Species Identification of Neglected Nontuberculous Mycobacteria in a Developing Country

Transcription

1 Jpn. J. Infect. Dis., 64, , 2011 Original Article Species Identification of Neglected Nontuberculous Mycobacteria in a Developing Country Hasan Shojaei*, Parvin Heidarieh 1, Abodolrazagh Hashemi 2, Mohamad Mehdi Feizabadi 1, and Abass Daei Naser Isfahan University of Medical Sciences, Isfahan; 1 Tehran University of Medical Sciences, Tehran; and 2 Ahwaz Jondi Shapour University of Medical Sciences, Ahwaz, Iran (Received January 17, Accepted April 5, 2011) SUMMARY: In developing countries where tuberculosis is still a health challenge, the prevalence of nontuberculous mycobacterial diseases is expected to rise as medical conditions that compromise the immune system become more widespread. In the current study, we aimed to determine the presence and diversity of nontuberculous mycobacteria (NTM) causing infections in Iranian patients. Sixty-seven clinical NTM isolates were identified using conventional and molecular methods, including PCR-restriction fragment length polymorphism analysis (PRA) and 16S rrna sequencing. Out of 67 patients with confirmed mycobacterial infection, 29 had an associated immunosuppressive syndrome, including 9 who were HIV-infected. Forty-nine NTM isolates were identified using PRA, and the remaining 18 isolates were identified using 16S rrna sequencing. We obtained the following results: Mycobacterium fortuitum, 30 isolates; M. kansasii, 12isolates;M. gordonae, 8isolates;M. porcinum, 3isolates;M. conceptionense, 3isolates;M. phlei, 2 isolates; and M. austroafricanum, M. avium, M. elephantis, M. intracellulare,m.lentiflavum,m.monacense,m.parascrofulaceum,andm. thermoresistibile,1isolate each; and 1 potentially novel mycobacterial species. With regard to the complexity of identification, it is recommended that laboratory diagnosis of NTM diseases be centralized by strengthening or setting up quality national and regional infrastructure. INTRODUCTION Nontuberculous mycobacteria (NTM), also known as atypical mycobacteria and mycobacteria other than tuberculosis, are environmental organisms that are normally found in soil and water. They have been recognized since late in the 19th century, when avian tuberculosis was first described in 1868 (1). However, they were not recognized as a cause of human disease until the 1950s (2). At that time, NTMs were rare, and were found almost exclusively in patients with underlying diseases. However, NTM disease patterns and epidemiology have changed since the 1980s and gradually emerged in previously unrecognized populations (3,4). Given that the clinical presentation of NTM is often difficult to differentiate from that of Mycobacterium tuberculosis complex (MTBC), it is important to accurately identify NTM so that the correct epidemiological controls and specific treatments can be implemented (5). To date, the genus Mycobacterium comprises over 150 species. Though M. tuberculosis is one of the most common, several other species are being increasingly recognized as significant human pathogens ( In countries like Iran, where tuberculosis is still a major health challenge that causes the suffering and *Corresponding author: Mailing address: Isfahan University of Medical Sciences, Isfahan, Iran. Tel: , Fax: , hasanshojaei@ msn.com deathofmanypeopleeachyear,theprevalenceofdiseases caused by NTM is expected to rise as both the AIDS epidemic and medical conditions that compromise the immune system become more widespread. Unavoidable health consequences of ongoing and unending natural and human disasters, including social unrest, ethnic conflict, sectarian bloodshed, war, earthquake, and floods in the turbulent region of the Middle East, seem to be considerably worsening the situation. In the current study, we aimed to assess the extent of NTM infection and the diversity of microorganisms found in an Iranian population using a combinatorial approach comprising