Parasitol Res (2000) 86: 1±7 Ó Springer-Verlag 2000

ORIGINAL PAPER

Martin Peters á Frank Wagner á Gereon Schares Canine neosporosis: clinical and pathological ®ndings and ®rst isolation of Neospora caninum in Germany

Received: 8 August 1999 / Accepted: 30 August 1999

Abstract Neosporosis was diagnosed in an 11-week-old species. In cattle, N. caninum has been found to be as- puppy of the breed Kleiner MuÈ nsterlaÈ nder with sociated with abortions. Clinical neosporosis in dogs is progressive hindlimb paresis. Pathohistological and most commonly observed in congenitally infected pup- immunohistological examinations revealed a dissemi- pies in the ®rst 6 months of life (Ruehlmann et al. 1995). nated infection with Neospora caninum. Parasitic stages Characteristically these animals show a progressively were demonstrated in the brain, spinal cord, retina, ascending paresis, usually starting at the pelvic limbs, muscles, thymus, heart, liver, kidney, stomach, adrenal caused by severe polymyositis, polyradiculitis, and dis- gland, and skin. Immunohistochemistry investigations seminated meningoencephalomyelitis. Generalized in- were carried out using polyclonal rabbit antisera devel- fection with multiple organ involvement in puppies has oped against N. caninum tachyzoites and the recombi- been reported (Dubey et al. 1988a; Barber et al. 1996). nant bradyzoite-speci®c antigen BAG-5 of Toxoplasma Some puppies su€ered and eventually died of myo- gondii, which is known to cross-react with N. caninum carditis caused by the parasite (Dubey et al. 1988a; Odin bradyzoites. BAG-5 antibodies recognized tissue cysts and Dubey 1993; Barber et al. 1996; WeissenboÈ ck et al. within the CNS and some protozoan stages that were 1997). Besides multifocal CNS involvement and not surrounded by a visible cyst wall. All parasite clus- polymyositis, pneumonia (Greig et al. 1995) or derma- ters in the retina and some in muscle tissue stained titis (Dubey et al. 1995; Fritz et al. 1997; Perl et al. 1998; positively with the BAG-5 antiserum. N. caninum was Poli et al. 1998) may be the predominant clinical feature isolated in cell culture and mice inoculated with brain in adult dogs. and spinal cord of the puppy. The new isolate is the ®rst Dogs are both de®nitive hosts (McAllister et al. reported in Germany and is designated NC-GER1. 1998) and intermediate hosts of N. caninum. Parasite stages in the intermediate host are the rapidly repli- cating tachyzoites and slowly replicating bradyzoites residing within tissue cysts. Intracellular multiplication Introduction of tachyzoites results in cell death with a subsequent in¯ammatory response (Dubey and Lindsay 1996). Neospora caninum is an apicomplexan parasite struc- In¯ammatory reactions do not seem to be caused by turally similar to Toxoplasma gondii. Natural infections intact tissue cysts as long as they do not rupture have been described worldwide in several mammalian (Mayhew et al. 1991). In dogs, tissue cysts have been described only within the CNS, including the retina (Dubey et al. 1990; Dubey and Lindsay 1996). Bradyzoites of N. caninum can be di€erentiated from M. Peters (&) tachyzoites by immunohistological examination Institut fuÈ r Pathologie, TieraÈ rztliche Hochschule Hannover, (McAllister et al. 1996; Fuchs et al. 1998; Schares BuÈ nteweg 17, D-30559 Hannover, Germany et al. 1999). e-mail: [email protected]; Fax: +49-511-9538675 This paper reports on the clinical signs and patho- F. Wagner morphological lesions observed in a puppy infected with Klinik fuÈ r kleine Haustiere, TieraÈ rztliche Hochschule Hannover, Bischofsholer Damm 15, D-30173 Hannover, Germany N. caninum. The distribution of the parasite in several tissues, including the eye; the immunohistochemical G. Schares di€erentiation between bradyzoites and tachyzoites in Bundesforschungsanstalt fuÈ r Viruskrankheiten der Tiere, Institut fuÈ r epidemiologische Diagnostik, Seestrasse 55, these tissues; and the ®rst isolation of the parasite in D-16868 Wusterhausen, Germany Germany are described. 2

