Journal of Microbiology (2018) Vol. 56, No. 5, pp.

300–311 eISSN 1976-3794

DOI 10.1007/s12275-018-8041-3 pISSN 1225-8873

MINIREVIEW

Importance of differential identification of Mycobacterium tuberculosis

strains for understanding differences in their prevalence, treatment
efficacy, and vaccine development

berculosis lineage, genetic diversity, diagnosis, genotyping

Hansong Chae and Sung Jae Shin*
Department of Microbiology, Institute for Immunology and
Immunological Diseases, Brain Korea 21 PLUS Project for Medical Introduction
Science, Yonsei University College of Medicine, Seoul 03722, Republic of
Korea Tuberculosis (TB) is a distressing infectious disease among
humans and remains one of the most important public health
(Received Jan 25, 2018 / Revised Feb 14, 2018 / Accepted Feb 14, 2018) problems worldwide. The World Health Organization re-
ports that 1–3 people die every minute from TB and the risk
Tuberculosis (TB), caused by Mycobacterium tuberculosis of multi-drug resistance (MDR) Mtb infection has increased
(Mtb), remains a serious global health problem in the 21st over time (World Health Organization, 2016). The causa-
century because of its high mortality. Mtb is an extremely tive agents of TB in humans are obligate pathogenic bac-
successful human-adapted pathogen that displays a multi- teria belonging to the Mycobacterium tuberculosis complex
factorial ability to control the host immune response and to (hereafter referred to as the Mtb complex), especially Myco-
evade killing by drugs, resulting in the breakdown of BCG bacterium tuberculosis (hereafter referred to as Mtb, the dis-
vaccine-conferred anti-TB immunity and development of covery of which was reported in 1882 by Robert Koch) (Koch,
multidrug-resistant (MDR) and extensively drug-resistant 1882; World Health Organization, 2016), M. africanum and
(XDR) Mtb. Although genetic components of the genomes M. bovis. Infection with a strain of the Mtb complex can re-
of the Mtb complex strains are highly conserved, showing sult in variable outcomes including clearance via innate im-
over 99% similarity to other bacterial genera, recently accu- munity, latent TB, or active extra-pulmonary or pulmonary
mulated evidence suggests that the genetic diversity of the TB (Coscolla and Gagneux, 2010). The variable disease out-
Mtb complex strains has implications for treatment outcomes, comes of TB have been attributed more to variation in host
development of MDR/XDR Mtb, BCG vaccine efficacy, immune responses and environmental factors than to ge-
transmissibility, and epidemiological outbreaks. Thus, new netic variation among Mtb strains, because of their low ge-
insights into the pathophysiological features of the Mtb com- netic diversity and the absence of conspicuous virulence fac-
plex strains are required for development of novel vaccines tors in comparison to other bacterial genera (Coscolla and
and for control of MDR/XDR Mtb infection, eventually lead- Gagneux, 2010; Gagneux, 2013). Nevertheless, particular
ing to refinement of treatment regimens and the health care strains belonging to specific Mtb complex species and Mtb
system. Many studies have focused on the differential iden- lineages display pronounced differences in their virulence
tification of Mtb complex strains belonging to different lin- phenotypes and pathogenesis, including the emergence of
eages because of differences in their virulence and geogra- drug resistance, breakdown of vaccine efficacy, and trans-
phical dominance. In this review, we discuss the impact of missibility among different ranges of hosts (Coscolla and
differing genetic characteristics among Mtb complex strains Gagneux, 2014). Due to increasing emphasis on the impor-
on vaccine efficacy, treatment outcome, development of MDR/ tance of differential identification of the Mtb complex strains,
XDR Mtb strains, and epidemiological outbreaks by focus- several molecular genotyping methods have recently been
ing on the best-adapted human Mtb lineages. We further introduced in clinical and experimental settings. They have
explore the rationale for differential identification of Mtb led to a much better understanding of the global phylogenetic
strains for more effective control of TB in clinical and labo- diversity of the Mtb complex and the potential factors influ-
ratory settings by scrutinizing current diagnostic methods. encing the different outcomes of TB (Coscolla and Gagneux,
2010, 2014; Couvin and Rastogi, 2015).
Keywords: Mycobacterium tuberculosis, Mycobacterium tu- Different species and strains of the Mtb complex exhibit
distinct responses in experimental studies and have been re-
*For correspondence. E-mail: sjshin@yuhs.ac; Tel.: +82-2-2228-1813; Fax:
ported to be able to affect clinical presentation (Coscolla and
+82-2-392-7088 Gagneux, 2010). Some clinical isolates have been reported
Copyright G2018, The Microbiological Society of Korea to display differences in their virulence depending on their
 Importance of genetic difference among the Mtb lineages 301

lineage (Coscolla and Gagneux, 2010; Couvin and Rastogi, tification of strains of the Mtb complex by summarizing data
2015). Especially, modern lineages such as the Beijing family from recent methods and experimental results. These data
(Lineage 2) and the Euro-American strains (Lineage 4) are support the relevance of studying the Mtb genetic diversity
believed to more virulent compared to the other ancient Mtb for understanding differences in treatment outcome and
lineages such as the East-African-Indian (Lineage 1) and M. vaccine efficacy.
africanum strains (Lineages 5 and 6) (Merker et al., 2015).
In addition, it could be shown that specific features [includ-
ing Bacillus Calmette-Guerin (BCG) vaccination efficacy and Genetic difference among the Mtb lineages
drug resistance] differ for each lineage or strain (Zhang et al.,
2016). However, how the genomic diversity of Mtb influ- In many bacterial pathogens, genetic diversity increases at the
ences TB epidemiology in clinical settings continues to be species or subspecies level through recombination, duplica-
debated. tion, insertion and deletion events. However, gene exchange
In this review, we discuss the potential implications of the in Mtb occurs rarely; therefore, Mtb displays a distinctly clo-
genomic and genetic diversity of Mtb lineages according to nal evolution (Nicol and Wilkinson, 2008). Seven distinct
recent studies conducted on humans and in experimental Mtb lineages have been generally identified by large sequence
settings. We also highlight the importance of differential iden- polymorphisms (LSPs): Lineage 1, Indo-Oceanic or East-

Fig. 1. Schematic diagram illustrating the evolutionary relationship among Mtb strains identified by large sequence polymorphisms [modified from Merker
et al. (2015)]. Genomic regions of difference (RD) can be used to classify human- and animal-adapted strains of the M. tuberculosis complex based on the
presence or absence of deletion sequences. The white boxes indicate which specific deletion sequence is associated with a lineage. M. tuberculosis strains are
grouped into seven major lineages, which are in turn associated with specific geographic regions. However, lineages 2 and 4 are much more geographically
widespread than other lineages.
302 Chae and Shin

Table 1. Genetic markers for differentiating major Mycobacterium tuberculosis lineages

Genetic markers Mycobacterium tuberculosis lineages
and corresponding
classification Lineage 1 Lineage 2 Lineage 3 Lineage 4 Lineage 5 Lineage 6 Lineage 7
TbD1 Ancestral Modern Modern Modern Ancestral Ancestral Intermediate
LSP marker
ΔRD239 ΔRD105 ΔRD750 pks15/1 Δ7bp ΔRD711 ΔRD702 ΔRD239
Brosch et al. (2002)
LSP-based East African-
Indo-Oceanic East Asian Euro-American West-African 1 West-African 2 Ethiopian
classification Indian
Spoligotyping marker T: Δ33-36
Brudey et al. (2006), Δ4-7 and H: Δ31 and 33-36 Δ33-36, Δ4-24, 28 and 29,
Δ29-32 and 34 Δ1-34 Δ7-9 and 39
Yimer et al. (2013), Δ23-34 X: Δ18 and 33-36 39 and 43 Δ4-24 and 26-28
Couvin and Rastogi (2015) LAM: Δ21-24 and 33-36
Classification by East African- Ghana, Haarlem,
Beijing Central Asian M. africanum M. africanum Ethiopian
spoligotyping (Spoligotype) Indian LAM and X etc.
SNP marker Rv3815c: G81A RpoB: T2646G KatG: T1388G
OxyR: C37T
Brosch et al. (2002) Rv0679c: C426G RpoB: C3243T
East Asia,
East Africa, East Africa,
Former Soviet Americas, Europe, Ghana, The Horn of
Geographical association South Asia, North India, Senegal
Union countries, North Africa Cameroon Africa
South India Central Asia
South Africa

