Cross-protection study of the nine serovars of Haemophilus paragallinarum in the Kume haemagglutinin scheme

Abstract

The cross-protection and haemagglutination-inhibition antibodies present in chickens vaccinated with one of the nine currently recognized Kume haemagglutinin serovars of Haemophilus paragallinarum were investigated. The results confirmed the widely accepted dogma that serogroups A, B, and C represent three distinct immunovars. Within Kume serogroup A, there was generally good cross-protection among all four serovars. However, within Kume serogroup C, there was evidence of a reduced level of cross-protection between some of the four serovars. The haemagglutination-inhibition antibody levels generally showed the same trend as with the cross-protection results. This study suggests that some apparent field failures of infectious coryza vaccines may be due to a lack of cross-protection between the vaccine strains and the field strains. Our results will help guide the selection of strains for inclusion in infectious coryza vaccines.

Il a été étudié la protection croisée et la présence des anticorps inhibant l'hémagglutination (HI) chez des poulets vaccinés avec un des neuf sérovars hémagglutinants d' Haemophilus paragallinarum de la classification de Kume. Les résultats ont confirmé le dogme largement accepté que les sérogroupes A, B et C représentent trois immunovars distincts. A l'intérieur du sérogroupe A de Kume, il y a généralement une bonne protection croisée entre les qautre sérovars. Cependant, à l'intérieur du sérogroupe C de Kume, une diminution du niveau de protection croisée a été mise en évidence entre certains des quatre sérovars. Les taux des anticorps HI a généralement montré la même tendance que les résultats de protection croisée. Cette étude suggère que certains échecs apparents de vaccination du coryza infectieux, observés sur le terrain, peuvent être dus à un manque de protection croisée entre les souches vaccinales et les souches sauvages. Nos résultats peuvent servir de guide à la sélection de souches à inclure dans les vaccins du coryza infectieux.

Es wurden die Kreuzschutz- und hämagglutinationshemmenden (HAH-) Antikörper in Hühnern untersucht, die mit einem der zur Zeit bekannten neun Kume-Hämagglutinin-Serovare von Hämophilus paragallinarum vakziniert worden waren. Die Ergebnisse bestätigten das allgemein anerkannte Dogma, dass die Serogruppen A, B und C drei verschiedene Immunovare darstellen. Innerhalb der Kume-Serogruppe A bestand generell ein guter Kreuzschutz gegen alle vier Serovare. In der Kume-Serogruppe C wurde jedoch eine Verminderung des Kreuzschutzes zwischen einigen der vier Serovare nachgewiesen. Im allgemeinen zeigten die HAH-Antikörper den gleichen Trend wie bei den Kreuzschutzergebnissen. Diese Studie lässt vermuten, dass ein offensichtliches Versagen von Impfstoffen gegen den ansteckenden Hühnerschnupfen im Feld auf das Fehlen eines Kreuzschutzes zwischen den Impf- und den Feldstämmen beruhen könnte. Unsere Ergebnisse werden bei der Auswahl der Stämme für Impfstoffe gegen den ansteckenden Hühnerschnupfen hilfreich sein.

Se investigó la protección cruzada y los anticuerpos de inhibición de la hemaglutinación (HI) presentes en pollos vacunados con una de los nueve serovares reconocidos actualmente de Haemophilus paragallinarum. Los resultados confirmaron el dogma extensamente aceptado que los serogroupos A, B y C representan tres distintas immunovariantes. Dentro del serogrupo A de Kume, generalmente había buena protección cruzada entre los cuatro serovares. Sin embargo, dentro del serogrupo C de Kume, hubo evidencia de un nivel reducido de protección cruzada entre algunos de los cuatro serovares. Los niveles de anticuerpos de HI demostraron generalmente la misma tendencia que los resultados de la protección cruzada. Este estudio sugiere que algunos fallos evidentes en el campo de las vacunas de coriza infeccioso, puedan ser debidos a una carencia de protección cruzada entre las cepas vacunales y las cepas de campo. Nuestros resultados ayudarán a dirigir la selección de cepas para la inclusión en vacunas de coriza.

