Autoagglutination and the specificity of the indirect fluorescent antibody test applied to the identification of Taylorella equigenitalis

Research in Veterinary Science /990.49, / /7-/ /9

Autoagglutination and the specificity of the indirect fluorescent antibody test applied to the identification of Taylorella equigenitalis E. A. TER LAAK, C. M. F. WAGENAARS, Central Veterinary Institute, PO Box 65, 8200 AB Lelystad, The Netherlands

Because the first strains of Tay/orella equigenitalis isolated in the Netherlands autoagglutinated, identification was difficult. The source of carbon dioxide to create a carbon dioxide atmosphere for incubation influenced the emulsifiability of these strains. Strains were emulsifiable when cultivated in a carbon dioxide incubator (7 per cent carbon dioxide in air), but were autoagglutinable when cultivated in a candle jar, or in a jar with a carbon dioxide system or anaerobic syi,em without the palladium catalyst. When strains autoagglutinated, they were identified by the indirect fluorescent antibody test. Although strains of other bacterial species, in particular Pasteurella haemolytlca, also showed fluorescence, which was partly caused by autofluorescence, the indirect fluorescent antibody test appeared to be a reliable additional test for identifying T equigenitalis. CONTAGIOUS equine metritis (CEM) was diagnosed in the Netherlands for the first time in 1987 (Ter Laak et al 1989). The causative bacterium, Taylorella equigenitalis, was isolated from five mares of the Dutch saddle horse breed in three separate outbreaks in 1987 as well as from a fetus and placenta aborted at seven months by a mare of the same breed in early 1988 (Fontijne and others 1989). A fourth outbreak occurred in trotters in 1988, when T equigenitalis was isolated from 14 mares which had been covered by one stallion. The fifth outbreak occurred in the Haflinger breed in 1988, when T equigenitalis was isolated from a mare and from the stallion that had covered the mare. Bacterial strains suspected of being T equigenitalis are identified by agglutinating them with a specific hyperimmune serum. It has been reported that T equigenitalis autoagglutinates when grown on Eugon chocolate agar; thus, strains of this species must be cultured on Columbia chocolate agar before an agglutination test can be performed (Platt et al 1978, Saxegaard 1979). For primary isolation, samples have to be cultured in an atmosphere enriched with carbon dioxide. In this study the carbon dioxide levels were raised by using the GasPak anaerobic system (BBL, Becton Dickinson) without the palladium catalyst, a method recommended by Fleming and Tribe (1977) and by Shreeve (1978). Because the first strains, isolated from Dutch saddle horses and grown on Columbia agar, autoagglutinated strongly, identification was difficult. Two strains isolated from the Dutch saddle horse breed could be emulsified: one was isolated from the only mare of the third outbreak, the other from the aborted fetus of the second outbreak. The other strains isolated from this breed, however, as well as those from the Haflinger breed autoagglutinated. In contrast, all 10 strains isolated from the trotters could be emulsified.

The present authors found that the source of carbon dioxide was important for emulsifiability. When strains were cultured in a carbon dioxide incubator (7 per cent carbon dioxide in air), autoagglutination diminished greatly. Thus, strains grown in carbon dioxide atmospheres generated in four different ways were compared: a carbon dioxide incubator (7 per cent carbon dioxide in air), a candle jar, and two GasPak envelope systems (BBL. Becton Dickinson). Of the latter, the first was the GasPak anaerobic system without the palladium catalyst, and the second was the GasPak carbon dioxide system. Lyophilised Tequigenitalis strains were reconstituted with distilled water and inoculated on to Columbia blood agar base (Oxoid) chocolate agar, supplemented with sodium sulphite and L-cystine (Atherton 1978) and on to Eugon (BBL) chocolate agar (Fleming and Tribe 1977). The cultures were passaged once before testing. Plates were incubated at 37°C for 48 hours. Eighteen strains of T equigenitalis isolated in the Netherlands were examined. The strains were compared with the type strain isolated in England in 1977 and obtained in 1978 from Dr J. E. Shreeve, Central Veterinary Laboratory, Weybridge, as well as a strain (CVI number 81178) isolated in Ireland in 1977 and obtained in 1978 from Dr P. J. Timoney, Veterinary Research Laboratory, Dublin (Table I). Goat hyperimmune serum was prepared in the Central Veterinary Institute, Lelystad, with type strain NCTC 11184 of T equigenitalis; the titre was 1600 in the slide agglutination test (SAT) and MOO in the indirect fluorescent antibody test (FAT) causing a strong fluorescence. It was diluted I: I00 in phosphate buffered saline (PBS) pH 7·4 before use. Horse hyperimmune serum was obtained from Dr F. A. Rommel, Plum Island Animal Disease Center, Long Island, New York, USA, and was used as a reference serum for some strains. Saline, PBS pH 7'4, and distilled water were used to determine autoagglutination. All strains of T equigenitalis were emulsifiable when grown on Columbia agar in 7 per cent carbon dioxide in air. Several strains isolated from the Dutch saddle horses and the Haflingers were autoagglutinable when grown on Columbia agar in candle jars or in jars with the GasPak systems. When T equigenitalis strains were grown for more than 48 hours in 7 per cent carbon dioxide in air, even those grown on Columbia agar tended to autoagglutinate and to form mucoid colonies. When colonies were grown for 48 hours on Eugon agar in 7 per cent carbon dioxide in air, colonies of many strains could be emulsified; the troller strains, however, formed mucoid colonies, which could not be emulsified. Strains emulsified equally in saline, PBS pH 7'4,

