Genomic Polymorphisms for Mycobacterium avium subsp. paratuberculosis Diagnostics
http://www.100md.com
微生物临床杂志 2005年第8期
McGill University Health Centre, Montreal, Quebec, Canada H3G 1A4
National Institute of Public Health and the Environment, 3720BA Bilthoven, The Netherlands
Australian Reference Laboratory for Bovine Tuberculosis, Department of Agriculture, South Perth, Western Australia, Australia 6151
ABSTRACT
Mycobacterium avium subsp. paratuberculosis is an emerging pathogen of mammals and is being actively investigated as a possible zoonotic agent. The lack of reliable diagnostic assays has hampered rational assessment of the prevalence of this organism in humans and animals. We have used a comparative genomic approach to reveal genomic differences between M. avium subsp. paratuberculosis and its close relative M. avium subsp. avium, a highly prevalent environmental organism. From computational and DNA microarray-based study of two prototype strains, M. avium subsp. avium strain 104 and M. avium subsp. paratuberculosis strain K10, we have uncovered two types of large sequence polymorphisms (LSPs): those present in the former but missing in the latter (LSPAs) and those only present in the latter (LSPPs). We examined the distribution of 3 LSPAs and 17 LSPPs across a panel of 383 M. avium complex isolates in order to determine their potential utility for the development of accurate diagnostic tests. Our results show that the absence of LSPA8 is 100% specific for the identification of M. avium subsp. paratuberculosis. Of the 17 LSPPs, 10 regions were not specific for M. avium subsp. paratuberculosis while 7 were shown to be highly specific (>98%) and, in some cases, highly sensitive as well (up to 95%). These data highlight the need to evaluate these regions across a diverse panel of clinical and environmental isolates and indicate the LSPs best suited for M. avium subsp. paratuberculosis diagnostics.
INTRODUCTION
Mycobacterium avium subsp. paratuberculosis is a serious pathogen of domestic ruminants, in which it causes paratuberculosis (Johne's disease), a chronic and eventually fatal inflammatory bowel disease. Its rising incidence in many regions of the world has resulted in significant economic losses to the livestock industry (13). This organism also infects free-ranging wildlife species (16). The impact of M. avium subsp. paratuberculosis in nonruminant wildlife species is largely unknown; however, the potential for interspecies transmission has important implications for paratuberculosis control programs. M. avium subsp. paratuberculosis is also being actively investigated as the possible cause of a debilitating human inflammatory bowel disease, Crohn's disease (4, 5, 18).
Effective control of Johne's disease and investigation of the potential link to Crohn's disease have been hampered by the lack of effective assays to easily and accurately diagnose M. avium subsp. paratuberculosis infections. Commercially available serological assays for bovine disease are convenient but offer poor sensitivity, especially during subclinical disease. Moreover, owing to the high degree of genetic similarity between M. avium subsp. paratuberculosis and the other members of the M. avium complex, many of the proteins that are recognized during the early stages of infection are not specific for M. avium subsp. paratuberculosis (14). Since M. avium subsp. avium is highly prevalent in the environment but not associated with a similar clinical syndrome in mammals, the capacity to differentiate between these closely related organisms is essential for rational clinical and epidemiologic assessment.
Previous work has identified a number of M. avium subsp. paratuberculosis genetic sequences that are absent from other mycobacteria, such as the insertion sequence IS900 (11), the F57 element (21), and the hspX gene (9), but the value of these sequences for diagnostics of M. avium subsp. paratuberculosis remains unclear. Although PCR testing for the multicopy insertion element IS900 is widely used, there are concerns about its specificity and its validity as a direct proxy for M. avium subsp. paratuberculosis. IS900-like elements have been found in unrelated organisms (10), and other M. avium complex insertion elements, including IS1311 (7) and IS1626 (22), share considerable sequence similarity. In the cases of F57 and hspX, their specificity for M. avium subsp. paratuberculosis has not to date been rigorously evaluated in a large sample of clinical isolates.
More recently, data derived from genome sequencing projects have suggested a number of polymorphic genomic regions that may serve in the specific diagnosis of M. avium subsp. paratuberculosis. Our microarray-based comparisons using M. avium subsp. avium strain 104 as the reference revealed 14 large sequence polymorphisms (LSPs) that are variably present among a small collection of M. avium isolates. Three of these regions appeared to be present in M. avium subsp. avium (LSPAs) and absent from M. avium subsp. paratuberculosis isolates (23). Conversely, computational comparisons of the genome sequences of M. avium subsp. paratuberculosis strain K10 (GenBank accession no. NC_002944) and M. avium subsp. avium strain 104 (http://www.tigr.org) have identified DNA sequences that are present in the former but missing or divergent in the latter (2, 3, 19). We have to date identified 17 regions varying in length from 2.9 to 66 kb that are unique to M. avium subsp. paratuberculosis strain K10; we call these LSPPs. Such sequences can be expected to lend themselves to nucleic acid-based diagnostic tests and, if they encode immunogenic proteins, to immunological assays as well. One of these, a 19-kb sequence that we call LSPP12, has been previously documented (8, 23). Another, a 98-kb segment of the genome that encompasses elements we call LSPP14 and LSPP15, also includes within it a 38-kb element that was recently described as an M. avium subsp. paratuberculosis-specific putative pathogenicity island (24).
The objective of our study was to assess the value of these LSPs for the specific diagnosis of M. avium subsp. paratuberculosis. To be diagnostically specific for M. avium subsp. paratuberculosis, an LSPP must be found only in M. avium subsp. paratuberculosis isolates and should be absent from non-M. avium subsp. paratuberculosis isolates. Conversely, for an LSPA to serve in the molecular diagnosis of M. avium subsp. paratuberculosis, it should be consistently present in non-M. avium subsp. paratuberculosis isolates but missing from a broad collection of M. avium subsp. paratuberculosis isolates. We therefore tested the distribution of the 17 LSPPs and 3 LSPAs across a panel of 383 M. avium complex isolates. Our results indicate that many LSP regions, although distinct between prototype genome sequences, are heterogeneously distributed across geographically diverse isolates and so lack the specificity required for diagnostics. However, a subset of LSPs do appear highly specific for M. avium subsp. paratuberculosis and should prove useful in the development of effective diagnostics for Johne's disease and evaluation of the Crohn's disease hypothesis.
MATERIALS AND METHODS
Bacterial isolates. A panel of 107 M. avium subsp. paratuberculosis, 260 M. avium subsp. avium, 4 M. avium subsp. silvaticum, and 12 M. intracellulare isolates was assembled (Table 1). Isolates were first identified as M. avium complex species by the laboratories providing them using the commercial nucleic acid probe AccuProbe (Gen-Probe Inc., San Diego, CA). Initial dependence on Mycobactin J and the presence of the insertion sequence IS900 were determined by these laboratories and used to identify M. avium strains to the subspecies level as M. avium subsp. paratuberculosis. Designation as M. avium subsp. avium was supported by phenotypic characteristics (including lack of Mycobactin J dependency) and the presence of IS1245. Additional extensive testing, including sequencing of the 16S-23S internal transcribed spacer (ITS) region, was completed for 160 of these isolates in the context of another study (17). M. avium subsp. silvaticum strains exhibited dependence on Mycobactin J on primary isolation and were IS900 negative. M. intracellulare strains were identified using the M. intracellulare AccuProbe, and species designation was confirmed by 16S rRNA gene sequencing.
Genomic DNA was extracted from each isolate using previously published methods (27). The amount of DNA in each sample was quantified using a spectrophotometer, and each sample was normalized to a concentration of 1 ng/μl prior to PCR testing.
Discovery of LSPP and testing across M. avium complex isolates. As part of our ongoing work to characterize M. avium complex organisms, we compared the sequenced genomes of M. avium subsp. avium strain 104 (http://www.tigr.org) and M. avium subsp. paratuberculosis strain K10 (GenBank accession no. NC_002944). Artemis software (http://www.sanger.ac.uk/Software/Artemis/) was used for visualization of the M. avium subsp. paratuberculosis strain K10 genome data and annotation of M. avium subsp. avium strain 104. Genome sequence comparison, via visual inspection in Artemis and with BLASTN, permitted identification of regions present in M. avium subsp. paratuberculosis strain K10 but lacking a homologous region in M. avium subsp. avium strain 104 (LSPP). Additional homology analyses (using BLASTN) were used to confirm that LSPP regions were truly missing and not simply displaced (e.g., due to genomic rearrangement). BLAST2 was used to precisely define the boundaries of these novel LSPPs. For the 17 regions and their locations in the M. avium subsp. paratuberculosis strain K10 genome, see Table 3.
To screen for the specificity of LSPPs, we initially tested for their presence by PCR targeting a short internal sequence across a panel of 96 M. avium subsp. avium isolates. LSPPs for which a screening PCR was negative for these M. avium subsp. avium strains were then tested across the extended panel of 383 M. avium complex isolates. Since LSPP14, which includes but is not restricted to a known 38-kb island (24), and LSPP12, a 19-kb sequence (8, 23), were previously postulated to be of potential diagnostic utility, we tested each of these LSPs with two and three different sets of primers targeting distinct loci, respectively. Primer sequences for each of the LSPPs tested are listed in Table 2.
Testing of LSPA across M. avium complex isolates. Our previous DNA microarray-based analysis revealed numerous large sequences variable among M. avium isolates. LSPA8, LSPA13, and LSPA14 appeared potentially specific for M. avium subsp. paratuberculosis (23). When aligned against the genome of M. avium subsp. paratuberculosis strain K10, these M. avium subsp. avium strain 104 sequences were confirmed to be absent from the M. avium subsp. paratuberculosis K10 genome. Testing for these regions employed PCR with primers designed toward the flanking regions of the sequence, such that an amplicon would be generated only if the region was indeed missing. The PCR test was considered positive (region missing) if amplification generated a product of the expected molecular weight. Furthermore, for each set of primers, a random selection of 25 PCR products, originating from 25 different strains, was sequenced using ABI dye terminator chemistry. The resulting sequences were aligned with the sequenced genomes of M. avium subsp. paratuberculosis strain K10 and of M. avium subsp. avium strain 104 to determine if the junction of the deleted regions was identical across the different isolates.
