Antigen identification and subunit vaccines

During the late 1970s and early 1980s, there was growing recognition that bacterial cells grown in vitro were quite distinct from the same organism grown in the host or under conditions that partially mimicked those found in the host. Many organisms have adapted to survival in environments with different nutrient sources and physical conditions and they have evolved to include mechanisms to sense their environment and produce components appropriate to the specific conditions. A classic example of this is growth under conditions where free iron is limited. Gram-negative organisms adapt to grow in the absence of free iron by producing a number of novel proteins on their surface which are able to sequester iron from host iron-binding proteins such as transferrin, lactoferrin, heme compounds and others (Bullen et al., Reference Bullen, Rogers and Griffiths1972, Reference Bullen, Rogers and Griffiths1978; Griffiths et al., Reference Griffiths, Rogers and Bullen1980, Reference Griffiths, Stevenson, Hale and Formal1985). Members of the Pasteurellaceae are excellent examples of this phenomenon (Deneer and Potter, Reference Deneer and Potter1989a, Reference Deneer and Potterb; Ogunnariwo and Schryvers, Reference Ogunnariwo and Schryvers1990; Ogunnariwo et al., Reference Ogunnariwo, Alcantara and Schryvers1991, Reference Ogunnariwo, Woo, Lo, Gonzalez and Schryvers1997) and a number of groups have demonstrated that vaccines enriched in these novel proteins were capable of inducing enhanced protection following immunization (Gilmour et al., Reference Gilmour, Donachie, Sutherland, Gilmour, Jones and Quirie1991; Gerlach et al., Reference Gerlach, Klashinsky, Anderson, Potter and Willson1992; Potter et al., Reference Potter, Schryvers, Ogunnariwo, Hutchins, Lo and Watts1999). Recombinant subunit vaccines based on the transferrin-binding protein TfbB are available for A. pleuropneumoniae and conventional vaccines have been developed for both bovine and ovine isolates of M. haemolytica. A number of laboratory techniques have since been developed to identify in vivo-expressed antigens, including in vivo expression technology (IVET) and conceptually related technologies (Slauch et al., Reference Slauch, Mahan and Mekalas1994; Slauch and Camilli, Reference Slauch and Camilli2000) and this has become a common methodology for antigen identification. Both IVET and signature-tagged mutagenesis have been successfully used in veterinary pathogens (Fuller et al., Reference Fuller, Shea, Thacker and Mulks1999, Reference Fuller, Kennedy and Lowery2000). In conjunction with traditional techniques for protein characterization, a number of virulence determinants with a role in infection have been targeted for veterinary vaccine development, including those based on adhesins, toxins, outer membrane proteins, capsular polysaccharide and others.

The field of genomics has revolutionized how bacterial antigens are identified since the first bacterial genome was sequenced over a decade ago (Fleischmann et al., Reference Fleischmann, Adams, White, Clayton, Kirkness, Kerlavage, Bult, Tomb, Dougherty, Merrick, Sutton, FitzHugh, Fields, Gocayne, Scott, Shirley, Liu, Glodek, Kelley, Weidman, Phillips, Spriggs, Hedblom, Cotton, Utterback, Hanna, Nguyen, Saudek, Brandon, Fine, Fritchman, Fuhrmann, Geoghagen, Gnehm, McDonald, Small, Fraser, Smith and Venter1995). Researchers now have access to sequences covering virtually all pathogens of animals and the bioinformatics tools to identify surface proteins, secreted products, B- and T-cell epitopes, etc. are commonly used in the vaccine development field (Scarselli et al., Reference Scarselli, Giuliani, Adu-Bobie, Pizza and Rappuoli2005). The first example of the use of a genome sequence to produce vaccine antigens was described by Pizza et al. (Reference Pizza, Scarlato, Masignani, Giuliani, Arico, Comanducci, Jennings, Baldi, Bartolini, Capecchi, Galeotti, Luzzi, Manetti, Marchetti, Mora, Nuti, Ratti, Santini, Savi, Scarselli, Storni, Zuo, Broeker, Hundt, Knapp, Blair, Mason, Tettelin, Hood, Jeffries, Saunders, Graff, Venter, Moxon, Grandi and Rappuoli2000) for Neisseria meningitidis and several other examples with veterinary application followed. The use of these techniques is only useful in the identification of putative protective antigens and they must still be screened for their ability to induce protective immunity in animal models. However, for organisms that are fastidious in nature or non-cultivatable, genomic approaches can cut vaccine development times by several years.

