Induction and Evasion of Type-I Interferon Responses during Influenza A Virus Infection

Abstract

Influenza A viruses (IAVs) are contagious pathogens and one of the leading causes of respiratory tract infections in both humans and animals worldwide. Upon infection, the innate immune system provides the first line of defense to neutralize or limit the replication of invading pathogens, creating a fast and broad response that brings the cells into an alerted state through the secretion of cytokines and the induction of the interferon (IFN) pathway. At the same time, IAVs have developed a plethora of immune evasion mechanisms in order to avoid or circumvent the host antiviral response, promoting viral replication. Herein, we will review and summarize already known and recently described innate immune mechanisms that host cells use to fight IAV viral infections as well as the main strategies developed by IAVs to overcome such powerful defenses during this fascinating virus–host interplay.

Among the respiratory pathogens that cause seasonal epidemics in humans worldwide, influenza viruses are particularly relevant because of their associated burden of disease. In addition, their ability to circulate in several mammalian and avian species provides an optimal scenario for zoonotic transmissions, which have resulted in severe pandemic outbreaks as a result of the introduction of novel influenza virus subtypes in the human population (Horimoto and Kawaoka 2005). The most devastating influenza pandemic ever recorded, the “Spanish Flu,” in 1918–1919 decimated ∼ 3% of the world population (Taubenberger et al. 2019). Although seasonal epidemics have significantly lower mortality rates, annual cases of severe illness oscillate between 3 and 5 million globally, with 300,000–650,000 deaths (WHO 2018).

Influenza viruses, as well as any other viral pathogen, must overcome the host innate immune response, which is rapidly mounted upon infection by recognizing conserved motifs in the virus. This ignites multiple downstream pathways of cytokine and interferon (IFN) activation to promote antiviral and pro-inflammatory responses, thus restricting early stages of infection before activation of adaptive immunity. In this review, we summarize the mechanisms used by the host innate immune response against influenza virus infection, and how the virus has evolved to counteract this first line of defense to successfully infect the host cell, with a special remark on recent advances in our knowledge about virus–host interactions. This review does not pretend to be an exhaustive compendium of the interplay between innate immunity and influenza A virus (IAV) replication, but rather it emphasizes some of the aspects of this interplay as examples of the complexity of the host–virus interactions taking place during IAV infection.

INNATE IMMUNE RESPONSE AND ITS ANTIVIRAL ACTION TO INFLUENZA INFECTION

Influenza viruses belong to the Orthomyxoviridae family, with a segmented, negative-sense, single-stranded RNA genome. They are classified into four genera (A, B, C, and D) (Shaw and Palese 2013). Influenza A and B viruses are responsible for all seasonal epidemics, whereas influenza C and D viruses are associated with mild disease and are restricted to specific hosts (Wagaman et al. 1989; Ferguson et al. 2016). The influenza A virus (IAV) genome consists of eight single-stranded segments, encapsidated as individual ribonucleotide complexes (RNPs). All segments encode for different proteins with specific roles during viral infection and replication. Among them, the viral hemagglutinin (HA) protein plays a key role in targeting airway and alveolar epithelial cells, mediating host cell recognition and viral entry (Shinya et al. 2006; van Riel et al. 2010). HA binds to sialic acid moieties on carbohydrate side chains of cell-surface glycoproteins and glycolipids (Gambaryan et al. 1997). This interaction induces internalization of the virion via endocytosis (Lakadamyali et al. 2004), and low pH conditions within the endosome trigger membrane fusion and subsequent release of the viral contents into the cytoplasm (Fontana et al. 2012). Although in humans, and in most mammalian species, influenza virus strains target the respiratory epithelium, in birds, most avian influenza viruses replicate predominantly in intestinal epithelia.

