1.1. IAV Viral Proteins

HA is a type I glycosylated protein, which is responsible for virus entry to host cell. Functional HA protein is a homotrimer structurally composed of a stem region and a globular head region in each monomer. The head region bearing N-acetylneuraminic acid (sialic acid, SA) binding pocket is critical for receptor attachment, and contains most antigenic determinants. The stem region undergoing conformational changes is responsible for low pH-triggered membrane fusion [10], and plays an important role in cross protection against heterosubtypic IAV infection [11]. N38 glycan at this region is critical for elicitation of cross-group antibody responses [12]. HA of diverse IAV subtypes that originated from different species presents distinct receptor-binding preference. For instance, human viruses prefer binding to SAs attached to cell-surface-associated α-2,6-linked galactose, whereas avian viruses prefer α-2,3-linked galactose [13,14,15]. Residue substitutions in the receptor-binding site (RBS) of HA is crucial in determining receptor-binding properties [16]. For instance, amino acid substitutions of S138A/G186V/T221P/Q226L within the RBS affected receptor-binding properties of avian H7N9 HA [17], while G186V was reported to be pivotal for the avian-specific strain to acquire human receptor-binding capacity [18].

NA is a type II glycoprotein with neuraminidase (sialidase) enzymatic activity. Each NA tetramer consists of four identical polypeptides, and each polypeptide contains an N-terminal, a hydrophobic membrane domain, a stalk region, and a globular head domain. NA can cleave SA from the mucus, cell surface, and from viral glycoproteins. While HA mediates virion-SA attachment and fusion, NA is responsible for terminal SA residues cleavage [19]. N-glycolyl and O-acetyl modification of SA could reduce binding affinities of both NA and HA [20]. In addition, NA possesses at least two calcium binding sites [21]. Gene analysis of these Ca²⁺ binding sites reveals that they are related to NA thermostability, further suggesting a correlation between NA thermostability and virus adaption [22]. Furthermore, NA is also the major antigenic target of the host humoral immunity, and NA-specific antibodies function in limiting virus egress via interfering with the sialidase activity have drawn wide attention for development of antiviral therapies [23,24].

The viral ribonucleoprotein complex (vRNP) is a rod-shaped structure composed of multiple copies of NP and a single trimeric RNA-dependent RNA polymerase complex (PB1, PB2, and PA) associated with viral genomic RNA [25]. NP mediates nuclear import of the vRNP complex, the PB1 subunit has the catalytic polymerase activity, the PB2 subunit contributes to cap binding, and the PA subunit is required for cleavage of the capped oligonucleotides. The complex is required for the transcription and replication of the viral genome [26]. The structure, functions, and modulation of the IAV RNA polymerase complex were further discussed by Te Velthuis and Fordor [27]. However, the mechanism of vRNP assembly remains largely unknown and several host proteins were reported to be involved during IAV infection [28,29,30]. Recently, eleuthe roside B₁ was demonstrated to be able to inhibit the vRNP in vitro [31].

M1 and M2 are encoded by the M gene on segment 7 of the viral genome [32]. The conserved M1 bears a positively charged nuclear localization sequence (NLS) motif RKLKR, which is essential for membrane binding, virus replication, nuclear export of the vRNPs, and self-polymerization [33]. M2 is a type III integral transmembrane proton channel, structurally containing three domains: an amino-terminal ectodomain, a transmembrane domain, and a cytoplasmic domain [34]. It is essential for uncoating the vRNP complex within the endosome, the process of which is essential for viral entry [35]. Moreover, in its retrograde migration route from the membrane to the cytoplasm, M2 was shown to prevent endosome fusion with the lysosome, suggesting new roles of the protein in the host cytoplasm after infection [36].

