Mem Inst Oswaldo Cruz, Rio de Janeiro, Vol.

105(6): 779-785, September 2010 779

Genotyping of Mycobacterium tuberculosis isolates from a

low-endemic setting in northwestern state of Paraná in southern Brazil
Erika Noda Noguti1, Clarice Queico Fujimura Leite2, Ana Carolina Malaspina2,
Adolfo Carlos Barreto Santos2, Rosário Dominguez Crespo Hirata3, Mario Hiroyuki Hirata3,
Elsa Massae Mamizuka3, Rosilene Fressatti Cardoso1/+
1
Departamento de Análises Clínicas, Universidade Estadual de Maringá, Av. Colombo 5790 bloco J90 sala 5, 78020-900 Maringá,
PR, Brasil 2Faculdade de Ciências Farmacêuticas, Universidade Estadual de São Paulo, Araraquara, SP, Brasil
3
Faculdade de Ciências Farmacêuticas, Universidade de São Paulo, São Paulo, SP, Brasil

The purpose of this study was to provide information about the genetic diversity and prevalent genotype of My-
cobacterium tuberculosis in a low-endemic setting in northwestern state of Paraná in southern Brazil. We employed
spoligotyping and mycobacterial interspersed repetitive units-variable number tandem repeat (MIRU-VNTR) tech-
niques to genotype M. tuberculosis isolates from patients with pulmonary tuberculosis (TB). The 93 isolates ana-
lyzed by spoligotyping were divided into 36 different patterns, 30 of which were described in the SITVIT database.
Latin American and Mediterranean, Haarlem and T families were responsible for 26.9%, 17.2% and 11.8% of TB
cases, respectively. From the 84 isolates analyzed by MIRU-VNTR, 58 shared a unique pattern and the remaining 26
belonged to nine clusters. The MIRU loci 40, 23, 10 and 16 were the most discriminatory. A combination of MIRU-
VNTR and spoligotyping resulted in 85.7% discriminatory power (Hunter-Gaston index = 0.995). Thus, combining
spoligotyping and MIRU-VNTR typing proved to be most useful for epidemiological study in this low-endemic setting
in southern Brazil. The current study demonstrated that there is significant diversity in circulating strains in the city
of Maringá and the surrounding regions, with no single genotype of M. tuberculosis predominating.

Key words: Mycobacterium tuberculosis - tuberculosis - molecular epidemiology - spoligotyping - MIRU

Tuberculosis (TB) is the leading cause of death by an isolates of the same strains. Furthermore, the typing
infectious agent and thus an important public health prob- methods employed must be reproducible, discriminatory
lem worldwide, with an estimated 9.27 million new cases and easy to perform.
in 2007. In Brazil, the estimated incidence rate is 48 cases Some polymerase chain reaction (PCR)-based tech-
per 100,000 inhabitants (WHO 2009). Brazil has five dis- niques are now being used to differentiate Mycobacte-
tinct geographic regions, among which the distribution rium tuberculosis isolates. Currently, genotyping ap-
of TB cases varies greatly. In the state of Paraná (PR) in proaches targeting the analysis of the variable number of
southern Brazil, the average rate of TB is 27.52 cases per tandem repeats (VNTR), based on mycobacterial inter-
100,000 inhabitants (Malaghini et al. 2009). Maringá and spersed repetitive units (MIRU), are the most promising.
the small cities surrounding it form a low-endemic area This technique is based on the variability found at 12
of TB in PR, with an incidence rate of 22.62 cases per specific loci interspersed throughout the mycobacterial
100,000 in 2007 (personal communication). genome (Supply et al. 2000). Recently, MIRU-VNTR
Genotyping methods have been applied extensively genotyping strategies using 15 or 24 loci (Supply et al.
in the epidemiological study of TB worldwide (Candia et 2006, Oelemann et al. 2007a, Alonso-Rodríguez et al.
al. 2007, Sharma et al. 2008, Valcheva et al. 2008). These 2008) were evaluated and applied to molecular epide-
studies have been based on the assumption that patients miological typing in mycobacteria.
with genotypically clustered strains are epidemiologi- Spoligotyping is the second most widely used meth-
cally linked and represent recent transmissions. In con- od for M. tuberculosis complex genotype after IS6110-
trast, patients infected with different types of strains are based fingerprinting (Prodinger 2007). It is based on the
not considered indicative of recent transmission. Genetic presence or absence of a set of target sequences in the
markers need to be both sufficiently polymorphic to dis- direct repeat (DR) locus in the M. tuberculosis complex
tinguish unrelated strains and stable enough to identify genome (Kamerbeek et al. 1997) that in combination
with MIRU has been used to replace typing via restric-
tion fragment length polymorphisms (RFLP) based on
the insertion sequence IS6110. The IS6110 method has
been considered the “gold standard” for genotyping M.
tuberculosis since 1993, but it is an expensive, labori-
Financial support: CAPES
ous and lengthy methodology that requires weeks of M.
+ Corresponding author: [email protected] tuberculosis culturing and specific software to analyze
Received 11 January 2010 the RFLP band-patterns, all of which make it difficult
Accepted 13 May 2010 to interpret and exchange data. In addition, the IS6110