ACTIVATION OF THE INNATE IMMUNE RESPONSE

Toll-Like Receptors and RIG-I-Like Receptors

The innate immune response starts when infected cells detect the presence of IAV RNA (Fig. 1). This rapid and nonspecific response is the first line of defense against viral pathogens and it is initially sensed through the recognition of pathogen-associated molecular patterns (PAMPs) via three different types of pattern recognition receptors (PRRs): Toll-like receptors (TLRs), the retinoid acid-inducible gene I (RIG-I), and the NOD-like receptor family member NOD-, LRR-, and pyrin domain-containing 3 (NLRP3) (Garcia-Sastre 2011; Pang and Iwasaki 2012; Iwasaki and Pillai 2014; Kuriakose and Kanneganti 2017). This leads to the innate immune signaling pathway activation, inducing cytokine production as well as the induction of different antiviral molecules (Ouyang et al. 2014; Cao 2016). PRRs discriminate self- from non-self-molecules in infected cells at the endosomal and cytoplasmic compartments thanks to unique viral features that are not found in cellular RNAs, such as the presence of a 5′-triphosphate group or regions of double-stranded RNA (dsRNA) (Hornung et al. 2006; Baum et al. 2010; Rehwinkel et al. 2010).

Figure 1.

Schematic diagram of the innate immune response signaling pathway and evasion strategies upon influenza A virus infection. Upon viral replication, 5′-triphosphate single-stranded RNA (ssRNA) and double-stranded RNA (dsRNA) products are recognized by host pathogen recognition receptors (PRRs), which include RIG-I. This leads to a conformational change in RIG-I and causes exposure of its CARD domains that are ubiquitinated by TRIM25. Then, RIG-I associates with MAVS at the mitochondrial membrane, starting a downstream signaling cascade that leads to IRF3, IRF7, and NF-κB transcription factor activation and expression of type-I IFN. Secreted IFN-α/β produced by influenza A virus (IAV)-infected cells bind to IFNAR receptors in the cell surface, leading to JAK1/TYK2 phosphorylation and followed by binding and phosphorylation of STAT1 and STAT2 that will form a complex with IRF9. This ISGF-3 complex will function as a transcription factor to activate and transcribe hundreds of genes, among them ISGs, with important roles in mediating the host antiviral response. Red boxes indicate specific sites at which IAV proteins interfere or block the host innate immune response. These mechanisms of evasion are more extensively explained in the text.

TLRs are transmembrane proteins that sense viral nucleic acids outside the cell membrane or internally at endosomes and lysosomes through their extracellular domain (Takeshita et al. 2006; Kawai and Akira 2011). TLR3 and TLR7/8 are particularly important for IAV infection, because they can recognize and bind to single-stranded RNA (ssRNA) and dsRNA, respectively (Lund et al. 2004; Blasius and Beutler 2010). Binding of TLRs to RNA occurs through the Toll/IL-1 Receptor (TIR) domains located in the TLR cytoplasmic tails and results in recruitment of adapter molecules: TRIF in the case of TLR3 and Myd88 in the case of TLR7/8 (Kawai and Akira 2011).

IAV intracellular ssRNA and associated transcriptional products are recognized by the RIG-I-like receptors (RLRs) RIG-I, MDA5, and LGP2, which can distinguish between viral and self-cellular RNAs in the cytoplasm of most cell types (Loo and Gale 2011). RLRs contain a central RNA helicase domain, as well as a carboxy-terminal repressor domain (RD), necessary for RNA binding. In addition, both RIG-I and MDA5 have tandem amino-terminal caspase activation and recruitment domains (CARDs), which mediate the antiviral transduction signal downstream the RNA recognition but are absent in LGP2 (Takahasi et al. 2009; Bruns and Horvath 2015). RIG-I and MDA5 structurally recognize different RNA species. Although RIG-I recognizes di- and triphosphate groups located at the end of dsRNA (Hornung et al. 2006; Goubau et al. 2014), MDA5 recognizes and binds to long dsRNAs that are potentially IAV replicative intermediates as well as web-like RNA aggregates with no end specificity (Kato et al. 2006; Pichlmair et al. 2006). Despite being structurally related to RIG-I and MDA-5, the specific role of LGP2 is still not clear. However, it is believed that this protein may act as a cofactor, thus making the viral RNA more accessible to MDA-5 (Venkataraman et al. 2007).

