INTRODUCTION Infectious coryza (IC) is an acute respiratory dis- ease of chickens and the cause of serious econom- ic losses as a result of an increased number of culls and a drop in egg production in laying flocks. The impact on broilers has been also shown during out- breaks in the USA (Droual, Bickford, Charlton, Cooper & Channing 1990). The causative agent A. paragallinarum, (Blackall, Christensen, Beckenham, Blackall & Bisgaard 2005) is a bacterial organism 299 Onderstepoort Journal of Veterinary Research, 76:299–309 (2009) Study on the efficacy and safety of different antigens and oil formulations of infectious coryza vaccines containing an NAD-independent strain of Avibacterium paragallinarum B. DUNGU1*, B. BRETT1, R. MacDONALD1, S. DEVILLE2, L. DUPUIS2, J. THERON3 and R.R. BRAGG4 ABSTRACT DUNGU, B., BRETT, B., MacDONALD, R., DEVILLE, S., DUPUIS, L., THERON, J. & BRAGG, R.R. 2009. Study on the efficacy and safety of different oil formulations of infectious coryza vaccines con- taining a NAD-independent strain of Avibacterium paragallinarum. Onderstepoort Journal of Veterinary Research, 76:299–309 The present study was designed to assess and compare three different formulations of the new Onderstepoort infectious coryza (IC) quadrivalent vaccine, which contain an NAD-independent strain of Avibacterium paragallinarum (previously known as Haemophilus paragallinarum), and a commer- cial IC vaccine, not containing an NAD-independent strain, for their safety and ability to protect chick- ens of varying ages against virulent challenges with four different serovars of A. paragallinarum, in- cluding the NAD-independent strain of the C-3 serovar. Four groups of 140 chickens each were vaccinated at the age of 17 weeks and revaccinated at the age of 19 weeks with each of the four vaccine formulations. A similar sized group of non-vaccinated chickens was used as control. Two rounds of challenge were conducted: a group of chicken in each vaccination group was challenged between 31 and 35 weeks of age, while another group was chal- lenged between 51 and 55 weeks of age. The “in-contact” challenge model was used in this experi- ment. For each vaccination group, the four challenge strains representing four local serovars were used in each challenge round. The efficacy of the vaccines was compared based on overall protection levels obtained and the duration of protection. The safety of the different vaccines was determined by the severity of post-vaccination reactions. The need for the incorporation of the NAD-independent strain in the vaccine was evidenced by the low protection level against NAD-independent challenge recorded in the group of birds vaccinated with the commercial vaccine. The results obtained confirmed not only the variation in virulence of dif- ferent South African serovars, with serovar C-3 being the most virulent and serovar B having almost no virulence but also the age related increase in susceptibility. The importance of a suitable formula- tion of the vaccine is discussed. Keywords: Avibacterium paragallinarum, C-3 serovar, chickens, coryza vaccine, NAD-independent strain * Author to whom correspondence is to be directed. Present address: GALVmed, Edinburgh EH26 OPZ Scotland, UK. E-mail: Baptiste.Dungu@galvmed.org 1 Onderstepoort Biological Products, Private Bag X07, Onder- stepoort, 0110 South Africa 2 SEPPIC, Tour Kupka C-7, Bld Franck Kupka, 92039 Paris, France 3 Department of Microbiology and Plant Pathology, University of Pretoria, Pretoria, 0002 South Africa 4 University of the Free State, P.O. Box 339, Bloemfontein, 9300 South Africa Accepted for publication 19 April 2009—Editor 300 Infectious coryza vaccines containing an NAD-independent strain of Avibacterium paragallinarum that can be either NAD (or V-factor) dependent or independent (Mouahid, Bisgaard, Morley, Mut ters & Mannheim 1992; Bragg, Coetzee & Verschoor 1993) for growth in vitro. Three serogroups are recognised, namely A, B and C (Page 1962), with up to four se- rovars within serogroup A and C (Kume, Sa wata, Nakai & Matsumoto 1983; Blackall, Eaves & Rogers 1990). The role played by NAD-independent sero vars has been extensively studied in South Afri ca (Miflin, Horner, Blackall, Chen, Bishop, Mor row, Ya ma guchi & Iritani 1995; Bragg, Coetzee & Ver schoor 1996), where they have been shown to be the most dom- inant serovars in certain provinces such as Kwa- Zulu-Natal. Since first recorded in South Africa in the late 1960s (Buys 1982), IC has occurred widely in different parts of the country. Despite the introduction of a vaccine in the mid-1970s, which seemed to decrease the incidence of the disease, more outbreaks were re- corded in the 1980s. Studies conducted by Bragg et al. (1996) demonstrated a significant shift in