Rev. sci. tech. Off. int. Epiz., 1984, 3 (1), 151-163

Two serotypes of Corynebacterium pseudotuberculosis isolated from different animal species

« A . A . B A R A K A T * , S.A. SELIM**, A . A T E F * , M.S. SABER**, E.K. NAFIE***

and A . A . EL-EBEEDY *

Summary : Hyperimmune sera against four strains of Corynebacterium pseudotuberculosis isolated from different animals (sheep, goat, buffalo and cattle) were obtained by inoculating rabbits with strains in different forms (live, dead plus live or dead). Immunization with live or dead bac­teria followed by live strains gave similar results with regard to antibody titres and number of precipitin lines.

Typing 21 strains of C. pseudotuberculosis by specific precipitin reac­tions in agar gel was performed by using antigens extracted with sodium deoxycholate. The authors concluded that they were divisible into two serotypes, I and II.

The nitrate reduction test and animal pathogenicity tests (by inocu­lating guinea-pigs subcutaneously) demonstrated additional differences between the two types.

Type I was nitrate negative, and produced local and general absces­ses in guinea-pigs, with a prolonged course (15 days) ending fatally. It included all the strains originating from sheep and goats, but only 3 of 5 cattle strains.

Type II comprised all 7 isolates from buffaloes and 2 of the cattle strains. All strains were nitrate positive and did not produce abscesses in guinea-pigs, but they caused tissue necrosis at the site of inoculation, with a rapid course (not more than five days) ending in death.

I N T R O D U C T I O N

The aim of the present investigation was to study the biochemical, patho­logical and antigenic interrelationships between different strains of Corynebac-

* Animal Health Research Institute, Dokki, Giza, Egypt. ** Faculty of Veterinary Medicine, Cairo University.

*** Faculty of Medicine, Assiut University. (Published in Annual Reports of U.S. Agricultural Research Program. Public Law 480—Project

No. EG-ARS-86.)

— 1 5 2 —

FIG. 1 A cow affected with C. pseudotuberculosis showing nodules along

the lymphatics of the leg

F I G . 2 A buffalo infected with C. pseudotuberculosis showing typical lesion

of skin oedematous disease (notice the oedematous dewlap)

terium pseudotuberculosis (C. ovis) isolated from sheep and goats with caseous lymphadenitis, cattle with suppurative lesions (Fig. 1) and from buffaloes affec-

— 153 —

ted by “oedematous skin disease”, which is characterized by oedematous swel­lings in the dewlap, flank and other locations. The swelling may exceed the size of a water-melon. On cutting or puncturing, such swellings constantly release a serous bloody exudate (Fig. 2). Coalesced abscesses and eroded areas are usually seen, measuring about 30 cm in diameter, mostly on the shoulders or abdominal regions. The disease occurs sporadically but occasional outbreaks have been seen (Carpano, 1934 ; 8, 4, 6, 1, 2).

Infection of buffaloes with the bacterium is of great interest in Egypt and its epidemiological relationship to infection in sheep is the subject of some speculation, so there was good reason to search for criteria useful to differen­tiate the isolated causal organisms.

There is agreement that the different strains of Corynebacterium pseudo­tuberculosis cannot be differentiated reliably from each other by their cultural and biochemical properties, except perhaps by the nitrate reduction test. Bibers-tein et al. (5) classified sheep, goat, horse and cattle strains of C. pseudotuber­culosis into 2 biotypes; a nitrate-negative one infecting sheep and goats, and a nitrate-positive one infecting cattle and horses.

Serologically, Shigidi (7) did not record any antigenic variation between strains of C. pseudotuberculosis irrespective of their source. There has been no report of comparisons between strains of ovine and bovine origin, specially those isolated from buffaloes, by means of animal pathogenicity tests or by serological methods for detecting intra-type variation or subspecies. Accor­dingly, comparative cultural, morphological , biochemical as well as antigenic studies were performed on 21 strains isolated from different sources.

M A T E R I A L S A N D M E T H O D S

Bacteria.

