Research in Microbiology 165 (2014) 82e90 www.elsevier.com/locate/resmic

Multilocus sequence typing scheme for the Mycobacterium abscessus complex Edouard Macheras a,c, Julie Konjek c, Anne-Laure Roux b,c, Jean-Michel Thiberge d, Sylvaine Bastian e,f, Sylvia Cardoso Lea˜o g, Moises Palaci h, Vale´rie Sivadon-Tardy a, Cristina Gutierrez i, Elvira Richter k, Sabine Ru¨sch-Gerdes k, Gaby E. Pfyffer l, Thomas Bodmer m, Vincent Jarlier e,f, Emmanuelle Cambau n, Sylvain Brisse d, Vale´rie Caro d, Nalin Rastogi o, Jean-Louis Gaillard a,b,c, Beate Heym a,c,* a

APHP Hoˆpitaux universitaires Paris Ile-de-France Ouest, Service de Microbiologie, Hoˆpital Ambroise Pare´, 9 avenue Charles de Gaulle, 92100 Boulogne-Billancourt, France b APHP Hoˆpitaux universitaires Paris Ile-de-France Ouest, Laboratoire de Microbiologie, Hoˆpital Raymond Poincare´, AP-HP, Garches, France c EA 3647, Universite´ de Versailles Saint-Quentin-en-Yvelines, 2 avenue de la Source de la Bie`vre, 78180 Montigny-le-Bretonneux, France d Institut Pasteur, Genotyping of Pathogens and Public Health, 25 rue du Docteur Roux, 75015 Paris, France e Centre national de Re´fe´rence des Mycobacte´ries et de la Re´sistance des Mycobacte´ries aux Antituberculeux, 47e83 Boulevard de l’Hoˆpital, 75013 Paris, France f APHP, Groupe Hospitalier Pitie´-Salpeˆtrie`re, Laboratoire de Bacte´riologie e Hygie`ne, 47e83 Boulevard de l’Hoˆpital, 75013 Paris, France g Departamento de Microbiologia, Imunologia e Parasitologia, Universidade Federal de Sa˜o Paulo, Rua Botucatu, 862, Vila Clementino, Sa˜o Paulo SP CEP 04023-062, Brazil h Nucleo de Doenc¸as Infecciosas, Universidade Federal do Espı´rito Santo, Av. Fernando Ferrari, 514, Goiabeiras Vito´ria e ES, CEP 29075-910, Brazil i Caribbean Public Health Agency (CARPHA), 16e18 Jamaica Boulevard, Federation Park, Port of Spain, Trinidad and Tobago k National Reference Center for Mycobacteria, Forschungszentrum Borstel, Parkalle 18, 23845 Borstel, Germany l Institut fu¨r Medizinische Mikrobiologie, Zentrum fu¨r LaborMedizin, Luzerner Kantonsspital, Spitalstrasse, 6004 Luzern, Switzerland m Institut fu¨r Infektionskrankheiten, Universita¨t Bern, Hochschulstrasse 4, 3010 Bern, Switzerland n APHP, Groupe hospitalier Lariboisie`re-Fernand Widal, Laboratoire de Microbiologie, 2 rue Ambroise Pare´, 75010 Paris, France o Institut Pasteur, Laboratoire de Recherche et de Re´fe´rence sur la Tuberculose et les Mycobacte´ries, BP 484 Morne Jolivie`re, Les Abymes 97183 Cedex, Guadeloupe, France Received 17 September 2013; accepted 19 December 2013 Available online 31 December 2013

Abstract We developed a multilocus sequence typing (MLST) scheme for Mycobacterium abscessus sensu lato, based on the partial sequencing of seven housekeeping genes: argH, cya, glpK, gnd, murC, pta and purH. This scheme was used to characterize a collection of 227 isolates recovered between 1994 and 2010 in France, Germany, Switzerland and Brazil. We identified 100 different sequence types (STs), which were distributed into three groups on the tree obtained by concatenating the sequences of the seven housekeeping gene fragments (3576 bp): the M. abscessus sensu stricto group (44 STs), the “M. massiliense” group (31 STs) and the “M. bolletii” group (25 STs). SplitTree analysis showed a degree of intergroup lateral transfers. There was also evidence of lateral transfer events involving rpoB. The most prevalent STs in our collection were ST1 (CC5; 20 isolates) and ST23 (CC3; 31 isolates). Both STs were found in Europe and Brazil, and the latter was implicated in a large

* Corresponding author. APHP Hoˆpitaux universitaires Paris Ile-de-France Ouest, Service de Microbiologie, Hoˆpital Ambroise Pare´, 9 avenue Charles de Gaulle, 92100 Boulogne-Billancourt, France. E-mail addresses: [email protected] (E. Macheras), [email protected] (J. Konjek), [email protected] (A.-L. Roux), jean-michel. [email protected] (J.-M. Thiberge), [email protected] (S. Bastian), [email protected] (S.C. Lea˜o), [email protected] (M. Palaci), valerie. [email protected] (V. Sivadon-Tardy), [email protected] (C. Gutierrez), [email protected] (E. Richter), [email protected] (S. Ru¨schGerdes), [email protected] (G.E. Pfyffer), [email protected] (T. Bodmer), [email protected] (V. Jarlier), [email protected] (E. Cambau), [email protected] (S. Brisse), [email protected] (V. Caro), [email protected] (N. Rastogi), jean-louis.gaillard@apr. aphp.fr (J.-L. Gaillard), [email protected] (B. Heym). 0923-2508/$ - see front matter Ó 2013 Institut Pasteur. Published by Elsevier Masson SAS. All rights reserved. http://dx.doi.org/10.1016/j.resmic.2013.12.003

E. Macheras et al. / Research in Microbiology 165 (2014) 82e90

83

post-surgical procedure outbreak in Brazil. Respiratory isolates from patients with cystic fibrosis belonged to a large variety of STs; however, ST2 was predominant in this group of patients. Our MLST scheme, publicly available at www.pasteur.fr/mlst, offers investigators a valuable typing tool for M. abscessus sensu lato in future epidemiological studies throughout the world. Ó 2013 Institut Pasteur. Published by Elsevier Masson SAS. All rights reserved. Keywords: Mycobacterium abscessus complex; Molecular typing; Multilocus sequence typing; Phylogenetic analysis; Molecular identification; rpoB; Cystic fibrosis; Opportunistic mycobacteria; Non-tuberculous mycobacteria

