Am. J. Trop. Med. Hyg., 83(5), 2010, pp. 1059–1065 doi:10.4269/ajtmh.2010.10-0263 Copyright © 2010 by The American Society of Tropical Medicine and Hygiene A Genotypic Approach for Detection, Identification, and Characterization of Drug Resistance in Mycobacterium ulcerans in Clinical Samples and Isolates from Ghana Marcus Beissner,* Nana-Yaa Awua-Boateng, William Thompson, Willemien A. Nienhuis, Erasmus Klutse, Pius Agbenorku, Joerg Nitschke, Karl-Heinz Herbinger, Vera Siegmund, Erna Fleischmann, Ohene Adjei, Bernhard Fleischer, Tjip S. van der Werf, Thomas Loscher, and Gisela Bretzel Department of Infectious Diseases and Tropical Medicine (DITM), University Hospital, Ludwig-Maximilians University of Munich, Munich, Germany; Kumasi Centre for Collaborative Research in Tropical Medicine (KCCR), Kwame Nkrumah University of Science and Technology (KNUST), Kumasi, Ghana; Agogo Presbyterian Hospital, Agogo, Ghana; University Medical Centre Groningen (UMCG), University of Groningen, Groningen, The Netherlands; Dunkwa Governmental Hospital, Dunkwa-on-Offin, Dunkwa, Ghana; Reconstructive Plastic Surgery and Burns Unit, Department of Surgery, Komfo Anokye Teaching Hospital, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana; Bernhard Nocht Institute for Tropical Medicine (BNITM), Hamburg, Germany Abstract. Standardized antimycobacterial therapy is considered the treatment of choice for Buruli ulcer disease. To assess the prevalence of drug resistance among clinical Mycobacterium ulcerans isolates in Ghana, we conducted a sequence-based approach to detect mutations associated with drug resistance. We subjected clinical samples to direct DNA sequencing of rpoB and rpsL genes and compared culture and whole-genome extracts regarding the efficiency of sequence analysis; 99.1% (rpoB) and 100% (rpsL) of the patients harbored M. ulcerans wild type. In one isolate (0.9%), a point mutation of the rpoB gene at codon Ser522 leading to an amino acid change was detected. Culture extracts yielded a significantly higher sequencing efficiency than whole-genome extracts. Our data suggest a low level of drug resistance in Ghana. However, mutations associated with drug resistance do occur and require monitoring. Improved techniques are necessary to enhance the efficiency of sequence analysis of whole-genome extracts. INTRODUCTION resistance surveillance for leprosy and implemented a genotypic drug-resistance surveillance system in 2009. So far, high concordance between conventional drug-susceptibility testing (mouse footpad technique) and DNA sequencing methods has been observed. In a large-scale study on sequence-based detection of drug resistance in human leprosy, rifampicinresistant M. leprae strains were detected among 2% of new cases and 8% of relapses in Southeast Asia.9–11 The majority of M. tuberculosis strains expressing phenotypic resistance to rifampicin shows mutations within the highly conserved 81-bp RMP resistance-determining region (RRDR) of the rpoB gene comprising codons 507–533, with codons Ser531 and His526 being involved most frequently.12 For M. leprae, rpoB gene mutations leading to phenotypic rifampicin resistance were likewise detected within the RRDR comprising codons 401–427 (equivalent to codons 507–533 in M. tuberculosis), predominantly affecting codon Ser425 (equivalent to Ser531 in M. tuberculosis).13 Resistance of M. tuberculosis to SM was most frequently associated with a single mutation in codon 43 of the rpsL gene and less commonly with mutations of the rrs gene.14,15 To date, clinical M. ulcerans strains resistant to RMP or SM have not yet been reported. However, mutations of the rpoB gene in codons Ser416 and His420 (Ser522 and His526 in M. tuberculosis numbering, respectively) of three M. ulcerans strains causing RMP-resistant phenotypes have been detected after RMP monotherapy of experimentally infected mice.16 With respect to the irregular use of antimycobacterial drugs before introduction of standardized antimycobacterial treatment, we conducted a pilot study in Ghana to evaluate the potential use of molecular tools, specifically gene sequencing, to obtain baseline data on the prevalence of mutations possibly associated with resistance of M. ulcerans to RMP and SM. Buruli ulcer disease (BUD), caused by Mycobacterium ulcerans, is the third most common mycobacterial infection in humans after tuberculosis and leprosy, and it has been reported from more than 30 countries worldwide, with dominant endemic foci in West Africa. BUD involves the skin and the subcutaneous adipose tissue. The disease starts as a painless nodule, papule, plaque, or edema and evolves into a painless ulcer with characteristically undermined edges. If left untreated, severe disability may occur. Previously, BUD was treated by wide surgical excision, and the World Health Organization (WHO) recommended antimycobacterial treatment of 56 days with streptomycin (SM) and rifampicin (RMP), if necessary, followed by surgical excision in 2004.1,2 Most West African countries implemented the standardized antibiotic therapy in 2006. However, non-standardized regimens of RMP and/or SM were also used as concomitant treatment before or after surgical interventions before 2006.3–5 Although effective and in many aspects advantageous over surgery, introduction of antimycobacterial treatment poses new challenges for the management of BUD.1 As is well-known from tuberculosis and leprosy, antimycobacterial