Current status of vaccines against infectious bursal disease

Abstract

Infectious bursal disease virus (IBDV) is the aetiological agent of the acute and highly contagious infectious bursal disease (IBD) or “Gumboro disease”. IBD is one of the economically most important diseases that affects commercially produced chickens worldwide. Along with strict hygiene management of poultry farms, vaccination programmes with inactivated and live attenuated viruses have been used to prevent IBD. Live vaccines show a different degree of attenuation; many of them may cause bursal atrophy and thus immunosuppression with poor immune response to vaccination against other pathogens and an increase in vulnerability to various types of infections as possible consequences. Depending on their intrinsic characteristics or on the vaccination procedures, some of the vaccines may not induce full protection against the very virulent IBDV strains and antigenic variants observed in the last three decades. As chickens are most susceptible to IBDV in their first weeks of life, active immunity to the virus has to be induced early after hatching. However, maternally derived IBDV-specific antibodies may interfere with early vaccination with live vaccines. Thus new technologies and second-generation vaccines including rationally designed and subunit vaccines have been developed. Recently, live viral vector vaccines have been licensed in several countries and are reaching the market. Here, the current status of IBD vaccines is discussed.

Introduction

Infectious bursal disease virus (IBDV) is the aetiological agent of an acute and highly contagious disease in young chickens. The disease, also named “Gumboro disease” according to the location of the first outbreaks in Gumboro, Delaware, USA, was initially described as avian nephrosis due to damage seen in the kidneys (Cosgrove, 1962) but was later designated infectious bursal disease (IBD) according to varying morphologic and histological changes observed in the bursa of Fabricius (Hitchner, 1970). Chickens infected with virulent classical IBDV between 3 and 6 weeks of age mostly show clinical signs and mortality accompanied with bursal atrophy. Chickens infected with virulent classical IBDV at less than 3 weeks of age usually have few or no clinical signs. Both infections would cause immunosuppression, which makes the birds vulnerable to a variety of secondary infections. Accordingly, chickens also develop a poor immune response to vaccination against other pathogens (Allan et al., 1972; Faragher et al., 1974; Rosales et al., 1989; Mazariegos et al., 1990). Thus, IBD is one of the most economically important diseases that affects commercially produced chickens worldwide (van den Berg, 2000; Müller et al., 2003; Eterradossi & Saif, 2008). IBDV has been shown to remain infectious for 122 days in a chicken house and for 52 days in feed and water. Along with strict hygiene management of poultry farms, vaccination with conventional live attenuated and inactivated viral vaccines has been used to prevent IBD.

IBDV is a member of the Birnaviridae family (Dobos et al., 1979; Müller et al., 1979) characterized by a non-enveloped icosahedral capsid (60 nm in diameter) containing a double-stranded RNA genome consisting of two segments (A and B). The intrinsic property of all RNA viruses is to evolve quickly due to the low proofreading activity of their viral replicases. Accordingly, in IBDV high variability of the viral genome is observed (for example, Zierenberg et al., 1999; Islam et al., 2001a; Delmas, 2008). Probably due to this inherent property and other additional evolutionary selection processes in the field that are not yet well understood, antigenic variants of IBDV were discovered in the early 1980s (Saif, 1984; Rosenberger et al., 1985) and very virulent IBDV (vvIBDV) strains emerged in the late 1980s (Box, 1989; Chettle et al., 1989). IBDV is evolving in the field (Eterradossi et al., 2004; Le Nouen et al., 2006; Jackwood et al., 2006; Letzel et al., 2007; Icard et al., 2008; Jeon et al., 2009; Jackwood & Sommer-Wagner, 2011). A recent study indicated that beside the usual highly virulent IBDV field virus other strains have also emerged; in Europe—and this may also hold true for other endemic regions—the Gumboro disease situation therefore appears to be growing in complexity (de Wit, personal communication, 2011). These changes in antigenicity and virulence made the task of controlling IBD by vaccination more challenging (for example, van den Berg, 2000; Müller et al., 2003; Eterradossi & Saif, 2008).