Among all the different RLRs described to date, RIG-I is the best characterized and it is considered to be the main viral RNA sensor that gets activated upon IAV infection, ultimately leading to IFN induction (Kato et al. 2006). Upon PAMP recognition, RIG-I RD binds to RNA 5′ triphosphate leading to RIG-I dimerization and activation. Further recognition of dsRNA stimulates ATP hydrolysis and induces RIG-I translocation in the RNA molecule (Myong et al. 2009). This results in a conformational change that leads to CARD domain exposure and ubiquitination by E3 ligases such as TRIM25 (Gack et al. 2007) and RIPLET (Oshiumi et al. 2010). Upon ubiquitination, RIG-I interacts with the mitochondrial antiviral-signaling (MAVS) protein, activating the interferon regulatory factor 3 (IRF-3), IRF7 and the nuclear factor kappa-light-chain-enhancer (NF-κB). This leads to the expression of a variety of cytokines, such as IFN-α and IFN-β. Moreover, NF-κB induces the activation of the NLRP3 inflammasome and mediates caspase-1 activation, subsequently releasing cytokines IL-18 and IL-1β, which play an important role in the pro-inflammatory response upon IAV infection (Kuriakose and Kanneganti 2017).

IFN ACTIVATION

Activation of IRF3, IRF7, and NF-κB drives the expression of pro-inflammatory cytokines, including TNF, IL6, IL1β, and IFN (Hiscott et al. 2006). The innate immune response against IAV infection heavily relies on activating type-I IFN (IFN-α and IFN-β) and type-III IFN (IFN-λ), mounting a response in infected cells in an autocrine manner, but also in noninfected cells through paracrine signaling. IFN-α/β and IFN-λ interact with their respective receptors IFNAR and IFNLR, respectively, inducing the Janus kinase-signal transducer and activator of transcription (JAK-STAT) signaling pathway (Aaronson and Horvath 2002). Downstream phosphorylation of STAT1 and STAT2 leads to the formation of the IFN-stimulated gene factor 3 (ISGF3) complex with the association to IRF9. Upon translocation into the nucleus, ISGF3 drives the transcription of several IFN-stimulated genes (ISGs) by interacting with the IFN-stimulated response element (ISRE) located in the promoter region (Platanias 2005; Schneider et al. 2014). Although both type-I and type-III IFNs were canonically considered to have overlapping functions, recent studies have revealed spatiotemporal differences between IFN-α/β and IFN-λ (Levy et al. 2011; Hemann et al. 2017). Although both are rapidly induced upon infection, type-I IFNs are promptly modulated to basal levels shortly after, whereas expression of IFN-λ is sustained for a longer period of time, localized in specific tissue compartments. Importantly, IFNAR is ubiquitously expressed, whereas IFNLR expression is mainly restricted to epithelial cells. This difference in expression translates into fundamentally different responses upon viral infection. Both IFN-α/β and IFN-λ have antiviral activity in respiratory epithelial cells. However, IFN-λ treatment against IAV has minimal inflammatory side effects in the lungs, contrary to what has been described for IFN-α/β (Davidson et al. 2016; Galani et al. 2017).

cGAS/STING

RIG-I and MDA5 can sense viral RNA to stimulate the signaling via MAVS and downstream to the TANK-binding kinase 1/IRF3 (TBK1-IRF3) pathway. Interestingly, an alternative mechanism that leads to IFN production in a RLR-MAVS-independent manner has been recently described—that is, through stimulator of interferon genes (STING) activation, leading to TBK1-IRF3 and IFN induction (Holm et al. 2016). STING is known to participate in the control of DNA viruses and bacteria, leading to type-I IFN induction upon activation by cyclic GMP-AMP (cGAMP). This molecule is produced after stimulation of the cGAMP synthase (cGAS) in response to cytoplasmic DNA. Moreover, a recent study identified a STING-dependent, cGAS-independent pathway of IFN induction that is triggered by membrane fusion during enveloped virus entry (Holm et al. 2016). This pathway is impaired by the IAV HA2 fusion peptide directly binding to STING, thus preventing STING dimerization and TBK1 activation (Holm et al. 2016).

