Molecular epidemiology of nontuberculous mycobacteria.

For reprint orders, please contact: [email protected]

Author for correspondence: Division of Infectious Diseases & Medical Microbiology, McGill University, A5.156, Montreal General Hospital, 1650 Cedar Avenue, Montreal H3G 1A4, Canada n Tel.: +1 514 934 1934 ext. 42815 n Fax: +1 514 934 8423 n [email protected] †

The emergence of nontuberculous mycobacteria (NTM) as impor tant environmental pathogens has stimulated the search for molecular markers to identify NTM sources, determine virulence mechanisms and describe their population structure. The availability of genome sequence data for a number of NTM isolates has permitted a more definitive approach to classification of these organisms based on sequence analysis of polymorphic targets, such as 16S rRNA, hsp65 and the internal transcribed spacer. An alternative approach, based on assessment of conserved inserted and deleted elements, also permits robust branding of clinical and laboratory isolates. Complementary to ‘top-down’ approaches that classify organisms at the species, subspecies and strain level, ‘bottom-up’ methods to determine the genetic similarity of pairs or groups of isolates have also been developed and used. Analysis of large restriction fragments by pulsed-field gel electrophoresis, restriction fragment length polymorphisms of insertion sequences, repetitive genetic elements, arbitrary primed PCR fragments and multilocus sequencing have largely supplanted phenotypic methods for strain comparison, such as serotyping, biotyping and multilocus enzyme electrophoresis. Together, these two sets of tools can provide an enhanced portrait of the NTM and be useful in epidemiologic investigations of the geographic and ecologic provenance of NTM infections. With further study, it is anticipated that the application of these genetic tools to well-defined collections of organisms will not only lead to an improved understanding of the source of NTM infection, but also help identify clinically relevant bacterial subtypes and eventually uncover genetic markers of bacterial virulence. Nontuberculous mycobacteria

The term nontuberculous mycobacteria (NTM) refers to a diverse collection of opportunistic mycobacteria that differ in habitat, growth requirements and metabolic capabilities [1] . Some also infect and cause disease in mammals (including humans), birds and fish [2] . Classically, this group has been defined by exclusion; indeed, in this review we will use NTM to refer to members of the genus that are not obligate host-associated pathogens, such as Mycobacterium tuberculosis and Mycobacterium leprae. The NTM have previously been called atypical mycobacteria but this moniker is inappropriate on a number of grounds. Statistically, host-associated pathogens are atypical among the genus, forming a small minority of recognized species. In addition, the observation that pathogenic members of the genus evolve through a process that includes genomic deletions (termed reductive genomics) may also be atypical [3] . As such, we will use the term NTM to refer to a mix of two conceptual sets of organisms: environmental bacteria that are normal inhabitants of soils, natural waters and drinking-water distribution systems; and host-associated pathogens that

may occasionally cause spill-over infections in humans, such as the poultry, animal and human pathogen Mycobacterium avium subsp. avium and the agent of Johne’s disease, Mycobacterium avium subsp. paratuberculosis [4] . The environmental NTM are normal inhabitants of soils, natural waters, and potable water from drinking-water distribution systems [1] and household plumbing [5] . As these NTM are oligotrophic, they grow in these ecosystems [6] . Because of the high cell-surface hydrophobicity of NTM [7] , numbers are high on particulates, surfaces of pipes (biofilms) and at air–water interfaces (surface slimes) [8,9] . As a consequence of their ubiquitous distribution, humans and other mammals are commonly surrounded by such NTM. Recent evidence has shown that NTM isolates from approximately 50% of households of NTM patients belong to the same species and share the same DNA fingerprint as the patient isolate [Falkinham JO III, Unpublished Data] . The host-associated pathogenic NTM are probably encountered less frequently and/or in fewer numbers than the environmental NTM; despite this, pathogenic NTM represent a disproportionate subject of concern, because of their proven

10.2217/FMB.09.75 © 2009 Future Medicine

Future Microbiol. (2009) 4(8), 1009–1020

Review

Marcel A Behr† & Joseph O Falkinham III

Future Microbiology

Molecular epidemiology of nontuberculous mycobacteria

Keywords molecular epidemiology n mycobacteria n typing

part of

ISSN 1746-0913

1009

Review

Behr & Falkinham

capacity to cause disease in eukaryotic hosts [10] . The reservoir of these organisms is their associated hosts; therefore, one can anticipate that the majority of biomass for these NTM will reside in the host, with some transient spill-over into their immediate environment. For instance, in the case of M. avium subsp. paratuberculosis, a diseased cow that sheds high numbers of organisms may produce over 1011 bacteria per day that are expelled via manure into the farm environment. With time, the number of organisms is expected to decline, at varying rates depending on soil conditions, such that a pasture that has been destocked can accept new uninfected animals after several years without risk of infection [10,11] . Therefore, while it is tempting to treat all NTM as ubiquitous organisms to which humans are regularly exposed, this is not likely to be an accurate representation of pathogenic NTM. Rather, the risk of exposure to any given NTM, and the degree of exposure, is expected to vary across the species as a function of the different environments that we inhabit and exploit. Most NTM infection of humans is thought to occur after we inhale or ingest these bacteria, either directly from their normal habitat or indirectly via contaminated food or water. Droplets suspended in the air (i.e., aerosols) [5] and dust particles [12] are NTM transmission vehicles [13] . In AIDS patients, acquisition of M. avium is most likely via the GI tract [14] . Cervical lymphadenitis in children aged 18 months to 3 years is likely due to ingestion of contaminated water or soil [15] . Host factors are major determinants of NTM infection. Elderly men and women of low body weight are at higher risk for NTM infections [16] . Cystic fibrosis, a‑1‑antitrypsin deficiency, chronic obstructive pulmonary disease, alterations in chest architecture (e.g., pectus excavatum) and other conditions predispose to pulmonary NTM disease [17] . The discovery that heterozygosity for mutations in the CFTR gene is a risk factor for disease [18] is a particularly troubling aspect of NTM epidemiology, as the frequency of individuals heterozygous for CFTR mutations is one in 25 (calculation based on a frequency of newborns with cystic fibrosis as one in 2000 new births). Likewise, the ongoing debate about whether M. avium subsp. paratuberculosis is implicated in the etiology of Crohn’s disease presents concerning statistics, both for the frequency of genetic polymorphisms associated with disease risk [19] and for the increasing prevalence of disease in many countries [20] . It is important to recognize 1010

Future Microbiol. (2009) 4(8)

that the appearance of NTM disease requires a combination of host factors and exposure [2] . In this manner, disease by NTM is clearly distinct from human TB, as humans represent the definitive host for M. tuberculosis. Even in the case of human exposure to pathogenic NTM, human-to-human transmission has not been demonstrated. Therefore, when considering bacterial virulence and host susceptibility, it is probably more realistic to move beyond a simplified paradigm, where bacteria are dichotomized as pathogenic versus nonpathogenic and hosts are likewise divided between the immune-competent or immune-suppressed. In the absence of direct evidence, we propose that isolation from a human sample of a low- virulence NTM is potentially an indicator of a profound host defect, while isolation of an NTM of greater pathogenicity indicates a more subtle defect in host resistance. In this manner, the accurate classification of clinical NTM isolates may serve as a critical stratification criteria for the assessment of host defects associated with NTM disease. Importance of molecular epidemiology for the study of NTM disease

