Abstract
Fluoroquinolone resistance in Mycobacterium tuberculosis can be conferred by mutations in gyrA or gyrB. The prevalence of resistance mutations outside the quinolone resistance-determining region (QRDR) of gyrA or gyrB is unclear, since such regions are rarely sequenced. M. tuberculosis isolates from 1,111 patients with newly diagnosed culture-confirmed tuberculosis diagnosed in Tennessee from 2002 to 2009 were screened for phenotypic ofloxacin resistance (>2 μg/ml). For each resistant isolate, two ofloxacin-susceptible isolates were selected: one with antecedent fluoroquinolone exposure and one without. The complete gyrA and gyrB genes were sequenced and compared with M. tuberculosis H37Rv. Of 25 ofloxacin-resistant isolates, 11 (44%) did not have previously reported resistance mutations. Of these, 10 had novel polymorphisms: 3 in the QRDR of gyrA, 1 in the QRDR of gyrB, and 6 outside the QRDR of gyrA or gyrB; 1 did not have any gyrase polymorphisms. Polymorphisms in gyrA codons 1 to 73 were more common in fluoroquinolone-susceptible than in fluoroquinolone-resistant strains (20% versus 0%; P = 0.016). In summary, almost half of fluoroquinolone-resistant M. tuberculosis isolates did not have previously described resistance mutations, which has implications for genotypic diagnostic tests.
INTRODUCTION
Fluoroquinolones (FQs) have potent in vitro and in vivo activity against Mycobacterium tuberculosis. They are recommended for the treatment of drug-resistant tuberculosis (TB) and for persons intolerant of current first-line therapy (4). Recent clinical trials have demonstrated that fluoroquinolones have the potential to shorten the duration of antituberculosis therapy (9, 23, 26).
Fluoroquinolones kill M. tuberculosis via double-stranded breaks in DNA, by binding to DNA gyrase. DNA gyrase consists of two A and two B subunits encoded by the gyrase A (gyrA) and gyrase B (gyrB) genes, respectively. The interaction between fluoroquinolones and DNA gyrase occurs in a conserved region of gyrA (codons 74 to 113) (32) and gyrB (codons 500 to 538) (8) known as the quinolone resistance-determining region (QRDR). Fluoroquinolone resistance in M. tuberculosis is frequently conferred by mutations in the QRDR of gyrA or gyrB (32). However, up to 58% of fluoroquinolone-resistant M. tuberculosis isolates lack known resistance mutations (12, 29, 33, 34). Most studies of fluoroquinolone-resistant M. tuberculosis have limited their assessment of mutations to the QRDR of gyrA and/or gyrB. A systematic review of all studies of genotypic fluoroquinolone resistance in M. tuberculosis published between January 1990 and June 2010 identified 42 studies (2,482 M. tuberculosis clinical isolates), of which only 1 (40 isolates) sequenced the gyrA and gyrB genes in their entirety (22a). The extent to which resistance mutations occur outside the QRDR of gyrA and gyrB is therefore unclear. Diagnostic tests such as line probe assays currently under evaluation for the detection of genotypic fluoroquinolone resistance in M. tuberculosis are limited to mutations in the QRDR of gyrA (6, 16, 19).
We have previously reported the prevalence of and risk factors for phenotypic fluoroquinolone resistance in culture-confirmed tuberculosis in Tennessee between 2002 and 2006 (10). We sought to extend those findings by characterizing genotypic mutations throughout all of gyrA and gyrB in both fluoroquinolone-resistant and -susceptible M. tuberculosis isolates and including isolates from three additional years of study.
MATERIALS AND METHODS
Study population.
The study was conducted among newly diagnosed culture-confirmed tuberculosis patients reported to the Tennessee Department of Health from January 2002 to December 2009. From 2002 to 2006, M. tuberculosis isolates were eligible for inclusion if the patient had ever enrolled in TennCare (Medicaid) (10). Between 2007 and 2009, all patients with culture-confirmed disease were eligible, regardless of TennCare status. Fluoroquinolone exposure in the 6 to 12 months prior to tuberculosis diagnosis was ascertained via the TennCare outpatient pharmacy database (2002 to 2009) as well as inpatient and outpatient medical record review, in-home patient interview, and a brief fluoroquinolone assessment form (2007 to 2009). All M. tuberculosis isolates were screened for phenotypic fluoroquinolone resistance (see below). Cases were defined as persons with M. tuberculosis resistant to ofloxacin at >2 μg/ml. For each resistant isolate, two ofloxacin-susceptible control isolates from different patients diagnosed in the same year as the case isolate were randomly selected: one with antecedent exposure to any fluoroquinolone (e.g., ciprofloxacin, ofloxacin, gatifloxacin, levofloxacin, or moxifloxacin) and one without.
The study was approved by the Institutional Review Boards of Vanderbilt University, the Tennessee Department of Health, and the Davidson County Metro Public Health Department and was also reviewed by the Bureau of TennCare. Participants in the prospective phase of the study (2007 to 2009) provided written informed consent.
Study definitions.
Mutations were defined as nucleotide changes that resulted in a change in amino acid (i.e., nonsynonymous) and that were associated with drug resistance. Base pair changes that did not affect the amino acid (i.e., synonymous) were not considered mutations. Polymorphisms were defined as variants resulting in amino acid changes for which the relationship with drug resistance was uncertain.
Laboratory methods.
