The rise and fall of the Mycobacterium tuberculosis genome

Opinion

The rise and fall of the Mycobacterium tuberculosis genome Fre´de´ric J. Veyrier1, Alexander Dufort2 and Marcel A. Behr2,3 1

Unite´ des Infections Bacte´riennes Invasives, De´partement Infection et Epide´miologie, Pasteur Institute, Paris, France Department of Microbiology and Immunology, McGill University, Montreal, QC, H3A 2B4, Canada 3 Department of Medicine, McGill University Health Centre, Montreal, QC, H3G 1A4, Canada 2

When studied from the perspective of non-tuberculous mycobacteria (NTM) it is apparent that Mycobacterium tuberculosis has undergone a biphasic evolutionary process involving genome expansion (gene acquisition and duplication) and reductive evolution (deletions). This scheme can instruct descriptive and experimental studies that determine the importance of ancestral events (including horizontal gene transfer) in shaping the present-day pathogen. For example, heterologous complementation in an NTM can test the functional importance of M. tuberculosis-specific genetic insertions. An appreciation of both phases of M. tuberculosis evolution is expected to improve our fundamental understanding of its pathogenicity and facilitate the evaluation of novel diagnostics and vaccines.

The debated origins of Mycobacterium tuberculosis Before the genomic era there was already a longstanding interest in understanding the origins of bacterial pathogens and the molecular attributes of virulence. Largescale genome sequencing has provided a rapid and unbiased means of uncovering the evolution of many pathogens, contributing to both fundamental microbiological insights and the development of new disease-control strategies [1]. For these reasons, the evolution of one of the most devastating human pathogens, Mycobacterium tuberculosis, has captivated researchers since its discovery in 1882. This interest was stimulated not only by the epidemiologic importance of the pathogen but also by the lack of consensus on its origins and its apparent exception to the stereotypes of bacterial evolution (e.g. acquisition of pathogenicity islands). The first debate concerning the evolution of M. tuberculosis centred on the origin of human tuberculosis (TB). Based on phenotypic similarities between M. tuberculosis (the human pathogen), M. bovis (the cause of bovine TB) and other members of the M. tuberculosis complex (Box 1), investigators had previously proposed that the human epidemic originally arose as a zoonosis from cows at the time of their domestication [2]. Somewhat surprisingly, genomic comparisons between these pathogens placed M. bovis in a more derivative position (based on its smaller genome), with M. tuberculosis occupying a more ancestral position [3,4]. Based on these findings the origin of TB was Corresponding author: Behr, M.A. ([email protected]).

156

now uncertain, but there was little evidence for a simple species jump from the bovine to human host [5]. A second debate then emerged regarding the nature of M. tuberculosis evolution. For most bacterial pathogens where the evolutionary process has been determined, the horizontal acquisition of virulence determinants has been a key process driving their rapid emergence from a closely related avirulent organism [6,7]. By contrast, M. tuberculosis presented as a longstanding human-associated pathogen that underwent reductive genomic evolution as it followed humans around the globe during paleomigration [8,9]. This same reductive trend was observed with subsequent sequenced mycobacterial genomes (i.e. M. leprae, M. bovis, M. avium subsp. paratuberculosis), leading to the suggestion that mycobacterial evolution was exceptional with regards to other pathogens and did not involve horizontal gene transfer (HGT) for the emergence of pathogenicity [10]. This hypothesis was not entirely implausible because gene deletions or mutations could have eliminated anti-virulence genes and/or changed the expression of virulence determinants. Indeed, within the M. tuberculosis complex, M. bovis overexpresses the antigens MPT70/83 because of a mutated anti-sigma factor [11], and the Beijing lineage of M. tuberculosis overexpresses genes of the dormancy regulon [12]. Nevertheless, this hypothesis was biased by the availability of genomic sequence data from terminally differentiated pathogens. The more recent availability of a number of non-tuberculous mycobacteria (NTM) sequences permitted reappraisal of M. tuberculosis evolution. Apparently we had missed a key part of the story, akin to studying the history of Rome from the assassination of Caesar without knowledge of the preceding events. The rise of M. tuberculosis How did Rome emerge as a regional power? How did M. tuberculosis evolve from a related NTM organism? Whereas the NTM have previously been termed atypical mycobacteria in comparison to pathogenic mycobacteria, it is the specialized pathogenic species that are atypical because environmental mycobacteria account for the majority of named species [13]. When considered from this vantage point, NTM sequences provide the optimal reference point for determining what is unique to pathogenic species, including M. tuberculosis. Indeed, independent analyses have detected the ancestral addition of genetic material in the M. tuberculosis genome, either by HGT