Since genotyping methods became available approximately two decades ago, molecular typing has become an adjunct method to understand the epidemiology of a number of bacterial diseases. However, for certain organisms, such as M. tuberculosis, it has become evident that molecular epidemiology serves not only to confirm what was previously suspected, but also to challenge and reverse prevailing beliefs [21] . This conceptual difference between confirmatory work and exploratory work in part stems from the long and variable incubation period of mycobacterial infections. Whereas exposure to many nonmycobacterial pathogens results in disease following a somewhat predictable incubation period, for most subjects with a mycobacterial disease, it is unlikely that the patient can assign with precision the time and place when they were exposed to the organism responsible for their clinical condition. This has been highlighted by studies in both San Francisco (USA) and Cape Town (South Africa) that show that members of the same household do not necessarily share the same bacterial genotype [22,23] . These same principles govern the potential utility of genotyping methods to provide a refined portrait of NTM epidemiology. Using appropriate classification methods, one can test in the first step whether the NTM species recovered future science group


from a patient sample is the same as that isolated from the environment (e.g., water) to which the patient was exposed. However, this is insufficient grounds to infer transmission and, hence, that this particular environment represents the source of exposure. Rather, following the establishment of an epidemiological link (e.g., the patient drank the water or showered in the water), genetic evidence of a match is required. However, one must be mindful of the possibility that the patient ‘infected’ the environmental site, rather than the reverse. Fingerprint matches are especially useful when the potential sources of infection harbor a wide diversity of different types, as is typically the case for environmental NTM. The recognition that genetically variable NTM are widely distributed in environments shared by humans provided the key prerequisite for molecular epidemiologic studies; namely, the ability to distinguish organisms unrelated to the event of interest from organisms likely to be implicated in the specific disease. Risk analysis for NTM infection requires predictive knowledge of the likelihood of disease. While we might consider it unfortunate to swallow a mouthful of Escherichia coli while swimming at a public beach, most of us would strongly advise against ingesting the equivalent number of bacteria in the case of E. coli O157. From this follows a key component of risk ana lysis; the assignment of an estimated infectious dose for a particular organism. In the case of NTM, an analysis of the risk posed by the presence of organisms in drinking water is premature without knowledge of which subspecies and strains are more likely associated with disease. In a first pass, as Mycobacterium gordonae is rarely associated with disease, yet is commonly recovered from water or sputum samples, it is widely accepted that the isolation of M. gordonae in the laboratory has little predictive value, as compared with, for example, Mycobacterium kansasii [24] . From this, clinicians and microbiologists familiar with the NTM have developed a semiquantitative sense of which NTM are more or less pathogenic. However, specific virulence genes or genetic elements whose presence/position is consistently associated with virulence for M. avium, Mycobacterium intracellulare or other NTM have not yet been identified. Recently, a limited number of genes potentially involved in NTM pathogenicity have been identified, such as a putative pathogenicity island in M. avium subsp. hominissuis [25] , M. avium subsp. paratuberculosis [26] and the macrophage-inducible gene [27] . However, until it is established future science group

Review

whether such genes are present uniquely in the pathogens and that organisms containing these elements are more likely to cause disease in humans, it remains difficult to assign these as virulence genes and predict their utility as molecular epidemiologic markers. A further complicating aspect is that metagenetic events are known to influence virulence in certain NTM. Specifically, M. avium subsp. hominissuis strains form interconvertible colony variants that differ markedly in virulence, susceptibility to antimicrobial agents, and growth and surface characteristics [28–30] . Transparent (SmT) colony types are more virulent, more resistant to antimicrobial agents, more hydrophobic and grow slower compared with the opaque (SmD) variants [29] . As these transparent colony variants appear at frequencies of one per 1000 upon laboratory cultivation that also selects for the opaque colony type [29] , assessment of virulence can be complicated by the composition of the colony types used in the experimental model. Useful ‘types’ of NTM

Before any risk analysis for NTM can be performed, methods to identify virulent types must be employed. The wide diversity of NTM types, particularly seen in M. avium and M. intracellulare (the M. avium complex; MAC), makes it important to be able to identify those habitats with virulent types. Although it might be assumed that isolates recovered from a patient’s environment whose marker fingerprint matches that of the patient represent a virulent type, no study of such isolates has been performed to identify common patterns of markers that would lead to identification of virulent types. Such a comparative study could aim not only to identify virulent types, but could also be used to identify geographic types (e.g., isolates unique to a particular continent or region), ecotypes (e.g., isolates unique to particular ecosystems or habitats) or environmental types (e.g., isolates capable of environmental persistence and degradation of recalcitrant compounds such as chlorinated hydrocarbon pollutants). Such a study would require thoughtful consideration of the typing tool; in particular, what level of discrimination is possible. Not all of the available typing markers will provide the requisite data [31] . In a first step, one can envision typing with a ‘top-down’ assay to provide a precise designation of the organisms of interest to the level of species, subspecies and strain. While this act may appear to merely represent the responsibility of the clinical laboratory, it is worth noting www.futuremedicine.com

1011

Review

Behr & Falkinham

that many hospital labs make no effort to distinguish M. avium from M. intracellulare, let alone to distinguish subspecies of M. avium. It can be argued that the successful epidemiologic study showing a phenotypic difference between these organisms is the prerequisite for requiring laboratories to make this distinction; if patients with M. avium and M. intracellulare are diagnosed and managed the same, there is no medical imperative to distinguish the organism responsible for the clinical syndrome. While it may appear intuitive to separate M. avium from M. intracellulare, given their considerable genetic differences [32] , the very methods that distinguish between these species have led to the identification of genetic variants that are considered to be similar to both, but identical to neither, including the following recently-named organisms: Mycobacterium chimaera, Mycobacterium colombiense, Mycobacterium arosiense, Mycobacterium bouchedurhonense, Mycobacterium marsiliense and Mycobacterium timonens [33–35] . Whether these variants warrant unique names is a matter for taxonomists to decide. For the purpose of NTM epidemiology, it must be remembered that the critical question should not be whether genetic differences are uncovered but, rather, whether these genetic differences predict a different epidemiologic outcome, such as exposure, infection or disease. Methods available for top-down classification or branding of organisms are numerous, resulting in a genetic profile that can be read as a ‘standalone’ attribute of the isolate. Thus, these methods contrast with ‘bottom-up’ tracking methods, which generate patterns that are read for their similarity or dissimilarity as compared with other isolates. Two techniques that serve for top-down classification are sequence analysis for variable regions in conserved genetic elements, and PCR-based identification of conserved inserted elements and genomic deletions. Because of the availability of genome sequence data for a number of prototype isolates, it is now relatively easy to design PCR primers to amplify any of a number of conserved housekeeping genes and assess for sequence variability as compared with a referent strain. The most conserved targets, such as 16S rRNA, provide the greatest range of organisms that can be identified with a given primer set, but typically present lower resolution. In contrast, by selecting a more variable element, the primers so designed may only amplify across a subset of NTM species, but the amplicon so obtained will provide much greater discrimination, to the level of 1012


