Evaluation of PCR–RFLP analysis targeting hsp65 and rpoB genes for the typing of mycobacterial isolates in Malaysia

Abstract

In this study, PCR–RFLP analysis (PRA) targeting hsp65 and rpoB gene regions was evaluated for the identification of mycobacterial species isolated from Malaysian patients. Overall, the hsp65 PRA identified 92.2 % of 90 isolates compared to 85.6 % by the rpoB PRA. With 47 rapidly growing species, the hsp65 PRA identified fewer (89.4 %) species than the rpoB PRA (95.7 %), but with 23 slow-growing species the reverse was true (91.3 % identification by the hsp65 PRA but only 52.5 % by the rpoB PRA). There were 16 isolates with discordant PRA results, which were resolved by 16S rRNA and hsp65 gene sequence analysis. The findings in this study suggest that the hsp65 PRA is more useful than the rpoB PRA for the identification of Mycobacterium species, particularly with the slow-growing members of the genus. In addition, this study reports 5 and 12 novel restriction patterns for inclusion in the hsp65 and rpoB PRA algorithms, respectively.

↵†Present address: Faculty of Information Science and Technology, Multimedia University, Melaka, Malaysia.
↵‡Present address: Department of Preclinical Sciences, Faculty of Medicine and Health Sciences, University Tunku Abdul Rahman, Kuala Lumpur, Malaysia.
The GenBank/EMBL/DDBJ accession numbers for the 16S rRNA and hsp65 gene sequences of M14, M15, M19, M26, M38, M41, M42, M44, M49, M60, M64, M65, M74, M81, M82 and M88 reported in this study are GQ184152−GQ184167, respectively, and GQ184168–GQ184183, respectively.
A table of isolate identification data is available as supplementary material with the online version of this paper.

INTRODUCTION

There are at least 130 species in the genus Mycobacterium, of which more and more are being found to be associated with human diseases (Tortoli et al., 2009). The timely diagnosis of mycobacterial infections is essential for early specific treatment and prevention of dissemination due to nosocomial (Bolan et al., 1985; Astagneau et al., 2001), occupational (Shelton et al., 1999) or domestic exposure (Winthrop et al., 2002; Lumb et al., 2004). This is particularly important in view of the resistance of most non-tuberculous mycobacteria (NTM) to one or more first-line antituberculosis drugs (Brown-Elliott & Wallace, 2002).

Conventionally, mycobacterial species are identified by their growth characteristics (such as growth rate, colonial morphology, pigmentation and photoreactivity) and biochemical reactions. However, phenotypic tests are extremely time-consuming and laborious, and the interpretation of test results could be problematic owing to the presence of inter-assay and intra-species variations (Springer et al., 1996). With the increasing diversity of mycobacterial species becoming known, conventional methods alone are no longer sufficient for diagnostic or research purposes. Mycolic acids analysis (Butler et al., 1996) using HPLC and direct gene sequence analysis (Böddinghaus et al., 1990; Ringuet et al., 1999; Adékambi & Drancourt, 2004) have shown higher accuracy and discriminative power for mycobacterial identification. However, the application of these approaches is limited to reference laboratories owing to the requirement for sophisticated analytical instruments. A variety of probe-based assays (Reisner et al., 1994; Kox et al., 1997; Tortoli et al., 2003) have been introduced, but their use is usually constrained by their relatively high running costs and the ability to identify only a limited number of mycobacterial species. However, PCR–RFLP assays (PRA) (Telenti et al., 1993; Lee et al., 2000) have been reported to provide rapid, accurate and cost-effective identification of a wide range of mycobacteria.

There is little information on the prevalence and distribution of mycobacterial species other than Mycobacterium tuberculosis in clinical settings in Malaysia. In this study, two PRAs, targeting hsp65 (Telenti et al., 1993) and rpoB (Lee et al., 2000) genes were evaluated for the identification of local isolates. Conventional biochemical tests and DNA sequence analyses (of 16S rRNA and hsp65 genes) were also performed to assist in the identification of these mycobacteria.

METHODS

Mycobacterial isolates.

A total of 90 clinical isolates (mostly from respiratory secretions but also from skin and other tissues) were obtained from the Microbiology Diagnostic Laboratory, University Malaya Medical Centre (UMMC), Kuala Lumpur, Malaysia. These isolates were from consecutive unique patients for whom a specific request for mycobacterial culture was made by their attending physicians to exclude tuberculous infections.

