EP1135523A1 - Detection of mycobacterium avium subspecies - Google Patents

Abstract

The invention relates to the field of microbiology, more specifically to the field of detection and identification of pathogenic micro-organisms, more specifically to the detection and identification of Mycobacterium avium subspecies paratuberculosis, an agent which causes Johne's disease in many ruminants. The invention provides a method by which a rapid and sensitive procedure for the detection and identification of Mycobacterium avium subspecies paratuberculosis in biological and natural samples is achieved, whereby Mycobacterium avium subspecies paratuberculosis can be discriminated from other Mycobacterium avium subspecies, such as subspecies avium.

Description

Title: Detection of Mycobacterium avium subspecies.

The invention relates to the field of microbiology, more specifically to the field of detection and identification of pathogenic micro-organisms, more specifically to the detection and identification of Mycobacterium avium subspecies para tuberculosis, an agent which causes Johne ' s disease in many ruminants.

Mycobacteria are aerobic, Gram-positive, acid-fast rod-shaped bacteria (0.2-0.6 x 1.0-10 μm) . The genus as a whole is characterised by long-chain mycolic acids in the cell-wall. Large amounts of lipids in the cell-wall result in remarkable resistance to de-staining of stained cells, which forms the basis of the Ziehl-Neelsen staining used to identify Mycobacteria. In general, Mycobacteria are slow- growing organisms. Mycobacterium avium subspecies para tuberculosis is ranked as an extremely slow growing organism. Many attempts have been made to improve the cultivation conditions. Further possibilities to significantly improve the growth rate of the bacterium in vitro are virtually absent. The taxono ic position of Mycobacterium avium subspecies para tuberculosis is defined on 16S ribosomal RNA sequence data (1) . The bacterium is a member of the Mycobacterium avium complex (MAC) . This complex is composed of Mycoba cterium avium, represented by the three subspecies avium, silva ticum and para tuberculosis , and of Mycobacterium intracell ulare (2) .

Mycobacterium avium subspecies para tuberculosis is, among others, characterised from the other bacteria in this complex by its requirement for mycobactin in the growth medium (3) . Paratuberculosis or Johne ' s disease, is a world-wide occuring disease caused by the bacterium Mycobacterium avium subspecies para tuberculosis . An estimated 30-40% of the cattle farms in The Netherlands is affected by this disease and approximately 2% of all cows in The Netherlands is infected with this bacterium. The total damage as a result of the disease (loss of milk production, mortality and restrictions in trade) is estimated at 45 million Dutch guilders per annum with a damage of 50 to 300 guilders per cow per annum on the affected farms. At this moment, infected animals generally are, once detected, culled or removed from the farm and generally destroyed.

Johne ' s disease is thought to be related to the inflammatory bowel syndrome in humans, also known as Crohn ' s disease. If the causal agent of Johne' s disaese is indeed also a casual of inflammatory bowel syndrome in humans, the disease may well be placed on the list of animal to human transmissible diseases or zoonoses.

In the mid-seventies a campaign was launched to eradicate the disease. This campaign was not successful mainly due to limited diagnostic possibilities. New attempts are initiated in which the efforts are directed towards identification, certification and canalisation of paratuberculosis-unsuspected farms, and detection and isolation of affected cattle and eradication of paratuberculosis on infected farms. In the latter programme both management measures and diagnosis play an important role. A solid diagnostic test for the detection of Mycobacterium avium subspecies para tuberculosis will play a pivotal role in such a campaign.

The detection of cattle infected with Mycobacterium avium subspecies para tuberculosis is hampered by two important factors: i) The clinical symptoms of illness are first detectable after an incubation period of four to five years. Due to the slow progression of the infection and slow development of an immune response, testing before the second year in the life of a cow is therefore in general not even considered useful, ii) The bacterium itself propagates extremely slow as a result of which the only currently reliable test to identify the bacterium - a conventional cultivation of faecal samples - requires up to 6 months. Furthermore, testing only once does not suffice. A minimal sampling rate of two times per year per farm is necessary to ensure effectiveness of the eradication program. In current methods, 4 separate cultures per faecal sample are monitored for growth over a period of up to half a year. This results in significant accumulation of cultures. Large-scale cultivation testing programs are therefore hampered by problems of capacity. Moreover, cultivation methods are frustrated by problems of cantamination with fungi and/or Bacillus species. The Netherlands alone harbours 30.000 cattle farms with a total of approximately 800.000 cows. If every cow in The Netherlands were to be tested by cultivation methods several millions of faecal sample cultures would have to be stored for up to half a year at 37° C.

Recent years have seen various attempts to develop alternative and more rapid methods of detection. Unfortunately, present alternative methods are considered unsuitable for large scale testing programs. Most contemporary diagnostic tests are simply not sensitive enough to reliably detect Mycobacterium avium subspecies para tuberculosis in faecal samples, also serological methods, albeit suitable for large scale screening programmes are nor sensitive enough.

Currently, there are at least 8 different diagnostic tests for the detection/demonstration of infection of Mycobacterium avium subspecies para tuberculosis . Three of these are direct methods for the detection of Mycobacterium avium subspecies para tuberculosis cells. The other five are tests that demonstrate the presence of an immune response by the host-animal. In general, direct methods of detection are considered superior with respect to reliability. The three tests used for direct detection of the bacterium are i) conventional cultivation (4), ii) BACTEC (radiometrical or fluorimetrical) cultivation (5) and iii) DNA probe testing (PCR) (6,7,8). The detection limit for conventional cultivation is approximately 100 bacteria per gram of faeces. As a result, conventional cultivation is still one of the most sensitive methods of detection. The BACTEC system for the detection of Mycobacterium avium subspecies para tuberculosis is a revision of the BACTEC system for the detection of M. tuberculosis in humans (Becton Dickinson Diagnostics) that is used in general hospitals. For paratuberculosis testing, the BACTEC growth medium is supplemented with specific components to allow growth of Mycobacterium avium subspecies para tuberculosis in the cultivation tubes. The method is more rapid (7 weeks) and more sensitive than conventional cultivation, but is also much more expensive. In order to apply this method on a large scale (1.000 samples/day) approximately 300 BACTEC incubators are required. This makes BACTEC cultivation, be it radiometric or fluorometric, not feasible. The 5 tests that are used to detect an immune response against the bacterium Myb . avium subsp. para tuberculosis are i) the complement fixation (CF) test, ii) agar gel immunodiffusion (AGID) test (specific for sheep) , iii) ELISA test on blood-serum (USDA-licensed) , iv) ELISA on milk, v) gamma interferon test (USDA-licensed) .

Immunological tests (among which are the serological and cellular immune-mediated assays) for the detection of paratuberculosis are generally not considered satisfactory with respect to sensitivity and specificity, especially in the case of diagnosis in sub-clinically infected animals. Often in serological tests, false-positive results are found due to cross-reactivity with similar antigens in other mycobacteria or related organisms. False-negative reactions occur in the case of so-called "serological non-responders"; animals in the terminal stage of the disease are often anergic. Also, antibody responses develop very slowly. Before three years of age, a very limited percentage of infected animals develop a humoral immune response, and some do not develop an immune response at all.

The Idexx Laboratories ELISA test on blood serum exhibits a specificity (percentage animals that is free of infection and tests negative) of 99% and a relative sensitivity (percentage of M. para tuberculosis-infected animals that tests positive) of 45% in "sub-clinically" infected animals (in this case cultivation results are taken as 100%) . In the clinical stage of infection the relative sensitivity increases to approximately 85%. Only animals of age 20 months and older can be tested reliably for paratuberculosis with this test. In younger animals the test is too insensitive unless the animals show clinical signs of infection. The interpretation of the test results is quantitative (increase in optical density in the ELISA) and is proportional to the antibody density in the animal's blood. High scores are strongly indicative of infection and possible shedding of bacteria in faeces and milk. A confirmation test (re-testing) must occur within 6 to 12 weeks. The test can, however, not be applied to animals that have received vaccination against Johne ' s disease as this results in false-positive test results. The system is suitable for large scale testing but cannot be considered as an indisputable evidence of infection. The detection of gamma interferon is a cellular immune response assay. The method for determination of interferon release by white blood-cells as a result of Johnin-PPD stimulation is available as a diagnostic kit licensed by the US Department of Agriculture (USDA) . Since the release of interferon precedes the formation of antibodies during pathogenesis of paratuberculosis, infection can be detected at an earlier stage compared to other immunological tests. However, the sampled blood that is treated with heparin to inhibit blood coagulation must be handled with care and tested within 12 hours after collection. Therefore, the interferon-assay is only performed on appointment and careful co-ordination between veterinary practitioner and testing laboratory is required. As a result, the test is not suited for large-scale testing.

The use of nucleic acid probe-tests for the detection of Myb avium subsp. para tuberculosis has, albeit at first sight a possible alternative for conventional cultivation, proven to be practically quite unfeasible. For one, this is due to the extreme nucleic acid homologies that exist between Myb avium subsp. para tuberculosis and other Myb avium subspecies, which are commonly found in samples such as feacal samples of ruminants (McFadden, J.J., et al. 1987b. J. Gen. Microbiol. 133:211-213; Saxegaard, F., and I. Baess. 1988. Acta Pathol. Microbiol. Immunol. Scand. 96:37-42;

Yoshimura, H.H., and D.Y. Graham. 1988. J. Clin. Microbiol. 26:1309-1312). For another reason, this is due to the low effective level of detectable nucleic acid due both to the extreme low cell densities of the target organism, the presence of a large excess of non-target organisms, as well as inhibition of enzymatic nucleic acid amplification reactions by inhibitory substances in the faecal matrix (Van der Giessen, J. .B., et al. 1992. J. Clin. Microbiol. 30(5) :1216-1219; Millar, D.S., et al.1995. Anal. Biochem. 226(2) :325-330) .

In general, nucleic acid-based tests can be directed towards (genomic) DNA or its RNA transcript. Two problems with using genomic DNA as target in DNA amplification methods for diagnostic purpose in general are the low copy number of target molecules per cell and the relative unpredictability of natural sequence variability within the total population of target organisms. More or less related organisms may suddenly show to possess very similar or homologous DNA sequences (Kunze, Z.M., et al . 1991. Mol . Microbiol. 5 (9) :2265-2272; Moss, M.T., et al. 1992. J. Gen. Microbiol. 138:139-145). Such unpredictability limits the reliability through the specificity. The specificity of a nucleic acid based testing .system is determined by the presence of the specific nucleic acid target sequence for hybridisation or the proper identification of the target sites by the probes or primers. The specificity of the test is therefore determined by the genetic variation in the target sequences over the total population of the bacterium to be detected. Knowledge about this variation is therefore important.

