Heterologous Expression of a Gene Encoding a 35 kDa Protein of Mycobacterium avium paratuberculosis in Escherichia coli

The full-length open reading frame coding for a potentially immunogenic 35 kDa protein of Mycobacterium avium paratuberculosis was generated using polymerase chain reaction technology. The gene was inserted in-frame into Escherichia coli expression plasmid pQE32. The resulting recombinant plasmid pPMP35 was transformed into E. coli M15. Analysis of the E. coli induced with isopropyl-beta-D-thiogalactopyranoside revealed that the protein accumulated into the cytoplasm as insoluble inclusion bodies. The level of expression of the recombinant 35 kDa protein (P35) was more than 30% of the total protein of E. coli cells. Expression of the recombinant protein was confirmed by immunoblotting. The P35 reacted with a rabbit antiserum raised against a sonicate of M. a. paratuberculosis. The protein was also recognized by serum from a goat with clinical paratuberculosis. Further, a polyclonal antiserum against P35 recognized a 35 kDa band in a membrane fraction of M. a. paratuberculosis. Also, the protein provoked a significant skin reaction in outbred guinea pigs sensitized with M. a. paratuberculosis, as well as in those sensitized with Mycobacterium avium. The results indicate that the 35 kDa protein of M. a. paratuberculosis is a membrane protein, having a role in the cellular immune response.     

Comparison of serological tests and faecal culture for the detection of Mycobacterium avium subsp. paratuberculosis infection in cattle and analysis of the antigens involved

Three hundred and forty-one sera from cattle in Western Australia and 106 sera from Mycobacterium paratuberculosis faecal culture positive cattle were used to evaluate the performance of two absorbed enzyme-linked immunosorbent assays (ELISA) (one locally produced, the other a commercial test) and a complement fixation test (CFT) for the detection of Johne's disease in cattle. The diagnostic sensitivity (47.2%) of the local ELISA was significantly higher than that of the commercial ELISA (31.1%), and significantly higher than that for the complement fixation test (17.9%) and immunoblot (20.8%). Diagnostic specificity for the two ELISAs was 99.7% and 97.9% and similar for CFT and immunoblot (97.1% and 97.7%, respectively). The diagnostic sensitivity rose for both ELISAs and the CFT as the number of M. paratuberculosis isolated from the faeces increased. The ELISA antigen was characterised by polyacrylamide gel electrophoresis and electrophoretic immunoblotting and was found to consist mostly of a carbohydrate-type macromolecule of 32-42 kDa. This macromolecule was identified as lipoarabinomannan (LAM) by using a LAM-specific monoclonal antibody in immunoblots and purified LAM in absorption experiments. By applying more complex antigen preparations in immunoblots, serum antibodies against proteins of 47, 37, 30, 24 and 21 kDa, and against the 32-42 kDa carbohydrate component were frequently found in infected cattle, and of these the 47 kDa protein and the 32-42 kDa antigen were immuno-dominant. Pre-absorption of the sera with M. phlei sonicate indicated that the protein antigens contributed markedly to non-specific serological cross-reactions, while the 32-42 kDa non-protein macromolecule appeared to be specific.     

Heterogeneity of Hanganutziu-Deicher antigen glycoproteins in different species animal sera.

A heterophilic Hanganutziu-Deicher (HD) antigen is present in many animal sera except human and chicken sera. To visualize the antigenic molecules, nine species animal and human sera were analyzed by SDS-PAGE, followed by Western blotting and immunostaining with avain anti-N-glycolylneuraminyl-lactosyl ceramide antibody which recognizes the terminal N-glycolylneuraminic acid moiety of glycoconjugates as an epitope of the HD antigen. Several HD antigen-active glycoprotein bands were detected in the sera of fetal calf, calf, horse, goat, monkey, rabbit, guinea pig, rat and mouse, except for human serum. The HD antigenic proteins showed heterogeneities in their molecular weights and were not identical with any major band visualized with silver-staining, indicating that they are minor components of serum proteins in each animal. Neuraminidase treatment destroyed the antigenicity of all proteins, confirming that N-glycolylneuraminic acid (NeuGc) at the non-reducing terminal of carbohydrate chains is the antigenic epitope in serum glycoprotein molecules as already confirmed in glycosphingolipid (GSL) antigens. The finding of HD-antigenic glycoproteins in animal sera suggests that they also stimulate HD antibody production in patients who received animal antiserum for therapeutic aim.     

