You are on page 1of 8

TIMI-1138; No.

of Pages 8

Review

Diversity and disease pathogenesis


in Mycobacterium tuberculosis
Digby F. Warner, Anastasia Koch, and Valerie Mizrahi
MRC/NHLS/UCT Molecular Mycobacteriology Research Unit, DST/NRF Centre of Excellence for Biomedical TB Research, Institute of
Infectious Disease and Molecular Medicine and Department of Clinical Laboratory Sciences, University of Cape Town, Cape Town,
South Africa

The increasing availability of whole-genome sequence


(WGS) data for Mycobacterium tuberculosis, the bacterium that causes tuberculosis (TB), suggests that circulating genotypes have been molded by three dominant
evolutionary forces: long-term persistence within the
human population, which requires a core programme
of infection, disease, and transmission; selective pressure on specific genomic loci, which provides evidence
of lineage-specific adaptation to host populations; and
drug exposure, which has driven the rapid emergence of
resistant isolates following the global implementation of
anti-TB chemotherapy. Here, we provide an overview of
these factors in considering the implications of genotypic diversity for disease pathogenesis, vaccine efficacy,
and drug treatment.

situated [9]. Until recently, the MTBC was considered clonal


or monomorphic [10]. As a result, the varied outcomes
following exposure of an individual to M. tuberculosis were
reasonably assumed to depend almost exclusively on host
genetics and environmental factors. Similarly, the efficacy
(and failure) of treatment and prophylaxis was in turn
understood as a function of the host (and compliance).
Bacillary genotypic variation was considered unimportant.
The recent availability of WGS technologies has rejected
this model: increasing evidence of strain diversity [7,11],
lineage-specific adaptation to host populations [12,13], and
microvariation within hosts and communities [1416] instead supports the idea that mycobacterial genetics and,
therefore, function, are a significant element in determining the heterogeneous outcomes of infection.

Genetic diversity in the Mycobacterium tuberculosis


complex
TB is a global problem, with recent reports estimating
approximately 8.6 million new cases and 1.3 million deaths
annually [1]. This is despite the existence of effective frontline combination chemotherapy, a widely administered vaccine, and the allocation over the past decade of massive
resources to develop improved interventions [2,3]. Co-infection with HIV, and the emergence of drug resistance have
amplified the problem; however, these represent relatively
recent, or modern (Figure 1), developments in the evolution
of the causative agent, Mycobacterium tuberculosis, as an
obligate human pathogen [4].
Modern bacteriology has been transformed by recent
advances in high-throughput DNA sequencing technology
[5] that have enabled the democratization of whole-genome
sequencing [6], and the impact on TB genomics has been
profound [7]. Mycobacterium tuberculosis is one member of a
group of closely related bacteria known as the Mycobacterium tuberculosis complex (MTBC), which comprises seven
closely related human lineages [8], animal-adapted strains
(including the TB vaccine strain, Mycobacterium bovis;
BCG), and the more distantly related Mycobacterium canettii group, in which the smooth tubercle bacilli (STB) are

The M. tuberculosis infection cycle


As an obligate pathogen, the persistence of M. tuberculosis
within the human population depends on the ability to
drive successive cycles of infection, disease (in some cases,
subclinical TB [17] followed by reactivation), and transmission. The reliance on a single host species necessarily
exposes the infecting pathogen to multiple potential evolutionary cul-de-sacs that might arise as a consequence of
the elimination of the bacillus (clearance) or the demise of
the organism within an infected individual (controlled
subclinical infection, or host death) before it is able to
ensure transmission to a new host. Moreover, the capacity
for the organism to remain viable during extended periods
of subclinical TB disease means that the infection cycle is
not defined by a uniform duration. For this reason, accurate dating of the MTBC remains a contentious issue: while
one study suggested that the complex emerged approximately 70 000 years ago [8], more recent work estimates
this occurrence at approximately 5000 years ago [18]. Nevertheless, circulating M. tuberculosis isolates represent the
genotypes that have successfully adapted to human colonization over a timescale of thousands of years [4,8], a
process that is marked by several historical events that
might have impacted the inferred coevolution of host and
pathogen (Figure 1) [13].
Many bacterial pathogens can accelerate evolution
through the enhanced activity of their DNA repair machinery (e.g., recombination) or aggressive sampling of the
immediate environment (e.g., fratricide, natural competence, and conjugation). Horizontal gene transfer (HGT)
had an important role in the emergence of M. tuberculosis

Corresponding authors: Warner, D.F. (Digby.warner@uct.ac.za);


Mizrahi, V. (valerie.mizrahi@uct.ac.za).
Keywords: Mycobacterium tuberculosis; genomics; epistasis; evolution; mutagenesis;
drugs; vaccine.
0966-842X/
2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tim.2014.10.005

Trends in Microbiology xx (2014) 18

TIMI-1138; No. of Pages 8

Review

Trends in Microbiology xxx xxxx, Vol. xxx, No. x

Informaon available from wholegenome sequencing

Human evoluon

M
T
BM
T
B
C

Evoluon of
MTBC from
last common
ancestor

A
n
c
i
e
n
t

Host nutrion

Lineage 4
(Europe)

Strain evoluon
to produce seven
main lineages of
MTB

Evoluonary pressures
impacng MTB

Lineage 2
(East Asia)
Lineage 3
(Central Asia)

Migraon of
humans out of
Africa
(70 000 years ago)

Lineage 5 and 6 Lineage 7


Lineage 1
(West Africa) (Ethiopia)
(Indian Ocean)

Ongoing transmission
between hosts
(Local and global)

