Peter Wark, Centre for Asthma and Respiratory Disease, Level 3 HMRI, John Hunter Hospital, Lookout Rd, New Lambton Height, NSW 2305, Australia. Email: email@example.com
Influenza presents a unique human infectious disease that has a substantial impact on the public health, in general, and especially for those with chronic airways diseases. People with asthma and chronic obstructive pulmonary disease (COPD) are particularly vulnerable to influenza infection and experience more severe symptoms with the worsening of their pre-existing conditions. Recent advances in reverse genetics and innate immunity has revealed several influenza virulence factors and host factors involved in influenza pathogenesis and the immune responses to infection. Early innate immunity plays a critical role of limiting viral infection and spread; however, the underlying mechanisms that lead to enhanced susceptibility to influenza infection and severe symptoms in those with asthma and COPD to infection remain un-investigated. This review will explore the importance of early innate antiviral responses to influenza infection and how these responses are altered by influenza virus and in those with chronic airways diseases.
Influenza viruses are major respiratory pathogens that cause enormous morbidity and mortality in annual influenza epidemics. Epidemics are caused by frequent mutation rates of influenza antigens that result in a reduced immunity in the general population. However, excess mortality is often seen in the elderly, children less than 5 years, and individuals with pre-existing pulmonary disease, such as asthma and chronic obstructive pulmonary disease (COPD). These groups are at higher risk of influenza infection and developing clinical complications, and patients with influenza-associated exacerbations of asthma and COPD have significantly altered pulmonary function.1–8 In addition, influenza viruses have the propensity to re-assort into new influenza strains with the potential to cause pandemics. The population has no or limited immunity to these new viruses, which may result in dramatic increases in morbidity and mortality. Indeed, in the last 15 years, two important new influenza strains have emerged: high-pathogenic avian H5N1 strain in 1997 and the swine-origin H1N1, which resulted in the 2009 pandemic (H1N1/2009). Both asthma and COPD were over-represented in those admitted to a hospital with H1N1/2009 during the pandemic period.9,10
Influenza virus is one of the infectious viruses commonly detected during acute asthma and COPD exacerbation.3,11,12 There are several studies demonstrating strong correlation between the frequency of seasonal influenza detection and acute exacerbation,3,7,13,14 and infection is often associated with rapid decline in lung function and worsening of pre-existing symptoms and increased risk of death.4 Despite the recommendation of annual vaccination, those with chronic airways diseases remain to be among the most vulnerable to adverse outcomes from infection. The underlying mechanisms of this increased susceptibility to the effect of influenza infection and severe symptoms in subjects with asthma and COPD remain largely unexplored. This review will consider the importance of influenza infection in acute exacerbations of asthma and COPD and will focus on the importance of early immune responses in the airways to infection, in particular how these are exploited by influenza and how these may be altered in chronic lung disease.
Influenza A virus has eight segmented, negative sense, single-stranded ribonucleic acids (ssRNAs), which are encapsulated within a viral envelope, and the structure of the virus particle is roughly spherical in shape with approximately 80–120 nm in diameter.15 The viral envelope contains surface glycoproteins, haemaglutinin (HA) and neuraminidase (NA), and M2 ion channels. Within the envelope are the eight segments of influenza RNA that encode for 11 viral proteins (Fig. 1). Segments 1, 3, 4, 5, and 6 encode for polymerase basic (PB)2, polymerase acidic (PA), HA, nucleoprotein (NP) and NA, respectively. Segment 2 encodes for PB1, and via a different reading frame, an accessory protein PB1-F2 is expressed in most influenza A strain. Segment 7 encodes for matrix 1 (M1) protein and also M2 ion channel via alternative splicing. Segment 8 encodes for non-structural protein (NS)-1 and nuclear export protein or also known as NS-2.