Introduction

Haemophilus paragallinarum is the causative agent of infectious coryza of chickens. This acute respiratory disease can occur in both growing chickens and layers, causing an increased culling rate in meat chickens and a reduction in egg production (10 to 40%) in laying and breeding hens, particularly on multiage farms (Blackall & Matsumoto, 2003).

Two inter-related schemes have been mainly used to serotype H. paragallinarum. The Page scheme was originally developed with the use of a slide agglutination test to recognize the three serovars, A, B, and C (Page, 1962). The Kume serotyping scheme was based on a haemagglutination-inhibition (HI) test and recognized seven serovars organized into three serogroups, termed I, II, and III (Kume et al., 1983). In further studies, Eaves et al. (1989) identified a new serovar within Kume serogroup I, and Blackall et al. (1990b) identified a further serovar within Kume serogroup II. Hence, Blackall et al. (1990b) proposed to alter the Kume scheme nomenclature to emphasize the fact that Kume serogroups I, II, and III corresponded to the Page serovars A, C, and B, respectively. Thus, the nine currently recognized serovars are termed serovars A-1, A-2, A-3, A-4 within Kume serogroup A (with all of the serovars corresponding to Page serovar A), serovar B-1 within Kume serogroup B (which corresponds to Page serovar B), and serovars C-1, C-2, C-3, and C-4 within Kume serogroup C (with all of the serovars corresponding to Page serovar C) (Blackall et al., 1990b).

It is generally accepted that the Page serovars, or Kume serogroups, represent three distinct immunovars (Blackall, 1999). The accepted dogma, with little experimental evidence, is that serovars within a Kume serogroup are cross-protective (Blackall, 1995). Specific evidence to date is: first, that Kume serovars C-1 and C-2 and C-2 and C-4 are cross-protective; and, second, that Kume serovars A-1 and C-1, A-1 and C-2, A-4 and C-2 and A-4 and C-4 are not cross-protective (Kume et al., 1980; Blackall & Reid, 1987; Blackall, 1991). Cross-protection between the remaining Kume serovars has not been examined and is thus unknown.

The protective antigens of H. paragallinarum have not been definitively identified. However, the haemagglutinin antigens have been proposed as the main pathogenic and immunogenic structures (Yamaguchi et al., 1993). Thus, HI antibody levels of both vaccinated and passive-immunized chickens have been closely correlated with protection against clinical signs and nasal clearance of the challenge organism (Kume et al., 1984; Takagi et al., 1991).

The aims of the present study were to determine the cross-protection between all nine recognized Kume serovars of H. paragallinarum in vaccination-challenge trials in chickens and to also examine the level of HI antibodies in the chickens.

Materials and Methods

Bacteria and growth conditions

The nine reference strains of the Kume scheme used were 221 (A-1), 2403 (A-2), E-3C (A-3), HP14 (A-4), 2671 (B-1), H-18 (C-1), Modesto (C-2), SA-3 (C-3), and HP60 (C-4) (Blackall et al., 1990b). The reference strains were all sourced from the culture collection held at the Animal Research Institute, Department of Primary Industries and Fisheries Queensland, Australia. The origin of strains is presented in Table 1. Both brain–heart infusion broth and agar plates, supplemented with 1% sodium chloride, 0.0025% (w/v) reduced nicotinamide adenine dinucleotide, and 1% (v/v) filter-sterilized, heat-inactivated horse serum, were used for propagation and maintenance of bacterial cultures. Also, 10% sheep blood agar with Staphylococcus epidermidis as feeder colony was used.

Table 1. Designations, serovars and origins of H. paragallinarum strains used in the present study

Strain	Serovar	Origin	Reference
221	A-1	Japan	Kume et al. (1983)
2403	A-2	Germany	Kume et al. (1983)
E-3C	A-3	Brazil	Kume et al. (1983)
HP14	A-4	Australia	Eaves et al. (1989)
2671	B-1	Germany	Kume et al. (1983)
H-18	C-1	Japan	Kume et al. (1983)
Modesto	C-2	USA	Kume et al. (1983)
SA-3	C-3	South Africa	Kume et al. (1983)
HP60	C-4	Australia	Blackall et al. (1990b)

