Electrophoresis for Typing
Mycobacterium abscessus
Isolates
Gabriel Esquitini Machado,aCristianne Kayoko Matsumoto,aErica Chimara,bRafael da Silva Duarte,cDenise de Freitas,d Moises Palaci,eDavid Jamil Hadad,eKarla Valéria Batista Lima,fMaria Luiza Lopes,fJesus Pais Ramos,gCarlos Eduardo Campos,g Paulo César Caldas,gBeate Heym,hSylvia Cardoso Leãoa
Departamento de Microbiologia, Imunologia e Parasitologia, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, Brazila; Núcleo de Tuberculose e Micobacterioses, Instituto Adolfo Lutz, São Paulo, Brazilb; Instituto de Microbiologia Paulo de Góes, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazilc; Departamento de Oftalmologia, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, Brazild; Nucleo de Doenças Infecciosas, Universidade Federal do Espírito Santo, Vitória, Espírito Santo, Brazile; Instituto Evandro Chagas, Belém, Pará, Brazilf; Centro de Referência Professor Hélio Fraga, Escola Nacional de Saúde Pública, Fiocruz, Rio de Janeiro, Brazilg; APHP, Hôpitaux Universitaires Paris Ile-de-France Ouest, Laboratoire de Microbiologie, Hôpital Ambroise Paré, Boulogne-Billancourt, Franceh
Outbreaks of infections by rapidly growing mycobacteria following invasive procedures, such as ophthalmological, laparoscopic, arthroscopic, plastic, and cardiac surgeries, mesotherapy, and vaccination, have been detected in Brazil since 1998. Members of theMycobacterium chelonae-Mycobacterium abscessusgroup have caused most of these outbreaks. As part of an epidemiologi-cal investigation, the isolates were typed by pulsed-field gel electrophoresis (PFGE). In this project, we performed a large-sepidemiologi-cale comparison of PFGE profiles with the results of a recently developed multilocus sequence typing (MLST) scheme forM. absces-sus. Ninety-three isolates were analyzed, with 40M. abscessussubsp.abscessusisolates, 47M. abscessussubsp.bolletiiisolates, and six isolates with no assigned subspecies. Forty-five isolates were obtained during five outbreaks, and 48 were sporadic iso-lates that were not associated with outbreaks. For MLST, seven housekeeping genes (argH,cya,glpK,gnd,murC,pta, andpurH) were sequenced, and each isolate was assigned a sequence type (ST) from the combination of obtained alleles. The PFGE patterns of DraI-digested DNA were compared with the MLST results. All isolates were analyzable by both methods. Isolates from mono-clonal outbreaks showed unique STs and indistinguishable or very similar PFGE patterns. Thirty-three STs and 49 unique PFGE patterns were identified among the 93 isolates. The Simpson’s index of diversity values for MLST and PFGE were 0.69 and 0.93, respectively, forM. abscessussubsp.abscessusand 0.96 and 0.97, respectively, forM. abscessussubsp.bolletii. In conclusion, the MLST scheme showed 100% typeability and grouped monoclonal outbreak isolates in agreement with PFGE, but it was less dis-criminative than PFGE forM. abscessus.
N
ontuberculous mycobacteria (NTM) are responsible for a broad spectrum of human diseases, such as pulmonary and disseminated infections, lymphadenopathy, and cutaneous and soft tissue infections, among others (1). The ubiquitous distribu-tion of these organisms facilitates the contaminadistribu-tion of medical equipment and solutions, leading to nosocomial infections and outbreaks (2–4). In recent years, there has been a significant in-crease in the detection of diseases caused by NTM, which can be attributed to an increase in the number of infections affecting immunocompromised patients and the number and type of inva-sive procedures used for cosmetic and therapeutic purposes, as well as to advances in molecular identification techniques (5). In Brazil, several outbreaks of infections by rapidly growing myco-bacteria (RGM) have been detected since 1998 in patients sub-jected to medical invasive procedures, such as ophthalmological, laparoscopic, arthroscopic, plastic and cardiac surgeries, or cos-metic procedures, such as mesotherapy (6–10). Between 2004 and 2008,⬎2,000 patients with surgical site infections caused by a unique strain ofMycobacterium abscessus were reported to the federal authorities in Brazil, and the problem was considered an epidemiological emergency (11). Moreover, M. abscessusis the most common RGM species isolated from pulmonary infections in some settings (12).M. abscessus sensu latois an RGM group that includes the for-merly described speciesM. abscessus,Mycobacterium massiliense, andMycobacterium bolletii(13). The reunion of these three
spe-cies into a single spespe-cies, that ofM. abscessus, was proposed in 2009 (13) because they could not be clearly discriminated by pheno-typic or molecular tests, including the gold standard, DNA-DNA hybridization. Small differences observed in the restriction pro-files of a 441-bp fragment of thehsp65gene, as analyzed by PCR-restriction enzyme analysis (PRA-hsp65) (14), in rpoBgene se-quences (15), and in susceptibility to few drugs led to the division ofM. abscessusin two subspecies,M. abscessussubsp.abscessusand
M. abscessussubsp.bolletii, with thebolletiisubspecies including formerM. massilienseandM. bolletii, which share the same
PRA-hsp65profile and highly similarrpoBsequences (16). This classi-fication may well include other subgroupings that have not yet been defined. Some researchers have proposed the existence of three subspecies, based on phylogenetic trees constructed with
Received7 March 2014Returned for modification14 April 2014 Accepted9 May 2014
Published ahead of print4 June 2014 Editor:S. A. Moser
Address correspondence to Sylvia Cardoso Leão, sylvia.leao@unifesp.br. Supplemental material for this article may be found athttp://dx.doi.org/10.1128 /JCM.00688-14.
Copyright © 2014, American Society for Microbiology. All Rights Reserved.
data from whole-genome sequencing and single-nucleotide poly-morphism (SNP) information contained in the core genomes (17–19). These studies evidenced deep genetic divisions corre-sponding to three different taxa, suggesting a subdivision in three subspecies, M. abscessus subsp. abscessus, M. abscessus subsp.
massiliense, andM. abscessus subsp.bolletii. Recently developed schemes for multispacer sequence typing (MST) (20) and multi-locus sequence analysis (MLSA) (21) confirmed the existence of three principal groups withinM. abscessus sensu lato, with each containing one of the type strains (M. abscessusCIP 104536,M. massilienseCIP 108297, andM. bolletiiCIP 108541). However, several reports have described isolates showing ambiguous iden-tification by partial sequencing of different DNA targets (7,10, 22–25), suggesting the occurrence of genetic exchange among members of theM. abscessusgroup and indicating that the taxon-omy ofM. abscessus sensu latois far from being resolved. To avoid confusing identifications, in this work, we classified theM. absces-susisolates into one of the two accepted subspecies (16) using PRA-hsp65andrpoBgene sequence analysis (13).
