To the editor:

A recent publication in Nature Immunology by Dahl et al.1 reports that influenza infection can enhance allergic responses in the lung, which in many respects models the clinical situation. This finding has important implications for how we should, or could, make use of the popular 'hygiene hypothesis' in developing plans for preventing allergic diseases.

There is no doubt that in certain clinical situations viral infections can exacerbate asthma symptoms, and the work by Dahl et al. reinforces this. However, the interactions between viral responses and asthma are complex, and indeed, recent epidemiological studies have indicated a possible protective effect of some viral infections against the development of asthma2. We have investigated the significance of how the timing of allergen exposure, in relation to influenza infection, affects the development of allergic airway inflammation. Unlike Dahl et al., we found that influenza infection can have both beneficial and detrimental effects upon the development of allergy.

If sensitized (allergic) mice inhale allergen during the acute stages of influenza infection and shortly afterwards (days 1 to 14 after infection), airway eosinophilia was exacerbated similar to what Dahl et al. have described1. In our model we found that dendritic cell (DC) migration to the draining lymph node, T helper cell type 2 (TH2) recruitment into the lung, TH2 cytokine production and airway hyperresponsiveness were similarly enhanced during this period3.

In contrast to the findings of Dahl and colleagues, we found that at later time points after influenza infection (14 to >100 days after infection), airway eosinophilia, TH2 cytokine production and airways hyperresponsiveness were suppressed4 (and Supplementary Fig. 1 online). The mechanism underlying this suppression was distinct from that mediating the exacerbation. We found that a long-lived population of lung-resident CD8+ T cells could suppress the local (but not systemic) response to allergen inhalation in an interferon-γ-dependent manner4.

It is important to note that although we see an influenza-mediated exacerbation of an allergic response, similar to Dahl et al., the duration of exacerbation in our model is not long lasting, but rather limited to the acute stage of infection and shortly afterwards. Dahl et al. propose that the genetic background of different mouse strains may influence the effect of viral infections on allergic responses1. However, our data shows that the key factor is the timing of allergen challenge with respect to the stage of the viral infection3. It is likely that the genetic background, sex and age of mice may influence the duration and intensity of infection, but these factors are unlikely to restrict a response to either exacerbation or suppression.

These two contrasting effects of influenza infection on asthma highlight the need for rigorous analysis of multiple time points and models before experimental findings are applied to a broader clinical and epidemiological setting.