NS1, structurally composed of an N-terminal dsRNA binding domain and a C-terminal effector domain, plays important roles in control of viral RNA synthesis, viral mRNA splicing, and virus particle morphogenesis during the virus life cycle [37,38]. It is also considered as the key factor to antagonize host immune system [39], since it could inhibit activation of retinoic acid-inducible gene-I proteins (RIG-I), interferon-inducible RNA-dependent protein kinase (PKR), and 2′-5′-oligoadenylate synthetase of the innate immunity via blocking nuclear export of host messenger RNAs (mRNAs) [40,41,42]. The mechanism underlying is revealed by determination of crystal structure of the NS1 and nuclear RNA export factor 1-nuclear transport factor 2-related export protein 1 (NXF1-NXT1) complex that NS1 could prevent mRNA export through the nuclear pore complex via binding of NXF1-NXT1 to nucleoporins [43]. Conserved RNA secondary structures at RNA positions 82-148 and 497-564 are essential for NS1 protein expression, thus affecting viral reproduction and virus–host interaction processes [44]. A detailed description of structure and function of the NS1 was reviewed by Han and his colleagues [45]. NS2, also known as NEP, contains a nuclear export signal and is responsible for chromosome region maintenance 1 (Crm1)-dependent nuclear export of RNPs in the infected cells [46,47]. Crm1 contains leucin-rich nuclear export signals (NES1 and NES2) [47,48], which are important for the NS2Crm1 interaction [46,48]. In addition, interaction of CHD3 and NES1 was identified to be important in the NS2-mediated vRNP nuclear export [49].

3.1. Immune Cells Involved in IAV Infection

Upon detection of infection, innate effector cells, including natural killer (NK) cells, neutrophils, and dendritic cells (DCs), etc., are recruited to the infected sites. NK cells are large granular lymphocytes, making up 10% of the resident lymphocytes in the lung. After recruitment from the blood, NK cells interact with DCs and macrophages to secret various cytokines and restrict infection via lysis of the IAV-infected cells. The lysis process is mediated by interaction between NK receptors p46 (most NKp46) and IAV HA protein expressed by the infected cell [116,117]. Interestingly, liver NK cells other than lung NK cells possessed a memory phenotype to protect mice against subsequent IAV infection, though the lung NK cells are important in control of primary IAV infection [118]. However, NK cells are also shown to exacerbate IAV pathology, since depletion of NK cells led to increased resistance to high dose H1N1 infection in mice [119,120]. The contribution of NK cells to anti-IAV defense in mouse models was later shown to be strain and dose dependent. In addition, the host genetic background also played an important role [121].

Neutrophils are key innate immune cells recruited to infection sites by cellular migration through vascular endothelium. They function in clearance of pathogens via phagocytosis, producing extracellular traps, and degranulation [122]. In addition, they also regulate adaptive immunity via guiding influenza specific CD8⁺ T cells to the infection sites [123].

The function of dendritic cells (DCs) is to monitor invading pathogens. After IAV infection, the conventional DCs migrate from lung to lymph nodes through interaction between CCR7 and its ligand, and present antigens to T cells [124,125]. One study based on a mouse model showed that, during IAV infection, immature and mature DCs were specialized in IAV HA processing, since both types of DCs could present one epitope of H1N1 HA (HA amino acids 107–119), whereas another epitope (HA amino acids 302–313) could only be processed by mature DCs [126]. The complex role of DCs in initiation of robust immunity against IAV infection is reviewed by Waithman and Mintern [127].

T cells and B cells are critical components in adaptive immunity against IAV infection. CD8⁺ T cells differentiate into cytotoxic T lymphocytes (CTLs) and defend IAV infection via producing cytokines and effector molecules, and cytotoxic effects (i.e., lysis) of infected cells mediated by MHC class I. CD4⁺ T cells target IAV-infected epithelial cells through binding with MHC class II molecules and contribute to B cell activation thus consequently promote antibody production. The activation of T cells and B cells in IAV infection will be exposited in Section 3.5.

3.2. Activation of Innate Immunity in IAV Infection

The reaction of innate immunity is nonspecific. It is triggered by recognition of pathogen associated molecular patterns (PAMPs) via host pathogen recognition receptors (PRRs). Toll-like receptors (TLRs), retinoic acid-inducible gene-I proteins (RIG-I), and NOD-like receptors are common PRRs, the activation of which leads to activation of innate immune signaling and further production of cytokines as well as other antiviral molecules.