In humans, the primary site of influenza infection is the epithelial cells of the respiratory mucosa. In contrast, in birds, the natural reservoir of influenza viruses, the primary site of infection is both the respiratory and intestinal tract.16 In order to enter the host cell epithelium, influenza binds by its surface glycoprotein HA to the sialyl sugar chain receptors on the host cell surface, allowing the virus to be internalized by endocytosis into the endosome in the host epithelial cell. The low pH environment of the endosome allows the viral HA to undergo conformational change and fuse with the endosomal membrane, thereby allowing the release of viral genes into the host cytoplasm17,18 (Fig. 2). Influenza viruses are different to other RNA viruses in that the transcription and replication of the viral genome takes place in the nucleus of the infected cells.19 The nuclear localization signal in the influenza genes allows the viral RNAs to be internalized into the host nucleus, where new viral genes are synthesized. The newly formed viral segments and structural proteins including HAs and NAs traffic to lipid rafts on the plasma membrane from which the virions are then released.20–23 Since the viral envelope is derived from the host membrane, which contains sialic acid (SA) glycoproteins, the newly formed virus remains intact on the host cell surface. The viral NA cleaves the host cell surface SA residues, releasing the newly formed virions free from the host cell surface.
THE CLINICAL EFFECT OF INFLUENZA INFECTION
Infection with seasonal influenza in healthy individuals usually results in symptoms ranging from asymptomatic infection to serious illnesses. The acute infection is usually associated with an abrupt onset of symptoms including fever (38–40°C), myalgia, anorexia, upper respiratory tract congestion and pharyngitis.24 The severity of the disease differs depending on the virus sub-types, the pre-existing health conditions of infected individuals and their immunity status. The immunological response and the virulence of influenza virus are critical for the outcome of infection. Host responses to influenza involve all arms of immune system, with the humoral and cell-mediated response being well characterized in viral clearance and recovery from infection.25 Annual influenza vaccination is designed to enhance the adaptive immune response by generating influenza-specific antibodies and priming the host-specific immune response.26 However, the frequent mutation rate of influenza viruses have made vaccine and antiviral drugs design difficult, thus a novel approach to antiviral drugs is important in preparation for future influenza pandemics. Even in the case of vaccination with known strains, efficacy is limited in subjects with asthma and COPD. Vaccination significantly reduces the severity of influenza infection and exacerbation rate but does not alter hospitalisation or mortality.27 An alternative to vaccination is the use of specific anti-influenza agents, such as the NA inhibitors, oseltamavir and zanamavir. While experience for the use of these antiviral agents has grown, especially since the 2009 pandemic, their clinical efficacy is limited28,29 and largely unknown in asthma or COPD. These drugs are relatively expensive and the development of antiviral resistance is of great concern.30 Innate immunity has recently gained much attention as it not only recognizes specific signatures of foreign pathogens, it also regulates antiviral and adaptive immunity. As it is the initial immune response to infection, enhancing the innate immunity has the theoretical advantage of limiting both the inflammatory response within the airway as well as reducing the systemic effects of infection. Understanding this early innate immune response hence provides novel therapeutic options to better prepare for future influenza pandemics such as the highly pathogenic avian influenza H5N1.
THE INNATE IMMUNE RESPONSE TO INFLUENZA
Innate immunity is an ancient component of the human immune system that provides the first line of defence against intrusion of any infectious pathogens. The most common route of transmission for respiratory pathogens is via the upper and lower airway; this therefore makes airway immune cells vitally important in detection and elimination of invading microbes. Airway epithelial cells (AECs), including of the upper and lower respiratory tract, are the primary site of infection for the majority of infectious respiratory viruses including influenza, and the interactions between host immune responses and influenza virus determine the outcome of infection.
