Pathology of Campylobacter jejuni Abortion in Sheep

Transcription

1 Vet. Pathol (1987) Pathology of Campylobacter jejuni Abortion in Sheep 0. R. HEDSTROM, R. J. So", E. D. LASSEN, B. D. HULTGREN, R. 0. CRISMAN, B. B. SMITH, AND S. P. SNYDER Veterinary Diagnostic Laboratory, Veterinary Teaching Hospital, Veterinary Basic Sciences, and College of Veterinary Medicine, Oregon State University, Corvallis, OR Abstract. Campylobacter jejuni was inoculated intravenously into pregnant ewes on gestation days 1 14 and 123 to reproduce ovine abortion. All ewes aborted days post-inoculation. High numbers of C. jejuni were isolated from ewe tissues (caruncle, bile, cecal feces), fetal tissues, and placenta. C. jejuni colonies were identified in caruncles and placenta by light microscopy and immunoperoxidase techniques. Histologically, inoculated ewes had a severe purulent endometritis with vasculitis. Placentas from inoculated ewes and field cases showed necrosis and purulent inflammation; however, placentas from inoculated ewes had large numbers of bacterial colonies compared to few bacteria found in field cases. Histologically, only one fetus from the inoculated ewes showed lesions (purulent bronchopneumonia), whereas all fetuses from field cases had a distinct bronchopneumonia, and one fetus showed multifocal hepatic necrosis. These results suggest that C. jejuni (serotypes Penner 1 and Lior 2) is an important abortifacient organism for sheep. Campylobacteriosis is a highly contagious and economically significant disease in sheep and is most often caused by Campylobacter fetus subspecies fetus (previously named C. fetus var. intestinalis) It is characterized by abortion (third trimester), stillbirths, premature births, weak lambs, and occasional ewe deaths due to metriti~.'~,~~ Grossly, aborted or stillborn fetuses may have no lesions or they may show subcutaneous serosanguineous edema, serosanguineous body fluids, discrete necrotic liver lesions, or bronchopneumonia. Placental cotyledons are enlarged, yellowish, and covered with blood-tinged exudate, while the intercotyledonary stroma is edematous. Microscopically, placental lesions, which are predominant in hilar zones, consist of septa1 necrosis, arteriolitis, leukocyte infiltration, and accumulation of high numbers of bacteria within lacunae, chorionic trophoblasts, and endothelial cells. In humans, Campylobacter jejuni is recognized to be a common cause of acute dia~-rhea,~j~ and is associated with abortion and neonatal sepsis.33 Campylobacter organisms are small, curved, highly motile, noncapsulated, microaerophilic, gram-negative bacilli. Campylobacter jejuni is easily separated from other Campylobacter species because it grows at 42 C, is resistant to cephalothin, and is inhibited by nalidixic acid. The genus Campylobacter has undergone extensive taxonomic changes in recent year^.^.'^.^^ According to previous classification schemes, C. jejuni appears to correspond to Marsh and Firehammer's Vibriofetus serotype 125 and Berg's et al. Vibrio fetus group C.3 Apparently, these classifications and another reports implicate C. jejuni as an abortifacient agent in sheep. There are no reports documenting the experi- 419 mental or natural pathologic lesions of ovine abortion caused by C. jejuni, and the strains of C. jejuni involved have not been well characterized by currently available techniques. The objective of this study was to experimentally reproduce ovine abortion with heat stable 1 (Penner) and heat labile 2 (Lior) serotype strains of C. jejuni. This organism was obtained from a field case of ovine abortion. Case Histories A fetus, placenta, and maternal serum sample from southem Oregon (flock 1) and three fetuses from three different ewes and one placenta from southern Washington (flock 2) were received by the Veterinary Diagnostic Laboratory, College of Veterinary Medicine, Oregon State University, Corvallis, Oregon. At the time of submission, four out of 200 breeding ewes from flock 1 had aborted term fetuses, and by the conclusion of lambing, approximately 15 ewes had aborted. This flock had been vaccinated 2 weeks prior to breeding with a commercial Campylobacter vaccine. Flock 2 had 40 out of 100 breeding adult ewes abort term fetuses by the time of submission, and by the conclusion of the lambing season, approximately 80 ewes had aborted. Ewes in flock 2 frequently had retained placentas and had not been vaccinated. Campylobacter jejuni was isolated from fetal tissues received from both submissions and was serotyped by the Centers for Disease Control, Atlanta, Georgia.28 Both isolates were determined to be Penner serotype 1 and Lior serotype 2. Materials and Methods The Campylobacter jejuni used in this experimental study was isolated from placenta from flock 1. Nine yearling, crossbred, virgin, nonvaccinated ewes selected from a dis-