For isolates in which amplification did not occur using the flanking primers, additional PCR testing was performed (using sets of primers targeting sequences within the LSPA) to determine whether the intervening sequence was present. An example is LSPA8 in Fig. 1, with corresponding positions on the genomes of M. avium subsp. avium strain 104, M. avium subsp. paratuberculosis strain K10, and M. tuberculosis strain H37Rv. The primer sequences used for all reactions are given in Table 2.
PCRs. All PCRs were performed in 50-μl volumes, using 5 μl of template DNA (5 ng) and 1 U Taq polymerase (MBI Fermentas), with 5 μl of 10x PCR buffer (MBI Fermentas), 2.5 mM MgCl2, 5 μl acetamide 50% (wt/vol), 0.2 mM deoxynucleoside triphosphates, and each primer at 0.5 μM. PCRs were conducted in 96-well plates. Amplification consisted of an initial denaturation step of 94°C for 5 min, followed by 30 cycles of denaturation at 94°C for 30s, annealing at 58°C for 45s, and extension at 72°C for 2 min. PCR products were separated by electrophoresis in precast 1% (wt/vol) agarose gels containing ethidium bromide (ReadyAgarose gels; Bio-Rad Laboratories, Hercules, CA). Samples that had bands of the expected molecular weight were considered positive, whereas those with bands of different size or where no bands were seen were considered negative. When results were ambiguous, PCRs were repeated using individual closed tubes.
M. avium subsp. paratuberculosis strain K10 and M. avium subsp. avium strain 104 were used as positive and negative controls for each set of primers used to test for the LSPs.
RESULTS
LSPPs: regions present in M. avium subsp. paratuberculosis K10 and absent from M. avium subsp. avium 104. LSPPs represent four types of polymorphisms with respect to the sequenced genome of M. avium subsp. avium 104. For certain regions, the M. avium subsp. paratuberculosis K10 genome simply contains an additional element when aligned against the corresponding positions on the genomes of M. avium subsp. avium strain 104 and M. tuberculosis strain H37Rv; the element is flanked by regions of high homology and synteny among the three genomes (schematically depicted by the example of LSPP4 in Fig. 2A). For other regions, the M. avium subsp. paratuberculosis K10 genome does not contain additional DNA, but rather, there is a region of extremely low sequence similarity (<20%) with the best hit against the M. avium subsp. avium 104 genome, flanked by regions of high homology (>97%) between the two sequences (depicted as the example of LSPP2 in Fig. 2B). Some regions are characterized by both additional DNA and sequence with low similarity (examples of LSPP14 and LSPP15 in Fig. 3A). Finally, one region is the result of additional DNA in the M. avium subsp. paratuberculosis K10 genome and loss of DNA from the M. avium subsp. avium genome and thus is characterized by the replacement of a DNA segment with another (example of LSPP12, Fig. 3B).
For the 17 LSPPs, a positive PCR using primers internal to these sequences indicated the presence of at least part of the region tested. From the screening panel of 96 M. avium subsp. avium isolates, 10 of these sequences were variably present in M. avium subsp. avium isolates while 7 could not be amplified from the screening panel (LSPP2, LSPP4, LSPP11, LSPP12, LSPP14, LSPP15, and LSPP16; Table 3). These regions were thus investigated using an extended panel of 383 M. avium complex isolates; results are summarized in Table 4, with more detailed descriptions below.
LSPP2, LSPP4, and LSPP15 could not be amplified from any of the 276 non-M. avium subsp. paratuberculosis isolates, demonstrating 100% specificity for this subspecies.
LSPP11 and LSPP16 also exhibited high specificity but unexpectedly amplified from a group of six non-M. avium subsp. paratuberculosis strains: one isolate gave amplification products for both LSPP11 and LSPP16; four isolates gave amplification products for LSPP11 only, and one gave a product for LSPP16 only. The five LSPP11 and two LSPP6 amplicons obtained from these isolates were identical to M. avium subsp. paratuberculosis strain K10 sequences. To further characterize these six isolates, all clinical human isolates collected in Montreal, we performed additional analysis, including 16S rRNA sequencing (to confirm species designation) and 16S-23S ITS sequencing to identify their sequevar. By 16S analysis, one isolate (13373) had a novel 16S rRNA sequence, similar to M. chimaera (previously known as MAC-A) (26) and a unique ITS sequence, closest to the MAC-C sequevar. Four isolates had a 16S sequence consistent with M. avium and had ITS sequences identical to the common Mav-B sequevar. Finally, we determined that one isolate (9033) was potentially mixed, based on the presence of double peaks in the DNA sequence chromatograms, with features of both M. avium subsp. avium and M. intracellulare (Min-A sequevar).
The presence of LSPP12, a 19-kb sequence that is found in lieu of a 197-kb sequence in M. avium subsp. avium 104, was evaluated with three distinct sets of internal primers targeting different genes. Across the entire panel of M. avium complex isolates, amplification of two target genes (MAP2182c and MAP2187c) was restricted to M. avium subsp. paratuberculosis strains, resulting in a calculated specificity of 100%. However, primers targeting MAP2181c (predicted to encode a transcriptional regulator of the TetR family) unexpectedly produced an amplicon from two non-M. avium subsp. paratuberculosis isolates, clinical specimens 13373 and 9033 noted above. Sequencing of the MAP2181c amplicons revealed numerous sequence level polymorphisms shared across these isolates and only 89% sequence identity to M. avium subsp. paratuberculosis strain K10.
LSPP14 is a heterogeneous region (Fig. 3A). Some portions are extra to M. avium subsp. paratuberculosis K10, including a previously described 38-kb island (24), while other segments exhibit moderate sequence similarity to M. avium subsp. avium 104. LSPP14 was tested using two different sets of primers targeting MAP3726 and MAP3750 within the additional segment of DNA. Seven non-M. avium subsp. paratuberculosis isolates tested positive for at least one of the two genes. One M. avium subsp. avium isolate was positive for both MAP3726 and MAP3750, while six other isolates were positive for one of the two genes tested. In the case of MAP3726, a sequence with 97% similarity to that of M. avium subsp. paratuberculosis strain K10 was amplified from three M. avium subsp. avium strains. One of these isolates was from a pig in The Netherlands and had an ITS sequence identifying it as a Mav-B sequevar, while the other two were isolated from birds and had Mav-A sequevars. For MAP3750, sequences with less than 93% similarity to that of M. avium subsp. paratuberculosis strain K10 were amplified from five M. avium subsp. avium isolates, including the pig isolate described above. The presence of polymorphic sequence within M. avium subsp. avium isolates, and the fact that only a subset of M. avium subsp. paratuberculosis isolates were positive for these two genes, prompted us to look for polymorphisms within M. avium subsp. paratuberculosis isolates. Amplicons obtained from eight different isolates for each of these two genes also revealed several single-nucleotide polymorphisms (SNPs) compared to the genome of M. avium subsp. paratuberculosis K10. Specifically, of the eight isolates tested, one (obtained from a deer) had a SNP noted within MAP3726 (at position 400 starting from the first base pair of the gene) and an additional two SNPs within MAP3750 (positions 331 and 396). Three other isolates (two from cows, one from a sheep) had a shared SNP within MAP3750 (at position 384). These findings suggest that LSPP14 is polymorphic within M. avium subsp. paratuberculosis.
Of the seven LSPPs tested extensively, only LSPP12 and LSPP15 could be amplified from the majority of M. avium subsp. paratuberculosis isolates, suggesting either that the other LSPPs were not universally present in this subspecies or that sequence level polymorphisms prevented their amplification across some isolates. Despite our efforts to identify a pattern, it was not possible to correlate the distribution of these variable LSPP regions with respect to the host source of the isolate tested.
LSPAs: regions present in M. avium subsp. avium 104 and absent from M. avium subsp. paratuberculosis K10. A positive PCR with primers flanking the LSPAs indicated the absence of the region. PCR with flanking primers demonstrated the absence of LSPA8 in 94 M. avium subsp. paratuberculosis isolates. For the 289 isolates (including 13 of M. avium subsp. paratuberculosis) which failed to amplify using the flanking primers, we performed internal PCR testing using two different sets of primers targeting sequences within LSPA8 and found the sequence was present in all 276 non-M. avium subsp. paratuberculosis isolates. The 13 M. avium subsp. paratuberculosis isolates did not generate PCR products with either set of primers internal to this region. Therefore, it appears that this sequence is consistently missing from M. avium subsp. paratuberculosis isolates and is present in every other M. avium complex strain tested. Sequencing of 25 of the amplicons generated from M. avium subsp. paratuberculosis strains using the flanking primers revealed the same polymorphism in all M. avium subsp. paratuberculosis strains, with the junction point truncating two genes (called Maa4183 [the ortholog of MAP3636] and Maa4172 [not annotated in M. avium subsp. paratuberculosis]), homologs of Rv0197 and Rv0201 in M. tuberculosis, respectively. These results strongly suggest that absence of LSPA8 is the result of a genomic deletion from the ancestral genome.
LSPA13, a 19.5-kb sequence, was confirmed as missing in 167 isolates by PCR using primers flanking this sequence. Of these, 102 were M. avium subsp. paratuberculosis isolates, but also 64 of M. avium subsp. avium and 1 of M. avium subsp. silvaticum were shown to lack this region. Sequencing of the amplicons generated using the flanking primers in 25 M. avium subsp. paratuberculosis and 25 M. avium subsp. avium isolates demonstrated the exact same sequence, which is also identical to that of the M. avium subsp. paratuberculosis K10 type strain. LSPA13 is located at the tRNA-argV gene and is flanked by direct repeats, features characteristic of mobile genetic elements. Whatever its origin, the absence of this region was not specific to M. avium subsp. paratuberculosis.
