INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Viral bronchiolitis in human infants has been associated with chronic airway dysfunction later in life, although it is not proven that viral illness can be implicated as a cause of asthma or other airway disorders (1, 2). To identify potential mechanisms by which viral illness early in life might be a factor in the development of the asthmatic phenotype, a rat model has been developed which has several features that parallel human asthma, including episodic, reversible airway obstruction, airway hyperresponsiveness to methacholine, chronic airway inflammation, and airway wall remodeling (3, 4). There appears to be a genetic predisposition to the development of the postviral asthmalike syndrome in rats. Some inbred strains, such as the Brown Norway (BN), readily develop the postviral sequelae, whereas others, such as the Fischer 344 (F344) strain, are highly resistant to postviral sequelae (3). The development of the postviral syndrome may be related to the immune response to acute viral illness, in that the BN strain differs from the F344 strain by having greater expression of interleukin-4 (IL-4) and IL-5, less interferon gamma (IFN-) production, fewer CD8-positive lymphocytes, and prolonged viral replication within the airways (5, 6). Because IFN- may have roles in controlling viral infections and in limiting collagen synthesis (7), we hypothesized that the lack of a vigorous IFN- response during acute viral illness might be an important factor in the development of the asthmalike syndrome after viral bronchiolitis in young BN rats. The purpose of this study was to test whether exogenous IFN-, given by the inhaled route during acute viral illness in weanling BN rats, could alter the course of the acute illness or alter the development of postviral airway sequelae.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Design

Weanling Brown Norway rats were infected with Sendai virus, and given daily treatments with aerosolized IFN- or sterile buffer, beginning 2 d before inoculation and ending on postinoculation Day 7; a sham-inoculated, saline-treated group served as a control. The acute viral illness was assessed during the first 7 d after inoculation, including lung viral titers, body weight changes, bronchoalveolar lavage (BAL) leukocytes, and whole-lung messenger RNA (mRNA) for IL-4 and IFN-. The postviral sequelae, including bronchiolar inflammation, bronchiolar fibrosis, and changes in pulmonary mechanics, were evaluated 4 to 10 wk postinoculation.

Animals

Male pathogen-free inbred BN rats were purchased as 3- to 4-wk-old weanlings from Charles River Breeding Laboratories (Raleigh, NC or Kingston, NY). All animals were housed in an American Association for Accreditation of Laboratory Animal Care (AAALAC)-accredited isolation facility, with noninfected and virus-infected groups having separate but identical rooms. Procedures were approved by the University of Wisconsin Animal Care and Use Committee.

Viral Procedures

Weanling rats were inoculated with parainfluenza type 1 (Sendai) virus strain P3193 by exposing them to an aerosol generated from stock fluid containing 10⁷ to 10⁸ plaque-forming units (pfu) virus/ml, 2 ml of which was delivered into a Glas-Col Aerosol Exposure Apparatus (Glas-Col, Terre Haute, IN) over 20 min. Noninfected controls were sham-inoculated in the same manner, using an aerosol of sterile vehicle (diluted chorioallantoic fluid). Viral titers were measured in homogenates of frozen lung, using a plaque assay described previously (8), and expressed as pfu/g lung tissue.

IFN- Treatments

The inhaled route was selected for IFN- treatment, based on successful use of this method in studies involving pulmonary allergen challenges in mice (9, 10). Recombinant rat IFN- was purchased from Biosource International (Camarillo, CA); 500,000 units were dissolved in 15 ml sterile phosphate-buffered saline (PBS), and placed in an ultrasonic nebulizer (Ultra-Neb 99; DeVilbiss, Somerset, PA). Compressed air flowing through the nebulizer cup at 2 L/min carried the aerosol to a 25-L plastic box containing the rats, delivering a total of 8.5 ± 0.5 (mean ± SD) ml of solution from the nebulizer cup during a 20-min treatment. Treatments consisted of 20 min exposure in the aerosol box each morning, starting 2 d before inoculation and ending on postinoculation Day 7. The rats were awake and unrestrained during aerosol treatments. Sterile PBS was used for treatments in the same manner as a control.