Influenza viruses first anchor to the AECs by binding its surface receptor HA to glycoproteins with terminal SAs of specific configurations on the surface of AECs. SAs are sugar molecules found ubiquitously as terminal residues on the glycan chain of many polysaccharide and glycoproteins on human epithelial cell surface and secreted glycoproteins.31 Human influenza viruses preferentially bind to glycoprotein with terminal SAα2,6Gal linkage mainly found in the upper airways, whereas avian influenza viruses bind to that with terminal SAα2,3Gal linkage predominantly found in the lower respiratory tract.32–35 This differential binding preference is thought to be the reason for the unsustainable human-to-human transmission of the highly pathogenic avian influenza H5N1.32,33,36 At present time, H5N1 infections in humans are mostly due to direct avian-to-human transmission, and this is attributed to close proximity with poultry.37 Direct human-to-human transmission of H5N1 has been reported in a few cases, while the unfavourable binding preference of avian HAs to SAα2,6Gal linkage in the upper airway is likely limiting this sustainable transmission.
Despite these long considered influenza binding targets, we have demonstrated that both human and avian influenza viruses were able to readily infect primary human bronchial epithelial cells (BECs) to a similar extent based on HA levels inside the infected cells immediately after infection and replicated to a similar titre after infection.38 The differential expression of these SA residues did not appear to influence the in vitro infectivity of influenza viruses, which indicates the possibility of other unknown factors involved in influenza endocytosis. A study found that mice without SAα2,6Gal linkage supported similarly productive human influenza replication compared with wild-type mice.39 In support of alternative influenza entry routes, phosphoinositide 3-kinase (PI3K), a ubiquitous signalling molecule that is involved in various cellular functions such as cell proliferation and metabolism40,41 was found to be implicated in influenza endocytosis. Ehrhardt et al. 2006 showed that influenza viruses were unable to endocytose into AECs when PI3K was blocked,42 further indicating the existence of alternative route of entry for influenza.
Following successful entry into the host cells, influenza takes over the host cellular machinery to drive viral replication in the cytoplasm.17,18 During this attack process, there are mechanisms in place that deploy defensive strategies to control infection. Pattern recognition receptors (PRRs) are sentinels that identify common conserved regions of viruses such as the viral genomes and initiate immune responses to infection. The primary PRRs in BECs include toll-like receptors (TLRs) and retinoic acid-inducible gene-I (RIG-I)-LRs (RLRs). TLR3 is located in the endosome where upon binding to viral double-stranded RNA (dsRNA) initiates signalling cascades that lead to inflammatory response via activation of the transcription factor, nuclear factor (NF)-κB (Fig. 3). Inflammatory cytokines such as CXC chemokine ligand (CXCL)-8 recruits neutrophils to the site of infection that phagocytose foreign pathogens in the airways. Neutrophils are also capable of releasing inflammatory cytokines including CXCL-8 via similar pathway that further amplify the inflammatory response to influenza.43,44
Influenza infection is known to cause inflammation in the lung,45–47 and this is especially true with the highly pathogenic avian influenza H5N1. A hallmark feature of H5N1 infection in the infected individuals is massive inflammatory cytokine storm, leading to severe toxic shock and multi-organ failure with a fatality rate of approximately 80%.48,49 Interestingly, influenza has been shown to only generate ssRNA and not dsRNA that is recognized by TLR3 during its replication process,50 indicating that there are other players involved in driving inflammation beside TLR3.
The second PRR family that also recognizes influenza RNAs is the nucleotide-binding domain and leucine-rich repeat containing protein 3 (NLRP3), a member of the Nod-like receptor family. NLRP3 protein has been shown to play an important role in inflammatory responses to influenza. NLRP3 induction was increased following infection and has been demonstrated to recognize influenza ssRNA;51 it then complexes with an adaptor protein apoptotic speck protein containing caspase recruitment domain (CARD) and forms a complex known as the inflammasome. This complex then leads to inductions of inflammatory cytokines such as interleukin (IL)-6 and tumour necrosis factor (TNF)-α.52–54 However, NLRP3 has also been shown to be activated by influenza-induced reactive oxygen species (ROS) during infection. Therefore, both influenza endocytosis and ROS generated during infection can activate NLRP3 inflammasome that leads to inflammatory responses.
