Understanding the Flu: Host and Environmental Factors Associated With Susceptibility to Influenza Virus-Induced Disease
IN THIS ARTICLE
Influenza virus infection is a worldwide public health burden because of the significant morbidity and mortality that result from seasonal outbreaks and devastating epidemics. Despite extensive research and therapeutic measures to combat influenza virus infections, the genetic and environmental factors that impact influenza virus-induced disease susceptibility remain poorly defined. Here we report the current state of knowledge regarding the host and environmental factors that impact influenza virus susceptibility. Mouse models of influenza virus infection have led to the identification of several candidate susceptibility genes, although many of these genes have not been verified in human cohort studies. Moreover, some studies have demonstrated that preexisting conditions correlate with enhanced influenza virus-associated morbidity and mortality, and chronic exposure to cigarette smoke and alcohol increases one’s risk of infection. However, there remains a lack of human-based evidence to corroborate the in vitro and animal model studies that demonstrate a role for environmental factors in the risk of influenza virus infection. Further studies are needed, particularly large cohort studies involving geographically dispersed individuals, to clarify the genetic and environmental risk factors for influenza virus-induced disease susceptibility and lead to more effective treatment strategies for infections.
Influenza, which is a respiratory illness often associated with fever, body aches and fatigue, is a disease of significant global health importance because of its ability to produce both devastating epidemics and continued seasonal illness (1). Of the three influenza virus subtypes, A, B and C, ianfluenza A viruses are clinically the most destructive because of the morbidity and mortality that they induce in humans and animals. The viruses within each subtype are classified based on the antigenic similarity of the viral proteins hemagglutinin (H) and neuraminidase (N), and the high mutation rate of the virus enables it to escape from preexisting immunity and targeted vaccine strategies (1). Worldwide, seasonal influenza causes severe illness in three to five million people, and it results in hundreds of thousands of deaths each year (2). Approximately 50 million fatalities have been attributed to the influenza pandemic that occurred between 1918-1919, and other H1N1 pandemics occurred throughout the 20th century that were associated with lesser but still significant morbidity and mortality (3). The mortality rate induced by highly pathogenic avian H5N1 infections is estimated to be between 50-60% (4), which is thought in part to be due to the excessive cytokine storm induced by these viruses compared to the immune response induced by infection with other influenza virus strains (5). Vaccine development against highly pathogenic H5N1 viruses is still in experimental stages, underscoring the continued threat that these viruses pose to public health (6).In the absence of a vaccine or pre-existing immunity, the emergence of a new strain of influenza virus places nearly every member of the population at risk for infection (7). However, in an immunologically naïve population, susceptibility to influenza is not equal among all persons. Susceptibility to and severity of influenza virus infection is mediated by an array of host characteristics and environmental exposures. With the 2009 swine-origin H1N1 pandemic and the continued threat of the emergence of H5N1 avian-origin influenza that is capable of sustained human-to-human transmission, it is increasingly important to understand the factors that affect susceptibility to influenza pathogenesis (8). In the early stages of a pandemic, before a vaccine is available, those at increased risk for severe disease can be specifically targeted for preventive measures (9). Although mouse models clearly demonstrate a role for genetics and several environmental factors in susceptibility to influenza virus infection, the results of human-based studies involving influenza virus infection have been less clear and somewhat controversial (9-14). In this paper, we review the experimental and epidemiologic literature on the host and environmental factors affecting influenza susceptibility and severity.
A review of the literature was conducted, which was limited to articles in English published from January 1, 1975 to January 16, 2013. Articles were identified by querying PubMed using search terms selected by the authors based on the intended scope of the review, followed by evaluation of the bibliographies of relevant articles. Search terms included “influenza” AND “susceptibility” AND one of the following: “stress” “nutrition,” “smoking,” “alcohol,” “pollution,” “pollutants,” “genetics,” “immunity,” “pregnancy,” and “respiratory infections.” Exclusion criteria: we chose to exclude articles pertaining to risk factors associated with co-infections with influenza and other viruses or bacterial species. Articles were not limited by host or animal models. Citations were reviewed, and relevant articles were included in the review, including epidemiologic, animal, and in vitro studies (Table 1).
1a. Epidemiological data
Several epidemiological and experimental studies suggest that influenza-induced disease has a significant genetic component. In-depth analysis of genealogical data for victims of the 1918 influenza pandemic found that the relative risk (RR) of death for relatives of infected individuals was significantly higher (P<0.0001) than the RR of unrelated individuals (14). The study revealed that the RR for first-degree relatives and spouses was 1.54 and 1.98 respectively, and second- and third-degree relatives were also significantly at risk for influenza-associated death (P<0.0001). Recent avian H5N1 influenza virus outbreaks reveal epidemiological clustering of infected individuals wherein over 90% of case clusters have occurred among blood-related family members (15). Moreover, a specific single nucleotide polymorphism (SNP) in the human Toll-like receptor 3 gene (TLR3), an antiviral gene involved in innate immunity, results in a loss-of-function mutation associated with increased susceptibility to influenza-induced encephalopathy (16). A study examining the role of SNPs in the tumor necrosis factor (TNF) gene during 2009 H1N1 influenza pandemic found that a specific SNP in the TNF gene was associated with increased incidences of influenza virus infections and increased risk for viral pneumonia (17).