1. Introduction Mycobacterium abscessus sensu lato (i.e., Mycobacterium abscessus sensu stricto, Mycobacterium massiliense and Mycobacterium bolletii) is a rapidly growing mycobacterium (RGM) that has emerged as an important pathogen following its recognition as an entity distinct from Mycobacterium chelonae in 1992 [28]. It is now recognized as the most frequent RGM causing lung disease in humans, far ahead of M. chelonae and Mycobacterium fortuitum [2]. M. abscessus sensu lato lung disease most often, but not exclusively, develops in subjects with underlying lung disorders [20,21]. The disease is particularly prevalent in patients with cystic fibrosis (CF), including young children, and is becoming a major issue in most CF centers in Europe, Israel and North America [21,25,33,40,46,51]. It has recently been shown that, likewise M. tuberculosis, the rough morphotype of M. abscessus, is more virulent and tends to persist. The molecular mechanisms of this finding are related to genetic modifications within the peptide gene cluster mps1-mps2-gap or mmpl4b implicated in the synthesis and transport of glyco-peptido-lipids [42,54]. M. abscessus sensu lato is also a leading cause of sporadic and epidemic cases of skin and soft-tissue (SST) infections following local trauma, the use of contaminated syringes or needles or after surgery [4,16,26]. Several large outbreaks of skin and soft-tissue infection have been reported following injection of adrenal cortex extract, mesotherapy, abdominoplasty, tattooing and piercing [2,7,30]. M. abscessus sensu lato may be responsible for disseminated, sometimes fatal, infections in immunodeficient patients, especially following organ transplantation [9]. Several methods have been developed for typing M. abscessus sensu lato isolates. These have been used to investigate outbreaks or pseudo-outbreaks of respiratory and skin and soft-tissue infections and to determine if M. abscessus sensu lato could be transmitted between patients in CF populations [10,25,39,52,58]. The typing methods described include multilocus enzyme electrophoresis (MLEE) [17,61,63], pulsed-field gel electrophoresis (PFGE); [62,67], random amplified polymorphic DNA (RAPD) [29,66], enterobacterial repetitive intergenic consensus PCR (ERIC-PCR) [48], repetitive sequence-based PCR (rep-PCR) [65] and multispacer sequence typing [49]. The reference typing method is PFGE, which is technically demanding and fastidious [67]. MLST is an unambiguous sequence-based typing method that has been used to investigate the population structure of

different bacterial pathogens belonging to various genera including Neisseria, Escherichia, Burkholderia, Listeria, Streptococcus and Staphylococcus [8,11,12,36,37,43]. MLST involves sequencing of 350- to 600-bp internal fragments of several housekeeping genes (typically seven). Data are stored in a central database and are freely available to laboratories worldwide (http://www.mlst.net/). MLST is thus particularly suitable for local and global epidemiological studies because it provides data that can be easily compared between laboratories via the Internet without exchanging strains. Infections with M. abscessus sensu lato are widespread as they are found in all continents and associated with a large panel of pathologies. Nevertheless, nothing is known as to the association of particular isolates of this species to epidemics or the tropism of this species for different anatomic sites or for CF patients. The MLST method constitutes an ideal tool for approaching these questions, but no MLST scheme is thus far available for typing of mycobacterial species, including isolates of the M. abscessus sensu lato species. In the present study, we developed a MLST scheme and applied it to a large collection of M. abscessus sensu lato isolates of various clinical and geographical origins. The use of this MLST scheme, which is publicly available (www.pasteur. fr/mlst), will lead to a better understanding of the epidemiology of M. abscessus sensu lato throughout the world. 2. Materials and methods 2.1. Bacterial strains and culture conditions We studied 224 isolates of M. abscessus sensu lato, including 223 clinical isolates from 197 patients and one environmental isolate (Table S1); 120 isolates were from our recent multilocus sequence analysis (MLSA) study [35]. The isolates were recovered between 1994 and 2010 and originated in France (N ¼ 181), Germany (N ¼ 6), Switzerland (N ¼ 7) and Brazil (N ¼ 30). Twenty-two of the Brazilian isolates were related to an outbreak and had been previously typed by pulsedfield electrophoresis (Fig. S1) [30]. Among the 181 “French” isolates, 4 were from the Reunion Island (Indian Ocean, southeast Africa) and 12 from Guadeloupe (Caribbean Ocean, Central America). Only one isolate was taken per subject (N ¼ 181), except for 43 serial isolates taken from 16 CF patients (2 isolates from 9 patients, 3 isolates from 4 patients, 4 isolates from 2 and 5 isolates from 1 patient) (Table S1). The type strains M. abscessus (sensu stricto) CIP 104536T (¼ATCC

