Many pathogenic mycobacteria are known to cause severe disease in humans and animals. M. avium subspecies paratuberculosis (Map) is the causative agent of Johne’s disease—a chronic wasting disease affecting ruminants such as cattle and sheep, responsible for significant economic losses in the dairy and beef industries. Due to the lack of treatment options or effective vaccines, mitigating losses can be difficult. In addition, the early stages of Map infection may occur in asymptomatic hosts that continue to shed viable bacteria in their faeces, leading to the infection of other healthy animals. Using multi-locus short sequence repeat (ML-SSR) analysis we previously reported that individual Johne’s positive dairy cattle from farms across the island of Newfoundland were infected by Map with multiple SSR-types simultaneously. The occurrence of multiple mixed genotype infections has the potential to change pathogen and disease dynamics as well as reduce the efficacy of treatments and vaccines. Therefore, we conducted whole genome sequencing (WGS) and single nucleotide polymorphism (SNP) analysis on a subset of these isolates for a more in-depth examination. We also implemented a PCR assay using two discriminatory SNPs and demonstrated the incidence of a mixed infection by three genotypically diverse Map isolates in a single animal. In addition, results show that WGS and SNP analysis can provide a better understanding of the relationship between Map isolates from individual and different animals. In the future such studies on the occurrence of mixed genotype infections could potentially lead to the identification of variable pathogenicity of different genotypes and allow for better tracking of Map isolates for epidemiological studies.
M. avium subsp. paratuberculosisGenome sequencingMixed genotype infectionDepartment of Natural ResourcesPARDP-1415-0011Government of Newfoundland and LabradorPARDP-1516-02Natural Sciences and Engineering Research Council of CanadaNSERC-DG-386417-2010Memorial University of NewfoundlandThis work was supported by grants from the Department of Natural Resources, Government of Newfoundland and Labrador (PARDP-1415-0011) and (PARDP-1516-02). Research in Kapil Tahlan’s laboratory is also funded by the Natural Sciences and Engineering Research Council of Canada (NSERC-DG-386417-2010) and Fraser Davidson received student support from Memorial University of Newfoundland. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.Introduction
The genus Mycobacterium is comprised of acid-fast bacilli, some of which are pathogenic and cause severe disease in humans and animals. For example, M. tuberculosis and M. leprae are the causative agents of tuberculosis and leprosy in humans, respectively. In addition, M. avium and M. intracellulare, two well defined species from the M. avium complex (MAC), have been linked to several diseases in humans and animals including Johne’s disease (JD) in ruminants (Mijs et al., 2002). JD is a chronic wasting disease affecting cattle, sheep, goats and other ruminants caused by M. avium subspecies paratuberculosis (Map) (Harris & Barletta, 2001). Rabbit, deer and other wildlife have also been shown to be infected by Map (Waddell et al., 2016). While various species of wildlife can act as a reservoir of Map (Nugent et al., 2011), the clinical symptoms of JD have only been observed in ruminants (Koets, Eda & Sreevatsan, 2015), camelids (Ghosh et al., 2012) and rabbits/hare (Beard et al., 2001). The symptoms of JD in infected animals are highly similar to those of Crohn’s disease in humans, where the absorptive surface of the gut is reduced due to thickening of the intestinal wall and other factors (Ott, Wells & Wagner, 1999). It has even been suggested that Map may play a role in Crohn’s disease (Eckburg & Relman, 2007), a notion that is still under debate (Bosca-Watts et al., 2015).
Map infections lead to significant economic losses in the dairy and beef industries due to lower yields of milk, reduced slaughter value and premature culling of infected animals (Ott, Wells & Wagner, 1999). Major obstacles in mitigating losses include the lack of affordable treatments that are licensed for food animals (Hermon-Taylor & Bull, 2002) or an effective vaccine that can guarantee complete protection of uninfected animals (Knust et al., 2013). It has also been shown that the early stages of Map infection can occur in asymptomatic hosts that continue to shed viable bacteria in their faeces, leading to subsequent infection of other susceptible individuals (Whitlock et al., 2000). Therefore, there has been a lot of recent interest in understanding the dissemination of Map using source tracking and epidemiological studies (Ahlstrom et al., 2015; Ahlstrom et al., 2016; Leão et al., 2016; Bryant et al., 2016).