Genetic susceptibility has also been linked to Human Leukocyte Antigen (HLA) type as HLA-A2+ individuals show better control of viral replication following influenza infection than individuals with other HLA types (18). Additionally, a study by Shashkov and colleagues found a correlation between ABO blood type and susceptibility to influenza virus infection (19). This study found more severe clinical manifestations of influenza virus infections in individuals with blood type A than individuals with blood type O or B. However, a study by Gottfredsson et al (20) challenges the proposed genetic link in susceptibility to influenza virus. This genealogical study looked at 455 Icelandic individuals that succumbed to 1918 Spanish influenza virus and found that the risk of mortality from influenza virus infection was similar within families of patients who succumbed to the illness and within families whose members survived the infection, suggesting that influenza-induced mortality is not influenced by genetic inheritance. Although some studies have challenged the genetic link to influenza virus infection (21), sufficient case study and genealogical data seems to suggest that susceptibility to influenza-induced illness has a strong heritable component.
1b. Animal models
Most of the evidence for genetic susceptibility to influenza virus infection has been generated through experimental studies, particularly those involving mouse models. Genetic mapping studies have mapped influenza virus resistance to the murine MX1 gene (22-24). Mice carrying the wild-type Mx1 gene were protected against 1918 H1N1 and H5N1 influenza virus challenge, while mice deficient in Mx1 succumbed to lethal infection. A mapping study by Boon et al identified five quantitative trait loci (QTL) associated with H5N1 influenza resistance in mice (25). Of the 121 genes located at the identified QTL, many of the genes, including IFN-alpha 4, TNF, CXCL11, CXCL2, and IRF-1, are associated with inflammation and antiviral immunity. The QTL-associated genes have been identified as candidate susceptibility genes due to the fact that polymorphisms in these genes may influence H5N1 influenza virus pathogenesis and disease. In a study evaluating the impact of influenza virus infection in inbred mouse strains, Trammell et al (26) found significant strain-specific differences in morbidity and mortality, viral titers and cytokine expression in the lungs of the different mouse strains following infection with an H3N2 influenza virus strain. A similar study by Srivastava et al (27) examined the susceptibility of seven inbred mouse strains to H1N1 and H7N7 influenza virus infections and found differences in influenza-induced morbidity and mortality, viral titers, and lung pathology across the different inbred mouse strains.
A follow-up study by the same group found that cytokine and chemokine responses differed between the susceptible and resistant mouse strains following influenza virus infection (28). Additionally, a study by Boivin et al (29) mapped genetic determinants of influenza susceptibility in mice to two sex-specific QTL. In this study, mice were infected with H3N2 influenza virus and then investigators utilized clinical QTL and lung RNA expression QTL (eQTL) to identify two sex-specific QTL on chromosomes 2 and 17. A more recent genetic mapping study by Bottomly et al identified 21 significant eQTL associated with host responses to mouse-adapted A/PR8/34 (PR8) H1N1 influenza virus infection using a novel panel of genetically diverse mouse lines derived from classically inbred and wild-derived mouse strains (30). The authors of the study identified extreme phenotypes within the 44 mouse lines examined, with some mice showing very mild and very severe immune responses to infection. The continued use of system-based genetic mapping to identify candidate susceptibility genes to influenza-induced disease, particularly in diverse animal populations, may be an effective approach to identify genes that are more likely to be important in a diverse human population. Furthermore, in vitro and knockout mouse studies demonstrated that interferon-inducible transmembrane 3 (IFITM3) is an essential host defense protein that restricts influenza virus replication and pathogenesis (31). Additionally, analyses of hospitalized influenza patients revealed that the infected patients showed enrichment for a polymorphism in the IFITM3 gene, and this mutation is associated with decreased influenza virus restriction in vitro. Taken together, this study demonstrates the ability for genetic mouse studies to effectively predict susceptibility factors for influenza virus pathogenesis in humans.
2a. Natural immunity
Knockout mouse studies have shown the importance of many toll-like receptor (TLR) genes and other innate immune genes in combating influenza-associated disease. Both TLR3 and TLR7 knockout mice show better survival rates than wild-type mice following lethal influenza A virus infection with H3N2 or H1N1 strains, respectively (32,33). Esposito et al. further demonstrated the importance of TLR3 to influenza virus susceptibility in a cohort study that examined the role of TLR polymorphisms in influenza-induced disease during the 2009 H1N1 pandemic (34). The results of the study showed that a genetic polymorphism in the TLR3 gene correlated with 2009 H1N1 influenza-induced pneumonia in children. Interestingly, the study by Esposito and colleagues found that viral loads in the blood of the patients were comparable between all of the study groups that were examined, despite differences in TLR polymorphisms and susceptibility to influenza-induced disease. An adaptor molecule involved in TLR signaling, MyD88, has also been found to be involved in the immune response to influenza virus infection (35). Similar to TLR3 and TLR7 knockout mice, MyD88-deficient mice also survive influenza A virus infection significantly better than wild-type mice following infections with pandemic and seasonal H1N1 strains (33). These knockout mouse studies and genetic cohort study indicate that several TLR and TLR-associated signaling molecules are important for the response to influenza virus infection, and these innate immune factors may be candidate genes to investigate for their role during human influenza virus infection.