RIG-I is a cytosolic PRR that specifically recognizes the ssRNA that is generated by influenza during replication process.46,50,55–57 This engagement leads to the activation of interferon (IFN) regulatory factor (IRF)-3, which is then translocated into the nucleus where they act as transcription factors for type I IFNs (IFN-α/-β), and type III IFNs (IFN-λ1/-λ2/-λ3)58–61 (Fig. 3). These IFN responses are the central component of the innate immunity in the control of viral infection. Once released, type I and III IFNs bind to their respective receptors (interferon alpha receptor 1/2 and IL-28Rα/IL-10β, respectively) on the same and neighbouring cells and induce an array of antiviral proteins known as IFN-stimulated genes (ISGs) against replicating viruses.62,63 The critical antiviral ISGs include protein kinase RNA-activated (PKR), MxA and 2′, 5′-oligoadenylate synthetase (OAS). ISGs that are positive regulators of type I IFNs include RIG-I, MDA-5 and IRF-7, when unregulated they further amplify the antiviral response. PKR, OAS and MxA have similar roles in their ability to sense IRF7, when unregulated they further amplify the anti-viral response and degrade viral RNAs and induce apoptosis of the infected cells.64–69 Type I IFNs can also directly cause apoptosis by up-regulating pro-apoptotic proteins TNF-related apoptosis-inducing ligand.70–72 RIG-I has also been shown to induce the expression of inflammatory cytokines via NF-κB signalling during influenza infection, which further amplifies inflammatory responses in the airways.46,56 Influenza infection thus appears to drive inflammatory response largely via TLR3 and NLRP3 inflammasome signalling pathway, and RIG-I is the mediator for inflammatory response and antiviral type I/III IFNs,46,60,61 the latter of which are critical in the fight against influenza at the site of infection.
The AEC is the first point of contact between the host and infecting viruses, but efficient containment and resolution of infection requires the activation of an adaptive immune response. In this regard, the dendritic cells (DCs) act as an important intermediary between the immediate innate immune response and the recruitment of an appropriate adaptive response. The population of DCs in the lung is dynamic with the rapid recruitment of immature DCs following exposure to bacteria, viruses or inflammatory stimuli (chemokines CCL2, 3, 5, 7 and 20).73,74 Several distinct DC phenotypes are present in the lung with varying immune functions. Myeloid DCs act as airway sentinel cells that capture influenza viral antigen released from virus-infected AECs, migrate to lymph nodes and activate an influenza-specific T cell response for efficient viral clearance.75 Another important subset of DCs is the plasmocytoid DCs (pDCs), which are powerful IFN inducers against influenza.76 TLR7 is the main PRR located in the endosome in pDCs that recognize influenza and initiates IFN responses via IRF7 and inflammatory response via NF-κB to influenza infection.77,78 The type I and III IFNs produced by pDCs and AECs not only help and limit viral infection but also enhance DC maturation and their ability to stimulate T cells for efficient viral clearance.79,80
DCs can also interact with a unique subset of T lymphocytes called invariant natural killer T (iNKT) cells that have been implicated in innate immunity to viral infections. Influenza infection in DCs leads to activation of RIG-I and TLR7 that results in the release of IL-1β and IL-23. These cytokines then synergistically prompt iNKT cells to produce immune-modulatory cytokines such as type II IFN, IFN-γ and IL-22, which enhanced innate and adaptive immune responses.81,82 IFN-γ is an immune-modulatory cytokine that signals DCs and BECs to produce inflammatory cytokines that recruit macrophages and neutrophils to the site of infection and inflammation.83 Influenza-activated/-infected DCs also release IL-1β, IL-6, TNF-α and IL-10, which expand and activate another subset of T lymphocytes called Th17 cells.82,84 Once activated by infected DCs, Th17 cells mainly play an inflammatory role in releasing IL-17 and IL-22 that then further enhance the inflammatory responses in DCs and BECs.85–89 This therefore contributes to the exaggerated inflammatory response in the airways of those infected with influenza, especially with H1N1/2009 pandemic virus.90 DCs therefore play an important role in driving both inflammatory and antiviral responses to influenza infection that further limits viral replication at the initial site of infection.
