1

Survey of 150 strains belonging to the Mycobacterium terrae complex and description of Mycobacterium engbaekii sp. nov., Mycobacterium heraklionense sp. nov. and Mycobacterium longobardum sp. nov. Tortoli Enrico,1 Gitti Zoe,2 Klenk Hans-Peter,3 Lauria Stefania,4 Mannino Roberta,4 Mantegani Paola,1 Mariottini Alessandro,5 Ioannis Neonakis.2 1 Emerging Bacterial Pathogens, San Raffaele Scientific Institute, Milan, Italy 2 Department of Clinical Bacteriology, Parasitology, Zoonoses and Geographical Medicine, University Hospital of Heraklion, Heraklion, Greece 3 DSMZ, German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany 4 Microbiology and Virology Laboratory, Careggi University Hospital, Florence, Italy 5 Cytogenetics and Genetics Unit, Careggi University Hospital, Florence, Italy Corresponding author: Enrico Tortoli, Emerging Bacterial Pathogens, San Raffaele Scientific Institute, via Olgettina 58, 20132 Milan, Italy; E-mail: e.tortoli@libero.it Running title: Mycobacterium terrae complex Keywords: Mycobacterium terrae complex, Mycobacterium engbaekii, Mycobacterium heraklionense, Mycobacterium longobardum, sp. nov. The GenBank accession number of the almost complete 16S rRNA sequences of M. engbaekii, M. heraklionense and M. longobardum are AF480577, GU084182 and JN571166 respectively. In the 16S rRNA gene the accession numbers of the different sequevars are JN571167-9 for M. arupense, JN571172 for M. engbaekii, JN571173 for M. hiberniae, JN571176-9 for M. kumamotonense, JN571171 for M. nonchromogenicum, JN571174-5 for M. senuense, JN571180-2 for M. terrae and JN571183-5 for other unassigned strains belonging to the M. terrae complex. In the hsp65 gene the accession numbers of the different sequevars are FJ263631 and JN571186-90 for M. arupense, JN571194-7 for M. engbaekii, JN571191-2 for M. heraklionense, JN571198 for M. hiberniae, JN571202-3 for M. kumamotonense, JN571199 for M. longobardum, JN571193 for M. nonchromogenicum, JN571200-1 for M. senuense, JN571204-7 for M. terrae and JN571208-14 for other unassigned strains belonging to the M. terrae complex. In the rpoB gene the accession numbers of the different sequevars are JN571215-29 for M. arupense, JN571242-5 for M. engbaekii, JN571230-6 for M. heraklionense, JN571246 for M. hiberniae, JN571251-58 and JN571290 for M. kumamotonense, JN571247-8 for M. longobardum, JN571237-41 for M. nonchromogenicum, JN571249-50 for M. senuense, JN571261-6 for M. terrae and JN571170 and JN571267-73 for other unassigned strains belonging to the M. terrae complex.

IJSEM Papers in Press. Published March 23, 2012 as doi:10.1099/ijs.0.038737-0

2

A thorough phenotypic and genotypic analysis of 150 strains belonging to the 1

Mycobacterium terrae complex resulted in the identification of a number of previously 2

unreported sequevars (sqvs.) within the species known to belong to the complex. In the 3

species Mycobacterium arupense were detected 3 sqvs. in the 16S rRNA gene, 6 sqvs. in 4

the hsp65 gene and 15 sqvs. in the rpoB gene; in Mycobacterium senuense 2 sqvs. were 5

present in each of the three genetic regions; in Mycobacterium kumamotonense 4, 2 and 9 6

sqvs. respectively and in M. terrae 3, 4 and 6 sqvs. respectively were found. The 7

extraneousness of Mycobacterium triviale with the M. terrae complex was confirmed. 8

The limited utility of biochemical tests and of mycolic acids analyses for the 9

differentiation of the members of M. terrae complex was confirmed as well. The survey 10

allowed furthermore to recognize three previously undescribed species characterized by 11

unique sequences in 16S rRNA, hsp65 and rpoB genes. Mycobacterium engbaekii sp. 12

nov., a species proposed 40 years ago but never validly published, is characterized by 13

pink photochromogenic pigmentation and rapid growth; phylogenetically it is related to 14

Mycobacterium hiberniae. The type strains of this species, of which eight strains were 15

investigated, is ATCC27353=DSM45694. A cluster of 24 strains was the basis for the 16

description of Mycobacterium heraklionense sp. nov. which has intermediate growth rate 17

and is unpigmented, the nitrate reductase activity is typically strong. Closely related to M. 18

arupense in the 16S rRNA gene, M. heraklionense is clearly differentiable from the latter 19

species in the other genetic regions. The type strain is NCTC13432=LMG2473= 20

CECT7509. Mycobacterium longobardum sp. nov., represented in the study from seven 21

strains, is characterized by a unique phylogenetic location within the M. terrae complex, 22

evidently divergent from any other species. The type strain is DSM45394=CCUG58460. 23

3

The Mycobacterium terrae complex (MTC) is a group, within the genus Mycobacterium, 24

created in the 1970s to gather Mycobacterium nonchromogenicum, Mycobacterium terrae 25

and Mycobacterium triviale, three species not differentiable with methods (biochemical 26

and cultural tests) available at that time. M. nonchromogenicum described in 1965 27

(Tsukamura, 1965), M. terrae described in 1966 (Wayne, 1966) and M. triviale described 28

in 1970 (Kubica et al., 1970) share important cultural features including intermediate 29

growth rate (from 5 to 15 days are required to develop clearly visible colonies from 30

diluted inocula on solid media) and lack of pigmentation. 31

In the early 1990s, the detection of a unique genetic signature: the presence of a two-32

nucleotide insertion in the helix 18 of the 16S rRNA gene (Kirschner et al., 33

1993,Springer et al., 1996), in comparison to other slow growing mycobacteria, 34

confirmed the consistence of the MTC. The signature above remains nowadays the most 35

reliable mark for the attribution of mycobacteria to MTC. At the same time the presence, 36

in M. triviale likewise in rapid growers, of an helix 18 14-nucleotide shorter, 37

unquestionably demonstrated the unrelatedness to this species to the complex. 38

A gap of more than 20 years separates recognition of the classical members of MTC from 39

the description of a new species related to such a group, Mycobacterium hiberniae 40

(Kazda et al., 1993). This novel mycobacterium is characterized by a unique phenotypic 41

feature; the pink pigmentation of the colonies, but the major role in its differentiation was 42

played by the genetic analysis that was, at that time, beginning to impose itself. The 43

boom years of MTC start however in 2006 with three new species described (Cloud et al., 44

2006, Masaki et al., 2006, Mun et al., 2008). 45

4

The identification at species level, within the members of the MTC still remains, almost 46

