Insights into the population structure of Mycobacterium tuberculosis using spoligotyping and RDRio in a southeastern Brazilian prison unit

1

3

4

5

6

7 Q1

8

9

10

111213 Q214151617181920

2122

2 4

2526272829

3031323334353637

3 8

60

61

62

63

Infection Genetics and Evolution xxx (2014) xxxndashxxx

MEEGID 1979 No of Pages 9 Model 5G

9 June 2014

Contents lists available at ScienceDirect

Infection Genetics and Evolution

journal homepage wwwelsevier comlocate meegid

Insights into the population structure of Mycobacterium tuberculosisusing spoligotyping and RDRio in a southeastern Brazilian prison unit

httpdxdoiorg101016jmeegid2014050311567-1348 2014 Published by Elsevier BV

uArr Corresponding author Address Laboratoacuterio de Microbiologia Celular(LAMICEL) Instituto Oswaldo Cruz Fiocruz Av Brasil 4365 21045-360 Rio deJaneiro RJ Brazil Telfax +55 21 2621598

E-mail address saadiocfiocruzbr (MHF Saad)

Please cite this article in press as Huber FD et al Insights into the population structure of Mycobacterium tuberculosis using spoligotyping and RDsoutheastern Brazilian prison unit Infect Genet Evol (2014) httpdxdoiorg101016jmeegid201405031

Feacute Dagmar Huber a Alexandra Saacutenchez b Harrison Magdinier Gomes c Sidra Vasconcellos cVeacuteronique Massari de Angela Barreto f Vanderci Cesconi b Silvia Maria de Almeida Machado aMichel K Gomgnimbou ghi Christophe Sola gh Bernard Larouzeacute dej Philip Noel Suffys cMaria Helena Feacuteres Saad auArra Laboratory of Cellular Microbiology Oswaldo Cruz Institute (IOC) Fiocruz Rio de Janeiro Brazilb Tuberculosis Control Program and Coordination Management in Prison Health State Department of Corrections Rio de Janeiro Brazilc Laboratory of Molecular Biology Applied to Mycobacteria IOC Fiocruz Brazild INSERM UMR_S 1136 Pierre Louis Institute of Epidemiology and Public Health Department of Social Epidemiology Francee Sorbonne Universiteacutes UPMC Univ Paris 06 UMR_S 1136 Pierre Louis Institute of Epidemiology and Public Health Department of Social Epidemiology Francef CRPHFENSP Fiocruz Rio de Janeiro Brazilg Univ Paris-Sud UMR8621 Orsay F-91405 Franceh CNRS Orsay F-91405 Francei Centre Muraz Bobo-Dioulasso Burkina Fasoj Department of Epidemiology and Quantitative Methods ENSPFiocruz Rio de Janeiro Brazil

a r t i c l e i n f o a b s t r a c t

39404142434445464748495051

Article historyReceived 7 March 2014Received in revised form 26 May 2014Accepted 27 May 2014Available online xxxx

KeywordsM tuberculosisPrisonSpoligotypingTuberculosisRv3062 (LigB)Large-sequence polymorphisms

52535455565758

Tuberculosis (TB) is still a serious public health problem continuing to be an important threat for con-fined populations We used spoligotyping to estimate the genotypic clades of Mycobacterium tuberculosisisolates from inmates in two blocks in a southeastern Brazilian prison unit with TB incidence rate of8185100000 The Latin American Mediterranean (LAM) clade is well represented in the country andthe LAM specific molecular markers RDRio large sequence polymorphism and the SNP on the Rv3062[ligB1212] were used to characterize spoligotype signatures from prison isolates Typing of RDRio and ligBincrease LAM clade from 667 (n = 72108) to 694 (n = 75) The LAM2 SIT17 (n = 23) and SIT179(n = 12) signatures comprised one third of all isolates followed by Haarlem (115 n = 12) T (87n = 9) and X (57 n = 6) clades Strains with unknown signatures represented 55 (n = 6) and four(37) did not match any lineage We observed RDRio among 64 (592) isolates and 54 (50) were ofthe LAM clade In particular the LAM2RDRio sub-lineage was significantly associated with clustering(p = 002) and its frequency was higher (32) when compared to that of the previous general TB casesin RJ (429) Overall cluster frequency defined by spoligotypingIS6110-RFLP was 62 The two evolu-tionary markers helped to evaluate some LAM signature misconceptions and demonstrate that LAM2RDRio was found with high frequency hitherto being unnoticed All these data allied to high clusteringimply that public health measures to minimize the escalation of TB in prison is essential and bothspoligotyping as well as RDRio would be useful tools to monitor the effects of the measures with respectto M tuberculosis lineage variation

2014 Published by Elsevier BV

59

64

65

66

67

1 Introduction

Tuberculosis (TB) is an infectious disease that continues to be aformidable public health problem and despite new tools and

68

69

70

71

72

interventions relentlessly spreads in low and middle income coun-tries leading to a significant impact worldwide This is even morepronounced when dealing with confined populations in an envi-ronment with high risk for TB infection such as prison (OPAS2008 Larouzeacute et al 2008)

In Rio de Janeiro (RJ) state prisons southeastern Brazil the gen-eral TB incidence rate in 2005 was over 3500100000 inhabitants(SEAPRJ 2010) the highest rate found in one unit (8185100000) while that of the general RJ population was 71100000

Rio in a

73 Q3

74

75

76 Q4

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

2 FD Huber et al Infection Genetics and Evolution xxx (2014) xxxndashxxx

MEEGID 1979 No of Pages 9 Model 5G

9 June 2014

(Saacutenchez et al 2005 2007 BRASIL 2006) As there is contactbetween prisoners and the general population it is essential toinvestigate the genotype of strains of the Mycobacterium tuberculo-sis complex (MTBC) for prevalence of specific patterns that mightindicate recent introduction in the prisons Molecular epidemiol-ogy could help to better understand the TB dynamics within theprison and the free population However data on M tuberculosismolecular typing in prison are limited in Brazil