Live vaccines

Live attenuated bacterial vaccines have several potential advantages over killed or subunit products, but their use in veterinary medicine has been limited to date. These advantages include the presence of the full repertoire of protective antigens, including those produced in vivo, the ability to induce balanced Th1/Th2 responses, the potential for single dose administration and lower production costs than subunit vaccines. Historically, attenuation has been carried out by using altered growth conditions, chemical mutagenesis or passage through alternative hosts. The use of Brucella abortis strain 19 is an excellent example of a traditional live vaccine that has proved effective in the control of brucellosis (Nicoletti and Milward, Reference Nicoletti and Milward1983). There are two areas that have limited the use of live vaccines in animals. First, the routine use of antibiotics in animal husbandry has precluded their use for obvious reasons, since there is a need for limited replication and transient colonization to occur. The second issue has been one of safety; since the attenuating mutation(s) were not known in many prototype vaccines, there have been concerns of reversion to virulence. Advances in molecular genetics in the 1970s and 1980s made it possible to rapidly construct defined insertions and deletions in specific genes, making rational attenuation possible and the first live attenuated bacterial vectors included transposon insertions of Salmonella enterica serovars (Stocker, Reference Stocker2000). These defined mutations resulted in safer vaccine strains, which could also be used for delivering a variety of foreign antigens to animals (Stocker, Reference Stocker2000; Thole et al., Reference Thole, van Dalen, Havenith, Pouwels, Seegers, Tielen, van der Zee, Zegers and Shaw2000; Garmory et al., Reference Garmory, Leary, Griffin, Williamson, Brown and Titball2003). Furthermore, by selecting the appropriate target genes for deletion, it is possible to tailor the degree of attenuation in a similar fashion to that routinely carried out in the development of modified live viral vaccines. For example, classical aroA mutants of S. enterica serovar Typhimurium (Salmonella Typhimurium) are able to persist in the host to a greater degree than the same strain carrying defects in purine metabolism (Charles and Dougan, Reference Charles and Dougan1990). Target genes for rational attenuation include those coding for proteins involved in physiological processes as well as those coding for virulence determinants. Numerous examples of each have been described and representative examples are listed in Table 2. The use of defined mutations is not without risk since vaccine strains are still capable of interacting with other microbes in the host. For example, the administration of Salmonella Typhimurium aroA mutants was observed to exacerbate infection with wild-type Salmonella Typhimurium in a mouse model of infection via a Toll-like receptor-4 (TLR-4)-dependent mechanism (Foster et al., Reference Foster, Barr, Grant, McKinley, Bryant, Macdonald, Gray, Yamamoto, Akira, Maskell and Mastroeni2008).

Table 2. Examples of attenuating mutations used in live bacterial vaccines

Other vaccine types

While live and killed vaccines account for the majority of licensed products currently available, a number of alternative technologies are also being employed for the development of new vaccines. These include DNA vaccines, bacterial spores, bacterial ghosts and peptide-based technologies. Immunization with plasmid DNA-containing genes coding for protective antigens from bacteria, viruses and parasites has been common practice in research laboratories for the past decade and a half. There has been significant interest in developing the technology for use in animals since it has a number of potential advantages over existing production methods. These include low production cost, correct post-translational modification for viral antigens, ability to induce balanced Th1/Th2 responses without an adjuvant and finally the ability to induce responses in the face of maternal antibodies. Most of the literature dealing with DNA vaccines has described the use of viral antigens delivered in mouse models and the results have been promising. However, when large animals have been used, results have been disappointing. Cox et al. (Reference Cox, Zamb and Babiuk1993) described the first example of vaccination of a non-rodent species with a DNA vaccine and they showed that immunization of cattle with nucleic acid encoding a bovine herpesvirus-1 glycoprotein was able to induce cellular and humoral immune responses. However, the dose of plasmid DNA was high, and repeated immunizations were required for the induction of immunity. Since that time, DNA vaccines have been shown to be effective in cattle, sheep, poultry, horses and fish and commercially available products have been produced for the latter two species. No vaccines for bacterial diseases are currently available, although experimental vaccines have been described for Mycobacterium avium subspecies paratuberculosis, Campylobacter jejuni, B. anthracis, Pasteurella multocida, Salmonella species and others. These are unlikely to see commercial development until the plasmid dose can be reduced and the method of delivery optimized (e.g. prime-boost strategies). Gerdts (Gerdts et al., Reference Gerdts, Babiuk, van Drunen and Griebel2000) demonstrated that in utero immunization of sheep resulted in protective immunity, a finding which has implications for potential fetal delivery of animal vaccines.

The development of oral vaccines for animals has been a goal of many groups, with most work focused on the use of live attenuated or live vectored vaccines. Subunit antigens containing mucosal adjuvants such as the cholera toxin or heat-labile enterotoxin have also been successfully employed, as have plant-derived vaccines (Rice et al., Reference Rice, Ainley and Shewen2005). However, the use of oral subunit vaccines in large animals remains problematic due to degradation of the antigen and poor absorption in the gut. There has been recent interest in the use of bacterial spores as delivery vehicles for heterologous antigens, with B. subtilis and B. anthracis representing two of the hosts used (Aloni-Grinstein et al., Reference Aloni-Grinstein, Gat, Altboum, Velan, Cohen and Shafferman2005; Zhou et al., Reference Zhou, Xia, Hu, Huang, Li, Li, Ma, Chen, Hu, Xu, Lu, Wu and Yu2008a, Reference Zhou, Xia, Hu, Huang, Ma, Chen, Hu, Xu, Lu, Wu and Yub). Spores have a number of advantages over conventional live vaccines due to their stability and resistance to degradation by chemical and physical means. Foreign antigens, including a fragment of tetanus toxin and a Clonorchis sinensis tegument protein, have been fused to a major spore coat protein (CotC) and, following oral administration in mice, shown to induce protective immunity (Zhou et al., Reference Zhou, Xia, Hu, Huang, Li, Li, Ma, Chen, Hu, Xu, Lu, Wu and Yu2008a, Reference Zhou, Xia, Hu, Huang, Ma, Chen, Hu, Xu, Lu, Wu and Yub). This represents an intriguing possibility for the production of low-cost, stable oral vaccines for animals.

