Mycobacterium paratuberculosis monoassociated nude mice as a paratuberculosis model

In this study, a paratuberculosis (Johne's disease) model was developed by intragastrically dosing gnotobiotic athymic nude mice with Mycobacterium paratuberculosis. The mice infrequently shed bacilli from their intestinal tracts during the first 4 months after inoculation. Following this time, increasing numbers of M. paratuberculosis (greater than 4.0 log10 bacilli per fecal pellet by 40 weeks) were recovered from the feces of the 12 mice that remained in the isolator. A similar pattern of recovery of M. paratuberculosis was obtained from the ileum, cecum, colon, and liver. Histopathologic lesions and acid-fast bacilli were rare during the first 4 months of infection and then, with time, increased in prevalence and severity. Mice maintained for 7 months or longer exhibited severe granulomatous inflammation and large numbers of acid-fast bacilli in the gastrointestinal tract and liver (up to 10(8) log10 colony forming units per gram wet weight). Five mice maintained for 7 months or more developed clinical signs consistent with those seen in paratuberculosis (weight loss, chronic diarrhea); three of these mice eventually died or became moribund and were euthanatized. M. paratuberculosis monoassociated mice released increased levels of tumor necrosis factor activity into their sera, as compared to uninfected control mice, when they were injected with bacterial lipopolysaccharide. The clinical signs, fecal shedding of M. paratuberculosis, granuloma formation, and progressive bacillary multiplication observed with these mice are consistent with naturally occurring M. paratuberculosis infection of ruminants (Johne's disease). This model will be useful for future studies of immunoregulation and antimicrobial therapy of paratuberculosis.     

Efficacy of immunologic assays for the detection of Johne's disease in dairy cows fed additional energy during the periparturient period.

This study was designed to evaluate if the immunosuppression typically observed during the immediate periparturient period (3 weeks before and after calving) in dairy cows influences the effectiveness of diagnostic tests for the detection of Johne's disease; and, if providing additional energy to the cows during this period would minimize any immunosuppressive effects. Twelve dairy cows naturally infected with Mycobacterium paratuberculosis were fitted with rumen cannulas in late gestation and assigned to treatment groups: control, n = 6; or stuffed, n = 6. Cows in the control group were allowed to consume feed ad libitum. Cows assigned to the stuffed treatment group were also fed ad libitum but received additional total mixed ration by manually stuffing their rumens with refused feed to maintain a dry matter intake of 2% body weight/day before calving and 2.5% body weight/day after calving. Parturition had a significant impact on immune function with significant reductions in M. paratuberculosis-specific antibodies detected in the serum and milk regardless of treatment group. Similarly, in vitro immunoglobulin production was decreased at calving for both treatment groups. In addition, stuffing cows modulated cell-mediated immune function by reducing antigen-specific lymphocyte proliferation and interferon-gamma production after calving. Shedding of M. paratuberculosis in the milk was apparent in 58% (7/12) of cows after parturition with no difference noted between control and stuffed animals. Parturition had no major effect on fecal shedding of cows regardless of treatment. These data suggest that parturition had a significant effect on immune function parameters including diagnostic tests for paratuberculosis. Furthermore, providing additional energy to cows with Johne's disease did not preclude immunosuppressive effects during the periparturient period.     

The detection of Mycobacterium paratuberculosis in bovine faeces by isolation and the comparison of isolation with the examination of stained smears by light microscopy

The detection of Mycobacterium paratuberculosis organisms in bovine faeces by isolation was compared with that by the microscopical examination of Ziehl-Neelsen stained faecal smears for the presence of clumps of acid-fast M. paratuberculosis organisms. Faeces were obtained from cattle naturally or experimentally infected with M. paratuberculosis as well as from uninfected cattle. Microscopical examination was an unreliable method for the detection of M. paratuberculosis organisms, since the organisms were only detected in 99 (=55.9%) of 177 culturally positive faecal samples. 1111 addition, clumps of acid-fast organisms indistinguishable from M. paratuberculosis were also observed iin three of 18 samples from cattle free from Johne's disease and in 18 of 37 culturally negative samples from paratuberculous cattle. When M. paratuberculosis organisms were added to faeces from an uninfected cow, results showed that isolation attempts should be positive when 15 or more M. paratuberculosis organisms per gram of faeces are present.     