Data Collection Procedures for Acute Studies

Infected and noninfected rats were studied separately to avoid exposure of the noninfected groups to virus. Animals of the viral inoculation group were coded so that all measurements were conducted with the investigator blinded with respect to the treatment grouping. For terminal studies, rats were anesthetized with urethane 1.5 g/kg, intraperitoneally (Sigma, St. Louis, MO) and exsanguinated via cardiac puncture. Lungs to be used for virus titers or cytokine mRNA were removed aseptically, frozen in liquid nitrogen, and stored at 80° C. In some of the rats, the lungs were lavaged with cold Hanks' balanced salt solution (HBSS) 5 times by filling to total lung capacity (20 cm H₂O fluid pressure) and then draining by gravity. The BAL fluid was centrifuged for 15 min, and the cell pellet was resuspended in 1.0 ml buffer. The number of cells obtained by BAL was measured using a Hemo-W Cell counter (Coulter Electronics, Hialeah, FL), and the leukocyte differential count was determined from 200 cells on a Cytospin (Shandon Lipshaw Inc., Pittsburgh, PA) slide stained with Diff-Quik (Baxter Healthcare Corp., Miami, FL).

Cytokine mRNA

Frozen lungs were weighed and RNA was recovered by lysis in guanidium thiocyanate and by phenol/chloroform extraction (11, 12). IL-4 and IFN- mRNA were detected in lungs by a competitive reverse transcriptase/polymerase chain reaction (RT-PCR) method as described (13) using a multispecific competitive fragment generously supplied by Angela Siegling (Institute for Medical Immunology, Humboldt University, Berlin, Germany). PCR primers for cytokines IL-4, IFN-) and for a housekeeping gene product hypoxanthine-guanine phosphoribosyltransferase (HPRT) were prepared as described (13) by the Interdisciplinary Center for Biotechnology and Research at the University of Florida. Each assay was optimized for temperature, Mg⁺⁺ concentration, and primers. For each complementary DNA (cDNA) sample, PCR reactions were performed containing 0, 0.5, 5, and 50 femtograms (fg) of competitive fragment. The cycler programs consisted of 1 min at 94° C, annealing temperature (56° C) for 15 min and 72° C for 2 min for 4 cycles, and then 36 cycles with the same temperatures and times except for duration changes to 2 min for the annealing step.

PCR products were stained with ethidium bromide and separated electrophoretically on 1.5% agarose gels. Gels were recorded and analyzed with a computerized video imaging system and software (Model 7000; UVP Inc., San Gabriel, CA). Endpoints were calculated as described (13). Data on cytokine mRNA abundance were recorded and analyzed with and without normalization to HPRT levels. There were no significant differences in the pattern and differences of mean values between the two methods of analysis. The data are reported as non-normalized mRNA abundance in competitive fragment units.

Histology

Previous studies found postviral abnormalities in lung histology to be present by postinoculation week 4 (3), and persisting at least 9 to 16 wk; we selected Week 4 as a representative time point to evaluate bronchiolar inflammation and fibrosis for this study. Rats were anesthetized and exsanguinated as described previously. A tracheal cannula was placed, and the lungs were filled with 10% buffered formalin (Fischer Scientific, Fair Lawn, NJ) at 20 to 30 cm H₂O pressure. Lungs were removed en bloc. Two blocks of the left lung were sectioned and embedded in paraffin. Serial paraffin microtome sections were stained with hematoxylin and eosin and with Van Gieson connective tissue stain. Each branch of bronchiole (airway generation) cut in transverse, longitudinal, or oblique planes in histologic sections was counted and was evaluated for the presence of inflammation and for the presence of fibrosis and/or other remodeling. Each slide was assigned a random number so that the pathologist had no knowledge of the experimental group the rat lung was from during the evaluation. An average of 34 bronchioles per rat were evaluated (range 20 to 52). A bronchiolar branch was evaluated as positive for inflammation if the wall of the bronchiolar branch contained three or more inflammatory cells (eosinophils, lymphocytes, or macrophages). A bronchiole was assessed as positive for fibrosis/remodeling if there was airway wall thickening that was associated with increased density of collagen that was detectable in the Van Gieson stain. Numbers of bronchioles with inflammation or fibrosis/remodeling were divided by the number of total bronchioles examined to calculate the percentage of bronchioles with each pathologic change.