treatment is prone to the development of drug resistance. Risk factors hereby encompass a lack of patients’ compliance as well as irregular and inadequate treatment regimens in terms of duration, dosage, and drug combination. The WHO estimates the current global level of drug resistance of M. tuberculosis complex (MTBC) to RMP and SM, as determined by the proportion method, to be 6.3% and 12.6%, respectively.6 Genotypic drug-resistance testing constitutes a reliable and expedient alternative to predict phenotypic drug resistance of M. tuberculosis to RMP and SM.7,8 In 2007, WHO also emphasized the importance of systematic global drug MATERIALS AND METHODS * Address correspondence to Marcus Beissner, Department of Infectious Diseases and Tropical Medicine (DITM), University Hospital, LudwigMaximilians University of Munich, Leopoldstrasse 5, 80802 Munich, Germany. E-mail: marcus.beissner@lrz.uni-muenchen.de Ethics statement. The study was approved by the Committee of Human Research Publication and Ethics, 1059 1060 BEISSNER AND OTHERS School of Medical Sciences, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana (CHRPE/07/01/05). For all patients who were enrolled in a drug trial, written informed consent was obtained from the study participants and/or their legal representatives if aged under 18 years. For patients who underwent routine surgical treatment (with or without concomitant drug therapy), written consent or in the case of the participants who were illiterate, verbal consent, according to the requirements of the ethics committee, for all treatment-related procedures including the collection of diagnostic specimens was obtained in the treatment centers and documented in the medical records of the respective patients. Patients. One hundred sixty-two laboratory-confirmed BUD patients with ulcerative (N = 99) and non-ulcerative (N = 63) lesions seeking treatment in five treatment centers in Ghana (Agogo Presbyterian Hospital, N = 116; Nkawie Government Hospital, N = 30; Dunkwa Government Hospital, N = 12; Global Evangelican Mission Hospital, Apromase-Ashanti, N = 3; Goaso Hospital, N = 1) were enrolled in the study. Between 2004 and 2007, 80 patients (49.4%) were treated by surgical excision with or without concomitant antimycobacterial treatment. At the time of collection of clinical samples, 29 of those patients (36.3%) had not received any antimycobac- terial drugs (ST). Thirty-two patients (40.0%) had received either a combination of RMP and SM (N = 30) or RMP alone (N = 2; ST+). Fourteen patients were treated for ≤ 7 days (short-term; ST+S), and two of those patients received RMP only. Twelve patients received 8–28 days of drug treatment (intermediate; ST+IN), and six were treated for 29–42 days (long term; ST+L). Nineteen patients (23.8%) had received antimycobacterial treatment; however, detailed information on the drugs applied and duration of treatment was not available (ST-NK). The mean duration of treatment before surgical excision and collection of clinical samples was 13.1 days (Table 1). Between 2005 and 2007, according to the clinical records, 82 patients (50.6%) were treated with antimycobacterial drugs alone (DT). Of these, 44 patients were enrolled in a randomized controlled drug trial.17 By the time of collection of clinical samples, 81 (98.8%) of these patients had not received antibiotics; one recurrent case, however, had received a full course of RMP and SM before the trial (Table 1). Clinical samples for follow-up analysis were available from seven patients (antimycobacterial treatment only, N = 6; antibiotics plus surgery, N = 1). At the time of collection of followup samples, five patients had received full-term treatment with RMP and SM for 56 days, and two patients were treated with Table 1 Sequencing results of 162 BUD patients in different treatment groups Results per patient|| rpoB gene results** Treatment group* DT ST Patients† Culture extracts‡ Swab extracts§ 50 37 MUT NR 40 82 56 (68.3) 0 26 (31.7) 82 53 (64.6) 0 29 (35.4) 14 22 29 19 (65.5) 0 10 (34.5) 7 3 (42.9) 0 4 (57.1) 9 8 14 7 (50.0) 0 7 (50.0) 13 5 (38.5) 0 8 (61.5) 4 10 12 4 (33.3) 0 8 (66.7) 11 4 (36.4) 0 7 (63.6) 6 5 (83.3) 0 1 (16.7) 6 5 (83.3) 0 1 (16.7) 1 (5.3) 1 (5.3) 0 2 (33.3) 6 2 5 19 16 Total|| || WT 12 5 ST+NK rpsL gene results†† NR¶¶ 14 1 ST+L MUT§§ 29 1 ST+IN WT‡‡ 82 14 ST+S Tissue extracts¶ 162 87 2 68 Definite results No results 3 88 19 17 (89.5) 162 (100%) 109 (67.3%) 108 (99.1%) 6 4 (66.7) 125 (100%) 74 (59.2%) 74 (59.2%) 0 1 (0.9%) 53 (32.7%) 51 (40.8%) Table 1 provides sequencing results of M. ulcerans isolates from 162 PCR-confirmed BUD patients. Two hundred forty-three DNA extracts were analyzed from these patients, and one final consensus result per patient is shown. From seven follow-up patients, sequencing results of isolates from visit one are listed only. * Patients were divided into different treatment groups as follows. DT = drug treatment [patients received a full course of rifampicin (RMP) and streptomycin (SM) for 56 days after specimen collection; one recurrent case, however, had received a full term of antimycobacterial treatment at the time of specimen collection]; ST = surgical treatment (patients were treated by surgical excision only); ST+S = surgical treatment plus short-term antimycobacterial treatment before specimen collection with RMP and SM for ≤ 7 days (two patients received RMP only); ST+IN = surgical treatment plus intermediate-term