Genotyping nowadays is a common tool in TB transmissionstudies and depending upon the study setting and objectives tech-niques such as IS6110-based restriction fragment length polymor-phism (RFLP) mycobacterial interspersed repetitive units-variablenumber tandem repeat (MIRU-VNTR) typing and spoligotypingare adopted Such standardized procedures on a worldwide levelhave deepened our knowledge on phylogeny and population struc-ture of organisms of MTBC Spoligotyping in particular an easy andcost-effective way for the assessment of M tuberculosis geneticdiversity has been widely applied for strain identification of theMTBC species and genotype family level and has been the basisfor the largest genotype database of M tuberculosis namelySpolDB4 and SITVITWEB (Demay et al 2012)

Recently a263 Kb deletion was described that characterizes apredominant MTBC genotype family initially observed in RJ andtherefore called RDRio present exclusively in isolates of the LatinAmerican Mediterranean (LAM) spoligotypes This genotype isidentified in about 15 of the TB cases worldwide and the deletioncomprises 10 genes including two of the PPE family (Lazzariniet al 2007 2008 Weisenberg et al 2012) In addition DosVultos et al (2008) have described polymorphisms in several genesinvolved in MTBC replication recombination and repair functions(3R genes) and among these the Rv3062 (ligB) gene coding foran ATP-dependent DNA ligase presents a single polymorphism atthe 1212 tcc (tcg) nucleotide (SNP) position exclusive for theLAM clade (Abadia et al 2010)

In this study spoligotyping was used to establish the genotypesof the circulating MTB complex isolated from patients confined intwo blocks of the C Unit of the Bangu Penitentiary Complex (BPC)with the highest TB incidence rate in Rio de Janeiro city Brazil(Saacutenchez et al 2005) In order to assess the contribution of theLAM clade in this burden set we adopted two evolutionary mark-ers the RDRio large deletion and the LAM-specific SNP on theRv3062 gene

2 Materials and methods

21 Population and study setting

Full description of the prison is available elsewhere (Saacutenchezet al 2012) Briefly in the Rio de Janeiro (RJ) State prison systemthere are around 20000 inmates distributed in 37 prison unitsand the 106 TB patients included in the present study were incar-cerates in one of these units the C Unit of the BPC the TB incidencerate in 2004 (8185100000) being the highest All inmates wereover 18 years old with an average incarceration of 4 years (maxi-mum 8 years) The C Unit accommodated 1418 inmates on a singlelevel two independent blocks organized in block A (n = 785inmates) with 12 cells and block B (n = 633) with 9 cells Each cellcontained around 60 to 75 inmates with a projected maximum of40 Transit from one block to another within the same block andfrom one cell to another occurs but records are not availableThe blocks are poorly ventilated and illuminated

22 Bacterial isolates

The isolates studied here were collected previously during a per-iod of 13 months as part of a prospective clinical-epidemiological

Please cite this article in press as Huber FD et al Insights into the populationsoutheastern Brazilian prison unit Infect Genet Evol (2014) httpdxdoiorg

study to evaluate X-ray as a screening procedure for TB diagnosisand control among prisoners (Saacutenchez et al 2007) All inmateswere screened for TB in July 2005 and again in July 2006 Thosediagnosed within this period due to individual spontaneous requestwere termed lsquolsquoincident casesrsquorsquo A total of 193 pulmonary cases werediagnosed and 108 isolates from 106 patients were available fortyping Strain identification for the MTBC level was performed bybiochemical tests (growth inhibition by p-nitrobenzoic acid resis-tance to thiophene-2-carboxylic acid hydrazide niacin nitratereductase and catalase thermo inactivation) while drug susceptibil-ity testing towards isoniazid (INH) rifampin (RIF) streptomycin(SM) and ethambutol (EMB) was carried out at the Reference CenterProf Helio Fraga (Fiocruz Curicica RJ) according to the standardproportion method (Canetti et al 1969)

23 Spoligotyping

Bacterial DNA was prepared by either CTAB based extraction(Van Soolingen et al 1991) or thermolysis (Abadia et al 2011)The PCR protocol for spoligotyping was conducted according toKamerbeek et al (1997) while the hybridization was effected ona microbead-based suspension as described by Zhang et al(2010) with modification In brief the DR region was amplifiedby PCR with 10ndash20 ng of DNA 10 lM of primers 50 biotinylatedDRa (50 GGTTTTGGGTCTGACGAC 30) and DRb (50 CCGAGAGGGGACGGAAAC 30) 5 M betaine and 1U of Taq DNA polymerase in a finalvolume of 25 lL For hybridization a microsphere mixture of the43 spacer format previously prepared was used adding 2 lL ofPCR product to 33 lL of the microsphere mixture The suspensionwas submitted to denaturation at 95 C for 10 min followed by ahybridization step of 52 C for 20 min and centrifugation of 1 minat 2250g The microbeads were resuspended in 33 lL of TE buffer(10 mM Tris 1 mM EDTA pH 80) and 25 lL of reporter mix(2ndash4 lglL of streptavidin-R-phycoerythrin diluted in 1X 3 Mtetramethylammonium chloride buffer) and then added to amicroplate After incubation the amount of hybridized ampliconson the beads was analyzed in a Luminex 200 analyzer accordingto the manufacturers recommendations

The spoligotype-based clusters were confirmed by their IS6110restriction fragment length polymorphism (IS6110-RFLP) patternsavailable from a previous study from our group (Saacutenchez et al2012) Laboratory cross-contamination was ruled out by verifyingin the laboratory register that isolates with an identical fingerprintpattern had not been processed the same day

24 Identification of the RDRio genotype by multiplex PCR

Identification of the RDRio deletion was established as describedpreviously (Lazzarini et al 2007 Gibson et al 2008) The RDRio

and wildtype (Wt) genotypes are represented by a PCR productof respectively 1175 and 530 bp while the presence of both bandssuggested infection with at least two isolates each having one ofthe genotypes hereafter referred to as RDRioWt