All M. tuberculosis isolates were stored at the Tennessee Department of Health mycobacteriology laboratory at −70°C after the initial clinical diagnosis was established. The first M. tuberculosis isolate obtained from each patient was used when available. Patients were excluded from screening if the M. tuberculosis isolate was unavailable. M. tuberculosis isolates were thawed and subcultured onto Lowenstein-Jensen (LJ) medium. A 1.0 McFarland (determined by nephelometer) suspension was prepared from freshly grown colonies on the LJ slant and served as the standard inoculum for all dilutions used in susceptibility testing. Screening for fluoroquinolone resistance was performed via agar proportion using ofloxacin as the representative fluoroquinolone (24). Isolates were identified as resistant if >1% colony growth occurred in the presence of the ofloxacin critical concentration (2 μg/ml) on Middlebrook 7H10 agar. The MIC was determined for each study isolate via an agar proportion method as described elsewhere (1). Inoculum size was confirmed by colony count for each test. The range of ofloxacin concentrations tested was 2 to 1,024 μg/ml for resistant isolates and 0.5 to 2 μg/ml for susceptible isolates. M. tuberculosis reference strain H37Rv (ATCC 27294) was the fluoroquinolone-susceptible control isolate, and a fluoroquinolone-resistant M. tuberculosis isolate from the Centers for Disease Control and Prevention was the resistant control; both were screened with each batch of patient isolates. Laboratory personnel were blinded to the fluoroquinolone exposure status of the patient.
Genotyping (spoligotyping and mycobacterial interspersed repetitive units [MIRU]) was performed on all fluoroquinolone-resistant M. tuberculosis isolates and all M. tuberculosis isolates in culture-confirmed tuberculosis cases in Tennessee from 2004 to 2009.
Extraction of chromosomal DNA from M. tuberculosis isolates.
DNA was extracted from M. tuberculosis isolates by transferring one 10-μg loopful of bacteria cultured on drug-free Lowenstein-Jensen slants to a 1.5-ml tube containing 1 ml of sterile water. Each tube was vortexed at high speed for 30 s and placed in a heat block at 94°C for 10 min. Tubes were then removed, immediately placed on ice, vortexed, and briefly spun down. The supernatant was pipetted into a new 1.5-ml tube.
PCR.
Separate 50-μl PCR mixtures for gyrA and gyrB were set up for each of the case and control M. tuberculosis isolates. The master mix contained 50 mM KCl, 10 mM Tris-HCl (pH 8.3), 2 mM MgCl2 (1.5 mM for gyrB), 0.2 mM deoxynucleoside triphosphate, 2.5 U Taq polymerase, and 0.5 μM gyrA primers F1 (5′-GCA ACC CTG CGT TCG ATT-3′) and R3 (5′-TAA TTG CCC GTC TGG TCT GC-3′) and gyrB primers F (5′-GGT ACA GTG GTG TGC GAC CC-3′) and R (5′-CTG TCA TCT ATT CCT CGT TTG CAA-3′). Amplification of the entire gyrA and gyrB genes was performed separately for each isolate. The amplified DNA fragment was purified with the Qiagen MinElute gel extraction kit (Qiagen, Valencia, CA).
Cloning.
Purified DNA was ligated into a pcR 2.1 vector using the TA cloning kit (Invitrogen, Carlsbad, CA) and transformed Escherichia coli via INVαF′-competent cells. The plasmid DNA was purified using the QIAprep Spin Miniprep kit (Qiagen, Valencia, CA), confirmed, and subsequently digested with EcoRI. The entire gyrA and gyrB genes were cloned.
DNA sequencing.
Plasmids were sequenced via an ABI Prism 3730xl DNA analyzer (Applied Biosystems, Foster City, CA) using BigDye chemistry per the manufacturer's instructions. Applied Biosystem's DNA sequencing software (version 5.0) was used to collect and analyze the raw data. The entire gyrA and gyrB genes were sequenced in both directions using 6 separate internal primer reactions, which included the T7 and M13R primers and 8 additional internal sequencing primers: GyrA-T7-1, 5′-ATG CCT ACA AAA CTG GCC G-3′; GyrA-M13-1, 5′-ACA GGA TCA AAT CGT GGG TG-3′; GyrA-T7-2, 5′-TTC GTG TGC TCC ACC CA-3′; GyrA-M13-2, 5′-ACC GCG GGA ATC CTC TT-3′; GyrB-T7-1, 5′-ACC ATC AAC CTG ACC GAC-3′; GyrB-M13-1, 5′-CGA GTC ACC TCT ACG ACA TAC A-3′; GyrB-T7-2, 5′-ACA CCG AAG TTC AGG CGA T-3′; and GyrB-M13-2, 5′-AGG AAC GCC ATC TCT TGC-3′.
Analysis of polymorphisms.
Sequences were analyzed using Polyphred software (University of Washington). The gyrA and gyrB sequences from the case and control M. tuberculosis isolates were assembled and compared to the M. tuberculosis reference strain sequence (H37Rv from the Pasteur Institute, France [S. Cole; accession number AL123456; annotation NC[00962.2; gyrA, gyrB]). A subset of sequences were sent to another research lab (A. Aubry, Faculté de Médecine Pitié-Salpêtrière, France); they found the same results as our lab.
Polymorphisms in gyrA codons 21, 95, and 668 were excluded from the analysis because they are common polymorphisms not associated with resistance (21, 28, 32). The frequency of polymorphisms and mutations was assessed in 3 regions of both gyrA and gyrB: before, within, and after the QRDR. The QRDR of gyrA was defined as codons 74 to 113 (32); the QRDR of gyrB was defined as codons 500 to 538 (GenBank CAB02426.1) (8).
Statistical analysis.
