Comparison of an Internally Controlled, Large-Volume LightCycler Assay for Detection of Mycobacterium tuberculosis in Clinical Samples with
http://www.100md.com
微生物临床杂志 2005年第4期
Labor Becker, Olgemller und Kollegen, München
Institut für Medizinische Mikrobiologie und Hygiene, University of Regensburg
Bayerisches Landesamt für Gesundheit und Lebensmittelsicherheit, Oberschleiheim, Germany
ABSTRACT
We present a sensitive and specific assay for reliable and flexible detection of members of the Mycobacterium tuberculosis complex (MTBC) in clinical samples. This real-time PCR assay, which uses the LightCycler 2.0 instrument and 100-μl glass capillaries, can provide a result within 1 h after DNA extraction. The primers amplify a 206-bp fragment of the MTBC 16S rRNA gene. The sensor hybridization probe targets a region highly specific to members of the MTBC. The assay also includes a novel type of internal control that monitors the function of the reaction components and can detect potential inhibitors. Template DNA was extracted by the same procedure used for the COBAS AMPLICOR M. tuberculosis assay, so the LightCycler assay could be directly compared to the COBAS AMPLICOR assay. The LightCycler assay was evaluated with 146 clinical samples of various types. Very good agreement (100% sensitivity, 98.6% specificity) could be shown between the LightCycler and COBAS AMPLICOR assays. Specificity was checked with a panel of nontuberculous mycobacteria, as well as a large panel of bacterial and fungal organisms.
INTRODUCTION
Tuberculosis (TB) infections, including those with multidrug-resistant strains, have increased in recent years, leading to a demand for rapid, sensitive, and specific diagnosis of TB. Nucleic acid amplification techniques (NAT) have therefore become an important tool for TB diagnosis. While culture is still the "gold standard," NAT-based assays have the enormous advantage of usually providing a result within a day. In contrast, culture-dependent methods that detect members of the Mycobacterium tuberculosis complex (MTBC) may require several weeks to produce results.
Numerous commercial and in-house NAT assays have been described (3, 5-9, 11, 13, 15, 17-20, 26-29, 31). However, because of differences in patient selection, specimen collection, sample preparation, and DNA isolation, it is difficult to compare and judge the quality of these assays (23).
The only NAT-based assays approved by the Food and Drug Administration are the Amplified Mycobacterium tuberculosis Direct Test (MTD; Gen-Probe, San Diego, Calif.) and the COBAS AMPLICOR M. tuberculosis assay (Roche Diagnostics, Mannheim, Germany). Evaluation of the latter with well-characterized clinical samples showed that it performed well with smear-positive respiratory samples (24). However, the flexibility of this assay is limited since it relies on a relatively slow block cycler amplification process and time-consuming colorimetric detection of amplification products. For economic reasons, most laboratories can make only one PCR run per day; thus, just-in-time testing of single samples is not feasible.
In contrast, real-time techniques, which use fluorescently labeled probes like exonuclease probes (TaqMan probes) or hybridization probes (FRET probes) to detect amplicons during the amplification process, avoid all of the time-consuming, labor-intensive, and contamination-prone postamplification steps. The LightCycler instrument (Roche Applied Science, Penzberg, Germany) can even speed up the amplification process (32), shortening the time required for real-time PCR.
Because of the uneven distribution of mycobacterial cells in the sample material, sensitive MTBC detection with an NAT assay requires a large initial volume of DNA. Proficiency testing has already shown that a large number of procedures lack sufficient sensitivity to be applied to smear-negative samples (22). Unfortunately, the restriction of the original LightCycler assay to a total reaction volume of 20 μl was a clear disadvantage. Here we describe a new MTBC LightCycler assay that uses 100-μl capillaries and at least 25 μl of template DNA. In addition, the new assay offers a novel type of internal control (IC) (2) that does not require a separate detection system.
MATERIALS AND METHODS
Clinical specimens. A total of 146 clinical samples that had already been analyzed by the COBAS AMPLICOR M. tuberculosis assay were used. All samples were from different patients and had been sent to the laboratory for routine MTBC testing. Samples included respiratory (sputum, bronchial and tracheal aspirates, bronchial and tracheal secretions, bronchoalveolar lavage fluid) and nonrespiratory (biopsy, blood, gastric fluid, stool, urine, cerebrospinal fluid, wound secretion) specimens. All 72 samples positive in the COBAS AMPLICOR assay were also positive in acid-fast stained smears (+ to ++++), and 57 of these samples were culture positive. In 15 cases, the investigated specimens were negative by culture. All of these culture-negative specimens originated from patients who received effective antimycobacterial treatment, and all of these patients had a positive culture result for M. tuberculosis from an earlier respiratory sample.
Total nucleic acid from each sample, which had initially been prepared for the COBAS AMPLICOR assay, was stored below –18°C.
Microbial strains. DNAs from various bacteria and fungi used for specificity testing were obtained from the culture collection of the Institut für Medizinische Mikrobiologie und Hygiene at the University of Regensburg, Regensburg, Germany. The concentration of the DNAs was not measured photometrically, but all of the DNAs produced a strong band in a PCR when amplified with universal 16/18S rRNA gene primers and/or pan-mycobacterium primers KY18 and KY75 (31). The species identities of all strains were verified by sequencing the 16/18S rRNA gene. Bacteria and fungi used for specificity testing included M. tuberculosis (strain H37Rv), M. bovis, M. africanum, M. avium, M. intracellulare, M. gordonae, M. chelonae, M. kansasii, M. fortuitum, M. malmoense, M. celatum, M. scrofulaceum, M. phlei, M. xenopi, M. terrae, Chlamydophila pneumoniae, Chlamydophila psittaci, Corynebacterium pseudodiphtheriticum, Neisseria meningitidis, Neisseria lactamica, Nocardia brasiliensis, Streptococcus pneumoniae, Streptococcus spp. (groups A, B, C, F, and G), Enterococcus spp., Staphylococcus aureus, Staphylococcus epidermidis, Staphylococcus warneri, Staphylococcus haemolyticus, Staphylococcus schleiferi, Escherichia coli, Citrobacter freundii, Klebsiella pneumoniae, Pseudomonas aeruginosa, Enterobacter aerogenes, Serratia marcescens, Pseudomonas fluorescens, Bordetella pertussis, Bordetella parapertussis, Mycoplasma pneumoniae, Aspergillus fumigatus, Aspergillus niger, Candida glabrata, and Candida albicans.
DNA extraction. Clinical samples of various types were decontaminated by the N-acetyl-L-cysteine (NALC)-NaOH method (16). Sample material with a volume of up to 10 ml was used directly. For samples with a larger volume, 30 ml was concentrated to 5 ml by centrifugation prior to the NALC-NaOH procedure. Each sample was incubated at ambient temperature for 25 min with an equal volume of NALC-NaOH solution (4% NaOH, 1.45% Na citrate, 0.5% NALC). Ten volumes of phosphate buffer (6.7 mM, pH 7.4) was then added, and the diluted sample was centrifuged at 3,000 x g for 15 min. The sediment was resuspended in 1 ml of the same phosphate buffer that had been supplemented with 0.5% Tween 80 (24). DNA was extracted from a 100-μl aliquot of the suspension. The remainder was used for acid-fast staining and inoculation of culture media.
An alkaline lysis procedure was used to extract DNA from the sample in accordance with the instructions for the COBAS AMPLICOR assay (Respiratory Specimen Preparation Kit; Roche Diagnostics). Briefly, a 100-μl suspension was added to 1 ml of washing buffer and then centrifuged for 10 min at 14,000 x g. The pellet was resuspended in 100 μl of lysis buffer and incubated at 60°C for 45 min. Finally, 100 μl of neutralization reagent was added to the lysed sample material. The nucleic acid preparation was directly used in either the COBAS AMPLICOR assay or the LightCycler assay.
To determine the sensitivity of the LightCycler assay, genomic DNA was isolated from M. tuberculosis H37Rv. After a loopful of cultured cells was lysed via the alkaline procedure, the preparation was subjected to several cycles of freezing and thawing, an additional proteinase K digest, and a final DNA purification step with silica columns (QIAamp DNA Blood Mini Kit; QIAGEN, Hilden, Germany). The result was a relatively pure genomic DNA solution with little tendency to clump. The number of copies of the M. tuberculosis genome present in the solution was estimated by measuring the A260 and assuming that 10 fg of mycobacterial DNA corresponds roughly to 3 genomes (31). Since the neutralization reagent used in the alkaline lysis procedure contains MgCl2, all dilutions of nucleic acids were performed with a 1:1 mixture of lysis and neutralization reagent. Thus, identical PCR conditions were maintained for all samples.
