Relationship of Upper and Lower Airway Cytokines to

Outcome of Experimental Rhinovirus Infection

JAMES E. GERN, ROSE VRTIS, KRISTINE A. GRINDLE, CHERI SWENSON, and WILLIAM W. BUSSE
Departments of Pediatrics and Medicine, University of Wisconsin Medical School, Madison, Wisconsin

To test the hypothesis that rhinovirus (RV)-induced immune responses influence the outcome of RV infections, we inoculated 22
subjects with allergic rhinitis or asthma with RV16. Nasal secretions and induced sputum were repeatedly sampled over the next
14 d. RV16 infection increased nasal granulocyte colony-stimulating factor (G-CSF) and interleukin (IL)-8, which was accompanied
by neutrophilia in blood and nasal secretions. Nasal G-CSF correlated closely with increased blood neutrophils (rs 0.69, p
0.005), whereas nasal neutrophils correlated with both G-CSF (rs
0.87, p 0.001) and IL-8 (rs 0.75, p 0.001). Although similar
relationships were present in sputum, changes in sputum neutrophils and G-CSF with RV16 infection were relatively modest. In addition, virus-induced changes in the sputum interferon--to-IL-5
messenger RNA ratio were inversely related to both peak cold
symptoms (rs 0.60, p 0.005) and the time to viral clearance
(undetectable picornavirus RNA). These results indicate that airway IL-8 and G-CSF are closely associated with virus-induced neutrophilic inflammation during an experimental RV infection in
atopic volunteers. In addition, the balance of airway T-helper cell
type 1 (Th1)- and Th2-like cytokines induced by RV infection may
help determine the clinical outcome of common cold infections,
raising the possibility that the individual subjects immune response, rather than atopic status per se, is important in this regard.

toms (5), and ex vivo secretion of IL-2 by peripheral blood

mononuclear cells (PBMC) correlates with the severity of experimental infection (6). Moreover, when PBMC from volunteers were evaluated ex vivo before experimental infection
with RV16, strong virus-induced proliferative responses and
interferon (IFN)- secretion were associated with reduced viral shedding and with cold symptoms, respectively, after inoculation (7).
These data suggest that variations in the generation of cytokines that regulate neutrophils, and the character of T-helper
cell type 1 (Th1) and Th2 responses to RV infection, could be
responsible for greater severity of illness in some individuals.
To test this hypothesis, we experimentally infected 22 seronegative subjects who had either allergic asthma or rhinitis with
RV16, and compared cytokine production and cell recruitment in the subjects nasal lavage fluid and sputum with outcomes of experimental infection as measured through viral titers in nasal secretions, symptom scores, and changes in
pulmonary function. The cytokines measured were granulocyte colony-stimulating factor (G-CSF) and IL-8, which have
been linked to the increase in blood and airway neutrophils,
respectively, in RV16 infection (8, 9), and IFN- and IL-5, as
examples of Th1- and Th2-like cytokines.

Respiratory infections with human rhinoviruses (RV) are frequent triggers of cough and wheezing in children and adults
with asthma. Soon after the onset of these illnesses, immune
responses are triggered that usually eradicate viral replication
within days. However, since RV does not destroy appreciable
amounts of airway epithelium, it is presumed that the immune
response to the virus also contributes to the pathogenesis of
respiratory symptoms, including dysfunction of the lower respiratory tract in asthma.
This hypothesis is supported by several observations in volunteers after experimental infection with RV. First, antiviral
medications are effective only if treatment is started in the
first 24 to 48 h after the onset of clinical illness (1), suggesting
that the immune response to the virus is thereafter responsible
for respiratory symptoms. In addition, many studies have
linked the intensity of the immune response to RV to the severity of cold symptoms. For example, neutrophil counts in
nasal secretions, as well as quantities of neutrophil chemoattractants such as interleukin (IL)-8, correlate with cold symptoms (2, 3) and, in subjects with asthma, correlate with changes
in airway responsiveness (4).
Lymphocyte responses have also been related to the outcome of RV infection. For example, the degree of T lymphopenia during the acute infection is inversely related to cold symp-

METHODS

(Received in original form March 2, 2000 and in revised form July 24, 2000)
Supported by grants AI40685 and AI34891 from the National Institutes of Health.
Correspondence and requests for reprints should be addressed to James E. Gern
M.D., H4/438 University of Wisconsin Hospital, Madison, WI 53792-4108. E-mail:
gern@medicine.wisc.edu
Am J Respir Crit Care Med Vol 162. pp 22262231, 2000
Internet address: www.atsjournals.org