50 years later, problematic. As well as biochemical and cultural tests (Wayne & Kubica, 47

1986), the analysis of cell wall lipids revealed poorly discriminative. And, more recently, 48

DNA-probes specific for the MTC species have not been implemented by any of the 49

commercial hybridization methods; probably as a consequence of the limited interest 50

aroused by organisms grossly labeled, since the beginning, as non pathogenic. 51

Unexpectedly, even the identification by means of genetic sequencing remains elusive as 52

in public domain databases hundreds of sequences related to MTC are crowded without, 53

or with unreliable, species allocation. 54

The aim of this study was to investigate the phylogenetic and taxonomic structure of the 55

MTC on a large number of isolates and to make available in GenBank a panel of species-56

specific genetic sequences characterized by certain labels. 57

58

Materials and methods 59

60

Strains 61

For this study, all the strains assigned to the MTC on the basis of the routine 62

identification performed in our laboratory as from the year 1996 onward were 63

investigated. All such strains (no. 156) had been grown from clinical specimens and 64

stored frozen at -80°C. While the large majority of them had been isolated either in 65

Careggi Hospital laboratory or in other Italian hospitals, 27 strains had been obtained 66

from laboratories of other countries. After thawing, each strain was grown on 67

Middlebrook 7H11 medium at 37°C. The type strains of the species “M. engbaekii” 68

5

(ATCC27353), M. hiberniae (ATCC49874), Mycobacterium kumamotonense 69

(DSM45093), Mycobacterium senuense (DSM44999) and M. terrae (CIP104321) were 70

also included in the study; other reference strains investigated were M. 71

nonchromogenicum (PI140330001), and M. triviale (PI141030004). 72

73

Genetic sequencing 74

Three different regions were chosen for genetic characterization, the genes coding for the 75

16S rRNA (16S), for the 65 kDa heath shock protein (hsp65) and for the β subunit of the 76

RNA polymerase (rpoB). Of the 16S, according to a previously reported procedure 77

(Reischl et al., 1998), a trait including 479 bp was sequenced starting from the 78

Escherichia coli-corresponding position 28. The almost complete 16S sequence was 79

determined for the strains for which the sp. nov. status is proposed here. 80

Three hundred and ninety nine bp were sequenced in the hypervariable trait of hsp65 81

(McNabb et al., 2004), starting from the Mycobacterium tuberculosis-homologous 82

position 443. 83

In the rpoB the stretch recently proposed for the differentiation of rapidly growing 84

mycobacteria (Adékambi et al., 2003) was investigated; the length of the nucleotide 85

sequence (starting at the Mycobacterium smegmatis-corresponding position 2,554) ranged 86

from 711 to 726 bp in different strains. 87

In all the regions above both the forward and reverse strands were determined using Big 88

Dye terminator chemistry and an AB3730 DNA sequencer (Applied Biosystems). 89

For the designation of sequevars (sqvs.) for which the assignation to a species was 90

possible, the first three letters (capitalized) of the species name were used followed, for 91

6

16S, by a small letter (a, b, . . .), for hsp65, by a number (1, 2, . . .) and, for rpoB, by a 92

Roman numeral (i, ii, . . .). 93

Sequences of protein-coding genes (hsp65, 133 codons; rpoB, 237-242 codons) were also 94

translated to the amino acid residues composition to distinguish silent mutations from the 95

ones affecting the protein structure. 96

97

Phylogenetic analysis 98

The phylogenetic analysis was conducted according to the neighbour-joining method 99

(Saitou & Nei, 1987) under the total gap removal and Kimura’s two-parameter 100

substitution model (Kimura, 1980), and was evaluated by bootstrap analysis based on 101

1,000 replicates using the MEGA software version 5 {Tamura, 2011 38699 /id}. The 102

trees were rooted using, as outgroup, Mycobacterium tuberculosis whose respective 103

sequences retrieved from GenBank had been added previously to various alignments. 104

The phylogenetic reconstruction based on 16S included 22 different sqvs. detected among 105

the strains investigated in this study. The sequences, downloaded from GenBank, of the 106

most closely related slowly- and rapidly-growing mycobacteria were added as well. 107

To improve the robustness of the tree (Devulder et al., 2005,Mignard & Flandrois, 108

2008,Stackebrandt et al., 2002) the sequences of the three genetic regions were 109

concatenated in a single filament including a number of nucleotides ranging, in different 110

strains, from 1,589 to 1,604 (16S, 479; hsp65, 399; rpoB, 711-726). In the investigation 111

concerning the concatenated sequences all the combinations (no. 75) of 16S, hsp65 and 112

rpoB sqvs. detected in our strains were included. 113

114

7

Lipid investigations 115

High performance liquid chromatography (HPLC) of cell-wall mycolic acids was carried 116

out on all the strains after esterification to bromophenacyl esters as reported (C.D.C., 117

1996). 118

119

Biochemical and cultural tests 120

For all the strains, nitrate reduction, growth rate and pigmentation of colonies were 121

investigated. For the strains considered to represent new species a number of randomly 122

selected strains (10 of M. heraklionense, 5 of M. engbaekii and 2 of M. longobardum) 123

were investigated with a wider panel of tests (Table 3) according to the standard 124

procedures (Kent & Kubica, 1985). 125

126

Susceptibility testing 127

The susceptibility testing was performed on randomly selected strains belonging to the 128

new species proposed here (4 of M. heraklionense, 4 of M. engbaekii and 2 of M. 129

longobardum). The minimal inhibitory concentrations (MIC) of drugs selected for their 130

activity on slowly growing mycobacteria were determined using commercially available 131

microdilution plates (SLOMYCO, VersaTREK) following the CLSI recommendations 132

{C.L.S.I., 2011 58541 /id} 133

134

Results 135

Genetic sequencing 136

8

The alignment of the 22 sqvs. detected in the 16S of the 156 strains investigated added to 137

the existing evidence that M. triviale does not belong to the MTC. The helix 18 of the 138

16S was in fact, in the six strains of M. triviale present in our panel, 14 nucleotides 139

shorter than in remaining strains. Such a feature, not only excludes M. triviale from MTC 140

but even places this species within rapid growers. These six strains were therefore 141

excluded from the study. 142

Six of the 21 remaining 16S sqvs. turned out to overlap the sequences of the type strains 143

of the six species known to belong to the MTC and this allowed to assign them to M. 144

arupense, M. nonchromogenicum, M. hiberniae, M. senuense, M. terrae and M. 145

kumamotonense. Another sqvs. was 100% identical to the sequence of the not officially 146

recognized species “Mycobacterium engbaekii”. Only nine of the remaining 16 sqvs. 147

were found in GenBank database, five of them had been previously deposed by one of us, 148

while the species assignation of two others was either lacking or incorrect. The pairwise 149

matrix of distances (Table S1) allowed the identification of four clusters of sqvs. which 150

were assigned to the species M. arupense, M. senuense, M. kumamotonense and M. terrae 151

on the basis of the inclusion of the sequences of the respective type strains. Within such 152

clusters the intraspecies variability was <1%. One sqv. shared by two strains (FI-153