Vaccination/challenge trials

A total of 900, 1-day-old, Mycoplasma gallisepticum-free and Mycoplasma synoviae-free, clinically healthy, ALPES Leghorn chickens were used in the study. All chickens were individually identified. Vaccines for each reference strain of H. paragallinarum were produced. Briefly, bacteria were grown overnight in brain–heart infusion broth, and supplemented with 1% sodium chloride, 0.0025% (w/v) reduced nicotinamide adenine dinucleotide, and 1% (v/v) filter-sterilized, heat-inactivated horse serum. A viable count was performed and the culture inactivated with 0.01% (w/v) thiomersal as reported by Blackall (1991). Once the viable count results were available, the cells suspensions were adjusted to 5×10⁸ colony-forming units/ml and aluminium hydroxide added to a final concentration of 10%. Groups of 90 chickens were inoculated subcutaneously at 6, and 9 weeks of age with 1 ml relevant vaccine. Three weeks after the second vaccination, all chickens were bled, relocated into nine groups containing 10 vaccinated chickens plus 10 unvaccinated control chickens and challenged by nasal instillation of 0.2 ml overnight broth culture of the relevant H. paragallinarum strain containing 5×10⁸ colony-forming units/ml as reported by Blackall (1991). Clinical signs of infectious coryza were recorded from the second to the seventh day post-challenge. All chickens were then humanely euthanized and both infraorbital sinuses cultured onto blood agar with a S. epidermidis nurse colony. Protection was defined as the absence of clinical signs of infectious coryza and a failure to re-isolate H. paragallinarum. A number of re-isolated bacteria were serotyped as previously reported (Soriano et al., 2001). Protection rates were compared by chi-square tests and considered significant at a probability of P<0.01. Furthermore, the sera of 30 chickens from each of the nine serovar-vaccinated groups were examined in HI tests using the haemagglutinins of the nine reference strains. The titres were expressed as the reciprocal of the highest dilution of serum sample that showed complete inhibition of the haemagglutinating activity. The serum HI antibody titres were transformed to base 10 logarithms to remove any skewness in the data. The HI titres of the various groups were compared by an analysis of variance and differences were tested for significance by the Tukey test. Results were considered significant at a probability of P<0.01.

Results

The results of the cross-protection trials are presented in Table 2. As noted earlier, a protected chicken was defined as one that showed no clinical signs during the observation period and yielded no haemophili from the sinus following culture. Within Kume serogroup A, serovars A-1, A-2 and A-3 were strongly cross-protective. The Kume serovar A-4 vaccine resulted in a significantly lower protection for the serovar A-2 challenge than for the other Kume serogroup A serovars. Within Kume serogroup C, there was a good level of cross-protection for serovars C-1, C-2 and C-3, with some exceptions. The Kume serovar C-1 vaccine provided significantly lower protection against the serovar C-2 challenge than for the other serovars of the group. Similar significant lower levels of protection resulted for the vaccine C-2/challenge C-3 and vaccine C-3/challenge C-2. Within the C-4 vaccine groups, significantly lower levels of protection were present for the challenge from serovars C-1 and C-3. The only instance of a vaccine being able to provide cross-protection that was at the same level as the homologous level was for the serovar C-4 vaccine and the serovar B-1 challenge.

Table 2. Protection in groups of vaccinated and challenged chickens with each serovar of H. paragallinarum

	Protection (%) against challenge with H. paragallinarum of serovar^a
Vaccine	A-1	A-2	A-3	A-4	B-1	C-1	C-2	C-3	C-4
A-1	100^A	90^A	90^A	80^A	30^B	10^B	20^B	30^B	50^B
A-2	90^A	90^A	90^A	30^B	11^B	20^B	50^B	0^B	60^B
A-3	100^A	90^A	100^A	50^B	30^B	20^B	20^B	10^B	50^B
A-4	100^A	60^B	70^A	90^A	50^B	30^B	20^B	30^B	40^B
B-1	10^B	50^B	30^B	10^B	100^A	20^B	60^B	0^B	50^B
C-1	10^B	60^B	60^B	0^B	50^B	100^A	50^B	80^B	100^A
C-2	0^B	40^B	50^B	10^B	33^B	70^A	70^A	40^B	90^B
C-3	30^B	50^B	40^B	11^B	20^B	100^A	40^B	90^A	80^B
C-4	40^B	30^B	50^B	44^B	70^A	50^B	70^A	40^B	100^A
Control	0^B	10^B	0^B	0^B	0^B	0^B	0^B	20^B	30^B
^a Values within a column followed by different uppercase superscripts differ at the 1% level of significance.