IMMUNE SUPPRESSION BY INFLUENZA NS1
The NS1 protein is a multifunctional protein expressed very rapidly to help the establishment of viral infection by interfering with host messenger RNA (mRNA) processing and translation, as well as inhibiting host immune responses, especially the antiviral system.
The NS1 protein is composed of an effector and an RNA-binding domain, the former of which binds and prevent the host mRNA from translation.65,91,92 Shutting down the host cellular protein synthesis in the infected cells helps the virus to gain control of the host machineries required for viral protein synthesis. The IFN antagonistic property of NS1 occurs at multiple stages of the IFN signalling cascade (Fig. 4). NS1 interacts with the viral sensor RIG-I and can also inhibit the downstream activation of IRF3, thereby inhibiting type I IFNs.93,94 The NS1 protein specifically inhibits activation of RIG-I, which is crucial for maximal type I IFNs expression during viral infection.95 Beside the inhibitory role in type I IFNs expression, NS1 protein also inhibits cellular proteins that establish the antiviral state of infected cells. The RNA-binding domain of NS1 can bind to viral RNA to prevent detection by PKR.96 It also binds to PKR itself via the effector domain (residue 123–127) and inhibits PKR-mediated viral mRNA suppression and PKR-induced apoptosis.97–100 OAS, which detects and cleaves viral RNA by activating RNase L, can also be blocked by the influenza NS1. The RNA-binding domain of NS1 can out-compete the RNA binding capacity of OAS, thereby inhibiting antiviral response.101
The NS1 protein of different strains of influenza appears to have variable strength in antiviral suppression. The NS1 of a human H3N2 strain was found to be more effective in IFN suppression than that of a low-pathogenic avian influenza, H11N9, and resulted in a higher viral replication.38 H5N1 encodes an NS1 protein that is highly effective in inhibiting host antiviral responses that contributes to it high fatality in humans.102,103 Influenza NS1 is no doubt a major virulence factor during infection due to its antiviral suppression. This allows influenza virus to replicate inside the host with minimal interference from the host immune system; nonetheless, the exact mechanisms by which NS1 exerts its function needs to be further investigated before novel therapeutics can be discovered.
INFLUENZA AND CHRONIC RESPIRATORY DISEASE
Chronic airway inflammation is a feature of both asthma and COPD.104–106 This chronic inflammation is often associated with increased oxidative stress, which can be caused by the constant inflammation or exposure to toxic inhaled gases such as cigarette smoke. IL-17 has also been shown to be elevated in sputum of those with asthma107 and COPD,108 which correlates with high inflammatory response and neutrophils infiltration in the lung of those with chronic airways diseases. Infection with influenza can further enhance the already elevated inflammatory responses and worsen symptoms in asthma and COPD. iNKT cells have also been shown to have increased in bronchial biopsies from asthmatic subjects;109 however, another study did not observe any difference in iNKT cell numbers in bronchial biopsies, sputum and bronchial lavage fluid of asthmatic and COPD subjects compared with healthy individuals.110 It is unclear what effect this pre-existent immune activation may have on acute influenza infection, though it may have the potential to considerably heighten the inflammatory response to infection within the airways of subjects with asthma and COPD.