antimycobacterial treatment before specimen collection with RMP and SM for 8–28 days; ST+L = surgical treatment plus long-term antimycobacterial treatment before specimen collection with RMP and SM for > 28 days; ST+NK = patients were surgically treated and had received antibiotics before specimen collection, but exact treatment data was missing in clinical records. † Number of patients per treatment group. ‡ Number of analyzed culture extracts from swab and/or tissue specimens. § Number of sequenced whole-genome extracts from swab specimens. ¶ Number of sequenced whole-genome extracts from tissue specimens. || Final sequencing result per patient. Percentages are given in brackets. ** RpoB is the gene for RNA polymerase β subunit (partial sequence = 342 bp). †† RpsL is the gene for ribosomal protein S12 (complete sequence = 375 bp). ‡‡ WT = wild type. The analyzed sequence corresponds 100% to the WT nucleotides of M. ulcerans strain Agy99 (accession number CP000325). §§ MUT = mutation. One isolate showed a point MUT at Ser522 of the rpoB gene. Species identification by sequence analysis of genes for 16S rRNA, rpsL, hsp65, and the ITS allowed the distinct allocation to M. ulcerans strain Agy99 (accession number CP000325) by 100% nucleotide concordance. ¶¶ NR = no result. Sequences were non-analyzable because of mixed sequences, deviation > 3% from M. ulcerans wild type, or inability to amplify. || || From 24 patients, a swab and a tissue extract were sequenced; additionally, 12 culture extracts were analyzed from these patients. For 45 patients, a swab or tissue extract was available with a corresponding culture extract inoculated from the respective clinical specimen. 1061 GENOTYPIC DRUG RESISTANCE TESTING IN M. ULCERANS RMP and SM for 21 days. Analysis of these specimens was done separately and is not provided in the attached tables. Clinical samples and laboratory confirmation. Clinical samples for laboratory confirmation were collected according to standardized procedures. Hereby, diagnostic swabs and tissue specimens from surgical patients were collected at the time of surgery, and diagnostic swabs and punch biopsies were collected from patients receiving antimycobacterial treatment before onset of treatment. As described elsewhere, the clinical samples were subjected to a dry-reagent based IS2404 polymerase chain reaction (DRB-PCR) and culture on LoewensteinJensen media at the Kumasi Center for Collaborative Research in Tropical Medicine, Kumasi, Ghana (KCCR) according to standardized procedures, including external quality assurance by standard IS2404 PCR. Mycobacterial cultures were subjected to a confirmatory IS2404 PCR.3,5,18–21 Samples subjected to sequence analysis. Suspensions of IS2404 PCR confirmed M. ulcerans cultures (N = 87) dissolved in 700 µL Cell Lysis Solution (Qiagen, Hilden, Germany) followed by inactivation at 80°C for 20 minutes, and IS2404 PCR-positive whole-genome extracts (total of 156 genome extracts) derived from swab (N = 68) and tissue samples (N = 88) were subjected to sequence analysis of rpoB and rpsL genes at the Department of Infectious Diseases and Tropical Medicine, University of Munich (DITM). Samples were stored at −20°C before shipment and on arrival at DITM, and shipment (courier service) was conducted at ambient temperature. Briefly, DNA was prepared using the Puregene DNA isolation kit (Gentra Systems) as described elsewhere (Table 1).19 Amplification of rpoB and rpsL genes. A partial sequence of the mycobacterial rpoB gene (342 bp) encompassing the RRDR was amplified by Mycobacterium genus-specific primers as described by Kim and others.22 A set of Mycobacterium genus-specific primers (rpsL-F: 5′-AAC AGC GAG AAC GAA AGC C- 3′; rpsL-R: 5′-TCA CCA GTT GCG TGA CCA G-3′) was used to amplify a sequence, including the entire rpsL gene (375 bp). The thermal-cycling protocol consisted of initial denaturation at 95°C (7 minutes) followed by 37 repeats [95°C (20 seconds), 52°C (25 seconds), and 72°C (45 seconds)] and a final extension at 72°C (5 minutes). Because of the very low frequency of rrs gene mutation reported from phenotypic SM-resistant isolates of the MTBC, we did not analyze this region in M. ulcerans.14,15 Gel extraction of PCR products. Amplicons were electrophoresed in a 1.2% agarose gel prepared with TrisAcetate EDTA (TAE) buffer light (Roth, Karlsruhe, Germany). Positive bands where cut out with a single sterile scalpel for each amplicon. Purification was carried out using the Millipore Ultrafree DA kit (Roth). DNA sequencing. Cycle sequencing was performed according to the manufacturer’s protocol on an ABI3730 automatic sequencer (Applied Biosystems, Darmstadt, Germany) at Helmholtz Research Center, Neuherberg, Germany. For each gene, a forward and a reverse sequence were generated. Species identification. In case deviant nucleotide sequences from M. ulcerans wild type (Agy99, complete genome; accession number CP000325) were detected, verification of the species M. ulcerans or identification of a contamination by another species was attempted by sequence analysis of the following genes using Mycobacterium genus-specific primers: 16S rRNA gene (924 bp), 65 kDa heat shock protein (HSP) gene (644 bp), and 16S-23S rRNA internal transcribed spacer gene (ITS; 220 bp) in accordance with the authors’ protocols. With respect to the observations regarding the inaccuracy of 16S rRNA sequencing results of non-tuberculous mycobacteria made