The results of the serological testing are presented in Table 3. As a general rule, the homologous HI titre was significantly higher than any other titre (including within the relevant Kume serogroup). This is shown by Kume serovars A-1, A-4, C-2, C-3 and C-4. A notable exception was the Kume serovar C-1 antigen. This antigen gave HI titres with antisera to serovars C-1, C-2, C-3 and C-4 that were not significantly different. The A-2 and A-3 antigens showed some degree of cross-reactivity within their serogroup (giving equivalent titres with sera from serovars A-1 and A-4, respectively).

Table 3. HI antibody titres to all nine Kume serovars of H. paragallinarum in vaccinated chickens

	Mean (log10) of HI antibody titres measured with haemagglutinin of H. paragallinarum of serovar^a
Vaccine	A-1	A-2	A-3	A-4	B-1	C-1	C-2	C-3	C-4
A-1	1.25^A	1.00^A	1.05^B	0.59^C	0.00^B	0.40^B	0.00^C	0.10^E	0.00^C
A-2	0.63^B	0.97^A	0.86^C	0.06^D	0.00^B	0.73^B	0.03^C	0.06^E	0.00^C
A-3	0.48^B	0.51^B	1.32^A	0.90^B	0.00^B	0.43^B	0.00^C	0.00^E	0.00^C
A-4	0.50^B	0.53^B	1.41^A	1.25^A	0.00^B	0.23^C	0.00^C	0.24^D	0.00^C
B-1	0.00^C	0.00^C	0.03^E	0.03^D	1.13^A	0.64^B	0.00^C	0.53^C	0.00^C
C-1	0.00^C	0.00^C	0.00^E	0.00^D	0.00^B	1.00^A	0.11^C	0.16^D	0.05^C
C-2	0.00^C	0.13^C	0.28^D	0.00^D	0.00^B	1.03^A	1.30^A	0.80^B	0.79^B
C-3	0.00^C	0.03^C	0.00^E	0.00^D	0.00^B	1.32^A	0.10^C	1.25^A	0.04^C
C-4	0.00^C	0.30^BC	0.33^D	0.00^D	0.00^B	0.96^A	0.68^B	0.41^C	1.35^A
Control	0.00^C	0.00^C	0.00^E	0.00^D	0.00^B	0.00^C	0.00^C	0.00^E	0.00^C
^a Mean values within a column followed by different uppercase superscripts differ at the 1% level of significance.

With a few exceptions, all HI titres for tests across Kume serogroups were significantly lower than any titre within a Kume serogroup. The exceptions were all associated with antigens that gave low cross-titres within their respective serogroup. For example, the A-4 antigen gave a very low HI titre to sera from the serovar A-2 vaccinated chickens. The titre was so low that the titres to the A-2 sera were not significantly different to the titres detected in the sera from the B-1, C-1, C-2, C-3 and C-4 vaccinated chickens. A similar result was obtained with the C-4 antigen, where the titres to C-1 sera were not significantly different to the titres obtained with the A-1, A-2, A-3, A-4 and B-1 sera.

Discussion

The present study appears to be first published investigation on cross-protection within all of the nine currently recognized serovars of H. paragallinarum of the Kume scheme. Our study has confirmed the widely accepted dogma that serogroups A, B, and C represent three distinct immunovars. With one exception, there was no significant cross-protection between Kume serogroups. The exception was that a vaccine based on the C-4 reference strain provided cross-protection against the challenge by the B-1 strain at a level that was not statistically different from the homologous protection. As Kume serovar C-4 has only been reported in Australia (Blackall et al., 1990b), this cross-protection is of little practical relevance, particularly as Kume serovar B-1/Page serovar B has never been reported in Australia-based serotyping studies (Thornton & Blackall, 1984; Blackall & Eaves, 1988; Eaves et al., 1989; Blackall et al., 1990a,b).