Isolation of specific peptides from Mycobacterium paratuberculosis protoplasm and their use in an enzyme-linked immunosorbent assay for the detection of paratuberculosis (Johne's disease) in cattle

An antigen was isolated from the protoplasm of Mycobacterium paratuberculosis by a combination of gel filtration, ion exchange, and affinity chromatography. The purified antigen constituted 7.8% of the total protein in the protoplasm. The specificity and sensitivity of the enzyme-linked immunosorbent assay (ELISA) for paratuberculosis, using the purified antigen, were evaluated with sera from 104 cattle which were examined (surveyed) for M paratuberculosis infection by fecal cultural technique. The ELISA was positive in 50 of 60 infected animals. Five of 44 noninfected animals were also test-positive. When a crude protoplasmic extract was used as antigen in the ELISA, sera from 37 infected and from 18 noninfected animals were test-positive. Cross-reactions were encountered in both complement-fixation test and the ELISA between crude or partially purified M paratuberculosis antigens and antisera to Nocardia asteroides, M avium, M phlei, and M fortuitum. The purified antigen gave no complement-fixation reaction with any of these antisera. In the ELISA, cross-reaction was not found when purified antigen was used and the sera were screened at 1:40 dilution.     

Protein antigens secreted by Mycobacterium paratuberculosis.

Proteins secreted by Mycobacterium paratuberculosis (M.ptb) during short-term cultivations were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western(Immuno) blotting. Cultivation in a defined medium containing 35S methionine allowed autoradiographic detection of proteins which had been secreted or passively released by actively metabolizing M.ptb organisms. After the first 3 days of cultivation, 4 proteins with molecular weights of approximately 38, 50, 65 and 110 kilodaltons (kd) were detected on SDS gels. Longer incubation up to 12 days resulted in an increased concentration of these proteins as well as in appearance of additional proteins ranging from 14 to over 90 kd. In long-term (8-10 weeks) culture filtrates only two prominent proteins with molecular weights of 30 and 65 kd proteins could be detected. Immunoblot analysis showed that some of the proteins secreted during short-term cultivations were recognized by sera from M.ptb-infected sheep and more significantly by sera from animals which had been immunized with a M.ptb live vaccine strain. The study indicates that during short incubation times M.ptb may secrete immunoreactive proteins which are not dominant in long-term cultures.     

HspX is present within Mycobacterium paratuberculosis-infected macrophages and is recognized by sera from some infected cattle

A portion of the gene encoding HspX has been previously identified as a sequence specific to Mycobacterium avium subspecies paratuberculosis (hereafter referred to as M. paratuberculosis) based on DNA hybridization experiments. In this study, rabbit antisera were raised against a recombinant protein of HspX fused to the Escherichia coli maltose binding protein (MBP/HspX). Immunoblots of lysates of M. paratuberculosis-infected macrophages probed with the rabbit antisera showed that HspX was present within infected macrophages of bovine and murine origin. This observation was confirmed by immunofluorescence microscopy of infected macrophages. Lysates of E. coli expressing HspX without the MBP fusion partner were loaded onto preparative SDS-PAGE gels and used to determine whether infected cattle generated a humoral immune response to the antigen. Sera from four of 24 paratuberculous cows (17%) detected HspX. No reactivity was present in sera from control cows. While HspX may be immunogenic during infection in some cows, the protein is not secreted and it does not stimulate cell-mediated immunity. Collectively, these data give a preliminary characterization of the first described M. paratuberculosis protein identified within infected macrophages.     

Antigenic analysis of Pasteurella haemolytica serovars 1 through 15 by crossed immunoelectrophoresis

