Antigenic relationships of Moraxella bovis isolates recovered from outbreaks of infectious bovine keratoconjunctivitis in Argentina, Brazil, and Uruguay between 1983 and 2000

Abstract

Cross-reactivity indices (CRIs) of 28 isolates of Moraxella bovis recovered from outbreaks of infectious bovine keratoconjunctivitis in Argentina (A, 11 isolates), Brazil (B, 7), and Uruguay (U, 10) between 1983 and 2000 were estimated. Hyperimmune sera were produced in rabbits and antibody titres determined with each isolate. Isolates showing CRIs³ 70 were placed in the same group. Group I had 13 isolates (A, 1; B, 6; U, 6); group II had 6 isolates (A, 4; U, 2); groups III, IV, and V had 2 isolates each, recovered in Argentina; group VI had 2 isolates, from Uruguay; and group VII had 1 isolate, from Brazil. The CRIs³ 70 between vaccine strains and isolates recovered before and after 1990 were 58% and 42%, 50% and 50%, and 33% and 67% with vaccine strains 2419, 2358, and 2439, respectively. Isolate 273, from Uruguay, showed CRIs > 70 with 78% of the isolates and is recommended as the vaccine strain.

Infectious bovine keratoconjunctivitis (IBK), the most important ocular disease of cattle (1), was considered 1 of the 8 diseases to be studied by the regional cooperative project PROCISUR (Scientific Development Program) of Mercosur by Argentina, Brazil, and Uruguay, 3 of the 4 countries belonging to Mercosur.

Fimbriae, outer membrane proteins, and lipopolysaccharides, among other surface antigens, are important virulence factors of Moraxella bovis, the gram-negative bacterium responsible for IBK. Adhesins are primary virulence factors that mediate adherence between bacteria and cellular targets and are, therefore, chosen as antigens for vaccines. Their antigenic diversity and their implication in the immunologic control of the disease have been reported (2,3,4,5,6,7).

The frequently reported variable rate of protection by vaccines against IBK among herds of different regions has been related to the existence of several strains of M. bovis (8,9). Lepper and colleagues (10) suggested that immunity against the disease is specific for the serogroup to which the causal strain belongs. Gil-Turnes and Aleixo (11) quantified adhesins of several M. bovis isolates using monoclonal antibodies and found that isolates from different regions could be antigenically different, suggesting that those recovered from diseased animals should be continually monitored to detect emerging isolates not included in vaccines.

Although vaccines containing adhesins as antigens were introduced in the Mercosur countries in 1983 (7), outbreaks in herds routinely vaccinated have been frequently reported since 1990. The objective of this work was to estimate the antigenic relationships of 28 M. bovis isolates recovered from diseased animals in Argentina, Brazil, and Uruguay between 1983 and 2000.

The isolates were characterized biochemically (12) and by hemagglutinated ovine erythrocytes (13). Isolates from Argentina were sent by the Estación Experimental Agropecuaria de Balcarce (INTA), those from Uruguay were sent by Ladivet, and those from Brazil belonged to the collection of the Veterinary Faculty of the Universidade Federal de Pelotas. Table I shows their identification and year of isolation.

Table I.

The hemagglutination titres of each strain were determined with the use of suspensions of bacteria grown on blood agar for 18 h at 37°C, as previously described (11). Briefly, to 50 μL of 2-fold suspensions of the cultures in phosphate-buffered saline the same volume of 0.5% ovine erythrocytes was added. The least quantity of bacteria showing hemagglutination was considered 1 hemagglutinating unit. Serum against each isolate was raised in pairs of rabbits, as previously described (3).

To perform enzyme-linked immunosorbent assays (ELISAs), we sensitized each well of 96-well plates (Nalge Nunc International, New York, USA) overnight with 1 hemagglutinating unit of a single isolate in carbonate–bicarbonate buffer, pH 9.6. Then 50 mL of each serum was added in triplicate to the wells, and the plates were incubated for 1.5 h. Conjugated goat antirabbit immunoglobulin G peroxidase (Hybridoma Subisotyping Kit Mouse; Calbiochem, San Diego, California, USA) was then added, and the plates were incubated for another 1.5 h. Orthophenyldiamine and hydrogen peroxide were used to develop the reaction. Optical densities were read at 480 nm in an ELISA Multiskan MCC/340 spectrophotometer (Titertek; Huntsville, Alabama, USA) after 10 min in the dark. The hyperimmune serum was used at the dilution in which the absorbance with its homologous antigen was 0.5. Serum from nonimmunized rabbits was used as the negative control.

Bilateral cross-reactivity indices (CRIs) were estimated by the equation CRI = 100 Ör × r', where r is the quotient of the OD₄₈₀ of serum A with antigen B and that of serum A with antigen A, and r' is the quotient of the OD₄₈₀ of serum B with antigen A and that of serum B with antigen B (14). Isolates whose CRIs were³ 70 were considered to be of the same serogroup. This threshold is routinely used to estimate the cross-reactivity of foot and mouth disease virus (15) and has also been used to estimate the antigenic relationships of B. bronchiseptica antigens (16). It is assumed that CRIs < 70 indicate that the protection conferred by 1 strain against another would be less than 70%, which is unacceptable when immunization of populations is sought.