Subjects
Twenty-two subjects with allergic rhinitis or mild allergic asthma participated in the study; the subjects characteristics are listed in Table I.
Only subjects with respiratory allergies or mild allergic asthma were
included in the study because these individuals are more likely to develop changes in lower airway physiology and/or lung function after
experimental or natural infections (10, 11). Subjects with moderate or
severe asthma, or those with a history of severe bronchospasm during
viral colds, were excluded from this initial trial because of safety concerns. Asthma was defined by a clinical history of (1) physician diagnosis; (2) prescription for asthma medication; (3) cough, wheezing or
shortness of breath and relief with appropriate asthma medication;
and (4) evidence of a 12% or greater improvement in FEV1 after inhalation of a -adrenergic agonist. In addition, methacholine testing
was performed (Table 1), and for the purposes of this study, subjects
with cumulative values for the provocative concentration of MCh
needed to reduce FEV1, by 20% (PC20) 64 mg/ml were classified as
having allergic rhinitis and not asthma.
At the screening visit, all subjects underwent a physical examination, skin prick tests to 12 common aeroallergens (including cat, house
dust mite, and an assortment of local pollens and molds), and spirometry. Atopy was defined as one or more positive skin tests (wheal size
histamine control). None of the subjects had neutralizing antibody
to RV16 at the time of the screening visit (1 to 3 mo before inoculation), although seven subjects had low titers (median: 1.4) of RV16neutralizing antibody detected in the serum obtained just before inoculation. Use of topical nasal corticosteroids was withheld for 30 d before
the study, and acetaminophen was the only medication allowed for
the treatment of cold symptoms during the study. The experimental
protocol was approved by the University of Wisconsin Hospital and
Clinics Human Subjects Committee, and informed consent was obtained from the study subjects before they were enrolled in the protocol. Data related to peripheral blood responses to viral infection in
the study subjects have previously been published (7).

2227

Gern, Vrtis, Grindle, et al.: Cytokines and Rhinovirus Infection

Overview of Study Design
Following the establishment of baseline values, subjects were inoculated with a standardized RV16 suspension on two successive days as
previously described (7). Nasal lavage and sputum induction were
performed at baseline and at 1, 2, 7, and 14 d after inoculation. Nasal
lavage fluids were subjected to quantitative viral culture as previously
described (10). Each of the subjects developed a respiratory infection,
as demonstrated either by virus cultured from nasal secretions or by a
rise in titer of RV16-specific neutralizing antibodies (Table 1). Symptoms (cough, nasal discharge, sneezing, stuffy nose, sore throat, headache, malaise, chilliness, shaking chills, fever, hoarseness, aching muscles or joints, and watering or burning of the eyes) were quantitated (0
not present, 1 mild, 2 moderate, 3 severe) daily by each subject
and recorded in a diary. The maximum possible symptom score was
39. In addition, all subjects were questioned at each clinic visit about
the presence of respiratory symptoms including cough and wheezing,
and were asked about increased use of asthma rescue medications
during the preceding phase of the study.

Sputum Induction
The method for obtaining sputum samples was based on that of Fahy
and colleagues (12). At the beginning of the sputum induction protocol, subjects were given 2 puffs (180 g) of albuterol, and spirometry
was performed after 10 min. The subjects were then given general instructions to produce a deep cough, to ignore the natural inclination
to swallow sputum, and to quickly blow the nose and rinse the mouth
before the sputum was collected. Subjects next breathed a 3% buffered saline solution mist from an DeVilbiss (Jackson, TN) 65 ultrasonic nebulizer for 20 min, with a deep breath taken once per minute.
After 5 and 10 min, spirometry was repeated to screen for bronchoconstrictive responses. The subjects were instructed to try to produce
sputum whenever they felt the need to expectorate and at the 10-, 15-,
and 20-min time points. To minimize contamination from the noses
and mouths, the subjects were asked to blow their noses, vigorously
rinse their mouths, and gargle with sterile water, followed by expectoration into a clean specimen container. If the subjects FEV1 fell beTABLE 1
SUBJECT CHARACTERISTICS

Subject Sex Age Disease FEV1*

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

M
M
M
F
M
M
M
F
M
M
M
F
F
F
F
F
F
F
F
M
F
F

42
40
24
25
23
23
21
32
25
20
18
20
18
21
21
25
28
21
25
20
23
22

AR
AA
AA
AA
AR
AA
AR
AR
AR
AA
AA
AA
AA
AA
AA
AR
AA
AR
AA
AA
AA
AA

86
84
106
112
111
93
88
130
110
83
104
78
80
92
94
107
120
103
99
92
104
97

PD20

59
41
1.6
19
200
3.2
200
200
200
75
70
2.6
37
8.5
37
200
31
99
5.4
8.7
33
13

Antibody Titer
Peak
Cold SS Acute Convalescent
23
19
7
14
12
9
4
5
4
2
10
1
15
9
15
17
15
14
10
8
9
10