conventional and molecular techniques. MATERIALS AND METHODS Isolates: From 2003 to 2009, a total of 67 clinical NTM isolates were collected by our molecular microbiology laboratory at the Infectious Diseases and Tropical Medicine Research Center (IDRC), Isfahan, Iran. The sources and clinical details of the isolates are summarized (see Table 1). They were isolated in or referred to our laboratory. The majority of cases (i.e., 53 strains, including M01, M02, M04 to M25, M27 to M34, M40 to M51, M61, M66, M70, M71, M79, M120, M137, M138, and M143) were received from peripheral laboratories in the region or from other laboratories across the country for the identification and characterization of the bacteria. These isolates had been grown in pure culture on L äowenstein-jensen (LJ) slants, and were verified to be 265

2 acid-fast bacteria (AFB) by Ziehl-Neelsen (ZN) staining. All isolates displayed growth characteristics and colony morphology consistent with those of Mycobacterium spp. The remaining isolates (i.e., M201, M202, M206, M210, M213, M214, M216, M219, M220 to M223, M226, and M228) were recovered from clinical specimens of the patients referred to our laboratory by their practicing physicians. Species characterization of these isolates was then analyzed using a battery of conventional and molecular tests as described below. Conventional identification: On specimens collected from nonsterile body sites, digestion and decontamination procedures were performed by Petroff's method in order to minimize the interference of contaminating bacteria in the recovery of mycobacteria (6). However, sterile bodily fluids such as blood, soft tissues, and lymph node biopsies were processed without decontamination. Examination and reporting of smears stained by auramine O dye and ZN stain were performed within 24 h of specimen receipt. LJ slants were loosely capped and incubated at 379C in an atmosphere of 5z CO 2. All culture tubes were examined daily for 30 days and twice weekly for 4 weeks thereafter. Slants containing pure cultures of AFB were subjected to a series of conventional phenotypic tests and standard biochemical assays, including assessments of growth rate, niacin accumulation, pigment production, growth in LJ medium (at 379C, 459C, and 529C), growth on MacConkey agar without crystal violet, semiquantitative and heat-stable (689C) catalase production, nitrate reduction, arylsulfatase activity, tolerance to 5z NaCl, iron uptake, urease activity, and tellurite reduction, according to standard procedures (6). Molecular identification: Strains that were initially identified as mycobacteria and were classified into Runyon groups (7) on the basis of phenotypic features (see Table 1), were subjected to more definitive identification by molecular approaches using the standard procedures described below. DNA extraction: Chromosomal DNA was extracted using the method described by Pitcher et al., with slight modifications to improve the susceptibility of cells to the standard digestion (8). Briefly, after thermal inactivation, pretreatment of biomass with lipase (Type VII; final concentration, 2 mg/ml [Sigma, Poole, UK]) and additional treatment with proteinase K (100 pg/ml) and sodium dodecyl sulfate (final concentration, 0.5z [wt/vol]) were performed. DNA was purified by phenol/chloroform/isoamyl alcohol and precipitated with isopropanol. The precipitate was washed in 70z ethanol, dehydrated, dissolved in 100 ml of Milli-Q water, and stored at -209C until use. Molecular assignment of clinical isolates to NTM group: A 228-base-pair (bp) segment corresponding to a sequence from the 65-kDa heat shock protein gene (hsp65) was used as a target and was amplified by polymerase chain reaction (PCR) using specific primers according to the procedures recommended by Khan and Yadav (9). PCR-restriction fragment length polymorphism analysis (PRA): PRA of hsp65 was performed as described by Kim et al. (10). Briefly, a 644-bp fragment of hsp65 was amplified and digested by three restriction enzymes, namely, AvaII, HphI, and HpaII. The enzyme-digested products were electrophoresed on a 3z agarose gel, and the pattern was compared to the restriction map algorithm constructed by Kim et al. (10) as well as our inhouse database. 