by Turenne and others,24 sequencing of the two additional regions of the M. ulcerans genome served as a quality-assurance measure.23–26 Sequence analysis. Sequences were analyzed using DNASIS Max software (MiraiBio, San Francisco, CA) and aligned with the M. ulcerans wild-type sequence (Agy99) of the respective gene. BLASTn analysis was performed on entries of GenBank. Quality assurance of 16S rRNA and ITS results was performed within ribosomal differentiation of microorganisms (RIDOM).24 Definite sequences were defined as wild type (WT; 100% nucleotide concordance with M. ulcerans, Agy99) or mutation (MUT; < 3% nucleotide deviation from the WT for the respective gene and positive species identification for M. ulcerans). No result (NR) subsumes non-analyzable sequences [i.e., negative (non-amplifiable) and contaminated (mixed) sequences as well as sequences deviating > 3% from the WT]. For each patient, the sequencing results of different specimens were aligned, and the consensus result (WT, MUT, or NR) is shown in Table 1. Comparison of whole-genome extracts with culture extracts. The efficiency of sequencing was defined as the number of analyzable sequences divided by the number of extracts subjected to sequencing. The overall efficiencies of sequencing of rpoB and rpsL genes from all whole-genome extracts analyzed in this study were compared with those from all available culture extracts of clinical specimens from all patients and types of lesions (Table 2). Comparison of whole-genome extracts with corresponding culture extracts. From 45 patients (ulcerative and nonulcerative lesions), whole-genome extracts and culture material derived from corresponding swab and/or tissue samples obtained from the same patients were available for comparison. The efficiencies of sequencing of rpoB and rpsL genes from whole-genome extracts were directly compared with those of the corresponding culture extracts (Table 3). Statistical analysis. For all statistical analyses, approximate tests (χ2 tests) and exact tests (Fisher’s exact tests) were conducted using Stata software (version 9.0; Stata Corp., College Station, TX). The results of statistical analyses were presented by means of P values, whereby significant differences were defined as P values below 0.05. P values did not serve only for hypothesis testing but as base for discussion. The study was cross-sectional, and no specific selection or randomization of study participants was performed. RESULTS Laboratory confirmation. From all 162 patients (100%), the IS2404 DRB-PCR result was positive. Of those, positive culture results (confirmed by standard IS2404 PCR) were obtained from 87 patients (53; 7%). RpoB sequencing results per patient. Definite sequencing results of the rpoB gene (obtained from culture isolates and/or whole-genome extracts from swab and/or tissue samples) were retrieved from 109 of 162 laboratory-confirmed cases (67.3%). The rpoB WT sequence of M. ulcerans was detected in 108 patients (99.1%; Table 1). The isolate of one patient (0.9%; treatment group S+NK) showed a mutation at codon Ser522 1062 BEISSNER AND OTHERS Table 2 Comparison of rpoB and rpsL gene-sequencing results of whole-genome extracts and culture extracts Culture extracts* Sequenced genes rpoB§ Total analyzed Definite sequences¶ Non-analyzable|| Negative** Mixed†† rpsL‡‡ Total analyzed Definite sequences¶ Non-analyzable|| Negative** Mixed†† Whole-genome extracts† Swab extracts Tissue extracts Total Swab extracts Tissue extracts Total 39 35 (89.7%) 4 (10.3%) 3 (75%) 1 (25%) 57 53 (93.0%) 4 (7.0%) 3 (75%) 1 (25%) 96 88 (91.7%) 8 (8.3%) 6 (75%) 2 (25%) 64 25 (39.1%) 39 (60.9%) 23 (59%) 16 (41%) 92 30 (32.6%) 62 (67.4%) 41 (66.1%) 21 (33.9%) 156 55 (35.3%) 101 (64.7%) 64 (63.4%) 37 (36.6%) 26 23 (88.5%) 3 (11.5%) 2 (66.7%) 1 (33.3%) 41 33 (80.5%) 8 (19.5%) 7 (87.5%) 1 (12.5%) 67 56 (83.6%) 11 (16.4%) 9 (81.8%) 2 (18.2%) 50 16 (32.0%) 34 (68.0%) 21 (61.8%) 13 (38.2%) 69 27 (39.1%) 42 (60.9%) 31 (73.8%) 11 (26.2%) 119 43 (36.1%) 76 (63.9%) 52 (68.4%) 24 (31.6%) P value‡ < 0.01 < 0.01 Table 2 shows rpoB- and rpsL-sequencing results from whole-genome extracts and culture extracts of 162 IS2404 PCR-confirmed BUD patients. * DNA was extracted from mycobacterial cultures that were inoculated from swab or tissue specimens. † DNA was extracted from tissue or swab specimens as whole-genome extracts. ‡ P value is the comparison of efficiencies of definite sequences from the total of whole-genome extracts and culture extracts. § rpoB is a 324-bp region of the mycobacterial rpoB gene comprising the rifampicin resistance-determining region (RRDR). ¶ Number of extracts with definite sequences. || Number of extracts with non-analyzable sequences (the result is a mixed or non-amplifiable sequence in PCR). ** Number of extracts with non-amplifiable sequences (negative). †† Mixed sequence is the number of extracts that could not be analyzed, because amplified sequences were contaminated. ‡‡ RpsL is the complete sequence of the gene encoding the ribosomal protein S12 (375 bp). (M. tuberculosis numbering) within the RRDR, leading to a transversion of thymine to guanine (TCC to GCC) and resulting in an amino acid change from serine to alanine. Species identification allowed the distinct allocation of the strain to M. ulcerans by 100% nucleotide concordance with the 16S rRNA, rpsL, hsp65 genes and the ITS of M. ulcerans strain Agy99. For 53 patients (32.7%), no result could be