the in- cidence of the four serovars occurring in the country over a 30-year period. Serovar C-3 had become pre- dominant, representing 73 % of isolated serovars in the 1990s, mainly in the period following the intro- duction of the vaccine, which did not include C-3. This work and the work by other researchers dem- onstrated and confirmed the importance of including locally occurring serovars in the vaccine for effec- tive control (Terzolo, Sandoval & Gonzalez Pondal 1997; Blackall 1999). Inactivated multivalent vaccines are used worldwide for the control of IC, most of them comprising sero- vars of serogroups A, B and C (Blackall1999). As for most inactivated vaccines, the adjuvant used for the formulation of the vaccine plays a critical role in de- fining the type of protection afforded by the antigen and the duration of immunity. Most IC vaccines are formulated with an oil emulsion. Ideally the vaccine should provide effective protection while being safe and having the least effect on chicken productivity. A long-lasting immunity is therefore preferred in or- der to avoid re-vaccinating layers during their pro- duction period. Water-in-oil emulsions are thus com- monly used as they are known to induce strong and long-term immunity (Aucouturier, Dupuis & Ganne 2001). One of their main disadvantages, however, is the possibility of local reactions when used with a crude antigen (Dupuis, Ascarateil, Aucouturier & Ganne 2006). Double oil emulsions, consisting of a water-in-oil-in-water emulsion are usually safer and also suitable for poultry (Dupuis et al. 2006). Depending on the type of formulations, type of emulsifying agent, the surfactant used and many other factors, different responses can be obtained with different formulation that will affect the protec- tion ability of the vaccine, hence the need for in vivo evaluations. An effective combination of a relevant vaccine strain and suitable adjuvant are critical for ensuring that the vaccine provides optimal protection. Since its introduction in the mid-1970s the Onder- stepoort Biological Products (OBP) IC vaccine has been adjusted to include the main serovars occur- ring in the country. The latest vaccine, CoryzaPlus vaccine, includes a C-3 NAD-independent strain to- geth er with locally occurring serovar A and C iso- lates. The present study was designed to evaluate the ef- ficacy and safety of alternative experimental oil vac- cines formulated with the same vaccine antigens as the OBP CoryzaPlus IC vaccine but using different oil adjuvants. The safety and ability of these formu- lations to provide long term protection to layers kept throughout a production cycle were compared to those of an international IC commercial vaccine registered in South Africa. MATERIAL AND METHODS Vaccines The OBP CoryzaPlus vaccine was used in the study, together with two experimental oil vaccines formu- lated with the same vaccine antigens as the OBP vaccine and a commercial vaccine, referred to as “ComA”. The OBP CoryzaPlus vaccine is a water- in-oil emulsion of formalin-inactivated A. paragalli­ na rum strains. It is a quadrivalent vaccine containing serotypes A and C (2 strains) and a C-3 NAD- independent vaccine strain. The same strains were used to formulate two experimental vaccines: one with the water-in-oil-in-water double emulsion Mon- tanide™ ISA 206 VG (Seppic) and the other with the water-in-oil emulsion, Montanide™ ISA 70 VG (Seppic). The commercial vaccine is a water-in-oil emulsion containing serovar A, B and C, but not the NAD-independent strain. The emulsions of the experimental vaccines were evaluated at the SEPPIC vaccine department labo- ratory in Castres (France) for physical and chemical properties using the KF values, the aspect, the par- ticle size and the stability at 4 °C, room temperature 301 B. DUNGU et al. and 37 °C up to 1 month. They were compared to a placebo emulsions made of a saline solution. Chickens In order to ensure homogeneity among the chick- ens to be studied, 750 fertilized eggs were procured and incubated simultaneously at the Agriculture Re- search Council (ARC) Poultry Research Centre at Glen Agriculture College, situated near Bloem fon- tein in the Free State Province, South Africa. Hatched chicks were kept for 16 weeks before being placed randomly into groups of 70 birds each. Two of the groups, totalling 140 chicks, were used for each vac cine, and the same number for non-vaccinated controls. For the challenge experiment, specific groups of birds were transferred to isolated layer facilities in the Animal House of the University of the Free State, Bloemfontein (UFS). This experiment was performed with the approval of the UFS Ethics Committee un- der project number 04/04. Bacterial isolates used for challenge The challenge