Molecular typing of outbreak isolates helps epidemiological investigation and the identification of possible sources of infection (26). Ultimately, typing can contribute to outbreak control. Pulsed-field gel electrophoresis (PFGE) is usually performed when specific methods are not available for the species being stud-ied. PFGE has good discriminatory power for separatingM. ab-scessussubsp.abscessusandM. abscessussubsp.bolletiistrains (24, 27) but is a cumbersome and time-consuming technique, requir-ing dedicated equipment that is not widely available. Moreover, DNA degradation associated with PFGE can occur withM. absces-susisolates (28). Schemes for MST and multilocus sequence typ-ing (MLST) were recently developed forM. abscessus(20,29). The MST scheme analyzes intergenic sequences, which, as a conse-quence of less evolutionary pressure, can show high variability (20). In the MLST scheme, single-copy housekeeping genes are sequenced, and the combination of the different gene alleles gen-erates a sequence type (ST) and defines a strain (29). A site for the interpretation and comparison of MLST profiles is available athttp: //www.pasteur.fr/recherche/genopole/PF8/mlst/Myco-abscessus .html.
This study evaluated the discriminatory power of the MLST scheme developed by a group at the Institut Pasteur (29), includ-ing seven housekeepinclud-ing genes, against a collection ofM. abscessus
isolates previously typed by PFGE, including isolates from out-breaks that occurred recently in Brazil and isolates from patients not connected to outbreaks.
MATERIALS AND METHODS
Mycobacterial isolates.A collection of 93 isolates ofM. abscessus sensu latowere studied. Forty-five isolates were obtained during five outbreaks that occurred in eight states in Brazil and were related to mesotherapy, laparoscopic and arthroscopic surgeries, liposurgery, mammoplasty, ab-dominoplasty, breast implant surgery, and vaccination. Forty-eight iso-lates from patients not related to outbreaks were included for comparison and calculation of the discriminatory power of the MLST scheme. These isolates were obtained from respiratory (30), ophthalmological (3), and injection abscess (1) samples between 2002 and 2013 in different states from Brazil (see Table SA1 in the supplemental material). Type strainsM. abscessusATCC 19977, M. massiliense CCUG 48898, and M. bolletii CCUG 50184 were used for comparison.
The original isolates were kept frozen at⫺80°C in Middlebrook 7H9 liquid medium (BD, Sparks, MD) supplemented with oleic
acid-albumin-dextrose-catalase (OADC) (BD) and 15% glycerol. From each isolate, the isolated colonies were obtained on Middlebrook 7H10-OADC agar plates, and single colonies were used in the molecular analyses.
Description of outbreaks.Outbreak 1 affected 14 patients submitted to mesotherapy, a technique involving multiple subcutaneous injections for cosmetic purposes. The outbreak occurred in a single hospital in the city of Belém, state of Pará (PA), between August and October 2004. Isolates from eight patients were obtained and identified asM. bolletii(M. abscessussubsp. bolletiiunder the new classification). Three different PFGE profiles were detected, indicating that it was a polyclonal outbreak (10). Isolated colonies from the same isolates were retyped by PFGE for this work, confirming the existence of three different PFGE patterns among the outbreak isolates.
Outbreak 2 was a nationwide epidemic of surgical site infections re-lated to laparoscopic and arthroscopic surgeries, which generated⬎2,000 notifications from several states to the Brazilian federal authorities be-tween 2004 and 2008 (10,31–33). The study of 157 isolates from the seven states that reported the most cases showed that the infections were caused by a single strain ofM. abscessussubsp.bolletii, initially identified asM. massiliense(7). This strain has a particularrpoBsequevar (GenBank ac-cession no.EU117207), and the PFGE patterns of the typed isolates were highly similar, with differences of only 1 to 3 bands. New localized out-breaks caused by this strain occurred in the states of Amazonas (AM) and Rio Grande do Sul (RS) in 2010, when the epidemic was considered con-trolled. Isolates from eight states were included in this study. One to three isolates from each state were included, according to the different PFGE profiles observed, comprising a total of 16 isolates.
Outbreak 3 occurred in a private clinic in the city of Vila Velha, state of Espírito Santo (ES), between February 12 and 29 July 2008. Fourteen patients subjected to liposurgery, breast implant surgery, and abdomino-plasty had infections caused byM. abscessussubsp.abscessus. Nine isolates were obtained and showed indistinguishable PFGE patterns. In January 2009, a new case was detected at the same clinic. This new isolate showed a PFGE pattern that was indistinguishable from that of the previous iso-lates and was included in this study.
Outbreak 4 occurred in the city of Campinas, state of São Paulo (SP), from 2002 to 2004, and was associated with breast implants. After the first six cases were reported, an epidemiological investigation was performed in different hospitals and 14 confirmed cases of mycobacterial infection, 14 possible cases, and one probable case were detected. Twelve of the 14 mycobacterial isolates obtained during that period were identified as My-cobacterium fortuitum, one asM. abscessus, and one asMycobacterium porcinum(34,35). TheM. fortuitumisolates showed different molecular patterns, indicating that the outbreak was polyclonal. In 2004, the out-break subsided until 2008, when new cases occurred in one of the studied hospitals. Most patients harbored isolates identified as M. fortuitum, showing the same molecular patterns detected in the previous period. However, three isolates were identified asM. abscessussubsp.abscessus and were included in the present study. TheM. abscessusisolate from the first outbreak period (34) was not recovered.
Outbreak 5 occurred in a single vaccination clinic in the city of Andra-dina, SP. Sixty cases were identified between January and October 2008. Infection byM. abscessussubsp.abscessuswas confirmed in 18 cases, and eight isolates were typed by PFGE, showing indistinguishable patterns. Samples from water, liquid soap, syringes, needles, and vaccine vials were collected and cultivated. Several mycobacterial species were recovered from water samples, but not the particular strain that was isolated from the patients, and the sources of infection were not identified (http://www .anvisa.gov.br/divulga/newsletter/boletim_vigipos/260710/boletim _nuvig.htm).
DNA extraction.A loopful of bacteria grown on Middlebrook 7H10-OADC plates was suspended in 300l of TET (10 mM Tris, 1 mM EDTA, 1% Triton X-100 [pH 8.0]), boiled for 10 min, centrifuged at 14,000⫻g for 2 min, and frozen at⫺20°C forⱖ16 h.
Bio-tecnologia, Brazil) as follows: 100l of the bacterial suspension in TET was mixed with 300l of the Brazol reagent and agitated in a vortex for 2 min. A volume of 100l of chloroform was added, and the content of the tube was homogenized and centrifuged for 12 min at 12,000⫻gand 4°C. The supernatant was mixed with the same volume of isopropanol in a new tube and incubated overnight at⫺20°C. After centrifugation at 12,000⫻gfor 20 min, the supernatant was discarded. A volume of 500l of 70% ethanol was added, and the material was centrifuged for 5 min at 12,000⫻g. The pellet was resuspended in 1⫻TE (10 mM Tris 10, 1 mM EDTA [pH 8.0]) with RNase (0.4 mg/ml).