16S rrna sequencing: The almost full length 16S rrna gene sequence was determined as described previously for representative isolates of clusters with the unknown PRA patterns of rare or newly described species of mycobacteria (11). Obtained sequences were aligned manually, compared with relevant GenBank sequences for the specific species of mycobacteria, and analyzed using jphydit and Mega4 software (12,13). The GenBank accession numbers for 16S rrna sequencing of representative Iranian mycobacteria clinical isolates determined in this study are as follows: M. avium (GU142929), M. austroafricanum (GU121552), M. conceptionense (GU142926), M. conceptionense-like (GU142922), M. elephantis (GU142921), M. fortuitum (HQ190907), M. fortuitum-like (HQ009482), M. gordonae (GU142923, GU142925, GU142930), M. intracellulare (GU142932), M. kansasii (GU142919), M. lentiflavum (GU142924), M. monacense (GU142931), M. parascrofulaceum (GU121551), M. phlei (GU142920), M. porcinum (GU142917), and M. thermoresistibile (GU142928). RESULTS Of the 67 patients with confirmed mycobacterial infections, 38 (56z) were female, 17 (25z) were between the ages of 50 and 60 years, 14 (21z) were between the ages of 30 and 40 years, 12 (18z) were between the ages of 40 and 50 years, 12 (18z) wereovertheageof70 years, 10 (15z) were between the ages of 60 and 70 years, and 2 (3z) were between the ages of 20 and 30 years. Twenty-nine patients were found to have a history of various immunosuppressive syndromes, including cystic fibrosis (5 cases), previously treated tuberculosis (1 case), malignancy (3 cases), chronic obstructive pulmonary disease (6 cases), diabetes mellitus (1 case), bronchitis (3 cases), chronic pelvic pain (1 case), or HIV infection (9 cases), while the remainder of the patients had no apparent history of such disorders (Table 1). Conventional identification: On the basis of growth characteristics and pigmentation (Table 1), the isolates studied were categorized into the four Runyon groups (7): 43 clinical isolates were categorized as rapidly growing (Runyon IV), and 24 isolates were categorized slowly growing mycobacteria, with 10 classified as scotochromogenic (Runyon II), 12 classified as photochromogenic (Runyon I), and 2 classified as nonchromogenic (Runyon III). On the basis of the biochemical properties, identification was made for all 67 isolates with variable levels of confidence. The M. fortuitum-like group was the most frequently encountered (36 isolates), followed by M. kansasii-like strains (12 isolates), and M. gordonae-like strains (8 isolates). The remaining strains were mostly rare mycobacteria found in only 1 to 3 isolates. Molecular identification: Genus-specific PCR amplification characteristically yielded a 228-bp fragment of the hsp65, confirming that all strains belonged to the genus Mycobacterium. Using PRA, we were able to definitively identify

3 Table 1. Clinical details and identification information of Iranian isolates of NTM No. Designation Origin 1) Year of isolation Immune defect 1) Runyon group Identification by conventional test PRA pattern (fragment size) by Species assignment by AvaII HphI HpaII PRA 16S rrna sequencing M01, M02, M04, M06 M10, M12, M13, M15 M25, M27 M34, M41 M42 M51, M216, M219 BAL (13), sputum (10), soft tissue (4), abscess (2), urine (1) BAL (6), lymphadenitis (1), sputum (5) HIV (3), COPD (4), cancer (3), CF (2) HIV (1), COPD (1), bronchitis (3), diabete mellitus (1), CF (1) IV M. fortuitum complex 501, , 207, , 161, 117 M. fortuitum M. fortuitum 2) I M. kansasii 456, , , 144, 126 M. kansasii type I M. kansasii 1 M214 BAL 2007 HIV III M. avium complex 240, 140, , 140, Unknown 3) M. avium 1 M220 