The Pearson chi-square test compared categorical variables and the Wilcoxon rank-sum test compared continuous variables. The proportions of fluoroquinolone-resistant and -susceptible M. tuberculosis isolates with polymorphisms in specific locations (before, within, and after the QRDR) of gyrA and gyrB were compared with the Pearson chi-square test. The global test controlled for inflation of type I error due to multiple comparisons by comparing the effect of the gyrase region. Two-sided P values of <0.05 were considered statistically significant.
Nucleotide sequence accession numbers.
The 75 new gene sequences for gyrA and gyrB identified in this study are available from GenBank under accession numbers JQ699102 to JQ699176 and JQ683939 to JQ684013, respectively.
RESULTS
There were 1,111 persons with culture-confirmed tuberculosis during the study period, of whom 25 (2.3%) had phenotypic fluoroquinolone resistance. Of the fluoroquinolone-susceptible M. tuberculosis isolates, 50 were included in this study. The clinical and demographic characteristics of the study population according to fluoroquinolone susceptibility test results are in Table 1. Among the 25 cases with fluoroquinolone resistance, 15 (60%) had fluoroquinolone exposure prior to diagnosis. By study design, 25 (50%) of the 50 controls with fluoroquinolone-susceptible tuberculosis had fluoroquinolone exposure prior to diagnosis. There were no significant differences in age, sex, race, country of birth, HIV infection, or frequency of pulmonary tuberculosis between the two groups. The median MIC was 64 μg/ml (interquartile range [IQR], 64 to 256 μg/ml) in fluoroquinolone-resistant M. tuberculosis isolates and 1.0 μg/ml (IQR, 1 to 1 μg/ml) in fluoroquinolone-susceptible isolates (P < 0.001). The organism inoculum sizes of both susceptible and resistant isolates ranged from 1.9 × 104 to 3.5 × 105 CFU/ml in the 10−2 dilution. None of the 75 M. tuberculosis isolates in this study were resistant to isoniazid or rifampin.
Table 1.
Demographic and clinical characteristics of the study populationa
| Characteristic | Value for indicated isolate type |
P valueb | |
|---|---|---|---|
| Fluoroquinolone resistant (n = 25) | Fluoroquinolone susceptible (n = 50) | ||
| Median age (yr) (IQRc) | 64 (40–80) | 58 (45–70) | 0.4 |
| Male sex | 13 (52) | 33 (66) | 0.2 |
| Black race | 6 (24) | 17 (34) | 0.4 |
| U.S. born | 21 (84) | 43 (86) | 0.8 |
| HIV infected | 3 (12) | 6 (12) | 1.0 |
| Site of disease: any pulmonary TB | 18 (72) | 43 (86) | 0.1 |
For each fluoroquinolone-resistant M. tuberculosis isolate, two fluoroquinolone-susceptible control isolates were selected: one with and one without fluoroquinolone exposure prior to tuberculosis diagnosis. Data are presented as number (%) except as noted.
Wilcoxon rank-sum test for continuous variables and the Pearson chi-square test for categorical variables.
IQR, interquartile range.
Of the 25 fluoroquinolone-resistant isolates, 10 had unique genotypes by spoligotyping or MIRU (i.e., were not in clusters) and 15 were in clusters with other tuberculosis cases. Of these 15, 12 were in clusters in which all other isolates were fluoroquinolone susceptible, and 3 shared the same cluster. There were no epidemiologic links identified between the 3 fluoroquinolone-resistant cases.
The prevalence of polymorphisms in gyrA and gyrB in fluoroquinolone-resistant and -susceptible M. tuberculosis isolates is presented in Table 2. There were no statistically significant differences in the proportions of resistant and susceptible isolates with polymorphisms in either gene combined, in gyrA alone, or in gyrB alone. When analysis was limited to the QRDR of gyrA, fluoroquinolone-resistant isolates were more likely to have a polymorphism, but only 68% of resistant isolates had such a polymorphism, and 6% of the susceptible isolates did. Fluoroquinolone-resistant isolates were also significantly more likely to have a previously described resistance mutation. All of these mutations were at codon 90, 91, or 94 of gyrA. Of note, however, is that 11 (44%) of the fluoroquinolone-resistant M. tuberculosis isolates did not have a previously described resistance mutation. Of these 11, 3 (27%) had documented fluoroquinolone exposure prior to diagnosis, compared with 12 of 14 (86%) resistant isolates with previously described resistance mutations (P = 0.003).
Table 2.
Association between phenotypic fluoroquinolone resistance and polymorphisms in gyrA or gyrB of M. tuberculosisa
| Polymorphism location | No. (%) of M. tuberculosis isolates that were: |
P valueb | |
|---|---|---|---|
| Fluoroquinolone resistant (n = 25) | Fluoroquinolone susceptible (n = 50) | ||
| ≥1 polymorphism in gyrA or gyrB | 24 (96) | 44 (88) | 0.26 |
| ≥1 polymorphism in gyrA | 21 (84) | 37 (74) | 0.33 |
| ≥1 polymorphism in gyrB | 20 (80) | 32 (64) | 0.16 |
| ≥1 polymorphism in QRDR of gyrA | 17 (68) | 3 (6) | <0.001 |
| ≥1 polymorphism in QRDR of gyrB | 1 (4) | 3 (6) | 0.72 |
| Any previously described resistance mutationc | 14 (56) | 0 (0) | <0.001 |
Polymorphisms at codons 21, 95, and 668 of gyrA were excluded because they are known to be unassociated with fluoroquinolone resistance. Data are presented as number (%). QRDR, quinolone resistance-determining region.
Pearson chi-square test or Fisher's exact test.
Mutations at codon 90, 91, or 94 in gyrA.