species, subspecies and strain. As an example, by 16S RNA analysis, all M. avium organisms have identical sequences. However, sequencing the variable 3´ part of hsp65 distinguishes pathogenic subspecies of M. avium (M. avium subsp. paratuberculosis and M. avium subsp. avium) and also provides unique genovars for cow and sheep strains of M. avium subsp. paratuberculosis [36] . This same approach has been employed with a number of different targets, either as part of a multilocus sequence approach [32] or by selecting one specific element (e.g., rpoB, internal transcribed spacer [ITS] or recF) to achieve specific classification of a collection of isolates. As an alternative to sequence-based typing, testing for the presence of specific insertion sequences (ISs) has already proven its utility in identifying subspecies associated with host range. Specifically, M. avium subsp. paratuberculosis carries IS900; no other M. avium sub species carries this IS [37] . Thus, detection of IS900 by PCR is used as evidence for the presence of this particular subspecies in the sample being tested. However, IS900-like sequences have been detected in Mycobacterium species other than M. avium subsp. paratuberculosis [38,39] . Likewise, the presence of a related IS, IS901, defines M. avium subsp. avium and M. avium subsp. silvaticum [40] . Conversely, it has also been noted through postgenomic analysis that certain subspecies of NTM can be defined by the absence of an element that has been lost through the process of reductive genomics. In this manner, it is now known that all M. avium subsp. paratuberculosis isolates lack a region of the M. avium genome called large sequence polymorphism of avium 8 (LSPA8), which is consistently present in other members of the species [41] . Based on this, a simple triple‑primer PCR has been designed to rapidly determine whether a particular isolate is M. avium subsp. paratuberculosis or not, at the same time permitting the lab to test for a mixed culture that would manifest as a double band on the PCR [42] . The use of triple‑primer PCR to distinguish relevant NTM has the potential to be as useful as a triple‑primer PCR assay that is 100% accurate in identifying an isolate as the vaccine strain, M. bovis BCG, as opposed to virulent M. tuberculosis [43] . Markers for molecular epidemiology

In considering typing a group of NTM isolates, it is important to identify the goals of the exercise. If it is to identify a source of NTM infection, then a highly discriminatory marker future science group


is needed to identify those isolates that are either identical or clonally related. In many instances, restriction fragment length polymorphism ana lysis (RFLP) of ISs provides the level of discrimination required for source tracking, as they are capable of transposition. However, too high a rate of transposition leads to a situation where each isolate has a unique fingerprint and unambiguous identification of a source is problematic. Unfortunately, the rates of variation of many of the markers employed in NTM typing (so-called ‘molecular clock’) have not been established. In contrast to source tracking, if the objective is to identify geographic (e.g., African vs European) or ecosystem types (e.g., natural water vs drinking water), the marker’s level of discrimination may not have to be high. For example, an investigation of the global population structure of M. tuberculosis utilized genomic deletion analysis of isolates from around the world [44] . For many markers utilized in typing NTM isolates, it is presumed, without data, that the markers are not subject to selection. In the absence of data comparing the types of NTM isolates from a large range of habitats and hosts, it is inappropriate to speculate that one marker is less subject to selection than another. Even at this present state of uncertainty, it is clear that some markers and the strains carrying those markers are likely subject to selection. For example, it is likely that only a fraction of the NTM isolates entering a water treatment plant are capable of colonizing the drinking-water distribution system and, furthermore, only a proportion of those resident in a drinking-water system might colonize and proliferate in household plumbing, and only a subfraction of those might be capable of causing infection in an exposed human. As most NTM are of relatively low pathogenicity, the relative number of those that can infect and cause disease is more important than their proportion of the NTM population. Based on that scenario, it would be predicted that the number of types would decrease; specifically, a wider range of types would be found in soil than in the drinking-water distribution system, in households and in patients. One major problem encountered in NTM typing is the definition of ‘related’. When fingerprints match, there is no problem in source tracking or placing an isolate in a particular class. One way to define related is to identify the level of relatedness of fingerprint patterns of isolates from a single outbreak. For repetitive sequence typing (rep‑PCR), a similarity of 93% was employed based on the percent similarity of patterns from future science group

Review

a single outbreak cluster [45] . Although Tenover et al. have published guidelines for the interpretation of relatedness for the large restriction fragments separated by pulsed-field gel electrophoresis (PFGE) [46] , those criteria have not been validated in NTM. However, it is important to know the level of variation in the source population. At one extreme is the case where there is little variation in types; too little to provide an adequate level of discrimination. Fortunately, for all the NTM species investigated in depth to date – M. avium, M. intracellulare, Mycobacterium xenopi, Mycobacterium haemophilum, Mycobacterium malmoense and Mycobacterium ulcerans – there is a great deal of variation of types. In fact, clonal variation in environmental sources and patients appears to be the rule. Specifically for members of MAC, it is rarely the case that identical fingerprints are obtained from a patient and their environment [5,12] . In the two instances cited, the fingerprint of a single patient isolate was compared with those from isolates recovered from the patient’s environment; either soil or water. There was considerable clonal variation amongst the environmental isolates and no one pattern matched that of the patient’s isolate [5,13] . However, that was likely due to clonal variation in the infected patient, a hypothesis supported by the demonstration of clonal variation in patients in another study [47] . As a consequence of clonal variation of NTM in both patients and in environmental compartments, it is necessary to collect a number of isolates (five–ten per patient or household) to ensure adequate possibility of recovering the full spectrum of genetic types in the population. Results of typing using different markers

A variety of different markers have been employed for typing, including serotyping, biotyping, multilocus enzyme electrophoresis (MLEE), plasmid typing, rep‑PCR, arbitrary primed PCR (AP‑PCR), PFGE and IS typing. In light of the ongoing discussion of ideal markers for molecular epidemiology, a number of characteristics of the available markers have been investigated. First, the presence or absence of the markers should not be subject to changes due to cultural conditions. This requirement excludes serotyping, biotyping and MLEE from consideration as each can be influenced by culture conditions. Serotyping suffers from a low percentage of typeability; not all isolates can be typed owing to a failure to agglutinate in the presence of antisera or their autoagglutination in the absence of www.futuremedicine.com