Mycobacteria are difficult to differentiate on the basis of their genetic makeup due to the presence of exceptional sequence conservation among species (Frothingham, R., et al. 1994. J. Clin. Microbiol. 32 (7) : 1639-1643) . A number of methodologies have been developed to accomplish the distinction between the various species and subspecies.

A DNA probe test for the detection of Myb. avium subsp. para tuberculosis is in one case based on the presence of an insertion sequence in the genome of Myb. avium subsp. para tuberculosis thought to be specific for this subspecies, the so-called IS900 sequence (McFadden, J.J., et al. 1987a. Mol. Microbiol. 1:283-291; Vary, P.H., et al. 1990. J. Clin. Microbiol. 28 (5) : 933-937 ) . For the detection of this para tuberculosis insertion element, which, when present, is present in 15-20 copies on the genome, a PCR reaction is applied. However, the IS900 sequence is not present in all strains of Myb . avium subsp. para tuberculosis due to the variation of genomic DNA as described above. Several studies report on the absence of IS900 sequences in between 3 and 20% of the clinical isolates of Mycobacterium para tuberculosis obtained from animals with classical Johne' s disease (Thoresen, O.F., and I. Olsaker. 1994. Vet. Microbiol. 40:293-303; Bauerfeind, R. , et al. 1996. J.Clin. Microbiol.

34 (7) :1617-1621) . Consequently, a test based on IS900 testing may bear the risk of false-negative results, thereby missing cows that should have been culled or destroyed. In 1992, Moss et al. demonstrated the presence of IS900-like elements (IS902) in the "wood pigeon strain" Mycobacterium avium subsp. silva ticum (Moss, M.T., et al. 1992. J. Gen. Microbiol. 138:139-145) which is closely related to Mycobacterium avium subsp. para tuberculosis . Other reports have confirmed the presence of such homologous genetic elements in Mycobacterium avium subspecies other than para tuberculosis (Kunze, Z.M., et al . 1991. Mol. Microbiol. 5 (9) :2265-2272; Kunimoto, D., et al. 1994. Am. Soc. Microbiol. 9:182; Roiz, M.P., et al . 1995. J. Clin. Microbiol. 33:1389-1391). More recently, while studying the etiology of the lung disease sarcoidosis, El-Zaatari et al. discovered a bacterium belonging to the Mycobacterium avium complex that failed to hybridise with a silvaticum-specific probe but was shown to contain the IS900 sequence, or a closely related sequence, as inferred from IS900 PCR (El- Zaatari, F.A.K., et al. 1997. Scand. J. Infect. Dis. 29:202- 204). According to test result for the IS900 sequences, the bacterium should be identified as Mycobacterium avium subsp. para tuberculosis . The possible socio-economic consequences of such findings, however, demand careful interpretation and the authors are accurate when accepting the possibilities of having detected a IS900 homologue or an organism of the avium complex not formerly known to possess the IS900 sequence. Clearly, results of the IS900 test are, although extremely useful in many cases, not unequivocally interpretable . In fact, careful selection of PCR primers is essential to discriminate between IS900 (Vary, P.H., et al. 1990. J. Clin. Microbiol. 28 (5) : 933-937 ) , IS901 (Kunze, Z.M., et al. 1991. Mol. Microbiol. 5 (9) : 2265-2272 ) and IS902 (Moss, M.T., et al . 1992. J. Gen. Microbiol. 138:139-145) insertion elements and IS900 hybridisation probes should be employed at utmost stringency to evade the possibilities of erroneously detecting homologous, but dissimilar genetic elements. Consequently, a test based on IS900 testing may also bear the risk of false-positive results, thereby marking cows as infected that are perfectly healthy.

A further great limitation of this IS900 test is that the PCR reaction itself is inhibited very strongly by substances that are naturally present in faeces (Van der

Giessen, J. .B., et al. 1992. J. Clin. Microbiol. 30(5):1216- 1219) . This inhibitory effect limits the use of PCR for the detection of para tuberculosis in faeces based on the genomic DNA targets, such as the IS900 sequence. To circumvent this problem, dilution (lOx) of the faeces sample prior to IS900 PCR testing can be applied. This will, to a limited extend reduce PCR inhibition ( idjojoatmodjo, M.N., et al. 1992. J Clin. Microbiol. 30:3195-3199; Varela, P., et al. 1994. J. Clin. Microbiol. 32:1246-1248). However, it also greatly reduces the sensitivity of the test. In recent years many improvements to the original protocol for PCR detection of the IS900 sequence have been made. Among these are so-called "hot start" PCR protocols to inactivate possible inhibitors, as well as protocols based on bead beating with zirconium beads in combination with direct zirconium adsorption and elution (Van der Giessen, J. .B. 1993. Academic thesis, Utrecht University, The Netherlands) . Despite these improvements, the PCR reaction for the direct detection of Myb . avium subsp. para tuberculosis in faeces is still not as sensitive as cultivation methods. Typically, only 10-30% of the samples that are considered positive by cultivation are diagnosed positive when direct IS900 PCR is used (Van der Giessen, J.W.B. , et al . 1992. J. Clin. Microbiol. 30(5):1216- 1219) . The combined application of BACTEC pre-cultivation and subsequent culture confirmation by IS900 PCR can reach sensitivities comparable to conventional cultivation, and such methods are frequently used (Evans, K.D., et al. 1992. J. Clin. Microbiol. 30:2427-2431; Sockett, D.C., et al. 1992. Can. J. Vet. Res. 56 (2 ): 148-153; Cousins, D.V., et al. 1995. Aust. Vet. J. 72 (12) :458-462; Whittington, R.J., et al . 1998. J. Clin. Microbiol. 36 (3) : 701-707) . Although this will reduce the time in which paratuberculosis can be diagnosed to about 1 week, it will not circumvent the problem of the culture logistics in large scale testing programs described earlier. To further improve the sensitivity of direct IS900 testing, samples may be "purified" by using immunomagnetic beads that can specifically "pull" the Myb . avium (subsp. para tuberculosis) cells (Grant, I.R., et al . 1998. Appl . Environ. Microbiol. 64 ( 9) : 3153-3158) or DNA (Millar, D.S., et al.1995. Anal. Biochem. 226 (2) : 325-330) from the sample that is under investigation. Such methods have been developed for para tuberculosis and the procedures for this can be automated. The eminent lack of accurate, rapid and reliable tests for differentiation of the various Mycobacterium avium complex bacteria and particularly para tuberculosis has resulted in the development of a series of alternative DNA probe tests. Among these are the one by Poupart et al . (Poupart, P., et al . 1993. J. Clin. Microbiol. 31(6):1601- 1605) in which a paratui>erculosis-specific sequence is used that was identified by screening a genomic library of para tuberculosis in a transcription vector. Such a unique sequence may be used as an RNA probe. Gormley et al .

(Gormley, E., et al . 1997. FEMS Microbiol. Lett. 147(1): 63- 68) identified the use of restriction fragment length polymorphisms in the PAN promotor region of pathogenic mycobacteria as a valuable tool in differentiation of para tuberculosis . More recently, Ellingson et al . (Ellingson, J.L., et al. 1998. Mol. Cell. Probes 12 (3) : 133-142 ) developed a so-called di-oligonucleotide hybridization (dOH) assay for the simultaneous detection of a Mycobacterium genus-specific recA gene sequence and a paratui-.erculosis-specific 30 bp hspX sequence. All of these alternative tests are based on genomic targets. An important drawback of these genomic methods, however, is related to the low copy number at which these genes or gene sequences occur in the cell. This limits the sensitivity of such tests. To facilitate large scale screening testing for

Johne ' s disease with the purpose of eradication several conditions must be met. The test system must be specific for Mycobacterium avium subspecies para tuberculosis and must therefore be able to make a distinction between this subspecies and the other subspecies of the Mycobacterium avium complex that do not cause Johne ' s disease. Secondly, the specificity of the test (no false-positives are wanted due to the fact that positive cows are all destroyed) must be close to a 100% (relative to conventional cultivation) . Thirdly, the sensitivity (no false-negatives) of the system must approach that of conventional cultivation methods, otherwise an eradication programme will not be successful. A sensitivity of 100 cells per gram of faeces must at least be reached. Although the IS900 probe test exhibits such sensitivities under artificial circumstances when tested on M. paratuberculosis-spiked culture-negative DNA samples (Vary et al. 1990. J. Clin. Microbiol. 28:933-937), in direct detection applications in crude bovine manure detection limits lower than 10,000-100,000 cells per gram of faeces are rarely met (Vary et al. 1990. J. Clin. Microbiol. 28:933-937; Van der Giessen et al. 1992. J. Clin. Microbiol. 30: 1216- 1219) . Fourthly, the throughput-rate of the process must be one working day (the total analysis does not necessarily be completed within one day; a certain accumulation of samples can be met) . Fifthly, the assay must allow automation. Finally, the system must reach capacities of approximately 1,000 analysis results per day.

Further conditions must be met in the case of high- throughput screening with respect to reliability of test results. Every method of analysis and specifically high- throughput screening requires an analysis of reliability. On the one hand, the reliability is determined by the chance of acquiring false-negative test results. This chance is expressed as sensitivity. On the other hand, chances exist on acquiring false-positive test results; this chance is expressed in the specificity. Although DNA target amplification technologies such as such as PCR (Mullis et al. 1987. US patent 4,683,195), TMA (Enns, R.K. 1987. Gen-Probe, Inc., San Diego, Calif.; Jonas, V., et al. 1993. J. Clin. Microbiol. 31 ( 9) : 2410-2416) or NASBA (Compton, J. 1991. Nature (London) 350:91-92), are extremely sensitive, a negative test result does not necessarily preclude the possibilities of isolating Myb . avium subsp. para tuberculosis from the sample material. The test result is influenced by method of sampling, and transport conditions, variability in the sampling process, laboratory procedural errors, sample miss-identification and, most importantly transcriptional errors. Furthermore, these target amplification methods do not present the possibilities to verify the test result. Such problems could be circumvented by using the technology of fluorescence in situ hybridization (FISH) in which whole bacterial cells are fluorescently stained with sequence- specific nucleic acid probes. FISH has as a unique advantage that cells are individually visible. However, the direct detection of very low numbers of intact cells of Myb . avium subsp. para tuberculosis in faecal samples is hampered by the high background population (10¹⁰-10¹¹ per gram) of other faecal micro-organisms.