Tables 3 and 4 include all of the mutations and polymorphisms identified in gyrA and gyrB in the 25 fluoroquinolone-resistant and 50 fluoroquinolone-susceptible M. tuberculosis isolates, respectively. Polymorphisms at codons 21, 95, and 668 in gyrA occurred with similar frequencies in resistant and susceptible isolates and were therefore not included in the tables. Their frequencies were as follows for resistant versus susceptible isolates: for codon 21, 96% versus 100%; for codon 95, 76% versus 78%; and for codon 668, 72% versus 76%.
Table 3.
Polymorphisms in the 25 fluoroquinolone-resistant M. tuberculosis isolatesa
| ID no. | FQ exposure prior to TB | Polymorphisms |
MIC (μg/ml) | |||
|---|---|---|---|---|---|---|
| In QRDR of gyrA | Outside QRDR of gyrA | In QRDR of gyrB | Outside QRDR of gyrB | |||
| 44 | Yes | Ala90Val | Ser536Asn, Lys542Arg | Lys284Glu | 8 | |
| 971 | No | Ser91Pro | Ala463Ser | Pro202Leu | 64 | |
| 134 | Yes | Asp94Gly | Gly646Asp, Thr721Ala | Tyr183His | 64 | |
| 14 | Yes | Asp94Gly | Ala175Thr | Lys443Glu | 64 | |
| 467 | Yes | Asp94Asn | Gln409Arg, Thr410Ala, Lys642Asn | 256 | ||
| 534 | Yes | Asp94Gly | Arg357Gln, Val750Ala | Trp213Arg, Ser452Leu | 64 | |
| 648 | Yes | Asp94Tyr | Thr778Ile | Lys635Arg | 64 | |
| 662 | Yes | Asp94Gly | Thr585Ala, Pro606Ser | 64 | ||
| 753 | Yes | Asp94Asn | Tyr622Cys | Asn438Asp, Arg592Gln | 128 | |
| 853 | Yes | Asp94Gly | Gly247Ser, Ile287Thr | 256 | ||
| 969 | Yes | Asp94Tyr | Lys325Stop, Thr378Ala, Thr519Ile | Glu662Gly | 64 | |
| 473 | No | Asp94Asn | Asp420Gly, Asn677Asp | Tyr495Cys | 256 | |
| 911 | Yes | Asp94Gly | Ala152Val, Tyr189His | 64 | ||
| 923 | Yes | Asp94Gly | Asp293Gly | Ser165Pro, Thr703Ala | 64 | |
| 354 | No | Met81Thr | Ala218Val | 256 | ||
| 378 | Yes | Leu109Pro | Thr267Ala, Lys793Arg, Thr798Ala | Phe343Leu | 256 | |
| 415 | No | Gln113Leu | Ala198Val, Lys471Glu | Leu393Pro | 64 | |
| 664 | Yes | Leu625Pro | Asn538Ile | Leu62Pro, Ser421Pro | 64 | |
| 31 | No | Ile302Thr, Arg495Gly | 256 | |||
| 357 | No | Asp639Gly | Leu610Pro | 16 | ||
| 494 | No | Thr272Ala | 64 | |||
| 584 | No | Leu548Pro | 256 | |||
| 1098 | No | Arg231His, Thr440Pro, Glu498Lys | 8 | |||
| 139 | No | His291Arg | 128 | |||
| 654 | Yes | 128 | ||||
Polymorphisms at codons 21, 95, and 668 of gyrA are not included. Polymorphisms in bold are previously reported resistance mutations. Isolates are ordered according to resistance mutations in the QRDR of gyrA and gyrB and then according to fluoroquinolone exposure prior to tuberculosis diagnosis. QRDR, quinolone resistance-determining region. gyrA QRDR, codons 74 to 113; gyrB QRDR, codons 500 to 538.
Table 4.
Polymorphisms in the 50 fluoroquinolone-susceptible M. tuberculosis isolatesa
| ID no. | FQ exposure prior to TB | Polymorphisms |
MIC (μg/ml) | |||
|---|---|---|---|---|---|---|
| In QRDR of gyrA | Outside QRDR of gyrA | In QRDR of gyrB | Outside QRDR of gyrB | |||
| 144 | Yes | Val77Ala, Arg98His | Met185Ile, Ile271Val, His595Arg | 1 | ||
| 190 | Yes | Met99Thr | Lys527Arg, Gly669Asp | Val262Ile | 2 | |
| 52 | Yes | Ser104Leu | Thr572Ala, Gly781Ser | 1 | ||
| 823 | No | Thr2Ile, Lys733Arg, Leu774Pro | Gly503Ser | Lys565Gln | 1 | |
| 537 | Yes | Met141Ile, Gln238Arg, Asp619Asn | Arg521His | 1 | ||
| 145 | No | Ile524Val | Asn91Ser, Glu235Gly | 1 | ||
| 29 | No | Arg292Stop | Phe212Leu | 1 | ||
| 55 | No | Thr222Ala, Thr585Met | 1 | |||
| 62 | No | Ile478Asn | Ser365Ala, Leu384Pro, Ile552Thr | 1 | ||
| 142 | No | Asn105Ser | 1 | |||
| 194 | No | Thr686Ile | Asn641Asp | 1 | ||
| 233 | No | Arg68His, Leu812Pro | Gln43Lys, Glu685Gly | 2 | ||
| 289 | No | Pro621Ser | Val423Ala | 2 | ||
| 320 | No | Lys707Glu | 1 | |||
| 402 | No | Gly247Ser, Asp515Gly, Arg575Gln | His30Tyr, Thr354Ala, Lys469Glu, Lys650Arg | 2 | ||
| 439 | No | <0.5 | ||||
| 446 | Yes | Gln22Arg, Thr285Pro | Tyr292Cys, Arg699Cys | 1 | ||
| 453 | No | Asp211Gly, Lys 295Glu | Ile579Val | 1 | ||
| 456 | No | Arg53His, Ser506Gly | Ile123Val, Glu498Val, Gly549Ser, Met655Val | 1 | ||
| 469 | No | Met185Thr, Phe202Ser, His280Tyr, Gln431Arg | Leu567Met, Ala683Asp | 2 | ||
| 522 | No | Ile362Ser | 1 | |||