Pasteurella haemolytica serovars 1 through 12, grown in broth and on agar plates, and 2 field isolates (types A1 and T10) were used to develop polyvalent crossed immunoelectrophoresis (XIE) reference systems. The maximal number of antigens was revealed by XIE when sonicates of agar plate-grown organisms were used as the immunogen (to produce antibodies) and as the soluble antigen for XIE. Antigens produced from agar plate-grown organisms were less contaminated (by antigenic components of the medium) than were those produced from organisms grown in broth. Seventy-two antigens were detected in sonicated preparations of agar plate-grown P haemolytica. The common antigen of gram-negative bacteria was identified in the P haemolytica XIE reference system; precipitation was observed with rabbit antiserum to the common antigen of gram-negative bacteria isolated from Escherichia coli, as well as with rabbit immunoglobulins (obtained from unvaccinated rabbits). Most preimmune sera from our vaccinated rabbits also precipitated the common antigen. Serovar-specific antigens in the P haemolytica XIE reference system were defined and presumptively identified as part of the bacterial lipopolysaccharide complex by use of the limulus amebocyte lysate test. Partial cross-reactions were found between serovar-specific antigens within each biovar (A and T). Pasteurella haemolytica biovar A-specific and biovar T-specific antigens were defined by crossed-line immunoelectrophoresis. When serovars A13, A14, and T15 were tested in the P haemolytica XIE reference system, they gave high matching coefficient values of 0.98, 0.98, and 0.87, respectively. The proposal to separate P haemolytica biovars A and T into 2 different species was supported by immunotaxonomic data obtained from crossed immunoelectrophoresis, but more extensive studies will be necessary to establish the appropriate taxonomic position of these 2 groups of organisms.     

Identification and partial purification of soluble antigens from culture-grown Besnoitia besnoiti endozoites

Soluble antigens from culture-grown Besnoitia besnoiti endozoites were identified following their partial purification by affinity chromatography. A specific eluate obtained after affinity chromatography on a column to which antibodies from serum of a naturally infected cow were bound exhibited seven polypeptide bands on sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE). Five bands were observed in the eluate from an immunoadsorbent to which antibodies from an experimentally infected calf were coupled. Eluted antigens were reactive in the enzyme-linked immunosorbent assay (ELISA). Reactivity of electrophoretically resolved antigens with sera of endozoite-inoculated cattle and sera from field cases of besnoitiosis were studied using the Western immunoblotting technique.     

Characterization of Brucella canis protein antigens and polypeptide antibody responses of infected dogs

The cytoplasmic protein antigens (CPAg) of Brucella canis were characterized by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and analysis of 35S-labeled polypeptides. Approximate molecular weights of the immunoreactive polypeptides were determined by migration patterns of the immunoprecipitated polypeptides after SDS-PAGE or Western immunoblotting of sera collected at various times after experimental infection of dogs. Polypeptides were specifically precipitated by sera of infected dogs, but not from the sera of normal or false-positive (seropositive, non-infected) animals. During the initial month after infection, proteins with molecular weight masses (MW) of approximately 18, 22, 31, 42 and 54 kDa were commonly recognized. A 20-kDa polypeptide was first recognized at 8-10 weeks after infection, but it was detected inconsistently after 6 months. Additional polypeptides detected from 2 to 12 months post-infection had MW of 22, 66-68 and, less regularly, 42, 60, 82, 100 and greater than 200 kDa. The polypeptides most consistently recognized in sera from B. canis-infected dogs had MW of 18, 22 and 68 kDa.     

Secreted antigens of Mycobacterium avium subspecies paratuberculosis as prominent immune targets

We here describe the identification and characterization of three novel secreted Mycobacterium avium subsp. paratuberculosis antigens of 9, 15 and 34 kDa (Map2609, Map2942c and Map0210c, respectively) by screening a genomic expression library with a serum of a naturally infected clinical cow. The 9, 15 and 34 kDa antigens display strong homology to previously described M. tuberculosis antigens, TB8.4, MPT53 and Erp, respectively. Furthermore, these antigens were shown to be recognized by antibodies from infected cattle, when tested with a limited number of sera from subclinical (n=7) and clinical (n=3) infected cattle.     

Paratuberculosis in cattle and free-living exotic deer.

Paratuberculosis was studied among dairy cows and exotic deer that shared grazing areas at Point Reyes National Seashore, California. Of the 10 dairy herds tested, 5 (50%) were infected with Mycobacterium paratuberculosis (based on results of fecal culture). Mycobacterium paratuberculosis was cultured from 9 (8.7%) of the 103 bovine fecal samples and from 4 (3.9%) of the 103 bovine rectal mucosa scapings tested. Of 89 fecal samples from 52 axis deer (Axis axis) and 37 fallow deer (Dama dama), 5 (9.6%) and 3 (8.1%), respectively, contained M paratuberculosis. Culture of intestinal necropsy samples from the same deer indicated that 3 (5.8%) of the axis deer and 2 (5.4%) of the fallow deer were infected with M paratuberculosis. The cows were tested for serum antibodies by the complement-fixation test and by radioimmunoassay. Of 95 sera tested by complement fixation, 15 (15.8%) were positive, as were 15 (14.7%) of 102 sera tested by radioimmunoassay. Culture results and serologic test results were compared on a herd basis.     