African; Lineage 2, East-Asian (includes Beijing family); Line- Lineage 2 and the Euro-American Lineage 4 are frequently
age 3, East-African-Indian or Central Asia strain; Lineage 4, isolated from patients with TB globally, while the prevalence
Euro- American; Lineage 5, West-African or M. africanum of the other Mtb lineages is more regional (Gagneux, 2013;
type 1; Lineage 6, West-African or M. africanum type 2 and Galagan, 2014). The Indo-Oceanic Lineage 1 and the Central-
Lineage 7, Ethiopia (Fig. 1) (Coll et al., 2014). Asian Lineage 3 are restricted to East Africa, Central, and
Of the seven Mtb lineages, Lineage 2, 3, and 4 have been South Asia. M. africanum type 1 and 2 (Lineages 5 and 6 re-
designated as modern lineages and Lineage 1, 5, 6, and 7 spectively) are the most geographically restricted and occur
have been designated as ancient lineages. This designation almost exclusively in West Africa (Hirsh et al., 2004; Gagneux
is based on the presence of the TbD1 sequence in their ge- et al., 2006; Smith et al., 2006). Similarly, the Ethiopian Line-
nome. TbD1-deleted lineages are considered evolutionarily age 7 was discovered recently and is limited to Ethiopia and
modern (Lineages 2, 3, and 4), and lineages where this ge- recent Ethiopian immigrants. Lineages 1–4 comprise 99.2%
nomic region is conserved are considered evolutionarily an- of all cases of Mtb and the remaining percentage of cases is
cient (Lineages 1, 5, 6, and 7; Table 1). The modern Mtb line- caused by other lineages that show restricted geographical
ages form a monophyletic group in contrast to the ancient and host distributions (Coscolla and Gagneux, 2014). The
Mtb lineages (Coll et al., 2014). reasons for why these three lineages (Lineages 5, 6 and 7) are
These human-adapted Mtb lineages are strongly associated highly confined to specific regions are unknown (Coscolla
with particular geographical regions (Hirsh et al., 2004; Perez- and Gagneux, 2014).
Osorio et al., 2012). Some lineages including the East-Asian

Table 2. The major outbreak strains of Mycobacterium tuberculosis

Mtb Lineage Outbreak strain Location (time) Characteristics
Lineage 1 43-16836 Thailand (2010s) Viratyosin et al. (2013)
PR05 Malaysia (2013) Ismail et al. (2013)
Sreevatsan et al. (1997),
GExpression of high lipid
Lineage 2 HN878 (210) Houston, Texas, and New York, USA (1990s) GRapid progression to death in mice Barczak et al. (2005),
Han et al. (2015)
GMost prevalent genotype in South Korea
K Kyunggi province, South Korea (1998) GDrug-susceptible Han et al. (2015)
Agerton et al. (1999),
W New York and other parts of USA (1990s) GStrong propensity for drug resistance Glynn et al. (2002)

CTRI-4 Tomsk, Siberia, Russian Federation (1998–1999)

Lineage 3 CH Leicester, UK (2001)
Rural area near the Kentucky-Tennessee
Lineage 4 CDC1551 border, USA (1994–1996)
Tugela Ferry, KwaZulu-Natal, South Africa
KZN GLargest global outbreak of XDR TB Cohen et al. (2015)
(2005)
F11 Western Cape communities of South Africa Victor et al. (2004)
(1990s)
C Los Angeles, USA (1994–1996) Coscolla and Gagneux (2014)
 Importance of genetic difference among the Mtb lineages 303

Aspects of epidemiology of Mtb lineages jor cause of an outbreak in Texas prisons in 1995 (Table 2)
(Sreevatsan et al., 1997). It is characterized by a more rapid
Differing virulence phenotypes of Mtb strains progression of the disease to death in an immunocompetent
Recent studies have explored how Mtb lineage variation is mouse model in comparison with the clinical strain CDC-
related to virulence, drug resistance, and BCG vaccination 1551 (Lineage 4) and the standard laboratory strain H37Rv
efficacy. Many experimental studies have produced evidence (Lineage 4). It is also characterized by an increased rate of
that clinical isolates of Mtb differ in virulence. Generally, Mtb relapse from the latent state (Barczak et al., 2005; De Jong
virulence is directly linked to its transmission ability in com- et al., 2008; Thwaites et al., 2008). Among the other strains
parison with many other bacteria; delayed immune response in the Beijing family, the K strain caused a large outbreak
is one of the distinct features in Mtb infection (Ernst, 2012). of TB in South Korea in 1998 (Table 2). This strain is also
More specifically, studies modeling infection with different characterized by a severe pathology and a high-level of re-
Mtb lineages show that the modern Mtb Lineages 2, 3, and activation. Moreover, Bacillus-Calmette-Guérin vaccination
4 induce a low level of inflammatory cytokines at early phase is less effective against the W-Beijing family than it is against
of infection (Table 2). They have shown that the expression other Mtb strains (including H37Rv) (Han et al., 2015).
of TNF-α, IFN-γ, IL-12, and IL-17 are important indicators The Mtb CH strain belonging to Lineage 3 was responsible
of a protective immune response against TB, particularly for a large outbreak of TB in Leicester, UK (Table 2). This
against highly virulent Mtb lineages in contrast to ancestral strain showed rapid progression to active infection within
Mtb Lineages 1 and 6 (Portevin et al., 2011). These modern a year and the immune response of the infected macrophages
Mtb lineages are more widespread globally than other line- is characterized by low expression of IL-12p40 and higher
ages. Therefore, many studies have reported that the modern expression of IL-10. The ability to avoid detection by innate
Mtb strains are associated with the high virulence phenotype immune responses due to the deletion of RD750 (Rv1519-
(i.e. they induce a delayed inflammatory response) and it is 1520) is specific to Lineage 3.
one of the factors linked to the global success of these line- As mentioned above, many studies have found that different
ages (Comas and Gagneux, 2011). Indeed, Reiling et al. (2013) Mtb lineages exhibit different virulence profiles and the vir-
have reported that modern Mtb lineages undergo fast rep- ulence is based on factors like bacterial uptake by host cells,
lication in human macrophages and in an aerosol-infected intracellular growth, pathological progression, and cytokine
mouse model. Other studies have also demonstrated induc- induction. We expect that, in agreement with other studies,
tion of lower and more delayed pro-inflammatory response the high virulence Mtb strains are related to the more mo-
upon infection by modern Mtb lineages compared to infec- dern lineages of Mtb and they need to be monitored for ef-
tion by other lineages (Coscolla and Gagneux, 2014). fective TB control.
Some studies have drawn comparisons within the group
of modern Mtb lineages. One study found that Lineage 3 Differences in vaccine efficacy against Mtb strains
exhibited a high anti-inflammatory response compared to There has been no vaccine licensed for or found to be effec-
Lineage 4 (Mihret et al., 2012; Newton et al., 2006), and the tive against TB except for the BCG vaccine. The BCG vac-
Beijing family (that belongs to Lineage 2) commonly induces cine is one of the most widely used vaccines in the world
a low-level immune response (lower expression of cytokines and over 80% of new born children and infants receive BCG
TNF-α, IL-6, IL-10, and GRP-α) compared to Mtb H37Rv vaccination in countries where its immunization policy is
(that belongs to Lineage 4) by macrophages and dendritic introduced (World Health Organization, 2016). The effi-
cells (Wang et al., 2010). In addition, Sarkar et al. (2012) cacy of BCG vaccine against TB can vary from the absence
have reported that Lineage 2 strains show a low pro-inflam- of any protection to 80% fewer cases of TB in different geo-
matory response compared to the other two modern Mtb graphical locations. This variation occurs due to a number
lineages. However, Krishnan et al. (2011) did not find any of factors including BCG vaccine-host interaction, host nu-
differences between Lineage 2 and Lineage 4. These disparate tritional status, environmental status, and genetic diversity
findings with respect to the immune response can be ex- of the Mtb strains (Abebe and Bjune, 2006). However, an
plained by differing experimental conditions and by differ- exact mechanism explaining the variation in BCG vaccine
ences in immune regulation by Mtb lineages that are expected efficacy has not been proposed to date. A putative reason
to occur at the sub-lineage level. Such differences have been for the variation in the protective efficacy of the BCG vac-
most widely studied in Lineage 2. Ribeiro et al. (2014) found cine is the difference in genetic characteristics of the Mtb
that the more modern Beijing strains cause more cytotoxi- strains (Zhang et al., 2016). The use of laboratory-adapted
city and lead to worse gross pathology in C57BL/6 mice. In Mtb strains, including Mtb H37Rv and Mtb Erdman, for
this context, Mtb strains belonging to the W-Beijing family evaluation of vaccine efficacy, was reported to be of limited
are the most widely studied because of their highly patho- success in recent vaccine studies because of differences in
genic phenotype. These strains are endemic in East Asia re- virulence between laboratory-adapted strains and clinical
gion but they have spread to New York, Russia, Latvia and strains (McShane and Williams, 2014; Henao-Tamayo et
South Africa since the late 1990s (Sreevatsan et al., 1997). al., 2015).
Many studies have adduced evidence from animal models In addition, there are two main approaches to the devel-
that the W-Beijing family is more hyper-virulent, charac- opment of a TB vaccine: the development of a whole myco-
terized by greater dissemination, a higher bacillary load in the bacterial priming agent to replace the BCG agent and/or the
sputum, and causes early death. For example of W-Beijing development of a booster agent to be administered after the
family, the HN878 strain belonging to Lineage 2 was a ma- BCG or another vaccine (Hatherill et al., 2016). Notably, the
304 Chae and Shin