IC. An IC oligonucleotide (ICO) basically contained just primer- and probe-binding regions (2). To allow this oligomer to be differentiated during LightCycler melting curve analysis, the synthesis process intentionally introduced three mismatches into the region that binds the fluorescein-labeled sensor probe. The ICO was prepared and used as described earlier (2). The sequence of the ICO (mismatches underlined) was 5'-ACGGAAAGGTCTCTTCGGAGATACTCGAGTGGCGAACGGGTGAGTAACACGTGGGTGGGAAGCATGTTTTGTGGTGTAAAGCGCTTTAGCGGTGTGGGATGAGCGTGACGGCCTACCAAG-3'.
PCR primers and probes. The oligonucleotide primers amplify a 206-bp region within the MTBC 16S rRNA gene. The sequence of the forward primer (MTB-F; Fig. 1) was 5'-ACGGAAAGGTCTCTTCG-3' (positions 88871 to 88887 in the genome of M. tuberculosis H37Rv) (4). It binds at approximately the same site as the KY18 primer used in the COBAS AMPLICOR assay (31). However, unlike the KY18 primer, its 3' end binds a somewhat more variable region. This change enhances the ability of the assay to distinguish between members of the MTBC and nontuberculous mycobacteria (NTM). This primer has already been used successfully to detect members of the MTBC in another LightCycler assay (27). The sequence of the reverse primer (MTB-R; Fig. 1) was 5'-CTTGGTAGGCCGTCAC-3' (positions 89076 to 89061) (4).
The hybridization probes bound a region highly specific to members of the MTBC. The same region is targeted by the COBAS AMPLICOR detection probe and a previously published LightCycler assay (27, 31). The sequences of the hybridization probes (MTB-FL and MTB-LC; Fig. 1) were 5'-GGATGCATGTCTTGTGGTGGAAA-(FL)-3'- (positions 88988 to 89010) (4), where FL = fluorescein, and 5'-(LC Red 640)-CGCTTTAGCGGTGTGGGATGAG-(Ph)-3' (positions 89012 to 89033) (4), where Ph is phosphate. When bound to the target, the two probes are separated by a gap of one nucleotide in order to allow efficient energy transfer between the two fluorophores. Since the gap nucleotide and, in many cases, a second, adjacent nucleotide are missing in NTM samples, quenching and/or steric hindrance of the fluorophores would be expected when the probes are bound to PCR products from NTM. In addition, there are at least two mismatches between one of the probes and the 16S rRNA gene target sequence of other (nontuberculous) mycobacterial species. Figure 1 shows the best alignment of the primers and probes with the ICO and the M. tuberculosis and NTM genomes.
For standard block cycler amplification of mycobacteria, primers KY18 (5'-CACATGCAAGTCGAACGGAAAGG-3') and KY75 (5'-GCCCGTATCGCCCGCACGCTCACA-3') were used. Primers 533F (5'-TGBCAGCMGCCGCGGTAA-3') and 1492R (5'-ACGGHTACCTTGTTACGACTT-3') were used for universal amplification of bacterial and fungal DNAs (1, 31).
The ICO and all of the primers and probes used were synthesized by Metabion, Martinsried, Germany.
PCR amplification. Rapid-cycling real-time PCR and melting temperature (Tm) analysis were performed with a LightCycler 2.0 instrument (Roche Applied Science). PCR was done with either 20- or 100-μl glass capillaries (Roche Applied Science). Each 10-μl reaction mixture contained 4 pmol of forward primer, 6 pmol of reverse primer, 2 pmol of each probe, 1.25 μl of ICO (10 copies per μl), 2 μl of LightCycler-FastStart DNA Master Hybridization ProbesPLUS (which includes reaction buffer, nucleotides, and Taq DNA polymerase; Roche Applied Science), and 2.5 μl of nucleic acid.
We decided to use no larger amounts of template DNA to minimize the risk of PCR inhibition. The DNA extracted with the Respiratory Specimen Preparation kit (Roche Diagnostics) also contains significant amounts of MgCl2, which might reduce PCR efficiency when used in a higher concentration.
In 20-μl capillaries, thermocycling was performed as follows: 95°C for 10 min for initial denaturation and activation of Taq polymerase, followed by 50 thermal cycles of 95°C for 0 s, 55°C for 10 s, 63°C for 5 s, and 72°C for 20 s. The ramping rate was 20°C/s. Fluorescence was measured during each 63°C step for specific real-time detection of M. tuberculosis PCR products. Because of the larger reaction volume, the incubation times of the thermocycling had to be increased for 100-μl capillaries as follows: 95°C for 15 s, 55°C for 20 s, 63°C for 10 s, and 72°C for 30 s.
After the amplification was complete, a melting curve analysis was performed as follows. The capillaries were heated at 95°C for 10 s, incubated at 40°C for 1 min, and finally slowly (0.2°C/s) heated to 80°C. Fluorescence was monitored continuously during the melting experiment. The LightCycler software (version. 4.0; Roche Molecular Biochemicals) converted melting curves to melting peaks by calculating the negative derivative of each measured fluorescence with respect to the temperature (–dF/dT) and plotting –dF/dT against temperature for the entire melting experiment.
Standard block cycler PCR amplification was performed with primers KY18/K75 or 533F/1492R on a GeneAmp PCR system 9600 thermocycler (Applied Biosystems, Foster City, Calif.). Initial denaturation was performed at 95°C for 15 min to activate Taq polymerase. This step was followed by 40 thermal cycles of 95°C for 20 s, 62°C for 20 s, and 72°C for 45 s. The 50-μl reaction mixtures contained 20 pmol of each primer oligonucleotide; dGTP, dATP, and dCTP (0.25 mM each); 0.75 mM dUTP; 3 mM MgCl2; 1.25 U of HotStarTaq DNA polymerase (QIAGEN); and 5 μl of nucleic acid in 1x HotStarTaq PCR buffer (QIAGEN).
COBAS AMPLICOR PCR. The Roche COBAS AMPLICOR PCR was performed in accordance with the manufacturer's instructions. Fifty microliters of nucleic acid was used for each reaction.
RESULTS
Sensitivity. A dilution series of DNA from M. tuberculosis H37Rv was assayed on the LightCycler in both 20- and 100-μl capillaries. Five experiments were performed in parallel for each dilution. The result is shown in Table 1. At a DNA concentration of 2.5 genomes per μl, two of five results were positive with the 100-μl capillary whereas none of the experiments was positive with the 20-μl capillary. At 5 genomes per μl, three of five results were positive with the 100-μl capillaries while one of five results was positive with the 20-μl capillaries.
The COBAS AMPLICOR assay could detect M. tuberculosis DNA down to 0.5 genome per μl, which corresponds to 25 genomes per assay (data not shown).
To investigate the influence of human DNA on the LightCycler assay, two identical dilution series of M. tuberculosis DNA were prepared, one with a diluent that contained 1.5 ng of human DNA per μl and the other with a diluent that lacked human DNA. Both dilution series gave the same result in the LightCycler assay, showing that the human DNA usually present in various amounts in clinical samples does not affect the sensitivity of the assay (data not shown).
Specificity of the assay. The specificity of the LightCycler assay was determined with a panel of different MTBC members, various NTM, and other bacteria and fungi. A real-time fluorescence signal was obtained only from members of the MTBC, i.e., M. bovis and M. africanum. All other investigated species were negative (data not shown).
The result of an experiment with M. tuberculosis, five different NTM, and two phylogenetically related, morphologically similar bacteria, i.e., C. pseudodiphtheriticum and N. brasiliensis, is shown in Fig. 2. A fluorescence signal was only obtained when M. tuberculosis DNA was the template (Fig. 2A, curve 9). Tm analysis showed a specific Tm of about 65°C for M. tuberculosis (Fig. 2B, curve 9). Melting peaks were also seen for M. avium and M. kansasii (Fig. 2B, curves 1 and 2), but since these Tms were below the temperature of the fluorescence acquisition step during PCR, those NTM did not produce a signal on the LightCycler amplification screen (Fig. 2A, curves 1 and 2). M. scrofulaceum also had a melting peak at about 58°C (data not shown). Agarose gel electrophoresis of the amplification products analyzed in Fig. 2 showed that all of the mycobacteria, as well as C. pseudodiphtheriticum and N. brasiliensis, were amplified (data not shown).
