CD8+ T Lymphocytes in Viral Hyperreactivity and M2 Muscarinic Receptor Dysfunction

Department of Environmental Health Sciences, School of Hygiene and Public Health, and Division of Clinical Immunology, Department of Internal Medicine, School of Medicine, Johns Hopkins University, Baltimore, Maryland; and Department of Pediatrics, University of Alberta, Edmonton, Alberta, Canada

	ABSTRACT

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
In the airways, inhibitory M₂ muscarinic receptors (M₂Rs) on parasympathetic nerves limit acetylcholine release. Viral infection causes M₂R dysfunction, which increases acetylcholine release and leads to airway hyperreactivity. In these studies we tested the role of CD8⁺ T cells in parainfluenza virus-induced hyperreactivity and M₂R dysfunction in normal guinea pigs and in guinea pigs previously sensitized to ovalbumin. Depleting CD8⁺ T cells prevented virus-induced M₂R dysfunction and hyperreactivity in sensitized animals, but not in nonsensitized animals. Sensitization increased the number of eosinophils in close relation to the airway nerves where, when activated, they release major basic protein, which binds to and blocks the M₂Rs. Regardless of sensitization, viral infection decreased the number of visible tissue eosinophils, likely reflecting eosinophil degranulation via cytolysis. Depleting CD8⁺ T cells prevented this virus-induced eosinophil degranulation. In addition, an antiviral effect of sensitization, which we previously showed to be eosinophil mediated, was again seen. This was prevented by depletion of CD8⁺ Tcells. Thus, CD8⁺ T cells play a role in airway hyperreactivity and M₂R dysfunction of sensitized virus-infected guinea pigs by mediating eosinophil degranulation near airway nerves. In contrast, CD8⁺ T cells are not necessary for virus-induced hyperreactivity and M₂R dysfunction in nonsensitized guinea pigs.

　

Key Words: asthma • eosinophils • muscarinic receptors • T lymphocytes • viral immunity

Viral infections are a significant cause of asthma exacerbations in both adults and children (1, 2). These infections also have an important role in exacerbations of chronic bronchitis (3, 4). Virus-induced airway hyperreactivity is vagally mediated (5, 6). Under normal conditions, the parasympathetic nerves release acetylcholine onto M₃ muscarinic receptors on airway smooth muscle, causing bronchoconstriction (7, 8). Release of acetylcholine is limited by inhibitory M₂ muscarinic receptors on the nerves (9). Viral infections cause loss of M₂ muscarinic receptor function, increasing acetylcholine release and potentiating vagally induced bronchoconstriction (10).

Multiple mechanisms participate in virus-induced M₂ receptor dysfunction. We have shown that, in nonsensitized guinea pigs, virus-induced hyperreactivity and M₂ receptor dysfunction are mediated by inflammatory cells other than eosinophils (11). Interferon- is produced by lymphocytes, including CD8⁺ T cells, in virus-infected airways (12, 13) and its increased expression has been noted in the CD8⁺ T cells of patients with asthma (14). This cytokine downregulates the expression of the M₂ receptor gene in airway parasympathetic neurons (15). Type I interferons ( and ß), may also participate in virus-induced M₂ receptor dysfunction and airway hyperreactivity (16). Thus, interferons are likely to be important mediators of M₂ receptor dysfunction and airway hyperreactivity in normal (nonsensitized) animals.

In contrast, we have shown that if guinea pigs are sensitized to a nonviral protein (ovalbumin) before viral infection, the development of airway hyperreactivity switches to an eosinophil-mediated pathway (11). In sensitized virus-infected animals, M₂ receptor dysfunction and associated hyperreactivity are mediated by release of eosinophil major basic protein (MBP), which binds to M₂ receptors (17), blocking their function. This switch to an eosinophil-dependent mechanism may be relevant to the response to viral infections in humans with asthma, who are frequently atopic (18).

Although it is clear that CD8⁺ T lymphocytes are activated by viral infections, and that they participate in the immune response, their roles in airway hyperreactivity and M₂ receptor dysfunction are not known. CD8⁺ T cells are present in increased numbers in the airways of patients with both atopic and nonatopic asthma (19), and in those who have died of acute asthma attacks (20). In mice sensitized to a nonviral protein, ovalbumin, virus-specific CD8⁺ T cells release interleukin (IL)-4 and IL-5 in response to viral antigen (21), leading to an eosinophilic response in the lung. The experiments described in this article were designed to test the role of the CD8⁺ T lymphocyte in virus-induced M₂ receptor dysfunction and airway hyperreactivity, and whether the CD8⁺ T lymphocyte is involved in the switch to an eosinophil-mediated pathway in ovalbumin-sensitized guinea pigs.

	METHODS

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
Sensitization and Viral Infection
Female Dunkin-Hartley guinea pigs were sensitized to ovalbumin by intraperitoneal injection as previously described (11). Parainfluenza type 1 (Sendai virus) was grown in rhesus monkey kidney cells (11), and animals were infected intranasally as previously described (11). Infections were done 21 days after the start of sensitization.