Vaccine targets

Vaccines for bacterial pathogens of livestock and poultry have historically been targeted at organisms that were significant disease threats in the target species. In the case of zoonotic diseases, these included brucellosis, anthrax, clostridial disease and others. Vaccination was able to protect not only the immunized host, but it also reduced the risk of transmission to humans through a reduction in the pathogen reservoir. However, many newer zoonotic diseases of interest do not cause clinical signs of disease in the target species. For example, the 33 million cases of food-borne illness reported each year in North America are due to a variety of bacteria, viruses and parasites with over one-half belonging to six bacterial species including C. jejuni, Shiga toxin-producing Escherichia coli, non-typhoid S. enterica species, Listeria monocytogenes, Staphylococcus aureus and Clostridium perfringens. Most chickens are exposed to C. jejuni and S. enteriditis during their life with no clinical signs of infection and E. coli O157:H7 colonizes cattle with no impact on health or productivity. From a vaccine development perspective, protection against colonization rather than disease presents several issues not encountered with traditional disease targets ranging from the choice of antigens to be included in the vaccine to regulatory issues regarding the choice of animal models. The first vaccines for non-disease-causing zoonotic pathogens were developed in the United Kingdom for Salmonella in poultry (Gast et al., Reference Gast, Stone and Holt1993). These killed vaccines were shown to significantly reduce contamination of eggs and, since then, live attenuated vaccines have also been commercialized. No vaccines are commercially available for C. jejuni or E. coli O157 although proof of concept for immunization against both bacterial species has been demonstrated in chickens and cattle, respectively. In the case of E. coli O157, vaccination with lipopolysaccharide (Conlan et al., Reference Conlan, KuoLee, Webb and Perry1999, Reference Conlan, KuoLee, Webb, Cox and Perry2000), intimin (Dean-Nystrom et al., Reference Dean-Nystrom, Gansheroff, Mills, Moon and O'Brien2002) and type III-secreted proteins (Potter et al., Reference Potter, Klashinsky, Li, Frey, Townsend, Rogan, Erickson, Hinkley, Klopfenstein, Moxley, Smith and Finlay2004) has been shown to reduce colonization, while immunization with bacterins has not shown any beneficial effect. We anticipate that this area will continue to see significant growth over the next decade.

Live vectors

Live vectored vaccines facilitate the delivery of recombinant proteins expressed within the vector itself, or genetic information either integrated into the genome or as plasmid DNA. Both viral and bacterial vectors have been developed for vaccine delivery in humans and animals that are incapable of reverting to full virulence and are therefore safe to use. Various bacterial delivery systems have been optimized for the delivery of recombinant antigens as well as plasmid DNA (Garmory et al., Reference Garmory, Leary, Griffin, Williamson, Brown and Titball2003). For example, S. enterica serovars or Shigella species can target mucosal tissues and deliver the foreign antigen specifically to antigen-presenting cells (APCs) via bacterial secretion systems (Gentschev et al., Reference Gentschev, Mollenkopf, Sokolovic, Hess, Kaufmann and Goebel1996, Reference Gentschev, Dietrich and Goebel2002; Autenrieth and Schmidt, Reference Autenrieth and Schmidt2000). Other vectors include commensal microorganisms such as Lactobacillus, Lactococcus, Staphylococcus, Streptococcus species and attenuated pathogenic microorganisms such as Salmonella, Shigella, BCG, Corynebacterium, Bacillus, Yersinia, Vibrio, Erysipelothrix and Bordetella species (Sizemore et al., Reference Sizemore, Branstrom and Sadoff1995, Reference Sizemore, Branstrom and Sadoff1997; Stocker, Reference Stocker2000; Darji et al., Reference Darji, zur Lage, Garbe, Chakraborty and Weiss2000; Locht, Reference Locht2000; Dietrich et al., Reference Dietrich, Spreng, Favre, Viret and Guzman2003; Roland et al., Reference Roland, Tinge, Killeen and Kochi2005). Other bacterial vectors have been demonstrated to carry plasmid DNA across the mucosal surfaces (Sizemore et al., 1996; Darji et al., Reference Darji, zur Lage, Garbe, Chakraborty and Weiss2000; Dietrich et al., Reference Dietrich, Spreng, Favre, Viret and Guzman2003; Loessner and Weiss, Reference Loessner and Weiss2004), and more recently, bacterial ghosts were demonstrated to represent effective means to deliver plasmid DNA (Ebensen et al., Reference Ebensen, Paukner, Link, Kudela, de Domenico, Lubitz and Guzman2004).