Performance of the IC. The optimal amount of ICO was determined in titration studies (data not shown). The goal was to keep the amount of ICO in the reaction mixture as small as possible to minimize competition with M. tuberculosis in low-titer samples, which could reduce the sensitivity of the assay. The optimal amount for 100-μl reaction mixtures proved to be 125 copies of the ICO (calculated from the concentration data provided by the manufacturer). In a dilution series of M. tuberculosis DNA, no significant difference in sensitivity could be seen between reaction mixtures with ICO and those without ICO (data not shown). Figure 3 shows an experiment with a fixed amount of M. tuberculosis DNA, a fixed amount of ICO, and a variable amount of inhibitor. NALC-NaOH solution was used as an inhibitor, since this reagent is used for initial decontamination of all clinical samples and traces of it might remain in inadequately washed DNA preparations. No fluorescence signal was obtained from reaction mixtures containing 1:5 and 1:10 dilutions of NALC-NaOH (Fig. 3A, curves 1 and 2). These samples were, however, clearly shown to be false negatives, since the ICO was not amplified in these reaction mixtures (as shown by the absence of an ICO-specific melting peak in curves 1 and 2 of Fig. 3B). In contrast, the negative control without inhibitor (Fig. 3B, curve 6) showed a strong ICO-specific peak at about 48°C. Reaction mixtures containing more dilute preparations of NALC-NaOH (Fig. 3, curves 3, 4, and 5) showed an increase in the M. tuberculosis-specific fluorescence signal, as well as an M. tuberculosis-specific melting peak at about 65°C. However, since the high concentration of M. tuberculosis outcompeted ICO amplification in sample 5, no ICO peak was seen in this sample (Fig. 3B, curve 5).
Evaluation with clinical samples. The LightCycler assay was evaluated with a variety of clinical samples that had previously been analyzed with the COBAS AMPLICOR assay (Table 2). All of the samples that initially tested positive in the COBAS AMPLICOR assay, i.e., 67 respiratory and 5 nonrespiratory specimens, were also positive in the LightCycler assay. Of the 73 samples that tested negative in the COBAS AMPLICOR assay (41 respiratory and 32 nonrespiratory), 72 were negative in the LightCycler assay.
For one bronchoalveolar lavage fluid sample that showed a positive smear, the result of culture analysis was M. malmoense. This sample was negative in the COBAS AMPLICOR assay but showed a weak positive signal in the LightCycler assay.
Because of inhibition, no valid COBAS AMPLICOR result could be obtained from one smear-positive, culture-positive stool sample. This sample, however, was clearly positive in the LightCycler assay.
DISCUSSION
We have developed a LightCycler assay for rapid detection of members of the MTBC. The forward primer and hybridization probes used in this assay target regions of the 16S rRNA gene similar to those targeted by the Food and Drug Administration-approved COBAS AMPLICOR assay. Since we intended to design an MTBC-specific assay, the forward primer was slightly extended to cover a more variable region of the 16S rRNA gene and produce more selective amplification. The reverse primer was shifted upstream to produce a shorter PCR product (206 bp for the LightCycler assay versus 584 bp for the COBAS AMPLICOR assay) because rapid-cycle PCR amplification works better when the PCR product is kept as short as possible. Since this region of the 16S rRNA gene lacked highly variable sequences, it was not possible to design a reverse primer that was more selective for the members of the MTBC.
Although the COBAS AMPLICOR assay is specific for the members of the MTBC, the assay will efficiently amplify the DNAs of all mycobacterial species (except M. simiae) (31). Despite several oligonucleotide mismatches, especially at the binding sites of the forward primer, the LightCycler assay also amplifies a corresponding segment within the 16S rRNA gene from most of the NTM species. However, only members of MTBC produced a real-time fluorescence signal.
Even during Tm analysis we had expected little or no signal from NTM since the hybridization probes bound to these samples should touch or even overlap, leading to steric problems and inefficient energy transfer. Interestingly, however, melting peaks were obtained for some NTM; these peaks could still be clearly differentiated from the MTBC peak by their lower Tm.
Inhibitors present in clinical samples, as well as inefficient extraction of nucleic acid, can cause false-negative results in amplification assays. Therefore, ICs should be required in all amplification assays used for diagnostic pathogen detection (14, 21).
For the COBAS AMPLICOR assay, a linearized, recombinant plasmid DNA serves as an IC (25). The plasmid contains the target sequence, which has been modified at the binding site of the detection probe. If the IC is successfully amplified, the amplicon will be detected with an IC-specific probe in a separate post-PCR hybridization reaction.
Similar ICs have been described for real-time amplification assays (10, 30, 33). In these, amplification of the IC was monitored with a separate probe that contained a fluorescent label different from that on the target-specific probe.
Recently, we have published a novel, simple technique for control of real-time amplification assays (2). In this technique, the ICs are single-stranded oligonucleotides that contain little more than the primer sites and a probe-binding site that contains mismatches to the actual target. An advantage of these ICOs is that they do not require separate control-specific detection probes. Because of the mismatches, probe-control hybridization does not occur during the fluorescence acquisition step of the PCR and therefore does not interfere with detection of the pathogen. However, ICO amplification can be detected during melting point analysis after PCR. Another advantage is that these ICOs can be obtained from many commercial providers of oligonucleotides, thereby eliminating the need for complicated, time-consuming, and expensive construction of plasmids (2). In the present study we have successfully applied this ICO technique to our LightCycler assay for the members of the MTBC. We have shown that the ICO does not reduce assay sensitivity but is able to detect impurities in the DNA preparation such as NALC, which inhibited amplification at concentrations as low as 0.8%.
Several in-house MTBC LightCycler assays have already been published (9, 12, 17, 19, 27). Recently, however, a meta-analysis of NAT used to diagnose tuberculous pleuritis showed significant heterogeneity in results from in-house ("home-brew") PCRs (23), while commercial assays did not show such heterogeneity. Since the DNA sample was prepared the same way for both assays, the commercial COBAS AMPLICOR test and our LightCycler assay could be directly compared with clinical samples. The agreement between the two assays was very high (100% sensitivity, 98.6% specificity). All samples positive in the COBAS AMPLICOR assay were also positive in the LightCycler assay, even though the COBAS AMPLICOR assay uses a larger volume (50 versus 25 μl) of sample DNA and is slightly more sensitive for analysis of DNA purified from cultured M. tuberculosis. The only discrepancy within 73 negative samples was one sample containing M. malmoense, which was negative in the COBAS AMPLICOR assay but produced a weak positive signal in the LightCycler assay. We cannot explain this result, since the LightCycler assay was clearly negative with DNA purified from a cultured M. malmoense strain.
When the sample is purified M. tuberculosis nucleic acid, the LightCycler assay appears to be slightly more sensitive when 100-μl capillaries are used instead of 20-μl capillaries. However, sensitivity tests based on purified DNA samples have only limited applicability to the assay of clinical samples. Mycobacteria usually are not evenly distributed in the sample material. In addition, cell clumping increases during the centrifugation steps required for decontamination and enrichment of clinical samples. Still, one can assume that increasing the volume of the initial sample is an important factor in increasing the sensitivity of the assay. Both the COBAS AMPLICOR assay and the LightCycler 100-μl assay allow relatively large sample volumes, i.e., 50 and 25 μl, respectively.
Compared with the COBAS AMPLICOR assay, the LightCycler assay is much more flexible. For example, if the sample contains an inhibitor the problem can usually be easily solved by repeating the PCR with a slight dilution of the sample DNA. With the COBAS instrument, it is almost impossible to do such a repeat assay on the same day as the initial assay. In contrast, since a LightCycler run takes only about an hour, rapid repeats are possible. Also, performing a COBAS AMPLICOR run with a single sample is very expensive since the assay still requires separate positive and negative controls. Because of the generally lower reagent costs, the LightCycler assay is more cost-effective, especially on runs with a single sample or only a few samples. It might even be possible to omit the positive control since the ICO can monitor the function of all of the components in the reaction mixture. In fact, when applying this strategy the efficiency of the nucleic acid extraction procedure would not be controlled. It is beyond dispute that coprocessing an artificial or real patient sample with a low number of M. tuberculosis cells would represent an optimal positive control. However, such a positive control is also not included in the COBAS AMPLICOR assay.
The use of commercial NAT assays to confirm TB, as well as their limited application to exclude a TB diagnosis, has already been discussed (23). Here we report a cost-effective in-house assay for fast and reliable detection of the members of the MTBC that compares favorably with the COBAS AMPLICOR assay. Since both assays use the same DNA preparation, the LightCycler assay can complement the latter assay, adding speed and flexibility to the routine diagnosis of TB infection.
ACKNOWLEDGMENTS
We thank M. Kühn and K. Stieber for providing cultures of M. tuberculosis and various bacterial strains and T. Becker for thoughtful review of the manuscript.