0966-842X/$ – see front matter ß 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.tim.2010.12.008 Trends in Microbiology, April 2011, Vol. 19, No. 4

Opinion

Trends in Microbiology April 2011, Vol. 19, No. 4

Box 1. Nomenclature of the Mycobacterium tuberculosis complex

Box 2. Mycobacterium tuberculosis: human pathogen, human symbiont or both?

Mycobacteria are dichotomously classified as belonging to the M. tuberculosis complex or being non-tuberculous mycobacteria (NTM). The latter include opportunistic pathogens, such as M. avium, M. marinum and M. kansasii. Of the M. tuberculosis complex, most human disease is due to M. tuberculosis, although up to half of cases in certain countries of West Africa are due to the related organism M. africanum. TB in other mammalian hosts is caused by a number of genetically related species – voles are infected with M. microti, dassies are infected with Dassie bacillus, seals are infected with M. pinnipedii, oryxes are infected with the Oryx bacillus, goats are infected with M. caprae and cattle are infected with M. bovis. These ecotypes of the M. tuberculosis complex are strictly subspecies, although by convention most refer to M. bovis rather than to M. tuberculosis subsp. bovis. The name M. prototuberculosis has been proposed for a group of organisms that grow as unusual smooth colonies, including M. canetti. Whether these organisms should retain this new designation and whether they should be considered to be members of the M. tuberculosis complex or of an out-group are both subjects of debate; ultimately, their taxonomy and placement should be facilitated by complete genome sequence data.

Pathogenicity refers to the ability of a microorganism to cause disease (i.e. a qualitative property). This contrasts with virulence, which is the degree of damage caused by the microorganism to the host (i.e. a quantitative property). A pathogen is therefore any microorganism that can cause disease. M. tuberculosis clearly fits this definition because the World Health Organization reported 9.4 million new cases of active TB and 1.7 million deaths in 2009. Symbiosis defines a long-term relationship between two dissimilar organisms. Subtypes of symbiosis are mutualism (a relationship beneficial to both organisms), commensalism [a relationship where an organism lives with (or in) another without injury], or parasitism (a relationship where one organism is detrimental to the other). The longstanding relationship of M. tuberculosis with humans also fits the definition of symbiosis. Whether this is strictly parasitism, based on a biomedical perspective, or whether M. tuberculosis provides an unrecognized benefit (e.g. protection against malaria) is currently unknown. In this review we use the terms symbiosis and symbiont, reserving judgment on the nature of the symbiotic relationship.