obtained (NR). Among these, five strains showed highly deviated rpoB sequences (> 3% aberrance) from the M. ulcerans WT. Species identification, however, did not allow the allocation to a disTable 3 Comparison of corresponding whole-genome extracts and culture extracts Corresponding specimens* Sequenced genes rpoB¶ Total analyzed Definite sequences|| Non-analyzable** Negative§ Mixed†† rpsL‡‡ Total analyzed Definite sequences|| Non-analyzable** Negative§ Mixed†† Culture extracts† Whole-genome extracts‡ 45 44 (97.8%) 1 (2.2%) 0 1 (100%) 45 19 (42.2%) 26 (57.8%) 12 (46.2%) 14 (53.8%) 45 41 (91.1%) 4 (8.9%) 4 (100%) 0 45 16 (35.6%) 29 (64.4%) 19 (65.5%) 10 (34.5%) P value§ < 0.01 < 0.01 Table 3 shows rpoB- and rpsL-sequencing results of corresponding whole-genome and culture extracts. * From 45 patients, culture extracts (inoculated from swab or tissue specimens) were analyzed, and results are compared with the corresponding whole-genome extract. † DNA was extracted from cultured strains that were inoculated from swab or tissue specimens (for ulcerative lesions) or solely from tissue specimens of non-ulcerative lesions. ‡ DNA was extracted from tissue or swab specimens as whole-genome extracts from ulcerative lesions and solely from tissue specimens of non-ulcerative lesions. Per patient, either a swab or a tissue specimen was analyzed corresponding to the specimen of which a mycobacterial culture was inoculated and a positive confirmatory IS2404 PCR result was obtained. ¶ rpoB is a 324-bp sequence of the mycobacterial rpoB gene comprising the rifampicin resistance-determining region (RRDR). || Number of extracts with definite sequences. ** Number of extracts with non-analyzable sequences (the result is a mixed or non-amplifiable sequence in PCR). § Number of extracts with non-amplifiable sequences (negative). †† Mixed sequence is the number of extracts that could not be analyzed, because amplified sequences were contaminated. ‡‡ RpsL is the sequence of the gene for ribosomal protein S12 (complete sequence = 375 bp). tinct mycobacterial species; therefore, a contamination with closely related species could not be excluded for these cases. RpsL sequencing results per patient. Definite sequencing results of the rpsL gene (obtained from culture isolates and/or whole-genome extracts from swab and/or tissue samples) were retrieved for 74 of 125 laboratory-confirmed patients (59.2%). All of these showed the M. ulcerans WT rpsL sequence. For 51 patients (40.8%), no results could be obtained (NR; Table 1). Comparison of whole-genome extracts with culture extracts. Among all clinical specimens analyzed in this study, the overall efficiency of rpoB sequencing of culture extracts (91.7%) was significantly higher (P < 0.01) than the efficiency of sequencing of whole-genome extracts (35.3%). Likewise, the overall efficiency of rpsL sequencing of culture extracts (83.6%) was significantly higher (P < 0.01) than the efficiency of sequencing of whole-genome extracts (36.1%). Among the whole-genome extracts, 64.7% of the rpoB sequences (63.4% negative and 36.6% mixed) and 63.9% of the rpsL sequences (68.4% negative and 31.6% mixed) were non-analyzable. Among the culture extracts, 8.3% of the rpoB sequences (75% negative and 25% mixed) and 16.4% of the rpsL sequences (81.8% negative and 18.2% mixed) were nonanalyzable (Table 2). Comparison of whole-genome extracts with corresponding culture extracts. Among clinical samples from patients with corresponding culture extracts and whole-genome extracts, the efficiency of rpoB sequencing of culture extracts (97.8%) was significantly higher (P < 0.01) than that of whole-genome extracts (44.2%). Likewise, the efficiency of rpsL sequencing from culture extracts (91.1%) was significantly higher (P < 0.01) than the efficiency of sequencing of whole-genome extracts (35.6%; Table 3). Follow-up patients. Among follow-up samples of five patients who had received full-term antibiotic treatment, WT rpoB and rpsL sequences were detected in isolates from three patients, whereas sequences from two patients’ isolates were not amplifiable; thus, no results were retrieved (NR). For two patients who had received 21 days of antimycobacterial treatment, rpoB and rpsL WT sequences were detected for GENOTYPIC DRUG RESISTANCE TESTING IN M. ULCERANS one patient and could not be amplified from specimens of the second patient (NR). DISCUSSION Conventional in vitro resistance testing according to the proportion method constitutes the most widespread technique within the WHO global surveillance system of antimycobacterial drug resistance of tuberculosis.6 Because of the inability to cultivate M. leprae, sequence-based detection of drug resistance has successfully been applied for leprosy.10 According to WHO recommendations, M. leprae isolates with an amino acid change in a drug resistance-determining region that have been confirmed by mouse footpad testing to confer phenotypic drug resistance are scored as resistant.11 Sequence-based resistance testing was also proven to be a rapid and reliable alternative to conventional resistance testing of M. tuberculosis.7,8 M. ulcerans strains expressing phenotypic and genotypic resistance were generated under RMP monotherapy in a mouse model, but systematic drug resistance surveillance in clinical M. ulcerans strains has not yet been conducted.16 To obtain baseline data on resistance to RMP and SM in Ghana, we screened clinical M. ulcerans