REFERENCES
Burggraf, S., H. Huber, and K. O. Stetter. 1997. Reclassification of the crenarchael orders and families in accordance with 16S rRNA sequence data. Int. J. Syst. Bacteriol. 47:657-660.
Burggraf, S., and B. Olgemller. 2004. Simple technique for internal control of real-time amplification assays. Clin. Chem. 50:819-825.
Cleary, T. J., G. Roudel, O. Casillas, and N. Miller. 2003. Rapid and specific detection of Mycobacterium tuberculosis by using the Smart Cycler instrument and a specific fluorogenic probe. J. Clin. Microbiol. 41:4783-4786.
Cole, S. T., R. Brosch, J. Parkhill, T. Garnier, C. Churcher, D. Harris, S. V. Gordon, K. Eiglmeier, S. Gas, C. E. Barry III, F. Tekaia, K. Badcock, D. Basham, D. Brown, T. Chillingworth, R. Connor, R. Davies, K. Devlin, T. Feltwell, S. Gentles, N. Hamlin, S. Holroyd, T. Hornsby, K. Jagels, and B. G. Barrell. 1998. Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393:537-544.
Cousins, D. V., S. D. Wilton, B. R. Francis, and B. L. Gow. 1992. Use of polymerase chain reaction for rapid diagnosis of tuberculosis. J. Clin. Microbiol. 30:255-258.
Del Portillo, P., L. A. Murillo, and M. E. Patarroyo. 1991. Amplification of a species-specific DNA fragment of Mycobacterium tuberculosis and its possible use in diagnosis. J. Clin. Microbiol. 29:2163-2168.
Desjardin, L. E., Y. Chen, M. D. Perkins, L. Teixeira, M. D. Cave, and K. D. Eisenach. 1998. Comparison of the ABI 7700 system (TaqMan) and competitive PCR for quantification of IS6110 DNA in sputum during treatment of tuberculosis. J. Clin. Microbiol. 36:1964-1968.
De Wit, D., L. Steyn, S. Shoemaker, and M. Sogin. 1990. Direct detection of Mycobacterium tuberculosis in clinical specimens by DNA amplification. J. Clin. Microbiol. 28:2437-2441.
Drosten, C., M. Panning, and S. Kramme. 2003. Detection of Mycobacterium tuberculosis by real-time PCR using pan-mycobacterial primers and a pair of fluorescence resonance energy transfer probes specific for the M. tuberculosis complex. Clin. Chem. 49:1659-1661.
Drosten, C., M. Weber, E. Seifried, and W. K. Roth. 2000. Evaluation of a new PCR assay with competitive internal control sequence for blood donor screening. Transfusion 40:718-724.
Eisenach, K. D., M. D. Cave, J. H. Bates, and J. T. Crawford. 1990. Polymerase chain reaction amplification of a repetitive DNA sequence specific for Mycobacterium tuberculosis. J. Infect. Dis. 161:977-981.
Heginbothom, M. L., J. T. Magee, and P. G. Flanagan. 2003. Evaluation of the Idaho Technology LightCycler PCR for the direct detection of Mycobacterium tuberculosis in respiratory specimens. Int. J. Tuberc. Lung Dis. 7:78-83.
Hermans, P. W., A. R. Schuitema, D. Van Soolingen, C. P. Verstynen, E. M. Bik, J. E. Thole, A. H. Kolk, and J. D. van Embden. 1990. Specific detection of Mycobacterium tuberculosis complex strains by polymerase chain reaction. J. Clin. Microbiol. 28:1204-1213.
Hoorfar, J., N. Cook, B. Malorny, M. Wagner, D. De Medici, A. Abdulmawjood, and P. Fach. 2003. Making internal amplification control mandatory for diagnostic PCR. J. Clin. Microbiol. 41:5835.
Kraus, G., T. Cleary, N. Miller, R. Seivright, A. K. Young, G. Spruill, and H. J. Hnatyszyn. 2001. Rapid and specific detection of the Mycobacterium tuberculosis complex using fluorogenic probes and real-time PCR. Mol. Cell. Probes 15:375-383.
Kubica, G. P., W. E. Dye, M. L. Cohn, and G. Middlebrook. 1963. Sputum digestion and decontamination with N-acetyl-L-cysteine-sodium hydroxide for culture of mycobacteria. Am. Rev. Respir. Dis. 87:775-779.
Lachnik, J., B. Ackermann, A. Bohrssen, S. Maass, C. Diephaus, A. Puncken, M. Stermann, and F. C. Bange. 2002. Rapid-cycle PCR and fluorimetry for detection of mycobacteria. J. Clin. Microbiol. 40:3364-3373.
Lemaitre, N., S. Armand, A. Vachee, O. Capilliez, C. Dumoulin, and R. J. Courcol. 2004. Comparison of the real-time PCR method and the Gen-Probe Amplified Mycobacterium Tuberculosis Direct Test for detection of Mycobacterium tuberculosis in pulmonary and nonpulmonary specimens. J. Clin. Microbiol. 42:4307-4309.
Miller, N., T. Cleary, G. Kraus, A. K. Young, G. Spruill, and H. J. Hnatyszyn. 2002. Rapid and specific detection of Mycobacterium tuberculosis from acid-fast bacillus smear-positive respiratory specimens and BacT/ALERT MP culture bottles by using fluorogenic probes and real-time PCR. J. Clin. Microbiol. 40:4143-4147.
Miller, N., S. G. Hernandez, and T. J. Cleary. 1994. Evaluation of Gen-Probe Amplified Mycobacterium Tuberculosis Direct Test and PCR for direct detection of Mycobacterium tuberculosis in clinical specimens. J. Clin. Microbiol. 32:393-397.
Nolte, F. S. 2004. Novel internal controls for real-time PCR assays. Clin. Chem. 50:801-802.
Noordhoek, G. T., S. Mulder, P. Wallace, and A. M. van Loon. 2004. Multicentre quality control study for detection of Mycobacterium tuberculosis in clinical samples by nucleic amplification methods. Clin. Microbiol. Infect. 10:295-301.
Pai, M., L. L. Flores, A. Hubbard, L. W. Riley, and J. M. Colford, Jr. 2004. Nucleic acid amplification tests in the diagnosis of tuberculous pleuritis: a systematic review and meta-analysis. BMC Infect. Dis. 4:6.
Reischl, U., N. Lehn, H. Wolf, and L. Naumann. 1998. Clinical evaluation of the automated COBAS AMPLICOR MTB assay for testing respiratory and nonrespiratory specimens. J. Clin. Microbiol. 36:2853-2860.
Rosenstraus, M., Z. Wang, S. Y. Chang, D. DeBonville, and J. P. Spadoro. 1998. An internal control for routine diagnostic PCR: design, properties, and effect on clinical performance. J. Clin. Microbiol. 36:191-197.
Shankar, P., N. Manjunath, K. K. Mohan, K. Prasad, M. Behari, Shriniwas, and G. K. Ahuja. 1991. Rapid diagnosis of tuberculous meningitis by polymerase chain reaction. Lancet 337:5-7.
Shrestha, N. K., M. J. Tuohy, G. S. Hall, U. Reischl, S. M. Gordon, and G. W. Procop. 2003. Detection and differentiation of Mycobacterium tuberculosis and nontuberculous mycobacterial isolates by real-time PCR. J. Clin. Microbiol. 41:5121-5126.
Sjobring, U., M. Mecklenburg, A. B. Andersen, and H. Miorner. 1990. Polymerase chain reaction for detection of Mycobacterium tuberculosis. J. Clin. Microbiol. 28:2200-2204.
Sritharan, V., and R. H. Barker, Jr. 1991. A simple method for diagnosing M. tuberculosis infection in clinical samples using PCR. Mol. Cell Probes 5:385-395.
Stocher, M., V. Leb, and J. Berg. 2003. A convenient approach to the generation of multiple internal control DNA for a panel of real-time PCR assays. J. Virol. Methods. 108:1-8.
Tevere, V. J., P. L. Hewitt, A. Dare, P. Hocknell, A. Keen, J. P. Spadoro, and K. K. Young. 1996. Detection of Mycobacterium tuberculosis by PCR amplification with pan-Mycobacterium primers and hybridization to an M. tuberculosis-specific probe. J. Clin. Microbiol. 34:918-923.
Wittwer, C. T., K. M. Ririe, R. V. Andrew, D. A. David, R. A. Gundry, and U. J. Balis. 1997. The LightCycler: a microvolume multisample fluorimeter with rapid temperature control. BioTechniques 22:176-181.
