Prediction of pyrazinamide resistance in Mycobacterium tuberculosis using structure-based machine-learning approaches.

Background: Pyrazinamide is one of four first-line antibiotics used to treat tuberculosis; however, antibiotic susceptibility testing for pyrazinamide is challenging. Resistance to pyrazinamide is primarily driven by genetic variation in pncA, encoding an enzyme that converts pyrazinamide into its active form. Methods: We curated a dataset of 664 non-redundant, missense amino acid mutations in PncA with associated high-confidence phenotypes from published studies and then trained three different machine-learning models to predict pyrazinamide resistance. All models had access to a range of protein structural-, chemical- and sequence-based features. Results: The best model, a gradient-boosted decision tree, achieved a sensitivity of 80.2% and a specificity of 76.9% on the hold-out test dataset. The clinical performance of the models was then estimated by predicting the binary pyrazinamide resistance phenotype of 4027 samples harbouring 367 unique missense mutations in pncA derived from 24 231 clinical isolates. Conclusions: This work demonstrates how machine learning can enhance the sensitivity/specificity of pyrazinamide resistance prediction in genetics-based clinical microbiology workflows, highlights novel mutations for future biochemical investigation, and is a proof of concept for using this approach in other drugs.

Using Reduced Inoculum Densities of Mycobacterium tuberculosis in MGIT Pyrazinamide Susceptibility Testing to Prevent False-Resistant Results and Improve Accuracy: A Multicenter Evaluation.

The primary platform used for pyrazinamide (PZA) susceptibility testing of Mycobacterium tuberculosis is the MGIT culture system (Becton Dickinson). Since false-resistant results have been associated with the use of this system, we conducted a multicenter evaluation to determine the effect of using a reduced cell density inoculum on the rate of false resistance. Two reduced inoculum densities were compared with that prescribed by the manufacturer (designated as "BD" method). The reduced inoculum methods (designated as "A" and "C") were identical to the manufacturer's protocol in all aspects with the exception of the cell density of the inoculum. Twenty genetically and phenotypically characterized M. tuberculosis isolates were tested in duplicate by ten independent laboratories using the three inoculum methods. False-resistant results declined from 21.1% using the standard "BD" method to 5.7% using the intermediate ("A") inoculum and further declined to 2.8% using the most dilute ("C") inoculum method. The percentages of the resistant results that were false-resistant declined from 55.2% for the "BD" test to 28.8% and 16.0% for the "A" and "C" tests, respectively. These results represent compelling evidence that the occurrence of false-resistant MGIT PZA susceptibility test results can be mitigated through the use of reduced inoculum densities.

Molecular detection of mutations associated with first- and second-line drug resistance compared with conventional drug susceptibility testing of Mycobacterium tuberculosis.

The emergence of multi- and extensively drug-resistant tuberculosis is a significant impediment to the control of this disease because treatment becomes more complex and costly. Reliable and timely drug susceptibility testing is critical to ensure that patients receive effective treatment and become noninfectious. Molecular methods can provide accurate and rapid drug susceptibility results. We used DNA sequencing to detect resistance to the first-line antituberculosis drugs isoniazid (INH), rifampin (RIF), pyrazinamide (PZA), and ethambutol (EMB) and the second-line drugs amikacin (AMK), capreomycin (CAP), kanamycin (KAN), ciprofloxacin (CIP), and ofloxacin (OFX). Nine loci were sequenced: rpoB (for resistance to RIF), katG and inhA (INH), pncA (PZA), embB (EMB), gyrA (CIP and OFX), and rrs, eis, and tlyA (KAN, AMK, and CAP). A total of 314 clinical Mycobacterium tuberculosis complex isolates representing a variety of antibiotic resistance patterns, genotypes, and geographical origins were analyzed. The molecular data were compared to the phenotypic data and the accuracy values were calculated. Sensitivity and specificity values for the first-line drug loci were 97.1% and 93.6% for rpoB, 85.4% and 100% for katG, 16.5% and 100% for inhA, 90.6% and 100% for katG and inhA together, 84.6% and 85.8% for pncA, and 78.6% and 93.1% for embB. The values for the second-line drugs were also calculated. The size and scope of this study, in numbers of loci and isolates examined, and the phenotypic diversity of those isolates support the use of DNA sequencing to detect drug resistance in the M. tuberculosis complex. Further, the results can be used to design diagnostic tests utilizing other mutation detection technologies.

Rapid detection of multidrug-resistant Mycobacterium tuberculosis by use of real-time PCR and high-resolution melt analysis.