events before its specialization [14–18] or by the duplication and modification of loci such as those coding for the type VII secretion system [19] and toxin–antitoxin dyads [20]. A new biphasic portrait for the evolution of M. tuberculosis comprises a rise (a dynamic phase with HGT, duplication and deletion) and fall (clonal conservation with deletions of presumably unnecessary genes) (Figure 1a). This revised paradigm of M. tuberculosis evolution permits one to revisit the question – what permits M. tuberculosis to infect one-third of humanity? To address this, one can imagine an experimental investigation re-creating the events that led to the emergence of M. tuberculosis. Macroevolution of the Mycobacterium genus toward M. tuberculosis For M. tuberculosis to evolve from a strictly environmental bacterium into a terminally differentiated symbiont (Box 2) several phenotypes must have been sequentially selected. The bacteria needed (i) protection from predators, probably free-living phagocytes such as amoebae [21], (ii) a survival strategy inside this predator that led to a capacity to multiply, thereby transforming the predator into a host, and (iii) the capacity to transmit between hosts. At this stage, equilibrium between within-host multiplication and between-host transmission would lead to a permissive relationship marked by co-evolution and terminal restriction of M. tuberculosis to its human host. Study of the latter co-evolutionary process can be accomplished by genomic analysis of circulating strains of M. tuberculosis [22]. By contrast, to infer the initial stages of evolution a cross-species comparison is required. Because mycobacteria span a large range of pathogenic potential it follows that many of the attributes selected for in the early steps of this process are commonly encoded in the genomes of several NTM organisms. Consequently, the events preceding each node of mycobacterial evolution could correlate with the appearance of a specific phenotype. Information generated by this approach will help to decipher the stepwise genesis of M. tuberculosis.

In silico multi-locus sequence analysis of 20 genes revealed that, of the sequenced mycobacterial species, M. tuberculosis is phylogenetically the most closely related to M. kansasii [16]. This latter species represents a heterogeneous group of environmental bacteria which can occasionally be associated with TB-like disease. Most human isolates are subtype 1, and this includes the type strain for which the genomic sequence has been determined [23]. M. kansasii is associated with lung disease in Silesian miners [24] and has been estimated to cause approximately 10% of diagnosed TB cases at a mine in South Africa [25]. Notably, following the evolutionary separation of M. kansasii from M. tuberculosis there have been both HGT events and deletions in the genome of the ancestor to M. tuberculosis (Figure 1b). These HGT events could explain the origin of at least 55 M. tuberculosis-specific genes. A small number of human TB cases from the Horn of Africa yielded isolates with unusual smooth colonies, termed M. canetti [26]. Recently, it has been recognized that these organisms have genetic features that are somewhat atypical for the M. tuberculosis complex, such as sequence variability and evidence of recombination [4,27]. In an evolutionary timeline, the M. tuberculosis– M. kansasii divergence occurred before the separation of M. canetti and M. tuberculosis (Figure 1b), after which point there have been only a few examples of HGT [14– 16,28]. Complete sequence data for M. canetti and related isolates (proposed name M. prototuberculosis) should help to determine how many other genomic events (deletions and duplications) occurred in the same interval. Relatively few cases of human disease are due to M. kansasii and exceedingly few cases of human TB are due to M. prototuberculosis or M. canetti; it would appear in both cases that these organisms can cause sporadic human disease, but this does not appear to result in a successful long-term association with the human population. In addition, person-to-person transmission of M. kansasii has not been described. Here lies the distinction between a nonprofessional pathogen (that only sporadically causes infection and disease) and a professional pathogen (causing a communicable infectious disease). We therefore propose that events after the node of evolution separating M. 157

()TD$FIG][ Opinion

Trends in Microbiology April 2011, Vol. 19, No. 4

(a)

Common ancestor of slow-growing mycobacteria HGT shared by pathogenic mycobacteria

HGT specific to each lineage/species

M. avium ~ 5.4 Mb

(b)

Common ancestor of M. kansasii and M. tuberculosis

M. marinum ~ 6.4 Mb

M. tuberculosis ~ 4.4 Mb

M. kansasii ~ 6.4 Mb

Common ancestor of M. canetti and M. tuberculosis

Common ancestor of M. tuberculosis and M. bovis

Gene acquisition via HGT

M. canetti ~ 4.5 Mb M. tuberculosis ~ 4.4 Mb

Genetic deletions

M. kansasii ~ 6.4 Mb

M. bovis ~ 4.3 Mb TRENDS in Microbiology

Figure 1. The rise and fall of M. tuberculosis. (a) After the common ancestor of slow-growing mycobacteria, horizontal gene transfer (HGT) has introduced genes shared across mycobacterial species (blue arrows at main axis) or alternatively species-specific genes found in only one or a few species (green arrows at secondary branches). At this scale it is difficult to appreciate the most proximal events preceding the genesis of M. tuberculosis. (b) M. tuberculosis evolution since the common ancestor with M. kansasii. Gene acquisitions (blue arrows) occurred before and after the divergence of M. tuberculosis and M. canetti, but after the M. tuberculosis–M. bovis node there is no evidence of HGT. Genomic deletions (red arrows) occurred at each stage, including the most recent phase, where they represent the most obvious distinction between the human and bovine forms of the tubercle bacillus.