Specific seroreactivity of Crohn's disease patients against p35 and p36 antigens of M. avium subsp. paratuberculosis

Crohn's disease (CD) is a chronic inflammatory bowel disease that is similar to Johne's disease in ruminants. Recent data have strengthened the association of M. avium subsp. paratuberculosis (M. paratuberculosis) with CD. To provide more evidence of an etiological association, antibody reactivities from CD patients were tested by immunoblotting against recombinant antigens that were identified previously from our M. paratuberculosis genomic library. Two clones (designated pMptb#40 (3.2-kb insert) and #48 (1.4-kb insert) expressing a 35K (p35)- and 36K(p36)-antigens showed specific reactivities with serum samples from CD patients.Serum samples from 75% of 53 CD patients, 14% of 35 normal individuals and 10% of 10 ulcerative colitis patients reacted to p35 antigen. Reactivities were also observed with serum samples from 89% of 89 CD patients, 14% of 50 normal controls and 15% of 29 ulcerative colitis patients reacted with p36 antigen. When the reactivity results from p35 and p36 were combined, the background from the controls was eliminated, i.e. only the CD patients reacted to both p35 and p36. The positive predictive value was 98% with specificity of 98% and the negative predictive value was 76% with sensitivity of 74% (39 positive out of 53). A statistical significance (p<0.0001) was observed when the results from CD serum samples reacting with either or both antigens were compared to the controls. The reactivity of anti-M. paratuberculosis (specifically against p35 and p36 antigens) antibodies in a significant proportion of CD patients would suggest a causal role for the organism in CD.     

Characterization of antigens from mycoplasmas of animal origin

Alcholeplasma laidlawii, Mycoplasma gallisepticum, M mycoides subsp mycoides, M agalactiae, M bovirhinis, mycoplasmal strain ST-6, and culture medium were compared with M bovis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, enzyme-linked immunosorbent assay (ELISA), and gel electrophoresis-derived ELISA. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis indicated there were areas of homology and areas of heterology among the species tested. Sera from rabbits hyperimmunized with the mycoplasma organisms and noninoculated culture medium demonstrated ELISA reactivity with M bovis antigens immobilized on polystyrene. Absorption of the serum from a rabbit hyperimmunized with M bovis reduced 65.9% of its reactivity with culture medium, 29.7% to 32.7% of its reactivity with the heterologous species, and 21.1% of its reactivity with the homologous species. Gel electrophoresis-derived ELISA performed on immobilized M bovis antigens separated by molecular weight, using sera from rabbits hyperimmunized with the mycoplasmal species under study and noninoculated culture medium revealed antigenic components which are shared among species or with the culture medium and several components which may be unique to M bovis.     

Mucosal immune response in cattle with subclinical Johne's disease

Mycobacterium avium subsp. paratuberculosis is the causative agent of Johne's disease, a chronic granulomatous enteritis of wild and domestic ruminants. During a long subclinical period, the organism persists in the intestine despite systemic cellular and humoral immune responses. To explore the mucosal immune response in Johne's disease, we isolated mononuclear leukocytes from the ileum of cows naturally infected with M. avium subsp. paratuberculosis and from cows that were not infected. We evaluated the immunophenotype of these cells and the proliferative responses after the addition of M. avium subsp. paratuberculosis sonicate or B-cell or T-cell mitogens. Although the percentage of T cells was increased in infected cows, these cells consisted mostly of memory (CD2+CD62L-) and regulatory (CD4+CD25+) T cells. Further evidence of immune hyporesponsiveness included a decrease in the percentage of T cells with an activated phenotype and a decrease in cells expressing major histocompatibility factor class II (MHC class II). Unlike the spleen, ileal lymphocytes from infected cows failed to proliferate in response to M. avium subsp. paratuberculosis sonicate. Additionally, ileal lymphocytes from infected cows proliferated poorly in response to concanavalin A and pokeweed mitogen, suggesting generalized T cell and B cell hyporesponsiveness. These results indicate that a state of tolerance may exist in the intestine of cows subclinically infected with M. avium subsp. paratuberculosis organisms in subclinically infected cows. This effect may be induced, at least in part, by proliferation of regulatory T cells that nonspecifically suppress mucosal immune responsiveness.     