07105/FI-11038) differed, for one nucleotide only, from the not officially recognized 154

“Mycobacterium paraterrae” (Lee et al., 2010). The three sqvs. remaining (NEW1, GN-155

9188, FI-09379), differed from any type strain and did not fit any cluster. 156

Thirty sqvs. were found in the hsp65. With the similarity matrix (Table S2) eight clusters 157

were recognized in this region, which were characterized by pairwise distances <3% (3 158

exceptions with values up to 3.36%). Five such clusters were immediately attributed to 159

9

the species M. arupense, M. senuense, M. kumamotonense, M. terrae and “M. engbaekii”, 160

because of the inclusion of the sequences of the respective type strains; they all were 161

detected in strains assigned to the same species by the 16S sequence. One of the clusters 162

remaining (NEW2, including 2 sqvs.), although clearly separated from the cluster of M. 163

arupense, had been detected in strains assigned, on the basis of 16S, to the latter species. 164

The other two clusters (NEW3 and NEW4) included, two and three, orphan (unassigned 165

to any known species) sqvs. respectively. Interestingly two (FI-07105 and FI-11038) of 166

the three sqvs. included in NEW4 belonged to the strains presenting, in 16S, close 167

similarity with “M. paraterrae” but clearly differed from the latter in the hsp65 region. In 168

the species M. nonchromogenicum and M. hiberniae one single sequevar was present. 169

Three sqvs. (NEW1, GN-9188 and FI-05196) differed from any type strain and did not fit 170

any of the clusters above. Interestingly the sequevar NEW1 was detected in the strains 171

classified as NEW1 also on the basis of the 16S. 172

As expected, a large variability was detected in the rpoB fragment, with the presence of 173

58 sqvs. The pairwise matrix of distances (data not shown) included ten clusters with 174

internal variability below the limit (3%) proposed for rpoB region (Adékambi & 175

Drancourt, 2004) in the large majority of cases (5 exceptions with values up to 5.3%). For 176

this gene, the presence in GenBank of only one MTC sequence overlapping to ours 177

(furthermore not assigned to any known species), did not allow us to attribute any 178

sequevar to a species; the attribution was therefore inferred from that achieved, for each 179

strain, on the basis of the 16S and hsp65 sqvs. Following this approach it was possible to 180

classify one cluster within each of the species: M. arupense (10 sqvs.), M. 181

nonchromogenicum (5 sqvs.), “M. engbaekii” (4 sqvs.), M. senuense (2 sqvs.), M. 182

10

kumamotonense (9 sqvs.) and M. terrae (6 sqvs.). Of the remaining four clusters, one 183

(NEW1, with 2 sqvs.) was detected in strains classified as NEW1 on the basis of 16S and 184

hsp65 sequences; and one (NEW2, with 7 sqvs.) was detected in strains classified as 185

NEW2 on the basis of hsp65 sequence). One cluster (NEW5, with 3 sqvs.) was detected 186

in strains assigned to the species M. arupense on the basis of 16S and hsp65 sequences, it 187

was however very distant from the rpoB sqvs. of the strains presenting in 16S, and hsp65 188

too, the typical sqvs. of this species. The last cluster (NEW6, with 2 sqvs.) remained 189

orphan. The sqvs. FI-06258, FI-07105/FI-09015/FI-09399, FI-09379, FI-05396/FI-190

06246/FI-05196 and FI-11038 did not fit any cluster. Interestingly the five cases in which 191

the intraspecies variability exceeded 3% were detected within the cluster of M. terrae 192

where the sqv. obtained from the type strain (CIP104321) clearly differed from all the 193

others. 194

Of the sqvs. we detected, apart from the ones of the type strains and the ones previously 195

deposed by some of us, only a limited number was already present in GenBank. As 196

regard to 16S, two of the five sqv. present were not assigned to any species while the 197

label of the remaining three was correct in two cases and incorrect in one. For what 198

concern the hsp65, there were three sqvs. correctly assigned, two lacking any species 199

attribution and three mislabeled; furthermore, in this region, our sequence of the type 200

strain of M. senuense presented one mismatch in comparison to the one present in the 201

database. For rpoB sqvs. only two, both assigned to the species M. terrae, were present in 202

GenBank although not identified. In this region too one type strain (M. terrae) presented 203

one nucleotide discordance. 204

11

The combinations of 16S, hsp65 and rpoB sqvs. detected in the strains investigated here 205

are reported in Table S3. 206

207

Phylogenetic analysis 208

The tree constructed using the sequences of 16S revealed a clear separation of the strains 209

included in the MTC from both the slow and the rapid growers (Fig. 1). The strains 210

investigated here were distributed among two major branches including, the first, M. 211

senuense, M. terrae and M. kumamotonense and, the second, M. arupense, M. 212

nonchromogenicum, M. hiberniae, “M. engbaekii”, and the clusters NEW1 and NEW2. 213

The branch including M. triviale appeared very distant from the species of MTC. 214

The phylogenetic tree inferred from hsp65 was fully in agreement with the one obtained 215

from 16S sequences (Fig. S1) while in the one based on the rpoB a polyphyletic 216

distribution of different species emerged (Fig. S2). 217

The phylogenetic analysis conducted on the concatenated sequences of the 75 different 218

combinations detected in the 137 MTC strains which was possible to assign to a species 219

produced a quite robust dendrogram characterized by a high percentage of nodes, 220

including the early ones, with very high bootstrap values (Fig. 2). Different species-221

specific clusters were clearly separated although belonging to two major groupings: the 222

first including M. arupense, NEW2, M. nonchromogenicum, M. hiberniae, “M. 223

engbaekii” and NEW1; the second comprising M. senuense, M. terrae and M. 224

kumamotonense. 225

12

The analysis of translated nucleotide sequences revealed a large number of synonymous 226

mutations with, in the hsp65 region, only 12 different amino acid sequences being coded 227

by 31 different sqvs. and, in rpoB, only 27 by 59 sqvs. 228

229

HPLC of cell wall mycolic acids 230

The HPLC of cell wall mycolic acids performed on the whole initial group of 156 strains, 231

revealed two different types of pattern (Fig. 3). The motif shared by the large majority of 232

the strains was characterized by the presence of two, clearly distinct, clusters of peaks, 233

with the first one, including three major peaks, starting to elute after 3 min, and the 234

second, including three minor peaks, eluting about 3 min later. The other pattern was 235

presented only by the six strains of M. triviale subsequently excluded from the study; in 236

this species only a late emerging cluster of peaks was present (Fig. 3D). 237

Because of the common motif shared by all the members of the MTC, with only a limited 238

variability in the relative height in the major peaks of the first cluster we tried to correlate 239

their profile to the different species. If the lower peaks were not taken into account, 240

almost all the strains presented three major peaks in the first clusters. The most common 241

motif was characterized by a highest central peak followed by the first and the third in 242

decreasing order (Fig. 3A). This pattern was presented by the strains belonging to the 243

species M. nonchromogenicum and to the group NEW1. It was furthermore shared by 244