Industrial revoluon
(19th century)
BCG
vaccinaon (1920s)

Diversity within
an individual

Diversity within
sites of infecon
in an individual

Anbioc therapy
(1950s)

M
o
d
e
r
n

HIV
(1970s)

TRENDS in Microbiology

Figure 1. The impact of whole-genome sequencing on reconstructing the evolutionary history of Mycobacterium tuberculosis (MTB). Abbreviation: MTBC, M. tuberculosis
complex.

as an exquisitely human-adapted pathogen [19] and ongoing recombination has been suggested as a source of
genetic variation [20]. However, little, if any, evidence
exists for a role of HGT in recent evolution in the MTBC
[4]. Instead, the modern evolution of M. tuberculosis has
been driven by chromosomal rearrangements and mutations, features that result in part from the ecological
isolation of the bacillus, as well as the bottlenecks that
occur during transmission [12]. Contrary to some other
bacterial and mycobacterial species [21], M. tuberculosis
does not have plasmids, and genetic drift is primarily
responsible for diversification and adaptation of this group
2

of organisms [12]. However, evidence that the population


structure of human MTBC is highly subdivided, both
geographically and within the lungs of infected individuals
[12], suggests the capacity for diversification.
Evidence for microdiversity
Numerous studies have identified significant genotypic diversity within bacilli isolated from single hosts [14,22
25]. In some cases, this has been attributed to mixed infection with two distinct strains [14,22], a phenomenon that is
likely to occur in high-burden settings with an elevated force
of infection [26,27] and, importantly, suggests the potential

TIMI-1138; No. of Pages 8

Review

Trends in Microbiology xxx xxxx, Vol. xxx, No. x

It is also possible that diversity is lost in sample collection, or during downstream manipulations required for
strain propagation and DNA isolation for WGS
(Figure 2). Clinical isolates are usually cultured from
sputum samples, which may not harbor bacteria that
are representative of the entire population residing within
the host, and might instead contain only a subset of the
phenotypic (and genetic) variants. Some strains might not
grow in laboratory media, while others might grow so well
as to dominate cultured bacillary populations [26]. Moreover, important new evidence suggests that propagation in
laboratory media induces genomic changes in cultured
isolates: a recent study reported a strong selective advantage for a large genomic duplication (approximately
350 kb) that arose in the bacillary population after only
five rounds of passage in broth media, and was associated
with attenuated virulence in mice [31]. In addition to the
implications for genotypephenotype analyses, this result
reinforces the potential to select inadvertently for laboratory-adapted variants, a possibility that is generally not
considered in genomic studies and might influence epidemiological inferences of strain prevalence and fitness [32].

for direct competition between infecting genotypes. In addition, there is increasing evidence of microdiversity within M.
tuberculosis populations that develop from a single infecting
strain [23,24]. These observations reinforce earlier work [28]
which demonstrated that different drug-resistance alleles
could arise in discrete pulmonary lesions from a single, drugsusceptible infecting genotype.
In some respects, it has been difficult to reconcile the
levels of intrapatient diversity with the lower levels of
genotypic variation detected within transmission clusters.
For example, while epidemiologically clustered strains
might differ by as few as five SNPs [16], as many as seven
SNPs separated strains isolated from the lungs of an
individual patient [24]. Therefore, it appears that the same
degree of heterogeneity can characterize intra- and interpatient diversity. By contrast, a separate report identified
only two SNPs (in katG, encoding the catalase-peroxidase
enzyme that activates the frontline anti-TB drug, isoniazid, and in rpoB, which encodes the b subunit of the RNA
polymerase complex, the target of another key frontline
agent, rifampicin) in serial isolates that developed sequential resistance to isoniazid and rifampicin over 12 years in a
noncompliant patient [29]. As noted elsewhere [30], the use
in this case of a reference genome comprising pooled data
from all the serial isolates might have obscured SNPs
present at low frequencies, highlighting the potential impact of the method used for genome assembly on the
interpretation of sequence data [19]. The same caveat is
likely to apply generally to studies that rely on the use of
reference genomes for sequence alignment and assembly:
as noted recently [6], improved methods to detect larger
genomic deletions and alterations are required to provide
better insight into the dynamics that might underlie microevolution.

Innate immunity

Adapve immunity

What are the implications of genotypic diversity for


pathogenesis?
The natural lifecycle of M. tuberculosis suggests a further
explanation for the apparent discrepancy between the
relative genetic stability of transmitted strains and the
potential intrapatient diversity. Mycobacterium tuberculosis is transmitted in infectious aerosols, which are inhaled
deep into the lung where the bacilli lodge in alveoli and are
engulfed by resident macrophages. Although the precise
details remain to be determined, it is assumed that successful transmission from a prevalent TB case to a new

Technical limitaons and biases

Tissue damage and cavitaon

Sampling

Potenal
sources of
genotoxic
stress

Immune eectors

Oxidave stress
Phagosomal acidicaon

Hypoxia

Only a (small) fracon of bacilli present in the clinical


specimen are culturable and/or selected for sequencing

Nutrient starvaon
Wgs and data analysis

Nitrosave stress

Infecon

Sequencing methodology and/or data analysis

Anbioc treatment

Replicaon within the host

Transmission
Whole-genome sequencing
Diagnosis
Symptoms
Smear microscopy
Culture
Gene Xpert RIF/MTB

How many bacilli are required to


establish an infecon?

How does the anatomical locaon


of a lesion impact genotoxic
stress on the bacilli?

How many bacilli are available for


transmission to a new host?

To what extent are bacilli cultured from a


clinical specimen representave of the
bacillary populaon within the host?