Zimmermann, K., and J. W. Mannhalter. 1996. Technical aspects of quantitative competitive PCR. BioTechniques 21:268-272, 274-279.(Siegfried Burggraf, Udo R)
Institut für Medizinische Mikrobiologie und Hygiene, University of Regensburg
Bayerisches Landesamt für Gesundheit und Lebensmittelsicherheit, Oberschleiheim, Germany
ABSTRACT
We present a sensitive and specific assay for reliable and flexible detection of members of the Mycobacterium tuberculosis complex (MTBC) in clinical samples. This real-time PCR assay, which uses the LightCycler 2.0 instrument and 100-μl glass capillaries, can provide a result within 1 h after DNA extraction. The primers amplify a 206-bp fragment of the MTBC 16S rRNA gene. The sensor hybridization probe targets a region highly specific to members of the MTBC. The assay also includes a novel type of internal control that monitors the function of the reaction components and can detect potential inhibitors. Template DNA was extracted by the same procedure used for the COBAS AMPLICOR M. tuberculosis assay, so the LightCycler assay could be directly compared to the COBAS AMPLICOR assay. The LightCycler assay was evaluated with 146 clinical samples of various types. Very good agreement (100% sensitivity, 98.6% specificity) could be shown between the LightCycler and COBAS AMPLICOR assays. Specificity was checked with a panel of nontuberculous mycobacteria, as well as a large panel of bacterial and fungal organisms.
INTRODUCTION
Tuberculosis (TB) infections, including those with multidrug-resistant strains, have increased in recent years, leading to a demand for rapid, sensitive, and specific diagnosis of TB. Nucleic acid amplification techniques (NAT) have therefore become an important tool for TB diagnosis. While culture is still the "gold standard," NAT-based assays have the enormous advantage of usually providing a result within a day. In contrast, culture-dependent methods that detect members of the Mycobacterium tuberculosis complex (MTBC) may require several weeks to produce results.
Numerous commercial and in-house NAT assays have been described (3, 5-9, 11, 13, 15, 17-20, 26-29, 31). However, because of differences in patient selection, specimen collection, sample preparation, and DNA isolation, it is difficult to compare and judge the quality of these assays (23).
The only NAT-based assays approved by the Food and Drug Administration are the Amplified Mycobacterium tuberculosis Direct Test (MTD; Gen-Probe, San Diego, Calif.) and the COBAS AMPLICOR M. tuberculosis assay (Roche Diagnostics, Mannheim, Germany). Evaluation of the latter with well-characterized clinical samples showed that it performed well with smear-positive respiratory samples (24). However, the flexibility of this assay is limited since it relies on a relatively slow block cycler amplification process and time-consuming colorimetric detection of amplification products. For economic reasons, most laboratories can make only one PCR run per day; thus, just-in-time testing of single samples is not feasible.
In contrast, real-time techniques, which use fluorescently labeled probes like exonuclease probes (TaqMan probes) or hybridization probes (FRET probes) to detect amplicons during the amplification process, avoid all of the time-consuming, labor-intensive, and contamination-prone postamplification steps. The LightCycler instrument (Roche Applied Science, Penzberg, Germany) can even speed up the amplification process (32), shortening the time required for real-time PCR.
Because of the uneven distribution of mycobacterial cells in the sample material, sensitive MTBC detection with an NAT assay requires a large initial volume of DNA. Proficiency testing has already shown that a large number of procedures lack sufficient sensitivity to be applied to smear-negative samples (22). Unfortunately, the restriction of the original LightCycler assay to a total reaction volume of 20 μl was a clear disadvantage. Here we describe a new MTBC LightCycler assay that uses 100-μl capillaries and at least 25 μl of template DNA. In addition, the new assay offers a novel type of internal control (IC) (2) that does not require a separate detection system.
MATERIALS AND METHODS
Clinical specimens. A total of 146 clinical samples that had already been analyzed by the COBAS AMPLICOR M. tuberculosis assay were used. All samples were from different patients and had been sent to the laboratory for routine MTBC testing. Samples included respiratory (sputum, bronchial and tracheal aspirates, bronchial and tracheal secretions, bronchoalveolar lavage fluid) and nonrespiratory (biopsy, blood, gastric fluid, stool, urine, cerebrospinal fluid, wound secretion) specimens. All 72 samples positive in the COBAS AMPLICOR assay were also positive in acid-fast stained smears (+ to ++++), and 57 of these samples were culture positive. In 15 cases, the investigated specimens were negative by culture. All of these culture-negative specimens originated from patients who received effective antimycobacterial treatment, and all of these patients had a positive culture result for M. tuberculosis from an earlier respiratory sample.
Total nucleic acid from each sample, which had initially been prepared for the COBAS AMPLICOR assay, was stored below –18°C.
Microbial strains. DNAs from various bacteria and fungi used for specificity testing were obtained from the culture collection of the Institut für Medizinische Mikrobiologie und Hygiene at the University of Regensburg, Regensburg, Germany. The concentration of the DNAs was not measured photometrically, but all of the DNAs produced a strong band in a PCR when amplified with universal 16/18S rRNA gene primers and/or pan-mycobacterium primers KY18 and KY75 (31). The species identities of all strains were verified by sequencing the 16/18S rRNA gene. Bacteria and fungi used for specificity testing included M. tuberculosis (strain H37Rv), M. bovis, M. africanum, M. avium, M. intracellulare, M. gordonae, M. chelonae, M. kansasii, M. fortuitum, M. malmoense, M. celatum, M. scrofulaceum, M. phlei, M. xenopi, M. terrae, Chlamydophila pneumoniae, Chlamydophila psittaci, Corynebacterium pseudodiphtheriticum, Neisseria meningitidis, Neisseria lactamica, Nocardia brasiliensis, Streptococcus pneumoniae, Streptococcus spp. (groups A, B, C, F, and G), Enterococcus spp., Staphylococcus aureus, Staphylococcus epidermidis, Staphylococcus warneri, Staphylococcus haemolyticus, Staphylococcus schleiferi, Escherichia coli, Citrobacter freundii, Klebsiella pneumoniae, Pseudomonas aeruginosa, Enterobacter aerogenes, Serratia marcescens, Pseudomonas fluorescens, Bordetella pertussis, Bordetella parapertussis, Mycoplasma pneumoniae, Aspergillus fumigatus, Aspergillus niger, Candida glabrata, and Candida albicans.
DNA extraction. Clinical samples of various types were decontaminated by the N-acetyl-L-cysteine (NALC)-NaOH method (16). Sample material with a volume of up to 10 ml was used directly. For samples with a larger volume, 30 ml was concentrated to 5 ml by centrifugation prior to the NALC-NaOH procedure. Each sample was incubated at ambient temperature for 25 min with an equal volume of NALC-NaOH solution (4% NaOH, 1.45% Na citrate, 0.5% NALC). Ten volumes of phosphate buffer (6.7 mM, pH 7.4) was then added, and the diluted sample was centrifuged at 3,000 x g for 15 min. The sediment was resuspended in 1 ml of the same phosphate buffer that had been supplemented with 0.5% Tween 80 (24). DNA was extracted from a 100-μl aliquot of the suspension. The remainder was used for acid-fast staining and inoculation of culture media.
An alkaline lysis procedure was used to extract DNA from the sample in accordance with the instructions for the COBAS AMPLICOR assay (Respiratory Specimen Preparation Kit; Roche Diagnostics). Briefly, a 100-μl suspension was added to 1 ml of washing buffer and then centrifuged for 10 min at 14,000 x g. The pellet was resuspended in 100 μl of lysis buffer and incubated at 60°C for 45 min. Finally, 100 μl of neutralization reagent was added to the lysed sample material. The nucleic acid preparation was directly used in either the COBAS AMPLICOR assay or the LightCycler assay.
To determine the sensitivity of the LightCycler assay, genomic DNA was isolated from M. tuberculosis H37Rv. After a loopful of cultured cells was lysed via the alkaline procedure, the preparation was subjected to several cycles of freezing and thawing, an additional proteinase K digest, and a final DNA purification step with silica columns (QIAamp DNA Blood Mini Kit; QIAGEN, Hilden, Germany). The result was a relatively pure genomic DNA solution with little tendency to clump. The number of copies of the M. tuberculosis genome present in the solution was estimated by measuring the A260 and assuming that 10 fg of mycobacterial DNA corresponds roughly to 3 genomes (31). Since the neutralization reagent used in the alkaline lysis procedure contains MgCl2, all dilutions of nucleic acids were performed with a 1:1 mixture of lysis and neutralization reagent. Thus, identical PCR conditions were maintained for all samples.