1.0
0
0
0
1.4
0
0
0
0
1.0
0
2.8
0
1.4
0
1.0
1.4
0
0
0
0
0

45
91
2.0
91
16
5.6
45
1.4
5.7
45
1.4
91
2.0
32
32
64
91
45
1.4
4.0
11
32

Definition of abbreviations: AA allergic asthma; AR allergic rhinitis; RV16 respiratory virus 16; PD20 provocative dose of methacholine causing a 20% decrease in
FEV1; SS symptom score.
* Percent predicted.

Cumulative dose units.

None of the subjects had detectable neutralizing antibody to RV16 at the time of
the initial screening visit. Serum was also obtained on the day of inoculation, and the titers in this table reflect the values on this day.

low 80% of the postbronchodilator FEV1 at any time during the induction, the procedure was interrupted until the FEV1 returned to
within 20% of the postbronchodilator baseline value.

Nasal Lavage
Nasal lavage was performed on each subject both before RV16 inoculation and during RV infection, as previously described (8). Five milliliters
of prewarmed Hanks balanced salt solution (HBSS) with 0.5% gelatin
was instilled into each nostril. The lavage fluid was then expelled consecutively from each nostril by having the subject blowing gently from one
nostril while the other nostril was closed by compression.

Processing of Nasal Lavage and Sputum

To release cells from mucus, sputum and nasal lavage samples were
mixed with an equal volume of Sputolysin (Calbiochem, La Jolla, CA;
0.1% dithiothreitol prepared in phosphate-buffered saline). This mixture was vortexed, rocked (20 C for 20 min), vortexed again and centrifuged (at 750 g for 10 min at 4 C). Aliquots of supernatant fluids
were saved at 80 C pending analysis for cytokines. The cell pellet
was washed with HBSS (4 C) containing 2% newborn calf serum and
was then filtered through 50-m nylon mesh before counting of the
total cell number. Both cell counts and differential counts were made.
RNA was isolated from samples of cells (5 105) with a reagent
containing phenol and chloroform (Trizol; Life Technologies, Rockville, MD), and the RNA extracts were frozen (80 C) until they
were analyzed. Enzyme-linked immunosorbent assays (ELISAs) for
IL-5, IL-8, and IFN- were performed as previously described (13,
14), and G-CSF was quantitated with a commercial kit (R&D Systems, Minneapolis, MN). The sensitivities of the ELISAs were as follows: IL-5 and IFN-, 5 pg/ml; IL-8, 15 pg/ml; and G-CSF, 20 pg/ml.

Cytokine Messenger RNA

The amount of IL-5 and IFN- messenger RNA (mRNA) in sputum
cells was quantitated with a semiquantitative reverse transcription
polymerase chain reaction (RT-PCR) assay as previously described
(13). Briefly, after RT of total cellular RNA, IL-5, IFN-, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) sequences were
amplified in 25 to 33 cycles of PCR (45 s at 94 C, 45 s at 60 C, 2 min
at 72 C, and 7 min at 72 C for terminal extension) with specific primers (IFN-: 5GCA TCG TTT TGG GTT CTC TTG GCT GTT
ACT GC3 and 5CTC CTT TTT CGC TTC CCT GTT TTA GCT
GCT GG3; IL-5: 5G CTT CTG CAT TTG AGT TTG CTA
GCT3 and 5TG GCC GTC AAT GTA TTT CTT TAT TAA G
3; GAPDH: 5TGA AGG TCG GAG TCA ACG GAT TTG GT
3 and 5CAT GTG GGC CAT GAG GTC CAC CAC3, (Clontech Laboratories, Palo Alto, CA). A standard curve generated from
dilutions of a highly positive reference sample was included in each
run of PCR. Negative controls included in each PCR run were samples containing reagents but no complementary DNA (cDNA). The
PCR products were resolved by gel electrophoresis and transferred to
a nylon membrane for Southern blotting with cDNA probes. Concentrations of cytokine mRNAs are expressed in units relative to a highly
positive reference standard, which was arbitrarily assigned a value of
1,200 U. Detection of GAPDH mRNA was done to ensure that each
sample contained cellular mRNA and also as a positive control for
RT and PCR. Concentrations of the mRNAs for IFN- and IL-5 were
not corrected for GAPDH mRNA because of the large shifts in the
number and relative proportions of cells in the airway and were instead expressed as a ratio.