antibody test (IFAT) and immunoblotting as described by Schares Materials and methods et al. (1998). Sera were examined for antibodies to T. gondii by immunoblotting (Schares et al. 1999). For PCR the CSF was cen- History trifuged; the pellet, digested with proteinase K; and the DNA, isolated using a DNA extraction kit (NucleoSpin C + T, Two of eight puppies of the fourth litter of a Kleiner MuÈ nster- Macherey & Nagel) according to the instructions of the manufac- laÈ nder bitch showed progressive hindlimb paresis. One male pup turer. Furthermore, DNA was extracted from tissues and examined developed clinical signs at 1 week of age and died at the age of for N. caninum DNA by PCR using primers Np6 and Np21 7 weeks. After postmortem examination, neosporosis was sus- (Yamage et al. 1996). pected. The second female pup showed sti€ness of the left hindlimb at 6 weeks of age and was treated with trimethoprim and sulfa- diazine (Tribrissen 20, Essex) for 12 days without clinical im- Isolation of N. caninum provement. During this period a serum sample was taken for serology. After a week without therapy the pup had become re- For isolation of protozoan parasites, brain and spinal cord tissues cumbent. At the age of 11 weeks, physical, neurological, and were treated separately. All solutions (bu€er, medium) used for ophthalmological examinations were done. Routine blood counts isolation were supplemented with penicillin (1,000 IU/ml) and and blood biochemistry were performed. A second serum sample streptomycin (1,000 lg/ml) to prevent bacterial growth. Tissues was taken from the pup and the bitch was bled for serological were homogenized in 100 ml PBS, ®ltered through sterile gauze, examination. A cerebrospinal ¯uid (CSF) sample was taken and centrifuged (800 g). The brain sediment and spinal cord sedi- atlantooccipitally for di€erential cell counts, serology, and poly- ment were resuspended in 100 and 50 ml RPMI supplemented with merase chain reaction (PCR). The pup was euthanized. 0.5% (w/v) trypsin (Sigma, Deisenhofen), respectively, and then incubated at 37 °C. After 30 min the homogenate was centrifuged three times (800 g). Pathological and immunohistochemical examination The resulting sediments were aliquoted and inoculated both onto 1-day-old Vero-cell monolayers (25 cm2) and intraperitone- Immediately after euthanasia a complete postmortem examination ally into mice. Cell cultures were examined microscopically every was carried out. Half of the longitudinally divided brain as well as day for the presence of parasite stages. Five and three Vero-cell the major parts of the spinal cord were removed for parasite iso- cultures were infected using aliquots of brain and spinal cord ho- lation. Native samples of the brain, spinal cord, heart, lung, liver, mogenate, respectively. Further aliquots of brain homogenate were spleen, kidney, and skeletal muscles were removed for Neospora inoculated into four BALB/c J-nu mice (nude mice; 3 months old), caninum PCR. Another set of tissue samples was ®xed in 4% buf- into two AKR mice (3 months old), and into four BALB/c mice fered formol saline and processed using routine techniques. Sec- (1 month old). Aliquots of spinal cord homogenate were inoculated tions were cut at 5 lm and stained with hematoxylin and eosin into three BALB/c J-nu mice (2±3 months old), into two AKR mice (H&E) for microscope examination. Sections were deparanized in (3 months old), and into two BALB/c mice (1 month old). At 9 h xylene for immunohistochemical examinations. After incubation in after inoculation, AKR and BALB/c mice were treated with a isopropanol and 96% (v/v) ethanol, endogenous peroxidase was subcutaneous injection of 2.5 mg prednisolone-21-acetate (Al- quenched with 0.5% (v/v) hydrogen peroxide in methanol for brecht, Aulendorf). Mice were examined daily for the development 30 min. After further rehydration in 75% (v/v) and 50% (v/v) of disease. Sick mice were euthanized, and their peritoneal exudate ethanol the slides were rinsed in phosphate-bu€ered saline (PBS) was examined microscopically for the presence of parasites. Posi- and then incubated in 0.05% (v/v) pronase E (Merck, Darmstadt) tive peritoneal exudate was washed once in a 20-fold volume of for 20 min at 37 °C. RPMI and inoculated onto a 1-day-old Vero-cell monolayer After another rinse in TRIS-PBS, sections were blocked in (25 cm2). TRIS-PBS containing 20% (v/v) normal goat serum and were in- cubated (45 min, 37 °C) with a rabbit antiserum (1:500 in PBS) developed against N. caninum (NC-1) tachyzoites (Schares et al. 1997). This antiserum detects bradyzoites and tachyzoites of Results N. caninum. To obtain exclusive immunolabeling of bradyzoites of N. caninum we used (45 min, 37 °C) a polyclonal rabbit antiserum (1:14,000 in PBS) developed against a BAG-5 fusion protein of Clinical ®ndings Toxoplasma gondii, which does not cross-react with tachyzoites. This antiserum does strongly cross-react with N. caninum brady- Clinical examination of the pup revealed paresis and zoites (McAllister et al. 1996). Furthermore, a mouse monoclonal hyperextension of the left hind leg along with a marked antibody (74.1.8.) against T. gondii BAG-5 antigen developed by atrophy and hardening of the thigh muscles of both Weiss et al. (1992) was used on a few representative sections of the CNS and skeletal muscle (1:25 in PBS). Antibodies were detected hindlimbs. The left tarsal joint could not be ¯exed using commercial kits (Vectastain Elite ABC Kit, Vector Labora- mechanically. The dog was not capable of getting up tories) as described by the manufacturer. without help but could stand after receiving passive For color development, 3,3¢-diaminobenzidine tetrahydro- support. Urination and defecation were under control. chloride (0.05 mg/ml in 0.05 M TRIS-PBS, pH 7.6) and 0.03% (v/ v) hydroxide peroxide were used. The enzyme reaction was stopped The proprioceptive positioning reaction was normal in after 5 min by thorough rinsing in tap water for 10 min. The sec- the forelimbs and in the right hindlimb. A moderately tions were counterstained with hematoxylin for 15 s, rinsed in tap delayed reaction was noted in the left hindlimb. All water for 15 min, dehydrated in increasing concentrations of eth- other postural reactions were within normal limits. The anol, transferred in xylene, and mounted. Substitutions of the primary antibodies by normal rabbit antiserum or normal mouse left hindlimb showed a depressed quadriceps, cranial serum were carried out as negative control procedures. tibial, and ¯exor re¯ex. No abnormality in the tested spinal re¯exes was observed for the other limbs. Deep- pain sensation was normal for all limbs, as was the Serological examinations and PCR perception of super®cial pain con®rmed by cutaneous Serum and CSF from the puppy and serum from the bitch were re¯ex. Radiographs of the vertebral column and examined for antibodies to N. caninum by an indirect ¯ourescence extremities were without pathological ®ndings. The 3 sensorium seemed to be undisturbed. De®cits of the the CNS but were more pronounced in the gray matter cranial nerves were not observed. than in the white matter. There was mild lympho- Examination of the eyes revealed a positive menace plasmacellular chorioiditis. In the brain stem and spinal re¯ex and a positive direct and indirect pupillary light cord, demyelination and neuronal degeneration with re¯ex in both eyes. Schirmer tear tests 1 and 2, appla- neuronophagia were present. The leptomeninx showed a nation tonometry (Tonopen XL II, Mentor), and slit- moderate degree of in®ltration, with mononuclear cells lamp biomicroscopy (Kowa SL-2, Kowa) showed no and a few neutrophils often being observed adjacent to pathology. By indirect ophthalmoscopy (Heine 2000 small vessels. Neuritis associated with peri- and intra- Ophthalmoskop, Heine; 30-D lens, Zeiss), several hy- fascicular lymphocytes, plasma cells, a few neutrophils, pore¯ective maculae were detected in the tapetal fundus and eosinophils was observed in the optic and trigeminal of both eyes (Fig. 1). These lesions were mainly associ- nerves as well as in the spinal nerve roots. Many pro- ated with the retinal vasculature. tozoan cysts were found throughout the brain, in the The hematological pro®le revealed 9.8 ´ 103 leuko- cranial nerves, and in the spinal cord. They measured cytes/ll, an 11% eosinophil content (reference range 0± 19.4 ‹ 5.2 ´ 16.2 ‹ 4.7 lm (n ˆ 40) and mainly had 6%), a 31% lymphocyte content (reference range 13± thin cyst walls (£1 lm). Immunohistochemically they 30%), an 8% monocyte content (reference range 0±4%), stained positively with the polyclonal N. caninum anti- and a 49% content of segmented neutrophils (reference serum, the polyclonal antiserum against the bradyzoite- range 55±75%). The hematocrit and erythrocyte indices speci®c antigen BAG-5 (Fig. 3a), and the mouse were within normal limits. The serum creatine kinase monoclonal BAG-5 antibody 74.1.8. Tachyzoites were activity was slightly elevated at 381 IU/l (reference range mainly associated with in¯ammatory lesions. A few £90 IU/l). Values recorded for alanine aminotransferase nonencysted parasites reacted with the polyclonal BAG- (ALT), alkaline phosphatase (AP), glutamic dehydro- 5 antiserum and the monoclonal BAG-5 antibody. The genase (GLDH), total bilirubin, and urea were within BAG-5-positive zoites were found scattered within small normal limits. numbers of randomly distinct foci (Fig. 3b). The CSF displayed marked pleocytosis of 397 cells/ The left eye showed moderate choroiditis of the cili- mm3, consisting of 70% neutrophils, 29% lymphocytes, ary body associated with in®ltration of lymphocytes, and 1% macrophages. The total protein concentration in histiocytes, and eosinophils. A few clusters of parasites the CSF was 214 g/dl, and the glucose value was 58 mg/dl. without visible cyst walls were detected in the inner nuclear layer, outer plexiform layer, and outer nuclear layer of the retina in the absence of in¯ammatory reac- Serological and PCR ®ndings tions. These stages reacted with the polyclonal BAG-5 antiserum (Fig. 4). Around some of the retinal parasites Both, the serum sample obtained from the pup during there was a