70% of the strains of M. arupense and was also represented in about half of the strains of 245

“M. engbaekii” and of the group NEW2. Less frequent, but scattered among various 246

groupings, was the motif in which the third peak was higher than the first with the second 247

being the highest. Unique to the species M. kumamotonense, although not presented by 248

13

all such strains, was a pattern characterized by three peaks almost equally high with the 249

height slightly rising from the first to the third (Fig. 3B). Equally unique, to the species 250

M. terrae, was the profile presenting in the first cluster four major peaks instead of three 251

(Fig. 3C). 252

253

Biochemical and cultural tests 254

The nitrate test was selected as the ability of reducing nitrate to nitrite is classically 255

considered the sole biochemical feature suitable to discriminate within the, otherwise 256

phenotypically homogeneous, species of the MTC (Wayne & Kubica, 1986). The nitrate 257

reductase was possessed by the strains of the of M. kumamotonense; it was absent in M. 258

arupense and M. terrae (Table S4). 259

The colonies of the strains investigated were buff and prevalently smooth; an exception 260

was represented by M. hiberniae and “M. engbaekii” which presented pink pigmentation 261

(Table S4). 262

At 37°C the growth on solid media, from standardized inocula, became distinctly visible, 263

on average, after 7-14 days, faster growth was occasionally observed, while M. senuense 264

grew typically slowly (Table S4). 265

266

Discussion 267

Several obvious conclusions seem to emerge from the results of genotypic and 268

phenotypic investigations carried on a large number of strains belonging to the MTC. 269

i. The legitimacy of the species of the MTC officially recognized at present is clearly 270

confirmed. 271

14

ii. The group we temporarily named NEW1 includes strains evidently different from any 272

other species of the MTC in all the three genetic regions investigated; for it (of which 273

seven strains were characterized in this study) the status of new species is proposed with 274

the name Mycobacterium longobardum. 275

iii. The strains included so far in the group NEW2, which had been initially assigned, on 276

the basis of the analysis of the 16S sequence, to the species M. arupense, are actually 277

clearly distant from the latter in the regions hsp65 and rpoB, and represent a previously 278

unreported species for which the name Mycobacterium heraklionense is proposed; of this 279

23 strains were characterized in this study. 280

iv. The seven strains investigated here and assigned to the not officially recognized 281

species “M. engbaekii” support the revival of M. engbaekii sp. nov. The name M. 282

engbaekii was proposed in 1972 for 15 rapidly growing strains presenting pink 283

pigmentation (Korsak & Boisvert, 1972); although the type strain of “M. engbaekii” was 284

deposed in the American Type Culture Collection the sp. nov. description did not follow 285

and it sank into oblivion. 286

v. Within the species M. arupense two variants can be clearly distinguished on the basis 287

of the rpoB sequence; sufficient evidence has not emerged so far to justify the elevation 288

of this new variant (indicated in this study as NEW6) to species rank, despite the 289

divergence, clearly over the proposed cut-off, in the rpoB region. 290

vi. Surprisingly, the recently described species M. arupense and M. kumamotonense are, 291

by far, the most frequently isolated members of the MTC. One of the reference strains 292

investigated here (M. nonchromogenicum PI140330001) turned out to be M. arupense, 293

which supports the hypothesis that the latter species and M. kumamotonense were 294

15

identified in the past (and, at times, at present too) as M. nonchromogenicum and M. 295

terrae respectively. In our study the isolations of M. nonchromogenicum and M. terrae 296

were not as frequent as generally acknowledged; which brings up for discussion the 297

widespread conviction considering M. nonchromogenicum the only member of MTC 298

potentially responsible for disease (Tsukamura et al., 1983). 299

vii. A number of strains have been detected presenting in 16S, hsp65 and rpoB either 300

conflicting sqvs. of MTC species (3 cases) or unreported sqvs. not related to other, 301

official or proposed, MTC species (10 cases; data not shown). Such strains (Table S5) 302

need further characterization and have not been included in the phylogenetic 303

investigations. 304

viii. Questions are raised by the type strain of M. terrae. While in the 16S its sequence 305

clusters with the ones of number of strains (Table S1) in hsp65 (Table S2) and, even more 306

evidently in rpoB, it clearly diverges from the others. A question remains so far 307

unanswered. Is the type strain of M. terrae a rare organism, poorly representative of the 308

species? Are, on the contrary, the other strains assigned here to M. terrae actually 309

members of a still undescribed new species? With the latter option implying that M. 310

terrae is so rare that no one was detected in our long-term study. 311

ix. A hoary, unresolved, problem is that of the lawfulness of describing a new species 312

based on a single isolate. It is not our aim to deal with this topic here; nevertheless in this 313

study a dozen unique strains emerged as potentially exploitable for the description of the 314

same number of new species! 315

316

Bacterial strains of the proposed new species 317

16

Seven strains of M. longobardum were isolated in two Italian hospitals between years 318

2006 and 2009 from the sputum of six Italian, and one Lebanese, patient. Microscopy 319

was negative and growth was obtained on solid media only (despite in four cases also 320

liquid media had been inoculated). In no case was the strain considered responsible for 321

disease. 322

Twenty three strains of M. heraklionense had been isolated in Greece (no. 10), in Italy 323

(no. 7) and in India (no. 6). The Greek strains were grown from one outpatient and nine 324

hospitalized patients in the island of Crete between 2002 and 2003. The Italian strains 325

were isolated between 2005 and 2011 in five different hospitals. No information is 326

available for the Indian strains apart from that they had been isolated before the year 327

