Respiratory viral infections as promoters of allergic sensitization and asthma in animal models

Influenza infection prevents induction of tolerance to allergens

Holt et al. 20 studied the effects of influenza infection in a model of tolerance induction. If mice were exposed to OVA aerosol prior to intraperitoneal sensitization with OVA and aluminium hydroxide, reduced levels of OVA-specific IgE and increased levels of OVA-specific IgG were detected. This phenomenon was interpreted as tolerance. Influenza infection at the time of OVA aerosol exposure prevented the development of tolerance, resulting in increased levels of OVA-specific IgE following subsequent intraperitoneal sensitization. Parallel findings were recently reported by Tsitoura et al. 21. They assessed the effects of influenza infection in a model of tolerance induction by intranasal application of OVA prior to intraperitoneal sensitization, which in turn was followed by intranasal OVA challenge. Influenza infection at the same time as the initial OVA application prevented tolerance induction, resulting in increased OVA-specific T-cell proliferation, increased OVA-specific IgE levels and increased type 2 T-helper cell (Th) cytokine production. When influenza infection did not coincide with the initial allergen exposure but preceded it by 15 or 30 days, tolerance was still prevented, as demonstrated by the lack of inhibition of OVA-induced T-cell proliferation. However, under these conditions, sensitization was associated with a strong Th1 immune response, with reduced production of interleukins (ILs)-4, -5 and -13 and increased production of interferon gamma (IFN-γ) upon restimulation with antigen. This Th1 response suppressed the production of OVA-specific IgE but resulted in enhanced OVA-specific IgG2a serum levels.

These findings suggest that infection with influenza prevents the development of tolerance and enhances IgE-mediated allergic sensitization if infection and sensitization coincide. If infection precedes primary allergen exposure, nonallergic Th1-driven sensitization is promoted.

Respiratory syncytial virus infection enhances sensitization to allergen

Leibovitz et al. 22 and Freihorst et al. 23 studied the virus primarily implicated in the promotion of allergic sensitization via the airways and asthma development in children. They determined, in a murine model, the impact of infection with RSV on sensitization to inhaled allergens (ragweed and OVA) during acute infection. In both sets of experiments, RSV infection resulted in increased levels of allergen-specific antibodies in serum (IgE and IgG) and bronchoalveolar lavage (BAL) fluid (IgA and IgG). In a recent publication of O'Donnell et al. 24, induction of anaphylactic antibodies by respiratory viral infections was reported. Following infection with influenza or RSV and concomitant sensitization to OVA aerosol, but not after sensitization without infection, a large proportion of mice collapsed upon subsequent cutaneous challenge with the antigen. In mice susceptible to features of anaphylaxis, OVA-specific IgG1 was detected. This class of antibody, which is known to mediate anaphylaxis in mice 25, was not detected in mice which failed to react to cutaneous antigen challenge.

Parainfluenza infection results in increased mucosal permeability

In guinea-pig models of PIV 3 and RSV infection studied by Riedel et al. 26 and Dakhama et al. 27, this ability of respiratory viruses to increase production of allergen-specific antibodies was confirmed. In both sets of studies, sensitization was begun during acute infection, when increased permeability of the respiratory mucosa was demonstrated by Riedel et al.26. Serum levels of horseradish peroxidase were increased following inhalation of horseradish peroxidase aerosol by acutely infected mice compared to noninfected animals.

The investigators suggested that the increased permeability of the airway mucosa may be an important factor in virus-induced enhancement of sensitization via the airways.

These studies and additional data from a rat model of influenza infection 28 and bovine model of RSV infection 29 demonstrate that viral lower respiratory tract infections can enhance allergic sensitization to inhaled antigens and prevent the development of tolerance which may be induced by allergen exposure via the airways in the absence of infection (table 1⇓). Among the mechanisms involved, increased permeability of the airway mucosa to allergens and recruitment of dendritic cells to the airway epithelium during acute infection may be important, facilitating sensitization through increased antigen uptake and more effective antigen presentation. The effects of respiratory viruses on sensitization appear to be dependent on timing, due to changes in the quality of immune responses during the course of infection. As a result, infection with a given respiratory virus could potentially result in either promotion or inhibition of allergic sensitization.