84

E. Macheras et al. / Research in Microbiology 165 (2014) 82e90

19977T), M. massiliense (CIP 108297T) and M. bolletii (CIP 108541T) were also included in the panel studied. One-hundred and ninety-three isolates were from respiratory specimens, 28 were from SST and bone disease, two were from other clinical origins (1 pericardial liquid, 1 lymph node) and one was from an environmental sample (sewer). The underlying disorder was known for 128 patients with respiratory isolates and was mostly CF (100 patients) and chronic obstructive pulmonary disease (COPD) (11 patients). All strains were stored at 70  C using cryopreservation beads (AES/bioMe´rieux, Craponne, France) and were grown on sheep-blood agar (bioMe´rieux, Craponne, France) at 37  C for 4 days prior to use. 2.2. Nomenclature used in the study It has recently been proposed that M. abscessus sensu lato be grouped into one species, M. abscessus, with two subspecies, M. abscessus subsp. abscessus and M. abscessus subsp. bolletii [31]. However, for reasons of simplicity, M. abscessus sensu lato isolates were classified into “M. abscessus” (for “M. abscessus sensu stricto), “M. massiliense” and “M. bolletii”. 2.3. PCR amplification and DNA sequencing Mycobacterial DNA was extracted using Tris-EDTA (SigmaeAldrich, Lyon, France), lysozyme (SigmaeAldrich, Lyon, France) and proteinase K (SigmaeAldrich, Lyon, France). Partial rpoB sequencing was performed as described previously with primers MYCOF1 and MYCOR2 [34]. rpoB sequences trimmed to 752 bp [1] were compared with corresponding rpoB sequences from the reference type strains (http://www.ncbi.nlm.nih.gov/). Fragments from the following housekeeping genes were amplified using the previously described sets of primers [35]: argH (argininosuccinate lyase), cya (adenylate cyclase), glpK (glycerol kinase), gnd (6phosphogluconate dehydrogenase), murC (UDP N-acetylmuramate-L-Ala ligase), pgm (phosphoglucomutase), pta (phosphate ace´tyltransfe´rase) and purH (phoshoribosylaminoimiazolcarboxylase ATPase subunit). Amplification was performed using 25 ml of ReddyMixÔ PCR Master Mix (Fisher Scientific SAS, Illkirch, France) and 1 ml of each primer (10 pmol). Dideoxy sequencing of the amplified gene fragments was performed on both strands using the Big Dye TerminatorÒ cycle sequencing kit (Applied Biosystems, Courtaboeuf, France). Sequencing products were purified by gel filtration (Bio-gel P100, Bio-Rad, Marnes-la-Coquette, France) and were run on a 3700 DNA Analyzer (Applied Biosystems, Courtaboeuf, France). The sequences were aligned and trimmed to defined start and end positions using BioEdit version 7.1.3 (http://www.mbio.ncsu.edu/bioedit/ bioedit.html). 2.4. Allele and sequence type assignment Sequence data were analyzed using the software Chromas Lite v2.01 (http://www.technelysium.com.au/chromas_lite.

html). For each MLST locus, each different sequence was assigned a distinct allele number regardless of whether the amino acid sequence was altered. For each isolate, the combination of alleles obtained at each locus defined its allelic profile or sequence type (ST). STs were numbered in order of their identification (ST-1, ST-2, etc.). The data have been deposited in the Institut Pasteur MLST web site at www. pasteur.fr/mlst. 2.5. Minimum spanning tree The most likely patterns of evolutionary descent in our collection were assessed using the minimum spanning tree (MST) algorithm [50]. STs were grouped into clonal complexes (CCs), which were defined as groups of STs matched at six of seven loci with at least one other ST of the group. A singleton ST corresponded to STs differing from every other ST at three or more of the seven loci. The statistical confidence in the assigned primary founders was determined by a bootstrap resampling procedure (1000 samples). An ancestral type (or group founder) was defined as the ST within the CC that differed from the highest number of other STs in the CC at only one locus out of seven, i.e. the ST with the highest number of single locus variants (SLVs). 2.6. Phylogenetic analysis The neighbor-joining [47] and neighbor-net algorithms [5] were applied to concatenated MLST sequences using MEGA 4.0.1 and SplitsTree 4.0 [22,56]. Codon positions were inframe and there was a total of 3576 bp in the final dataset. Bootstrap confidence values were based on 1000 replications. Evolutionary distances were computed using the maximum composite likelihood method [55]. Trees were drawn to scale with branch lengths representing the inferred evolutionary distances. Mean evolutionary diversity was determined using MEGA 4.0.1 [56]. The number of polymorphic sites, the index of association (IA) and the number of non-synonymous versus synonymous mutations (dN/dS ) were evaluated using the software START2 [24]. 3. Results 3.1. Selection of loci for MLST We recently studied the phylogeny of M. abscessus sensu lato using an MLSA approach based on the partial sequences of eight housekeeping genes: argH, cya, glpK, gnd, murC, pgm, pta and purH [35]. This first study, initiated by the finding that the same isolate, identified as M. abscessus by rpoB sequencing, could be identified as another species of the M. abscessus group by sequencing of another housekeeping gene, showed that lateral gene transfer is frequent among species of the M. abscessus group, and the use of a single gene may not be sufficient to correctly identify the isolate to the species level.

E. Macheras et al. / Research in Microbiology 165 (2014) 82e90

Taking advantage of this previous work, we sequenced the same eight genes in 223 clinical isolates, one environmental isolate and the three type strains M. abscessus CIP 104536T, M. massiliense CIP 108297T and M. bolletii CIP 108541T. The number of allelic profiles obtained with the set of eight genes was unchanged regardless of whether pgm sequences were included or not (data not shown). We therefore only included seven loci in the final MLST scheme: argH, cya, glpK, gnd, murC, pta and purH (Fig. 1). 3.2. Nucleotide sequence variation for each of the seven MLST loci Between 18 (murC ) and 26 (cya and pta) alleles were found for each locus, with an average of 23 alleles per locus (Table 1). No deletions or insertions were observed. All differences between alleles were due to point mutations at 25 ( pta) to 43 (argH ) polymorphic sites, with mean evolutionary diversity values ranging from 0.01357 to 0.02755. We identified a total of 239 single polymorphic (SNP) sites (6.68% of the 3576-bp concatenated sequence) among the seven loci. Most SNPs were silent, with ratios of non-synonymous substitutions and synonymous mutations ranging from 0.0029 (cya) to 0.1579 (argH ) (mean 0.0603). 3.3. STs and analysis of population structure MLST analysis of the 224 isolates and 3 type strains identified 100 different STs (Table S2). Neighbor-joining analysis of concatenated sequences (3576 bp) showed that these 100 STs were distributed into three distinct groups, each containing the ST assigned to one of the three type strains (M. abscessus CIP 104536T, ST 1; M. massiliense CIP 108297T, ST 37; M. bolletii CIP 108541T, ST 38) (Fig. 2). We could thus individualize a “M. abscessus” group (44 STs), a “M. massiliense” group (31 STs) and a “M. bolletii” group (25 STs). Consistent with our previous MLSA analysis [35], a number of STs were close to the main branch point. The most representative of these “borderline” STs were ST9 and ST57 (“M. massiliense” group), ST64 (“M. bolletii” group) and ST76 0

glpK cya

pta rpoB

5,067,172 bp CIP 104536T

purH

85

Table 1 Nucleotide sequence variation for each locus of the final MLST scheme. MLST locus