isolates obtained from BUD patients treated by surgery and/or antimycobacterial drugs. At the time of the study, conventional susceptibility testing for M. ulcerans was not established in Ghana; therefore, we applied a sequence-based approach for the detection of mutations associated with drug resistance. Our findings showed no mutations among patients without previous antimycobacterial treatment. One strain isolated from a patient treated by surgery and concomitant antibiotic therapy as early as 2004 (information on drug combination, dosage, and duration of treatment could not be retrieved from the files) expressed a mutation at codon Ser522 of the rpoB gene. The respective mutation was also described by Marsollier et al.16 after RMP monotherapy of experimentally infected mice, and phenotypic resistance was confirmed.16 These findings suggest that antimycobacterial treatment, especially if administered as monotherapy or in an irregular, nonstandardized fashion, may also lead to rpoB mutations of human M. ulcerans isolates. To establish phenotypic correlates of the mutation detected in the respective strain, conventional susceptibility testing was attempted. However, we did not succeed in obtaining subcultures from the original isolate. Mutations of the rpsL gene were not detected. This may be related to the fact that streptomycin is applied intramuscularly and has presumably not been administered in monotherapy. According to current WHO recommendations, all new BUD cases are subjected to drug treatment. Patients who develop a new BUD lesion after complete healing of the initial lesion (recurrences) and BUD patients who missed a total of 14 days since the start of treatment (defaulters) may receive a second course of antimycobacterial therapy with regard of the SM lifetime dose (90 g in adults).27 Since the introduction of antimycobacterial treatment in 2006, more than 3,000 BUD cases have been reported in Ghana, and presumably, the majority of these cases have been subjected to drug treatment (Asiedu K, personal communication). Whereas human-to-human transmission plays a crucial role for the spread of resistant MTBC and M. leprae strains, according to current knowledge, M. ulcerans is acquired from the environment.1,28–30 With only a few reported cases of infections contracted from humans, the risk 1063 of transmitting resistant strains among populations afflicted with BUD may be considered minimal.31,32 The rate of drug resistance detected in our study among clinical M. ulcerans isolates obtained between 2004 and 2007 was low (0.9%). Nevertheless, the emergence of drug-resistant strains is possible and will, in the first place, affect the treatment outcome of individual patients under antimycobacterial treatment.16,33,34 Therefore, monitoring of drug resistance will facilitate individual clinical management decisions, especially in recurrences, defaulters, and patients with non-healing lesions. With respect to the long generation time of M. ulcerans and the limited sensitivity of the method, especially in pre-treated patients, cultures alone are not ideal for genotypic drug resistance testing.5 Especially for clinical management decisions, whole-genome extracts constitute a better diagnostic target to obtain rapid results. In our study, analysis of cultured isolates yielded definite sequencing results for rpoB and rpsL genes in > 80%, whereas the number of definite results obtained from whole-genome extracts was significantly lower (< 40% for rpoB and rpsL genes). The percentage of non-analyzable sequences among whole-genome extracts was > 60%. Among these, the respective sequences of 63.4% of the rpoB and 68.4% of the rpsL genes were non-amplifiable. In general, the presence of only a low amount of mycobacterial DNA in clinical samples often hampers diagnostic procedures and may also constitute a source of error in the present study.35 Amplification of mixed sequences from related species existing as commensals on the human skin represents another challenge in sequencing of whole-genome extracts. According to the results of this study, the authors consider it essential to apply refined techniques in further studies. The design of more specific primers can improve the efficiency of sequencing of whole-genome extracts. Optimization of extraction procedures (e.g., mechanical homogenization by zirconium beads followed by enzymatic lysis) can augment the yield of M. ulcerans DNA recovered from swab and tissue samples.36 Immunomagnetic separation, developed and successfully applied in environmental studies by Marsollier and others,37 may be a promising tool to concentrate and purify M. ulcerans from clinical samples; however, this technique is currently restricted to BUD reference laboratories or collaborative research programs. From 12% of the study subjects, documentation of previous antimycobacterial treatment could not be retrieved from the clinical records. These partially incomplete sets of data, therefore, constitute a weakness of this study. With respect to correct interpretation of results, further studies on drug resistance in M. ulcerans should aim at obtaining complete and detailed patient-related information. WHO provides a form (BU01) for the recording of clinical, epidemiological, and treatment data, which, in general, is available in all endemic countries. A consequent use of the BU01 form will facilitate the collection of relevant information.1 Received May 9, 2010. Accepted for publication July 30, 2010. Acknowledgments: Primer sequences (rpsL-F and rpsL-R) were kindly provided by F. Portaels and P. Stragier (Institute for Tropical Medicine, Antwerp, Belgium). The authors thank David Schenavsky for proofreading of the manuscript. The article contains parts of the doctoral dissertation of Marcus Beissner. There is no conflict of interest among authors. 