IC. An IC oligonucleotide (ICO) basically contained just primer- and probe-binding regions (2). To allow this oligomer to be differentiated during LightCycler melting curve analysis, the synthesis process intentionally introduced three mismatches into the region that binds the fluorescein-labeled sensor probe. The ICO was prepared and used as described earlier (2). The sequence of the ICO (mismatches underlined) was 5'-ACGGAAAGGTCTCTTCGGAGATACTCGAGTGGCGAACGGGTGAGTAACACGTGGGTGGGAAGCATGTTTTGTGGTGTAAAGCGCTTTAGCGGTGTGGGATGAGCGTGACGGCCTACCAAG-3'.
PCR primers and probes. The oligonucleotide primers amplify a 206-bp region within the MTBC 16S rRNA gene. The sequence of the forward primer (MTB-F; Fig. 1) was 5'-ACGGAAAGGTCTCTTCG-3' (positions 88871 to 88887 in the genome of M. tuberculosis H37Rv) (4). It binds at approximately the same site as the KY18 primer used in the COBAS AMPLICOR assay (31). However, unlike the KY18 primer, its 3' end binds a somewhat more variable region. This change enhances the ability of the assay to distinguish between members of the MTBC and nontuberculous mycobacteria (NTM). This primer has already been used successfully to detect members of the MTBC in another LightCycler assay (27). The sequence of the reverse primer (MTB-R; Fig. 1) was 5'-CTTGGTAGGCCGTCAC-3' (positions 89076 to 89061) (4).
The hybridization probes bound a region highly specific to members of the MTBC. The same region is targeted by the COBAS AMPLICOR detection probe and a previously published LightCycler assay (27, 31). The sequences of the hybridization probes (MTB-FL and MTB-LC; Fig. 1) were 5'-GGATGCATGTCTTGTGGTGGAAA-(FL)-3'- (positions 88988 to 89010) (4), where FL = fluorescein, and 5'-(LC Red 640)-CGCTTTAGCGGTGTGGGATGAG-(Ph)-3' (positions 89012 to 89033) (4), where Ph is phosphate. When bound to the target, the two probes are separated by a gap of one nucleotide in order to allow efficient energy transfer between the two fluorophores. Since the gap nucleotide and, in many cases, a second, adjacent nucleotide are missing in NTM samples, quenching and/or steric hindrance of the fluorophores would be expected when the probes are bound to PCR products from NTM. In addition, there are at least two mismatches between one of the probes and the 16S rRNA gene target sequence of other (nontuberculous) mycobacterial species. Figure 1 shows the best alignment of the primers and probes with the ICO and the M. tuberculosis and NTM genomes.
For standard block cycler amplification of mycobacteria, primers KY18 (5'-CACATGCAAGTCGAACGGAAAGG-3') and KY75 (5'-GCCCGTATCGCCCGCACGCTCACA-3') were used. Primers 533F (5'-TGBCAGCMGCCGCGGTAA-3') and 1492R (5'-ACGGHTACCTTGTTACGACTT-3') were used for universal amplification of bacterial and fungal DNAs (1, 31).
The ICO and all of the primers and probes used were synthesized by Metabion, Martinsried, Germany.
PCR amplification. Rapid-cycling real-time PCR and melting temperature (Tm) analysis were performed with a LightCycler 2.0 instrument (Roche Applied Science). PCR was done with either 20- or 100-μl glass capillaries (Roche Applied Science). Each 10-μl reaction mixture contained 4 pmol of forward primer, 6 pmol of reverse primer, 2 pmol of each probe, 1.25 μl of ICO (10 copies per μl), 2 μl of LightCycler-FastStart DNA Master Hybridization ProbesPLUS (which includes reaction buffer, nucleotides, and Taq DNA polymerase; Roche Applied Science), and 2.5 μl of nucleic acid.
We decided to use no larger amounts of template DNA to minimize the risk of PCR inhibition. The DNA extracted with the Respiratory Specimen Preparation kit (Roche Diagnostics) also contains significant amounts of MgCl2, which might reduce PCR efficiency when used in a higher concentration.
In 20-μl capillaries, thermocycling was performed as follows: 95°C for 10 min for initial denaturation and activation of Taq polymerase, followed by 50 thermal cycles of 95°C for 0 s, 55°C for 10 s, 63°C for 5 s, and 72°C for 20 s. The ramping rate was 20°C/s. Fluorescence was measured during each 63°C step for specific real-time detection of M. tuberculosis PCR products. Because of the larger reaction volume, the incubation times of the thermocycling had to be increased for 100-μl capillaries as follows: 95°C for 15 s, 55°C for 20 s, 63°C for 10 s, and 72°C for 30 s.
After the amplification was complete, a melting curve analysis was performed as follows. The capillaries were heated at 95°C for 10 s, incubated at 40°C for 1 min, and finally slowly (0.2°C/s) heated to 80°C. Fluorescence was monitored continuously during the melting experiment. The LightCycler software (version. 4.0; Roche Molecular Biochemicals) converted melting curves to melting peaks by calculating the negative derivative of each measured fluorescence with respect to the temperature (–dF/dT) and plotting –dF/dT against temperature for the entire melting experiment.
Standard block cycler PCR amplification was performed with primers KY18/K75 or 533F/1492R on a GeneAmp PCR system 9600 thermocycler (Applied Biosystems, Foster City, Calif.). Initial denaturation was performed at 95°C for 15 min to activate Taq polymerase. This step was followed by 40 thermal cycles of 95°C for 20 s, 62°C for 20 s, and 72°C for 45 s. The 50-μl reaction mixtures contained 20 pmol of each primer oligonucleotide; dGTP, dATP, and dCTP (0.25 mM each); 0.75 mM dUTP; 3 mM MgCl2; 1.25 U of HotStarTaq DNA polymerase (QIAGEN); and 5 μl of nucleic acid in 1x HotStarTaq PCR buffer (QIAGEN).
COBAS AMPLICOR PCR. The Roche COBAS AMPLICOR PCR was performed in accordance with the manufacturer's instructions. Fifty microliters of nucleic acid was used for each reaction.
RESULTS
Sensitivity. A dilution series of DNA from M. tuberculosis H37Rv was assayed on the LightCycler in both 20- and 100-μl capillaries. Five experiments were performed in parallel for each dilution. The result is shown in Table 1. At a DNA concentration of 2.5 genomes per μl, two of five results were positive with the 100-μl capillary whereas none of the experiments was positive with the 20-μl capillary. At 5 genomes per μl, three of five results were positive with the 100-μl capillaries while one of five results was positive with the 20-μl capillaries.
The COBAS AMPLICOR assay could detect M. tuberculosis DNA down to 0.5 genome per μl, which corresponds to 25 genomes per assay (data not shown).
To investigate the influence of human DNA on the LightCycler assay, two identical dilution series of M. tuberculosis DNA were prepared, one with a diluent that contained 1.5 ng of human DNA per μl and the other with a diluent that lacked human DNA. Both dilution series gave the same result in the LightCycler assay, showing that the human DNA usually present in various amounts in clinical samples does not affect the sensitivity of the assay (data not shown).
Specificity of the assay. The specificity of the LightCycler assay was determined with a panel of different MTBC members, various NTM, and other bacteria and fungi. A real-time fluorescence signal was obtained only from members of the MTBC, i.e., M. bovis and M. africanum. All other investigated species were negative (data not shown).
The result of an experiment with M. tuberculosis, five different NTM, and two phylogenetically related, morphologically similar bacteria, i.e., C. pseudodiphtheriticum and N. brasiliensis, is shown in Fig. 2. A fluorescence signal was only obtained when M. tuberculosis DNA was the template (Fig. 2A, curve 9). Tm analysis showed a specific Tm of about 65°C for M. tuberculosis (Fig. 2B, curve 9). Melting peaks were also seen for M. avium and M. kansasii (Fig. 2B, curves 1 and 2), but since these Tms were below the temperature of the fluorescence acquisition step during PCR, those NTM did not produce a signal on the LightCycler amplification screen (Fig. 2A, curves 1 and 2). M. scrofulaceum also had a melting peak at about 58°C (data not shown). Agarose gel electrophoresis of the amplification products analyzed in Fig. 2 showed that all of the mycobacteria, as well as C. pseudodiphtheriticum and N. brasiliensis, were amplified (data not shown).