2 420 Hedstrom et al. Table 1. Ewe serum progesterone levels measured at 48 hour intervals or at day of abortion following experimental inoculation with Campylobacter jejuni. Post-inoculation Num- Abortion Day Day ber o* Post-inoculation progesterone levels (ngml) Day 2 Day 4 Day 6 Day I Day 8 Day 10 Day o 4 Control Control * Day 0 represents inoculation day. ease-free flock were used. Ewe feces were cultured for C. jejuni using Campylobacter agar with incorporation of Cefoperazone, Vancomycin, and Amphoterian B (CVA media, PML Microbiologicals, Tualatin, OR) and appropriate isolation techniques.21 Ewe serums were also serologically evaluated for C. jejuni, I I Leptospira sp., and toxoplasmosis (Toxo-Screen DA, Hynson, Westcott, Dunning, Baltimore, MD) 1 month prior to and at breeding. Virgin, yearling rams were tested for C. jejuni with fecal cultures and serologically for C. jejuni, leptospirosis, and toxoplasmosis similar to ewes. In ewes or rams, serum antibody was not detected for leptospirosis, toxoplasmosis, or C. jejuni, and C. jejuni was not recovered from feces. Estrus was synchronized with two subcutaneous injections of 2 mg of fenprostalene (Bovilene, Syntex Animal Health, Des Moines, IA) given 10 days apart, and then bred at estrus following the second injection. At post-breeding day 2 1, ewes were confirmed pregnant if serum progesterone levels were greater than 1 ng/ml (Diagnostic Products Corp., Los Angeles, CA). Pregnant ewes were randomly divided into two groups; group I contained four ewes and group I1 contained five ewes. The ewes were pastured in strict isolation facilities at Oregon State University until inoculated. At post-breeding day 114, three pregnant ewes in group I were inoculated via the jugular vein with 5 ml of lo8 colony forming unitdm1 of C. jejuni in phosphate buffered saline. At post-breeding day 123, four pregnant ewes from group I1 were inoculated via the jugular vein with 5 ml of lo9 colony forming unitdm1 of C. jejuni in phosphate buffered saline. One control ewe from each group was inoculated via the jugular vein with 5 ml of phosphate buffered saline. The C. jejuni inoculum was prepared from a 48-hour brucella broth culture washed 3 times in phosphate buffered saline and concentrated by centrifugation. Groups I and I1 and control ewes were housed separately on concrete floor stalls in strict isolation facilities. All ewes were examined twice daily, and rectal temperatures were measured. Serum progesterone was measured at the time of inoculation and at 48-hour intervals thereafter until necropsy. Ewes and fetuses were necropsied immediately following abortion, and control ewes were necropsied concurrently with the last aborting ewe in each group. Maternal and fetal tissues were cultured on CVA agar and incubated at 42 C under microaerophilic conditions (BBL Campy Pak, Becton Dickinson, Cockeysville, MD). Maternal tissues cultured included: caruncle, liver, bile, spleen, lung, kidney, and cecal feces. Fetal tissues cultured included: liver, lung, spleen, stomach contents, and placental cotyledons. C. jejuni colonies were identified by colony morphology, growth at 42 C, and resistance to cephalothin. Colony numbers were roughly quantitated by estimation of plate quadrant growth. Impression smears prepared from placental cotyledons were stained with Gimenez and examined for Chlamydia organisms. Immunoglobulin G levels, using single radial immunodiffusion plate^,^.