Detection of RV by RTPCR
Picornavirus RNA was detected with RTPCR (35 cycles) as previously described (15), using primers OL26 (5GCA CTT CTG TTT
CCC C3) and OL27 (5CGG ACA CCC AAA GTA G3) (16).
PCR products were resolved by gel electrophoresis, stained with
ethidium bromide, and photographed. The presence of a 380-bp band
was interpreted as a positive test for RV.

Statistical Analysis
Of the 22 subjects in the study, 15 had mild asthma. Subgroup analyses of subjects with allergic asthma versus those with allergic rhinitis

2228

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

VOL 162

2000

Cells and Cytokines in Nasal Secretions

and Peripheral Blood

Figure 1. Relationship between nasal cytokines and peripheral blood

neutrophils. One day after inoculation with RV16, levels of (A) granulocyte colony-stimulating factor (G-CSF) and (B) interleukin (IL)-8 in nasal lavage fluid were compared with peripheral blood neutrophil counts
24 h later.

revealed no group-specific significant differences in values for G-CSF

or IL-8 protein, or for IFN- or IL-5 mRNA levels. Therefore, the cytokine values for these subgroups were pooled in subsequent statistical analyses. Spearmans rank correlations test was used to evaluate
correlations of cytokine secretion or mRNA, symptom scores, and viral shedding with FEV1. Cold-related changes in these parameters
were evaluated with Wilcoxons signed ranks test. All analyses were
done with the aid of computer software (SigmaStat 2.0; Jandel Scientific, San Rafael, CA).

RESULTS
Clinical Effects of RV16 Inoculation

Of the 22 subjects who had no detectable antibody to RV16 at

the time of screening, seven had detectable neutralizing antibody to RV16 (median titer: 1.4, Table 1) approximately 1 mo
later, at the time of inoculation. After inoculation with RV16,
21 subjects developed symptoms of upper respiratory infection and had infection verified by viral culture. The remaining
subject neither developed cold symptoms nor shed virus after
inoculation, but had an increase in the titer of neutralizing antibody to RV16 and was included in the analysis. The mean
baseline symptom score was 0.95 (range: 0 to 3), and the mean
peak symptom score, which occurred from 1 to 3 d after inoculation, was 10.2 (range: 1 to 23, Table 1). The mean peak
symptom score (11.3) in the subgroup of subjects who had low
titers of RV16-neutralizing antibody in their serum was similar
to that of the group as a whole. There were significant correlations between viral shedding in nasal secretions and both
symptom scores (rs 0.58, p 0.005, Day 2) and peripheral
blood neutrophil count (rs 0.61, p 0.005, Day 2). None of
the study subjects reported wheezing or increased use of albuterol associated with the induced colds, and there were no
cold-related changes in mean FEV1 for the group as a whole.

Figure 2. Relationship between cytokines and neutrophils in nasal lavage fluid after RV16 inoculation. Two days after inoculation with RV16,
the levels of (A) G-CSF and (B) IL-8 were compared with neutrophil
counts in nasal lavage fluid.

Sufficient amounts of nasal lavage fluid for both quantitative

viral culture and cytokine analysis were obtained from 18 of the
22 study subjects. By the first day after inoculation, RV infection
significantly increased nasal lavage neutrophil counts (median before RV16-induced colds: 32,200/ml [interquartile range
(IQR): 5,344 to 94,600/ml]; median during colds: 363,000/ml
[IQR: 13,700 to 2,830,000/ml], p 0.05), and this coincided
with increased concentrations of G-CSF (median before colds:
53 pg/ml [IQR: 33 to 114 pg/ml], median during colds: 868 pg/ml
[IQR: 146 to 3,960] pg/ml p 0.05) and IL-8 (median before
colds: 288 pg/ml [IQR: 128 to 522]; median during colds: 519
pg/ml [IQR: 366 to 827 pg/ml], p 0.05). Levels of G-CSF
and IL-8 correlated closely with the amount of viral shedding
in nasal secretions (rs 0.89 and rs 0.72, respectively, p
0.001). In addition, there was a strong correlation between the
amount of G-CSF in nasal secretions at 1 d after inoculation
and the peripheral blood neutrophil count on the following
day (Figure 1A). In contrast, the correlation between IL-8 and
blood neutrophils was relatively weak (Figure 1B).
By the second day after inoculation, the relationships between nasal IL-8 and G-CSF and blood PMN were weaker (rs
0.55 and rs 0.32, respectively), but the amounts of these cytokines in nasal secretions correlated closely with nasal PMN
(Figure 2). In addition, there was a significant correlation between G-CSF in nasal secretions and cold symptom scores
(rs 0.58, p 0.01, n 18), and a trend toward a positive association between nasal IL-8 and symptom scores (rs 0.41,
p 0.09, n 18). There was no correlation between nasal
IL-8 or G-CSF and changes in FEV1 (data not shown).
Sputum Analysis