In conclusion, immunological tests do not present indisputable evidence of infection with the bacterium. The sole proof is formed by demonstrating the presence of the bacterium Myb . avium subsp. para tuberculosis in milk, blood or faeces or other relevant samples. For large-scale analysis, cultivation methods are unsuited. The only options for large scale screening lie in the application of diagnostic methods, which rely on the demonstration of genetic material of the bacterium Myb . avium subsp. para tuberculosis . However, the diagnosis of the disease, i.e., the reliable demonstration of bacterial nucleic acid, is among other things, hampered by the enormous similarity between this subspecies and for example Mycobacterium avium subspecies avium . Mycobacterium avium subspecies avium is prevalent as a normal commensal bacterium in birds, and is relatively widespread. The currently available most sensitive nucleic acid test is the demonstration of the presence of a specific insertion element, IS900, present in the genome of the subspecies para tuberculosis . However, the sensitivity of this test is too low to allow direct detection of the bacterium in faecal samples or other natural habitats of the bacterium that contain substances inhibitory to enzymatic nucleic acid amplification reactions. Furthermore, as explained above, the IS900 sequence is sometimes absent in the genome of Mycobacterium avium subspecies para tuberculosis, thereby creating the possibility of a false-negative diagnosis and the IS900 sequence can sometimes be found in the genome of other Mycobacterium avium subspecies, thereby creating the risk of a false-positive diagnosis .

The invention provides a method by which a rapid and sensitive procedure for the detection and identification of Mycobacterium avium subspecies para tuberculosis in biological and natural samples is achieved, whereby Mycobacterium avium subspecies para tuberculosis can be discriminated from other Mycobacterium avium subspecies, such as subspecies avium . The invention provides a nucleic acid probe or primer allowing detecting nucleic acid derived from Mycobacterium avium whereby nucleic acid derived from Mycobacterium avium subspecies para tuberculosis can be discriminated from nucleic acid derived from other Mycobacterium avium subspecies. In particular, the invention provides a nucleic acid probe or primer allowing detecting nucleic acid derived from Mycobacterium avium whereby Mycobacterium avium subspecies para tuberculosis can be discriminated from other Mycobacterium avium subspecies, said nucleic acid comprising a stable and conserved mutation specific for Mycobacterium avium subspecies paratuberculosis .

The invention provides a method for specifically detecting nucleic acid derived from a causal agent of Johne ' s disease in a sample whereby nucleic acid from Mycobacterium avium subspecies para tuberculosis is discriminated from nucleic acid from other Mycobacterium avium subspecies comprising detecting in said nucleic acid a mutation specifically conserved for Mycobacterium avium subspecies para tuberculosis . Said mutation is specifically conserved throughout Mycobacterium avium subspecies para tuberculosis , as opposed to the IS900 insertion sequence, which is sometimes missing or is found in other subspecies.

In a preferred embodiment, the invention provides a method wherein said nucleic acid is derived from 23S ribosomal RNA, for example wherein said ribosomal RNA comprises a nucleic acid as shown in figure 1, or wherein said conserved mutation is located at position 754, 1363 or 3093 as shown in figure 2. The invention provides three conserved mutations that are present in the functional paratuberculosis 23S rRNA gene sequence. These can be used as diagnostic targets to distinguish between the paratuberculosis and avium subspecies. As such they are part of the present invention in which they are referred to as mutations 754, 1363 and 3093 (figure 2) .

Of all genes in the genome, the ribosomal RNA genes belong to the most stable - or least variant. Due to their role as key elements in protein synthesis, ribosomal RNA's, and consequently the genes encoding them, are highly conserved both in structure and sequence (Pace, N.R., et al . 1985. ASM News 51:4-12). The ribosomal RNA genes comprise a mosaic of variable regions, which allow for discrimination between lower taxa, alternated by sequences that are well conserved, thus allowing for differentiation between higher order taxa (Woese, C.R. 1987. Microbiol. Rev. 51:221-271;

Olsen, G.J., et al . 1986. Ann. Rev. Microbiol. 40:337-365). Due to the high copy number of mature ribosomal RNA molecules, as a part of the cellular ribosomes, and their predictable, evolutionary regulated, sequence variability, ribosomal RNA' s have become a well established tool in the identification of bacteria. However, among all bacteria, the mycobacteria possess exceptional sequence conservation among species (Frothingham, R. , et al. 1994. J. Clin. Microbiol. 32 (7) :1639-1643) . And the ribosomal RNA's do not represent feasible targets to differentiate between subspecies. Van der Giessen et al. (Van der Giessen, J.W.B., et al. 1992. J. Med. Microbiol. 36:255-263) attempted to developed a PCR test for Myb . avium subsp. para tuberculosis based on the 16S ribosomal RNA gene. Although several sequence differences between the various para tuberculosis strains and avium strains were found, the sequence comparisons between the 16S rRNA gene of para tuberculosis and its most closely related subspecies avi um did not reveal a single stable base in difference that could be used as a distinction between para tuberculosis and avium . The 16S ribosomal RNA test thus detected both para tuberculosis and avium together and exhibited a sensitivity that was between one and two orders of magnitude lower than the IS900 test (Van der Giessen, J.W.B., et al . 1992. J. Clin. Microbiol. 30 (5) : 1216-1219) . This may be due to the difference in copy number between the ribosomal RNA operon (max 1-3 copies per genome) and the IS900 sequence (15-20 copies). Although the presence of strain-specific sequence differences, based on the comparison of gene sequences from two separate isolates (Van Der Giessen et al. 1994. Microbiology 140:1103-1108), may sometimes mistakenly be interpreted as to yield diagnostically valuable differentiation criteria for closely related species, they do not provide reliable subspecies-level sequence information and may not be used as such.

In a follow-up study, Van der Giessen et al. performed a sequence comparison of the 23S rRNA genes of one strain of para tuberculosis and one strain of avium (Van der Giessen, J.W.B., et al . 1994. Microbiology (UK) 140:1103- 1108). The presence of 9 base differences between the two sequences (3000 nucleotides each) was reported and the possibilities to differentiate the two strains on the bases of these differences was proposed. Also, Stone et al . (Stone, B.B., et al. 1995. Int. J. System Bacteriol. 45:811-819) performed a partial sequence analysis (226 bases) of the 23S rRNA of a large number of different Mycobacteria, among which were one avium and one para tuberculosis strain. Their results indicated the presence of even more sequence differences between the 23S rRNA gene sequence of their avium and para tuberculosis strains. Among the 226 nucleotides for which the sequence was determined, no less than 4 differences were reported between the two strains. However, both the study of Stone et al. and Van der Giessen et al. could not demonstrate the presence of stable and conserved discriminatory signatures or mutations in the para tuberculosis 23S rRNA gene compared to the avium gene that could be used to differentiate para tuberculosis from avium .

The natural genetic variability of bacterial cells of the same species but of different clonal origin within a consortium or from geographically separated populations is not well understood. Frequently, lack of knowledge in this domain results in genetic tests that produce a prodigality of false negative results due to the fact that many field isolates do not contain the presumed distinctive sequence characteristics .

Development of a diagnostic test for paratuberculosis on the basis of the diagnostic targets proposed by Van der Giessen et al . would have resulted in a diagnostic test system with a high rate of false negative test results due to the proposition of target- or signature sequences that are not conserved in paratuberculosis strains of different clonal origin. No less than six of the nine mutations suggested by Van der Giessen et al . can not be used to make a distinction between avium and paratuberculosis cells since these mutations are not conserved. These six mutations do not occur in the type strain of paratuberculosis (ATCC 19698) . These mutations are sequence positions in the 23S rRNA gene of the investigated strains that represent random mutations or natural sequence variability in so far that these mutations do not exhibit evolutionary or phylogenetic significance and for that reason they can not be used for distinctive testing. More such mutations can be found by inter-strain sequence comparison. However, they serve no diagnostic purpose. Such mutations mask the subspecies-specific mutations that are significant to subspecies-specific diagnostics of paratuberculosis .

The present invention reveals three point mutations as potential subspecies-specific target-sites that are conserved in all of the 25 investigated reference strains and field isolates of paratuberculosis . Two of these mutations, situated at positions 754 and 1363, are specifically proposed for combined use in a diagnostic test for the subspecies paratuberculosis as these targets are situated in relatively close proximity of one another.

Van der Giessen et al . suggested the presence of a total of 9 mutations or "mismatches' between the 23S rRNA gene sequences of avium and paratuberculosis. They identified two additional differences between the obtained sequences but these were positioned immediately next to one another in a non-transcribed spacer region and can therefore not be used for sensitive detection as they are non transcribed in functional ribosomal RNA. The conserved status of these "mismatches' remains presently unknown. The nine "mismatches' in the 23S rRNA gene that were identified by Van der Giessen et al . were proposed as potential diagnostic targets. The present invention however reveals that only three conserved mutations are present in the functional paratuberculosis 23S rRNA gene sequence. These can be used as diagnostic targets to distinguish between the paratuberculosis and avium subspecies. As such they are part of the present invention in which they are referred to as mutations 754, 1363 and 3093 (figure 2) . The six remaining targets suggested by Van der Giessen et al . are not conserved (positions 1746,1747, 1843, 2718, 2810, and 3126 respectively in figure 2) . A final mutation identified through the present study is displayed in figure 2 as position 3188. This mutation is part of the internal transcribed spacer region between the 23S and 5S rRNA genes and can not be used for sensitive detection. Its conserved status was recently confirmed by Scheibl and Gerlach (Vet. Microbiol. (1997) 57:151-158).

In all, over 60 differences between an individual avium and an individual paratuberculosis strain can be identified in the alignment presented in figure 2, of which only three can be used for subspecies-specific diagnostics and are part of the present invention. This illustrates that the present invention reveals paratuberculosis targets not identified before.

Differences in the ribosomal RNA genes are in general not considered feasible to serve as a basis for differentiation between closely related species or subspecies. In order to distinguish different species of the genus Mycobacterium, the internal transcribed spacer (ITS) region between the 16S and 23S rRNA genes that is not transcribed into functional RNA has recently gained more attention (Frothingham. R, et al. 1993. J. Bacteriol .

175 (10) :2818-2825; Glennon, M., et al. 1994. Tuber. Lung Dis. 75 (5) :353-360; Frothingham, R, and K.H. Wilson. 1994. J. Infect. Dis. 169 (2 ): 305-312. ; Ji, Y.E., et al. 1994. Microbiology. 140 (Pt7 ): 1763-1773) . Due to the absence of evolutionary consequences of mutations in the ITS region, mutations occur more frequently in this non-transcribed spacer than in the functional rRNA genes themselves. One problem with such mutations is therefore that they are not fixed during evolution and that they may even differ between strains of the same species. For differentiation of subspecies, however, even ITS region sequences may sometimes contain insufficient differences (Bourque, S.N., et al. 1995. Appl. Environ. Microbiol. 61 (4) : 1623-1626) . Isolates of Mycobacterium leprae have also been found to contain identical sequences in the ITS region (De Wit, M.Y.L., and O 00/34517

20

P.R. Klatser. 1994. Microbiology 140:1983-1987). Recently, Scheibl and Gerlach (Scheibl, P. and G.F. Gerlach. 1997. Vet. Microbiol. 57 (2-3) : 151-158 ) demonstrated that despite its high potential mutation rate, the ITS region of all Mycobacterium avium subsp. para tuberculosis strains investigated exhibited one common base difference compared to other Mycobacterium avium subspecies. Two problems exist with trying to use this difference as a differentiation criteria between avi um and para tuberculosis . For one the ITS region is not transcribed into functional RNA as a result of which the copy number of this target remains too low to generate improvement of the sensitivity of detection when compared to available tests such as the IS900 probe test. But more importantly, and unlike in ribosomal RNA's, base mutations in the ITS region encounter no evolutionary pressure and mutations can occur quite random and frequent. ITS region mutations are therefore less stable that rRNA gene mutations. This limits the use of ITS sequences to strain level differentiation of Myc. avium complex bacteria (Frothingham. R, et al. 1993. J. Bacteriol. 175 (10) : 2818-2825; Frothingham, R, and K.H. Wilson. 1994. J. Infect. Dis. 169 (2) : 305-312) .