Molecular identification of the isolates. (i) PRA-hsp65.A 441-bp fragment of thehsp65gene was amplified using the primers Tb11 and Tb12 (Table 1), as described by Telenti et al. (14). The amplicons were digested in two separate tubes with BstEII and HaeIII restriction enzymes. The digestion products were visualized after electrophoresis in 3% aga-rose gels stained with ethidium bromide, using a 50-bp ladder as the molecular size standard. The restriction fragment sizes were estimated by visual analysis and were compared to the patterns reported at PRASITE (http://app.chuv.ch/prasite/index.html). Isolates showing theM. absces-sus1 pattern (BstEII, 235/210 bp; HaeIII, 145/70/60/55 bp) were identified asM. abscessussubsp.abscessus, and isolates showing theM. abscessus2 pattern (BstEII, 235/210 bp; HaeIII, 200/70/60/50 bp) were identified as M. abscessussubsp.bolletii.
(ii)rpoBsequencing.Region V of therpoBgene was amplified using the MycoF and MycoR primers (Table 1) and sequenced as described by Adékambi et al. (15). The sequences were analyzed by similarity using the Basic Local Alignment Search Tool (BLAST) (http://www.ncbi.nlm.nih .gov/BLAST) against the sequences fromM. abscessusATCC 19977T,M. massilienseCCUG 48898T, andM. bolletiiCCUG 50184T. The sequences showing the highest similarity with therpoBsequence ofM. abscessus ATCC 19977Tled to the identification of the isolate asM. abscessussubsp. abscessus. When therpoBsequence showed the highest similarity to the sequences from M. massiliense CCUG 48898T or M. bolletii CCUG 50184T, the isolate was identified asM. abscessussubsp.bolletii.
Multilocus sequence typing.Seven housekeeping genes were in-cluded in the MLST scheme:argH(argininosuccinate lyase),cya (adenyl-ate cyclase),glpK(glycerol kinase),gnd(6-phosphogluconate dehydroge-nase), murC (UDP N-acetylmuramate-L-Ala ligase), pta (phosphate acetyltransferase), andpurH(phosphoribosylaminoimidazole carboxy-lase ATPase subunit). The genes were amplified using the primers de-picted inTable 1.
PCRs were prepared in 15l containing 1⫻PCR buffer (75 mM Tris-HCl, 20 mM (NH4)2SO4, 0.01% [vol/vol] Tween 20 [pH 8.8]), 2 mM MgCl2, 0.4M each primer, and 2l of bacterial lysates or 0.5 to 1l of purified DNA.
The obtained sequences were trimmed to the sizes indicated in the MLST site (http://www.pasteur.fr/recherche/genopole/PF8/mlst/primers _abscessus.html). Consensus sequences were constructed using the BioEdit program v.7.1.3.0 and were analyzed using the single-locus query from the MLST site (http://www.pasteur.fr/cgi-bin/genopole/PF8 /mlstdbnet.pl?page⫽oneseq&file⫽Mabscessus_profiles.xml).
For each isolate, a combination of the different obtained alleles was submitted to the allelic profile query (http://www.pasteur.fr/cgi-bin /genopole/PF8/mlstdbnet.pl?page⫽profile-query&file⫽Mabscessus _profiles.xml) to obtain a sequence type (ST) number. Novel alleles and STs identified in this work were numbered and included in the MLST database. Isolates were grouped in a clonal complex (CC) when all alleles except one were identical.
The efficiency of the MLST scheme was estimated through quantifying typeability, reproducibility, and discriminatory power, which were eval-uated by calculating the numerical index of discrimination (D) based on Simpson’s index of diversity, as previously described (36).
Pulsed-field gel electrophoresis.All isolates included in this study were previously analyzed by PFGE, according to protocols described in previous publications (7,10,24). With some isolates, PFGE was repeated or carried out for this work according to previously described protocols (24). The gel images were analyzed using the BioNumerics program v.5.1 (Applied Maths, Sint-Martens-Latem, Belgium). Dendrograms were pre-TABLE 1Primers used for molecular identification and MLST
Gene Primer Sequence (5=to 3=) Position (gene)a Reference
hsp65 Tb11 ACCAACGATGGTGTGTCCAT 145–164 (MAB_0650) 14
Tb12 CTTGTCGAACCGCATACCCT 565–585 (MAB_0650)
rpoB MycoF GGCAAGGTCACCCCGAAGGG 1100–1119 (MAB_3869c) 15
MycoR AGCGGCTGCTGGGTGATCATC 368–388 (MAB_3869c)
argH ARGHF GACGAGGGCGACAGCTTC 1114–1131 (MAB_2342) 23
ARGHSR1 GTGCGCGAGCAGATGATG 514–531 (MAB_2343)
cya ACF GTGAAGCGGGCCAAGAAG 2701–2718 (MAB_0489) 23
ACSR1 AACTGGGAGGCCAGGAGC 1004–1021 (MAB_0490c)
glpK GLPKSF1 AATCTCACCGGCGGTGTC 511–528 (MAB_0382) 23
GLPKSFR2 GGACAGACCCACGATGGC 1102–1119 (MAB_0382)
gnd GNDF GTGACGTCGGAGTGGTTGG 65–83 (MAB_2412c) 23
GNDSR1 CTTCGCCTCAGGTCAGCTC 930–948 (MAB_2413c)
murC MURCSF1 CGGACGAAAGCGACGGCT 512–529 (MAB_2007) 23
MURCSR2 CCAAAACCCTGCTGAGCC 1101–1118 (MAB_2007)
pta PTASF1 GATCGGGCGTCATGCCCT 1367–1384 (MAB_4226c) 21
PTASR2 ACGAGGCACTGCTCTCCC 665–682 (MAB_4226c)
purH PURHSF1 CGGAGGCTTCACCCTGGA 516–533 (MAB_1064) 21
PURHSR2 CAGGCCACCGCTGATCTG 1132–1149 (MAB_1064)
pared using the band-based Dice unweighted-pair group method using average linkages (UPGMA) method using 1% to 2% optimization and position tolerance.
RESULTS AND DISCUSSION
Identification of isolates included in the study.The identifica-tion of the 93 isolates studied using PRA-hsp65andrpoB sequenc-ing in combination was congruent with all but six isolates. Forty isolates were identified asM. abscessus subsp.abscessusand 47 were identified asM. abscessussubsp.bolletii. Six isolates were not assigned to a subspecies because of divergence between the
PRA-hsp65profile and therpoBsequence analysis. They were classified asM. abscessussubsp.abscessusbyrpoBsequencing andM. absces-sussubsp.bolletiiby PRA-hsp65(Table 2).
