Ducks are reported to be relatively resistant to common poultry diseases, including viral disease, compared to chicken (Pal et al., 2017), and are commonly asymptomatic to avian influenza virus infection (Kim et al., 2009; Fleming-Canepa et al., 2019; CDC, Centre for Disease control and Prevention, 2021). There is clear lack of further systematic characterization of the indigenous ducks at the molecular level. In an effort to understand, we have studied this as a first step. Hence, there is an urgent need to explore the innate immune response genes, particularly against viral infection.

Avian influenza is caused by single-stranded RNA virus, which is negatively stranded, and belongs to Orthomyxoviridae family (WHO, 2019). It is commonly known as bird flu since birds are the main host. Based on the antigenic differences, two surface proteins, namely, hemagglutinin and neuraminidase of avian influenza virus have been mostly subtyped and the nomenclature is provided accordingly. Till date, 18 subtypes of HA (H1–H18) and 11NA (N1–N11) have been detected (Tong et al., 2013; Ying et al., 2014). H5, H7, and H9 were observed to be the most pathogenic subtypes of bird. Most of the H5 and H7 subtypes were regarded as highly pathogenic avian influenza (HPAI) virus, owing to the higher incidence and mortality of birds. The greatest concern is the lack of definite treatment or vaccination due to frequent mutation and reassortment of viral strain, regarded as antigenic shift and antigenic drift (Webster and Govorkova, 2014; WHO, 2020). Due to massive culling of birds in the affected area and the ban on the export of poultry products, the WHO has regarded avian influenza as one of the most economically effected zoonotic diseases (WHO, 2018b). It has been reported that in West Bengal during the 2008 outbreak, there was loss of 500.42 crores, 6 percent of the poultry population was culled, and five lakh families were affected. In Manipur at the 2007 outbreak, the loss due to the disease has been found to be 14 per cent of the total value of livestock outputs in the entire state. More than three lakh birds were culled, and 24 tonnes of poultry feed was destroyed post-flu (Otte et al., 2008a). Vietnam estimated the direct losses of 109 village and backyard producers with flock sizes smaller than 50 birds at US$69 (VND1084000) per farm (Otte et al., 2008a). An average loss of US$22 was estimated per household from the loss of birds in Egypt (Otte et al., 2008a).

The basic mechanism of host immunity against viral infection is generally different from that of other infectious agents such as bacteria and protozoa. Viruses utilize the host immune mechanism for their infection and further survival, thus allowing it to act as hijackers. Accordingly, viruses employ the host cellular machinery for living normal cells through the process of invasion; multiplication within the host might in turn kill, damage, or change the cells, and make the individual sick (Maarouf et al., 2018). The molecular mechanism of replication of avian influenza virus involves certain proteins. The HA protein present from the surface of the AI virus aids in recognition and binding to sialic acid on the surface of host cells, thereby aiding in the entry of the virus in the host cell (Matrosovich et al., 2009a). Following the binding, virus particles are endocytosed, leading to endosome maturation, and pH is lowered, resulting in the conformational change in HA, thereby causing fusion of the endosome and virion membranes. The viral M2 protein (matrix 2) acts as an ion channel for further lowering of the pH of the viral particle. This leads to the dissociation of the M1 protein (matrix 1) virion “shell” in such a way that the eight vRNPs [NP (nucleoprotein)-coated and polymerase complex (PB1, PA, and PB2)-bound viral RNAs] are released into the cytosol (Basler and Aguilar, 2008a). In the next step, the viral RNPs transport into the cell nucleus wherein accessory cellular components necessary for influenza viral replication and transcription are present. Following the process of genome replication, transcription, and protein synthesis, NEP (nuclear export protein) and M1 act to traffic newly synthesized vRNPs out of the nucleus, into the cytoplasm, and to the plasma membrane, leading to the assembly of progeny virions. At this stage, several viral proteins contribute to budding, including M1 and M2. In the next step, NA aids in the removal of sialic acid from glycoproteins in both the viral and cell membranes, resulting in the prevention of the interaction between HA and host cell receptors, and release of new infectious virus particles (Medina and García-Sastre, 2011). NS1 (nonstructural protein 1) plays a role within the infected cell in order to counteract innate host–cell defense systems, for example, interferon (IFN), which may otherwise limit efficient virus replication (Hale et al., 2008).