To determine the kinetics of cell influx into the lower airway

after RV infection, we analyzed induced sputum for its content of cells and cytokines at 2, 7, and 14 d after RV inoculation, and compared the results with baseline values. The quantities of total cells, lymphocytes, and eosinophils in sputum did
not change significantly after RV inoculation (data not shown).
Neutrophils in the sputum were significantly increased at 7 d
after inoculation, with numbers after the passing of a further
week being similar to those before cold induction (Figure 3).
To identify factors that could contribute to virus-induced
recruitment of neutrophils into the airway, we measured IL-8
and G-CSF levels in sputum supernatant fluids. Sputum G-CSF
was increased at 2 d after inoculation, and gradually drifted
back toward its baseline value (Figure 4A). No significant
changes were noted for sputum both IL-8 (data not shown).
The levels of both IL-8 and G-CSF correlated with numbers of

Figure 3. Cold-related changes in sputum neutrophils. Neutrophils were

quantitated in specimens of induced sputum before (pre-cold) and 2,
7, and 14 d after inoculation with RV16.

2229

Gern, Vrtis, Grindle, et al.: Cytokines and Rhinovirus Infection

Figure 4. Relationship of sputum cytokines to influx of neutrophils. (A)

Sputum G-CSF was quantitated before (pre-cold) and 2, 7, and 14 d after inoculation with RV16. The boxes
represent the 25th to 75th percentile
values, with the medians indicated by
horizontal lines, and error bars designate the 10th and 90th percentiles.
Seven days after inoculation, levels of
sputum (B) G-CSF and (C) IL-8 were
compared with neutrophil counts in
the sputum.

sputum neutrophils during the acute cold (Figures 4B and 4C).

Neither IL-8 nor G-CSF levels in sputum correlated with cold
symptom scores or changes in FEV1 (data not shown).
We measured protein and mRNA for both IL-5 and IFN-
to determine the relationship between virus-induced Th1 and
Th2 cytokine production and the effects of viral infection. Little or no IL-5 or IFN- protein was detected in sputum supernatants, and add-back experiments established that one or
more factors (e.g. sputum proteoglycan) masked the detection
of IL-5 and IFN- by ELISA (data not shown). In contrast,
both IFN- mRNA and IL-5 mRNA were detectable both
before colds and after inoculation. Expression of both IL-5
mRNA and IFN- mRNA was enhanced at 2 d after inoculation, and remained slightly increased at 7 d after inoculation
(Figure 5).
The mRNAs of IFN- and IL-5 were also expressed as a ratio. An increase in the IFN-/IL-5 ratio would represent a deviation toward a Th1-like pattern, whereas lower ratios would
suggest a drift toward a Th2-like pattern. Although RV infection increased the mRNA concentrations for both IL-5 and
IFN-, the ratio of IFN- mRNA to IL-5 mRNA did not
change during colds (data not shown). However, during the
acute phase of the illness there was an inverse relationship between the IFN- mRNA IL-5 mRNA ratio and peak cold
symptoms (Figure 6A). There was no significant relationship
between the IFN- mRNA/IL-5 mRNA ratio and changes in
airway eosinophils during the acute cold (data not shown).
In addition, we compared the IFN-/IL-5 ratio with the
presence of RV in nasal secretions and sputum. The percentage of subjects with picornavirus RNA detectable by RTPCR

Figure 5. Changes in (A) interferon (IFN)- and (B) IL-5 during acute
rhiovirus (RV) colds. Just before inoculation and 2 d and 7 d after inoculation, sputum cells were analyzed for IFN- and IL-5 mRNA. The horizontal
lines represent median values. *p 0.05 versus values before colds.

in sputum decreased with the time after inoculation: virus was

detected in three (14%) subjects before inoculation, in 21
(95%) subjects at 2 and 7 d after inoculation, and in 11 (50%)
subjects at 14 d after inoculation. Subjects without detectable
virus in their sputum at 14 d after inoculation had a significantly higher IFN- mRNA/IL-5 mRNA ratio than did the
rest of the group (Figure 6B). There was no relationship, however, between the IFN- mRNA/IL-5 mRNA ratio in the sputum and peak viral titers in nasal lavage fluid specimens (data
not shown).