Sequence length (bp)

No. of alleles

No. (%) of polymorphic sites

Mean evolutionary diversity

dN/dS

argH cya glpK gnd murC pta purH

480 510 534 480 537 486 549

23 26 19 23 18 26 25

43 31 27 38 41 25 34

0.02755 0.01873 0.01311 0.02358 0.02339 0.01357 0.01746

0.1579 0.0029 0.0402 0.0267 0.1226 0.0127 0.0553

(8.95) (6.08) (5.05) (7.91) (7.44) (5.14) (6.19)

(“M. abscessus” group). The analysis of their allelic profile showed that these STs displayed inconsistent alleles at one or two loci, most likely due to intergroup exchange within M. abscessus sensu lato. 3.4. Clonal lineages Minimum spanning tree analysis showed that 49 of the 100 identified STs formed 11 CCs, 6 belonging to the “M. abscessus” group and 5 to the “M. massiliense” group (Fig. 3). The largest “M. abscessus” CC, CC1, comprised 14 STs, with ST63 being the most likely ancestor. However, it was not this complex, but CC5, that included ST1 (the ST of M. abscessus CIP 104536T). CC2 and CC3, the largest CCs from the “M. massiliense” group, consisted of 9 and 5 STs, respectively, the most likely ancestor of CC2 being ST69 and that of CC3 being ST23. CC2 also included ST37, the ST of M. massiliense CIP 108297T. The proportion of singleton STs was 39% (17/44) for the “M. abscessus” group, 29% (9/31) for the “M. massiliense” group and 100% (25/25) for the “M. bolletii” group (including ST-38, the ST of M. bolletii CIP 108541T). 3.5. Recombination within M. abscessus sensu lato The split graph obtained with the seven MLST loci displayed a network-like rather than tree-like structure, implying a certain degree of lateral exchanges between the “M. abscessus”, “M. massiliense” and “M. bolletii” groups (Fig. 4). For example, STs 9 and 57 were the only “M. massiliense” STs to display the argH allele 4, which was the most frequent allele in the “M. bolletii” group (18 out of the 25 “M. bolletii” STs). Furthermore, ST9 displayed the glpK allele 1, which was the most frequently encountered allele in the “M. abscessus” group (38 out of 44 “M. abscessus” STs). Most genetic exchange events were found between the “M. abscessus” and “M. massiliense” groups and between the “M. massiliense” and “M. bolletii” groups. Any of the seven MLST loci as well as rpoB were involved, suggesting that these events may occur in any part of the genome.

murC gnd

argH

Fig. 1. Genomic location of housekeeping genes included in the final MLST scheme. Location of rpoB is also indicated.

3.6. STs and rpoB-based identification We compared MLST and rpoB sequences from each of the 224 isolates. The results of this analysis were inconsistent for

86

E. Macheras et al. / Research in Microbiology 165 (2014) 82e90

Fig. 2. Phylogenetic relationships of the 100 STs identified in this study. The tree was constructed from concatenated MLST gene sequences of the 224 isolates and the three M. abscessus sensu lato type strains using the neighbor-joining method. Bootstrap support values (%) are indicated for each node. Type strains: *M. abscessus CIP 104536T; **M. massiliense CIP 108297T; ***M. bolletii CIP 108541T.

36 8

CC11

7

9

CC9

7

35

57 2 2

34

73

67 71

89

42

80 23

CC1 93 93

23

8383 41

17

16 5 5

22

107 107 47 47 79 79

CC8

22

20

74 81

11

87 11

1

24

99

CC2

104 75 75

CC3

64

10

86

12

84 84

85

30

43

90

46

43

50 50

32

CC10 70

92

102

4

68

102 52

66

105 106

1

CC5

21

37

51

CC6

15

37

CC7 91

72 26

103

69

38 38

49

53

18

33 33

27

69

44

82

40

94

31

18

88

19 63

77 96 95 14 96 65 6565 14

29

97

33

6

76

25

48

61 54

78

101 28

CC4

98

56

13

55

Fig. 3. Minimum spanning tree analysis. Each of the 100 STs is represented by a circle. Clonal complexes CC1 to CC11 correspond to groups of connected STs. Central circles are group founders. : M. abscessus sensu stricto (N ¼ 101). :“M. massiliense” group (N ¼ 90). : “M. bolletii” group (N ¼ 36). Number of mismatches: 1 ,2 ,3 ,4 .

E. Macheras et al. / Research in Microbiology 165 (2014) 82e90

87

Fig. 4. SplitTree analysis. The tree was constructed from concatenated MLST sequences of the 224 isolates and the three M. abscessus sensu lato type strains using the neighbor-net method. Type strains: *M. abscessus CIP 104536T; **M. massiliense CIP 108297T; *** M. bolletii CIP 108541T.

15 (6.7%) isolates (Table 2). The most frequent inconsistency was an “M. massiliense” MLST sequence with an “M. abscessus” rpoB sequence (N ¼ 9) or an “M. bolletii” rpoB sequence (N ¼ 3). Other inconsistencies like an “M. abscessus” MLST sequence with an “M. massiliense” rpoB sequence and an “M. bolletii” MLST sequence with an “M. abscessus” rpoB sequence were less frequent. 3.7. Clinical and geographical aspects When including only one isolate per patient and only one isolate out of the 22 Brazilian outbreak isolates, we found that ST1 (18 patients), ST2 (14 patients) and ST23 (9 patients) were the most prevalent STs in our study (Table S2). ST1 and ST23 comprised both respiratory and SST isolates, whereas Table 2 Strains with evidence of rpoB lateral transfer. ST

Isolate number

ST9 ST15 ST28 ST34 ST36 ST57 ST65 ST66 ST80 ST83 ST88

19 80 47 63, 91, 124, 154, 161 31 113 123 125 165 179 192

a

Group assignment>a MLST sequence

rpoB sequence

“M.massiliense” “M. massiliense” “M. massiliense” “M. massiliense” “M. massiliense” “M. massiliense” “M. abscessus” “M. bolletii” “M. massiliense” “M. abscessus” “M. massiliense”

“M. “M. “M. “M. “M. “M. “M. “M. “M. “M. “M.

bolletii” abscessus” bolletii” abscessus” abscessus” bolletii” massiliense” abscessus” abscessus” massiliense” abscessus”

Among the “M. abscessus”, “M. massiliense” and “M. bolletii” groups.