Depletion of CD8⁺ Cells
CD8⁺ cells were depleted with a monoclonal anti-guinea pig CD8 IgG₁ (AbCD8, 125 µl) administered intraperitoneally 3 days before infection (CT6; Biosource International, Camarillo, CA). Depletion of CD8⁺ T cells was confirmed by flow cytometry. Whole blood from each animal was incubated for 30 minutes on ice with various primary antibodies. A saturating concentration of AbCD8 (diluted 1:1,000; Biosource International) was used to detect CD8⁺ T cells. Mouse IgG₁ served as the negative control whereas the positive controls consisted of saturating concentrations of mouse anti-human very late activation antigen-4 (VLA₄) IgG₁ (HP1/2, diluted 1:100) (22) and mouse anti-guinea pig CD4 IgG₁ (CT7, diluted 1:1,000; Biosource International). After being washed with phosphate-buffered saline containing 1% bovine serum albumin, the aliquots were incubated in the dark on ice for 30 minutes with a 1:50 dilution of fluorescein isothiocyanate-labeled polyclonal F(ab')₂ goat anti-mouse IgG (Biosource International). Erythrocytes were lysed, and the leukocytes were suspended in 1% paraformaldehyde in phosphate-buffered saline. Fluorescein isothiocyanate-labeled cells were detected with a FACSCalibur flow cytometer with an excitation wavelength of 488 nm and CellQuest software (BD Biosciences Immunocytometry Systems, San Jose, CA).

Physiologic Studies
Physiologic experiments were done 4 days after viral infection. Anesthetized animals were paralyzed with succinylcholine (10 µg/kg per minute, intravenous) and ventilated via a tracheal cannula (tidal volume, 1 ml/100 g body weight; 100 breaths/minute; Harvard Apparatus, South Natick, MA). Pulmonary inflation pressure was recorded as a measure of bronchoconstriction (11). Catecholamines were depleted with guanethidine (10 mg · kg^-1, intraperitoneal).

To determine the response to vagal stimulation, the cut nerves were stimulated electrically (2.0–25.0 Hz for 5 seconds, 0.1-millisecond pulse duration, 10 V) every 2 minutes. At the end of each experiment, atropine (1 mg · kg^-1, intravenous) was given to confirm that bronchoconstrictions were due to the release of acetylcholine. M₂ receptor function was determined by the ability of cumulative doses of the muscarinic agonist pilocarpine (1–100 µg · kg^-1, intravenous) to inhibit vagally mediated bronchoconstriction. In these studies, the vagus nerves were stimulated at 1-minute intervals (2 Hz, 0.2 millisecond, 2.5–30 V, 44 pulses per train). The voltage was chosen at the beginning of each experiment (mean, 9.11 ± 0.92 V) to give an increase in pulmonary inflation pressure of 20 mm H₂O. Inhibition of vagally induced bronchoconstriction by pilocarpine demonstrates a functional M₂ muscarinic receptor.

The function of M₃ muscarinic receptors in airway smooth muscle was tested by measuring bronchoconstriction in response to increasing doses of acetylcholine (1–10 µg · kg^-1, intravenous).

Lung Lavage
At the end of each experiment, lungs were lavaged five times with 10-ml aliquots of warm phosphate-buffered saline. Cells were counted and differentials were obtained using cytospun slides stained with Diff-Quik (American Scientific Products, McGaw, IL).

Histology
Lungs were inflated and fixed with 3.7% formaldehyde in 0.1 M phosphate buffer. After 24 hours, lungs were embedded in paraffin for sectioning (6.0-µm sections).

Identification of Airway Nerves and Eosinophils
Staining with a polyclonal rabbit antibody to protein gene product 9.5 (Biogenesis, Sandtown, NH) was used to identify airway nerves as previously described (23). Eosinophils were subsequently stained with 1% chromotrope 2R for 30 minutes.

The number of eosinophils and their relation to airway nerves were determined in cartilaginous bronchi. To ensure that the eosinophils counted were from similar-sized airways and from similar-sized areas of examination, all airways used for analysis were measured for circumference and area by an image analysis program (Image-Pro Plus; Media Cybernetics, Silver Spring, MD). Eosinophils within the submucosa (from the inner border of the smooth muscle to the lung adventitia) (24) were counted in terms of their relation to stained nerves because of the interest in efferent, not afferent, parasympathetic nerves. Eosinophils were considered to be related to a nerve if they were within 8 µm of the nerve. The total eosinophil count included all eosinophils from both the mucosa (from the epithelium to the inner border of the smooth muscle [24]) and the submucosa. Both the total number of eosinophils and the number of those specifically related to nerves was determined from an averaged count of 10 high-power fields for each airway examined.

Virus Isolation and Titration
Viral content was determined by infecting rhesus monkey kidney cells with dilutions of lung homogenates as previously described (11). Viral content was determined as the amount of lung homogenate required to produce infection in 50% of rhesus monkey kidney monolayers (the TCID₅₀).

Statistical Analysis
Acetylcholine, frequency, and pilocarpine responses were analyzed by two-way analyses of variance for repeated measures. Baseline heart rates, blood pressures, weights, pulmonary inflation pressures, histologic measurements, and bronchoalveolar lavage were evaluated by analysis of variance.

	RESULTS

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
CD8⁺ T Cell Depletion
In all guinea pigs treated with AbCD8, flow cytometry data confirmed circulating CD8⁺ T cell depletion  . Before intraperitoneal injection with the antibody to CD8, incubation of blood lymphocytes with the antibody to CD8 confirmed the presence of CD8⁺ T cells. Animals treated with CD8 antibody were depleted of circulating CD8⁺ T cells 2 days after pretreatment, which was the day before viral infection. Circulating CD8⁺ T cells remained depleted until at least Day 7, the day of in vivo airway response measurements. In addition, treatment with AbCD8 did not deplete circulating CD4⁺ T cells as determine by flow cytometry (data not shown).

fig.ommitted Figure 1. Flow cytometry of guinea pigs treated with an antibody to CD8 (AbCD8) confirms CD8+ T cell depletion. Blood lymphocytes from an untreated animal (top) and from an animal 3 days after treatment with AbCD8 (bottom) were incubated with IgG1, a negative control antibody (gray line), or with AbCD8 (black line). Log fluorescence intensity (LFI) on the x axis.