ISG15

Among all the ISG that are induced upon type-I IFN stimulation, interferon-stimulated gene 15 (ISG15) is one of the most rapidly up-regulated upon viral infection (Korant et al. 1984; Der et al. 1998). ISG15 is composed of two domains connected through a proline-containing linker that are structurally similar to ubiquitin (Narasimhan et al. 2005). This protein can be released from cells as an extracellular cytokine or proteolytically processed at the carboxy-terminal domain and further covalently conjugated into proteins through its lysine residues (D'Cunha et al. 1996). ISGylation of the target protein occurs through sequential enzymatic reactions similar to ubiquitin conjugation, involving E1 activating, E2 conjugating, and E3 ligase enzymes (Kim et al. 2004; Zhao et al. 2004; Krug et al. 2005; Dastur et al. 2006; Oudshoorn et al. 2012). Through the years, numerous proteins have been identified as targets of ISG15 (Giannakopoulos et al. 2005) even though the consequences of host protein ISGylation still remain poorly understood. ISG15 was first described as a molecule with antiviral activity, because ISG15-deficient mice become more susceptible to viral infection and show increased morbidity and mortality rates for both IAVs and influenza B viruses (IBVs) (Lenschow et al. 2007). Interestingly, ISG15 has been found to play different roles during IAV or IBV infection based on its ability to bind to NS1. For example, IBV NS1 is able to bind ISG15, inhibiting its interaction with its E1 enzyme and preventing the formation of ISG15 conjugates (Yuan and Krug 2001). On the other hand, ISGylation of NS1 of certain IAV strains by ISG15 conjugation impedes its interaction with importin-α, restricting nuclear import in detriment to viral replication (Zhao et al. 2010). In any case, although some NS1 of IBV do not bind directly to ISG15, the virus still can inhibit both the IFN response and ISG induction—including ISG15—during infection (Yuan and Krug 2001; Sridharan et al. 2010). This suggests that ISG15 has exerted a different evolutionary pressure of IBV or that IAVs have been able to adapt to such evolutionary pressure through a different mechanism. In any case, the antiviral effects of ISG15 in humans are unclear, as ISG15 in humans acts as a negative regulator of IFN signaling by stabilizing its deconjugating enzyme USP18, which in its turn inhibits IFNAR activation (Zhang et al. 2015; Honke et al. 2016).

Mx1

One of the first ISG that was described to specifically inhibit IAV replication was the Myxovirus resistance gene (Mx1) (Lindenmann 1962). Studies involving the mouse Mx1 locus revealed that the majority of conventional laboratory strains have large deletions or mutations that block proper expression of the Mx1 protein (Staeheli et al. 1986, 1988; Haller et al. 1987; Jin et al. 1998). Different Mx1 gene orthologs were described in different classes of vertebrates, including birds, reptiles, fish, and mammals, with different degrees of involvement with the antiviral response against IAV (Bazzigher et al. 1993; Robertsen 2006). Identification of the human Mx genes provided further insights in the evolution of Mx1 and its role in the innate immune response (Aebi et al. 1989). Being a dynamin-like GTPase protein, Mx1 can inhibit the replication of many viruses, including IAV (Haller et al. 2015). More details are needed to fully understand how this is accomplished, yet it is known that Mx1 interacts with the ribonucleoprotein complex of IAV and of other viruses, interfering with proper viral assembly and eventually blocking replication (Zimmermann et al. 2011; Verhelst et al. 2012). Interestingly, it has been shown that the sequence of NP, the most abundant component of the viral ribonucleoprotein complex, has been selected in human IAV to reduce the strength of binding, which has diminished the human Mx1-mediated antiviral effect (Zimmermann et al. 2011; Manz et al. 2013).

PKR

The protein kinase RNA-activated (PKR) is a serine/threonine protein kinase constitutively expressed at basal levels in mammalian cells and up-regulated by type-I and type-III IFNs (Meurs et al. 1990; Ank et al. 2006). PKR is a potent antiviral factor that becomes activated upon dsRNA binding during viral infection (Galabru and Hovanessian 1987). The best-characterized PKR cellular substrate is the α-subunit of eukaryotic initiation factor eIF2 (eIF2α). Upon viral infection, activated PKR phosphorylates eIF2α, causing translational shutdown of both viral and cellular genes (Su et al. 2006). Levels of PKR are also regulated at a post-transcriptional level. At high protein levels, activated PKR inhibits global translation initiation in order to autoregulate the expression of its own mRNA levels (Thomis and Samuel 1992). In this way, PKR can tailor the IFN response by repressing ISG translation in infected cells with IFN-activated PKR while elevating surveillance in uninfected cells in which PKR remains inactive even though its total protein levels are increased by IFN. The importance of PKR in the antiviral response is underscored by the plethora of viral PKR antagonists found in every virus family (Domingo-Gil et al. 2011). For IAV, the NS1 protein is known to directly bind to PKR, preventing its activation (Bergmann et al. 2000).

