|Home | About | Journals | Submit | Contact Us | Français|
Rapid and accurate identification of mycobacterial species is essential for patient management. We describe the use of the Invader assay in conjunction with the BACTEC MGIT 960 system that together provide an efficient procedure for clinical use. This assay discriminates single-base differences (e.g., genotyping single-nucleotide polymorphisms) under homogeneous and isothermal conditions and can measure directly on genomic DNA without prior target DNA amplification. To identify a wide variety of mycobacterial species, 20 Invader probes were designed to target the 16S rRNA gene and the 16S-23S rRNA gene internal transcribed spacer 1 (ITS-1) region. To validate the Invader probes, we used 78 ATCC strains, and 607 clinical mycobacterial strains, which were identified by DNA sequencing of the 16S rRNA gene and ITS-1. The Invader assay could accurately identify and differentiate these strains according to target sequences. Moreover, it could detect and identify 116 (95.1%) of 122 positive liquid cultures from the BACTEC MGIT 960 system and did not react to 83 contaminated MGIT cultures. Species identification takes 6.5 h by the Invader assay: 2.0 h for DNA extraction, 0.5 h for handling, and up to 4 h for the Invader reaction. The Invader assay has the speed, ease of use, and accuracy to be an effective procedure for the bacteriological diagnosis of mycobacterial infections.
Rapid and accurate identification of mycobacteria is essential for determining appropriate therapies and for epidemiological studies (7, 11, 14, 37). For example, the U.S. Centers for Disease Control and Prevention (CDC) recommends turnaround times of 2 to 3 weeks for processing of Mycobacterium tuberculosis (4, 33, 36). A definitive diagnosis of mycobacterial infection depends on growth and identification of the bacteria (2, 3). To speed the bacterial culturing, time-consuming cultures on egg-based solid media, such as Löwenstein-Jensen and Ogawa slants, are being replaced by faster liquid culture methods, such as the BACTEC MGIT 960 system (Becton-Dickinson, Sparks, MD) and the MB/BacT system (Organon Teknika, Boxtel, The Netherlands) (1, 13, 15, 21, 22, 33).
Now a method is urgently needed to rapidly identify a wide variety of Mycobacterium species directly from liquid cultures. Unfortunately, the available methods have several limitations. For example, the widely used AccuProbe system (GenProbe, San Diego, CA) (8, 21, 25, 29) identifies only a limited number of the many mycobacterial species seen in a clinical laboratory. Although much faster and more accurate, DNA sequencing (16, 26, 39), PCR restriction fragment length polymorphism assays (6, 28, 35) and InnoLiPA Mycobacteria (Innogenetics, Ghent, Belgium) (23, 38) require expensive equipment and an exclusive workspace for PCR. In addition, DNA sequencing and PCR-restriction fragment length polymorphism assay require pure cultures, and if the sample was the mixed culture, separate culture on a solid medium would further slow these assays.
Many routine procedures in clinical laboratories have been simplified by homogeneous fluorescent detection systems. Here we report our study of one of these, the Invader assay (Third Wave Technologies, Madison, WI) (17). The Invader assay can accurately discriminate single-base differences, such as single nucleotide polymorphisms, and can measure directly on genomic DNA without prior target amplification. We evaluated the suitability of the Invader assay for directly identifying mycobacteria from the MGIT 960 system.
The validation of the Invader probes was performed by testing 78 ATCC strains (Table (Table2)2) and 607 clinical mycobacterial strains (Table (Table3),3), which were identified and confirmed the target sequences by DNA sequencing of the 16S rRNA gene and the 16S-23S rRNA gene internal transcribed spacer 1 (ITS-1). These clinical strains cultured on Ogawa slants mainly for nontuberculous Mycobacterium infections were tested in the BML general laboratory from December 2003 to June 2005 (excluding overlapping patients).
Between 21 and 25 November 2004, 1,390 consecutive clinical specimens were received for routine mycobacterial detection in the BML general laboratory. These included 1,040 sputum and 155 other respiratory specimens, 28 digestive samples, 15 sterile body fluids, 14 urine samples, 6 wound samples, and 132 samples from unspecified sources.