1064 BEISSNER AND OTHERS Financial support: This project was supported by the European Commission (Project INCO-CT-2005-015476-BURULICO). Authors’ addresses: Marcus Beissner, Joerg Nitschke, Karl-Heinz Herbinger, Vera Siegmund, Erna Fleischmann, Thomas Loscher, and Gisela Bretzel, Department of Infectious Diseases and Tropical Medicine (DITM), University Hospital, Ludwig-Maximilians University of Munich, Munich, Germany, E-mails: marcus.beissner@lrz.unimuenchen.de, j.a.nitschke@web.de, herbinger@lrz.uni-muenchen.de, v.siegmund@eurice.de, erna.fleischmann@lrz.uni-muenchen.de, loe scher@lrz.uni-muenchen.de, and bretzel@lrz.uni-muenchen.de. NanaYaa Awua-Boateng and Ohene Adjei, Kumasi Centre for Collaborative Research in Tropical Medicine (KCCR), Kwame Nkrumah University of Science and Technology, Kumasi, Ghana, E-mails: awua.boateng@ kccr.de and oadjei@africaonline.com.gh. William Thompson, Agogo Presbyterian Hospital, Agogo, Ghana, E-mail: wnat@agogohospital .org.Willemien A. Nienhuis and Tjip S. van der Werf, University Medical Centre Groningen (UMCG), University of Groningen, Groningen, The Netherlands, E-mails: wiannix@hotmail.com and t.s.van.der .werf@int.umcg.nl. Erasmus Klutse, Dunkwa Governmental Hospital, Dunkwa-on-Offin, Dunkwa, Ghana, E-mail: eyklutse@yahoo.com. Pius Agbenorku, Reconstructive Plastic Surgery and Burns Unit, Department of Surgery, Komfo Anokye Teaching Hospital, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana, E-mail: pimagben@yahoo.com. Bernhard Fleischer, Bernhard Nocht Institute for Tropical Medicine (BNITM), Hamburg, Germany, E-mail: fleischer@bni-hamburg.de. Reprints requests: Gisela Bretzel, Department of Infectious Diseases and Tropical Medicine, University Hospital, Ludwig-Maximilians University of Munich, Leopoldstrasse 5, 80802 Munich, Germany, E-mail: bretzel@lrz.uni-muenchen.de. REFERENCES 1. World Health Organization, 2008. Buruli ulcer: progress report, 2004–2008. Wkly Epidemiol Rec 17: 145–154. 2. World Health Organization, 2004. Provisional Guidance on the Role of Specific Antibiotics in the Management of Mycobacterium ulcerans Disease (Buruli Ulcer). Geneva, Switzerland: World Health Organization. 3. Siegmund V, Adjei O, Nitschke J, Thompson W, Klutse E, Herbinger KH, Thompson R, van Vloten F, Racz P, Fleischer B, Loescher T, Bretzel G, 2007. Dry reagent-based polymerase chain reaction compared with other laboratory methods available for the diagnosis of Buruli ulcer disease. Clin Infect Dis 45: 68–75. 4. Schunk M, Thompson W, Klutse E, Nitschke J, Opare-Asamoah K, Thompson R, Fleischmann E, Siegmund V, Herbinger KH, Adjei O, Fleischer B, Loscher T, Bretzel G, 2009. Outcome of patients with Buruli ulcer after surgical treatment with or without antimycobacterial treatment in Ghana. Am J Trop Med Hyg 81: 75–81. 5. Herbinger KH, Adjei O, Awua-Boateng NY, Nienhuis WA, Kunaa L, Siegmund V, Nitschke J, Thompson W, Klutse E, Agbenorku P, Schipf A, Reu S, Racz P, Fleischer B, Beissner M, Fleischmann E, Helfrich K, van der Werf TS, Loscher T, Bretzel G, 2009. Comparative study of the sensitivity of different diagnostic methods for the laboratory diagnosis of Buruli ulcer disease. Clin Infect Dis 48: 1055–1064. 6. World Health Organization, 2008. Anti-Tuberculosis Drug Resistance in the World. Report No. 4. Geneva, Switzerland: World Health Organization. 7. Van Rie A, Warren R, Mshanga I, Jordaan AM, van der Spuy GD, Richardson M, Simpson J, Gie RP, Enarson DA, Beyers N, van Helden PD, Victor TC, 2001. Analysis for a limited number of gene codons can predict drug resistance of Mycobacterium tuberculosis in a high-incidence community. J Clin Microbiol 39: 636–641. 8. Telenti A, Honore N, Bernasconi C, March J, Ortega A, Heym B, Takiff HE, Cole ST, 1997. Genotypic assessment of isoniazid and rifampin resistance in Mycobacterium tuberculosis: a blind study at reference laboratory level. J Clin Microbiol 35: 719–723. 9. World Health Organization, 2007. Rifampicin Resistance in Leprosy. Report of an Informal Consultation. Geneva, Switzerland: World Health Organization. 10. Matsuoka M, Budiawan T, Aye KS, Kyaw K, Tan EV, Cruz ED, Gelber R, Saunderson P, Balagon V, Pannikar V, 2007. The frequency of drug resistance mutations in Mycobacterium leprae isolates in untreated and relapsed leprosy patients from Myanmar, Indonesia and the Philippines. Lepr Rev 78: 343–352. 11. World Health Organization, 2009. Guidelines for Global Surveillance of Drug Resistance in Leprosy. Geneva, Switzerland: World Health Organization. 12. Morlock GP, Plikaytis BB, Crawford JT, 2000. Characterization of spontaneous in vitro-selected, rifampin-resistant mutants of Mycobacterium tuberculosis strain H37Rv. Antimicrob Agents Chemother 44: 3298–3301. 13. Sapkota BR, Ranjit C, Neupane KD, Macdonald M, 2008. Development and evaluation of a novel multiple-primer PCR amplification refractory mutation system for the rapid detection of mutations conferring rifampicin resistance in codon 425 of the rpoB gene of Mycobacterium leprae. J Med Microbiol 57: 179–184. 