| 636 | No | <0.5 | ||||
| 677 | No | 1 | ||||
| 686 | No | Gly117Asp, Gln545Stop, Val596Ala, Arg691Gln | Ser55Pro, Val135Ala, Ser165Leu | 1 | ||
| 696 | No | Asp3Asn | Val423Ala | 1 | ||
| 825 | No | Asn176Tyr | Asp554Asn | 1 | ||
| 829 | No | 1 | ||||
| 832 | No | Ala347Val, Arg789His | 1 | |||
| 834 | No | Gln475Stop | Asn308Asp, Ala309Val, Lys425Arg | 1 | ||
| 1046 | No | 1 | ||||
| 64 | Yes | Lys425Arg, His553Arg, Asn597Ser | 1 | |||
| 69 | Yes | Asn656Tyr | 2 | |||
| 141 | Yes | Leu331Pro | 1 | |||
| 243 | Yes | Val423Ala, Ala681Thr | 1 | |||
| 317 | Yes | Asp209Asn, Gly247Ser, Ile328Thr, Arg409Gln | 1 | |||
| 358 | Yes | Gln508Gly | Asn105Ser, Ala620Thr, Glu631Gly | 2 | ||
| 408 | Yes | Tyr617His | Val326Glu | 1 | ||
| 421 | Yes | Leu6Ser | Lys642Glu | 1 | ||
| 464 | Yes | Met33Thr, Ala601Thr | Leu465Ser | <0.5 | ||
| 540 | Yes | 1 | ||||
| 548 | Yes | Asp426Gly, Tyr775Cys | 2 | |||
| 667 | Yes | Asp67Gly, Arg710Gln | 1 | |||
| 693 | Yes | Leu407Pro, Ser666Pro | Ser273Asn, Thr664Ala | 1 | ||
| 741 | Yes | Gln161Arg, Gly247Ser | 2 | |||
| 822 | Yes | Tyr57His, Lys733Gln | Gly335Arg, Val450Gly | 2 | ||
| 824 | Yes | Asp13Asn, Arg320His | Phe359Ser | <0.5 | ||
| 826 | Yes | Gly744Ser, Val783Met | 1 | |||
| 827 | Yes | Gln594Stop | 2 | |||
| 830 | Yes | Pro136Leu, Val323Met | Arg614Cys | 1 | ||
| 1085 | Yes | Gly576Asp, Gly653Asp | <0.5 | |||
Polymorphisms at codons 21, 95, and 668 of gyrA are not included. There were no previously reported resistance mutations identified. Isolates are ordered according to resistance mutations in the QRDR of gyrA and gyrB and then according to fluoroquinolone exposure prior to tuberculosis diagnosis. QRDR, quinolone resistance-determining region. gyrA QRDR, codons 74 to 113; gyrB QRDR, codons 500 to 538.
There were three polymorphisms in the QRDR of gyrA in fluoroquinolone-resistant isolates that have not previously been associated with fluoroquinolone resistance (Met81Thr, Leu109Pro, and Gln113Leu). Of the eight remaining fluoroquinolone-resistant isolates, 7 had polymorphisms outside the QRDR (one had no polymorphisms) and one had a polymorphism at Asn538Ile in gyrB.
Among the 50 fluoroquinolone-susceptible M. tuberculosis isolates, 3 had polymorphisms in the QRDR of gyrA (all 3 had prior fluoroquinolone exposure) and 3 had polymorphisms in the QRDR of gyrB (1 had fluoroquinolone exposure) (Table 4).
Table 5 demonstrates the proportion of fluoroquinolone-resistant and -susceptible M. tuberculosis isolates with polymorphisms according to region of gyrA and gyrB. As expected, fluoroquinolone-resistant isolates were more likely to have mutations in the QRDR of gyrA (68% versus 6% in susceptible isolates). However, resistant isolates were not more likely to have mutations in the QRDR of gyrB (4% versus 6% in susceptible isolates). Interestingly, fluoroquinolone-susceptible isolates were significantly more likely to have polymorphisms in gyrA at codons 1 to 73 than fluoroquinolone-resistant isolates (20% versus 0%, respectively). Of the 10 isolates with polymorphisms in this region, each had a polymorphism at a different codon. The global test suggested that at least one region had a significant difference in mutation rate, thus assuring that the observed difference was not due to inflation of type I error by multiple comparisons. There were no other regions that had significantly more polymorphisms in fluoroquinolone-resistant versus -susceptible isolates.
Table 5.
Proportions of fluoroquinolone-resistant and -susceptible M. tuberculosis isolates with polymorphisms, listed according to location within gyrA and gyrBa
| Polymorphism location | No. (%) of M. tuberculosis isolates that were: |
P valueb | |
|---|---|---|---|
| Fluoroquinolone resistant (n = 25) | Fluoroquinolone susceptible (n = 50) | ||
| gyrA codons | |||
| 1–73 | 0 (0) | 10 (20) | 0.016 |
| 74–113 (QRDR) | 17 (68) | 3 (6) | <0.001 |
| 114–900 | 18 (72) | 35 (70) | 0.86 |
| gyrB codons | |||
| 1–499 | 15 (60) | 22 (44) | 0.19 |
| 500–538 (QRDR) | 1 (4) | 3 (6) | 0.72 |
| 539–800 | 8 (32) | 19 (38) | 0.61 |
Polymorphisms at codons 21, 95, and 668 of gyrA are not included.