 
Baseline Physiologic Measurements
There were no differences between the groups for baseline heart rate, systolic blood pressure, diastolic blood pressure, or animal weight (data not shown). Sensitization of pathogen-free guinea pigs did not alter baseline pulmonary inflation pressure (80.0 ± 3.1 mm H₂O, n = 7) compared with nonsensitized controls (76.0 ± 3.4 mm H₂O, n = 10). Regardless of sensitization, viral infection increased baseline pulmonary inflation pressure compared with respective controls (nonsensitized: 130.0 ± 6.0 mm H₂O, n = 14, p < 0.0001; sensitized: 114.2 ± 6.8 mm H₂O, n = 12, p = 0.0014). This increased baseline was not prevented by pretreatment with antibody to CD8 (nonsensitized: 120.0 ± 7.6 mm H₂O, n = 15; sensitized: 100.9 ± 7.4 mm H₂O, n = 11). Treatment with antibody to CD8 in uninfected animals did not affect baseline pulmonary inflation pressures.

Response to Vagal Stimulation
Vagal stimulation (1–25 Hz) caused frequency-dependent bronchoconstriction that was unaffected by sensitization  . Viral infection increased vagally induced bronchoconstriction in both sensitized and nonsensitized animals . CD8⁺ T cell depletion prevented virus-induced vagal hyperreactivity in guinea pigs sensitized to ovalbumin, but not in nonsensitized animals . Depletion of CD8⁺ T cells had no effect on baseline vagal reactivity in either sensitized or nonsensitized groups (data not shown). Acetylcholine-induced bronchoconstriction (2–10 µg · kg^–1, intravenous) was not affected by sensitization, viral infection, or antibody to CD8 (peak values [mm H₂O]: control, 188 ± 31; sensitized, 145 ± 35; virus, 191 ± 23; sensitized/virus, 183 ± 28; CD8 depleted sensitized, 220 ± 37; CD8 depleted virus, 216 ± 26; CD8 depleted sensitized/virus, 201 ± 31). ED₅₀ (the dose producing 50% of the maximal response) for acetylcholine was likewise unaffected (data not shown).

fig.ommitted   Figure 2. Viral infection causes vagal hyperreactivity, which can be prevented in sensitized virus-infected guinea pigs. Vagal stimulation produced frequency-dependent bronchoconstriction measured as an increase in pulmonary inflation pressure. Viral infection significantly potentiated this in both sensitized (closed circles, n = 5) and nonsensitized (open circles, n = 6) animals compared with their respective controls (open squares, n = 4 and closed squares, n = 5) (p < 0.0001). Pretreatment of nonsensitized animals with AbCD8 did not prevent virus-induced vagal hyperreactivity (open triangles, n = 5). However, depletion of CD8+ T cells decreased reactivity (closed triangles, n = 4, p = 0.004) to a level similar to that of sensitized controls. Results are expressed as the mean increase in pulmonary inflation pressure (mm H2O) ± SEM.

 
M₂ Muscarinic Receptor Function
Pilocarpine inhibited vagally induced bronchoconstriction in uninfected guinea pigs, demonstrating that there are functional M₂ muscarinic receptors on the parasympathetic nerves  . This was unaffected by sensitization. In virus-infected animals, irrespective of sensitization, pilocarpine no longer inhibited vagally induced bronchoconstriction, demonstrating M₂ receptor dysfunction .

fig.ommitted Figure 3. Viral infection causes M2 muscarinic receptor dysfunction, which is prevented in sensitized virus-infected guinea pigs. Pilocarpine (1–100 µg · kg-1, intravenous) inhibits vagally induced bronchoconstriction in pathogen-free guinea pigs regardless of whether they are sensitized (closed squares, n = 4) or not (open squares, n = 4). In contrast, pilocarpine does not inhibit vagally induced bronchoconstriction in virus-infected animals whether they are sensitized (closed circles, n = 7) or not (open circles, n = 4). In sensitized guinea pigs, AbCD8 given before viral infection protected the ability of pilocarpine to inhibit vagally induced bronchoconstriction (closed triangles, n = 7; p = 0.0006). In contrast, in virus-infected animals pretreated with AbCD8 before infection, pilocarpine did not inhibit vagally induced bronchoconstriction (open triangles, n = 8). Results are expressed as the ratio of vagally induced bronchoconstriction in the presence of pilocarpine to the response of vagal stimulation in the absence of pilocarpine. There was a significant difference between the pilocarpine dose–response curves in noninfected versus virus-infected guinea pigs (p < 0.0001). Each point is the mean of at least four animals with the standard error of the mean shown by vertical bars.

 
CD8⁺ depletion prevented virus-induced M₂ receptor dysfunction in virus-infected guinea pigs sensitized to ovalbumin but not in nonsensitized guinea pigs . CD8 depletion did not affect baseline M₂ muscarinic receptor function in either the sensitized or nonsensitized groups (data not shown).