Do bacilli that establish infecon


have specic genotypic and/or
epigenec characteriscs?

How many bacilli are present in a


lesion and how does the lesional
bacillary burden change with
disease progression?

Do transmied bacilli have parcular genotypic and/or epigenec


characteriscs?

To what extent are cultured bacilli


representave of transmied bacilli?

Key quesons

How does the anatomical locaon


of a lesion impact genotypic diversity within the lesion?

TRENDS in Microbiology

Figure 2. Genetic diversification of Mycobacterium tuberculosis within a host: key questions, technical limitations, and biases. Abbreviation: WGS, whole-genome
sequencing.

TIMI-1138; No. of Pages 8

Review
host requires that multiple conditions be met. First, bacilli
must escape in sufficient numbers and in a physiological
state(s) that will ensure transient survival in the environment before inhalation by the new host. Inhaled organisms
must then overcome (or subvert) the barrier defence systems of the host to gain access to the alveoli. Again, the
details are not clear, but it is assumed that at least a single
M. tuberculosis bacillus must then establish infection and,
subsequently, overcome immune defences to replicate and
produce TB disease capable of driving a new infection cycle
[33].
Even in high-burden settings, the probability of TB
infection is relatively rare (approximately 45% per
annum [27]), which, as noted above, reflects the multiple
obstacles at which a potential infection is thwarted. These
inherent bottlenecks are likely to have significant implications for the apparent monomorphism of M. tuberculosis
[10]. Given that the infecting (transmitted) bacillus must
replicate until a population size is reached that is sufficiently large to establish a foothold in the new host, the
founder genotype will necessarily dominate the expanding population. Moreover, the M. tuberculosis lifecycle is
not dependent on achieving maximal bacillary numbers
within a given microenvironment: recent evidence from the
nonhuman primate model indicates that, in immune competent hosts, bacterial populations within individual
lesions consistently achieve a maximum size of approximately 2105 bacilli per lesion before onset of the adaptive
immune response and, following depletion owing to immune-mediated killing, stabilize at approximately 102
bacilli per lesion during active disease [34]. Permissive
lesions that exceed this carrying capacity and spread
locally, or result in TB pneumonia, are rare. Therefore,
the microenvironmental and molecular mechanisms that
might enable small bacillary populations to accumulate
the mutational diversity suggested by inferred in vivo
mutation rates (reviewed in [30]) require elucidation.
The bottlenecks described above imply that any mutations that are generated during host infection are likely to
achieve relatively low frequencies within discrete populations. That is, while numerous SNPs might arise during the
course of infection, a complex interplay of factors will determine whether specific individual mutations ultimately become fixed alleles that are transmitted as distinct strains
and sublineages. Critically, allelic fixation will depend on
the ability of the bacillus to overcome those same barriers as
the infection cycle progresses. However, there are two important exceptions: first, because drug treatment represents
an immediate threat to survival, resistance-conferring
mutations will be rapidly fixed in any bacillary population
exposed to extended therapy. Accordingly, comparative genomics studies are united in identifying drug resistance
polymorphisms regardless of strain diversity or geographic
region [15,35,36]. Paradoxically, selection of resistance
mutations is exacerbated in the presence of functioning
TB control programs, which allow even low-fitness drugresistant strains to outcompete both drug-sensitive and
other, less-fit drug-resistant strains [37,38]. Where strains
acquire compensatory mutations, fixation in the circulating
population is accelerated [39]. As a result of their close
association with drug resistance, compensatory mutations
4

Trends in Microbiology xxx xxxx, Vol. xxx, No. x

constitute the second exception: whether preceding (enabling) or following (modifying) the acquisition of the drug
resistance mutation, this form of epistasis is the subject of
intense research to understand the development and propagation of resistance and, potentially, to identify alternative
counteracting therapies and interventions.
In combination, these observations suggest that the
intrapatient diversity might be greater than expected.
They also imply that the capacity to generate diversity
might be critical to disease progression within individual
hosts, even though the resulting SNPs are not necessarily
broadly selected (transmitted) within a population. Some
evidence to support this hypothesis stems from the observation that MTBC strains are associated with a high
proportion of nonsynonymous SNPs within the 3R genes
involved in DNA replication, repair, and recombination
[40]. Therefore, it is tempting to speculate that the identified polymorphisms result in a relaxation of 3R function
and fidelity that facilitates the rapid generation of genetic
diversity during host infection, perhaps as a strategy to
enable adaptation to allopatric hosts. That is, while M.
tuberculosis maintains a core gene set that enables infection and transmission, it retains the capacity for microdiversity through mutations in other genes. The
epidemiological success of modern strains might indicate
the exploitation of this capacity to develop increased virulence against a background of greater host population
density and comorbidities, such as HIV (Figure 1).
Evidence for a conserved interaction between host and
pathogen
The contention that coevolution might have resulted in a
core M. tuberculosishost interaction is supported by
several observations that derive from independent analyses of both bacillary and host genotypes and functions.
For example, a key study [41] showed that T cell epitopes
are highly conserved across M. tuberculosis lineages,
suggesting that selective pressure acts against sequence
diversity in immunogenic regions. This is reinforced by a
more recent analysis [42] that revealed that sequence
variation in pe_pgrs genes (thought to be involved in
antigenic variation) is restricted to regions that are distinct from the known T cell epitopes, suggesting instead
that another selective pressure drives sequence variation
in these loci.
At a functional level, evidence that macrophage infection triggers a conserved, core mycobacterial transcriptional response [43] (with some scope for lineage-specific
effects) appears to have a corollary in the corresponding
host response, which has elements consistent with both
conserved and lineage-dependent function [4446]. Moreover, the observation that hypervirulent mutants often
contain causal mutations in structural and regulatory
genes [47] perhaps indicates that virulence in M. tuberculosis is under tight control [48]. An additional line of
support is provided by the specific hostpathogen interactions that occur in the different members of the MTBC:
despite close phylogenetic relations, human TB is caused
almost exclusively by M. tuberculosis and Mycobacterium
africanum, with little evidence of zoonotic transmission of
any of the other MTBC members.