LSPA14, a 23-kb region, was initially judged to be missing in 101 isolates, all of which were M. avium subsp. paratuberculosis, by PCR using flanking primers (set A). Upon sequencing, the amplicons obtained from 25 of these M. avium subsp. paratuberculosis isolates were identical to that of the prototype, M. avium subsp. paratuberculosis K10. When testing for the presence of LSPA14, an intervening sequence was shown to be present in 215 isolates (all non-M. avium subsp. paratuberculosis). However, 67 isolates did not give products with either the flanking or the internal primers, leading us to hypothesize that the sequence may be missing (as in M. avium subsp. paratuberculosis) but associated with polymorphisms in the flanking regions. To test the latter possibility, we designed a new set of primers (set B) to amplify across LSPA14, such that the resulting PCR products would encompass the sequences of the original flanking primers (set A). Using flanking primer set B, we determined that the region was indeed missing in these isolates. Upon sequencing the resulting PCR products, we noted that the LSPA14 junction point in these isolates was identical to that in M. avium subsp. paratuberculosis. However, the sequence of the flanking region in these isolates was only 97% identical to that of M. avium subsp. paratuberculosis. Further, two of the SNPs noted to be common to these non-M. avium subsp. paratuberculosis isolates occurred along the recognition site for one of the original flanking primers (set A). These results indicate that a positive PCR using flanking primer set A appears to be specific for M. avium subsp. paratuberculosis; however, absence of this region is not specific to M. avium subsp. paratuberculosis.
DISCUSSION
Members of the M. avium complex form a closely related group of bacteria, with M. avium subsp. avium and M. avium subsp. paratuberculosis sharing the same species designation. Despite their taxonomic relationship, M. avium subsp. avium and M. avium subsp. paratuberculosis exhibit markedly different capacities to cause disease across a number of hosts. In particular, M. avium subsp. paratuberculosis is emerging as a serious threat to many mammalian species, but control programs to detect and eradicate this pathogen have been hampered by the lack of sensitive, specific, and practical assays. The availability of sequence information for two prototype strains of the M. avium complex offers the opportunity to uncover M. avium subsp. paratuberculosis-specific sequences that may be employed for nucleic acid-based tests and development of immunological assays.
Using a comparative genomic approach, we have identified two types of markers: large sequences present in M. avium subsp. paratuberculosis K10 but not in M. avium subsp. avium 104 (LSPPs) and large sequences present in M. avium subsp. avium strain 104 but not in M. avium subsp. paratuberculosis (LSPAs). We evaluated the distribution of these markers across a large panel of M. avium complex isolates and determined that, of the latter, PCR-based testing for the loss of LSPA8 was 100% specific for the identification of M. avium subsp. paratuberculosis. Akin to the RD1 region of the M. tuberculosis complex, which permits accurate differentiation between virulent M. bovis and BCG strains (25), our results for LSPA8 have immediate applicability in the diagnostics of M. avium subsp. paratuberculosis. First, the presence of this region (by PCR testing of a sequence internal to LSPA8) can be used to determine that an isolate is highly unlikely to be M. avium subsp. paratuberculosis. Second, the absence of LSPA8 (by PCR testing using primers flanking these sequences) can be used to diagnose M. avium subsp. paratuberculosis with 100% specificity.
Of the 17 LSPPs, 10 had poor specificity for M. avium subsp. paratuberculosis, based on the amplification of genomic targets from these regions in non-M. avium subsp. paratuberculosis isolates, while 7 of the LSPPs showed a high degree of specificity for M. avium subsp. paratuberculosis. In recent microarray-based genomic comparisons, Paustian et al. also identified genomic regions divergent between M. avium subsp. paratuberculosis and other members of the M. avium complex (20). The seven regions they describe as potentially unique to M. avium subsp. paratuberculosis represent a subset of our LSPPs. However, our confirmatory testing of these regions across an extended panel revealed that sequences identical or nearly identical to LSPP11, LSPP14, and LSPP16 (which encompass the regions called MAP_RD3, MAP_RD6, and MAP_RD7 by Paustian et al.) can be occasionally found in other closely related bacteria, which were isolated from clinical specimens from various hosts.
LSPP14, a 66-kb sequence rich in genes encoding proteins of the PE/PPE family, includes a previously described 38-kb element (24). It has recently been suggested that one gene from this region, MAP3732, encodes an immunogenic protein of potential utility for the diagnosis of M. avium subsp. paratuberculosis infection (19). By PCR testing of two different gene loci within LSPP14, we were able to detect sequences with homology to M. avium subsp. paratuberculosis K10 in some M. avium subsp. avium isolates, with one isolate being positive for two of the genes tested. Additionally, only 67 of the 107 (63%) M. avium subsp. paratuberculosis isolates tested positive for both targets; 30 isolates amplified with only parts of this sequence, and 10 isolates did not amplify with any part of this sequence. These findings, combined with variable sequence data obtained from M. avium subsp. paratuberculosis isolates, confirm that there is sequence divergence in this region, which may undermine its potential utility in the accurate diagnosis of M. avium subsp. paratuberculosis infection.
LSPP2, LSPP4, and LSPP15 were 100% specific for M. avium subsp. paratuberculosis. The presence of LSPP12, a previously described 19-kb island (8, 23), was also highly predictive of the identification of M. avium subsp. paratuberculosis. Two of the targeted genes for this region, MAP2187c and MAP2182c, were only detected in M. avium subsp. paratuberculosis, while an element similar but not identical to MAP2181c, a putative transcriptional regulator of the TetR family, can be present in a minority of non-M. avium subsp. paratuberculosis isolates. MAP2182c encodes a protein (HspX) possibly involved in mediating cell attachment and has been shown in a previous PCR-based study to be specific for M. avium subsp. paratuberculosis (9). In that study, where PCR tests were done in two different laboratories, hspX was detected in 10 out of 12 (83.3%) M. avium subsp. paratuberculosis isolates, which is comparable to the sensitivity of our assay (92.5%). Of note, three M. avium subsp. paratuberculosis isolates failed to amplify with all three primer pairs used in testing for LSPP12 yet had an LSPA profile suggestive of M. avium subsp. paratuberculosis (absence of LSPA8); this suggests that a small proportion of M. avium subsp. paratuberculosis isolates may be truly devoid of this sequence or that there is sequence level polymorphism in certain isolates.
Of the sequences noted to be highly specific for M. avium subsp. paratuberculosis, LSPP12 and LSPP15 consistently gave positive results across M. avium subsp. paratuberculosis isolates, while amplification of most other LSPPs was variable. The low observed sensitivity of PCR-based testing for these LSPPs may reflect technical limitations of a PCR-based approach. However, in this work we deliberately opted for PCR parameters (such as low annealing temperatures) that maximize sensitivity and thereby achieved successful and consistent amplification of numerous other targets across a wide range of isolates. Moreover, in preliminary microarray-based studies of M. avium subsp. paratuberculosis strains, certain isolates that gave negative amplification results for the LSPP regions studied in this investigation also failed to hybridize to probes designed for these same regions (unpublished observations). Together, these observations argue that certain elements are truly missing from certain M. avium subsp. paratuberculosis strains. In fact, many of the LSPPs we describe have characteristics of mobile genetic elements and include genes annotated as encoding transposases and phage-like integrases. LSPP1 is most likely a prophage. LSPP13 is composed almost entirely of IS elements. The extra sequence represented by LSPP4 is, as is common with mobile elements, inserted at a tRNA gene and flanked by long direct repeats. Testing for such sequences may be specific, but consistent with their mobile nature, they may be variably present in M. avium subsp. paratuberculosis.
LSPP12 was the most widely conserved M. avium subsp. paratuberculosis-specific region. Besides the previously described hspX gene, this sequence also includes a cluster of six homologs of the mammalian cell entry gene (mce) family (MAP2189 to MAP2194), as well as genes predicted to encode potentially useful metabolic functions, more specifically, genes involved in lipid metabolism. The former have previously been identified as important for invasion, survival within macrophages, and possibly virulence (6), although some homologs of mce genes present in the opportunistic pathogen M. avium subsp. avium were noted to be missing from M. avium subsp. paratuberculosis (23). Our analysis indicates that there are nine mce gene clusters in the M. avium subsp. avium strain 104 genome, including four that are homologous to those identified in M. tuberculosis. The genes present in LSPP12 represent an mce cluster unique to M. avium subsp. paratuberculosis. Although the functional importance of this requires further study, work done by others has shown that there are immunodominant epitopes within mce genes (12), suggesting that these could potentially be exploited as a source of antigenic proteins for the diagnosis of M. avium subsp. paratuberculosis. LSPP12 is also of interest because it truncates mbtA, which mediates the first step in mycobactin biosynthesis (Fig. 3B). This truncation is potentially sufficient to prevent siderophore biosynthesis and may account for the dependence on Mycobactin J for the in vitro growth of M. avium subsp. paratuberculosis.
LSPP15, for which amplification was observed in over 90% of M. avium subsp. paratuberculosis isolates, includes genes likely required for the transport of iron. Since iron plays a key role in the balance between pathogen survival and host immunity, and since M. avium subsp. paratuberculosis, unlike other mycobacteria, lacks cell-wall associated mycobactin and siderophores, the proteins encoded by these genes may serve in iron acquisition in the intracellular environment. These findings suggest that LSPP12 and LSPP15 are of great potential not only for diagnostics but possibly also for studies of the pathogenesis of M. avium subsp. paratuberculosis.