Sputum 2006 HIV III M. avium complex 213, 126, 119, , M. intracellulare M. intracellulare 1 M206 BAL 2004 Healthy IV Mycobacterium sp. 525, , 141, , 161, 129 M. phlei M. phlei 1 M226 BAL 2004 Healthy IV Mycobacterium sp. 470, , 141, Unknown M. phlei 1 M210 BAL 2003 Healthy IV Mycobacterium sp. 520, , 206, , 103 Unknown M. austroafricanum 1 M201 Sputum 2003 Treated tuberculosis IV Mycobacterium sp , 144, , 129 M. thermoresistibile M. thermoresistibile 3 M71, M 213, M228 BAL (1), gastric washing (1), sputum (1) CF (1) IV M. fortuitum complex 501, , 141, , 161, 117 M. porcinum M. porcinum 2 M137,M70 Urine(1),blood(1) 2009 HIV,COPD IV M. fortuitum complex 470, , , 161, 117 Unknown M. conceptionense 1 M138 Urine (1) 2003 HIV II Mycobacterium sp. 470, , 170, , 100 Unknown M. gordonae 2 M222, M223 Sputum, BAL 2003 Healthy II Mycobacterium sp. 470, , 170, , 100 Unknown M. gordonae 5 M14, M40, M61, M120, M143 BAL (3), urine (1), Lymph node (1) HIV (1), CF (1), Healthy (3) II Mycobacterium sp. 470, , , 117 Unknown M. gordonae 1 M202 BAL 2003 Healthy IV M. flavescens 270, , 206, , 210, 97 Unknown M. elephantis 1 M79 Vaginal discharge and urine 2008 Chronic pelvic pain II M. scrofulaceum 240, 220, , 143, Unknown M. parascrofulaceum 1 M11 Sputum and BAL 2008 Healthy IV Mycobacterium sp. 456, , , 160, 117 Unknown M. monacense 1 M221 BAL 2003 Healthy II M. simiae 456, , 112, M. lentiflavum M. lentiflavum 1 M05 BAL 2008 Healthy IV Mycobacterium sp. 520, , 140, , 160, 126 Unknown M. fortuitum-like 1 M66 BAL 2007 Healthy IV M. fortuitum complex 240, 220, , 140, Unknown M. conceptionense 1) : The digits in the parentheses indicate the number of cases. 2) : Representative isolates from each cluster was subjected to 16S rrna gene sequencing. 3) : Unknown patterns were designated based on algorithm described by Kim et al. (10). NTM, nontuberculous mycobacterium; PRA, PCR-restriction fragment length polymorphism analysis; BAL, bronchoalveolar lavage; COPD, chronic obstructive pulmonary disease; CF, cystic fibrosis. 267

4 Fig. 1. The differentiation of the Iranian clinical isolates of nontuberculous mycobacteria by hsp65 PRA. The amplified hsp65 DNAs (644 bp) of the isolates were digested with AvaII (A), HpaII (B), and HphI (C), and electrophoresed on a 3z agarose gel. Restriction fragments shorter than 97 bp were not considered to reduce confusion with primer-dimer bands. M, marker DNA (50 bp ladder); 1, M. fortuitum ATCC 6841; 2 and 8, M. fortuitum (M02 and M15); 3, M. tuberculosis H37Rv ATCC 27294; 4, Mycobacterium sp.(m05);5,m. monacense (M11); 6, M. kansasii (M42); 7, M. gordonae (M14). out of 67 Iranian isolates, since they produced a PRA pattern identical to those of the reference type established species according to algorithm described by Kim et al. (10). These isolates comprised 30 M. fortuitum isolates, 12 M. kansasii isolates, 3 M. porcinum isolates, and 4 single isolates of M. intracellulare, M. lentiflavum, M. phlei, and M. thermoresistibile (Table 1). However, 18 of the isolates, namely, M05, M11, M14, M40, M61, M66, M70, M79, M120, M137, M138, M143, M202, M210, M214, M222, M223, and M226, presented a restriction pattern that was not comparable to those of the reference strains (10). These isolates were further analyzed by determining the nucleotide sequence of the 16S rrna gene (Table 1 and Fig. 1). In addition, several of the Iranian isolates with known PRA identities, including M. kansasii (1 isolate), M. fortuitum (1 isolate), M. porcinum (3 isolates), M. intracellulare (1 isolate), M. phlei (1 isolate), M. lentiflavum (1 isolate), and M. thermoresistibile (1 isolate), were reevaluated by 16S rrna sequencing to investigate the possibility of intraspecies variation when compared to the reference type species of closely related mycobacteria. All the test strains analyzed by 16S rrna sequencing showed mycobacterial nucleotide signatures; specifically, they shared sequences at positions (A T), (G T), 307 (C), 328 (T), (A T), 