. This particular rhinovirus strain, however, has
been shown to increase airway responsiveness in previous studies by us5,6 and others.20,25 Because the major focus of our study was to evaluate the cellular and mediator responses in the lower airway in relationship to nasal inoculation with RV16, bronchial challenges either with histamine or methacholine were not performed to avoid any influence this intervention may have on the lower airway independent of the viral infections. Second, we did not use a placebo inoculation; rather, we performed a cross-over study with baseline evaluations and waited at least 4 weeks between the initial evaluation and virus inoculation. The precold and cold initial lavage values (BL1, precold and BL2, cold) were similar, supporting our contention that lower airway effects of the RV16 inoculation were related to the cold and not previous instrumentation before the cold. There was also an increase in peripheral blood eosinophils 96 hours after RV16 inoculation (Table III). Although we did not find increases in lavage fluid eosinophils or eosinophil mediators (EDN), it is possible that RV16-associated changes in this cell occur later or require a concomitant interaction with antigen to be detected in the lower airway.7 Finally, our study was not designed to evaluate the important question as to whether healthy subjects and those with allergic disease respond differently to a rhinovirus infection.26 In previous studies we have found that nonallergic subjects do not show changes in airway responsiveness8 or markers of inflammation in lavage fluid after RV16 inoculation, whereas allergic and asthmatic subjects have increased airway inflammation and responsiveness.7 To enhance the likelihood that we could
J ALLERGY CLIN IMMUNOL VOLUME 105, NUMBER 6, PART 1
determine effects of RV16 on lower airway inflammation and to elucidate the kinetics of viral infection on upper and lower airway responses, this study was limited to individuals with allergic airway disease. Whether a similar profile of response is seen in healthy subjects is the focus of additional studies. In summary, there is increasing evidence that neutrophils are an important component of RV-induced inflammation and may contribute to both upper and lower airway responses to the viral infection. Our findings suggest a temporal progression of responses to an RV infection that eventually increase airway neutrophilic inflammation and could, in the susceptible host, contribute to exacerbations of asthma (Fig 4). First, there is an acute upper respiratory response to the RV infection, which generates G-CSF and IL-8, and these cytokines are likely to contribute to increases in circulating neutrophils and their recruitment, respectively. The eventual development of a bronchial neutrophilic inflammatory response occurs later in the cold (ie, 4 days after inoculation). The mechanisms that regulate neutrophil recruitment to the lung are not established but may involve the presence of RV-infected cells in the lower airway and generation of chemoattractants, as suggested by previous work,27 or the migration of activated leukocytes from the upper airway to the lung. Additional studies will be needed to determine the significance of changes in lower airway neutrophils and modifications in lung function that can lead to an asthma exacerbation. We thank Dr Sally Wenzel for analysis of lavage fluids for leukotrienes, Dr Gerald Gleich for EDN assays, Ms Mary Jo Jackson and Ms Ann Dodge for research-nursing assistance in the conduct of this study, and Ms Kathryn Purcell for assistance in preparation of the manuscript.
REFERENCES 1. Folkerts G, Busse WW, Nijkamp FP, Sorkness R, Gern JE. Virus-induced airway hyperresponsiveness and asthma. Am J Respir Crit Care Med 1998;157:1708-20. 2. Johnston SL, Pattemore PK, Sanderson G, Smith S, Lampe F, Josephs L,et al. Community study of role of viral infections in exacerbations of asthma in 9-11 year old children. BMJ 1995;310:1225-9. 3. Nicholson KG, Kent J, Ireland DC. Respiratory viruses and exacerbations of asthma in adults. BMJ 1993;307:982-6. 4. Pattemore PK, Johnston SL, Bardin PG. Viruses as precipitants of asthma symptoms, I:epidemiology. Clin Exp Allergy 1992;22:324-36. 5. Lemanske RF Jr, Dick EC, Swenson CA, Vrtis RF, Busse WW. Rhinovirus upper respiratory infection increases airway hyperreactivity and late asthmatic reactions. J Clin Invest 1989;83:1-10. 6. Calhoun WJ, Swenson CA, Dick EC, Schwartz LB, Lemanske RF Jr, Busse WW. Experimental rhinovirus 16 infection potentiates histamine release after antigen bronchoprovocation in allergic subjects. Am Rev Respir Dis 1991;144:1267-73. 7. Calhoun WJ, Dick EC, Schwartz LB, Busse WW. A common cold virus, rhinovirus 16, potentiates airway inflammation after segmental antigen bronchoprovocation in allergic subjects. J Clin Invest 1994;94:2200-8.