25 LigB1212 SNP detection

Strains of the RDRio sublineage that had a spoligotype not repre-sentative of the LAM lineage were submitted to SNP detection asdescribed previously (Dos Vultos et al 2008) with some modifica-tions the entire gene sequence was amplified with 40 mM of theprimers ligB (Rv3062) F (50 CCAGGCGGTATTGGTAAAGA 30) and ligBR (50 GTCGACGAGCGTGAATCTG 30) in reactions containing 10 ng ofDNA 1875 lL KAPA2G Robust HotStart Readymix (Kapa BioSys-tems MA USA) and 1 DMSO Amplifications were performed withthe following program initial denaturation of 1 min at 95 C fol-lowed by 35 cycles of 3000 at 95 C 3000 at 50 C 3000 at 72 C and

structure of Mycobacterium tuberculosis using spoligotyping and RDRio in a101016jmeegid201405031

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

FD Huber et al Infection Genetics and Evolution xxx (2014) xxxndashxxx 3

MEEGID 1979 No of Pages 9 Model 5G

9 June 2014

a final extension of 50 at 72 C The amplicons of 1924 bp were ana-lyzed on agarose gel in 1X TAE (40 mM Trisndashacetate and 1 mMEDTA) and stained with ethidium bromide For the sequencingreaction we used the ABI PRISM kit BigDyeTM version Terminator3 (Applied Biosystems) and internal primers ligB Fi (50 GGTCAGGATCTACACCCAA 30) and ligB Ri (50 TTCGAGCAAGTCGGTACCAT 30)Sequence reading was effected on an ABI3730 96 capillary sequen-cer (Applied Biosystems) The sequences were aligned with thepublished sequence of the M tuberculosis H37Rv genome depositedin GenBank (httpwww sangeracukProjectsMtuberculosis)The SNP at position 1212 is tcg representing LAM isolates or tccrepresentative for other genotype families

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

26 Data analysis

Demographic data spoligotyping and RDRio patterns were intro-duced in Excel 2007 Spreadsheets and analyzed with EPInfo 6 (CDCAtlanta GA USA) Phylogenetic clades identification was performedby searching in the SITVIT WEB database (httpwwwpasteur-guadeloupefr8081SITVIT_ON LINE) which defined genotypiclineagesublineages including specific signatures for the speciesof MTBC as well as norms defining M tuberculosis specifically (LatinAmerican Mediterranean-LAM clade with 12 sub lineages East Afri-can-India-EAI clade with 9 sublineages Ill-defined T clade with 5sublineages X clade that is known to have IS6110-low bandingwith 4 sublineages Harleen clade and Manu family with 3 sub lin-eages each Central Asian clade and 2 sub lineages finally Beijingand S clades) In this database Spoligotype International Type(SIT) designates patterns shared by two or more patient isolateswhereas an orphan defines patterns reported for a single isolateNew spoligopatterns whether orphan or encountered in multipleisolates in the present population but not present in SpolDB4 andSITVIT2 WEB were repeated Spoligotypes were also compared withthose in the SpolBR1 database harboring genotypes from the gen-eral Brazilian population half of these from Rio de Janeiro (Gomeset al 2012) Cluster analysis by Spoligotypes andor IS6110-RFLPpatterns was carried out with the software BioNumerics 53 soft-ware (Applied Maths NV Sint-Martens-Latem Belgium) After-wards for correlation of clustering with patients and isolatecharacteristics cluster was then defined as two or more patientsinfected with M tuberculosis harboring the same genotypes by bothspoligotypeIS6110-RFLP Categorical variables were compared bythe v2 test The Fisher exact test with the Yates correction whenapplicable was applied and statistical significance was p 6 005

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

3 Results

31 Spoligotype frequencies

Spoligotypes were distributed in four clades the most frequentbeing the Latin American Mediterranean (LAM 72108 666) fol-lowed by Haarlem (H 12108 11) T (8108 74) and X (610855) Unknown (Uk) signatures were observed in six isolates(55) while four isolates (37) were assigned as new patterns

A total of 28 different international spoligotypes (SIT) wereobserved among the 108 isolates (Fig 1) and 86 (796) belongedto 14 clusters ranging in size from 2 to 23 isolates Within theLAM clade SIT17 was the predominant genotype and also definedthe largest cluster (n = 23 213) followed by SIT179 (n = 12111) one third of the isolates therefore being of the LAM2 sub-lineage Other frequent sub-lineages were LAM9 (n = 15 16) rep-resented by SIT42 (n = 9 83) and SIT1800 (n = 6 55) LAM3SIT33 (n = 8 74) and LAM6SIT 64 (n = 6 55) Interestingly therewere two isolates with the spoligotype T4-CEU SIT39 (2)described for the first time in Brazil

Please cite this article in press as Huber FD et al Insights into the populationsoutheastern Brazilian prison unit Infect Genet Evol (2014) httpdxdoiorg

32 Distribution of the RDRio lineage

The RDRio deletion was observed in 64 of the 108 isolates(592) and 54 of those belonged to the LAM clade (50) includingthe sub-lineage LAM1 andor LAM4 (n = 3) LAM2 (n = 33) LAM5(n = 1) LAM9 (n = 15) and surprisingly two LAM3 isolates TheRDRio genotype was however also witnessed in other clades suchas T (n = 5) and H (n = 1) as well as the isolates of Uk (n = 1) andnew (n = 3) patterns Upon further genetic analysis five of these12 isolates (10 non LAM and two LAM3) had a LAM definingligB1212 SNP including one isolate of the LAM3 sub-lineage twoof T clade and one each of Uk and new signature (Table 1) there-fore increasing the number of RDRio isolates among members ofthe LAM clade to 57 (527) and increasing the frequency of LAMin this study to 694 (75108) In eight isolates (74) we docu-mented a mixed profile of RDRioWt two of these being assignedas new and the T1ndashT4 clade three of the LAM2 and the other threeas LAM9 The main Wt sub-lineages were LAM3 LAM6 and H3 allin 7 of 8 isolates respectively Uk (56) T1 (34) H1 and X1 (33)T4CEU1 and X3 (22) and finally the new (23)