117

E. A. Ter Look, C. M. F. Wagenaars

118

TABLE 1: Agglutination responses in slide agglutination tests using T equigenita/is strains grown in two different atmospheric systems

Origin (number of strainsl Dutch saddle horse (4)t Dutch saddle horse (11: Dutch saddle horse (1) § Dutch trotter (10) Dutch Haflinger (2) Irish thoroughbred (1) English thoroughbred (type strainl P A M

t

Agglutination responses after 4B h incubation in 7 per cent C02 in air GasPak· Columbia agar Eugon agar Columbia agar Eugon agar P P P

P, A·· P p. M

A P P

A P M

P

M

P

M

P P

PAM P P P

p. M

P

p. A, M

P, A

Positive agglutination Autoagglutination Mucoid culture Anaerobic system without palladium catalyst From first and second CEM outbreak From fetus. second CEM outbreak From mare. third CEM outbreak Results were not reproducible

or distilled water. The SAT results were not reproducible for some strains. The type strain varied in its emulsifiability, possibly because it had been passaged so many times (Table I). Only a few investigators have reported isolating strains of T equigenitalis that autoagglutinated. Flatscher and others (1984a) in Austria reported that many (a 'not small' number) of the 57 strains isolated in their study were viscous and produced a coarse, flaky suspension in saline after incubation in a candle jar; thus, it was impossible to test these strains for agglutination. Most strains, however, could be emulsified and identified as T'equigenitalis. This finding was mentioned in the English summary of the original article (Flat scher and others 1984a), but was omitted in the summary reported in the Veterinary Bulletin (Flat scher and others 1984b). Neill and others (1984) reported that some strains grown in 10 per cent carbon dioxide in air autoagglutinated. Although the type of carbon dioxide generator and agar medium were unimportant for emulsifiability of the Irish strain and several Dutch strains, the present authors concluded that these factors were important for other strains of T equigenitalis. Therefore, in order to avoid autoagglutination, it is recommended that a carbon dioxide incubator be used for culturing strains of T equigenitalis, that Columbia chocolate agar is preferable as has been recommended by other investigators, and that cultures should not be older than 48 hours when tested in the SAT. When strains autoagglutinated, they were identified by the indirect FAT. Although Tainturier and others (1981) recommended the direct FAT for identifying T'equigenitalis, the specificity of the test has not been investigated until now. In addition to the 20 T equigenitalis strains the following bacterial species were also examined in the FAT (number of strains examined are in brackets): Pasteurella haemolytica (19), P multocida (10), Actinobacillus pleuropneumoniae (two), A equuli (two), A suis (two), A Iignieresii (two), and Brucella abortus (four). These species were selected because they are culturally or biochemically similar or are genetically related to T equigenitalis (Taylor and others 1978). The strains were cultured on Columbia chocolate agar for 48