The 21 strains of C. pseudotuberculosis comprised 4 from sheep and 5 from goats affected with caseous lymphadenitis; 5 from cows with suppurative lesions in different locations, and 7 from buffaloes with the so-called “oedematous skin disease”. Biochemical reactions were performed on all strains, using the standard methods described by Cowan and Steel (1966).

Pathogenicity tests.

A dose of 1 0 1 3 bacteria was inoculated subcutaneously inside the thigh of guinea-pigs, according to the procedure described by Barakat et al. (4).

Serological studies.

For studying antigenic interrelationships between the different strains, hyper-

— 154 —

immune sera were prepared against 4 strains representing the different species, i.e. one each from sheep, goat, cattle and buffalo.

Production of antisera.

Antisera against C. pseudotuberculosis were prepared by three different tech­niques: immunization of rabbits with formalin-killed organisms according to Shigidi (7); inoculation with formalin-killed cultures followed by live organisms by the method of Yanagawa et al. (9); immunization of rabbits with live cultu­res only.

Antigens were prepared by growing bacteria in Roux bottles containing brain-heart infusion agar. After 48 hours of incubation at 37°C, cells were har­vested from each bottle by adding a suitable amount of sterile saline. Harvested cells were transferred into screw-capped bottles, washed 3 times by saline and adjusted to the standard Macferland opacity tube No . 3. As the strains Nos. 6 (buffalo) and 19 (goat) were highly toxic and fatal to rabbits in the original concentration, the suspensions were diluted with sterile saline to 1/10 before inoculation.

All groups were inoculated with increasing doses (0.2, 0.4, 0.6, 0.8 and 1 ml) at weekly intervals. All rabbits were bled one week after the last injection.

Gel double diffusion test.

Purified agar (Difco) was dissolved in equal volumes of distilled water and barbital buffer (pH 8.2), warmed to make a 1% agar solution. Merthiolate in a concentration of 1:10 000 or sodium azide in a concentration of 0 .05% was added as a preservative. The melted agar was kept in a water bath at 60°C until needed.

Preparation of corynebacterial precipitating antigens was based on the pro­cedure described by Yanagawa et al. (9). Strains were cultured on brain-heart infusion agar in Roux bottles and incubated at 37°C for 48 hours , harvested by sterile saline, washed 3 times then resuspended in 2 ml of saline. To the suspension, 1 ml of 2 % sodium deoxycholate was added at 45°C for 3 hours, and used as treated antigen. Wells were filled with the test materials and the plates were incubated at room temperature for 72 hours.

Titration of hyperimmune sera.

All prepared immune sera were titrated by the agar gel diffusion method. Serial dilutions of serum prepared in saline from 1/2-1/64 were titrated against the homologous prepared antigen.

— 155 —

R E S U L T S

The results of comparative studies on growth, morphological, biochemical properties and pathogenicity of all investigated strains of ovine (sheep or goat) or bovine (buffalo or cattle) origin are shown in Table I.

T A B L E I

Main differential characters of ovine (sheep and goat) and bovine (buffalo and cattle) C. pseudotuberculosis strains

Character Ovine origin Bovine origin

Character sheep goat buffalo cattle strains

— Time of growth after 24 hours within 24 h within 24 h on blood agar

— Definite haemo­ after 24 hours within 24 h within 24 h some­lysis with wider zone times wide zone

— Average diamet­ 0.2-0.5 mm 0.5-2.3 mm 0.3-2 mm er of colony

— Microscopic coccobacilli larger coccobacilli coccobacilli appearance

— Nitrate test negative positive variable — Pathogenicity:

Inoculation site tense hot after much tenser after variable 24 h lesion 24 h lesion is circumscribed somewhat diffuse

Clinical course prolonged course rapid course variable of disease (about 15 days) (about 5 days)

ends with death ends with death P.M. changes local large abscess large swelling at variable

at site of inocula­ site containing tion containing macerated tissues creamy pus; and exudative abscesses in all material; no internal organs abscesses in

internal organs

Hyperimmunization.

Antibody titres and the number of precipitin lines for sera prepared in rab­bits by the 3 different schedules of immunization are shown in Table II .