While classically defined co-infection refers to cases involving two or more different species of pathogens, it can also include instances where genotypically different strains of the same species of pathogen are involved (referred to as mixed genotype infection) (Cox, 2001). Mixed genotype infections caused by a single bacterial species appear to be quite common (Balmer & Tanner, 2011), but are often overlooked or missed (Read & Taylor, 2001). Therefore, studying JD transmission and dissemination could be further complicated by intra-host evolution of Map or by the co-infection of hosts by multiple genetically divergent Map strains. Whole genome sequencing (WGS) offers a rapid and more precise tool for investigating infectious disease epidemiology compared to the traditionally used methods (Eyre et al., 2012; Eyre et al., 2013). However, WGS is often performed on a single isolate/colony from an individual due to time and financial constraints. If a mixed genotype infection is present, the analysis of a single isolate can completely miss identical, similar or divergent strains infecting the donor and recipient, leading to inaccurate conclusions about transmission (Van den Berg et al., 2005). Therefore, it has been suggested that epidemiological studies require the analysis of multiple isolates from an individual to accurately trace transmission (Döpfer et al., 2008).
Many bacterial infection outbreak and surveillance studies have employed molecular techniques, including WGS, for detecting varying degrees of mixed genotype infections for other bacterial pathogens (Wang et al., 1999; St Sauver et al., 2000; Aranda, Fagundes-Neto & Scaletsky, 2004; Cespedes et al., 2005; Ugolotti et al., 2016). There have also been numerous WGS studies that have examined the genetic diversity of Map from dairy animals (Ahlstrom et al., 2015; Bryant et al., 2016; Ahlstrom et al., 2016; Yue et al., 2016; Leão et al., 2016), but none which have analyzed or addressed mixed Map genotype infections using multiple isolates from a single animal. We recently reported that Map with different short sequence repeat (SSR) types could be isolated from individual dairy animals from the island of Newfoundland (Podder et al., 2015). Here we use WGS, single nucleotide polymorphism (SNP) calling and phylogenetics to analyze six Map isolates, three of which were isolated from a single animal. Our results demonstrate the occurrence of co-infection by genetically distinct Map isolates in Newfoundland dairy cattle.
Materials and MethodsEthics statement
The study was approved by the Institutional Animal Care Committee (IACC, Memorial University of Newfoundland) as an “A” rated protocol (Number: 15-01-KT) because the samples used in the study were obtained from routine veterinary diagnostic submissions unrelated to this research. The report describes molecular and WGS analysis on previously isolated bacteria and did not directly involve any animals.
WGS and molecular typing of isolates
Sample collection, isolation/culture of Map isolates, DNA extraction and subsequent manipulations were carried out as described previously (Podder et al., 2015). Six genetically distinct Map isolates were selected based on SSR profiles (Podder et al., 2015) and DNA was sent for WGS to The Centre for Applied Genomics (The Hospital for Sick Children, Toronto, Canada). Nextera XT libraries were prepared and sequence data was gathered using the Illumina HiSeq platform (Illumina, Inc. USA) with an average read depth of 1,000× coverage. De novo assembly of the six genomes was carried out using the A5 pipeline (Tritt et al., 2012), yielding 98, 90, 94, 90, 97 and 90 contigs for Map isolates 89C, 93B, 95A, 95B, 95E and 96E, respectively. Annotations were performed using the National Center for Biotechnology Information (NCBI) Prokaryotic Genome Annotation Pipeline (PGAP) and the Rapid Annotations based on Subsystem Technology (RAST) servers (Aziz et al., 2008; Brettin et al., 2015; Overbeek et al., 2014), which were analyzed using the Geneious R8 software package (Biomatters Ltd., Auckland, New Zealand). The Whole Genome Shotgun sequences for the six Map isolates have been deposited in GenBank under the BioSample accession numbers LGRY00000000 (89C), LGRZ00000000 (93B), LGSA00000000 (95A), LGSB00000000 (95B), LGSC00000000 (95E) and LGSD00000000 (96E). Using a previously established protocol, raw reads were analyzed at the SNP level for comparison against the reference Map K10 genome (Table S1) and previously sequenced Canadian isolates representative of different phylogenetic clades present in the country (Ahlstrom et al., 2016). The isolate with the highest depth of coverage from each Canadian clade containing more than one isolate was selected for comparison. An alignment of concatenated SNPs was used to produce a maximum likelihood phylogenetic dendrogram using PhyML (Guindon & Gascuel, 2003) with the TPM1uf nucleotide substitution model as determined by jModelTest (Darriba et al., 2012), and 100 bootstrap pseudo-replications to evaluate node support.