Before inoculation, the MGIT 960 tubes were prepared as described by the manufacturer (Becton Dickinson). A 0.5-ml portion of the processed specimen was inoculated into the MGIT, and the tubes were introduced into the MGIT 960 instrument and incubated until they were found to be positive by the instrument or for 42 days. A 0.1-ml portion of the processed specimen was inoculated onto a solid slant of Ogawa egg medium (Kyokuto Pharmaceutical, Tokyo, Japan), and the slants were incubated in 5 to 10% CO2. For 56 days, the growth on the slants was examined for visible colonies. All positive media were examined by Ziehl-Neelsen and Gram staining to confirm the presence of only acid-fast bacteria, and the colonies were subcultured onto Trypticase soy agar II with 5% sheep blood (TSA II; Becton Dickinson) to check for contaminants.
For Ogawa slants, a sample of DNA was extracted from a loopful (3-mm3 sphere) of bacterial colony. Bacterial cells were mechanically disrupted with glass beads. After phenol-chloroform treatment (DDH Mycobacteria; Kyokuto Pharmaceuticals), DNA in the aqueous phase was extracted and purified on a robotic liquid handler AGE-96 (Biotec, Tokyo, Japan) with magnetic silica particles (MagneSil blood genomic max yield system; Promega, Madison, WI). For the MGIT 960 system, a 4.0-ml aliquot of culture broth was centrifuged for 10 min at 13,000 × g. The pellet was extracted with the bacterial DNA/RNA extraction kit (MORA-EXTRACT; Kyokuto Pharmaceuticals). Bacterial cells were mechanically disrupted with zirconia beads, and phenol-chloroform treatment was performed, as recommended by the manufacturer. A 100-μl aliquot of TE buffer (10 mM Tris-HCl [pH 8.0], 0.1 mM EDTA) was added to the extracted DNA pellet, and DNA concentrations were determined by using the PicoGreen system (Molecular Probes, Eugene, OR) as recommended by the manufacturer.
For an MGIT-positive culture, an aliquot of each culture was tested in parallel by the Amplicor PCR (Roche Diagnostic Systems) for M. tuberculosis complex, M. avium, and M. intracellulare as described in the manufacturer's instructions.
The 25-μl reaction mixture contained ExTaq HS buffer (Takara Shuzo, Ohtsu, Japan) with 2.0 mM MgCl2, 200 μM concentrations of each of the deoxynucleoside triphosphates, 1.0 U of ExTaq HS DNA polymerase, 10 ng of template, 10 pmol of each of the primers SSU-bact-27f (5′-AGA GTT TGA TCM TGG CTC AG-3′) and SSU-bact-907r (5′-CCG TCA ATT CMT TTR AGT TT-3′) for the 16S rRNA gene and 16S-1511f (5′-AAG TCG TAA CAA GGT ARC CG-3′) and 23S-23r (5′-TCG CCA AGG CAT CCA CC-3′) for the ITS-1 region (18). Amplification was performed with a GeneAmp PCR system 9700 thermocycler (Applied Biosystems, Foster City, CA) for 30 cycles (30 s at 94°C, 30 s at 53°C, and 90 s at 72°C), followed by an extension step at 72°C for 7 min. The PCR products were visualized with ethidium bromide staining and UV illumination. Purification of the amplicons was performed with the AMPure PCR purification system (Agencourt, Beverly, MA), according to the manufacturer's instructions. The ABI Prism BigDye Terminator v1.1 cycle sequencing ready reaction kit (Applied Biosystems) was used for the sequencing of the PCR products. The sequencing reaction mixture contained 0.5 μl of BigDye premix, 1.75 μl of 5× sequencing buffer, 1.6 pmol of sequencing primer, and approximately 10 ng of PCR product template in a total volume of 10 μl. SSU-bact-27f, SSU-bact-907r, 16S-1511f, and 23S-23r were used for sequencing primers for both DNA strands. The sequencing reaction was performed with a GeneAmp PCR system 9700 thermocycler (Applied Biosystems). A denaturation step at 95°C for 2 min proceeded 25 cycles (10 s at 96°C, 15 s at 53°C, and 2.5 min at 60°C). Sequencing products were purified with a CleanSEQ Sequencing Reaction Clean-Up system (Agencourt) and analyzed with a 3130xl genetic analyzer (Applied Biosystems), according to the manufacturer's instructions. Raw sequencing data were edited to resolve discrepancies by evaluating the electrophoretograms with sequencing analysis software (v3.3; Applied Biosystems). The edited sequence data from both strands were aligned with the DNASIS Pro (Hitachi Software Engineering, Yokohama, Japan). We analyzed the consensus sequence of approximately 500 bp of the 5′ end of the 16S rRNA gene for comparison with the sequence databases stored in GenBank (BLAST; http://www.ncbi.nlm.nih.gov/BLAST/) or the Ribosomal Differentiation of Medical Microorganisms (RIDOM; http://www.ridom-rdna.de/) (9, 10).