online | memorias.ioc.fiocruz.br
780 Genotyping of M. tuberculosis isolates from Brazil • Erika Noda Noguti et al.

method has limited use in genotyping M. tuberculosis were detected by chemiluminescence after incubation
isolates that contain few copies of IS6110. with a streptavidin-peroxidase conjugate (Boehringer
The aim of our study was to use spoligotyping and Ingelheim, Germany) and assessed by an enhanced
MIRU-VNTR typing to provide preliminary informa- chemiluminescence system (GE Healthcare UK Limit-
tion about the genetic diversity and prevalent genotype ed, Buckinghamshire, UK). Spoligotypes were reported
of M. tuberculosis isolates in a low-endemic setting in using an octal code in which the 43-digit binary repre-
northwestern PR. senting the 43 spacers (“1” is hybridization and “0” is
PATIENTS, MATERIALS AND METHODS
no hybridization) was divided into 14 sets of three digits
(spacers 1-42) plus one additional digit (spacer 43). Each
Study isolates - From November 2005-June 2008, a three-digit set was converted to octal code (000 = 0, 001
total of 93 M. tuberculosis isolates were obtained from = 1, 010 = 2, 011 = 3, 100 = 4, 101 = 5, 110 = 6 and 111 =
sputum of patients suspected of TB who were seen at the 7), with the final digit remaining either 1 or 0, yielding a
Clinical Bacteriology Laboratory, Department of Clini- 15-digit octal designation (Dale et al. 2001).
cal Analysis of the State University of Maringá. This lab
is a reference TB laboratory that handles patients from MIRU-VNTR typing - MIRU-VNTR typing was
Maringá and other cities in northwestern PR. All isolates performed in only 84 of the 93 M. tuberculosis isolates
were cultured in DifcoTM Lowenstein Medium Base (Bec- at the Laboratory of Mycobacteria Dr. Hugo David in
ton, Dickinson and Company, Sparks, MD, USA) and the School of Pharmaceutical Science, São Paulo State
identified as M. tuberculosis using the conventional bio- University, Araraquara, São Paulo (SP), Brazil. MIRU-
chemical tests (Kent & Kubica 1985) and molecular biol- VNTR was not performed in nine of the isolates because
ogy (van Embden et al. 1993). The following retrospective there was insufficient DNA for analysis.
demographic and epidemiological data were collected for The isolates were genotyped by PCR amplification
all patients through a review of a national TB notification of the original 12 MIRU-VNTR loci (2, 4, 10, 16, 20, 23,
database (National Diseases Notification System): city 24, 26, 27, 31, 39, 40) as described by Supply et al. (2000)
and zip code of residence at the time of diagnosis, age, and Mazars et al. (2001) in a PTC-100 thermal cycler (MJ
sex, ethnicity, HIV status, sample susceptibility profile, Research, Ramsey, Minnesota, USA). Each locus was
alcoholism and the occurrence of other diseases. amplified individually from 2 µL of mycobacterial DNA
(20 ng) in 23 µL of a reaction mixture containing 0.4 µM
DNA extraction - DNA from M. tuberculosis was of loci-respective primers and PCR Master Mix (Pro-
extracted from a subculture on DifcoTM Lowenstein Me- mega Corporation, Madison, Wisconsin, USA) according
dium Base, as described by González-y-Merchand et al. to the manufacturer’s instructions. The PCR conditions
(1996), with minor modifications. Briefly, a loopful of for each set of primers have been described elsewhere
bacteria was suspended in 6 M guanidine hydrochloride (Pandolfi 2006). PCR products were subjected to electro-
(Sigma Chemical Co, St. Louis, MO, USA) and bacilli phoresis in a 2% weight/volume agarose gel (Invitrogen
were lyzed by freezing at -20°C for 30 min followed by Life Technologies, SP, Brazil). Both 50 and 100-bp DNA
heating at 100ºC for 10 min. This procedure was repeat- Ladders (Invitrogen Life Technologies, SP, Brazil) were
ed twice. DNA was further extracted with two volumes used as molecular markers. The gels were stained with
of phenol-chloroform-isoamyl alcohol (25:24:1, v/v) ethidium bromide and visualized under ultraviolet light,
followed by two steps of extractions with chloroform- then photodocumented with an Alpha-imager 2200 (Al-
isoamyl alcohol (24:1, v/v). DNA was purified by etha- pha Innotech Corporation, San Leandro, CA, USA). PCR
nol precipitation, dissolved in 50 µL of Tris-ethylenedi- fragment size was determined by visual comparison with
amine tetraacetic acid (TE buffer), pH 8.0, and stored at the molecular markers and the MIRU allele scoring was
-20ºC until use. DNA concentration was determined by determined according to Mazars et al. (2001) and Supply
ultraviolet spectrophotometry. et al. (2001). The results from each of the 12 loci were
Spoligotyping - Spoligotyping was performed in all combined to create the 12-digit allelic profiles.
93 M. tuberculosis isolates, using the standard method, Interpretation of genotyping results - The observed
to detect the presence or absence of 43 spacers (Mo- spoligotypes were compared to the international data-
lhuizem et al. 1998). Briefly, the DR region was ampli- base, SITVIT, which is an updated version of the pub-
fied from 1 µL of mycobacterial DNA (10 ng) in 24 µL lished SpolDB4.0 database (Brudey et al. 2006) and it is
of a reaction mixture containing 0.4 µM of each primer available at pasteur-guadeloupe.fr:8081/SITVITDemo/.
and PCR Master Mix (Promega Corporation, Madison, For previously unreported spoligopatterns in the Spol-
WI, USA) according to manufacturer’s instructions. The DB4.0, the “Spotclust” database was used (Vitol et al.
primer sequences were Dra, 5’-GGTTTTGGGTCT- 2006). This model takes into account knowledge of the
GACGAC-3’ (biotinylated 5’ end) and DRb, 5’-CCGA- evolution of the DR region and assigns spoligopatterns
GAGGGGACGGAAAC-3’ (Integrated DNA Technolo- to families and subfamilies, using a computer algorithm
gies, Inc Coralville, USA). The DNA amplification was based on studies of SpolDB3 (cgi2.cs.rpi.edu/~bennek/
carried out in a TC-512 thermal cycler (Techne, UK). SPOTCLUST.html). BioNumerics software (version
PCR products were hybridized with a set of 43 spacer 4.45; Applied Maths, Sint-Martens-Latem, Belgium) was
oligonucleotides covalently linked to the spoligo-mem- used for analysis of spoligotyping and MIRU-VNTR pat-
brane (Isogen Life Sciences, The Netherlands) accord- terns. Dendrograms were constructed for spoligotyping,
ing to the manufacturer’s instructions. Bound fragments MIRU-VNTR and the combination of both methodolo-
 Mem Inst Oswaldo Cruz, Rio de Janeiro, Vol. 105(6), September 2010 781