kansasii from M. tuberculosis are responsible for the establishment of a permissive relationship with humans leading to the TB epidemic. A more complete sampling of both M. kansasii and M. prototuberculosis, including environmental sources and non-human reservoirs, could explain the relatively paucity of human disease in both cases, and potentially identify animal models by using these organisms to test for their virulence. To date, while there have been intriguing reports on the bacterial genetics of M. prototuberculosis, there is little epidemiologic information regarding the human cases, such as occupational risks, exposure to pets and molecular epidemiologic evidence of transmission between contacts. Genomic sequence data for organisms genetically related to M. tuberculosis 158

will undoubtedly help to define the genetic differences between these species. In concert, an enhanced understanding of the biology of these organisms – optimally including their primary reservoir – is expected to provide the phenotypic basis for experimental studies (outlined below) aimed at understanding the functional consequence of these genetic differences. Experimental investigation of M. tuberculosis evolution The determination that attenuated M. bovis strains known as bacillus Calmette–Gue´rin (BCG) lack genomic regions present in virulent M. bovis provided candidate genetic events to explain mechanistically the attenuation of the vaccine strain [29,30]. This led to a series of investigations

Opinion

Trends in Microbiology April 2011, Vol. 19, No. 4

aimed at documenting the functional relevance of BCGassociated deletions [31,32], beginning with region of difference 1 (RD1) which is now known to code for a type VII secretion system [33]. In retrospect, the core observation that enabled the discovery of this virulence determinant was probably not the detection of a 9-gene deletion, but was instead the longstanding observation that M. tuberculosis [()TD$FIG]is at least 5 logs more likely to cause disease after infection

than BCG vaccines (1 in 10 versus 1 in 1 million) [34]. In an analogous manner, one can imagine the experimental dissection of ancestral events (HGT, duplication and deletion) coupled with an epidemiologic understanding of the pathogenicity of organisms before and after these events. At present it is probably not possible to estimate the percentage of individuals infected with M. kansasii or M. prototuberculosis who progress to active disease, but