Effects of altered dietary iron intake in Mycobacterium paratuberculosis-infected dairy cattle: sequential observations on growth, iron and copper metabolism and development of paratuberculosis

Twenty calves were orally inoculated with Mycobacterium paratuberculosis at six weeks old. At six months old, 10 of these, plus four uninfected controls were maintained on limited dietary copper and supplemented iron intake for a further 27 months. During this time all these animals, together with a further four untreated controls, were bred before being killed and examined for evidence of paratuberculosis. Despite significant reduction in weight gain, attributable to both iron supplementation and infection, no significant difference was found in the numbers of iron-supplemented and unsupplemented animals that developed clinical signs nor in the extent and severity of intestinal lesions between groups. Accumulation of iron in paratuberculosis lesions was not affected by iron supplementation but was positively correlated with the frequency of shedding of M paratuberculosis in faeces (P less than 0.05). Dietary iron supplementation alone resulted in serum hyperferraemia, hepatic siderosis and slight hypocuprosis, whereas, in infected animals, this resulted in marked hypocuprosis and anaemia within groups (P less than 0.05). Infection alone resulted in serum hypoferraemia and intestinal and hepatic siderosis which was positively correlated with the severity of infection within groups (P less than 0.05). Susceptibility to paratuberculosis may result from failure ultimately to limit monokine-mediated iron sequestration in intestinal tissue.     

Sensitivity of test strategies used in the Voluntary Johne's Disease Herd Status Program for detection of Mycobacterium paratuberculosis infection in dairy cattle herds

OBJECTIVE: To evaluate sensitivities at the herd level of test strategies used in the Voluntary Johne's Disease Herd Status Program (VJDHSP) and alternative test strategies for detecting dairy cattle herds infected with Mycobacterium paratuberculosis. DESIGN: Nonrandom cross-sectional study. SAMPLE POPULATION: 64 dairy herds from Pennsylvania, Minnesota, Colorado, Ohio, and Wisconsin. Fifty-six herds had at least 1 cow shedding M. paratuberculosis in feces; the other 8 herds were free from paratuberculosis. PROCEDURE: For all adult cows in each herd, serum samples were tested for antibodies to M. paratuberculosis with an ELISA, and fecal samples were submitted for bacterial culture for M. paratuberculosis. Sensitivities at the herd level (probability of detecting infected herd) of various testing strategies were then evaluated. RESULTS: Sensitivity at the herd level of the testing strategy used in level 1 of the VJDHSP (use of the ELISA to test samples from 30 cows followed by confirmatory bacterial culture of feces from cows with positive ELISA result) ranged from 33 to 84% for infected herds, depending on percentage of cows in the herd with positive bacterial culture results. If follow-up bacterial culture was not used to confirm positive ELISA results, sensitivity ranged from 70 to 93%, but probability of identifying uninfected herds as infected was 89%. CONCLUSIONS AND CLINICAL RELEVANCE: Results suggest that the testing strategy used in the VJDHSP will fail to identify as infected most dairy herds with a low prevalence of paratuberculosis. A higher percentage of infected herds was detected if follow-up bacterial culture was not used, but this test strategy was associated with a high probability of misclassifying uninfected herds.     

The effect of experimental infection with Rhodococcus equi on immunodeficient mice

To investigate the pathogenesis of respiratory lesions caused by the facultative intracellular pathogen, Rhodococcus equi, pulmonary clearance was compared in four groups of genetically defined mice, chosen for their specific deficits in immune and inflammatory responses. Complement-deficient A/J, immunodeficient nu/nu (nude), scid/scid.bg/bg (SCID/beige), C57BI/6J.bg/bg (beige) and normal Swiss mice (SW) received approximately 10(7) R. equi intranasally on day 0. Bacterial clearance was assessed in lung, liver and spleen on days 1, 4, 7 and 14. Pulmonary clearance was not significantly different between SW and A/J mice. Beige mice cleared R. equi more rapidly and completely than A/J and SW, indicating that deficits in phagocytic and NK cell function associated with the bg/bg gene did not compromise clearance. Pulmonary clearance in immunodeficient SCID/beige mice paralleled that of the SW and A/J mice initially but bacterial proliferation produced significant differences from SW mice at day 14. Nude mice were unable to clear R. equi from day 1, resulting in the death of two nude mice at day 11. Both SCID/beige and nude mice developed severe pyogranulomatous bronchopneumonia, whereas A/J and SW mice developed transient pulmonary lesions. Beige mice developed minimal lung lesions. Significant systemic bacterial proliferation occurred only in nude and SCID/beige mice. We conclude that deficiencies in complement components, phagocytic and NK cells do not impair the pulmonary clearance of R. equi but that a competent cellular immune system is required to prevent pneumonia and death. The difference in early phase pulmonary clearance in nude and SCID/beige mice indicates two phases are important for clearance. An acapsular mutant of R. equi was completely cleared from the lungs of SCID/beige mice suggesting an important role for the capsule in virulence for mice.     