Table 3. Characteristics of major Mycobacterium tuberculosis lineages

Characteristics
Lineage a b
In vitro characteristics In vivo and clinical characteristics
GInduce relatively weak levels of Th1-associated
cytokines (IL-1β, IL-6, IL-12p40, except TNF)
GLow uptake of bacteria by human macrophages. GLower bacillary load c
than Beijing family Reiling et al. (2013),
Lineage 1 GLess virulent and more susceptible to oxidative GEvolve into MDR or XDR Mtb strains at the second
stress than Lineage 4 highest rate following the Beijing family (Lineage 2) Nguyen et al. (2016)
GLower growth in human alveolar lung cells than
modern lineages
GLow vaccination efficacy
GInduce relatively weak levels Th1-associated GRapid progression to death in animal models
cytokines GHigh bacillary load and larger granulomatous
GLow uptake of bacteria by human macrophages lesions in lung of infected mice Bifani et al. (2002),
Abebe and Bjune (2006),
Lineage 2 GHigher intracellular growth in human alveolar lung GMore often isolated from the young who have been Reiling et al. (2013),
cells than ancestral lineages vaccinated before
GHigh levels of necrosis, and low levels of apoptosis GEvolve into MDR or XDR strains at the highest rate Ribeiro et al. (2014),
Merker et al. (2015)
in infected macrophages GOutbreaks generally caused by MDR/XDR strains,
especially from the modern Beijing family
GHigher number of cellular infiltrates
Lineage 3 GInduce less IL-12p40 and more IL-10; rapid growth Reiling et al. (2013)
in immune cells.
GSome strains (especially, CDC1551) less virulent
GInduce high levels of Th1-associated cytokines
GHigh uptake of bacteria by human macrophages. than other clinical and laboratory strains, and show
Lineage 4 slow growth in infected animals. Reiling et al. (2013)
GHigh growth in human alveolar lung cells than GKNZ strains known to evolve into MDR and XDR
ancestral lineages
strains
GDepending on the strains, different levels of GOnly confirmed outbreak of MDR strains occurred
cytokines may be induced; produce less IFN-γ than in France and Brazil (1989–1992)
Other lineages other lineages GGenerally, no difference in propensity to evolve into Winglee et al. (2016)
(5, 6, and 7) GLow uptake of bacteria by human macrophages MDR/ XDR strains than other lineages, or lower
GLow growth in human alveolar lung cells than MDR/ XDR evolution in isolates
modern lineages GLineage 5 and 6 strains have decreased virulence
a
Including inflammatory response characteristics and intracellular growth
b
Including pathology and virulence characteristics in animal models, and drug resistance
c
MDR, Multidrug-resistant; XDR, Extensively drug-resistant

MVA85A vaccine given as a single-dose parenteral booster Coscolla and Gagneux, 2014). In addition, the Beijing family
after BCG priming, which was being developed as a new- strains consistently induce accelerated bacterial multiplica-
generation vaccine by Oxford University, showed insignifi- tion and death compared to the M. canettii, Haarlem, and
cant prevention efficacy in infants. One of the reasons be- Somali clades (Reiling et al., 2013).
hind the failure might be the underestimation of Mtb genetic As mentioned above, various diagnostic techniques and
diversity (McShane and Williams, 2014). Especially, as ex- genotyping methods are being commercialized and utilized
emplified by the Mtb Beijing family. As mentioned above, against Mtb. However, most of these simply distinguish my-
the Beijing family is related to one of the seven major Mtb cobacteria from other bacterial genera, or they only classify
lineages and is spread widely in Russia and East Asia, espe- isolates as belonging to the Mtb complex, or as NTM. Tech-
cially in countries neighboring China. Several studies cover- niques like multiplex PCR (both conventional PCR and real-
ing these areas have reported that typical Beijing family time PCR) can be used to further identify the species of the
isolates were more likely to be identified from patients im- isolates, but cannot be used to distinguish the lineage in Mtb.
munized with the BCG vaccine and from areas where the Therefore, a variety of diagnostic methods might yet be de-
BCG vaccine was prescribed (Rook et al., 1981; Abebe and veloped to take advantage of the specific properties of Mtb
Bjune, 2006). Thus, several epidemiological studies have sug- strains, for effective TB control.
gested that extensive BCG vaccination may be one of driv-
ing factors behind the emergence of the Beijing family (Abebe Different proportions of drug resistance among Mtb strains
and Bjune, 2006; Jeon et al., 2008).
The molecular epidemiology of TB can be used to under-
Different lineages of Mtb are characterized by different
stand infection routes and outbreaks of drug-resistant TB.
aspects of infection in animal models. The Beijing family,
Multidrug-resistant (MDR) TB (which comprises 5% of TB
which has been known to avoid BCG-induced immune re-
cases globally) and extensively drug-resistant (XDR)-TB have
sponses, is an example of this. Results showed that infection
been observed in various regions around the world, espe-
by a strain of the Beijing family rapidly induces pneumonia,
cially East Asia, countries of the former Soviet Union, and
and TNF-α and iNOS expression when compared to infec-
South Africa. Generally, development of MDR and XDR
tion by Mtb H37Rv. Furthermore, mice immunized with
Mtb appears to result from the acquisition of genomic mu-
BCG immunization die rapidly and at a higher rate upon
tations (Lanzas et al., 2013). Since drug-resistant (DR) Mtb
such an infection. On the other hand, when infected with
develops upon mutations in individual resistance genes (in-
M. canettii or a Lineage 1 strain, the mice showed limited
cluding rpoA, katG, and rrs) that cause a defect in the bac-
pneumonia and continuous expression of TNF-α, and iNOS
terium’s life cycle, it is known that DR Mtb strains show low
and an almost 100% survival rate (Abebe and Bjune, 2006;
 Importance of genetic difference among the Mtb lineages 305

infectivity (Ford et al., 2013; Coscolla and Gagneux, 2014). veral studies have supported the view that strains of Mtb
However, some groups have reported that infectious DR Lineage 2 and KwaZulu-Natal (KZN) strain have a higher
Mtb strains do exist. It has been suggested that DR Mtb propensity to develop drug-resistance than other Mtb line-
strains are able to spread (Table 3) (Ford et al., 2013; Coscolla ages and strains (Table 3) (Buu et al., 2012; Merker et al.,
and Gagneux, 2014). Finally, though MDR-TB seems to be 2015). Merker et al. (2015) analyzed strains of the Beijing
prevalent in areas with a high number of TB outbreaks, a family found globally for drug resistance. Their data involved
WHO report has shown that MDR-TB occurs only in spe- the calculation of genetic distances among Beijing sub-lin-
cific areas (Galagan, 2014; World Health Organization, 2016). eage genomes. The Beijing sub-lineages CC1 and CC2 ex-
As mentioned above, since each Mtb lineage is almost ex- hibit a higher proportion of drug-resistant strains, and also
clusively associated with a specific region, it was hypothe- a lower pairwise distance than other sub-lineages (Couvin
sized that DR, MDR, and XDR TB strains found in a spe- and Rastogi, 2015; Merker et al., 2015). In addition, the
cific area are related to a specific Mtb lineage. However, the KZN strain belonging to Lineage 4 is known to be respon-
global discovery of such strains has led to the belief that DR sible for an outbreak of XDR-TB in 2005 in the KwaZulu-
strains can be transmitted between people, and patients who Natal region of South Africa. Genomic and proteomic an-
carry DR strains need to be managed more thoroughly. Se- alyses may provide opportunities for the discovery of poten-

(A) (B)

Fig. 2. Various molecular genotyping markers found in the Mycobacterium genome [modified from Comas et al. (2009)]. (1) Insertion sequence (IS) 6110–
Restriction fragment length polymorphisms arise among M. tuberculosis complex strains because of the variable occurrence of a specific transposase gene
that carries a restriction enzyme site (pvuII); (2) Direct repeats (DR) of length 36 base pairs that are interspersed by unique spacer sequences (35–43 base pairs)
are used in spoligotyping methods. Spoligotyping patterns are generally represented by black (signifying the presence of particular spacers) and white
(signifying the absence of particular spacers) boxes; (3) Variable number tandem repeats (VNTR) of mycobacterial interspersed repetitive units (MIRU)
are also employed as unique markers; (4) insertion and deletion of genes (or gene clusters) and (5) single nucleotide polymorphisms (nonsynonymous SNPs
and synonymous SNPs) identified through whole-genome sequencing are also used for differentiating M. tuberculosis strains. Almost all methods for estimating
genetic diversity are based on the use of these five genetic markers.
306 Chae and Shin