631 (G), (G C), (T A), (A T), 843 (C), and (A T) (14,15). The 16S rrna sequences of all the rapidly growing mycobacteria test strains contained a characteristic short helix at the position, while 16S rrna sequences of all the slowly growing test strains had an extended helix at the position, characteristic of slowly growing mycobacteria. However, among the slowly growing mycobacteria, two strains, M79 and M221, presented with sequences that were somewhat deviant. These isolates were characterized by a short helix 18 (12 nucleotides shorter than that in M. tuberculosis) in hypervariable region B, placing them in an intermediate position between rapid and slow growers in the Mycobacterium phylogenetic tree (Figs. 2 and 3). The nucleotide signature sequences hypervariable A (positions ) and hypervariable B (positions ) for the test strains, together with those of closely related rapidly and slowly growing mycobacteria, are shown in Fig. 2. Analysis of the 16S rrna sequences of 18 isolates showing unknown PRA patterns revealed that six isolates, M11, M66, M70, M79, M137, and M202, belong to the type species of a recently described mycobacterial species that is not included in the algorithm described by Kim et al. (10) (Table 1). The remaining isolates that showed unique PRA patterns (i.e., M14, M40, M61, M120, M138, M143, M210, M214, M222, M223, and M226) were found to be members of previously established species since their sequence similarity values fell within the cut-off range of relevant established species (Table 1). M. gordonae strains produced three different patterns: ``pattern i'' for strains M222 and M223, ``pattern ii'' for strain M138, and ``pattern iii'' for the remaining strains M14, M40, M61, M120, and M143. The nucleotide sequence of isolate M05 showed substantial differences from the corresponding sequence of its nearest neighbor, M. fortuitum, within the hypervariable regions A and B (Fig. 2). Isolate M05 formed a distinct line of descent in the neighbor-joining based phylogenetic tree (Fig. 3). DISCUSSION The clinical presentation of NTM is often hard to differentiate from that of MTBC; therefore, it is important to accurately identify NTM, since the management and treatment of infected patients, as well as the epidemiological control methods implemented, must reflect the specific mycobacterial species encountered (16). In low-income countries like Iran, where the microbiological diagnosis of tuberculosis relies mostly on microscopy and where access to culture and drug susceptibility testing is virtually non-existent, physicians may neglect to consider the possibility that infections found in patients with smear-positive specimens may be due to NTM (16,17). Isolation and identification of NTM species from clinical specimens can often require specific preparatory 268

5 Fig. 2. Alignment of selected stretches of 16S rrna sequencing of Iranian NTM isolates with those of type strains of closely related mycobacteria. 16S rrna positions are according to Escherichia coli numbering system. `` '' indicates that the base pair was identical to that of type strain of M. tuberculosis. procedures, such as decontamination steps or selective incubation methods (e.g., the use of a CO 2 -enriched atmosphere or a specific temperature for optimal growth). Since most clinical laboratories do not use these techniques, infections caused by these organisms are likely to be underdiagnosed. On the other hand, extensive identification schemes comprising a minimum of 10 phenotypic tests have been used to identify the entire spectrum of mycobacteria, wich generally inconclusive results (17). These tests are difficult to interpret and are limited by subjectivity and low specificity (18). Consequently, attempts to identify mycobacteria by conventional phenotypic tests alone may result in erroneous or incomplete identification if a sufficient number of discriminative tests are not available. However, in response to the need for a more