Serodiagnosis of bovine paratuberculosis by use of a dot enzyme-linked immunosorbent assay

A dot ELISA was developed for detection of antibodies to Mycobacterium paratuberculosis. The assay was evaluated by testing sera from cattle that were determined, by bacteriologic culturing of feces, to be infected with M paratuberculosis and were suspected of having clinical disease. Further evaluation involved testing sera from cattle in which M paratuberculosis had not been isolated from feces on several attempts. Results of the dot ELISA were positive for sera from 86 of 101 infected cattle, and results were negative for sera from 64 of 64 noninfected cattle. Results of conventional ELISA and agar gel immunodiffusion (AGID) tests were positive for 79 of 99 and for 51 of 101 infected cattle, respectively. The dot ELISA also was evaluated by comparing results of testing 708 sera with results of bacteriologic culturing of matched fecal samples from 262 cattle in 3 central Ohio dairy herds known to include cattle infected with M paratuberculosis. Results of the dot ELISA were positive for 25 of 39 sera from cattle with positive results on culturing of concurrently obtained fecal specimens. The dot ELISA results were negative for 661 of 669 sera from cattle with negative results to culturing of concurrently obtained fecal specimens. The 39 sera from cattle with positive results on bacteriologic culturing of matched fecal specimens had positive results for ELISA and the AGID test 25 and 14 times, respectively. The 669 sera from cattle with concurrently negative results on bacteriologic culturing of feces had negative results to ELISA and the AGID test 559 and 668 times, respectively.     

Detection of stable diagnostic antigen from bile and feces of Fasciola hepatica infected cattle.

Diagnostic antigens in bile and feces from Fasciola hepatica infected cattle were detected and characterized by enzyme-linked immunotransfer blot (EITB) techniques. As sources of antigen, samples of bile, intestinal contents and feces were collected from five uninfected calves and from 10 calves with known Fasciola hepatica burdens. A band detected by EITB using a densitometer in the area corresponding to 26 kDa reacted with rabbit anti-fresh fluke antigen and infected cattle sera but not with fluke-negative rabbit sera, rabbit anti-Fasciola hepatica egg sera, Fascioloides magna positive or negative cattle sera. This band was not detected by Coomassie blue in sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE) gels or by Ponceau-S stained nitrocellulose strips. Band groups located at 104-66, 66-42, 42-26 and 25-16 kDa reacted inconsistently with the above sera. Sera from mice hyperimmunized with Fasciola hepatica excretory-secretory (ES) products detected only the 26 kDa band by EITB, without cross-reactivity with bands in the other molecular weight (MW) ranges. The results suggest that the 26 kDa antigen may consist of a stable component of ES products and/or tegument-related worm antigen. Diagnosis of Fasciola hepatica through detection of specific, stable antigens in feces of infected animals offers potential advantages over serum-based tests of better sample accessibility, discrimination between previous and current infections, and possible semi-quantitation of fluke burdens.     

Antigen specificity of antibody responses to Corynebacterium pseudotuberculosis in naturally infected sheep with caseous lymphadenitis.

Constituents of Corynebacterium pseudotuberculosis were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Analysis of sonicated whole bacterial cells and ether-extracted cells revealed more than 35 bands in silver-stained gels. SDS-PAGE analysis of concentrated culture filtrates with exotoxin activity demonstrated more than 15 bands. Sera from sheep with C. pseudotuberculosis-induced disease of variable severity were used to probe immunoblots of electrophoresed ether-extracted cells and culture filtrates. Twenty or more corynebacterial molecules, ranging in molecular weight from 20 to 112 kDa, in ether-extracted cells were recognized by antibodies in the sera of naturally exposed sheep with positive ELISA titers. These sera also recognized up to six molecules, ranging from, 20 to 68.1 kDa, on immunoblots of ammonium sulfate-concentrated culture filtrate. There was no apparent relationship between the stage of disease and the response to specific corynebacterial antigens in these animals.     

Development and initial evaluation of an indirect enzyme-linked immunosorbent assay for the detection of Leptospira interrogans serovar hardjo antibodies in bovine sera.