It is clear from Table II that rabbits immunized with formalin-killed cul­tures of C. pseudotuberculosis gave immune sera having a titre of 1/32-1/64,

— 1 5 6 —

T A B L E I I

Antibody titration and number of precipitin lines of differently prepared hyperimmune sera against C. pseudotuberculosis

No. and origin of strain

dead ¡

titre

I

intigen

No. of lines

mmunizati

dead s

titre

ion by using

and live

No. of lines

live a

titre

ntigen

No. of lines

1 (sheep) 1/64 1-2 1/32 2-3 1/32 2-3 6 (buffalo) 1/32 1-2 1/16 2-3 1/16 2-3

14 (cow) 1/32 1-2 1/16 2-3 1/16 2-3 19 (goat) 1/16 1-2 1/32 2-3 1/32 2-3

but the number of precipitin lines with homologous antigens did not exceed one or two. While immunization by formalin-killed strains followed by live bacteria, or by live bacteria alone resulted in a rather lower antibody titre (1/16-1/32), there were more precipitin lines, never less than 2 or 3.

Classification of C. pseudotuberculosis by the agar-gel diffusion technique.

Four hyperimmune sera prepared in rabbits against strains 1 (sheep origin),

F I G . 3

Gel precipitation reaction of C. pseudotuberculosis antigens (peripheral wells) with antiserum against sheep strain

(central well) showing lines of identity

— 1 5 7

TABLE III

Immunodiffusion of antisera to four strains, Nos. 1 (sheep), 19 (goat), 6 (buffalo) and 14 (cow)

of C. pseudotuberculosis, against 21 strains of different origins

Antigen No.

1

Origin of

strain 1 (¡

sheep

sheep)

3

Antisen

19 (go

3

i against strain No.

at) 6 (buffalo)

0

14 (cow)

0 2 sheep 2 3 (1) (1) 3 sheep 2 2 0 (D 4 sheep 3 2 0 0

12 cow 2 2 0 0 13 cow 2 2 0 0 14 cow 2 2 0« 0 17 goat 2 2 0 0 18 goat 2 2 1(F) (1) 19 goat 2 3 F F 20 goat 2 2 0 0 21 goat 2 2 F F

5 buffalo 2 3 3 3 6 buffalo 2 2 3 2 7 buffalo 2 3 2 3 8 buffalo 2 1 2 2 9 buffalo 3 3 2 2

10 buffalo 3 4 3 3 11 buffalo 2 3 2 2 15 cow 2 2 1 2 16 cow 2 2 1 2

19 (goat), 6 (buffalo) and 14 (cattle origin) were tested against the 21 bacterial strains by the precipitation reaction in agar-gel.

Results are given in Table III , which shows that strains of C. pseudotuber­culosis from different animals can be divided into two serological groups, designated I and II. Reactions between homologous antisera and antigens gave distinct lines, while between those belonging to different types there were usually no line or only faint ones.

It is seen from the table that all strains of ovine origin (sheep and goat) and 3 strains of cattle origin belonged to type I, and only reacted with hyper­immune sera prepared against strains 1 (sheep) and 19 (goat), while no line or only faint lines developed with hyperimmune sera against buffalo and cattle strains (Fig. 3).

— 1 5 8 —

F I G . 4

Gel precipitation reaction of C. pseudotuberculosis of sheep (upper 3 wells) and buffalo (lower 3 wells) antigens with antiserum

against sheep strain showing more than one line of identity with the homologous antigens only

The other strains, seven of buffalo origin and two cattle strains, gave 2-3 distinct lines with all 4 hyperimmune sera, and were grouped as type II. No strains of ovine origin were included in type II (Fig. 4).

D I S C U S S I O N

A bovine skin disease, reported in Egypt by many authors, affects mainly buffaloes, and is characterized by large swellings accompanied by rise in tem­perature. The disease may last for a long time, resulting in emaciation and serious economic losses for the farmers. This disease has been regarded as infec­tious by Egyptian veterinary authorities since 1960 and was named “oedematous skin disease". The cause of the disease was obscure until reports by Barakat and Eid (3), and Anon. (1, 2) of the isolation of Corynebacterium pseudotuber­culosis (C. ovis) from infected buffaloes, and experimental reproduction of the disease by intradermal inoculation of the isolated organisms into buffalo calves. It is known that C. pseudotuberculosis mainly causes caseous lym­phadenitis in sheep and goats, while in cattle it causes suppurative conditions.