Table 4. Evaluation of various molecular genotyping methods

Genotyping method a Principle b Strength Weakness
GLow resolution of Mtb strains
harboring less than five copies of
GUse of IS6110, a mobile element GBetter resolution than the IS6110
IS6110-RFLP found frequently in the Mtb genome spoligotyping method
Thierry et al. (1990), GAnalysis of digestion pattern of GAbility to identify Mtb lineages and GRequirement of large amounts of
good quality genomic DNA (2–3 g),
Kato-Maeda et al. (2011) IS6110 produced by the restriction strains expensive computer software, and
enzyme pvuII
personnel with technical expertise
GLaborious and time-consuming
GRelatively cost-effective and simple
GIdentification of Mtb genotype by GLow resolution compared to
length polymorphisms of the spacer compared to other methods IS6110-RFLP
GHighly reproducible, PCR-based
Spoligotyping regions between direct repeats (DRs) method, requires small amount of GIneffective discrimination of Mtb
Brudey et al. (2006), GAnalysis of binding patterns of strains belonging to the Beijing
probes specific to the spacer regions DNA family
Kato-Maeda et al. (2011) GFree accessibility to the international
present between the GGCCA and spoligotyping database (SpolDB4) GLack of ability to differentiate
GGCCA sequences (DRs) Lineage 7
for referencing
GRequirement of specific equipment
MIRU-VNTR GAnalysis of copy number variation GBetter resolution than both such as a sequencer, electro-capillary
Weniger et al. (2010), of repetitive sequences known as IS6110-RFLP and spoligotyping
MIRU GA simple PCR-based method apparatus, and a specialized in silico
Kato-Maeda et al. (2011) automatic analytical system
GOccurrence of inconsistent results
due individual settings and target
preferences
Deletion analysis GDetection of LSPs, especially GFast and labor-saving GRequirement of confirmatory
identification of irreversible GPCR-based simple method
Coscolla and Gagneux (2014) deletions GMost frequently used method methods
GRelatively less reproducible than
other methods
GHigh frequency of exceptional cases
GThe most advanced and reliable
GIn silico identification of SNPs method
GMore expensive than other methods
WGS present either in the whole genome GUnrivalled resolution GRequirement of specialized software
or in specific genes by comparison GProvision of several adjunctive
Coll et al. (2014) with sequences of reference strains pieces of information such as drug and skilled personnel for data
manipulation
resistance and genetic features of the
targeted Mtb
a
IS6110-RFLP, Insertion sequence 6110-restriction fragment length polymorphism; Spoligotyping, Spacer oligonucleotide typing; MIRU-VNTR, Mycobacterial Interspersed
Repetitive Unit-Variable Number of Tandem Repeat; WGS: Whole genome sequencing.
b
Mtb, Mycobacterium tuberculosis; LSPs, Large sequence polymorphisms.

tial drug targets. In particular, the availability of genomic based on the different number of copies and the different lo-
data of MDR and XDR Mtb strains (include the KZN strain) cations of IS6110 insertion in the whole genome. Thus, clas-
might make it possible to find a clue regarding the origin of sification based IS6110 is performed by digestion with pvuII
drug resistance in bacteria that may lead to the discovery of followed by electrophoretic separation and hybridization
new drug targets. with an IS6110-derived probe (Thierry et al., 1990; Comas
et al., 2009). This method is highly informative when there
are over five IS6110 copies in the whole genome, and it is
Differential identification of Mtb strains and used to determine whether isolates collected during the same
genotyping methods outbreak come from the same strain or not (Comas et al.,
2009; Kato-Maeda et al., 2011; Sinha et al., 2016). However,
Genomic variations IS6110-RFLP requires a large amount of good quality geno-
Mobile elements and copy number repeats mic DNA (about 2–3 g per reaction), and this assay is also
time-consuming and difficult to reproduce because it re-
A) Insertion Sequence (IS) 6110 quires experienced personnel with high technical expertise.
In the Mtb complex, IS6110 is a mobile genetic element Due to these limitations, PCR-based assays have been devel-
that has been utilized as the gold standard for genotyping oped for genotyping Mtb complex strains that require only
since the 1990s (Fig. 2 and Table 4) (Thierry et al., 1990). a small amount of DNA (Tortoli et al., 2012; Yimer et al.,
IS6110 can be present in multiple copies in various parts of 2013; Chae et al., 2017). Additionally, IS6110-RFLP cannot
the whole genome of an Mtb complex strain. The number be performed for some strains that have less than five IS6110
of copies of IS6110 varies from strain to strain because IS is copies. Furthermore, the lack of reliability of this assay pre-
highly unstable. However, experimental evidence shows that vents the comparison of genotyping results among labo-
although IS6110 is an unstable element, transmission events ratories (Thierry et al., 1990; Warren et al., 2002).
involving it are very rare in mycobacteria (Warren et al.,
2002; Comas et al., 2009). Based on this characteristic, many B) Spacer oligonucleotide typing (spoligotyping)
studies have established that IS6110-restriction fragment Spoligotyping is a PCR-based technique for differentiating
length polymorphisms (IS6110-RFLP) can be used to classify M. tuberculosis strains (Fig. 2 and Table 4). It is based on the
Mtb complex strains. RFLP classification using IS6110 is existence of unique spacer regions between direct repeats,
 Importance of genetic difference among the Mtb lineages 307

which are known as the ‘clustered regularly interspaced short are modern lineage and Lineages 1, 5, and 6 are ancient lin-
palindromic repeats’, in the Mtb genome (Fig. 2) (Groenen eage based on the presence of TbD1 (Brosch et al., 2002;
et al., 1993). During spoligotyping, a primer pair amplifies Achtman, 2008).
the spacer present in the direct repeat locus and the products Besides RDs, there are several new genetic markers known
of the amplification are covalently hybridized to oligonu- under the category of LSPs. For example, the insertion of the
cleotide probes derived from direct repeat spacers specific IS6110 gene into a specific region of the genome can lead
to a strain (Brudey et al., 2006; Oelemann et al., 2007). The to the development of an LSP. In 1991, B0/W 148 strains were
“spoligo-type” pattern so generated is compared to patterns isolated in New York; the characteristic of these strains is
of spoligo-type found in the international spoligotyping data- that the IS6110 gene is inserted into the noise transfer func-
base (SpolDB4) that includes thousands of spoligo-type pat- tion (NTF) region of these strains. Within the Mtb Beijing
terns (Brudey et al., 2006). This assay was originally developed family, the strains whose NTF region is found to carry the
to identify the route of infection by an Mtb complex strain. IS6110 gene are classified as the modern type, whereas the
However, this genotyping method is of limited utility when other strains are classified as the ancient type (Bifani et al.,
drawing a phylogenetic tree and for accurate strain discrim- 2002; Tsolaki et al., 2005; Merker et al., 2015). In addition,
ination (particularly, discrimination of the Lineage 2/Beijing our group recently found a unique gene cluster enabling to
family) because many spoligo-types share similar patterns identify the Beijing family among the Mtb complex using
(especially Lineage 3/ CAS family) (Devi et al., 2015). whole genome sequencing. Since this region was highly con-
served within the Beijing family, we developed a multiplex
C) Mycobacterial Interspersed Repetitive Unit-Variable Num- PCR assay for distinguishing the Beijing family not only from
ber Tandem Repeat (MIRU-VNTR) other major nontuberculous mycobacteria but also from
The first whole genome sequence of Mtb was finished in 1998 other Mtb complex (Chae et al., 2017).
(Cole et al., 1998). Afterwards, comparative genome analysis
was performed on several Mtb genomes leading to the disco- Single nucleotide polymorphisms (SNPs)
very of VNTRs known as MIRUs in the Mtb complex (Fig. LSPs can serve as robust markers for phylogenetic classifi-
2 and Table 4). These were soon utilized as a new method cation. However, LSPs are not different enough to allow dis-
for molecular genotyping of Mtb. An MIRU is a repeating tinction between closely related Mtb strains (Galagan, 2014).
unit composed of dozens of base pairs and is used to iden- On the other hand, SNPs encompass all types of mutations
tify strains by variation in its number (Fig. 2) (Weniger et al., in the whole genome and provide much better resolution
2010; Couvin and Rastogi, 2015). These days, it has become than LSPs (Fig. 2 and Table 4) (Coll et al., 2014). For exam-
a preferred genotyping technique over spoligotyping in many ple, as mention above, the Beijing family is associated with
countries. Strain genotyping using VNTR has also developed a unique spoligo-type pattern and the unique presence of
rapidly as an alternative to IS6110-RFLP because it is a PCR- RD 207. However, the use of these markers still seems to carry
based assay and thus needs only a small amount of DNA. many disadvantages, since it relies on the unstable insertion
This simple process also provides high speed and accuracy and deletion of elements. Therefore, a mutation in the rv0679c
(Weniger et al., 2010). In addition, this technique lends it- locus allows distinguishing Lineage 2 in Mtb from other
self to digitization and has allowed the development of an lineages which supports the evidence of LSPs marker with
online database system called VNTR plus (Oelemann et al., the reported advantages of SNPs (Nakajima et al., 2013). The
2007). However, this technique possesses less discrimination discovery of new SNPs by whole-genome sequencing ap-
ability than IS6110-RFLP, because different VNTR loci are proaches has the potential to produce more extensive insights
used by each country and laboratory. This means that the into the molecular and epidemiological mechanisms of Mtb
Mtb complex strains found in each country are identified transmission, as well as allow the development of better di-
by different VNTR characteristics. Therefore, further experi- agnostic tools and therapies (Niemann and Supply, 2014).
ments with the appropriate combination of VNTR loci are In addition, SNPs analysis can allow the calculation of ge-
needed to establish the standard VNTR method (Kato-Maeda netic distances between Mtb strains and lead to the inference
et al., 2011). of their overall phylogeny, and the widespread occurrence
of SNPs in the Mtb complex genome has been used for pro-
Large Sequence Polymorphisms (LSPs) ducing a phylogenetic classification of the Mtb complex (Coll
The analysis results of Mtb genome sequencing led to the dis- et al., 2014). Real-time PCR and sequencing are known tech-
covery of LSPs (including region of difference; RD) (Comas niques for SNP detection. These methods are important to
et al., 2009). Even though horizontal gene transmission events understand the current state of the Mtb complex, as well as
have been detected in other mycobacteria species, the genetic to help predict the direction of its evolution. Coscolla and
components among the Mtb complex is highly conserved with Gagneux (2014) have reported that the strains belonging to
99.9% similarity at the nucleotide level (Galagan, 2014). Since the modern Lineage 2, 3, and 4 differ by 970 SNPs on aver-
horizontal gene transfer rarely occurs in the Mtb complex, age. Strains belonging to the ancient Lineage 1, 5, and 6 are
LSPs could be suitable phylogenetic markers for lineage and more distantly related, with an average distance of 1500 SNPs
sub-lineage classification. For example, RDs in mycobacteria between them. This information can be used to construct a
have been used as major genetic markers to classify the strains phylogenetic tree and identify which Mtb strains have led to
of the Mtb complex into main lineages and sub-lineages (Fig. recent outbreaks. In addition, because horizontal gene trans-
2 and Table 4) (Hirsh et al., 2004; Mostowy et al., 2004a, fer among Mtb complex strains is rare, SNPs can serve as a
2004b). In additions, representatively Lineages 2, 3, and 4 robust marker for identifying Mtb strains according to their
308 Chae and Shin