The current study describes the development of a unique real-time PCR assay for the detection of mutations conferring drug resistance in Mycobacterium tuberculosis. The rifampicin resistance determinant region (RRDR) of rpoB and specific regions of katG and the inhA promoter were targeted for the detection of rifampin (RIF) and isoniazid (INH) resistance, respectively. Additionally, this assay was multiplexed to discriminate Mycobacterium tuberculosis complex (MTC) strains from nontuberculous Mycobacteria (NTM) strains by targeting the IS6110 insertion element. High-resolution melting (HRM) analysis following real-time PCR was used to identify M. tuberculosis strains containing mutations at the targeted loci, and locked nucleic acid (LNA) probes were used to enhance the detection of strains containing specific single-nucleotide polymorphism (SNP) transversion mutations. This method was used to screen 252 M. tuberculosis clinical isolates, including 154 RIF-resistant strains and 174 INH-resistant strains based on the agar proportion method of drug susceptibility testing (DST). Of the 154 RIF-resistant strains, 148 were also resistant to INH and therefore classified as multidrug resistant (MDR). The assay demonstrated sensitivity and specificity of 91% and 98%, respectively, for the detection of RIF resistance and 87% and 100% for the detection of INH resistance. Overall, this assay showed a sensitivity of 85% and a specificity of 98% for the detection of MDR strains. This method provides a rapid, robust, and inexpensive way to detect the dominant mutations known to confer MDR in M. tuberculosis strains and offers several advantages over current molecular and culture-based techniques.

Human tuberculosis due to Mycobacterium bovis in the United States, 1995-2005.

BACKGROUND: Understanding the epidemiology of human Mycobacterium bovis tuberculosis (TB) in the United States is imperative; this disease can be foodborne or airborne, and current US control strategies are focused on TB due to Mycobacterium tuberculosis and airborne transmission. The National TB Genotyping Service's work has allowed systematic identification of M. tuberculosis-complex isolates and enabled the first US-wide study of M. bovis TB. METHODS: Results of spacer oligonucleotide and mycobacterial interspersed repetitive units typing were linked to corresponding national surveillance data for TB cases reported for the period 2004-2005 and select cases for the period 1995-2003. We also used National TB Genotyping Service data to evaluate the traditional antituberculous drug resistance-based case definition of M. bovis TB. RESULTS: Isolates from 165 (1.4%) of 11,860 linked cases were identified as M. bovis. Patients who were not born in the United States, Hispanic patients, patients <15 years of age, patients reported to be HIV infected, and patients with extrapulmonary disease each had increased adjusted odds of having M. bovis versus M. tuberculosis TB. Most US-born, Hispanic patients with TB due to M. bovis (29 [90.6%] of 32) had extrapulmonary disease, and their overall median age was 9.5 years. The National TB Genotyping Service's data indicated that the pyrazinamide-based case definition's sensitivity was 82.5% (95% confidence interval; 75.3%-87.9%) and that data identified 14 errors in pyrazinamide-susceptibility testing or reporting. CONCLUSIONS: The prevalence of extrapulmonary disease in the young, US-born Hispanic population suggests recent transmission of M. bovis, possibly related to foodborne exposure. Because of its significantly different epidemiologic profile, compared with that of M. tuberculosis TB, we recommend routine surveillance of M. bovis TB. Routine surveillance and an improved understanding of M. bovis TB transmission dynamics would help direct the development of additional control measures.

Multiphasic approach reveals genetic diversity of environmental and patient isolates of Mycobacterium mucogenicum and Mycobacterium phocaicum associated with an outbreak of bacteremias at a Texas hospital.