Evaluation of bacteriologic culture of pooled fecal samples for detection of Mycobacterium paratuberculosis

OBJECTIVES: To compare sensitivity of several methods of bacteriologic culture of pooled bovine fecal samples for detection of Mycobacterium paratuberculosis and evaluate homogeneity in number of M paratuberculosis in pooled fecal samples. SAMPLE POPULATION: Feces from 10 dairy cows that shed M paratuberculosis at various concentrations and 1 dairy cow known to be free of infection with M paratuberculosis. PROCEDURE: 5 fecal pooling methods, 2 culture methods, and 2 pool sizes were evaluated. Each pooled sample contained 1 infected sample and 4 or 9 uninfected samples. RESULTS: Sensitivity of detection of M paratuberculosis was greater with smaller pool size (5 vs 10 samples/pool). Detection sensitivity was also associated with concentration of bacteria in the infected sample. Results indicated that, compared with concurrent bacterial culture of individual infected samples, 37 to 44% of pooled samples with low bacterial concentrations yielded positive culture results and 94% of pooled samples with high bacterial concentrations yielded positive results. CONCLUSIONS AND CLINICAL RELEVANCE: Bacteriologic culture of pooled fecal samples may provide a valid and cost-effective method of detecting M paratuberculosis infection in cattle herds.     

Use of the polymerase chain reaction assay for the detection of Mycobacterium avium subspecies paratuberculosis in blood and liver biopsies from experimentally infected sheep

OBJECTIVE: To evaluate a polymerase chain reaction-based assay for the detection of Mycobacterium avium subsp paratuberculosis in blood and liver biopsies from subclinically infected sheep. STUDY DESIGN: A direct PCR assay for the detection of M a paratuberculosis was applied to liver biopsy specimens and to samples of blood that were sequentially collected over 53 weeks from 14 sheep infected experimentally with the organism. RESULTS: Of 117 blood samples from the 14 experimentally infected sheep, two tested positive in the PCR assay. Both positive results were obtained in two subclinically infected sheep that had paratuberculosis later confirmed by histological examination at necropsy. However, the assay failed to detect the target DNA in samples of blood from five other sheep with histologically confirmed paratuberculosis. Similarly, the PCR assay on liver biopsy specimens collected 32 weeks after administration of M a paratuberculosis gave only two positive results, both of which were obtained in sheep with histological evidence of paratuberculosis. CONCLUSIONS: The PCR assay on blood and liver biopsies does not provide a useful method for the diagnosis of M a paratuberculosis infection in subclinically infected sheep.     

Chronic infection with Trichostrongylus vitrinus in sheep. Some effects on food utilisation, skeletal growth and certain serum constituents

The effect on food utilisation and skeletal growth of daily infection for 90 days with 2500 Trichostrongylus vitrinus larvae was investigated. Controls were killed initially or fed ad libitum. Food intake, digestibility of dry matter and nitrogen and content of body fat, protein, water and minerals were determined, and selected bones examined chemically and histologically. Serum constituents were determined weekly. Digestibility of food dry matter and nitrogen were reduced during weeks 3 and 5 after initial infection but subsequently recovered. Deposition of fat, protein, calcium and phosphorus in the body was reduced by 36, 58, 67 and 78 per cent respectively. Gross efficiency of utilisation of metabolisable energy was reduced by 36 per cent. Bone changes were associated with a reduction in osteoblastic activity. Hypoalbuminaemia, hyperglobulinaemia and hypophosphataemia were present in infected sheep. There were no significant changes in serum calcium or urea concentrations. It is concluded that the effects of continuous intakes of T vitrinus and T colubriformis larvae on host metabolism are essentially the same.     

Polymerase chain reaction analysis of TNF-alpha and IL-6 mRNA levels in whole blood from cattle naturally or experimentally infected with Mycobacterium paratuberculosis.