TIMI-1138; No. of Pages 8

Review
Genotypephenotype variability in a host-adapted
pathogen
Given the significant bottlenecks to allelic fixation within
the M. tuberculosis population, what factors drive the spread
of SNPs not associated with drug resistance? A recent study
conducted in a low-density setting indicated that there is a
sympatric relation between specific M. tuberculosis strains
and cognate hosts [49], suggesting that host genotypes have
some influence on bacillary diversity. However, in highdensity settings with significant bacterial and host genomic
diversity, there is likely to be less selective pressure: although interstrain competition may be strong, the population of susceptible hosts is large and so able to accommodate
reduced fitness variants, including multidrug-resistant
strains (reviewed in [50]). This effect will be exacerbated
if infection with one bacillary genotype favours re-infection
with another different genotype [14], and could result in an
explosion of genotypic variation, as suggested by recent
evidence of significant strain and lineage diversity within
a well-defined setting in an endemic region [51]. Of course,
frequent bottlenecks imposed by transmission raise the
possibility that chance, not selection, is a major factor
determining circulating genotypes [52], which is again consistent with observation that most SNPs in MTBC occur as
singletons. However, an important consequence is that
elucidating genotypephenotype associations becomes difficult: alleles suggesting convergent evolution across different
lineages offer rare glimpses into functional adaptations [53].
Implications of genotypic diversity: transmission of
hypervirulent strains
The conserved hostpathogen interaction proposed above
assumes that M. tuberculosis is primarily infecting immune competent individuals. As noted elsewhere [54], a
functional adaptive immune response is essential for M.
tuberculosis to complete its lifecycle. When infection and
disease occur against a background of compromised immunity, TB disease manifestation and, therefore, the infection
cycle, are corrupted, consistent with the finding that HIVpositive individuals are poor TB transmitters [55]. The
dependence on an immune-competent host for optimal
transmission also suggests that the ability to cause active
disease (i.e., strain virulence) will directly impact transmissibility. Where a positive correlation exists between
pathogen virulence and transmission, selection acts to
increase virulence and reduce latency to maximize exposure to potential new hosts (reviewed in [13]). For M.
tuberculosis, whose natural evolution has occurred in the
context of increasing human population density [8], the
selective pressure for transmission is likely to be associated with an increase in strain virulence. The inevitable, and
concerning, consequence is that the combination in highburden TB settings of elevated strain diversity, direct
competition between genotypes, and strong drug pressure
will drive the emergence of increasingly virulent drugresistant isolates with low to no short-term fitness costs.
In some respects, the recent expansion of the Beijing family
of strains (a sublineage of Lineage 2) in diverse geographic
settings, and their association with drug resistance and
hypervirulence in animal models, provides a cogent realization of these combined selective pressures [56].

Trends in Microbiology xxx xxxx, Vol. xxx, No. x

What are the processes underlying genome dynamics in


M. tuberculosis?
The observed intrapatient microdiversity implies that the
M. tuberculosis mutation rate might be elevated during
host infection, a possibility that has also been invoked to
explain the emergence of multidrug resistance in the presence of combination therapy (reviewed in [30]). To date,
however, evidence from both animal [57] and human studies [58] suggests that, during active disease, mutations
accumulate at rates that are within the ranges calculated
in vitro. Determining the mutation rate during latent
infection is more complex, as reflected by the fact that
values calculated using M. tuberculosis isolates obtained
from patients presenting with reactivation disease [58]
differ from those predicted in a nonhuman primate model
of infection [57]. Importantly, a recent clinical study estimated that the mutation rate during extended latency
(>20 years) is at least 30 times lower than the rate that
occurs during active disease, a result that is consistent
with separate analyses reporting apparent genetic stasis
in a well-characterized panel of reactivation TB isolates
[59]. While this may indicate that there is little host
pressure on the organism during latent or subclinical
infection, fixation of mutations requires chromosomal replication which in turn, raises important questions regarding the assumed rate at which bacilli divide during host
infection.
Various lines of evidence led to the assumption that the
bacillary population remains stable during chronic TB, and
comprises slow or nonreplicating organisms. However, the
application of a clock plasmid that is lost from daughter
cells during division has established that, during chronic
infection in the mouse model, a stable balance is established between bacillary replication and death [60]. Profound differences in TB pathology, particularly with
respect to the formation of hypoxic microenvironments
within granulomatous lesions [61], mean that extrapolation of findings from the chronic mouse model to humans
must be made with caution. Nonetheless, in addition to
suggesting that the bacilli are under constant immune
surveillance, these observations imply that bacilli may
be replicating at a rate higher than previously thought,
a possibility that is consistent with current models that
propose that a continuum of mycobacterial growth states
prevails during host infection [62]. A compelling mathematical model [63] utilized the possibility of an elevated
bacillary replication rate to demonstrate that the likelihood of emergence of drug resistance before initiation of
anti-TB therapy is higher than previously expected, even
when based on established in vitro mutation rates.
Is there any evidence of mutator strains of M. tuberculosis? For an obligate pathogen, the benefits of a mutator
phenotype for the development of drug resistance are likely
to be outweighed by negative effects on virulence and the
susceptibility of mutators to extinction as a result of bottlenecks. Nevertheless, the existence of strain-specific mutation rates was suggested in a recent study [64] that reported
that M. tuberculosis strains from the East Asian lineage
acquire drug resistance SNPs more rapidly compared
with strains from the Euro-American lineage under the
same conditions in vitro. Importantly, these experiments
5