MLST.All genes were amplified with the isolates of the study, confirming the 100% typeability of the proposed MLST scheme. During the study, several isolates were typed two or three times, and the results were confirmed. Moreover, isolates BRA09, BRA25, BRA26, BRA28, and BRA29 were previously typed in a different laboratory and generated the same ST, indicating that MLST has good reproducibility (21). The choice of using bacterial lysates or DNA purified using a commercial kit did not seem to affect the amplification and sequencing results.
In this study, amplifications were carried out without the use of the PCR kits used in the work of Macheras et al. (29). The ampli-fication of some genes required modiampli-fications to the PCR mix. GenemurCwas amplified only after adding 10% glycerol to the ultrapure water used to complete the PCR mix volume. The MgCl2concentration was adjusted to 1.5 mM to reduce nonspe-cific amplifications with theglpKandargHprimers. ST23 isolates generated nonspecific amplifications of thecyagene, and the se-quencing performed with the forward primer was often contam-inated. We located the sequence of the forward primer in the ge-nome of isolate GO-06 (GenBank accession no. CP003699), which is an isolate from the strain that caused the Brazilian surgi-cal site epidemics and belongs to ST23. Compared to thecya for-ward primer sequence (GTG AAG CGG GCC AAG AAG), the corresponding sequence of GO-06 has two mismatches (indicated by bold type) (GTG AAG CGG GCTAAAAAG) that might favor
the annealing of the forward primer to five adjacent regions that have similar sequences, which were detected with Oligo 7 Primer Analysis software (Molecular Biology Insights, Inc., CO, USA) (data not shown). Spurious annealing of thecyaforward primer generated a collection of amplicons of similar sizes and explain the contamination of the sequences generated with this primer. For-tunately, sequences obtained with thecyareverse primer were used for unambiguous allele assignment of ST23 isolates. Tettelin et al. (18) observed byin silicoanalysis that thecyaforward primer ACF would not generate amplicons with some strains and de-signed a new primer that was thought to be conserved across the
M. abscessusgroup. In the MLST scheme developed by Kim et al. (37), the authors designed new primers to amplifycyaandmurC, but no explanation was given to justify these changes.
Table 3presents the MLST results. Theptagene showed the highest number of alleles withM. abscessussubsp.abscessus iso-lates. With isolates identified asM. abscessussubsp.bolletii, the genesglpK,pta, andpurHshowed more alleles than other genes. TheargHgene showed the largest number of polymorphic sites and the highest percentage of nucleotide divergence in the ob-served alleles. AllM. abscessussubsp.abscessusisolates showed the sameglpKallele (allele 1), and no polymorphic sites were detected. Kim et al. (37) excluded theglpKgene from the MLST scheme because of the low frequency of polymorphic sites. No deletions or insertions were observed in the analyzed genes, as previously ob-served by Macheras et al. (29).
In total, 33 STs were identified among the 93 isolates and were specific for each subspecies. Eight STs were specific toM. abscessus
subsp.abscessus, 21 were specific toM. abscessussubsp.bolletii, and four were detected that were specific to the six isolates not assigned to a subspecies. Alleles fromM. abscessussubsp.abscessus
were not found inM. abscessussubsp.bolletiiisolates and vice versa, with the exception ofargHallele 3, which is specific toM. abscessussubsp.abscessus, which was observed with isolate IAL 013, which was identified asM. abscessussubsp.bolletii. The pres-ence of alleles from oneM. abscessussubspecies in isolates from another subspecies was previously observed by Macheras et al. (29).
TABLE 2Isolates included in the study
Origin n
Location of isolationa
Yr of
isolation PRA-hsp65pattern (no. of isolates)
Outbreak 1 8 Belém, PA 2004 M. abscessustype 2
Outbreak 2 2 Belém, PA 2004 M. abscessustype 2
2 Rio de Janeiro, RJ 2006
2 Cuiabá, MT 2006
1 Vitória, ES 2007
1 Santo Ângelo, RS 2007
2 Curitiba, PR 2008
2 Assis, SP 2008
1 Carazinho, RS 2010
3 Manaus, AM 2010
Outbreak 3 10 Vila Velha, ES 2008 M. abscessustype 1
Outbreak 4 3 Campinas, SP 2008 M. abscessustype 1
Outbreak 5 8 Andradina, SP 2008 M. abscessustype 1
Not related to outbreaks 48 M. abscessustype 1 (19),M. abscessustype 2 (23),
M. abscessustype 2 (6)b aAM, Amazonas; ES, Espírito Santo; MT, Mato Grosso; PA, Pará; PR, Paraná; RJ, Rio de Janeiro; RS, Rio Grande do Sul; SP, São Paulo.
bAmbiguous identification;
The six isolates not identified at the subspecies level carryM. abscessussubsp.bolletii-specific alleles, except forglpkallele 1 from isolate CRPHF 3932 andcyaallele 12 from CRPHF 3452, which are specific toM. abscessussubsp.abscessus. This is a strong indi-cation that the six isolates should be included in subspeciesbolletii, thereby confirming the classification obtained by PRA-hsp65and not that obtained byrpoBsequence analysis. The ambiguous iden-tification ofM. abscessusisolates obtained by sequence analysis of different genes was reported in several publications (22–25), sug-gesting the occurrence of homologous recombination after lateral gene transfer events between members of theM. abscessusgroup. Genetic exchange events affecting the housekeeping genescyaand
glpKwere observed for the first time by Macheras et al. (21,23). We detected probable horizontal gene transfer events affecting the
argH,cya, andglpKgenes, and Kreutzfeldt et al. (38) observed evidence of the same events with thecya,pta, andpurHgenes, confirming that lateral transfer events of housekeeping genes are somewhat frequent in this group of bacteria.
Comparison of MLST and PFGE patterns of outbreak iso-lates.This is the first large-scale comparison ofM. abscessusstrains typed by MLST and PFGE. The results obtained with the outbreak isolates were mostly congruent, indicating that outbreak isolates belonging to the same ST often showed indistinguishable PFGE patterns. However, there were some exceptions that will be dis-cussed below.
We showed in a previous study (10) that the eight isolates from the mesotherapy outbreak (outbreak 1) were distributed in three
rpoBsequevars, all with the highest similarity to the corresponding sequence ofM. bolletiiCIP 108541Tand three clusters by PFGE. We repeated the PFGE with isolated colonies and confirmed the three PFGE profiles, which corresponded to the three different novel STs identified in this study (Fig. 1A).