DISCUSSION
In this study of the immune response to induced colds, the severity of symptoms and amount of viral shedding were quite
variable, despite careful control of the inoculation procedure
and use of the same lot of virus for all subjects. Since the subjects were seronegative at the time of screening (although a few
individuals had low titers of neutralizing antibody at the time
of inoculation), the severity of the viral infection was largely
determined by host immune factors other than the quantity of
virus-specific antibody. Analysis of cells and cytokines in the
blood, nasal lavage fluid, and sputum suggested that two different types of immune responses were related to the clinical
and virologic outcome of the infection. First, our findings support previous studies suggesting that airway neutrophilia and
factors such as IL-8 and G-CSF, which regulate neutrophil recruitment and activation, may be important in the pathogenesis of virus-induced respiratory symptoms and changes in airway responsiveness (3, 4, 9, 17). In addition, a novel finding of
our study was the association between relatively weak Th1like responses (low IFN-/IL-5 ratio) and more severe respira-

Figure 6. Association of IFN- mRNA/IL-5 mRNA ratio to outcomes of

respiratory infection. The IFN- mRNA/IL-5 mRNA ratios were calculated during the acute cold (48 h after inoculation) for each individual.
These ratios were (A) compared with the peak symptom score during
the acute cold, and (B) grouped according to viral clearance at 14 d after inoculation with RV16.

2230

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

tory symptoms as well as a longer duration of viral shedding in

the sputum.
In the nose, IL-8 and G-CSF were rapidly induced after viral inoculation, and appeared to be related to neutrophil trafficking in the airway. On the first day after inoculation, increases in nasal G-CSF correlated with increases in blood
neutrophils, suggesting that G-CSF produced in the nose enters the systemic circulation and activates hematopoeisis by
the bone marrow. On the second day, there was a close relationship between IL-8 and nasal neutrophils. Although an association between IL-8 and neutrophils or neutrophil myeloperoxidase activity has been previously observed (3, 9) and is
consistent with the activity of IL-8 as a neutrophil chemoattractant, we observed an even closer correlation between G-CSF
and neutrophils in nasal lavage fluid. This novel observation
suggests either that G-CSF contributes to neutrophil recruitment to the airway or that airway neutrophils are a source of
G-CSF during viral infection. Additional studies will be required to determine the mechanism responsible for this effect.
Measurement of cells and cytokines in the sputum was
done in an effort to further define the lower airway response
to RV infection and the timing of its development. Although
the changes in cellularity were small, increased neutrophils
were noted 1 wk after inoculation, suggesting that sputum
neutrophilia follows the recruitment of neutrophils into the
upper airway, perhaps indicating that RV infection first occurs
in the upper airway and then gradually moves into the chest.
Furthermore, changes in lower airway neutrophil numbers
correlated with changes in G-CSF and IL-8, suggesting that
the same factors are responsible for regulating neutrophil
recruitment in the upper and lower airways. However, we
found no relationship between changes in sputum neutrophils,
G-CSF, or IL-8 and alterations in lung function. The lack of
correlation could have been due to the relatively small changes
in lung function, sputum cellularity, and cytokines, and our
findings were similar to those reported in two previous studies
of sputum physiology after experimental infection with RV16
(9, 18). One limitation common to studies utilizing RV inoculation is that few study subjects actually develop exacerbations
of asthma, even when small changes in lung function are noted
(9, 19). Examination of sputum during naturally occurring virus-induced exacerbations of asthma may reveal more dramatic changes in sputum physiology, and thus may have
greater potential for establishing links with the pathogenesis
of acute airway obstruction.
A novel aspect of our study was its comparison of the pattern of Th1- and Th2-like cytokine responses, as indicated
by changes in sputum IFN- mRNA and IL-5 mRNA, with
the clinical and virologic outcome of RV infections. The ratio
of IFN- mRMA to IL-5 mRNA was determined in order to
identify tendencies toward a Th1-like (high ratio) or Th2-like
(low ratio) immune response to viral infection. RV infection
increased IFN- mRNA, which is expected of a virus infection, but IL-5 mRNA was also increased. This indicates that
RV infection causes activation of both Th1- and Th2-like cytokines, and this is supported by the absence of a net shift in
the IFN- mRNA/IL-5 mRNA ratio during acute colds.
The IFN- mRNA/IL-5 mRNA ratio during the acute
phase of infection was inversely related to two different outcome measures: peak cold symptom scores and time to virus
clearance from the sputum. These findings suggest that strong
Th1-like cytokine responses in the airway may play an important role in limiting the amount of viral replication and respiratory symptoms in colds caused by RV. This hypothesis is
consistent with our previous observations that vigorous increases in RV-induced IFN- secretion by PBMC incubated