Physiology

Postviral abnormalities in lung mechanics have been detected in previous studies at time points ranging from 4 to 18 wk postinoculation (3, 4). Weeks 8 and 10 were selected for measurements of pulmonary resistance and dynamic pulmonary compliance; at this time point the rats are > 200 g and are easily intubated via the oral route. Rats were sedated with pentobarbital (40 mg/kg, intraperitoneally; Abbott Laboratories, North Chicago, IL), instrumented with an orotracheal cannula and an esophageal pressure probe, and placed in a total body rodent plethysmograph, as described previously (4). Resistance was determined from simultaneous changes in esophageal pressure and in airflow, and compliance was determined from changes in esophageal pressure and lung volume, both computed for each breath by a Buxco Model 6 Pulmonary Mechanics Analyzer and DA16 data logger (Buxco Electronics, Sharon, CT). Pulmonary resistance was determined by subtracting the resistance of the tracheal cannula and connector from the total measured resistance. Because of the episodic nature of postviral airway obstruction in this rat model (4), resistance and compliance were measured on two occasions in each rat, separated by a 2-wk interval; the average of the two measurements was used for analyses.

Data Analysis

One-way analysis of variance (ANOVA) was used to test for differences among groups for variables that met the assumptions for parametric tests; when significant differences were present, planned post hoc paired comparisons were done with Fisher's least significant difference (LSD) test. The independent sample t test was used to test parametric data from protocols involving two treatment groups. Pulmonary resistance was transformed to its inverse (conductance) to conform to parametric assumptions for ANOVA. Body weight was used as a covariate for the analysis of dynamic compliance. Variables that did not conform readily to parametric assumptions were tested analogously using Kruskal-Wallis and Mann-Whitney tests. SYSTAT version 7.0 (SPSS Inc., Chicago, IL) software was used for analyses.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Studies during Acute Viral Infection

Treatment with IFN- had minimal effects on the course of the acute viral illness. Lung viral titers (Figure 1) were not affected by daily treatments with aerosolized IFN-, being similar for the two treatment groups of virus-infected rats with respect to both peak titer and timing of viral clearance. Similarly, virus-induced growth retardation (Figure 2) and numbers of BAL leukocytes (Figure 3) were not significantly altered by the IFN- treatments in the infected rats. The BAL leukocyte differential counts also were similar for the two treatment groups of virus-infected rats, except that the IFN- group had an earlier increase in BAL neutrophils, the percentage of neutrophils being significantly higher at postinoculation Day 3 (20.4 ± 6.0 versus 7.9 ± 5.3 [mean ± SD] %neutrophils, p = 0.02, t test).

View larger version (11K): [in this window] [in a new window]   Figure 1.   Viral titers. Viral titers in IFN-- (open circles) and PBS- (open squares) treated rats were similar for the two treatment groups at all time points after inoculation with virus. Each symbol represents one rat.

View larger version (12K): [in this window] [in a new window]   Figure 2.   Weight change after inoculation. Compared with noninfected rats (open triangles, n = 10), both groups of virus-infected rats had transient growth retardation (p < 0.0001 at Day 7), but IFN- (open circles, n = 34) and PBS- (open squares, n = 35) treated rats were not significantly different from each other (p = 0.15). Symbols are mean ± SD for each group.

View larger version (10K): [in this window] [in a new window]   Figure 3.   BAL fluid leukocytes. The numbers of BAL leukocytes were similar for IFN-- (open circles) and PBS- (open squares) treated rats at all time points after inoculation with virus. Each symbol represents one rat.

Relative amounts of IFN- mRNA from lung homogenates obtained on Days 3, 5, and 7 after viral inoculation were consistently higher compared with samples from noninfected rats (p = 0.002, Mann-Whitney test), but IFN- mRNA expression was not altered by treatment with aerosolized IFN- (p > 0.4, Mann-Whitney). IL-4 mRNA was variable, but tended to be more abundant in virus-infected compared with noninfected rat lungs (p = 0.06, Mann-Whitney). The IL-4 mRNA was not significantly altered by aerosolized IFN- treatment when the treatment effect was tested for all time points combined, but on postinoculation Day 5 the IFN- treated rats had significantly less lung IL-4 mRNA compared with that of PBS-treated rats (p = 0.015, Mann-Whitney).