Johne's disease is characterized by a chronic enteritis that results in granulomatous inflammation, cachexia, and eventual death of cattle infected with Mycobacterium paratuberculosis. The cytokines tumor necrosis factor-alpha (TNF-alpha) and interleukin-6 (IL-6) have been associated with granuloma formation and wasting in other disease syndromes. The potential role of these cytokines in the development and progression of Johne's disease has not been investigated. Using the polymerase chain reaction (PCR) and specific bovine oligonucleotide cytokine primers and probes, we examined the expression of messenger RNA for these cytokines in whole blood from M. paratuberculosis infected and uninfected cattle. Cytokine mRNA levels were examined before and after in vitro incubation with E.coli lipopolysaccharide (LPS) and lipoarabinomannan (LAM) purified from M. paratuberculosis. Uninfected calves, experimentally infected calves, and naturally infected cattle all displayed similar cytokine mRNA expression patterns. However, individual animals demonstrated variability in the levels of IL-6 and TNF-alpha mRNA expression as determined by a semiquantitative PCR method using 32P-labelled oligonucleotide probes.     

Temporal Mycobacterium paratuberculosis infection in T-cell receptor (TCR)-alpha and TCR-delta-deficient mice

Mycobacterium paratuberculosis is the causative agent of paratuberculosis (Johne's disease), a chronic inflammation of the terminal portion of the ileum in ruminants. The predominance of cell-mediated immunity in early stages of the disease suggests that T lymphocytes are essential to protect the host from infection with M. paratuberculosis. In this study, we investigated the role of alphabeta and gammadelta T cells in resistance to M. paratuberculosis infection using a T-cell receptor (TCR) knockout mouse model. Weanling TCR-alpha-deficient, TCR-delta-deficient, and C57BL/6 control mice (5-6 weeks of age) were acclimated for 2 weeks and then inoculated intraperitoneally with 10(8)CFU/ml of M. paratuberculosis (either strain 19698 or strain Ben). Groups of mice within each treatment group were euthanized at 1, 3 and 6 months post-inoculation. Sections of spleen, liver, ileum and mesenteric lymph node were prepared for bacterial culture and histologic examination. At all time points of infection and regardless of bacterial strain, TCR-alpha-deficient mice had higher levels of M. paratuberculosis colonization in their tissues compared to TCR-delta-deficient mice or C57BL/6 control mice. Lesions were located predominately in the liver and the ileum, depending upon period of infection, and lesion scores were higher for TCR-alpha-deficient mice compared to the other treatment groups. These results suggest that alphabeta T cells play a major role in resistance to infection with M. paratuberculosis and that gammadelta T cells may play a lesser role and potentially confound protective immune responses.     

Studies of infectivity and pathogenicity of an isolate of Trypanosoma evansi in Yankasa sheep.

The course of experimental infection and pathogenicity of an isolate of Trypanosoma evansi were investigated using eight infected and six uninfected control Yankasa sheep. The sheep were each infected intravenously via the jugular vein with approximately 2.0 x 10(6) T. evansi parasites. The effects of the parasite on body temperature, packed cell volume (PCV), haemoglobin, erythrocytes, total protein, were monitored three times a week for approximately 9 weeks. Body weights were determined once every week for the duration of the experiment. The results showed that all the infected sheep were positive for the parasite. The prepatent period varied between 3 and 6 days. T. evansi produced parasitaemic waves at an average of 8.3 days interval. Two distinct forms of the disease were produced namely, acute (4-14 days postinfection), and chronic (43-59 days postinfection). Anaemia was a distinct feature of the disease. While the mean rectal temperatures were significantly elevated (P < 0.05), the mean values of the haematological parameters of the infected sheep dropped significantly (P < 0.05) compared to the preinfection levels. Observed clinical signs included pale mucous membrane, epiphora, loss of appetite, emaciation, dullness and rough hair coat together with fluctuating pyrexia which in most cases coincided with rise in parasitaemia. It is suggested that the isolate of T. evansi is pathogenic for Yankasa sheep.     