Between March and May 2006, a Texas hospital identified five Mycobacterium mucogenicum bloodstream infections among hospitalized oncology patients using fluorescence high-performance liquid chromatography analysis of mycolic acids. Isolates from blood cultures were compared to 16 isolates from environmental sites or water associated with this ward. These isolates were further characterized by hsp65, 16S rRNA, and rpoB gene sequencing, hsp65 PCR restriction analysis, and molecular typing methods, including repetitive element PCR, random amplified polymorphic DNA PCR, and pulsed-field gel electrophoresis (PFGE) of large restriction fragments. Three of five patient isolates were confirmed as M. mucogenicum and were in a single cluster as determined by all identification and typing methods. The remaining two patient isolates were identified as different strains of Mycobacterium phocaicum by rpoB sequence analysis. One of these matched an environmental isolate from a swab of a hand shower in the patient's room, while none of the clinical isolates of M. mucogenicum matched environmental strains. Among the other 15 environmental isolates, 11 were identified as M. mucogenicum and 4 as M. phocaicum strains, all of which were unrelated by typing methods. Although the 16S rRNA gene sequences matched for all 14 M. mucogenicum isolates, there were two each of the hsp65 and rpoB sequevars, seven PCR typing patterns, and 12 PFGE patterns. Among the seven M. phocaicum isolates were three 16S rRNA sequevars, two hsp65 sequevars, two rpoB sequevars, six PCR typing patterns, and six PFGE patterns. This outbreak represents the first case of catheter-associated bacteremia caused by M. phocaicum and the first report of clinical isolates from a U.S. hospital. The investigation highlights important differences in the available typing methods for mycobacteria and demonstrates the genetic diversity of these organisms even within narrow confines of time and space.

Mycobacterium cosmeticum sp. nov., a novel rapidly growing species isolated from a cosmetic infection and from a nail salon.

Four isolates of a rapidly growing Mycobacterium species had a mycolic acid pattern similar to that of Mycobacterium smegmatis, as determined by HPLC analyses. Three of the isolates were from footbath drains and a sink at a nail salon located in Atlanta, GA, USA; the fourth was obtained from a granulomatous subdermal lesion of a female patient in Venezuela who was undergoing mesotherapy. By random amplified polymorphic DNA electrophoresis and PFGE of large restriction fragments, the three isolates from the nail salon were shown to be the same strain but different from the strain from the patient in Venezuela. Polymorphisms in regions of the rpoB, hsp65 and 16S rRNA genes that were shown to be useful for species identification matched for the two strains but were different from those of other Mycobacterium species. The 16S rRNA gene sequence placed the strains in a taxonomic group along with Mycobacterium frederiksbergense, Mycobacterium hodleri, Mycobacterium diernhoferi and Mycobacterium neoaurum. The strains produced a pale-yellow pigment when grown in the dark at the optimal temperature of 35 degrees C. Biochemical testing showed that the strains were positive for iron uptake, nitrate reduction and utilization of d-mannitol, d-xylose, iso-myo-inositol, l-arabinose, citrate and d-trehalose. The strains were negative for d-sorbitol utilization and production of niacin and 3-day arylsulfatase, although arylsulfatase activity was observed after 14 days. The isolates grew on MacConkey agar without crystal violet but not on media containing 5 % (w/v) NaCl or at 45 degrees C. They were susceptible to ciprofloxacin, amikacin, tobramycin, cefoxitin, clarithromycin, doxycycline, sulfamethoxazole and imipenem. The name Mycobacterium cosmeticum sp. nov. is proposed for this novel species; two strains, LTA-388(T) (=ATCC BAA-878(T)=CIP 108170(T)) (the type strain) and 2003-11-06 (=ATCC BAA-879=CIP 108169) have been designated, respectively, for the strains of the patient in Venezuela and from the nail salon in Atlanta, GA, USA.

Screening and characterization of mutations in isoniazid-resistant Mycobacterium tuberculosis isolates obtained in Brazil.

We investigated mutations in the genes katG, inhA (regulatory and structural regions), and kasA and the oxyR-ahpC intergenic region of 97 isoniazid (INH)-resistant and 60 INH-susceptible Mycobacterium tuberculosis isolates obtained in two states in Brazil: São Paulo and Paraná. PCR-single-strand conformational polymorphism (PCR-SSCP) was evaluated for screening mutations in regions of prevalence, including codons 315 and 463 of katG, the regulatory region and codons 16 and 94 of inhA, kasA, and the oxyR-ahpC intergenic region. DNA sequencing of PCR amplicons was performed for all isolates with altered PCR-SSCP profiles. Mutations in katG were found in 83 (85.6%) of the 97 INH-resistant isolates, including mutations in codon 315 that occurred in 60 (61.9%) of the INH-resistant isolates and 23 previously unreported katG mutations. Mutations in the inhA promoter region occurred in 25 (25.8%) of the INH-resistant isolates; 6.2% of the isolates had inhA structural gene mutations, and 10.3% had mutations in the oxyR-ahpC intergenic region (one, nucleotide -48, previously unreported). Polymorphisms in the kasA gene occurred in both INH-resistant and INH-susceptible isolates. The most frequent polymorphism encoded a G(269)A substitution. Although KatG(315) substitutions are predominant, novel mutations also appear to be responsible for INH resistance in the two states in Brazil. Since ca. 90.7% of the INH-resistant isolates had mutations identified by SSCP electrophoresis, this method may be a useful genotypic screen for INH resistance.