Toll-like receptors are responsible for sensing pathogens at cell membranes, endosomes, and lysosome [128]. TLR3 and TLR7 are shown to be involved in IAV detection at endosomes [105]. TLR3 recognizes double stranded RNA (dsRNA) which may be released by cellular stress and cell death [105] and unidentified RNA structures in phagocytosed cells infected with IAVs [129]. In macrophages and dendritic cells, TLR3 interacted with TIR-domain-containing adapter, then activated the serine-threonine kinase IκKε (IKKε) and TANK binding kinase 1 (TBK1) to phosphorylate interferon regulatory factor 3 (IRF3), the process of which further led to expression of IFN-β [130]. In addition, an over-reacting TLR3 activation promoted IAV pathogenesis, which could be reduced by a single-stranded oligonucleotide (ssON) functioning as a TLR3 inhibitor, resulting in restrained viral loads both in vitro and in vivo [131]. TLR7 recognizes single stranded RNA (ssRNA). In plasmacytoid dendritic cells (pDCs), after activation of TLR7 during IAV infection, IRF7 or nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) were activated via myeloid differentiation factor 88 (MyD88) to induce type I IFNs [132]. In avian macrophages, activation of TLR7 produced pro-inflammatory molecules such as interleukin (IL)-1β [133]. In addition, in mouse models, TLR7 played an important role in activation of NK cells [134]. It was also shown to be involved in development of adaptive immunity to prevent IAV infection [135,136].

RIG-I recognizes ssRNAs and transcriptional products of IAVs, which triggers activation of the caspase activation and recruitment domains (CARDs) via dephosphorylation or ubiquitination by E3 ligases, resulting in activation of transcription factors including IRFs and NF-κB [137]. OTUB1 played an essential role in regulation of RIG-I [138]. In addition, melanoma differentiation-associated gene 5 (MDA5) was also involved in sensing transcriptional products of IAVs in the cytoplasm [139].

For NOD-like receptor family, pyrin domain containing 3 (NLRP3) and NLR apoptosis inhibitory protein 5 were activated after IAV infection [140]. IAV M2 ion channel and PB1-F2 were involved in activation of NLRP3 inflammasome and stimulate IL-1β secretion subsequently [141,142]. The role of the NLRP3 inflammasome in regulation of anti-IAV responses is discussed in detail by Sarvestani and his colleagues [143]. Delayed oseltamivir and sirolimus combined treatment could suppress NLRP3 inflammasome mediated secretion of IL-1β and IL-18, resulting in attenuation of H1N1-induced lung injury [144].

3.3. The Host Interferon (IFN) Response in IAV Infection

After detecting viral components, transcription factors including NF-κB and IRFs are activated, leading to transcription of IFNs and pro-inflammatory cytokines. IFNs bind to receptors, resulting in upregulation of multiple interferon-stimulated genes (ISGs) [145]. It is well known that type I IFNs (IFN-α and IFN-β) and type III IFNs (IFN-λ 1-4) play critical roles in antiviral responses. Mice failed to restrict non-pathogenic IAV when both type I and type III IFN receptors were knocked out [146].

The expressed IFNs consequentially bind to different receptors. Type I IFNs interact with IFN-α/β receptors (IFNAR), whereas type III IFNs interact with IFN-λ receptors (IFNLR). Janus kinase-signal transducer and activator of transcription (JAK-STAT) signaling pathway is then activated, resulting in transcription of numerous IFN-stimulated genes (ISGs) [147,148]. Though IFN-λs share many characteristics such as expression patterns, signaling pathways, etc. with type I IFNs, they are the first IFNs produced at the infected epithelial sites to block virus spread [149]. Furthermore, IFN-λs served an important role in programming DCs to direct effective T cell immunity against IAV infection [150].