gies. The genetic distance was built using the unweight-

ed pair group method with arithmetic mean (UPGMA)
algorithm (Sneath & Sokal 1973). The evaluation of the
discriminative power of each typing method alone as
well as the two in combinations was undertaken using
the Hunter-Gaston index (HGI) (Hunter & Gaston 1988).
HGI is based on the probability that two unrelated strains
sampled from the population will be placed into differ-
ent typing groups. Allelic diversity of each locus’ MIRU-
VNTR was classified as “highly discriminant” (HGI >
0.6), “moderately discriminant” (0.3 ≤ HGI ≤ 0.6) and
“poorly discriminant” (HGI < 0.3) (Sola et al. 2003).
RESULTS
The study population belonged to eight cities (Ma-
ringá and seven other small, neighboring cities) and con-
sisted of 78 males (83.9%) and 15 females (16.1%). The
mean patient age was 40 years. The results of HIV-test-
ing were available for 43 patients (46.2%) and, among
these, four (9.3%) tested positive for HIV. Data regard-
ing the use of alcohol and the presence of other diseases
were not obtained.
Spoligotyping - Of the 93 M. tuberculosis isolates, 36
different spoligotyping patterns were observed. Based
on these spoligotypes, distinct families of TB were iden-
tified: Haarlem (H), Latin American and Mediterranean
(LAM), undesignated (U), T family (modern TB strains),
East-African Indian (EAI) and S lineage (Figure).
The spoligopatterns of 83 isolates (89.2%) were clas-
sified according to the SITVIT database and, of these,
65 isolates were clustered into 12 Shared International
Types (ST), with each cluster containing between three
and 13 isolates. The remaining 18 isolates showed unique
STs. The main STs found in this study were ST46 (n = 13, combined numerical analysis of 84 Mycobacterium tuberculosis
15.7%), ST20 (n = 6, 7.2%), ST42 (n = 6, 7.2%) and ST47 clinical isolates from Maringá and other cities in the northwest of
the state of Paraná, Brazil, using spoligotyping and mycobacterial
(n = 6, 7.2%).
interspersed repetitive units-variable number of tandem repeats
Of the 83 M. tuberculosis isolates identified in the (MIRU-VNTR) typing. The dendrogram was obtained using the
SITVIT database, 72 were classified into families and BioNumerics. CAM: ���������������������������������������
Cameroun�������������������������������
; H: Haarlem; LAM: Latin Ameri-
sublineages according to SpolDB4. The number of these can and Mediterranean; S: S lineage; Snd: spoligotype not de-
72 isolates that ranked in the LAM, U, H and T families scribed; st: Shared International Types; T: modern tuberculosis
were, respectively, 25 (34.7%), 17 (23.6%), 16 (22.2%) strains; U: undesignated.