Temperature-mediated heteroduplex analysis for detection of pncA mutations associated with pyrazinamide resistance and differentiation between Mycobacterium tuberculosis and Mycobacterium bovis by denaturing high- performance liquid chromatography.

The goal of this study was to apply temperature-mediated heteroduplex analysis using denaturing high-performance liquid chromatography to identify pyrazinamide (PZA) resistance in Mycobacterium tuberculosis isolates and simultaneously differentiate between M. tuberculosis and Mycobacterium bovis. Features that contributed to an optimal assay included the use of two different reference probes for the pncA gene targets from wild-type M. tuberculosis and wild-type M. bovis, optimization of the column temperature, increasing the starting concentration of the elution buffer, and reducing the rate of elution buffer increase (slope). A total of 69 strains were studied, including 48 wild-type M. tuberculosis strains (13 were PZA-resistant strains) and 21 M. bovis strains (8 were BCG strains). In all isolates tested, wild-type M. tuberculosis generated a single-peak pattern when mixed with the M. tuberculosis probe and a double-peak pattern with the M. bovis probe. In contrast, all M. bovis isolates generated a double-peak pattern when mixed with the M. tuberculosis probe and a single-peak pattern with the M. bovis probe. PZA-resistant mutant M. tuberculosis isolates generated characteristic patterns that were easily distinguishable from both wild-type M. tuberculosis and M. bovis isolates. Chromatographic patterns generated by the two reference probes allowed the rapid detection of PZA resistance with the simultaneous ability to distinguish between M. tuberculosis and M. bovis. This approach may allow the detection of drug resistance-associated mutations, with potential application to clinical and epidemiological aspects of tuberculosis control.

ethA, inhA, and katG loci of ethionamide-resistant clinical Mycobacterium tuberculosis isolates.

Ethionamide (ETH) is a structural analog of the antituberculosis drug isoniazid (INH). Both of these drugs target InhA, an enzyme involved in mycolic acid biosynthesis. INH requires catalase-peroxidase (KatG) activation, and mutations in katG are a major INH resistance mechanism. Recently an enzyme (EthA) capable of activating ETH has been identified. We sequenced the entire ethA structural gene of 41 ETH-resistant Mycobacterium tuberculosis isolates. We also sequenced two regions of inhA and all or part of katG. The MICs of ETH and INH were determined in order to associate the mutations identified with a resistance phenotype. Fifteen isolates were found to possess ethA mutations, for all of which the ETH MICs were > or =50 microg/ml. The ethA mutations were all different, previously unreported, and distributed throughout the gene. In eight of the isolates, a missense mutation in the inhA structural gene occurred. The ETH MICs for seven of the InhA mutants were > or =100 microg/ml, and these isolates were also resistant to > or =8 microg of INH per ml. Only a single point mutation in the inhA promoter was identified in 14 isolates. A katG mutation occurred in 15 isolates, for which the INH MICs for all but 1 were > or =32 microg/ml. As expected, we found no association between katG mutation and the level of ETH resistance. Mutations within the ethA and inhA structural genes were associated with relatively high levels of ETH resistance. Approximately 76% of isolates resistant to > or =50 microg of ETH per ml had such mutations.

Identifying Mycobacterium species and strain typing using a microfluidic labchip instrument.

We developed schemes for rapid identification of Mycobacterium species and strain typing using a microfluidic labchip instrument. A 439-bp region of the gene that codes for the 65-kDa heat shock protein (hsp65), which has sequence polymorphisms specific for most mycobacterial species, was examined using PCR-restriction analysis (PRA). We performed PRA in duplicate, using 2 strains each of 12 species, and observed that fragment sizes (bp) determined automatically by the instrument were consistently smaller than the correct sizes for each of the species as determined by sequence analysis (mean variance, < 7 bp). Mycobacterium tuberculosis isolates were typed with the labchip instrument using mycobacterial interspersed repetitive unit-variable number tandem repeat (MIRU-VNTR) typing, which determines the number of copies of repeated units at 12 loci in the genome based on product size after PCR amplification. Seven strains with one to six repeat copies at each locus were examined. Sizes were smaller by a mean of 13.47 bp compared with correct sizes predicted by sequence analysis, but could be used to correctly identify all strains types. Isolates of Mycobacterium chelonae and Mycobacterium abscessus were typed using randomly amplified polymorphic DNA (RAPD) electrophoresis, and patterns obtained using the labchip instrument were compared with multilocus enzyme electrophoresis (MEE) types. Patterns were distinct and reproducible for all strains except those with closely related MEE types. The labchip instrument is a versatile alternative for sizing mycobacterial DNA fragments.