Bronchoalveolar Lavage
Total cells, including neutrophils and macrophages in lung lavage, were increased by viral infection regardless of whether the animals were sensitized or not  . Lymphocyte numbers were not significantly affected by either sensitization or viral infection. Sensitization significantly increased lavage eosinophils compared with nonsensitized controls. This increase with sensitization was not potentiated further by viral infection.

fig.ommitted Figure 4. Airway leukocyte populations measured in bronchoalveolar lavage. Viral infection increased total cells (T), macrophages (M), and neutrophils (N) in both nonsensitized and sensitized animals. Sensitization itself increased eosinophils (E); this was not affected by viral infections. None of the treatments affected lymphocyte numbers (L). AbCD8 did not alter leukocyte numbers in uninfected animals, regardless of sensitization. In contrast, AbCD8 decreased total leukocytes, macrophages, and neutrophils in virus-infected animals, regardless of sensitization. AbCD8 increased the number of eosinophils in nonsensitized virus-infected animals. In contrast, pretreatment of sensitized virus-infected animals with AbCD8 decreased the number of eosinophils compared with sensitized virus-infected animals not pretreated with AbCD8.

 
AbCD8 decreased total cells, macrophages, and neutrophils in the lavage fluid of virus-infected animals regardless of whether the animals were sensitized or not . AbCD8 increased the number of eosinophils in the nonsensitized virus-infected animals, but decreased the number of eosinophils in the sensitized virus-infected animals. AbCD8 did not alter leukocyte numbers in uninfected animals, regardless of sensitization. AbCD8 did not significantly alter the number of lymphocytes recovered in the lavage fluid of any of the groups of animals.

Eosinophil Number and Relation to Nerves by Airway Histology
In contrast to the findings in lavage fluid, histologic studies showed that viral infection of both nonsensitized and sensitized guinea pigs markedly decreased the number of airway eosinophils compared with respective nonsensitized and sensitized controls  . CD8⁺ T cell depletion prevented this virus-induced depletion of eosinophils in both nonsensitized and sensitized guinea pigs compared with respective nonsensitized and sensitized virus-infected controls. Pretreatment with antibody to CD8 did not affect baseline total eosinophil numbers in either the sensitized or nonsensitized groups (data not shown).

fig.ommitted Figure 5. The decrease in tissue eosinophil number and eosinophils related to nerves seen after viral infection is prevented by pretreatment with AbCD8. Total eosinophils (left): Airway histology sections were stained with chromotrope 2R (red) to quantitate the average number of eosinophils per high-power field (HPF) over the entire thickness of the airway. Viral infection of both nonsensitized (Virus column, n = 29) and sensitized (Sens/virus column, n = 53) guinea pigs caused a large decrease in the number of airway eosinophils (p < 0.0001) as compared with nonsensitized (Control column, n = 29) and sensitized (Sensitized column, n = 29) animals. AbCD8 prevented this virus-induced depletion of eosinophils in both nonsensitized (n = 24; data not shown) and sensitized (AbCD8/Sens/virus column, n = 28) guinea pigs (p < 0.001). Eosinophils related to nerves (right): Airway sections were also counterstained with protein gene product 9.5 (stains nerves black) to localize airway nerves. The number of eosinophils in close relation to the airway nerves was determined. Sensitized guinea pigs had an increased number of eosinophils in close relation to airway nerves (Sensitized column, n = 22) as compared with nonsensitized animals (Control column, n = 22; p < 0.0001). Viral infection of both nonsensitized (Virus column, n = 29) and sensitized guinea pigs (Sens/virus column, n = 53) caused a large decrease in the number of airway eosinophils related to nerves. AbCD8 prevented virus-induced depletion of eosinophils related to nerves in both nonsensitized (n = 24; data not shown) and sensitized (AbCD8/Sens/virus column, n = 28) guinea pigs. Each column represents the mean number of airways observed for each group of animals. At least five animals per group were included and the standard error of the mean is shown by vertical bars.

 
Sensitized guinea pigs had an increased number of eosinophils in close relation to airway nerves as compared with nonsensitized animals . Viral infection of both nonsensitized and sensitized guinea pigs markedly decreased the number of airway eosinophils related to nerves. Pretreatment with AbCD8 prevented virus-induced depletion of eosinophils related to nerves in both nonsensitized and sensitized guinea pigs. Pretreatment with antibody to CD8 did not affect baseline eosinophil numbers in relation to nerves in either the sensitized or nonsensitized groups (data not shown).

Quantification of Viral infection
All guinea pigs that were exposed to parainfluenza became infected  . As we have previously shown (11), the viral content of the lungs was significantly decreased in sensitized guinea pigs compared with nonsensitized virus-infected guinea pigs. Pretreatment with AbCD8 prevented this antiviral effect of sensitization. In contrast, in nonsensitized animals antibody to CD8 decreased viral content.

fig.ommitted Figure 6. Effects of sensitization and CD8+ T cell depletion on viral titers. Viral titers from the lungs of all virus-exposed guinea pigs were quantified. Sensitized virus-infected guinea pigs (Sensitized column; n = 14) had a large decrease in viral titer compared with nonsensitized virus-infected guinea pigs (Unsensitized column; n = 12; p = 0.0027). Pretreatment with AbCD8 caused a significant decrease in the viral titer of nonsensitized virus-infected guinea pigs (Unsensitized/AbCD8 column; n = 16), but caused a significant increase in the viral titers of sensitized virus-infected animals (Sensitized/AbCD8 column; n = 9; p < 0.0001). Each column represents the mean of at least five animals, with the standard error of the mean shown by vertical bars.

 

	DISCUSSION

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
These experiments demonstrate that CD8⁺ T lymphocytes are involved in the virus-induced M₂ receptor dysfunction and associated hyperreactivity of ovalbumin-sensitized guinea pigs, but not of nonsensitized guinea pigs. Viral infection causes M₂ receptor dysfunction and airway hyperreactivity whether the animal is sensitized or not. However, in animals sensitized before viral infection, CD8⁺ T cell depletion prevents M₂ receptor dysfunction and airway hyperreactivity.