Pearson chi-square test or Fisher's exact test.
DISCUSSION
Most (56%) of the fluoroquinolone-resistant M. tuberculosis isolates in this study had a previously described resistance mutation—at codon 90, 91, or 94 in the QRDR of gyrA. However, the relatively high proportion (44%) of isolates without one of these mutations has implications for the ability of current rapid genotypic diagnostic tests to detect such resistance. The GenoType Mycobacterium tuberculosis Drug Resistance second-line (MTBDRsl) DNA strip assay (Hain Lifescience, Nehren, Germany) is a rapid diagnostic test that utilizes mutations at codons 90, 91, and 94 and wild-type probes for codons 85 to 97 in gyrA to detect fluoroquinolone resistance in M. tuberculosis (16). Such a test would have had low sensitivity in our patient population. It should be noted that most of the patient populations studied to date with this diagnostic test have had multidrug-resistant tuberculosis (MDR-TB; defined as resistance to at least isoniazid and rifampin) (6, 16). In MDR-TB patients, the MTBDRsl assay has been 76 to 91% sensitive in detecting genotypic fluoroquinolone resistance (6, 16, 17, 19, 20). Many such patients likely received extensive prior treatment, including fluoroquinolones. In contrast, our patient population was comprised of newly diagnosed fluoroquinolone-monoresistant tuberculosis patients in whom the proportion with fluoroquinolone exposure was high—but presumably with shorter treatment courses for presumptive nontuberculous bacterial infections rather than known chronic tuberculosis (10, 11). Our patient population more closely reflects the population in whom fluoroquinolones might be used in the future as a first-line short-course antituberculosis therapy (9, 26).
It is important to determine the resistance mutation or mechanism in the 11 M. tuberculosis isolates without a previously described mutation. The mutations of highest priority for evaluation would be the three novel polymorphisms in the QRDR of gyrA (Met81Thr, Leu109Pro, and Gln113Leu). Of the 8 remaining fluoroquinolone-resistant isolates, 7 had polymorphisms outside the QRDR (one had no polymorphisms) and 1 had a polymorphism at codon 538 of gyrB. These novel polymorphisms merit evaluation as potentially novel resistance mutations. None of these polymorphisms occurred in more than one strain. If these polymorphisms do not confer fluoroquinolone resistance, alternative mechanisms such as efflux pumps and pentapeptide repeat proteins should be assessed (15, 25, 31).
In 22 of the 25 fluoroquinolone-resistant M. tuberculosis isolates, the ofloxacin MIC was ≥64 μg/ml (Table 3). Although fluoroquinolone-resistant isolates often have a MIC of 4 to 16 μg/ml, there have been reports of isolates with ofloxacin MICs of >32 μg/ml (3, 7, 22, 30). MIC depends in part on inoculum size, which should be standardized but is often not reported in studies of phenotypic and genotypic fluoroquinolone resistance in M. tuberculosis. However, the inoculum size used in this study was comparable to that used by other investigators (2, 13), including those reporting high ofloxacin MICs (3). The very high MICs in our study raise the possibility of additional resistance mechanisms besides gyrase mutations. This warrants further investigation.
There were 11 fluoroquinolone-susceptible M. tuberculosis isolates with an ofloxacin MIC of 2 μg/ml (Table 4). Although this still met our definition of susceptibility, it raises the possibility of borderline resistance. All 11 isolates had at least 1 gyrase polymorphism (1 had one in the QRDR of gyrA), but none of the polymorphisms have previously been demonstrated to confer resistance, and none were present in more than one isolate. Of the 11 patients, 7 had reported antecedent fluoroquinolone exposure.
Of note is that patients with fluoroquinolone-resistant M. tuberculosis isolates with previously described resistance mutations were significantly more likely to have documented fluoroquinolone exposure prior to tuberculosis diagnosis than persons with fluoroquinolone-resistant M. tuberculosis without such mutations (86% versus 27%; P = 0.003). Further evaluation is warranted, including an assessment as to whether greater duration of exposure is associated with an increased likelihood of mutations in codons 90, 91, and 94 of gyrA.
Of the 25 patients with fluoroquinolone-resistant M. tuberculosis, 10 did not have documented fluoroquinolone exposure prior to tuberculosis diagnosis. Some fluoroquinolone exposure could have been missed, such as inpatient exposure among patients from 2002 to 2006. Alternatively, these patients may have been initially infected with fluoroquinolone-resistant M. tuberculosis rather than acquiring the resistance via exposure after they had developed disease. However, this would seem unlikely since 22 resistant isolates were not clustered with other fluoroquinolone-resistant isolates, and the 3 that were clustered did not have epidemiologic links.
Among the 50 fluoroquinolone-susceptible M. tuberculosis isolates, 3 had polymorphisms in the QRDR of gyrA and 3 had polymorphisms in the QRDR of gyrB (Table 4). These polymorphisms have not previously been reported (22a), but polymorphisms have been reported in the QRDR of gyrA in fluoroquinolone-susceptible M. tuberculosis isolates in other studies—Pro102His and Leu109Val (18, 30). While it is unlikely that the polymorphisms identified in our study confer fluoroquinolone resistance given the low MIC of the organisms, they should be evaluated to see whether they predispose to resistance, perhaps in the presence of greater fluoroquinolone exposure. Polymorphisms outside of the QRDR could also potentially predispose to fluoroquinolone resistance, but any in vitro evaluation should prioritize the evaluation of polymorphisms within the QRDR, particularly in gyrA.