A confirmatory PCR assay was designed to amplify a 110 bp DNA fragment using the primer pair F: CTCCTTTCGGCCGCTGTA and R: AGCCCATTCGCTCCGTAT. Two differentiating SNPs identified in the WGS analysis (SNP171 and SNP172), which are in close physical proximity to one another in the genome, were targeted with this single PCR and Sanger sequencing assay using the same set of primers described above (Fig. S1). The assay was first tested using chromosomal DNA from each isolate as a template. Total DNA was extracted from four primary liquid cultures (89, 93, 95 and 96, the same ones that led to the six Map isolates) for subsequent PCR/sequence analysis using the MagMax Total Nucleic Acid Isolation Kit (Thermo Fisher Scientific, Waltham, MA, USA). A no-template DNA negative control was included in the assay, however no positive control was utilized because the amplicons were subsequently sequenced to determine if SNPs were present or not. This assay was utilized to confirm that specific SNPs associated with each isolate could be detected in the original sample and that sample 95 actually contained DNA from the three individual Map isolates used in the WGS analysis.
Results and Discussion
We recently reported that on more than one occasion, Map with different SSR types could be isolated from single animals from dairy farms from Newfoundland (NL), Canada (Podder et al., 2015), suggesting mixed genotype infections. Our previous SSR study only analyzed four genetic loci/repeats to examine genetic diversity. These repetitive loci are known to be unstable (Van Belkum et al., 1998; Kasnitz et al., 2013); therefore, whether the different SSR types indeed represented distinct strains was yet to be confirmed. We used WGS, SNP calling and phylogenetics to analyze six NL Map isolates with different SSR-types (Podder et al., 2015) with the purpose examining their genetic relatedness/diversity at a higher resolution. The NL Map isolates 95A, 95B and 95E all originated from the same animal, whereas the isolates 89C, 93B and 96E came from different animals located on separate farms (Podder et al., 2015). The use of NCBI’s PGAP and the RAST servers allowed for annotation of the six NL Map genome sequences and their comparison with the revised Map K10 reference sequence (Wynne et al., 2010) (Table 1).
10.7717/peerj.2793/table-1
Characteristics of genome sequences of Map isolates from Newfoundland and their comparison with the reference K10 strain.
The table shows the sizes of the respective genomes (in Mb or megabases) and the numbers of SNPs that were identified in the current study. Details regarding sequencing and analysis are provided in the text of the manuscript.
Isolate/Straina
Characteristics
K10b
89C
93B
95Ac
95Bc
95Ec
96E
Genome size (Mb)
4.829
4.777
4.768
4.776
4.772
4.774
4.771
SNPs relative to K10
NAd
196
74
94
74
84
72
Strain-specific (unique) SNPs e
NAd
157
4
64
3
46
1
Notes.
The WGS data for the NL Map isolates represent genomes at the contiguous sequence level, which are not closed or completed.
The revised (Wynne et al., 2010) genome sequence of the K10 strain in the public database was used for comparison.
The three isolates were derived from a single animal whereas all others came from separate animals from different farms. For the Newfoundland isolates, the first number refers to the identity of the animal sampled followed by a letter assigned to a specific isolated Map colony used in the subsequent analysis (Podder et al., 2015).
NA, not applicable as the sequence was used for identifying variant SNPs (single nucleotide polymorphisms) in the other isolates.
These SNPs were only present in the one isolate each analyzed in the current study.
WGS results showed that all of the isolates have a highly similar genome size of approximately 4.77 Mb and are smaller than the ∼4.83 Mb K10 reference genome, which can be explained as our sequences do not represent fully closed genomes (Table 1). The notable similarity between our isolates and K10 was not surprising as Map is known to exhibit low genetic heterogeneity (Collins & De Lisle, 1986; Möbius et al., 2015). SNP analysis showed that Map 89C has nearly 2.0–2.6 times the number of total SNPs relative to K10 when compared to the other NL Map isolates included in the study (Table 1). From the total SNPs we calculated the number of SNPs that were unique to each of the six NL Map isolates within our dataset. It was observed that 89C has by far the largest number of unique SNPs (n = 157), whereas isolates 93B, 95B and 96E range from only 1–4 (Table 1). Of particular interest are isolates 95A, 95B and 95E from the same animal. The pairwise SNP differences between these isolates ranged from 115 to 125 (Table S2). This was considered to be a significant finding because it has been estimated that Map accumulates SNPs at a slower rate than M. tuberculosis (0.3 SNPs per genome per year) (Bryant et al., 2013; Bryant et al., 2016). As such, the molecular clock of Map is too slow to account for within-host evolution based on the numbers of unique SNPs observed in isolates 95A, 95B and 95E, supporting that a mixed genotype infection by genetically distinct strains had occurred.