The present invention reveals the presence of three stable and conserved mutations in the 23S rRNA of Mycobacterium avium subspecies paratuberculosis which allow the differentiation of this bacterium from other

Mycobacteri um avium complex bacteria, while at the same time providing for a high copy number target in the form of functionally transcribed RNA.

For example, and further explained in the experimental part, the invention provides a nucleic acid probe or primer wherein said ribosomal RNA comprises a nucleic acid sequence as shown in figure 1. Herein are identified at least 4 unique mutations in the 23S rRNA gene of subspecies para tuberculosis, of which three are transcribed into the functional ribosomal RNA of this bacterium. The ribosomal RNA is present in some 1,000 to 10,000 copies in growing cells and constitute a far more powerful target for direct enzymatic nucleic amplification than the IS900 sequence.

The present invention reveals three point mutations as potential subspecies-specific target-sites that are conserved in reference strains and field isolates of paratuberculosis . Two of these mutations, situated at positions 754 and 1363, are specifically proposed for combined use in a diagnostic test for the subspecies paratuberculosis as these targets are situated in relatively close proximity of one another. The present invention shows that these three mutations are conserved and present in the functional paratuberculosis 23S rRNA gene sequence. These can be used as diagnostic targets to distinguish between the paratuberculosis and avium subspecies. As such they are part of the present invention in which they are referred to as mutations 754, 1363 and 3093 (figure 2) . A final mutation identified through the present study is displayed in figure 2 as position 3188. This mutation is part of the internal transcribed spacer region between the 23S and 5S rRNA genes and can not be used for sensitive detection. Its conserved status was recently confirmed by Scheibl and Gerlach (Vet . Microbiol. (1997) 57:151-158). In all, over 60 differences between an individual avium and an individual paratuberculosis strain are identified in the alignment presented in figure 2, of which only three can be used for subspecies-specific diagnostics. Nucleic acid detection generally utilises specific hybridisation of a probe or primer to the nucleic acid to be detected. Such nucleic detection in general is known in the art and can for example be achieved with classical hybridisation techniques, such as Northern or Southern Blotting on nucleic acid derived directly from a sample or on nucleic acid derived via an amplification process, for example via recombinant DNA techniques wherein said nucleic acid is first amplified, be it indirectly in a host cell, for example in a bacterium such as E. coli using a suitable plasmid system or directly in a suitable amplification system such as PCR or NASBA. Nucleic acid to be detected can also be sequenced directly from the starting material, but more often from nucleic acid that has first been amplified as described above. A probe or primer according to the invention comprises a DNA, RNA or PNA oligonucleotide sequence specific for a Mycobacterium avium subspecies with which said species can be detected in a nucleic acid detection assay. Primers, as defined herein, are understood to be unlabelled oligonucleotides, which are selected to start (prime) sequencing or amplification techniques. By selecting a primer for a specific match, or, alternatively, for a specific mismatch, for example for a nucleotide sequence having a specific mutation, deletion, insertion or other discriminatory feature, said sequencing or amplification techniques can be employed to specifically detect a nucleic acid that may or may not be present in a sample, and discriminate it from other, related, nucleic acid sequence that may be present. Probes, as defined herein, are understood to be labelled oligonucleotides, which are selected to detect and label the desired nucleic acid, i.e. to report its presence. Labelling is achieved with reporter molecules, which are widely known in the art, such as radioactive labels, enzymes, particles such as gold or silver particles, chromophores, fluorochromes, excitation or quencher molecules, and other reporter molecules known in the art. Probes are in general used to detect a nucleic acid by hybridisation to said nucleic acid, be it by hybridisation to a nucleic acid present in a sample, for example by in si tu hybridisation such as by FISH (fluorescent in si tu hybridisation, or by hybridisation to an amplified fragment of said nucleic acid (also known as an amplicon) . PCR or NASBA derived amplicons, for example, can be detected by probing with such a probe, for example by blotting techniques, or by more sophisticated techniques such as Taqman or FRET technology.

The invention provides a nucleic acid probe or primer allowing detecting nucleic acid derived from any Mycobacterium avium , such as derived from its various subspecies avium , silva ticum or para tuberculosis , whereby nucleic acid derived from Mycobacterium avium subspecies para tuberculosis can be discriminated from nucleic acid derived from other Mycobacterium avium subspecies, however, in a preferred embodiment, the invention provides a nucleic acid probe or primer according to the invention for detecting a causal agent of Johne ' s disease in a sample, preferably in a sample derived from a ruminant, most preferably from a cow. In a most preferred embodiment, the invention provides a nucleic acid probe or primer allowing detecting nucleic acid derived from Mycobacterium avium derived from a faecal, blood or milk sample. Herewith the invention provides methods and means to detect Mycobacterium avium in samples obtained from animals that may or may not be suspected of having Johne' s disease, allowing discriminating between the various

Mycobacterium avium subspecies, and positively identifying those animals for example infected with subspecies para tuberculosis . n general, both RT-PCR and NASBA are sensitive methods for detection of bacterial ribosomal RNA's. As a result, cells of pure cultures suspended in common buffers, such as PBS, can be detected at sensitivities of only a few cells per milliliter or gram (see figure 8 of the present invention) . However, nucleic acid extracts obtained from crude manure of fecal samples exert very strong inhibitory effects on enzymatic amplification of specific nucleic acid sequences both by (RT-)PCR and NASBA (see figure 6 of the present invention) . The reason for this is not fully understood by the art. The result of this, however, is a very low sensitivity of diagnostic assays in fecal samples, due to the necessity of extensive sample dilution.

Selective recovery of the target cells from the matrix, i.e. by immunomagnetic capture, is one way to circumvent this problem as demonstrated by Grant et al . for detection of M. para tuberculosis in milk (Appl. Environ. Microbiol. 64, 1998, pp. 3153-3158) and by Widjojoatmodjo et al. for detection of salmonella in fecal samples (J Clin Microbiol 30, 1992, pp. 3195-3199) . Another way is to attempt excluding co-extraction of inhibitory compounds, such as bile-acids, by cellulose adsorption purification of extracted RNA's as demonstrated by Wilde et al . for detection of viral RNA's from fecal samples (J. Clin. Microbiol. 28, 1990, pp. 1300-1307) . Yet another way, and one that presents an embodiment of the present invention, is to use selective lysis and pre- extraction procedures of non-target materials while retaining the target cells or the target nucleic acids in the sample. Due to their extraordinary cellular composition, cells of Mycoba cterium species are very resistant to lysis, at least more resistant than most other bacterial organisms. Specific procedures, separately known to the art to severely disturb cellular integrity, such as repeated freeze-thawing or bead beating with small beads, such as glass beads, are effective ways to lyse most or all cells including those of M. para tuberculosis . In a preferred embodiment of the invention its is provided that certain treatments that result in at least partial lysis of most fecal bacterial cells do not effectively or only little lyse cells of M. paratuberculosis or other Mycobacteria . In a preferred embodiment, the invention thus provides selective lysis of non-mycobacterial organisms or matter to allow better detection of a mycobacterial organism. Therefor, the resistance to lysis of M. para tuberculosis can be used to increase assay sensitivity. These treatments may include, but are not limited to, TRI-reagent processing of samples (see below) or NaOH exposure. Further examples for the use of such selective lysis treatment and their effect on assay sensitivity are given below. It is an embodiment of the present invention that such selective or partial lysis treatments can be used to reduce the presence of non-target cells, non-target nucleic acids or compounds otherwise inhibitory to or interfering with the desired specific enzymatic nucleic acid amplification prior to extraction and purification of target nucleic acids, thereby increasing the sensitivity of diagnostic systems for detection of specific micro-organisms at very low concentrations in complex matrices, such as M. para tuberculosis in fecal samples, milk, sputum or blood. In a preferred embodiment, the invention provides a nucleic acid probe or primer derived from ribosomal RNA, such as 23S ribosomal RNA. In the case of ribosomal RNA 1,000 to 10,000 copies are normally present in growing cells. This copy number is much higher than that of other genomic targets that are generally present in 1 to 20 copies. The unique advantage of RNA target amplification by, e.g., reverse transcriptase PCR (RT-PCR) or NASBA is therefore that these methods are superior in sensitivity than the DNA amplification methods for detection of bacterial cells, e.g., by PCR. However, also for hybridisation techniques, such as in si tu hybridisation, this high copy number is advantageous, since it again allows superior sensitivity.

The invention provides a method for detecting nucleic acid derived from Mycobacterium avium in a sample comprising using at least one nucleic acid probe or primer according to the invention. An assay or method as provided by the invention is for example based on the detection of one or more specific point-mutations in the 23S ribosomal RNA of the bacterium. The high copy number of ribosomal RNA's relative to other genomic targets provides for a very high sensitivity of the assay, thereby allowing for performance of the assay in "difficult" matrices , such as faecal, sputum, blood or milk samples. The invention provides a method or assay to detect the bacterium in milk, faeces, soil, feed, and any other habitat in which the bacterium can be found. The speed and ease with which the assay can be performed relative to conventional methods enables one to use it for routine analyses of a large number of samples. Samples can comprise those taken from individuals, such as cows, however, it is also feasible to test bulk or tank milk samples, pooled faecal or pooled blood samples with a method provided by the invention.

Herewith, the invention provides a method for detecting a causal agent of Johne ' s disease. Especially, a method or assay as provided by the invention can be used to discriminate between Mycobacterium avium subspecies para tuberculosis and its closest relative Mycobacterium avium subspecies avium . The detection of Myb . avium subsp. para tuberculosis cells in faeces is greatly simplified by a method or assay according to the invention because whole bacterial cells need for example not be detected but DNA or RNA liberated from the cells in the sample is analysed for the presence of specific target or signature sequences. In order to develop a functional test for the detection and identification of Myb . avium subsp. paratuberculosis cells in faeces, the test must be able to make a distinction between the three subspecies in the Mycobacterium avium complex. This test has now been provided by the invention. A method as provided by the invention has sufficiently high sensitivity to test faecal samples. Said method can be performed using nucleic acid amplification techniques, or hybridisation techniques, such as si tu hybridisation, for example as described herein.