Several viral vectors including poxviruses, herpesviruses and adenoviruses have been used as vectors for veterinary vaccines (Yokoyama et al., Reference Yokoyama, Maeda and Mikami1997). The usefulness of viral vectors for livestock species and poultry greatly depends on their host range and tropism. Ideally, a vector has a broad host range, infects and replicates within a wide range of tissues and ensures optimal post-translational modifications of the recombinant antigen including proper folding and processing (Yokoyama et al., Reference Yokoyama, Maeda and Mikami1997; Sheppard, Reference Sheppard1999). Canarypox-based vaccines are already licensed for the use in both large and small animals, and other viral vectors are currently being evaluated for vaccine delivery including DNA viruses such as adenoviruses, herpesviruses and RNA viruses such as Newcastle disease virus (NDV), Venezuelan equine encephalitis (VEE), Semliki forest virus (SFV) and a few retroviruses. For example, a recombinant alpha-herpesvirus, pseudorabies virus (PRV), carrying the E1 glycoprotein of the classical swine fever virus (CSFV), was used by van Zijl et al. (Reference van Zijl, Wensvoort, de Kluyver, Hulst, van der Gulden, Gielkens, Berns and Moormann1991) to protect pigs against challenge infection with CSFV. Kit et al. (Reference Kit, Kit, Little, Di Marchi and Gale1991) used the bovine herpesvirus type 1 (BHV-1) to deliver the viral protein (VP) 1 of foot-and-mouth disease virus (FMDV) to young calves, and Kweon et al. (Reference Kweon, Kang, Choi and Kang1999) used BHV-1 as a vector for delivering the glycoprotein E2 of the bovine virus diarrhea virus (BVDV) to calves. These and numerous other examples demonstrate the potency of live viral vectors for vaccine delivery.

Particulate delivery systems

Particulate delivery systems have been developed as effective delivery vehicles for whole cells, subunit antigens and DNA. These systems can significantly improve the immunogenicity of both mucosally and systemically administered antigens by protecting the antigen from degradation, especially at the mucosal surfaces. Antigen uptake into specialized mucosa-associated lymphoid tissues (O'Hagan, Reference O'Hagan1996) as well as by APCs is increased, resulting in a better and more balanced immune response. Most particulate delivery systems are based on liposomes, polymers or a combination of the above. For example, a variety of polymers can be formed into microparticles including poly-lactide-co-glycolide (PLG), alginate, chitosan, polyphosphazenes and starch (Bowersock et al., Reference Bowersock, HogenEsch, Suckow, Guimond, Martin, Borie, Torregrosa, Park and Park1999; Illum et al., Reference Illum, Jabbal-Gill, Hinchcliffe, Fisher and Davis2001; Mutwiri et al., Reference Morein, Sundquist, Hoglund, Dalsgaard and Osterhaus2005). The former have been extensively investigated for the delivery of micro-encapsulated vaccines with successful induction of protective immune responses against a variety of pathogens (Jones et al., Reference Jones, McBride, Thornton, O'Hagan, Robinson and Farrar1996; Allaoui-Attarki et al., Reference Allaoui-Attarki, Pecquet, Fattal, Trolle, Chachaty, Couvreur and Andremont1997; Chen et al., Reference Chen, Jones, Fynan, Farrar, Clegg, Greenberg and Herrmann1998; O'Hagan, Reference O'Hagan1998; Herrmann et al., Reference Herrmann, Chen, Jones, Tinsley-Bown, Fynan, Greenberg and Farrar1999; Vajdy and O'Hagan, Reference Vajdy and O'Hagan2001; Fattal et al., Reference Fattal, Pecquet, Couvreur and Andremont2002) and have also been used to encapsulate plasmid DNA for delivery (Tacket et al., Reference Tacket, Reid, Boedeker, Losonsky, Nataro, Bhagat and Edelman1994; Felder et al., Reference Felder, Vorlaender, Gander, Merkle and Bertschinger2000; Singh et al., Reference Singh, Vajdy, Gardner, Briones and O'Hagan2001; Carcaboso et al., Reference Carcaboso, Hernandez, Igartua, Rosas, Patarroyo and Pedraz2004). However, its use may be limited by the fact that PLG is soluble only in organic solvents, which may denature critical epitopes required for inducing protection. Derivatives of PLG, including biodegradable polyesters, circumvent this problem and might represent better alternatives for future vaccine delivery.