As the virus gets an entry in the body, an immune response is triggered, followed by local inflammatory signaling. Innate immune reaction is initially activated by conserved pathogen-associated molecular pattern (PAMP), pattern recognition receptors (PRRs), retinoic acid–inducible gene (RIG)-I like receptors, MDA5, LGP 2, and toll-like receptor (TLRs) such as TLR3 and TLR7 (Kannaki et al., 2010). Viral nucleic acid binds to these receptors expressed on macrophages, microglia, dendritic cells, and astrocytes; releases type-I interferon (IFN-I); and helps in the production of interferon-stimulated genes (ISGs) (Ali et al., 2019). Interferon-I upregulates antiviral proteins, and accordingly, peripheral immune cells are stimulated and alter endothelial tight junction (Otte et al., 2008b). It has been observed that the absence of IFN_I signaling leads to the prevention of microglial differentiation and decrease of peripheral myeloid cell patrolling (Otte et al., 2008b).

TLR7 is a member of the Toll-like receptor family, which recognizes single-stranded RNA in endosomes, which is a common feature of viral genomes (Manglani and McGavern, 2018). TLR7 can recognize GU-rich single-stranded RNA. TRL7 was reported to have influences on viral infection in poultry and has been regarded as a vital component of antiviral immunity, particularly in ducks (Manglani and McGavern, 2018). RIG-I (retinoic acid–inducible gene I) or RIG-I–like receptor dsRNA helicase enzyme is part of the RIG-I–like receptor family, which also includes MDA5 and LGP2. These have been reported to function as a pattern recognition receptor that is a sensor for viruses such as influenza A, others such as Sendai virus, and Flavivirus (Matrosovich et al., 2009b). RIG-I typically recognizes short 5′ triphosphate uncapped double-stranded or single-stranded RNA (Basler and Aguilar, 2008b). RIG-I and MDA5 are the viral receptors, acting through a common adapter MAVS and trigger an antiviral response through type-I interferon response. RIG1 is an important gene conferring antiviral immunity for ducks, particularly avian influenza (Matrosovich et al., 2009b).

TLR3 is another member of the toll-like receptor (TLR) family. Infectious agents express PAMP (pathogen-associated molecular patterns), which is readily recognized by TLR3, which in turn secretes cytokines responsible for effective immunity. It recognizes dsRNA associated with a viral infection, and induces the activation of IRF3, unlike all other toll-like receptors which activate NF-κB (Kannakiet al., 2010). IRF3 ultimately induces the production of type I interferons, which is ultimately responsible for host defense against viruses (Kell and Gale, 2015). In our lab, earlier we had studied immunogenetics against bacterial disease with identified immune response molecule such as CD14 gene in goat (Vercammen et al., 2008a; Schlee et al., 2009), cattle (Stetson and Medzhitov, 2006), and buffalo (Pal and Chatterjee, 2009; Pal et al., 2013). We reported for the first time the role of mitochondrial cytochrome B gene for immunity in sheep (Pal et al., 2011) and immune-response genes in Haringhata Black chicken (Pal et al., 2014).

Indigenous duck population in the Indian subcontinent was observed to have better immunity against viral infections, and so far, no systematic studies were undertaken. Certain reports revealed that ducks were mostly asymptomatic and were better resistant to avian influenza infection. Thus, the present study was conducted with the aim of molecular characterization of immune response genes (TLR3, TLR7, and RIGI) of duck, providing the initial proteomics study and prediction of the binding site with multiple strains of avian influenza virus through in silico studies (molecular docking), establishment of disease-resistant genes of ducks through quantitative PCR, and experimental validation of the identified genes through differential mRNA expression profiling of the identified gene with respect to healthy and challenged embryonated eggs as in vitro studies.