Identification of the isolates used in our study to the species level was based partly on the presence of IS900, a multicopy insertion element widely used in the molecular diagnosis of M. avium subsp. paratuberculosis. Therefore, the sensitivity and specificity of our LSPs have not been directly compared to those of this element. However, insertion elements of different bacterial species share sequence similarities, a feature which can confound diagnostic testing relying on PCR amplification of small fragments of these elements. This issue has previously been raised in the molecular diagnostics of M. avium subsp. paratuberculosis relying on IS900 (7, 10) and also for the use of IS6110 as a marker of M. tuberculosis (15). Furthermore, as has been shown to be the case with some members of the M. tuberculosis complex which do not contain IS6110 (1), it is possible some strains of M. avium subsp. paratuberculosis are either low copy number or IS900 negative, providing additional argument for the need to develop molecular tests that rely on genomic characteristics other than mobile elements.
In summary, we have evaluated the distribution of numerous LSPs across a large panel of M. avium complex isolates. Our findings emphasize that in silico identification of a species-specific sequence in a prototype genome does not guarantee its utility as a diagnostic marker. Although most polymorphisms lack the sensitivity and specificity necessary for reliable diagnosis of M. avium subsp. paratuberculosis infection, the proteins they encode may contribute to the virulence of M. avium subsp. paratuberculosis and still prove valuable as markers of infection or disease. In contrast, seven of the LSPPs and LSPA8 revealed themselves to be exquisitely specific for M. avium subsp. paratuberculosis. Further work geared toward the characterization of these and other polymorphisms should facilitate the development of effective diagnostics.
ACKNOWLEDGMENTS
This study was supported by a grant from the Natural Science and Engineering Research Council (GEN2282399). M.S. is funded by the Fonds de la Recherche en Sante du Quebec (FRSQ). M.B. is a New Investigator of the Canadian Institutes of Health Research.
None of us has a conflict of interest or any commercial association that may pose a conflict of interest.
Sequence data for M. avium subsp. paratuberculosis K10 were obtained from the National Center for Biotechnology Information (accession number NC_002944), and sequence data for M. avium subsp. avium 104 were obtained from The Institute for Genomic Research website at http://www.tigr.org. We thank G. Cangelosi and D. M. Collins for supplying isolates.
REFERENCES
Agasino, C. B., A. Ponce de Leon, R. M. Jasmer, and P. M. Small. 1998. Epidemiology of Mycobacterium tuberculosis strains in San Francisco that do not contain IS6110. Int. J. Tuberc. Lung Dis. 2:518-520.
Bannantine, J. P., E. Baechler, Q. Zhang, L. Li, and V. Kapur. 2002. Genome scale comparison of Mycobacterium avium subsp. paratuberculosis with Mycobacterium avium subsp. avium reveals potential diagnostic sequences. J. Clin. Microbiol. 40:1303-1310.
Bannantine, J. P., J. K. Hansen, M. L. Paustian, A. Amonsin, L. L. Li, J. R. Stabel, and V. Kapur. 2004. Expression and immunogenicity of proteins encoded by sequences specific to Mycobacterium avium subsp. paratuberculosis. J. Clin. Microbiol. 42:106-114.
Behr, M. A., M. Semret, A. Poon, and E. Schurr. 2004. Crohn's disease, mycobacteria, and NOD2. Lancet Infect. Dis. 4:136-137.
Bull, T. J., E. J. McMinn, K. Sidi-Boumedine, A. Skull, D. Durkin, P. Neild, G. Rhodes, R. Pickup, and J. Hermon-Taylor. 2003. Detection and verification of Mycobacterium avium subsp. paratuberculosis in fresh ileocolonic mucosal biopsy specimens from individuals with and without Crohn's disease. J. Clin. Microbiol. 41:2915-2923.
Chitale, S., S. Ehrt, I. Kawamura, T. Fujimura, N. Shimono, N. Anand, S. Lu, L. Cohen-Gould, and L. W. Riley. 2001. Recombinant Mycobacterium tuberculosis protein associated with mammalian cell entry. Cell Microbiol. 3:247-254.
Cousins, D. V., R. Whittington, I. Marsh, A. Masters, R. J. Evans, and P. Kluver. 1999. Mycobacteria distinct from Mycobacterium avium subsp. paratuberculosis isolated from the faeces of ruminants possess IS900-like sequences detectable IS900 polymerase chain reaction: implications for diagnosis. Mol. Cell. Probes 13:431-442.
Dohmann, K., B. Strommenger, K. Stevenson, L. de Juan, J. Stratmann, V. Kapur, T. J. Bull, and G. F. Gerlach. 2003. Characterization of genetic differences between Mycobacterium avium subsp. paratuberculosis type I and type II isolates. J. Clin. Microbiol. 41:5215-5223.
Ellingson, J. L., C. A. Bolin, and J. R. Stabel. 1998. Identification of a gene unique to Mycobacterium avium subspecies paratuberculosis and application to diagnosis of paratuberculosis. Mol. Cell. Probes 12:133-142.
Englund, S., G. Bolske, and K. E. Johansson. 2002. An IS900-like sequence found in a Mycobacterium sp. other than Mycobacterium avium subsp. paratuberculosis. FEMS Microbiol. Lett. 209:267-271.
Green, E. P., M. L. Tizard, M. T. Moss, J. Thompson, D. J. Winterbourne, J. J. McFadden, and J. Hermon-Taylor. 1989. Sequence and characteristics of IS900, an insertion element identified in a human Crohn's disease isolate of Mycobacterium paratuberculosis. Nucleic Acids Res. 17:9063-9073.
Harboe, M., A. K. Das, D. Mitra, G. Ulvund, S. Ahmad, R. E. Harkness, D. Das, A. S. Mustafa, and H. G. Wiker. 2004. Immunodominant B-cell epitope in the Mce1A mammalian cell entry protein of Mycobacterium tuberculosis cross-reacting with glutathione S-transferase. Scand. J. Immunol. 59:190-197.
Harris, N. B., and R. G. Barletta. 2001. Mycobacterium avium subsp. paratuberculosis in veterinary medicine. Clin. Microbiol. Rev. 14:489-512.
Kalis, C. H., M. T. Collins, J. W. Hesselink, and H. W. Barkema. 2003. Specificity of two tests for the early diagnosis of bovine paratuberculosis based on cell-mediated immunity: the Johnin skin test and the gamma interferon assay. Vet. Microbiol. 97:73-86.
Kent, L., T. D. McHugh, O. Billington, J. W. Dale, and S. H. Gillespie. 1995. Demonstration of homology between IS6110 of Mycobacterium tuberculosis and DNAs of other Mycobacterium spp. J. Clin. Microbiol. 33:2290-2293.
Manning, E. J. 2001. Mycobacterium avium subspecies paratuberculosis: a review of current knowledge. J. Zoo Wildl. Med. 32:293-304.
Mijs, W., P. de Haas, R. Rossau, T. Van der Laan, L. Rigouts, F. Portaels, and D. Van Soolingen. 2002. Molecular evidence to support a proposal to reserve the designation Mycobacterium avium subsp. avium for bird-type isolates and ‘M. avium subsp. hominissuis’ for the human/porcine type of M. avium. Int. J. Syst. Evol. Microbiol. 52:1505-1518.
Naser, S. A., G. Ghobrial, C. Romero, and J. F. Valentine. 2004. Culture of Mycobacterium avium subspecies paratuberculosis from the blood of patients with Crohn's disease. Lancet 364:1039-1044.
Paustian, M. L., A. Amonsin, V. Kapur, and J. P. Bannantine. 2004. Characterization of novel coding sequences specific to Mycobacterium avium subsp. paratuberculosis: implications for diagnosis of Johne's disease. J. Clin. Microbiol. 42:2675-2681.
Paustian, M. L., V. Kapur, and J. P. Bannantine. 2005. Comparative genomic hybridizations reveal genetic regions within the Mycobacterium avium complex that are divergent from Mycobacterium avium subsp. paratuberculosis isolates. J. Bacteriol. 187:2406-2415.
Poupart, P., M. Coene, H. Van Heuverswyn, and C. Cocito. 1993. Preparation of a specific RNA probe for detection of Mycobacterium paratuberculosis and diagnosis of Johne's disease. J. Clin. Microbiol. 31:1601-1605.
Puyang, X., K. Lee, C. Pawlichuk, and D. Y. Kunimoto. 1999. IS1626, a new IS900-related Mycobacterium avium insertion sequence. Microbiology 145(Pt. 11):3163-3168.
Semret, M., G. Zhai, S. Mostowy, C. Cleto, D. Alexander, G. Cangelosi, D. Cousins, D. M. Collins, D. van Soolingen, and M. A. Behr. 2004. Extensive genomic polymorphism within Mycobacterium avium. J. Bacteriol. 186:6332-6334.
Stratmann, J., B. Strommenger, R. Goethe, K. Dohmann, G. F. Gerlach, K. Stevenson, L. L. Li, Q. Zhang, V. Kapur, and T. J. Bull. 2004. A 38-kilobase pathogenicity island specific for Mycobacterium avium subsp. paratuberculosis encodes cell surface proteins expressed in the host. Infect. Immun. 72:1265-1274.
Talbot, E. A., D. L. Williams, and R. Frothingham. 1997. PCR identification of Mycobacterium bovis BCG. J. Clin. Microbiol. 35:566-569.