Studies of Postviral Sequelae

At 4 wk after inoculation, lungs were evaluated for the extent of involvement of bronchiolar fibrosis and bronchiolar inflammatory cell aggregates (Figures 4 and 5). Compared with the noninfected control group, postviral rats from the PBS treatment group had significant bronchiolar fibrosis and inflammation; these postviral effects were attenuated markedly in the IFN- treatment group, resulting in bronchiolar fibrosis and inflammation scores that were not significantly different from those of the noninfected control group (Figure 5).

View larger version (76K): [in this window] [in a new window]   Figure 4.   (A) Terminal bronchiole from a noninfected rat demonstrating normal airway wall structure. Hematoxylin-eosin stain; original magnification ×260. (B) Terminal bronchiole from a virus-infected rat treated with PBS demonstrating persistent thickening (asterisks) and fibrosis of the airway wall and accumulation of inflammatory cells in the wall at 4 wk after inoculation. Inflammatory cells in this airway are predominantly eosinophils. Original magnification: ×260.

View larger version (16K): [in this window] [in a new window]   Figure 5.   Bronchiolar fibrosis and inflammation at 4 wk after inoculation. Postviral rats from the PBS-treated group (V-PBS) had significantly greater bronchiolar fibrosis and inflammation compared with both noninfected control and IFN--treated postviral (V-IFN-) groups. The p values for pairwise group comparisons are above each plot. Each symbol represents one rat; bars indicate group medians.

At 8 to 10 wk after inoculation, pulmonary resistance and dynamic compliance were measured to evaluate postviral changes in lung physiology. Postviral abnormalities were present in the PBS-treated virus group, measured as elevated resistance and decreased dynamic compliance compared with those of the noninfected group (Figure 6). The IFN- treated postviral group, however, had no significant abnormalities in either resistance or dynamic compliance (Figure 6), both variables being significantly different than those of the PBS-treated postviral group, while not significantly different from those of the noninfected group.

View larger version (16K): [in this window] [in a new window]   Figure 6.   Pulmonary resistance and dynamic compliance at 8 to 10 wk after inoculation. Postviral rats from the PBS-treated group (V-PBS) had significantly greater resistance and reduced dynamic compliance compared with both noninfected control and IFN--treated postviral (V-IFN-) groups. The p values for pairwise group comparisons are above each plot. Each symbol represents the average of two measurements in one rat; bars indicate group medians.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Treatment of weanling BN rats with aerosolized IFN- during acute Sendai virus bronchiolitis was effective in attenuating the persistent postviral bronchiolar fibrosis, bronchiolar inflammation, and airway obstruction that was present in the PBS-treated postviral group. This protective effect was not associated with an alteration of the acute viral illness, as measured by peak viral titers, duration of viral titers, numbers of BAL leukocytes, and virus-associated growth retardation. Thus, while the viral illness initiated the processes that created the postviral airway sequelae, these processes were not linked inexorably with the intensity or duration of the acute viral illness. This is a potentially important observation, in that it supports the notion that chronic airway disease is not necessarily determined by the intensity or duration of the initial injury, but instead by a failure to control and terminate the postinjury repair processes. The results of this study suggest that an early intervention aimed at modulating the inflammation and repair process may allow resolution, avoiding the eventual evolution to chronic inflammation, remodeling, and physiological dysfunction of the airways.

