Strategies for enhancing immunity against avian influenza virus in chickens: a review

ABSTRACT

Poultry infection with avian influenza viruses (AIV) is a continuous source of concern for poultry production and human health. Uncontrolled infection and transmission of AIV in poultry increase the potential for viral mutation and reassortment, possibly resulting in the emergence of zoonotic viruses. To this end, implementing strategies to disrupt the transmission of AIV in poultry, including a wide array of traditional and novel methods, is much needed. Vaccination of poultry is a targeted approach to reduce clinical signs and shedding in infected birds. Strategies aimed at enhancing the effectiveness of AIV vaccines are multi-pronged and include methods directed towards eliciting immune responses in poultry. Strategies include producing vaccines of greater immunogenicity via vaccine type and adjuvant application, and increasing bird responsiveness to vaccines by modification of the gastrointestinal tract (GIT) microbiome and dietary interventions. This review provides an in-depth discussion of recent findings surrounding novel AIV vaccines for poultry, including reverse genetics vaccines, vectors, protein vaccines and virus-like particles, highlighting their experimental efficacy among other factors such as safety and potential for use in the field. In addition to the type of vaccine employed, vaccine adjuvants also provide an effective way to enhance AIV vaccine efficacy; therefore, research on different types of vaccine adjuvants and vaccine adjuvant delivery strategies is discussed. Finally, the poultry gastrointestinal microbiome is emerging as an important factor in the effectiveness of prophylactic treatments. In this regard, current findings on the effects of the chicken GIT microbiome on AIV vaccine efficacy are summarized here.

KEYWORDS:

• Vaccine
• influenza
• H9N2
• adjuvants
• microbiome
• prebiotics
• probiotics
• immune response

Introduction

Globally, avian influenza viruses (AIV) pose a constant threat for the poultry industry. This threat includes low (LPAIV) and highly (HPAIV) pathogenic strains of AIV, which both can have detrimental outcomes for poultry operations and human health. HPAIVs such as H5N1 virus have resulted in the mortality of hundreds of millions of poultry through culling and infection, which comes at a tremendous financial cost (FAO, 2012). In addition to economic losses, human infections with some strains of highly pathogenic virus are associated with high mortality levels, including viruses that express H5 and H7 proteins, responsible for hundreds of infections per year (Sutton, 2018). Although human-to-human transmission does not routinely occur for HPAIV strains, there is a constant threat for the emergence of these viruses into human populations, highlighted by certain H5-expressing clades which have potential for human pandemics (Yamaji et al., 2020). While LPAIV strains do not directly affect poultry and human health at the same level, uncontrolled infection in poultry is not without consequences. For example, H9N2 AIV is a LPAIV subtype that is highly prevalent in poultry worldwide. This subtype can be associated with morbidity and mortality if confounding factors are present and, importantly, this virus might donate internal genes to HPAIV viruses leading to enhanced pathogenicity (Peacock et al., 2019). Ultimately, uncontrolled infection and transmission of poultry with AIV increases the chance of human infection, results in economic losses, and drives the ability for virus mutations and reassortment, promoting the emergence of infectious and pathogenic viruses. An example is the H7N9 virus, a highly lethal virus in humans, which has acquired some of its internal genes, including PB2 and PA (Pu et al., 2021) from an H9N2 virus, and has been shown to have accumulated novel mutations in viral polymerase genes that could further enhance its transmission capabilities (Huang & Wang, 2020). Therefore, it is of utmost importance to limit poultry infection with AIV.

Disrupting the infection and transmission of AIV in poultry is likely to be achieved with a multi-pronged approach. This approach includes the continued use of biosecurity practices and culling of infected birds. However, novel technologies can enhance the surveillance efforts to identify and predict emergence and severity of AIV outbreaks (Steinfath et al., 2018; Yousefinaghani et al., 2019, 2020) and decrease the time it takes to detect infected birds (Astill, Fraser et al., 2018). Poultry vaccination is also a candidate strategy to protect poultry from lethal AIV infections and prevent infection and transmission. The efficacy of poultry vaccine use has been demonstrated recently. A dual H5 and H7 inactivated virus vaccine for poultry seems to have led to a significant reduction in human infections with H7N9 virus (Zeng et al., 2018). Poultry AIV vaccine strategies are continuously being explored to search for more immunogenic or cross-protective formulations, including novel approaches in reverse genetics, vector strategies, proteins and peptides, and virus-like particles (VLP). The effectiveness of vaccines depends on many factors, highlighted by the targeted AIV entities that are incorporated into the vaccine, and the resultant immunogenicity or quantity of responses that are produced. While vaccine design can affect immune responses, adjuvants have shown beneficial effects on poultry vaccine immune responses. In addition to vaccine and adjuvant formulations, recent research efforts have uncovered the importance of the poultry gastrointestinal tract (GIT) microbiota in immune system function. Therefore, efforts that drive the maintenance or development of healthy GIT microbiota, such as probiotics and other dietary factors, are a key area of research to enhance antiviral immune responses and AIV vaccine efficacy. This review sets out to summarize key findings surrounding poultry AIV vaccines, highlighting novel and existing vaccine strategies, adjuvants and vaccine delivery, and the effects of diet and the microbiota on immune responses to AIV vaccines in poultry.

AIV vaccine strategies for poultry

Vaccination is a well-studied method that can be used against infection with AIV in poultry. Commercial vaccine strategies against highly pathogenic H5 and H7 viruses and low pathogenic viruses, specifically H9N2 AIV, are employed in some countries. These include China, Egypt, Hong Kong, Indonesia, Vietnam and Mexico (Yoo et al., 2018). Licensed poultry vaccines for AIV vary and include whole inactivated virus (WIV) and haemagglutinin (HA)-expressing viral vectors (Suarez & Pantin-Jackwood, 2017). The majority (∼95%) of AIV vaccines employed in the field are WIV vaccines suspended in an oil emulsion or surfactant adjuvant, while the remainder of used vaccines are viral vectors that express an influenza virus protein, usually HA (Swayne, 2020). Vector vaccine models licensed in at least one country worldwide have been reviewed recently (Suarez & Pantin-Jackwood, 2017) and include fowlpox virus, herpesvirus of turkeys and Newcastle disease virus vectors.

Some novel AIV vaccines for poultry have recently been licensed in certain countries and have been tested experimentally. For example, an alphavirus replicon RNA particle vaccine has been licensed and produced for emergency use in the USA. The vaccine contains an RNA sequence of an H5 HA protein and has been tested in chickens, demonstrating complete protection after H5N2 challenge compared to 100% mortality in unvaccinated birds (Ladman et al., 2019). Additionally, this vaccine significantly reduced cloacal virus shedding in challenged chickens. Despite differences in immunogenicity, other factors such as the required time for production and the ability to differentiate between vaccinated and infected birds (DIVA) must be considered when comparing novel vaccines to WIV vaccine strategies. A recombinantly expressed H5 HA protein produced by a baculovirus vector insect cell expression system has also recently been registered in Mexico and Egypt. In a series of challenge trials with non-clade-matched H5N1 virus strains, the vaccine has been shown to offer complete protection in chickens while also significantly reducing virus shedding. However, some challenge viruses were still orally shed by birds 1 week after the challenge (Oliveira Cavalcanti et al., 2017).

Multiple factors must be considered when developing influenza vaccines for poultry. Several recent reviews have summarized these items (Spackman & Swayne, 2013; Hasan et al., 2016; Yoo et al., 2018), and they include inducing appropriate and targeted immunogenicity, safety or lack of replication, the ability for large-scale production, and the mechanisms to facilitate DIVA. Many experimental vaccines attempt to direct antibody responses against the HA protein by designing vaccines that include the HA protein or lead to expression of the HA protein, and these responses can be quantified with the haemagglutination inhibition (HI) assay. The effectiveness of vaccine-induced protection can be altered if vaccine HA and challenge virus HA sequences differ. It is understood that a negative correlation exists between mortality in chickens and the amino acid percentage difference between HA proteins in a vaccine and challenge virus (Spackman & Swayne, 2013). Likewise, meta-analysis of vaccine studies against highly pathogenic Indonesian strains of H5N1 suggests that, rather than selecting vaccine candidates based on vaccine-induced HI titres, HA protein homogeneity in vaccines and challenge viruses is a better indicator of protection (Villanueva-Cabezas et al., 2017). This report suggests the importance of surveying circulating influenza viruses for novel viruses or HA gene sequences. Antigenic drift in human influenza viruses has led to large-scale research efforts towards the development of universal vaccines (Sautto et al., 2018; Yedidia & Rudolph, 2018). Likewise, many experimental poultry vaccines have attempted to generate broad protection against different influenza viruses or against all influenza viruses that express the same HA protein. Many mechanisms exist to facilitate DIVA (Hasan et al., 2016) including (1) Non-vaccine strategies, such as maintaining a proportion of unvaccinated birds that should lack antibodies to influenza virus proteins, diagnostic approaches that employ ELISA-based detection of non-structural proteins or functional protein epitopes that are not present in WIV or vectored vaccines, and (2) Vaccination strategies, including the use of purified proteins or vectors that lack expression of many influenza virus proteins. Based on these strategies, most novel poultry influenza vaccine platforms do facilitate DIVA. For this review, research surrounding influenza virus vaccine strategies can be grouped into inactivated viruses, bacterial and viral vectors, recombinant proteins and virus-like particles. These vaccine methodologies are associated with different benefits; however, immune responses must be considered a top priority for candidate vaccines, including cell- and antibody-mediated responses.

Inactivated AIV vaccine strategies

WIVs are the most commonly used AIV vaccines for poultry. They are preferable to attenuated or replicating vaccines for their safety profile and to recombinant vaccines for their simple production methodology that also allows for the development of autogenous vaccines. Virus inactivation strategies can also be used to produce reverse-engineered AIV vaccines that contain specific HA sequences, which is a promising technique in poultry for the generation of vaccines against specific viral strains. Multiple strategies exist for inactivating influenza viruses and it is a current course of research. For example, analysis of formaldehyde and beta-propiolactone (BPL) inactivation of different influenza virus strains has revealed different effects on viral protein structures (Pawar et al., 2015). Using the HI assay, it was shown that these two chemicals affect low and highly pathogenic AIV differently when used at different concentrations, and, for some viruses such as H9N2 virus, formaldehyde inactivation was found to affect protein structures more significantly than BPL (Pawar et al., 2015). This is an important consideration for vaccine-induced immune responses, as alterations in vaccine protein structures could affect antibody-mediated responses in chickens against important protein epitopes. The effects of different inactivation strategies on immune responses in chickens have been assessed experimentally; Astill, Alkie, Yitbarek, Abdelaziz, et al. (2018) demonstrated elevated levels of antibody- and cell-mediated immune responses in chickens administered BPL-inactivated H9N2 virus vaccines compared to formaldehyde-inactivated vaccines.