Performance of the IC. The optimal amount of ICO was determined in titration studies (data not shown). The goal was to keep the amount of ICO in the reaction mixture as small as possible to minimize competition with M. tuberculosis in low-titer samples, which could reduce the sensitivity of the assay. The optimal amount for 100-μl reaction mixtures proved to be 125 copies of the ICO (calculated from the concentration data provided by the manufacturer). In a dilution series of M. tuberculosis DNA, no significant difference in sensitivity could be seen between reaction mixtures with ICO and those without ICO (data not shown). Figure 3 shows an experiment with a fixed amount of M. tuberculosis DNA, a fixed amount of ICO, and a variable amount of inhibitor. NALC-NaOH solution was used as an inhibitor, since this reagent is used for initial decontamination of all clinical samples and traces of it might remain in inadequately washed DNA preparations. No fluorescence signal was obtained from reaction mixtures containing 1:5 and 1:10 dilutions of NALC-NaOH (Fig. 3A, curves 1 and 2). These samples were, however, clearly shown to be false negatives, since the ICO was not amplified in these reaction mixtures (as shown by the absence of an ICO-specific melting peak in curves 1 and 2 of Fig. 3B). In contrast, the negative control without inhibitor (Fig. 3B, curve 6) showed a strong ICO-specific peak at about 48°C. Reaction mixtures containing more dilute preparations of NALC-NaOH (Fig. 3, curves 3, 4, and 5) showed an increase in the M. tuberculosis-specific fluorescence signal, as well as an M. tuberculosis-specific melting peak at about 65°C. However, since the high concentration of M. tuberculosis outcompeted ICO amplification in sample 5, no ICO peak was seen in this sample (Fig. 3B, curve 5).
Evaluation with clinical samples. The LightCycler assay was evaluated with a variety of clinical samples that had previously been analyzed with the COBAS AMPLICOR assay (Table 2). All of the samples that initially tested positive in the COBAS AMPLICOR assay, i.e., 67 respiratory and 5 nonrespiratory specimens, were also positive in the LightCycler assay. Of the 73 samples that tested negative in the COBAS AMPLICOR assay (41 respiratory and 32 nonrespiratory), 72 were negative in the LightCycler assay.
For one bronchoalveolar lavage fluid sample that showed a positive smear, the result of culture analysis was M. malmoense. This sample was negative in the COBAS AMPLICOR assay but showed a weak positive signal in the LightCycler assay.
Because of inhibition, no valid COBAS AMPLICOR result could be obtained from one smear-positive, culture-positive stool sample. This sample, however, was clearly positive in the LightCycler assay.
DISCUSSION
We have developed a LightCycler assay for rapid detection of members of the MTBC. The forward primer and hybridization probes used in this assay target regions of the 16S rRNA gene similar to those targeted by the Food and Drug Administration-approved COBAS AMPLICOR assay. Since we intended to design an MTBC-specific assay, the forward primer was slightly extended to cover a more variable region of the 16S rRNA gene and produce more selective amplification. The reverse primer was shifted upstream to produce a shorter PCR product (206 bp for the LightCycler assay versus 584 bp for the COBAS AMPLICOR assay) because rapid-cycle PCR amplification works better when the PCR product is kept as short as possible. Since this region of the 16S rRNA gene lacked highly variable sequences, it was not possible to design a reverse primer that was more selective for the members of the MTBC.
Although the COBAS AMPLICOR assay is specific for the members of the MTBC, the assay will efficiently amplify the DNAs of all mycobacterial species (except M. simiae) (31). Despite several oligonucleotide mismatches, especially at the binding sites of the forward primer, the LightCycler assay also amplifies a corresponding segment within the 16S rRNA gene from most of the NTM species. However, only members of MTBC produced a real-time fluorescence signal.
Even during Tm analysis we had expected little or no signal from NTM since the hybridization probes bound to these samples should touch or even overlap, leading to steric problems and inefficient energy transfer. Interestingly, however, melting peaks were obtained for some NTM; these peaks could still be clearly differentiated from the MTBC peak by their lower Tm.
Inhibitors present in clinical samples, as well as inefficient extraction of nucleic acid, can cause false-negative results in amplification assays. Therefore, ICs should be required in all amplification assays used for diagnostic pathogen detection (14, 21).
For the COBAS AMPLICOR assay, a linearized, recombinant plasmid DNA serves as an IC (25). The plasmid contains the target sequence, which has been modified at the binding site of the detection probe. If the IC is successfully amplified, the amplicon will be detected with an IC-specific probe in a separate post-PCR hybridization reaction.
Similar ICs have been described for real-time amplification assays (10, 30, 33). In these, amplification of the IC was monitored with a separate probe that contained a fluorescent label different from that on the target-specific probe.
Recently, we have published a novel, simple technique for control of real-time amplification assays (2). In this technique, the ICs are single-stranded oligonucleotides that contain little more than the primer sites and a probe-binding site that contains mismatches to the actual target. An advantage of these ICOs is that they do not require separate control-specific detection probes. Because of the mismatches, probe-control hybridization does not occur during the fluorescence acquisition step of the PCR and therefore does not interfere with detection of the pathogen. However, ICO amplification can be detected during melting point analysis after PCR. Another advantage is that these ICOs can be obtained from many commercial providers of oligonucleotides, thereby eliminating the need for complicated, time-consuming, and expensive construction of plasmids (2). In the present study we have successfully applied this ICO technique to our LightCycler assay for the members of the MTBC. We have shown that the ICO does not reduce assay sensitivity but is able to detect impurities in the DNA preparation such as NALC, which inhibited amplification at concentrations as low as 0.8%.
Several in-house MTBC LightCycler assays have already been published (9, 12, 17, 19, 27). Recently, however, a meta-analysis of NAT used to diagnose tuberculous pleuritis showed significant heterogeneity in results from in-house ("home-brew") PCRs (23), while commercial assays did not show such heterogeneity. Since the DNA sample was prepared the same way for both assays, the commercial COBAS AMPLICOR test and our LightCycler assay could be directly compared with clinical samples. The agreement between the two assays was very high (100% sensitivity, 98.6% specificity). All samples positive in the COBAS AMPLICOR assay were also positive in the LightCycler assay, even though the COBAS AMPLICOR assay uses a larger volume (50 versus 25 μl) of sample DNA and is slightly more sensitive for analysis of DNA purified from cultured M. tuberculosis. The only discrepancy within 73 negative samples was one sample containing M. malmoense, which was negative in the COBAS AMPLICOR assay but produced a weak positive signal in the LightCycler assay. We cannot explain this result, since the LightCycler assay was clearly negative with DNA purified from a cultured M. malmoense strain.
When the sample is purified M. tuberculosis nucleic acid, the LightCycler assay appears to be slightly more sensitive when 100-μl capillaries are used instead of 20-μl capillaries. However, sensitivity tests based on purified DNA samples have only limited applicability to the assay of clinical samples. Mycobacteria usually are not evenly distributed in the sample material. In addition, cell clumping increases during the centrifugation steps required for decontamination and enrichment of clinical samples. Still, one can assume that increasing the volume of the initial sample is an important factor in increasing the sensitivity of the assay. Both the COBAS AMPLICOR assay and the LightCycler 100-μl assay allow relatively large sample volumes, i.e., 50 and 25 μl, respectively.
Compared with the COBAS AMPLICOR assay, the LightCycler assay is much more flexible. For example, if the sample contains an inhibitor the problem can usually be easily solved by repeating the PCR with a slight dilution of the sample DNA. With the COBAS instrument, it is almost impossible to do such a repeat assay on the same day as the initial assay. In contrast, since a LightCycler run takes only about an hour, rapid repeats are possible. Also, performing a COBAS AMPLICOR run with a single sample is very expensive since the assay still requires separate positive and negative controls. Because of the generally lower reagent costs, the LightCycler assay is more cost-effective, especially on runs with a single sample or only a few samples. It might even be possible to omit the positive control since the ICO can monitor the function of all of the components in the reaction mixture. In fact, when applying this strategy the efficiency of the nucleic acid extraction procedure would not be controlled. It is beyond dispute that coprocessing an artificial or real patient sample with a low number of M. tuberculosis cells would represent an optimal positive control. However, such a positive control is also not included in the COBAS AMPLICOR assay.
The use of commercial NAT assays to confirm TB, as well as their limited application to exclude a TB diagnosis, has already been discussed (23). Here we report a cost-effective in-house assay for fast and reliable detection of the members of the MTBC that compares favorably with the COBAS AMPLICOR assay. Since both assays use the same DNA preparation, the LightCycler assay can complement the latter assay, adding speed and flexibility to the routine diagnosis of TB infection.
ACKNOWLEDGMENTS
We thank M. Kühn and K. Stieber for providing cultures of M. tuberculosis and various bacterial strains and T. Becker for thoughtful review of the manuscript.
REFERENCES
Burggraf, S., H. Huber, and K. O. Stetter. 1997. Reclassification of the crenarchael orders and families in accordance with 16S rRNA sequence data. Int. J. Syst. Bacteriol. 47:657-660.