Furthermore, the invention provides a diagnostic kit comprising at least nucleic acid probe or primer according to the invention, and optionally other means, such as buffers and other reagents or instructions, to detect Mycobacterium avium, preferably subspecies paratuberculosis , the causal agent of Johne ' s disease.

The invention provides method for detecting a ruminant infected with a causal agent of Johne ' s disease comprising obtaining a sample from said cow and testing said sample for the presence of nucleic acid using at least one nucleic probe or primer, or a method or diagnostic kit according to the invention. Herewith, the invention provides a method for eradicating Johne ' s disease from a herd of ruminants comprising using a method according to the invention and further comprising culling or removing a ruminant infected with a causal agent of Johne ' s disease ruminant from said herd, after which said ruminant, for example an infected cow, may be destroyed. Such a testing and remove system, generally called a control or eradication programme, is best performed under strict supervision of or even prescribed by (veterinary) governmental authorities, but may also very well be achieved in a voluntary effort by combined farmers or others involved in the agricultural community. In summary, the invention provides use of a nucleic acid probe or primer allowing detecting nucleic acid derived from Mycobacterium avium whereby nucleic acid derived from Mycobacterium avium subspecies para tuberculosis can be discriminated from nucleic acid derived from other Mycobacterium avium subspecies. Such use is for example to detect a causal agent of Johne' s disease in faecal samples, for example by using a nucleic acid detection method as provided by the invention. Such use, as provided by the invention, for example allows governmental authorities and individual farmers, veterinarians, and concerned agricultural organisations to eradicate Johne ' s disease from a herd of ruminants, more specifically of cows infected with Mycobacterium avium subspecies para tuberculosis . The invention is further explained in the experimental part of the description without limiting the invention thereto.

Experimental part

Detection of paratuberculosis-specific gene sequences by PCR.

Source and identi ty of bacteria used in this study. The source and identity as well as the condition of the pure cultures from which nucleic acids were extracted is presented in table 1, in which DSMZ is the Deutsche Sammlung von Mikro-organismen und Zellculturen, ATCC is the American Type Culture Collection.

TNA extraction and 23S rDNA sequencing

Total nucleic acids were isolated from colonies grown on Lowenstein-Jensen medium or directly from lyophilized cell pellets as obtained from culture collections. Methods used were as described elsewhere (Aznar et al., 1994. Int. J. System. Bacteriol. 44:330-337). The 23S rRNA genes were amplified by using primers directed towards conserved regions and the genes were sequenced commercially. The 23S rRNA genes were aligned and checked for the presence of Myb . avium subsp. paratαberculosis-specific sequences. Part of the alignment is displayed in Fig. 1.

The paratuberculosis sequence is characterised by the presence of 3 single point-mutations as compared to the avium and silvaticum sequences. These three point-mutations comprise of a transition of C in avium and silvaticum to a T in the paratuberculosis gene. These point mutations can be detected by using the polymerase chain reaction such as described now.

PCR detection of point-muta tions . PCR was performed on a Eppendorf Mastercycler gradient with a temperature gradient from 55 to 68 degrees centigrade. The PCR reaction mixture (100 μl) consisted of the following components: 10 μl of lOx reaction buffer; 0.3 μl of Taq DNA polymerase (Promega corp.); 2 μl of both forward and reverse primer @ 100 μM; 1 μl of DNA template @ 25 ng/μl; 4 μl of dNTP's and 80,7 μl of H20. Thermal cycles consisted of a denaturing step at 95°C for 5 min, followed by 35 cycles of 95°C, gradient temp 55-68°C and 72 °C for 1, 2, and 3 min respectively. PCR products were checked by standard agarose gel-electrophoresis in 0.8% agarose gels and stained with ethidium bromide.

A series of primers was tested to determine the optimal site of the "mismatch" position within the probe sequence relative to the avium sequence (as displayed in table 2). When used in a PCR reaction several of these primers were found to result in reliable discrimination between subspecies avium and subspecies para tuberculosis (see figure 2) .

This invention forms the key element of the diagnostic procedure which comprises at least one of the following steps:

1. Sampling of faeces

2. Transport and storage of the sample in RNA- stabilizing solution. 3. Selective lysis of non-mycobacterial matter.

4. Liberation of the desired nucleic acid by lysis of the bacterial cells.

5. Amplification of the bacterium-specific sequences by a method known to the art 6. Detection of the bacterium-specific sequences by nucleic acid hybridization methods Further detailed description

Example 1 : Cell lysis , ribonucleic acid extraction and RNA amplification from pure cultures of Mycobacterium avium subspecies paratuberculosis .

Sample preparation. Pure cultures of M. para tuberculosis strain GDR were grown on Lόwenstein-Jensen gradient agars. A loop of cells was collected from the agar surface and resuspended into glycerol (20% v/v) PBS. The cell-suspensions were stored at -20°C until their use in spiking experiments to determine the sensitivity of the various detection assays. Frozen cell-suspensions were rapidly thawed, washed twice in water and resuspended into 100 μl of water. A volume of 10 μl was fixed in 0.4% formalin for subsequent determination of cell number. For cell counting, Sybr Green (Molecular Probes, Leiden, The Netherlands) was used as a DNA counter-stain. Briefly, a 10 μl volume of the washed and fixed cell suspension was added to 1.0 ml of PBS. To this diluted suspension, Sybr Green was added to yield a final concentration of 1/10,000-th of the original manufacturer-stock. Staining occurred at 37 °C for 10 min. After staining, the entire suspension was drawn over a 0,2 μm pore-size black polycarbonate membrane filter (Millipore, Etten-Leur, The Netherlands). The filters were mounted on a microscope slide and cell counts were determined using an Olympus BX-60 epifluorescence microscope (Paes Nederland BV, Zoeterwoude, The Netherlands) . A minimum of 15 fields-of-view per sample were counted or until the total number of counted cells was 300.

RNA extraction and quantification. RNA was extracted from pure culture cell suspensions by using three basic methods. Modifications were made to the procedures in case cell lysis of M. para tuberculosis was incomplete. These modifications comprised the inclusion of freeze/thawing cycles or bead-beating procedures. Basic procedures were as directed by the manufacturers of the RNA extraction systems. The following methods were tested for RNA extraction from M. para tuberculosis cells: i) RNA extraction according to the silica-adsorption principle described by Boom et al. (J. Clin. Microbiol. 28, 1990, pp. 495-503) which was purchased commercially as part of the NucliSens Basic Kit system for NASBA diagnostics (Organon Teknika, Boxtel, The Netherlands), ii) RNA extraction according to the acid- phenol/guanidinium principle described by Chomczynski and Sacchi (Anal Biochem 162, 1987, pp. 156-159) which was purchased commercially as TRI-reagent (Sigma Chemical Comp., Zwijndrecht, The Netherlands) and iii) a solid phase RNA extraction with the RNeasy Plant kit purchased from Qiagen (Hilden, Germany) , which includes a protoplast disrupting shredder filter. Extracts were dissolved in DEPC-treated water containing the RNase inhibitor RNasin (Promega, Leiden, The Netherlands) to prevent degradation.

Quantification of extracted nucleic acids was performed by using the fluorochrome RiboGreen (Molecular Probes, Leiden, The Netherlands) as an nucleic acid stain as directed by the manufacturer. Fluorescence readings were made on a Fluoroskan Ascent FL (Labsystems Oy, Helsinki, Finland) with excitation filter at 485 nm and emission filter at 530 nm. The efficiency of RNA extraction from M. para tuberculosis cells and the quality of the extract for nucleic acid amplification purpose was further evaluated empirically by NASBA and RT-PCR.

The yield of cellular RNA obtained from 1 x 10 5 cells of M. para tuberculosis strain GDR by various methods as well as NASBA results with such extracts as determined fluorimetrically with RiboGreen as a nucleic acid stain and by NASBA ECL detection is presented hereunder.

Method pg of RNA extracted NASBA ECL counts

NucliSens 4,790 638, 625

TRI 739 46,586

TRI + bead beating n. d. 725,521

RNeasy Plant 6,476 n. d.

NASBA amplification of 23S rRNA

NASBA was performed by using the commercial NucliSens Basic Kit system for NASBA diagnostics (Organon Teknika) . Reaction conditions for NASBA amplification of RNA were as directed by the manufacturer. Detection of M. para tuberculosis-specific 23S-rRNA fragments was performed by ECL detection using a biotin labeled probe. Measurements were performed on a NucliSens reader (Organon Teknika) . NASBA primers and probes specific for M. para tuberculosis were designed based on multiple alignments of the complete 23S rRNA gene sequences from 5 M. para tuberculosis and 5 M. avium strains, together with sequences from M. intracellulare and M. silvaticum. Alignments were constructed by using the Multiple Sequence alignment tool from the BCM Search Launcher

(www.hgsc.bcm.tmc.edu/ SearchLauncher/) . Confirmation of the existence of conserved M. para tuberculosis-specific sequences in the 23S rRNA gene was further evaluated by partial sequencing and by point-mutation PCR (primers with wild-type position at 3' -end) from 12 additional para tuberculosis strains and 3 additional avium strains. Various NASBA primer sets specific for M. para tuberculosis were designed and their performance was tested empirically. The optimized primers and probes are presented in figure 6. Optimized results were obtained with one particular primer set (see figure 7). RT-PCR amplification of 23S rRNA. Primers specific for M. paratuberculosis were designed based on multiple alignments of the complete 23S rRNA gene sequences from 5 M. para tuberculosis and 5 M. avium strains, together with sequences from M. intracellulare and M. silva ti cum . Alignments were constructed by using the Multiple Sequence alignment tool from the BCM Search Launcher (www.hgsc.bcm.tmc.edu/ SearchLauncher/) . Confirmation of the existence of conserved M. para tuberculosis-specific sequences in the 23S rRNA gene was further evaluated by partial sequencing and by point-mutation PCR (primers with wild-type position at 3' -end) from 12 additional para tuberculosis strains and 3 additional avium strains. The primers used are presented in figure 6. A total of 41 thermal cycles was applied.

Sensitivity of the various amplification reactions was determined by using standardized cell-suspensions of M. para tuberculosis in 20% glycerol. M. para tuberculosis- specific 23S-rRNA from TRI-reagent extracted and CF11 purified total RNA was transcribed into cDNA with AMV Reverse Transcriptase (Promega Inc., Leiden, The Netherlands) at 50°C for 90 min. The reaction conditions were as directed by the manufacturer. cDNA sequences were then amplified by PCR. Reaction products were checked for correct length on 1.5% agarose gels using ethidium bromide staining. The agarose gels were blotted onto hybond N+ membranes (Amersham Pharmacia) using a vacuum manifold and standard procedures (Sambrook J, Fritch EF and Maniatis T. Molecular cloning: A laboratory manual. 2nd ed. Cold Spring Harbor Labi. Press, New York, 1989) . DNA was crosslinked by UV irradiation using a UV crosslinker (Amersham Inc., s Hertogenbosch, The Netherlands) at 700 Joules/cm2. Blots were stored dry until use. For confirmation of PCR reaction product identity, blots were hybridized with a fluorescein labeled probe (see figure 6) and stained by NBT-BCIP chromogenic staining and anti- FITC-AP F_ab fragments (Roche Diagnostics, Almere, The Netherlands) .