DIM Rapidly growing mycobacteria DIM M. avium DIM

HO

Rv2957

M. ulcerans Rv2949 Rv2962, ML0126m Rv2958/59m

Δ Rv2962*

M. marinum

Δ Rv2958/59m

O

ML2348/ Me HO

46m/47m Rv2957

Rv2957

MeO

M. leprae O Me HO

OMe HO MeO

Me O

OH

Rv2954m/55m

DIM ML0126m

O

O MeO

OH

O

O MeO

O

DIM

OMe

MkanA1_010100020855 /45m/35m

M. kansasii M. tuberculosis O Me

DIM O

O

DIM

MeO O Me

Me OMe

O

HO

O O O

Me

OMe

OMe

Me

OMe

O

HO

OAc O

O OH

O

Me MeO

O

HO

Me

O

OMe

HO MeO MeO

TRENDS in Microbiology

Figure 2. The construction of the empire: evolution of phenolic glycolipids. In the slow-growing mycobacteria sublineage, phenolic glycolipids (PGLs) evolved from the ancestral molecule phthiocerol dimycoserate (DIM). This schematic representation of mycobacterial phylogeny depicts the gene acquisitions (arrows) predicted to participate in the creation of PGL. Each genetic event is color-coded to match the corresponding modification to the PGL structure shown in the same color. The genes are named with respect to M. tuberculosis H37Rv or the corresponding species annotation. Names followed by ‘m’ (e.g. Rv2954m) encode methyltransferases. Non-underlined genes are predicted to be acquired through HGT, whereas underlined genes have been generated through duplication. Deletions or mutations by transposition are represented by arrows leaving the tree. The blue asterisk next to DRv2962 indicates that the Rv2962 ortholog is mutated in the sequenced strain of M. ulcerans [17], explaining the production of phenolphthiodiolone as shown [50]; however, some isolates of M. ulcerans produce a PGL identical to that of M. marinum [50], suggesting that a subset of strains harbor either an intact glycosyltransferase or a new enzyme with similar activity.

159

Opinion these organisms can readily be tested for virulence in animal models. In parallel, a clearer sequence of the evolutionary events from M. kansasii to M. tuberculosis should provide a refined list of gene acquisitions that correlate with a relevant phenotype, setting the stage for experimental investigation of the functional role of candidate genes. To illustrate this, we have traced the evolution (predicted via genomic comparison) of a locus coding for the production of phenolic glycolipids (PGLs) that are documented to have a role in oxidative stress resistance [35], immunomodulation [36] and cell tropism [37]. PGLs are synthesized on the phthiocerol dimycoserate (DIM) backbone, which is common to all mycobacteria, by adding a phenol ring and methylated sugars (reviewed in [38]). These complex lipids are only produced by slow-growing mycobacteria with species-specific differences in the carbohydrate moiety [38]. By considering the stepwise evolution of this locus (Figure 2) one observes that the introduction via HGT of a chorismate lyase (Rv2949) was essential for the generation of the phenol ring [39], followed later by the addition of species-specific glycosyland methyl-transferases. Experiments using this locus (and others) that have been the site of HGT events can serve to validate bioinformatic predictions by generating a M. tuberculosis-specific product in an NTM organism via heterologous expression of the M. tuberculosis gene(s). Interestingly, Tabouret et al. have proven this concept by expressing the PGL from M. leprae in BCG, in so doing reconstructing some of the steps in the evolution of PGL and demonstrating the role of PGL in modulating early interactions with phagocytes [40]. More generally, a considerable number of HGT genes in M. tuberculosis are predicted to encode transferases [16]. Because these enzymes are predicted to catalyze the transfer of a functional group from one molecule to another, it might be that a key role of these acquired genes is to fine-tune host responses to already existing mycobacterial products, ultimately promoting the stability of the organism within the host. Concluding remarks and future directions Based on the genomic differences between M. kansasii (shotgun sequence available), M. canetti (sequence expected shortly) and M. tuberculosis (genome sequence available), we expect that it will be possible to identify candidate genes that might explain the difference between organisms that sporadically cause human disease and professional pathogens that successfully spread among humans. Beyond the use of this new evolutionary paradigm to guide laboratory experiments it is also conceivable that the study of NTM might help to reframe the conceptual notions of mycobacterial pathogenicity that are often attributed to M. tuberculosis. Pathogenic potential is present in several slow-growing mycobacteria that are able to cause localized disease, however the capacity to establish a long-term host-specific relationship is restricted to a few terminally differentiated mycobacteria. What are the common attributes in different mycobacteria that permit survival in phagocytic cells, impaired phagosome–lysosome fusion, induction of autophagy, and subversion of innate 160