In terms of cross-protection within the two Kume serogroups that contain multiple serovars, we found differences between serogroup A and serogroup C. Within Kume serogroup A, serovars A-1, A-2, and A-3 showed a high level of cross-protection. When serovar A-4 is considered, there is a lower level of cross-protection. The vaccines based on A-2 and A-3 gave a protection level against the A-4 challenge that was significantly lower than the respective homologous challenge. Similarly, the vaccine based on A-4 gave a significantly lower protection against a challenge from the A-2 than from the homologous (A-4) challenge. To date, Kume serovar A-4 has only been reported in Australia, and indeed is the only form of Kume serogroup A present in that country (Eaves et al., 1989). Hence, in practical terms, this means that for most of the world inactivated infectious coryza vaccines need contain only one strain within the Kume serogroup A to provide good levels of protection against the recognized diversity within Kume serogroup A. Within Australia, any entry into the national poultry flock of an isolate of H. paragallinarum of Kume serovar A-2 may result in lowered vaccine efficacy in the field because all Australian-based infectious coryza vaccines are based on Kume serovar A-4.

Within Kume serovars C-1, C-2, C-3 and C-4, there was no vaccine that could give the same level of cross-protection for all four serovars. As an example, the serovar C-1 vaccine gave significantly lower protection against a C-2 challenge than it gave against the C-1, C-3 and C-4 challenges. Similarly, while the C-2-based vaccine gave equivalent protection for challenge from serovars C-1, C-2 and C-4, the protection against a C-3 challenge was significantly lower. This means that inactivated infectious coryza vaccines may need more that one Kume serogroup C strain depending upon the range of Kume serovar C types in the field. Our results should help vaccine manufacturers select the best possible combination of strains to be included in infectious coryza vaccines.

A rational selection of vaccine strains is only possible in those areas where there is a good knowledge of the distribution of the various Kume serovars—particularly the serovars within Kume serogroup C. Unfortunately, there have been few studies performed using the Kume serotyping scheme. Kume serovars A-1, B-1, and C-2 have been recognized in the USA (Kume et al., 1983; Eaves et al., 1989), A-1, A-2, B-1, and C-2 in Mexico and Germany (Kume et al., 1983; Eaves et al., 1989; Soriano et al., 2001), A-3 in Brazil (Kume et al., 1983), A-1, B-1, C-2, and C-3 in South Africa (Kume et al., 1983; Eaves et al., 1989), A-1 and C-1 in Japan (Kume et al., 1983; Eaves et al., 1989), C-3 in Zimbabwe (Bragg, 2002a), and A-4, C-2, and C-4 in Australia (Eaves et al., 1989; Blackall et al., 1990b). At the moment, most infectious coryza vaccines contain only a C-1 or C-2 strain and not both (Blackall, 1995). Our results suggest that, depending upon the range of Kume serovars in the field, vaccines with a single serovar C strain may not be optimal. A similar suggestion has been made by Bragg et al. (1996) who have suggested that increased prevalence of Kume serovar C-3 in South Africa poultry has occurred despite the extensive use of infectious coryza vaccines as the vaccines do not contain a serovar C-3 strain.

Our results for cross-protection within the Kume serogroup C match the previous reports that a C-2 vaccine protects against both a C-1 and a C-4 challenge (Kume et al., 1980; Blackall, 1991). Jacobs & van der Werf (2000) have reported that a commercial vaccine provided protection against challenge from several South African serovar C-3 field isolates. Unfortunately, it is difficult to compare this prior work with our current study as Jacobs & van der Werf (2000) provided no information on the Page serovar C in the vaccine they used.