TABLE I
Spoligopatterns and family assignment of 10 isolates of Mycobacterium tuberculosis not described in SITVIT database

Isolates with identical pattern

Spoligotype Spotclust ID Spotclust probability n

SND1 607777677560731 T1 0.99 1

SND2 777777607700171 LAM9 0.99 4
SND3 777777207760631 LAM9 0.99 1
SND4 577777607560771 LAM9 0.99 2
SND5 774337606560771 LAM9 0.99 1
SND6 777757777760471 T1 0.99 1

ID: identification; LAM: Latin American and Mediterranean; SND: spoligotype not described; T: modern tuberculosis strains.
782 Genotyping of M. tuberculosis isolates from Brazil • Erika Noda Noguti et al.

and 11 (15.3%). The minor families observed in our study Allele polymorphism analysis of 12 MIRU loci re-
were EAI (n = 1, 1.4%), S lineage (n = 1, 1.4%) and H1-S (n vealed that MIRU locus 40 was the most discriminatory
= 1, 1.4%). Eleven M. tuberculosis isolates were not clas- locus with eight alleles, followed by MIRU loci 23, 10
sified into families according to SpolDB4, but did have and 16. In MIRU locus 40, the presence of three alleles
attributed ST numbers according to the SITVIT database was most frequent, followed by a single copy of an al-
(ST2508, ST2512, ST2525, ST2563 and ST2654). lele. The MIRU 20, 26 and 31 loci were moderately dis-
Ten isolates (10.8%), comprising six spoligotyping criminant. Other loci were less polymorphic, with only
patterns not found in the SITVIT database, were as- two alleles for MIRU loci 4, 20 and 39 and one allele for
signed to families and subfamilies by “SpotClust”. The MIRU locus 24, for which a single copy was present in
family assignment of these remaining isolates revealed all 84 M. tuberculosis isolates analyzed (Table II).
that 80% belonged to the LAM family and 20% belonged
Combining spoligotyping and MIRU-VNTR typing -
to the T family (Table I). These six new spoligotypes
Among the 84 M. tuberculosis isolates analyzed by spoligo-
have been submitted to the SITVIT database (pasteur-
typing and MIRU-VNTR typing, 72 (85.7%) distinct geno-
guadeloupe.fr:8081/SITVITDemo/).
types were obtained, resulting in eight clusters with 100%
MIRU-VNTR typing - MIRU-VNTR typing was con- similarity. By considering similarity indices of at least 69%,
ducted for 84 clinical isolates and a total of 67 distinct we observed three distinct clonal groups that represented
MIRU patterns were obtained. Of the 84 isolates, 58 (69%) 92.9% of the isolates analyzed (Figure). Table III summa-
were orphans and the remaining 26 (31%) were grouped rizes the discriminative power of each typing method alone
into nine clusters, each comprising 2-6 isolates (Figure). and of the two methods combined, according to the HGI.