In susceptible chickens, IBDV initially replicates in macrophages of the gut-associated tissues, then primarily in proliferating B lymphocytes of the bursa of Fabricius (Müller et al., 1979). In the avian immune system, the bursa of Fabricius is an epithelial and lymphoid organ where lymphocyte stem cells acquire the characteristics of mature, immunocompetent B lymphocytes. Depending on the development of the bursa, chickens are most susceptible to IBDV in their first weeks of life, and active immunity to the virus has to be induced early after hatching. However, maternally derived IBDV-specific antibodies that are transmitted to the offspring via the yolk of embryonated eggs may interfere with early vaccination. Whereas innate immune mechanisms seem to be fully functional in the newly hatched chick, optimal adaptive immune responses only develop following the first weeks after hatch, and transfer of maternal antibodies helps to protect the offspring until adaptive immune response becomes fully effective (Davison et al., 2008). In newly hatched layer type chicks, maternally derived antibodies exhibit a linear or curvilinear decline with a mean half-life of 5 to 6 days. Fahey et al. (1987) reported a half-life of 6.7 days for IBDV-specific maternal antibodies. It is generally thought that the half-life of maternally derived antibodies in broiler lines is much shorter, approximately 3 days; however, to the authors’ knowledge there is no published evaluation for this figure, which has nevertheless been used extensively when designing IBD vaccination programmes for broilers (Gardin, 1994; Block et al., 2007). Under field conditions, however, the decay pattern of IBDV-specific maternally derived antibodies proved to be more complex, depends largely on initial antibody levels, and the levels may vary between hatches and also within a hatch, making it difficult to predict the optimal time point for vaccination (de Wit, 1998). Interference of maternally derived antibodies with vaccine uptake still remains a major problem in vaccination against IBD using live vaccines (Block et al., 2007). In order to help prevent IBD more effectively, new technologies and next-generation vaccines have been developed and introduced into the market (Meeusen et al., 2007).

Conventional live and inactivated IBDV vaccines

Live viral vaccines mimic infection in the target host. They can replicate and induce both cellular and humoral immunity. They do not require an adjuvant to be effective and are suitable for mass administration to the chicken, but they may also have undesirable side-effects. These include horizontal and vertical transmission (although the latter not in the case of IBD vaccines), reversion to virulence and vaccine reactions that may result in disease or production loss. In general, the live IBDV vaccines in use by the poultry industry have been attenuated by serial passage in tissue culture, eggs or embryo-derived tissues, with the aim of maintaining the immune response induced by the parent virus whilst attenuating the ability of the vaccine virus to cause clinical disease or significant immunosuppression (Schijns et al., 2008).

Most commercially available conventional live IBDV vaccines are based on classical virulent strains. Those classified as “mild” vaccines exhibit only poor efficacy in the presence of certain levels of maternally derived antibodies and against vvIBDV. “Intermediate” and “intermediate plus” or “hot” vaccines have a much better efficacy and may break through higher levels of maternally derived antibodies, but they can induce moderate to severe bursal lesions and, thus, cause corresponding levels of immunosuppression (Mazariegos et al., 1990; Tsukamoto et al., 1995; Kumar et al., 2000; Rautenschlein et al., 2005). They may not fully protect chickens against infection by the vvIBDV strains (Rautenschlein et al., 2005) or by antigenic variants. Safety and efficacy of this type of vaccine still remain a major concern. In addition, it should be kept in mind that the practical on-farm administration of the conventional live vaccines to a large number of animals is also a technically demanding process, with difficulties inherent to farm-to-farm variability (variable chicks, variable farming conditions, variable skills in vaccination crews, etc.) that should not be underestimated when assessing the results of vaccination programmes.

Non-replicating antigens, such as inactivated whole viruses, viral subunits or recombinant viral antigens, lack efficient immunogenicity unless they are combined with supporting adjuvants and administered in repeated injections, or follow a suitable priming with a replicating antigen. The need for possibly repeated individual injections obviously contributes to the implementation costs of these vaccines, and their use is usually restricted to highly valuable birds such as future breeder birds, where vaccination before lay provides passive immunity to the offspring by means of maternally derived antibodies. However, such vaccines have also been used occasionally in birds as young as 10 days old, particularly in areas heavily contaminated with virulent viruses (Wyeth & Chettle, 1990).