strains used in this experiment repre- sented all three recognised serogroups of A. para­ gallinarum and major strains known to occur in South Africa. Serovar C-3 (Tongaat) strain of A. par­ a gallinarum, known to be highly virulent and very prevalent in South Africa was used, in parallel with Strain C-2 (serovar C-2), isolate 0083 (serogroup A), Strain B (serogroup B) and the NAD-independent strain 1750 (serovar C-3). The purity of each of the challenge strains was es- tablished by plating out the challenge bacteria onto BTA plates, streaking with Staphylococcus aureus and incubating at 37 °C. Vaccination and challenge experiments Chickens in the four study groups were vaccinated at 17 weeks of age, and boosted 3 weeks later. The birds were vaccinated by the subcutaneous injec- tion of 0.5 mℓ of the respective vaccines during both vaccination procedures. All vaccinations were con- ducted by the same operator. The safety of each vaccine was assessed by evaluating local reactions in vaccinated chickens during the first week post- inoculation. This evaluation was conducted on a number of chickens randomly selected in each group, during the first and the second vaccination. Post vaccination signs recorded were principally swellings of different sizes. They were graded as “none”, when there was no detectable swelling, “mild” for a small lesion and “severe” when there was a prominent swelling, sometimes affecting the gener- al health of the chicken. Two rounds of challenges were conducted. The first round of challenges with virulent strains of A. para­ gallinarum was conducted when the birds were be- tween 31 and 35 weeks of age (from 15 weeks post vaccination). The second round of challenge, conducted in the remaining unchallenged vaccinated chickens, took place when they were between 51 and 55 weeks of age. As controls, unvaccinated chickens of similar age were included in all challenge experiments. The challenge method used was according to the “in-contact” challenge model established by Bragg (2002a), in which one bird in a group of ten birds is directly challenged by intra-sinus injection with 0.1 mℓ of a bacterial suspension. The remaining birds in the group are challenged through natural in-contact route as the ten birds in each group are in adjoining cages with a communal water supply. The clinical signs were recorded and scored over the 20 days post-challenge observation period, according to the method described by Bragg (2002a), and used to calculate percentage protection according to the above method. Due to space constraints at the animal laboratory facility of the UFS, each round of vaccination was divided into two phases. For the first round of chal- lenge (at Week 31 of age), the first phase involved the Tongaat strain, the C-2 strain and the 0083 strain. The cage were then cleaned, disinfected and left empty for a week before conducting the second phase with the NAD-independent strain and the se- rogroup B strain. For the second round, the first phase (Week 51 of age) involved the Tongaat and the C-2 strains, while the second phase involved strain 0083 and the NAD-independent strain 1750. RESULTS Stability of the experimental IC oil vaccines Tables 1A and 1B summarize the results obtained for the physical characteristic evaluation and stabil- ity of the two experimental vaccines formulated with Montanide™ ISA 206 VG and Montanide™ ISA 70 VG respectively. Each one of the two experimental coryza vaccines had comparable physical charac- teristics to the equivalent SEPPIC placebo in terms 302 Infectious coryza vaccines containing an NAD-independent strain of Avibacterium paragallinarum of homogeneity and particle size. They all were also stable for a month at 4 °C and room temperature. Post-vaccinal safety of the vaccines Vaccination reactions were recorded during the first week post-vaccination in all groups, following the first and second vaccination. The results of the vac- cination reaction observations are given in Table 2 and Fig. 1 and 2. After the first vaccination, no vaccination reactions were seen in the control group (which were not vac- cinated) and in the birds vaccinated with the Montanide™ ISA 206 VG vaccine. The most severe clinical signs were observed in the birds vaccinated with the OBP IC vaccine and the Montanide™ ISA 70 VG formulation. After the second vaccination, no reactions were ob- served in the control group or in the birds vaccinat- ed with the Montanide™ ISA 206 VG formulation. The worst vaccination reactions were seen in the group of birds vaccinated with the ComA vaccine, with 20 % of the birds showing very severe vaccina- tion reactions characterized by subcutaneous swell- ing of varying degree. In some cases, the vaccina- tion reactions were so severe that