TIMI-1138; No. of Pages 8

Review
eliminated the possibility that this effect results from an
increased ability to adapt to antibiotic pressure, and instead
indicated that East Asian lineage strains are associated
with an elevated mutation rate in the absence of antibiotic
pressure, although the causative mechanism is unknown.
Linking strain genotypes with disease phenotypes
The complex genotypes associated with drug resistance
[35,36], as well as emerging evidence of the impact of
compensatory mutations on the acquisition and maintenance of resistance alleles [32], highlight the importance of
determining epistatic interactions. For those mutations
that occur in the absence of drug resistance, it is even
more challenging to determine the functional consequences of different mutations: as noted elsewhere [11],
the absence of HGT means that all SNPs in an individual
M. tuberculosis genome are in linkage disequilibrium.
Therefore, while low-level homoplasy means that SNPs
can be usefully applied to measure evolutionary distances
among isolates, determining their impact on bacillary
function and pathogenesis is not trivial. For this reason,
despite the massive increase in sequence data, there is still
a need to obtain additional genomic information for carefully selected panels of clinical M. tuberculosis isolates as
well as related nontuberculous mycobacteria and other
Actinobacteria [9,19,65]. As discussed below, alternative
approaches to sampling bacilli from different microenvironments and anatomical loci will be critical to future
efforts to determine the degree of heterogeneity within
bacillary subpopulations, as well the impact of the pan
genome on bacterial pathogenesis and disease outcome.
A powerful example of the utility of diverse genomes for
comparative analyses was recently provided by the demonstration that SNPs in the two-component regulator,
PhoPR, contribute directly to the reduced virulence and
transmissibility of animal-adapted and M. africanum
strains by reducing the export of virulence factors, such
as the major secreted antigen, ESAT-6, and decreasing the
synthesis of polyacyltrehalose lipids and sulfolipids
[53]. Given recent evidence implicating PhoPR in the
metabolic adaptation of M. tuberculosis to low pH [66], it
is likely that a compromised ability to cope with this
important antimicrobial defence exacerbates the phenotype of phoPR mutants, a possibility that is reinforced by
the observation that the PhoPR regulon also includes the
pH-responsive aprABC locus [67] which is limited to members of the MTBC.
In addition to known drug-resistance alleles [15,35,36],
some nonsynonymous mutations and insertion-deletion
events are likely to be inactivating [68]; however, for most
genomic mutations and rearrangements, predicting the
impact of specific polymorphisms on gene expression, protein function, and strain fitness remains a major challenge
[69,70], and is exacerbated where multiple mutations differentiate the strain of interest from the parental isolate.
Moreover, evidence that synonymous SNPs can influence
function [69,71] suggests that, for many genomic mutations,
inferring the potential impact from sequence data alone is
not possible and will require experimental investigation
by means of allelic exchange mutagenesis [72] as well as
additional multiletter acronym or MLA-seq applications
6

Trends in Microbiology xxx xxxx, Vol. xxx, No. x

(e.g., RNA-seq) that can provide a comprehensive inventory


of the consequences of specific mutations for information
pathway function [6].
Concluding remarks: approaching a systems biology of
TB
Mycobacterium tuberculosis has a 4.4-Mb genome that
harbors evidence of the reductive evolution characteristic
of an obligate pathogen [4,65]; however, the bacillus remains
a formidable prototroph capable of colonizing diverse host
environments and resisting the associated stresses. We
have argued here that at least part of the success of the
organism appears to reside in the stable interaction with its
obligate human host while retaining the capacity to generate phenotypic diversity. Specifically, arming discrete bacillary subpopulations with genotypic variability might enable
a small infecting population to explore a huge fitness
landscape. Although beyond the scope of the current review,
increasing evidence of stochastic behavior [73] as well as
potential for epigenetic modifications, such as DNA methylation, to alter bacillary physiology [74], further supports
the application of novel sampling methods and sequencing
technologies [75] to catalog the full diversity of physiological
states in clinical and experimental TB infection. The recent
use of shotgun metagenomics to detect and characterize
M. tuberculosis in clinical samples [76] might herald the
widespread application of culture-free techniques to this

Box 1. Outstanding questions


Transmission
 When is Mycobacterium tuberculosis transmitted during the
infection cycle?
 How many bacilli are transmitted?
 What is the anatomical and microenvironmental origin of
transmitted bacilli?
Colonization
 What factors determine lineage-specific immune responses?
 What is the impact of the host microbiome on M. tuberculosis
infection?
 How sterile is the M. tuberculosis niche?
 What is the impact of mixed M. tuberculosis infection?
Disease





What is the size of the infecting M. tuberculosis population?


How much diversity is there within individual lesions?
What are the correlates of host specificity?
How do host and pathogen genotypes interact?

Latent and/or subclinical TB infection


 How big is the M. tuberculosis reservoir?
 Does reactivation occur in only approximately 10% of cases
because these are the only individuals who harbor viable bacilli?
 Do bacilli replicate throughout clinical latency?
Microdiversity
 What is the impact of intrapatient diversity on disease progression?
 What determines lineage-specific mutation rates?
 What factors determine strain success?