rarefaction of the outer plexiform layer chemotherapy and the second sample taken at 11 days along with a fusion of the inner and outer nuclear layers. after the end of chemotherapy revealed a Neospora Cavitation was observed between the retinal pigment caninum IFAT titer of 1:1,600, whereas that of the bitch epithelium and the rods and cones. was 1:50. Immunoblotting showed the pup and the bitch In¯ammatory changes were detected in all skeletal to be seropositive for N. caninum. Antibodies against muscles (masseter, supraspinatus, triceps brachii, biceps N. caninum were detected in CSF by IFAT (titer 1:200) femoris, semitendinosus, gastrocnemius, extraocular and immunoblotting. In contrast, antibodies to Toxo- muscles, and diaphragm) that were examined histologi- plasma gondii were not demonstrable by immunoblot- cally. These consisted of multifocal in®ltration with ting in serum of the puppy or the bitch. PCR revealed N. lymphocytes, histiocytes, and plasma cells, combined caninum DNA in the CSF and all tissues examined. with a moderate degree of interstitial ®brosis and acute myonecrosis (Fig. 5). In some foci the predominant in- ¯ammatory cells were neutrophils and eosinophils. Pathomorphological and immunohistological ®ndings Lesions and ®brosis were most severe in the muscles of the left pelvic limb. Many muscle ®bers contained par- On necropsy the only gross lesions consisted of a gray to asite clusters. A few of these stained positively with the yellow streaky discoloration of the atrophic thigh and polyclonal BAG-5 antiserum (Fig. 6). Parasites were gastrocnemius muscles, especially in the left pelvic limb. also detected in muscle ®bers by the mouse monoclonal Although the esophagus was not dilated, the thoracic antibody 74.1.8. Individual zoites stained with the pol- section contained some food. yclonal BAG-5 antiserum were not observed. Cyst wall Histologically, in the CNS of the pup a multifocal structures were not visible. granulomatous, partially suppurative encephalomyelitis The left sciatic nerve showed a mild in®ltration with was present (Fig. 2) in association with focal necrosis, eosinophils, whereas the right sciatic nerve was not in- gliosis, vascular proliferation, vasculitis, and perivascu- ¯amed; parasites were not detected. There was reactive lar cung, predominantly with lymphocytes, histiocytes, hyperplasia in both the super®cial cervical lymph nodes and plasma cells and occasionally with neutrophils and and the popliteal lymph nodes. The cytoplasm of some eosinophils. The lesions were disseminated throughout macrophages in the lymph nodes stained immunohisto- 4

chemically positive for N. caninum antigen. The tongue with hepatocellular necrosis. The spleen was hyperemic. and esophagus showed a moderate degree of myositis Parasites were not detected. In the kidneys, a few soli- associated with multiple protozoan clusters in myo®bers. tary parasites were identi®ed immunohistochemically in In¯ammatory changes in the heart consisted of mild tubular epithelial cells, and a cluster of protozoan stages multifocal subendocardial and myocardial in®ltrations was observed in a glomerulus in the absence of an in- with lymphocytes, histiocytes, plasma cells, and granu- ¯ammatory reaction. Both adrenal glands showed focal locytes. Some myo®bers showed acute necrosis, but very necrosis in the zona fasciculata along with in®ltration of few contained groups of parasites. In the lung, in¯am- granulocytes and clusters of protozoan stages. There matory changes and parasites were not observed. A few was a mild degree of focal in®ltration by eosinophils, clusters of parasites were scattered throughout the cor- lymphocytes, and plasma cells in the lamina propria of tex, medulla, and interstitium of the thymus in the ab- the stomach. A few parasites were detected in chief cells sence of in¯ammatory reactions. of the glandular epithelium of the fundus in the absence In the liver, minute focal hepatocellular necrosis was of an in¯ammatory response. Neither in¯ammation nor seen associated with mild in®ltration of mononuclear parasites were detected in the small or large intestines. cells and granulocytes. Very few solitary tachyzoites The pancreas showed perivascular and mild interstitial were detected immunohistochemically in association in®ltrations with histiocytes and neutrophils without 5 b sporosis. The pup had a serum IFAT titer of 1:1,600, Fig. 1 Tapetal fundus of the left eye with several hypore¯ective which is in accordance with the ®ndings of Dubey and maculae Lindsay (1996), who observed IFAT titers of ³1:200 in Fig. 2 Granulomatous encephalitis with gliosis, perivascular cung, and a solitary Neospora caninum cyst (¬). H&E, ´200 clinical cases of proven neosporosis. Antibodies to