Regarding other characteristics associated with RDRio statussuch as drug resistance our data indicated that the most resistantisolates (68) including the MDR were Wt (Table 2) In additionthere was no association between positive AFB in sputum andRDRio status

33 Distribution of clades according to both spoligotyping and RFLPIS6110 analysis

Although combined spoligotypingIS6110-RFLP analysis fordefinition of clustering reduced the number of clusters to 62 (67108 and ranging in size from two to 14 isolates) the numberof clustered patterns increased to 17 since some spoligoclusterswere discriminated by RFLP Results including demographic epide-miologic drug resistance and genotypic data in 17 clusters com-prising 67 isolates of 66 patients are shown in Table 2 Threespoligoclusters were subdivided by RFLP (i) the cluster formedby SIT17 into two RFLP clusters containing 14 and 2 isolates andseven strains with unique patterns (ii) SIT33 in three clustersone with three and the other two with two isolates each and(iii) SIT64 subdivided in two clusters one with four and the otherwith two isolates plus another two isolates with unique patternsThe second biggest spoligotype cluster (SIT179) as well as clustersformed by SIT1800 SIT119 SIT2511 SIT92 SIT47 SIT50 and SIT53remained clustered by RFLP The two clusters formed by membersof the X clade SIT92 and SIT119 displayed a four band RFLPpattern one isolate presenting a unique pattern Finally clustersSIT42 and SIT2498 of a total of 12 isolates were reduced to twoisolates each One must be aware that SIT179 and SIT17 ofLAM2RDRio contributed to 42 of the clustered isolates (clustersC1 and D2)

Among drug resistant isolates 50 (48) were LAM6SIT64 twowere H1SIT47 and one each of LAM2179 and LAM9SIT42 sub-lineage The IS6110-RFLP typing confirmed clustering for all exceptfor the LAM9SIT42 and one LAM6SIT64 isolate

Notable was a patient diagnosed in July 2005 infected with theLAM2 SIT179 isolate (cluster CC1 Table 2) INH resistant and pre-senting a mixed RDRioWt profile This patient was treated with thefirst line specific chemotherapy for TB Despite good clinical evolu-tion 5 months later (December 2005) a Wt (for RDRio) LAM6 SIT64 (Cluster AA1) isolate harboring the MDR profile was evi-denced An isolate with the same genotype had been obtained inSeptember 2005 from another patient sharing the same B blockalthough in different cells (Table 2) and also treated for TB inthe past In addition a third patient of the same block yielded anINH mono resistantWt isolate belonging to this AA1 cluster It

structure of Mycobacterium tuberculosis using spoligotyping and RDRio in a101016jmeegid201405031

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

Fig 1 Spoligotype patterns and number of shared international typing (SIT) of 108 M tuberculosis strains isolated from 106 inmates incarcerated in a southeastern Brazilianprison unit

Table 1Distribution of RDRio lineage and ligB1212 mutation among M tuberculosis isolated from inmates incarcerated in a southeastern Brazilian prison unit assigned as LAM3 and non-LAM clade by spoligotyping

Spoligotype pattern SpolDB4 clade (SIT) RDRio lineagea ligB SNPb

LAM3(SIT33) 2 1T1 (SIT53) 2 1T1 (SIT291) 1 0T3 (orphan) 1 1T3-T4 (SIT73) 1 NDH3 (SIT50) 1 0Uk (2535) 1 1New 1 1New 1 0New 1 0

ND not donea Number of strains with RDRio deletionb LigB1212 number of isolates with tcc (tcg) single nucleotide polymorphism (SNP) at the gene Rv3062 (ligB) gene associated with Latin American Mediterranean (LAM)

clade

4 FD Huber et al Infection Genetics and Evolution xxx (2014) xxxndashxxx

MEEGID 1979 No of Pages 9 Model 5G

9 June 2014

is noteworthy that all INH or MDR strains were infecting inmatesof the block B

34 Correlation of clustering according to both spoligotypingIS6110-RFLP with patients and isolate characteristics

When comparing clustering with patient and isolate character-istics no correlation was established with drug resistance and spe-cific variables such as age interval RDRio status positive AFBsputum stain incarceration block isolate data and the LAM cladeHowever the frequency of LAM2 clustered isolates was slightlymore frequent when compared to those of other LAM types(p = 006) and significantly higher than those in the non-LAM clade(p = 001) Similarly there was no association for the LAMRDRio

lineage but there was significance when comparing LAM2RDRio

(p = 002) against other sub lineage of LAMRDRio isolates (Table 3)

Please cite this article in press as Huber FD et al Insights into the populationsoutheastern Brazilian prison unit Infect Genet Evol (2014) httpdxdoiorg

4 Discussion

The Euro American clade comprises the families Haarlem LAMT S and X as defined by spoligotyping (Gagneux and Small 2007)and predominates in South America Central America Europe andAfrica (Brudey et al 2006) The LAM clade is the most representa-tive in South American countries with frequencies of 493 inColombia (Cerezo et al 2012) 52 in Paraguay (Candia et al2007) 60 in Venezuela (Aristimuntildeo et al 2006) and 46 in Brazil(Gomes et al 2012) Within Brazil the distribution of this lineageis not homogeneous and presents lower frequencies in the moresoutheastern Brazilian states (35) and a considerably higher fre-quency in the northeast (gt70) of the country possibly due tothe high frequency of LAM in Portugal (David et al 2005) andthe migration of the Portuguese in early 16th century to thisregion In the present study the 694 rate of LAM is slightly higher

structure of Mycobacterium tuberculosis using spoligotyping and RDRio in a101016jmeegid201405031

Tabl

e2

Dem

ogra

phic

data

of66

inm

ates

inca

rcer

ated

ina

sout

heas

tern

Braz

ilian

pris

onun

it

drug

resi

stan

cean

dge

noty

ping

data

ofth

eir

67M

tu

berc

ulos

isis

olat

esbe

long

ing

toth

e17

clus

ters

defi

ned

byco

ncor

dant

spol

igot

ypin

gIS

6110

-RFL

Pan

alyz

ean

dRD

Rio

typi

ng

(con

tinu

edon

next

page

)