VOL 162

2000

with RV ex vivo were associated with reduced viral replication

after experimental inoculation (7). Studies with rodents also
suggest that Th1-like cytokines such as IFN- are important in
the clearance of viral infections (20, 21) and that administration of exogenous IFN- can reduce chronic airway dysfunction after Sendai virus infection in a susceptible strain of
weanling rats (22).
These findings raise questions about the origin of the IFN-
mRNA and IL-5 mRNA in sputum, and the mechanism(s) for
their generation. A likely source of these cytokine mRNAs
are lymphocytes. RV-specific T-cell clones have been shown
to secrete large amounts of IFN- in vitro, but some clones
also secrete Th2-like cytokines such as IL-4 and IL-5 (14), and
mouse models of viral infection also support the concept that
virus-activated lymphocytes can produce a broad range of
cytokines (23). In addition to antigen-specific activation of
lymphocytes, pathways have been described for antigen-independent activation of lymphocytes through the virus-induced
generation of soluble cytokines or mediators that secondarily
stimulate lymphocyte IFN- production (24, 25). Regardless
of the mechanism of their production, concurrent viral stimulation of IFN- and IL-5 could be a potent inflammatory stimulus. For example, in a mouse model of allergen-induced inflammation, passive transfer of Th1 clones enhanced, rather
than counteracted, allergen-induced inflammation and airway
responsiveness mediated by Th2 lymphocyte activation (26).
In summary, in the group of allergic individuals in our study,
experimental infection with RV16 caused rapid increases in
nasal G-CSF and IL-8, and these upper airway events were
closely associated with neutrophilic responses in the blood
(G-CSF) and nasal secretions (IL-8 and possibly G-CSF). Although corresponding changes in sputum neutrophils and
G-CSF were relatively modest, analysis of cytokine mRNA in
the sputum suggests that the balance of Th1- to Th2like
mRNA may play an important role in controlling viral clearance and respiratory symptoms in RV infections. There is accumulating evidence that markers of atopy are risk factors for
more severe clinical manifestations during RV infections (10,
27, 28). Since allergy and asthma are associated with an immune response that is skewed toward Th2-like responses, the
resulting cytokine imbalance may not only produce a distinct
response to allergens but may also impair immune responses
to viral infections. Additional controlled studies, to compare
Th1- and Th2-like immune responses with measures of clinical
outcome during naturally occurring RV infection, will be
needed to determine whether this could contribute to the
greater lower airway effects of RV infections in patients with
asthma.
References
1. Turner RB, Wecker MT, Pohl G, Witek TJ, Mcnally E, St George R,
Winther B, Hayden FG. Efficacy of tremacamra, a soluble intercellular adhesion molecule 1, for experimental rhinovirus infectiona randomized clinical trial. JAMA 1999;281:17971804.
2. Levandowski RA, Weaver CW, Jackson GG. Nasal secretion leukocyte
populations determined by flow cytometry during acute rhinovirus infection. J Med Virol 1988;25:423432.
3. Teran LM, Johnston SL, Schroder 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:
13621366.
4. Grnberg K, Timmers MC, Smits HH, De Klerk EPA, Dick EC, Spaan
WJM, Hiemstra PS, Sterk PJ. 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;27:3645.
5. Levandowski RA, Ou DW, Jackson GG. Acute-phase decrease of T
lymphocyte subsets in rhinovirus infection. J Infect Dis 1986;153:743748.
6. Hsia J, Goldstein AL, Simon GL, Sztein M, Hayden FG. Peripheral

