A recurring theme in Gothic literature is that the human condition is an enigmatic mixture of good and evil. These themes are central in classics such as Robert Louis Stevenson's Dr. Jekyll and Mr. Hyde, in which the protagonist begins as a respected pillar of the community, but his darker side becomes evident as the story unfolds. In this issue of the Journal, Miller and colleagues (pp. 508–516) suggest that IFN-λ1, a prototypic antiviral cytokine, may also have an injurious duality (1).
Previous studies have demonstrated that rhinovirus (HRV) infections are closely associated with exacerbations of childhood asthma. Since HRV infections often cause mild or asymptomatic illnesses, this raises questions as to mechanisms that differentiate mild colds from severe episodes of wheezing and shortness of breath in children with asthma. Viral factors could relate to the great diversity among HRVs, which consist of over 150 different types in three species (A, B, and C). In fact, there is some evidence that infections with HRV-C species viruses may be more likely to cause wheezing and exacerbations of asthma, while B species viruses may be less likely to do so (2, 3). Environmental risk factors for virus-induced wheeze include exposure to pollutants, and for sensitized children, specific allergens. Finally, host factors that increase the risk of wheezing with HRV and other respiratory viruses include low baseline lung function and individual variations in immunologic responses to infection (4–6).
The role of immunologic factors in determining illness severity has been difficult to ascertain, partly due to the difficulties in sampling the lower airway during virus-induced asthma exacerbations. As a result, most studies have instead sampled the upper airway or blood, assuming that these compartments will provide insight into pulmonary infectious and inflammatory mechanisms. This approach has provided insights, yet no definitive answers, about the nature of viral respiratory infections and their role in acute airway obstruction. Generally, the severity of both upper and lower respiratory illnesses is positively associated with inflammatory indicators, including kinins, cytokines (e.g., IL-1, IL-8, CXCL10), and inflammatory cells (mononuclear cells and neutrophils) in airway secretions (7). Because HRV causes relatively little tissue destruction, even in severe colds, it is assumed that these inflammatory responses are the major contributors to the severity of both upper and lower respiratory illnesses. In contrast, most studies have shown that interferons (IFNs) moderate the severity of viral respiratory infections. Type I IFN includes IFN-α and IFN-β, type II interferon is IFN-γ, and type III interferons include IFN-λ1, IFN-λ2, and IFN-λ3 (also known as IL-29, IL-28A, and IL-28B).
To identify mechanisms of virus-induced wheezing, Miller and colleagues conducted a prospective study comparing children with asthma who presented to clinics with an upper respiratory illness (URI) and wheezing (n = 200) to those who presented with a URI without wheezing (n = 209). The protocol included assessments of viral factors (species, quantity of viral shedding) and immunologic responses (cytokine and interferon levels) in nasal washes. Nearly all eligible children participated in the study, and several findings are noteworthy. Viruses were detected in 82% of the children, and HRVs were most often detected. Infections with HRV compared with other viruses were significantly associated with wheezing illnesses, but the species of HRV or quantity of shedding were similar in wheezing illnesses versus URI. Nasal wash cytokine levels in the wheezing versus URI groups were similar, with the interesting exception of selected interferons. Both IFN-λ1 and IFN-α were increased in children with wheezing illnesses; however, only IFN-λ1 was associated with the severity of illness and appeared to mediate the effects of HRV infection on asthma exacerbations. The authors conclude that IFN-λ1 is uniquely associated with exacerbations of asthma in their study, and suggest that IFN-λ1 should be considered as a target for prevention or therapy against HRV-associated asthma exacerbations.
In contrast, results from some (8, 9) but not all (10, 11) previous studies have suggested that deficient production of IFN-β and IFN-λ by virus-infected airway epithelial cells could be a key feature of asthma leading to impaired apoptosis of virus-infected cells, increased viral replication, and more severe airway clinical symptoms. Miller and colleagues also cultured airway epithelial cells derived from nasal brushings, and reported that HRV-induced IFN-λ1 was lower in children with asthma. This represents an interesting duality: HRV-induced IFN-λ1 secretion from epithelial cells in vitro was lower in asthma, but HRV-induced IFN-λ1 in vivo was increased in expression during wheezing illnesses.
The concept that interferons can contribute to illness is familiar to rheumatologists, since chronic IFN-α production has been implicated in the pathogenesis of arthritis and lupus (12). Furthermore, in a mouse model of respiratory viral illness, virus-induced type I interferon production was linked to expression of the high-affinity IgE receptor on lung dendritic cells, which in turn recruit to the airways alternatively activated macrophages that overproduce IL-13 and drive airway inflammation (13). Analysis of patterns of gene expression in children with asthma exacerbations supports the idea that similar mechanisms might be operative in humans (14). These findings suggest that a positive association between overproduction of IFN-λ1 and airway pathology is plausible.
The study by Miller and colleagues has several limitations, and the authors acknowledge most of these nuances. First, causality cannot be adequately assessed in an observational trial, and it is important to consider the alternative possibility that greater viral replication drives more severe illness and increased interferon responses. The authors measured viral RNA in nasal secretions, and the lack of correlation between viral shedding and wheezing led them to discount this possibility. However, estimation of viral shedding based on a single upper airway sample per illness is not likely to be accurate, and there are technical difficulties in quantitating RNA from the large number of genetically diverse HRVs. These shortcomings cast some doubt on their conclusion. Furthermore, the study did not include routine sampling of nasal secretions, and so asymptomatic colds were not included. Since up 25% of HRV infections are asymptomatic, this is an important consideration in analyzing the relationship between infection, immune responses, and illness severity.
Regardless of these technical limitations, the results of this large and well-designed study are thought provoking, and cast a shadow on the reputation of IFN-λ1, heretofore a well-respected defender against the ravages of viral respiratory infections in children. Are the effects of this antiviral cytokine more complicated than initially appreciated? Should we be equally suspicious of other pillars of the antiviral community? To resolve these questions, additional studies are needed to more clearly define mechanisms of potential adverse effects of overproduction of IFN-λ1, and to establish temporal relationships and kinetics of viral replication, IFN responses, and respiratory outcomes. These additional studies, perhaps utilizing experimental inoculation techniques or animal models of viral respiratory infection, should provide the information that is required to decide whether intervention studies to neutralize IFN-λ1 are truly warranted. The next chapter should be interesting.