Fig. 3 a N. caninum cysts in the CNS as viewed after immunohisto- Toxoplasma gondii were not detected. However, serology chemical staining with the polyclonal BAG-5 antiserum (ABC). Note alone is of limited diagnostic value because clinically the thin cyst wall (¯). Oil immersion, di€erential interference contrast normal dogs may have IFAT titers of >1:800 (Dubey (DIC), ´1,000. b N. caninum stages in the CNS show immunohisto- and Lindsay 1993). Antibodies in the CSF have been chemical staining with the polyclonal BAG-5 antiserum in a sprinkled pattern without a cyst wall structure (ABC). Oil immersion, DIC, reported in cases of canine neosporosis by Hay et al. ´1,000 (1990), Jackson et al. (1995), and Van Ham et al. (1996). Fig. 4 a Cluster of N. caninum stages in the inner nuclear layer of the Demonstration by PCR of N. caninum DNA in the CSF retina. Protozoa show immunohistochemical staining with the poly- con®rmed actual infection with the parasite. clonal BAG-5 antiserum (ABC). (pe Pigment epithelium, rc rods and cones, onl outer nuclear layer, opl outer plexiform layer, inl inner The puppy was born in the fourth litter of a clinically nuclear layer, ipl inner plexiform layer, gcl ganglion cell layer) Note the healthy bitch. One of seven littermates had died with lack of in¯ammatory reaction, rarefaction of the outer plexiform layer similar clinical symptoms, and neosporosis was sus- (¬), and cavitation between the pigment epithelium and the layer of pected on postmortem examination. According to the rods and cones (¯). DIC, ´200. b Higher magni®cation of a. N. caninum stages are located within the cytoplasm of a cell with an eccentric owner, comparable disease had not been seen in former nucleus. A cyst wall is not visible. Oil immersion, DIC, ´1,000 litters. Transplacental infection is believed to be the Fig. 5 Myositis with a mixed in¯ammatory-cell in®ltration and an major route of transmission in canine neosporosis N. caninum cluster within a myo®ber (®). H&E, ´400 (Ruehlmann et al. 1995). Repeated vertical transmission Fig. 6 Group of N. caninum stages within a myo®ber stained with the has been reported, but not all pups of a litter have to be polyclonal BAG-5 antiserum (ABC). Oil immersion, DIC, ´1,000 infected, and not all infected pups develop clinical dis- ease (Barber and Trees 1998). In the present case the detectable protozoan stages. In the ovaries, uterus, uri- seropositivity of the bitch and the presence of the disease nary bladder, and thyroid gland, in¯ammatory lesions in two young pups of the same litter indicate congenital were absent and parasites were not observed. In the skin, infection. Nevertheless, postnatal infection cannot be parasites were detected immunohistochemically in the ruled out. epithelium of hair follicles. Moderate pleocytosis with dominating neutrophils indicated meningoencephalitis, which was not obvious from the results of the neurological examination. Eo- Isolation of N. caninum sinophils were not demonstrated in the CSF of the pup. The presence of eosinophils in the CSF is not a constant After inoculation of cell cultures with brain and spinal ®nding in protozoan encephalomyelitis, although it has cord homogenates, parasites were ®rst detected after 4 previously been reported in several cases of neosporosis and 15 days, respectively. Prednisolone-acetate-treated (Cuddon et al. 1992; Jackson et al. 1995; Van Ham et al. mice (AKR, BALB/c) became sick and were euthanized 1996). Cytological examination revealed no evidence of at 10±13 days after inoculation. BALB/c J-nu mice be- tachyzoites in the CSF, as in a case of canine neo- came sick and were euthanized at 21±31 days after in- sporosis reported by Dubey et al. (1988a). Biochemical jection. The peritoneal exudate of a prednisolone-treated serum analysis showed a slight elevation in creatine ki- AKR mouse was positive at 11 days after inoculation nase activity, which is commonly observed in association with spinal cord material (the peritoneal exudate con- with N. caninum infection (Ruehlmann et al. 1995) due tained 5 ´ 106 parasites) and was used to infect a Vero- to polymyositis. The complete blood count was char- cell culture. In all infected cell cultures, parasites grew acterized by increased numbers of eosinophils and rapidly, and isolates were stabilized in liquid nitrogen at monocytes. Although increased numbers of eosinophils between 17 and 25 days after the dog had been eu- were observed in the present case and have been seen in thanized. Isolates were designated NC-GER1.1 (brain; other cases (Dubey et al. 1988a; Wolf et al. 1991), this is 17 days of cell culture), NC-GER1.2 (spinal cord; not a constant ®nding in canine neosporosis, and he- 11 days in a mouse followed by 6 days of cell