Inactivated IBD vaccines are mostly formulated as water-in-oil emulsions, usually combining several antigens. It has been observed that inactivated IBD vaccines were also able to induce IBDV-specific T-cell and inflammatory responses in chickens (Rautenschlein et al., 2002). It has been reported that inactivated IBD vaccines must have either a high or an optimized antigenic content in order to induce in breeders an immunity that helps protect the progeny from infection by variant IBDV strains (Rosenberger et al., 1987; Müller et al., 1992). Inactivated vaccines are most efficiently used in a prime-boost regimen, using attenuated live IBDV as priming vaccine.

Genetically engineered live IBDV vaccines

The viral capsid protein VP2, encoded by genomic segment A and derived from a large precursor protein VP0 by a series of proteolytic processes (for a recent review, see Delmas, 2008), carries immune determinants that control antibody-dependent neutralization and protection (Azad et al., 1987; Becht et al., 1988; Müller et al., 1992; Schnitzler et al., 1993). Determining the complete nucleotide sequences of both segments of the IBDV double-stranded RNA genome (Mundt & Müller, 1995) allowed development of a reverse genetics system (Mundt & Vakharia, 1996). Meanwhile this approach has been widely used with the aim of generating attenuated IBDV potentially applicable as vaccine. Among others, attenuated mutant IBDV was generated from vvIBDV by site-directed mutagenesis of nucleotide sequences encoding specific amino acids in IBDV structural protein VP2 (Islam et al., 2001b; van Loon et al., 2002; Raue et al., 2004; Noor, 2009). However, and not quite unexpected, reversion to virulence was regularly observed in infected chickens (Raue et al., 2004; Noor, 2009), and the engineering of safe live viral vaccines is therefore likely to require a number of attenuating mutations distributed throughout the genome (Meeusen et al., 2007; Noor, 2009). A mutant IBDV was rescued by reverse genetics, which provided protection against both classical as well as antigenic variant IBDV (Mundt et al., 2003). Intrasegmental chimeric viruses, with genomic regions in virulent serotype 1 IBDV replaced by the corresponding ones of the apathogenic serotype 2 IBDV, were shown to have no impact on the most prominent property of serotype 1 IBDV infection, the damage of follicles in the bursa of Fabricius (Schröder et al., 2000, 2001). Interserotypic reassortants, generated with virulent serotype 1 and apathogenic serotype 2 IBDV strains as the parental viruses (Oberländer, 2004; Zierenberg et al., 2004), did not cause severe damage in the bursa of Fabricius and induced high titres of serotype-specific neutralizing antibodies, making them potential vaccine candidates. As with other live IBDV vaccines, these experimental vaccines are still subject to interference with maternal antibodies. None of them have yet reached the market.