To visualize the genetic relatedness/divergence between the six Map isolates from Newfoundland and the reference K10 strain, phylogenetic analysis was performed using a concatenated sequence of SNPs from each isolate (Fig. 1A). Results show that the three isolates from the single animal (95A, B and E) do not cluster together in a phylogenetic tree, which further supports a mixed genotype infection. While 95A, B and E are genetically distinct from one another, 95B clustered with 93B and 96E. This would seem to suggest that 95A, 95B and 95E are co-infecting as genetically divergent strains, and that 95B may be epidemiologically related to 93B and 96E isolated from dairy cattle on two different farms.
10.7717/peerj.2793/fig-1
Maximum likelihood phylogenetic trees based on concatenated SNPs using the TPM1uf nucleotide substitution model (Ahlstrom et al., 2015).
(A) Unrooted phylogenetic tree of the Newfoundland Map isolates and the K10 reference strain. (B) Phylogenetic rooted tree including the Newfoundland isolates and representative Canadian Map isolates identified in a previous study (Ahlstrom et al., 2016). The tree was rooted using a divergent Map isolate (represented by the star) unrelated to the current study and is denoted by an asterisk. Bootstrap values with branch support equal to or greater than 70% are displayed in black. (A and B) Newfoundland isolates are displayed in red, the K10 reference strain is displayed in green and representative Canadian Map isolates are displayed in blue.
Similar phylogenetic analysis was also conducted using SNP profiles from divergent Map subtypes previously described from Canada, one of which encompasses 86% of all Canadian isolates (Ahlstrom et al., 2016). Figure 1B shows the genetic relationship between the NL Map isolates and also places them among six isolates A1_075, A1_194, A1_139, A1_377, A1_092 and A1_067 which belong to the Canadian Map clades A, C, D, E, F and H, respectively (Ahlstrom et al., 2016). As expected, 95A, 95B and 95E do not cluster together, but isolates 93B, 95B and 96E do cluster together despite the fact that each one originated from a different farm, suggesting a recent transmission event occurred between these farms, either directly or indirectly. The movement of cattle within and between provinces in the Canadian dairy industry is extensive and is thought to be a major contributor to herd level Map transmission (Ahlstrom et al., 2016). Furthermore, over the last decade more than 200,000 cattle have been imported from the United States of America (Canadian Dairy Information Centre (CDIC), 2013). It should therefore not be surprising that genetic relatedness between geographically distant isolates is observed. Further information regarding a link between these farms, such as cattle purchases, could elucidate how such genetically related Map isolates were geographically dispersed.
In addition to the WGS and SNP analysis, a confirmatory PCR and sequencing assay was developed to detect unique differentiating SNPs (SNP171 and SNP172) found in the individual isolates (Table 2). SNP171 and SNP172 are located at positions 2,316,042 and 2,316,081 in the revised K10 genome (Wynne et al., 2010) respectively, within a non-coding region between the genes mapk_2034 (PPE-repeat protein) and mapk_2035 (Ren71). The assay was tested using chromosomal DNA isolated from axenic Map cultures as a template (Table 2) and was then applied to DNA extracted from the primary cultures inoculated with fecal samples from respective infected animals (Table 2). These primary cultures were the ones that were later plated onto solid agar media to pick individual isolated colonies for establishing axenic cultures for WGS analysis (Podder et al., 2015). The purpose of the PCR/sequencing assay was to ensure that the respective SNP variants found in the individual isolates could also be detected in the original primary cultures. For instance, the 95A, 95B and 95E isolates have T/T/G as SNP172 and A/G/A at SNP171, respectively (Table 2); thus, these alleles should also be detected in DNA extracts from the primary mixed culture (Table 2). In such cases, more than one nucleotide/allele was identified at both loci in DNA extracts from primary cultures (Table 2), which appeared as “doublet” peaks during sequencing. The results confirmed that primary culture 95 contained DNA associated with the 95A, 95B and 95E isolates and that a mixed genotype infection occurred.
10.7717/peerj.2793/table-2
Analysis of two discriminatory SNPs in chromosomal DNA isolated from individual Map isolates and DNA extracted from non-axenic primary cultures.
The identities of the nucleotides were determined using the PCR and sequencing assay that is described under the materials and methods section. The primary cultures were the animal derived mixed cultures that led to the isolation of the individual Map isolates for subsequent analysis.