Assay sensitivity. By using NASBA of 23S rRNA targets, cells of M. para tuberculosis in pure culture cell suspensions could effectively be detected at numbers as low as 10 cells per ml of sample (the lowest cell number tested) , provided that cells were disrupted by bead beating (see figure 8). Similar detection sensitivities were obtained by RT-PCR (see figure 9) .

Example 2 : NASBA amplification and detection of Mycobacterium. paratuberculosis-specific RNA in total RNA extracts from bovine fecal samples.

Sample preparation. Fresh cow manure samples were collected from the rectum of individual animals from disease- free farms. The manure was diluted 1:1 with PBS. After thorough vortexing, samples were centrifuged at low speed (100 x g) and aliquots of the supernatant representing 0,5 gram of the crude manure bacterial fractions were stored at - 20°C until use. Total RNA extraction from M. paratύberculosis-free bovine manure. For NASBA amplification, RNA was extracted by using the commercial NucliSens Basic Kit (Organon Teknika) based on silica adsorption. Briefly, to a cell pellet obtained from 0,5 grams of a crude manure bacterial cell fraction a volume of 10,0 ml of Lysis Buffer (NucliSens Basic Kit) was added. Extraction of the RNA occurred with silica (Basic Kit) as directed by the manufacturer. Further purification was achieved by boiling of the extract (10 min, 100°C) . Following extraction, the RNA was even further purified by the method of Wilde et al. (J. Clin. Microbiol. 28, 1990, pp. 1300-1307) using the cellulose fiber compound CF-11 (Whatman International Ltd., Clifton NJ, USA). Briefly, isopropanol precipitated total RNA was washed with 75% ethanol and dried for 5 min at room temperature. The pellet was resuspended in a 100 μl volume of DEPC-treated water and heated to 55°C for 10 min. Upon cooling, extracts were diluted 1:1 with 2 x STE containing 66% ethanol and an amount of 30 mg of CF-11 was added. Samples were placed on a rotating mixer for 45 min at room temperature. After centrifugation and removal of the supernatant, the pellet was washed 3 times with 1 x STE containing 20% ethanol. The RNA was eluted in a 20 μl volume of 10-50% formamide in DEPC water. The eluate was stored at -20°C until use.

RNA extraction from pure cultures of M. paratuberculosis .

RNA was extracted from pure culture cell suspensions by using the NucliSens Basic Kit system for NASBA diagnostics (Organon Teknika) . NASBA amplification and detection. Non-purified total RNA extracts from crude bovine manure were mixed with RNA extracts obtained from pure cultures of M. para tuberculosis in order to investigate the inhibiting effect of fecal matrix compounds on the amplification reaction. Reaction conditions for NASBA amplification were as described in example 1. Various dilutions of non-purified crude manure RNA extract were mixed with RNA extracted from 1,000,000 cells of M. para tuberculosis strain GDR. Non-purified crude manure extract showed a severe inhibitory effect on NASBA amplification of M. para tuberculosis RNA (see figure 10) . This effect could be reduced by administering very low amounts of bovine fecal matrix extract in the amplification reaction or by purification of RNA extract as described.

Example 3 : NASBA amplification and detection of Mycobacterium paratuberculosis cells in bovine fecal samples .

Sample preparation. Sample preparation was as described in example 2.

Total RNA extraction from bovine manure spiked with M. paratuberculosis cells. For NASBA amplification, RNA was extracted by using the commercial TRI-reagent (Sigma) . Briefly, to a cell pellet obtained from 500, 50 and 5 mg of a crude manure bacterial cell fraction, various amounts of M. paratuberculosis cells ranging from 10,000 to 100 cells were added. The pellets were resuspended in a 1 ml volume of 4% NaOH. The partial lysate was centrifuged at 10,000 x g and the remaining cell pellet was resuspended in 1 ml of TRI- reagent. The TRI-reagent suspended cell fraction was transferred to a mini bead beater vial, which contained 0,5 ml of 0,1 μm sized glass beads (Biospec, Bartlesville OK, USA) . Cells were lysed by bead beating in a mini beat beater (Biospec) for 3 x 60 sec and cooled on ice. RNA was extracted as advised by reagent the manufacturer. The RNA extract was further purified by CF-11 adsorption as described in example 2.

NASBA amplification and detection. M. para tuberculosis-specific 23S rRNA' s in the extracts were measured by NASBA amplification and detection as described in example 1. Sensitivity of the assay was between 100 and 1,000 cells per gram of feces (see figure 11) .

Example 4 : Detection of Mycobacterium paratuberculosis in spiked bovine fecal samples by reverse transcription PCR (RT-PCR) and reverse transcription nested PCR (RT-nested-PCR) .

Sample preparation. Sample preparation was as described in example 2.

RNA extraction from bovine manure. Total RNA was extracted from 0,5 grams of the crude manure bacterial fraction described in example 3.

RT-PCR and RT-nested-PCR amplification. RT-PCR procedures were as described in example 1. For RT-nested-PCR, a 2 μl volume of PCR reaction product was added to a second PCR reaction containing the nested primerset (see figure 7) . An additional 41 thermal cycles were applied.

Following CF-11 purification and RT-PCR amplification of extracted rRNA's, a detection limit of between 100 and 1,000 cells of M. para tuberculosis per gram of manure was achieved. However, using the nested PCR approach a further improvement of sensitivity was achieved. As little as 10 cells per gram of crude manure could successfully be detected with the RT-nested-PCR procedure (see figure 12) .

When nucleic acid amplification is applied on a large scale, effective programs for decontamination or the prevention of cross-contamination must be installed. The large amounts of amplicon that are generated in the rooms in which the actual amplification step takes place must be contained. Physical separation must be realised between activities of reagent preparation, sample handling and amplification & detection. Another approach is the containment of the source of contamination: the amplification reaction itself. This can be achieved by applying closed- system amplification and detection by which the reaction vessel is never opened again. Alternatively, and possible when applying PCR, or DNA amplification is the incorporation of uracil-DNA glycosylase in the system. This enzyme specifically degradates uracyl containing sequences that are incorporated in the amplicon prior to the initiation of new amplification reactions (Ferre et al . , 1996 In : A labora tory guide to RNA . Isola tion , analysis and synthesis . P. A. Krieg, ed. pp. 175-221. Wiley-Liss, New York) The sensitivity of the system is primarily determined by the amount of detectable targets. Such targets can then be amplified by PCR or NASBA. Other methods of amplification are so-called signal amplification methods of which branched DNA (bDNA) and LCR (ligase chain reaction) are examples. Amplification targets can comprise RNA or DNA. In the case of ribosomal RNA 1,000 to 10,000 copies are normally present in growing cells. This copy number is much higher than that of genomic targets that are generally present in 1 to 20 copies. The unique advantage of RNA target amplification by, e.g., reverse transcriptase PCR (RT-PCR) or NASBA is therefore that these methods are superior in sensitivity than the DNA amplification methods for detection of bacterial cells, e.g., by PCR. Not only the actual assay (DNA amplification method) or amplicon detection parts are important in the method of detection of M. para tuberculosis, also the data analysis and specifically the sample pre-treatment are critical processes to be assigned. Sample preparation relates to the liberation of DNA or RNA from M. paratuberculosis cells present in sample .

RNA targets are present at much higher frequencies then are DNA targets. Sensitivity is thus improved by using RNA as the starting material for nucleic acid amplification based detection methods. As a result, RT-PCR, NASBA, or TMA technologies are preferred. The availability of unique RNA targets allows the combination of such amplification technologies with RMNA targeted fluorescence in situ hybridization (FISH) technology. Strong fluorescent signals are required for the detection of Myb . avium subsp. para tuberculosis cells in faecal samples by FISH. It may be expected that the slowly growing cells contain very few ribosomes. Signal amplification protocols such as proposed in EP 97.20.2618 can be used to improve the signal in FISH procedures. The use of FISH in the detection of M. para tuberculosis in faecal samples may require the selective enrichment of M. para tuberculosis cells from the autochthonous background population (e.g., by cultivation or immunomagnetic capture) . Rapid microscopic confirmation of samples that have tested positive by nucleic acid amplification methods can be realised through FISH. The high reliability and confirmation possibilities of FISH and the high sensitivity and speed of the nucleic acid amplification technologies makes parallel development of both FISH and NASBA or RT-PCR extremely valuable.

For the extraction of RNA from pure cultures of bacteria or from relatively simple matrices, high-throughput solid phase extraction procedures are commercially available. An important aspect of the RNA extraction reaction is the purity of the final extract. The faecal matrix is known to contain substances that exert significant inhibitory action on the enzymes of nucleic acid amplification technologies that cannot be easily removed from the extract. Bilirubin and bile-salts are known to inhibit the PCR reaction at concentrations as low as 10 to 50 milligrams per millilitre (Widjojoatmodjo et al . , 1992. J Clin Microbiol 30:3195-3199). Several procedures have been developed to improve nucleic acid amplification from the fecal matrix. Among these are

100-fold (Varela et al . , 1994. J Clin Microbiol 32:1246-1248) to 500 fold (Widjojoatmodjo et al . , 1992. J Clin Microbiol 30:3195-3199) dilutions of the fecal samples to reduce inhibition. Alternatively, ion-exchange column purification (Kato et al . _f 1993. J Infect Dis 167:455-458) or cetyltrimethylammonium bromide treatment of the extract (Jiang et al . , 1992. J Clin Microbiol 30:2529-2534) can be used to reduce the amount of inhibiting compounds. Also glass-matrix precipitation (Stacy-Phipps et al . , 1995. J Clin Microbiol 33:1054-1059) or alternative resins or the use of chaotropic compounds in the extraction (Shieh et al . , 1995. J Virol Methods 54:51-66) can improve enzymatic nucleic acid amplification. The reliability of accurately discriminating between point mutations in amplification reactions can substantially be increased by using competitor primers or, preferably, PCR clamping with PNA (ørum, H., et al. 1993. Nucleic Acids Res. 21:5332-5336).