TABLE II
Allelic polymorphism of 12 mycobacterial interspersed repetitive units (MIRU) loci from the 84 Mycobacterium tuberculosis
isolates from patients with tuberculosis in Maringá and region in the northwest of the state of Paraná, Brazil

Allele number

MIRU number 0 1 2 3 4 5 6 7 8 HGI Conclusion

2 - 1 10 72 1 - - - - 0.254 Poorly discriminant

4 - - 82 2 - - - - - 0.047 Poorly discriminant
10 - - 6 14 40 23 1 - - 0.673 Highly discriminant
16 1 9 9 45 20 - - - - 0.641 Highly discriminant
20 - 21 63 - - - - - - 0.380 Moderately discriminant
23 - - 1 16 - 29 35 3 - 0.678 Highly discriminant
24 - 84 - - - - - - - 0.000 Poorly discriminant
26 - - - 2 16 61 4 1 - 0.439 Moderately discriminant
27 - 4 6 74 - - - - - 0.219 Poorly discriminant
31 - 4 12 68 - - - - - 0.326 Moderately discriminant
39 - 1 83 - - - - - 0.024 Poorly discriminant
40 - 26 5 34 8 2 4 3 2 0.732 Highly discriminant

the allelic diversity of the loci was classified as highly discriminant [Hunter-Gaston index (HGI) > 0.6], moderately discriminant
(0.3 ≤ HGI ≤ 0.6) and poorly discriminant (HGI < 0.3), according to Sola et al. (2003).

TABLE III
Discriminatory power of spoligotyping and mycobacterial interspersed repetitive units (MIRU) typing, alone and in association

distinct patterns clusters Clustered isolates

Methodologies n n n HGI
Spoligotyping 33 13 64 0.954
MIRU 67 9 26 0.991
Spoligotyping + MIRU 72 8 20 0.995

HGI: Hunter-Gaston index.