Isolate/Samplea
SNP172 isolate (primary)b
SNP171 isolate (primary)b
K10
T
T
89C/89
NDc (T)
G (G)
93B/93
C (T)
G (G)
95A/95
T (T/G)d
A (G/A)d
95B/95
T (T/G)d
G (G/A)d
95E/95
G (T/G)d
A (G/A)d
96E/96
T (T)
G (G)
Notes.
Identity of axenic isolate or primary culture used to extract template DNA used in the analysis. The assigned number refers to the identity of the animal from which the primary sample was derived followed by a letter assigned to a specific isolated Map colony used in the subsequent analysis (Podder et al., 2015).
After PCR amplification the identity of the nucleotide associated with the SNP in chromosomal DNA from each isolate (underlined) and in total DNA from primary cultures (shown in parenthesis) was determined using Sanger sequencing.
ND, none detected. A variant SNP was not detected at this location in the isolate when compared to the reference K10 strain.
In some cases more than one nucleotide was detected during analysis of total DNA from primary cultures, which corresponds with the SNPs identified in the separate isolates.
There is evidence for mixed genotype infections by mycobacteria, including other subspecies of M. avium. One such example is a study showing that HIV-positive inmates in a Spanish prison were at risk of exogenous reinfection with multiple strains of M. tuberculosis, some of which had different drug susceptibilities (Chaves et al., 1999). In a separate study (Theisen et al., 1995), infection by two strains of M. tuberculosis with different drug susceptibilities was also reported in an immunocompetent patient. Additionally, multiple strains of M. tuberculosis have been identified in single sputum specimens from patients with active tuberculosis (Warren et al., 2004). Mixed genotype infections by M. avium subspecies avium have also been observed. It was found that certain AIDS patients were sometimes infected with multiple strains of M. avium subspecies avium (Arbeit et al., 1993). Similarly, a mixed M. avium subspecies avium infection in chickens was also reported following examination of DNA isolated from fecal samples (Shitaye et al., 2008). Recently, a mixed infection by M. bovis was reported in cattle diagnosed with bovine tuberculosis and microevolution of the isolates was also characterized (Navarro et al., 2015; Navarro et al., 2016). However, it should be noted that all of the studies mentioned above used conventional molecular typing methods and not WGS analysis.
Overall, our results on the six Newfoundland Map isolates suggest that co-infection of a single dairy cow with at least three different isolates (95A, 95B and 95E) occurred. The biological and clinical significance of understanding mixed genotype infections should not be understated (Cox, 2001). Distinguishing individual strains playing a role during an infection can be problematic, making full understanding of the evolution of pathogens and disease progression difficult. Lacking this understanding, the predicted outcome of mixed genotype infections on disease dynamics and pathogenesis, as well as on treatments and vaccines is limited (Balmer & Tanner, 2011; Navarro et al., 2015). The decision to analyze/sequence a single colony from each sample/specimen is typically made because of financial and logistical considerations in clinical laboratories or in large-scale transmission studies (Van den Berg et al., 2005; Eyre et al., 2013), but the value of analyzing multiple isolates is clear. In the future, it will be interesting to conduct an extensive study on the prevalence of Map mixed genotype infections, which we predict are frequent (Podder et al., 2015). Our results also demonstrate the application of WGS and SNP analysis as a high resolution tool for analyzing mixed genotype infections and for studying intra-host evolution of a pathogen during the process.
Supplemental Information10.7717/peerj.2793/supp-1
Nucleotide sequence of the genetic locus including the SNPs targeted in the PCR assay used to discriminate between the different Map isolates described in the study
The DNA sequence of the genetic locus targeted in the PCR assay that was used to discriminate between the different Map isolates is shown. The sequence corresponds to that present in the Map K-10 reference strain (positive control) and the two SNPs (SNP171 and SNP172) used in the analysis are marked. The arrows indicate the sequences of the forward (F) and reverse (R) primers used for PCR amplification and subsequent DNA sequence analysis.
10.7717/peerj.2793/supp-2
List of different SNPs identified in the NL Map isolates when compared to the reference K-10 strain
Details regarding SNP identification and analysis are described in the text of the manuscript.
10.7717/peerj.2793/supp-3
Distance matrix based on pairwise SNP differences between NL Map isolates, representative Canadian isolates and the reference K-10 strain
Details regarding SNP identification and analysis are described in the text of the manuscript.
We would like to thank Dr. Suzanne Dufour and Joost Verhoeven (Department of Biology, Memorial University of Newfoundland) for providing access to computational resources and for help with managing sequencing data, respectively.
Additional Information and DeclarationsCompeting Interests
The authors declare there are no competing interests.