Experimental infection of weaner sheep with S strain Mycobacterium avium subsp. paratuberculosis

We sought to determine whether infection of recently weaned 12-16-week-old Merino lambs with an Australian S strain M. a. paratuberculosis, at doses consistent with natural exposure, could be detected in the first few months post-inoculation. Such detection would facilitate the use of weaner sheep as sentinel animals for the presence of infectious doses of M. a. paratuberculosis on pastures. In controlled pen trials, oral doses of approximately 10(7)-10(8) viable organisms were demonstrated to be infective, whereas doses below 10(4) organisms failed to produce detectable infection. Mycobacterium avium subsp. paratuberculosis (M. a. paratuberculosis) was isolated from intestinal and/or lymphoid tissues collected at necropsy 7 or 14 weeks after first infection, but there were no associated gross or microscopic lesions. Skin testing with intradermal Johnin detected all three infected lambs at 13 weeks post-infection, and one of the three infected lambs at 6 weeks post-infection, with 100% specificity. Results for whole blood IFN-gamma assay showed some correlation with infection status but lacked specificity. One infected lamb gave a positive result in an ELISA for antibodies to M. a. paratuberculosis, 14 weeks post-infection and 1 week after skin testing. This was the first demonstration of experimental infection with S strain M. a. paratuberculosis in Australian Merino sheep at doses likely to be representative of natural infection. Culture from tissues in the first few months post-exposure could facilitate the use of naive weaner sheep as tracer animals to detect heavy contamination of pastures with M. a. paratuberculosis, but low-level contamination may not be detected in such a system.     

Cytokine gene expression in ileal tissues of cattle infected with Mycobacterium paratuberculosis

Cytokine gene expression in ileal tissues of cattle infected with Mycobacterium paratuberculosis was evaluated. The effects of infection with M. paratuberculosis on cytokine production may influence immune regulation at the site of colonization, resulting in the chronic inflammatory state associated with the latter stages of this disease. Ileal samples were obtained at necropsy from noninfected control cows (n=8) and clinically infected cows (n=7) and processed for immunohistochemistry and in situ hybridization. Cows infected with M. paratuberculosis were in the latter stages of disease with clinical signs such as weight loss, watery diarrhea, and inappetence. Among cytokines we studied, interleukin-1alpha (IL-1alpha), IL-1beta, IL-6, and interferon-gamma (IFN-gamma) were expressed significantly more in infected animals than in noninfected control animals. The expression of tumor necrosis factor-alpha (TNF-alpha), however, was not different between the two groups of cattle. In addition, immunohistochemical staining demonstrated that the number of resident macrophages in the ileum of infected animals was three times greater than that of noninfected cows. In contrast to this, ileal tissues from noninfected control animals contained 1.5 times more neutrophils than the ileal tissues from cows infected with M. paratuberculosis. These data demonstrate that localized ileal cytokine production is different between cows chronically infected with M. paratuberculosis and noninfected control cows.     

Clearance of Theileria sergenti-infected bovine red blood cells in severe combined immune deficiency mice

Clearance of Theileria sergenti-infected bovine red blood cells (Bo-RBCs) from the blood circulation of severe combined immune deficiency (SCID) mice was studied to help understand the mechanisms of anemia developing in cattle infected with T. sergenti. For the clearance test, Bo-RBC samples having 2%, 58%, and 76% parasitemia and, as a control, parasite-free Bo-RBCs were prepared in the Bo-RBC-SCID mouse model. The T. sergenti-infected Bo-RBCs and the uninfected control Bo-RBCs were separately labeled with two, green and red, fluorescent dyes, mixed together, and injected intravenously into SCID mice. The blood samples collected at various time points were observed under a fluorescent microscope, and the numbers of green and red fluorescing RBCs were counted differentially to determine the clearance rates of T. sergenti-infected and uninfected Bo-RBCs. This test clearly demonstrated that the Bo-RBC samples having higher parasitemias were cleared faster from the blood circulation of SCID mice. The results suggest that the intravascular clearance system in SCID mice may have a mechanism by which T. sergenti-parasitized and non-parasitized Bo-RBCs are recognized and cleared differentially.     

Distribution of Mycobacterium avium subsp. paratuberculosis in organs of naturally infected bull-calves and breeding bulls

Paratuberculosis, caused by Mycobacterium avium subsp. paratuberculosis, has particular importance in cattle due to the resulting chronic diarrhoea, weight loss, decreased production, infertility and eventual death. While faecal oral route of infection is generally recognised, reports about semen-derived infection are rare. The objective of this work was to assess whether M.a. paratuberculosis may disseminate from the gastrointestinal tract to reproductive organs, and compare this event between naturally infected bull-calves and breeding bulls. Ten bull-calves, aged 6-28 weeks and four breeding bulls were tested by serology, faecal and tissue culture, IS900 PCR and RFLP. In seven bull-calves M.a. paratuberculosis was isolated predominantly from mesenteric lymph nodes (75%); isolates from mucosa of the intestine constituted 25%. In three breeding bulls, M.a. paratuberculosis was isolated both from intestinal mucosa and mesenteric lymph nodes. Head and mediastinal lymph nodes, liver, spleen and semen of bull no. 1 (Holstein-Friesian); testes and epididymis of bull no. 2 (Piemonte); testes, epididymides and seminal vesicle of bull No. 3 (Hereford); and seminal vesicle of bull No. 4 (Simmental) tested positive by culture. Hot-start PCR revealed M.a. paratuberculosis in semen, seminal vesicle and intestinal tissue where culture isolation was difficult. Isolates from bull-calves and breeding bulls were of RFLP types B-C9 and B-C1, respectively. Bull-calves born in infected herd can be sources of infection when later used for natural mating or artificial insemination. Sub-clinically infected bulls release M.a. paratuberculosis into semen, consequently infecting the uterine environment of cows.     