We have previously shown that in ovalbumin-sensitized animals, virus-induced hyperreactivity and loss of M₂ receptor function are mediated by eosinophils (11). Eosinophil major basic protein (MBP) is an M₂ receptor antagonist (17), and an antibody to MBP prevents loss of M₂ receptor function and associated hyperreactivity in sensitized virus-infected animals (11). Further, heparin, which binds and neutralizes MBP, reverses M₂ receptor dysfunction in the sensitized virus-infected animals (11). Likewise, depleting eosinophils with an antibody to IL-5 prevents virus-induced hyperreactivity and M₂ receptor dysfunction in sensitized animals (11). Thus, in guinea pigs sensitized before viral infection, MBP release from eosinophils causes M₂ receptor dysfunction and vagal hyperreactivity. In contrast, M₂ receptor dysfunction and airway hyperreactivity after viral infection in nonsensitized animals are not eosinophil mediated. None of the treatments that reverse or prevent virus-induced airway hyperreactivity and M₂ receptor dysfunction in sensitized animals works in nonsensitized animals (11).

The reason for this difference may be the unique relation of eosinophils to the airway nerves of sensitized animals, and the response of these eosinophils to viral infection. Eosinophils are clustered around the airway nerves in both antigen-challenged guinea pigs and in humans dying of asthma (23). The current study shows that sensitization alone, without antigen challenge, increases the number of eosinophils in the airway wall and specifically around the airway nerves . Viral infection dramatically decreases the number of eosinophils in both sensitized and nonsensitized airways. As we have discussed above, after viral infection, MBP released from the eosinophil mediates M₂ receptor dysfunction and airway hyperreactivity in sensitized animals (11). Thus, it is plausible that the loss of visible eosinophils in the airway wall is due to the eosinophil's response to viral infection by degranulation via cytolysis. Degranulation via cytolysis involves eosinophil rupture and exudation of cellular contents, making eosinophils difficult to stain and recognize histologically. Although viral infection appears to degranulate eosinophils in both sensitized and nonsensitized animals, because the eosinophils of sensitized animals are clustered around the airway nerves, the MBP that is released will be in a unique position to bind to the M₂ receptors on the nerves. Depletion of CD8⁺ T cells prevented this decrease in eosinophil number after viral infection in both sensitized and nonsensitized animals . In only the sensitized animals did depletion of CD8⁺ T cells prevent virus-induced M₂ receptor dysfunction and hyperreactivity and . It is in these sensitized animals, where eosinophils are clustered around the nerves , that eosinophil activation and degranulation mediate airway hyperreactivity via M₂ receptor dysfunction.

The mechanism by which CD8⁺ T cells could mediate virus-induced eosinophil degranulation is not known. However, CD8⁺ T cells are responsible for a variety of cytokines capable of activating eosinophils. CD8⁺ T cells produce interferon- in response to viral infection (12, 13), which has been shown to directly induce eosinophil degranulation (25, 26). Similarly, tumor necrosis factor- and granulocyte-macrophage colony-stimulating factor, both produced by CD8⁺ T cells (27, 28), induce eosinophil degranulation (29). In addition to cytokines, T cells may also be able to regulate eosinophil degranulation via cell-to-cell contact. It has been shown that T cells can bind to eosinophils via VLA₄ and CD18 and cause eosinophil activation (30). Eosinophils also appear to have a role in antigen presentation to lymphocytes (31), and can specifically present rhinovirus antigens to T cells, activating interferon- release (32). This interaction stimulates increased expression of eosinophil CD18, which could lead to increased eosinophil–T cell cross-talk. Thus, viral infection under appropriate conditions could lead to a bidirectional activation of both cells.

The effects of CD8⁺ T cells on eosinophils have also been shown to be important in other models of virus-induced airway hyperreactivity. Depleting CD8⁺ T cells prevents respiratory syncytial virus-induced airway hyperreactivity and lung eosinophilia in BALB/c mice (33). Administering IL-5 to these mice restores eosinophilia and airway hyperreactivity. These investigators suggested that CD8⁺ T cell production of IL-5 is required for virus-induced airway hyperreactivity. In atopic humans, CD8⁺ T cells have been shown to be as good as CD4⁺ T cells in producing IL-5 (34).

Whereas the mechanism of virus-induced M₂ receptor dysfunction and hyperreactivity in sensitized guinea pigs is eosinophil mediated, the mechanism in nonsensitized animals is still unclear. Unlike the BALB/c mouse model, IL-5 depletion in virus-infected guinea pigs did not prevent virus-induced hyperreactivity (11). If all inflammatory cells are depleted with cyclophosphamide virus-induced hyperreactivity can be prevented in nonsensitized guinea pigs, suggesting that an inflammatory cell other than the eosinophil may play a role (35). Interferon- decreases M₂ receptor expression in cultured airway parasympathetic nerves (15). Interferon- would still be produced by other lymphocytes despite depletion of CD8⁺ T cells in the nonsensitized virus-infected guinea pigs. Thus, if interferon- is the key mediator of virus-induced M₂ receptor dysfunction in nonsensitized animals, it would not be surprising that CD8⁺ T cell depletion alone cannot prevent virus-induced airway hyperreactivity in these guinea pigs.