Author Contributions
Fraser W. Davidson performed the experiments, analyzed the data, wrote the paper, prepared figures and/or tables, reviewed drafts of the paper.
Christina Ahlstrom performed the experiments, analyzed the data, prepared figures and/or tables, reviewed drafts of the paper.
Jeroen De Buck and Hugh G. Whitney conceived and designed the experiments, contributed reagents/materials/analysis tools, reviewed drafts of the paper.
Kapil Tahlan conceived and designed the experiments, contributed reagents/materials/analysis tools, wrote the paper, reviewed drafts of the paper.
Ethics
The following information was supplied relating to ethical approvals (i.e., approving body and any reference numbers):
The study was approved by the Institutional Animal Care Committee (IACC, Memorial University of Newfoundland) as an “A” rated protocol (Number: 15-01-KT) because the samples used in the study were obtained from routine veterinary diagnostic submissions unrelated to this research. The report describes molecular and WGS analysis and did not directly involve any animals.
DNA Deposition
The following information was supplied regarding the deposition of DNA sequences:
The Whole Genome Shotgun sequences for the six Map isolates have been deposited in GenBank under the accession numbers LGRY00000000 (89C), LGRZ00000000 (93B), LGSA00000000 (95A), LGSB00000000 (95B), LGSC00000000 (95E) and LGSD00000000 (96E).
ReferencesAhlstromCBarkemaHWStevensonKZadoksRNBiekRKaoRTrewbyHHaupsteinDKeltonDFFecteauGLabrecqueOKeefeGPMcKennaSLBDe BuckJ2015Limitations of variable number of tandem repeat typing identified through whole genome sequencing of Mycobacterium avium subsp. paratuberculosis on a national and herd level1616110.1186/s12864-015-1387-6AhlstromCBarkemaHWStevensonKZadoksRNBiekRKaoRTrewbyHHaupsteinDKeltonDFFecteauGLabrecqueOKeefeGPMcKennaSLTahlanKDe BuckJ2016Genome-wide diversity and phylogeography of Mycobacterium avium subsp. paratuberculosis in Canadian dairy cattle112e014901710.1371/journal.pone.0149017ArandaKRSFagundes-NetoUScaletskyICA2004Evaluation of multiplex PCRs for diagnosis of infection with diarrheagenic Escherichia coli and Shigella spp42125849585310.1128/JCM.42.12.5849-5853.2004ArbeitRDSlutskyABarberTWMaslowJNNiemczykSFalkinhamJOO’ConnorGTVon ReynCF1993Genetic diversity among strains of Mycobacterium avium causing monoclonal and polyclonal bacteremia in patients with AIDS16761384139010.1093/infdis/167.6.1384AzizRKBartelsDBestAADeJonghMDiszTEdwardsRAFormsmaKGerdesSGlassEMKubalMMeyerFOlsenGJOlsonROstermanALOverbeekRAMcNeilLKPaarmannDPaczianTParrelloBPuschGDReichCStevensRVassievaOVonsteinVWilkeAZagnitkoO2008The RAST server: rapid annotations using subsystems technology97510.1186/1471-2164-9-75BalmerOTannerM2011Prevalence and implications of multiple-strain infections111186887810.1016/S1473-3099(11)70241-9BeardPMRhindSMBuxtonDDanielsMJHendersonDPirieARudgeKGreigAHutchingsMRStevensonKSharpJM2001Natural paratuberculosis infection in rabbits in Scotland124429029910.1053/jcpa.2001.0466Bosca-WattsMToscaJAntonRMoraMMinguezMMoraF2015Pathogenesis of Crohn’s disease: bug or no bug6111210.4291/wjgp.v6.i1.1BrettinTDavisJJDiszTEdwardsRAGerdesSOlsenGJOlsonROverbeekRParrelloBPuschGDShuklaMThomasonJAStevensRVonsteinVWattamARXiaF2015RASTtk: a modular and extensible implementation of the RAST algorithm for building custom annotation pipelines and annotating batches of genomes5Article 836510.1038/srep08365BryantJMSchürchACVan DeutekomHHarrisSRDe BeerJLDe JagerVKremerKVan HijumSASiezenRJBorgdorffMBentleySDParkhillJVan SoolingenD2013Inferring patient to patient transmission of Mycobacterium tuberculosis from whole genome sequencing data1311010.1186/1471-2334-13-110BryantJMThibaultVCSmithDGEMcluckieJHeronISevillaIABietFHarrisSRMaskellDJBentleySDParkhillJStevensonK2016Phylogenomic exploration of the