Reduced formation of granulomata in gamma(delta) T cell knockout BALB/c mice inoculated with Mycobacterium avium subsp. paratuberculosis

The role of gamma(delta) T cells in the bovine immune response to Mycobacterium avium subsp. paratuberculosis (M. paratuberculosis) infection is poorly understood. Accordingly, using BALB/c mice that are innately susceptible to M. paratuberculosis, we compared wild-type and gamma(delta) T cell knockout BALB/c mice to study the protective roles of gamma(delta) T cells in M. paratuberculosis infection. Ten-week-old mice were inoculated intraperitoneally with either a low dose (4 x 10(6) colony-forming units [CFU]/mouse) or a high dose (4 x 10(9) CFU/mouse) of M. paratuberculosis strain ATCC 19698. Histopathologic and morphometric examinations showed reductions in the number and area of granulomatous lesions in the liver of the knockout mice at 18 weeks after inoculation with either the low or the high dose of the mycobacteria. Furthermore, at 18 weeks after inoculation, the bacterial load in the spleens of the knockout mice inoculated with the high dose was significantly lower than that of wild-type mice. No differences were found in bacterial load between the knockout and the wild-type mice in the low-dose groups. In contrast, in the livers of wild-type mice inoculated with either the low or high mycobacterial dose, increased areas of epithelioid granulomata were observed and the granulomata became disseminated widely during the experimental period. These findings in model mice suggest that gamma(delta) T cells, rather than restricting mycobacterial growth, may play a crucial role in development of epithelioid granulomata similar to those seen consistently in bovine paratuberculosis. The results of this study may have relevance to our understanding of the pathogenesis of paratuberculosis in ruminants, in which a prominent number of gamma(delta) T cells exist in the lymphoid system.     

Body water content and turnover in cats fed dry and canned rations.

Body water content and total body water turnover in cats fed commercially available dry food and then given canned rations were determined with tritiated water. Cats during the feeding of either ration did not differ in body water content or turnover. Cats during the feeding of the dry ration derived a greater fraction of their total water turnover from drinking water, and these cats drank more water per gram of dry matter intake than when fed the canned ration. On the basis of total water intake, however, those given the canned ration had significantly greater water intake per gram of dry matter; also, their total water turnover per gram of dry matter was greater. Compared with other animals, cats have a similar ratio of body water to body solids, but the rate of water turnover per unit of body weight is slower.     

Intrauterine and transmammary transmission of Mycobacterium avium subsp paratuberculosis in sheep

OBJECTIVE: To investigate intrauterine infection of foetuses with Mycobacterium avium subsp paratuberculosis and the presence of infection in mammary secretions of sheep. DESIGN: A study of 142 late-pregnant ewes and their foetuses from two heavily infected flocks. PROCEDURE: Infection of ewes was determined at necropsy by histopathology and culture of tissues and mammary secretions. Antemortem tests (clinical assessment, faecal culture and serology) were also applied. Foetuses from 59 infected ewes and 47 apparently uninfected ewes were examined by culture and histopathology. RESULTS: Five of five ewes with clinical ovine Johne's disease had infected foetuses. Only one of 54 subclinically affected ewes, and none of 47 uninfected ewes had an infected foetus. M a paratuberculosis was cultured from mammary secretions or mammary glands of only two of 76 ewes, both of which were clinical cases and had infected foetuses. CONCLUSION: Although intrauterine or transmammary transmission of Mycobacterium avium subsp paratuberculosis may occur frequently in clinically affected sheep, these are less common in subclinically infected ewes. Therefore these modes of transmission are unlikely to compromise existing control programs for ovine Johne's disease on most farms, especially if programs include the immediate culling of clinically affected sheep.