It was notable that fluoroquinolone-susceptible M. tuberculosis isolates were more likely to have polymorphisms in the region of gyrA just prior to the QRDR (i.e., in codons 1 to 73). There was not one particular codon location but rather 10 different codons (in 10 different isolates) within this region that contained polymorphisms. This suggests that it is this region of gyrA rather than a specific location within the region that is important. This should be evaluated in other populations of M. tuberculosis isolates and, if confirmed, investigated further as a region of gyrA in which polymorphisms perhaps confer fluoroquinolone susceptibility or protect against resistance. A previous report of mutations at codons 80 and 90 in gyrA that conferred fluoroquinolone hypersusceptibility suggested that such mutations could improve the flexibility of protein folding and enhance quinolone binding to the gyrase complex (3).
There were some limitations of this study. First, there were relatively few fluoroquinolone-resistant M. tuberculosis isolates. However, almost half of the resistant isolates did not have previously described resistance mutations, allowing the identification of several potentially novel resistance mutations. In addition, there were twice as many susceptible control isolates, and all isolates underwent sequencing of the entire gyrA and gyrB rather than just the QRDR. This also increased our ability to detect potentially novel mutations and polymorphisms. Second, there were no MDR-TB (and therefore, extremely drug-resistant tuberculosis [XDR-TB]) isolates in this study population. However, this may be considered a strength, since we evaluated isolates that were resistant only to fluoroquinolones. Resistance to other antituberculosis drugs could potentially influence resistance to fluoroquinolones (5, 14, 27). The study population also represents the broader population of newly diagnosed tuberculosis patients rather than the smaller MDR- and XDR-TB populations. Finally, inpatient fluoroquinolone exposure was not obtained from 2002 to 2006. However, four separate methods were used to ascertain exposure from 2007 to 2009.
In conclusion, there was a high proportion of fluoroquinolone-resistant M. tuberculosis isolates that did not have previously described resistance mutations. Once the novel mutations and/or alternative mechanisms that confer resistance in these isolates are characterized, this information will need to be incorporated into the development of rapid genotypic diagnostic tests that detect fluoroquinolone resistance in M. tuberculosis. More broadly, however, these results suggest that genotypic tests that rely primarily on detecting mutations in the QRDR of gyrA (and codons 90, 91, and 94 in particular) may have decreased sensitivity, particularly in newly diagnosed tuberculosis patients.
ACKNOWLEDGMENTS
We acknowledge Sue May for assistance with phenotypic resistance testing, Ed Mitchel for assistance with fluoroquinolone exposure data, and Howard Price for agar preparation.
This work was supported by the National Institutes of Health (K12 RR017697-05), including the National Institute of Allergy and Infectious Diseases (NIAID R01 AI 063200, NIAID K24 AI 65298).
We report no financial conflict of interest related to the subject of this paper.
Footnotes
Published ahead of print 21 December 2011
REFERENCES
- 1. Andrews JM. 2001. Determination of minimum inhibitory concentrations. J. Antimicrob. Chemother. 48(Suppl. 1):5–16 [DOI] [PubMed] [Google Scholar]
- 2. Aubry A, Pan XS, Fisher LM, Jarlier V, Cambau E. 2004. Mycobacterium tuberculosis DNA gyrase: interaction with quinolones and correlation with antimycobacterial drug activity. Antimicrob. Agents Chemother. 48:1281–1288 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Aubry A, et al. 2006. Novel gyrase mutations in quinolone-resistant and -hypersusceptible clinical isolates of Mycobacterium tuberculosis: functional analysis of mutant enzymes. Antimicrob. Agents Chemother. 50:104–112 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Blumberg HM, et al. 2003. American Thoracic Society/Centers for Disease Control and Prevention/Infectious Diseases Society of America: treatment of tuberculosis. Am. J. Respir. Crit. Care Med. 167:603–662 [DOI] [PubMed] [Google Scholar]
- 5. Borrell S, Gagneux S. 2011. Strain diversity, epistasis and the evolution of drug resistance in Mycobacterium tuberculosis. Clin. Microbiol. Infect. 17:815–820 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Brossier F, Veziris N, Aubry A, Jarlier V, Sougakoff W. 2010. Detection by GenoType MTBDRsl test of complex mechanisms of resistance to second-line drugs and ethambutol in multidrug-resistant Mycobacterium tuberculosis complex isolates. J. Clin. Microbiol. 48:1683–1689 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Cambau E, et al. 1994. Selection of a gyrA mutant of Mycobacterium tuberculosis resistant to fluoroquinolones during treatment with ofloxacin. J. Infect. Dis. 170:479–483 [DOI] [PubMed] [Google Scholar]
- 8. Cole ST, et al. 1998. Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393:537–544 [DOI] [PubMed] [Google Scholar]
- 9. Conde MB, et al. 2009. Moxifloxacin versus ethambutol in the initial treatment of tuberculosis: a double-blind, randomised, controlled phase II trial. Lancet 373:1183–1189 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Devasia RA, et al. 2009. Fluoroquinolone resistance in M. tuberculosis: the effect of duration and timing of fluoroquinolone exposure. Am. J. Respir. Crit. Care Med. 180:365–370 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Gaba PD, et al. 2007. Increasing outpatient fluoroquinolone exposure before tuberculosis diagnosis and impact on culture-negative disease. Arch. Intern. Med. 167:2317–2322 [DOI] [PubMed] [Google Scholar]