Tortoli, E., L. Rindi, M. J. Garcia, P. Chiaradonna, R. Dei, C. Garzelli, R. M. Kroppenstedt, N. Lari, R. Mattei, A. Mariottini, G. Mazzarelli, M. I. Murcia, A. Nanetti, P. Piccoli, and C. Scarparo. 2004. Proposal to elevate the genetic variant MAC-A, included in the Mycobacterium avium complex, to species rank as Mycobacterium chimaera sp. nov. Int. J. Syst. Evol. Microbiol. 54:1277-1285.
van Soolingen, D., P. W. Hermans, P. E. de Haas, D. R. Soll, and J. D. Van Embden. 1991. Occurrence and stability of insertion sequences in Mycobacterium tuberculosis complex strains: evaluation of an insertion sequence-dependent DNA polymorphism as a tool in the epidemiology of tuberculosis. J. Clin. Microbiol. 29:2578-2586.(Makeda Semret, David C. A)
National Institute of Public Health and the Environment, 3720BA Bilthoven, The Netherlands
Australian Reference Laboratory for Bovine Tuberculosis, Department of Agriculture, South Perth, Western Australia, Australia 6151
ABSTRACT
Mycobacterium avium subsp. paratuberculosis is an emerging pathogen of mammals and is being actively investigated as a possible zoonotic agent. The lack of reliable diagnostic assays has hampered rational assessment of the prevalence of this organism in humans and animals. We have used a comparative genomic approach to reveal genomic differences between M. avium subsp. paratuberculosis and its close relative M. avium subsp. avium, a highly prevalent environmental organism. From computational and DNA microarray-based study of two prototype strains, M. avium subsp. avium strain 104 and M. avium subsp. paratuberculosis strain K10, we have uncovered two types of large sequence polymorphisms (LSPs): those present in the former but missing in the latter (LSPAs) and those only present in the latter (LSPPs). We examined the distribution of 3 LSPAs and 17 LSPPs across a panel of 383 M. avium complex isolates in order to determine their potential utility for the development of accurate diagnostic tests. Our results show that the absence of LSPA8 is 100% specific for the identification of M. avium subsp. paratuberculosis. Of the 17 LSPPs, 10 regions were not specific for M. avium subsp. paratuberculosis while 7 were shown to be highly specific (>98%) and, in some cases, highly sensitive as well (up to 95%). These data highlight the need to evaluate these regions across a diverse panel of clinical and environmental isolates and indicate the LSPs best suited for M. avium subsp. paratuberculosis diagnostics.
INTRODUCTION
Mycobacterium avium subsp. paratuberculosis is a serious pathogen of domestic ruminants, in which it causes paratuberculosis (Johne's disease), a chronic and eventually fatal inflammatory bowel disease. Its rising incidence in many regions of the world has resulted in significant economic losses to the livestock industry (13). This organism also infects free-ranging wildlife species (16). The impact of M. avium subsp. paratuberculosis in nonruminant wildlife species is largely unknown; however, the potential for interspecies transmission has important implications for paratuberculosis control programs. M. avium subsp. paratuberculosis is also being actively investigated as the possible cause of a debilitating human inflammatory bowel disease, Crohn's disease (4, 5, 18).
Effective control of Johne's disease and investigation of the potential link to Crohn's disease have been hampered by the lack of effective assays to easily and accurately diagnose M. avium subsp. paratuberculosis infections. Commercially available serological assays for bovine disease are convenient but offer poor sensitivity, especially during subclinical disease. Moreover, owing to the high degree of genetic similarity between M. avium subsp. paratuberculosis and the other members of the M. avium complex, many of the proteins that are recognized during the early stages of infection are not specific for M. avium subsp. paratuberculosis (14). Since M. avium subsp. avium is highly prevalent in the environment but not associated with a similar clinical syndrome in mammals, the capacity to differentiate between these closely related organisms is essential for rational clinical and epidemiologic assessment.
Previous work has identified a number of M. avium subsp. paratuberculosis genetic sequences that are absent from other mycobacteria, such as the insertion sequence IS900 (11), the F57 element (21), and the hspX gene (9), but the value of these sequences for diagnostics of M. avium subsp. paratuberculosis remains unclear. Although PCR testing for the multicopy insertion element IS900 is widely used, there are concerns about its specificity and its validity as a direct proxy for M. avium subsp. paratuberculosis. IS900-like elements have been found in unrelated organisms (10), and other M. avium complex insertion elements, including IS1311 (7) and IS1626 (22), share considerable sequence similarity. In the cases of F57 and hspX, their specificity for M. avium subsp. paratuberculosis has not to date been rigorously evaluated in a large sample of clinical isolates.
More recently, data derived from genome sequencing projects have suggested a number of polymorphic genomic regions that may serve in the specific diagnosis of M. avium subsp. paratuberculosis. Our microarray-based comparisons using M. avium subsp. avium strain 104 as the reference revealed 14 large sequence polymorphisms (LSPs) that are variably present among a small collection of M. avium isolates. Three of these regions appeared to be present in M. avium subsp. avium (LSPAs) and absent from M. avium subsp. paratuberculosis isolates (23). Conversely, computational comparisons of the genome sequences of M. avium subsp. paratuberculosis strain K10 (GenBank accession no. NC_002944) and M. avium subsp. avium strain 104 (http://www.tigr.org) have identified DNA sequences that are present in the former but missing or divergent in the latter (2, 3, 19). We have to date identified 17 regions varying in length from 2.9 to 66 kb that are unique to M. avium subsp. paratuberculosis strain K10; we call these LSPPs. Such sequences can be expected to lend themselves to nucleic acid-based diagnostic tests and, if they encode immunogenic proteins, to immunological assays as well. One of these, a 19-kb sequence that we call LSPP12, has been previously documented (8, 23). Another, a 98-kb segment of the genome that encompasses elements we call LSPP14 and LSPP15, also includes within it a 38-kb element that was recently described as an M. avium subsp. paratuberculosis-specific putative pathogenicity island (24).
The objective of our study was to assess the value of these LSPs for the specific diagnosis of M. avium subsp. paratuberculosis. To be diagnostically specific for M. avium subsp. paratuberculosis, an LSPP must be found only in M. avium subsp. paratuberculosis isolates and should be absent from non-M. avium subsp. paratuberculosis isolates. Conversely, for an LSPA to serve in the molecular diagnosis of M. avium subsp. paratuberculosis, it should be consistently present in non-M. avium subsp. paratuberculosis isolates but missing from a broad collection of M. avium subsp. paratuberculosis isolates. We therefore tested the distribution of the 17 LSPPs and 3 LSPAs across a panel of 383 M. avium complex isolates. Our results indicate that many LSP regions, although distinct between prototype genome sequences, are heterogeneously distributed across geographically diverse isolates and so lack the specificity required for diagnostics. However, a subset of LSPs do appear highly specific for M. avium subsp. paratuberculosis and should prove useful in the development of effective diagnostics for Johne's disease and evaluation of the Crohn's disease hypothesis.
MATERIALS AND METHODS
Bacterial isolates. A panel of 107 M. avium subsp. paratuberculosis, 260 M. avium subsp. avium, 4 M. avium subsp. silvaticum, and 12 M. intracellulare isolates was assembled (Table 1). Isolates were first identified as M. avium complex species by the laboratories providing them using the commercial nucleic acid probe AccuProbe (Gen-Probe Inc., San Diego, CA). Initial dependence on Mycobactin J and the presence of the insertion sequence IS900 were determined by these laboratories and used to identify M. avium strains to the subspecies level as M. avium subsp. paratuberculosis. Designation as M. avium subsp. avium was supported by phenotypic characteristics (including lack of Mycobactin J dependency) and the presence of IS1245. Additional extensive testing, including sequencing of the 16S-23S internal transcribed spacer (ITS) region, was completed for 160 of these isolates in the context of another study (17). M. avium subsp. silvaticum strains exhibited dependence on Mycobactin J on primary isolation and were IS900 negative. M. intracellulare strains were identified using the M. intracellulare AccuProbe, and species designation was confirmed by 16S rRNA gene sequencing.
Genomic DNA was extracted from each isolate using previously published methods (27). The amount of DNA in each sample was quantified using a spectrophotometer, and each sample was normalized to a concentration of 1 ng/μl prior to PCR testing.
Discovery of LSPP and testing across M. avium complex isolates. As part of our ongoing work to characterize M. avium complex organisms, we compared the sequenced genomes of M. avium subsp. avium strain 104 (http://www.tigr.org) and M. avium subsp. paratuberculosis strain K10 (GenBank accession no. NC_002944). Artemis software (http://www.sanger.ac.uk/Software/Artemis/) was used for visualization of the M. avium subsp. paratuberculosis strain K10 genome data and annotation of M. avium subsp. avium strain 104. Genome sequence comparison, via visual inspection in Artemis and with BLASTN, permitted identification of regions present in M. avium subsp. paratuberculosis strain K10 but lacking a homologous region in M. avium subsp. avium strain 104 (LSPP). Additional homology analyses (using BLASTN) were used to confirm that LSPP regions were truly missing and not simply displaced (e.g., due to genomic rearrangement). BLAST2 was used to precisely define the boundaries of these novel LSPPs. For the 17 regions and their locations in the M. avium subsp. paratuberculosis strain K10 genome, see Table 3.
To screen for the specificity of LSPPs, we initially tested for their presence by PCR targeting a short internal sequence across a panel of 96 M. avium subsp. avium isolates. LSPPs for which a screening PCR was negative for these M. avium subsp. avium strains were then tested across the extended panel of 383 M. avium complex isolates. Since LSPP14, which includes but is not restricted to a known 38-kb island (24), and LSPP12, a 19-kb sequence (8, 23), were previously postulated to be of potential diagnostic utility, we tested each of these LSPs with two and three different sets of primers targeting distinct loci, respectively. Primer sequences for each of the LSPPs tested are listed in Table 2.
Testing of LSPA across M. avium complex isolates. Our previous DNA microarray-based analysis revealed numerous large sequences variable among M. avium isolates. LSPA8, LSPA13, and LSPA14 appeared potentially specific for M. avium subsp. paratuberculosis (23). When aligned against the genome of M. avium subsp. paratuberculosis strain K10, these M. avium subsp. avium strain 104 sequences were confirmed to be absent from the M. avium subsp. paratuberculosis K10 genome. Testing for these regions employed PCR with primers designed toward the flanking regions of the sequence, such that an amplicon would be generated only if the region was indeed missing. The PCR test was considered positive (region missing) if amplification generated a product of the expected molecular weight. Furthermore, for each set of primers, a random selection of 25 PCR products, originating from 25 different strains, was sequenced using ABI dye terminator chemistry. The resulting sequences were aligned with the sequenced genomes of M. avium subsp. paratuberculosis strain K10 and of M. avium subsp. avium strain 104 to determine if the junction of the deleted regions was identical across the different isolates.