Trends in Microbiology April 2011, Vol. 19, No. 4

defence mechanisms? Are these features shared, or do they differ across mycobacterial species? What are the common attributes that allow prolonged persistence in a eukaryotic host, and do certain attributes, such as the modality of host cell death, differ between species? Further support for the use of NTM to understand M. tuberculosis is provided by the tremendous insights gained from the use of M. marinum as a model organism to dissect the type VII secretion system, to understand the role of the granuloma in mycobacterial infection, and to identify host determinants of resistance via genome-wide screening [41,42]. Perhaps model systems that currently emphasize the first steps of infection (entry) should be recalibrated to address disease (exit) because this is the requisite stage that sets M. tuberculosis apart from most NTM. On a more pragmatic level, understanding the unique genetic attributes of M. tuberculosis will guide our interpretation of novel diagnostic tests. Newer immunodiagnostic tests, termed interferon-g release assays (IGRAs), test for a cell-mediated immune response to antigens (ESAT-6 and CFP-10) that are often described as specific for M. tuberculosis infection. In fact, orthologs of these antigens are found in the completely avirulent organism M. smegmatis [43], and the genomic region encoding these antigens is conserved in many slow-growing mycobacteria with noted exceptions (e.g. absent from M. bovis BCG, M. microti, the Dassie bacillus, M. ulcerans, and M. avium) [44]. In addition, the protein sequences of these antigens are highly conserved (e.g. ESAT-6 from M. kansasii is 99% identical to that of M. tuberculosis). This conservation could imply cross-reactivity for immunological diagnostic assays which detect these antigens and therefore could explain the unexpected false-positive IGRA results in patients from whom M. kansasii was isolated [45]. Furthermore, understanding the genetically unique attributes of M. tuberculosis could conceivably instruct the development of a new vaccine. It has been suggested that a conceptual goal for a vaccine against a virulent pathogen might be to train the immune response to combat virulence factors, such that less-virulent strains of the organism are selected [46]. This paradigm is best exemplified by Corynebacterium diphtheriae where the toxin-based vaccine permits circulation of toxin-negative organisms in the community [47,48]. Based on this model, a live attenuated vaccine that contains structural elements of a bacterium, but lacks genes needed for full pathogenicity, might provide protection against many strains of M. tuberculosis – but inadvertently select for a virulent strain, therefore potentially explaining the rise of Beijing strain organisms with BCG immunization campaigns [49]. A refined understanding of the unique attributes of M. tuberculosis might help to orient vaccination efforts away from obvious candidates, based on the strength of the adaptive immune response they elicit, towards novel host pathways that resist the establishment of permissive symbiosis by M. tuberculosis. The genomic record suggests that the M. tuberculosis genome was not built in a day and that its reductive evolution has been long and complicated, probably taking place on a geologic rather than biologic timescale. It is our contention that a broader appreciation of M. tuberculosis