1013

Review

Behr & Falkinham

antisera [48] . Biotyping and MLEE both suffer from the fact that the synthesis of pigments and enzymes are influenced by medium composition, growth conditions and growth rate [49,50] . Furthermore, the traits are quite variable, with some isolates being highly pigmented or producing high levels of enzyme activity, and it is unclear where to establish a line between presence and absence of activities. Second, the markers should provide stable patterns, particularly upon repeated cultivation. In many instances, NTM strains have gone through many laboratory passages before a typing laboratory has had a chance to type. Routinely, once an environmental or patient isolate has been recovered, it should be immediately placed in frozen storage (-70°C). Stable fingerprint patterns have been obtained for rep‑PCR [45] , PFGE [51] and IS1245 RFLP [52] by analysis of sequential M. avium isolates recovered from patients for periods of up to 2 years. In one report, among eight patients whose M. avium isolates were typed by IS1245 RFLP, seven had stable isolates, while the RFLP pattern from one patient changed with time [53] . Analysis of PFGE patterns from AIDS patients led to discovery of polyclonal M. avium infections [54] . It is likely that the extent of polyclonal NTM infections or variation in the different fingerprinting pattern has been underestimated to date as it is common practice to recover a single mycobacterial isolates from a patient specimen. Discovery of polyclonal infection was aided by the fact that five–ten isolates were recovered from each patient sample [54] . With respect to the individual typing markers, each has drawbacks. Isolates of M. avium, M. intracellulare and Mycobacterium scrofulaceum (MAIS group) share homologous plasmids, providing evidence for horizontal gene transmission of these elements [55] . Although MAIS plasmids are quite stable and six plasmid homology groups have been identified, the discriminatory power of plasmid typing is limited by the fact that a substantial proportion of MAIS isolates lack plasmids entirely [55] . Rep‑PCR, AP‑PCR, PFGE and IS1245 typing all offer high percentages of typeability and the number of types to provide sufficient discriminatory power. Although it is relatively time consuming and requires expensive instrumentation, PFGE appears to be the accepted gold standard for typing. AP‑PCR is considerably simpler, but its reproducibility between individuals and between laboratories is low. Accuracy in IS1245 typing is dependent upon developing probes or PCR primers that are specific 1014


for IS1245 and do not cross-react or amplify IS1311 – a related NTM IS [56] . Furthermore, the number of IS1245 copies per genome is quite variable, ranging from none to over 20. As is the case for markers, the discriminatory power of IS1245 typing is limited by copy number [57] . Results of typing isolates of M. avium and M. intracellulare using a variety of markers are summarized in Table 1. For each study, an index of discrimination (D) has been calculated following the guidance of Hunter and Gaston:

D =1-

S 1 n j ^n j -1h / N^ N -1h j =1

where N is the number of strains typed, s is the total number of types identified and nj is the number of strains belonging to the jth type [57] . Ribotyping is of little utility for typing members of the genus Mycobacterium as they carry only one or two rRNA genes [58] . The discrimination index for all the methods, save MLEE, is quite high (Table 1) . Where it was possible to separate isolates from different sources (i.e., AIDS patients, non-AIDS patients or the environment), the discrimination indices were similar, as were the number of types, percent typed and percent of isolates in the largest class (Table 1) . This suggests that all the methods are rather robust and useful for typing. Although the discrimination index for serotyping is high, this method carries the disadvantage that the percent of isolates typed is the lowest of all the methods, owing to the fact that some isolates auto-agglutinate or fail to react with the antisera [48] . One limitation of the typing and hence analysis presented in Table 1 is that many of the studies included a proportion of M. intracellulare isolates. In a number of studies, the authors investigated the relationship between results of several typing methods. For example, MLEE type and serotype [59] and PFGE type and serotype [51] failed to correlate. In a few studies, there has been a comparison of typing methods using the same isolates; for example, rep‑PCR and IS1245 [45] , and IS1245 and PFGE [60] . In the study investigating rep‑PCR typing, of 17 M. avium isolates that were typed by both rep‑PCR and IS1245, there was only one instance of an isolate that matched one group by the IS1245 pattern but belonged to a different rep‑PCR group [45] . Likewise, of 16 M. avium isolates that belonged to three different IS1245 RFLP groups (A, B and C), future science group


three isolates (19%) shared a unique but different PFGE band pattern from those shared by members of the other three groups [60] . Thus, it appears that there is no absolute correlation between results of the different typing methods. However, there has not been a comprehensive inter- or intralaboratory study comparing the different typing methods. Furthermore, there have not been a sufficient number of isolates representing different environmental compartments (e.g., water, biofilm, aerosol and soil) and geographic regions to determine their type and relationship to patient isolates. In the following sections, we review information on those Mycobacterium species where typing has been productively employed, when available (Box 1) . Molecular epidemiology of M. kansasii

Mycobacterium kansasii is an environmental opportunistic mycobacterium that has been isolated from tap water collected from hospitals and households. Using a variety of markers, including IS1652 and the major polymorphic tandem repeat (MPTR), PFGE, AP‑PCR and 16S rRNA RFLP, it was shown that isolates of M. kansasii could be placed into five subspecies or biotypes [61,62] . The value of this typing lies in the fact that isolates from AIDS patients were limited to two of the types, yet all five types were represented amongst isolates from immunocompetent patients. Isolates belonging to one group were seldom, if ever, recovered from water, whereas three groups were principally recovered from water and seldom from patients [61,62] . Such information is quite useful for identification of sources that harbor potentially pathogenic M. kansasii populations. The sequencing of the type strain of M. kansasii, which is ongoing, may provide further genetic data to uncover the basis for these different biotypes and clarify the relationship between these biotypes of M. kansasii and the closely related M. gastri. Molecular epidemiology of the M. avium complex

As the results of typing members of MAC have been employed above to characterize the strengths and weaknesses of typing methods, there is no need to duplicate the information here. It is sufficient to indicate that a variety of methods appear to be useful in source tracking; namely rep‑PCR [45] , PFGE [51,54] and IS1245 RFLP [5] . Although none appears to provide more information than another, there has been no side-by-side comparison as was cited above for MLEE type and serotype [59] , future science group

Review

Table 1. Discrimination index for nontuberculous mycobacteria typing methods. Marker

Strains tested

Index*

Number of types

Typed (%) Largest class (%)

Serotype

55AIDS 28non-AIDS 106,550AIDS 119non-AIDS 34AIDS 28non-AIDS 151 47 115 23PT 39PT+ENV 28 104 147 39 93 43

0.83 0.84 0.80 0.95 0.88 0.84 0.80 0.38 0.29 0.76 0.76 0.96 0.99 0.94 0.99 0.99 0.95

8 8 10 25 12 8 8 23 58 4 4 23 98 38 39 89 9

78 79 87 77 82 79 100 100 100 82 72 100 100 80 100 100 100

Biotyping MLEE Plasmids rep-PCR AP-PCR PFGE IS1245

49 47 46 18 39 47 29 4 15 35 31 21 3 20 3 2 21

Ref. [77] [77] [48] [78] [78] [78] [49] [50] [59] [55] [55] [45] [79] [80] [51] [81] [60]

*Discrimination index calculated as described by Hunter and Gaston [57]. AP-PCR: Arbitrary primed PCR; ENV: Environmental; MLEE: Multilocus enzyme electrophoresis; PFGE: Pulsed-field gel electrophoresis; PT: Patient isolate; rep-PCR: Repetitive sequence typing PCR.