- 12. Ginsburg AS, Grosset JH, Bishai WR. 2003. Fluoroquinolones, tuberculosis, and resistance. Lancet Infect. Dis. 3:432–442 [DOI] [PubMed] [Google Scholar]
- 13. Guillemin I, Cambau E, Jarlier V. 1995. Sequences of conserved region in the A subunit of DNA gyrase from nine species of the genus Mycobacterium: phylogenetic analysis and implication for intrinsic susceptibility to quinolones. Antimicrob. Agents Chemother. 39:2145–2149 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Hazbon MH, et al. 2006. Population genetics study of isoniazid resistance mutations and evolution of multidrug-resistant Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 50:2640–2649 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Hegde SS, et al. 2005. A fluoroquinolone resistance protein from Mycobacterium tuberculosis that mimics DNA. Science 308:1480–1483 [DOI] [PubMed] [Google Scholar]
- 16. Hillemann D, Rusch-Gerdes S, Richter E. 2009. Feasibility of the GenoType MTBDRsl assay for fluoroquinolone, amikacin-capreomycin, and ethambutol resistance testing of Mycobacterium tuberculosis strains and clinical specimens. J. Clin. Microbiol. 47:1767–1772 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Huang WL, Chi TL, Wu MH, Jou R. 2011. Performance assessment of the GenoType MTBDRsl test and DNA sequencing for detection of second-line and ethambutol drug resistance among patients infected with multidrug-resistant Mycobacterium tuberculosis. J. Clin. Microbiol. 49:2502–2508 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Kam KM, et al. 2006. Stepwise decrease in moxifloxacin susceptibility amongst clinical isolates of multidrug-resistant Mycobacterium tuberculosis: correlation with ofloxacin susceptibility. Microb. Drug Resist. 12:7–11 [DOI] [PubMed] [Google Scholar]
- 19. Kiet VS, et al. 2010. Evaluation of the MTBDRsl test for detection of second line drug resistance in Mycobacterium tuberculosis. J. Clin. Microbiol. 48:2934–2939 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Kontsevaya I, et al. 2011. Evaluation of two molecular assays for rapid detection of mycobacterium tuberculosis resistance to fluoroquinolones in high-tuberculosis and -multidrug-resistance settings. J. Clin. Microbiol. 49:2832–2837 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Lau RW, et al. 2011. Molecular characterization of fluoroquinolone resistance in Mycobacterium tuberculosis: functional analysis of gyrA mutant at position 74. Antimicrob. Agents Chemother. 55:608–614 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Lee AS, Tang LL, Lim IH, Wong SY. 2002. Characterization of pyrazinamide and ofloxacin resistance among drug resistant Mycobacterium tuberculosis isolates from Singapore. Int. J. Infect. Dis. 6:48–51 [DOI] [PubMed] [Google Scholar]
- 22a. Maruri F, et al. A systematic review of gyrase mutations associated with fluoroquinolone-resistant Mycobacterium tuberculosis and a proposed gyrase numbering system. J. Antimicrob. Chemother., in press [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Narayanan Pand Tuberculosis Research Center 2002. Shortening short-course chemotherapy: a randomized clinical trial for treatment of smear-positive pulmonary tuberculosis with regimens using ofloxacin in the intensive phase. Indian J. Tuberc. 49:27–38 [Google Scholar]
- 24. NCCLS 2003. Susceptibility testing of mycobacteria, nocardiae, and other aerobic actinomycetes; approved standard. NCCLS document M24-A. NCCLS, Wayne, PA: [PubMed] [Google Scholar]
- 25. Pasca MR, et al. 2004. Rv2686c-Rv2687c-Rv2688c, an ABC fluoroquinolone efflux pump in Mycobacterium tuberculosis. Antimicrob. Agents Chemother. 48:3175–3178 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Rustomjee R, et al. 2008. A phase II study of the sterilising activities of ofloxacin, gatifloxacin and moxifloxacin in pulmonary tuberculosis. Int. J. Tuberc. Lung Dis. 12:128–138 [PubMed] [Google Scholar]
- 27. Safi H, Sayers B, Hazbon MH, Alland D. 2008. Transfer of embB codon 306 mutations into clinical Mycobacterium tuberculosis strains alters susceptibility to ethambutol, isoniazid, and rifampin. Antimicrob. Agents Chemother. 52:2027–2034 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Sandgren A, et al. 2009. Tuberculosis drug resistance mutation database. PLoS Med. 6:e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Sullivan EA, et al. 1995. Emergence of fluoroquinolone-resistant tuberculosis in New York City. Lancet 345:1148–1150 [DOI] [PubMed] [Google Scholar]
- 30. Sulochana S, Narayanan S, Paramasivan CN, Suganthi C, Narayanan PR. 2007. Analysis of fluoroquinolone resistance in clinical isolates of Mycobacterium tuberculosis from India. J. Chemother. 19:166–171 [DOI] [PubMed] [Google Scholar]
- 31. Takiff HE, et al. 1996. Efflux pump of the proton antiporter family confers low-level fluoroquinolone resistance in Mycobacterium smegmatis. Proc. Natl. Acad. Sci. U. S. A. 93:362–366 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Takiff HE, et al. 1994. Cloning and nucleotide sequence of Mycobacterium tuberculosis gyrA and gyrB genes and detection of quinolone resistance mutations. Antimicrob. Agents Chemother. 38:773–780 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Yew WW, Chan E, Chan CY, Cheng AF. 2002. Genotypic and phenotypic resistance of Mycobacterium tuberculosis to rifamycins and fluoroquinolones. Int. J. Tuberc. Lung Dis. 6:936–937 [PubMed] [Google Scholar]
- 34. Zhang Y, Yew WW. 2009. Mechanisms of drug resistance in Mycobacterium tuberculosis. Int. J. Tuberc. Lung Dis. 13:1320–1330 [PubMed] [Google Scholar]