relationships between strains of Mycobacterium avium subspecies paratuberculosis177910.1186/s12864-015-2234-5Canadian Dairy Information Centre (CDIC)2013Animal Industry Market Information System trade databasehttp://www.dairyinfo.gc.ca20 June 2016CespedesCSaïd-SalimBMillerMLoSKreiswirthBNGordonRJVavagiakisPKleinRSLowyFD2005The clonality of Staphylococcus aureus nasal carriage191344445210.1086/427240ChavesFDrondaFAlonso-SanzMNoriegaAR1999Evidence of exogenous reinfection and mixed infection with more than one strain of Mycobacterium tuberculosis among Spanish HIV-infected inmates13561562010.1097/00002030-199904010-00011CollinsDMDe LisleGW1986Restriction endonuclease analysis of various strains of Mycobacterium paratuberculosis isolated from cattle471022262229CoxF2001Concomitant infections, parasites and immune responses122SupplS23S3810.1017/S003118200001698XDarribaDGuillermoLTDoalloRamónPosadaD2012jModelTest 2: more models, new heuristics and parallel computing9877210.1038/nmeth.2109DöpferDBuistWSoyerYMunozMAZadoksRNGeueLEngelB2008Assessing genetic heterogeneity within bacterial species isolated from gastrointestinal and environmental samples: how many isolates does it take?74113490349610.1128/AEM.02789-07EckburgPBRelmanDA2007The role of microbes in Crohn’s disease44225626210.1086/510385EyreDWCuleMLGriffithsDCrookDWPetoTEAWalkerASWilsonDJ2013Detection of mixed infection from bacterial whole genome sequence data allows assessment of its role in Clostridium difficile transmission95e100305910.1371/journal.pcbi.1003059EyreDWGolubchikTGordonNCBowdenRPiazzaPBattyEMIpCLWilsonDJDidelotXO’ConnorLLayRBuckDKearnsAMShawAPaulJWilcoxMHDonnellyPJPetoTEWalkerASCrookDW2012A pilot study of rapid benchtop sequencing of Staphylococcus aureus and Clostridium difficile for outbreak detection and surveillance2e00112410.1136/bmjopen-2012-001124GhoshPHsuCAlyamaniEJShehataMMAl-DubaibMAl-NaeemAHashadMMahmoudOMAlharbiKBJAl-BusadahKAl-SwailemATalaatAM2012Genome-wide analysis of the emerging infection with Mycobacterium avium subspecies paratuberculosis in the Arabian camels (Camelus dromedarius)72e3194710.1371/journal.pone.0031947GuindonSGascuelO2003A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood52569670410.1080/10635150390235520HarrisNBBarlettaRG2001Mycobacterium avium subsp. paratuberculosis in veterinary medicine14348951210.1128/CMR.14.3.489-512.2001Hermon-TaylorJBullT2002Crohn’s disease caused by Mycobacterium avium subspecies paratuberculosis: a public health tragedy whose resolution is long overdue5113610.1099/0022-1317-51-1-3KasnitzNKöhlerHWeigoldtMGerlachGFMöbiusP2013Stability of genotyping target sequences of Mycobacterium avium subsp. paratuberculosis upon cultivation on different media, in vitro and in vivo passage, and natural infection1673–457358310.1016/j.vetmic.2013.09.008KnustBPattonERibeiro-LimaJBohnJJWellsSJ2013Evaluation of the effects of a killed whole-cell vaccine against Mycobacterium avium subsp. paratuberculosis in 3 herds of dairy cattle with natural exposure to the organism242566366910.2460/javma.242.5.663KoetsAPEdaSSreevatsanS2015The within host dynamics of Mycobacterium avium ssp. paratuberculosis infection in cattle: where time and place matter46Article 6110.1186/s13567-015-0185-0LeãoCGoldstoneRJBryantJMcluckieJInácioJSmithDGEStevensonK2016Novel single nucleotide polymorphism-based assay for genotyping Mycobacterium avium subsp. paratuberculosis54355656410.1128/JCM.01958-15MijsWDe HaasPRossauRVanDLRigoutsLPortaelsFVan SoolingenD2002Molecular 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. avium52Pt 515051518MöbiusPHölzerMFelderMNordsiekGGrothMKöhlerHReichwaldKPlatzerMMarzM2015Comprehensive insights in the Mycobacterium avium subsp. paratuberculosis genome using new WGS data of sheep strain JIII-386 from Germany792585260110.1093/gbe/evv154NavarroYRomeroBBouzaEDomínguezLLdJuanGarcía-de-ViedmaD2016Detailed