Guinea-pig models

In addition to allergen-specific Igs as markers of sensitization, the consequences of virus-induced allergic airway sensitization have become a major focus of research based on the hypothesis that respiratory virus infections can predispose to the development of asthma. Airway inflammation and airway responsiveness to bronchoprovocation, both features of bronchial asthma, were first monitored in guinea-pig models of virus infection with allergen sensitization 26, 33. Following infection with both PIV 3 and RSV, enhanced sensitization to OVA aerosol was associated with increased airway inflammation, marked eosinophilia and airway hyperresponsiveness (AHR) to allergen challenge and to nonspecific provocation with methacholine 26, 27. Kudlacz and Knippenberg 33, in contrast, did not find any consequences of enhanced sensitization following infection with PIV 3. In their model, PIV 3 infection prior to sensitization to OVA prevented the increases in histamine release, numbers of leukocytes in BAL fluid and methacholine-induced dyspnoea which were observed in noninfected sensitized guinea-pigs. The discrepancy between these observations and the findings reported by Riedel et al. 26 may be due to the fact that Kudlacz and Knippenberg 33 used 10-fold higher amounts of PIV 3 for infection, possibly inducing a long-lasting and significant Th1 response, which may have modified the initial sensitization response. Unfortunately, allergen-specific antibodies, as parameters of sensitization, were not monitored in this study.

Murine models

In mice, enhanced sensitization to aeroallergens following respiratory virus infection is also associated with increased airway inflammation and AHR. Suzuki et al. 18 and Yamamoto et al. 19 sensitized mice with OVA aerosol and aluminium hydroxide during the acute phase of influenza infection and then challenged them with the allergen 2–3 weeks after primary sensitization. In both studies, this resulted in increased OVA-specific IgE levels, increased airway inflammation and AHR compared to sensitization in the absence of infection or infection without sensitization. The composition of inflammatory cells was assessed in BAL fluid. Both groups observed a predominant increase in CD8+ T-cell numbers following sensitization in infected mice. In addition, Yamamoto et al. 19 also reported increases in numbers of CD4+ T-cells, eosinophils and macrophages.

The present authors have investigated a model of RSV infection and subsequent allergic sensitization to OVA via the airways in mice. In order to investigate whether RSV infection can predispose to allergen sensitization and the development of obstructive airway disease through modulation of immune responses, independently of mechanisms associated with the acute infection such as increased mucosal permeability, an approach differing from that of previously established models was chosen. In this approach, sensitization was begun only after the acute phase of the infection or even after resolution of the consequences of acute respiratory virus disease, i.e. on days 11–21 or 21–30 postinfection. Following RSV infection in mice, maximal viral replication occurs on day 4 and maximal inflammatory changes and AHR can be detected on days 6–8. By day 21, postinfection airway inflammation and AHR have completely resolved. Exposure to OVA aerosol alone on 10 consecutive days resulted in sensitization, as indicated by OVA-specific IgE and IgG1, but these mice did not develop altered airway responses to inhaled methacholine. Acute RSV infection prior to sensitization did not affect serum levels of allergen-specific antibodies 30. This is in contrast to the models discussed earlier, in which exposure to allergen was begun during the acute phase of infection and resulted in increased allergen-specific antibody production.

Following RSV infection and subsequent exposure to allergen, airway responsiveness to nonspecific provocation with inhaled methacholine was assessed by barometric body plethysmography, monitoring “enhanced pause”, a calculated value which correlates with airway resistance 34. Twenty-four hours later, lungs and peribronchial lymph nodes (PBLNs), the regional lymph nodes of the lung, were harvested. Lung cells were isolated by collagenase digestion and inflammatory cells identified using light microscopy. Mononuclear cells from PBLNs were cultured and concentrations of the cytokines IFN-γ, IL-4 and IL-5 measured in supernatants by enzyme-linked immunosorbent assay (ELISA). RSV infection but not sham infection or administration of ultraviolet light-inactivated virus prior to sensitization resulted in lung eosinophilia and neutrophilia and the development of AHR to inhaled methacholine. These effects of RSV infection on subsequent exposure to allergen were associated with a decrease in IFN-γ and increase in IL-4 production in cultures of mononuclear cells from PBLNs. IL-5 concentrations did not change significantly 30.

Cytokines in respiratory syncytial virus-induced effects on allergic sensitization

Due to the pivotal role of IL-5 in the development of airway eosinophilia and AHR, both in allergic sensitization 35–37 and acute RSV infection 38, the present authors sought to delineate the role of IL-5 in RSV-induced effects on allergic airway sensitization. In order to do this, both anti-IL-5 treatment and mice genetically deficient in IL-5 were studied. Treatment with anti-IL-5 during the allergen sensitization phase prevented eosinophil influx into the lungs and development of AHR 30. This is in keeping with the effects of anti-IL-5 treatment reported in models of allergen-induced airway inflammation and AHR 35–37. Anti-IL-5 treatment during the RSV infection phase alone significantly reduced lung eosinophilia and AHR following subsequent allergen exposure in the presence of IL-5 39. Mice genetically deficient in IL-5 failed to develop lung eosinophilia and AHR following RSV infection and allergen sensitization. Both were restored by administration of exogenous IL-5 during the acute infection. In contrast, reconstitution with IL-5 during the allergen sensitization phase alone did not restore AHR, despite inducing significant lung eosinophilia 39. These data illustrate that the presence of IL-5, and possibly of eosinophils, during the acute infection phase is critical if not essential to the development of lung eosinophilia and AHR following subsequent allergen sensitization via the airways.