CD8⁺ T cell depletion also had an effect on viral titers recovered from lungs, depending on the sensitization status of the animal. Depleting CD8⁺ T cells increased viral titers in sensitized animals and decreased viral titers in nonsensitized animals . This effect may be related to the antiviral properties of eosinophils. We previously showed that viral titers in the lungs of sensitized guinea pigs were decreased by an order of magnitude compared with those in nonsensitized animals (11). Depleting eosinophils reversed this antiviral effect. The antiviral effect of eosinophils in sensitized animals was seen again in our current study . Whereas CD8⁺ T cell depletion decreased the number of eosinophils in the lung lavage of sensitized animals , CD8⁺ T cell depletion increased airway eosinophils in the lung lavage of nonsensitized animals. Thus, the antiviral effect of CD8⁺ T cell depletion is related to the effect of CD8⁺ T cell depletion on eosinophil number in the lung lavage.

Other groups have shown an antiviral role for eosinophils. Klebanoff and Coombs found that eosinophils, and particularly the production of hypobromous acid by eosinophil peroxidase, could inhibit HIV infection in vitro (36). We also have preliminary data confirming this mechanism with parainfluenza virus (37). Rosenberg and Domachowske showed an antiviral effect of eosinophil-derived neurotoxin and eosinophil cationic protein, both of which are RNases (38). As the primary site of viral infection and replication is the airway epithelial cell, it is likely that eosinophils in the airways and in the epithelial layers are responsible for this antiviral role, rather than eosinophils in the vicinity of the nerves. Thus, the antiviral effects of eosinophils are distinct from their ability to inhibit M₂ receptor function via MBP release.

We suggest that virus-induced asthma exacerbation involves both the presence of eosinophils in the airways and their activation by CD8⁺ T cells. Guinea pigs sensitized to ovalbumin have an increased number of eosinophils and CD8⁺ T cells in their airways before developing postchallenge airway hyperreactivity (39). Similarly, individuals with asymptomatic asthma have increased airway eosinophils and lymphocytes (40), including CD8⁺ T cells (41), compared with nonasthmatic control subjects. O'Sullivan and coworkers studied the relation of asthma deaths and viral infection (42). The total number of CD8⁺ T cells in the airway tissue was increased, and CD8⁺ T cells were producing increased amounts of IL-4 compared with interferon-. This production of IL-4 suggests that the CD8⁺ T cell may participate in the development of airway eosinophilia.

During asthma attacks, airway eosinophils are activated (43, 44). There is growing evidence of an interaction between eosinophils and viruses. In vitro, eosinophils can interact with viruses as antigen-presenting cells (32), with signs of direct activation (45) and degranulation (46). Virus-induced eosinophil degranulation has also been demonstrated in humans (47, 48). After intranasal infection with rhinovirus, biopsies of the lower airways of asthmatic individuals contain increased numbers of eosinophils (49). Rhinoviral infection in atopic hosts causes increased airway hyperreactivity compared with nonatopic hosts (50). Such hyperreactivity correlates with increased eosinophil cationic protein levels (51).

M₂ receptors are present in humans (52) and M₂ muscarinic receptor function is lost in some patients with asthma (53, 54). We believe viral infection induces eosinophil degranulation via an effect of CD8⁺ T cells. The subsequent release of MBP onto the airway nerves causes M₂ receptor dysfunction and vagal hyperreactivity. Although depleting CD8⁺ T cells prevents virus-induced eosinophil degranulation in both sensitized and nonsensitized animals, M₂ receptor dysfunction and hyperreactivity are prevented only in the sensitized animals. This is likely due to the clustering of eosinophils around the nerves in the airways of sensitized animals , a phenomenon also seen in patients with asthma (23). Thus, viral infection in sensitized animals causes release of eosinophil MBP onto the M₂ receptors of the airway nerves, inducing airway hyperreactivity. Furthermore, there is an eosinophil-mediated antiviral effect of sensitization, which is reversed by CD8⁺ T cell depletion.

	Acknowledgments

 
The authors thank Sherry Hudson for technical assistance with flow cytometry.

	REFERENCES

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 

1. Nicholson KG, Kent J, Ireland DC. Respiratory viruses and exacerbations of asthma in adults. BMJ 1993;307:982–986.

2. Johnston SL, Pattemore PK, Sanderson G, Smith S, Lampe F, Josephs L, Symington P, O'Toole S, Myint SH, Tyrrell DA. Community study of role of viral infections in exacerbations of asthma in 9–11 year old children. BMJ 1995;310:1225–1229.

3. Greenberg SB, Allen M, Wilson J, Atmar RL. Respiratory viral infections in adults with and without chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2000;162:167–173.

4. Hogg JC. Chronic bronchitis: the role of viruses. Semin Respir Infect 2000;15:32–40.

5. Empey DW, Laitinen LA, Jacobs L, Gold WM, Nadel JA. Mechanisms of bronchial hyperreactivity in normal subjects after upper respiratory tract infection. Am Rev Respir Dis 1976;113:131–139.

6. Aquilina AT, Hall WJ, Douglas RG Jr, Utell MJ. Airway reactivity in subjects with viral upper respiratory tract infections: the effects of exercise and cold air. Am Rev Respir Dis 1980;122:3–10.

7. Roffel AF, Elzinga CR, Zaagsma J. Muscarinic M3 receptors mediate contraction of human central and peripheral airway smooth muscle. Pulm Pharmacol 1990;3:47–51.

8. Howell RE, Laemont K, Gaudette R, Raynor M, Warner A, Noronha-Blob L. Characterization of the airway smooth muscle muscarinic receptor in vivo. Eur J Pharmacol 1991;197:109–112.

9. Fryer AD, Maclagan J. Muscarinic inhibitory receptors in pulmonary parasympathetic nerves in the guinea-pig. Br J Pharmacol 1984;83:73–978.

10. Fryer AD, Jacoby DB. Parainfluenza virus infection damages inhibitory M₂ muscarinic receptors on pulmonary parasympathetic nerves in the guinea-pig. Br J Pharmacol 1991;102:267–271.