Indigenous duck population was characterized to be very hardy, and usually asymptomatic to common avian diseases. But there is a paucity of information regarding the systemic genetic studies on duck involved in its unique immune status. It is evident that duck possesses some unique genetic makeup which enables it to provide innate immunity against viral infection, particularly avian influenza.

In this current study, we identified three immune response molecules, which were earlier known to have immunity against viral infection such as RIGI, TLR7, and TLR3. We characterized these proteins of duck, and attempted to identify the SNPs or variations in nucleotide among duck and chicken. Through molecular docking, the most promising IR molecules conferring innate immunity against avian influenza have been identified, which were later on confirmed through the wet lab study as differential mRNA expression profiling. We attempt to explore the unique genetic constitution of duck immune response with respect to that of chicken. However, RIGI was reported to be expressed only in duck, not in chicken; hence, comparison was not available. TLR7 expression profiling was observed to be significantly better in duck than in chicken and other poultry species, indicative of better antiviral immunity in ducks. 53 non-synonymous mutations with amino acid variations were observed while comparing amino acid sequence of duck with other poultry species, including chicken, most of which are confined to the LRR domain. We had already depicted that LRR is an important domain for pathogen binding site as in this case of avian influenza. For the effective antiviral activity, binding of viral protein with the immune response molecule is the primary criterion. Leucine-rich repeats (LRRs) were observed to be important domain involved in binding with hemagglutinin and neuraminidase surface protein in case of both TLR3 and TLR7. Similar studies have also reported LRR as the important domain against bacterial infections in case of CD14 molecule in cattle (Stetson and Medzhitov, 2006), goat (Vercammen et al., 2008a; Schlee et al., 2009), and buffalo (Pal and Chatterjee, 2009; Pal et al., 2013). Other important domains identified were LRRNT, LRRCT, site for TIR, and the sites for leucine-rich receptor-like protein kinase, including certain posttranslational modification sites. Similar reports were also identified in different species (Stetson and Medzhitov, 2006; Vercammen et al., 2008a; Pal and Chatterjee, 2009; Schlee et al., 2009; Pal et al., 2013). An important observation identified was that although the CARD domain was believed to be an important binding site for RIGI for some identified virus, it has no binding ability with avian influenza virus. Other studies have reported the role of the CARD domain in binding with the MAVS domain as a part of antiviral immunity (Kelley and Sternberg, 2009; Šali et al., 2020).

TLR3 (CD283 or cluster of differentiation 283) is a pattern recognition receptor rich in leucine-rich repeats as revealed in duck TLR3 of the current study. Other important domains include LRRNT, LRRCT, TIR, leucine-rich receptor-like protein kinase, leucine zipper, GPI anchor, and leucine-rich nuclear export signal. The other sites for posttranslational modification as observed were N-linked glycosylation, casein kinase, phosphorylation, myristoylation, and phosphokinase phosphorylation. Variability of amino acids in important domains was observed for duck TLR3 compared to that of chicken, goose, and turkey. It was observed that genetic similarity between duck and goose was more than that of chicken and turkey. Some amino acids which are conserved for ducks and denote for important domains such as LRR, LRRCT, and TIR have been identified. In the LRRCT domain, valine is present in a duck in contrast to alanine in chicken, turkey, and goose. The current study identified 45 sites of non-synonymous substitutions between duck and chicken, which affect important domains for TLR3 as a pattern recognition receptor. Although it is the first report of characterization of TLR3 in indigenous duck, earlier studies were conducted in Muscovy duck, when full-length cDNA of TLR3 was characterized to be of 2836 bp encoding polypeptide of 895 amino acids (Zhang and Skolnick, 2005). The characterization of the deduced amino acid sequence contained 4 main structural domains: a signal peptide, an extracellular leucine-rich repeats domain, a transmembrane domain, and a Toll/IL-1 receptor domain (Zhang and Skolnick, 2005), which is in agreement to our current study. It is to be noted that TLR3 is a PRR, with the secondary structure being visualized as helix, loop, and sheet with the sites for disulfide bond being depicted in blue spheres. It recognizes dsRNA associated with a viral infection, and induces the activation of IRF3, unlike all other Toll-like receptors which activate NF-κB. IRF3 ultimately induces the production of type I interferons, which aid in host antiviral immunity (Morgan et al., 2019). In the current study, we observed sites for leucine zipper in duck TLR3, which is an inherent characteristic for dimerization. Earlier studies have also reported that TLR3 forms a large horseshoe shape that contacts with a neighboring horseshoe, forming a “dimer” of two horseshoes (Niu et al., 2019). As already explained that glycosylation is an important PTM (posttranslational modification site), it acts a glycoprotein. But in the proposed interface between the two horseshoe structures, two distinct patches which are rich in positively charged amino acids and may be responsible for the binding of negatively charged viral dsRNA were observed.