Members of the genus Corynebacterium have been grouped by morpho-

— 159 —

logical and biochemical characteristics. In the present investigation, morpho­logical, biochemical, biological and immunological techniques were applied to 21 strains of different origins (4 from sheep, 5 from goats, 5 from cows and 7 from buffaloes).

The authors came to the conclusion that the strains could not be differ­entiated morphologically or by using established biochemical reactions, except the nitrate reduction test, the results of which agreed with serological tests.

It was noticed that all strains from sheep and goats were nitrate negative, while all those isolated from buffaloes were nitrate positive. In the case of strains isolated from cows, 3 of 5 strains were nitrate negative and 2 strains were nitrate positive. It is apparent from these results that two biotypes exist: a nitrate negative one infecting sheep and goats, and a nitrate positive one infecting buffaloes.

In cows no host correlation was observed, as both types (nitrate positive and negative) were isolated. The results of nitrate reduction agreed with those of laboratory animal inoculation. When guinea-pigs were inoculated sub-cutaneously with the 21 strains, two different pathological pictures were obser­ved. The first group was composed of all strains of ovine origin in addition to 3 cow strains (Nos. 13, 14 and 15; all nitrate negative). Animals injected inside the thigh developed tense, hot circumscribed lesions at the site of inocula­tion for 24 hours . The disease in guinea-pigs was prolonged and ended with death after 15 days. Local and general abscesses were found in internal organs post mor tem. The picture was completely different for the second group of organisms (all nitrate positive) which contained all the buffalo strains and 2 cow strains (Nos. 15 and 16). At the site of inoculation, much tenser and more dif­fuse lesions developed within 24 hours . The course of the disease was shorter and ended with death after 5 days. Large swellings were present at the site of inoculation, containing macerated tissue and exudate without any pus formation either locally or in internal organs.

Apart from biochemical reactions and pathogenicity testing in guinea-pigs, all 21 strains were studied serologically and could be differentiated into two serotypes by means of the agar-gel precipitation technique. Shigidi (7) used the same technique and specific hyperimmune sera prepared in rabbits inoculated with formalin-killed antigens of C. pseudotuberculosis, but was unable to detect antigenic differences between strains isolated from sheep, goats, horses and donkeys.

However, Yanagawa et al. (9) used hyperimmune sera prepared by inoculat­ing rabbits with formalin-killed cultures followed by live ones, and they dif­ferentiated 78 strains of C. renale into 3 serotypes.

To avoid any influence of the form of inoculated strains on the specificity of antibodies produced, we prepared hyperimmune sera by immunizing rab­bits with formalin-killed cultures according to Shigidi (7), or killed cultures

— 1 6 0 —

followed by live ones according to Yanagawa et al. (9), or live cultures alone. Tests on hyperimmune serum prepared by inoculating killed C. ovis against homologous antigens showed that the titre of antibodies was relatively high (1/32 to 1/64), exceeding those of serum prepared by the other 2 schedules. However, their specificity was lower because they gave only 1 or 2 thick lines of precipitation, while those sera prepared by inoculating killed plus live cul­tures, or live cultures alone gave at least 2 or 3 sharp lines of precipitation with the homologous antigens.

By comparing the results of the three techniques of immunization, it might be thought that multiplication of bacteria in vivo is important for stimulating antibody formation against the present antigens. Moreover, it is known that short courses of immunization produce more specific antibodies.

Typing 21 strains of C. pseudotuberculosis by precipitation reactions in gel was performed with antigen extracted by sodium deoxycholate. It was found that they were divisible into 2 types. 12 strains (4 from sheep, 5 from goats and 3 from cows) gave 2 or 3 lines with hyperimmune serum prepared against strains Nos. 1 (sheep) and 19 (goat), and no line or only faint lines with serum prepared against the other 2 strains (buffalo and cow strains); these comprised one group or type I. The other 9 strains which included all 7 buffalo strains and 2 cow strains (Nos. 15 and 16) reacted with the 4 types of immune sera, giving 2 or 3 lines of precipitation and grouped serologically as type II . These results indicate that strains of type II may share some of type I antigen frac­tions in addition to their own antigens.