The 16S rRNA gene and ITS-1 sequences were aligned with those from the GenBank database and quality-controlled RIDOM database. With Invader Creator software (Third Wave Technologies), 20 specific probes were designed to identify and differentiate mycobacteria in the conserved regions of ITS-1 and 16S rRNA gene hypervariable regions A and B, including species-specific sites. In addition, mixed genus Mycobacteria (GNS) probes for the mycobacteria genus-specific region and broad-range bacterial (BRB) probe for the bacteria universal region were designed from conserved regions of 16S rRNA gene (Table (Table1)1) . Signal probes and Invader oligonucleotide for the specific nucleotide sequences were designed to have theoretical annealing temperatures of 63 and 77°C, respectively, using a nearest-neighbor algorithm on the basis of final probe and target concentrations. The signal probes and Invader oligonucleotides used to detect Mycobacterium species by the Invader assay are shown in Table Table11.
The Invader assay utilizes the thermostable flap endonuclease Cleavase XI, which cleaves invasive structures formed from single-base overlap between the Invader oligonucleotide and the signal probe when hybridized to a complementary target DNA. This method can accurately discriminate single-base differences, such as single-nucleotide polymorphisms, and can measure genomic DNA (≥104 copies/assay). The Invader assay combines structure-specific cleavage enzymes and a universal FRET (Förster resonance energy transfer) system, and the biplex format of the Invader assay enables simultaneous detection of two kinds of species in a single well. At first, 3 μl of genomic DNA (0.03 to 0.33 ng/μl) was added to a 384-well plate, 6 μl of mineral oil (Sigma) was overlaid into all reaction wells, the mixtures were denatured by incubation at 95°C for 10 min, and 3 μl of the appropriate reaction mixture was added. The reaction mixture contained 32 ng of Cleavase XI enzyme, 0.817 μmol of each signal probe/liter, 0.163 μmol of each Invader oligonucleotide/liter, 0.65 μmol of each of FAM dye and Redmond Red dye FRET cassettes (Epoch Bioscience, Redmond, WA)/liter, 5.04% PEG 8000, 30.7 mmol of MgCl2/liter, and 24.5 mmol of morpholinepropanesulfonic acid/liter. After the reagent was dispensed, the plate was spun for 10 s at 400 × g and then sequentially incubated isothermally at 64°C up to 4 h in the thermal fluorescence microtiter plate reader (Fluodia; Otsuka Electronics, Osaka, Japan), whereas the fluorescence intensities were measured at 15-min intervals for FAM (excitation, 486 nm; emission, 530 nm) and Redmond Red (RED) (excitation, 560 nm; emission, 620 nm).
Raw data were analyzed by using a Microsoft Excel-based spreadsheet (Microsoft, Redmond, WA). For each specific signal, fold-over-zero (FOZ) values were calculated as follows for the signal obtained with each dye: FOZ = raw counts from sample/raw counts from no target control.