Mutations in katG, inhA, and ahpC genes of Brazilian isoniazid-resistant isolates of Mycobacterium tuberculosis.

The presence of mutations in specific regions of the katG, inhA, and ahpC genes was analyzed with 69 Mycobacterium tuberculosis isoniazid-resistant isolates from three Brazilian states. Point mutations in codon 315 of the katG gene were observed in 87.1, 60.9, and 60% of the isolates from Rio Grande do Sul, Rio de Janeiro, and São Paulo, respectively. Mutations in the inhA gene were identified only in one isolate from RJ State, and the ahpC promoter region revealed mutations in distinct positions in 12.9, 21.7, and 6.7% of the isolates from RS, RJ and SP, respectively.

Isolation and identification of mycobacteria from livestock specimens and milk obtained in Brazil.

The prevalence of Mycobacterium bovis and other mycobacterial species in livestock specimens and milk was evaluated. An emphasis was placed upon the distribution of these organisms in milk that is readily available to the public that was either untreated, pasteurized, or treated using ultra high temperature. Twenty-two pathologic specimens from livestock (bovine, swine and bubaline) in five Brazilian states and 128 bovine milk samples from retail markets in the State of S o Paulo were examined for mycobacteria. Identification was made by classical biochemical tests, thin layer chromatography of mycolic acids and polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis. Mycobacteria were isolated from 15 (68.2%) caseous lesions and from 23 (18%) milk samples. Eleven isolates were identified as M. bovis, and the remaining 27 nontuberculous mycobacterial isolates were represented by five species and six unidentified rapidly growing mycobacterial strains. The data demonstrate that animal products in Brazil are frequent reservoirs of mycobacteria and may pose a risk to the public.

Isolation and identification of mycobacteria from livestock specimens and milk obtained in Brazil

The prevalence of Mycobacterium bovis and other mycobacterial species in livestock specimens and milk was evaluated. An emphasis was placed upon the distribution of these organisms in milk that is readily available to the public that was either untreated, pasteurized, or treated using ultra high temperature. Twenty-two pathologic specimens from livestock (bovine, swine and bubaline) in five Brazilian states and 128 bovine milk samples from retail markets in the State of São Paulo were examined for mycobacteria. Identification was made by classical biochemical tests, thin layer chromatography of mycolic acids and polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis. Mycobacteria were isolated from 15 (68.2 percent) caseous lesions and from 23 (18 percent) milk samples. Eleven isolates were identified as M. bovis, and the remaining 27 nontuberculous mycobacterial isolates were represented by five species and six unidentified rapidly growing mycobacterial strains. The data demonstrate that animal products in Brazil are frequent reservoirs of mycobacteria and may pose a risk to the public

Temperature-mediated heteroduplex analysis performed by using denaturing high-performance liquid chromatography to identify sequence polymorphisms in Mycobacterium tuberculosis complex organisms.

PCR products containing sequence polymorphisms were prepared from six mycobacterial genes, denatured, mixed with reference PCR products, and reannealed; the mixtures were then examined with a denaturing high-performance liquid chromatography system (WAVE) equipped with a temperature-controlled alkalated polystyrene divinyl benzene column. Mismatching of bases in heteroduplexes of the PCR products causes elution patterns of the DNA from the column to be altered. The six mycobacterial genes studied were oxyR, in which a specific polymorphism (G(1031)A) is found only in certain species of the Mycobacterium tuberculosis complex, and five genes in which mutations associated with antituberculosis drug resistance have been found. The resistance genes (with affected drug and PCR product sizes given parenthetically) were rpoB (rifampin; 258 bp), katG (isoniazid; 205 bp), pncA (pyrazinamide; 579 bp); rpsL (streptomycin; 196 bp), and embB (ethambutol; 185 bp). Elution patterns of heteroduplexes of all 20 polymorphisms studied shifted detectably at column temperatures ranging from 65.3 to 68 degrees C and elution times of 3.5 to 6 min. These results show that temperature-mediated heteroduplex analysis is a potentially useful genotypic screen for mutations associated with antituberculosis drug resistance and for the G(1031)A polymorphism in oxyR. The method may allow users to detect novel as well as heterogeneous mutations without using expensive kits or detection labels.