In order to further delineate the mechanisms involved, the roles of IL-4 and IFN-γ were studied in the above model utilizing mice deficient in either of these cytokines. Mice deficient in IL-4 developed neither lung eosinophilia nor AHR following RSV infection and sensitization. The absence of IL-4 was compensated for by administration of IL-5, indicating that availability of IL-4 may be essential for IL-5 production, at least sufficient for the development of RSV-induced effects on airway sensitization 39. IFN-γ did not appear to contribute to the effects of RSV infection on allergen sensitization. Conversely, mice deficient in IFN-γ developed increased lung eosinophilia and AHR, exceeding the effects of RSV infection and sensitization in wild-type controls 39.

T-cells in respiratory syncytial virus-induced effects on allergic sensitization

In order to define the role of T-cells as possible regulatory cells mediating the effects of RSV infection on airway allergen sensitization, two approaches were utilized: 1) depletion of T-cells during RSV infection and sensitization; and 2) adoptive transfer of T-cells which had been isolated from the PBLNs of RSV-infected mice and transferred (intravenously) to naive mice prior to allergen sensitization via the airways. Treatment of mice with anti-CD8 resulted in almost complete depletion of CD8+ T-cells (>96%). This prevented the influx of neutrophils and eosinophils into the lungs and the development of AHR following allergen exposure. Depletion of CD4+ T-cells was also almost complete during RSV infection (>98%), but could not be sustained during the allergen sensitization phase (72% depletion). Although resulting in a reduction in lung eosinophilia and AHR, CD4+ T-cell depletion did not prevent these consequences of RSV infection on allergen sensitization. Adoptive transfer of CD3+ T-cells, harvested from PBLNs 14 days after infection resulted in eosinophil influx into the lungs and development of AHR associated with an increase in IL-5 production in noninfected recipients following allergen sensitization. In contrast, no such effects were observed following transfer of T-cells from noninfected mice or following transfer of T-cells from RSV-infected animals into recipients without subsequent allergen sensitization. Further, the effects of RSV infection on allergen sensitization could also be transferred by CD8+ T-cells isolated from PBLNs. Transfer of isolated CD4+ T-cells, in contrast, did not result in the influx of eosinophils and AHR following allergen sensitization 40.

These observations indicate that T-cells, in particular CD8+ T-cells, are critical in mediating the effects of RSV infection on subsequent exposure of the airways to allergen. It may be speculated that IL-5-producing CD8+ T-cells (Tc2), play a pivotal role in this interaction, in contrast to the cytotoxic IFN-γ-producing CD8+ T-cells (Tc1), which dominate the acute response to viral infection between virus exposure and subsequent exposure to allergen. This former Tc2 subset may expand and become more dominant after the acute phase of the RSV infection has resolved, thus favouring eosinophilic airway inflammation and the development of AHR in response to allergic sensitization via the airways (fig. 1⇓). This hypothesis needs to be tested by direct assessment of cytokine expression and production at the single cell level in T-cell subpopulations over the course of RSV infection and subsequent resolution.

Taken together, studies from a number of laboratories have demonstrated that enhanced sensitization to allergen following respiratory virus infections can be associated with increases in airway inflammation and airway responsiveness. In addition, RSV infection appears capable of enhancing the consequences of subsequent allergen sensitization independently of increasing levels of allergen-specific antibodies. The effects of RSV infection and also of infections with other respiratory viruses on allergic airway sensitization are probably mediated and regulated by T-cells and, under certain conditions, CD8+ T-cells. Interestingly, CD8+ T-cells are also pivotal in the development of AHR in both acute RSV infection 41 and allergic sensitization 42. Further, the presence of IL-5 during acute RSV infection is essential for eosinophil influx into the airways and the development of AHR following subsequent allergen sensitization. In the light of these observations, it appears that RSV infection can prime and induce differentiation of a subset of CD8+ T-cells to produce IL-5 both during acute infection and following subsequent allergen exposure. This results in increased levels of IL-5 in the airways, favouring the development of eosinophilic inflammatory responses. The eosinophils, in turn, appear to be critical effector cells of the development of AHR, both in acute RSV infection and following subsequent allergic sensitization. A corollary of the studies in IL-4-deficient mice is that IL-4 in the lung milieu is essential for this capacity of CD8+ T-cells to differentiate into IL-5 producers.