Polyphosphazenes are synthetic biodegradable and water-soluble polymers that are easy to produce and that are stable at room temperature eliminating the need for refrigeration. Poly[di(sodium carboxylatophenoxy)phosphazene] (PCPP) has previously been shown to have adjuvant activity with a variety of viral and bacterial antigens in mice (Payne et al., Reference Payne, Jenkins, Woods, Grund, Geribo, Loebelenz, Andriav and Roberts1998; McNeal et al., Reference McNeal, Rae and Ward1999; Wu et al., Reference Wu, Wade and Taylor2001) and sheep (Mutwiri et al., Reference Mutwiri, Benjamin, Soita, Townsend, Yost, Roberts, Andriav and Babiuk2007a, b). In a direct comparison using the hepatitis B surface antigen (HBsAG), PCEP was more potent than the conventional adjuvant alum (Mutwiri et al., Reference Mutwiri, Benjamin, Soita, Townsend, Yost, Roberts, Andriav and Babiuk2007a). Polyphosphazenes not only enhance the magnitude of the immune response but also modulate the quality of the responses, resulting in a more balanced type of response (Mutwiri et al., Reference Mutwiri, Gerdts, Lopez and Babiuk2007b). Other examples of particulate delivery systems include immune stimulating complexes (ISCOMs), which are small 40 nm nanoparticles composed of saponin (adjuvant), lipids and antigen (Morein et al., Reference Morein, Sundquist, Hoglund, Dalsgaard and Osterhaus1984). ISCOMs have adjuvant activity and also the ability to target APCs (Morein et al., Reference Morein, Hu and Abusugra2004). Numerous studies indicate that ISCOMs enhance immune responses to a variety of vaccine antigens in laboratory and domestic animals, and ISCOMs are now included in injectable commercial veterinary vaccines (Bowersock et al., Reference Bowersock, HogenEsch, Suckow, Guimond, Martin, Borie, Torregrosa, Park and Park1999; Morein et al., Reference Morein, Hu and Abusugra2004).

Similarly, liposomes are phospholipid vesicles which enhance immune responses primarily by increasing antigen uptake and presentation of the antigen (Gregoriadis, Reference Gregoriadis1990). Liposomes have been used extensively in experimental vaccines for both human and veterinary vaccines using a variety of viral and bacterial pathogens. Liposomes have also been used to incorporate additional immune modulators, such as cholera toxin or recombinant cytokines such as interleukin-2 (IL-2) (Zhou et al., Reference Zhou, Kraehenbuhl and Neutra1995; Harokopakis et al., Reference Harokopakis, Hajishengallis and Michalek1998; Baca-Estrada et al., Reference Baca-Estrada, Foldvari and Snider1999, Reference Baca-Estrada, Foldvari, Snider, Harding, Kournikakis, Babiuk and Griebel2000; Fukutome et al., Reference Fukutome, Watarai, Mukamoto and Kodama2001; Tana et al., Reference Tana, Isogai and Oguma2003; Wang et al., Reference Wang, Christopher, Nagata, Zabielski, Li, Wong and Samuel2004). An alternative to co-delivering cytokines is to use liposomes containing cytokine-inducing immune modulators such as CpG oligodeoxyribonucleotides (CpG ODN). This would reduce the potential toxicity often associated with large doses of cytokines. For example, the adjuvant activity of CpG ODN was augmented by liposomal delivery in mice immunized intranasally against influenza and hepatitis B viruses (Joseph et al., Reference Joseph, Louria-Hayon, Plis-Finarov, Zeira, Zakay-Rones, Raz, Hayashi, Takabayashi, Barenholz and Kedar2002).

Other particulate delivery systems include alginate and chitosan, naturally occurring carbohydrates that polymerize into particles upon contact with divalent cations (Bowersock et al., Reference Bowersock, HogenEsch, Suckow, Guimond, Martin, Borie, Torregrosa, Park and Park1999; Illum et al., Reference Illum, Jabbal-Gill, Hinchcliffe, Fisher and Davis2001; van der Lubben et al., Reference van der Lubben, Verhoef, van Aelst, Borchard and Junginger2001; Mutwiri et al., Reference Morein, Sundquist, Hoglund, Dalsgaard and Osterhaus2005; Douglas et al., Reference Douglas, Piccirillo and Tabrizian2006; Greimel et al., Reference Greimel, Werle and Bernkop-Schnurch2007). Oral immunization of cattle with the model antigen ovalbumin formulated in alginate microparticles enhanced IgA- and IgG-immune responses in the respiratory tract, providing evidence that they can stimulate the common mucosal immune system in a large animal (Bowersock et al., Reference Bowersock, HogenEsch, Torregrosa, Borie, Wang, Park and Park1998). Calves immunized intranasally with porcine serum albumin (PSA) in alginate microparticles developed high levels of anti-PSA IgG1 antibodies in serum, nasal secretions and saliva (Rebelatto et al., Reference Rebelatto, Guimond, Bowersock and HogenEsch2001). In contrast, oral immunization did not induce any significant responses, suggesting that IN was more efficient for induction of nasal and serum responses. Rabbits orally immunized with P. multocida antigens encapsulated in alginate microparticles developed significantly higher nasal IgA responses (Bowersock et al., Reference Bowersock, HogenEsch, Suckow, Guimond, Martin, Borie, Torregrosa, Park and Park1999). Increased protection against M. haemolytica and Streptococcus pneumoniae challenge in mice was demonstrated by encapsulating antigens in alginate microparticles following oral immunization (Kidane et al., Reference Kidane, Guimond, Ju, Sanchez, Gibson and Bowersock2001). These studies clearly show that alginate microparticles are effective for mucosal delivery of vaccines in small and large animals.