ST2 comprised only respiratory isolates, mostly from CF patients. ST7 and ST49 were the only other STs that included isolates from SST and bone disease. Six out of the 12 most prevalent STs (3 patients) included isolates from different countries: the most geographically widespread were ST1 (France, Brazil, USA, Switzerland) and ST23 (Brazil, France, Germany, Switzerland). Six of the most prevalent STs were isolated only from France: ST2, ST12, ST20, ST33, ST41 and ST49. With the exception of ST52, the most prevalent STs were recovered in different years (Table S1). 4. Discussion During the last decade, the development of MLST has provided a robust tool for studying bacterial populations that belong to the main human pathogens (e.g., Neisseria meningitidis, group B streptococcus, Staphylococcus aureus, Campylobacter jejuni, Listeria monocytogenes) [11,12,32,36,43,45]. There are numerous MLST databases providing information on thousands of isolates worldwide. Approaches based on sequencing of housekeeping genes have already been used in phylogenetic studies of environmental mycobacteria [19], species of the M. tuberculosis complex [38,54] and M. prototuberculosis [23]. In spite of the fact that techniques for molecular typing of M. tuberculosis were much more frequently based on analysis of insertion sequences or repetitive sequences, since isolates of M. tuberculosis show high clonality, the multigenic sequence approach and whole genome sequencing seem to be gaining interest for these species [54]. Here we describe the development of an MLST typing scheme for members of the M. abscessus complex,

88

E. Macheras et al. / Research in Microbiology 165 (2014) 82e90

which is the first MLST scheme applied to a large number of mycobacterial isolates. This scheme is based on partial sequencing of seven housekeeping genes, representing a total sequence of 3576 bp. The percentage of polymorphic sites was around 6.5% (range, 4.7e8.95%), indicating a relatively high level of genetic diversity. Similar data have been reported for L. monocytogenes, which is also an opportunistic human pathogen mainly present in soil and aquatic environments [45]. The dN/dS ratios for the seven loci were all much less than 1, which is consistent with most of the variations being selectively neutral. In the collection of strains that we tested, both French strains (w75% of strains) and strains isolated from CF patients (w75% of strains of respiratory origin with known case history) were overrepresented. It should also be noted that almost all of the isolates were from clinical samples because, in contrast to related RGMs such as M. chelonae, M. abscessus is only rarely found in environmental samples [14,15,44,57]. The genetic diversity of M. abscessus sensu lato may therefore, in fact, be higher than that determined in our analysis. We identified 100 different STs among the 227 isolates included in our study which formed three groups corresponding to the groups “M. abscessus”, “M. massiliense” and “M. bolletii”. The number of STs was higher in the “M. abscessus” group (44 STs) than in “M. massiliense” (31 STs) and “M. bolletii” (25 STs). However, considering the number of isolates tested in each of these groups and only accounting for those isolates that were epidemiologically unrelated to each other (N ¼ 184), the proportion of different STs was similar for “M. abscessus” (54%) and “M. massiliense” (44%), whereas the proportion was much higher for “M. bolletii” (81%). Analysis of the MST confirmed this particular population structure for “M. bolletii” isolates, as all of the STs were singletons. This could be due to a higher rate of recombination or less substantial clonal expansion of the “M. bolletii” isolates. Nevertheless, a larger number of strains from this group need to be studied to confirm these findings. Our study shows the importance of horizontal gene transfer between the three groups, especially between “M. abscessus” and “M. massiliense” and between “M. massiliense” and “M. bolletii”. We were able to demonstrate this type of process for the MLST genes as well as for the rpoB gene. This confirms the unreliability of methods using a single gene to determine whether a M. abscessus sensu lato strain belongs to “M. abscessus”, “M. massiliense” or “M. bolletii” [34]. In certain cases, we were able to “trace” the genetic transfer event involving rpoB. For example, within the CC3 complex of the “M. massiliense” group, ST23 and other closely related STs (ST42, ST48, ST97) have rpoB sequences consistent with belonging to the “M. massiliense” group, whereas STs 34 and 80, belonging to the very close CC9, have a rpoB sequence typical of the “M. abscessus” group, indicating that the members of CC9 (ST34 and ST80) may have evolved by acquisition of the rpoB gene from the “M. abscessus” group by members of the CC3 complex. A recent study on M. tuberculosis using MLST and whole genome sequencing has shown that horizontal gene transfer is prominent between members of

the M. tuberculosis complex [54] and several studies have highlighted the importance of horizontal transfer in the evolution of microbial genomes [6,23,60]. Our study was biased by the collection of strains that we analyzed, with a large proportion of French isolates and isolates obtained from CF patients. This shortcoming should be corrected in the near future by the availability of data for a larger number of strains isolated worldwide and collected in the MLST website (http://www.pasteur.fr/recherche/genopole/ PF8/mlst/Myco-abscessus.html). However, despite these problems, our study already provides very important information about the prevalence of certain STs and their geographical distribution. ST1, the most frequent “M. abscessus” ST, and ST23, the most frequent “M. massiliense” ST, are found in several western European countries and on the American continent. This is also the case for ST34, a “M. massiliense” ST which is less prevalent than ST23. For the other STs, we currently have studied only European strains, although they have been isolated in at least two different countries. The same applies to the “M. abscessus” (ST18) and “M. massiliense” (ST7) groups, as well as the “M. bolletii” (ST52, ST75) group, although this group is substantially underrepresented in our collection. It is premature to conclude as to whether certain STs are preferentially associated with particular infections, because there are insufficient isolates per ST in our collection to establish any correlation between ST and disease. However, we found respiratory and SST isolates in the two most prevalent STs, ST1 and ST23, as well as in other STs such as ST7 and ST49. Therefore, most of the STs for which we have large numbers of strains include isolates from both respiratory and SST infections, which does not suggest any specialization. This is consistent with previous PFGE typing studies showing that respiratory and SST strains may share the same pulse-type [59,65]. Several studies using different typing methods have shown a wide diversity among isolates obtained from populations of CF patients, and even if patient-to-patient transmission cannot formally be proven, it cannot be excluded and in same cases even seems probable [3,6,13,25,27,51]. Our study confirms the wide diversity of isolates from CF cases; the 127 isolates from CF patients were grouped into 63 different STs belonging to three groups: “M. abscessus” (30 STs), “M. massiliense” (19 STs) and “M. bolletii” (14 STs). Indeed, 48 of 63 STs included an isolate from only one CF patient; however, certain STs were found in several different patients, particularly ST1 (9 patients), ST2 (11 patients), ST23 (4 patients), ST33 (4 patients) and ST41 (3 patients). The most likely explanation is that these STs are widespread in the environment, consistent with the broad geographic distribution of some of them, notably ST1 and ST23, and that patients are infected from different sources containing the same ST. Nevertheless, in some cases, nosocomial or patient-to-patient transmission cannot be completely excluded. This applies, in particular, to several cases involving ST2 in pediatric CF centers in Paris. A study is currently in progress to resolve this issue. One of the major strengths of the MLST approach is to identify epidemic clones and evaluate their spread throughout