Depending upon agreement between nitrate reduction, pathogenicity and serological testing, the authors concluded that the 21 strains were divisible into 2 separate types. Similar results were obtained by Biberstein et al. (5) who

T A B L E I V

The main criteria for differentiating C. pseudotuberculosis into two types

No. and origin of strain

Nitrate test Pathogenicity for guinea-pigs Serotype

Sheep strains negative Prolonged course (15 days) Goat strains negative with local and internal abscesses. I Cow strains negative

(Nos. 12, 13, 14)

Buffalo strains

Cow strains (Nos. 15 and 16)

positive

positive

Short course (5 days) with macerated tissues at site of inoculation. No internal abscesses.

II

- 161 —

divided C. pseudotuberculosis into 2 biotypes depending on host predilection and nitrate reducing activity.

All strains of ovine origin were included in type I, while all buffalo strains were included in serotype II, the cattle strains being divided between the 2 sero­types. This information corresponds to field observations that organisms of serotype I infect mainly sheep and goats, which can be considered natural hosts of that serotype, which causes a specific disease (caseous lymphadenitis), while members of serotype II infect mainly buffaloes causing a characteristic disease (oedematous skin disease) and they can be considered as the natural host of serotype II.

Since both serotypes can be isolated from cattle, they can be considered as a non-specific host for bo th types, the lesions consisting of abscesses in various locations.

This proposal for host specificity needs experimental infection of different animals for further confirmation, and this has been done by experimentally infecting buffaloes, cows and sheep with typical strains of serotype I (ovine origin) and serotype II (buffalo origin).

When serotype I is injected intradermally, it produces typical caseous lymphadenitis in sheep, while in cattle and buffalo the result is an inflammatory swelling at site of inoculation, which rapidly heals. On the other hand, injec­tion of serotype II produces the typical “oedematous skin disease" in buffaloes, but only local inflammation at site of injection in sheep and cattle.

* * *

DEUX SÉROTYPES DE CORYNEBACTERIUM PSEUDOTUBERCULOSIS ISO­LÉS CHEZ DIFFÉRENTES ESPÈCES ANIMALES - A.A. Barakat, S.A. Selim, A. Atef, M.S. Saber, E.K. Nafie et A.A. El-Ebeedy.

Résumé : Des sérums hyperimmuns contre quatre souches de Cory­nebacterium pseudotuberculosis isolées chez un mouton, une chèvre, un buffle et un bovin ont été obtenus en inoculant les souches à des lapins au moyen de trois techniques différentes : utilisation de bactéries vivan­tes, de bactéries inactivées puis de bactéries vivantes, de bactéries inacti­vées. Avec les deux premières techniques, l'immunisation a donné des résultats semblables quant aux titres des anticorps et au nombre des lignes de précipitation.

Le typage de 21 souches de C. pseudotuberculosis par la réaction spé­cifique de précipitation en gélose a été réalisé en utilisant des antigènes extraits au désoxycholate de sodium. Les auteurs ont ainsi distingué deux sérotypes, I et II, parmi ces souches.

L'épreuve de réduction par le nitrate et l'épreuve du pouvoir patho­gène in vivo, par l'inoculation sous-cutanée à des cobayes, ont révélé d'au­tres différences entre ces deux types.

— 1 6 2 —

Le type I était négatif à l'épreuve de réduction par le nitrate ; il a produit des abcès locaux et généralisés chez les cobayes, avec une évolu­tion lente (15 jours) et une issue fatale. Ce type correspondait à toutes les souches provenant d'ovins et de caprins, mais seulement à trois des cinq souches bovines.

Le type II correspondait aux sept souches isolées chez des buffles et à deux des souches bovines. Toutes les souches étaient positives à l'épreuve de réduction par le nitrate ; elles n'ont pas produit d'abcès chez les cobayes mais provoqué une nécrose des tissus au point d'inoculation, avec une évolution rapide (cinq jours au plus) et une issue fatale.