and PFGE type and serotype [51] . Clearly, such a study is needed using a collection of environmental and patient isolates; particularly of MAC, as they are responsible for the majority of NTM infections. In the case of M. avium subsp. paratuberculosis, a number of methods have been employed to look for genetic variability among strains in the context of epidemiologic studies. Classically, the use of IS900 RFLP, following protocols established for IS6110-based study of M. tuberculosis, has revealed strain variability, along with distinct profiles for cow and sheep strains of this organism [63,64] . A number of alternative methods have Box 1. Nontuberculous Mycobacterium species where typing has been productively utilized. Slowly growing Mycobacterium species n Mycobacterium kansasii n Mycobacterium avium n Mycobacterium intracellulare n Mycobacterium xenopi n Mycobacterium malmoense n Mycobacterium haemophilum n Mycobacterium ulcerans Rapidly growing Mycobacterium species n Mycobacterium abscessus n Mycobacterium chelonae n Mycobacterium fortuitum

www.futuremedicine.com

1015

Review

Behr & Falkinham

been proposed, each with potential advantages and disadvantages. These include PFGE, ana lysis of short sequence repeats and mycobacterial interspersed repetitive units (MIRU) [65,66] . Beyond the technical aspects of these methods (typeability, reproducibility and standardization), a key research priority is to document the background level of genetic variability among herds subject to epidemiologic study. So long as different patterns can be obtained from animals on the same herd, and identical patterns can be obtained from animals on different herds [67] , it will remain challenging to use these methods in an attempt to track the spread of this organism across space and time. Although multilocus sequence typing (MLS) typically lacks the level of discrimination necessary for source tracking, it produces unambiguous data on strains that can effectively rule out an epidemiologic link when different patterns are observed. Moreover, this method can provide valuable information on population structure that is not readily discernible from classical tracking methods, such as PFGE and RFLP. For example, the use of MLS illustrated narrow diversity amongst isolates of M. avium subsp. avium, M. avium subsp. silvaticum and M. avium subsp. paratuberculosis compared with that observed in M. avium subsp. hominissuis and M. intracellulare [32] . Furthermore, there are differences in the ratio of nonsynonymous to synonymous codon changes (dN/dS), with the ratio lower (0.08) in M. avium subsp. hominissuis compared with the values for M. avium subsp. silvaticum (0.50) and M. avium subsp. paratuberculosis (0.67). It is tempting to speculate that the narrow diversity of the pathogenic clones is due to their limited host range, whereas M. avium subsp. hominissuis and M. intracellulare are less restricted, leading to their widespread distribution in the environment. In the case of M. intracellulare, the availability of a genomic sequence for the type strain of the species is expected to help reveal newer molecular epidemiologic markers and serve as the basis for further studies of its population structure, using MLS or similar approaches. Molecular epidemiology of M. xenopi

An IS, IS1395, has been shown to be of use in typing M. xenopi strains [68] . However, care must be taken as probes for IS1395 can also hybridize to IS1081 of Mycobacterium bovis and IS1245/IS1311 of MAC [68] . Band patterns of IS1395 RFLP and PFGE were compared and shown to have the same level of discrimination, with slightly higher 1016


discrimination shown by IS1395 RFLP owing to its ability to differentiate two strains that shared the same PFGE pattern [68] . To date, IS1395 typing has not been employed for M. xenopi outbreaks, which are typically associated with multiple cases in the same period of time in dwellings with recirculating hot water systems [13] . Molecular epidemiology of M. malmoense

Employment of ribosomal (rRNA) RFLP or ribotyping for strains of M. malmoense illustrated the limitations of such an approach. Using both 16S and 23S rRNA probes, only five different patterns were obtained amongst M. malmoense isolates from Britain and Finland [69] . Better discrimination was attained using AP‑PCR. Seven distinct band patterns were obtained amongst 45 M. malmoense isolates (six from British and 39 from Finnish patients) [70] . In the absence of clearly defined criteria for the authors’ definition of types, it is difficult to judge the significance of the individual types. Inspection of the band patterns in gels suggests that the majority of the strains were clonally related. Molecular epidemiology of M. haemophilum

Two studies have described methods for typing M. haemophilum isolates, neither of which involved an outbreak or source-tracking investigation but rather were demonstrations of possible utility. PFGE showed a rather narrow range of variability of large restriction fragments of genomic DNA amongst a collection of isolates from different hospitals in the New York metropolitan area, Albany (NY), Florida and Texas [71] . A second study used a clone from a recombinant library of a strain of M. haemophilum that was present in multiple copies in the genome [72] . As shown in PFGE gels, there appears to be little variation in the pattern of restriction fragments hybridizing with the clone [72] . The data suggest that genotypic diversity amongst the available isolates of M. haemophilum is limited, as was shown for M. avium subsp. paratuberculosis [32] . As there are few environmental isolates of M. haemophilum available, neither method has been applied to source tracking. Molecular epidemiology of M. ulcerans

Fingerprinting isolates of M. ulcerans was performed using a plasmid, pTBN12, that has been employed for typing M. tuberculosis and future science group


M. kansasii [73] . The RFLP types appeared stable over time and of 70 M. ulcerans isolates collected from patients throughout the world (e.g., Australia and Africa), 11 different patterns were found [73] . However, two different banding patterns – V1 and V2 – only differed by migration of a single band, again illustrating the absence of validated criteria for the definition of a type. Molecular epidemiology of rapidly growing mycobacteria

Pulsed-field gel electrophoresis has proven to be the most useful method for typing rapidly growing mycobacteria: Mycobacterium fortuitum, Mycobacterium chelonae and Mycobacterium abscessus. The investigation of a M. fortuitum outbreak involving positive sputum cultures from 16 patients in a hospital ward by PFGE showed that all patients harbored the isolates of the same PFGE pattern and that of an M. fortuitum isolate recovered from the water-line supplying the showers in the ward [74] . Similarly, PFGE proved of use in studying outbreaks of M. chelonae and M. abscessus, although a discrimination index was not calculated. The PFGE patterns of isolates from five patients collected over 8 months to 13 years showed that the patterns were unchanged [75] . Furthermore, PFGE could be employed to track the source of nosocomial outbreaks [75] . The demonstration of the utility of PFGE for typing these rapidly growing

Review

mycobacteria species also identified a potential problem; fragmentation of DNA in lysates (DNA smears) of a significant proportion of isolates, particularly of M. abscessus [75] . Arbitrary primed PCR was investigated as a method for rapid and simple typing of M. abscessus [76] , as the DNA from a substantial proportion of M. abscessus isolates was substantially fragmented [75] . In spite of the fragmentation of DNA from some isolates (43%), AP‑PCR did provide band patterns that could be interpreted using the criteria developed and used for fingerprinting and source tracking [76] . However, as highlighted by others, there is a potential for misinterpretation owing to experimental variation in band mobility and the need for preparing reagents freshly and running analyses in parallel [51] . Future perspective

It is anticipated that the future will bring forth identification of NTM population structure, geographic types and perhaps even virulent types. The latter is needed to assist the development of a risk analysis for NTM more generally, and MAC in particular. The authors predict that the bottleneck in NTM molecular epidemiology is organisms being subjected to molecular analysis; specifically, whether they are representative of the epidemiologic question, whether they have been collected and stored with care

Executive summary Objectives of molecular epidemiology Importance of defining goals. Genotypic methods are superior to phenotypic-based methods. n Not all typing methods address all questions appropriately. n Identify typing method(s) with appropriate level of discrimination. n n