E. Macheras et al. / Research in Microbiology 165 (2014) 82e90

the world. Several bacterial clones with worldwide distribution have thus been highlighted in recent years, particularly the methicillin-resistant ST8 clone of S. aureus (USA300 strain) [41] and the ST17 clone of group B streptococcus [53]. For the first time, we have demonstrated the involvement of an ST23 isolate from the “M. massiliense” group in a large postsurgical procedure outbreak of SST infection in Brazil [30] This ST, which is present in Europe and the Americas, may be an epidemic clone with worldwide distribution. It would be of interest to determine the ST of isolates responsible for the large epidemics of SST infection involving M. abscessus sensu lato that have been reported in the last decade [7,18,59,64]. This would reveal whether there exists an epidemic “M. massiliense” ST23 clone distributed worldwide and whether there exist other epidemic clones in this group and in the “M. abscessus” and “M. bolletii” groups. Over recent years, there has been a large increase in the number of cases of M. abscessus sensu lato infections worldwide, and the reasons for this are poorly understood. The MLST typing scheme that we propose should enable a better understanding of the epidemiology of this microorganism and help to establish preventative strategies for the most vulnerable populations, especially CF patients. Acknowledgments We thank the association “Vaincre la Mucoviscidose” for financial support of this work.

Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.resmic.2013.12.003. References [1] Adekambi T, Colson P, Drancourt M. rpoB-based identification of nonpigmented and late-pigmenting rapidly growing mycobacteria. J Clin Microbiol 2003;41:5699e708. [2] Appelgren P, Farnebo F, Dotevall L, Studahl M, Jonsson B, Petrini B. Lateonset posttraumatic skin and soft-tissue infections caused by rapid-growing mycobacteria in tsunami survivors. Clin Infect Dis 2008;47:e11e16. [3] Bange FC, Brown BA, Smaczny C, Wallace Jr RJ, Bottger EC. Lack of transmission of Mycobacterium abscessus among patients with cystic fibrosis attending a single clinic. Clin Infect Dis 2001;32:1648e50. [4] Bechara C, Macheras E, Heym B, Pages A, Auffret N. Mycobacterium abscessus skin infection after tattooing: first case report and review of the literature. Dermatology 2010;221:1e4. [5] Bryant D, Moulton V. Neighbor-net: an agglomerative method for the construction of phylogenetic networks. Mol Biol Evol 2004;21:255e65. [6] Bryant JM, Grogono DM, Greaves D, Foweraker J, Roddick I, Inns T, et al. Whole-genome sequencing to identify transmission of Mycobacterium abscessus between patients with cystic fibrosis: a retrospective cohort study. Lancet 2013;381:1551e60. [7] Cardoso AM, Martins de Sousa E, Viana-Niero C, Bonfim de Bortoli F, Pereira das Neves ZC, Leao SC, et al. Emergence of nosocomial Mycobacterium massiliense infection in Goias, Brazil. Microbes Infect 2008;10:1552e7.