RIG1 [also known as DEAD-box protein 58 (DDX58)] is an important molecule conferring antiviral immunity. Various important domains have been identified, such as CARD_RIG1 (caspase activation and recruitment domain found in RIG1), CARD2 interaction site, CARD1 interface, helicase insert domain, double-stranded RNA binding site, RIG-I-C (C terminal domain of retinoic acid–inducible gene, RIG-I protein, and a cytoplasmic viral RNA receptor), RD interface, zinc-binding domain, and RNA binding. CARD proteins were observed to be responsible for the recognition of intracellular double-stranded RNA, a common constituent of a number of viral genomes. Unlike NLRs, these proteins, RIG-I contain twin N-terminal CARD domains and C-terminal RNA helicase domains that directly interact with and process the double-stranded viral RNA. CARD domains act through the interaction with the CARD motif (IPS-1/MAVS/VISA/Cardiff) which is a downstream adapter anchored in the mitochondria (Vercammen et al., 2008b; Botos et al., 2011).

Through the in silico alignment study, it was clearly observed that TLR3 binds with the H antigen of the avian influenza virus (H5N1). It was known that avian influenza virus (H5N1) strain is 100–200 nm spherical, enveloped, and includes 500 projecting spikes containing 80% hemagglutinin and 20% neuraminidase (Bouvier and Palese, 2008), with genome being segmented antisense ssRNA. In order to combat the infection, TLR3 binds with the hemagglutinin spikes of the influenza virus. TLR3 gene was observed to have a role to combat against Marek’s disease (Said et al., 2018): seven amino acid polymorphism sites in ChTLR3 with 6 outer part sites and 1 inner part site (Barber et al., 2010). TLR3 cannot act alone. It acts while interacting with a series of molecules such as TICAM1, MAP3K7, TAB2, TRAF6, Myd88, IRAK4, IFIH1, and even TLR7. TLR3 acts through the RIG1-like receptor signaling pathway and acts through TRIF (Said et al., 2018). Along with in silico studies, we validated these findings with experimental challenge with avian influenza virus in embryonated fibroblast cell. The expression was observed to be more in infected egg than in healthy control for these genes in duck and chicken.

RIG1 is an important molecule which is only expressed in ducks, not in a chicken. Duck RIGI transfected cells were observed to recognize the RIG-I ligand, and a series of antiviral genes were expressed such as IFN-β, MX1, PKR, IFIt5, and OASI, and consequently, HPAIV (highly pathogenic avian influenza virus) titers were reduced significantly (Haunshi and Cheng, 2014a; Ruan et al., 2015). RIG-I belongs to the IFN-stimulated gene family, and it acts through the RIG-I–like receptor signaling pathway. RIG1 detects dsRNA virus in the cytoplasm and initiates an antiviral response by producing type-I and Type-III IFN, through the activation of the downstream signaling cascade. RIG-I is an IFN-inducible viral sensor and is critical for amplifying the antiviral response (Potter et al., 2008; Barber et al., 2012). Although RIG1 expression is absent in chicken; it can produce INFα by another pathway. It has been observed that IFN-β expression upon influenza infection is mediated principally by RIG1 (Yoneyama et al., 2004). INFα expression induced by chicken is unable to protect the host from avian influenza infection as IFN-β, produced in ducks (Bouvier and Palese, 2008; Takahasi et al., 2008). This may be one of the major reasons why ducks are resistant to HPAIV, but not chicken. Since the RIG1 gene is not expressed in chicken, a comparative study was not possible. Avian influenza virus was observed to have surface glycoproteins as hemagglutinin and neuraminidase spikes on its outer surface (Loo et al., 2008). It was observed that duck RIG1 can bind with both H antigen and NA antigen of avian influenza virus. It is interesting to note how RIGI acts on virus and causes destruction. RIG-I acts through the RIG-I–like signaling pathway, and secretes NRLX1 and IPS1, leading to the production of IKKβ, which in turn causes the secretion of NFκβ and Iκβ (Koemer et al., 2007). These substances ultimately cause viral myocarditis and the destruction of the virus (Loo and Gale, 2011). A sequence of reactions occurs such as high fever, acute respiratory distress syndrome, chemoattraction of monocytes and macrophages, T-cell activation, and antibody response (Haunshi and Cheng, 2014b).