Possibilities for detection of the produced amplicon are infinite. Technologies such as the chemiluminescence based hybridization protection (HP) assay with acridinium- esters (AE) (Gen-Probe) , fluorescence based technologies such as the use of molecular beacons (Tyagi and Kramer. 1996. Nature Biotechnol. 14:303-308), Taq-man procedures (Perkin Elmer) , the FRET principle (Roche Diagnostics) or the use of intercalating dyes can be used to detect the amplicons in various ways that all have their own specificities. Sandwich hybridization assays in combination with magnetic bead capture formats can also be applied. Apart from the fact that an RNA-based approach results in a much higher amount of initial target or amplification-template this approach has another unique advantage: contamination problems are significantly reduced since RNA amplicons are much more labile than DNA amplicons and will naturally deteriorate.

Legends to the figures.

Figure 1

Comparative sequence alignment at the four sites in the 23S rRNA gene region that contain mutations specific for Mycobacterium avium subsp. para tuberculosis .

The short names represent the following strains: paratGDR, Mycobacterium avium subspecies para tuberculosis strain GDR; par44135, Mycobacterium avium subspecies para tuberculosis strain DSM 44135; parbovE5, Mycobacterium avium subspecies para tuberculosis strain Spbov E5; paratGie, Mycobacterium avium subspecies para tuberculosis strain J2A; parl9698, Mycobacterium avium subspecies para tuberculosis strain ATCC 19698 (Type strain); avi44157, Mycobacterium avium subspecies avium strain DSM 44157; avi25291,

Mycobacterium avium subspecies avium strain strain ATCC 25291 (Type strain) ; avi43216, Mycobacterium avium subspecies avium strain DSM 43216; avi44158, Mycobacterium avium subspecies avium strain DSM 44158; MYCAVIUM, Mycobacterium avium subspecies avium strain strain 23435; intl3950, Mycobacterium intracell ulare strain ATTC 13950 (Type strain) ; sil44175, Mycobacterium avium subspecies silvaticum strain DSM 44175 (Type strain) ; 23SEcoli, Escherichia coli 23S rRNA gene sequence Genbank J01695. The first 5 species represent well characterized strains of Mycobacterium avium subsp. para tuberculosis . The next 5 strains represent well characterized strains of Mycobacterium avium subsp. avium . The four mutations that form part of this invention are numbered 754, 1363, 3093 and 3188. The 3188 mutation is positioned in the ITS region between the 23S and 5S rRNA genes. The boxed regions represent the thymine residue in the rRNA operon of Mycobacterium avium subsp. para tuberculosis that form the basis of subspecies-specific detection methods as embodied in the present invention.

Figure 2

Sequence alignment of the complete 23S rRNA gene of selected Mycoba cterium species. The data presented in this alignment show that only positions numbered 754, 1363, 3093 and 3188 harbour sequence identities unique to the subspecies para tuberculosis of the Mycobacterium avium complex. The species abbreviations are as in figure 1.

Figure 3.

Comparative sequence alignment of a 656 basepair fragment of the 23S rRNA gene of selected Mycobacterium species that overlaps with mutation 1363 in figs. 1 and 2. The first 19 sequences represent the fragment of well characterized strains as well as field isolates of (presumably) Mycobacterium avium subsp. para tuberculosis . The next 8 sequences represent the fragment of well characterized strains as well as field isolates of (presumably) Mycobacterium avium subsp. avium . This figure shows that the mutation is also present in field isolates of Myco-acteriu-r- avium subsp. para tuberculosis . The presence of mutation 1363 for para tuberculosis is visible on position 150 of this alignment.

Figure 4

Temperature gradient PCR performed with the two point-mutation primers that anneal specifically with the

Mycobacterium avium subsp. paratuberculosis 23S rRNA gene at the position of mutations 754 and 1363. The figure represents an agarose gel of PCR products obtained after amplification of a portion of the 23S rRNA genes from DNA from Mycobacteriuir- avium subspecies paratuberculosis strain GDR (all upper slots of the gel) and Mycobacterium avium subspecies avium strain 97-613 (all lower slots of the gel) . The primer pair consisted of primer 20F and primer 22R (as described in table 2) . A series of 10 different annealing temperatures was applied by using the eppendorf Master cycler gradient apparatus. From left to right over the gel, PCR products from reactions with increasing annealing temperatures were applied. At an annealing temperature of

66.5 °C, no PCR product is generated from Mycobacterium avium subspecies avium, whereas a PCR product with this primer set can still be obtained even at annealing temperatures of 68.9 °C

Figure 5

Paratu-berculosis-specific PCR performed with the two 23S rRNA point-mutation primers used in the experiment of fig. 4. The figure represents agarose gels of PCR products obtained after amplification of a portion of the 23S rRNA genes from DNA isolated from a large number of well characterized strains as well as field isolates of (presumably) Mycobacterium avium subsp. paratuberculosis (top) and of (presumably) MycoJbacterium avium subsp. avium (bottom) . An annealing temperature of 68 °C was used. No product of the specific length (approximately 600 bases) was obtained from DNA isolated from Mycobacterium avium subsp. avium strains.

In the upper gel, the products of paratG14, paratG53, paratG63, parat390, parat437, parat442, parat444, parat434, parat421, parat412, parat423, parat424, parat415, and paratG32 were obtained from PCR reactions with DNA from field isolates of Mycobacterium avium subsp. paratuberculosis . The other para tuberculosis strains are coded as described in the legend to figure 1. The codes of the strains used in the lower gel were as described in the legend to figure 1. The codes avi97613 and avi97675 represent field isolates of Mycobacterium avium subsp. avium.

Figure 6. Probes and primers used for the enzymatic amplification of 23S ribosomal RNA sequences of M. paratuberculosis by NASBA or RT- (nested) -PCR.

Figure 7. Effect of primer design on NASBA amplification of 23S rRNA sequences from M. para tuberculosis . Primerset A was chosen for further studies. By using the optimized primer set, a detection sentitivity of less than 10 cells per ml of buffer could be attained (see text) .

Figure 8. Effect of cell lysis on detection of M. para tuberculosis cells by NASBA. By using forced cell disruption or selective lysis much higher detection sensitivities could be attained. TRI-reagent itself could be used as a selective lysis environment.

Figure 9. Southern blots from RT-PCR amplification of RNA extracted from pure cultures of M. para tuberculosis strain GDR by TRI reagent with (above) and without (below) additional bead beating procedures. Detection by RT-PCR is as sensitive as NASBA (i.e., less than 10 cells per ml).

Figure 10. Effect of matrix inhibition on the enzymatic amplification of 23S ribosomal RNA sequences of M. para tuberculosis by NASBA. A total amount of RNA extracted from 1,000,000 cells of M. paratuberculosis strain GDR was added to crude manure RNA extract (see text) . Figure 11. NASBA amplification and detection of M. para tuberculosis cells in bovine manure. Non-target cell lysis was achieved by 4% NaOH treatment. Assay sensitivity was further improved by CF-11 adsorption (see text) .

Figure 12. Southern blot from RT-nested-PCR amplification of M. para tuberculosis cells in bovine manure. As little as 10 cells per gram of manure could be detected by the nested PCR approach. P.C., positive control.

References

1. Thorel, M.F., et al. 1990. Numerical taxonomy of mycobactin-dependent' mycobacteria, emended description of Mycobacterium avium, and description of Mycobacterium avium subsp. avium subsp. nov., Mycobacterium avium subsp. paratuberculosis subsp. nov. and Mycobacterium avium subsp. silva ticum subsp. nov. Int. J. System. Bacteriol. 40:254-260.

2. Good, R.C. 1992. The genus Mycobacterium - Medical. In : The Prokaryotes : a handbook on the biology of bacteria : ecophysiology, isola tion , iden tifica tion , applica tions . 2^nd ed. A. Balows et al., editors. Vol 2. pp. 1243. Springer Verlag, New York.

3. Goodfellow, M. and L.G. Wayne. 1982. Taxonomy and nomenclature, p. 471-521. In : C. Ratledge, and J. Stanford (ed.), The biology of the mycobacteria, vol. 1. Physiology, identification and classification. Academic Press, London.

4. Kent, P.T., and G.P. Kubica. 1985. Public health mycobacteriology : a guide for the level III laboratory. U.S. Department of Health and Human Services, Centers for Disease Control, Atlanta. 5. Siddiqi, S.H. 1988. BACTEC TB system: product and procedure manual. Becton Dickinson Diagnostic Instrument Systems, Towson. MD.

6. McFadden, J.J., et al . 1987a. The use of DNA probes identifying restriction-fragment-length-polymorphisms to examine the Mycobacterium avium complex. Mol. Microbiol. 1:283-291.

7. Vary, P.H., et al. 1990. Use of highly specific DNA probes and the polymerase chain reaction to detect Mycobacterium para tuberculosis in Johne' s disease. J. Clin. Microbiol. 28 (5) : 933-937. 8. Moss, M.T., et al. 1991. Specific detection of Mycobacterium para tuberculosis by DNA hybridisation with a fragment of the insertion element IS900. Gut 32 (4 ): 395-398.

9. Benedictus, G. 1985. Some aspects of bovine paratuberculosis. Academic thesis, Utrecht University, The Netherlands .

10. Wayne, L.G., et al . 1993. Serovar determination and molecular taxonomic correlation in Mycobacterium avium, Mycobacterium intracellulare, and Mycobacterium scrofulaceum : a cooperative study on the international working group on mycobacterial taxonomy. Int. J. System. Bacteriol. 43(3) :482- 489.

11. McFadden, J.J., et al . 1987b. Determination of genome size and DNA homology between unclassified mycobacterium species isolated from patients with Crohn' s disease and other mycobacteria. J. Gen. Microbiol. 133:211-213.

12. Saxegaard, F., and I. Baess. 1988. Relationship between Mycobacterium avium, Mycobacterium para tuberculosis , and ^λwood pigeon mycobacteri-i . ' Acta Pathol. Microbiol. Immunol. Scand. 96:37-42.

13. Yoshimura, H.H., and D.Y. Graham. 1988. Nucleic acid hybridization studies of mycobactin-dependent mycobacteria. J. Clin. Microbiol. 26:1309-1312.

14. Van der Giessen, J.W.B., et al. 1992. Evaluation of the abilities of three diagnostic tests based on the polymerase chain reaction to detect Mycobacterium para tuberculosis in cattle: application in a control program^'. J. Clin. Microbiol. 30 (5) : 1216-1219.

15. Millar, D.S., et al. 1995. Solid-phase hybridization capture of low-abundance target DNA sequences: application to the polymerase chain reaction detection of Mycobacterium para tuberculosis and Mycobacterium avium subsp. silva ticum . Anal. Bioche . 226 (2) : 325-330. 16. Kunze, Z.M., et al. 1991. IS901, a new member of a widespread class of atypical insertion sequences, is associated with pathogenicity in Mycobacterium avium . Mol. Microbiol. 5 (9) : 2265-2272. 17. Moss, M.T., et al. 1992. IS902, an insertion element in the chronic enteritidis-causing Mycobacterium avium subsp. silva ticum . J. Gen. Microbiol. 138:139-145.