2005. 328

Of the seven strains of M. engbaekii six had been isolated in Italy, in three different 329

hospitals between 1998 and 2011, and one in Guadeloupe. 330

Clinical and epidemiological information available for part of the strains above are 331

reported in Table 1. 332

333

Genetic sequences of the proposed new species 334

In the 16S, M. arupense was the officially recognized species most closely related, to M. 335

longobardum from which however 25 bp out 1,488 differed (similarity 98.3). M. 336

longobardum presented the best resemblance in the hsp65 (11 mismatches in 399 bp, 337

similarity 97.2%) with M. kumamotonense. 338

17

As far as M. heraklionense is concerned, it was characterized by high similarity with M. 339

arupense both in the 16S (only 7 discrepancies out of 1,427 bp, similarity 99.5%) and in 340

the hsp65 (99.2% similarity, 3 mismatches out of 399 bp). 341

For the three proposed new mycobacteria, the species most closely related in the rpoB 342

presented very low similarity (<94%), this value is however affected by the limited 343

coverage of the database of this genetic region. The comparison of their sequences with 344

those determined in this study for the species belonging to the MTC revealed, for each of 345

them, a clear divergence with the closer species; M. heraklionense differed 2.9% from M. 346

nonchromogenicum; M. longobardum differed 3.6% from M. heraklionense and M. 347

engbaekii differed 4.2% from M. heraklionense. 348

349

PCR restriction analysis (PRA) of the proposed new species 350

PRA patterns (Telenti et al., 1993) inferred on the basis of restriction sites present in the 351

hsp65 sequences were different for each of the three potential sp. nov. (Table 2); two 352

biotypes were detected for M. engbaekii. The PRA patterns above were unique and 353

suitable to differentiate M. engbaekii, M. heraklionense and M. longobardum from any 354

other known Mycobacterium species (http://app.chuv.ch/prasite/). 355

356

Phylogenetic analysis of the proposed new species 357

In the phylogenetic reconstructions based on 16S, hsp65 and rpoB the three proposed 358

new species belonged to a sub-branch of the MTC including also M. arupense, M. 359

nonchromogenicum and M. hiberniae; with M. kumamotonense, M. senuense and M. 360

terrae located on a different branch. The tree emerging from the concatenated sequences 361

18

clearly showed M. longobardum separated from the others species; M. heraklionense 362

closer to M. nonchromogenicum and M. engbaekii closer to M. hiberniae (Fig. 2). 363

364

HPLC of cell wall mycolic acids of the proposed new species 365

The low discriminating power, within the MTC, of HPLC of cell wall mycolic acids was 366

confirmed also by the strains belonging to the species M. engbaekii, M. heraklionense 367

and M. longobardum which presented the typical profile characterized by an early major 368

and a late minor clusters of peaks (Fig. 3). 369

370

Biochemical and cultural tests and susceptibility testing results of the proposed new 371

species 372

The major differences, revealed by biochemical and cultural tests (Table 3) among the 373

three species, concerned morphology and pigmentation of the colonies (rough pink 374

scotochromogenic in M. engbaekii, rough and unpigmented in M. longobardum and 375

smooth unpigmented in M. heraklionense), the Tween 80 hydrolysis (negative in M. 376

longobardum only), the arylsulfatase at three days (positive in M. longobardum only) and 377

the β-glucosidase (positive in M. heraklionense only). With respect to other known 378

members of the MTC the discriminative power of biochemical and cultural tests was, as 379

expected, very limited (data not shown) such confirming the perception which lead to 380

propose, about 40 years ago, the inclusion in a complex. 381

The susceptibility testing revealed susceptibility of the three species to clarithromycin 382

and resistance to quinolones with the activity of other molecules being variable (Table 4). 383

384

19

Description of M. engbaekii sp. nov. 385

M. engbaekii sp. nov. (eng.ba.e’ki.i N.L. gen. m. engbaekii of Engbaek, to honor of the 386

Danish mycobacteriologist H.C. Engbaek). 387

Cells are typically acid fast, rod-shaped, with some coccoid forms, but without branches 388

or aerial hyphae; spores are not produced. Mature growth is obtained in solid media at 389

temperatures between 25 and 37 °C in less than 10 days. Colonies grow rough and 390

unpigmented in the dark but after light exposure develop a pink pigmentation. The 391

organism is positive for nitrate reductase, Tween 80 hydrolysis and tellurite reduction, 392

and is negative for niacin accumulation, arylsulfatase at 3 days, urease and β-glucosidase. 393

Catalase at 68 °C is present and more than 45 mm foam in the semi-quantitative test are 394

produced. Isolates are susceptible to amikacin, clarithromycin, ethambutol, linezolid and 395

rifabutin, and resistant to doxycycline and sulfamethoxazole.. Mycolic acids produce an 396

early major and a late minor cluster of peaks. The gene sequences are unique in the 16S, 397

in the hsp65 (where 4 sqvs. are present) and in rpoB (4 sqvs.). Phylogenetically the 398

species is included in the MTC more closely related to M. hiberniae. The type strain is 399

ATCC27353=DSM45694. 400

401

Description of M. heraklionense sp. nov. 402

M. heraklionense sp. nov. (he.ra.kli.on.en’se N.L. adj. n. heraklionense from Heraklion 403

the city in Crete island where many such strains were isolated). 404

Cells are typically acid fast, rod-shaped, with some coccoid forms, but without branches 405

or aerial hyphae; spores are not produced. Mature growth is obtained in solid media at 406

temperatures between 25° and 37°C in 5 to 12 days. Colonies grow smooth and 407

20

unpigmented both in the dark and after light exposure. The organism is positive for 408

nitrate reductase, Tween 80 hydrolysis and β-glucosidase, and is negative for niacin 409

accumulation, and arylsulfatase at 3 days, tellurite reduction, and urease. Catalase at 68°C 410

is present and more than 45 mm foam in the semi-quantitative test are produced. Isolates 411

are susceptible to clarithromycin and resistant to quinolones, rifampicin, 412

sulfamethoxazole and doxycycline. Mycolic acids produce an early major and a late 413

minor cluster of peaks. The gene sequences are unique in the 16S, where two sqvs. 414

closely related to M. arupense are present, in the hsp65 (3 sqvs.), where best resembles 415

M. nonchromogenicum, and in rpoB (7 sqvs.), equally divergent from M. arupense and 416

M. nonchromogenicum. Phylogenetically the species is included in the MTC more closely 417

related to M. nonchromogenicum. The type strain is GN-1T=10803T (NCTC13432= 418

LMG24735=CECT7509). 419

420

Description of M. longobardum sp. nov. M. longobardum sp. nov. (lon.go’bar.dum N.L. 421

adj. n. from Lombardy the region were the strains were isolated). 422

Cells are typically acid fast, rod-shaped, with some coccoid forms, but without branches 423

or aerial hyphae; spores are not produced. Mature growth is obtained in solid media at 424

temperatures between 25° and 37°C in 7 to 14 days. Colonies grow rough and 425

unpigmented both in the dark and after light exposure. The organism is positive for 426

nitrate reductase and arylsulfatase at 3 days, and is negative for niacin accumulation, 427

Tween 80 hydrolysis, tellurite reduction, urease and β-glucosidase. Catalase at 68 °C is 428

present and more than 45 mm foam in the semi-quantitative test are produced. Isolates are 429

susceptible to sulfamethoxazole and clarithromycin, and resistant to quinolones, linezolid 430