 Mem Inst Oswaldo Cruz, Rio de Janeiro, Vol. 105(6), September 2010 783

DISCUSSION MIRU locus 23 to have the second-lowest discrimina-

Molecular methods have been used for epidemiologi- tory power, which differs from our finding that MIRU
cal studies of TB in developed countries, but this kind of locus 39 was the second-least discriminating.
study is scarce in developing countries. Particularly, there MIRU-VNTR significantly reduced the number of
are few studies of this nature in Brazil (Borsuk et al. 2005, epidemiological links among the isolates studied, while
Cafrune et al. 2006, Oelemann et al. 2007b, Malaspina et spoligotyping overestimated these links. Thus, MIRU-
al. 2008, Malaghini et al. 2009). To explore the molecular VNTR demonstrated a superior capacity to differenti-
epidemiology of TB in northwestern PR, we chose to work ate M. tuberculosis clinical isolates in our study. We
with two effective and fast methodologies: spoligotyping believe that using the newly developed 15 or 24-locus
and MIRU-VNTR typing (Malaghini et al. 2009). MIRU-VNTR (Supply et al. 2006, Oelemann et al. 2007a,
The distribution of 93 isolates analyzed by spoligotyp- Alonso-Rodríguez et al. 2008) could further increase the
ing in our study revealed 36 distinct patterns, six of which discriminatory power of this method.
have not yet been described in SITVIT (10.8% of isolates). Despite the high discriminatory power of MIRU-
These new STs have been submitted to the database by our VNTR typing observed in our study, a few isolates clus-
group. The LAM, H and T families were the largest fami- tered by this method and, in some cases, the clustered
lies observed in our study and they are the three genotypic isolates were discriminated by spoligotyping. This result
families most frequently found in Africa, Central Ameri- demonstrates the need to combine the two methodologies
ca, Europe and South America (Brudey et al. 2006). to provide a higher discriminatory power for epidemio-
Previous studies analyzing M. tuberculosis isolates logical study. According to Cowan et al. (2005), when
from Araraquara and the state of Rio Grande do Sul (RS) combining these two methodologies, the clustering rate
in southern Brazil reported the prevalence of ST53 (T1 is similar to that of IS6110-RFLP fingerprinting.
sublineage) (Borsuk et al. 2005, Malaspina et al. 2008). The results obtained with spoligotyping in our study
In contrast, in our study, ST53 was only representative suggested a possible epidemiological link between two
of 5.4% (5/93) of the isolates, while ST46 (U, likely H patients who were prisoners in the same prison at the
lineage) was the most prevalent and accounted for ap- same time and whose isolates displayed the same mo-
proximately 14% (13/93) of the isolates. lecular pattern (ST46). However, when these isolates
One isolate detected in our study belonged to the EAI were typed by MIRU-VNTR, they showed different
lineage (ST48, EAI1_SOM), which is typically more patterns. In other M. tuberculosis isolates that were
prevalent in southeast Asia. This lineage has also been clustered by combining spoligotyping and MIRU-VN-
detected in another study in northern Brazil. ST1892 (U TR typing, no correlation with retrospective epidemio-
lineage) and ST2512, both characterized in our study, logical data was apparent.
are known to have exclusive geographic distribution in Significant diversity in circulating strains was ob-
Brazil. ST61 (LAM10_CAM sublineage), ST563 (U lin- served in our study of M. tuberculosis isolates from what
eage) and ST2654 were also identified in our study but is considered to be a low-endemic TB setting in Brazil,
had not yet been described in our country. The other STs, using a combination of spoligotyping and MIRU-VNTR
which had already been described in the SITVIT data- typing. These methodologies allowed 85.7% differentia-
base (pasteur-guadeloupe.fr:8081/SITVITDemo/), have tion of the M. tuberculosis isolates analyzed, which is
been previously identified in Brazil, but this is the first in agreement with a recent study conducted in the city
report of their presence in our state, probably because of of Curitiba (the capital of PR) that found 93.5% differ-
the limited study in our region previously. entiation using Mixed-linker PCR DNA fingerprinting
When MIRU-VNTR was applied to the study of M. tu- (Malaghini et al. 2009). Alternatively, an analysis of M.
berculosis isolates, the most allelic diversity was observed tuberculosis isolates from three regions of RS found
for MIRU loci 40, 23, 10 and 16 (HGI = 0.732, 0.678, 0.673 66% clonal differentiation using IS6110-RFLP and spo-
and 0.641, respectively), while moderate polymorphisms ligotyping (Cafrune et al. 2006), while isolates from two
were found in MIRU loci 26, 20 and 31 (HGI = 0.439, cities in RS that were analyzed by IS6110-RFLP and
0.380 and 0.326, respectively). Conversely, Kovalev et al.
spoligotyping revealed high clonal diversity of M. tuber-
(2005) observed the most allelic diversity for MIRU loci
culosis (Borsuk et al. 2005).
26, 31 and 10 in M. tuberculosis isolated from the Ural
In summary, the present study offers the first in-
region of the Russian Federation. Sharma et al. (2008),
studying isolates from Kanpur, India, and working with sight about the genetic diversity of M. tuberculosis iso-
only six MIRU loci (MIRU 4, 10, 16, 26, 39 and 40), ob- lates from patients with TB in cities from the northwest
served that the most discriminatory loci were, in order of region of PR. This kind of study is important for trac-
diversity, 26, 10, 16 and 40. Despite the high discrimina- ing relationships among the strains and may also help
tory power of MIRU locus 40 as observed in this study (8 to interrupt the chain of transmission between differ-
alleles), it is possible that a higher discriminatory power ent communities and aid in recognition of the principal
can be obtained with another study population. clades responsible for spreading the disease. Of course,
MIRU locus 24 was present in a single copy in all additional studies with more isolates over longer time
isolates analyzed in our study, which is in agreement periods will be critical if we are to fully understand the
with Kovalev et al. (2005). However, that study found epidemiology of TB in this region.
784 Genotyping of M. tuberculosis isolates from Brazil • Erika Noda Noguti et al.