Jarjour et al 1177
8. Gern JE, Calhoun WJ, Swenson C, Shen G, Busse WW. Rhinovirus infection preferentially increases lower airway responsiveness in allergic subjects. Am J Respir Crit Care Med 1997;155:1872-6. 9. D’Alessio DJ, Peterson JA, Dick EC. Transmission of experimental rhinovirus colds in volunteer married couples. J Infect Dis 1976;133:28-36. 10. D’Alessio DJ, Meschievitz CK, Schultz SB, Dick EC. A model for obtaining predictable natural transmission of rhinoviruses in human volunteers. J Infect Dis 1984;150:195-201. 11. Jackson GG, Dowling HF, Spresman IG, Boand AV. Transmission of the common cold to volunteers under controlled conditions. Arch Intern Med 1958;101:267-73. 12. Kelly EAB, Rodriguez RR, Busse WW, Jarjour NN. The effect of segmental bronchoprovocation with antigen on airway lymphocyte function. Am J Respir Crit Care Med 1997;156:1421-8. 13. Sedgwick JB, Calhoun WJ, Gleich GJ, Kita H, Abrams JS, Schwartz LB, et al. Immediate and late airway response of allergic rhinitis patients to segmental antigen challenge: characterization of eosinophil and mast cell mediators. Am Rev Respir Dis 1991;144:1274-81. 14. Wenzel SE, Trudeau JB, Kaminsky DA, Cohn J, Martin RJ, Westcott JY. Effect of 5-lipoxygenase inhibition on bronchoconstruction and airway inflammation in nocturnal asthma. Am J Respir Crit Care Med 1995; 52:897-905. 15. Turner RB. The role of neutrophils in the pathogenesis of rhinovirus infections. Pediatr Infect Dis J 1990; 9:832-35. 16. Levandowski RA, Weaver CW, Jackson GG. Nasal secretion leukocyte populations determined by flow cytometry during acute rhinovirus infection. J Med Virol 1988; 25:423-32. 17. Jarjour NN, Calhoun WJ, Kelly EAB, Gleich GJ, Schwartz LB, Busse WW. The immediate and late allergic responses to segmental bronchopulmonary provocation in asthma. Am J Respir Crit Care Med 1997;155:1515-21. 18. Fahy JV, Kim KW, Liu J, Boushey HA. Respiratory pathophysiologic responses: prominent neutrophilic inflammation in sputum from subjects with asthma exacerbation. J Allergy Clin Immunol 1995;95:843-52. 19. Teran LM, Johnston SL, Schröder JM, Church MK, Holgate ST. Role of nasal interleukin-8 in neutrophil recruitment and activation in children with virus-induced asthma. Am J Respir Crit Care Med 1997;155:1362-6. 20. Grünberg K, Timmers MC, Smits HH, De Klerk EPA, Dick EC, Spaan WJM, et al. Effect of experimental rhinovirus 16 colds on airway hyperresponsiveness to histamine and interleukin-8 in nasal lavage in asthmatic subjects in vivo. Clin Exp Allergy 1997;27:36-45. 21. Teran LM, Carroll MP, Frew AJ, Redington AE, Davies DE, Lindley I, et al. Leukocyte recruitment after local endobronchial allergen challenge in asthma: relationship to procedure and to airway interleukin-8 release. Am J Respir Crit Care Med 1996;154:469-76. 22. Grünberg K, Smits HD, Timmers MC, de Klerk MPA, Dolhain RJEM, Dick EC, et al. Experimental rhinovirus 16 infection effects on cell differentials and soluble markers in sputum in asthmatic subjects. Am J Respir Crit Care Med 1997;156:609-16. 23. Khwaja A, Linch DC. Effects of granulocyte colony-stimulating factor and granulocyte-macrophage colony-stimulating factor on neutrophil formation and function. Curr Opin Hematol 1994;1:216-20. 24. Schmitz S, Franke H, Brusis J, Wichmann HE. Quantification of the cell kinetic effects of G-CSF using a model of human granulopoiesis. Exp Hematol 1993;21:755-60. 25. Cheung D, Dick EC, Timmers MC, De Klerk EPA, Spaan WJM,Sterk PJ. Rhinovirus inhalation causes long-lasting excessive airway narrowing in response to methacholine in asthmatic subjects in vivo. Am J Respir Crit Care Med 1995;152:1490-6. 26. Fleming HE, Little FF, Schurr D, Avila PC, Wong H, Liu J, et al. Rhinovirus-16 colds in healthy and in asthmatic subjects. Am J Respir Crit Care Med 1999;160:100-8. 27. Gern JE, Galagan DM, Jarjour NN, Dick EC, Busse WW. Detection of rhinovirus RNA in lower airway cells during experimentally induced infection. Am J Respir Crit Care Med 1997; 155:1159-61.