TIMI-1138; No. of Pages 8

Review
end and, critically, offers one approach to avoid the biases
inherent in strain sampling and propagation.
Understanding the evolutionary processes that have
shaped and enabled the exquisite adaptation of M. tuberculosis may provide clues to biological processes that are
important for pathogenesis and, therefore, potential targets
for novel therapeutics [77]. Similarly, the influence on the
adaptive immune response of exposure to complex circulating genotypes in endemic settings suggests that future
vaccine designs and studies will have to consider the potential impact of strain diversity on efficacy. From a diagnostic
perspective, advances in the application of metabolomics
techniques to analyze sputum for both mycobacterial and
host markers of disease [78] suggest that further refinements will enable the differentiation of major lineages by
analogy with recent reports from Salmonella [79].
Finally, the model presented here is consistent with
emerging evidence from several other bacterial systems
in which the application of advanced genomics techniques
has revealed a similarly unexpected ability of a founding
strain to drive high levels of genotypic and phenotypic
diversification (reviewed in [5]). Further research will be
required to ascertain the implications of diversity for M.
tuberculosis pathogenesis and future interventions (Box 1).
However, there is an urgent need to develop systems
biology approaches to determine the emergent properties
of discrete, genotypically diverse bacterial populations on
the single infected host.
Acknowledgments
We apologize to all those authors whose work was not cited owing to space
limitations. We acknowledge funding from the South African Medical
Research Council (SA MRC), the National Research Foundation of South
Africa, and the Howard Hughes Medical Institute (Senior International
Research Scholars grant to V.M.). Work in our laboratory on TB
transmission is funded by the SA MRC with funds from National
Treasury under the Economic Competitiveness and Support Package
(MRC-RFA-UFSP-01-2013/CCAMP).

References
1 Zumla, A. et al. (2013) WHOs 2013 global report on tuberculosis:
successes, threats, and opportunities. Lancet 382, 17651767
2 Zumla, A. et al. (2013) Advances in the development of new tuberculosis
drugs and treatment regimens. Nat. Rev. Drug Discov. 12, 388404
3 Weiner, J., III and Kaufmann, S.H. (2014) Recent advances towards
tuberculosis control: vaccines and biomarkers. J. Intern. Med. 275,
467480
4 Galagan, J.E. (2014) Genomic insights into tuberculosis. Nat. Rev.
Genet. 15, 307320
5 McAdam, P.R. et al. (2014) High-throughput sequencing for the study of
bacterial pathogen biology. Curr. Opin. Microbiol. 19C, 106113
6 McPherson, J.D. (2014) A defining decade in DNA sequencing. Nat.
Methods 11, 10031005
7 Gagneux, S. (2013) Genetic diversity in Mycobacterium tuberculosis.
Curr. Top. Microbiol. Immunol. 374, 125
8 Comas, I. et al. (2013) Out-of-Africa migration and Neolithic
coexpansion of Mycobacterium tuberculosis with modern humans.
Nat. Genet. 45, 11761182
9 Supply, P. et al. (2013) Genomic analysis of smooth tubercle bacilli
provides insights into ancestry and pathoadaptation of Mycobacterium
tuberculosis. Nat. Genet. 45, 172179
10 Achtman, M. (2012) Insights from genomic comparisons of genetically
monomorphic bacterial pathogens. Philos. Trans. R. Soc. Lond. B: Biol.
Sci. 367, 860867
11 Stucki, D. and Gagneux, S. (2013) Single nucleotide polymorphisms in
Mycobacterium tuberculosis and the need for a curated database.
Tuberculosis 93, 3039