Subunit vaccines

Over the last three decades many reports have been published on the recombinant expression of the viral structural protein VP2 (rVP2), the major protective IBDV antigen where neutralizing epitopes are conformation dependent (Becht et al., 1988; Fahey et al., 1989). Denatured VP2 does not induce protection in chickens (Fahey et al., 1989), and denatured and renatured VP2 also lost the ability to induce neutralizing antibodies in chickens (Öppling et al., 1989). Different expression systems have been used, such as Escherichia coli (Rong et al., 2007), yeast (Fahey et al., 1991), fowlpoxvirus (Bayliss et al., 1991; Heine & Boyle, 1993), baculovirus (Vakharia et al., 1993), Semliki Forest virus (Phenix et al., 2001), and even plant expression systems (Wu et al., 2004). Either the VP2 encoding region alone or the polyprotein gene was used for expression, and polyhistidine tags have been used for effective protein purification. It was observed that IBDV empty capsids offered better protection than did the tubular structures that formed after expression in a baculovirus-based system (Martinez-Torrecuadrada et al., 2003). A fusion protein consisting of VP2 and chicken interleukin-2 considered to enhance immunogenicity has been developed and was tested as a potential vaccine (Liu et al., 2005). In a mimotope vaccine approach, the multi-mimotope protein r5EPIS—generated to comprise IBDV-specific monoclonal antibody binding peptides—promised to be a novel subunit vaccine candidate for IBDV (Wang et al., 2007). In experimental vaccination studies using rVP2, partial to 100% protection has been reported. To date, three vaccines based on the VP2 subunit have been placed on the market in some countries, with VP2 expressed either in the baculovirus system, or in E. coli or in the yeast Pichia pastoris (Pitcovski et al., 2003). As with inactivated vaccines, the requirements of parenteral administration and booster immunizations are limiting factors for the use of such vaccines. Recombinant subunit vaccines based on the expression of VP2 alone could allow the development of a DIVA strategy to differentiate vaccinated flocks from infected ones, as VP3 of IBDV is also a potent antigen and could induce anti-VP3 antibodies only detected in infected flocks (once possible maternally derived anti-VP3 antibodies have waned).

IBDV immune complex vaccines

A new concept of vaccines was the establishment of immune complex vaccines (Icx). These vaccines consist of a mixture of a certain amount of IBDV-specific antibodies obtained from the sera of hyperimmunized chickens and infectious IBD vaccine virus (Whitfill et al., 1995). Their major advantage is that they are suitable for in ovo vaccination at day 18 of incubation with commercial egg-injection machines, which initially had been used for vaccination against Marek's disease virus and fowlpoxvirus. Alternatively, the Icx vaccines can be delivered by subcutaneous injection at 1 day old in the hatchery (Ivan et al., 2005). In addition it has been shown that these vaccines were effective in the presence of maternally derived antibodies (Haddad et al., 1997; Giambrone et al., 2001). Both methods of administration therefore allowed a more automated and systematic administration process than with conventional live vaccines that are usually given via the drinking water or by eye-drop in some rare cases. At challenge, the experimental efficacy of the Icx vaccines was identical to or better than that induced by vaccination with live IBDV vaccines. The working mechanism was investigated by comparing the infectivity of the IBDV-Icx and the virus alone at various time points after in ovo injection (Jeurissen et al., 1998). With both vaccines, IBDV was associated with B lymphocytes, macrophages and follicular dendritic cells in the bursa of Fabricius and spleen, although IBDV complexing with specific antibodies caused a delay in virus detection of approximately 5 days (Jeurissen et al., 1998). In another study (Ivan et al., 2005), the virus was first detected in the bursa of vaccinated specific pathogen free chickens at day 14 post vaccination and on days 17 to 21 in broilers with maternally derived IBDV-specific antibodies. Most remarkable was the low level of depletion of bursal and splenic B lymphocytes in chickens vaccinated experimentally with IBDV–Icx (Jeurissen et al., 1998). Furthermore, in ovo inoculation with the IBDV–Icx vaccine induced more germinal centres in the spleen, and larger amounts of IBDV were localized on both splenic and bursal follicular dendritic cells. Recently, recombinant neutralizing antibodies have been developed and used in an experimental IBDV-Icx vaccine (Sapats et al., 2006; Ignjatovic et al., 2006).

DNA vaccines

Immune response to a foreign antigen can be induced by transfer of naked DNA, encoding the target gene, into host cells. This procedure overcomes the problems posed by the presence of specific antibodies to the target antigen in the organism and promotes the induction of specific antibodies and also of cytotoxic T cells after intracellular expression of the antigen (Hsieh et al., 2010). A number of studies were conducted on the development of a DNA vaccine, with variable success, to induce an efficient immune response in chickens. IBDV polyprotein encoding cDNA appeared to perform better than VP2 encoding cDNA (Fodor et al., 1999; Li et al., 2003), and co-administration of IBDV-specific cDNA and interleukin-2 (Hulse & Romero, 2004; Li et al., 2004) or interleukin-6 (Sun et al., 2005) encoding DNA increased vaccine efficacy. A potential use of DNA vaccine for priming in ovo or at 1 day old followed by boosting with inactivated vaccine or vectored vaccine has been described (Haygreen et al., 2006; Hsieh et al., 2007) and induced protection from a virulent challenge infection. The use of DNA vaccination via the in ovo route was also described (Oshop et al., 2003; Haygreen et al., 2006; Park et al., 2009). The results showed that the in ovo delivery without a boost vaccination was not sufficient to induce protective immunity. Bacteria, including Lactococcus lactis, Salmonella typhimurium and E. coli, have been used to deliver IBDV cDNA vaccine orally, but with variable success. Here, difficulties in the secretion or translocation of the expressed viral protein across the bacterial cell wall could be the limiting factor (Li et al., 2006; Mahmood et al., 2007).