All isolates from outbreak 2 showed 100% similar rpoB se-quences and were included in ST23 by MLST. The PFGE patterns showed differences of 1 to 3 bands (Fig. 1B). At least one of these
bands, approximately 50 kb in length, corresponds to a plasmid that is absent in some isolates (39). Tenover et al. (40) and van Belkum et al. (26) stated that isolates differing by one to four PFGE bands could be considered closely related and therefore most likely part of the (same) outbreak. This is most likely valid in the context of habitual health care-associated outbreaks of limited duration (ⱕ6 months), which is not the case here, because surgical site infections related to this outbreak were detected from 2004 to 2010. The maintenance ofrpoBsequences and MLST patterns throughout the epidemic period confirmed once more that this prolonged outbreak was caused by a unique strain. This strain may have suffered genomic changes caused by deletions, insertions, or by nucleotide substitutions in DraI restriction sites between 2004 and 2010, which caused limited band shifts in the PFGE patterns. This was not investigated here. The sequences of the MLST house-keeping genes did not change in the course of this outbreak, sug-gesting that MLST profiles may be more stable than PFGE patterns withinM. abscessussubsp.bolletii. Interestingly, MLST and whole-genome sequencing approaches helped demonstrate that ST23 is a global strain ofM. abscessussubsp.bolletiithat causes pulmonary and soft tissue infections, which have been detected in Brazil, the United States, France, Germany, Malaysia, Scotland, Switzerland, and the United Kingdom (17,29,38).
All isolates from outbreak 3 were found to belong to ST142, which was detected for the first time in this work, and displayed indistinguishable PFGE profiles, confirming that this outbreak was caused by a single strain ofM. abscessus subsp.abscessus
(Fig. 1C).
Two STs, ST1 and ST49, were detected among the three isolates from outbreak 4. These isolates showed nonrelated PFGE profiles, indicating that PFGE was more discriminative than MLST, which detected only two different STs (Fig. 1D). These findings were expected because this polyclonal outbreak was caused byM. for-tuitum, and M. abscessus was only occasionally isolated, most likely as a result of intensive laboratory analysis or due to the TABLE 3Variability in the genomiclociof the MLST scheme
Gene
M. abscessus subspecies
DNA fragment size (bp)
No. of alleles
No. of
polymorphic sites
% nucleotide divergence among alleles
Minimum Maximum
argH abscessus 480 3 25 1.67 4.58
bolletii 480 7 31 0.21 5.21
cya abscessus 510 3 2 0.20 0.39
bolletii 510 9 20 0.20 2.94
glpK abscessus 534 1 0 0 0
bolletii 534 10 15 0.19 1.87
gnd abscessus 480 5 21 0.21 3.96
bolletii 480 9 27 0.21 4.17
murC abscessus 537 3 3 0.19 0.56
bolletii 537 8 26 0.19 3.91
pta abscessus 486 6 6 0.21 0.82
bolletii 486 10 17 0.21 2.68
purH abscessus 549 5 14 0.18 2.19
contamination of samples or cultures. Otherwise, M. abscessus
may have caused a secondary polyclonal outbreak concomitant with the majorM. fortuitumoutbreak.
Outbreak 5 occurred in a vaccination clinic in the city of
Ireland. These results confirmed that this outbreak was caused by a single strain ofM. abscessussubsp.abscessus(Fig. 1E).
Comparison of MLST and PFGE patterns of isolates from patients not associated with outbreaks.Among the 48 isolates from patients not related to outbreaks, 19 were identified asM. abscessussubsp.abscessus. These isolates were found to belong to seven STs, including ST22 and especially ST1, the most frequent ST detected in this study (Fig. 2A). Half of the isolates were in-cluded in ST1, while PFGE distinguished four patterns, with 81.96% similarity, among these isolates (Fig. 2A, box 1). The high
break 5 isolates were found to belong to ST22, and their PFGE profiles are indistinguishable from that of isolate IAL 063, which was obtained from the sputum sample of a patient in a different city and, therefore, was not epidemiologically related to the out-break 5 isolates (Fig. 2C). Isolates EPM 25694 and EPM 13219 were found to belong to ST49, as were isolates IAL 045 and IAL 046 from outbreak 4. All ST49 isolates showed different PFGE patterns, with 64.99% overall similarity (Fig. 2D). ST49 has been detected in Malaysia and France, where it is one of the most prev-alent STs (29).
TheDvalue for MLST and PFGE was calculated based on the 19 nonrelatedM. abscessussubsp.abscessusisolates plus the type strainM. abscessusATCC 19977. MLST had aDvalue of 0.69, which is far below the desirable value ofⱖ0.95 considered accept-able for a good typing method (26). TheDvalue of PFGE was 0.93, showing that PFGE has a higher discriminatory capacity than MLST forM. abscessussubsp.abscessus. In a previous study, aD
value of 0.97 was obtained forM. abscessustyped by PFGE, con-firming the good discriminatory power of PFGE forM. abscessus
(27).
The 23 isolates identified asM. abscessussubsp.bolletiiwere distributed in 17 different STs, and ST117 and ST151 were the most common (Fig. 3). Five isolates, CRM 473, IAL 010, CRM 375, CRM 642, and B67, belong to ST117, and IAL 035 belongs to ST167, which differs from ST177 in theglpKallele, thus compris-ing a CC. These isolates showed three different PFGE patterns (84.93% similarity) (Fig. 3, box 1). They were obtained in four different states (Mato Grosso [MT], SP, Rio de Janeiro [RJ], and PA), and the patients were not epidemiologically related. It is pos-sible that this is a circulating strain that has independently infected different patients. ST117 isolates were also detected in Malaysia.
Three isolates, IAL 021, IAL 028, and IAL 025, all belong to ST151, which was identified for the first time in this work, and have indistinguishable PFGE patterns (100% similarity), but no epidemiological relationship was identified between the patients carrying them (Fig. 3, box 2). They lived in different cities in the SP state, and the isolates were obtained in different years. Again, this might be a frequently occurring strain that is circulating in SP.
TheDvalues for MLST and PFGE, calculated based on the 23 epidemiologically unrelatedM. abscessus subsp.bolletiiisolates and the type strainsM. massilienseCCUG 48898 andM. bolletii
CCUG 50184, were 0.96 and 0.97, respectively, indicating that the two methods have similar discriminatory capacities for separating
M. abscessussubsp.bolletiiisolates. The sameDvalue of 0.97 was obtained for PFGE in a previous study from our group, with 41 isolates showing theM. abscessustype 2 PRA-hsp65pattern that is characteristic ofM. abscessussubsp.bolletii(24).
The last six isolates, for which precise subspecies identifica-tions were not achieved by PRA-hsp65andrpoBsequence analysis, belong to four STs. Isolates P54, CRPHF 3850, and BRA26 belong to ST34 and form a CC together with isolates CRPHF 3932 (ST170) and CRPHF 3452 (ST171). They have different PFGE patterns, with 69.25% similarity. Four ST34 isolates detected in France hadrpoBsequences more similar to therpoBsequence of
M. abscessusATCC 19977T, as did the isolates reported here. This confirms that ST34 comprises a particular group of isolates show-ing evidence ofrpoBlateral transfer. The six isolates showed un-related PFGE patterns, thus confirming the superior discrimina-tory power of PFGE for this group of isolates (Fig. 4).