Trends in Microbiology xxx xxxx, Vol. xxx, No. x

12 Hershberg, R. et al. (2008) High functional diversity in Mycobacterium


tuberculosis driven by genetic drift and human demography. PLoS
Biol. 6, e311
13 Gagneux, S. (2012) Host-pathogen coevolution in human tuberculosis.
Philos. Trans. R. Soc. Lond. B: Biol. Sci. 367, 850859
14 Bryant, J.M. et al. (2013) Whole-genome sequencing to establish
relapse or re-infection with Mycobacterium tuberculosis: a
retrospective observational study. Lancet Respir. Med. 1, 786792
15 Casali, N. et al. (2014) Evolution and transmission of drug-resistant
tuberculosis in a Russian population. Nat. Genet. 46, 279286
16 Walker, T.M. et al. (2013) Whole-genome sequencing to delineate
Mycobacterium tuberculosis outbreaks: a retrospective observational
study. Lancet Infect. Dis. 13, 137146
17 Robertson, B.D. et al. (2012) Detection and treatment of subclinical
tuberculosis. Tuberculosis 92, 447452
18 Bos, K.I. et al. (2014) Pre-Columbian mycobacterial genomes reveal seals
as a source of New World human tuberculosis. Nature 514, 494497
19 Boritsch, E.C. et al. (2014) A glimpse into the past and predictions for
the future: the molecular evolution of the tuberculosis agent. Mol.
Microbiol. 93, 835852
20 Namouchi, A. et al. (2012) After the bottleneck: genome-wide
diversification of the Mycobacterium tuberculosis complex by mutation,
recombination, and natural selection. Genome Res. 22, 721734
21 Stinear, T.P. et al. (2004) Giant plasmid-encoded polyketide synthases
produce the macrolide toxin of Mycobacterium ulcerans. Proc. Natl.
Acad. Sci. U.S.A. 101, 13451349
22 Koser, C.U. et al. (2013) Whole-genome sequencing for rapid
susceptibility testing of M. tuberculosis. N. Engl. J. Med. 369, 290292
23 Sun, G. et al. (2012) Dynamic population changes in Mycobacterium
tuberculosis during acquisition and fixation of drug resistance in
patients. J. Infect. Dis. 206, 17241733
24 Perez-Lago, L. et al. (2014) Whole genome sequencing analysis of
intrapatient microevolution in Mycobacterium tuberculosis: potential
impact on the inference of tuberculosis transmission. J. Infect. Dis.
209, 98108
25 Merker, M. et al. (2013) Whole genome sequencing reveals complex
evolution patterns of multidrug-resistant Mycobacterium tuberculosis
Beijing strains in patients. PLoS ONE 8, e82551
26 Hanekom, M. et al. (2013) Population structure of mixed
Mycobacterium tuberculosis infection is strain genotype and culture
medium dependent. PLoS ONE 8, e70178
27 Wood, R. et al. (2011) Tuberculosis control has failed in South Africa:
time to reappraise strategy. S. Afr. Med. J. 101, 111114
28 Kaplan, G. et al. (2003) Mycobacterium tuberculosis growth at the
cavity surface: a microenvironment with failed immunity. Infect.
Immun. 71, 70997108
29 Saunders, N.J. et al. (2011) Deep resequencing of serial sputum isolates
of Mycobacterium tuberculosis during therapeutic failure due to poor
compliance reveals stepwise mutation of key resistance genes on an
otherwise stable genetic background. J. Infect. 62, 212217
30 McGrath, M. et al. (2014) Mutation rate and the emergence of drug
resistance in Mycobacterium tuberculosis. J. Antimicrob. Chemother.
69, 292302
31 Domenech, P. et al. (2014) The origins of a 350-kilobase genomic
duplication in Mycobacterium tuberculosis and its impact on
virulence. Infect. Immun. 82, 29022912
32 Koch, A. et al. (2014) The impact of drug resistance on Mycobacterium
tuberculosis physiology: what can we learn from rifampicin? Emerg.
Microb. Infect. 3, e17
33 OGarra, A. et al. (2013) The immune response in tuberculosis. Annu.
Rev. Immunol. 31, 475527
34 Lin, P.L. et al. (2014) Sterilization of granulomas is common in active
and latent tuberculosis despite within-host variability in bacterial
killing. Nat. Med. 20, 7579
35 Farhat, M.R. et al. (2013) Genomic analysis identifies targets of
convergent positive selection in drug-resistant Mycobacterium
tuberculosis. Nat. Genet. 45, 11831189
36 Zhang, H. et al. (2013) Genome sequencing of 161 Mycobacterium
tuberculosis isolates from China identifies genes and intergenic
regions associated with drug resistance. Nat. Genet. 45, 12551260
37 Blower, S.M. and Chou, T. (2004) Modeling the emergence of the hot
zones: tuberculosis and the amplification dynamics of drug resistance.
Nat. Med. 10, 11111116
7

TIMI-1138; No. of Pages 8

Review
38 Heaton, B.E. et al. (2014) Deficiency of double-strand DNA break repair
does not impair Mycobacterium tuberculosis virulence in multiple
animal models of infection. Infect. Immun. 82, 31773185
39 Comas, I. et al. (2012) Whole-genome sequencing of rifampicin-resistant
Mycobacterium tuberculosis strains identifies compensatory mutations
in RNA polymerase genes. Nat. Genet. 44, 106110
40 Dos Vultos, T. et al. (2008) Evolution and diversity of clonal bacteria:
the paradigm of Mycobacterium tuberculosis. PLoS ONE 3, e1538
41 Comas, I. et al. (2010) Human T cell epitopes of Mycobacterium
tuberculosis are evolutionarily hyperconserved. Nat. Genet. 42,
498503
42 Copin, R. et al. (2014) Sequence diversity in the pe_pgrs genes of
Mycobacterium tuberculosis is independent of human T cell
recognition. MBio 5, e00960-13
43 Homolka, S. et al. (2010) Functional genetic diversity among
Mycobacterium tuberculosis complex clinical isolates: delineation of
conserved core and lineage-specific transcriptomes during intracellular
survival. PLoS Pathog. 6, e1000988
44 Portevin, D. et al. (2011) Human macrophage responses to clinical
isolates from the Mycobacterium tuberculosis complex discriminate
between ancient and modern lineages. PLoS Pathog. 7, e1001307
45 Krishnan, N. et al. (2011) Mycobacterium tuberculosis lineage
influences innate immune response and virulence and is associated
with distinct cell envelope lipid profiles. PLoS ONE 6, e23870
46 Reiling, N. et al. (2013) Clade-specific virulence patterns of
Mycobacterium tuberculosis complex strains in human primary
macrophages and aerogenically infected mice. MBio 4, e00250-13
47 Parish, T. et al. (2003) Deletion of two-component regulatory systems
increases the virulence of Mycobacterium tuberculosis. Infect. Immun.
71, 11341140
48 Warner, D.F. (2014) Mycobacterium tuberculosis metabolism. Cold
Spring Harb. Perspect. Med. Published online November 2014. http://
dx.doi.org/10.1101/cshperspect. a021121
49 Fenner, L. et al. (2013) HIV infection disrupts the sympatric host
pathogen relationship in human tuberculosis. PLoS Genet. 9,
e1003318
50 Brites, D. and Gagneux, S. (2012) Old and new selective pressures on
Mycobacterium tuberculosis. Infect. Genet. Evol. 12, 678685
51 Middelkoop, K. et al. (2014) Factors affecting tuberculosis strain
success over 10 years in a high TB- and HIV-burdened community.
Int. J. Epidemiol. 43, 11141122
52 Comas, I. and Gagneux, S. (2009) The past and future of tuberculosis
research. PLoS Pathog. 5, e1000600
53 Gonzalo-Asensio, J. et al. (2014) Evolutionary history of tuberculosis
shaped by conserved mutations in the PhoPR virulence regulator. Proc.
Natl. Acad. Sci. U.S.A. 111, 1149111496
54 Russell, D.G. (2013) The evolutionary pressures that have molded
Mycobacterium tuberculosis into an infectious adjuvant. Curr. Opin.
Microbiol. 16, 7884
55 Middelkoop, K. et al. (2014) Transmission of tuberculosis in a South
African community with a high prevalence of HIV infection. J. Infect.
Dis. Published online July 22, 2014. http://dx.doi.org/10.1093/infdis/
jiu403
56 Niemann, S. and Supply, P. (2014) Diversity and evolution of
Mycobacterium tuberculosis: moving to whole-genome-based
approaches. Cold Spring Harb. Perspect. Med. Published online
September 4, 2014. http://dx.doi.org/10.1101/cshperspect.a021188
57 Ford, C.B. et al. (2011) Use of whole genome sequencing to estimate the
mutation rate of Mycobacterium tuberculosis during latent infection.
Nat. Genet. 43, 482486
58 Colangeli, R. et al. (2014) Whole genome sequencing of Mycobacterium
tuberculosis reveals slow growth and low mutation rates during latent
infections in humans. PLoS ONE 9, e91024