DNA samples from 54 mycobacterial reference strains were used to determine the cutoff value of FOZ. The mean plus five standard deviations of the FOZ value of the nontarget sequence in each well was calculated to 1.22 to 1.95 (probe set 1 to 10) of FAM-FOZ and 1.23 to 1.70 (probe set 1 to 10) of RED-FOZ. Therefore, the cutoff level was set at 2.00.
The most widely accepted gene for bacterial identification is the 16S rRNA gene, but this gene alone lacks sufficient resolution to identify all mycobacterial species (5, 32, 37, 39). For example, the 16S rRNA gene could not distinguish between M. kansasii and M. gastri, and also between M. chelonae and M. abscessus. However, since the ITS-1 region had greater diversity than the 16S rRNA gene (27, 18), the Invader probes designed in the ITS-1 region could distinguish these strains. Therefore, the Invader assay was set up with a more effective combination of the 16S rRNA gene and the ITS-1 region.
To validate the Invader probes, we evaluated 65 type and 13 reference strains and 607 mycobacterial clinical strains. Specific signals were obtained for the target sequences of the type and reference strains, according to the criteria of the Invader assay. The GNS and the BRB probes were designed for the mycobacterial genus-specific region and the bacteria universal regions, respectively. Using the Invader assay, strains that reacted only to both the BRB and GNS probes were mycobacteria other than target species, and strains that reacted to only the BRB probe were bacteria other than mycobacteria (Table (Table2).2). Next, we confirmed the variation and conservation of target sequences in 607 clinical mycobacterial strains, which were identified by DNA sequencing of the 16S rRNA gene and ITS-1. For all 607 clinical strains, the identities determined by DNA sequencing corresponded exactly to those determined by the Invader assay (Table (Table33).
By examining a serial dilution of reference strains, the detection limit of genomic DNAs for all Invader probes was found to be 0.03 ng/μl. DNA extracted from 122 positive samples of the MGIT 960 system had a minimum concentration of <0.01 ng/μl (less than the lower limit of the PicoGreen system), a maximum value 24.38 ng/μl, and an average of 4.40 ng/μl. Zirconia beads efficiently extracted DNA for M. tuberculosis and NTM. Six MGIT samples were negative in the Invader assay, and the DNA concentrations of four MGIT samples were below the detection limit (<0.01 ng/μl). Although the remaining two MGIT samples had sufficient DNA for the Invader assay (from ≥1 ng/μl to <10 ng/μl), smears and cultures of these samples showed contamination with S. aureus (≥107 CFU/ml). Also, the Invader probes of these two samples reacted only with the BRB probe. Overall, when more than 0.10 ng of mycobacterial genomic DNA/μl was obtained from the MGIT, the Invader assay detected individual species in a simultaneous assay (Table (Table44).
After MGIT cultivation of the 1,390 samples, 122 (8.8%) samples were found to be positive for mycobacteria. The breakdown of 122 MGIT-positive samples was 63 (51.6%) M. tuberculosis, 29 (23.8%) M. avium, 17 (13.9%) M. intracellulare, 4 (3.3%) M. kansasii, 2 (1.6%) M. abscessus, 2 (1.6%) M. chelonae, 2 (1.6%) M. fortuitum, 1 (0.8%) M. gordonae, and 6 (4.9%) negative isolates. Two (1.6%) samples had both M. avium and M. abscessus. Although six samples were negative as determined by the Invader assay, they were positive by the Amplicor M. tuberculosis complex PCR. The four negative samples except for two negative samples contaminated with S. aureus changed to positive for M. tuberculosis complex by the Invader assay after additional cultivation of the remaining MGIT broth for 1 week (Table (Table4).4). In addition, 83 (6.0%) MGIT cultures showed contamination of bacteria other than mycobacteria (Table (Table5).5). In the Invader reactions, these samples were positive only for the BRB probe.