Among the three principal groups generated by MLSA, each con-taining one of the type strains ofM. abscessus,M. massiliense, and
M. bolletii, theM. abscessusbranch was robust, while theM. bolletii
andM. massiliensebranches had relatively low bootstrap values (65 and 71%, respectively) and a diffuse appearance, particularly theM. massiliensebranch.M. abscessussubsp.abscessusinfections are more often respiratory than areM. abscessussubsp.bolletii
infections (which are more often skin and soft tissue infections), and antibiotic treatment is more important in respiratory infec-tion (mostly 2 or 3 antibiotics) than in skin infecinfec-tions (mostly macrolide treatment as monotherapy), which may imply different mutational pressures on clinical isolates from each subspecies.
In conclusion, a good typing method should be capable of typ-ing all isolates from a species, grouptyp-ing isolates belongtyp-ing to a single strain and discriminating between nonrelated isolates. The
M. abscessusMLST scheme showed 100% typeability and grouped most isolates based on outbreak settings that were grouped by PFGE, but it was less discriminative than PFGE forM. abscessus
subsp. abscessus. MLST was less discriminative than PFGE for other species, such asSalmonella enterica(41),Lactobacillus san-franciscensis(42),Staphylococcus aureus(43), and Pseudomonas aeruginosa(44). The possibility of increasing the discriminatory power of theM. abscessusMLST scheme by including additional genes has to be evaluated in the future.
The advantages of the MLST scheme cannot be overlooked. PCR and sequencing are currently available in most settings and can overcome the need to cultivate the mycobacterium. When the STs of strains responsible for monoclonal outbreaks are known, MLST can be of great help for epidemiological surveillance. Nev-ertheless, the implications of MLST for studies of population ge-netics and dynamics may be more significant than those for epi-demiological investigation. We have detected strains that share MLST profiles in patients with no suspected epidemiological rela-tionship. MLST can be used to track these strains and observe their distribution and dynamics.
ACKNOWLEDGMENTS
This study received financial support from Fundação de Amparo à Pes-quisa do Estado de São Paulo (www.fapesp.br) (FAPESP) (grant 2011/ 18326-4). G.E.M. and C.K.M. received fellowships from FAPESP (2012/ 18038-1, 2012/07335-5, and 2013/16018-6).
REFERENCES
1.Brown-Elliott BA, Wallace RJ, Jr.2002. Clinical and taxonomic status of pathogenic nonpigmented or late-pigmenting rapidly growing
mycobac-teria. Clin. Microbiol. Rev.15:716 –746.http://dx.doi.org/10.1128/CMR .15.4.716-746.2002.
2.Tiwari TS, Ray B, Jost KC, Jr, Rathod MK, Zhang Y, Brown-Elliott BA, Hendricks K, Wallace RJ, Jr.2003. Forty years of disinfectant failure: outbreak of postinjectionMycobacterium abscessusinfection caused by contamination of benzalkonium chloride. Clin. Infect. Dis.36:954 –962.
http://dx.doi.org/10.1086/368192.
3.Villanueva A, Calderon RV, Vargas BA, Ruiz F, Aguero S, Zhang Y, Brown BA, Wallace RJ, Jr.1997. Report on an outbreak of postinjection abscesses due toMycobacterium abscessus, including management with surgery and clarithromycin therapy and comparison of strains by random amplified polymorphic DNA polymerase chain reaction. Clin. Infect. Dis. 24:1147–1153.http://dx.doi.org/10.1086/513656.
4.Wallace RJ, Jr, Brown BA, Griffith DE.1998. Nosocomial outbreaks/ pseudo-outbreaks caused by nontuberculous mycobacteria. Annu. Rev. Microbiol. 52:453– 490. http://dx.doi.org/10.1146/annurev.micro.52.1 .453.
5.Tortoli E.2009. Clinical manifestations of nontuberculous mycobacteria infections. Clin. Microbiol. Infect.15:906 –910.http://dx.doi.org/10.1111 /j.1469-0691.2009.03014.x.
6.Freitas D, Alvarenga L, Sampaio J, Mannis M, Sato E, Sousa L, Vieira L, Yu MC, Martins MC, Hoffling-Lima A, Belfort R, Jr. 2003. An outbreak ofMycobacterium chelonaeinfection after LASIK. Ophthalmol-ogy110:276 –285.http://dx.doi.org/10.1016/S0161-6420(02)01643-3. 7.Leão SC, Viana-Niero C, Matsumoto CK, Lima KV, Lopes ML, Palaci
M, Hadad DJ, Vinhas S, Duarte RS, Lourenço MC, Kipnis A, das Neves ZC, Gabardo BM, Ribeiro MO, Baethgen L, de Assis DB, Madalosso G, Chimara E, Dalcolmo MP.2010. Epidemic of surgical-site infections by a single clone of rapidly growing mycobacteria in Brazil. Future Microbiol. 5:971–980.http://dx.doi.org/10.2217/fmb.10.49.
8.Sampaio JLM, Junior DN, de Freitas D, Höfling-Lima AL, Miyashiro K, Alberto FL, Leão SC.2006. An outbreak of keratitis caused by Mycobac-terium immunogenum. J. Clin. Microbiol.44:3201–3207.http://dx.doi.org /10.1128/JCM.00656-06.
9.Strabelli TM, Siciliano RF, Castelli JB, Demarchi LM, Leão SC, Viana-Niero C, Miyashiro K, Sampaio RO, Grinberg M, Uip DE. 2010. Mycobacterium chelonaevalve endocarditis resulting from contaminated biological prostheses. J. Infect.60:467– 473.http://dx.doi.org/10.1016/j .jinf.2010.03.008.
10. Viana-Niero C, Lima KV, Lopes ML, Rabello MC, Marsola LR, Bril-hante VC, Durham AM, Leão SC.2008. Molecular characterization of Mycobacterium massilienseandMycobacterium bolletiiin isolates collected from outbreaks of infections after laparoscopic surgeries and cosmetic procedures. J. Clin. Microbiol. 46:850 – 855. http://dx.doi.org/10.1128 /JCM.02052-07.
11. ANVISA.2011. Relatório descrito de investigação de casos de infecções por micobactérias não tuberculosis de crescimento rápido (MCR) no Bra-sil no período de 1998 a 2009. Agência Nacional de Vigilância Sanitária, Brasilia, Brazil. (In Portuguese.)http://www.anvisa.gov.br/hotsite/hotsite _micobacteria/relatorio_descrito_mcr_16_02_11.pdf.