culture), matological values are usually not altered (Dubey and and NC-GER1.3 (spinal cord; 25 days of cell culture). Lindsay 1996). Detection of N. caninum DNA in all tissues examined by PCR as well as by pathohistological and immuno- Discussion histological examination revealed that N. caninum in- fection of the puppy was disseminated. In¯ammatory Neosporosis was clinically suspected in an 11-week-old lesions combined with the presence of parasites were female Kleiner MuÈ nsterlaÈ nder pup presented with pa- identi®ed in the CNS, skeletal muscles, tongue, eso- resis of the left hindlimb in association with extensor phageal muscle, heart, liver, adrenal glands, and skin. rigidity. Clinical signs and the presence of antibodies to The histopathological lesions detected in the CNS and Neospora caninum in serum and CSF as detected by skeletal muscles of the puppy were similar to those de- immunoblotting and IFAT were indicative of neo- scribed in the literature (Ruehlmann et al. 1995). The 6 simultaneous presence of granulomatous lesions and of McAllister et al. (1996) indicates that antibodies foci with acute suppurative in¯ammation in the CNS as against BAG-5 cross-react with N. caninum bradyzoites well as the clinical deterioration observed after the ter- but have no anity for tachyzoites. As expected, BAG-5 mination of treatment indicate rapid progression, per- antibodies recognized tissue-cyst stages within the CNS. haps even acute exacerbation, of the infection. However, However, BAG-5 antibodies also reacted with stages this was not re¯ected by an alteration in the IFAT titer that were not surrounded by a detectable cyst wall in the of the puppy. The severity of the in¯ammatory lesions, CNS and retina. Remarkably, a few of the parasite especially in the CNS, and the high density of protozoan clusters without visible cyst walls observed in myo®bers stages are in marked contrast to the observed neuro- of the skeletal muscles, diaphragm, and tongue reacted logical symptoms. The high density of protozoan stages, with the polyclonal BAG-5 antiserum. BAG-5-positive on the other hand, presumably explains the rapid iso- parasitic stages in skeletal myo®bers were also detected lation of the parasite. N. caninum was ®rst detected in using the mouse monoclonal antibody 74.1.8, which is cell cultures that had been inoculated with infected brain directed against the same antigen. Thus far, N. caninum and spinal cord after 4 and 15 days, respectively. Mice cysts have been shown only in neural tissues and, once, became sick at 10±31 days after inoculation. The new in an ocular muscle of a congenitally infected foal isolate is the ®rst reported in Germany and was desig- (Lindsay et al. 1996). The observation that nonencysted nated NC-GER1. To the knowledge of the authors, in stages in the skeletal muscles, tongue, and diaphragm addition to several bovine isolates and one equine iso- exhibit bradyzoite markers is surprising. In vitro studies late, six canine isolates of the parasite have been de- with T. gondii indicate that the BAG-5 antigen is an scribed. Five isolates from dogs have been reported in early developing bradyzoite marker (Dubey et al. the United States (Dubey et al. 1988b, 1998a; Hay et al. 1998b). Therefore, stages expressing the BAG-5 antigen 1990; Lindsay et al. 1991) and named NC-1 to NC-5, might not be fully mature bradyzoites. This view might and one isolate has been described in the United King- be supported by our ®nding that the BAG-5-speci®c dom (Barber et al. 1995) and designated NC-Liverpool reactivities observed in nonencysted extraneural stages (NC-Liv). Studies on the pheno- and genotypic charac- did not have the same intensity as those observed in terization of the new German isolate are in progress. encysted zoites. To our knowledge this is the ®rst time Protozoa stained immunohistochemically positively that BAG-5-positive N. caninum stages have been ob- with a polyclonal rabbit antiserum developed against served outside neural tissues. It is not known whether this N. caninum tachyzoites (Schares et al. 1997). Cysts and is a common ®nding in N. caninum-infected hosts or what tachyzoites were detected at all of the examined levels the role of these stages might be in the life cycle of N. of the brain and spinal chord. Cysts measured an av- caninum. Observations on Hammondia heydorni, an api- erage of 19.4 ´ 16.2 lm and had predominantly thin complexan parasite that seems to be more closely related cyst walls (£1 lm). This observation is in accordance to N. caninum than is T. gondii (Ellis et al. 1999), indicate with the size of tissue cysts reported in the CNS of that in addition to the brain, other tissues from inter- puppies by Dubey et al. (1998a) and Jardine (1996). mediate