In addition, seven reference strains (M. tuberculosis H37Rv ATCC 27294, Mycobacterium avium ATCC 25291, Mycobacterium intracellulare ATCC 13950, Mycobacterium kansasii ATCC 12478, Mycobacterium scrofulaceum ATCC 19981, Mycobacterium fortuitum ATCC 6841, Mycobacterium bovis ATCC 19210) and five atypical mycobacterial isolates (Mycobacterium gordonae, M. fortuitum, Mycobacterium chelonae, Mycobacterium marinum and Mycobacterium xenopi) kindly provided by Professor Peter Hawkey, Public Health Laboratory, Heartlands Hospital, Birmingham, UK, were also included in the evaluation. All isolates were maintained on Löwenstein–Jensen (LJ) media.

DNA extraction.

Colonies from LJ medium were suspended in 200 μl double-distilled water and boiled for 10 min. The supernatant collected after centrifugation at 8000 g for 10 min was used for PCR amplification.

PRA targeting hsp65 and rpoB genes.

The partial hsp65 gene (439 bp) was amplified using primers Tb11 (5′-ACCAACGATGGTGTGTCCAT-3′) and Tb12 (5′-CTTGTCGAACCGCATACCCT-3′) as described by Telenti et al. (1993). Amplification of the partial rpoB gene (360 bp) was performed using primers Rpo5′ (5′-TCAAGGAGAAGCGCTACGA-3′) and Rpo3′ (5′-GGATGTTGATCAGGGTCTGC-3′), as described by Lee et al. (2000). A 10 μl aliquot of amplified hsp65 products was digested with 5 units BstEII (New England Biolabs) and HaeIII (New England Biolabs) in separate reaction tubes. The reaction mixtures were incubated for 3 h at 60 °C for BstEII and at 37 °C for the HaeIII reaction. Similarly, amplified rpoB products were digested with 5 units MspI (New England Biolabs) and HaeIII (New England Biolabs) at 37 °C for 3 h. The digestion mixtures were analysed by Nusieve 3 : 1 agarose gel (3 % w/v) electrophoresis. DNA size markers, pBR322-MspI-digested DNA (New England Biolabs) and 100 bp DNA size markers (Fermentas), were used to enable estimation of the DNA fragment size.

For the hsp65 PRA, mycobacterial isolates were identified by comparing their restriction patterns with those available in the online database PRAsite (prasite/index.html); for the rpoB PRA, the restriction patterns described by Lee et al. (2000) were referred to. Isolates with discordant identification results by the PRAs were further studied by conventional biochemical tests and direct DNA sequence analysis of 16S rRNA and hsp65 genes.

Growth and biochemical tests.

The mycobacterial properties studied included growth rate and pigment production on LJ medium, ability to grow on MacConkey agar, sodium chloride tolerance, niacin accumulation, Tween 80 hydrolysis, tellurite reduction (Lutz, 1992), and the activity of nitrate reductase, urease, arylsulfatase and catalase (semiquantitative and heat-stable) enzymes.

Direct DNA sequence analysis.

The 564 bp long 5′-region of the 16S rRNA gene was amplified from the isolates using primers 16MycF (5′-CGTGCTTAACACATGCAAGTCG-3′) and 16MycR (5′-GTGAGATTTCACGAACA-ACGC-3′) (Devulder et al., 2005). Amplified hsp65 and 16S rRNA gene products were purified using a MinElute PCR purification kit (Qiagen) and sequenced using a BigDye terminator v3.1 cycle sequencing kit (Applied Biosystems) on an ABI-377 genetic analyzer (Applied Biosystems). The sequences were analysed by referring to public gene sequence repositories: the National Center for Biotechnology Information GenBank database (Altschul et al., 1994; Benson et al., 2007) and the Ribosomal Differentiation of Medical Micro-organisms (RIDOM) database (Harmsen et al., 2002).

RESULTS

Identification of mycobacteria by hsp65 PRA and rpoB PRA

All seven American Type Culture Collection reference strains were identified correctly by the two PRAs. With the five reference strains from the UK Public Health Laboratory, however, there were discrepancies between the two assays. The hsp65 PRA identified M. fortuitum as Mycobacterium peregrinum, a closely related species in the M. fortuitum complex. M. chelonae was identified as Mycobacterium abscessus in the hsp65 assay but as the M. chelonae/M. abscessus complex in the rpoB PRA, which does not distinguish between the two species (Lee et al., 2000). M. gordonae was assigned to different subtypes (type IX by the hsp65 PRA and type IV by the rpoB PRA). Neither M. marinum or M. xenopi were identified in the rpoB PRA.