12. Hoefsloot W, van Ingen J, Andrejak C, Angeby K, Bauriaud R, Bemer P, Beylis N, Boeree MJ, Cacho J, Chihota V, Chimara E, Churchyard G, Cias R, Daza R, Daley CL, Dekhuijzen PN, Domingo D, Drobniewski F, Esteban J, Fauville-Dufaux M, Folkvardsen DB, Gibbons N, Gómez-FIG 4PFGE patterns and dendrograms prepared using the BioNumerics program v.5.1 and the corresponding MLST alleles and STs fromM. abscessusisolates not assigned to a subspecies. Numbers in the upper left corner indicate percent similarity of PFGE profiles. (Some PFGE profiles were reproduced from references
Mampaso E, Gonzalez R, Hoffmann H, Hsueh PR, Indra A, Jagielski T, Jamieson F, Jankovic M, Jong E, Keane J, Koh WJ, Lange B, Leao S, Macedo R, Mannsåker T, Marras TK, Maugein J, Milburn HJ, Mlinkó T, Morcillo N, Morimoto K, Papaventsis D, Palenque E, Paez-Peña M, Piersimoni C, Polanová M, Rastogi N, Richter E, et al. 2013. The geographic diversity of nontuberculous mycobacteria isolated from pul-monary samples: an NTM-NET collaborative study. Eur. Respir. J.42: 1604 –1613.http://dx.doi.org/10.1183/09031936.00149212.
13. Leão SC, Tortoli E, Viana-Niero C, Ueki SY, Lima KV, Lopes ML, Yubero J, Menendez MC, Garcia MJ.2009. Characterization of myco-bacteria from a major Brazilian outbreak suggests that revision of the taxonomic status of members of theMycobacterium chelonae-M. abscessus group is needed. J. Clin. Microbiol.47:2691–2698.http://dx.doi.org/10 .1128/JCM.00808-09.
14. Telenti A, Marchesi F, Balz M, Bally F, Böttger EC, Bodmer T.1993. Rapid identification of mycobacteria to the species level by polymerase chain reaction and restriction enzyme analysis. J. Clin. Microbiol.31:175– 178.
15. Adékambi T, Colson P, Drancourt M.2003.rpoB-based identification of nonpigmented and late-pigmenting rapidly growing mycobacteria. J. Clin. Microbiol. 41:5699 –5708. http://dx.doi.org/10.1128/JCM.41.12 .5699-5708.2003.
16. Leão SC, Tortoli E, Euzéby JP, Garcia MJ.2011. Proposal that Myco-bacterium massilienseandMycobacterium bolletiibe united and reclassified asMycobacterium abscessussubsp.bolletiicomb. nov., designation of My-cobacterium abscessussubsp.abscessussubsp. nov. and emended descrip-tion ofMycobacterium abscessus. Int. J. Syst. Evol. Microbiol.61:2311– 2313.http://dx.doi.org/10.1099/ijs.0.023770-0.
17. Bryant JM, Grogono DM, Greaves D, Foweraker J, Roddick I, Inns T, Reacher M, Haworth CS, Curran MD, Harris SR, Peacock SJ, Parkhill J, Floto RA.2013. Whole-genome sequencing to identify transmission of Mycobacterium abscessusbetween patients with cystic fibrosis: a retrospec-tive cohort study. Lancet 381:1551–1560. http://dx.doi.org/10.1016 /S0140-6736(13)60632-7.
18. Tettelin H, Davidson RM, Agrawal S, Aitken ML, Shallom S, Hasan NA, Strong M, de Moura VC, De Groote MA, Duarte RS, Hine E, Parankush S, Su Q, Daugherty SC, Fraser CM, Brown-Elliott BA, Wallace RJ, Jr, Holland SM, Sampaio EP, Olivier KN, Jackson M, Zelazny AM.2014. High-level relatedness amongMycobacterium abscessussubsp.massiliense strains from widely separated outbreaks. Emerg. Infect. Dis.20:364 –371.
http://dx.doi.org/10.3201/eid2003.131106.
19. Heydari H, Wee WY, Lokanathan N, Hari R, Mohamed Yusoff A, Beh CY, Yazdi AH, Wong GJ, Ngeow YF, Choo SW.2013. MabsBase: a Mycobacterium abscessusgenome and annotation database. PLoS One 8:e62443.http://dx.doi.org/10.1371/journal.pone.0062443.
20. Sassi M, Ben Kahla I, Drancourt M. 2013. Mycobacterium abscessus multispacer sequence typing. BMC Microbiol.13:3.http://dx.doi.org/10 .1186/1471-2180-13-3.
21. Macheras E, Roux AL, Bastian S, Leão SC, Palaci M, Sivadon-Tardy V, Gutierrez C, Richter E, Rüsch-Gerdes S, Pfyffer G, Bodmer T, Cambau E, Gaillard JL, Heym B.2011. Multilocus sequence analysis andrpoB sequencing ofMycobacterium abscessus(sensu lato) strains. J. Clin. Micro-biol.49:491– 499.http://dx.doi.org/10.1128/JCM.01274-10.
22. Kim HY, Kook Y, Yun YJ, Park CG, Lee NY, Shim TS, Kim BJ, Kook YH.2008. Proportions ofMycobacterium massilienseandMycobacterium bolletiistrains among KoreanMycobacterium chelonae-Mycobacterium ab-scessusgroup isolates. J. Clin. Microbiol.46:3384 –3390.http://dx.doi.org /10.1128/JCM.00319-08.
23. Macheras E, Roux AL, Ripoll F, Sivadon-Tardy V, Gutierrez C, Gaillard JL, Heym B.2009. Inaccuracy of single-target sequencing for discrimi-nating species of theMycobacterium abscessusgroup. J. Clin. Microbiol. 47:2596 –2600.http://dx.doi.org/10.1128/JCM.00037-09.
24. Matsumoto CK, Chimara E, Bombarda S, Duarte RS, Leão SC.2011. Diversity of pulsed-field gel electrophoresis patterns ofMycobacterium abscessustype 2 clinical isolates. J. Clin. Microbiol.49:62– 68.http://dx.doi .org/10.1128/JCM.01665-10.
25. Zelazny AM, Root JM, Shea YR, Colombo RE, Shamputa IC, Stock F, Conlan S, McNulty S, Brown-Elliott BA, Wallace RJ, Jr, Olivier KN, Holland SM, Sampaio EP.2009. Cohort study of molecular identification and typing ofMycobacterium abscessus,Mycobacterium massiliense, and Mycobacterium bolletii. J. Clin. Microbiol.47:1985–1995.http://dx.doi .org/10.1128/JCM.01688-08.
26. van Belkum A, Tassios PT, Dijkshoorn L, Haeggman S, Cookson B, Fry NK, Fussing V, Green J, Feil E, Gerner-Smidt P, Brisse S, Struelens M, European Society of Clinical Microbiology and Infectious Diseases (ESCMID) Study Group on Epidemiological Markers (ESGEM).2007. Guidelines for the validation and application of typing methods for use in bacterial epidemiology. Clin. Microbiol. Infect.13(Suppl 3):S1–S46.http: //dx.doi.org/10.1111/j.1469-0691.2007.01732.x.