ISGs encode various antiviral proteins functioning in different ways to defend IAV infection. For instance, MxA GTPase from the Mx family could retain viral genome from entry to the cytoplasm via blocking the function of IAV NP. In addition, in vitro research found that avian IAVs were more sensitive to MxA than human IAVs [151,152]. Cholesterol 25-hydroxylase (CH25H) were identified to block IAV entry via altering the cellular membrane properties to interfere with viral fusion, and amplified the activation of immune cells [153]. Guanylate-binding protein 3 (GBP3) of IFN-inducible GTPases inhibited IAV replication via binding to the viral polymerase complex [154]. Members of the tripartite motif-containing (TRIM) family are also involved in cellular anti-IAV processes. For instance, TRIM14 could interact with IAV NP for ubiquitination and proteasomal degradation, thus restricting IAV replication in a type I IFN and NF-κB independent manner [155]. TRIM22 degraded IAV NP via polyubiquitination, thus resulting in inhibition of IAV infection [156]. TRIM 25 regulated the re-localization of RIG-I and was responsible for RIG-I ubiquitination as well as RIG-I-mediated IFN production [157]. TRIM32 recognized IAV PB1 protein and reduced its polymerase activity [158]. TRIM41 targeted NP for ubiquitination and degradation in vitro [159]. For further reading on other ISGs, several reviews regarding IFN responses during IAV infection are recommended here [160,161]. A general description of activation of innate immunity and IFN signaling pathway after IAV infection is illustrated in Figure 1.

Open in a separate window

Figure 1

Schematic diagram for activation of innate immune responses and interferon (IFN) signaling pathway after influenza A virus (IAV) infection. Intracellular detection conducted by pathogen recognition receptors (PRRs), including retinoic acid-inducible gene-I proteins (RIG-I), melanoma differentiation-associated gene 5 (MDA5), and toll-like receptors (TLRs), activate transcription factors such as IRFs and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), resulting in expression of IFNs and interferon-stimulated genes (ISGs).

In order to counter IFN-stimulated antiviral proteins, IAV viral proteins apply multiple strategies. For instance, HA protein was shown to trigger ubiquitination of IFNAR to attenuate the type I IFN signaling pathway [162]. The follow-up work showed that poly (ADP-ribose) polymerase 1 (PARP1) functions as an interacting partner of HA protein to mediate the HA-induced IFNAR degradation [163]. NS1 is the most important IFNs antagonist protein via mechanisms including inhibition of the TRIM25-mediated RIG-I ubiquitination, suppression of protein kinase R (PKR), phosphorylation of IκB kinases (IKK) α and β in the NF-κB pathway, interruption of the phosphorylation of STAT1, STAT2, and STAT3 [39,115], and degradation of OTUB1 [138]. Phosphorylation of NS1 is crucial for its function of antagonizing IFN-β expression, since dephosphorylation at position 73 and 83 of the protein induced a high level of IFN-β [91]. Non-structural protein PB1-F2, identified from a+1 open reading frame (ORF) of PB1 gene segment [164], is multifunctional in deregulation of type I interferon [165,166]. It counteracted RLR-mediated activation of IFN pathway not only by targeting mitochondrial MAVS [115,165,167], but also by binding to the DEAD-box helicase DDX3 to induce proteasome-dependent degradation [166]. Furthermore, PB1-F2 interacted with mitochondrial Tu translation elongation factor (TUFM) to mediate formation of autophagosome, thus inducing complete mitophagy, which is critical for MAVS degradation [167]. Novel PA-X protein could also modulate innate immune responses. A review regarding the function of NS1 and PA-X proteins in antagonizing host innate immunity is recommended here [114].

3.4. Autophagy

Though autophagy is essential for cellular metabolism and homeostasis, it also plays important roles in innate immune responses against pathogen infection. For cellular homeostasis, the mTOR pathway is one of the most conserved autophagic pathways. The mTOR complex 1 (MTORC1) negatively regulates the ULK1 kinase activity, thus affecting the autophagy induction [168]. c-Jun N-terminal protein kinase 1 (JNK1) disrupts the Bcl-2/Beclin-1 complex through phosphorylation, thus regulating the autophagy induction [169,170]. JNK1 is also reported to upregulate Beclin-1 expression through phosphorylation of transcription factor c-Jun in vitro [171].

In contrast to the autophagic pathways for cellular metabolism and homeostasis, less is known about autophagosome formation after IAV infection [172]. To restrict infection of multiple viruses including IAVs, TRIM23 is essential to mediate autophagy via its RING E3 ligase and ADP-ribosylation factor (ARF) GTPase activity [173]. Beclin-1 and TUFM-regulated autophagy also inhibited IAV replication [112]. In HeLa cells and A549 cells, IAV infection activated JNK1 to induce autophagosome formation and TGF-β-activated kinase 1 might contribute to the process [174,175]. Furthermore, autophagy was involved in maintaining memory B cells to counteract IAV infection [176].