DOS SEROTIPOS DE CORYNEBACTERIUM PSEUDOTUBERCULOSIS AISLA­DOS EN DISTINTAS ESPECIES ANIMALES. - A.A. Barakat, S.A. Selim, A. Atef, M.S. Saber, E.K. Nafie y A.A. El-Ebeedy.

Resumen : Se consiguieron sueros hiperinmunes contra cuatro cepas de Corynebacterium pseudotuberculosis aisladas en una oveja, una cabra, un búfalo y un bovino, inoculando las cepas en conejos mediante tres técnicas distintas : uso de bacterias vivas, de bacterias inactivadas y des­pués de bacterias vivas, de bacterias inactivadas. Con las dos primeras técnicas, la inmunización dió resultados semejantes en cuanto a los títu­los de los anticuerpos y al número de las líneas de precipitación.

Se realizó la tipificación de 21 cepas de C. pseudotuberculosis mediante la reacción específica de precipitación en gelosa usando antí-genos extraídos con desoxicolato de sodio. Los autores distinguieron así dos serotipos, I y II, entre las referidas cepas.

La prueba de reducción con el nitrato y la prueba del poder patogé­nico in vivo, mediante la inoculación subcutánea en cobayos, revelaron otras diferencias entre ambos tipos.

El tipo I era negativo a la prueba de reducción con el nitrato ; pro­dujo abcesos locales y generalizados en los cobayos, con un curso lento (15 días) y un fatal desenlace. Este tipo correspondía a todas las cepas procedentes de ovinos y caprinos, aunque únicamente a tres de las cinco cepas bovinas.

El tipo II correspondía a las siete cepas aisladas en búfalos y a dos de las cepas bovinas. Todas las cepas eran positivas a la prueba de reduc­ción con el nitrato ; no produjeron abcesos en los cobayos, aunque pro­vocaron una necrosis de los tejidos en el punto de inoculación, con un curso rápido (cinco días como más) y un fatal desenlace.

* * *

— 1 6 3 —

REFERENCES

1. ANON. (1979). — Oedematous skin disease. In: First Annual Report U.S. Agric. Res. Program. Ed. by Anim. Hlth. Res. Inst., Cairo, April 1979. Grant No. FG-EG-189, Project No. EG-ARS-86.

2. ANON. (1980). — Oedematous skin disease. In: Second Annual Report U.S. Agric. Res. Program. Ed. by Anim. Hlth. Res. Inst., Cairo, April 1980. Grant No. FG-EG-189, Project No. EG-ARS-86.

3. BARAKAT A . A . and EID F . I . A . (1971). — Bovine lymphangitis. Monthly Report, Vet. Pive, ARE, Jan. 1-10.

4. BARAKAT A . A . , SABER S. and AWAD H.H. (1973). — The average standard sub­cutaneous C. ovis dose to produce either morbidity or lethality in guinea-pigs. 11th Ann. Arab Vet. Med. Cong., Cairo.

5. BIBERSTEIN E . L . , KNIGHT H .D . and JANG S. (1971). — Two biotypes of Coryne­bacterium pseudotuberculosis. Vet. Rec., 89, 691-692.

6. FOUAD K . , SALEH M . , KHAMIS K . , SHOUMAN T . and FAHMY L. (1972). — Further investigation on the so-called “oedematous skin disease” of buffaloes and cattle. 10th Ann. Arab Vet. Congr., Cairo. J. Egyptian Vet. Med. Ass., 34, 154-170.

7. SHIGIDI M.T .A. (1975). — Antigenic relationship of various isolates of Coryne­bacterium pseudotuberculosis. Bull. Epiz. Dis. Afr., 22 (3), 263-269.

8. SOLIMAN K.N. , Eid F.I.A. and SAYOUR E . M . (1963). — Ulcerative lymphangitis in buffaloes in Egypt. 4th Ann. Arab Vet. Congr., Cairo, 283-295.

9. YANAGAWA T., BASRI H. and OTSUKI K. (1967). — Three types of Corynebacter­ium renale classified by precipitin reactions in gels. Jap. J. vet. Res., 15, 111-120.