OASL

The ISG family of 2–5′ oligoadenylate synthetases (OASs) promote degradation of viral dsRNA through activation of RNase L (Kristiansen et al. 2011). Later studies unveiled additional antiviral mechanisms independent of RNaseL activity, including the inhibition of IAV replication by the human oligoadenylate synthetase-like (OASL) protein (Zhu et al. 2015). OASL is related to the OAS proteins because it contains the amino-terminal OAS-like domain but cannot induce RNaseL recruitment. Instead, OASL contributes to the antiviral activity by enhancing RIG-I activation (Zhu et al. 2014). In the absence of viral RNA, RIG-I remains inactivated in a folded conformation (O'Neill and Bowie 2011). When RNA is recognized by the carboxy-terminal domain, the amino-terminal caspase activation and recruitment (CARD) domains bind the K63-linked polyubiquitin complexes (pUb) via the E3 ubiquitin ligase TRIM25. This leads to a conformational change and activation of RIG-I, which starts the aggregation to MAVS via CARD, eventually leading to the IRF3 signaling cascade (Gack et al. 2009). Recent studies have shown that OASL can mediate RIG-I activation in the presence of viral RNA without TRIM25 (Zhu et al. 2014). This is accomplished by mimicking the action of pUb, thus binding to RIG-I directly without additional ligands. Although more details are needed to fully understand its role in the innate immune response, the interaction of OASL with RIG-I opens a new avenue of future antiviral therapies against viruses that are primarily sensed through this pathway.

IFITM

Several members of the interferon-induced transmembrane (IFITM) family of proteins are key factors in the inhibition of a wide variety of RNA and DNA viruses, including IAV (Brass et al. 2009; Perreira et al. 2013; Li et al. 2018). IFITM proteins are dynamically distributed in the cell membrane and in different organelles, such as lysosomes and endosomes (Bailey et al. 2014). Their mechanism of action is still not fully understood, yet many studies provide enough evidence to affirm that restriction of the infectious cycle by IFITM happens at early stages of the viral entry, by altering membrane fusion either at the cell membrane and/or during endosomal maturation (Perreira et al. 2013; Desai et al. 2014). So far, IFITM1, IFITM2, and IFITM3 have been described to possess antiviral activity, with different degrees of involvement depending on the virus and the site of viral fusion. In regard to IAV infection, IFITM3 appears to be the major contributor of the three to inhibit viral replication (Brass et al. 2009; Feeley et al. 2011; Huang et al. 2011; Li et al. 2013). Although there are many hypotheses as to how IFITM3 mediates this inhibition, recent studies have postulated that IFITMs recruit other cellular proteins to modulate viral fusion (Fu et al. 2017). Moreover, evidence suggests that IAV has already evolved to counteract the antiviral action of IFITMs by inhibiting their expression via p53 activation (Wang et al. 2018). Because of the potency of IFITM3 to restrict viral replication, natural variations of the IFITM genes within the human population can modulate the response and possibly contribute to higher sensitivity to infections (Zhao et al. 2019). A particular IFITM3 single-nucleotide polymorphism (SNP), rs12252, was found to be significantly enriched in hospitalized subjects because of IAV infection (Everitt et al. 2012). This confirms the importance of IFITM3 in the innate immune response and how any alteration can dramatically alter its antiviral action.