Trends in Microbiology xxx xxxx, Vol. xxx, No. x

59 Yang, Z. et al. (2011) How dormant is Mycobacterium tuberculosis


during latency? A study integrating genomics and molecular
epidemiology. Infect. Genet. Evol. 11, 11641167
60 Gill, W.P. et al. (2009) A replication clock for Mycobacterium
tuberculosis. Nat. Med. 15, 211214
61 Via, L.E. et al. (2008) Tuberculous granulomas are hypoxic in guinea
pigs, rabbits, and nonhuman primates. Infect. Immun. 76, 23332340
62 Barry, C.E., III et al. (2009) The spectrum of latent tuberculosis:
rethinking the biology and intervention strategies. Nat. Rev.
Microbiol. 7, 845855
63 Colijn, C. et al. (2011) Spontaneous emergence of multiple drug
resistance in tuberculosis before and during therapy. PLoS ONE 6,
e18327
64 Ford, C.B. et al. (2013) Mycobacterium tuberculosis mutation rate
estimates from different lineages predict substantial differences in
the emergence of drug-resistant tuberculosis. Nat. Genet. 45, 784790
65 McGuire, A.M. et al. (2012) Comparative analysis of mycobacterium
and related actinomycetes yields insight into the evolution of
Mycobacterium tuberculosis pathogenesis. BMC Genomics 13, 120
66 Baker, J.J. et al. (2014) Slow growth of Mycobacterium tuberculosis at
acidic pH is regulated by phoPR and host-associated carbon sources.
Mol. Microbiol. 94, 5669
67 Abramovitch, R.B. et al. (2011) aprABC: a Mycobacterium tuberculosis
complex-specific locus that modulates pH-driven adaptation to the
macrophage phagosome. Mol. Microbiol. 80, 678694
68 Tsolaki, A.G. et al. (2004) Functional and evolutionary genomics of
Mycobacterium tuberculosis: insights from genomic deletions in
100 strains. Proc. Natl. Acad. Sci. U.S.A. 101, 48654870
69 Golby, P. et al. (2013) Genome-level analyses of Mycobacterium bovis
lineages reveal the role of SNPs and antisense transcription in
differential gene expression. BMC Genomics 14, 710
70 Rose, G. et al. (2013) Mapping of genotype-phenotype diversity among
clinical isolates of Mycobacterium tuberculosis by sequence-based
transcriptional profiling. Genome Biol. Evol. 5, 18491862
71 Safi, H. et al. (2013) Evolution of high-level ethambutol-resistant
tuberculosis through interacting mutations in decaprenylphosphorylb-D-arabinose biosynthetic and utilization pathway genes. Nat. Genet.
45, 11901197
72 Nebenzahl-Guimaraes, H. et al. (2014) Systematic review of allelic
exchange experiments aimed at identifying mutations that confer drug
resistance in Mycobacterium tuberculosis. J. Antimicrob. Chemother.
69, 331342
73 Wakamoto, Y. et al. (2013) Dynamic persistence of antibiotic-stressed
mycobacteria. Science 339, 9195
74 Shell, S.S. et al. (2013) DNA methylation impacts gene expression and
ensures hypoxic survival of Mycobacterium tuberculosis. PLoS Pathog.
9, e1003419
75 Fang, G. et al. (2012) Genome-wide mapping of methylated adenine
residues in pathogenic Escherichia coli using single-molecule real-time
sequencing. Nat. Biotechnol. 30, 12321239
76 Doughty, E.L. et al. (2014) Culture-independent detection and
characterisation of Mycobacterium tuberculosis and M. africanum in
sputum samples using shotgun metagenomics on a benchtop
sequencer. PeerJ 2, e585
77 Gagneux, S. and Small, P.M. (2007) Global phylogeography of
Mycobacterium tuberculosis and implications for tuberculosis
product development. Lancet Infect. Dis. 7, 328337
78 du Preez, I. and Loots, D.T. (2013) New sputum metabolite markers
implicating adaptations of the host to Mycobacterium tuberculosis, and
vice versa. Tuberculosis 93, 330337
79 Nasstrom, E. et al. (2014) Salmonella Typhi and Salmonella Paratyphi
A elaborate distinct systemic metabolite signatures during enteric
fever. Elife e03100

You might also like