The subtype of influenza virus can alter the effects of inactivating agents. Recent research comparing BPL and formaldehyde inactivation of H1, H7, H3, and H5 viruses demonstrated that BPL more effectively inactivates the viruses; however, evidence of increased effects on protein structures was noted (Herrera-Rodriguez et al., 2019). It is essential to point out the concentration of formaldehyde used in the study was 10 times less than that of BPL. Nevertheless, Herrera-Rodriguez et al. (2019) also demonstrated different levels of immune system stimulation induced by the WIVs by assessing in vitro Toll-like receptor (TLR) 7 stimulation in a human cell line, demonstrating generally elevated stimulation from BPL WIVs. Additionally, novel inactivation mechanisms, such as ultrashort pulse laser and low energy electron irradiation techniques, have been shown to maintain virus protein structures in inactivated virions (Sabbaghi et al., 2019).

Reverse genetics AIV vaccines

Inactivation is also a crucial aspect of creating vaccines produced using reverse genetics. The segmented genome of AIV allows for the generation of viruses that contain internal gene segments of non-pathogenic viruses while displaying the surface proteins of highly pathogenic viruses. This system can additionally be used to modify HA sequences to enhance replication abilities in embryonated chicken eggs (ECE) or cell lines to enhance production capacity (Lu et al., 2005). However, despite attenuation of the virus after modifying internal influenza genes, inactivation is still important to inhibit any replication in the host. Many examples of inactivated reverse genetics designed AIV vaccines for chickens have recently been documented. For example, using consensus HA and neuraminidase (NA) sequences from H5N8 AIVs expressed in a PR8 backbone expression system, which includes an influenza virus with six internal gene segments, specific point mutations were induced in HA and NA gene sequences to enhance growth characteristics in ECE and decrease mammalian pathogenicity (An et al., 2019). The H5N8 PR8 virus was inactivated and administered to chickens prior to challenge with a lethal dose of H5N6 AIV. It was observed that chickens were protected from infection, yet over half of the birds were shedding the challenge virus 7 days post-infection. Reverse genetics vaccine platforms also allow for HA proteins of outbreak viruses to be included in vaccine preparations. For example, Hoang et al. (2020) produced an H5N1 vaccine that contained HA and NA proteins from viruses responsible for outbreaks from 2012 to 2014 in Vietnam, demonstrating >90% protection after H5 clade-matched viral challenge. Also, reverse genetics has been used to produce five vaccines containing different H5 proteins from recent Korean AIV outbreak strains, all of which demonstrated 100% protection against challenge with the HA-matched viral strain (Kang et al., 2020). Importantly, this study showed that when chickens were administered a vaccine dose of 512 HA units, cloacal and oral shedding was completely inhibited after challenge, while 10-fold lower doses did not prevent shedding. Reverse genetics also allows for HA and NA proteins from different AIV viruses to be combined. For example, an AIV vaccine displaying an H5 HA protein along with an N2 NA protein from an H9N2 virus resulted in 100% protection in chickens after challenge with an H5N1 virus at 28 days post-vaccination, and 90% protection at 200 days post-vaccination (Bhatia et al., 2016). Protection against H9N2 AIV due to immune responses to the N2 NA protein were not assessed; however, the N2 NA protein was suggested to enhance viral replication in addition to allowing for DIVA. In addition to H5 vaccines, reverse genetics approaches have been used to generate candidate H9N2 vaccines by modifying the H9 HA and N2 NA proteins leading to enhanced replication in ECE in addition to greater serum neutralization titres against field H9N2 viruses in chickens vaccinated with the inactivated virus (Wang et al., 2019). Interestingly, an influenza virus has been produced that expresses an H5 HA protein of one virus clade (2.3.4.4) in addition to the HA1 subunit of an H5 HA protein of another clade (2.3.2.1), ultimately producing HI titres against both HA proteins in chickens vaccinated with the inactivated dual HA-expressing virus (Hou et al., 2018). This group has also made a similar virus expressing both H5 HA and H7 HA1 subunit proteins, producing protection against a lethal challenge of both viruses (Li et al., 2018). These dual HA-expressing viruses have been generated via insertion of an HA gene into the RNA segment that codes for the influenza virus non-structural protein (NS1).

Overall, reverse genetics allows influenza viruses to be generated expressing HA proteins or other proteins from specific viruses of concern, in addition to allowing researchers to enhance viral replication in ECE and cell lines by modifying HA and NA proteins. Additionally, novel cloning methods that reduce the need for restriction enzymes have also recently been described, potentially reducing difficulties associated with recombination during the production of reverse genetics influenza viruses (Bhat et al., 2020). Regardless of the production methods or virus produced, inactivation is required for chicken vaccines. Therefore, it is of paramount importance to select an inactivation method that does not alter virus protein structures and induces the highest levels of immunogenicity.

Recombinant vector AIV vaccine strategies

Alongside inactivated virus vaccines, vector vaccines are also used in some parts of the world for poultry vaccination against AIV. Like inactivated virus vaccines, vector vaccines do not result in AIV replication in the host but result in the expression or display of influenza virus proteins. Vector vaccines are often immunogenic and, due to their inherent structure, they can contain multiple components that stimulate innate responses leading to more potent adaptive immune responses. Strategies to generate vectored AIV vaccines include both bacterial and viral methods. Vector vaccines can result in the production of a foreign vaccine protein, including surface display of the protein, or can result in expression and production of the foreign vaccine gene following infection of host cells. Many strategies have been explored recently for AIV vaccines for chickens, with varying results in immunogenicity and protective efficacy.

Recombinant bacterial AIV vector vaccines

Recombinant bacteria can be generated to produce various influenza virus protein antigens and delivered to chickens. As opposed to most viral vectors, bacterial vectors that carry genetic sequences for vaccine antigens often produce the formed protein, which can then be targeted by the chicken immune response after vaccine delivery. A large variety of vectors have been used to produce many different influenza virus proteins and peptides. The following section provides a summary of research strategies focused on expression of HA and other AIV proteins.

Many bacterial vector strategies have focused on the HA protein of influenza virus. For example, Bacillus subtilis engineered to express an H5 protein was orally administered to chickens leading to significantly increased tracheal IgA and serum IgG titres in addition to enhanced in vitro T cell proliferation after stimulation with purified H5 protein (Mou et al., 2016). Lactobacillus plantarum has also been used to express an HA protein. Briefly, Yang, Yang, Yang, et al. (2017) produced recombinant bacteria that expressed H9 on their surface. The bacteria were used to orally vaccinate chickens, producing systemic antibody-mediated immune responses that matched levels induced by an intramuscularly administered inactivated H9N2 AIV vaccine. Both the inactivated and L. plantarum H9N2 vaccines led to similar decreases in viral shedding after challenge, despite significantly greater levels of IgA in faecal droppings of the recombinant L. plantarum group. A Salmonella enterica Serovar Enteritidis ghost vaccine expressing an H1 protein was shown to induce antibody responses against the H1 protein and Salmonella outer membrane proteins in chickens, in addition to decreasing titres of both S. enteritidis and H1N1 virus following experimental challenges (Won et al., 2020). Ghost vaccine platforms such as this permit safe delivery of an immunostimulatory vector, initiating immune responses against a prominent foodborne pathogen while also boosting responses against the recombinant protein expressed in the vector. Nonetheless, pre-existing immunity to bacteria that commonly infect chickens, such as Salmonella, has been shown to affect anti-HA immune responses after vaccination with a vector-matched bacterial vaccine (Hajam & Lee, 2017). This effect was noted specifically when anti-Salmonella antibody titres were elevated at the time of vaccine administration.

Expression and production of other AIV proteins in bacterial vectors have also been developed to immunize chickens against influenza viruses. While immune responses directed against the HA protein can inhibit viral entry into host cells, there are other important factors to consider for immune responses directed against other influenza proteins. For example, the matrix (M) proteins, including M1 and the ion channel protein, M2, in addition to the nucleoprotein (NP), are often conserved among influenza viruses. This means cell-mediated immune responses directed against these proteins can potentially be protective against multiple different viruses. Additionally, vaccines that target the extracellular domain of M2 (M2e) permit cross-protective antibody-mediated responses. Due to the large DNA coding capacity of bacterial plasmids, vectors can also be designed to express large fusion proteins that contain multiple influenza virus proteins. As earlier mentioned, L. plantarum has been used to produce mucosal responses against a recombinantly expressed HA protein; this effect was also observed for a recombinant L. Plantarum vaccine expressing a fusion protein containing the H9N2 M1 protein, NP proteins and an immunostimulatory peptide (Yang, Yang, Shi, et al., 2017). Briefly, the L. plantarum vaccine was found to induce mucosal antibody responses and systemic cell- and antibody-mediated immune responses in chickens after oral or nasal administration. This bacterial vector has since also been engineered to express a fusion protein containing both the M2e and HA proteins of an H9N2 virus (Yang et al., 2018), where it again produced mucosal antibody responses as previously observed. It also resulted in systemic anti-HA antibody responses that were significantly elevated compared to an intramuscularly-administered commercial inactivated H9N2 virus vaccine. It should be noted that the above studies using L. plantarum as a vector administered the vaccine a total of six times, a difficult practice to replicate in commercial settings.

Salmonella bacteria have been used to express various influenza virus proteins, in addition to HA, for chicken vaccines. Salmonella enterica Serovar Typhimurium vectors expressing a consensus H7 protein, consensus N9 protein or a tetramer of M2e were combined into a single oral dose for chickens, demonstrating cell- and antibody-mediated immune responses and a significant reduction in cloacal H7N9 virus shedding post-challenge (Kim et al., 2018). It should be noted that chickens administered with just the vector expressing H7 developed non-significantly different levels of immune responses and decreases in virus shedding. Salmonella enterica Serovar Gallinarum has additionally been engineered to contain sequences for M2e fused to CD40 ligand. Hajam et al. (2018) developed two vectors, one that produced the fusion protein and one that contained the plasmid DNA sequence, designed to transfer genetic material to host cells. Oral immunization with the protein-producing vector resulted in anti-M2e protein systemic and mucosal antibody responses in addition to cell-mediated responses. However, co-administration of both vectors produced significantly greater antibody- and cell-mediated responses in addition to significantly decreased H9N2 virus shedding when compared to single administration. More recently, S. gallinarum vectors have been produced to express HA1 or HA2 of H9 in addition to M2e (Hajam et al., 2020), demonstrating enhanced cell-mediated responses induced by the HA1 and M2e vectors compared to an intramuscularly administered H9N2 virus inactivated vaccine. All of the S. gallinarum vectored vaccines significantly reduced virus shedding after experimental challenge; however, no differences in virus shedding were observed between vaccine groups.