chronological analysis of microevolution events in herds infected persistently by Mycobacterium bovis1839710210.1016/j.vetmic.2015.11.032NavarroYRomeroBCopanoMFBouzaEDomínguezLDe JuanLGarcía-de-ViedmaD2015Multiple sampling and discriminatory fingerprinting reveals clonally complex and compartmentalized infections by M. bovis in cattle7519910410.1016/j.vetmic.2014.11.004NugentGWhitfordEHunnamJWilsonPCrossMDe LisleG2011Mycobacterium avium subsp. paratuberculosis infection in wildlife on three deer farms with a history of Johne’s disease59629329810.1080/00480169.2011.605747OttSLWellsSJWagnerBA1999Herd-level economic losses associated with Johne’s disease on US dairy operations403–417919210.1016/S0167-5877(99)00037-9OverbeekROlsonRPuschGDOlsenGJDavisJJDiszTEdwardsRAGerdesSParrelloBShuklaMVonsteinVWattamARXiaFStevensR2014The SEED and the rapid annotation of microbial genomes using subsystems technology (RAST)42Database issueD206D21410.1093/nar/gkt1226PodderMPBanfieldSEKeefeGPWhitneyHGTahlanK2015Typing of Mycobacterium avium subspecies paratuberculosis isolates from Newfoundland using fragment analysis104e012607110.1371/journal.pone.0126071ReadAFTaylorLH2001The ecology of genetically diverse infections29255191099110210.1126/science.1059410ShitayeJEMatlovaLHorvathovaAMoravkovaMDvorska-BartosovaLTremlFLamkaJPavlikI2008Mycobacterium avium subsp. avium distribution studied in a naturally infected hen flock and in the environment by culture, serotyping and IS901 RFLP methods1271–215516410.1016/j.vetmic.2007.07.026St SauverJMarrsCFFoxmanBSomselPMaderaRGilsdorfJR2000Risk factors for otitis media and carriage of multiple strains of Haemophilus influenzae and Streptococcus pneumoniae6662263010.3201/eid0606.000611TheisenAReichelCRüsch-GerdesSHaasWHRockstrohJKSpenglerUSauerbruchT1995Mixed-strain infection with a drug-sensitive and multidrug-resistant strain of Mycobacterium tuberculosis345896315121513TrittAEisenJAFacciottiMTDarlingAE2012An integrated pipeline for de novo assembly of microbial genomes (the A5 assembly pipeline)79e4230410.1371/journal.pone.0042304UgolottiELargheroPVanniIBandettiniRTripodiGMelioliGDi MarcoERasoABiassoniR2016Whole-genome sequencing as standard practice for the analysis of clonality in outbreaks of meticillin-resistant Staphylococcus aureus in a paediatric setting93437538110.1016/j.jhin.2016.04.003Van BelkumASchererSVan AlphenLVerbrughH1998Short-sequence DNA repeats in prokaryotic genomes622275293Van den BergRJAmeenHAFurusawaTClaasECVan der VormERKuijperEJ2005Coexistence of multiple PCR-ribotype strains of Clostridium difficile in faecal samples limits epidemiological studies54pt 2173179WaddellLRajićAStärkKMcEwenSA2016Mycobacterium avium ssp. paratuberculosis detection in animals, food, water and other sources or vehicles of human exposure: a scoping review of the existing evidence132324810.1016/j.prevetmed.2016.08.003WangINDykhuizenDEQiuWDunnJJBoslerEMLuftBJ1999Genetic diversity of ospC in a local population of Borrelia burgdorferi sensu stricto15111530WarrenRMVictorTCStreicherEMRichardsonMBeyersNGey Van PittiusNCVan HeldenPD2004Patients with active tuberculosis often have different strains in the same sputum specimen169561061410.1164/rccm.200305-714OCWhitlockRHWellsSJSweeneyRWVan TiemJ2000ELISA and fecal culture for paratuberculosis (Johne’s disease): sensitivity and specificity of each method773–438739810.1016/S0378-1135(00)00324-2WynneJWSeemannTBulachDMCouttsSATalaatAMMichalskiWP2010Resequencing the Mycobacterium avium subsp. paratuberculosis K10 Genome: improved annotation and revised Genome sequence192236319632010.1128/JB.00972-10YueRLiuCBarrowPLiuFCuiYYangLZhaoDZhouX2016The isolation and molecular characterization of Mycobacterium avium subsp. paratuberculosis in Shandong province, China8Article 910.1186/s13099-016-0092-6