Opinion evolution, including both its rise to prominence and its long reign of destruction, could help to hasten its ultimate demise. References 1 Yang, X. et al. (2008) Infectious disease in the genomic era. Annu. Rev. Genomics Hum. Genet. 9, 21–48 2 Stead, W.W. (1997) The origin and erratic global spread of tuberculosis. How the past explains the present and is the key to the future. Clin. Chest Med. 18, 65–77 3 Mostowy, S. et al. (2002) Genomic deletions suggest a phylogeny for the Mycobacterium tuberculosis complex. J. Infect. Dis. 186, 74–80 4 Brosch, R. et al. (2002) A new evolutionary scenario for the Mycobacterium tuberculosis complex. Proc. Natl. Acad. Sci. U.S.A. 99, 3684–3689 5 Smith, N.H. et al. (2009) Myths and misconceptions: the origin and evolution of Mycobacterium tuberculosis. Nat. Rev. Microbiol. 7, 537– 544 6 Wren, B.W. (2003) The Yersiniae – a model genus to study the rapid evolution of bacterial pathogens. Nat. Rev. Microbiol. 1, 55–64 7 Schoen, C. et al. (2008) Whole-genome comparison of disease and carriage strains provides insights into virulence evolution in Neisseria meningitidis. Proc. Natl. Acad. Sci. U.S.A. 105, 3473–3478 8 Mostowy, S. and Behr, M.A. (2005) The origin and evolution of Mycobacterium tuberculosis. Clin. Chest Med. 26, 207–216 v–vi 9 Hershberg, R. et al. (2008) High functional diversity in Mycobacterium tuberculosis driven by genetic drift and human demography. PLoS Biol. 6, e311 10 Gal-Mor, O. and Finlay, B.B. (2006) Pathogenicity islands: a molecular toolbox for bacterial virulence. Cell. Microbiol. 8, 1707–1719 11 Said-Salim, B. et al. (2006) Mutations in Mycobacterium tuberculosis Rv0444c, the gene encoding anti-SigK, explain high level expression of MPB70 and MPB83 in Mycobacterium bovis. Mol. Microbiol. 62, 1251– 1263 12 Fallow, A. et al. (2010) Strains of the East Asian (W/Beijing) lineage of Mycobacterium tuberculosis are DosS/DosT-DosR two-component regulatory system natural mutants. J. Bacteriol. 192, 2228–2238 13 Falkinham, J.O., 3rd (2009) Surrounded by mycobacteria: nontuberculous mycobacteria in the human environment. J. Appl. Microbiol. 107, 356–367 14 Jang, J. et al. (2008) Horizontally acquired genomic islands in the tubercle bacilli. Trends Microbiol. 16, 303–308 15 Becq, J. et al. (2007) Contribution of horizontally acquired genomic islands to the evolution of the tubercle bacilli. Mol. Biol. Evol. 24, 1861– 1871 16 Veyrier, F. et al. (2009) Phylogenetic detection of horizontal gene transfer during the step-wise genesis of Mycobacterium tuberculosis. BMC Evol. Biol. 9, 196 17 Stinear, T.P. et al. (2008) Insights from the complete genome sequence of Mycobacterium marinum on the evolution of Mycobacterium tuberculosis. Genome Res. 18, 729–741 18 Gordon, S.V. et al. (2009) Pathogenicity in the tubercle bacillus: molecular and evolutionary determinants. Bioessays 31, 378–388 19 Gey van Pittius, N.C. et al. (2006) Evolution and expansion of the Mycobacterium tuberculosis PE and PPE multigene families and their association with the duplication of the ESAT-6 (esx) gene cluster regions. BMC Evol. Biol. 6, 95 20 Ramage, H.R. et al. (2009) Comprehensive functional analysis of Mycobacterium tuberculosis toxin–antitoxin systems: implications for pathogenesis, stress responses, and evolution. PLoS Genet. 5, e1000767 21 Cirillo, J.D. et al. (1997) Interaction of Mycobacterium avium with environmental amoebae enhances virulence. Infect. Immun. 65, 3759– 3767 22 Comas, I. et al. (2009) Genotyping of genetically monomorphic bacteria: DNA sequencing in Mycobacterium tuberculosis highlights the limitations of current methodologies. PLoS One 4, e7815 23 Alcaide, F. et al. (1997) Heterogeneity and clonality among isolates of Mycobacterium kansasii: implications for epidemiological and pathogenicity studies. J. Clin. Microbiol. 35, 1959–1964 24 Chobot, S. et al. (1997) Endemic incidence of infections caused by Mycobacterium kansasii in the Karvina district in 1968–1995