FD Huber et al Infection Genetics and Evolution xxx (2014) xxxndashxxx 5

MEEGID 1979 No of Pages 9 Model 5G

9 June 2014

Please cite this article in press as Huber FD et al Insights into the population structure of Mycobacterium tuberculosis using spoligotyping and RDRio in asoutheastern Brazilian prison unit Infect Genet Evol (2014) httpdxdoiorg101016jmeegid201405031

Tab

le2

(con

tinu

ed)

6 FD Huber et al Infection Genetics and Evolution xxx (2014) xxxndashxxx

MEEGID 1979 No of Pages 9 Model 5G

9 June 2014

Please cite this article in press as Huber FD et al Insights into the population structure of Mycobacterium tuberculosis using spoligotyping and RDRio in asoutheastern Brazilian prison unit Infect Genet Evol (2014) httpdxdoiorg101016jmeegid201405031

351351351351351351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

Table 3Clustered and non-clustered isolates as defined by combining Spoligotyping IS6110-RFLP in relation to gender bock of incarceration strain isolate data drug resistanceand genotype lineages identified in a southeastern Brazilian prison unit

Variables Clustered N = 67()

Non-clusteredN = 41 ()

Pvaluea

Age630 years 56 (83) 33 (80) 074gt30 years 6 (9) 5 (12)Missing 5 3

Block of incarcerationA 38 (567) 18 (439) 099B 29 (433) 21 (512)Missing 0 2

Strain isolate data2005 22 (328) 11 (268)Incidentb 25 (373) 14 (341) 0982006 18 (268) 14 (341) 054Missing 2 2

Strain drug susceptibility testSusceptible 60 (896) 39 (95) 025Resistant 7 (104)1 1 (24)2

Missing 1

LineagesLAM 50 (746) 26 (634) 030Other clades 17 (253) 15 (366)

LAMRDRio 36 (537) 21 (512) 056LAMRDWT 12 (179) 4 (98)

LAM2 28 (418) 7 (17) 001Other clades 39 (582) 34 (829)

LAM2RDRio 26 (389) 7 (17) 002Other LAMRDRio

sublineage9 (134) 11 (268)

LAM2 28 (418) 7 (17) 006Other LAM sublineage 21 (313) 16 (39)

All RDRioAFB positive 15 (223) 12 (292) 017All Non-RDRioAFB

positive9 (134) 18 (439)

All RDRio isolates 37 (552) 27 (658) 091All non-RDRio isolates 22 (328) 14 (341)

LAMQ6 Latin American Mediterranean clade RD region deleted Wt wild type AFBacid fast bacilli sputum stain

a v Square Fisher exact test significant P005b TB diagnosed after 2005 first screening

FD Huber et al Infection Genetics and Evolution xxx (2014) xxxndashxxx 7

MEEGID 1979 No of Pages 9 Model 5G

9 June 2014

(p = 012) than that for Rio de Janeiro southeastern Brazil (58Gomes et al 2012) and significantly higher than that of a prisonin southern Brazil (40 p = 0001) (Kuhleis et al 2012) indicatingthat LAM has adapted to our geographic area and this is reflectingin the prison

Although there was no significant association between LAM iso-lates and clustering significant clustering was apparent for bothLAM2 and LAM2RDRio sub-lineage isolates as well as In an earlierstudy on spoligotypes of M tuberculosis isolates collected between1996 and 2005 in 11 states of Brazil over 50 of the strains werefrom Rio de Janeiro the most frequent genotypes being LAM9(SIT42 103) T1 (SIT53 758) H3 (SIT50 542) LAM2 (SIT17427) LAM6 (SIT64 366) and H1 (SIT47 276) (Gomes et al2012) Most of these SITs represent micro-clusters in our prisonpopulation as confirmed by IS6110-RFLP (Table 2) and this is con-cordant with the worldwide prevalence suggesting easy transmis-sion (Gibson et al 2008) However LAM2 genotypes SIT179 andSIT17 were detected in 32 of the inmate TB cases against 429in the Gomes et al (2012) study in the same period and in thesame region contributing to 42 of the clustered strains (clustersC1 and D2 Table 2) in the prison Besides Brazil SIT179 has rarelybeen reported outside French Guyana USA and Portugal all havinghistorical commercial or geographic links with Brazil Although

Please cite this article in press as Huber FD et al Insights into the populationsoutheastern Brazilian prison unit Infect Genet Evol (2014) httpdxdoiorg

SIT17 isolates were subdivided by IS6110-RFLP in two sub clustersmost remained clustered but all of the SIT179 isolates were con-firmed as IS6110-RFLP clustered (Table 2)

The RDRio sub lineage was initially described as a predominantcause of TB in Rio de Janeiro (Lazzarini et al 2007) and later on inmany other countries making it an important cause of tuberculosisin the world (Weisenberg et al 2012 Gibson et al 2008) Thisclonal sub-lineage of LAM has been associated with different fac-tors for TB severity including increase in sputum bacillary load(Lazzarini et al 2007 2008) an association neither proven signif-icant in our study (p = 017) nor in another study in Brazil (Vinhaset al 2013) LAM1 (SIT20) and LAM2 (SIT17) have been reported asbeing exclusively RDRio (Lazzarini et al 2007) which was con-firmed in our study Although LAM1 strain TB cases were docu-mented with considerable frequency in Rio de Janeiro between1996ndash2005 (3185 364) (SITVITWEB Gomes et al 2012) SIT20was isolated merely from one patient only among inmates Inter-estingly SIT179 was seldom described in Brazil (nine strainsaccording to SITVITWEB) and until 2005 only four of these in Riode Janeiro (Gomes et al 2012) Therefore at that time the highfrequency of LAM2RDRio in the prison population might havereflected a more efficient transmission of this sublineage in thisparticular scenario (overcrowded low level of hygiene poor illu-minated and ventilated) when compared to the general populationin Rio de Janeiro Even though recent data did not evidence a fit-ness difference between Brazilian RDRio LAM and non RDRio strains(Von Groll et al 2010) this was concluded after observing replica-tion in growth medium and direct evidence for more efficienttransmission of RDRio strains reported by evaluation transmissiondynamics in the city of New York (Weisenberg et al 2012) How-ever recently it was reported that the spread of RDRio strains isdetermined mainly by person to person transmission rather thanby virulence factors of this strains (Narvskaya et al 2014) In addi-tion LAM2RDRio was a successful clustered sub-lineage in thisprison environment As far as we know this is the first popula-tion-based study evaluating the presence of RDRio in a prison pop-ulation in Rio de Janeiro and further investigation is necessary todetermine if this particular LAM2 sub-lineage whose clonalitywas confirmed by a more discriminatory typing method (IS6110-RFLP ndash C1 and D2 clusters Table 2) was successfully introducedin the other units of the BCP