21

and streptomycin. Mycolic acids produce an early major and a late minor cluster of 431

peaks. The gene sequences are unique in the 16S, in the hsp65 and in rpoB (where 2 sqvs. 432

are present). Phylogenetically the species is included in the MTC clearly separated from 433

other species. The type strain is FI-07034T (DSM45394=CCUG58460). 434

435

Acknowledgement 436

437

We thank Antonella Grottola (Laboratory of Microbiology and Virology, Modena 438

University Hospital, Modena, Italy) for her valuable comments on the manuscript. 439

We are grateful to Meg Lafferty for reviewing the language of the manuscript. 440

441

References 442

443

Adékambi, T., Colson, P., & Drancourt, M. (2003). rpoB-based identification of 444

nonpigmented and late-pigmenting rapidly growing mycobacteria. J Clin Microbiol 41, 445

5699-5708. 446

Adékambi, T. & Drancourt, M. (2004). Dissection of phylogenetic relationships among 447

19 rapidly growing Mycobacterium species by 16S rRNA. hsp65, sodA, recA, and rpoB 448

gene sequencing. Int J Syst Evol Microbiol 54, 2095-2105. 449

C.D.C. (1996). Standardized method for HPLC identification of mycobacteria. U.S. 450

Department of Health and Human Services, Public Health Service, Atlanta. 451

22

Cloud, J.L., Meyer, J.J., Pounder, J.I., Jost, K.C., Jr., Sweeney, A., Carrol, K.C., & 452

Woods, G.L. (2006). Mycobacterium arupense sp. nov., a non-chromogenic bacterium 453

isolated from clinical specimens. Int J Syst Evol Microbiol 56, 1413-1418. 454

C.L.S.I. (2011). Susceptibility testing of mycobacteria, nocardiae and other aerobic 455

actinomycetes; approved standard - Second Edition. C.L.S.I., Wayne. 456

457 Devulder, G., Perouse, d.M., & Flandrois, J.P. (2005). A multigene approach to 458

phylogenetic analysis using the genus Mycobacterium as a model. Int J Syst Evol 459

Microbiol 55, 293-302. 460

Kent, P.T. and Kubica, G.P. (1985). Public health mycobacteriology. A guide for the 461

level III laboratory. U.S. Department of Health and Human Services, Atlanta. 462

Kimura, M. (1980). A simple method for estimating evolutionary rates of base 463

substitutions through comparative studies of nucleotide sequences. J Mol Evol 16, 111-464

120. 465

Kirschner, P., Springer, B., Vogel, U., Meier, A., Wrede, A., Kiekenbeck, M., Bange, 466

F.C., & Böttger, E.C. (1993). Genotypic identification of mycobacteria by nucleic acid 467

sequence determination: report of a 2-year experience in a clinical laboratory. J Clin 468

Microbiol 31, 2882-2889. 469

Korsak, T. & Boisvert, H. (1972). Mycobactéries a pigment rose. Ann Inst Pasteur 122, 470

31-41. 471

23

Kubica, G.P., Silcox, V.A., Kilburn, J.O., Smithwick, R.W., Beam, R.E., Jones, 472

W.D., & Stottmeier, K.D. (1970). Differential identification of mycobacteria. VI. 473

Mycobacterium triviale Kubica sp. nov. Int J Syst Bacteriol 20, 161-174. 474

Lee, H., Lee, S.A., Lee, I.K., Yu, H.K., Park, Y.G., Jeong, J., Lee, S.H., Kim, S.R., 475

Hyun, J.W. & other authors. (2010). Mycobacterium paraterrae sp. nov. recovered 476

from a clinical specimen: novel chromogenic slow growing mycobacteria related to 477

Mycobacterium terrae complex. Microbiol Immunol 54, 46-53. 478

Masaki, T., Ohkusu, K., Hata, H., Fujiwara, N., Iihara, H., Yamada-Noda, M., 479

Nhung, P.H.H.M., Asano, Y., Kawamura, Y., & Ezaki, T. (2006). Mycobacterium 480

kumamotonense sp. nov. recovered from clinical specimen and the first isolation report of 481

Mycobacterium arupense in Japan: novel slowly growing, nonchromogenic clinical 482

isolates relaed to Mycobacterium terrae complex. Microbiol Immunol 50, 889-897. 483

McNabb, A., Eisler, D., Adie, K., Amos, M., Rodrigues, M., Stephens, G., Black, 484

W.A., & Isaac-Renton, J. (2004). Assessment of partial sequencing of the 65-kiloDalton 485

heat shock protein gene (hsp65) for routine identification of mycobacterium species 486

isolated from clinical sources. J Clin Microbiol 42, 3000-3011. 487

Mignard, S. & Flandrois, J.P. (2008). A seven-gene, multilocus, genus-wide approach 488

to the phylogeny of mycobacteria using supertrees. Int J Syst Evol Microbiol 58, 1432-489

1441. 490

Mun, H.S., Park, J.H., Kim, H., Yu, H.K., Park, Y.G., Cha, C.Y., Kook, Y.H., & 491

Kim, B.J. (2008). Mycobacterium senuense sp. nov., a slowly growing non-chromogenic 492

24

species closely related to the Mycobacterium terrae complex. Int J Syst Evol Microbiol 493

58, 641-646. 494

N.C.C.L.S. (2003). Susceptibility testing for mycobacteria, nocardiae and other aerobic 495

actinomycetes; approved standard M24-A. N.C.C.L.S., Wayne, Pa. 496

Reischl, U., Emler, S., Horak, Z., Kaustova, J., Kroppenstedt, R.M., Lehn, N., & 497

Naumann, L. (1998). Mycobacterium bohemicum sp. nov., a new slow-growing 498

scotochromogenic mycobacterium. Int J Syst Bacteriol 48, 1349-1355. 499

Saitou, N. & Nei, M. (1987). The neighbor-joining method: a new method for 500

reconstructing phylogenetic trees. Mol Biol Evol 4, 406-425. 501

Springer, B., Stockman, L., Teschner, K., Roberts, G.D., & Böttger, E.C. (1996). 502

Two-laboratory collaborative study on identification of mycobacteria: molecular versus 503

phenotypic methods. J Clin Microbiol 34, 296-303. 504

Stackebrandt, E., Frederiksen, W., Garrity, G.M., Grimont, P.A., Kampfer, P., 505

Maiden, M.C.J., Nesme, X., Rossellò-Mora, R., Swings, J. & other authors. (2002). 506

Report of the ad hoc committee for the re-evaluation of the species definition in 507

bacteriology. Int J Syst Evol Microbiol 52, 1043-1047. 508

Tamura, K., Peterson, D., Peterson, N., Steker, G., Nei, M., & Kumar, S. (2011). 509