Lactococcus lactis, a food-grade species of bacteria, has also been recently engineered as a vector for chicken vaccines. Secretion of the stem region of an H9 protein (HA2) fused to M1 by L. lactis was engineered and administered a total of six times to chickens via oral administration, resulting in significantly greater titres of mucosal anti-influenza antibody compared to an intramuscularly-administered inactivated H9N2 virus vaccine (Sha et al., 2020). Similar vectors expressing either NA of H1N1 or a tetramer M2e protein were orally administered to chickens, demonstrating mucosal and systemic antibody responses against the targeted protein antigens, which possessed neutralizing capacities against an H1N1 virus (Lahiri et al., 2019). Recent advances in bacterial vector influenza vaccines for chickens demonstrate that mucosal immune responses are an advantage of these systems. Mucosal responses are important for defense against influenza viruses in chickens, and many experimental bacterial vector vaccines have shown the ability to reduce viral shedding after challenge. However, it is important to note that many of the experimental results generated have used multiple oral administrations of vaccine. For large numbers of birds in commercial settings, it may be difficult to implement these methods of administration.

Recombinant viral vector AIV vaccines

Unlike recombinant bacterial vectors, the capacity to express large fusion proteins is hindered due to limits on the amount of genetic material that can be accommodated by viral vectors. However, viral vectors can infect chicken cells and express influenza proteins or protein subunits, a process that can produce high levels of immunogenicity. Nevertheless, immune responses induced by vaccines must be directed at specific viral proteins of influenza virus, or they risk losing efficacy. For example, following the 2014/2015 H5N2 AIV outbreak in North America, existing anti-H5 vaccines were tested experimentally in chickens followed by challenge with influenza viruses that contained a clade-matched H5 protein. It has been identified that inactivated vaccines produced greater protection than the existing vector vaccines (Kapczynski et al., 2017); however, the best predictor of efficacy in reducing viral shedding and mortality was homology of vaccine and challenge HA proteins, not HI titres directed against the vaccine HA protein. Therefore, importance must be placed on the HA gene that is incorporated into recombinant viral vector influenza vaccines for chickens.

Vectored AIV vaccines for chickens have recently been reviewed, highlighting their immunostimulatory capacity and the potential for these vaccines to be administered to chickens at 1 day of age or in combination with a booster vaccine consisting of an inactivated HA strain-matched vaccine (Suarez & Pantin-Jackwood, 2017). Viral vector vaccines are generally capable of inducing potent immune responses without need for adjuvants as they can mimic natural infection of host cells (Rauch et al., 2018), leading to stimulation of innate signalling pathways that can enhance subsequent adaptive responses. This also offers the advantage of antigen presentation via major histocompatibility complex (MHC)-I proteins and the induction of cell-mediated responses. Additionally, some viral vectors have tropism for mucosal tissues, allowing for more efficacious responses when administered to mucosal surfaces (Guroff, 2007). This is especially advantageous for AIV vaccines, as mucosal responses act as an initial barrier to infection. Many types of recombinant viruses have been licensed commercially for poultry influenza vaccines, and research efforts continue to elucidate vector strategies that produce enhanced immunity and protection in chickens against AIV infection. Newcastle disease virus (NDV) vectors are one type of recombinant vaccine that have been licensed for chickens in some countries and a review has recently been produced regarding their effectiveness (Kim & Samal, 2019). The authors noted that the potential for pre-existing immunity to the vector has been overcome experimentally by production of chimeric NDV vectors that express and display avian paramyxovirus surface proteins in place of native NDV surface proteins, along with expression of an HA protein. This strategy circumvents pre-existing immunity that could thwart efficacy of an NDV vector; however, this comes at a cost, as NDV vectors for poultry can offer bivalent protection against NDV. Dual protection against NDV and AIV has recently been demonstrated experimentally for both H9N2 and H5N1 virus vaccines. Briefly, incorporation of the ectodomain of H9 fused to the transmembrane domain of the NDV F protein showed enhanced HA expression in NDV virions, and vaccination with this vector demonstrated enhanced anti-H9 HI titres and decreased H9N2 virus shedding in vaccinated chickens compared to expression of the entire H9 open reading frame (Xu et al., 2019). For H5N1, a thermostable NDV vector has been used to express H5, an H5 HA1 subunit, or an H5 HA1 subunit fused to a tissue plasminogen activator signal sequence (Xu , Qin et al., 2020); the latter vector demonstrated significantly greater anti-H5 HI titres and levels of protection against lethal NDV and H5N1 virus challenge in vaccinated chickens. Interestingly, expression of HA1 or HA2 subunits in an NDV vector was also shown to provide no protective effects against H5N1 challenge (Shirvani, Paldurai et al., 2020). In addition, despite producing low HI titres and seroconversion rates, an NDV vectored H7N9 virus vaccine produced 100% protection following a lethal challenge in vaccinated chickens (Hu et al., 2020). Chicken vaccine strategies directed against H7N9 vaccines have thus far demonstrated low levels of neutralization and HI titres; however, studies such as this demonstrate the potential benefits of antibodies directed against non-neutralizing regions of the H7 protein. The genome of NDV also permits larger inserts, allowing for multiple AIV genes to be expressed; for example, expression of both the HA and NA proteins of H5N2 (Cho et al., 2018). Dual protein expression of AIV proteins has also recently been used to generate an NDV-vector expressing both soluble and membrane-bound H5 proteins, demonstrating 5.25 times greater HA protein production than singular expression of membrane-bound H5 protein (Murr et al., 2020). The authors suggest this vector could result in enhanced immune responses in animals, including chickens, although in vivo trials were not performed.

In addition to NDV, other viruses in the family Paramyxoviridae have also been used as vectors for AIV vaccines. Due to the restricted replication of NDV to the respiratory tract in chickens and common pre-existing immunity to NDV, the use of other avirulent viruses that have tropism for additional organs and tissues could potentially enhance the immune response elicited by the vector. Therefore, paramyxoviruses have been engineered to express influenza proteins as vaccines for chickens, with varying degrees of success. For example, avian paramyxovirus serotypes (APMV) 2, 6 and 10 have been shown to have low antibody-mediated cross-reactivity to NDV and have been experimentally used to produce recombinant chicken vaccines. APMVs 2, 6 and 10 were designed to express an H5 protein and were administered oculonasally to chickens, where they showed increased efficacy compared to an NDV vector expressing the same H5 protein in pre-NDV immunized birds against H5N1 virus challenge; however, the opposite was observed when birds did not have pre-existing immunity to the vector, indicating decreased immunogenicity (Tsunekuni et al., 2017). APMV 10, now taxonomically referred to as avian avulavirus 10, has since been genetically modified by flanking H5 expression cassettes on the untranslated regions of different virus genes (Tsunekuni et al., 2020). By flanking the NP protein, HA protein production was maximized, resulting in complete protection from a lethal challenge and no shedding of the challenge H5N1 virus from the throat or cloaca of vaccinated birds. The efficacy of APMV vectors could also be impacted by the serotype; therefore, Shirvani, Varghese, et al. (2020) selected APMV 3 for experimental production of an H5N1 vaccine due to its increased replication in tissues beyond the respiratory system. The APMV 3 vector was compared to an NDV vector after oculonasal vaccination in day-old chickens; both resulted in complete protection against H5N1 challenge, however, the APMV 3 vector induced significantly greater neutralizing antibody titres. Importantly, the APMV 3 vaccine provided complete protection against lethal challenge with NDV.

Of the licensed vaccines for AIV in poultry, herpesvirus of turkeys (HVT) vaccine has been a focus of research. HVT vectors that express many different AIV proteins have been generated, including a vaccine that reduced mortality from infection with H7N1 virus (Li et al., 2011) and different strains of H5 viruses (Kapczynski et al., 2015). In addition, a HVT vector expressing H9 produced immune responses in vaccinated chickens while, importantly, also completely inhibiting virus shedding from H9N2-challenged birds (Liu et al., 2019). Similar to NDV vectors, HVT-based vaccines are especially beneficial for poultry as bivalent vaccines, protecting against AIV and Marek’s disease virus (MDV). The efficacy of experimental HVT-vectored AIV vaccines is often tested against both MDV and AIV challenge, and it has been shown that insertion of HA genes into the HVT genome in different regions can alter these responses (Gao et al., 2011). Specifically, HA gene insertion into two different unique short genes found in the HVT genome has been shown to significantly impact HI titres and protection levels in chickens while not significantly affecting protection against MD when both vectors were compared to the unmodified HVT vaccine. An avirulent MDV strain has also been explored as a vector for expression of H5. After being administered to chickens, the avirulent vector provided significantly greater protection levels against lethal challenges of both H5N1 virus and MDV (Cui et al., 2013) when compared to the HVT vector. The currently licensed HVT H5N1 vaccine has been demonstrated by the manufacturer to reduce mortality and decrease shedding after infection with a number of antigenically diverse H5-expressing AIVs (Gardin et al., 2016). The engineering of HVT vaccines which express HA proteins that antigenically match circulating AIV will be important for continued efficacy of these vaccines. In addition, modifications to HVT expression systems must also be considered. For example, Balzli et al. (2018) compared two vectors that employed matching expression cassettes; however, one also contained a separate promoter region, leading to a significant reduction in protection and shedding from lethal challenge with H5N2 virus in vaccinated chickens. Recent innovations in genetic engineering technology have also shown potential to more rapidly develop HVT vaccines for poultry (Tang, Zhang et al., 2019).

Selection of appropriate HA proteins for AIV vector vaccines will be an important focus for future vaccines. For the most part, research in the development of viral vector vaccines has focused exclusively on inducing immune responses against HA proteins, although potential exists for other proteins to be included in vaccines. This includes proteins or, potentially, peptides, which are more often conserved between AIVs. For example, immune responses were observed in broilers that received a bacteriophage that displayed the N-terminal region of the M2 protein fused to the bacteriophage capsid protein (Lotfi et al., 2019). This system is beneficial as numerous capsid proteins are present in one viral particle, thereby presenting high quantities of influenza peptide to the chicken immune system; however, the vaccine has yet to be tested for protection against AIV in vivo.