As observed in other studies (Gibson et al 2008 David et al2012 Weisenberg et al 2012 Vinhas et al 2013) a small amount(65) of the RDRio isolates were not of the LAM lineage This war-rants further evaluation with other associated LAM markers suchas Ag85C103 since Weisenberg et al (2012) reported that all non-LAMRDRio were categorized as members of the LAM family underAg85C103 SNP In a recent publication on RDRio in Vitoria the Stateof Espirito Santo also located in southeastern Brazil no SNP anal-ysis was performed to confirm the LAM nature of the RDRio strains(Vinhas et al 2013)

The RDRio genotype has been associated with drug resistanceand the LAM subfamily of LAM9 LAM1 LAM2 and LAM5 (Davidet al 2012 Dalla Costa et al 2013) However in our study as wellas others (Weisenberg et al 2012 Cardoso Oelemann et al 2011)most of the drug and multi-drug resistant strains were of the Wtgenotype Notwithstanding the lack of association here may relateto the small number of resistant strains found in this particularprison and the fact that 50 of the drug resistant were in LAM6 iso-lates One of the MDR patients was initially infected with a mixedRDRioWt isolate exhibiting INH resistance only (cluster CC1Table 2) and treatment may have selected a fraction of a WtMDR strain (Cluster AA1) infecting a second patient both inmatesof block B One must be aware that this possible dual infectionseems not to have been detected by the other genotyping assaysfor which we do not have a reasonable explanation On the other

structure of Mycobacterium tuberculosis using spoligotyping and RDRio in a101016jmeegid201405031

440

441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462 Q5

463

464

465

466

467

468

469

470

471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507

508509510511512513514515516517518519520521522523524525526527528529530531532533534535

8 FD Huber et al Infection Genetics and Evolution xxx (2014) xxxndashxxx

MEEGID 1979 No of Pages 9 Model 5G

9 June 2014

hand this second event of TB could have been transmitted by thesecond patient diagnosed early in September 2005 and the RDRio

wild type features just a coincidence (Table 2) A higher numberof isolates with RDRio mixed profiles were described elsewherebut no further investigation was effected (Weisenberg et al2012) Identification of a mixed isolate from the same patientmay be a concern for possible new strain transmission or diseaserelapse in a confined environment such as prison However thisrequires further investigation

In conclusion LAM was the most frequent clade found and thetwo evolutionary markers helped to evaluate some misleading sig-natures as well as displaying that LAM2RDRio was significantlyassociated to clustering and as this was hitherto unnoticed itmight account for favorable conditions for the spread in this partic-ular prison However as we did not study the isolates from theentire penitentiary complex we cannot confirm that LAM2RDRio

is also the most frequent isolate in the other units of BPC All thesedata allied to high clustering in this prison imply that public healthmeasures to minimize the escalation of TB in prison are essentialand both spoligotyping and RDRio would be useful tools to monitorthe effect of the measures in M tuberculosis lineage variation

536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593

Uncited references

Mokrousov et al (2014) Sauacutede (2014) and SEAPRJ (2006 2004)

Acknowledgements

This work was partially supported by the Brazilian ConselhoNacional de Desenvolvimento Cientiacutefico e Tecnoloacutegico (Universaland PROEP-CNPq) and Fundaccedilatildeo de Amparo agrave Pesquisa do Estadodo Rio de Janeiro (FAPERJ) Rio de Janeiro Brazil English reviewand revision by Mitchell Raymond Lisbon natural of ChicagoIllinois USA UCLA 1969

References

Abadia E Zhang J dos Vultos T Ritacco V Kremer K Aktas E Matsumoto TRefregier G van Soolingen D Gicquel B Sola C 2010 Resolving lineageassignation on Mycobacterium tuberculosis clinical isolates classified byspoligotyping with a new high-throughput 3R SNPs based method InfectGenet Evol 10 1066ndash1074

Abadia E Zhang J Ritacco V Kremer K Ruimy R Rigouts L Gomes HM EliasAR Fauville-Dufaux M Stoffels K Rasolofo-Razanamparany V Garcia deViedma D Herranz M Al-Hajoj S Rastogi N Garzelli C Tortoli E SuffysPN van Soolingen D Refreacutegier G Sola C 2011 The use of microbead-basedspoligotyping for Mycobacterium tuberculosis complex to evaluate the quality ofthe conventional method providing guidelines for quality assurance whenworking on membranes BMC Infect Dis 11 110ndash123

Aristimuntildeo L Armengol R Cebollada A Espantildea M Guilarte A Lafoz CLezcano MA Revillo MJ Martiacuten C Ramiacuterez C Rastogi N Rojas J de SalasAV Sola C Samper S 2006 Molecular characterisation of Mycobacteriumtuberculosis isolates in the first national survey of anti-tuberculosis drugresistance from Venezuela BMC Microbiol 10 (6) 18