^^ were determined from pooled peritoneal and thoracic fluid for all fetuses. Multiple 1 -cm-thick sections of ewe uterus, caruncle, liver, and lung, and fetal perio- and cotyledon placenta, lung, liver, kidney, thymus, spleen, brain, adrenal, thyroid, heart, and biceps femoris muscles were fixed in 10% neutral buffered formalin. Tissues were dehydrated in graded alcohols, cleared, embedded in paraffin, cut at 6 pm, and stained with hematoxylin and eosin (HE) for light microscopy. Localization of C. jejuni in ewe caruncle and fetal cotyledon was performed by immunoperoxidase techniques,13 using a commercially available biotinylated secondary antibody (goat anti-rabbit) and an avidin-biotin-peroxidase complex (Vectastain, ABC kit 4001, Vector Laboratories, Burlingame, CA). Hyperimmune serum against a washed, whole cell antigen of C. jejuni (Penner 1 and Lior 2) was obtained from immunized New Zealand white rabbits as described previously.2o This serum was diluted 1 : 500 and was used as the primary antibody. Tissues used for immunohistochemistry were fixed in 10% formalin for 24 hours and paraffin embedded. Controls included: omission of primary antibody, use of normal rabbit serum, and use of noninoculated ewe and fetal tissues. Fetuses obtained from field cases were necropsied, and fetal tissues and fetal peritoneal and thoracic fluid were processed as described above for the experimentally inoculated ewes for light microscopy, Gimenez-stained cotyledon smears, and immunoglobulin G levels. Fetal lung, liver, stomach content, and cotyledons were cultured on 5% sheep blood agar,

3 McConkey s agar, and thioglycolate enrichment broth. Leptospira sp. and toxoplasmosis serology tests were done on flock 1 ewe serum similar to inoculated ewes.32 C. jejuni Abortion in Sheep 42 1 Results Abortion occurred in all ewes (1 00%) 7 to 12 days post-inoculation of C. jejuni. Ewes rectal temperatures became elevated ( C) 12 to 24 hours prior to abortion, but beyond this, they exhibited no other clinical or preparturient signs. Serum progesterone levels were normal 48 hours prior to abortion, but then declined rapidly at abortion (Table 1). No abortions or clinical signs were observed in control ewes, and serum progesterone remained elevated (Table 1). At necropsy, gross lesions were present in caruncles of all C. jejuni inoculated ewes and two ewes, one from each group, had retained placentas. Uterine walls were thickened with edema fluid, which was most prominent in caruncle and pericaruncle stromal regions. Individual caruncles were swollen, and on transverse section, hemorrhagic streaks extended deep into the caruncles. Mucoid-serosanguineous and fibrinous exudate covered the caruncles superficial surfaces (Fig. 1). Gross lesions were not seen in inoculated or control ewes lungs, hearts, gastrointestinal tracts, mesenteric lymph nodes, kidneys, mammary glands, adrenal glands, or ovaries. Occasionally, control and inoculated ewe livers had a white fibrous subcapsular tracts, and associated bile ducts were thickened. These changes were suggestive of previous liver fluke migration. Placentas from inoculated ewes showed mottled, yellowtan swollen cotyledons with no evidence of surface exudate. Intercotyledon chorioallantoic and amnionic membranes were clear with no evidence of edema or exudate (Fig. 2). Aborted fetuses usually had undergone in utero autolysis, except for two, which were fresh. Meconium-stained fetuses had variable amounts of serosanguineous subcutaneous edema, serosanguineous peritoneal and thoracic fluids, and variable degrees of hemoglobin-stained viscera. Gross lesions were not seen in visceral organs. Fetuses received from the field showed postmortem autolysis, minimal serosanguineous peritoneal and thoracic fluid, and hemoglobinstained viscera; gross lesions were not found in visceral organs or placentas. Histologically, caruncles showed severe widespread lesions which involved maternal septa, smooth muscle capsule, and adjacent endometrium. Multifocally, maternal septa contained necrotic foci, leukocyte infiltration, purulent arteriolitis, and fibrin thrombi (Figs. 3, 4). Multifocally, uterine spaces, usually adjacent to affected septa, contained necrotic cellular debris, neutrophils, fibrinous exudate, and some sloughed necrotic Fig. 1. Caruncle from ewe 5 inoculated intravenously with C. jejuni and necropsied following abortion (day 7 postinoculation). Caruncle swollen. Hemorrhage (arrows) and mucoid-serosanguineous and fibrinous exudate present on surface. Fig. 2. Placenta from ewe 5, Fig. 1. Cotyledon swollen and mottled. Intercotyledon areas normal. maternal epithelium. Endometrial stroma was edematous, and endometrial glands were filled with inflammatory exudate; however, intercaruncular endometrial epithelium usually remained intact (Fig. 3). Histologcally, placental villi from inoculated ewes were necrotic, and numerous large bacterial colonies were found within trophoblasts and adjacent stroma (Fig. 5). Intracellular C. jejuni distended trophoblastic cytoplasm and surrounded or displaced nuclei. Ne-