these birds were removed from the experiment for ethical reasons as their general health condition was seriously deterio- rating. TABLE 1A Comparative evaluation of the Coryza special vaccine formulated with ISA 206 adjuvant, and a SEPPIC placebo emulsi- fied in the same adjuvant. Karl Fisher method (KF) defines the emulsion water content while D (v) is the mean of particles diameter Type of emulsion KF ( %) D (v, 0.5) D (v, 0.9) in µm Stability after 15 days 4 °C/RT/37 °C Stability after one month 4 °C/RT/37 °C SEPPIC placebo emulsion W/O/W 47.5 0.26 0.62 OK/OK/Deph OK/OK/Deph Coryza ISA 206 vaccine W/O/W 49.0 0.24 0.42 OK/OK/BR OK/OK/BR TABLE 1B Comparative evaluation of the coryza special vaccine formulated with ISA 70 adjuvant and a SEPPIC placebo emulsified in the same adjuvant. Karl Fisher method (KF) defines the emulsion water content while D (v) is the mean of particles di- ameter Type of emulsion KF ( %) D (v, 0.5) D (v, 0.9) in µm Stability after 15 days 4 °C/RT/37 °C Stability 1 month 4 °C/RT/37 °C SEPPIC placebo emulsion W/O 29.0 0.33 0.84 OK/OK/OK OK/OK/OK Coryza ISA 70 vaccine W/O 29.8 0.26 0.55 OK/OK/OK OK/OK/OK 303 B. DUNGU et al. Control ComA Isa 206 OBP IC ISA 70 No signs 0 4 0 11 9 Mild 0 24 8 38 30 Severe 100 72 92 51 61 100 90 80 70 60 50 40 30 20 10 0 P er ce n ta g e FIG. 1 Graphic representation of the percentage of vaccination reactions obtained after the first vaccination with the different vaccines used in this experiment Control ComA Isa 206 OBP IC ISA 70 No signs 100 52 100 84 97 Mild 0 28 0 3 2 Severe 0 20 0 12 1 100 90 80 70 60 50 40 30 20 10 0 P er ce n ta g e FIG. 2 Graphic representation of the vaccination reactions obtained after the second vaccination with the different vaccines used in this experiment TABLE 2 The number of birds showing vaccination reactions for each of the different vaccines Vaccine 1st vaccination 2nd vaccination No. birds No Mild Severe No. birds No Mild Severe Control 60 60 (100 %) 0 0 60 60 (100 %) 0 0 ComA 50 36 (72 %) 12 (24 %) 2 (4 %) 83 43 (52 %) 23 (28 %) 17 (20 %) ISA 206 50 46 (92 %) 4 (8 %) 0 (0 %) 78 78 (100 %) 0 0 OBP IC 45 23 (51 %) 17 (38 %) 5 (11 %) 122 103 (84 %) 4 (3 %) 15 (12 %) ISA 70 46 28 (61 %) 14 (30 %) 4 (9 %) 154 149 (97 %) 3 (2 %) 2 (1 %) 304 Infectious coryza vaccines containing an NAD-independent strain of Avibacterium paragallinarum 1: Tongaat C3 strain Vaccine Mean score % protection Control ComA ISA 206 OBP ISA 70 1.010 0.191 0.360 0.150 0.130 – 81.1 64.4 85.1 87.1 2: Tongaat C-2 strain Vaccine Mean score % protection Control ComA ISA 206 OBP ISA 70 0.58 0.04 0.09 0.29 0.00 – 93.1 84.5 50.0 100.0 3: A-1 isolate 0083 Vaccine Mean score % protection Control ComA ISA 206 OBP ISA 70 0.55 0.19 0.17 0.20 0.26 – 65.5 69.1 63.6 52.7 4: NAD-independent C-3 isolate 1750 Vaccine Mean score % protection Control ComA ISA 206 OBP ISA 70 0.15 0.125 0.04 0.04 0.08 – 16.7 73.3 73.3 46.7 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ISA 70 OBP Com A ISA 206 Control 0.0 0.5 1.0 1.5 2.0 2.5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ISA 70 OBP Com A ISA 206 Control 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ISA 70 OBP Com A ISA 206 Control 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ISA 70 OBP Com A ISA 206 Control TABLE 3A First vaccination: Graphic representation of the mean daily disease scores obtained from each of the vaccines for chick- ens challenged with different A. paragallinarum challenge strains and table summarizing the corresponding mean lesion score and percentage protection—the Tongaat (C-3) strain (1); C-2 strain (2); the A-1 isolate 0083 (3); and the NAD- independent C-3 isolate 1750 (4) 305 B. DUNGU et al. 1: Tongaat C3 strain Vaccine Mean score % protection Control ComA ISA 206 OBP ISA 70 1.65 0.595 0.065 0.245 0.055 – 64.0 96.1 85.2 96.7 2: Tongaat C-2 strain Vaccine Mean score % protection Control ComA ISA 206 OBP ISA 70 0.27 0.14 0.195 0.10 0.10 – 48.1 27.8 63.0 63.0 3: A-1 isolate 0083 Vaccine Mean score % protection Control ComA ISA 206 OBP ISA 70 0.130 0.075 0.090 0.090 0.080 – 42.3 30.8 30.8 38.5 4: NAD-independent C-3 isolate 1750 Vaccine Mean score % protection Control ComA ISA 206 OBP ISA 70 0.56 0.13 0.24 0.01 0.03 – 76.8 57.1 98.2 94.6 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ISA 70 ISA 206 OBP ComA Control 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ISA 70 OBP ComA ISA 206 Control 0.0 0.1 0.2 0.3 0.4 0.5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ISA 70 OBP ComA ISA 206 Control 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ISA 70 OBP ComA ISA 206 Control TABLE 3B Second vaccination: Graphic representation of the daily protection scores provided by each of the vaccines for chickens challenged with different A. paragallinarum challenge