phylogenetic classification. The publication of whole genome says are higher than that of other commercial methods. On
sequencing (WGS) results by Rose et al. (2013) led to the average, a sensitivity of 75.5% is obtained with conventional
identification of all SNPs specific to the different lineages PCR and a sensitivity of 77.3% is obtained with real-time
of the Mtb complex. The number of SNPs ranges from 124 PCR (not a significant difference). The specificity of both
in Lineage 2 (a modern Mtb lineage) to 698 in Lineage 5 (an methods with sputum samples is 99.5%. However, cerebro-
ancient Mtb lineage). In all cases, over 40% of the non-syn- spinal fluid, pleural fluid, and urine samples show a lower
onymous SNPs were fixed within their respective Mtb line- sensitivity because of inhibitory substances (< 50%) (Sar-
ages and therefore they can be predicted to be involved in miento et al., 2003; Tortoli et al., 2012).
gene function. These data reveal that although the general GenXpert Mtb/RIF is a cartridge-based nucleotide amplifi-
genomic diversity of Mtb is lower than other bacterial gen- cation assay and simultaneously detects Mtb and rifampicin
era (Achtman, 2008), a lot of SNPs have accumulated within (RIF) resistance. Since 2010, this assay has been used in TB
each Mtb lineage and they are likely to contribute to pheno- endemic countries (including China and countries in Africa)
typic differences among them. The most common SNPs in (Detjen et al., 2015). This assay is based on the amplification
Mtb are mutations associated with drug resistance and vir- of Mtb and RIF specific DNA sequences by PCR; genomic
ulence (Perez-Osorio et al., 2012; Casali et al., 2014). Due to DNA from the sputum samples of patients is first purified
these mutations, certain gene products cannot be expressed, and concentrated and then clinically relevant RIF resistance-
leading to an interruption in the metabolism of certain an- inducing mutations (located in the RNA polymerase beta
tibiotics and attenuation of virulence (Coll et al., 2014). gene) are identified. The assay requires only 90 min for ob-
The katG and phoPR genes provide particularly illustrative taining results and it reduces biohazard risk and technical
examples. First, the Ser375Thr mutation in katG confers re- training time (Detjen et al., 2015).
sistance to isoniazid (Mokrousov et al., 2002). katG encodes The Line probe assay (LiPA) is similarly able to identify
a catalase-peroxidase that activates the isoniazid pro-drug. an Mtb complex strain and detect genetic mutations in the
However, the Ser375Thr mutation prevents activation of iso- katG gene region, which is related to isoniazid (INH)-resis-
niazid, resulting in resistance to it. Second, phoPR, which tance, and the rpoB gene region, which is related to RIF-
regulates a two-component virulence system, is essential for resistance. This assay is based on the use of several nitro-
Mtb complex virulence including the regulation of secretion cellulose paper strips layered with specific 10-oligonucleotide
of ESAT-6, acyltrehalose-based lipids, and the modulation probes. After the katG and rpoB regions are amplified using
of antigen export (Coscolla and Gagneux, 2014). SNPs in PCR, the PCR products are hybridized with the oligonucleo-
this gene have been reported to lead to critical Mtb pheno- tide probes present on the paper strips. The results are deter-
type changes. If an amino acid changes at position 219 of mined by the absence or presence of various colored lines
phoP in the Mtb H37Rv strain, its virulence is remarkably (Morgan et al., 2005).
attenuated and it is then referred as the Mtb H37Ra strain.
(Coll et al., 2014; Coscolla and Gagneux, 2014). B) Enzyme-linked immunosorbent assay (ELISA)-based me-
thods
Genotyping and diagnosis methods Interferon-γ release assay (IGRAs) can detect latent TB in-
Commercialized assays fection based on the cellular immune response. After plac-
ing a sample of whole-blood in a tube coated with the pep-
The molecular diagnosis of TB directly from clinical isolates tides including ESAT-6, CFP-10, and TB7.7, interferon-γ
has been performed since the early 1990s. These methods released by T cells is measured. The assay has better sensi-
compensate for defects in microscopy and culture-based me- tivity and specificity than the tuberculin skin test and shows
thods, the major disadvantages of which are poor sensitivity no reaction with samples of BCG-vaccinated individuals
3
(requiring 10 CFU/ml within sputum) and a narrow detec- (Lalvani and Pareek, 2010; Starke, 2014).
tion range (Sarmiento et al., 2003). Many studies have intro- For diagnosis of TB in HIV-infected patients, recent studies
duced various genotyping methods and molecular diagnostic have described the use of the lipoarabinomannan (LAM)-
assays based on several different technologies. ELISA assay. LAM is a glycolipid found in the Mtb cell wall
component and serves as a major virulence factor. Since
A) PCR-based methods this assay employs urine-based testing for identification of
One of the first commercialized assays was performed us- LAM in the urine of patients, samples for it are easier to col-
ing multiplex PCR (both conventional PCR and real-time lect and store than samples for sputum-based testing. The
PCR), which was based on the amplification of housekeeping sensitivity of LAM-ELISA assay falls between 23–84%, and
genes (e.g. 16S rRNA genes common to all mycobacteria) its specificity falls between 75–99%. On average, the speci-
and Mtb-specific insertion sequences (e.g. IS6110 or RD se- ficity of this assay is higher than its sensitivity. A previous
quences) (Tevere et al., 1996). Conventional multiplex PCR study tried to combine urine and sputum samples for per-
is based on the simultaneous use of multiple specific primers forming the LAM assay. It reported a poorer specificity (15–
and real-time multiplex PCR use primers with fluorescent 21%) but higher sensitivity (86–90%) compared to LAM as-
probes. Unlike microscopy-based tests, this assay can clas- says performed on urine-only samples (Dheda et al., 2010).
sify species and subspecies belonging to the Mtb complex Although LAM is still considered unnecessary for diagnos-
and the NTM group (Kim et al., 2013; Mutingwende et al., ing TB patients in a clinical setting, its utility has gradually
2015; Chae et al., 2017). advanced for diagnosing TB in HIV-infected patients.
In general, sensitivity and specificity of multiplex PCR as-
 Importance of genetic difference among the Mtb lineages 309