27. Sampaio JL, Viana-Niero C, de Freitas D, Höfling-Lima AL, Leão SC. 2006. Enterobacterial repetitive intergenic consensus PCR is a useful tool for typingMycobacterium chelonaeandMycobacterium abscessusisolates. Diagn. Microbiol. Infect. Dis. 55:107–118. http://dx.doi.org/10.1016/j .diagmicrobio.2006.01.006.
28. Zhang Y, Yakrus MA, Graviss EA, Williams-Bouyer N, Turenne C, Kabani A, Wallace RJ, Jr.2004. Pulsed-field gel electrophoresis study of Mycobacterium abscessusisolates previously affected by DNA degradation. J. Clin. Microbiol.42:5582–5587.http://dx.doi.org/10.1128/JCM.42.12 .5582-5587.2004.
29. Macheras E, Konjek J, Roux AL, Thiberge JM, Bastian S, Leão SC, Palaci M, Sivadon-Tardy V, Gutierrez C, Richter E, Rüsch-Gerdes S, Pfyffer GE, Bodmer T, Jarlier V, Cambau E, Brisse S, Caro V, Rastogi N, Gaillard JL, Heym B.2014. Multilocus sequence typing scheme for the Mycobacterium abscessuscomplex. Res. Microbiol.165:82–90.http://dx .doi.org/10.1016/j.resmic.2013.12.003.
30. Matsumoto CK, Chimara E, Ramos JP, Campos CE, Caldas PC, Lima KV, Lopes ML, Duarte RS, Leão SC.2012. Rapid tests for the detection of theMycobacterium abscessussubsp.bolletiistrain responsible for an epidemic of surgical-site infections in Brazil. Mem. Instit. Oswaldo Cruz, 107:969 –977.http://dx.doi.org/10.1590/S0074-02762012000800002. 31. Cardoso AM, Martins de Sousa E, Viana-Niero C, Bonfim de Bortoli F,
Pereira das Neves ZC, Leão SC, Junqueira-Kipnis AP, Kipnis A.2008. Emergence of nosocomialMycobacterium massilienseinfection in Goiás, Brazil. Microbes Infect.10:1552–1557.http://dx.doi.org/10.1016/j.micinf .2008.09.008.
32. Duarte RS, Lourenço MC, Fonseca LdS, Leão SC, Amorim EdLT, Rocha ILL, Coelho FS, Viana-Niero C, Gomes KM, da Silva MG, Lorena NSdO, Pitombo MB, Ferreira RMC, Garcia MHdO, de Oliveira GP, Lupi O, Vilaça BR, Serradas LR, Chebabo A, Marques EA, Teixeira LM, Dalcolmo M, Senna SG, Sampaio JLM.2009. Epidemic of postsurgical infections caused byMycobacterium massiliense. J. Clin. Microbiol.47: 2149 –2155.http://dx.doi.org/10.1128/JCM.00027-09.
33. Monego F, Duarte RS, Nakatani SM, Araújo WN, Riediger IN, Brockelt S, Souza V, Cataldo JI, Dias RC, Biondo AW.2011. Molecular identi-fication and typing ofMycobacterium massilienseisolated from postsurgi-cal infections in Brazil. Braz. J. Infect. Dis.15:436 – 441.http://dx.doi.org /10.1016/S1413-8670(11)70224-0.
34. Padoveze MC, Fortaleza CM, Freire MP, Brandão de Assis D, Mada-losso G, Pellini AC, César ML, Pisani Neto V, Beltramelli MM, Chimara E, Ferrazoli L, da Silva Telles MA, Sampaio JL, Leão SC.2007. Outbreak of surgical infection caused by non-tuberculous mycobacteria in breast implants in Brazil. J. Hosp. Infect.67:161–167.http://dx.doi.org/10.1016 /j.jhin.2007.07.007.
35. Sampaio JL, Chimara E, Ferrazoli L, da Silva Telles MA, Del Guercio VM, Jericó ZV, Miyashiro K, Fortaleza CM, Padoveze MC, Leão SC. 2006. Application of four molecular typing methods for analysis of Myco-bacterium fortuitumgroup strains causing post-mammaplasty infections. Clin. Microbiol. Infect. 12:142–149. http://dx.doi.org/10.1111/j.1469 -0691.2005.01312.x.
36. Hunter PR, Gaston MA.1988. Numerical index of the discriminatory ability of typing systems: an application of Simpson’s index of diversity. J. Clin. Microbiol.26:2465–2466.
37. Kim SY, Kang YA, Bae IK, Yim JJ, Park MS, Kim YS, Kim SK, Chang J, Jeong SH.2013. Standardization of multilocus sequence typing scheme forMycobacterium abscessusandMycobacterium massiliense. Diagn. Mi-crobiol. Infect. Dis.77:143–149.http://dx.doi.org/10.1016/j.diagmicrobio .2013.06.023.
38. Kreutzfeldt KM, McAdam PR, Claxton P, Holmes A, Seagar AL, Lau-renson IF, Fitzgerald JR.2013. Molecular longitudinal tracking of Myco-bacterium abscessusspp. during chronic infection of the human lung. PLoS One8:e63237.http://dx.doi.org/10.1371/journal.pone.0063237. 39. Leão SC, Matsumoto CK, Carneiro A, Ramos RT, Nogueira CL, Lima
absces-sussubsp.bolletii. PLoS One8:e60746.http://dx.doi.org/10.1371/journal .pone.0060746.
40. Tenover FC, Arbeit RD, Goering RV, Mickelsen PA, Murray BE, Persing DH, Swaminathan B.1995. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J. Clin. Microbiol.33:2233–2239.
41. Torpdahl M, Skov MN, Sandvang D, Baggesen DL.2005. Genotypic characterization ofSalmonellaby multilocus sequence typing, pulsed-field gel electrophoresis and amplified fragment length polymorphism. J. Mi-crobiol. Methods63:173–184.http://dx.doi.org/10.1016/j.mimet.2005.03 .006.
42. Picozzi C, Bonacina G, Vigentini I, Foschino R.2010. Genetic diversity in ItalianLactobacillus sanfranciscensisstrains assessed by multilocus se-quence typing and pulsed-field gel electrophoresis analyses. Microbiology 156:2035–2045.http://dx.doi.org/10.1099/mic.0.037341-0.
43. Xie Y, He Y, Gehring A, Hu Y, Li Q, Tu SI, Shi X.2011. Genotypes and toxin gene profiles ofStaphylococcus aureusclinical isolates from China. PLoS One6:e28276.http://dx.doi.org/10.1371/journal.pone.0028276. 44. Johnson JK, Arduino SM, Stine OC, Johnson JA, Harris AD. 2007.