TLR7 is another important molecule responsible for antiviral immunity; it recognizes single-stranded RNA as the genetic material. It acts through the Toll-like receptor signaling pathway. However, from the current study, it was observed that TLR7 binds well with the NA antigen. The identified domains for TLR7 are mainly LRR (leucine-rich repeat), TIR, cysteine-rich flanking region, LRRNT (leucine-rich repeat N-terminal domain), and TPKR-C2 (tyrosine-protein kinase receptor C2 Ig-like domain). TLR7 releases Myd88, which in turn releases IRAK (Yamada et al., 2018). Ultimately, IRF7 is released, which causes viral myocarditis. It is interesting to note that in human and mice, TLR7 is alternatively spliced and expressed as two protein isoforms (Brisse and Ly, 2019). Another interesting observation was that chicken erythrocytes do not express TLR7 (Heil et al., 2003). While studying the TLR7 expression pattern in different avian species, an interesting observation in our current study was that TLR 7 gene expression was significantly better in duck than in other poultry species, such as indigenous chicken breeds (Aseel and Haringhata Black chicken), goose, and guineafowl. This is the first report for such a comparative study. It was reported that chicken TLR7 follow a restricted expression pattern. TLR7 expression was better in a macrophage cell line, chicken B-cell–like cell line, but the expression was observed to be lower in kidney cell line (Philbin et al., 2005). Following Marek’s disease virus expression, TLR7 expression was observed to be increased in the lungs (St Paul et al., 2013).

Similarly, increased TLR7 expression was noted in IBDV (infectious bursal disease virus) (Iqbal et al., 2005). With regard to avian influenza infection, it was observed that at the early stage of low pathogenic avian influenza virus (LPAIV) infection of H11N9, in both duck and chicken, TLR7 is transiently expressed in peripheral blood mononuclear cells (PBMCs), while as infection progresses, the expression declines. Hence, it was observed that in chicken, TLR7 expression depended on the interaction between host and RNA virus (Abdul-Careem et al., 2009). Thus, differences in the expression pattern of TLR7 in chicken and duck were suggested (Abdul-Careem et al., 2009). Even in chicken, the TLR7 expression pattern was found to vary between HPAIV and LPAIV. Thus, TLR7 was observed to be an important immune response gene for avian influenza; TLR7 ligands show considerable potential for antivirals in chicken (Abdul-Careem et al., 2009). Although no direct report was available for better TLR7 expression in a duck than in a chicken, it was reported that tissue tropism and immune function of duck TLR7 are different from those of chicken TLR7, which result in a difference in susceptibility between chicken and duck, when infected by the same pathogen (Abdul-Careem et al., 2009). A high expression pattern of duck TLR7 in respiratory and lymphoid tissue was observed to be different from that of chicken.