Conclusions and perspectives types of Mycobacterium tuberculosis and low-level protection

by bacille Calmette-Guérin (BCG) vaccines: Is there a link? Clin.
During the past decades, the importance of understanding Exp. Immunol. 145, 389–397.
the genetic diversity of Mtb lineages has been underscored Achtman, M. 2008. Evolution, population structure, and phylogeo-
graphy of genetically monomorphic bacterial pathogens. Annu.
by differences in their virulence and geographic dominance. Rev. Microbiol. 62, 53–70.
It is plausible that with the development of more efficient Agerton, T.B., Valway, S.E., Blinkhorn, R.J., Shilkret, K.L., Reves, R.,
molecular techniques and genetic tools, our knowledge of Schluter, W.W., Gore, B., Pozsik, C.J., Plikaytis, B.B., Woodley,
the genetic diversity of the Mtb complex strains and func- C., et al. 1999. Spread of strain w, a highly drug-resistant strain
tions of the specific components of their genomes will ex- of Mycobacterium tuberculosis, across the United States. Clin.
pand. Unsurprisingly, human genetic variation is also known Infect. Dis. 29, 85–92; discussion 93–85.
to play an important role in the spread of TB since the suc- Barczak, A.K., Domenech, P., Boshoff, H.I., Reed, M.B., Manca, C.,
cess of an Mtb infection is reciprocally influenced by genetic Kaplan, G., and Barry, C.E. 3rd. 2005. In vivo phenotypic domi-
variations of the host and the pathogen. Accumulating evi- nance in mouse mixed infections with Mycobacterium tuberculosis
clinical isolates. J. Infect. Dis. 192, 600–606.
dence suggests that pathogenic, phenotypic, and genotypic Bifani, P.J., Mathema, B., Kurepina, N.E., and Kreiswirth, B.N. 2002.
variation among Mtb strains contributes to differences in Global dissemination of the Mycobacterium tuberculosis W-bei-
vaccine efficacy, treatment outcomes, and the development jing family strains. Trends Microbiol. 10, 45–52.
of MDR/XDR Mtb strains. Thus, a better understanding of Brosch, R., Gordon, S.V., Marmiesse, M., Brodin, P., Buchrieser, C.,
the genetic diversity of Mtb isolates found in clinical settings Eiglmeier, K., Garnier, T., Gutierrez, C., Hewinson, G., Kremer, K.,
is needed. In addition, since most TB patients undergo latent et al. 2002. A new evolutionary scenario for the Mycobacterium
infection prior to getting the active disease, further research tuberculosis complex. Proc. Natl. Acad. Sci. USA 99, 3684–3689.
is needed to find out the genotypes and lineages of Mtb which Brudey, K., Driscoll, J.R., Rigouts, L., Prodinger, W.M., Gori, A., Al-
reactivate most frequently from the latent state; such research Hajoj, S.A., Allix, C., Aristimuno, L., Arora, J., Baumanis, V., et al.
2006. Mycobacterium tuberculosis complex genetic diversity: Mining
is important for controlling the spread of TB. Recently, se-
the fourth international spoligotyping database (SpolDB4) for
veral molecular genotyping methods have been introduced classification, population genetics and epidemiology. BMC Micro-
in experimental and clinical settings to identify specific geno- biol. 6, 23.
types of the Mtb complex. Although differential identifica- Buu, T.N., van Soolingen, D., Huyen, M.N., Lan, N.T., Quy, H.T.,
tion of Mtb genotypes and lineages has not been conducted Tiemersma, E.W., Kremer, K., Borgdorff, M.W., and Cobelens,
routinely in clinical settings, some studies have reported on F.G. 2012. Increased transmission of Mycobacterium tuberculosis
the association between the phylogenetic age of Mtb lineages beijing genotype strains associated with resistance to strepto-
and differences in their epidemiology. Epidemiological pat- mycin: A population-based study. PLoS One 7, e42323.
terns show that the more modern lineages of Mtb are more Casali, N., Nikolayevskyy, V., Balabanova, Y., Harris, S.R., Ignatyeva,
O., Kontsevaya, I., Corander, J., Bryant, J., Parkhill, J., Nejentsev,
virulent and exhibit more successful transmission globally.
S., et al. 2014. Evolution and transmission of drug-resistant tu-
According to multiple criteria (immune response, virulence, berculosis in a Russian population. Nat. Genet. 46, 279–286.
drug resistance and BCG vaccine efficacy) Lineage 2 and Chae, H., Han, S.J., Kim, S.Y., Ki, C.S., Huh, H.J., Yong, D., Koh, W.J.,
Lineage 4 are more virulent phenotype than Lineage 1 and and Shin, S.J. 2017. Development of a one-step multiplex PCR
M. africanum (Lineage 6). Other lineages (Lineage 3, Line- assay for differential detection of major Mycobacterium species.
age 5, and Lineage 7) need to undergo more studies. J. Clin. Microbiol. 55, 2736–2751.
A better understanding of the phenotypic and pathogenic Cohen, K.A., Abeel, T., Manson McGuire, A., Desjardins, C.A., Mun-
variations arising from the genetic diversity of Mtb strains samy, V., Shea, T.P., Walker, B.J., Bantubani, N., Almeida, D.V.,
will facilitate the development of effective prevention and Alvarado, L., et al. 2015. Evolution of extensively drug-resistant
treatment strategies, including the development of new drugs tuberculosis over four decades: Whole genome sequencing and
dating analysis of Mycobacterium tuberculosis isolates from Kwa-
and more effective vaccines against Mtb infection. Further- Zulu-Natal. PLoS Med. 12, e1001880.
more, the identification of genetic factors encoding virulence Cole, S.T., Brosch, R., Parkhill, J., Garnier, T., Churcher, C., Harris,
in clinically significant strains is necessary for differentiating D., Gordon, S.V., Eiglmeier, K., Gas, S., Barry, C.E. 3rd, et al. 1998.
Mtb strains and for improving public health measures. Deciphering the biology of Mycobacterium tuberculosis from the
complete genome sequence. Nature 393, 537–544.
Coll, F., McNerney, R., Guerra-Assuncao, J.A., Glynn, J.R., Perdigao,
Acknowledgements J., Viveiros, M., Portugal, I., Pain, A., Martin, N., and Clark, T.G.
2014. A robust SNP barcode for typing Mycobacterium tuber-
This research was supported by a grant from Basic Science culosis complex strains. Nat. Commun. 5, 4812.
Comas, I. and Gagneux, S. 2011. A role for systems epidemiology in
Research Program (NRF-2016R1A2A1A05005263) and by tuberculosis research. Trends Microbiol. 19, 492–500.
the Korea Mouse Phenotyping Project (NRF-2017M3A9D- Comas, I., Homolka, S., Niemann, S., and Gagneux, S. 2009. Geno-
5A01052446) through the National Research Foundation typing of genetically monomorphic bacteria: DNA sequencing
of Korea (NRF) funded by the Ministry of science, ICT & in Mycobacterium tuberculosis highlights the limitations of cur-
Future Planning. rent methodologies. PLoS One 4, e7815.
Coscolla, M. and Gagneux, S. 2010. Does M. tuberculosis genomic
diversity explain disease diversity? Drug Discov. Today Dis. Mech.
References 7, e43–e59.
Coscolla, M. and Gagneux, S. 2014. Consequences of genomic diver-
Abebe, F. and Bjune, G. 2006. The emergence of Beijing family geno- sity in Mycobacterium tuberculosis. Semin. Immunol. 26, 431–444.
Couvin, D. and Rastogi, N. 2015. Tuberculosis - a global emergency:
310 Chae and Shin