89

[8] Chenal-Francisque V, Lopez J, Cantinelli T, Caro V, Tran C, Leclercq A, et al. Worldwide distribution of major clones of Listeria monocytogenes. Emerg Infect Dis 2011;17:1110e2. [9] Chernenko SM, Humar A, Hutcheon M, Chow CW, Chaparro C, Keshavjee S, et al. Mycobacterium abscessus infections in lung transplant recipients: the international experience. J Heart Lung Transplant 2006;25:1447e55. [10] Dytoc MT, Honish L, Shandro C, Ting PT, Chui L, Fiorillo L, et al. Clinical, microbiological and epidemiological findings of an outbreak of Mycobacterium abscessus hand-and-foot disease. Diagn Microbiol Infect Dis 2005;53:39e45. [11] Enright MC, Day NP, Davies CE, Peacock SJ, Spratt BG. Multilocus sequence typing for characterization of methicillin-resistant and methicillin-susceptible clones of Staphylococcus aureus. J Clin Microbiol 2000;38:1008e15. [12] Enright MC, Spratt BG, Kalia A, Cross JH, Bessen DE. Multilocus sequence typing of Streptococcus pyogenes and the relationships between emm type and clone. Infect Immun 2001;69:2416e27. [13] Esther Jr CR, Henry MM, Molina PL, Leigh MW. Nontuberculous mycobacterial infection in young children with cystic fibrosis. Pediatr Pulmonol 2005;40:39e44. [14] Falkinham 3rd JO. Nontuberculous mycobacteria in the environment. Clin Chest Med 2002;23:529e51. [15] Feazel LM, Baumgartner LK, Peterson KL, Frank DN, Harris JK, Pace NR. Opportunistic pathogens enriched in showerhead biofilms. Proc Nat Acad Sci U S A 2009;106:16393e9. [16] Fisher EJ, Gloster Jr HM. Infection with Mycobacterium abscessus after Mohs micrographic surgery in an immunocompetent patient. Dermatol Surg 2005;31:790e4. [17] Galil K, Miller LA, Yakrus MA, Wallace Jr RJ, Mosley DG, England B, et al. Abscesses due to Mycobacterium abscessus linked to injection of unapproved alternative medication. Emerg Infect Dis 1999;5:681e7. [18] Garcia-Navarro X, Barnadas MA, Dalmau J, Coll P, Gurgui M, Alomar A. Mycobacterium abscessus infection secondary to mesotherapy. Clin Exp Dermatol 2008;33:658e9. [19] Gomila M, Ramirez A, Lalucat J. Diversity of environmental Mycobacterium isolates from hemodialysis water as shown by a multigene sequencing approach. Appl Environ Microbiol 2007;73:3787e97. [20] Griffith DE, Aksamit T, Brown-Elliott BA, Catanzaro A, Daley C, Gordin F, et al. An official ATS/IDSA statement: diagnosis, treatment and prevention of nontuberculous mycobacterial diseases. Am J Respir Crit Care Med 2007;175:367e416. [21] Griffith DE. Emergence of nontuberculous mycobacteria as pathogens in cystic fibrosis. Am J Respir Crit Care Med 2003;167:810e2. [22] Huson DH, Bryant D. Application of phylogenetic networks in evolutionary studies. Mol Biol Evol 2006;23:254e67. [23] Gutierrez MC, Brisse S, Brosch R, Fabre M, Omaı¨s B, Marmiesse M, et al. Ancient origin and gene mosaicism of the progenitor of Mycobacterium tuberculosis. PLoS Pathog 2005;1(e5). [24] Jolley KA, Feil EJ, Chan MS, Maiden MC. Sequence type analysis and recombinational tests (START). Bioinformatics 2001;17:1230e1. [25] Jonsson BE, Gilljam M, Lindblad A, Ridell M, Wold AE, WelinderOlsson C. Molecular epidemiology of Mycobacterium abscessus, with focus on cystic fibrosis. J Clin Microbiol 2007;45:1497e504. [26] Koh WJ, Jeon K, Lee NY, Kim BJ, Kook YH, Lee SH, et al. Clinical significance of differentiation of Mycobacterium massiliense from Mycobacterium abscessus. Am J Respir Crit Care Med 2011;183:405e10. [27] Konig B, Tammer I, Sollich V, Konig W. Intra- and interpatient variability of the hsp65 and 16S-23S intergenic gene region in Mycobacterium abscessus strains from patients with cystic fibrosis. J Clin Microbiol 2005;43:3500e3. [28] Kusunoki S, Ezaki T. Proposal of Mycobacterium peregrinum sp. nov., nom. rev. and elevation of Mycobacterium chelonae subsp. abscessus (Kubica et al.) to species status: Mycobacterium abscessus comb. nov. Int J Syst Bacteriol 1992;42:240e5. [29] Lai KK, Brown BA, Westerling JA, Fontecchio SA, Zhang Y, Wallace Jr RJ. Long-term laboratory contamination by Mycobacterium

90

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37] [38]

[39]

[40]

[41]

[42]

[43]

[44] [45]

[46]

[47] [48]

E. Macheras et al. / Research in Microbiology 165 (2014) 82e90 abscessus resulting in two pseudo-outbreaks: recognition with use of random amplified polymorphic DNA (RAPD) polymerase chain reaction. Clin Infect Dis 1998;27:169e75. Leao SC, Viana-Niero C, Matsumoto CK, Lima KV, Lopes ML, Palaci M, et al. Epidemic of surgical-site infections by a single clone of rapidly growing mycobacteria in Brazil. Future Microbiol 2009;5:971e80. Leao SC, Tortoli E, Euzeby JP, Garcia MJ. Proposal that Mycobacterium massiliense and Mycobacterium bolletii be united and reclassified as Mycobacterium abscessus subsp. bolletii comb. nov., designation of Mycobacterium abscessus subsp. abscessus subsp. nov. and emended description of Mycobacterium abscessus. Int J Syst Evol Microbiol 2011;61:2311e3. Levesque S, Michaud S, Arbeit RD, Frost EH. High-resolution melting system to perform multilocus sequence typing of Campylobacter jejuni. PLoS One 2011;6:e16167. Levy I, Grisaru-Soen G, Lerner-Geva L, Kerem E, Blau H, Bentur L, et al. Multicenter cross-sectional study of nontuberculous mycobacterial infections among cystic fibrosis patients, Israel. Emerg Infect Dis 2008;14:378e84. Macheras E, Roux A-L, Ripoll F, Sivadon-Tardy V, Gutierrez C, Gaillard J-L, et al. Inaccuracy of single-target sequencing for discriminating species of the Mycobacterium abscessus group. J Clin Microbiol 2009;47:2596e600. Macheras E, Roux A-L, Bastian S, Leao SC, Palaci M, Sivadon-Tardy V, et al. Multilocus sequence analysis and rpoB sequencing of Mycobacterium abscessus (sensu lato) strains. J Clin Microbiol 2011;49:491e9. Maiden MC, Bygraves JA, Feil E, Morelli G, Russell JE, Urwin R, et al. Multilocus sequence typing: a portable approach to the identification of clones within populations of pathogenic microorganisms. Proc Natl Acad Sci U S A 1998;95:3140e5. Maiden M. Multilocus sequence typing of bacteria. Annu Rev Microbiol 2006;60:561e88. Mestre O, Luo T, Dos Vultos T, Kremer K, Murray A, Namouchi A, et al. Phylogeny of Mycobacterium tuberculosis Beijing strains constructed from polymorphisms in genes involved in DNA replication, recombination and repair. PLoS One 2011;6:e16020. Newman MI, Camberos AE, Ascherman J. Mycobacterium abscessus outbreak in US patients linked to offshore surgicenter. Ann Plast Surg 2005;55:107e10. Olivier KN, Weber DJ, Wallace Jr RJ, Faiz AR, Lee JH, Zhang Y, et al. Nontuberculous mycobacteria. I: multicenter prevalence study in cystic fibrosis. Am J Respir Crit Care Med 2003;167:828e34. Pasquale TR, Jabrocki B, Salstrom SJ, Wiemken TL, Peyrani P, Haque NZ, et al. Emergence of methicillin-resistant Staphylococcus aureus USA300 genotype as a major cause of late-onset nosocomial pneumonia in intensive care patients in the USA. Int J Infect Dis 2013;17:e398e403. Pawlik A, Garnier G, Orgeur M, Tong P, Lohan A, Le Chevalier F, et al. Identification and characterization of the genetic changes responsible for the characteristic smooth-to-rough morphotype alterations of clinically persistent Mycobacterium abscessus. Mol Microbiol 2013;90:612e29. Pope CE, Short P, Carter PE. Species distribution of Burkholderia cepacia complex isolates in cystic fibrosis and non-cystic fibrosis patients in New Zealand. J Cyst Fibros 2010;9:442e6. Primm TP, Lucero CA, Falkinham 3rd JO. Health impacts of environmental mycobacteria. Clin Microbiol Rev 2004;17:98e106. Ragon M, Wirth T, Hollandt F, Lavenir R, Lecuit M, Le Monnier A, et al. A new perspective on Listeria monocytogenes evolution. PLoS Pathog 2008;4:e1000146. Roux AL, Catherinot E, Ripoll F, Soismier N, Macheras E, Ravilly S, et al. Multicenter study of prevalence of nontuberculous mycobacteria in patients with cystic fibrosis in France. J Clin Microbiol 2009;47:4124e8. Saitou N, Nei M. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 1987;4:406e25. Sampaio JL, Viana-Niero C, de Freitas D, Hofling-Lima AL, Leao SC. Enterobacterial repetitive intergenic consensus PCR is a useful tool for