For isolates in which amplification did not occur using the flanking primers, additional PCR testing was performed (using sets of primers targeting sequences within the LSPA) to determine whether the intervening sequence was present. An example is LSPA8 in Fig. 1, with corresponding positions on the genomes of M. avium subsp. avium strain 104, M. avium subsp. paratuberculosis strain K10, and M. tuberculosis strain H37Rv. The primer sequences used for all reactions are given in Table 2.
PCRs. All PCRs were performed in 50-μl volumes, using 5 μl of template DNA (5 ng) and 1 U Taq polymerase (MBI Fermentas), with 5 μl of 10x PCR buffer (MBI Fermentas), 2.5 mM MgCl2, 5 μl acetamide 50% (wt/vol), 0.2 mM deoxynucleoside triphosphates, and each primer at 0.5 μM. PCRs were conducted in 96-well plates. Amplification consisted of an initial denaturation step of 94°C for 5 min, followed by 30 cycles of denaturation at 94°C for 30s, annealing at 58°C for 45s, and extension at 72°C for 2 min. PCR products were separated by electrophoresis in precast 1% (wt/vol) agarose gels containing ethidium bromide (ReadyAgarose gels; Bio-Rad Laboratories, Hercules, CA). Samples that had bands of the expected molecular weight were considered positive, whereas those with bands of different size or where no bands were seen were considered negative. When results were ambiguous, PCRs were repeated using individual closed tubes.
M. avium subsp. paratuberculosis strain K10 and M. avium subsp. avium strain 104 were used as positive and negative controls for each set of primers used to test for the LSPs.
RESULTS
LSPPs: regions present in M. avium subsp. paratuberculosis K10 and absent from M. avium subsp. avium 104. LSPPs represent four types of polymorphisms with respect to the sequenced genome of M. avium subsp. avium 104. For certain regions, the M. avium subsp. paratuberculosis K10 genome simply contains an additional element when aligned against the corresponding positions on the genomes of M. avium subsp. avium strain 104 and M. tuberculosis strain H37Rv; the element is flanked by regions of high homology and synteny among the three genomes (schematically depicted by the example of LSPP4 in Fig. 2A). For other regions, the M. avium subsp. paratuberculosis K10 genome does not contain additional DNA, but rather, there is a region of extremely low sequence similarity (<20%) with the best hit against the M. avium subsp. avium 104 genome, flanked by regions of high homology (>97%) between the two sequences (depicted as the example of LSPP2 in Fig. 2B). Some regions are characterized by both additional DNA and sequence with low similarity (examples of LSPP14 and LSPP15 in Fig. 3A). Finally, one region is the result of additional DNA in the M. avium subsp. paratuberculosis K10 genome and loss of DNA from the M. avium subsp. avium genome and thus is characterized by the replacement of a DNA segment with another (example of LSPP12, Fig. 3B).
For the 17 LSPPs, a positive PCR using primers internal to these sequences indicated the presence of at least part of the region tested. From the screening panel of 96 M. avium subsp. avium isolates, 10 of these sequences were variably present in M. avium subsp. avium isolates while 7 could not be amplified from the screening panel (LSPP2, LSPP4, LSPP11, LSPP12, LSPP14, LSPP15, and LSPP16; Table 3). These regions were thus investigated using an extended panel of 383 M. avium complex isolates; results are summarized in Table 4, with more detailed descriptions below.
LSPP2, LSPP4, and LSPP15 could not be amplified from any of the 276 non-M. avium subsp. paratuberculosis isolates, demonstrating 100% specificity for this subspecies.
LSPP11 and LSPP16 also exhibited high specificity but unexpectedly amplified from a group of six non-M. avium subsp. paratuberculosis strains: one isolate gave amplification products for both LSPP11 and LSPP16; four isolates gave amplification products for LSPP11 only, and one gave a product for LSPP16 only. The five LSPP11 and two LSPP6 amplicons obtained from these isolates were identical to M. avium subsp. paratuberculosis strain K10 sequences. To further characterize these six isolates, all clinical human isolates collected in Montreal, we performed additional analysis, including 16S rRNA sequencing (to confirm species designation) and 16S-23S ITS sequencing to identify their sequevar. By 16S analysis, one isolate (13373) had a novel 16S rRNA sequence, similar to M. chimaera (previously known as MAC-A) (26) and a unique ITS sequence, closest to the MAC-C sequevar. Four isolates had a 16S sequence consistent with M. avium and had ITS sequences identical to the common Mav-B sequevar. Finally, we determined that one isolate (9033) was potentially mixed, based on the presence of double peaks in the DNA sequence chromatograms, with features of both M. avium subsp. avium and M. intracellulare (Min-A sequevar).
The presence of LSPP12, a 19-kb sequence that is found in lieu of a 197-kb sequence in M. avium subsp. avium 104, was evaluated with three distinct sets of internal primers targeting different genes. Across the entire panel of M. avium complex isolates, amplification of two target genes (MAP2182c and MAP2187c) was restricted to M. avium subsp. paratuberculosis strains, resulting in a calculated specificity of 100%. However, primers targeting MAP2181c (predicted to encode a transcriptional regulator of the TetR family) unexpectedly produced an amplicon from two non-M. avium subsp. paratuberculosis isolates, clinical specimens 13373 and 9033 noted above. Sequencing of the MAP2181c amplicons revealed numerous sequence level polymorphisms shared across these isolates and only 89% sequence identity to M. avium subsp. paratuberculosis strain K10.
LSPP14 is a heterogeneous region (Fig. 3A). Some portions are extra to M. avium subsp. paratuberculosis K10, including a previously described 38-kb island (24), while other segments exhibit moderate sequence similarity to M. avium subsp. avium 104. LSPP14 was tested using two different sets of primers targeting MAP3726 and MAP3750 within the additional segment of DNA. Seven non-M. avium subsp. paratuberculosis isolates tested positive for at least one of the two genes. One M. avium subsp. avium isolate was positive for both MAP3726 and MAP3750, while six other isolates were positive for one of the two genes tested. In the case of MAP3726, a sequence with 97% similarity to that of M. avium subsp. paratuberculosis strain K10 was amplified from three M. avium subsp. avium strains. One of these isolates was from a pig in The Netherlands and had an ITS sequence identifying it as a Mav-B sequevar, while the other two were isolated from birds and had Mav-A sequevars. For MAP3750, sequences with less than 93% similarity to that of M. avium subsp. paratuberculosis strain K10 were amplified from five M. avium subsp. avium isolates, including the pig isolate described above. The presence of polymorphic sequence within M. avium subsp. avium isolates, and the fact that only a subset of M. avium subsp. paratuberculosis isolates were positive for these two genes, prompted us to look for polymorphisms within M. avium subsp. paratuberculosis isolates. Amplicons obtained from eight different isolates for each of these two genes also revealed several single-nucleotide polymorphisms (SNPs) compared to the genome of M. avium subsp. paratuberculosis K10. Specifically, of the eight isolates tested, one (obtained from a deer) had a SNP noted within MAP3726 (at position 400 starting from the first base pair of the gene) and an additional two SNPs within MAP3750 (positions 331 and 396). Three other isolates (two from cows, one from a sheep) had a shared SNP within MAP3750 (at position 384). These findings suggest that LSPP14 is polymorphic within M. avium subsp. paratuberculosis.
Of the seven LSPPs tested extensively, only LSPP12 and LSPP15 could be amplified from the majority of M. avium subsp. paratuberculosis isolates, suggesting either that the other LSPPs were not universally present in this subspecies or that sequence level polymorphisms prevented their amplification across some isolates. Despite our efforts to identify a pattern, it was not possible to correlate the distribution of these variable LSPP regions with respect to the host source of the isolate tested.
LSPAs: regions present in M. avium subsp. avium 104 and absent from M. avium subsp. paratuberculosis K10. A positive PCR with primers flanking the LSPAs indicated the absence of the region. PCR with flanking primers demonstrated the absence of LSPA8 in 94 M. avium subsp. paratuberculosis isolates. For the 289 isolates (including 13 of M. avium subsp. paratuberculosis) which failed to amplify using the flanking primers, we performed internal PCR testing using two different sets of primers targeting sequences within LSPA8 and found the sequence was present in all 276 non-M. avium subsp. paratuberculosis isolates. The 13 M. avium subsp. paratuberculosis isolates did not generate PCR products with either set of primers internal to this region. Therefore, it appears that this sequence is consistently missing from M. avium subsp. paratuberculosis isolates and is present in every other M. avium complex strain tested. Sequencing of 25 of the amplicons generated from M. avium subsp. paratuberculosis strains using the flanking primers revealed the same polymorphism in all M. avium subsp. paratuberculosis strains, with the junction point truncating two genes (called Maa4183 [the ortholog of MAP3636] and Maa4172 [not annotated in M. avium subsp. paratuberculosis]), homologs of Rv0197 and Rv0201 in M. tuberculosis, respectively. These results strongly suggest that absence of LSPA8 is the result of a genomic deletion from the ancestral genome.
LSPA13, a 19.5-kb sequence, was confirmed as missing in 167 isolates by PCR using primers flanking this sequence. Of these, 102 were M. avium subsp. paratuberculosis isolates, but also 64 of M. avium subsp. avium and 1 of M. avium subsp. silvaticum were shown to lack this region. Sequencing of the amplicons generated using the flanking primers in 25 M. avium subsp. paratuberculosis and 25 M. avium subsp. avium isolates demonstrated the exact same sequence, which is also identical to that of the M. avium subsp. paratuberculosis K10 type strain. LSPA13 is located at the tRNA-argV gene and is flanked by direct repeats, features characteristic of mobile genetic elements. Whatever its origin, the absence of this region was not specific to M. avium subsp. paratuberculosis.