strains and table summarizing the corresponding mean lesion score and percentage protection—the Tongaat (C-3) strain (1); C-2 strain (2); the A-1 isolate 0083 (3); and the NAD- independent C-3 isolate 1750 (4) 306 Infectious coryza vaccines containing an NAD-independent strain of Avibacterium paragallinarum Efficacy of the vaccine formulations to different serovar after the first challenges The mean daily disease and protection scores for each of the groups of birds are summarized in Tables 3A and B for the first and second challenge rounds presented below. No clinical signs occurred in birds challenged with the strain of serovar B, thus the efficacy of the vac- cine against serogroup B could not be determined. The percentage protection calculated from the dis- ease challenge and a comparison of the mean dis- ease scores in the first and second rounds are pre- sented in Table 4. DISCUSSION Vaccine reactions of varying degrees occurred in all groups after the first vaccination, irrespective of the type of vaccine. The least severe vaccination reac- tions were seen in the chickens which received the Montanide™ ISA 206 VG formulation, which can be attributed to the fact that this was the only formula- tion in which the continuous phase is aqueous (wa- ter-in-oil-in-water), therefore causing the least in- flammatory local reaction as compared to water in oil emulsions (Jansen, Hofmans, Theelen, Manders & Schijns 2006). After the second round of vaccina- tion, the birds vaccinated with the ComA vaccine showed the most severe vaccination reactions: a total of 14 birds were removed from the experiment for ethical reasons based on the severity of their re- actions. The two experimental formulations induced vaccination reactions in only 3 % of the birds with the Montanide™ ISA 70 VG formulation and none of the birds inoculated with the Montanide™ ISA 206 VG formulation. Overall, the Montanide™ ISA formulations showed the least number of local post- vaccination reactions. A quadrivalent vaccine needs a potent adjuvant to induce protection. The reactogenic properties of the inactivated bacteria used as antigen in the present vaccines require a safe adjuvant. The acceptable balance between efficacy and safety is obtained through the use of a specific adjuvant formulation. Montanide™ ISA adjuvants are based on a homog- enous ready to use mix of purified mineral oils and refined oleic esters of anhydrous mannitol of vege- table origin. Emulsifying properties of specific surfac- tants can only be obtained through strict synthesis parameters, and have a direct impact on the vac- cine safety and efficacy (Stone 1988). Furthermore, surfactants used in Montanide™ formulations are manufactured in dedicated equipment which avoids any cross contamination. The hydrophilic parts of these amphiphile molecules are made of mannitol sugar known to have better injectability than sorbi- tol-based surfactants. Adjuvant formulations of this quality do not induce adverse reactions observed when multipurpose industrial sorbitan oleate-based formulation (Tween/Span adjuvant formulation) is used. The Montanide™ ISA 206 VG renders a wa- ter-in-oil-in-water emulsion in a one step process, giving a very fluid vaccine when containing 50 % of adjuvant. The Montanide™ ISA 70 VG renders a water-in-oil emulsion containing 70 % of adjuvant. In the later case, the antigenic media are entrapped in the oily phase. The different antigenic release profiles (Aucouturier et al. 2001) and the ratio of ad- juvant to antigenic media can explain the perfect safety profile obtained with Montanide™ ISA 206 VG. Indeed, aqueous-based formulations are quick- ly eliminated from the injection site while water-in-oil TABLE 4 Comparison of all of the vaccines based on percentage protection against the NAD-dependent, NAD-independent strain and for all of the challenge strains used (excluding serovar B-1, which did not result in any clinical signs when used to challenge chickens in this experiment, including the unvaccinated controls). Mean protections obtained for the first and second round of challenges for all of the vaccines and challenged strains used in this experiment are also included Challenge ComA OBP Isa 206 Isa 70 1st 2nd 1st 2nd 1st 2nd 1st 2nd C-3 81.1 64.0 85.1 85.2 64.4 96.1 87.1 96.7 C-2 93.1 48.1 50.0 63.0 84.5 27.8 100.0 63.0 A-1 65.5 42.3 63.6 30.8 69.1 30.8 54.7 38.5 Mean NAD-dependent protection 79.9 51.5 66.2 59.7 72.7 51.6 80.6 66.1 NAD-independent C-3 16.7 76.8 73.3 98.2 73.3 57.1 46.7 94.6 Mean overall protection (all isolate) 64.1 64.2 68.0 78.9 72.8 54.3 72.1 80.3 307 B. DUNGU et al. emulsions ensure a long-lasting release of the anti- gens (Jansen et al. 2005). The virulence of different challenge strains could be assessed through the