11. Adamko DJ, Yost BL, Gleich GJ, Fryer AD, Jacoby DB. Ovalbumin sensitization changes the inflammatory response to subsequent parainfluenza infection: eosinophils mediate airway hyperresponsiveness, M₂ muscarinic receptor dysfunction, and antiviral effects. J Exp Med 1999;190:1465–1478.

12. Mbawuike IN, Wells J, Byrd R, Cron SG, Glezen WP, Piedra PA. HLA-restricted CD8⁺ cytotoxic T lymphocyte, interferon-, and interleukin-4 responses to respiratory syncytial virus infection in infants and children. J Infect Dis 2001;183:687–696.

13. Seder RA, Le Gros GG. The functional role of CD8⁺ T helper type 2 cells. J Exp Med 1995;181:5–7.

14. Magnan AO, Mely LG, Camilla CA, Badie MM, Montero-Julian FA, Guillot CM, Casano BB, Prato SJ, Fert V, Bongrand P, et al. Assessment of the Th1/Th2 paradigm in whole blood in atopy and asthma. Increased IFN--producing CD8⁺ T cells in asthma. Am J Respir Crit Care Med 2000;161:1790–1796.

15. Jacoby DB, Xiao HQ, Lee NH, Chan-Li Y, Fryer AD. Virus- and interferon-induced loss of inhibitory M₂ muscarinic receptor function and gene expression in cultured airway parasympathetic neurons. J Clin Invest 1998;102:242–248.

16. Bowerfind WM, Fryer AD, Jacoby DB. Double-stranded RNA causes airway hyperreactivity and neuronal M₂ muscarinic receptor dysfunction. J Appl Physiol 2002;92:1417–1422.

17. Jacoby DB, Gleich GJ, Fryer AD. Human eosinophil major basic protein is an endogenous allosteric antagonist at the inhibitory muscarinic M₂ receptor. J Clin Invest 1993;91:1314–1318.

18. Kalliel JN, Goldstein BM, Braman SS, Settipane GA. High frequency of atopic asthma in a pulmonary clinic population. Chest 1989;96:1336–1340.

19. Shiota Y, Arikita H, Horita N, Hiyama J, Ono T, Yamakido M. Intracellular IL-5 and T-lymphocyte subsets in atopic and nonatopic bronchial asthma. J Allergy Clin Immunol 2002;109:294–298.

20. O'Sullivan S, Cormican L, Faul JL, Ichinohe S, Johnston SL, Burke CM, Poulter LW. Activated, cytotoxic CD8⁺ T lymphocytes contribute to the pathology of asthma death. Am J Respir Crit Care Med 2001;164:560–564.

21. Coyle AJ, Erard F, Bertrand C, Walti S, Pircher H, Le Gros G. Virus-specific CD8⁺ cells can switch to interleukin 5 production and induce airway eosinophilia. J Exp Med 1995;181:1229–1233.

22. Costello RW, Jacoby DB, Fryer AD. Pulmonary neuronal M₂ muscarinic receptor function in asthma and animal models of hyperreactivity. Thorax 1998;53:613–616.

23. Costello RW, Schofield BH, Kephart GM, Gleich GJ, Jacoby DB, Fryer AD. Localization of eosinophils to airway nerves and effect on neuronal M₂ muscarinic receptor function. Am J Physiol 1997;273:L93–L103.

24. Bai A, Eidelman DH, Hogg JC, James AL, Lambert RK, Ludwig MS, Martin J, McDonald DM, Mitzner WA, Okazawa M. Proposed nomenclature for quantifying subdivisions of the bronchial wall. J Appl Physiol 1994;77:1011–1014.

25. Lacy P, Mahmudi-Azer S, Bablitz B, Hagen SC, Velazquez JR, Man SF, Moqbel R. Rapid mobilization of intracellularly stored RANTES in response to interferon- in human eosinophils. Blood 1999;94:23–32.

26. Bandeira-Melo C, Gillard G, Ghiran I, Weller PF. EliCell: a gel-phase dual antibody capture and detection assay to measure cytokine release from eosinophils. J Immunol Methods 2000;244:105–115.

27. Till S, Li B, Durham S, Humbert M, Assoufi B, Huston D, Dickason R, Jeannin P, Kay AB, Corrigan C. Secretion of the eosinophil-active cytokines interleukin-5, granulocyte/macrophage colony-stimulating factor and interleukin-3 by bronchoalveolar lavage CD4⁺ and CD8⁺ T cell lines in atopic asthmatics, and atopic and non-atopic controls. Eur J Immunol 1995;25:2727–2731.

28. Ratner A, Clark WR. Role of TNF- in CD8⁺ cytotoxic T lymphocyte-mediated lysis. J Immunol 1993;150:4303–4314.

29. Horie S, Gleich GJ, Kita H. Cytokines directly induce degranulation and superoxide production from human eosinophils. J Allergy Clin Immunol 1996;98:371–381.

30. Mengelers HJ, Maikoe T, Raaijmakers JA, Lammers JW, Koenderman L. Cognate interaction between human lymphocytes and eosinophils is mediated by ß₂-integrins and very late antigen-4. J Lab Clin Med 1995;126:261–268.

31. Weller PF, Rand TH, Barrett T, Elovic A, Wong DT, Finberg RW. Accessory cell function of human eosinophils: HLA-DR-dependent, MHC- restricted antigen-presentation and IL-1 expression. J Immunol 1993;150:2554–2562.