Objectives for typing nontuberculous mycobacteria Identification of virulent types for risk analysis. Description of population structure of nontuberculous mycobacteria populations. n Identification of forces acting on nontuberculous mycobacteria populations. n Strain tracking to identify source of exposure. n n

Methods for molecular typing of nontuberculous mycobacteria Serotype, biotype and bacteriophage type. Multilocus enzyme electrophoresis. n Pulsed-field gel electrophoresis. n Multilocus sequence typing. n Insertion sequence restriction fragment length polymorphism. n Repetitive-sequence PCR. n n

Molecular epidemiology of nontuberculous mycobacteria Wide diversity within species. Clonal diversity of species within patients and habitats. n Reduced diversity amongst diverged subspecies. n Need for definition of ‘related’ within context of clonal diversity. n n

future science group


1017

Review

Behr & Falkinham

and whether there is sufficient accompanying information (clinical, geographical and ecological) to test the relevant hypotheses about exposure, infection and disease. Only through an understanding of whether different NTM are more or less likely to result in human disease will it be possible to contemplate a switch from genetic markers that permit tracking organisms to genetic markers of disease; a key step forward in risk management for NTM infection. Bibliography

9.

Papers of special note have been highlighted as: n of interest nn of considerable interest 1.

n

2.

n

3.

4.

5.

nn

6.

7.

n

8.

Falkinham JO III: Nontuberculous mycobacteria in the environment. Clin. Chest Med. 23, 529–551 (2002).

11. Whittington RJ, Marsh IB, Taylor PJ,

Marshall DJ, Taragel C, Reddacliff LA: Isolation of Mycobacterium avium subsp paratuberculosis from environmental samples collected from farms before and after destocking sheep with paratuberculosis. Aust. Vet. J. 81, 559–563 (2003). 12. De Groote MA, Pace NR, Fulton K,

Falkinham JO III: Relationship between Mycobacterium isolates from patients with pulmonary mycobacterial infection and potting soils. Appl. Environ. Microbiol. 72, 7602–7606 (2006).

Turenne CY, Wallace RJ Jr, Behr MA: Mycobacterium avium in the postgenomic era. Clin. Microbiol. Rev. 20, 205–229 (2007). Falkinham JO III, Iseman MD, de Haas P, van Soolingen D: Mycobacterium avium in a shower linked to pulmonary disease. J. Water Health 6, 209–213 (2008). Report linking Mycobacterium avium exposure from a showerhead with disease in an individual. Norton CD, LeChevallier MW, Falkinham JO III: Survival of Mycobacterium avium in a model distribution system. Water Res. 38, 1457–1466 (2004). Brennan PJ, Nikaido H: The envelope of mycobacteria. Annu. Rev. Biochem. 64, 29–63 (1995). A review of the characteristics of the envelope of NTM. Falkinham JO III, Norton CD, LeChevallier MW: Factors influencing numbers of Mycobacterium avium, Mycobacterium intracellulare, and other mycobacteria in drinking water distribution systems. Appl. Environ. Microbiol. 67, 1225–1231 (2001).

1018

Wendt SL, George KL, Parker BC, Gruft H, Falkinham JO III: Epidemiology of infection by nontuberculous mycobacteria. III. Isolation of potentially pathogenic mycobacteria from aerosols. Am. Rev. Resp. Dis. 122, 259–263 (1980). Marsh IB, Reddacliff LA: Survival and dormancy of Mycobacterium avium subsp. paratuberculosis in the environment. Appl. Environ. Microbiol. 70, 2989–3004 (2004).

A review of the epidemiology, risk factors and rates of NTM disease. Alexander DC, Turenne CY, Behr MA: Insertion and deletion events that define the pathogen Mycobacterium avium subspecies paratuberculosis. J. Bacteriol. 190, 2479–2487 (2009).

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript.

10. Whittington RJ, Marshall DJ, Nicholls PJ,

Review of the environmental habitats of the major nontuberculous mycobacteria (NTM). Marras TK, Daley CL: Epidemiology of human pulmonary infection with nontuberculous mycobacteria. Clin. Chest Med. 23, 553–567 (2002).

Financial & competing interests disclosure

nn

Report linking NTM exposure in potting soil with disease in individuals.

13. Falkinham JO III: Epidemiology of infection

by nontuberculous mycobacteria. Clin. Microbiol. Rev. 9, 177–215 (1996). n

A review of the environmental habitats and epidemiology of the major NTM.

14. von Reyn CF, Maslow JN, Barber TW,

Falkinham JO III, Arbeit RD: Persistent colonisation of potable water as a source of Mycobacterium avium infection in AIDS. Lancet 343, 1137–1141 (1994). nn

Report of M. avium disease in AIDS patients with exposure to M. avium in water.

15. Wolinsky E: Mycobacterial lymphadenitis in

children: a prospective study of 105 nontuberculous cases with long-term follow up. Clin. Infect. Dis. 20, 954–963 (1995). 16. Prince DS, Peterson DD, Steiner RM

et al.: Infection with Mycobacterium avium complex in patients with predisposing conditions. N. Engl. J. Med. 321, 863–868 (1989).


17. Rodman DM, Polis JM, Heltshe SL et al.: Late

diagnosis defines a unique population of long-term survivors of cystic fibrosis. Am. J. Respir. Crit. Care Med. 171, 621–626 (2005). 18. Kim JS, Tanaka N, Newell JD et al.:

Nontuberculous mycobacterial infection. CT scan findings, genotype, and treatment responsiveness. Chest 128, 3863–3869 (2005). 19. Hugot JP, Zaccaria I, Cavanaugh J et al.:

Prevalence of CARD15/NOD2 mutations in Caucasian healthy people. Am. J. Gastroenterol. 102, 1259–1267 (2007). 20. Economou M, Pappas G: New global map of

Crohn’s disease: genetic, environmental, and socioeconomic correlations. Inflamm. Bowel Dis. 14, 709–720 (2008). 21. Small PM, Hopewell PC, Singh SP et al.:

The epidemiology of tuberculosis in San Francisco. A population-based study using conventional and molecular methods. N. Engl. J. Med. 330, 1703–1709 (1994). 22. Behr MA, Hopewell PC, Paz EA,

Kawamura LM, Schecter GF, Small PM: Predictive value of contact investigation for identifying recent transmission of Mycobacterium tuberculosis. Am. J. Respir. Crit. Care Med. 158, 465–469 (1998). 23. Verver S, Warren RM, Munch Z et al.:

Proportion of tuberculosis transmission that takes place in households in a high-incidence area. Lancet 363, 212–214 (2004). 24. Bodle EE, Cunningham JA, Della-Latta P,

Schluger NW, Saiman L: Epidemiology of nontuberculous mycobacteria in patients without HIV infection, New York City. Emerg. Infect. Dis. 14, 390–396 (2008). 25. Danelishvili L, Wu M, Stang B et al.:

Identification of Mycobacterium avium pathogenicity island important for macrophage and amoeba infection. Proc. Natl Acad. Sci. USA 104, 11038–11043 (2007). 26. Stratmann J, Strommenger B, Goethe R

et al.: A 38-kilobase pathogenicity island specific for Mycobacterium avium subsp. paratuberculosis encodes cell surface proteins expressed in the host. Infect. Immun. 72, 1265–1274 (2004).