monitoring of clinical scores in the unvaccinated control animals. The Tongaat C-3 strain demonstrated the highest virulence, fol- lowed by the C-2 and the A-1 strains. Serovar B did not induce any adverse reactions. The high clinical score recorded with serovar C-3 (1.01 and 1.65 after the first and the second challenge, respectively) is consistent with previous results (Bragg 2002a) and confirms the association of the C-3 serovars with most severe outbreaks in South Africa. The lack or poor virulence of South African serovar B has also been recorded previously (Bragg 2002a.) Good levels of protection were, however, recorded for the C-3 serovar for all the vaccine formulations, at both challenges, which strengthen the value of vaccination for the control of IC in South Africa. One of the main objectives of this experiment was to investigate the duration of protection provided by the different vaccines. In order to do this, the chick- ens were kept until they were older than 50 weeks. Vaccinated birds were then challenged with the same challenge strains as those used at approxi- mately 30 weeks of age. The results of the second challenge with serovar C-3 in unvaccinated birds showed an increase of the mean disease score from 1.01 in the first challenge to 1.65 in the second challenge, indicating a possible worsening in bird susceptibility to the bacterium associated with in- creasing age. A similar increase in clinical signs as the birds aged was also recorded with the NAD- independent C-3 strain, while the opposite was ob- served with the C-2 and the A-1 strains. Further work is required to confirm these observations. Dif- ferences of occurrence of IC in different age groups have been reported elsewhere with variations be- tween different countries and production systems (Blackall 1999) who reported on a study of village chickens in Thailand where it was reported that IC was the most common cause of death in chickens less than 2 months old and in those over 6 months of age. While the protection level with the ComA vaccine decreased from 81.1 % to 64 % from the first to the second challenge, the OBP CoryzaPlus vaccine generated an unchanged level of protection (85.1 % for the first challenge and 85.2 % in the second chal- lenge). The protection levels were found to increase to 90 % when the Montanide™ ISA 70 VG and Mon- tanide™ ISA 206 VG formulations were used. The Montanide™ ISA 70 VG vaccine generated the best protection overall to all challenge strains, close- ly followed by the OVI vaccine. This can be attrib- uted to the similarity in their formulation, water-in-oil and the inclusion of the NAD-independent C-3 strain. Montanide™ ISA 70 VG also showed the best safety results. Larger field experiments are re- quired to confirm the findings of the present study. During the first round of challenge experiments, the clinical signs associated with IC in the unvaccinated control birds challenged with the Tongaat strain (C-3) were very severe. The levels of protection obtained when the different vaccinated groups of birds were challenged with serovar C-3 ranged from 64.4 % (ISA 206) to 87.1 % for the Montanide™ ISA 70 VG formulation. The clinical signs obtained when unvaccinated birds were challenged with the serovar C-2 strain were less severe than those obtained when the unvacci- nated birds were challenged with serovar C-3. These results confirm previous findings by Bragg (2002a). The levels of protection obtained by the different vaccines against the serovar C-2 challenge ranged from 50.0 % for the OBP vaccine to 100 % for the ISA 70 formulation. When unvaccinated birds were challenged with strain 0083 (serovar A-1), the clinical signs were less severe than those obtained in previous chal- lenge experiments. The levels of protection obtained against challenge with serovar A-1 ranged from 52.7 % for Montanide™ ISA 70 VG and 69.1 % ob- tained with the Montanide™ ISA 206 VG formula- tion. When unvaccinated birds were challenged with se- rogroup B, no clinical signs were seen. Previous work has demonstrated that the South African sero- group B strains are of very low virulence (Bragg 2002b), but in those experiments, some clinical signs were recorded. The fact that no clinical signs were seen in the unvaccinated birds makes any comparison of the protection impossible. The levels of virulence recorded with the different serovars correspond to the findings in South African strains of A. paragallinarum recorded by Bragg (2002a). The ability to protect birds through vaccina- tion confirms the need for well-structured vaccina- tion programmes. The birds were also challenged with an NAD-inde- pendent strain of A. paragallinarum. In previous ex- periments, the virulence of both naturally occurring NAD-independent