Among the 90 clinical isolates examined, there were 20 strains of the M. tuberculosis complex (MTBC), 47 NTM rapid growers and 23 slow growers (Table 1⇓). All MTBC isolates were concordantly identified by the two assays. Among the rapid growers, the rpoB assay identified 28 M. fortuitum (21 type I and 7 type II) and 17 strains classified as M. chelonae/M. abscessus complex because the restriction patterns generated for the two species are indistinguishable. In contrast, the hsp65 assay identified only 24 M. fortuitum (type I) but was able to differentiate M abscessus (8 type I and 9 type II) from M. chelonae, and M. peregrinum (type III) from M. fortuitum. One isolate (M15) was unidentifiable in both assays and four isolates (one Mycobacterium austroafricanum and three M. fortuitum) were identified by the rpoB assay but not by the hsp65 assay.

Table 1.

hsp65 and rpoB PRA results for mycobacterial clinical isolates

Among the slow-growing NTM species, the rpoB assay performed dismally compared to the hsp65 assay, identifying only 12 of 23 isolates even with the use of two restriction enzymes. The 11 isolates not identifiable included M. marinum, Mycobacterium simiae, M. gordonae, Mycobacterium parascrofulaceum, M. avium and M. intracellulare. However, with the hsp65 PRA, there were only two unidentifiable patterns, both of which were also unidentifiable with the rpoB assay.

Table 2⇓ summarizes the results obtained for all 90 clinical isolates. Overall, the hsp65 PRA identified 83 (92.2 %) isolates, while the rpoB PRA identified only 78 (86.7 %). The two assays gave concordant results in 73 (81.1 %) isolates. Among the 70 NTM isolates, there was only 1 discordant result (a M. peregrinum/M. fortuitum difference that was not considered a significant discrepancy); however, there were 16 (22.9 %) unidentifiable PRA patterns, 9 from the rpoB assay, 4 from hsp65 and 3 from both assays.

Table 2.

Summary of mycobacterial species identification by hsp65 and rpoB PRAs

Resolving undeterminable results

The 16 NTM isolates that gave unidentifiable PRA patterns with either PRA were further analysed by taking into consideration their growth characteristics, the results of biochemical tests and DNA sequence information. Nucleotide sequences 389–535 bp in length were obtained from the direct sequence determination of the 16S rRNA gene and sequences 365–407 bp long from the hsp65 gene. All the DNA sequences obtained for these gene sequences of M14, M15, M19, M26, M38, M41, M42, M44, M49, M60, M64, M65, M74, M81, M82 and M88 have been deposited in the National Center for Biotechnology Information GenBank with accession numbers GQ184152 to GQ184183.

Based on the results of conventional tests and direct DNA sequence analysis, each of the 16 NTM isolates could be assigned to a Mycobacterium species (Supplementary Table S1 available with the online journal). Nine of these (M14, M19, M26, M41, M44, M60, M81, M82 and M88) were consistent with the species identified by the hsp65 PRA and three (M38, M42 and M65) were similar to the rpoB PRA identification. M49, which was identified as M. austroafricanum by the rpoB PRA, was redesignated M. gordonae based on the sequence analysis, although phenotypically, it was a rapid growing scotochromogen more like M. austroafricanum (Tsukamura et al., 1983). The remaining three isolates (M15, M64 and M74) exhibited novel restriction patterns that have not been reported before in the published algorithms of both PRAs. They were identified as Mycobacterium insubricum, Mycobacterium terrae and M. gordonae based on sequence analysis and, in the case of the latter two, on biochemical reactions as well.

Intraspecies variation

The intraspecies genetic variation was roughly assessed by the number of banding patterns noted for each species of NTM studied. As the number of isolates was small, only M. abscessus (17 strains), M. fortuitum (28 strains), M. avium (6 strains) and M. gordonae (7 strains) were analysed. The number of banding patterns observed for these isolates were 1, 2, 4 and 3, respectively, in the rpoB PRA, and 2, 3, 2 and 3, respectively, in the hsp65 PRA.