Recombinant protein AIV vaccines

Problems associated with WIV vaccines include issues with DIVA and effects of inactivation mechanisms on protein structures, while live vector vaccines have potential problems such as risks associated with administration of replicating pathogens and pre-existing vector immunity. Therefore, recombinant protein vaccines offer a potential remedy. AIV proteins can be generated in a variety of recombinant expression systems, harvested, and administered to chickens as vaccines. Although protein vaccines can provide safe and targeted immune responses, they often lack immunogenicity as they do not contain the other immunostimulatory elements that WIVs and vectors contain, often requiring adjuvants or other strategies to enhance responses. Also, protein vaccines are often experimentally administered intramuscularly or subcutaneously as opposed to vector strategies that can induce immune responses when mucosally administered. Many methods exist to generate recombinant proteins, including bacterial, plant, and eukaryotic expression systems. In addition, multiple AIV proteins, peptides, or fusion proteins have been engineered for chicken AIV vaccines.

When developing HA protein expression systems, it is important to consider post-translational modifications such as protein folding and glycosylation to make proteins that are epitope-matched to wild-type AIVs. An experimental H5 subunit vaccine for chickens has been produced in transgenic tobacco seeds, and subcutaneous injection of the purified HA protein resulted in systemic immune responses (Ceballo et al., 2017). This system offers potential for a high-yield cost-effective method to produce AIV protein vaccines, although extensive purification procedures could hinder future use of plant-produced vaccines. In addition, novel strategies employing bacterial expression systems for AIV protein production have been examined for chicken vaccines. Escherichia coli has been developed to produce the ectodomain of the H5 protein in inclusion bodies (Saczyńska et al., 2017). This recombinant protein vaccine resulted in complete protection against a heterologous viral challenge and inhibited any transmission from vaccinated to naïve chickens; however, to finalize HA protein folding, chromatography-based protein refolding steps were required before subcutaneous vaccination in chickens. Insect cell baculovirus expression systems have also been explored as a cost-effective method to produce AIV proteins, as they permit protein folding in a similar fashion to mammalian expression systems (Schaly et al., 2021). Recombinant baculoviruses expressing AIV proteins can be used to infect insect cells resulting in protein production. This system has been used to produce an H5 protein fused to an E. coli enterotoxin, which was purified and administered via intranasal injection to chickens, producing systemic antibody-mediated immune responses (Tang, Lin et al., 2019). This expression system has also been used to produce baculoviral vectors that express H5 fused to the cytoplasmic domain of the baculovirus g64 protein (Tung et al., 2020), leading to in vitro surface expression of H5 in infected insect cells. Intramuscular injection of the insect cell lysates in chickens produced HI antibody responses and resulted in complete protection from lethal H5N2 virus challenge.

Other novel peptides and proteins have also been tested in chickens, including mosaic proteins. Generation of mosaic proteins involves applying algorithms that use multiple similar protein sequences to generate one unique consensus protein while also focusing on the inclusion of preferred T cell peptide sequences. A mosaic H5 protein has been developed from multiple H5 protein sequences and tested as a vaccine in chickens (Kingstad-Bakke et al., 2019). The mosaic protein was delivered in nanoparticle formulations, demonstrating cell- and antibody-mediated immune responses, in addition to significantly decreasing virus shedding after challenge with an H5N2 virus. The H5 mosaic protein was also expressed in a modified vaccinia Ankara (MVA) vector, demonstrating elevated levels of immunogenicity; however, it is important to note that both the nanoparticle and MVA vaccines induced antibody responses against multiple different H5 viruses. Mosaic proteins allow for immunity to be developed against multiple H5 proteins; other strategies also exist to facilitate this, including the use of peptide sequences from highly conserved regions of the influenza HA protein. A dual peptide vaccine that contained two linked HA peptides which are highly conserved has recently demonstrated the potential for efficacious cross-protective peptide vaccines. Sisteré-Oró et al. (2020) used a baculovirus-insect expression system that produced HA peptide sequences from H1N1 and H5N1 viruses linked to a flagellin protein for enhanced immunogenicity. Interestingly, two subcutaneous injections in chickens led to significantly enhanced protection against a lethal H7N1 virus challenge. Neutralizing antibody titres could not be distinguished between control and vaccinated birds, and anti-H7 ELISA results also showed no differences. The protection afforded by the vaccine suggests the involvement of non-neutralizing antibodies in addition to cell-mediated responses directed against the conserved peptides. Developing peptide vaccines that target cytotoxic T cell responses is a potentially effective strategy for chicken vaccines; therefore, work has been performed to assess the potential for peptide binding in chicken MHC-I and T cell receptor (TCR) complexes (Li et al., 2020). This has led to the identification of a conserved H5 peptide that offers excellent potential for recognition by specific chicken TCRs and has also demonstrated significant increases of peptide-specific CD8+ T cells in peptide-immunized chickens compared to chickens administered a WIV vaccine. Likewise, a peptide within the HA1 subunit of H5N1 virus has also been identified and found to be presented via MHC-I and MHC-II proteins, leading to in vitro activation of both CD4+ and CD8+ chicken spleen cells from B19 MHC haplotype chickens vaccinated against H5 HA (Steinfath et al., 2018).

Developing vaccines that target T cell responses often focus on proteins from conserved AIV proteins, including M1 and M2e, due to the potential to produce immune responses that are protective against multiple different AIVs. For example, a consensus M2e protein based on sequences from H1N1, H2N3 and H2N1 human influenza viruses was produced as a quadruplicate fusion protein also linked to heat shock protein 70 for chicken vaccination (Dabaghian et al., 2014). Compared to an H9N2 WIV vaccine, intramuscular immunization with the fusion protein demonstrated significantly elevated antibody– and cell-mediated responses against M2e and decreased viral shedding after H9N2 virus challenge, although the difference was non-significant. It would be important to determine if the M2e fusion protein offers protective efficacy in chickens against other clinically relevant strains of AIV. Developing mucosally-administered vaccines for poultry carries importance as the administration of vaccines is more feasible, and mucosal immune responses are primarily involved in AIV infection. Therefore, mucosal M2e vaccines have also been demonstrated in poultry for protection against H9N2. Specifically, a 4xM2e fusion protein was intranasally-administered to chickens with polyethyleneimine as an adjuvant (Hajam et al., 2019). The vaccine was able to produce cell-mediated responses in PBMCs and antibody responses in mucosal secretions and serum while eliminating shedding in greater numbers of H9N2 virus-challenged birds compared to the non-vaccinated group. Overall, protein vaccines represent a safe and targeted approach for developing AIV vaccines; however, enhancing their immunogenicity is important, especially if administration to mucosal surfaces is used.

Virus-like particle AIV vaccines

Another type of chicken AIV vaccine includes virus-like particle (VLP) vaccines. Due to the nature of virion formation, the production of virus particles that contain important immunogenic proteins, while not containing any viral genes or internal proteins, is possible. This offers many benefits, including the potential for DIVA, inability for recombination with wild-type AIVs, no need for virus inactivation and lack of any pre-existing vector immunity in chickens. Production of virus-like particles differs, including the capsid-based VLPs, which rely on multi-protein complexes, or envelope-based influenza VLPs that closely mimic influenza virion structures. Influenza VLPs typically contain HA or NA proteins which associate with viral membranes and the M1 protein, which helps form virus particles when expressed in cells or different expression systems. In addition, because these particles are similar to influenza virus virions when considering surface proteins, they interact with host cells in a similar fashion, leading to the formation of efficacious adaptive immune responses. VLPs for chicken AIV vaccines have been described recently (Rahn et al., 2015), highlighting existing production systems, including baculoviral-insect cell systems, production in transfected or stably expressing mammalian cell lines and production in plant cells.

Recent research on VLP vaccines has primarily employed baculovirus-insect cell systems for production. Insect cells, primarily Spodoptera frugiperda (Sf9) cells, can be infected with recombinant baculovirus for production of recombinant proteins or VLPs that can then be isolated from culture supernatant and used for immunization. The use of insect cells has promise for the generation of chicken and human influenza vaccines due to the decreased production costs compared with mammalian cells, and research efforts have demonstrated the ability to scale up production of influenza VLPs from insect cell expression systems (Lai et al., 2019). Experimentally, many different AIV VLP vaccines have been developed using this system and have been shown to be highly immunogenic in chickens, including an H9N2 VLP that contained H9 and N2 (Li et al., 2017), an H5N1 VLP with H5 (Wu et al., 2017), an H6N1 VLP that included H6 and M1 (Zhu et al., 2020) and an H7N9 VLP that contained H7, N9 and M1 (Hu et al., 2017). Baculovirus-insect cell systems have also been used to produce VLPs that express multiple influenza proteins or structurally modified influenza proteins. For example, Kang et al. (2019) demonstrated that an intramuscularly administered VLP vaccine that contained five M2e tandem sequences of avian, swine and human influenza virus origins efficaciously reduced virus shedding after H5N1 virus challenge in mice. In addition, a VLP vaccine that contained H5, H7, H9, N1 and a retroviral gag protein in place of M1 has been produced (Pushko et al., 2017). After subcutaneous administration, challenge with H5N1, H7N3 and H9N2 viruses demonstrated complete protection and decreased virus shedding in vaccinated birds. Inclusion of the nucleoprotein in an H5-, N1- and M1-containing VLP has also shown enhanced protection against lethal H5N1 virus challenge in chickens (Xue et al., 2015). Finally, baculovirus-insect cell expression systems have been manipulated to generate bivalent VLP vaccines against NDV and AIV (Xu, Qian et al., 2020). Importantly, this work showed that, when the bivalent vaccine also contained a dendritic cell stimulatory peptide fused to the HA protein, the efficacy of the vaccine, when administered intranasally to chickens, was increased. In addition to infection of insect cell cultures, recombinant baculovirus expressing H5 or H7 proteins have been used to infect silkworm pupae for subsequent processing and extraction of H5 and H7 VLPs (Nerome et al., 2017), both of which were capable of generating systemic HI antibody titres after oral immunization of chickens. Another candidate system for large-scale generation of influenza VLPs is via plant production. Recombinant bacteria expressing influenza proteins can infect leaves of plants such as Nicotiana benthamiana, leading to VLP accumulation in leaf tissue. Subsequent processing and extraction techniques can be employed to harvest influenza VLPs. For example, a plant-produced VLP that contained H6 and M1 was used to immunize chickens intramuscularly, resulting in significantly decreased respiratory and gastrointestinal H6N2 virus shedding after challenge when compared to chickens administered a conventional WIV H6N1 vaccine (Smith et al., 2020).