Brudey K Driscoll JR Rigouts L Prodinger WM Gori A Al-Hajoj SA Allix CAristimuntildeo L Arora J Baumanis V Binder L Cafrune P Cataldi A CheongS Diel R Ellermeier C Evans JT Fauville-Dufaux M Ferdinand S Garcia deViedma D Garzelli C Gazzola L Gomes HM Guttierez MC Hawkey PMvan Helden PD Kadival GV Kreiswirth BN Kremer K Kubin M KulkarniSP Liens B Lillebaek T Ho ML Martin C Martin C Mokrousov INarvskaiumla O Ngeow YF Naumann L Niemann S Parwati I Rahim ZRasolofo-Razanamparany V Rasolonavalona T Rossetti ML Ruumlsch-GerdesS Sajduda A Samper S Shemyakin IG Singh UB Somoskovi A SkuceRA van Soolingen D Streicher EM Suffys PN Tortoli E Tracevska TVincent V Victor TC Warren RM Yap SF Zaman K Portaels F RastogiN Sola C 2006 Mycobacterium tuberculosis complex genetic diversitymining the fourth international spoligotyping databases (Spol DB4) forclassification population genetics and epidemiology BMC Microbiol 6 23ndash40

Candia N Lopez B Zozio T Carrivale M Diaz C Russomando G de RomeroNJ Jara JC Barrera L Rastogi N Ritacco V 2007 First insight intoMycobacterium tuberculosis genetic diversity in Paraguay BMC Microbiol 8 (7)17

Canetti G Fox W Khomenko A Mahler HT Menon NK Mitchison DA RistN Smelev NA 1969 Advances in techniques of testing mycobacterial drug

Please cite this article in press as Huber FD et al Insights into the populationsoutheastern Brazilian prison unit Infect Genet Evol (2014) httpdxdoiorg

sensitivity and the use of sensitivity tests in tuberculosis control programmesBull World Health Organ 41 21ndash43

Cardoso Oelemann M Gomes HM Willery E Possuelo L Batista Lima KVAllix-Beacuteguec C Locht C Goguet de la Salmoniegravere YO Gutierrez MC SuffysP Supply P 2011 The forest behind the tree phylogenetic exploration of adominant Mycobacterium tuberculosis strain lineage from a high tuberculosisburden country PLoS One 6 e18256

Cerezo I Jimeacutenez Y Hernandez J Zozio T Murcia MI Rastogi N 2012 A firstinsight on the population structure of Mycobacterium tuberculosis complex asstudied by spoligotyping and MIRU-VNTRs in Bogotaacute Colombia Infect GenetEvol 12 657ndash663

Dalla Costa ER Lazzarini LC Perizzolo PF Diacuteaz CA Spies FS Costa LLRibeiro AW Barroco C Schuh SJ da Silva Pereira MA Dias CF GomesHM Unis G Zaha A Almeida da Silva PE Suffys PN Rossetti ML 2013Mycobacterium tuberculosis of the RDRio genotype is the predominant cause oftuberculosis and associated with multidrug resistance in Porto Alegre CitySouth Brazil J Clin Microbiol 51 1071ndash1077

David S Barros V Portugal C Antunes A Cardoso A Calado A Sancho L deSousa JG 2005 Update on the spoligotypes of Mycobacterium tuberculosiscomplex isolates from the Fernando Fonseca Hospital (Amadora-SintraPortugal) Rev Port Pneumol 11 513ndash531

David S Duarte EL Leite CQ Ribeiro JN Maio JN Paixatildeo E Portugal CSancho L Germano de Sousa J 2012 Implication of the RD(Rio)Mycobacterium tuberculosis sublineage in multidrug resistant tuberculosis inPortugal Infect Genet Evol 12 1362ndash1367

Demay C Liens B Burguiegravere T Hill V Couvin D Millet J Mokrousov I Sola CZozio T Rastogi N 2012 SITVITWEB A publicly available internationalmultimarker database for studying Mycobacterium tuberculosis genetic diversityand molecular epidemiology Infect Genet Evol lthttpwwwpasteur-guadeloupefr8081SITVIT_ONLINEgt

Dos Vultos T Mestre O Rauzier J Golec M Rastogi N Rasolofo V Tonjum TSola C Matic I Gicquel B 2008 Evolution and diversity of clonal bacteriathe paradigm of Mycobacterium tuberculosis PLoS One 6 e1538

Gagneux S Small PM 2007 Global phylogeography of Mycobacterium tuberculosisand implications for tuberculosis product development Lancet Infect Dis 7328ndash337

Gibson AL Huard RC Gey van Pittius NC Lazzarini LC Driscoll J KurepinaN Zozio T Sola C Spindola SM Kritski AL Fitzgerald D Kremer KMardassi H Chitale P Brinkworth J Garcia de Viedma D Gicquel B PapeJW van Soolingen D Kreiswirth BN Warren RM van Helden PD RastogiN Suffys PN Lapa e Silva J Ho JL 2008 Application of sensitive and specificmolecular methods to uncover global dissemination of the major RDRiosublineage of the Latin American-Mediterranean Mycobacterium tuberculosisspoligotype family J Clin Microbiol 46 1259ndash1267

Gomes HM Elias AR Oelemann MA Pereira MA Montes FF Marsico AGKritski AL Filho Ldos A Caldas PC Possuelo LG Cafrune P Rossetti MLLucena N Saad MH Cavalcanti HR Leite CQ de Brito RC Lima K LopesML Souza M Trindade Rde C Zozio T Sola C Rastogi N Suffys PN 2012Spoligotypes of Mycobacterium tuberculosis complex isolates from patientsresidents of 11 states of Brazil Infect Genet Evol 12 649ndash656

Kamerbeek J Schouls L Kolk A van Agterveld M van Soolingen D Kuijper SBunschoten A Molhuizen H Shaw R Goyal M van Embden J 1997Simultaneous detection and strain differentiation of Mycobacterium tuberculosisfor diagnosis and epidemiology J Clin Microbiol 35 907ndash914