strains (Bragg 2002b) and ex- 308 Infectious coryza vaccines containing an NAD-independent strain of Avibacterium paragallinarum perimentally produced NAD-independent strains (Taole et al. 2002) were found to be of low virulence. It has also been demonstrated that the levels of pro- tection obtained when vaccinated birds were chal- lenged with the NAD-independent strains were very low (Bragg 2004). In the present study the preva- lence of clinical signs noted in the unvaccinated birds was indeed very low: a mean disease score of only 0.15 was obtained in the unvaccinated birds challenged with the NAD-independent serovar C-3 strain as compared to a mean disease score of 1.01 obtained with the NAD-dependent serovar C-3 strain of A. paragallinarum. The need for the inclusion of a NAD-independent strain in the vaccine was shown by the low protec- tion level afforded by the commercial vaccine ComA, which lacks such a strain. The protection afforded by the commercial vaccine to a challenge with the NAD-independent strain improved however in late challenge, possibly as a result of an improved cross- protection over time with the other strains in the vac- cine. These findings are in contrast to those of Ja- cobs et al. (2000) who demonstrated that the Nobilis coryza vaccine (Intervet International BV) provided protection against the NAD-independent strains in 9-week-old chickens vaccinated 2 weeks earlier, despite the fact that it does not include an NAD- independent vaccine antigen. Bragg (2004), on the other hand, demonstrated that there was evidence of immune evasion by the NAD-independent strains when an experimental vaccine only containing NAD- dependent strains was used to vaccinate the birds. The difference between the finding of Jacobs et al. (2000) and Bragg (2004) could be attributed to the different challenge models used. Bragg et al. (2002b) demonstrated that the naturally occurring NAD-in de- pendent strains are of low virulence. Taole et al. (2002) showed that there is a substantial decrease in virulence when NAD-dependent strains of A. par­ agallinarum are experimentally converted to NAD- independent strains. This low virulence level could account for the perceived efficacy of a vaccine against the NAD-independent strains as reported by Jacobs et al. (2000) Given the high prevalence of NAD-independent serovars in South Africa, their in- volvement in a number of outbreaks throughout the country, and previous observation of the limited ability of the NAD-dependent vaccine strain to pro- tect birds against them, it is critical that vaccines in use in the country contain NAD-independent strains. The two experimental vaccines, as well as the OBP CoryzaPlus vaccines contained an NAD-independ- ent strain of A. paragallinarum, while the ComA vac- cine did not. Protection levels of only 16.7 % were recorded with ComA vaccine while it was between 46.7 % (Montanide™ ISA 70 VG) and 73.3 % (Mon- tanide™ ISA 206 VG and OBP) for the vaccines containing an NAD-independent strain. These data clearly indicate the ability of a vaccine containing an NAD-independent strain to protect against these strains. In a previous study conducted in South Afri- ca it was demonstrated that a commercial vaccine without an NAD-independent strain generated good protection against local NAD-independent C-3 strain (Jacobs et al. 2000). The fact that the chickens used in the experiment were vaccinated at 3 and 7 weeks of age and challenged 2 weeks later could explain the good level of protection recorded. In the present study the chickens were vaccinated and challenged much later (vaccination at 17 and 20 weeks, chal- lenge at 33 or 55 weeks of age), in order to evaluate the level of protection at different stages of the pro- ductive life of layers. The results of the present ex- periment are certainly much closer to the layer pro- duction system and suggest the need for the use of a vaccine that includes NAD-independent strains in areas where NAD-independent variants are known to occur. The low virulence recorded for the NAD- independent strains discussed above could also contribute to this discrepancy. The results obtained in the present study confirm the variation in the virulence of different serovars occurring in South Africa, with serovar C-3 being the most virulent and serovar B having almost no virulence. The results indicate an age related in- crease in susceptibility to infection, as illustrated in the increased disease scores. The study highlights the importance of a suitable vaccine formulation that generates protection throughout the productive life of chickens and the need for incorporation of lo- cal dominant strains in the vaccine, including NAD- independent strains. REFERENCES AUCOUTURIER J., DUPUIS L. & GANNE V. 