Other types of VLP vaccines also have potential as chicken AIV vaccines. Certain viral proteins react when combined to create geometric particles that contain specific numbers of monomeric viral proteins. Expression of these specific proteins in expression systems results in the formulation of protein-based VLPs. These VLPs can be genetically modified to have certain proteins or peptides fused to each monomeric protein. For example, an M2e-norovirus P particle vaccine has been generated, consisting of M2e fused to the norovirus P protein. When expressed in bacterial expression systems, the norovirus P protein forms a 24-mer VLP; in this case, the VLP also contains 24 copies of the M2e protein. This M2e-P particle vaccine has demonstrated efficacy in decreasing shedding in chickens after challenge with H5N2, H7N2 and H6N2 viruses (Elaish et al., 2015). Also, when combined with WIV vaccines, the M2e-P particle vaccine enhanced HI titres and decreased shedding of H5N2, H7N2 and H6N2 viruses in chickens (Elaish et al., 2017). Most recently, the M2e-norovirus VLP vaccine was administered in combination with a fusion H5 HA2-cyanase protein, demonstrating increased protection from H5N2 virus compared to administration of either vaccine alone (Elaish et al., 2019). This group also demonstrated that tripling the number of M2e copies on each norovirus P protein did not enhance M2e antibody responses; however, it did significantly reduce responses against the P protein.

Vaccine adjuvants and adjuvant delivery systems for chicken AIV vaccines

Novel adjuvants are being developed to elicit specific, protective, and long-term immunity after vaccination. Adjuvants can be chemical or biological substances that have diverse immunomodulatory roles. They are able to alter the nature of immune responses against specific antigens in vaccines. Many adjuvants have been licenced in human influenza vaccines including oil-based emulsions such as AS03, AF03 and MF59 in addition to other types such as potassium aluminium sulphate (alum) and heat-labile enterotoxin (LT) (Tregoning et al., 2018). In chickens, many experimental adjuvants are currently being studied, including immunostimulants such as TLR ligands, bacterial components, and recombinant proteins such as chicken cytokines.

Fundamentally, vaccine adjuvants enhance the ability of the host’s immune system to recognize an antigen and facilitate its elimination by the induction of innate and adaptive immune responses. Currently used vaccines mostly require a booster vaccination to induce protective responses in hosts, whereas adjuvants can potentially increase responses after a single administration, decreasing cost and labour associated with subsequent vaccinations. The mechanism of action employed by adjuvants can depend on their structural characteristics (Coffman et al., 2010). Adjuvants enhance adaptive immune responses by facilitating antigen recognition and presentation of antigenic peptides to T and B lymphocytes in secondary lymphoid organs, leading to induction of antibody- and cell-mediated immune responses (Calabro et al., 2011). Recognition of soluble antigens may be facilitated when they are adsorbed onto adjuvants such as alum, which prolongs the ability for antigens to be taken up by antigen-presenting cells. Adjuvants can also directly or indirectly recruit innate immune system cells to the site of vaccination leading to the production of local inflammatory reactions which later skew adaptive responses towards T helper (Th) 1 or Th2 responses.

Delivery of vaccine adjuvants faces several barriers, such as lack of stability and limited induction of local and mucosal responses (Lavelle & O’Hagan, 2006; Criscuolo et al., 2019). In order to overcome such limitations, adjuvant delivery platforms have been investigated and employed for more effective immunization. A number of delivery systems coupled with novel formulations have been used in chicken AIV vaccines, including synthetic and natural polymer-based micro- and nanoparticles, liposomes (Christensen et al., 2007), and virosomes (Mallick et al., 2011). Delivery systems function to increase immune responses by multiple mechanisms depending on their particular characteristics (Alving et al., 2012). Adjuvant delivery platforms are hypothesized to enhance vaccine efficacy by improving uptake of the adjuvant and increasing its stability, leading to successful induction of adaptive immune responses and circumventing the need for repeated administration to chickens (Engel et al., 2011). The following section summarizes recent findings surrounding different types of adjuvants and adjuvant delivery systems in poultry, specifically highlighting novel and existing adjuvants employed in chicken AIV vaccines and their effects on immunogenicity and protective efficacy.

Adjuvants for AIV vaccines

TLR ligands

The host immune system recognizes microbially conserved motifs, pathogen-associated molecular patterns (PAMPs), with the help of a number of germline-encoded pattern recognition receptors (PRRs). The interaction of PAMPs with PRRs facilitates intracellular signalling pathways in various immune system cells followed by the production of pro-inflammatory and anti-viral cytokines (Hopkins & Sriskandan, 2005). PRRs are expressed by different immune system cells and are located on cellular and endosomal membranes and in the cytoplasm. This feature enables them to recognize microbial invasion via many different mechanisms (Kawai & Akira, 2011). In addition to TLRs, PRR families include NOD-like receptors (NLRs), retinoic acid-inducible gene (RIG-I)-like receptors (RLRs), and C-type lectin receptors (CLRs) (Akira et al., 2006).

In recent years, microbial ligands or TLR ligands have been shown to act as prophylactic agents for protection against AIV infection in poultry. Due to the diverse immunostimulatory properties conferred by natural or synthetic TLR ligands, they are promising candidates for inclusion in poultry vaccines as adjuvants, and have been demonstrated as effective adjuvants for chicken AIV vaccines. For example, Pam3CSK4 and bacterial flagellin, ligands for TLR2 and TLR5, respectively, have been reported to enhance secretory IgA and serum IgY antibody titres when combined with an inactivated AIV vaccine in chickens. Secretory IgA has been shown to enhance mucosal immunity in the host by neutralizing influenza virus and inhibiting viral replication (Krammer et al., 2018). Therefore, the use of TLR2 and TLR5 ligand adjuvants can potentially contribute to vaccine-mediated protection from AIV infection in chickens.

TLR21 ligands have also been a focus of research in chicken AIV studies. For example, it was shown that the use of CpG ODN 2007, a synthetic TLR21 ligand, when combined with formaldehyde-inactivated H9N2 AIV vaccine, was associated with elevated serum antibody titres in chickens. Additionally, use of a lower dose of CpG ODN 2007 was reported to have higher HI and virus neutralization (VN) titres compared to a higher dose of CpG ODN 2007 (Singh et al., 2015). Combining CpG ODN 2007 with an inactivated AIV vaccine has also been shown to be superior compared to flagellin, a TLR5 ligand, in the generation of serum and lachrymal antibody-mediated responses (Astill, Alkie, Yitbarek, Taha-Abdelaziz, et al., 2018). When combined with a BPL-inactivated H9N2 vaccine, CpG ODN 2007 has been shown to produce cell-mediated immune responses in chickens (Astill, Alkie, Yitbarek, Abdelaziz, et al., 2018). In part, these responses could be explained by up-regulated expression of type II interferons (IFN) and MHC-II transcripts in the spleen, which has been observed in chickens after administration of TLR ligands (St. Paul et al., 2011). Specifically, St. Paul et al. (2011) demonstrated that intramuscular administration with LPS in chickens induced significantly up-regulated expression of type I IFNs, IFN-γ, interleukin (IL)-10, and IL-1β in the spleen, while treatment of chickens with CpG ODN 2007 led to upregulation of IL-10 and IFN-γ at later time-points post-treatment.

Furthermore, studies suggest that TLR ligands are effective in reducing H4N6 AIV shedding after experimental challenge. However, this effect varies depending on the type of ligand, dose and route of administration (Barjesteh et al., 2015). Specifically, intranasal and intramuscular treatment with two different doses of TLR2, 4, 7 and 21 in chickens showed variable reduction in virus shedding at different time-points. Intranasal treatment with CpG-ODN 1826, a TLR21 ligand, was the most effective at reducing oral shedding of the virus, while lipopolysaccharide (LPS) from Escherichia coli 026: B6, a TLR4 ligand, showed the maximum decrease in virus shedding through the cloacal route. The antiviral response observed was attributed to enhanced tracheal expression levels of type I IFNs and IFN-stimulated genes such as 2′,5′- oligoadenylate synthetase (OAS), protein kinase R (PKR), viperin and IFN-induced transmembrane protein 3 (IFITM3). It has been shown that respiratory epithelial cells can generate an anti-viral milieu and subsequently limit AIV replication in the trachea (Barjesteh et al., 2016). In accordance with the above study, activated dendritic cells and macrophages under the influence of TLR ligand treatment are also associated with enhanced immunity in the respiratory tract mucosa (Tamura & Kurata, 2004). In addition, stimulating chicken caecal tonsil mononuclear cells with CpG ODN 2007 in vitro also elicited enhanced innate responses compared to other TLR ligands (Taha-abdelaziz et al., 2016; Alqazlan et al., 2020). These lines of evidence raise the possibility of TLR ligands to serve as adjuvants for use in mucosally administered chicken AIV vaccines. This has been demonstrated, via the use of ocular AIV vaccines containing Pam2CSK4 or poly I:C, TLR2 and TLR3 ligands, in chickens (Takaki et al., 2016). Application of Pam2CSK4 in the chicken conjunctiva resulted in up-regulated expression of IL-1β. The above study also highlighted the ocular administration of TLR ligands as a potential way to increase serum IgY production in chickens.

A report by Hung et al. (2011) has demonstrated an alternative way to deliver certain TLR ligands as vaccine adjuvants, specifically by encoding ligand sequences within a plasmid. Multiple copies of a CpG motif (64CpG-plasmid) were encoded in a plasmid and administered with an inactivated H5N2 AIV vaccine to chickens. The results from the study revealed significantly higher expression of IFN-γ, TLR3 and TLR7 transcripts in chicken splenocytes (Hung et al., 2011). Encoding copies of CpG in plasmids offers a cost-effective method for large-scale production; however, the stability of plasmid approaches must be considered.

Other stimulatory microbial components, such as bacterial spores, can act as stimulatory adjuvants for AIV vaccines. For example, Lee et al. (2020) explored the potential of B. subtilis spores to be used as adjuvants with an inactivated H9N2 AIV vaccine. Administration of B. subtilis spores with an inactivated H9N2 AIV vaccine induced high serum antibody titres and enhanced IFN-γ and IL-17-producing T cells when compared to an oil emulsion adjuvanted H9N2 AIV vaccine. Other naturally derived microbial ligands have demonstrated efficacy in AIV vaccines. For example, Xiaowen et al. (2009) revealed that the intranasal delivery of an inactivated H5N2 AIV vaccine, adjuvanted with CpG ODN derived from E. coli, effectively induced systemic immunity and mucosal immunity in the upper respiratory tract and mucosal-associated lymphoid tissues in chickens. This study highlighted that CpG ODN can elicit both Th1 and Th2 responses, in addition to an increased number of CD3+ T lymphocytes, intraepithelial lymphocytes (IELs) and mast cells in the respiratory tract and increased concentrations of IgA and IgG antibodies.