Kuhleis D Ribeiro AW Costa ER Cafrune PI Schmid KB Costa LL RibeiroMO Zaha A Rossetti ML 2012 Tuberculosis in a southern Brazilian prisonMem Inst Oswaldo Cruz 107 909ndash915

Larouzeacute B Saacutenchez A Diuana V 2008 Tuberculosis behind bars in developingcountries a hidden shame to public health Trans R Soc Tropic Med Hyg 102841ndash842

Lazzarini LC Huard RC Boechat NL Gomes HM Oelemann MC Kurepina NShashkina E Mello FC Gibson AL Virginio MJ Marsico AG Butler WRKreiswirth BN Suffys PN Lapa e Silva JR Ho JL 2007 Discovery of a novelMycobacterium tuberculosis lineage that is a major cause of tuberculosis in Riode Janeiro Brazil J Clin Microbiol 45 3891ndash3902

Lazzarini LC Spindola SM Bang H Gibson AL Weisenberg S da SilvaCarvalho W Augusto CJ Huard RC Kritski AL Ho JL 2008 RDRioMycobacterium tuberculosis infection is associated with a higher frequency ofcavitary pulmonary disease J Clin Microbiol 46 2175ndash2183

Mokrousov I Vyazovaya A Narvskaya O 2014 Mycobacterium tuberculosisLatin American-Mediterranean family and its sublineages in the light of robustevolutionary markers J Bacteriol 196 1833ndash1841

OPAS Organizacioacuten Panamericana de la Salud Guiacutea para el control de latuberculosis en poblaciones privadas de libertad de Ameacuterica Latina y elCaribersquorsquoWashington DC OPS 2008 lthttpwww1pahoorgProjectaspSEL=PRampLNG=SPAampID=352ampPRGRP=docs_gengt (accessed March 10 2013)

Saacutenchez AR Gerhardt G Natal S Capone D Espinola A Costa W Pires JBarreto A Biond E Larouzeacute B 2005 Prevalence of pulmonary tuberculosisand comparative evaluation of screening strategies in a Brazilian prison Int JTuberc Lung Dis 9 633ndash639

Saacutenchez A Huber FD Massari V Barreto A Camacho LA Cesconi V SaadMH Larouze B 2012 Extensive Mycobacterium tuberculosis circulation in ahighly endemic prison Need for urgent environmental interventionsEpidemiol Infect 140 1853ndash1857

Saacutenchez AR Massari V Gerhardt G Barreto AW Cesconi V Pires J EspiacutenolaAB Biondi E Larouzeacute B Camacho LA 2007 Tuberculosis in Rio de Janeiro

structure of Mycobacterium tuberculosis using spoligotyping and RDRio in a101016jmeegid201405031

594595596597598599600601602603604605606607608609610611612

613614615616617618619620621622623624625626627628629

630

FD Huber et al Infection Genetics and Evolution xxx (2014) xxxndashxxx 9

MEEGID 1979 No of Pages 9 Model 5G

9 June 2014

prisons Brazil an urgent public health problem Cad Saude Publica 23 545ndash552

Sauacutede Secretaria de Estado de Tuberculose RJ ainda eacute o estado do Brasil com maiorincidecircncia Rio de Janeiro Brasil lthttpwwwsauderjgovbrimprensa -noticias18288gt (accessed February 2014)

SEAPRJ ndash Secretaria de Estado de Administraccedilatildeo Penitenciaacuteria do Rio de Janeiro(2006) Relatoacuterio Teacutecnico Anual do Programa de Controle de Tuberculose Rio deJaneiro Brasil lthttpwwwgooglecomsearchq=+Rio+de+Janeiro+28RJ29++prisC3 A3o2Ca+taxa+de+incidencia+de+TBampie=utf-8ampoe=utf-8ampaq=tamprls=orgmozillapt-BRofficialampclient=firefox-aampchannel=sbgt (accessed February2014)

SEAPRJ ndash Secretaria de Estado de Administraccedilatildeo Penitenciaacuteria do Rio de Janeiro2004 Relatorio Tecnico Anual do Programa de Controle da Tuberculose Rio deJaneiro Brasil

van Soolingen D Hermans PW de Haas PE Soll DR van Embden JD 1991Occurrence and stability of insertion sequences in Mycobacterium tuberculosiscomplex strains evaluation of an insertion sequence-dependent DNApolymorphism as a tool in the epidemiology of tuberculosis J Clin Microbiol29 2578ndash2586

Please cite this article in press as Huber FD et al Insights into the populationsoutheastern Brazilian prison unit Infect Genet Evol (2014) httpdxdoiorg

Vinhas SA Moiseacutes P Hebert SM Paola PLA Fabiacuteola KR Renata LP DietzeR Gomes HM Suffys PN Golub JE Riley LW Maciel ELN 2013Mycobacterium tuberculosis DNA fingerprint clusters and its relationship withRDRio genotype in Brazil Tuberculosis 93 207ndash221

Von Groll A Martin A Felix C Prata PF Honscha G Portaels F Vandame P daSilva PE Palomino JC 2010 Fitness study of the RDRio lineage and LatinAmerican-Mediterranean family of Mycobacterium tuberculosis in the city of RioGrande Brazil FEMS Immunol Med Microbiol 58 119ndash127

Weisenberg SA Gibson AL Huard RC Kurepina N Bang H Lazzarini LCChiu Y Li J Ahuja S Driscoll J Kreiswirth BN Ho JL 2012 Distinct clinicaland epidemiological features of tuberculosis in New York City caused by theRD(Rio) Mycobacterium tuberculosis sublineage Infect Genet Evol 12 664ndash670

Zhang J Abadia E Refregier G Tafaj S Boschiroli ML Guillard B AndremontA Ruimy R Sola C 2010 Mycobacterium tuberculosis complex CRISPRgenotyping improving efficiency throughput and discriminative power ofspoligotyping with new spacers and a microbead-based hybridization assay JMed Microbiol 59 285ndash294

structure of Mycobacterium tuberculosis using spoligotyping and RDRio in a101016jmeegid201405031