hosts are infectious to the de®nitive host, the Some protozoan clusters were also detected in the ret- dog, via the oral route (Matsui et al. 1981, 1986). How- ina of the left eye. Focal cystic spaces between the ever, only two reports exist on tissue cysts in H. heydorni- retina and the pigment epithelium as well as the rar- infected intermediate hosts (Heydorn 1973; Matsui efaction of the outer plexiform layer in the vicinity of 1991), and it is tempting to speculate that in addition to the protozoa were probably sequelae of a previous encysted stages, nonencysted stages might be responsible retinitis. These lesions were visible by indirect for the oral transmission of H. heydorni from interme- ophthalmoscopy as maculae in the tapetal fundus. diate to de®nitive hosts (Matsui 1991). Therefore, further Focal retinochoroiditis associated with neosporosis has studies are needed to elucidate the capability of non- previously been reported only by Dubey et al. (1990). encysted but BAG-5-positive N. caninum stages to infect They observed focal retinitis, primarily of the inner de®nitive or intermediate carnivorous hosts. retinal layers, in four puppies and found tissue cysts Acknowledgements The authors would like to thank M.M. and tachyzoites associated with retinal lesions in three McAllister for providing the polyclonal BAG-5 antiserum, for of the dogs. The localization of tissue cysts is strikingly reading the manuscript, and for helpful suggestions. We thank similar in ocular toxoplasmosis. In murine congenital L.M. Weiss for providing the mouse monoclonal BAG-5 antibody ocular toxoplasmosis, cysts are most frequently found 74.1.8. We are indebted to A. BaÈ rwald (BFAV Wusterhausen) for in the inner retinal layers at the border of retinal scars her help in isolating NC-GER1 and to B. Buck (Institut fuÈ r Pathologie, TieraÈ rztliche Hochschule Hannover) for her technical or remote from scars in the absence of an in¯ammatory assistance. Many thanks are due W. Thiel (Staatliches Vet- reaction (McMenamin et al. 1986). erinaÈ runtersuchungsamt Detmold) and C. Suschka (pTA in GuÈ - Bradyzoites were immunohistochemically di€erenti- tersloh) for their help in discovering the dog. ated from tachyzoites using a polyclonal rabbit antise- rum against the bradyzoite-speci®c BAG-5 antigen References (McAllister et al. 1996). BAG-5 is a 28- to 30-kDa, small heat-shock protein of T. gondii that is speci®cally ex- Barber JS, Trees AJ (1998) Naturally occurring vertical transmis- pressed in the cytosol of bradyzoites. The investigation sion of Neospora caninum in dogs. Int J Parasitol 28: 57±64 7

Barber JS, Holmdahl OJM, Owen MR, Guy F, Uggla A (1995) Lindsay DS, Steinberg H, Dubielzig RR, Semrad SD, Konkle DM, Characterization of the ®rst European isolate of Neospora Miller PE, Blagburn BL (1996) Central nervous system caninum (Dubey, Carpenter, Speer, Topper, and Uggla). Para- neosporosis in a foal. J Vet Diagn Invest 8: 507±510 sitology 111: 563±568 Matsui T (1991) The tissue stages of Isospora heydorni in the guinea Barber JS, Payne-Johnson CE, Trees AJ (1996) Distribution of pig as an intermediate host. Jpn J Parasitol 40: 581±586 Neospora caninum within the central nervous system and other Matsui T, Morii T, Iikijma T, Ito S, Tsunoda K, Correa WM, tissues of six dogs with clinical neosporosis. J Small Anim Pract Fujino T (1981) Cyclic transmission of the small type of 37: 568±574 Isospora bigemina of the dog. Jpn J Parasitol 30: 179±186 Cuddon P, Lin DS, Bowman DD, Lindsay DS, Miller TK, Duncan Matsui T, Morii T, Iikijma T, Ito S, Tsunoda K, Kobayashi F, ID, Lahunta A de, Cummings J, Suter M, Cooper B, King JM, Fujino T (1986) Isospora heydorni isolated in Brazil: endoge- Dubey JP (1992) Neospora caninum infection in English nous stages in dogs. Jpn J Parasitol 35: 215±222 springer spaniel littermates. Diagnostic evaluation and organ- Mayhew IG, Smith KC, Dubey JP, Gatward LK, McGlennon NJ ism isolation. J Vet Intern Med 6: 325±332 (1991) Treatment of encephalomyelitis due to Neospora cani- Dubey JP, Lindsay DS (1993) Neosporosis. Parasitol Today 9: num in a litter of puppies. J Small Anim Pract 32: 609±612 452±458 McAllister M, Parmley SF, Weiss LM, Welch VJ, McGuire M Dubey JP, Lindsay DS (1996) A review of Neospora caninum and (1996) An immunohistochemical method for detecting brady- neosporosis. Vet Parasitol 67: 1±59 zoite antigen (BAG5) in Toxoplasma gondii-infected tissues Dubey JP, Carpenter JL, Speer CA, Topper MJ, Uggla A (1988a) cross-reacts with Neospora caninum bradyzoite antigen. Newly recognized fatal protozoan disease of dogs. J Am Vet J Parasitol 82: 354±355 Med Assoc 192: 1269±1285 McAllister MM, Dubey JP, Lindsay DS, Jolly WR, Wills RA, Dubey JP, Hattel AL, Lindsay DS, Topper