2001. Adjuvants de signed for veterinary and human vaccines. Vaccine, 19: 2666–2672. BLACKALL, P.J. 1999. Infectious coryza: overview of the dis- ease and new diagnostic options. Clinical Microbiological Review, 12:627-32. BLACKALL, P.J., EAVES, L.E. & ROGERS D.G. 1990. Proposal of a new serovar and altered nomenclature for Haemophilus paragallinarum in the Kume hemagglutinin scheme. Journal of Clinical Microbiology, 28:1185–1187. BLACKALL, P.J., CHRISTENSEN, H., BECKENHAM, T., BLACK- ALL, L.L. & BISGAARD, M. 2005. Reclassification of Pas teur­ ella gallinarum, Haemophilus paragallinarum, Pas teur ella avi um and Pasteurella volantium as Avibacterium gall i na rum gen. nov. comb. nov., Avibacterium paragallina rum comb. 309 B. DUNGU et al. nov., Avibacterium avium comb. nov. and Avibacterium vol­ antium comb. nov. International Journal of Systematic and Evolutionary Microbiology, 55:353–362. BRAGG, R.R. 2002a. Virulence of South African isolates of Haemophilus paragallinarum. Part 1: NAD-dependent field isolates. Onderstepoort Journal of Veterinary Research, 69: 163–169. BRAGG, R.R. 2002b. Virulence of South African isolates of Haemophilus paragallinarum. Part 2: Naturally occurring NAD-independent field isolates. Onderstepoort Journal of Veterinary Research, 69:171–175. BRAGG, R.R., COETZEE, L. & VERSCHOOR, J.A. 1993. Plasmid encoded NAD independence in some South African isolates of Haemophilus paragallinarum. Onderstepoort Jour nal of Veterinary Research, 60:147–152. BRAGG, R.R., COETZEE L. & VERSCHOOR, J.A. 1996. Changes in the incidences of the different serovars of Haemophilus paragallinarum in South Africa; a possible ex- planation of vaccination failures. Onderstepoort Journal of Veterinary Research, 63:217–226. BRAGG, R.R. 2004. Evidence of possible evasion of protective immunity by NAD-independent isolates of Haemophilus par­ agallinarum in poultry. Onderstepoort Journal of Veterinary Research, 71:53–58. BUYS, S.B. 1982. Die bereiding van bakterieë teen Haemophilus paragallinarum besmetting, geskik vir Suid-Afrikaanse om- standighede. M.Med.Vet. thesis, University of Pretoria. DROUAL, R., BICKFORD A.A., CHARLTON B.R., COOPER G.L. & CHANNING, S.E. 1990. Infectious coryza in meat chickens in the San Joaquin Valley of California. Avian Dis­ eases, 34:1009–1016. DUPUIS, L., ASCARATEIL, S., AUCOUTURIER, J. & GANNE, V. 2006. SEPPIC vaccine adjuvants for poultry. Annals of the New York Academy of Science, 1081:202-205. HORNER, R.F., BISHOP, G.C., JARVIS, C.J. & COETZEE, T.H.T. 1995. NAD (V-factor)-independent and typical Hae m­ ophilus paragallinarum infecting commercial chickens: a five year field study. Avian Pathology, 24:453–463. JACOBS, A.A. & VAN DER WERF, J. 2000. Efficacy of a com- mercially available coryza vaccine against challenge with re- cent South African NAD-independent isolates of Haemophilus paragallinarum in chickens. Journal of the South African Vet­ erinary Association, 71:109–110. JANSEN, T., HOFMANS, M.P.M., THEELEN, M.J.G., MAN- DERS, F. & SCHIJNS V.E.J.C. 2006. Structure- and oil type- based efficacy of emulsion adjuvants. Vaccine, 24:5400– 5405. JANSEN, T., HOFMANS, M.P.M., THEELEN, M.J.G. & SCHIJNS, V.E.J.C. 2005. Structure-activity relations of water-in-oil vac- cine formulations and induced antigen-specific antibody re- sponses Vaccine, 23:1053–1060. KUME, K., SAWATA, A., NAKAI, T. & MATSUMOTO, ? 1983. Serological classification of Haemophilus paragallinarum with a hemagglutinin system. Journal of Clinical Microbiology, 17: 958–964. MIFLIN, J.K., HORNER, R.F., BLACKALL, P.J., CHEN, X., BISHOP, G.C., MORROW, C.J., YAMAGUCHI, T. &. IRI TANI, Y. 1995. Phenotypic and molecular characterization of V-fac- tor (NAD)-independ ent Haemophilus paragallinarum. Avian Diseases, 39:304–308. MOUAHID, M., BISGAARD, M., MORLEY, A.J., MUTTERS, R. & MANNHEIM, W. 1992. Occurrence of V-factor (NAD) inde- pendent strains of Haemophilus paragallinarum. Veterinary Microbiology, 31:363–368. PAGE, L.A. 1962. Haemophilus infections in chickens: I. Char- acteristics of 12 Haemophilus isolates recovered from dis- eased chickens. American Journal of Veterinary Research, 23:85–95. TAOLE, M., ALBERTYN, J., VAN HEERDEN, E. & BRAGG, R.R. 2002. Virulence of South African isolates of Haemophilus paragallinarum. Part 3: experimentally produced NAD-inde- pendent isolates. Onderstepoort Journal of Veterinary Res­ earch, 69:189–196. TERZOLO, H.R., SANDOVAL, V.E. & GONZALEZ PONDAL, F. 1997. Evaluation of inactivated infectious coryza vaccines in chickens challenged by serovar B strains of Haemophilus paragallinarum. Avian Pathology, 26:365–376. STONE, H.D.1988. Optimization of hydrophile-lipophile balance for improved efficacy of Newcastle disease and avian influ- enza oil-emulsion vaccines. Avian Diseases, 32:68–73.