DISCUSSION

Current knowledge on the complexity and diversity within the genus Mycobacterium has come largely from the use of molecular techniques for the differentiation of species and subspecies. Phenotypic characteristics often fail to provide a definitive identification, as illustrated in this study where biochemical tests were inconclusive in 7 of 16 isolates eventually identified by DNA sequence analysis. Nevertheless, these tests can still be useful for the differentiation of certain clusters of mycobacteria that are genetically similar but phenotypically distinct (such as M. marinum–Mycobacterium ulcerans and M. kansasii– Mycobacterium gastri clusters).

In Malaysia, few laboratories have facilities for mycobacterial speciation. Reference and state hospital laboratories use liquid culture for isolation and commercial tests such as the AccuProbe system (Gen-Probe) for the identification of M. tuberculosis, M. kansasii and the M. avium–M. intracellulare complex. In teaching hospitals, commercial line probe assays such as GenoType Mycobacterium CM/AS (Hain Lifescience) and INNO-LiPA Mycobacteria (Innogenetics) are sometimes used to identify NTM species. None of these commercial tests have been adequately compared for test performance and value for money. This study has shown that the PRA [at a cost of £5.33 (US $ 8) per test compared with about £23.33 (US $ 35) for a line probe assay] could be an inexpensive alternative to the commercial tests in laboratories with smaller budgets. It requires only basic facilities for DNA amplification and gel electrophoresis, but is capable of identifying a wide spectrum of mycobacteria, especially when multiple targets and digestion enzymes are used. Many of the NTM species identified in this study are reported for what is believed to be the first time from Malaysian patients, for instance, M. simiae, Mycobacterium szulgai, Mycobacterium interjectum, M. terrae, M. parascrofulaceum and M. insubricum. The role they played in the clinical conditions they were associated with is not known, but their identification is expected to lead to an increased awareness and interest that will lead to more studies and a better understanding of their epidemiology and pathogenicity. A multi-country retrospective survey conducted by the International Union Against Tuberculosis and Lung Disease found M. avium complex, M. gordonae, M. xenopi, M. kansasii and M. fortuitum to be the five species most frequently isolated from patient samples, and concluded that NTM diseases appeared to be on the rise, with a changing pattern of infecting species over time and in different geographical areas (Martín-Casabona et al., 2004). The UMMC is a teaching hospital that also serves as a major referral hospital for patients from all over Malaysia. Each year, the laboratory cultures about 8000 specimens for mycobacteria and identifies M. tuberculosis and NTM in 7–8 % and 2–3 % of specimens, respectively. The 70 NTM isolates examined in this study would represent close to 35 % of all NTM strains isolated in the medical centre in 1 year. The paucity of reports on NTM infections in the country does not allow us to estimate how representative our isolates are of all isolates circulating in the country. Our data suggest that in Malaysia M. fortuitum is by far the most frequently encountered NTM in clinical samples, followed by M. abscessus, M. avium–M. intracellulare and M. gordonae. However, as in other parts of the world, a different pattern may emerge following the introduction and wider use of better diagnostic techniques for the identification of NTM species.

Telenti et al. (1993) and Lee et al. (2000) showed that hsp65 and rpoB PRAs provided sufficient information to identify most clinically important Mycobacterium species and subspecies. The hsp65 PRA has been evaluated by many other workers (Brunello et al., 2001; Häfner et al., 2004; Leão et al., 2005) and there are 135 restriction patterns described for 174 mycobacterial species and subspecies in the PRA site database (). There are fewer restriction patterns documented in the rpoB PRA algorithm. Lee et al. (2000) used 43 species and subspecies for the initial evaluation of this PRA and described only 36 restriction patterns. This smaller database could have contributed to the larger number of unidentifiable slow-growing NTM species obtained with the rpoB PRA in this study. In both assays, intraspecies genetic variation gives rise to more than one restriction pattern for a single mycobacterial species, and this, together with similarity in band sizes that are critical for species discrimination, has caused difficulty in the interpretation of restriction patterns for some mycobacteria.

There is no major disagreement in the results obtained from the sequence analysis of 16S rRNA and hsp65 genes for the 16 NTM isolates with different hsp65 and rpoB PRA results. Both assays contributed novel restriction patterns, 5 (7.1 %) by the hsp65 PRA (2 for M. gordonae and 1 each for M. fortuitum, M. terrae and M. insubricum) and 12 (17.1 %) by the rpoB PRA (3 for M. avium, 2 for M. parascrofulaceum, 1 each for M. gordonae, M. marinum, M. simiae, M. intracellulare, M. interjectum, M. terrae, and M. insubricum) (Supplementary Table S1 available with the online journal). These novel patterns illustrate the immensity of intraspecies variation that can be revealed with the testing of mycobacterial species from different geographical areas.