Oxidative stress and cigarette smoke exposure has been shown to activate NF-κB and up-regulate inflammatory cytokines in BECs and macrophages,111,112 and the resulting oxidative stress can lead to oxidation of DNA and protein and cause epithelial and tissue injury with amplified inflammation.113 Oxidative stress has also been linked to reduced histone deacetylase (HDAC) 2, an enzyme generally associated with the epigenetic regulation of gene expression. HDAC2 is involved in the negative regulation of inflammatory cytokine expression,114,115 and its expression was shown to be decreased in alveolar macrophages from asthmatic patients116,117 and in peripheral lung tissue, alveolar macrophages and lung biopsies from COPD subjects.114 This indicates that chronic exposure to oxidative stress may have tremendous impact on factors that regulate inflammation and together with frequent viral infections leading to chronic damage to the lung of those with chronic airways diseases.
Respiratory viral infection accounts for approximately 40–60% of all acute exacerbation of COPD and up to 80% of asthma exacerbations,5,118 and influenza virus is one of the infectious viruses commonly detected during acute exacerbation.11,12 Influenza infection accounts for approximately 16–25% of all virus-induced acute COPD exacerbation.7,119,120 Infection may further amplify pre-existent airway inflammation, but this increase does not appear to directly contribute to viral clearance as recovery period from infection in those with asthma and COPD is longer than that in healthy individuals.
RIG-I- and IFN-initiated antiviral responses in infected BECs are well known in its ability to reduce viral replication, and we also have shown that both type I and type III IFNs can suppress influenza replication. This inhibition was even more pronounced when both types of IFNs were used in combination.38 While the antiviral response profile in the airways of chronic airways diseases is largely uncharacterized, two recent studies provided a glimpse of impairment of innate immunity in asthma. Wark et al. 2005 and Contoli et al. 2006 demonstrated for the first time that primary BECs (pBECs) from subjects with asthma have impaired innate immune response to infection with rhinovirus (the common cold virus), which leads to increased susceptibility to infection and more severe clinical disease.121,122 Rhinovirus infection in asthmatic pBECs induced a heightened release of inflammatory cytokines including CXCL-8 but produced a lower level of IFN-β and IFN-λ1 compared with healthy controls. This indicates that while pBECs also actively participate in inflammation, the reduced antiviral response directly impairs the ability of the infected host cells to undergo early apoptosis, leading to increased virus replication. Given the similar inflammatory conditions in the upper airways of those with asthma and COPD, individuals with COPD may similarly be deficient in eliciting a robust antiviral response upon exposure to viral infection, leading to acute exacerbation and increased mortality. Furthermore, while influenza infection can further promote inflammatory responses in those with chronic airways diseases, influenza viruses also produce NS1 protein that effectively suppresses RIG-I- and IFN-driven antiviral responses to maximize viral replication in the host.38,95,123,124 This strongly indicates that those with asthma and COPD not only suffer excessive inflammation in the airways following influenza infection, but their ability to mount robust antiviral responses is incapacitated, leading to highly efficient viral replication and prolonged recovery.
Cigarette smoke has also been shown to inhibit RIG-I and lead to reduced antiviral responses to influenza infection in lung fibroblast and epithelial cells.125,126 This correlated with oxidative stress such as that induced by cigarette smoking, which decreased type I IFN-mediated antiviral responses.127 In addition, cigarette smoke was also shown to reduce HDAC1 expression in macrophages,112 which is involved in positive regulation of type I IFN responses.128 This therefore raises the possibility that chronic exposure to cigarette smoke may progressively modify epigenetic and antiviral profiles in asthma and COPD pBECs and result in unregulated inflammatory and impaired antiviral responses. In combination with frequent viral infections, this leads to chronic damages to the lung of COPD subjects.
Influenza remains an important infectious disease that presents a great concern to all individuals, especially those with chronic inflammatory airways diseases such as COPD. Despite vigorous research in innate immunity in recent years, it remains unclear why subjects with COPD have increased susceptibility to infection with influenza and consequently suffer more severe complications. Recent advances in innate immunology have identified important host factors in controlling influenza infections. Understanding the interaction between influenza and innate immunity will not only unveil the abnormalities that exist in asthma and COPD, but this may also potentially lead to better treatment and therapeutics.