REFERENCES strains isolated in Ural region, Russian Federation, by MIRU-

VNTR genotyping. Int J Tuberc Lung Dis 9: 746-752.
Alonso-Rodríguez N, Martínez-Lirola M, Herránz M, Sanchez-
Benitez M, Barroso P, INDAL-TB group, Bouza E, García de Malaghini M, Brockelt SR, Burger M, Kritski A, Thomaz-Soccol V
Viedma D 2008. Evaluation of the new advanced 15-loci MIRU- 2009. Molecular characterization of Mycobacterium tuberculo-
VNTR genotyping tool in Mycobacterium tuberculosis molecu- sis isolated in the state of Parana in southern Brazil. Tuberculo-
lar epidemiology studies. BMC Microbiol 8: 34. sis (Edinb) 89: 101-105.
Borsuk S, Dellagostin MM, Madeira S de G, Lima C, Boffo M, Mattos Malaspina AC, Cavalcanti HR, Leite CQ, Machado SM, Viana BH,
I, Almeida da Silva PE, Dellagostin OA 2005. Molecular char- Silva RM, Hage EF, Figueiredo WM, Marques E, Ferrazoli L,
acterization of Mycobacterium tuberculosis isolates in a region Arbex M, Lessi M, Fonseca LS, Rigouts L, Saad MH 2008. Use-
of Brazil with a high incidence of tuberculosis. Microbes Infec fulness of Mycobacterium tuberculosis molecular typing in a
7: 1338-1344. tuberculosis low-endemic agro-industrial setting of Brazil. Jpn
J Infect Dis 61: 231-233.
Brudey K, Driscoll JR, Rigouts L, Prodinger WM, Gori A, Al-Hajoj
SA, Allix C, Aristimuño L, Arora J, Baumanis V, Binder L, Mazars E, Lesjean S, Banuls AL, Gilbert M, Vincent V, Gicquel B,
Cafrune P, Cataldi A, Cheong S, Diel R, Ellermeier C, Evans Tibayrenc M, Locht C, Supply P 2001. High-resolution minisat-
JT, Fauville-Dufaux M, Ferdinand S, Garcia de Viedma D, ellite-based typing as a portable approach to global analysis
Garzelli C, Gazzola L, Gomes HM, Guttierez MC, Hawkey PM, of Mycobacterium tuberculosis molecular epidemiology. Proc
van Helden PD, Kadival GV, Kreiswirth BN, Kremer K, Kubin M, Natl Acad Sci USA 98: 1901-1906.
Kulkarni SP, Liens B, Lillebaek T, Ho ML, Martin C, Martin C,
Mokrousov I, Narvskaïa O, Ngeow YF, Naumann L, Niemann Molhuizem HO, Bunschoten AE, Schouls LM, van Embden JDA
S, Parwati I, Rahim Z, Rasolofo-Razanamparany V, Rasolo- 1998. Rapid detection and simultaneous strain differentiation of
navalona T, Rossetti ML, Rüsch-Gerdes S, Sajduda A, Samper Mycobacterium tuberculosis complex bacteria by spoligotyping.
S, Shemyakin IG, Singh UB, Somoskovi A, Skuce RA, van In T Parish, NG Stoker, Methods in molecular biology, Mycobac-
Soolingen D, Streicher EM, Suffys PN, Tortoli E, Tracevska T, teria protocols, vol. 101, Humana Press, New Jersey, p. 381-394.
Vincent V, Victor TC, Warren RM, Yap SF, Zaman K, Portaels Oelemann MC, Diel R, Vatin V, Haas W, Rüsch-Gerdes S, Locht
F, Rastogi N, Sola C 2006. Mycobacterium tuberculosis com- C, Niemann S, Supply P 2007a. Assessment of an optimized
plex genetic diversity: mining the fourth international spoligo- mycobacterial interspersed repetitive-unit-variable-number-
typing database (SpolDB4) for classification, population genet- tandem-repeat typing system combined with spoligotyping for
ics and epidemiology. BMC Microbiol 6: 23. population-based molecular epidemiology studies of tuberculo-
Cafrune PI, Riley LW, Possuelo LG, Valim AR, Borges M, Ribeiro sis. J Clin Microbiol 45: 691-697.
MO, Rossetti ML, Zaha A 2006. Recent transmission of tuber- Oelemann MC, Fontes ANB, Pereira MAS, Bravin Y, Silva G,
culosis involving retired patients. J Infect 53: 370-376. Degrave W, Carvalho ACC, Brito RC, Kritski AL, Suffys PN
2007b. Typing of Mycobacterium tuberculosis strains isolated
Candia N, Lopez B, Zozio T, Carrivale M, Diaz C, Russomando G,
in community health centers of Rio de Janeiro city, Brazil. Mem
de Romero NJ, Jara JC, Barrera L, Rastogi N, Ritacco V 2007.
Inst Oswaldo Cruz 102: 455-462.
First insight into Mycobacterium tuberculosis genetic diversity
in Paraguay. BMC Microbiol 7: 75. Pandolfi JR 2006. Otimização da técnica de MIRU (mycobacterial
interspersed repetitive units) para o estudo epidemiológico de
Cowan LS, Diem L, Monson T, Wand P, Temporado D, Oemig TV,
pacientes com tuberculose, Thesis, Universidade Estadual Pau-
Crawford JT 2005. Evaluation of a two-step approach for large-
lista, 82 pp.
scale, prospective genotyping of Mycobacterium tuberculosis
isolates in the United States. J Clin Microbiol 43: 688-695. Prodinger WM 2007. Molecular epidemiology of tuberculosis: toy
or tool? A review of the literature and examples from Central
Dale JW, Brittain D, Cataldi AA, Cousins D, Crawford JT, Driscoll
Europe. Wien Klin Wochenschr 119: 80-89.
J, Heersma H, Lillebaek T, Quitugua T, Rastogi N, Skuce RA,
Sola C, Van Soolingen D, Vincent V 2001. Spacer oligonucle- Sharma P, Chauhan DS, Upadhyay P, Faujdar J, Lavania M, Sachan
otide typing of bacteria of the Mycobacterium tuberculosis S, Katoch K, Katoch VM 2008. Molecular typing of Mycobac-
complex: recommendations for standardised nomenclature. Int terium tuberculosis isolates from a rural area of Kanpur by
J Tuberc Lung Dis 5: 216-219. spoligotyping and mycobacterial interspersed repetitive units
(MIRUs) typing. Infect Genet Evol 8: 621-626.
González-y-Merchand JA, Estrada-García I, Colston MJ, Cox RA
1996. A novel method for the isolation of mycobacterial DNA. Sneath PHA, Sokal RR 1973. Numerical taxonomy: the principles
FEMS Microbiol Lett 135: 71-77. and practice of numerical classification, WH Freeman & Co,
San Francisco, 573 pp.
Hunter PR, Gaston MA 1988. Numerical index of the discrimina-
tory ability to typing systems: an application of Simpson’s index Sola C, Filliol I, Legrand E, Lesjean S, Locht C, Supply P, Rastogi
of diversity. J Clin Microbiol 26: 2465-2466. N 2003. Genotyping of the Mycobacterium tuberculosis com-
plex using MIRUs: association with VNTR and spoligotyping
Kamerbeek J, Schouls L, Kolk A, van Agterveld M, van Soolingen
for molecular epidemiology and evolutionary genetics. Infect
D, Kuijper S, Bunschoten A, Molhuizen H, Shaw R, Goyal M,
Genet Evol 3: 125-133.
van Embden J 1997. Simultaneous detection and strain differen-
tiation of Mycobacterium tuberculosis for diagnosis and epide- Supply P, Allix C, Lesjean S, Cardoso-Oelemann M, Rüsch-Gerdes
miology. J Clin Microbiol 35: 907-914. S, Willery E, Savine E, de Haas P, van Deutekom H, Roring S,
Kent PT, Kubica GP 1985. Public health mycobacteriology. A guide Bifani P, Kurepina N, Kreiswirth B, Sola C, Rastogi N, Vatin
for the level III laboratory, Centers for Disease Control, At- V, Gutierrez MC, Fauville M, Niemann S, Skuce R, Kremer K,
lanta, 207 pp. Locht C, van Soolingen D 2006. Proposal for standardization of
optimized mycobacterial interspersed repetitive unit-variable-
Kovalev SY, Kamaev EY, Kravchenko MA, Kurepina NE, Skornia- number tandem repeat typing of Mycobacterium tuberculosis.
kov SN 2005. Genetic analysis of Mycobacterium tuberculosis J Clin Microbiol 44: 4498-4510.
 Mem Inst Oswaldo Cruz, Rio de Janeiro, Vol. 105(6), September 2010 785