With an increasing recognition of mycobacterial species as human pathogens, mycobacterial diagnostic laboratories must be able to establish a minimum set of tests that will allow them to classify and correctly identify the most commonly found mycobacteria in clinical practice. Where facilities exist for molecular testing, a cost-effective protocol would be to apply the PRA for first line identification and reserve DNA sequencing for isolates not identifiable by PRA patterns. Using representative strains in the country, the construction of a national database (to include clinical information, PRA restriction patterns and other laboratory test findings) will facilitate species identification and eventually improve the epidemiological surveillance for these important micro-organisms.

Acknowledgments

The authors are grateful to Professor Peter Hawkey from the Public Health Laboratory, Heartlands Hospital, UK, for the generous gift of atypical mycobacterial species, and would like to thank the tuberculosis laboratory staff in the UMMC, Kuala Lumpur, for assistance with the isolation and initial identification of clinical isolates.

References

Adékambi, T. & Drancourt, M. (2004). Dissection of phylogenetic relationships among 19 rapidly growing Mycobacterium species by 16S rRNA, hsp65, sodA, recA and rpoB gene sequencing. Int J Syst Evol Microbiol 54, 2095–2105.
Altschul, S. F., Boguski, M. S., Gish, W. & Wootton, J. C. (1994). Issues in searching molecular sequence databases. Nat Genet 6, 119–129.
Astagneau, P., Desplaces, N., Vincent, V., Chicheportiche, V., Botherel, A., Maugat, S., Lebascle, K., Léonard, P., Desenclos, J. & other authors (2001). Mycobacterium xenopi spinal infections after discovertebral surgery: investigation and screening of a large outbreak. Lancet 358, 747–751.
Benson, D. A., Karsch-Mizrachi, I., Lipman, D. J., Ostell, J. & Wheeler, D. L. (2007). GenBank. Nucleic Acids Res 35, D21–D25.
Böddinghaus, B., Rogall, T., Flohr, T., Blöcker, H. & Böttger, E. C. (1990). Detection and identification of mycobacteria by amplification of rRNA. J Clin Microbiol 28, 1751–1759.
Bolan, G., Reingold, A. L., Carson, L. A., Silcox, V. A., Woodley, C. L., Hayes, P. S., Hightower, A. W., McFarland, L., Brown, J. W., III & other authors (1985). Infections with Mycobacterium chelonei in patients receiving dialysis and using processed hemodialyzers. J Infect Dis 152, 1013–1019.
Brown-Elliott, B. A. & Wallace, R. J., Jr (2002). Clinical and taxonomic status of pathogenic nonpigmented or late-pigmenting rapidly growing mycobacteria. Clin Microbiol Rev 15, 716–746.
Brunello, F., Ligozzi, M., Cristelli, E., Bonora, S., Tortoli, E. & Fontana, R. (2001). Identification of 54 mycobacterial species by PCR-restriction fragment length polymorphism analysis of the hsp65 gene. J Clin Microbiol 39, 2799–2806.
Butler, W. R., Floyd, M. M., Silcos, V., Cage, G., Desmond, E., Duffey, P. S., Guthertz, L. S., Gross, W. M., Jost, K. C., Jr & other authors (1996). Standardized Method for HPLC Identification of Mycobacteria. Atlanta, GA. : Centers for Disease Control and Prevention.
Devulder, G., de Montclos, M. P. & Flandrois, J. P. (2005). A multigene approach to phylogenetic analysis using the genus Mycobacterium as a model. Int J Syst Evol Microbiol 55, 293–302.
Häfner, B., Haag, H., Geiss, H.-K. & Nolte, O. (2004). Different molecular methods for the identification of rarely isolated non-tuberculous mycobacteria and description of new hsp65 restriction fragment length polymorphism patterns. Mol Cell Probes 18, 59–65.