IAV also utilizes autophagy to complete its life cycle. NS1 protein is proposed to suppress JNK1-mediated autophagy induction [174]. M2 could also block autophagosome maturation and mediate microtubule-associated protein 1 light chain 3 (LC3)-bound membrane redistribution, thus allowing filamentous budding of IAV [177,178,179]. Circ-GATAD2A (GATA zinc finger domain containing 2A), induced by IAV infection, could inhibit autophagy and promote IAV replication [180]. For a comprehensive reading on IAV-induced apoptosis, a review is recommended here [181].

3.5. Adaptive Immunity against IAV Infection

Upon detection of IAVs, DCs trigger production of IFNs and cytokines, which in turns assist maturation of the DCs into antigen presenting cells (APCs), and initiate T cell immune responses. Through the activation of Ag-bearing DCs, naïve CD4⁺ T cells differentiate into Th1, Th2, Th7, regulatory T cells (Treg cells), follicular helper T cells, and killer cells. Th1 and follicular helper T cells are the most abundant CD4⁺ T helper cells. They can secret antiviral cytokines, regulate CD8⁺ T cell differentiation, promote B cell activation, and maintain immunological memory [182,183]. Th17 cells induced pulmonary pathogenesis and could decrease mortality of IAV-infected mice [184,185]. In addition, γδ T cells, expanding in the late stage of IAV infection with a T cell receptor (TCR)-independent manner, could efficient eliminate IAV-infected airway epithelial cells, resulting in lower viral titers [186]. New surrogate markers CD49d and CD11a were used to explore the kinetics of IAV-specific CD4 T cells responses, revealing endogenous CD4 T cell response to primary IAV infection is predominantly composed of T-bet+ cells [187].

CD8⁺ T cells are major components for virus clearance in adaptive immunity. After activated by DCs, CD8⁺ T cells undergo rapid expansion, differentiation, and migration to the infected sites. In general, to establish effective primary cytotoxic T lymphocyte (CTL) responses, CD4⁺ T cells play an essential role, with a mouse model as an exception [188]. CTLs produce cytotoxic granules containing perforin and granzymes (GrA and GrB) to induce apoptosis and interrupt IAV replication [189]. In addition, CTLs produce cytokines, such as TNF, FASL, and TRAIL, which recruit death receptors to induce apoptosis [190]. In addition, IL16 deficiency enhanced the Th1 and CTL responses upon IAV infection [191]. Furthermore, as CD8⁺ cells could last for two years in murine models, IAV-specific memory CTLs reacted specific to epitopes in conserved IAV proteins [192]. In the nasal epithelia, they could prevent the spread of the virus from the URT to the lung [193]. To establish memory CD8⁺ T cells, autophagy plays an important role [194], while the function of CD4⁺ T cells in memory CTL responses is “context-dependent”. A recent study showed that CD4+ T cells promoted IAV-specific CTL memory at the initial priming stage of viral infection [195]. Grant and her colleagues summarized and discussed the importance of CD8⁺ T cell immunity against IAVs [192], and this review is recommended for further reading.

With the help of CD40 ligand (CD40L), CD4⁺ cells contribute to B cell activation [183]. With the help of memory T cells, naïve B cells could reduce morbidity and promote recovery on heterosubtypic infection [196]. For different types of antibodies, IgG could inhibit pathogenesis, while IgA functions in blocking IAV transmission [197]. In addition, IAV-specific antibody-dependent cell-mediated cytotoxicity (CDCC) also plays a role in cross-protection against IAV infection. A general description of adaptive immunity against primary IAV infection is illustrated in Figure 2.

Open in a separate window

Figure 2

Schematic diagram for adaptive immune responses against primary IAV infection.

Antigenic shift and drift, resulting in reassorted and mutated HA and/or NA, are responsible for AIV escaping from host immunity [50,51,52]. Furthermore, additional glycosylation on H5 HA could also induce virus escape from neutralizing antibodies [198].