Supply P, Lesjean S, Savine E, Kremer K, van Soolingen D, Locht van Embden JD, Cave MD, Crawford JT, Dale JW, Eisenach KD,
C 2001. Automated high-throughput genotyping for study of Gicquel B, Hermans P, Martin C, McAdam R, Shinnick TM,
global epidemiology of Mycobacterium tuberculosis based on Small PM 1993. Strain identification of Mycobacterium tuber-
mycobacterial interspersed repetitive units. J Clin Microbiol culosis by DNA fingerprinting: recommendations for a stan-
39: 3563-3571. dardized methodology. J Clin Microbiol 31: 406-409.
Supply P, Mazars E, Lesjean S, Vincent V, Gicquel B, Locht C 2000. Vitol I, Driscoll J, Kreiswirth B, Kurepina N, Bennett KP 2006.
Variable human minisatellite-like regions in the Mycobacteri- Identifying Mycobacterium tuberculosis complex strains fami-
um tuberculosis genome. Mol Microbiol 36: 762-771. lies using spoligotypes. Infect Genet Evol 6: 491-504.
Valcheva V, Mokrousov I, Rastogi N, Narvskaya O, Markova N WHO - World Health Organization 2009. Global tuberculosis con-
2008. Molecular characterization of Mycobacterium tuberculo- trol: epidemiology, strategy, financing. WHO report WHO/
sis isolates from different regions of Bulgaria. J Clin Microbiol HTM/TB/2009.411). Available from: http://www.who.int/tb/
46: 1014-1018. publications/global_report/2009/en/index.html.