A rapid and accurate method for diagnosing mycobacterial infection is urgently needed in the clinical laboratory. This study describes the combined use of the BACTEC MGIT 960 system and the Invader assay. The Invader assay correctly identified and differentiated total 888 clinical and reference strains according to a target sequence. In particular, it detected and identified 116 (95.1%) of 122 positive liquid cultures soon after automatic positive signal of the MGIT 960 system. Thus, the Invader assay has the accuracy and sensitivity to be an effective procedure for a definitive diagnosis of mycobacterial infection in a clinical setting.
The key to any assay based on DNA sequences is the probe. The target region for DNA probes must be sufficiently conserved among clinical strains of the species, and the sequence data should be reliable, definitive, and commonly used. To identify mycobacteria, probes have traditionally been based on sequences from the 16S rRNA gene that corresponds to Escherichia coli positions 130 to 210 and positions 430 to 500. However, the 16S rRNA genes of some species have the same or very similar sequences (5, 32, 37, 39). For example, using the first 500 bases of the 5′ end of the 16S rRNA gene to analyze sequences, RIDOM or MicroSeq (Applied Biosystems) (24) cannot differentiate M. abscessus and M. chelonae; M. kansasii and M. gastri; M. marinum and M. ulcerans; M. senegalense, M. houstonense ATCC 49403, and M. farcinogenes; M. porcinum and M. neworleansense ATCC 49404; and M. septicum and M. peregrinum. In the present study, probes were based on a combination of the 16S rRNA gene and the ITS-1 region. The specificity of these probes was confirmed by extensive comparison with well-studied databases.
Although the 16S rRNA gene has a low mutation rate, it does display microheterogeneity within a species. Microheterogeneity was examined by mixed probes within the variable regions of M. intracellulare (sequevars [sequence variants] I to V), and M. gordonae (sequevar I to V) (Table (Table1).1). For example, M. intracellulare (sequevar III and IV) displayed microheterogeneity for type strain and were determined to be negative by the Amplicor M. intracellulare PCR. As an additional control for the accuracy of mycobacterial identification, mixed GNS probes were designed to select for the mycobacteria genus-specific region. These probes allowed differentiation of families closely related to the Mycobacteriaceae, including the Gordoniaceae, Nocardiaceae, and Tsukamurellaceae. A BRB probe based on the bacterial conserved region was designed for confirmation of the quantity and quality of sample DNA. In addition, mixed mycobacteria could be recognized by comparing the signal of a species probe with the signals of the BRB and GNS probes.
In the clinical laboratory, the MicroSeq 500 assay is a commercial sequencing assay which can be used for the routine identification of clinical mycobacterial isolates. The turnaround time for this assay is 2 days and requires approximately 4 h of a technologist's time. On the other hand, the turnaround time for an Invader assay of 20 samples was <6.5 h and required only about 0.5 h of a technologist's time. DNA sequencing requires less judgment on the part of technologists for interpretation and can identify a wide range mycobacterial species. However, labor, the reagents, equipment, and software necessary for this assay are significantly more expensive than the ribosomal probe hybridization. These factors limit the ability of hospital-based laboratories to use this assay (24). Moreover, DNA sequencing requires more time and effort to target mixed cultures and multiple copy regions, such as ITS-1, since subcloning or separate cultures are needed. Setup of the Invader assay system requires no expensive measuring equipment, such as an automated DNA sequencer, and no exclusive workspace for PCR.
In the present study, the Invader probe setting cost-effectively detected more than 90% of the mycobacteria isolated in a routine clinical laboratory. On the other hand, each species probe could also perform an independent assay as needed.
In supplemental studies, although M. tuberculosis complexes were difficult to identify at the species or strain level, these complex species and BCG strains could be discriminated by designing the Invader probes for gyrB and RD1 (12, 19, 34; data not shown). Since the Invader assay can measure genes with different GC contents, future versions may measure drug resistance genes simultaneously (41, 40, 20, 30, 31).
We thank Yuko Kazumi and Isamu Sugawara (Mycobacterial Reference Center, The Research Institute of Tuberculosis, Tokyo, Japan) for helpful discussions. We thank Fuminori Hoshino for help in preparing the manuscript and Gary Howard for critical reading of the manuscript.
Published ahead of print on 8 August 2007.