Adjuvants

Adjuvants were originally described by Ramon (Ramon, Reference Ramon1924) as ‘helper’ substances that can enhance the immune responses to the antigen alone. Adjuvants have been developed mainly by trial and error, screening a wide variety of natural and synthetic molecules. Adjuvants can be classified into delivery systems and immunostimulators (Singh and O'Hagan, Reference Singh and O'Hagan2003) and include a wide variety of molecules and carrier systems. Several of these adjuvants have been developed for veterinary applications, including liposomes, oil-in-water emulsions (squalenes), saponins, carbomers, monophosphoryl lipid A, Quil A, ISCOMS, heat-labile enterotoxins, cholera toxin and QS21 to name a few and the advances in this field were recently summarized in an excellent review by Singh and O'Hagan (Singh and O'Hagan, Reference Singh and O'Hagan2003). Here, we will only focus on three adjuvant systems, which currently show great promise for adjuvanting vaccine responses.

Bacterial DNA as well as synthetic DNA containing CpG motifs (CpG ODN) represent very potent danger signals resulting in the induction of innate and adaptive immune responses. Signaling is mediated by TLR-9, which so far is the only identified receptor for synthetic double-stranded DNA. Numerous investigators have shown that treatment of animals with CpG ODN can protect against a variety of experimental infectious and non-infectious diseases (Wilson et al., Reference Wilson, Dar, Napper, Marianela, Babiuk and Mutwiri2006). CpG ODNs promote predominantly Th1-type immune responses (Chu et al., Reference Chu, Targoni, Krieg, Lehmann and Harding1997; Krieg, Reference Krieg2006) and clinical studies are in various stages in humans to evaluate therapy against infectious disease, cancer, asthma and allergy (Krieg, Reference Krieg2006). Successful use of synthetic CpG DNA as a vaccine adjuvant has been demonstrated in various animal species including large animals such as cattle and pigs. For example, CpG ODN was shown to be an excellent adjuvant for stimulating immune responses against an experimental vaccine based on a subunit protein (gD antigen) of the BHV-1 in mice, sheep and cattle by producing enhanced serum Ig2a levels and interferon-γ (IFN-γ) in splenocytes or peripheral blood lymphocytes, indicating a more balanced or Th1-type response (Ioannou et al., Reference Ioanu, Griebel, Hecker, Babiuk and van Drunen2000b, Reference Ioanu, Gomis, Karvonen, Hecker, Babiuk and van Drunen2002a). Interestingly, the use of CpG ODNs in combination with low levels of mineral oil enhanced the immune response and reduced the amount of tissue damage associated with conventional vaccine adjuvants in sheep (Ioannou et al., Reference Ioanu, Gomis, Karvonen, Hecker, Babiuk and van Drunen2002a) and cattle (Ioannou et al., Reference Ioanu, Griebel, Hecker, Babiuk and van Drunen2002b). Similarly, incorporation of CpG ODN in a commercial equine influenza virus vaccine resulted in significant enhancement of antibodies against influenza virus (Lopez et al., Reference Lopez, Hecker, Mutwiri, van Drunen, Babiuk and Townsend2006).

Cationic host defense peptides (HDPs) are endogenous effectors found in virtually every life form. They are short peptides that can be grouped as defensins and cathelicidins (Oppenheim et al., Reference Oppenheim, Biragyn, Kwak and Yang2003; Finlay and Hancock, Reference Finlay and Hancock2004). These molecules are fundamental components of the innate immune response with a broad spectrum of functions including direct antimicrobial activities, immunostimulatory functions of both innate and acquired immunity, as well as involvement in wound healing, cell trafficking and vascular growth (Bowdish et al., Reference Bowdish, Davidson and Hancock2006; Brown and Hancock, Reference Brown and Hancock2006; Mookherjee et al., Reference Mookherjee, Brown, Bowdish, Doria, Falsafi, Hokamp, Roche, Mu, Doho, Pistolic, Powers, Bryan, Brinkman and Hancock2006). While the antimicrobial activities of HDPs have been known for a long time, more recent evidence suggests that HDPs are immunomodulatory resulting in recruitment of immature DCs and T-cells, glucocorticoid production, phagocytosis, mast cell degranulation, complement activation, and IL-8 production by epithelial cells (Yang et al., Reference Yang, Chertov, Bykovskaia, Chen, Buffo, Shogan, Anderson, Schroder, Wang, Howard and Oppenheim1999, Reference Yang, Chertov and Oppenheim2001, Reference Yang, Biragyn, Hoover, Lubkowski and Oppenheim2004). The immunoenhancing activity of HDPs has been demonstrated in several studies. For example, ovalbumin-specific immune responses were enhanced in mice when human neutrophil peptides 1, 2 and 3 (HNP1–3) were co-administered intranasally (Lillard et al., Reference Lillard, Boyaka, Chertov, Oppenheim and McGhee1999). It was also demonstrated that HIV gp120-specific mucosal, systemic and CTL immune responses could be achieved after immunization with a fusion DNA vaccine encoding the murine β-defensin 2 and the HIV gp120 gene (Biragyn et al., Reference Biragyn, Surenhu, Yang, Ruffini, Haines, Klyushnenkova, Oppenheim and Kwak2001, Reference Biragyn, Ruffini, Leifer, Klyushnenkova, Shakhov, Chertov, Shirakawa, Farber, Segal, Oppenheim and Kwak2002). These examples provide evidence that HDPs have been successfully used as adjuvants to enhance vaccine-specific immunity.