Outer sheath antigen from Leptospira interrogans serovar hardjo type hardjoprajitno and acetic acid extracted antigens from serovar hardjo types hardjoprajitno and hardjobovis were evaluated in an immunoassay for ability to detect hyperimmune rabbit serum to serovar hardjo. The degree of cross-reactivity with hyperimmune rabbit sera to L. interrogans serovars pomona, copenhageni, grippotyphosa, canicola and sejroe, and Leptospira biflexa serovar patoc was also measured for each antigen. All of the antigens reacted with the antiserum to L. interrogans serovar hardjo. The outer sheath antigen however, also showed wide cross-reactivity with the antisera to all of the serovars of L. interrogans tested and with the antiserum to L. biflexa serovar patoc. The acetic acid extracted antigen from either type hardjoprajitno, or type hardjobovis, showed a high degree of specificity for serovar hardjo antiserum. The hardjobovis acetic acid extracted antigen was characterized by sodium dodecyl sulfate polyacrylamide gel electrophoresis and immunoblotting, and was incorporated into an indirect ELISA for detection of anti-serovar hardjo antibodies in bovine serum. This ELISA showed a relative specificity of 100% with 156 bovine sera which were negative at a dilution of 1:100 in the microscopic agglutination test (MAT) for L. interrogans serovars hardjo, pomona, sejroe, icterohaemorrhagiae, copenhageni, canicola, and grippotyphosa. The relative sensitivity of this assay with 192 bovine sera which had serovar hardjo MAT titres of > or = 100 was 95.3% (95% confidence limit = 2.99%). The degree of cross-reactivity with 289 bovine sera which had serovar pomona MAT titres of > or = 100 (with no detectable serovar hardjo MAT titres) was approximately 1.0%. This assay was: easily standardized, scored objectively, repeatable, semi-automated and used a non-hazardous antigen that can be routinely prepared in gram amounts.     

Isolation and characterization of immunoglobulin binding proteins from Staphylococcus intermedius and Staphylococcus hyicus.

125-I-IgG binding activities were observed with 15 (17%) of 90 S. intermedius isolates from dogs and 39 (95%) of 41 S. hyicus isolates from pigs. Binding activities were not detected with S. hyicus isolates from cows. The IgG binding proteins of 2 S. intermedius, 2 S. hyicus, and protein A from S. aureus Cowan I were isolated from their cell surfaces. The proteins precipitated with IgG preparations from human, rabbit, pig, dog and horse, but not with IgG from cow, mouse and chicken. This indicated that these IgG binding proteins could be classified as type I receptors. In addition, the isolated proteins from all 3 staphylococcal species precipitated with polyclonal chicken anti-protein A antiserum. SDS-PAGE, Western blotting and gel isoelectric focussing of the proteins revealed numerous bands in the 42,000 D range and acid isoelectric points. The isoelectric point of the isolated proteins from both S. intermedius cultures was slightly more acidic than those from S. hyicus and S. aureus. The present results indicate a close functional and antigenic similarity, if not identity, between IgG binding proteins of S. intermedius and S. hyicus, and protein A of S. aureus.     

Analysis of antigens in Mycobacterium paratubercuolsis.

Analysis of antigens in Mycobacterium paratuberculosis. Acta vet. scand. 1979, 20, 200–215. — Using crossed immunoelectrophoresis (GIE) and crossed line immunoelectrophoresis (GLIE), antigens from different strains and variants of Mycobacterium paratuberculosis were compared, and cross-reactions between 1 of these strains and Mycobacterium avium and BGG studied. In each of 4 bovine laboratory strains of M. paratuberculosis examined, altogether 44 different antigens were demonstrated. This is the largest number of antigens in M. paratuberculosis which has been described so far. No important difference in the antigenic structure of the strains was found. The 4 laboratory strains are being used routinely in the production of vaccine against Johne’s disease in Norway and Iceland. One of the aims of the present work was to investigate the antigenic relationship between these strains and the goat-pathogenic Norwegian and the Icelandic variant of M. paratuberculosis. Out of 44 different antigens demonstrated in the laboratory strains, 39 and 31 gave cross-reactions against the Norwegian and the Icelandic variant, respectively. This is in accordance with practical experience, as the results of vaccination against Johne’s disease, performed in Norway for many years, are very good.Twenty-seven and 24 cross-reacting antigens between M. paratuberculosis and strains of M. avium and BGG, respectively, were observed. This finding agrees with clinical observations.Another aim of the investigation was to identify species-specific antigens as regards M. paratuberculosis. One antigen showed a marked cross-reaction between the strains of M. paratuberculosis examined, but did not react with antisera against M. avium and BGG. Some other antigens showed partial specificity.The results obtained stress the complicated antigenic situation in mycobacteria which is of decisive significance as regards the diagnosis and classification of mycobacterial infections.