Cytokines

Cytokines are mediators of cellular signalling which can be produced by the host in response to antigens and microbial insults. The induction of cytokines is a key mechanism of action for various adjuvants; therefore, it is possible to directly administer cytokines as adjuvants to enhance vaccine-induced responses directly. Moreover, cytokines may be used to directly trigger the host immune responses towards Th1 or Th2 responses. A variety of methods for the administration of cytokines has been explored, including encapsulation within liposomes (van Slooten et al., 2000) or biodegradable polymers (Mallick et al., 2011), conjugated with vaccine antigens, or cloned into different vaccine vectors such as vaccinia virus, facilitating the prolonged expression of the cytokine in the host.

Various chicken cytokines have been recognized to have immunomodulatory and antiviral properties against different viruses (Asif et al., 2004; Gan et al., 2019). For example, Schijns et al. (2000) assessed the ability of recombinant chicken cytokines such as IFN-α/ß, IFN-γ and IL-1ß to exert immunostimulatory activities on antibody-mediated responses in chickens. It was observed that recombinant E. coli expressing the above-mentioned cytokines facilitated elevated antibody responses when co-administered with a formalin-detoxified tetanus toxoid as a bacterial model agent. However, no improvement was noticed in antibody response when recombinant chicken IFNs (α/ß or γ) were co-administered with a formalin-inactivated infectious bursal disease virus (IBDV) vaccine. Thus, it can be inferred that recombinant bacteria expressing cytokines may enhance antibody responses in chickens more when compared to directly administered recombinant cytokines (Schijns et al., 2000). A report by Gan et al. (2019) highlighted the experimental application of prokaryotic expression systems to produce chicken IFN-α (rchIFN-α) and further tested its efficacy as a vaccine adjuvant when combined with an inactivated H9N2 AIV vaccine. The treatment resulted in enhanced antibody- and Th1-biased responses in chickens that received the recombinant cytokine (rchIFN-α) when compared to the chickens that received just the H9N2 AIV vaccine. The protective efficacy of the vaccine and cytokine adjuvant combination was assessed by challenging the chickens with a high dose of H9N2 AIV. The co-administration of rchIFN-α and H9N2 AIV vaccine resulted in a marked reduction in cloacal shedding of H9N2 AIV in challenged chickens (Gan et al., 2019).

IL-2 is recognized to have a significant role in the activation of conventional T cells and natural killer (NK) cells. A study by Xiaowen et al. (2009) tested the adjuvant potential of recombinant cytokine IL-2 (rIL-2) along with an H5N2 inactivated vaccine when delivered intranasally in 1-day-old chickens. The results from the study revealed enhanced levels of IgA and IgG in the trachea and lungs of the immunized chickens along with an increased number of mast cells (Xiaowen et al., 2009).

Expression of recombinant chicken cytokines fused to influenza virus proteins offers a candidate strategy to produce effective vaccines. Fusion constructs have been made using IL-2 and chicken granulocyte–macrophage colony-stimulating factor (GM-CSF), fused to influenza virus neuraminidase and haemagglutinin genes, respectively (Yang et al., 2009). A Madin–Darby canine kidney (MDCK) cell line, expressing the cytokine fusion proteins on the MDCK cell membrane, was established using the fusion constructs. A complete H3N2 influenza virus that contained membrane-embedded IL-2 and GM-CSF was harvested. Importantly, cytokine bioactivity was conserved after using various inactivation methods. Chickens that received the IL-2 bearing influenza virus vaccine showed enhanced systemic antibody levels, compared to chickens immunized with a conventional inactivated H3N2 influenza vaccine.

A novel approach exploring the adjuvant potential of IL-15 and IL-18 was evaluated in chickens to determine the induction of immune responses stimulated by an H5 DNA vaccine (Lim et al., 2012). The genes for IL-15 and IL-18 were amplified and then ligated into a plasmid pDisplay (pDis) vector. Serum HI titres in chickens injected with pDis/H5 + pDis/IL-15 were higher than the chickens vaccinated with pDis/H5 alone. Further, the chickens vaccinated with pDis/H5 + pDis/IL-15 showed an increase in blood CD8+ T cells when compared to the control chickens. These results indicate that the pDis/IL-15 adjuvant can facilitate cell- and antibody-mediated responses, which may have a protective role against influenza virus (Lim et al., 2012). The above findings suggest that cytokines possess an adjuvant potential and can provide improved protection in chicken AIV vaccines.

Delivery platforms for adjuvants

Alternative strategies aimed at enhancing the effectiveness of AIV vaccine adjuvants in chickens have also focused on mechanisms of delivery. For example, in an effort to focus the delivery of cytokine adjuvants to specific cells and tissues, various delivery platforms such as encapsulation within virosomes, nanoparticle systems, and liposomes have been explored. Other adjuvants, such as TLR ligands, can also be delivered this way along with AIV vaccines. Generating novel delivery systems has allowed for existing limitations associated with certain vaccine adjuvants to be overcome, such as a lack of stability and limited induction of local and mucosal immune responses. Most of the work on adjuvant delivery systems primarily focused on the encapsulation of adjuvants.

Polymeric particles

Synthetic and natural polymer-based nanoparticles (NP) are extensively used in the delivery of vaccine adjuvants. One polymer that has become well established for experimental adjuvant delivery in poultry AIV vaccines is poly D, L-lactic-co-glycolic acid (PLGA) based particles. PLGA particles possess several advantages, including being biodegradable and biocompatible, cost-effective, and they can be engineered to possess unique physicochemical properties (Malyala & O’Hagan, 2017). Polymer-based carrier systems have demonstrated versatility in encapsulating different types of adjuvants, leading to enhanced immune responses and effective mucosal vaccines (Swayne, 2009).

The use of NPs for the delivery of TLR ligands has previously been shown to enhance immune responses in chickens. TLR ligands can be delivered via NPs in different strategic ways, for example, via surface adsorption (Fischer et al., 2009), encapsulation (Alkie et al., 2017), or surface chemical conjugation (de Titta et al., 2013), to boost cell- and antibody-mediated immune responses. Studies focused on the administration of CpG ODN 2007 encapsulated within PLGA NPs combined with an inactivated H9N2 AIV vaccine have revealed increases in systemic IgY and mucosal IgA antibody titres in chickens (Singh et al., 2015).

The efficacy of PLGA-encapsulated CpG ODN 2007 and Pam3CSK4 was determined in vitro in chicken macrophages and in vivo in chickens (Alkie et al., 2017). CpG ODN 2007 encapsulated in PLGA NPs indicated enhanced responses in chicken macrophages signified by up-regulated expression of IL-1β, IFN-β and IFN-γ in chicken splenocytes when compared with encapsulated Pam3CSK4. This could be attributed to the stability of CpG ODN 2007 when encapsulated within PLGA compared to Pam3CSK4; however, differences in TLR-binding profiles of these ligands must also be considered.

There is great potential for the use of NP-based adjuvant delivery systems for mucosal AIV vaccines in chickens. Alkie et al. (2018) demonstrated an increase in mucosal immune responses against H4N6 AIV using NP-encapsulated TLR ligands (Alkie et al., 2018). It was revealed that the immunogenicity of an inactivated H4N6 AIV vaccine was increased when the inactivated H4N6 virus and CpG ODN 2007 were incorporated within PLGA NPs. Specifically, the PLGA vaccine formulation was administered subcutaneously in chickens, resulting in higher antibody titres when compared to the group treated with only AIV encapsulated PLGA NPs. Given these findings, there is potential for the use of PLGA NPs to encapsulate TLR ligand adjuvants in AIV vaccines for chickens.

The delivery of cytokine adjuvants within NPs has also been reported in mice. A proprietary calcium phosphate (CaP) nanoparticle system was utilized to deliver biologically active IFN-α via intratracheal instillation into the lungs of male BALB/c mice (Morçöl et al., 2018). The results showed that IFN-α formulated with CaP nanoparticles retained its biological activity in mice. Importantly, CaP NPs are non-toxic, non-inflammatory, and cause no adverse reactions when administered by injection or via mucosal routes. Therefore, the use of CaP NPs to deliver recombinant cytokines or TLR ligands as vaccine adjuvants has potential for chicken AIV vaccines, especially for mucosally-administered vaccines.

Liposomes

Liposomes are lipid bilayer membranes which enclose aqueous compartments, and their function is largely dependent on their physio-chemical properties. These particles have been shown to exhibit immunogenic activity and have been referred to as potential immunostimulatory carriers for vaccine adjuvants (Christensen et al., 2007). By the virtue of their slow and sustained release of entrapped molecules (antigen/cytokine/TLR agonists) at the vaccination site, liposomes have been suggested to enhance the delivery of antigens to mucosal surfaces (Dhakal et al., 2018).

Research by Chiou et al. (2009) explored intranasal delivery of bio-adhesive liposomes in chickens. The bio-adhesive liposomes were developed using low-viscosity bio-adhesive gels such as xanthan gum (XG) and tremella (T) lipopolysaccharides in conjunction with an inactivated H5N3 virus vaccine. The study showed that chickens vaccinated with a lower dose of bio-adhesive liposomal vaccine had higher serum and mucosal antibody levels compared to those immunized with a high dose of the bio-adhesive liposomal vaccine. The liposomal vaccine when combined with a low-viscosity gel resulted in enhanced antibody titres compared to liposomes combined with a high-viscosity gel (Chiou et al., 2009). The study suggested that when delivered mucosally, the low-viscosity gel liposomal vaccine induces elevated production of IgA in mucosal surfaces and IgG levels in the serum compared to high-viscosity gel formulation. Importantly, adding a bio-adhesive gel increases stability and prolongs antigen recognition on mucosal sites of the respiratory system.

In addition, liposomal-delivered divalent subunit vaccine formulations have been tested in mice to monitor immune responses generated against mouse tumour cell lines (Van Slooten et al., 2001). These vaccine preparations constituted either the native form of murine IFN-γ (m-IFN-γ) or m-IFN-γ encapsulated within a liposome as adjuvants. The encapsulated version proved to be significantly more effective than free m-IFN-γ, demonstrating an increase in immune responses to implanted tumour cells. While anti-tumour immune responses differ from anti-AIV responses, the potential of liposomal encapsulation of cytokines or other adjuvants has implications for chicken AIV vaccines.

Virosomes

Influenza virus virosomes represent another platform for AIV vaccine adjuvant delivery in chickens. Influenza virus virosomes are comparable in structure to influenza virus, possessing surface glycoproteins HA and NA on their surface but lacking internal genes and proteins. This structural makeup promotes virosome attachment and entry into host cells without replication (Huckriede et al., 2003). The presence of HA and NA proteins on influenza virosomes is of additional advantage over other delivery platforms which do not possess any self-antigenic properties (Almeida et al., 1975). Additionally, due to their ability to enter host cells, virosomes can lead to the presentation of antigenic peptides via both MHC class I and II pathways (Bungener et al., 2002), and therefore can efficiently cause stimulation of cytotoxic T lymphocytes and antibody-mediated immune responses.