Trends in Microbiology April 2011, Vol. 19, No. 4

25

26

27 28

29

30 31

32

33 34

35

36

37 38 39 40

41

42

43

44

45

46 47

48

49

50

(analysis of epidemiological data – review). Cent. Eur. J. Public Health 5, 164–173 Churchyard, G.J. et al. (1999) Mycobacterial disease in South African gold miners in the era of HIV infection. Int. J. Tuberc. Lung Dis. 3, 791– 798 van Soolingen, D. et al. (1997) A novel pathogenic taxon of the Mycobacterium tuberculosis complex, Canetti: characterization of an exceptional isolate from Africa. Int. J. Syst. Bacteriol. 47, 1236–1245 Gutierrez, M.C. et al. (2005) Ancient origin and gene mosaicism of the progenitor of Mycobacterium tuberculosis. PLoS Pathog. 1, e5 Rosas-Magallanes, V. et al. (2006) Horizontal transfer of a virulence operon to the ancestor of Mycobacterium tuberculosis. Mol. Biol. Evol. 23, 1129–1135 Mahairas, G.G. et al. (1996) Molecular analysis of genetic differences between Mycobacterium bovis BCG and virulent M. bovis. J. Bacteriol. 178, 1274–1282 Behr, M.A. et al. (1999) Comparative genomics of BCG vaccines by whole-genome DNA microarray. Science 284, 1520–1523 Pym, A.S. et al. (2002) Loss of RD1 contributed to the attenuation of the live tuberculosis vaccines Mycobacterium bovis BCG and Mycobacterium microti. Mol. Microbiol. 46, 709–717 Lewis, K.N. et al. (2003) Deletion of RD1 from Mycobacterium tuberculosis mimics bacille Calmette–Gue´rin attenuation. J. Infect. Dis. 187, 117–123 Abdallah, A.M. et al. (2007) Type VII secretion – mycobacteria show the way. Nat. Rev. Microbiol. 5, 883–891 Casanova, J.L. et al. (1996) Idiopathic disseminated bacillus Calmette– Gue´rin infection: a French national retrospective study. Pediatrics 98, 774–778 Chan, J. et al. (1989) Microbial glycolipids: possible virulence factors that scavenge oxygen radicals. Proc. Natl. Acad. Sci. U.S.A. 86, 2453– 2457 Guenin-Mace, L. et al. (2009) Lipids of pathogenic mycobacteria: contributions to virulence and host immune suppression. Transbound Emerg. Dis. 56, 255–268 Rambukkana, A. (2001) Molecular basis for the peripheral nerve predilection of Mycobacterium leprae. Curr. Opin. Microbiol. 4, 21–27 Onwueme, K.C. et al. (2005) The dimycocerosate ester polyketide virulence factors of mycobacteria. Prog. Lipid Res. 44, 259–302 Stadthagen, G. et al. (2005) p-Hydroxybenzoic acid synthesis in Mycobacterium tuberculosis. J. Biol. Chem. 280, 40699–40706 Tabouret, G. et al. (2010) Mycobacterium leprae phenolglycolipid-1 expressed by engineered M. bovis BCG modulates early interaction with human phagocytes. PLoS Pathog. 6, e1001159 Davis, J.M. and Ramakrishnan, L. (2009) The role of the granuloma in expansion and dissemination of early tuberculous infection. Cell 136, 37–49 Volkman, H.E. et al. (2010) Tuberculous granuloma induction via interaction of a bacterial secreted protein with host epithelium. Science 327, 466–469 Converse, S.E. and Cox, J.S. (2005) A protein secretion pathway critical for Mycobacterium tuberculosis virulence is conserved and functional in Mycobacterium smegmatis. J. Bacteriol. 187, 1238–1245 Gey Van Pittius, N.C. et al. (2001) The ESAT-6 gene cluster of Mycobacterium tuberculosis and other high G + C Gram-positive bacteria. Genome Biol. 2, RESEARCH0044 Kobashi, Y. et al. (2009) Clinical evaluation of the QuantiFERON-TB Gold test in patients with non-tuberculous mycobacterial disease. Int. J. Tuberc. Lung Dis. 13, 1422–1426 Ewald, P.W. (1998) The evolution of virulence and emerging diseases. J. Urban Health 75, 480–491 Mokrousov, I. (2009) Corynebacterium diphtheriae: genome diversity, population structure and genotyping perspectives. Infect. Genet. Evol. 9, 1–15 Walther, B.A. and Ewald, P.W. (2004) Pathogen survival in the external environment and the evolution of virulence. Biol. Rev. Camb. Philos. Soc. 79, 849–869 Parwati, I. et al. (2010) Possible underlying mechanisms for successful emergence of the Mycobacterium tuberculosis Beijing genotype strains. Lancet Infect. Dis. 10, 103–111 Daffe, M. et al. (1992) The phenolic mycoside of Mycobacterium ulcerans: structure and taxonomic implications. J. Gen. Microbiol. 138, 131–137

161