Harmsen, D., Rothganger, J., Frosch, M. & Albert, J. (2002). RIDOM: Ribosomal Differentiation of Medical Microorganisms Database. Nucleic Acids Res 30, 416–417.
Kox, L. F. F., Jansen, H. M., Kuijper, S. & Kolk, A. H. J. (1997). Multiplex PCR assay for immediate identification of the infecting species in patients with mycobacterial disease. J Clin Microbiol 35, 1492–1498.
Leão, S. C., Bernardelli, A., Cataldi, A., Zumarraga, M., Robledo, J., Realpe, T., Mejía, G. I., da Silva Telles, M. A., Chimara, E. & other authors (2005). Multicenter evaluation of mycobacteria identification by PCR restriction enzyme analysis in laboratories from Latin America and the Carribbean. J Microbiol Methods 61, 193–199.
Lee, H., Park, H.-J., Cho, S.-N., Bai, G.-H. & Kim, S.-J. (2000). Species identification of mycobacteria by PCR-restriction fragment length polymorphism of the rpoB gene. J Clin Microbiol 38, 2966–2971.
Lumb, R., Stapledon, R., Scroop, A., Bond, P., Cunliffe, D., Goodwin, A., Doyle, R. & Bastian, I. (2004). Investigation of spa pools associated with lung disorders caused by Mycobacterium avium complex in immunocompetent adults. Appl Environ Microbiol 70, 4906–4910.
Lutz, B. (1992). Identification tests for mycobacteria. In Clinical Microbiology Procedures Handbook, pp. 3.12.1–3.12.17. Edited by Isenberg, H. D.. Washington, DC. : American Society for Microbiology.
Martín-Casabona, N., Bahrmand, A. R., Bennedsen, J., Østergaard Thomsen, V., Curcio, M., Fauville-Dufaux, M., Feldman, K., Havelkova, M., Katila, M.-L. & other authors (2004). Non-tuberculous mycobacteria: patterns of isolation. A multi-country retrospective survey. Int J Tuberc Lung Dis 8, 1186–1193.
Reisner, B. S., Gatson, A. M. & Woods, G. L. (1994). Use of Gen-Probe AccuProbes to identify Mycobacterium avium complex, Mycobacterium tuberculosis complex, Mycobacterium kansasii, and Mycobacterium gordonae directly from BACTEC TB broth cultures. J Clin Microbiol 32, 2995–2998.
Ringuet, H., Akoua-Koffi, C., Honore, S., Varnerot, A., Vincent, V., Berche, P., Gaillard, J. L. & Pierre-Audigier, C. (1999). hsp65 sequencing for identification of rapidly growing mycobacteria. J Clin Microbiol 37, 852–857.
Shelton, B. G., Flanders, W. D. & Morris, G. K. (1999). Mycobacterium sp. as a possible cause of hypersensitivity pneumonitis in machine workers. Emerg Infect Dis 5, 270–273.
Springer, B., Stockman, L., Teschner, K., Roberts, G. D. & Bottger, E. C. (1996). Two-laboratory collaborative study on identification of mycobacteria: molecular versus phenotypic methods. J Clin Microbiol 34, 296–303.
Telenti, A., Marchesi, F., Balz, M., Bally, F., Bottger, E. C. & Bodmer, T. (1993). Rapid identification of mycobacteria to the species level by polymerase chain reaction and restriction enzyme analysis. J Clin Microbiol 31, 175–178.
Tortoli, E., Mariottini, A. & Mazzarelli, G. (2003). Evaluation of INNO-LiPA MYCOBACTERIA v2: improved reverse hybridization multiple DNA probe assay for mycobacteria identification. J Clin Microbiol 41, 4418–4420.
Tortoli, E., Baruzzo, S., Heijdra, Y., Klenk, H. P., Lauria, S., Mariottini, A. & van Ingen, J. (2009). Mycobacterium insubricum sp. nov. Int J Syst Evol Microbiol 59, 1518–1523.
Tsukamura, M., Van Der Meulen, H. J. & Grabow, W. O. K. (1983). Numerical taxonomy of rapidly growing, scotochromogenic mycobacteria of the Mycobacterium parafortuitum complex: Mycobacterium austroafricanum sp. nov. and Mycobacterium diernhoferi sp. nov., nom. rev. Int J Syst Bacteriol 33, 460–469.
Winthrop, K. L., Abrams, M., Yakrus, M., Schwartz, I., Ely, J., Gillies, D. & Vugia, D. J. (2002). An outbreak of mycobacterial furunculosis associated with footbaths at a nail salon. N Engl J Med 346, 1366–1371.