Reduction of bronchiolar fibrosis may be central to the protective effects of IFN- treatment in this study. IFN- treatment has been shown to attenuate increases of transforming growth factor-beta (TGF-) mRNA, procollagen mRNA, and total lung collagen after bleomycin pulmonary challenge in mice (14). In the rat bronchiolitis model, we observed previously that the F344 strain had a transient increase in lung TGF-₁ mRNA and in bronchiolar macrophages positive for TGF-₁ protein, both being back to control levels by Day 14 after viral inoculation (3). In contrast, BN rats had larger and more persistent increases in lung TGF-₁ mRNA, and in TGF-positive bronchiolar macrophages, having continued elevated numbers of TGF-positive bronchiolar macrophages at 30 d after viral inoculation (3). Similarly, the profibrotic cytokine tumor necrosis factor-alpha (TNF-) was more prominent in virus-infected BN rats, with lung TNF- mRNA and numbers of bronchiolar fibroblasts positive for TNF- protein both having greater magnitudes and longer persistence in response to viral illness compared with those of F344 rats (12). In the current study, it is possible that the IFN- treatment modulated a profibrotic cytokine environment in the BN rats and/or directly inhibited collagen production, thereby preventing excessive deposition of extracellular matrix in the bronchiolar walls after repair of the viral airway injury. Because the IFN- treatments ended on postinoculation Day 7, when virus was still present and acute airway inflammation was maximal, a direct inhibition of collagen deposition during the subsequent period of airway repair and remodeling seems less likely a mechanism than a more fundamental modulation of the cells and cytokines involved in the response to viral illness. Bronchiolar wall thickening theoretically could be responsible for the airway obstruction and hyperresponsiveness of the postviral syndrome in rats (15, 16), and thus it is plausible that reduced postviral bronchiolar inflammation and fibrosis in the IFN--treated rats could account for the absence of abnormalities in lung mechanics in this group.

Although the results of this study are consistent with our hypothesis that the postviral airway sequelae in BN rats is a consequence of a suboptimal IFN- response during viral illness, it is possible that the efficacy of IFN- treatment was due to a supraphysiologic pharmacological effect of the treatment, rather than due to correction of a relative IFN- insufficiency. Future studies will address whether selective depletion of IFN- is sufficient to produce postviral airway sequelae in the normally resistant F344 rats.

A number of observations in human infants suggest that factors related to age and genetic background may affect IFN- production by stimulated mononuclear leukocytes. Cord blood mononuclear cells of newborns who go on to develop atopic dermatitis or asthmatic symptoms in the first year of life produce significantly less IFN- compared with the cord blood cells of those who do not (17). Similarly, peripheral blood lymphocytes from infants and children having a strong family history for asthma or other atopic diseases produce less IFN- compared with lymphocytes from matched control subjects without such family history (18). Finally, IFN- production by blood mononuclear cells increases with age (18, 19), and this may be due in part to an age-dependent change in IL-12 expression (20). The rat bronchiolitis model is also consistent with these observations in that BN rats are high IgE producers (21) and young BN rats produce less IL-12 during viral illness compared with F344 rats (22).

Taken together, these observations suggest that a dysregulation in cytokine responses to various environmental stimuli early in life may be associated with the development of atopic diseases, including asthma. These data also imply that at least one abnormality may relate to a relative imbalance in the production of T helper cell, type 1 (Th1; IFN-, IL-2) and Th2 (IL-4, IL-5) cytokines. Because a relative decrease in Th1 response may augment IL-4 regulation of IgE antibody production (23), the observed links between diminished IFN- production and the development of atopic diseases appear to have an immunologic basis. The strong association between elevated IgE antibody levels measured in children at 9 mo of age and the development of persistent wheezing or asthma further reinforces these potential relationships (24).

Because not all infants with reduced IFN- production develop asthma, and because the persistence into childhood of a Th1/Th2 cytokine imbalance alone is more commonly associated with atopy rather than asthma (25), it seems likely that some other process must occur that ultimately actuates the chronic airway inflammatory processes and their physiologic consequences; that is, the asthmatic phenotype seems to result from a combination of both an inheritable component (cytokine dysregulation or atopic trait) and an environmental component (viral infections and allergens). Consistent with this concept are preliminary reports suggesting that viral lower respiratory tract illness, particularly respiratory syncytial virus or parainfluenza viral infections, may play a role in the development of the asthmatic phenotype (26).

In conclusion, the ability of IFN- treatment of weanling BN rats during acute viral bronchiolitis to prevent the development of the postviral asthmalike syndrome is consistent with the idea that the chronic airway dysfunction of human asthma may be the result of subtle imbalances in the control of the inflammation and repair responses to viral airway epithelial injury. These results further suggest that it may be possible to interrupt the evolution of acute airway injury to chronic airway disease with an early immunomodulatory intervention. With its many parallels to the immunologic, microbiologic, and physiologic factors that are thought to contribute to human asthma, the rat bronchiolitis model should provide a means to investigate the relative importance of cytokine dysregulation and viral infections early in life to the initiation of this common human respiratory tract disease.