Tools and methods to monitor, understand, and control the epi- Jeon, B.Y., Derrick, S.C., Lim, J., Kolibab, K., Dheenadhayalan, V.,
demic with specific example of the beijing lineage. Tuberculosis Yang, A.L., Kreiswirth, B., and Morris, S.L. 2008. Mycobacterium
(Edinb.). 95 Suppl 1, S177–S189. bovis BCG immunization induces protective immunity against
De Jong, B.C., Hill, P.C., Aiken, A., Awine, T., Antonio, M., Adetifa, nine different Mycobacterium tuberculosis strains in mice. Infect.
I.M., Jackson-Sillah, D.J., Fox, A., Deriemer, K., Gagneux, S., et al. Immun. 76, 5173–5180.
2008. Progression to active tuberculosis, but not transmission, Kato-Maeda, M., Metcalfe, J.Z., and Flores, L. 2011. Genotyping of
varies by Mycobacterium tuberculosis lineage in the gambia. J. Mycobacterium tuberculosis: Application in epidemiologic studies.
Infect. Dis. 198, 1037–1043. Future Microbiol. 6, 203–216.
Detjen, A.K., DiNardo, A.R., Leyden, J., Steingart, K.R., Menzies, D., Kim, Y., Choi, Y., Jeon, B.Y., Jin, H., Cho, S.N., and Lee, H. 2013. A
Schiller, I., Dendukuri, N., and Mandalakas, A.M. 2015. Xpert simple and efficient multiplex PCR assay for the identification
MTB/Rif assay for the diagnosis of pulmonary tuberculosis in of Mycobacterium genus and Mycobacterium tuberculosis com-
children: A systematic review and meta-analysis. Lancet Respir. plex to the species level. Yonsei Med. J. 54, 1220–1226.
Med. 3, 451–461. Koch, R. 1882. Die Ätiologie der Tuberkulose. Berliner Klin Wo-
Devi, K.R., Bhutia, R., Bhowmick, S., Mukherjee, K., Mahanta, J., and chenschr 19, 221–230.
Narain, K. 2015. Genetic diversity of Mycobacterium tuberculosis Krishnan, N., Malaga, W., Constant, P., Caws, M., Tran, T.H., Sal-
isolates from Assam, India: Dominance of beijing family and mons, J., Nguyen, T.N., Nguyen, D.B., Daffe, M., Young, D.B.,
discovery of two new clades related to CAS1_Delhi and EAI et al. 2011. Mycobacterium tuberculosis lineage influences innate
family based on spoligotyping and MIRU-VNTR typing. PLoS immune response and virulence and is associated with distinct
One 10, e0145860. cell envelope lipid profiles. PLoS One 6, e23870.
Dheda, K., Davids, V., Lenders, L., Roberts, T., Meldau, R., Ling, D., Lalvani, A. and Pareek, M. 2010. Interferon gamma release assays:
Brunet, L., van Zyl Smit, R., Peter, J., Green, C., et al. 2010. Clinical Principles and practice. Enferm. Infecc. Microbiol. Clin. 28, 245–
utility of a commercial LAM-ELISA assay for TB diagnosis in 252.
HIV-infected patients using urine and sputum samples. PLoS Lanzas, F., Karakousis, P.C., Sacchettini, J.C., and Ioerger, T.R. 2013.
One 5, e9848. Multidrug-resistant tuberculosis in panama is driven by clonal
Ernst, J.D. 2012. The immunological life cycle of tuberculosis. Nat. expansion of a multidrug-resistant Mycobacterium tuberculosis
Rev. Immunol. 12, 581–591. strain related to the KZN extensively drug-resistant M. tuber-
Ford, C.B., Shah, R.R., Maeda, M.K., Gagneux, S., Murray, M.B., culosis strain from South Africa. J. Clin. Microbiol. 51, 3277–3285.
Cohen, T., Johnston, J.C., Gardy, J., Lipsitch, M., and Fortune, S.M. McShane, H. and Williams, A. 2014. A review of preclinical animal
2013. Mycobacterium tuberculosis mutation rate estimates from models utilised for TB vaccine evaluation in the context of recent
different lineages predict substantial differences in the emergence human efficacy data. Tuberculosis (Edinb). 94, 105–110.
of drug-resistant tuberculosis. Nat. Genet. 45, 784–790. Merker, M., Blin, C., Mona, S., Duforet-Frebourg, N., Lecher, S.,
Gagneux, S. 2013. Genetic diversity in Mycobacterium tuberculosis. Willery, E., Blum, M.G., Rusch-Gerdes, S., Mokrousov, I., Aleksic,
Curr. Top. Microbiol. Immunol. 374, 1–25. E., et al. 2015. Evolutionary history and global spread of the My-
Gagneux, S., DeRiemer, K., Van, T., Kato-Maeda, M., de Jong, B.C., cobacterium tuberculosis Beijing lineage. Nat. Genet. 47, 242–249.
Narayanan, S., Nicol, M., Niemann, S., Kremer, K., Gutierrez, Mihret, A., Bekele, Y., Loxton, A.G., Aseffa, A., Howe, R., and Walzl,
M.C., et al. 2006. Variable host-pathogen compatibility in Myco- G. 2012. Plasma level of IL-4 differs in patients infected with
bacterium tuberculosis. Proc. Natl. Acad. Sci. USA 103, 2869–2873. different modern lineages of M. tuberculosis. J. Trop. Med. 2012,
Galagan, J.E. 2014. Genomic insights into tuberculosis. Nat. Rev. 518564.
Genet. 15, 307–320. Mokrousov, I., Otten, T., Filipenko, M., Vyazovaya, A., Chrapov, E.,
Glynn, J.R., Whiteley, J., Bifani, P.J., Kremer, K., and van Soolingen, Limeschenko, E., Steklova, L., Vyshnevskiy, B., and Narvskaya,
D. 2002. Worldwide occurrence of Beijing/W strains of Myco- O. 2002. Detection of isoniazid-resistant Mycobacterium tuber-
bacterium tuberculosis: A systematic review. Emerg. Infect. Dis. culosis strains by a multiplex allele-specific PCR assay targeting
8, 843–849. katG codon 315 variation. J. Clin. Microbiol. 40, 2509–2512.
Groenen, P.M., Bunschoten, A.E., van Soolingen, D., and van Emb- Morgan, M., Kalantri, S., Flores, L., and Pai, M. 2005. A commercial
den, J.D. 1993. Nature of DNA polymorphism in the direct re- line probe assay for the rapid detection of rifampicin resistance
peat cluster of Mycobacterium tuberculosis; application for strain in Mycobacterium tuberculosis: A systematic review and meta-
differentiation by a novel typing method. Mol. Microbiol. 10, analysis. BMC Infect. Dis. 5, 62.
1057–1065. Mostowy, S., Cousins, D., and Behr, M.A. 2004a. Genomic interro-
Han, S.J., Song, T., Cho, Y.J., Kim, J.S., Choi, S.Y., Bang, H.E., Chun, gation of the dassie bacillus reveals it as a unique RD1 mutant
J., Bai, G.H., Cho, S.N., and Shin, S.J. 2015. Complete genome within the Mycobacterium tuberculosis complex. J. Bacteriol. 186,
sequence of Mycobacterium tuberculosis K from a korean high 104–109.
school outbreak, belonging to the Beijing family. Stand. Genomic Mostowy, S., Onipede, A., Gagneux, S., Niemann, S., Kremer, K.,
Sci. 10, 78. Desmond, E.P., Kato-Maeda, M., and Behr, M. 2004b. Genomic
Hatherill, M., Tait, D., and McShane, H. 2016. Clinical testing of analysis distinguishes Mycobacterium africanum. J. Clin. Micro-
tuberculosis vaccine candidates. Microbiol. Spectr. 4, 5. biol. 42, 3594–3599.
Henao-Tamayo, M., Shanley, C.A., Verma, D., Zilavy, A., Stapleton, Mutingwende, I., Vermeulen, U., Steyn, F., Viljoen, H., and Grobler,
M.C., Furney, S.K., Podell, B., and Orme, I.M. 2015. The efficacy A. 2015. Development and evaluation of a rapid multiplex-PCR
of the BCG vaccine against newly emerging clinical strains of based system for Mycobacterium tuberculosis diagnosis using
Mycobacterium tuberculosis. PLoS One 10, e0136500. sputum samples. J. Microbiol. Methods 116, 37–43.
Hirsh, A.E., Tsolaki, A.G., DeRiemer, K., Feldman, M.W., and Small, Nakajima, C., Tamaru, A., Rahim, Z., Poudel, A., Maharjan, B., Khin
P.M. 2004. Stable association between strains of Mycobacterium Saw, A., Ling, H., Hattori, T., Iwamoto, T., Fukushima, Y., et al.
tuberculosis and their human host populations. Proc. Natl. Acad. 2013. Simple multiplex PCR assay for identification of Beijing
Sci. USA 101, 4871–4876. family Mycobacterium tuberculosis isolates with a lineage-specific
Ismail, A., Teh, L.K., Ngeow, Y.F., Norazmi, M.N., Zainul, Z.F., Tang, mutation in Rv0679c. J. Clin. Microbiol. 51, 2025–2032.
T.H., Najimudin, N., and Salleh, M.Z. 2013. Draft genome se- Newton, S.M., Smith, R.J., Wilkinson, K.A., Nicol, M.P., Garton, N.J.,
quence of a clinical isolate of Mycobacterium tuberculosis strain Staples, K.J., Stewart, G.R., Wain, J.R., Martineau, A.R., Fandrich,
PR05. Genome Announc. 1, 3. S., et al. 2006. A deletion defining a common asian lineage of
 Importance of genetic difference among the Mtb lineages 311