LSPA14, a 23-kb region, was initially judged to be missing in 101 isolates, all of which were M. avium subsp. paratuberculosis, by PCR using flanking primers (set A). Upon sequencing, the amplicons obtained from 25 of these M. avium subsp. paratuberculosis isolates were identical to that of the prototype, M. avium subsp. paratuberculosis K10. When testing for the presence of LSPA14, an intervening sequence was shown to be present in 215 isolates (all non-M. avium subsp. paratuberculosis). However, 67 isolates did not give products with either the flanking or the internal primers, leading us to hypothesize that the sequence may be missing (as in M. avium subsp. paratuberculosis) but associated with polymorphisms in the flanking regions. To test the latter possibility, we designed a new set of primers (set B) to amplify across LSPA14, such that the resulting PCR products would encompass the sequences of the original flanking primers (set A). Using flanking primer set B, we determined that the region was indeed missing in these isolates. Upon sequencing the resulting PCR products, we noted that the LSPA14 junction point in these isolates was identical to that in M. avium subsp. paratuberculosis. However, the sequence of the flanking region in these isolates was only 97% identical to that of M. avium subsp. paratuberculosis. Further, two of the SNPs noted to be common to these non-M. avium subsp. paratuberculosis isolates occurred along the recognition site for one of the original flanking primers (set A). These results indicate that a positive PCR using flanking primer set A appears to be specific for M. avium subsp. paratuberculosis; however, absence of this region is not specific to M. avium subsp. paratuberculosis.
DISCUSSION
Members of the M. avium complex form a closely related group of bacteria, with M. avium subsp. avium and M. avium subsp. paratuberculosis sharing the same species designation. Despite their taxonomic relationship, M. avium subsp. avium and M. avium subsp. paratuberculosis exhibit markedly different capacities to cause disease across a number of hosts. In particular, M. avium subsp. paratuberculosis is emerging as a serious threat to many mammalian species, but control programs to detect and eradicate this pathogen have been hampered by the lack of sensitive, specific, and practical assays. The availability of sequence information for two prototype strains of the M. avium complex offers the opportunity to uncover M. avium subsp. paratuberculosis-specific sequences that may be employed for nucleic acid-based tests and development of immunological assays.
Using a comparative genomic approach, we have identified two types of markers: large sequences present in M. avium subsp. paratuberculosis K10 but not in M. avium subsp. avium 104 (LSPPs) and large sequences present in M. avium subsp. avium strain 104 but not in M. avium subsp. paratuberculosis (LSPAs). We evaluated the distribution of these markers across a large panel of M. avium complex isolates and determined that, of the latter, PCR-based testing for the loss of LSPA8 was 100% specific for the identification of M. avium subsp. paratuberculosis. Akin to the RD1 region of the M. tuberculosis complex, which permits accurate differentiation between virulent M. bovis and BCG strains (25), our results for LSPA8 have immediate applicability in the diagnostics of M. avium subsp. paratuberculosis. First, the presence of this region (by PCR testing of a sequence internal to LSPA8) can be used to determine that an isolate is highly unlikely to be M. avium subsp. paratuberculosis. Second, the absence of LSPA8 (by PCR testing using primers flanking these sequences) can be used to diagnose M. avium subsp. paratuberculosis with 100% specificity.
Of the 17 LSPPs, 10 had poor specificity for M. avium subsp. paratuberculosis, based on the amplification of genomic targets from these regions in non-M. avium subsp. paratuberculosis isolates, while 7 of the LSPPs showed a high degree of specificity for M. avium subsp. paratuberculosis. In recent microarray-based genomic comparisons, Paustian et al. also identified genomic regions divergent between M. avium subsp. paratuberculosis and other members of the M. avium complex (20). The seven regions they describe as potentially unique to M. avium subsp. paratuberculosis represent a subset of our LSPPs. However, our confirmatory testing of these regions across an extended panel revealed that sequences identical or nearly identical to LSPP11, LSPP14, and LSPP16 (which encompass the regions called MAP_RD3, MAP_RD6, and MAP_RD7 by Paustian et al.) can be occasionally found in other closely related bacteria, which were isolated from clinical specimens from various hosts.
LSPP14, a 66-kb sequence rich in genes encoding proteins of the PE/PPE family, includes a previously described 38-kb element (24). It has recently been suggested that one gene from this region, MAP3732, encodes an immunogenic protein of potential utility for the diagnosis of M. avium subsp. paratuberculosis infection (19). By PCR testing of two different gene loci within LSPP14, we were able to detect sequences with homology to M. avium subsp. paratuberculosis K10 in some M. avium subsp. avium isolates, with one isolate being positive for two of the genes tested. Additionally, only 67 of the 107 (63%) M. avium subsp. paratuberculosis isolates tested positive for both targets; 30 isolates amplified with only parts of this sequence, and 10 isolates did not amplify with any part of this sequence. These findings, combined with variable sequence data obtained from M. avium subsp. paratuberculosis isolates, confirm that there is sequence divergence in this region, which may undermine its potential utility in the accurate diagnosis of M. avium subsp. paratuberculosis infection.
LSPP2, LSPP4, and LSPP15 were 100% specific for M. avium subsp. paratuberculosis. The presence of LSPP12, a previously described 19-kb island (8, 23), was also highly predictive of the identification of M. avium subsp. paratuberculosis. Two of the targeted genes for this region, MAP2187c and MAP2182c, were only detected in M. avium subsp. paratuberculosis, while an element similar but not identical to MAP2181c, a putative transcriptional regulator of the TetR family, can be present in a minority of non-M. avium subsp. paratuberculosis isolates. MAP2182c encodes a protein (HspX) possibly involved in mediating cell attachment and has been shown in a previous PCR-based study to be specific for M. avium subsp. paratuberculosis (9). In that study, where PCR tests were done in two different laboratories, hspX was detected in 10 out of 12 (83.3%) M. avium subsp. paratuberculosis isolates, which is comparable to the sensitivity of our assay (92.5%). Of note, three M. avium subsp. paratuberculosis isolates failed to amplify with all three primer pairs used in testing for LSPP12 yet had an LSPA profile suggestive of M. avium subsp. paratuberculosis (absence of LSPA8); this suggests that a small proportion of M. avium subsp. paratuberculosis isolates may be truly devoid of this sequence or that there is sequence level polymorphism in certain isolates.
Of the sequences noted to be highly specific for M. avium subsp. paratuberculosis, LSPP12 and LSPP15 consistently gave positive results across M. avium subsp. paratuberculosis isolates, while amplification of most other LSPPs was variable. The low observed sensitivity of PCR-based testing for these LSPPs may reflect technical limitations of a PCR-based approach. However, in this work we deliberately opted for PCR parameters (such as low annealing temperatures) that maximize sensitivity and thereby achieved successful and consistent amplification of numerous other targets across a wide range of isolates. Moreover, in preliminary microarray-based studies of M. avium subsp. paratuberculosis strains, certain isolates that gave negative amplification results for the LSPP regions studied in this investigation also failed to hybridize to probes designed for these same regions (unpublished observations). Together, these observations argue that certain elements are truly missing from certain M. avium subsp. paratuberculosis strains. In fact, many of the LSPPs we describe have characteristics of mobile genetic elements and include genes annotated as encoding transposases and phage-like integrases. LSPP1 is most likely a prophage. LSPP13 is composed almost entirely of IS elements. The extra sequence represented by LSPP4 is, as is common with mobile elements, inserted at a tRNA gene and flanked by long direct repeats. Testing for such sequences may be specific, but consistent with their mobile nature, they may be variably present in M. avium subsp. paratuberculosis.
LSPP12 was the most widely conserved M. avium subsp. paratuberculosis-specific region. Besides the previously described hspX gene, this sequence also includes a cluster of six homologs of the mammalian cell entry gene (mce) family (MAP2189 to MAP2194), as well as genes predicted to encode potentially useful metabolic functions, more specifically, genes involved in lipid metabolism. The former have previously been identified as important for invasion, survival within macrophages, and possibly virulence (6), although some homologs of mce genes present in the opportunistic pathogen M. avium subsp. avium were noted to be missing from M. avium subsp. paratuberculosis (23). Our analysis indicates that there are nine mce gene clusters in the M. avium subsp. avium strain 104 genome, including four that are homologous to those identified in M. tuberculosis. The genes present in LSPP12 represent an mce cluster unique to M. avium subsp. paratuberculosis. Although the functional importance of this requires further study, work done by others has shown that there are immunodominant epitopes within mce genes (12), suggesting that these could potentially be exploited as a source of antigenic proteins for the diagnosis of M. avium subsp. paratuberculosis. LSPP12 is also of interest because it truncates mbtA, which mediates the first step in mycobactin biosynthesis (Fig. 3B). This truncation is potentially sufficient to prevent siderophore biosynthesis and may account for the dependence on Mycobactin J for the in vitro growth of M. avium subsp. paratuberculosis.
LSPP15, for which amplification was observed in over 90% of M. avium subsp. paratuberculosis isolates, includes genes likely required for the transport of iron. Since iron plays a key role in the balance between pathogen survival and host immunity, and since M. avium subsp. paratuberculosis, unlike other mycobacteria, lacks cell-wall associated mycobactin and siderophores, the proteins encoded by these genes may serve in iron acquisition in the intracellular environment. These findings suggest that LSPP12 and LSPP15 are of great potential not only for diagnostics but possibly also for studies of the pathogenesis of M. avium subsp. paratuberculosis.