rapid and accurate identification of clinically relevant NTM, appropriate use of molecular testing, in conjunction with the key phenotypic tests, has shown significant promise (16,18). The importance of molecular data in the accurate identification of NTM cannot be stressed enough, since in our study a simple PRA provided accurate identification for 73z of isolates (49 out of 67). The identification of the 17 remaining anonymous mycobacterial isolates was not accomplished by preliminary PRA due to unique or non-identical PRA patterns. Instead, conclusive identification of each of these remaining isolates was achieved by 16S rrna gene sequencing, demonstrating the significance of this sequencing approach in the diagnosis of NTM infections. The substantial differences between the nucleotide signature sequences of the isolate M05 and that of its nearest neighbor, M. fortuitum, aswellastheformation of a recognizable phyletic subline, suggests that this organism exhibits features of a potentially novel species. Further studies, including hsp65, rpob, and 16S-23S internal transcribed spacer (ITS) gene sequencing, are being conducted in order to reach an unambiguous identification for this isolate. Among the 67 NTM isolates, M. fortuitum (n = 30, 45z) was the most prevalent species, followed by M. kansasii (n = 12, 18z), and M. gordonae (n = 8, 12z). These findings are in accordance with those of a multi-country retrospective survey conducted in 1996 by the Working Group of the Bacteriology and Immunology Section of the International Union Against Tuberculosis and Lung Disease, which contacted 50 laboratories in many different countries, including Iran (19). According to this report, which included evaluation of 63 Iranian NTM isolates, M. fortuitum wasfoundtobe predominant in Iran and Turkey. In fact, with the exception of these two countries and countries like Belgium and the Czech Republic, the M. avium complex (MAC) has been the species most frequently isolated from clinical specimens. The authors have suggested that the increase in identification of MAC is most likely due to the rising incidence of HIV infection (19). This might explain the low frequency of MAC in our study, 269

6 Fig. 3. The phylogenetic tree of almost complete 16S rrna sequences of representative type strains of genus Mycobacterium and 22 Iranian clinical NTM isolates. The numbers at the nodes indicate the level of bootstrap support based on 1,000 resampled data sets; only values greater than 40z are given. especially since, according to UNAIDS and the World Health Organization's Report on the global AIDS epidemic, the prevalence of HIV infection in Iran has remained rather low (0.2z of the population aged years) during the past 4 years (20). Unfotunately, our data does not permit us to estimate the true prevalence of unknown mycobacteria species in the laboratory isolates because the majority of the strains examined in our study were collected from other laboratories, and therefore, a selection bias might exist. However, the high number of inconclsively identified NTM strains among a collection of isolates from 270

7 patients with initial presumptive tuberculosis diagnosis is disturbing. The clinical relevance of the strains that were isolated from patients with pulmonary infection (54 out of 67 cases, 81z) was demonstrated on the basis of the fact that they were all obtained from pure cultures from at least two independent clinical samples with a preliminary positive smear of AFB. The etiologic role of the remaining extrapulmonary or rare mycobacterial isolates (7 cases, namely, M05, M11, M66, M70, M79, M137, and M202) might be inferred from the fact that AFB were microscopically observed in all clinical specimens. Furthermore, all isolates were recovered from the pure culture. In conclusion, the high number of NTM found in a collection of unidentified isolates and the diversity of species involved in the various clinical manifestations of healthy