27. Yoon JH, Kim E-C, Kim JS, Song EY, Yi J,

Shin S: Possession of the macrophageinduced gene by isolates of the Mycobacterium avium complex is not associated with significant clinical disease. J. Med. Microbiol. 58, 256–260 (2009).

37. Semret M, Turenne CY, Behr MA: Insertion

Pathogenicity of transparent, opaque, and rough variants of Mycobacterium avium in chickens and mice. Am. Rev. Respir. Dis. 102, 499–506 (1970).

IS900-like sequence found in a Mycobacterium sp. other than Mycobacterium avium subsp. paratuberculosis. FEMS Microbiol. Lett. 209, 267–271 (2002).

30. Kansal RG, Gomez-Foores R, Hehta RT:

n

Review of methods, approaches and results of different typing methods.

32. Turenne CY, Collins DM, Alexander DC,

nn

Identification of members of Mycobacterium avium species by Accu-Probes, serotyping, and single IS900, IS901, IS1245 and IS901flanking region PCR with internal standards. J. Microbiol. Methods 64, 333–345 (2006). Extensive genomic polymorphism within Mycobacterium avium. J. Bacteriol. 186, 6332–6334 (2004). nn

Report establishing the genetic relatedness of members of the M. avium complex.

43. Talbot EA, Williams DL, Frothingham R:

34. Murcia MI, Tortoli E, Menendez MC,

Palenque E, Garcia MJ: Mycobacterium colombiense sp. nov., a novel member of the Mycobacterium avium complex and description of MAC-X as a new ITS genetic variant. Int. J. Syst. Evol. Microbiol. 56, 2049–2054 (2006). 35. Tortoli E, Rindi L, Garcia MJ et al.: 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 (2004). 36. Turenne CY, Semret M, Cousins DV,

Collins DM, Behr MA: Sequencing of hsp65 distinguishes among subsets of the Mycobacterium avium complex. J. Clin. Microbiol. 44, 433–440 (2006).


Multilocus enzyme electrophoresis analysis of the Mycobacterium avium complex and other mycobacteria. J. Clin. Microbiol. 29, 264–271 (1991). 51. Mazurek GH, Harman S, Zhang Y et al.:

Large DNA restriction fragment polymorphism in the Mycobacterium avium–M. intracellulare complex: a potential epidemiologic tool. J. Clin. Microbiol. 31, 390–394 (1993). 52. Pestel-Caron M, Arbeit RD: Characterization

of IS1245 for strain typing of Mycobacterium avium. J. Clin. Microbiol. 36, 1859–1863 (1998).

Report documenting the extensive genetic polymorphism in the M. avium complex.

42. Semret M, Turenne CY, de Haas P,

Mycobacterium arosiense sp. nov., a slowly growing, scotochromogenic species causing osteomyelitis in an immunocompromised child. Int. J. Syst. Evol. Microbiol. 58, 2398–2402 (2008).

50. Wasem CF, McCarthy CM, Murray LW:

41. Semret M, Zhai G, Mostowy S et al.:

Behr MA: Mycobacterium avium subsp. paratuberculosis and M. avium subsp. avium are independently evolved pathogenic clones of a much broader group of M. avium organisms. J. Bacteriol. 190, 2479–2487 (2008).

33. Bang D, Herlin T, Stegger M et al.:

Epidemiology of infection by nontuberculous mycobacteria. VI. Identification and use of epidemiologic markers for studies of Mycobacterium avium, M. intracellulare, and M. scrofulaceum. Am. Rev. Respir. Dis. 132, 39–43 (1986).

40. Bartos M, Hlozek P, Svastova P et al.:

31. Maslow JN, Mulligan ME, Arbeit RD:

Molecular epidemiology: application of contemporary techniques to the typing of microorganisms. Clin. Infect. Dis. 17, 153–164 (1993).

49. Fry KL, Meissner PS, Falkinham JO III:

39. Englund S, Bolski G, Johansson KE: An

mutation in Mycobacterium avium. Infect. Immun. 2, 223–228 (1970). Change in colony morphology influences the virulence as well as the biochemical properties of the Mycobacterium avium complex. Microb. Pathog. 25, 203–214 (1998).

distribution, frequency, and specimen source of Mycobacterium avium complex serotypes isolates from patients with acquired immunodeficiency syndrome. J. Clin. Microbiol. 28, 926–929 (1990).

38. Cousins DV, Whittington R, Marsh I,

28. Schaefer WB, Davis CL, Cohn ML:

29. McCarthy C: Spontaneous and induced

48. Yakrus MA, Good RC: Geographic

sequence IS900 revisited. J. Clin. Microbiol. 44, 1081–1083 (2006). Masters A, Evans RJ, Kluver P: 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 (1999).

Collins DM, Behr MA: Differentiating host-associated variants of Mycobacterium avium by PCR for detection of large sequence polymorphisms. J. Clin. Microbiol. 44, 881–887 (2006).

53. Bauer J, Andersen AB: Stability of insertion

sequence IS1245, a marker for differentiation of Mycobacterium avium strains. J. Clin. Microbiol. 37, 442–444 (1999). 54. Slutsky AM, Arbeit RD, Barber TW et al.:

Polyclonal infections due to Mycobacterium avium complex in patients with AIDS detected by pulsed-field gel electrophoresis of sequential clinical isolates. J. Clin. Microbiol. 32, 1773–1778 (1994).

PCR identification of Mycobacterium bovis BCG. J. Clin. Microbiol. 35, 566–569 (1997). 44. Gagneux S, DeRiemer K, Van T et al.:

Variable host–pathogen compatibility in Mycobacterium tuberculosis. Proc. Natl Acad. Sci. USA 103, 2869–2873 (2006). 45. Cangelosi GA, Freeman RJ, Lewis KN et al.:

Evaluation of a high-throughput repetitivesequence-based PCR system for DNA fingerprinting of Mycobacterium tuberculosis and Mycobacterium avium complex strains. J. Clin. Microbiol. 42, 2685–2693 (2004). 46. Tenover FC, Arbeit RD, Goering RV et al.:

Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J. Clin. Microbiol. 33, 2233–2239 (1995). 47. von Reyn CF, Jacobs NJ, Arbeit RD,

Maslow JN, Niemczyk S: Polyclonal Mycobacterium avium complex infection in patients with AIDS: variation in antimicrobial susceptibilities of different strains of M. avium isolated from the same patient. J. Clin. Microbiol. 33, 1008–1100 (1995).


Review

nn

Report documenting polyclonal NTM infections in AIDS patients.