Identification of the isolates used in our study to the species level was based partly on the presence of IS900, a multicopy insertion element widely used in the molecular diagnosis of M. avium subsp. paratuberculosis. Therefore, the sensitivity and specificity of our LSPs have not been directly compared to those of this element. However, insertion elements of different bacterial species share sequence similarities, a feature which can confound diagnostic testing relying on PCR amplification of small fragments of these elements. This issue has previously been raised in the molecular diagnostics of M. avium subsp. paratuberculosis relying on IS900 (7, 10) and also for the use of IS6110 as a marker of M. tuberculosis (15). Furthermore, as has been shown to be the case with some members of the M. tuberculosis complex which do not contain IS6110 (1), it is possible some strains of M. avium subsp. paratuberculosis are either low copy number or IS900 negative, providing additional argument for the need to develop molecular tests that rely on genomic characteristics other than mobile elements.
In summary, we have evaluated the distribution of numerous LSPs across a large panel of M. avium complex isolates. Our findings emphasize that in silico identification of a species-specific sequence in a prototype genome does not guarantee its utility as a diagnostic marker. Although most polymorphisms lack the sensitivity and specificity necessary for reliable diagnosis of M. avium subsp. paratuberculosis infection, the proteins they encode may contribute to the virulence of M. avium subsp. paratuberculosis and still prove valuable as markers of infection or disease. In contrast, seven of the LSPPs and LSPA8 revealed themselves to be exquisitely specific for M. avium subsp. paratuberculosis. Further work geared toward the characterization of these and other polymorphisms should facilitate the development of effective diagnostics.
ACKNOWLEDGMENTS
This study was supported by a grant from the Natural Science and Engineering Research Council (GEN2282399). M.S. is funded by the Fonds de la Recherche en Sante du Quebec (FRSQ). M.B. is a New Investigator of the Canadian Institutes of Health Research.
None of us has a conflict of interest or any commercial association that may pose a conflict of interest.
Sequence data for M. avium subsp. paratuberculosis K10 were obtained from the National Center for Biotechnology Information (accession number NC_002944), and sequence data for M. avium subsp. avium 104 were obtained from The Institute for Genomic Research website at http://www.tigr.org. We thank G. Cangelosi and D. M. Collins for supplying isolates.
REFERENCES
Agasino, C. B., A. Ponce de Leon, R. M. Jasmer, and P. M. Small. 1998. Epidemiology of Mycobacterium tuberculosis strains in San Francisco that do not contain IS6110. Int. J. Tuberc. Lung Dis. 2:518-520.
Bannantine, J. P., E. Baechler, Q. Zhang, L. Li, and V. Kapur. 2002. Genome scale comparison of Mycobacterium avium subsp. paratuberculosis with Mycobacterium avium subsp. avium reveals potential diagnostic sequences. J. Clin. Microbiol. 40:1303-1310.
Bannantine, J. P., J. K. Hansen, M. L. Paustian, A. Amonsin, L. L. Li, J. R. Stabel, and V. Kapur. 2004. Expression and immunogenicity of proteins encoded by sequences specific to Mycobacterium avium subsp. paratuberculosis. J. Clin. Microbiol. 42:106-114.
Behr, M. A., M. Semret, A. Poon, and E. Schurr. 2004. Crohn's disease, mycobacteria, and NOD2. Lancet Infect. Dis. 4:136-137.
Bull, T. J., E. J. McMinn, K. Sidi-Boumedine, A. Skull, D. Durkin, P. Neild, G. Rhodes, R. Pickup, and J. Hermon-Taylor. 2003. Detection and verification of Mycobacterium avium subsp. paratuberculosis in fresh ileocolonic mucosal biopsy specimens from individuals with and without Crohn's disease. J. Clin. Microbiol. 41:2915-2923.
Chitale, S., S. Ehrt, I. Kawamura, T. Fujimura, N. Shimono, N. Anand, S. Lu, L. Cohen-Gould, and L. W. Riley. 2001. Recombinant Mycobacterium tuberculosis protein associated with mammalian cell entry. Cell Microbiol. 3:247-254.
Cousins, D. V., R. Whittington, I. Marsh, A. Masters, R. J. Evans, and P. Kluver. 1999. Mycobacteria distinct from Mycobacterium avium subsp. paratuberculosis isolated from the faeces of ruminants possess IS900-like sequences detectable IS900 polymerase chain reaction: implications for diagnosis. Mol. Cell. Probes 13:431-442.
Dohmann, K., B. Strommenger, K. Stevenson, L. de Juan, J. Stratmann, V. Kapur, T. J. Bull, and G. F. Gerlach. 2003. Characterization of genetic differences between Mycobacterium avium subsp. paratuberculosis type I and type II isolates. J. Clin. Microbiol. 41:5215-5223.
Ellingson, J. L., C. A. Bolin, and J. R. Stabel. 1998. Identification of a gene unique to Mycobacterium avium subspecies paratuberculosis and application to diagnosis of paratuberculosis. Mol. Cell. Probes 12:133-142.
Englund, S., G. Bolske, and K. E. Johansson. 2002. An IS900-like sequence found in a Mycobacterium sp. other than Mycobacterium avium subsp. paratuberculosis. FEMS Microbiol. Lett. 209:267-271.
Green, E. P., M. L. Tizard, M. T. Moss, J. Thompson, D. J. Winterbourne, J. J. McFadden, and J. Hermon-Taylor. 1989. Sequence and characteristics of IS900, an insertion element identified in a human Crohn's disease isolate of Mycobacterium paratuberculosis. Nucleic Acids Res. 17:9063-9073.
Harboe, M., A. K. Das, D. Mitra, G. Ulvund, S. Ahmad, R. E. Harkness, D. Das, A. S. Mustafa, and H. G. Wiker. 2004. Immunodominant B-cell epitope in the Mce1A mammalian cell entry protein of Mycobacterium tuberculosis cross-reacting with glutathione S-transferase. Scand. J. Immunol. 59:190-197.
Harris, N. B., and R. G. Barletta. 2001. Mycobacterium avium subsp. paratuberculosis in veterinary medicine. Clin. Microbiol. Rev. 14:489-512.
Kalis, C. H., M. T. Collins, J. W. Hesselink, and H. W. Barkema. 2003. Specificity of two tests for the early diagnosis of bovine paratuberculosis based on cell-mediated immunity: the Johnin skin test and the gamma interferon assay. Vet. Microbiol. 97:73-86.
Kent, L., T. D. McHugh, O. Billington, J. W. Dale, and S. H. Gillespie. 1995. Demonstration of homology between IS6110 of Mycobacterium tuberculosis and DNAs of other Mycobacterium spp. J. Clin. Microbiol. 33:2290-2293.
Manning, E. J. 2001. Mycobacterium avium subspecies paratuberculosis: a review of current knowledge. J. Zoo Wildl. Med. 32:293-304.
Mijs, W., P. de Haas, R. Rossau, T. Van der Laan, L. Rigouts, F. Portaels, and D. Van Soolingen. 2002. Molecular evidence to support a proposal to reserve the designation Mycobacterium avium subsp. avium for bird-type isolates and ‘M. avium subsp. hominissuis’ for the human/porcine type of M. avium. Int. J. Syst. Evol. Microbiol. 52:1505-1518.
Naser, S. A., G. Ghobrial, C. Romero, and J. F. Valentine. 2004. Culture of Mycobacterium avium subspecies paratuberculosis from the blood of patients with Crohn's disease. Lancet 364:1039-1044.
Paustian, M. L., A. Amonsin, V. Kapur, and J. P. Bannantine. 2004. Characterization of novel coding sequences specific to Mycobacterium avium subsp. paratuberculosis: implications for diagnosis of Johne's disease. J. Clin. Microbiol. 42:2675-2681.
Paustian, M. L., V. Kapur, and J. P. Bannantine. 2005. Comparative genomic hybridizations reveal genetic regions within the Mycobacterium avium complex that are divergent from Mycobacterium avium subsp. paratuberculosis isolates. J. Bacteriol. 187:2406-2415.
Poupart, P., M. Coene, H. Van Heuverswyn, and C. Cocito. 1993. Preparation of a specific RNA probe for detection of Mycobacterium paratuberculosis and diagnosis of Johne's disease. J. Clin. Microbiol. 31:1601-1605.
Puyang, X., K. Lee, C. Pawlichuk, and D. Y. Kunimoto. 1999. IS1626, a new IS900-related Mycobacterium avium insertion sequence. Microbiology 145(Pt. 11):3163-3168.
Semret, M., G. Zhai, S. Mostowy, C. Cleto, D. Alexander, G. Cangelosi, D. Cousins, D. M. Collins, D. van Soolingen, and M. A. Behr. 2004. Extensive genomic polymorphism within Mycobacterium avium. J. Bacteriol. 186:6332-6334.
Stratmann, J., B. Strommenger, R. Goethe, K. Dohmann, G. F. Gerlach, K. Stevenson, L. L. Li, Q. Zhang, V. Kapur, and T. J. Bull. 2004. A 38-kilobase pathogenicity island specific for Mycobacterium avium subsp. paratuberculosis encodes cell surface proteins expressed in the host. Infect. Immun. 72:1265-1274.
Talbot, E. A., D. L. Williams, and R. Frothingham. 1997. PCR identification of Mycobacterium bovis BCG. J. Clin. Microbiol. 35:566-569.
Tortoli, E., L. Rindi, M. J. Garcia, P. Chiaradonna, R. Dei, C. Garzelli, R. M. Kroppenstedt, N. Lari, R. Mattei, A. Mariottini, G. Mazzarelli, M. I. Murcia, A. Nanetti, P. Piccoli, and C. Scarparo. 2004. Proposal to elevate the genetic variant MAC-A, included in the Mycobacterium avium complex, to species rank as Mycobacterium chimaera sp. nov. Int. J. Syst. Evol. Microbiol. 54:1277-1285.
van Soolingen, D., P. W. Hermans, P. E. de Haas, D. R. Soll, and J. D. Van Embden. 1991. Occurrence and stability of insertion sequences in Mycobacterium tuberculosis complex strains: evaluation of an insertion sequence-dependent DNA polymorphism as a tool in the epidemiology of tuberculosis. J. Clin. Microbiol. 29:2578-2586.(Makeda Semret, David C. A)