Molecular characterization of Mycobacterium avium subspecies paratuberculosis isolates. Epidemiological distribution in Spain
Elena Castellanos Rizaldos defended the PhD Thesis at the Faculty of Veterinary Medicine of the Complutense University of Madrid
July 9th, 2010
Paratuberculosis or Johne’s disease is a chronic granulomatous inflammation that mainly affects the gastrointestinal tract of domestic ruminants, although it also affects other animal species. The etiological agent is M. a. paratuberculosis, a member of the Mycobacterium avium complex.
The main objective of this thesis, which has been compiled in this summary, is the development and application of new rapid molecular tools to characterize M. a. paratuberculosis. This research is mainly focused on the description of new methodologies capable of discriminating the most slow-growing phenotypes (types I and III). All of these studies have been conducted with financial support from two projects, AGL2005-07792 (Ministry of Science and Innovation, Spain) and ParaTBTools FP6-2004-FOOD-3B-023106 (European Union).
Traditional molecular techniques, such as PFGE and RFLP-IS900 have divided M. a. paratuberculosis into three types; type I (also named “ovine”), type II (“bovine”) and type III (“intermediate”). Nonetheless, there are some isolates (especially the most slow-growing phenotypes) that are not typable by these techniques due to DNA quality and quantity requirements. In addition, these techniques are complex and difficult to apply as routine tests to characterize M. a. paratuberculosis.
Hence, the deficiency of rapid, inexpensive, high resolution molecular tools expensive, capable of discriminating between the three types of M. a. paratuberculosis led us to this line of research. Initially, we began designing different PCR-based techniques to amplify house-keeping genes, which were then sequenced to detect possible type-specific differences at the nucleotide level. From these genes, gyrA and gyrB, topoisomerases involved in the regulation of the helicoidal structure of DNA, presented SNPs that were type-specific (I, II and III). At the same time, one of the SNPs present in the gyrB gene in type III strains revealed a recognition site for the restriction enzyme Hpy188III. The enzymatic digestion of type III strains product (at that moment only distinguishable from type I strains by PFGE and RFLP-IS900) resulted in a different pattern from those of type I and II.
Continuing with this research, we also examined the inh-A gene, whose expression product is the enzyme enoyl-acyl carrier protein reductase, involved in the elongation of fatty acids to mycolic acids, also showed type-specific SNPs. Similar to the gyrB gene, one of the SNPs present in the sequence of the inh-A gene of type III strains constituted a recognition site for the restriction enzyme SinI, which produced a different restriction pattern for type III isolates compared to those obtained for types I and II.
Furthermore, we developed another technique based on PCR and sequencing directed to IS900, one of the most widely exploited target regions in M. a. paratuberculosis for diagnosis and identification purposes. The analysis of the different IS900 sequences also resulted in type-specific SNPs for each of the types (I, II and III). This finding constitutes another tool for the identification of the three M. a. paratuberculosis types from a small DNA quantity, such as DNA recovered from biological samples (blood, tissues, etc.).
Despite the clear division of M. a. paratuberculosis into three types, the difficulty of isolation of the most slow-growing phenotypes and the limited geographical distribution of some of the types has resulted in the consideration of just two M. a. paratuberculosis types in most of the studies (leaving out type III or intermediate strains). As a result, in collaboration with Saint Georges University (London, United Kingdom), we performed a genomic hybridization comparison using microarray technology on a selection of M. a. paratuberculosis isolates obtained from Spain and representing the three types. This study highlighted the genetic differences present between the three types, where it was found that M. a. paratuberculosis types I and III showed 62 ORFs homologous to M. a. hominissuis 104 genome but divergent from the M. a. paratuberculosis K-10 genome.
In addition, the comparison of types I and III strains with the M. a. paratuberculosis K-10 genome revealed deletions of two ORFs in the type I strains analyzed and seven ORFs in the type III strains (whereas six were consecutive loci). This previous data served to optimize PCR-based techniques directed to types I and III specific genomic regions, and constituted another technique to differentiate between these two types. However, the microarray comprises an expensive method that can only be conducted in the correspondent facilities and also requires trained personnel.
Taking this into account, considering the number of laboratories that use real time PCR technology nowadays and in collaboration with the University of Calgary (Canada) we accomplished a novel molecular approach derived from real time PCR principles and the analysis of differences in melting curves. This technique was able to type M. a. paratuberculosis isolates in 1 hour and 30 minutes. Complimentary to this study we also applied these principles to discriminate between MAC members with environmental and clinical relevance, especially in the case of immunocompromised patients.
In addition to this, the different outbreaks of paratuberculosis that occur worldwide and the need to establish links between them have created a demand for a rapid molecular technique able to subtype within the M. a. paratuberculosis types. For this reason, and complementing previous studies we analyzed different MIRU-VNTR loci in a panel of M. a. paratuberculosis isolates of Spanish origin. Consequently, we described a novel locus (VNTR-259) and proposed the standardization of the interpretation of the results in terms of numbers of copies from agarose gels as previously attained for other members of the genus Mycobacterium. The combination of loci used in the study (MIRU-2, MIRU-3, VNTR-25, VNTR-32 and VNTR-259) yielded an index of discrimination HGDI of 0.84, a high value indicative of the utility of this technique for the study of the epidemiology of the different outbreaks of the disease. Moreover, this technique was applied for the first time in Spain in other members of MAC different from M. a. paratuberculosis, such as M. a. avium recovered from birds of prey.
To conclude, the development of these new approaches resulted in the identification, characterization and description of novel type II strains of M. a. paratuberculosis from Spanish origin, sampled from Guadarrama goats. These strains revealed a novel deletion of 16 Kb. Included within this deletion, virulence-related operons were identified, such as the mammalian cell entry genes or mce among others. To date, efforts have been directed to the creation of mce knock-out strains in other members of the genus Mycobacterium. The overall result of this study ended in the description of the first natural mce mutant strain of M. a. paratuberculosis ever reported. The importance of this deletion, that it contains different virulence genes (mce genes and genes members of the PE, Proline-Glutamic acid and PPE, Proline-Proline-Glutamic acid families), was determined by in vitro infections in four cell lines (MOCL-4, BOMAC, THP-1 and CACO-2) and quantitative amplification of pre-rRNA 16S gene. Afterwards, the differential regulation of the genes within the 16 Kb region was evaluated by the analysis of the transcriptome of M. a. paratuberculosis K-10 (reference strain) after in vitro infections in THP-1 and MOCL-4 cell lines.
In summary, the application of the techniques described within this thesis should depend on the purpose of the study and the capabilities of the laboratory where they are to be applied. Therefore, in laboratories with real time PCR and high resolution melt analysis technology, the use of the techniques described in chapter IV to differentiate between M. a. paratuberculosis types and MAC members would be ideal due to the speed and reproducibility of this molecular tool.
For those laboratories where the above technology is not available, a PCR targeting specific regions for types I and III strains of M. a. paratuberculosis (MAV4125: 306 bp or MAV4126: 303 bp) and type III strains (MAP3584: 633 bp or MAP1435: 265 bp) could be used. Ideally, the application of a multiplex PCR, targeting genes MAV4125 or MAV4126 and MAP3584 would differentiate between the three types. Thus, type I strains would show amplification for both genes, type II for just MAP3584 gene and type III strains just for genes MAV4125 or MAV4126.
Conversely, if the aim of the technique was to establish epidemiological links between paratuberculosis outbreaks in different countries and farms, the use of a MIRU-VNTR typing approach would be useful due to its elevated discriminatory power among highly homogeneous isolates.