55. Jucker MT, Falkinham JO III: Epidemiology

of infection by nontuberculous mycobacteria. IX. Evidence for two DNA homology groups among small plasmids in Mycobacterium avium, Mycobacterium intracellulare, and Mycobacterium scrofulaceum. Am. Rev. Respir. Dis. 142, 858–862 (1990). 56. Picardeau M, Vincent V: Typing of

Mycobacterium avium isolates by PCR. J. Clin. Microbiol. 34, 389–392 (1996). 57. Hunter PR, Gaston MA: Numerical index of

the discriminatory ability of typing systems: an application of Simpson’s index of diversity. J. Clin. Microbiol. 26, 2465–2466 (1988). 58. Bercovier H, Kafri O, Sela S: Mycobacteria

possess a surprisingly small number of ribosomal RNA genes in relation to the size of their genome. Biochem. Biophys. Res. Commun. 136, 1136–1141 (1986).

1019

Review

Behr & Falkinham

59. Feizabadi MM, Robertson ID, Cousins DV

et al.: Genetic characterization of Mycobacterium avium isolates recovered from humans and animals in Australia. Epidemiol. Infect. 116, 41–49 (1996). 60. Legrand E, Sola C, Verdol B. Rastogi N:

Genetic diversity of Mycobacterium avium recovered from AIDS patients in the Caribbean as studied by a consensus IS1245-RFLP method and pulsed-field gel electrophoresis. Res. Microbiol. 151, 271–283 (2000). 61. Picardeau M, Prod’Hom G, Raskine L,

LePennec MP, Vincent V: Genotypic characterization of five subspecies of Mycobacterium kansasii. J. Clin. Microbiol. 35, 25–32 (1997). 62. Alcaide F, Richter I, Bernasconi C et al.:

Heterogeneity and clonality among isolates of Mycobacterium kansasii: implications for epidemiological and pathogenicity studies. J. Clin. Microbiol. 35, 1959–1964 (1997). 63. Collins DM, Gabric DM, de Lisle GW:

Identification of two groups of Mycobacterium paratuberculosis strains by restriction endonuclease analysis and DNA hybridization. J. Clin. Microbiol. 28, 1591–1596 (1990). 64. Pavlik I, Bejckova L, Pavlas M, Rozsypalova Z,

Koskova S: Characterization by restriction endonuclease analysis and DNA hybridization using IS900 of bovine, ovine, caprine and human dependent strains of Mycobacterium paratuberculosis isolated in various localities. Vet. Microbiol. 45, 311–318 (1995). 65. Amonsin A, Li LL, Zhang Q et al.: Multilocus

short sequence repeat sequencing approach for differentiating among Mycobacterium avium subsp. paratuberculosis strains. J. Clin. Microbiol. 42, 1694–1702 (2004). 66. Sevilla I, Li L, Amonsin A et al.: Comparative

analysis of Mycobacterium avium subsp. paratuberculosis isolates from cattle, sheep and goats by short sequence repeat and pulsedfield gel electrophoresis typing. BMC Microbiol. 8, 204 (2008). 67. Motiwala AS, Strother M, Amonsin A et al.:

Molecular epidemiology of Mycobacterium avium subsp. paratuberculosis: evidence for

1020

limited strain diversity, strain sharing, and identification of unique targets for diagnosis. J. Clin. Microbiol. 41, 2015–2026 (2003). 68. Picardeau M, Varnerot A, Rauzier J,

Mycobacterium abscessus strains from nosocomial outbreaks. J. Clin. Microbiol. 35, 3132–3239 (1997). 77. Horsburgh CR Jr, Cohn DL, Roberts RB

Gicquel B, Vincent V: Mycobacterium xenopi IS1395, a novel insertion sequence expanding the IS256 family. Microbiology 142, 2453–2461 (1996).

et al.: Mycobacterium avium–M. intracellulare isolates from patients with or without acquired immunodeficiency syndrome. Antimicrob. Agents Chemother. 30, 955–957 (1986).

69. Kauppinen J, Pelkonen J, Katila M-J: RFLP

analysis of Mycobacterium malmoense strains using ribosomal RNA gene probes: an additional tool to examine intraspecies variation. J. Microbiol. Meth. 19, 261–267 (1994). 70. Kauppinen J, Mantyjarvi R, Katila M-J:

Random amplified polymorphic DBA genotyping of Mycobacterium malmoense. J. Clin. Microbiol. 32, 1827–1829 (1994).

78. Hoffner SE, Kallenius G, Petrini B,

Brennan PJ, Tsang AY: Serovars of Mycobacterium avium complex isolated from patients in Sweden. J. Clin. Microbiol. 28, 1105–1107 (1990). 79. Pfaller SL, Aronson TW, Holtzman AE,

Covert TC: Amplified fragment length polymorphism analysis of Mycobacterium avium complex isolates recovered from southern California. J. Med. Microbiol. 56, 1152–1160 (2007).

71. Yakrus MA, Straus WL: DNA

polymorphisms detected in Mycobacterium haemophilum by pulsed-field gel electrophoresis. J. Clin. Microbiol. 32, 1083–1084 (1994).

80. Burki DR, Bernasconi C, Bodmer T,

Telenti A: Evaluation of the relatedness of strains of Mycobacterium avium using pulsed-field gel electrophoresis. Eur. J. Clin. Microbiol. Infect. Dis. 14, 212–217 (1995).

72. Kikuchi K, Bernard EM, Kiehn TE,

Armstrong D, Riley LW: Restriction fragment length polymorphism analysis of clinical isolates of Mycobacterium haemophilum. J. Clin. Microbiol. 32, 1763–1767 (1994).

81. Picardeau M, Varnerot A, Lecompte T,

Brel F, May T, Vincent V: Use of different molecular typing techniques for bacteriological follow-up in a clinical trial with AIDs patients with Mycobacterium avium bacteremia. J. Clin. Microbiol. 35, 2503–2510 (1997).

73. Jackson K, Edwards R, Leslie DE, Hayman J:

Molecular typing for Mycobacterium ulcerans. J. Clin. Microbiol. 33, 2250–2253 (1995). 74. Burns DN, Wallace RJ Jr, Schultz ME et al.:

Nosocomial outbreak of respiratory tract colonization with Mycobacterium fortuitum: demonstration of the usefulness of pulsedfield gel electrophoresis in an epidemiologic investigation. Am. Rev. Respir. Dis. 144, 1153–1159 (1991).

Affiliations n

75. Wallace RJ Jr, Zhang Y, Brown BA, Fraser V,

Mazurek GH, Maloney S: DNA large restriction fragment patterns of sporadic and epidemic nosocomial strains of Mycobacterium chelonae and Mycobacterium abscessus. J. Clin. Microbiol. 31, 2697–2701 (1993). 76. Zhang Y, Rajagopalan M, Brown BA,

Wallace RJ Jr: Randomly amplified polymorphic DNA PCR for comparison of


n

Marcel A Behr Division of Infectious Diseases & Medical Microbiology, McGill University, A5.156, Montreal General Hospital, 1650 Cedar Avenue, Montreal H3G 1A4, Canada Tel.: +1 514 934 1934 ext. 42815 Fax: +1 514 934 8423 [email protected] Joseph O Falkinham III Department of Biological Sciences, Virginia Polytechnic Institute & State University, Blacksburg, VA 24061–20406, USA