4

5 C. jejuni Abortion in Sheep 423 Table 2. Campylobacter jejuni recovered from experimentally inoculated ewe and fetal tissues and fluids.* Ewe Fetus Animal Number Feces Uterus Bile Spleen Liver Lung Placenta Contents Stomach Lung Liver Spleen Kidney Group It Group 11* $ I l * Grading: + 1 to +4 indicates increasing numbers of colony forming units on direct plating I t Group I, ewes (1-3) were inoculated with 5 ml of lo8 colony forming units/ml of C. jejuni, and Control (4) was inoculated with 5 ml sterile phosphate buffered saline on gestation day 114. * Group 11, ewes (5-8) were inoculated with 5 ml of lo9 colony forming units of C. jejuni, and Control (9) was inoculated with 5 ml sterile phosphate buffered saline on gestation day 123. crotic cellular debris was present, and the stroma contained a few scattered leukocytes. Postmortem autolysis was also present. Placentas from field cases had large areas of necrosis in chorionic villi, varying degrees of arteriolitis, abundant necrotic debris, and high numbers of leukocytes within the stroma. Postmortem autolysis was also present. However, in contrast to the inoculated ewes, hematoxylin and eosin (HE)-stained sections did not show bacterial colonies within stroma or trophoblastic epithelium. Fetuses from inoculated ewes did not have histologic lesions in liver, kidney, spleen, striated muscle, heart, adrenal, thymus, thyroid, brain, and lung, except for some fetuses which had meconium within lung airways, and one fetus (group I) which had a purulent bronchopneumonia. All fetuses received from the field had a purulent bronchopneumonia and meconium within airways, and one fetal liver had multifocal areas of coagulative necrosis. Field fetuses did not show histologic lesions in heart, striated muscle, kidney, adrenal, thyroid, or brain. In inoculated ewes, immunoperoxidase techniques demonstrated extremely high numbers of C. jejuni within placental trophoblasts lining most chorionic villi (Fig. 5). Subadjacent stroma, vascular endothelium, and lumens also contained C. jejuni. Placentas obtained from field cases had fewer numbers of immunoperoxidase-stained C. jejuni when compared to inoculated ewes, but bacteria were found to be located similarly as in inoculated ewes. Small colonies of C. jejuni were found scattered throughout caruncles of inoculated ewes. Bacteria were primarily found within maternal epithelium or within leukocytes in uterine spaces (Fig. 6). Septa and vascular endothelium adjacent to affected uterine spaces also contained varying amounts of C. jejuni. High numbers of C. jejuni were consistently isolated from inoculated ewes' feces, uteri, and bile (Table 2). Placentas from inoculated ewes and field cases usually contained higher numbers of C. jejuni than fetal visceral organs and stomach contents. From flock 1, C. jejuni was isolated only from placenta, whereas in flock 2, it was recovered from placenta, stomach contents, lung, and liver from all three fetuses. No other significant bacterial organisms were isolated from placentas or visceral organs from field cases. Gimenez-stained placental smears, obtained from field cases, did not demonstrate Chlamydia organisms. Immunoglobulin levels were less than 20 mg/dl in pooled peritoneal and thoracic fluids obtained from all fetuses. Following abortion, Leptospira sp. and toxoplasmosis serum antibody titers were negative in all inoculated ewes and the ewe from flock 1. Ewe serum antibody titers to C. jejuni were not determined. Discussion This experimental study confirms that our field isolate, heat stable 1 (Penner)28 and heat labile 2 (Lior)2s strain of C. jejuni, is an abortifacient organism in sheep. t Fig. 6. C. jejuni localized in uterine caruncle with immunoperoxidase techniques, ewe intravenously inoculated. C, jejuni colonies located primarily within uterine luminal space (arrow) and sloughed maternal epithelium (arrowhead). Bar = mm.