TRIMs

The superfamily of tripartite motif-containing (TRIM) proteins is made up of E3 ubiquitin-ligating enzymes that mediate the last of the three sequential enzymatic reactions necessary for protein ubiquitination. Ubiquitination is a post-translational modification (PTM) of proteins required for diverse cellular processes such as cell cycle progression, transcriptional regulation, and protein degradation by the proteasome, which ultimately consists on allowing reversible ubiquitin linkage to a substrate to activate or deactivate a specific target function (Morreale and Walden 2016). TRIM proteins are structurally conserved and they contain a conserved RBCC domain, which includes a RING (R) domain, one or two B-boxes (B), and a coiled-coil (CC) domain (Ebner et al. 2017; Esposito et al. 2017). Several TRIM proteins are known to be induced by IFN and some TRIM family members have been reported to limit viral replication acting indirectly as components of the innate immune pathway. Such is the case of TRIM56, which efficiently targets both IAV and IBV and elicits an antiviral function through a nondegradative mechanism (Liu et al. 2016). Upon infection, a portion of TRIM56 relocates from the nucleus to the cytoplasm and its presence leads to a reduction of both viral transcription and cRNA synthesis, suggesting that TRIM56 may target a component of the vRNP or the viral RNA itself. Although the specific mechanism of restriction for TRIM56 is still poorly understood, its antiviral activity does not rely on its conserved RBCC domain but on a specific region located on the carboxy-terminal tail. In addition to this, other TRIMs directly target viral proteins for proteasomal degradation. For instance, TRIM22 is significantly up-regulated upon IAV infection or IFN treatment and directly targets viral NP for polyubiquitination and further proteasomal degradation (Di Pietro et al. 2013; Lian and Sun 2017). Similarly, TRIM14 has been found to target NP for polyubiquitination and proteasomal degradation through its PRY-SPRY domain (Wu et al. 2019). Contrary to its role in IFN signaling, TRIM14 does not require IFN production to inhibit IAV replication, because TBK1^−/− and IFNAR^−/− cell lines do not prevent TRIM14-mediated restriction of IAV. Other TRIMs have been recently reported to exert an antiviral role against IAV without requiring any prior IFN induction. Such is the case of TRIM41, which can also target viral NP through its SPRY domain for subsequent polyubiquitin-mediated degradation (Patil et al. 2018). Overall, these findings show that TRIM proteins can redundantly target NP in order to orchestrate and mount an efficient antiviral response. However, even though NP is a major target for TRIM proteins, other IAV components have been described to be TRIMs substrates. For instance, TRIM32 is a novel PB1 host interactor that is conserved among many different IAV strains. TRIM32 selects PB1 for polyubiquitination and proteasomal degradation leading to decreased polymerase activity and viral replication (Fu et al. 2015). This process does not require prior IFN induction, thus allowing early target and detection of PB1 that is present in uncoated viral particles.

ESCAPING THE HOST INNATE IMMUNE RESPONSE

Despite having all the mechanisms ready to counteract external pathogens, the host innate immune response is not always successful in preventing viral infections, including IAV. Selective pressure has pushed the immune system to develop more effective strategies that confer protection. But at the same time, viruses like IAV have adapted to overcome these new host barriers, by modulating the innate immune response and thus facilitating their replication. Several IAV proteins have been reported to inhibit different components of the host immunity.