32. Handzel ZT, Busse WW, Sedgwick JB, Vrtis R, Lee WM, Kelly EA, Gern JE. Eosinophils bind rhinovirus and activate virus-specific T cells. J Immunol 1998;160:1279–1284.

33. Schwarze J, Cieslewicz G, Joetham A, Ikemura T, Hamelmann E, Gelfand EW. CD8 T cells are essential in the development of respiratory syncytial virus-induced lung eosinophilia and airway hyperresponsiveness. J Immunol 1999;162:4207–4211.

34. Walker C, Virchow JC Jr, Bruijnzeel PL, Blaser K. T cell subsets and their soluble products regulate eosinophilia in allergic and nonallergic asthma. J Immunol 1991;146:1829–1835.

35. Fryer AD, Yarkony KA, Jacoby DB. The effect of leukocyte depletion on pulmonary M2 muscarinic receptor function in parainfluenza virus-infected guinea-pigs. Br J Pharmacol 1994;112:588–594.

36. Klebanoff SJ, Coombs RW. Virucidal effect of stimulated eosinophils on human immunodeficiency virus type 1. AIDS Res Hum Retroviruses 1996;12:25–29.

37. Adamko DJ, Yost BL, Gleich GJ, Fryer AD, Jacoby DB. Antiviral activity of eosinophil peroxidase mediated via sodium bromide . Am J Respir Crit Care Med 2000;163:A952.

38. Rosenberg HF, Domachowske JB. Eosinophils, eosinophil ribonucleases, and their role in host defense against respiratory virus pathogens. J Leukoc Biol 2001;70:691–698.

39. Frew AJ, Moqbel R, Azzawi M, Hartnell A, Barkans J, Jeffery PK, Kay AB, Scheper RJ, Varley J, Church MK. T lymphocytes and eosinophils in allergen-induced late-phase asthmatic reactions in the guinea pig. Am Rev Respir Dis 1990;141:407–413.

40. Kirby JG, Hargreave FE, Gleich GJ, O'Byrne PM. Bronchoalveolar cell profiles of asthmatic and nonasthmatic subjects. Am Rev Respir Dis 1987;136:379–383.

41. Amin K, Ludviksdottir D, Janson C, Nettelbladt O, Bjornsson E, Roomans GM, Boman G, Seveus L, Venge P. Inflammation and structural changes in the airways of patients with atopic and nonatopic asthma. BHR Group. Am J Respir Crit Care Med 2000;162:2295–2301.

42. O'Sullivan S, Cormican L, Faul JL, Ichinohe S, Johnston SL, Burke CM, Poulter LW. Activated, cytotoxic CD8⁺ T lymphocytes contribute to the pathology of asthma death. Am J Respir Crit Care Med 2001;164:560–564.

43. Frigas E, Loegering DA, Solley GO, Farrow GM, Gleich GJ. Elevated levels of the eosinophil granule major basic protein in the sputum of patients with bronchial asthma. Mayo Clin Proc 1981;56:345–353.

44. Bousquet J, Chanez P, Lacoste JY, Barneon G, Ghavanian N, Enander I, Venge P, Ahlstedt S, Simony-Lafontaine J, Godard P. Eosinophilic inflammation in asthma. N Engl J Med 1990;323:1033–1039.

45. Kimpen JL, Garofalo R, Welliver RC, Ogra PL. Activation of human eosinophils in vitro by respiratory syncytial virus. Pediatr Res 1992;32:160–164.

46. Olszewska-Pazdrak B, Pazdrak K, Ogra PL, Garofalo RP. Respiratory syncytial virus-infected pulmonary epithelial cells induce eosinophil degranulation by a CD18-mediated mechanism. J Immunol 1998;160:4889–4895.

47. Teran LM, Seminario MC, Shute JK, Papi A, Compton SJ, Low JL, Gleich GJ, Johnston SL. RANTES, macrophage-inhibitory protein 1, and the eosinophil product major basic protein are released into upper respiratory secretions during virus-induced asthma exacerbations in children. J Infect Dis 1999;179:677–681.

48. Oh JW, Lee HB, Yum MK, Kim CR, Kang JO, Park IK. ECP level in nasopharyngeal secretions and serum from children with respiratory virus infections and asthmatic children. Allergy Asthma Proc 2000;21:97–100.

49. Fraenkel DJ, Bardin PG, Sanderson G, Lampe F, Johnston SL, Holgate ST. Lower airways inflammation during rhinovirus colds in normal and in asthmatic subjects. Am J Respir Crit Care Med 1995;151:879–886.

50. Gern JE, Calhoun W, Swenson C, Shen G, Busse WW. Rhinovirus infection preferentially increases lower airway responsiveness in allergic subjects. Am J Respir Crit Care Med 1997;155:1872–1876.

51. Grunberg K, Smits HH, Timmers MC, de Klerk EP, Dolhain RJ, Dick EC, Hiemstra PS. Experimental rhinovirus 16 infection: effects on cell differentials and soluble markers in sputum in asthmatic subjects. Am J Respir Crit Care Med 1997;156:609–616.

52. Minette PA, Barnes PJ. Prejunctional inhibitory muscarinic receptors on cholinergic nerves in human and guinea pig airways. J Appl Physiol 1988;64:2532–2537.

53. Minette PA, Lammers JW, Dixon CM, McCusker MT, Barnes PJ. A muscarinic agonist inhibits reflex bronchoconstriction in normal but not in asthmatic subjects. J Appl Physiol 1989;67:2461–2465.

54. Ayala LE, Ahmed T. Is there loss of protective muscarinic receptor mechanism in asthma? Chest 1989;96:1285–1291.