The 28 isolates were distributed into 7 serogroups (Table I). Group I, the most numerous, and the only group in which isolates had been recovered in all 3 countries, consisted of 13 isolates (46% of those tested): 1 from Argentina, 6 from Brazil, and 6 from Uruguay. Of these 13 isolates, 8 (6 from Brazil and 2 from Uruguay) had been recovered before 1990. Group II had 6 isolates (21% of those tested): 4 from Argentina (1 isolated before 1990) and 2 from Uruguay (both isolated before 1990). Groups III, IV, and V were formed by 2 isolates each from Argentina (7% of those tested in each group), group VI was formed by 2 isolates from Uruguay (7%), and group VII was formed by 1 isolate from Brazil (4%).

Forty-five percent of the CRIs between the isolates of group I and those of the other 6 groups were 0, whereas only 8% of the CRIs between the isolates of groups II to VII were zero. Three isolates of group I (23%) and 4 isolates of group II (33%) had CRIs >70 with isolates of the other groups (Table II).

Table II.

CRIs between the 3 strains of a vaccine extensively used in the region (7) and isolates recovered before and after 1990 were 58% and 42% with strain 2419, 50% and 50% with strain 2358, and 33% and 67% with strain 2439, respectively (Table II). These strains form 85%, 10%, and 5% of the antigenic mass of the vaccine, respectively. The changes in the ratios could be related to the concentrations of the strains in the vaccine, suggesting that the vaccine's extensive use could have induced some kind of immunologic selection.

The 11 Argentinean isolates belonged to 5 serogroups, denoting wide antigenic diversity, even though they were recovered from 4 herds. One isolate, which showed very low CRIs with other isolates from that country, belonged to group I, and 1 isolate recovered before and 3 recovered after 1990 were in group II, the group with more isolates from Argentina than from the other countries. The other isolates, 1 recovered before and 5 after 1990, were in 3 groups of 2 isolates each.

Of the 7 Brazilian isolates, 6 belonged to group I and had been recovered before 1990. Three isolates recovered from a single animal had CRIs between 83 and 101, indicating very close relationships, even though marked differences were detected with monoclonal antibodies (11). The Torres isolate, recovered in 2000 and placed in group VII, had CRIs³ 70 with 2 Uruguayan isolates.

The 10 Uruguayan isolates belonged to 3 groups: 6 to group I, 2 to group II, and 2 to group VI. They were recovered from 10 herds, which suggests less antigenic diversity than among the isolates from Argentina.

Some antigens seemed to be highly preserved in the region. Of the 16 isolates recovered after 1990, 5 belonging to group I and 3 to group II showed CRIs > 70 with isolates recovered before 1990. Isolate 273, recovered in Uruguay in 1996, cross-reacted with all the isolates tested: CRIs with the members of its own group varied between 106 and 174, and CRIs with the members of the other groups varied between 38 and 107; isolate 273 showed CRIs >70 with 21 (78%) of all the other isolates, including 2 vaccine strains, 5 of the 6 isolates in group II, 1 of the 2 isolates in each of groups III, IV, and V, and the 1 isolate in group 7 (Table II). This comprehensive cross-reactivity, resembling the "dominant" strains of foot and mouth disease virus (16), makes this isolate a strong candidate as a vaccine strain.

Three isolates recovered from the same animal (R1, R2, and R7) showed CRIs >70 among themselves, but their CRIs with isolates recovered from other animals were not always identical, thus showing that they were not the same strain. Variations in the antigenic profiles of isolates recovered in the same outbreak were also registered. Six isolates recovered from herd E2 in Argentina during an outbreak belonged to 3 serogroups, and 3 isolates recovered from herd Campomar, also in Argentina, belonged to 2 serogroups. Table II shows that several isolates had very low CRIs, denoting unrelated antigenic profiles. Gil-Turnes and Albuquerque (17) reported a similar finding in a Brazilian herd. The microbiota of bovine eyes contains antigenically different M. bovis isolates (18), and it is possible that the immune response induced by the isolates that are prevalent at the beginning of an outbreak cause the emergence of isolates with unrelated antigenic composition (17). On the other hand, considering that M. bovis may have more than 1 gene coding for pilin near an inversion region (19), it is possible that secretory antibodies of tears or ocular tissues induce genetic inversion (20) and, thus, the emergence of new antigens.

Our results demonstrate the need to continually survey the antigenic characteristics of M. bovis recovered from routinely vaccinated herds, in an effort to formulate vaccines with the antigens prevalent in a geographic region. They also stress the need to use autochthonous isolates in the efficiency testing of vaccines produced with antigens recovered in distant countries.