MEGA5: molecular evolutionary genetics analysis using maximum likelihood, 510

evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28, 2731-2739. 511

25

Telenti, A., Marchesi, F., Balz, M., Bally, F., Böttger, E.C., & Bodmer, T. (1993). 512

Rapid identification of mycobacteria to the species level by polymerase chain reaction 513

and restriction enzyme analysis. J Clin Microbiol 31, 175-178. 514

Tsukamura, M. (1965). A group of mycobacteria from soil sources resembling 515

nonphotochromogens (group 3). A description of Mycobacterium nonchromogenicum. 516

Med Biol 71, 110-113. 517

Tsukamura, M., Kita, N., Otsuka, W., & Shimoide, H. (1983). A study of the 518

taxonomy of the Mycobacterium nonchromogenicum complex and report of six cases of 519

lung infection due to Mycobacterium nonchromogenicum. Microbiol Immunol 27, 219-520

236. 521

Wayne, L.G. (1966). Classification of mycobacteria. III. Species within Group III. Am 522

Rev Respir Dis 93, 919-928. 523

Wayne, L.G. & Kubica, G.P. (1986). Family Mycobacteriaceae CHESTER 1897, 63AL, 524

p. 1435-1457. In: P. H. A. Sneath, N. S. Mair, M. E. Sharpe, and J. G. Holt (eds.), 525

Bergey's manual of systematic bacteriology. The Williams & Wilkins Co., Baltimore. 526

527

528

529 530 531

26

Table 1. Epidemiological, microbiological and clinical characteristics of 6 strains of M. engbaekii 16 strains of M. heraklionense and 7 strains of M. longobardum.

Speciesa Strain Age Sex Specimen type Microscopy Culture (pos/done)

Disease Sitee Year Sequevars (16S/hsp65/rpoB)

E FI-04007 74 M Gastric washing Neg NA NA GL 1998 a/2/iii E FI-98002 NA M NA NA NA NA MI 1998 a/1/ii E FI-06007 62 M Urine Neg 1/3 NA VI 2006 a/3/iii E FI-98058 NA M NA NA NA NA TR 1998 a/1/i E FI-98001 NA M NA NA NA NA MI 1998 a/1/iv E HSR-11012 75 M Bronchial aspirate Neg 1/1 NA VI 2011 a/1/ii H GN01T 74 M NA Neg 1/1 Renal failure, heart failure H 2002 a/2/ii H GN02 76 M NA Neg 1/1 Myelodisplastic syndrome H 2003 a/2/ii H GN04 83 M NA Neg 1/1 COPDc H 2003 a/2/ii H GN05 42 M NA Neg 1/1 Tuberculosis H 2003 a/2/ii H GN06 70 M NA Neg 1/1 Lung cancer H 2003 a/2/ii H GN08 59 M NA Neg 1/1 Non-Hodgkin lymphoma H 2003 a/2/ii H GN09 35 M NA Neg 1/1 Rheumatoid arthritis H 2003 a/2/ii H GN10 82 M NA Neg 1/1 Pulmonary fibrosis H 2003 a/2/ii H GN12 77 M NA Neg 1/1 Lung cancer H 2003 a/2/ii H FI-10248 61 M NA NA NA NA AN 2010 a/1/iii H HRS-11013 82 M Bronchial lavage Neg 1/2 NA NO 2011 a/1/iii H FI-06009 62 M Sputum Neg 1/3 NA NO 2006 a/1/i H FI-06255 NA F Sputum NA NA NA CO 2007 a/2/v H FI-08098 NA M NA Neg 1/1 NA MI 2008 a/1/iii H FI-08101 NA M NA NA NA NA MI2 2008 a/1/ii H FI-05158 74 F Sputum Neg 1/2 NA NO 2005 a/1/iii L FI-09110 71 M NA Neg 1/2b TB VA 2009 a/1/ii L FI-07054 65 M NA Neg 1/3b COPDc VA 2006 a/1/ii L FI-06254 78 M NA Neg 1/1b Pneumonia VA 2006 a/1/i L FI-09059 51 M NA Neg 1/9b Bronchitis VA 2008 a/1/i L FI-07034T 72 F NA Neg 1/4 Broncho-pneumonitis VA1 2006 a/1/i L FI-07020 76 M NA Neg 1/2 Lung cancer VA1 2006 a/1/ii L FI-07089 39 F NA Neg 2/3 Suspected TB in LESd patient VA1 2006 a/1/i

a E, M. engbaekii; H, M. heraklionense; L, M. longobardum b positive in liquid- but negative in solid-medium c chronic obstructive pulmonary disease d systemic lupus erythematosus e H, Greece (Heraklion); GL, Guadeloupe; other acronyms, different Italian cities.

27

Table 2. Patterns detected by PCR restriction analysis with enzymes BstEII and HaeIII in the strains for

which the status of sp. nov. is proposed.

Sequevars Restriction patterns

BstEII HaeIII

ENG1 289-96 124-58-54

ENG2-3-4 304-96 118-87-58

HER1-2 210-96-94 118-87-77

LON1 210-190 118-112-69

28

Table 3. Results of biochemical and cultural tests investigated for three proposed new species.

a photochromogenic

Test M. engbaekii M. heraklionense M. longobardum

Growth rate rapid intermediate intermediate

Growth at 25°C positive positive positive

Growth at 45°C variable negative negative

Pigmentation pinka absent absent

Colony morphology rough smooth rough

Niacin accumulation negative negative negative

Nitrate reduction negative positive positive

Catalase 68°C positive positive positive

Catalase >45 mm positive positive positive

Tween 80 hydrolysis variable positive negative

Tellurite reduction positive negative negative

Arylsulfatase 3 day negative negative positive

Urea hydrolysis negative negative negative

β-glucosidase positive positive negative

Tolerance of

MacConkey negative negative variable

p-nitrobenzoic acid positive positive positive

Thiophene carboxylic acid positive positive positive

Tiacetazone positive positive positive

Hydroxylamine positive variable positive

Isoniazid positive positive positive

Oleate positive negative positive

29

Table 4. Minimal inhibitory concentrations of several strains of M. engbaekii, M. heraklionense and M.

longobardum to antimycobacterial drugs.

a I, intermediate; R, resistant, S, susceptible.