Increased airway inflammation and airway hyperresponsiveness

Robinson et al. 43 studied the effects of RSV infection in guinea-pigs previously sensitized to OVA via the airways, monitoring airway responsiveness to acetylcholine and airway inflammation. Both sensitization alone and RSV infection alone resulted in AHR and airway inflammation with epithelial necrosis, airway wall oedema, mononuclear and granulocytic infiltrates, bronchoalveolar lymphoid tissue hyperplasia and goblet cell metaplasia. RSV infection in allergic animals further increased AHR and airway inflammation, particularly promoting airway epithelial necrosis. The authors concluded that prior sensitization potentiated the physiological and structural changes associated with RSV infection and posited that an established allergic diathesis may increase the severity of RSV infection in children.

Prolonged airway hyperresponsiveness

Peebles et al. 31 utilized a mouse model of intraperitoneal sensitization to OVA with aluminium hydroxide followed by OVA aerosol challenge to assess the effects of RSV infection during the challenge. Monitoring airway responsiveness to methacholine at weekly intervals, they observed that RSV infection prolonged AHR. Eight days after infection, methacholine provocation resulted in AHR to the same degree in sensitized and challenged mice with or without RSV infection. However, on day 15 after infection, AHR was still present in mice that had been infected with RSV but not in those sensitized and challenged without infection. This prolongation of AHR was associated with an increase in the numbers of lymphocytes in the BAL fluid on day 15. Eosinophils were the predominant cell type in the BAL fluid, but their numbers did not differ between RSV-infected and noninfected animals. In addition, on histological examination, increased alveolar and interstitial inflammation and increased mucus production were noted in mice infected with RSV. Prior allergen sensitization did not impact the virus load as determined by plaque assays from lung homogenates 4 days postinfection.

Recurrent respiratory syncytial virus infection in allergic airways

A study recently reported by Matsuse et al. 32 focused on the effects of recurrent RSV infection in mice sensitized to Dermatophagoides farinae (Df). Following intraperitoneal sensitization (together with aluminium hydroxide), mice were challenged intranasally with the allergen 2, 6 and 14 weeks after sensitization. The day following each 3-day challenge period, mice were infected with RSV or sham-infected. Controls were neither sensitized nor challenged and only infected with RSV or sham-infected. Following each challenge period, airway responsiveness to methacholine was measured using barometric body plethysmography. Primary RSV infection resulted in increased airway responsiveness in nonsensitized mice and lasted for 10 days. In these animals, the duration of AHR was shortened after the second RSV infection and was no longer detectable following tertiary infection. In contrast, RSV infection in mice sensitized to Df resulted in AHR that increased in duration following the secondary infection and was even more pronounced after tertiary infection. The magnitude of airway responsiveness in these mice exceeded the levels measured in animals sensitized without infection or infected without sensitization. The enhancement of AHR in mice sensitized and infected with RSV was associated with increased airway inflammation with bronchial exudates composed of mononuclear cells and eosinophils as well as massive tissue eosinophilia. Further, RSV infection resulted in increased levels of macrophage inflammatory protein 1α (MIP-1α) in the lungs of Df-sensitized mice and in a Th2 cytokine shift with increased production of IL-4 and IL-5 and a decrease in production of IFN-γ by anti-CD3 stimulated thoracic lymph node mononuclear cells. Increased levels of total serum IgE, but no changes in allergen-specific IgE or IgG1, were detectable. Measurements of RSV antigen by ELISA in lung homogenates revealed an increased virus load in sensitized compared to nonsensitized mice following tertiary infection.

These observations suggest that recurrent RSV infections augment the synthesis of Th2-derived cytokines, total IgE and MIP-1α following allergen sensitization, thus inducing persistent and increased airway inflammation and AHR. In the absence of sensitization, the effects of RSV infection were attenuated on subsequent reinfection.

Hence, three studies indicate that there are consequences of respiratory virus infection in previously sensitized mice which lead to enhanced airway inflammation and increased AHR that can persist for long periods of time. Factors involved in this potentiation in allergic hosts could be increased production of chemokines in response to infection, increased recruitment of T-lymphocytes, possibly Th2, resulting in increased production of Th2-derived cytokines such as IL-4 and IL-5. These changes support an inflammatory response dominated by eosinophils as well as the development of persistent AHR (fig. 2⇓). Less effective clearance of virus during an immune response with a Th2 bias may also have contributed to the changes seen. These findings in rodent models suggest that similar immune mechanisms may also be of importance in humans. On the one hand, allergic airway inflammation may, indeed, predispose to more severe obstructive airway disease following respiratory virus infection. On the other hand, respiratory virus infections are likely to play a role in both potentiation and perpetuation of airway inflammation and airway obstruction in allergic asthma.