Studies have highlighted the incorporation of CpG-ODN 2007 within virosomes. Mallick et al. (2011) reported that the immunization with virosomes prepared from inactivated H4N6 AIV which encapsulated CpG-ODN 2007 elicited high anti-H4N6 HI and virus neutralization titres with increased IgG and IgA serum antibodies when compared to groups that received only the H4N6 virosome (Mallick et al., 2011). Further, immunization with CpG-ODN 2007 encapsulated in H4N6 virosomes resulted in antigen-specific cell-mediated responses in chicken splenocytes, as evidenced by increased IFN-γ expression when splenocytes from vaccinated birds were re-stimulated in vitro with inactivated H4N6 AIV.

Cytokines have also been co-delivered in encapsulated form within virosomes along with AIV vaccine antigens. Mallick et al. (2011) investigated a virosome-based H4N6 AIV vaccine and assessed its immunogenicity in chickens when encapsulating recombinant chicken IFN-γ protein (Mallick et al., 2011). The encapsulation of recombinant chicken IFN-γ within the virosome augmented systemic antibody-mediated immune responses leading to B cell activation and class switching, ultimately increasing antigen-specific IgA and IgG responses.

Multiple factors should be considered in designing effective AIV vaccines for poultry, including the use of novel adjuvants along with adjuvant delivery systems. Microbial ligands and cytokines are highly immunostimulatory and have demonstrated efficacy in experimental poultry AIV vaccines. To further increase the effectiveness of these adjuvants in poultry AIV vaccines, delivery mechanisms that increase adjuvant stability and prolong adjuvant interaction with immune system cells must be further explored.

Probiotics and dietary factors for enhancing immune responses to influenza viruses

The GIT microbiota plays an essential role in viral pathogenesis. Most viral infections occur through mucosal tissues and encounter commensal microbiota on these surfaces, including the respiratory system and GIT (Yitbarek, Alkie, et al., 2018; Yitbarek, Taha-Abdelaziz, et al., 2018). It is also known that the induction and regulation of host immune responses to different respiratory pathogens, including viruses, are mediated via GIT microbiota. Further, GIT microbiota is critical for maintaining basal levels of type I IFN, an essential factor of antiviral responses (Schaupp et al., 2020).

Innate responses to different antigens in the GIT are mediated through GIT-associated lymphoid tissues (GALTs) such as the caecal tonsils and Peyer’s batches (Casteleyn et al., 2010; Lammers et al., 2010). The mechanisms involved in GIT microbiota-mediated immunity to respiratory pathogens involve PRRs, including TLRs (Ichinohe et al., 2011; Fagundes et al., 2012; Clarke, 2014) and NLRs (Samuelson et al., 2015; Kim & Kim, 2017). Various adjuvants mimic commensal microbiota molecular structures which can stimulate PRRs, resulting in enhanced vaccine efficacy and reduced virus shedding in infected chickens (Barjesteh et al., 2015; Singh et al., 2016). This indicates that GIT microbiota can contribute to vaccine efficacy.

Induction of immune responses against mucosal vaccines can be challenging due to the phenomenon of immune regulation at mucosal surfaces in the mucosa-associated lymphoid tissues (MALTs). TLR ligands are often used as adjuvants to induce protective innate and adaptive immune responses. Although TLR ligand application requires the formulation of an oral or injectable form of these ligands, they have the potential for cost-effective use in poultry. As part of MALTs, GALTs play a significant role in developing and modulating local and systemic immune responses through interaction with various antigens from food, GIT microbiota and pathogens (Brisbin et al., 2010; Lammers et al., 2010; Tokuhara et al., 2019). Furthermore, manipulating GALTs with treatments such as probiotics, prebiotics, or vitamins is a potential approach for preventing disease and enhancing immune responses to AIV infection and vaccination.

Probiotics

The most utilized lactic acid-producing bacteria in the meat and dairy industry are Lactobacillus, Lactococcus and Bifidobacterium. Some strains of these bacteria have probiotic capacities. Although mechanisms are not entirely understood, lactic acid-producing bacteria can confer health benefits onto the host (Bajaj et al., 2015; Ren et al., 2021), such as improved immune responses, in a variety of species (Shah, 2007; Khalesi et al., 2019), including chickens (Brisbin et al., 2010). Evidence also suggests that dietary inclusion of lactic acid bacteria can help reduce enteric diseases in farm animals by various mechanisms, including competitive exclusion mechanisms for pathogenic bacteria and production of bioactive molecules with bactericidal activities (Axelsson & Ahrné, 2000; Avila et al., 2010; Argyri et al., 2013).

There is now accumulating evidence that the chicken intestinal microbiome has a profound effect on the development of immune responses to viruses, such as AIV. For example, it has been determined that interactions of chicken macrophages with lactobacilli induce antiviral responses against avian influenza virus (Shojadoost, Kulkarni, Brisbin et al., 2019). Also, it has become clear that virus infections could impact the intestinal microbiome composition. In a recent study, infection of chickens with H9N2 AIV disrupted GIT microbiota leading to higher host susceptibility to viral replication, higher virus shedding and dysregulated type I IFNs (Yitbarek, Alkie, et al., 2018; Yitbarek, Weese, et al., 2018). Type I IFNs play an essential role in innate responses and blocking of virus replication (Iwasaki & Pillai, 2014), leading to virus clearance (Abt et al., 2012; Barjesteh et al., 2015). If the virus evades the activated innate responses, adaptive immune responses are induced, including cell- and antibody-mediated responses (Pang & Iwasaki, 2012; van de Sandt et al., 2012). Double-stranded RNA originating from lactic acid bacteria induces basal levels of type I IFNs, in contrast to double-stranded RNA from pathogenic bacteria where this effect is not observed (Kawashima et al., 2013).

There have been attempts to employ the microbiome to enhance host responses to AIV. For instance, several studies have recently investigated the effects of probiotic lactobacilli on immune responses after AIV infection and vaccination in chickens. To this end, Yitbarek and colleagues showed that the complex GIT microbiota in chickens can influence responses to AIV infection (Yitbarek, Taha-Abdelaziz, et al., 2018). Several members of the microbiota may exert such functions, among which lactobacilli are of interest. Probiotic lactobacilli regulate GIT microbiota composition, representing a potential explanation for different responses to influenza virus infection and vaccination (Yitbarek, Taha-Abdelaziz, et al., 2018; Yitbarek et al., 2019). For example, chickens with depleted GIT microbiota show higher virus shedding and impaired type I IFN responses compared to chickens with intact microbiota following AIV infection. However, type I IFN responses were shown to be restored in GIT microbiota-depleted AIV-infected chickens that received probiotic lactobacilli treatment or faecal microbiota transplant (FMT) (Yitbarek, Alkie, et al., 2018).

Administration of probiotic lactobacilli in chickens has been shown to modulate antigen-specific antibody responses (Yitbarek, Alkie, et al., 2018; Yitbarek et al., 2019), mainly due to the presence of microbe-associated molecular patterns (MAMPs) and microbial metabolites, such as short-chain fatty acids and ATP (Kim & Kim, 2017). Administration of probiotics to non-microbiota-depleted chickens resulted in significantly higher HI titres compared to antibiotic-treated chickens after vaccination (Yitbarek et al., 2019). Further antibody analysis showed significantly higher IgM and IgG titres in non-microbiota-depleted chickens treated with probiotics compared to microbiota-depleted chickens. Also, significantly higher IgG titres were observed in chickens treated administered FMT following antibiotic-mediated microbiota depletion compared to those not administered FMT (Yitbarek et al., 2019). These results indicate the involvement of GIT microbiota in adaptive immune responses through the induction of antibody production and possibly cell-mediated responses.

Regarding the underlying mechanisms that connect the GIT microbial composition to immunity against AIV, significantly downregulated levels of IL-22, a cytokine involved in GIT tissue integrity regulation, have been observed in multiple tissues of antibiotic-depleted chickens, including in the caecal tonsils and especially in the ileum. However, upon treatment with probiotic lactobacilli or FMT, the expression of IL-22 was restored to similar levels observed in non-microbiota-depleted chickens (Yitbarek, Taha-Abdelaziz, et al., 2018). These results suggest the involvement of GIT microbiota in modulating the expression of IL-22 in GIT-associated tissues. More importantly, after H9N2 virus infection, the expression of IL-22 in the probiotic- or FMT-treated groups was significantly higher than that of the non-microbiota-depleted chickens. However, the direct effects on the immune system should be considered in addition to potential effects on the respiratory system microbiota when examining results from this antibiotic-depletion model (Yitbarek, Taha-Abdelaziz, et al., 2018).

There is evidence that probiotics can decrease AIV replication in mucosal tissues; for example, treatment of chicken macrophages with lactobacilli resulted in the induction of antiviral responses towards AIV (Shojadoost, Kulkarni, Brisbin et al., 2019). Further, in vitro treatment with lactobacilli has been shown to induce antiviral and immunostimulatory responses in chicken caecal tonsil cells against H9N2 virus, which resulted in a reduction of AIV replication in these cells after infection (Alqazlan et al., 2020). In addition, Alqazlan et al. (2020) demonstrated that lactobacilli alone, or when combined with CpG ODN, induce robust mRNA expression of Th1-type cytokines (IL-2, IL-12, IFN-γ), crucial for host defence against intracellular microbial agents. Also, lactobacilli alone, or combined with CpG ODN, stimulate mRNA expression of pro-inflammatory cytokines (IL-1β and IL-6), the immunoregulatory cytokine IL-10, which plays a vital role in maintaining immune homeostasis and B cell survival and differentiation (Alqazlan et al., 2020), in addition to interferon-stimulated genes, including viperin. These results raised the possibility that these effects could be attributed to the immunostimulatory effects of probiotic lactobacilli, which induced caecal tonsil cells to mount innate responses against AIV.

When administered in ovo, lactobacilli treatment downregulates the expression of inflammatory cytokines IL-2, IL-6 and IL-8 in the caecal tonsils, which further confirms the immunomodulatory properties of lactobacilli and their role in maintaining homeostasis of the immune system (Alizadeh et al., 2020). However, the aforementioned study also demonstrated up-regulated cytokine expression in the spleen. In contrast, IL-13 was up-regulated in the bursa of Fabricius, suggesting that lactobacilli might differentially modulate cytokine expression profiles in systemic (spleen) and local (caecal tonsils) secondary lymphoid organs (Alizadeh et al., 2020). Higher expression of IL-13 in the bursa of Fabricius of lactobacilli-treated birds also indicates the role of lactobacilli as beneficial commensal bacteria in B cell development (Alizadeh et al., 2020). Overall, the immunomodulatory effects observed from treating chickens with probiotics suggest their potential use to enhance the effectiveness of AIV vaccines for poultry.