Synthetic single-stranded RNA (ssRNA) and small antiviral compounds (imidazoquinolines) have been shown to also promote Th1 rather than Th2 immune responses (Vasilakos et al., Reference Vasilakos, Smith, Gibson, Lindh, Pederson, Reiter, Smith and Tomai2000). Studies in our own laboratories have indicated that similar to bacterial DNA, ssRNA and imidazoquinolones signal through specific TLRs resulting in maturation and activation of effector cells such as porcine dendritic cells (G. Auray, 2008). Evidence from studies in mice and humans has confirmed that these molecules have adjuvant activity in vivo; however, extensive studies in livestock species have not been conducted.

Mechanisms of viral–bacterial synergism

One of the more extensively studied synergisms between a viral and bacterial infection is that between BHV-1 and M. haemolytica. BHV-1 infects the epithelium of the upper respiratory tract, which results in rhinotracheitis. Lesions include loss of cilia and goblet cells leading to epithelial erosions, which could progress to necrosis of epithelium and adjacent lymphoid tissue (Schuh et al., Reference Schuh, Bielefeldt, Babiuk and Doige1992). This facilitates migration and colonization of the lower respiratory tract by bacterial pathogens such as M. haemolytica. Furthermore, alveolar macrophages and peripheral blood leukocytes from cattle with active BVH-1 infection are more susceptible to the lethal effects of leukotoxin ex vivo than leukocytes from uninfected cattle. Czuprynski et al. (Reference Czuprynski, Leite, Sylte, Kuckleburg, Schultz, Inzana, Behling-Kelly and Corbeil2004) reported that BHV-1 infection affects the response of bovine lung and peripheral blood leukocytes to M. haemolytica leukotoxin ex vivo. Active viral infection increased the susceptibility of bovine alveolar macrophages and leukocytes to M. haemolytica leukotoxin by increasing the production of proinflammatory cytokines (i.e. IL-1β, tumor necrosis factor-α and IFN-γ). This then leads to expression or activation of the β2-integrin CD11a/CD18 (LFA-1) on the leukocyte surface, which correlates with increased binding and responsiveness to leukotoxin. In support of this observation, ex vivo exposure to proinflammatory cytokines (i.e. IL-1β, tumor necrosis factor-α and IFN-γ) increased LFA-1 expression on bovine leukocytes, and in vitro incubation of bovine leukocytes with BHV-1 potentiated LFA-1 expression making them more responsive to leukotoxin (Leite et al., Reference Leite, Kuckleburg, Atapattu, Schultz and Czuprynski2004).

BVDV is also known to contribute to shipping fever. Recently, evidence was presented by Al-Haddawi et al. (Reference Al-Haddawi, Mitchell, Clark, Wood and Caswell2007), who demonstrated that infection with a type 2 BVDV isolate inhibited LPS-induced up-regulation of tracheal antimicrobial peptide (TAP) mRNA expression and lactoferrin secretion by tracheal epithelial cells. This suggests that BVDV-2 abrogates respiratory innate immune responses and thus predisposes to bacterial pneumonia in cattle.

BRSV was found to synergistically enhance H. somni infection (Gershwin et al., Reference Gershwin, Berghaus, Arld, Anderson and Corbeil2005). When calves were either dually infected or singly infected with H. somni or BRSV, disease was significantly more severe in dually infected calves and high IgE and IgG responses were detected to H. somni antigens. An increase in IgE antibodies specific for antigens of H. somni presents a possible mechanism for pathogenesis of the disease enhancement. A recent report by Corbeil (Reference Corbeil2008) showed that the pathogenic synergy of BRSV and H. somni in calves can be attributed, in part at least, to the higher IgE anti-major membrane protein (MOMP) antibody responses in dually infected calves.

A number of strategies are used by the viruses involved in BRD to compromise the immune response to pathogens. BVDV infects immune cells and inhibits cell proliferation, whereas BHV-1 and BRSV only inhibit proliferation. BHV-1 induces apoptosis of CD4+ Th cells, which impairs cytokine production and the development of both humoral and cell-mediated immune responses. The ability of BHV-1 to down-regulate MHC class I expression compromises the development of CTL response against BHV-1 as well as other pathogens involved in BRD (Srikumaran et al., Reference Srikumaran, Kelling and Ambagala2007).