6 424 Hedstrom et al. Eighty percent of ewes in flock 2 aborted, which also suggests that this strain is highly abortifacient. However, only three fetuses and one placenta were examined from flock 2, and other causes of abortion cannot be entirely excluded. In addition, we cannot exclude the nonclassified anaerobic flagellated bacterium recently described by Kirkbride et a1.18 as an additional abortifacient organism in the field cases; however, our experimental results clearly fulfill Koch's postulates with C. jejuni. Penner serotype 1 was the only serotype of C. jejuni found in a recent study of C. jejuni isolates recovered from the reproductive tract of but, in this study, the heat labile serotypes were not determined. The heat stable serotype of C. jejuni isolated from an aborted caprine fetus was Penner 1.' Penner 1 was the second to third most frequent serotype encountered in the United States when C. jejuni was serotyped from hospitalized human patients,28 or when all C. jejuni isolates were studied.29 Gross and histologic findings in our study are similar to reported placentome pathology in sheep intrave- nously inoculated with Vibrio fetus. l4 Compared with this latter study, we did not find intercotyledonary edema in placentas from either natural or experimental disease or uterine discharge. Histologic lesions were not stratified in hilar zones of caruncles, but we found more exudation of leukocytes into uterine spaces than found in placentomes infected with Vibrio fetus. We studied ewe caruncles and placentas after ewes had aborted; consequently, our lesions may be more severe. Placentas had varying degrees of postmortem autolysis, making it more difficult to determine precise gross and histologic lesions than in the previous study. Furthermore, nomenclature of the genus Campylobacter from older literature is very confusing and the exact species of Campylobacter strains previously reported are uncertain, since organisms were not classified according to presently accepted riter ria.^^,^',^^ Consequently, direct comparison of our results with previous literature may not be appropriate. The differences seen in fetal and placental lesions, when natural and experimental diseases were compared, might be due to overwhelming bacterial infection in the experimental condition. Fetuses may have died before they had the opportunity to mount a significant inflammatory response in placenta and lungs, as seen in the naturally infected fetuses. Fetuses of days of gestational age are capable of producing an inflammatory re- ~ponse.~l The intravenous route of exposure used in this experiment represents a severe challenge and may not be the most appropriate route to determine the pathogenicity of this organism, because the natural route of exposure is thought to be ~ ral.'~.~~ In sheep, serum progesterone levels fell sharply hours prior to abortion caused by Chlamydia psittaci, which is analogous to the findings in our study. Our observations support Jensen's proposed pathogenesis for V. fetus, l4 i.e., that following V. fetus septicemia, maternal septa develop arteriolitis and necrosis, allowing V. fetus to traverse necrotic vessels into uterochorionic spaces. Organisms may also pass through maternal terminal septa into lacuna hematomas. Erythrophagocytic, chorioallantoic, and placentome trophoblast infection occurs, followed by replication and formation of large intracellular bacterial colonies. Trophoblast entry and intracellular replication of high numbers of B. abortus is important for spread of this organism within placentomes and to fe- Concurrently, chemotaxis and exudation of high numbers of leukocytes occurs. Trophoblast necrosis and spread of organisms to chorionic villi and vasculature then permit vascular dissemination of the organisms to the fetus. Fetal death occurs because of fetal sepsis, toxemia, and hypoxia. Intravenous administration of C. fetus toxins can cause abortion in sheep and other animal^.