[49] [50]

[51]

[52]

[53]

[54]

[55]

[56]

[57]

[58] [59]

[60]

[61]

[62]

[63]

[64]

[65]

[66]

[67]

typing Mycobacterium chelonae and Mycobacterium abscessus isolates. Diagn Microbiol Infect Dis 2006;55:107e18. Sassi M, Ben Kahla I, Drancourt M. Mycobacterium abscessus multispacer sequence typing. BMC Microbiol 2013;13:3. Schouls LM, van der Heide HG, Vauterin L, Vauterin P, Mooi FR. Multiple-locus variable-number tandem repeat analysis of Dutch Bordetella pertussis strains reveals rapid genetic changes with clonal expansion during the late 1990s. J Bacteriol 2004;186:5496e505. Sermet-Gaudelus I, Le Bourgeois M, Pierre-Audigier C, Offredo C, Guillemot D, Halley S, et al. Mycobacterium abscessus and children with cystic fibrosis. Emerg Infect Dis 2003;9:1587e91. Song JY, Sohn JW, Jeong HW, Cheong HJ, Kim WJ, Kim MJ. An outbreak of post-acupuncture cutaneous infection due to Mycobacterium abscessus. BMC Infect Dis 2006;6:6. Sørensen UB, Poulsen K, Ghezzo C, Margarit I, Kilian M. Emergence and global dissemination of host-specific Streptococcus agalactiae clones. MBio 2010;1:e00178e80. Supply P, Marceau M, Mangenot S, Roche D, Rouanet C, Khanna V, et al. Genomic analysis of smooth tubercle bacilli provides insight into ancestry and pathoadaptation of Mycobacterium tuberculosis. Nature Genet 2013;45:172e81. Tamura K, Nei M, Kumar S. Prospects for inferring very large phylogenies by using the neighbor-joining method. Proc Natl Acad Sci U S A 2004;101:11030e5. Tamura K, Dudley J, Nei M, Kumar S. MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol Biol Evol 2007;24:1596e9. Thomson R, Tolson C, Carter R, Coulter C, Huygens F, Hargreaves M. Isolation of nontuberculous mycobacteria (NTM) from household water and shower arosols in patients with pulmonary disease caused by NTM. J Clin Microbiol 2013;51:3006e11. Toy BR, Frank PJ. Outbreak of Mycobacterium abscessus infection after soft tissue augmentation. Dermatol Surg 2003;29:971e3. Viana-Niero C, Lima KV, Lopes ML, Rabello MC, Marsola LR, Brilhante VC, et al. Molecular characterization of Mycobacterium massiliense and Mycobacterium bolletii in isolates collected from outbreaks of infections after laparoscopic surgeries and cosmetic procedures. J Clin Microbiol 2008;46:850e5. Veyrier F, Pletzer D, Turenne C, Behr MA. Phylogenetic detection of horizontal gene transfer during the step-wise genesis of Mycobacterium tuberculosis. BMC Evol Biol 2009;9:196. Wallace Jr RJ, Musser JM, Hull S, Silcox VA, Steele LC, Forrester GD, et al. Diversity and sources of rapidly growing mycobacteria associated with infections following cardiac surgery. J Infect Dis 1989;159:708e16. Wallace Jr RJ, Zhang Y, Brown BA, Fraser V, Mazurek GH, Maloney S. DNA large restriction fragment patterns of sporadic and epidemic nosocomial strains of Mycobacterium chelonae and Mycobacterium abscessus. J Clin Microbiol 1993;31:2697e701. Yakrus MA, Hernandez SM, Floyd MM, Sikes D, Butler WR, Metchock B. Comparison of methods for identification of Mycobacterium abscessus and M. chelonae isolates. J Clin Microbiol 2001;39:4103e10. Yuan J, Liu Y, Yang Z, Cai Y, Deng Z, Qin P, et al. Mycobacterium abscessus post-injection abscesses from extrinsic contamination of multiple-dose bottles of normal saline in a rural clinic. Int J Infect Dis 2009;13:537e42. Zelazny AM, Root JM, Shea YR, Colombo RE, Shamputa IC, Stock F, et al. Cohort study of molecular identification and typing of Mycobacterium abscessus, Mycobacterium massiliense and Mycobacterium bolletii. J Clin Microbiol 2009;47:1985e95. Zhang Y, Rajagopalan M, Brown BA, Wallace Jr RJ. Randomly amplified polymorphic DNA PCR for comparison of Mycobacterium abscessus strains from nosocomial outbreaks. J Clin Microbiol 1997;35:3132e9. Zhang Y, Yakrus MA, Graviss EA, Williams-Bouyer N, Turenne C, Kabani A, et al. Pulsed-field gel electrophoresis study of Mycobacterium abscessus isolates previously affected by DNA degradation. J Clin Microbiol 2004;42:5582e7.