hours in two different atmospheres: a carbon dioxide incubator (7 per cent carbon dioxide in air) and the V GasPak anaerobic system without the palladium catalyst. Smears of cultures were fixed on to microscope slides, which were covered with T equigenitaiis goat antiserum I: 100; normal goat serum I; 100and PBS pH 7·4 were used as controls. After incubation for 30 minutes at 37°C slides were rinsed with PBS pH 7·4 and covered with rabbit-antigoat immunoglobulin conjugated to fluorescein isothiocyanate (Red Cross Blood Transfusion Service, Amsterdam) in a dilution of I :20. Slides were incubated and washed as above and examined by fluorescence microscopy. Strains of T equigenitalis fluoresced strongly with T equigenitalis antiserum but did not fluoresce with normal serum. Most strains of other bacterial species, in particular P haemolytica, also fluoresced with T'equigenitaiis antiserum. Fluorescence was generally less intense in organisms treated with normal serum (Table 2). Results were identical for the two atmospheric conditions used. FAT results in control slides did not differ when normal goat serum or PBS were used. Part of the fluorescence was caused by autofluorescence, which can be detected on untreated control slides. TABLE 2: Responses of T equlgenita/is and other bacterial species in the indirect FAT

Bacterial species

Intensity of fluorescence response to T equigeniralis Antiserum Normal serum

Tequigenira/is (201· + + + +t P beemotvtice (19) + + (111. + + + (BI Pmulrocida (10) + (41. + + (61 A pleuropneumoniae (2)+ A equu/i (2) (11. + + (11 (1) A suis (2) (1). + (1) A /ignieres;; (21 (1}. + (2). + (2) 8 abortus (41

+(111. + +(B) + + - (11. + (11 -

(2),

+

(2)

• Figures in brackets indicate number of strains t Range of fluorescence response: - (negative I to + + + + (strongly positivel

FA T for

T equigenitalis

When the hyperimmune serum was absorbed with a strain of P haemolytica that reacted strongly with unabsorbed serum in the FAT, fluorescence of this species diminished only slightly, because autofluorescence was constant. Fluorescence of T equigenitalis strains with absorbed serum was as strong as with unabsorbed serum. P haemolytica, which was isolated from a sample collected from the clitoral sinuses of a Dutch mare, is easily distinguished from T equigenitalis, because it reacts weakly in the oxidase test and can ferment glucose. T equigenitalis, however, is strongly oxidase positive. Moreover, the flora of the equine genital tract consists mostly of oxidase-negative bacteria. In conclusion, the present authors found that the indirect FAT is a reliable additional test for identifying T equigenitalis. Acknowledgements We thank Dr M. E. Mackintosh, Equine Research Station, Animal Health Trust, Newmarket, for confirming the diagnosis of the first cases of CEM in the Netherlands and for helpful discussions, and V. Thatcher for Englishlanguage editing.

119

FLATSCHER, J., AWAD-MASALMEH, M., SAGMEISTER, H. & WILLINGER, H. (l984a) Wiener Tierdrztliche Monatsschrift7I,85-90 FLATSCHER, J., AWAD-MASALMEH, M., SAGMEISTER, H. & WILLINGER, H. (l984b) Veterinary Bulletin 54, 542 FLEMING, M. P. & TRIBE, G. W. (1977) Veterinary Record 101,470-471 FONTIJNE, P., TER LAAK, E. A. & HARTMAN, E. G. (1989) Veterinary Record 125, 485 NEILL, S. D., O'BRIEN, J. J., McMURRAY, C. H. & BLANCHFLOWER, W. J. (1984) Equine Veterinary Journal 16, 430-434 PLATT, H., ATHERTON, J. G. & SIMPSON, D. J. (1978) Equine Veterinary Journal 10, 153-159 SAXEGAARD, F. (1979) Acta Veterinaria Scandinavica 20, 145-147 SHREEVE, J. E. (1978) Veterinary Record 102, 20 TAINTURIER, D. J., DELMAS,C. F.&DABERNAT, H. J.(l981) Journal of Ctinical Microbiology 14, 355-360 TAYLOR, C. E. D., ROSENTHAL, R. 0., BROWN, D. F. J., LAPAGE, S. P., HILL, L. R. & LEGROS, R. M. (1978) Equine Veterinary Journal 10, 136-144 TER LAAK, E. A., FENNEMA, G. & JAARTSVELD, F. H. J. (1989) Tijdschrift voor Diergeneeskunde 114, 189-201

! References ATHERTON, J. G. (1978) Veterinary Record 103, 432

Received September 28, /989 Accepted January 29, /990