MJ (1988b) Neonatal McGuire AM (1998) Dogs are de®nitive hosts of Neospora Neospora caninum infection in dogs: isolation of the causative caninum. Int J Parasitol 28: 1473±1478 agent and experimental transmission. J Am Vet Med Assoc 193: McMenamin PG, Dutton GN, Hay J, Cameron S (1986) The 1259±1263 ultrastructural pathology of congenital murine toxoplasmic Dubey JP, Koestner A, Piper RC (1990) Repeated transplacental retinochoroiditis. I. The localization and morphology of transmission of Neospora caninum in dogs. J Am Vet Med Toxoplasma cysts in the retina. Exp Eye Res 43: 529±543 Assoc 197: 857±860 Odin M, Dubey JP (1993) Sudden death associated with Neospora Dubey JP, Metzger FL, Hattel AL, Lindsay DS, Fritz DL (1995) caninum-myocarditis in a dog. J Am Vet Med Assoc 293: 831± Canine cutaneous neosporosis: clinical improvement with clin- 833 damycin. Vet Dermatol 6: 37±44 Perl S, Harrus S, Satuchne C, Yakobson B, Haines D (1998) Cu- Dubey JP, Dorough KR, Jenkins MC, Liddell S, Speer CA, Kwok taneous neosporosis in a dog in Israel. Vet Parasitol 79: 257± OCH, Shen SK (1998a) Canine neosporosis: clinical signs, di- 261 agnosis, treatment and isolation of Neospora caninum in mice Poli A, Mancianti F, Carli MA, Stroscio MC, Kramer L (1998) and cell culture. Int J Parasitol 28: 1293±1304 Neospora caninum infection in a Bernese cattle dog from Italy. Dubey JP, Lindsay DS, Speer CA (1998b) Structures of Toxo- Vet Parasitol 78: 79±85 plasma gondii tachyzoites, bradyzoites, and sporozoites and Ruehlmann D, Podell M, Oglesbee M, Dubey JP (1995) Canine biology and development of tissue cysts. Clin Microbiol Rev 11: neosporosis: a case report and literature review. J Am Anim 267±299 Hosp Assoc 31: 174±183 Ellis JT, Morrison DA, Liddell S, Jenkins MC, Mohammed OB, Schares G, Peters M, Wurm R, Tackmann K, Henning K, Con- Ryce C, Dubey JP (1999) The genus Hammondia is para- raths FJ (1997) Neospora caninum verursacht Aborte in einem phyletic. Parasitology 118: 357±362 Rinderbestand in Nordrhein-Westfalen. Dtsch TieraÈ rztl Fritz D, George C, Dubey JP, Trees AJ, Barber JS, Hopfner CL, Wochenschr 104: 208±212 Mehaut S, Le Net JL, Longeart L (1997) Neospora associated Schares G, Peters M, Wurm R, BaÈ rwald A, Conraths FJ (1998) The nodular dermatitis in a middle aged dog. Canine Pract 22: 21±24 eciency of vertical transmission of Neospora caninum in dairy Fuchs N, Sonda S, Gottstein B, Hemphill A (1998) Di€erential cattle analysed by serological techniques. Vet Parasitol 80: 87± expression of cell surface- and dense granule-associated 98 Neospora caninum proteins in tachyzoites and bradyzoites. Schares G, Dubremetz JF, Dubey JP, BaÈ rwald A, Loyens A, J Parasitol 84: 753±758 Conraths JF (1999) Neospora caninum: identi®cation of 19-, Greig B, Rossow KD, Collins JE, Dubey JP (1995) Neospora 38- and 40-kDa surface antigens and a 33-kDa dense granule caninum pneumonia in an adult dog. J Am Vet Med Assoc 206: antigen using monoclonal antibodies. Exp Parasitol 92 (in 1000±1001 press) Hay WH, Shell LG, Lindsay DS, Dubey JP (1990) Diagnosis and Van Ham LML, Thoonen H, Barber JS, Trees AJ, Polis I, De Cock treatment of Neospora caninum infection in a dog. J Am Vet H, Hoorens JK (1996) Neospora caninum in the dog: typical and Med Assoc 197: 87±89 atypical cases. Vlaams Diergeneeskd Tijdschr 65: 326±334 Heydorn AO (1973) Zum Lebenszyklus der kleinen Form von Weiss LM, LaPlace D, Tanowitz HB, Wittner M (1992) Identi®- Isospora bigemina des Hundes. 1. Rind und Hund als moÈ gliche cation of Toxoplasma gondii bradyzoite-speci®c monoclonal Zwischenwirte. Berl MuÈ nch TieraÈ rztl Wochenschr 86: 323±329 antibodies. J Infect Dis 166: 213±215 Jackson W, Lahunta A de, Adaska J, Cooper B, Dubey JP (1995) WeissenboÈ ck H, Dubey JP, Suchy A, Sturm E (1997) Neosporose Neospora caninum in an adult dog with progressive cerebellar als Ursache von Encephalomalazie und Myocarditis bei signs. Prog Vet Neurol 6: 124±127 Hundewelpen. Wien TieraÈ rztl Monatsschr 84: 233±237 Jardine JE (1996) The ultrastructure of bradyzoites and tissue cysts Wolf M, Cachin M, Vandevelde M, Tipold A, Dubey JP (1991) Zur of Neospora caninum in dogs: absence of distinguishing mor- klinischen Diagnostik des protozoaÈ ren Myositissyndroms phological features between parasites of canine and bovine (Neospora caninum) des Welpen. TieraÈ rztl Prax 19: 302±306 origin. Vet Parasitol 62: 231±240 Yamage M, Fletchner O, Gottstein B (1996) Neospora caninum: Lindsay DS, Dubey JP, Blagburn BL (1991) Characterisation of a speci®c oligonucleotide primers for the detection of brain cysts Neospora caninum (Protozoa: Apicomplexa) isolate (NC-3) in DNA of experimentally infected nude mice by the polymerase mice. J Ala Acad Sci 62: 1±8 chain reaction (PCR). J Parasitol 82: 272±279