When considering the GIT microbiota and influenza vaccine responses, it has been shown in both germ-free and antibiotic-depleted mice that antibody-mediated responses to influenza virus infection and vaccination with a trivalent inactivated influenza virus vaccine are compromised (Abt et al., 2012; Oh et al., 2014). In addition, in chickens, administration of five Lactobacillus species to non-microbiota-depleted birds resulted in higher HI and IgM titres compared to chickens treated with antibiotics (Yitbarek et al., 2019). Systems biology approaches have demonstrated correlations between the expression of genes related to innate responses, such as TLRs, and HI titres after vaccination in humans (Nakaya et al., 2011), suggesting the importance of innate stimulation following vaccination. To address this, a study was conducted to combine probiotics and innate stimulation to explore the effects of this combination on the generation of immune responses after AIV vaccination. This study showed that probiotic lactobacilli enhance the immunogenicity of an inactivated H9N2 AIV vaccine when combined with CpG ODN 2007 in chickens (Alqazlan et al., 2021). In this study, probiotic lactobacilli administration combined with a CpG ODN 2007 adjuvanted vaccine induced significantly higher HI titres compared to lactobacilli or CpG ODN 2007 alone. Additionally, probiotic lactobacilli administration enhanced cell-mediated immune responses, as demonstrated by a significant increase in IFN-γ production in chicken splenocytes (Alqazlan et al., 2021). These findings suggest further investigation to elucidate the interactions between innate and adaptive responses to help understand how modulation of the GIT microbiota can be used to enhance protective responses to influenza virus vaccines. Future studies on probiotics may enable the development of adjuvants from probiotic bacteria, including ligands generated from cell wall constituents, cytoplasmic components, or metabolites.

Prebiotics

Beyond probiotics, prebiotics, such as yeast-derived carbohydrates (YDC), are also known for health benefits and potential immunomodulatory properties (Alizadeh et al., 2017). Prebiotics are compounds which induce the development or activity of beneficial microorganisms and are able to alter GIT microbiota. Therefore, it is conceivable that prebiotic administration can affect the innate and adaptive responses; for example, dietary supplementation with YDC and symbiotic bacteria (SNB) (Lactobacillus acidophilus, L. casei, Streptococcus faecium, and Bacillus subtilis) has been found to modulate innate responses of pullet chicks (Yitbarek et al., 2015). YDC and SNB can enhance antibody-mediated immune responses and extend the decay rate of maternal antibodies (Guo et al., 2003; Ghosh et al., 2012). Despite the commercial use of prebiotics, there is a paucity of information about the efficacy of these compounds to enhance vaccine-induced immune responses in chickens.

Prebiotics and cell wall polysaccharides present in YDC (β1,3–1,6-glucans and mannan) can function as MAMPs which can modulate immune responses through interaction with PRRs. Also, it has been shown that MAMPs modulate the production of cytokines (IL-4, IL-13, and IL-10) that are involved in antibody production (Christensen et al., 2002; Shashidhara & Devegowda, 2003; Jawhara et al., 2012; Dibaji et al., 2014). The combination of SNB and YDC in poultry diets has shown synergistic effects, including increased IgG concentrations in the serum of pullet chickens, implying possible protection of pullet chickens against invading pathogens and, consequently, lowering mortality in the early stages of life (Alizadeh et al., 2017). Moreover, SNB-supplemented diets increase the expression of IL-10 and IL-4, which are essential cytokines involved in B cell proliferation and differentiation to antibody-producing plasma cells (Yitbarek et al., 2015). Nevertheless, the effects of dietary YDC or SNB on serum IgG titres against infectious bronchitis virus after week 3 were found to be not significantly different compared to birds not administered these prebiotics (Alizadeh et al., 2017). This result has also been observed for total serum IgG in broilers (Midilli et al., 2008). Alternatively, treatment with a synbiotic preparation (Biomin® IMBO), enhances antibody responses following vaccination against H9N2 AIV compared to chickens that only received the vaccine (Talebi et al., 2015). These findings suggest that prebiotic supplementation may be most effective when given to birds along with probiotics. However, further research is required to elucidate the ability of prebiotics to alter immune responses to AIV vaccines in chickens.

Dietary supplements

Immunosuppression in the poultry industry often occurs due to poor management conditions and nutritional status, intensive production systems and high-density rearing, leaving birds susceptible to infectious diseases. Nutritional supplementation of the chicken diet can enhance immune system function and improve responses to different pathogens. For example, feed supplementation with select vitamins, minerals and amino acids can enhance immunity by augmenting antibody- and cell-mediated immune responses, mainly when applied in amounts that exceed basic recommended requirements. Recommended doses of many nutrients are currently sufficient to prevent deficiency-related diseases, but are not high enough to stimulate enhanced levels of immune responses (Nockels et al., 1996). Also, poultry gain little or no benefit from microbial synthesis of vitamins in the gastrointestinal tract, therefore, there is a higher requirement for vitamins and minerals in the diet (McDowell & Ward, 2009). Chicken diets supplemented with minerals and vitamins have shown benefits to the immune system, demonstrated by responses to infectious antigens and experimental antigens.

Feed supplementation strategies have been applied to enhance innate responses against viruses, including AIV, and augment vaccine-induced protective immune responses (Wu et al., 2019). The effects of vitamins A, D, E, and C on the poultry immune system have recently been reviewed (Shojadoost et al., 2021) and many aspects of the effects of vitamins are highly relevant to poultry and AIV vaccination. For example, vitamin E has been widely used in animal feeds due to its high efficiency as a natural antioxidant (Chen et al., 1998). It is especially crucial for the poultry industry since chickens cannot synthesize vitamin E (Chan & Decker, 1994). Vitamin E is known to regulate immune system cell function (Moriguchi & Muraga, 2000), including prevention of inflammatory reactions by suppressing activation of nuclear factor (NF)-κB (Calfee-Mason et al., 2004; Trushenski & Kohler, 2007). Vitamin E has also shown immunostimulatory effects on antibody- and cell-mediated immune responses in chickens vaccinated with an inactivated H9N2 AIV vaccine (El-Ela et al., 2016). Another important antioxidant is vitamin C (ascorbic acid), necessary for the biosynthesis of various vital compounds such as collagen, carnitine, 1,25-dihydroxy vitamin D and adrenaline. It also regulates diverse reactions (e.g. secretion of corticosterone and regulation of body temperature) and activation of the immune system (Kutlu, 2001; McDowell & Ward, 2009). A positive synergistic effect of vitamin E and vitamin C on the immune response has been observed due to their high antioxidant activities (Yin et al., 1993). Interestingly, a synergistic effect of these combined vitamins has been documented in mice and birds in their ability to reduce influenza virus infection (Tantcheva et al., 2003; Barbour et al., 2007).

In addition to vitamins, minerals, such as selenium, have shown beneficial effects on bird health, performance, and immune system function, and selenium is widely used as a feed supplement for chicken (Shojadoost, Kulkarni, Yitbarek et al., 2019). Selenium exhibits high antioxidant activity which is beneficial in broiler chickens during heat stress (Rao et al., 2016). Furthermore, selenium is required for normal function of innate and adaptive immune responses (Turner & Finch, 1991; McKenzie et al., 2001; Arthur et al., 2003). The roles of selenium in the immune system are attributed mainly to its antioxidant effects as it regulates the function of glutathione peroxidase (Baker et al., 1993). Importantly, selenium has been shown to enhance immune responses in chickens against AIV and AIV vaccines. For example, dietary supplementation with selenium was shown to enhance vaccine-induced antibody responses against an H9N2 AIV vaccine (Shojadoost et al., 2020). In this regard, Shojadoost et al. (2020) found that the source of dietary selenium affected the outcome of vaccination; specifically, an organic source of selenium (selenium-enriched yeast) enhanced HI titres compared to the effects of an inorganic source (sodium selenite). This difference might be due to the enhanced bioavailability of organic selenium (Delezie et al., 2014). Nevertheless, H9N2 virus shedding in vaccinated birds post experimental infection was decreased in birds that received both sources of selenium (Shojadoost et al., 2020). These same organic and inorganic sources of selenium have been used in an H9N2 virus challenge study, where it was demonstrated that both types of selenium led to decreased oral and cloacal virus shedding compared to the control, in addition to increased expression of IFN-stimulated genes and cytokines (IFN-α, IFN-β and IFN-γ) in the caecal tonsils and spleen, respectively (Shojadoost, Kulkarni, Yitbarek et al., 2019).

Among alternative dietary strategies for enhancing immune responses to AIV in chickens, Hypericum perforatum extract has been investigated in hens as a dietary supplement (Jiang et al., 2012). Specifically, the extract was associated with increased antibody titres to an inactivated H5N1 reassortant virus vaccine. Jiang et al. (2012) observed that different doses of Hypericum perforatum extract led to different effects on antibody response; specifically, 500 mg/kg in feed was found more effective than 250 mg/kg or 1000 mg/kg.

Conclusions

In summary, vaccinating poultry is an essential part of a multi-pronged strategy to decrease infection and transmission of AIV in poultry. Vaccines that include HA proteins must ensure that the HA protein homology matches that of wild-type AIVs. Nevertheless, many vaccines have targeted conserved AIV proteins or peptides such as the matrix protein and nucleoprotein, and have demonstrated potential efficacy as universal vaccines. It will be crucial for AIV vaccine strategies to ensure that large-scale production is feasible and facilitates DIVA. Along with targeting vaccine strategies, the immune response induced in poultry can be enhanced further by including immunostimulatory adjuvants and by developing novel mechanisms to deliver adjuvants and vaccines with enhanced stability to mucosal tissues. Furthermore, the effectiveness of vaccines and adjuvants relies on multiple parameters of host health and function, which includes the GIT microbiota. Maintaining poultry GIT microbiota through the use of probiotics and prebiotics, along with the use of well-characterized dietary supplements, can increase the effectiveness of current poultry AIV vaccines. Together, the strategies presented in this review provide a better understanding of the complex factors that contribute to the immunogenicity and protective efficacy of poultry AIV vaccines.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The support for the current work was provided by the Ontario Ministry of Agriculture, Food and Rural Affairs, and Natural Sciences and Engineering Research Council of Canada (NSERC). We are also thankful to Chicken Farmers of Saskatchewan, Egg Farmers of Canada, and Canadian Poultry Research Council for funding. Funding was also provided by the University of Guelph’s Food from Thought initiative. S.R. was recipient of an Arrell Graduate Scholarships. N.A. was a recipient of the Saudi Ministry of Education Scholarship.