Live viral vector vaccines

Vector vaccines are genetically engineered vaccines in which a gene from one organism—the donor—is inserted into the genome of another organism—the vector—to elicit a protective immune response against both organisms. Among others, fowlpoxvirus (Heine & Boyle, 1993), Newcastle disease virus (Huang et al., 2004), herpesvirus of turkey (HVT) (Darteil et al., 1995), Marek's disease virus (Tsukamoto et al., 1999), avian adenovirus (Francois et al., 2004) and T4 bacteriophage (Cao et al., 2005) have been used as vector viruses for expressing VP2, the only antigen inducing protective immunity to IBDV. HVT has been used as a safe and effective vaccine against Marek's disease for decades; as it is poorly sensitive to interference with maternally derived antibodies, HVT has been proposed as a vector for IBD (Darteil et al., 1995; Tsukamoto et al., 2002) and other diseases. Meanwhile, several “HVT plus IBDV-VP2” vector vaccines have been developed for application in ovo or by the subcutaneous route in 1-day-old chickens. Some have been licensed in various countries, and data on field efficacy trials have been reported (Bublot et al., 2007; Le Gros et al., 2009). In the study described by Le Gros et al. (2009), maternal immunity—interfering with live IBDV vaccine replication in a control group—had no detectable effect on the efficacy of the vector vaccine. At the age of 30 days, 93% protection against a challenge with vvIBDV was observed. Remarkably, it was also observed that this vector vaccine provided protection from a challenge with variant IBDV (Perozo et al., 2009). As with recombinant subunit vaccines, vector vaccines that express VP2 alone could also allow the development of a DIVA strategy.

Conclusion

Although identified more than 40 years ago, IBDV continues to be a major threat to commercial poultry all over the world. Along with strict bio-security, the use of conventional inactivated and live IBDV vaccines to control IBD had been a success story until the early 1980s, when antigenic variants emerged in the USA. The emergence of very virulent IBDV strains in the late 1980s made the task of controlling IBD by vaccination even more challenging. Furthermore, until today, interference of maternally derived antibodies with vaccine uptake is a major problem in early vaccination against IBD with live vaccines. Recently, new technologies such as in ovo vaccination and live viral vector vaccines proved effective even in the presence of high levels of maternally derived antibodies. In more and more countries, in ovo vaccination is rapidly being adopted as the method of choice for immunizing chickens against Marek's disease and other poultry diseases including IBD; systems have been developed that allow automated mass application. The vectored vaccines may pave the way to a sustained and successful IBD prevention and control regimen in the near future. It has been speculated, however, that in the field it may be difficult to maintain high efficacy “… because these tightly regulated recombinant vaccines cannot easily adapt to meet the emergence of very virulent strains of both IBDV and MDV, apparently induced by the comprehensive numbers of vaccinations performed against these diseases” (Meeusen et al., 2007). This statement could be debated for IBD, as the epidemiological process for the emergence of new IBDV strains with an increased virulence is still poorly understood and as the available vaccines might be more readily adapted by replacing the VP2 gene insert in some of the genetically engineered vaccines than by attenuating the emerging virus. However, vigilance in monitoring field viruses and the natural history of the disease is constantly required (Saif, 2004).

Acknowledgements

The authors wish to thank Thomas W. Vahlenkamp for critical reading of the manuscript.