2231

Gern, Vrtis, Grindle, et al.: Cytokines and Rhinovirus Infection

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

blood mononuclear cell interleukin-2 and interferon- gamma production, cytotoxicity, and antigen-stimulated blastogenesis during experimental rhinovirus infection. J Infect Dis 1990;162:591597.
Parry DE, Busse WW, Sukow KA, Dick CR, Swenson CA, Gern JE.
Rhinovirus-induced peripheral blood mononuclear cell responses and
outcome of experimental infection in allergic subjects. J Allergy Clin
Immunol 2000;105:692698.
Jarjour NN, Gern JE, Kelly EAB, Swenson CA, Dick CR, Busse WW.
The effect of an experimental rhinovirus 16 infection on bronchial lavage neutrophils. J Allergy Clin Immunol 2000;105:11691177.
Fleming HE, Little FF, Schnurr D, Avila PC, Wong H, Liu J, Yagi S,
Boushey HA. Rhinovirus-16 colds in healthy and in asthmatic subjects: similar changes in upper and lower airways. Am J Respir Crit
Care Med 1999;160:100108.
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;155:18721876.
Folkerts G, Busse WW, Nijkamp FP, Sorkness R, Gern JE. State of the
art: virus-induced airway hyperresponsiveness and asthma. Am J Respir
Crit Care Med 1998;157:17081720.
Fahy JV, Liu J, Wong H, Boushey HA. Analysis of cellular and biochemical constituents of induced sputum after allergen challenge: a
method for studying allergic airway inflammation. J Allergy Clin Immunol 2000;93:10311039.
Schroth MK, Grimm E, Frindt P, Galagan DM, Konno S, Love R, Gern
JE. Rhinovirus replication causes RANTES production in primary
bronchial epithelial cells. Am J Respir Cell Mol Biol 1999;20:1220
1228.
Gern JE, Dick EC, Kelly EAB, Vrtis R, Klein B. Rhinovirus-specific T
cells recognize both shared and serotype-restricted viral epitopes. J
Infect Dis 1997;175:11081114.
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;155:11591161.
Gama RE, Horsnell PR, Hughes PJ, North C, Bruce CB, al-Nakib W,
Stanway G. Amplification of rhinovirus specific nucleic acids from
clinical samples using the polymerase chain reaction. J Med Virol 1989;
28:7377.
Turner RB, Weingand KW, Yeh CH, Leedy DW. Association between

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

interleukin-8 concentration in nasal secretions and severity of symptoms of experimental rhinovirus colds. Clin Infect Dis 1998;26:840846.
Grnberg K, Smits HH, Timmers MC, de Klerk EPA, Dolhain RJ, Dick
EC, Hiemstra PS, Sterk PJ. Experimental rhinovirus 16 infection: effects on cell differentials and soluble markers in sputum in asthmatic
subjects. Am J Respir Crit Care Med 1997;156:609616.
Grnberg K, Timmers MC, De Klerk EPA, Dick EC, Sterk PJ. Experimental rhinovirus 16 infections causes variable airway obstruction in
subjects with atopic asthma. Am J Respir Crit Care Med 1999;160:
13751380.
Tishon A, Lewicki H, Rall G, VonherrathM, Oldstone MBA. An essential rote for type 1 interferon-gamma in terminating persistent viral infection. Virology 1995;212:244250.
Fiette L, Aubert C, Mller U, Huang S, Aguet M, Brahic M, Bureau J-F.
Theilers virus infection of 129Sv mice that lack the interferon / or
interferon receptors. J Exp Med 1995;181:20692076.
Sorkness RL, Castleman WL, Kumar A, Kaplan MR, Lemanske RF Jr.
Prevention of chronic postbronchiolitis airway sequelae with IFN-
treatment in rats. Am J Respir Crit Care Med 1999;160:705710.
Coyle AJ, Erard F, Bertrand C, Walti S, Pircher H, Legros G. Virus-specific CD8(+) cells can switch to interleukin 5 production and induce
airway eosinophilia. J Exp Med 1995;181:12291233.
Gern JE, Vrtis R, Kelly EAB, Dick EC, Busse WW. Rhinovirus produces nonspecific activation of lymphocytes through a monocytedependent mechanism. J Immunol 1996;157:16051612.
Tough DF, Borrow P, Sprent J. Induction of bystander T cell proliferation by viruses and type I interferon in vivo. Science 1996;272:1947
1950.
Hansen G, Berry G, DeKruyff RH, Umetsu DT. Allergen-specific Th1
cells fail to counterbalance Th2 cell-induced airway hyperreactivity
but cause severe airway inflammation. J Clin Invest 1999;103:175183.
Rakes GP, Arruda E, Ingram JM, Hoover GE, Zambrano JC, Hayden
FG, Platts-Mills TA, Heymann PW. Rhinovirus and respiratory syncytial virus in wheezing children requiring emergency care: IgE and
eosinophil analyses. Am J Respir Crit Care Med 1999;159:785790.
Duff AL, Pomeranz ES, Gelber LE, Price HW, Farris H, Hayden FG,
Platts-Mills TAE, Heymann PW. Risk factors for acute wheezing in
infants in infants and children: viruses, passive smoke, and IgE antibodies to inhalant allergens. Pediatrics 1993;92:535540.