NS1

The IAV nonstructural protein NS1 is nowadays considered the main antagonist of the host innate immune response during viral replication (Greenspan et al. 1988; García-Sastre et al. 1998; Kochs et al. 2007). It is encoded by the smallest IAV RNA segment in conjunction with NEP, a spliced variant of the NS mRNA (Baez et al. 1980). Several functional domains of NS1 confer a wide range of interaction mechanisms with cellular and viral proteins, in order to benefit IAV replication. The amino terminal end of NS1 contains a RNA-binding domain (RBD) that allows its interaction with dsRNA molecules, blocking the activation of RLR sensors such as PKR and OASL (Liu et al. 1997). This mechanism was initially described to be the main function of NS1 to antagonize the host immune response, yet further analysis revealed several other interactions that contribute to this goal (Klemm et al. 2018). NS1 directly interacts with RIG-I to inhibit IFN induction (Mibayashi et al. 2007; Jureka et al. 2015), as well as with TRIM25 and RIPLET to block RIG-I activation (Gack et al. 2009; Rajsbaum et al. 2012). Moreover, NS1 also interacts with PKR, thus preventing IFN activation (Li et al. 2006a; Min et al. 2007). The RBD domain of NS1 contains a nuclear localization signal (NLS) and a selection of viral strains also contain an additional NLS in the carboxy-terminal region (Melén et al. 2007). The NS1 also contains nuclear export signals (NESs) and nucleolar localization domains (Hale et al. 2008). This contributes to the dynamic and diverse localization of NS1 during viral replication both in the nucleus and in the cytoplasm. Because of this, NS1 can also interact with other cellular components, both in the cytoplasm and in the nucleus. NS1 interacts with the 30-kDa subunit of the cleavage and polyadenylation specificity factor (CPSF30) to inhibit cellular mRNA processing and thus inhibit the host antiviral activity (Nemeroff et al. 1998; Noah et al. 2003). This inhibitory activity also promotes host transcriptional antitermination and chromatin remodeling in infected cells (Twu et al. 2006; Bauer et al. 2018). NS1 also targets the inhibitor of kappa B kinase (IKK), eventually inhibiting the nuclear factor kappa-light-chain-enhancer of activated B cells (NK-κB) pathway and therefore preventing the expression of several antiviral genes (Gao et al. 2012). Interestingly, specific IAV strains contain a histone-like sequence in the NS1 protein that suppresses the host antiviral response by binding to the RNA polymerase-associated factor 1 complex (PAF1C) (Marazzi et al. 2012). Other NS1 activities that contribute to overcome the antiviral response include the inhibition of nuclear export of host mRNA (Zhang et al. 2019), as well as activation of transcriptional repressors, such as the CCAAT/Enhancer Binding Protein beta (C/EBPβ), which NS1 recruits to act as a repressor element in the RIG-I promoter (Kumari et al. 2020). Moreover, almost every NS1 protein contains a carboxy-terminal binding motif that targets class 1 PDZ domains (Obenauer et al. 2006), which are mediating protein–protein interactions in a wide variety of cellular pathways (Fanning and Anderson 1999). This PDZ domain binding motif (PBM) in the NS1 protein has been suggested to be a virulence determinant, because carboxy-terminal truncations or extensions result in attenuation of the virus (Li et al. 2006b; Soubies et al. 2010).

The mechanisms by which IAV NS1 modulates the cellular machinery to promote viral replication include such a wide range of pathways that highlights its relevance as a multifunctional virulence factor. Because of the intrinsic mutational rate of IAV during replication and its ability to infect different hosts, NS1 is in constant evolutionary pressure. Our recent research efforts have analyzed the NS1 evolutionary landscape and host tropism using a barcoded NS1 library of recombinant NS1 IAV, providing new insights on the plasticity of the protein and its ability to adapt to different hosts (Muñoz-Moreno et al. 2019). Following studies may describe new details about NS1 evolution, discovering new interactors to advance our understanding of how IAV counteracts the host innate immunity. In any case, NS1 mutated recombinant IAV represent potential novel live attenuated vaccines against human IAV (Wang et al. 2019a), and they are currently being used in the field as vaccines in pigs against swine influenza because of their ability to induce broad protective responses against multiple swine influenza virus strains (Genzow et al. 2018).

Viral Polymerase

The IAV polymerase also plays a role in regulating the host antiviral response. Early after viral infection, when a reduced number of viral components are still present in infected cells, the viral polymerase complex can inhibit IFN-β promoter activity in a RIG-I- and MAVS-dependent manner. This inhibition of the type-I IFN response does not compete with the function of NS1, being more efficiently promoted by the PB2 polymerase subunit (Graef et al. 2010; Iwai et al. 2010). In previous years, the role of the viral polymerase in inhibiting the IFN response had been partially supported through studies using UV-inactivated IAV (Marcus et al. 2005) and through systems biology analysis to elucidate novel viral–host interactors (Shapira et al. 2009). Recent studies in the field have shown that the IAV PA protein also interacts with IRF3 to suppress IFN-β production (Yi et al. 2017). The amino-terminal endonuclease activity of PA is required for this interaction, revealing a new strategy by which IAV blocks IFN-β signaling. It is also worth mentioning that some IAV proteins—including the viral polymerase—are known virulence and pathogenicity contributors. Therefore, it is not surprising that viruses carrying highly efficient polymerases have increased fitness abilities that can better outcompete the host antiviral response (Grimm et al. 2007).