Burggraf, S., and B. Olgemller. 2004. Simple technique for internal control of real-time amplification assays. Clin. Chem. 50:819-825.
Cleary, T. J., G. Roudel, O. Casillas, and N. Miller. 2003. Rapid and specific detection of Mycobacterium tuberculosis by using the Smart Cycler instrument and a specific fluorogenic probe. J. Clin. Microbiol. 41:4783-4786.
Cole, S. T., R. Brosch, J. Parkhill, T. Garnier, C. Churcher, D. Harris, S. V. Gordon, K. Eiglmeier, S. Gas, C. E. Barry III, F. Tekaia, K. Badcock, D. Basham, D. Brown, T. Chillingworth, R. Connor, R. Davies, K. Devlin, T. Feltwell, S. Gentles, N. Hamlin, S. Holroyd, T. Hornsby, K. Jagels, and B. G. Barrell. 1998. Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393:537-544.
Cousins, D. V., S. D. Wilton, B. R. Francis, and B. L. Gow. 1992. Use of polymerase chain reaction for rapid diagnosis of tuberculosis. J. Clin. Microbiol. 30:255-258.
Del Portillo, P., L. A. Murillo, and M. E. Patarroyo. 1991. Amplification of a species-specific DNA fragment of Mycobacterium tuberculosis and its possible use in diagnosis. J. Clin. Microbiol. 29:2163-2168.
Desjardin, L. E., Y. Chen, M. D. Perkins, L. Teixeira, M. D. Cave, and K. D. Eisenach. 1998. Comparison of the ABI 7700 system (TaqMan) and competitive PCR for quantification of IS6110 DNA in sputum during treatment of tuberculosis. J. Clin. Microbiol. 36:1964-1968.
De Wit, D., L. Steyn, S. Shoemaker, and M. Sogin. 1990. Direct detection of Mycobacterium tuberculosis in clinical specimens by DNA amplification. J. Clin. Microbiol. 28:2437-2441.
Drosten, C., M. Panning, and S. Kramme. 2003. Detection of Mycobacterium tuberculosis by real-time PCR using pan-mycobacterial primers and a pair of fluorescence resonance energy transfer probes specific for the M. tuberculosis complex. Clin. Chem. 49:1659-1661.
Drosten, C., M. Weber, E. Seifried, and W. K. Roth. 2000. Evaluation of a new PCR assay with competitive internal control sequence for blood donor screening. Transfusion 40:718-724.
Eisenach, K. D., M. D. Cave, J. H. Bates, and J. T. Crawford. 1990. Polymerase chain reaction amplification of a repetitive DNA sequence specific for Mycobacterium tuberculosis. J. Infect. Dis. 161:977-981.
Heginbothom, M. L., J. T. Magee, and P. G. Flanagan. 2003. Evaluation of the Idaho Technology LightCycler PCR for the direct detection of Mycobacterium tuberculosis in respiratory specimens. Int. J. Tuberc. Lung Dis. 7:78-83.
Hermans, P. W., A. R. Schuitema, D. Van Soolingen, C. P. Verstynen, E. M. Bik, J. E. Thole, A. H. Kolk, and J. D. van Embden. 1990. Specific detection of Mycobacterium tuberculosis complex strains by polymerase chain reaction. J. Clin. Microbiol. 28:1204-1213.
Hoorfar, J., N. Cook, B. Malorny, M. Wagner, D. De Medici, A. Abdulmawjood, and P. Fach. 2003. Making internal amplification control mandatory for diagnostic PCR. J. Clin. Microbiol. 41:5835.
Kraus, G., T. Cleary, N. Miller, R. Seivright, A. K. Young, G. Spruill, and H. J. Hnatyszyn. 2001. Rapid and specific detection of the Mycobacterium tuberculosis complex using fluorogenic probes and real-time PCR. Mol. Cell. Probes 15:375-383.
Kubica, G. P., W. E. Dye, M. L. Cohn, and G. Middlebrook. 1963. Sputum digestion and decontamination with N-acetyl-L-cysteine-sodium hydroxide for culture of mycobacteria. Am. Rev. Respir. Dis. 87:775-779.
Lachnik, J., B. Ackermann, A. Bohrssen, S. Maass, C. Diephaus, A. Puncken, M. Stermann, and F. C. Bange. 2002. Rapid-cycle PCR and fluorimetry for detection of mycobacteria. J. Clin. Microbiol. 40:3364-3373.
Lemaitre, N., S. Armand, A. Vachee, O. Capilliez, C. Dumoulin, and R. J. Courcol. 2004. Comparison of the real-time PCR method and the Gen-Probe Amplified Mycobacterium Tuberculosis Direct Test for detection of Mycobacterium tuberculosis in pulmonary and nonpulmonary specimens. J. Clin. Microbiol. 42:4307-4309.
Miller, N., T. Cleary, G. Kraus, A. K. Young, G. Spruill, and H. J. Hnatyszyn. 2002. Rapid and specific detection of Mycobacterium tuberculosis from acid-fast bacillus smear-positive respiratory specimens and BacT/ALERT MP culture bottles by using fluorogenic probes and real-time PCR. J. Clin. Microbiol. 40:4143-4147.
Miller, N., S. G. Hernandez, and T. J. Cleary. 1994. Evaluation of Gen-Probe Amplified Mycobacterium Tuberculosis Direct Test and PCR for direct detection of Mycobacterium tuberculosis in clinical specimens. J. Clin. Microbiol. 32:393-397.
Nolte, F. S. 2004. Novel internal controls for real-time PCR assays. Clin. Chem. 50:801-802.
Noordhoek, G. T., S. Mulder, P. Wallace, and A. M. van Loon. 2004. Multicentre quality control study for detection of Mycobacterium tuberculosis in clinical samples by nucleic amplification methods. Clin. Microbiol. Infect. 10:295-301.
Pai, M., L. L. Flores, A. Hubbard, L. W. Riley, and J. M. Colford, Jr. 2004. Nucleic acid amplification tests in the diagnosis of tuberculous pleuritis: a systematic review and meta-analysis. BMC Infect. Dis. 4:6.
Reischl, U., N. Lehn, H. Wolf, and L. Naumann. 1998. Clinical evaluation of the automated COBAS AMPLICOR MTB assay for testing respiratory and nonrespiratory specimens. J. Clin. Microbiol. 36:2853-2860.
Rosenstraus, M., Z. Wang, S. Y. Chang, D. DeBonville, and J. P. Spadoro. 1998. An internal control for routine diagnostic PCR: design, properties, and effect on clinical performance. J. Clin. Microbiol. 36:191-197.
Shankar, P., N. Manjunath, K. K. Mohan, K. Prasad, M. Behari, Shriniwas, and G. K. Ahuja. 1991. Rapid diagnosis of tuberculous meningitis by polymerase chain reaction. Lancet 337:5-7.
Shrestha, N. K., M. J. Tuohy, G. S. Hall, U. Reischl, S. M. Gordon, and G. W. Procop. 2003. Detection and differentiation of Mycobacterium tuberculosis and nontuberculous mycobacterial isolates by real-time PCR. J. Clin. Microbiol. 41:5121-5126.
Sjobring, U., M. Mecklenburg, A. B. Andersen, and H. Miorner. 1990. Polymerase chain reaction for detection of Mycobacterium tuberculosis. J. Clin. Microbiol. 28:2200-2204.
Sritharan, V., and R. H. Barker, Jr. 1991. A simple method for diagnosing M. tuberculosis infection in clinical samples using PCR. Mol. Cell Probes 5:385-395.
Stocher, M., V. Leb, and J. Berg. 2003. A convenient approach to the generation of multiple internal control DNA for a panel of real-time PCR assays. J. Virol. Methods. 108:1-8.
Tevere, V. J., P. L. Hewitt, A. Dare, P. Hocknell, A. Keen, J. P. Spadoro, and K. K. Young. 1996. Detection of Mycobacterium tuberculosis by PCR amplification with pan-Mycobacterium primers and hybridization to an M. tuberculosis-specific probe. J. Clin. Microbiol. 34:918-923.
Wittwer, C. T., K. M. Ririe, R. V. Andrew, D. A. David, R. A. Gundry, and U. J. Balis. 1997. The LightCycler: a microvolume multisample fluorimeter with rapid temperature control. BioTechniques 22:176-181.
Zimmermann, K., and J. W. Mannhalter. 1996. Technical aspects of quantitative competitive PCR. BioTechniques 21:268-272, 274-279.(Siegfried Burggraf, Udo R)