14. Das R, Gupta P, Singh P, Chauhan DS, Katoch K, Katoch VM, 2009. Association of mutations in rpsL gene with high degree of streptomycin resistance in clinical isolates of Mycobacterium tuberculosis in India. Indian J Med Res 129: 108–110. 15. Mieskes KT, Rusch-Gerdes S, Truffot-Pernot C, Feldmann K, Tortoli E, Casal M, Loscher T, Rinder H, 2000. Rapid, simple, and culture-independent detection of rpsL codon 43 mutations that are highly predictive of streptomycin resistance in Mycobacterium tuberculosis. Am J Trop Med Hyg 63: 56–60. 16. Marsollier L, Honore N, Legras P, Manceau AL, Kouakou H, Carbonnelle B, Cole ST, 2003. Isolation of three Mycobacterium ulcerans strains resistant to rifampin after experimental chemotherapy of mice. Antimicrob Agents Chemother 47: 1228–1232. 17. Nienhuis WA, Stienstra Y, Thompson WA, Awuah PC, Abass KM, Tuah W, Awua-Boateng NY, Ampadu EO, Siegmund V, Schouten JP, Adjei O, Bretzel G, van der Werf TS, 2010. Antimicrobial treatment for early, limited Mycobacterium ulcerans infection: a randomized controlled trial. Lancet 375: 664–672. 18. Stinear T, Ross BC, Davies JK, Marino L, Robins-Browne RM, Oppedisano F, Sievers A, Johnson PD, 1999. Identification and characterization of IS2404 and IS2606: two distinct repeated sequences for detection of Mycobacterium ulcerans by PCR. J Clin Microbiol 37: 1018–1023. 19. Siegmund V, Adjei O, Racz P, Berberich C, Klutse E, van Vloten F, Kruppa T, Fleischer B, Bretzel G, 2005. Dry reagent-based PCR as a novel tool for laboratory confirmation of clinically diagnosed Mycobacterium ulcerans-associated disease in areas in the tropics where M. ulcerans is endemic. J Clin Microbiol 43: 271–276. 20. Bretzel G, Siegmund V, Nitschke J, Herbinger KH, Thompson R, Fleischmann E, Fleischer B, Adjei O, 2006. External quality assurance for the laboratory diagnosis of Buruli ulcer disease in Ghana. Trop Med Int Health 11: 1688–1693. 21. Bretzel G, Siegmund V, Nitschke J, Herbinger KH, Thompson W, Klutse E, Crofts K, Massavon W, Etuaful S, Thompson R, Asamoah-Opare K, Racz P, Vloten F, van Berberich C, Kruppa T, Ampadu E, Fleischer B, Adjei O, 2007. A stepwise approach to the laboratory diagnosis of Buruli ulcer disease. Trop Med Int Health 12: 89–96. 22. Kim BJ, Lee SH, Lyu MA, Kim SJ, Bai GH, Chae GT, Kim EC, Cha CY, Kook YH, 1999. Identification of mycobacterial species by comparative sequence analysis of the RNA polymerase gene (rpoB). J Clin Microbiol 37: 1714–1720. 23. Talaat AM, Reimschuessel R, Trucksis M, 1997. Identification of mycobacteria infecting fish to the species level using polymerase chain reaction and restriction enzyme analysis. Vet Microbiol 58: 229–237. 24. Turenne CY, Tschetter L, Wolfe J, Kabani A, 2001. Necessity of quality-controlled 16S rRNA gene sequence databases: identifying nontuberculous Mycobacterium species. J Clin Microbiol 39: 3637–3648. 25. Kim H, Kim SH, Shim TS, Kim MN, Bai GH, Park YG, Lee SH, Cha CY, Kook YH, Kim BJ, 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: 199–209. 26. Roth A, Reischl U, Streubel A, Naumann L, Kroppenstedt RM, Habicht M, Fischer M, Mauch H, 2000. Novel diagnostic algorithm for identification of mycobacteria using genus-specific GENOTYPIC DRUG RESISTANCE TESTING IN M. ULCERANS 27. 28. 29. 30. 31. amplification of the 16S-23S rRNA gene spacer and restriction endonucleases. J Clin Microbiol 38: 1094–1104. World Health Organization, 2007. Summary Report of the WHO Annual Meeting on Buruli Ulcer. Geneva, Switzerland: World Health Organization. Eddyani M, Ofori-Adjei D, Teugels G, De Weirdt D, Boakye D, Meyers WM, Portaels F, 2004. Potential role for fish in transmission of Mycobacterium ulcerans disease (Buruli ulcer): an environmental study. Appl Environ Microbiol 70: 5679–5681. Johnson PD, Azuolas J, Lavender CJ, Wishart E, Stinear TP, Hayman JA, Brown L, Jenkin GA, Fyfe JA, 2007. Mycobacterium ulcerans in mosquitoes captured during outbreak of Buruli ulcer, southeastern Australia. Emerg Infect Dis 13: 1653–1660. van der Werf TS, Stienstra Y, Johnson RC, Phillips R, Adjei O, Fleischer B, Wansbrough-Jones MH, Johnson PD, Portaels F, van der Graaf WT, Asiedu K, 2005. Mycobacterium ulcerans disease. Bull World Health Organ 83: 785–791. Debacker M, Zinsou C, Aguiar J, Meyers WM, Portaels F, 2003. First case of Mycobacterium ulcerans disease (Buruli ulcer) following a human bite. Clin Infect Dis 36: e67–e68. 1065 32. Maranes Antonanzas I, Lopez Gallardo Y, Garcia Pardo A, Abella Vazquez ML, 2006. Transmission of Buruli ulcer from person to person. Rev Clin Esp 206: 256–257. 33. Asiedu K, Wansbrough-Jones M, 2007. Mycobacterium ulcerans infection (Buruli or Bairnsdale ulcer): challenges in developing management strategies. Med J Aust 186: 55–56. 34. Eddyani M, Debacker M, Martin A, Aguiar J, Johnson CR, Uwizeye C, Fissette K, Portaels F, 2008. Primary culture of Mycobacterium ulcerans from human tissue specimens after storage in semisolid transport medium. J Clin Microbiol 46: 69–72. 35. Pierre-Audigier C, 1994. Recent biological techniques for the diagnosis of mycobacterial infections. Presse Med 23: 665–670. 36. Hosek J, Svastova P, Moravkova M, Pavlik I, Bartos M, 2006. Methods of mycobacterial DNA isolation from different biological material: a review. Vet Med (Praha) 51: 180–192. 37. Marsollier L, Robert R, Aubry J, Saint Andre JP, Kouakou H, Legras P, Manceau AL, Mahaza C, Carbonnelle B, 2002. Aquatic insects as a vector for Mycobacterium ulcerans. Appl Environ Microbiol 68: 4623–4628.