PA-X

The polymerase acidic (PA) protein is the canonical product translated from IAV RNA segment 3. However, an alternative open reading frame (ORF) was discovered to produce a highly conserved fusion protein among strains, called PA-X (Jagger et al. 2012). Initial studies revealed PA-X to be influencing the host immune response, thus affecting viral replication and pathogenicity (Gao et al. 2015; Hu et al. 2018). However, its action seems to be strain-specific and the reasoning behind it remains elusive. In any case, PA-X is known to cleave host transcripts in the nucleus, degrading them via the host RNase Xrn1 and facilitating the processing of IAV mRNAs (Khaperskyy et al. 2016).

PB1-F2

Segment 2 of the IAV genome encodes for the RNA polymerase basic protein 1 (PB1), which is a subunit of the viral RNA polymerase complex. There is an alternative ORF within segment 2 of specific IAV strains that produces a second protein, called PB1-F2, originally described as a promoter of cell apoptosis, localized in the mitochondria of immune cells (Chen et al. 2001). Increased levels of reactive oxygen species (ROS) have been associated with the inhibition of the superoxide anion dismutase 1 (SOD1) by PB1-F2 (Shin et al. 2015). Further studies revealed a specific point mutation in the PB1-F2 sequence of highly virulent H5N1 IAV strains, which increases its pro-apoptotic properties and thus its viral pathogenicity (Conenello et al. 2007). Moreover, natural killer (NK) cell inhibition and neutrophil recruitment have been associated with the presence of PB1-F2, which also exacerbates pathogenicity (Vidy et al. 2016). Finally, PB1-F2 is also inhibiting the IFN response by modulation of MAVS through the host protein calcium-binding and coiled-coil domain 2 (NDP52) (Varga et al. 2011; Leymarie et al. 2017).

M2

The influenza virus matrix protein 2 (M2) is a small proton ion channel transmembrane protein encoded in segment 7 through alternative splicing that is important in several steps of IAV replication cycle. During viral entry, the low pH in endosomes activates the ion channel activity of M2, pumping protons toward the interior of the virion and promoting the release of vRNP into the cytoplasm after membrane fusion (Pinto et al. 1992). In addition, the ion channel activity of M2 also regulates the pH balance between the cytoplasm and the acidic lumen of the trans-Golgi network (TGN), in order to protect HA from pH-induced premature conformational changes (Ciampor et al. 1992; Sakaguchi et al. 1996). To date, how IAV M2 protein can evade the host immune response remains poorly understood. However, recent studies have shown that M2 is able to anchor to mitochondria in both virus-infected and M2-overexpressing cells, accelerating MAVS self-association and aggregate formation, thus promoting a MAVS-mediated antiviral immunity (Wang et al. 2019b). Moreover, M2 can increase levels of Ca⁺² into the cytoplasm, ultimately resulting in ROS production and autophagy activation. Such autophagy stimulation leads to MAVS-mediated innate immune signaling inhibition and MAVS aggregates regulation though direct binding of MAVS to ATG5 (Jounai et al. 2007; Zhao et al. 2012) and LC3B (Sun et al. 2016; Cheng et al. 2017). Meanwhile, M2 can also block the last step of the autolysosome formation during autophagy, reducing excessive elimination of MAVS aggregates and ROS (Wang et al. 2019b) and preventing the antiviral consequences of autophagy (Rossman and Lamb 2009; Zhang et al. 2014). In summary, M2 can be defined as a tight regulator of the innate immune response that is able to trigger MAVS signaling activation while preventing excessive activation of RLR signaling and inhibiting antiviral autophagic processes.

FINAL REMARKS

There is a complex and dynamic interplay between IAVs to efficiently promote virus replication and the host innate immune response to counteract and fight against viral infections. The molecular basis of these virus–host interactions has been extensively investigated over the last two decades, providing further insights into the mechanisms that the host cell uses upon IAV infection to efficiently control the disease outcome. Given the high diversity of IAV and the relatively conserved nature of the host innate immune response, further research efforts in this field are needed to provide a deeper understanding and to enable the development of more effective antiviral treatments that could be used against multiple IAV strains/subtypes and the prevention of future pandemics.