Our study provides further evidence that Page serovar B/Kume serovar B-1 isolates represent a distinct immunovar—meaning that vaccines that lack a serovar B component are unlikely to provide protection. This observation of a lack of cross-protection between Page serovars A and C and Page serovar B has been reported previously (Jacobs et al., 1992). We have also reported previously that a bivalent (serovar A-1/serovar C-1) vaccine provided no protection against a Mexican serovar B-1 isolate (Fernández et al., 2003). While the Kume serotyping scheme recognizes only one serovar (B-1) within serogroup B, this should not be regarded as evidence of antigenic homogeneity. Few studies have attempted to examine the antigenic heterogeneity within Page serovar B/Kume serogroup B. The little information that is available tends to suggest that further serovars will be recognized in the serogroup B as only partial cross-protection has been reported within three serovar B strains (Yamaguchi et al., 1991). This evidence of antigenic diversity within Page serovar B/Kume serogroup B is further supported by the recent report that a standard commercial A, B, C trivalent infectious coryza vaccine provided a weaker level of protection against so-called ‘variant’ Page serovar B field isolates (Jacobs et al., 2003). Indeed, Jacobs et al. (2003) reported that the inclusion of one of the ‘variant’ Page serovar B isolates, resulting in a tetravalent vaccine, was necessary to achieve good protection.

We have used a vaccination/challenge format that is commonly used in the evaluation of infectious coryza vaccines (Kume et al., 1980; Blackall & Reid, 1987; Blackall, 1991, 1995). This widely accepted format involves the use of vaccines that are adjusted to a live cell count, an observation period for clinical signs that is limited just to day 2 or to days 2 to 7 post-challenge, and a necropsy at 7 days post-challenge. Our use of a live cell count for vaccine standardization means that there is a potential that the total cells per dose (i.e. live and dead cells) for each vaccine may have varied. A recent study that examined virulence of H. paragallinarum has proposed a longer period for the observation of clinical signs (Bragg, 2002b) but there is no evidence that such an extended observation period is of any relevance in vaccination/challenge trials. The nature of the work required that a single adjuvant had to be selected for use. Aluminium hydroxide is an adjuvant that has been shown repeatedly to be both safe and effective for infectious coryza vaccines (Blackall & Reid, 1987; Blackall, 1995).

The HI antibodies to H. paragallinarum are regarded as the main protective immune response against infectious coryza in chickens (Kume et al., 1984; Takagi et al., 1991). In our study, HI antibody titres were consistently highest when the vaccine strain and the serological antigen were the same Kume serovar (Table 3). Hence, in general, HI antibody titres were correlated with protection levels. There was one notable exception. The C-4 vaccinated chickens showed a level of protection against the B-1 challenge that was not statistically different from the protection conferred by the same vaccine against challenge by Kume serovars C-2 and C-4 (Table 2). However, the C-4 vaccinated chickens contained no detectable HI antibodies to serovar B-1 (Table 3). It is possible that other antigens than haemagglutinins could be involved in protection; for example, the capsule as suggested by others (Sawata et al., 1984).

There are important implications in our work for those laboratories that use HI tests to monitor vaccine efficacy in vaccinated flocks. We found that the use of a C-2-based haemagglutinin gave a very poor level of HI antibodies in C-1 vaccinated birds (Table 3). In contrast, the C-1 haemagglutinin gave titres that were not significantly different in any of the C-1, C-2, C-3 and C-4 vaccinated groups. This means that laboratories must select the serovar C antigen used in the HI test to monitor vaccines to match the serovar in the vaccine. As a general rule, our results suggest that a C-1 strain is better overall choice as a haemagglutinin antigen than a C-2 or a C-3 strain. We found no such marked variation in HI detection within Kume serogroup A.

In conclusion, the cross-protection and HI antibody titres of inactivated, aluminium-hydroxide-adsorbed, infectious coryza vaccines are dependent on the serovars included in the vaccines. Based on our results and a knowledge of the global Kume serovar distribution, most poultry regions of the world would be best served by an inactivated infectious coryza vaccine containing reference strains of serovars A-1, B-1, and C-2 of H. paragallinarum. In those regions where Kume serovar C-3 is present, other serovar combinations may be necessary. The available information on cross-protection within Kume serogroup B indicates that it may be difficult to predict levels of protection.

Translations of the abstract in French, German and Spanish are available on the Avian Pathology website.

Acknowledgments

The present research work forms part of the Ph.D. studies of the senior author (EVS) at the Universidad Nacional Autónoma de México (UNAM). The technical asistance of MVZ Enrique Q. Velásquez, Alberto J. Guadarrama and Germán C. Guadarrama is gratefully aknowledged.