^' The pathogenesis of these abortions was hypothesized to be caused by an allergic inflammatory reaction to the Campylobacter toxin; the microvasculature is involved, and this reaction is intensified by pregnancy. In human campylobacteriosis, the pathogenic properties of C. jejuni may be due to in- vasiveness, enterotoxins, or cytot~xins;~~.~~ however, the role of specific toxins is not clear because some authors have not been able to detect enterotoxins among different strains of C. jejuni Reisolating high numbers of C. jejuni from ewe and fetal tissues implies that this disease has potential to be a serious zoonosis. Campylobacter jejuni is one of the most common agents causing diarrhea in hu- man~.~~.~~ Human abortion and perinatal sepsis can also be caused by C. jejuni infection.33 Women who abort Campylobacter-infected fetuses are fecal culture positive but often do not have diarrhea,33 which is analogous to what we observed in the sheep in this study. Also, the high number of C. jejuni within ewe feces, bile, and uterus and fetal tissues could contaminate the environment and serve as a source of infection for other ewes, potentially resulting in an abortion storm. Campylobacter jejuni can survive in bile for 2 months maintained at 37 C and for at least 3 weeks in feces and other fluids maintained at 4 C.4 Isolating high numbers of C. jejuni from experimentally inoculated ewe bile, but low numbers from ewe liver and parenchymal organs, plus the lack of significant histologic lesions in ewe parenchymal organs, suggests that C. jejuni can replicate within bile or gallbladder. Furthermore, this event may occur separately from replication within the uterus and may be

7 an important means of spreading this disease to other susceptible ewes. Lambs fed C. jejuni continued to shed this organism 17 days post-feeding.lo Campylobacter fetus subspecies fetus can be isolated from sheep gallbladder and intestines up to 41 days post-intravenous inoculation,6 and the liver and gallbladder are the primary sites of Campylobacter infection in carrier sheep. Our results emphasize the importance of using CVA agar and incubation at 42 C as a part of routine diagnostic procedures when attempting to determine the cause of ovine abortions. Without these selective procedures, C. jejuni could be overgrown with secondary bacteria from heavily contaminated fetuses and placentas, and these techniques can be used to separate C. jejuni from C. fetus subspecies fetus. Further research is needed to determine the prevalence and pathogenesis of C. jejuni abortion in sheep, to determine which strains of C. jejuni are abortifacient in sheep, and to develop a more efficacious vaccine. Acknowledgements The authors acknowledge Ms. Charlotte Patton of Centers for Disease Control, Atlanta, Georgia, for serotyping the C. jejuni used in this study. The technical assistance of P. Allision, J. G. Armstrong, D. Holthofer, and H. Allen is acknowledged. The authors are grateful to Dr. Don Bailey and Ms. Judy Harris for specimen submissions. References Anderson KL, Hamoud MM, Urbance JW, Rhoades HE, Bryner JH: Isolation of Campylobacterjejuni from an aborted caprine fetus. J Am Vet Med Assoc 183:90-92, 1983 Anderson TD, Meador VP, Cheville NF Pathogenesis of placentitis in the goat inoculated with Brucella abortus. Vet Pathol23: , 1986 Berg RL, Jutila W, Firehammer BD: A revised classification of V. fetus. Am J Vet Res 32: 11-20, 1971 Blaser MJ, Hardesty HL, Powers B, Wang WL Survival of Campylobacter fetus subsp. jejuni in biological milieus. J Clin Microbiol 11: , 1980 Bosc M J, Djiane J, Nicolle A, Rodolakis A Evaluation of progesterone, estrogens and placental lactogen in pregnant ewes following experimental infection with an abortive agent Chlamydia psittaci Var-Ovis. Thenogen- OlOgy 16~ , 1981 Bryner JH, Estes PC, Foley JW, O Berry PA: Infectivity of three Vibrio fetus biotypes for gallbladder and intestines of cattle, sheep, rabbits, guinea pigs and mice. Am J Vet Res , 1971 Bryner JH, O Berry PA, Estes PC, Foley JW: Studies of vibrios from gallbladder of market sheep and cattle. J Vet