and immunocompromised patients merit special attention by health authorities, doctors, and microbiologists in developing countries. In health centers that deal with tuberculosis, action should be taken to increase awareness of appropriate diagnostic criteria and management guidelines for NTM diseases. Moreover, expanding the number and quality of national and regional reference laboratories may facilitate more accurate laboratory diagnosis of NTM diseases. Acknowledgments The authors are grateful to the Office of Vicechancellor for Research Isfahan University of Medical Sciences for financial support of the current study. Conflict of interest None to declare. REFERENCES 1. Wallace, R. (2010): History of MAC. Online at gdisease.org. 2. Crow, H.E., King, C.T., Smith, C.E., et al. (1957): A limited clinical, pathologic, and epidemiologic study of patients with pulmonary lesions associated with atypical acid-fast bacilli in the sputum. Am. Rev. Tuberc., 75, Falkinham, J.O., 3rd. (1996): Epidemiology of infection by nontuberculous mycobacteria. Clin. Microbiol. Rev., 9, Wolinsky, E. (1979): Nontuberculous mycobacteria and associated diseases. Am. Rev. Respir. Dis., 119, Griffith, D.E., Aksamit, T., Brown-Elliott, B.A., et al. (2007): An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases. Am. J. Respir. Crit. Care Med., 175, Kent, P.T. and Kubica, G.P. (1985): Public Health Mycobacteriology: a Guide for the Level III Laboratory. Centers for Disease Control, U.S. Department of Health and Human Services, Atlanta, Ga. 7. Timpe, A. and Runyon, E.H. (1954): The relationship of atypical acid-fast bacteria to human disease; a preliminary report. J. Lab. Clin. Med., 44, Pitcher, D.G., Saunders, N.A. and Owen, R.J. (1989): Rapid extraction of bacterial genomic DNA with guanidium thiocyanate. Lett. Appl. Microbiol., 8, Khan, I.U. and Yadav, J.S. (2004): Development of a single-tube, cell lysis-based, genus-specific PCR method for rapid identification of mycobacteria: optimization of cell lysis, PCR primers and conditions, and restriction pattern analysis. J. Clin. Microbiol., 42, Kim, H., Kim, S.H., Shim, T.S., et al. (2005): PCR restriction fragment length polymorphism analysis (PRA)-algorithm targeting 644 bp heat shock protein 65 (hsp65) gene for differentiation of Mycobacterium spp. J. Microbiol. Methods, 62, Shojaei, H., Magee, J.G., Freeman, R., et al. (2000): Mycobacterium elephantis sp. nov., a rapidly growing non-chromogenic Mycobacterium isolated from an elephant. Int. J. Syst. Evol. Microbiol., 50, Jeon, Y.S., Chung, H., Park, S., et al. (2005): jphydit: a JAVA-based integrated environment for molecular phylogeny of ribosomal RNA sequences. Bioinformatics., 21, Tamura, K., Dudley, J., Nei, M., et al. (2007): MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol. Biol. Evol., 24, Brosius, J., Palmer, M.L., Kennedy, P.J., et al. (1978): Complete nucleotide sequence of a 16S ribosomal RNA gene from Escherichia coli. Proc. Natl. Acad. Sci. USA, 75, Stackberandt, E., Rainey, F.A. and Ward-Rainey, N.L. (1997): Proposal for a new hierarchic classification system, Actinobacteria classis nov. Int. J. Syst. Evol. Microbiol., 47, Tortoli, E., Rogasi, P.G., Fantoni, E., et al. (2010): Infection due to a novel mycobacterium, mimicking multidrug-resistant Mycobacterium tuberculosis. Clin. Microbiol. Infect., 16, Springer, B., Stockman, L., Teschner, K., et al. (1996): Twolaboratory collaborative study on identification of mycobacteria: molecular versus phenotypic methods. J. Clin. Microbiol., 34, Tortoli, E., Bartoloni, A., Bottger, E.C., et al. (2001): Burden of unidentifiable mycobacteria in a reference laboratory. J. Clin. Microbiol., 39, Martáƒn-Casabona, N., Bahrmand, A.R., Bennedsen, J., et al. (2004): Non-tuberculous mycobacteria: patterns of isolation. A multi-country retrospective survey. Int. J. Tuberc. Lung Dis., 8, UNAIDS/WHO Epidemiological Fact Sheets on HIV and AIDS, 2008 Update. Online at Accessed 23 February