Drug M. engbaekii M. heraklionense M. longobardum

MIC Interpretationa MIC Interpretationa MIC Interpretationa

Ciprofloxacin 2 I ≥16 R 16 R

Moxifloxacin 2-4 R >8 R ≥8 R

Linezolid ≤1 S 4-16 I/S 64 R

Rifampicin 4 I ≥8 R 4-8 I/R

Rifabutin ≤0.25 S ≤0.25 S 0.5 I

Sulfamethoxazole ≥76 R >152 R 5-9.5 S

Doxycycline 16 R ≥16 R 4 I

Ethambutol ≤0.5 S 2 S/I 4 I

Clarithromycin 1 S 1 S 2 S

Amikacin 2-8 S 1-32 S/I 32 I

Streptomycin 2-8 S/I 2->64 R/S 32 R

30

Table S3. Combinations of 16S, hsp65 and rpoB sqvs. detected in the strains investigated.

Genetic region Combinations number Species 16S hsp65 rpoB

ARUa ARU1 ARUi 7 ARU ARUa ARU1 ARUii 9 ARU ARUa ARU1 ARUiii 4 ARU ARUa ARU1 ARUix 1 ARU ARUa ARU1 ARUvi 2 ARU ARUa ARU1 ARUvii 1 ARU ARUa ARU1 ARUviii 4 ARU ARUa ARU1 ARUx 3 ARU ARUa ARU1 ARUxii 1 ARU ARUa ARU1 ARUxiii 2 ARU ARUa ARU1 ARUxiv 1 ARU ARUa ARU2 ARUi 2 ARU ARUa ARU2 ARUii 2 ARU ARUa ARU2 ARUiv 1 ARU ARUa ARU2 ARUviii 2 ARU ARUa ARU2 ARUx 1 ARU ARUa ARU2 ARUxv 1 ARU ARUa ARU3 ARUiv 1 ARU ARUa ARU3 ARUvi 1 ARU ARUa ARU4 ARUi 1 ARU ARUa ARU4 ARUv 1 ARU ARUa ARU5 ARUviii 1 ARU ARUb ARU6 ARUxi 1 ARU ARUc ARU1 ARUx 1 ARU

ENGa ENG1 ENGi 2 ENG ENGa ENG1 ENGii 2 ENG ENGa ENG1 ENGiv 1 ENG ENGa ENG2 ENGiii 1 ENG ENGa ENG3 ENGiii 1 ENG

HERa HER1 HERi 6 HER HERa HER1 HERii 1 HER HERa HER1 HERiii 4 HER HERa HER1 HERiv 1 HER HERa HER2 HERii 11 HER HERa HER2 HERv 1 HER

HIBa HIB1 HIBi 1 HIB

KUMa KUM1 KUMi 1 KUM KUMa KUM1 KUMv 1 KUM KUMa KUM1 KUMvii 5 KUM KUMa KUM1 KUMx 1 KUM KUMa KUM2 KUMi 1 KUM KUMa KUM2 KUMviii 1 KUM KUMb KUM1 KUMi 3 KUM

31

KUMb KUM1 KUMv 1 KUM KUMb KUM1 KUMviii 2 KUM KUMb KUM1 kUMiii 1 KUM KUMb KUM2 KUMii 3 KUM KUMb KUM2 kUMiii 1 KUM KUMb KUM2 kUMiv 1 KUM KUMb KUM2 KUMv 1 KUM KUMc KUM1 kUMix 1 KUM KUMc KUM2 KUMi 1 KUM KUMc KUM2 kUMii 3 KUM KUMc KUM2 kUMiii 1 KUM KUMc KUM2 kUMiv 1 KUM KUMc KUM2 KUMvi 1 KUM KUMc KUM2 KUMvii 1 KUM KUMd KUM1 KUMviii 1 KUM KUMd KUM2 KUMii 1 KUM

LONa LON1 LONi 4 LON LONa LON1 LONii 3 LON

NONa NON1 NONi 3 NON NONa NON1 NONii 3 NON NONa NON1 NONiii 1 NON NONa NON1 NONiv 1 NON NONa NON1 NONv 1 NON

SENa SEN1 SENi 1 SEN SENb SEN2 sENii 1 SEN

TERa TER1 TERi 1 TER TERa TER2 TERii 1 TER TERa TER2 TERiv 1 TER TERa TER3 TERiii 1 TER TERa TER4 TERvi 1 TER TERb TER4 TERv 1 TER

32

Table S4. Phenotypic features detected in different species of M. terrae complex.

Species Number of

strains

Nitrate

reductase

(% positive)

Pigmentation

(% positive)

Growth rate

M. arupense 51 negative (4%) none (0%) intermediate

M. engbaekii 7 positive (43%) pink (85%) rapid

M. hiberniae 1 negative (0%) pink (100%) intermediate

M. kumamotonense 34 positive (55%) none (6%) intermediate

M. nonchromogenicum 9 variable (55%) none (0%) intermediate

M. senuense 2 variable (50%) none (0%) slow

M. terrae 6 negative (0%) none (0%) intermediate

33

Table S5. Strains presenting conflicting or new sqvs, not assigned to any of the species of the M. terrae

complex.

Genetic region Combinations number 16S hsp65 rpoB

FI-07105 FI-07105 FI-07105 1 FI-07105 FI-11038 FI-11038 1 FI-09379 FI-09379 FI-09379 1 FI-09015 FI-09015 ARUx 1

ARUa ARU1 FI-07105 2 ARUa ARU1 HERvi 1 HERa FI-05196 FI-05196 1 HERa ENG4 FI-05196 2

FI-10193 FI-10193 FI-10193 1 KUMa FI-06258 FI-09100 1 KUMa KUM1 HERvii 1 TERc FI-06258 FI-06258 1 TERc FI-06258 FI-09100 1

34

Figure 1. Phylogenetic tree based on 16S rRNA sequences constructed using the neighbour-joining method

(bootstrapped 1,000 times). Bootstrap values >50% are given at nodes. Bar, 0.005 substitutions per

nucleotide position.

Figure 2. Phylogenetic tree based on concatenated sequences of 16S rRNA, hsp65 and rpoB constructed

using the neighbour-joining method, every combination of sequevars detected in the strains investigated was

included. Bootstrap was replicated 1,000 times; only values >50% are given at nodes. Bar, 0.01 substitutions

per nucleotide position.

Figure 3. Most frequently detected HPLC profiles of mycolic acids. A, pattern presented by M.

nonchromogenicum, M. longobardum, most M. arupense (70%) but occasionally found also within other

species of the M. terrae complex; B, pattern prevalent in the species M. kumamotonense; C, most frequent

profile of the species M. terrae; D, pattern of M. triviale; HMMIS, high molecular mass internal standard.

Figure S1. Phylogenetic tree based on hsp65 sequevars constructed using the neighbour-joining method

(bootstrapped 1,000 times). Bootstrap values >50% are given at nodes. Bar, 0.01 substitutions per nucleotide

position.

Figure S2. Phylogenetic tree based on rpoB sequevars constructed using the neighbour-joining method

(bootstrapped 1,000 times). Bootstrap values >50% are given at nodes. Bar, 0.02 substitutions per nucleotide

position.