Other host factors, including those involved in interferon (IFN) induction, have also been noted for their antiviral role following influenza virus infection. The RNA helicase RIG-I, a molecule involved in IFN induction during viral infection, is directly inhibited by the influenza viral protein NS1 during PR8 influenza A virus infection (36). Inhibition of RIG-I and other host factors by NS1 prevents IFN signaling during influenza virus infection, a mechanism of immune evasion elicited by the virus (37). Recent studies have also begun to investigate the role of inflammasome proteins, such as NLRP3, and suggest a protective role for these proteins in identifying and protecting from influenza A virus infection (38). Additionally, several cytokines and chemokines induced by influenza virus have been found to have protective and pathologic effects in the infected host. Interleukin-17 (IL-17), a pro-inflammatory cytokine involved in neutrophil activation and migration, has been found to have pathologic effects in the influenza-infected host (39). A study examining the role of IL-7 during mouse-adapted influenza A virus infection (PR8) found that IL-17 receptor knockout mice showed reduced weight loss and increased survival rates compared to wild-type mice following influenza virus challenge. In contrast, mice lacking the protease-activated receptor-2 (PAR2), a receptor expressed in the respiratory tract and associated with IFN-gamma production, showed increased susceptibility to influenza virus infection compared to wild-type infected mice (40). Other cytokines and chemokines induced by influenza virus include CCL2 (MCP-1), CCL5 (RANTES), IL-1 beta, IL-6, TNF-alpha, and IFN, resulting in a primarily pro-inflammatory Th1-type immune response in the infected host (41). A small cohort study demonstrated a correlation between a naturally occurring polymorphism in the CCR5 gene and susceptibility to the 2009 H1N1 pandemic influenza virus strain (42), although further studies will be required to determine whether this correlation is maintained in larger studies.
Furthermore, examination of inflammatory infiltrates in the lung following influenza virus infection reveals protective and pathologic roles for immune cells during infection. Infection of macaques with 1918 influenza virus results in enhanced neutrophil migration into the lung and corresponds with severe immunopathology (43), while NK T cells, CD8 T cells, NK cells and macrophages have been shown to protect infected mice from severe influenza-induced disease by promoting viral clearance (44-46). Respiratory epithelial cells and macrophages have also been shown to play important roles in the innate immune responses to influenza virus infection, and the differential immune responses induced in these cell types in humans and pigs correlate with differences in susceptibility to infection (47). An in vitro study also demonstrated that complement proteins, which are important mediators of innate immunity, from guinea pig sera effectively lysed influenza A virus-infected hamster kidney cells (48), although the importance of complement-mediated influenza virus clearance in vivo is unclear. Moreover, deficiency in immunoglobulin A (IgA) secretion during H1N1 influenza A virus infection (PR8) results in increased susceptibility to illness in mice and in guinea pigs (49,50), and virus-specific IgG responses, particularly IgG2 responses in humans, have been shown to be important for protection from pandemic 2009 H1N1 virus-induced disease in both humans and mice (42,51). However, earlier studies in ferrets showed that passive transfer of H0N1 PR8-like influenza virus-specific antisera alone does not protect ferrets from de novo infection with homologous virus, suggesting that antibody alone is not sufficient for protection from influenza virus infection in ferrets (52).
Another aspect of the influenza virus-specific immune response that has only recently been investigated is the role of small RNAs in the development of influenza-induced disease. A recent study by Peng and colleagues demonstrated that host-derived small RNA expression, which includes microRNAs and non-microRNAs, following influenza virus infection varied according to the viral strain that the mice were exposed to, and the targets of the small RNAs were predicted to impact innate immunity and cytokine production during infection (53). Thus, the immune response in the infected host, both at the innate and adaptive levels, greatly influences host susceptibility to influenza-associated disease.
Immunosenescence, a phenomenon accompanied by immune system dysregulation in aged individuals, results in depletion of adaptive immunity (i.e. memory T and B cell responses) while maintaining innate immune function (54). The increased Th2-type environment induced by immune cells, as well as poor CD8 T cell activation during viral infection, may affect the ability for older individuals to clear influenza virus from the lungs following infection. Increased susceptibility to influenza virus infection in immunosenescent individuals may also be a result of increased incidences of underlying medical conditions in aged populations (54). Elderly individuals have increased rates of respiratory problems because of airway hyperactivity, chronic obstructive pulmonary disease (COPD), lung neoplasms, pulmonary fibrosis and pneumonia. Establishment of influenza infection in the lungs of aged individuals with a comorbid condition is correlated with increased influenza-associated disease and mortality (55). Repeated influenza vaccinations in elderly humans and in mice have been shown to improve humoral responses (and thus protective vaccine-induced antibody responses) against influenza virus (56,57).Continued on Next Page »