TLR3, TLR7, and TLR21 localize mainly in the ER in the steady-state and traffic to the endosome, where they engage with their ligands. The recognition triggers the downstream signal transduction to activate NF-κв or IRF3/7, finally induces interferon and inflammatory cytokine production (Abdul-Careem et al., 2009). We can explore these identified and characterized genes for production of transgenic or gene-edited chicken resistant to avian influenza as a future control strategy against avian influenza through immunomodulation, devoid of side effects as in case of use of drugs (Rauf et al., 2011). It is to be noted that since the control of avian influenza virus has been difficult and challenging either through vaccination (Chen et al., 2013; Rios et al., 2017; Rx List, 2019) or treatment through antiviral drugs (FDA, 2020; Science News, 2016; Principi et al., 2019; WebMD, 2020) due to frequent mutation and genetic reassortment (regarded as antigenic shift or antigenic drift) of the single stranded RNA genome which is prone to mutations (NHS, 2014; Zhang et al., 2019; Richard et al., 2002a). An interesting observation revealed that unlike antibodies (comprising of immunoglobulins) which were highly specific, arising due to variability of Fab site and variable region (Nguyen, 2020), immune response molecules for innate immunity can bind Avian influenza virus (H, N antigen), irrespective of strains. As we analyze the binding sites, some important domains were identified, which may be involved in antiviral activity. This led to the finding that the therapeutic approach may be attempted with the recombinant product corresponding to the identified domain. Gene editing with gene insert from identified gene may lead to the evolution of disease-resistant strains/lines of chicken or duck.

Although the receptors for human influenza and avian influenza are different, mutations may overcome the barrier. As a case report in 2018, a human infection with a novel H7N4 avian influenza virus was reported in Jiangsu, China. Circulating avian H9N2 viruses were reported to be the origin of the H7N4 internal segments, unlike both the human H5N1 and H7N9 viruses that had H9N2 backbones. The major concern is that genetic reassortment and adaptive mutation of avian influenza virus give rise to human influenza virus strain H7N4 (Jiao et al., 2012; Ayora-Talavera, 2018; Karen, 2018; Li et al., 2020). The WHO has also warned about the pandemic on human flu resulting from genetic reassortment of avian influenza (Qu et al., 2020). We observed in this study that the H5N1 strain of avian influenza, a highly pathogenic strain, was genetically closer to H6N2. Recent reports revealed that H6N2 is continuously evolving in different countries such as South Africa (Quan et al., 2019), Egypt (WHO, 2018a), India (Worldometer, 2020), and North America (Zanaty et al., 2019) due to genetic reassortment. It is gradually evolving from the low pathogenic form to the high pathogenic form and observed to overcome species barrier with interspecies genetic assortment (Kumar et al., 2018; Richard et al., 2002b; Gillim-Ross et al., 2008; FAO, 2019) and every possibility to evolve as a pandemic for human. Reports are available depicting human influenza virus arising due to genetic reassortment of avian influenza in China (Jiao et al., 2012; Ayora-Talavera, 2018; Karen, 2018; Li et al., 2020). These findings highlight the growing importance of the study in the current era, when the world is suffering from a pandemic.

Although molecular docking analysis is available for the identification of various drug molecules with avian influenza virus (Amir et al., 2011; Liu et al., 2015; Khan et al., 2017; Pal and Chakravarty, 2019b), this is the first report of molecular docking analysis with the immune response molecules responsible for antiviral immunity against avian influenza and is the basis for finding drug for a disease. A series of immune response molecules are responsible for providing antiviral immunity with their respective interaction in various pathways as we depicted through String and KEGG pathway analyses in our current study.

The future outcome for the current study is the possible utilization of the identified disease resistant genes (RIGI, TLR3, and TLR7) for the development of avian influenza–resistant chicken with the identified gene insert from duck through gene editing or a transgenic approach.

RIG1 detects the virus that is present within the cytosol of infected cells (cell intrinsic recognition), whereas TLR3 detects virus-infected cells, and TLR7 detects viral RNA that has taken up into the endosomes of sentinel cells (cell-extrinsic recognition). TLR7 may be regarded as the promising gene for antiviral immunity with pronounced expression profiling in duck in contrast to other poultry birds. Molecular docking revealed RIGI, TLR3, and TLR7 as the promising genes conferring antiviral immunity against avian influenza. Point mutations have been detected in chicken TLR3 with respect to that of duck indicative of reduced antiviral immunity in chicken in comparison to duck utilization of the identified disease-resistant genes (RIGI, TLR3, and TLR7) for the development of avian influenza–resistant chicken with the identified gene insert from duck.