Effect of viral vaccines on disease

Vaccination against viral infections is expected to limit the chance of bacterial infections and thus reduce the need for antibiotic use. Although few controlled studies have been performed to assess the ability of viral vaccination to prevent bacterial infections, there are several reports on vaccination against the viruses involved in BRD that support this contention. According to Jericho et al. (Reference Jericho, Loewen, Smithson and Kozub1991), prevention of viral infection by vaccination can reduce the severity of disease after subsequent M. haemolytica exposure. We showed that calves vaccinated with one of the three major glycoproteins of BHV-1, gB, gC or gD, or various combinations experienced a significantly lower number of sick days and reduced lung pathology when they were subsequently challenged with BHV-1 followed by M. haemolytica. All calves in the vaccinated groups survived, whereas 60% of the animals in the control group died. Among these glycoproteins, gD was the superior protective antigen (Babiuk et al., Reference Babiuk, LItalien, van Drunen, Zamb, Lawman, Hughes and Gifford1987; van Drunen Littel-van den Hurk et al., Reference van Drunen, Gifford and Babiuk1990). Since it usually is not clear which virus is involved in field situations, calves need to be ultimately vaccinated against all four common viruses associated with BRD in order to have an impact on secondary bacterial infection. Stilwell et al. (Reference Stilwell, Matos, Caroli and Lima2008) reported that the use of Rispoval 41, a quadrivalent vaccine against respiratory viruses, reduced morbidity and mortality due to BRD in beef calves after weaning. This observation provides support for the value of viral vaccines in preventing overall disease, thus reducing the need for antibiotics.

Vaccination has also been evaluated as an approach to reduce the severity of porcine respiratory disease (PRD) in pigs. In one study, pigs vaccinated with a PRRSV vaccine did not show decreased potentiation of PRRSV pneumonia by Mycoplasma hyopneumoniae, whereas vaccination with M. hyopneumoniae bacterin decreased the enhancement of PRRSV-induced pneumonia in the dually infected pigs (Thacker et al., Reference Thacker, Thacker, Young and Halbur2000). In contrast, in one field study, pigs vaccinated with a live PPRSV vaccine performed better with respect to feed conversion ratio and average daily gain as compared with non-vaccinated animals, and decreased incidence of respiratory bacterial infection was observed in vaccinated pigs compared with untreated ones, demonstrating that indeed a viral vaccine can protect from subsequent bacterial infection in pigs (Mavromatis et al., Reference Mavromatis, Kritas, Alexopoulos, Tsinas and Kyriakis1999).

In chickens, live IBV vaccine can protect against intranasal challenge with a mixture of E. coli strains and IBV strains of the same serotype; this live IBV vaccine usually also protected against challenge with the E. coli pool and IBV strains of other serological types. Some protection, but less efficient, against challenge with IBV and the E. coli pool was also observed in chickens vaccinated with an inactivated IBV strain (Cook et al., Reference Cook, Smith and Huggins1986). In another study, vaccination against IBV resulted in protection against predisposure for colibacillosis after a subsequent IBV infection. When 1-day-old commercial broilers were vaccinated ocularly with IBV vaccine and then challenged with the virulent IBV strain M41 followed by E. coli 506, it was apparent that IBV vaccination reduced both the number of broilers with E. coli airsacculitis and the severity of airsacculitis. In contrast, in spray-vaccinated groups, no significant reduction of the number of birds with systemic colibacillosis or the severity of this infection was obtained, whereas eye-drop vaccination resulted in conflicting results (Matthijs et al., Reference Matthijs, van Eck, de Wit, Bouma and Stegeman2005).

An interesting result of vaccination of chickens with NDV vaccine was reported by Huang and Matsumoto (Reference Huang and Matsumoto2000). When chickens at 5 weeks of age were vaccinated with a NDV vaccine at various days before challenge exposure with O1:K1 strain of E. coli via an intra-air sac route, viable numbers of E. coli in the spleen 24 h after the infection were reduced. This was effective for a period of 2–8 days post-vaccination. This suggests that innate immunity induced by NDV vaccination may suppress the multiplication of E. coli in chickens.

One may have to be careful with the use of live viral vaccines, as administration of a live vaccine containing IBV and NDV together with exposure to Mycoplasma gallisepticum and E. coli in chickens actually resulted in extensive multiplication of E. coli and severe and persistent histological lesions and mortality. This effect was less severe when the chickens were exposed to only the vaccine and E. coli, while chickens inoculated with only vaccine, or with M. gallisepticum and/or E. coli, did not show any lesions (Nakamura et al., Reference Nakamura, Ueda, Tanimura and Guchi1994).

Overall, however, the benefits of vaccination against the viral pathogens involved in mixed viral–bacterial infections have been clearly demonstrated and the results indicate that vaccination against viral components involved in diseases such as BRD can reduce the incidence of bacterial infection. In view of the numerous instances of viral–bacterial synergisms, additional controlled studies in this area would be beneficial in further reducing the use of antibiotics.