18. Frothingham, R. , et al. 1994. Extensive DNA sequence conservation throughout the Mycobacterium tuberculosis complex. J. Clin. Microbiol. 32 (7 ): 1639-1643.

19. Thoresen, O.F. and I. Olsaker. 1994. Distribution and hybridization patterns of the insertion element IS900 in clinical isolates of Mycobacterium para tuberculosis . Vet. Microbiol. 40:293-303. 20. Bauerfeind, R. , et al . 1996. Molecular characterization of Mycobacterium para tuberculosis isolates from sheep, goats, and cattle by hybridization with a DNA probe to insertion element IS900. J.Clin. Microbiol. 34 (7 ): 1617-1621.

21. Kunimoto, D., et al . 1994. Sequence homologous to IS900 present in some M. avium (abstract). Am. Soc. Microbiol.

9:182.

22. Roiz, .M.P., et al. 1995. Use of restriction fragment length polymorphism as a genetic marker for typing Mycoba cterium avium strains. J. Clin. Microbiol. 33:1389- 1391.

23. El-Zaatari, F.A.K., et al . 1997. Detection of Mycobacterium avium complex in cerebrospinal fluid of a sarcoid patient by specific polymerase chain reaction. Scand. J. Infect. Dis. 29:202-204. 24. Widjojoatmodjo, M.N., et al . 1992. The magnetic immuno polymerase chain reaction assay for direct detection of salmonellae in fecal samples. J Clin Microbiol 30:3195-3199. 25. Varela, P., et al . 1994. Direct detection of Vibrio cholerae in stool samples. J Clin Microbiol 32:1246-1248 26. Van der Giessen, J.W.B. 1993. A molecular approach to the diagnosis and control of bovine paratuberculosis. Academic thesis, Utrecht University, The Netherlands.

27. Evans, K.D., et al. 1992. Identification of Mycobacterium tuberculosis and Mycojbacterium avium - M. intracellulare directly from primary BACTEC cultures by using acridinium ester labeled DNA probes. J. Clin. Microbiol. 30:2427-2431.

28. Sockett, D.C., et al. 1992. Evaluation of conventional and radiometric fecal culture and a commercial DNA probe for diagnosis of Mycobacterium para tuberculosis infections in cattle. Can. J. Vet. Res. 56 (2) : 148-153.

29. Cousins, D.V., et al. 1995. Use of BACTEC radiometric culture method and polymerase chain reaction for the rapid screening of faeces and tissues for Mycobacterium para tuberculosis . Aust. Vet. J. 72 (12) : 458-462

30. Whittington, R.J., et al. 1998. Rapid detection of Mycobacterium para tuberculosis in clinical samples from ruminants and in spiked environmental samples by modified BACTEC 12B radiometric culture and direct confirmation by IS900 PCR. J. Clin. Microbiol. 36 (3) : 701-707.

31. Grant, I.R., et al. 1998. Isolation of Mycobacterium para tuberculosis from milk by immunomagnetic separation. Appl. Environ. Microbiol. 64 (9) : 3153-3158. 32. Poupart, P., et al. 1993. Preparation of a specific RNA probe for detection of Mycobacterium para tuberculosis and diagnosis of Johne's disease. J. Clin. Microbiol. 31(6):1601- 1605.

33. Gormley, E., et al . 1997. Identification and differentiation of mycobacteria using the PAN promotor sequence from Mycobacterium para tuberculosis as a DNA probe. FEMS Microbiol. Lett. 147 (1) : 63-68.

34. Ellingson, J.L., et al. 1998. Identification of a gene unique to Mycobacterium avium subspecies paratujbercuiosis and application to diagnosis of paratuberculosis. Mol. Cell. Probes 12 (3) :133-142.

35. Van der Giessen, J.W.B., et al . 1992. Amplification of 16S ribosomal RNA sequences to detect Mycobacterium para tuberculosis . J. Med. Microbiol. 36:255-263.

36. Van der Giessen, J.W.B., et al . 1994. Comparison of the 23S ribosomal RNA genes and the spacer region between the 16S en 23S rRNA genes of the closely related Mycobacterium avium and Mycobacterium paratuberculosis and the fast-growing Mycobacterium phlei . Microbiology (UK) 140:1103-1108.

37. Stone, B.B., et al. 1995. Comparison of Mycobacterium 23S rRNA sequences by high-temperature reverse transcription and PCR. Int. J. System Bacteriol. 45:811-819.

38. Pace, N.R., et al. 1985. Analyzing natural microbial populations by rRNA sequences. ASM News 51:4-12.

39. Woese, C.R. 1987. Bacterial evolution. Microbiol. Rev. 51:221-271.

40. Olsen, G.J., et al . 1986. Microbial ecology and evolution: a ribosomal RNA approach. Ann. Rev. Microbiol. 40:337-365.

41. Frothingham. R, et al. 1993. Sequence-based differentiation of strains in the Mycobacterium avium complex. J. Bacteriol. 175 (10) : 2818-2825.

42. Glennon, M., et al . 1994. The ribosomal intergenic spacer region: a target for the PCR based diagnosis of tuberculosis. Tuber. Lung Dis. 75 (5) : 353-360.

43. Frothingham, R and K.H. Wilson. 1994. Molecular phylogeny of the Mycobacterium avium complex demonstrates clinically meaningful divisions. J. Infect. Dis. 169(2) :305- 312.

44. Ji, Y.E., et al. 1994. Nucleotide sequences of the spacer-1, spacer-2 and trailer regions of the rrn operons and secondary structures of precursor 23S rRNAs and precursor 5S rRNAs of slow-growing mycobacteria. Microbiology. 140 (Pt7) :1763-1773.

45. Bourque, S.N., et al. 1995. Comparative analysis of the 16S to 23S ribosomal intergenic spacer sequences of Bacillus thuringiensis strains and subspecies and of closely related species. Appl . Environ. Microbiol. 61 (4 ): 1623-1626.

46. De Wit, M.Y.L. and P.R. Klatser. 1994. Mycobacterium leprae isolates from different sources have identical sequences of the spacer region between the 16S and 23S ribosomal RNA genes. Microbiology 140:1983-1987.

47. Scheibl, P. and G.F. Gerlach. 1997. Differentiation of Mycobacterium para tuberculosis isolates by rDNA-spacer analysis and random amplified polymorphic DNA patterns. Vet. Microbiol. 57 (2-3) : 151-158. 48. Mullis et al . 1987. Process for amplifying, detecting, and/or-cloning nucleic acid sequences. US patent 4,683,195. 49.Enns, R.K. 1987. Clinical studies summary report: the Gen- Probe® rapid diagnostic system for the Mycobacterium avium complex. Gen-Probe, Inc., San Diego, Calif. 50. Jonas, V., et al. 1993. Detection and identification of Mycobacterium tuberculosis directly from sputum sediments by amplification of rRNA. J. Clin. Microbiol. 31 (9) : 2410-2416. 51. Compton, J. 1991. Nucleic acid sequence-based amplification. Nature (London) 350:91-92. 52. øru , H. , et al. 1993. Single base pair mutation analysis by PNA directed PCR clamping. Nucleic Acids Res. 21:5332-5336. Table 1. Organisms used in this study

Mycobacterium avium subspecies paratujbercuiosis strain GDR Mycobacterium avium subspecies para tuberculosis strain DSM 44135

Mycobacterium avium subspecies para tuberculosis strain Spbov

E5

Mycobacterium avium subspecies paratuberculosis strain ATCC

19698 (Type strain) Mycobacterium avium subspecies avium strain 23435

Mycobacterium avium subspecies avium strain DSM 44158 Mycobacterium avium subspecies avium strain DSM 43216 Mycobacterium avium subspecies avium strain DSM 44157 Mycobacterium avium subspecies avium strain ATCC 25291 (Type strain)

Mycobacterium avium subspecies avium strain 97-613 Mycobacterium avium subspecies silva ticum strain DSM 44175 (Type strain) ycojbacteriufl- intracellulare strain ATTC 13950 (Type strain)

Table 2

Primers for point-mutation PCR. The site of the "mutation" is indicated. Primers with code "F" are forward primers, those with code "R" are reverse primers.

Code sequence (5'-> 3') Td (°C)

20F CTGAATAGGGCGCATCCCfjj 60

19F TGAATAGGGCGCATCCCffT 58 IF GAATAGGGCGCATCCC|TT 58

3F AATAGGGCGCATCCCf|TTT 56

5F ATAGGGCGCATCCCffTTTG 58

7F TAGGGCGCATCCC|TTTGG 60

9F AGGGCGCATCCClJTTTGGG 62 11F GGGCGCATCCC|TTTGGGG 64

13F GGCGCATCCC|TTTGGGGT 62

22R CTCCCTCCACCACCGg 54

21R TCCCTCCACCACCGlT 52

15R CCCTCCACCACCG|TC 54 16R CCTCCACCACCG|TCA 52

17R CTCCACCACCGffJTCAC 52

18R CACCACCG-f§TCACCCG 54

22Ra CACCCTCCACCACCGl 54

21Ra ACCCTCCACCACCGiT 52

Claims

1. A method for specifically detecting nucleic acid derived from a causal agent of Johne ' s disease in a sample whereby nucleic acid from Mycobacterium avium subspecies para tuberculosis is discriminated from nucleic acid from other Mycobacterium avium subspecies comprising detecting in said nucleic acid a mutation specifically conserved for Mycobacterium avium subspecies paratuberculosis .

2. A method according to claim 1 wherein said nucleic acid is derived from 23S ribosomal RNA.

3. A method according to claim 2 wherein said ribosomal RNA comprises a nucleic acid as shown in figure 1.

4. A method according to claim 2 or 3 wherein said conserved mutation is located at position 754, 1363 or 3093 as shown in figure 2.

5. A method according to anyone of claims 1 to 4 further comprising treatment of said sample to selectively lyse at least a part of non-mycor-acterial matter.

6. A method according to claim 1 to 5 wherein said sample is a ruminant sample, preferably a faecal sample.

7. A method according to claim 6 further comprising nucleic acid amplification.

8. A method according to claim 6 or 7 further comprising hybridisation.

9. A method according to claim 6 further comprising in situ hybridisation.

10. A nucleic acid probe or primer for use in a method according to claims 6 to 9.

11. A diagnostic kit comprising a probe or primer according to claim 10.

12. A method for detecting at least one ruminant infected with a causal agent of Johne ' s disease comprising obtaining a sample from said cow and testing said sample for the presence of nucleic acid using a method according to anyone of claims 1 to 9.

13. A method according to claim 12 further comprising culling said ruminant from said herd.

14. A method according to claim 12 or 13 wherein said ruminant is a cow.

15. Use of a method according to anyone of claims 1 to 9 or of a probe or primer according to claim 10 for detecting a ruminant infected with a causal agent of Johne ' s disease.