Res 33: , 1972 Diker KS, Istanbulluoglu E: Ovine abortion associated with Campylobacter jejuni. Vet Rec 118:307, 1986 Fahey JL, McKelvey EM: Quantitative determination C. jejuni Aboi tion in Sheep 425 of serum immunoglobulins in antibody-agar plates. J Immunol94:84-90, Firehammer BD, Myers LL Campylobacterfetus subsp. jejuni: its possible significance in enteric disease of calves and lambs. Am J Vet Res 42: , Firehammer BD, Border MM: Bulk growth procedures and a button agglutination test for Campylobacter. Am J Vet Res , Garcia MM, Eaglesome MD, Rigby C Campylobacter s importance in veterinary medicine. Vet Bull , Hsu SM, Rake L, Fanger H: The use of avidin-biotinperoxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 29: , Jensen R, Miller VA, Molello JA: Placental pathology of sheep with vibriosis. Am J Vet Res 22: , Jensen R, Swift BL: Diseases of Sheep, 2nd ed., pp Lea and Febiger, Philadelphia, Jubb KVF, Kennedy PC, Palmer N: Pathology of Domestic Animals, 3rd ed., vol. 3, pp Academic Press, London, Karmali MA, Skirrow MB: Taxonomy of the genus Campylobacter. In: Campylobacter Infection in Man and Animals, ed. Butzler JP, pp CRC Press, Boca Raton, FL, Kirkbride CA, Gates CE, Collins JE Abortion in sheep caused by a nonclassified, anaerobic, flagellated bacterium. Am J Vet Res , Klipstein FA, Engert RF, Short H, Schenk EA: Pathogenic properties of Campylobacter jejuni: assay and correlation with clinical manifestations. Infect Immun 50: 43-49, Lior H: Serotyping of Campylobacter jejuni and E. coli by slide agglutination based on heat-labile antigenic factors. In: Campylobacter Infections in Man and Animals, ed. Butzler JP, pp CRC Press, Boca Raton, FL, Luechtefeld NW, Wang W-LL, Blaser MJ, Reller LB: Campylobacter fetus subsp. jejuni: background and laboratory diagnosis. Lab Med 12(8): , Mancini G, Carbonara AO, Heremans JF: Immunochemical quantitation of antigens by single radial immunodiffusion. Immunochem , Mandal BK, DeMol P, Butzler JP Clinical aspects of Campylobacter infection in humans. In: Campylobacter Infection in Man and Animals, ed. Butzler JP, pp CRC Press, Boca Raton, FL, Manninen KI, Prescott JF, Dohoo IR: Pathogenicity of Campylobacter jejuni isolates from animals and humans. Infect Immun 38:46-52, Marsh H, Firehammer BD: Serological relationships of twenty-three ovine and three bovine strains of Vibrio fetus. Am J Vet Res , Miller VA, Jensen R, Gilroy J J: Bacteremia in pregnant sheep following oral administration of Vibrio fetus. Am J Vet Res 20: , 1959

8 426 Hedstrom et al. 27 Osborne J C Pathologic responses in animals after Vibriofetus toxin shock. Am J Vet Res 26: , Patton CM, Barrett TJ, Morris G K Comparison of the Penner and Lior methods for serotyping Campylobacter spp. J Clin Microbiol22(4): , Penner JL, Hennessy JN, Goodbody MM: Development of a scheme for serotyping Campylobacter jejuni. In: Campylobacter Epidemiology Pathogenesis and Biochemistry, ed. Newel1 DG, pp MTP Press Ltd, Lancaster, UK, Prescott JF, Munroe DL: Campylobacter jejuni enteritis in man and domestic animals. J Am Vet Med Assoc 181( 12): , Sawyer MM, Moe JB, Osburn BI: Ontogeny of immunity and leukocytes in the ovine fetus and elevation of immunoglobulins related to congenital infection. Am J Vet Res 39: , Schmitz JA, Sonn RJ: A study of ewe abortions in Oregon. Proceedings 26th Annual Meeting Amer Assoc Vet Lab Diagnosticians, pp , Simor AE, Karmali MA, Jadavji T, Roscoe M: Abortion and perinatal sepsis associated with Campylobacter infection. Rev Infec Dis 8(3): , Waldstrom T, Baloda SB, Drovacek R, Faris A, Bongston S, Walder M: Swedish isolates of Campylobacter jejunilcoli do not produce cytotonic or cytotoxic enterotoxins. Lancet ik911, Warner DP, Bryner JH, Beran GW Epidemiologic study of campylobacteriosis in Iowa cattle and the possible role of unpasteurized milk as a vehicle of infection. Am J Vet Res 47: , 1986 Request reprints from Dr. 0. R. Hedstrom, Veterinary Diagnostic Laboratory, College of Veterinary Medicine, Oregon State University, Corvallis, OR (USA).