Mycobacterium tuberculosis associates with immune subversion. Acad. Sci. USA 94, 9869–9874.
Proc. Natl. Acad. Sci. USA 103, 15594–15598. Starke, J.R. 2014. Interferon-gamma release assays for diagnosis
Nguyen, V.A., Banuls, A.L., Tran, T.H., Pham, K.L., Nguyen, T.S., of tuberculosis infection and disease in children. Pediatr. 134,
Nguyen, H.V., Nguyen, N.L., Nguyen, N.L., Dang, D.A., Marks, e1763–e1773.
G.B., et al. 2016. Mycobacterium tuberculosis lineages and anti- Tevere, V.J., Hewitt, P.L., Dare, A., Hocknell, P., Keen, A., Spadoro,
tuberculosis drug resistance in reference hospitals across Viet J.P., and Young, K.K. 1996. Detection of Mycobacterium tuber-
Nam. BMC Microbiol. 16, 167. culosis by PCR amplification with pan-Mycobacterium primers
Nicol, M.P. and Wilkinson, R.J. 2008. The clinical consequences and hybridization to an M. tuberculosis-specific probe. J. Clin.
of strain diversity in Mycobacterium tuberculosis. Trans. R. Soc. Microbiol. 34, 918–923.
Trop. Med. Hyg. 102, 955–965. Thierry, D., Brisson-Noel, A., Vincent-Levy-Frebault, V., Nguyen,
Niemann, S. and Supply, P. 2014. Diversity and evolution of Myco- S., Guesdon, J.L., and Gicquel, B. 1990. Characterization of a
bacterium tuberculosis: Moving to whole-genome-based appro- Mycobacterium tuberculosis insertion sequence, IS6110, and its
aches. Cold Spring Harb. Perspect. Med. 4, a021188. application in diagnosis. J. Clin. Microbiol. 28, 2668–2673.
Oelemann, M.C., Diel, R., Vatin, V., Haas, W., Rusch-Gerdes, S., Thwaites, G., Caws, M., Chau, T.T., D'Sa, A., Lan, N.T., Huyen, M.N.,
Locht, C., Niemann, S., and Supply, P. 2007. Assessment of an Gagneux, S., Anh, P.T., Tho, D.Q., Torok, E., et al. 2008. Rela-
optimized mycobacterial interspersed repetitive-unit-variable- tionship between Mycobacterium tuberculosis genotype and the
number tandem-repeat typing system combined with spoligo- clinical phenotype of pulmonary and meningeal tuberculosis. J.
typing for population-based molecular epidemiology studies of Clin. Microbiol. 46, 1363–1368.
tuberculosis. J. Clin. Microbiol. 45, 691–697. Tortoli, E., Urbano, P., Marcelli, F., Simonetti, T.M., and Cirillo,
Perez-Osorio, A.C., Boyle, D.S., Ingham, Z.K., Ostash, A., Gautom, D.M. 2012. Is real-time PCR better than conventional PCR for
R.K., Colombel, C., Houze, Y., and Leader, B.T. 2012. Rapid Mycobacterium tuberculosis complex detection in clinical sam-
identification of mycobacteria and drug-resistant Mycobacterium ples? J. Clin. Microbiol. 50, 2810–2813.
tuberculosis by use of a single multiplex PCR and DNA sequen- Tsolaki, A.G., Gagneux, S., Pym, A.S., Goguet de la Salmoniere, Y.O.,
cing. J. Clin. Microbiol. 50, 326–336. Kreiswirth, B.N., Van Soolingen, D., and Small, P.M. 2005. Geno-
Portevin, D., Gagneux, S., Comas, I., and Young, D. 2011. Human mic deletions classify the Beijing/W strains as a distinct genetic
macrophage responses to clinical isolates from the Mycobacte- lineage of Mycobacterium tuberculosis. J. Clin. Microbiol. 43,
rium tuberculosis complex discriminate between ancient and 3185–3191.
modern lineages. PLoS Pathog. 7, e1001307. Victor, T.C., de Haas, P.E., Jordaan, A.M., van der Spuy, G.D., Ri-
Reiling, N., Homolka, S., Walter, K., Brandenburg, J., Niwinski, L., chardson, M., van Soolingen, D., van Helden, P.D., and Warren,
Ernst, M., Herzmann, C., Lange, C., Diel, R., Ehlers, S., et al. 2013. R. 2004. Molecular characteristics and global spread of Mycobac-
Clade-specific virulence patterns of Mycobacterium tuberculosis terium tuberculosis with a western cape F11 genotype. J. Clin.
complex strains in human primary macrophages and aerogeni- Microbiol. 42, 769–772.
cally infected mice. mBio 4, e00250–13. Viratyosin, W., Kulawonganunchai, S., Smittipat, N., Juthayothin,
Ribeiro, S.C., Gomes, L.L., Amaral, E.P., Andrade, M.R., Almeida, T., Penpassakarn, P., Pasomsub, E., Chantratita, W., Chaiprasert,
F.M., Rezende, A.L., Lanes, V.R., Carvalho, E.C., Suffys, P.N., A., and Palittapongarnpim, P. 2013. Draft genome sequence of the
Mokrousov, I., et al. 2014. Mycobacterium tuberculosis strains of Mycobacterium tuberculosis strain 43-16836, belonging to the
the modern sublineage of the Beijing family are more likely to Indo-Oceanic lineage, isolated from tuberculous meningitis in
display increased virulence than strains of the ancient sublineage. Thailand. Genome Announc. 1, e00801-13.
J. Clin. Microbiol. 52, 2615–2624. Wang, C., Peyron, P., Mestre, O., Kaplan, G., van Soolingen, D., Gao,
Rook, G.A., Bahr, G.M., and Stanford, J.L. 1981. The effect of two Q., Gicquel, B., and Neyrolles, O. 2010. Innate immune response
distinct forms of cell-mediated response to mycobacteria on to Mycobacterium tuberculosis Beijing and other genotypes. PLoS
the protective efficacy of BCG. Tubercle. 62, 63–68. One 5, e13594.
Rose, G., Cortes, T., Comas, I., Coscolla, M., Gagneux, S., and Young, Warren, R.M., van der Spuy, G.D., Richardson, M., Beyers, N., Borg-
D.B. 2013. Mapping of genotype-phenotype diversity among cli- dorff, M.W., Behr, M.A., and van Helden, P.D. 2002. Calculation
nical isolates of Mycobacterium tuberculosis by sequence-based of the stability of the IS6110 banding pattern in patients with
transcriptional profiling. Genome Biol. Evol. 5, 1849–1862. persistent Mycobacterium tuberculosis disease. J. Clin. Microbiol.
Sarkar, R., Lenders, L., Wilkinson, K.A., Wilkinson, R.J., and Nicol, 40, 1705–1708.
M.P. 2012. Modern lineages of Mycobacterium tuberculosis ex- Weniger, T., Krawczyk, J., Supply, P., Niemann, S., and Harmsen, D.
hibit lineage-specific patterns of growth and cytokine induction 2010. MIRU-VNTRplus: A web tool for polyphasic genotyping
in human monocyte-derived macrophages. PLoS One 7, e43170. of Mycobacterium tuberculosis complex bacteria. Nucleic Acids
Sarmiento, O.L., Weigle, K.A., Alexander, J., Weber, D.J., and Miller, Res. 38, W326–W331.
W.C. 2003. Assessment by meta-analysis of PCR for diagnosis Winglee, K., Manson McGuire, A., Maiga, M., Abeel, T., Shea, T.,
of smear-negative pulmonary tuberculosis. J. Clin. Microbiol. 41, Desjardins, C.A., Diarra, B., Baya, B., Sanogo, M., Diallo, S., et al.
3233–3240. 2016. Whole genome sequencing of Mycobacterium africanum
Sinha, P., Gupta, A., Prakash, P., Anupurba, S., Tripathi, R., and Sri- strains from Mali provides insights into the mechanisms of geo-
vastava, G.N. 2016. Differentiation of Mycobacterium tuberculosis graphic restriction. PLoS Negl. Trop. Dis. 10, e0004332.
complex from non-tubercular mycobacteria by nested multiplex World Health Organization. 2016. Global tuberculosis report 2016.
PCR targeting IS6110, MTP40 and 32kD alpha antigen encoding Yimer, S.A., Hailu, E., Derese, Y., Bjune, G.A., and Holm-Hansen, C.
gene fragments. BMC Infect. Dis. 16, 123. 2013. Spoligotyping of Mycobacterium tuberculosis isolates among
Smith, N.H., Kremer, K., Inwald, J., Dale, J., Driscoll, J.R., Gordon, pulmonary tuberculosis patients in Amhara Region, Ethiopia.
S.V., van Soolingen, D., Hewinson, R.G., and Smith, J.M. 2006. Acta Pathol. Microbiol. Immunol. Scand. 121, 878–885.
Ecotypes of the Mycobacterium tuberculosis complex. J. Theor. Zhang, L., Ru, H.W., Chen, F.Z., Jin, C.Y., Sun, R.F., Fan, X.Y., Guo,
Biol. 239, 220–225. M., Mai, J.T., Xu, W.X., Lin, Q.X., et al. 2016. Variable virulence
Sreevatsan, S., Pan, X., Stockbauer, K.E., Connell, N.D., Kreiswirth, and efficacy of BCG vaccine strains in mice and correlation with
B.N., Whittam, T.S., and Musser, J.M. 1997. Restricted structural genome polymorphisms. Mol. Ther. 24, 398–405.
gene polymorphism in the Mycobacterium tuberculosis complex
indicates evolutionarily recent global dissemination. Proc. Natl.