Peptide Nucleic Acid-Mediated Competitive PCR Clamping for Detection of Rifampin-Resistant Mycobacterium tuberculosis

Abstract

Peptide nucleic acid-mediated competitive PCR clamping, which can selectively amplify mutant alleles, was developed to detect mutations in four codons (513, 516, 526, and 531) of the rpoB gene in Mycobacterium tuberculosis strains. This simple method successfully identified the mutations in all 40 of the M. tuberculosis strains tested.

The emergence of and increase in multidrug-resistant tuberculosis are serious threats to public health. A variety of genotypic methods that can detect the mutations in a 68-bp rifampin resistance-determining region in the rpoB gene (13-15, 17) have been reported for the rapid diagnosis of Rif^r Mycobacterium tuberculosis (4-9, 11, 18-20). These tests could have tremendous potential; however, it should be noted that their predictive values would be compromised when drug-resistant organisms are present as subpopulations among predominantly susceptible tubercle bacilli (1, 4, 16). If the subpopulation exceeds 1% of the total when tested against rifampin, successful therapeutic treatment of the patient may be compromised (3).

Here we describe the development of a peptide nucleic acid (PNA)-mediated competitive PCR clamping (PMCPC) technique for the detection of mutations in the rifampin resistance-determining region. An outline of this assay is depicted in Table 1. Because PCR clamping was done with wild-type complementary PNA probes, the mutant allele was selectively amplified and could easily be characterized by the presence of amplicon 1. The performance of the assay was evaluated for specificity, sensitivity, and subpopulation detection.

TABLE 1.

Schematic view of rpoB gene fragment targeted by PMCPC assays^a

^aShort arrows depict the primers, long double-headed arrows represent PCR fragments, either universal (100 bp) or mutant specific (517 bp). Interpretation of the results is summarized below the scheme. OPF, IPF, and OPR denote outer primer forward, inner primer forward, and outer primer reverse, respectively. The sequences of OPF and OPR are specific for the M. tuberculosis complex.

Forty M. tuberculosis strains (wild type, 16 strains; 531TTG mutant, 12 strains; 531TGG mutant, 2 strains; 526CGC mutant, 4 strains; 526TAC mutant, 2 strains; 526GAC mutant, 1 strain; 526GGC mutant, 1 strain; 516TAC mutant, 1 strain; 513CGA mutant, 1 strain) were used for evaluation of the PMCPC assay. Two outer primers (OPF [12] and OPR) and one inner primer (IPF), together with one of the wild-type allele-specific PNA probes, were used for the PMCPC assays (Table 2). The interpretation of the results is summarized in Table 1. The PCR mixture (final volume of 50 μl) consisted of genomic DNA (1 ng) and 10× buffer with 15 mM MgCl₂, 200 μM each deoxynucleoside triphosphate, enzyme mix (0.6 μl) (Expand high-fidelity PCR system; Roche Diagnostics GmbH), 20 pmol each of the OPF, OPR, and IPF primers, and 100 pmol of an allele-specific PNA probe. PMCPC reactions were performed with a GeneAmp PCR System 9700 (Applied Biosystems Japan Ltd., Tokyo, Japan) under the following conditions: initial denaturation at 95°C for 3 min; 45 cycles of 94°C for 30 s (denaturation), 67°C for 30 s (PNA hybridization), and 60°C for 30 s (primer annealing and extension); and a final elongation at 72°C for 7 min. The amplified fragments were detected by ethidium bromide staining of agarose gels. Mixtures of mutant and wild-type target DNAs containing 50, 20, 10, 4, 2, or 1% mutant DNA were prepared and used for subpopulation analysis by PMCPC (see Fig. 2).

TABLE 2.

Primers and probes used in this study

Primer or probe	Sequence
OPF	5′-CATGTCGGCGAGCCCAT
OPR	5′-CGATGTTGGGCCCCTCA
IPF	5′-GGTCTGTCACGTGAGCGT
PNA 513_16	NH2-GCCAATTCATGGACCA-CONH2
PNA 526	NH2-CCCACAAGCGCCGA-CONH2
PNA 531	NH2-CGCCGACTGTCGGC-CONH2

FIG. 2.

PMCPC assay of heterogeneous populations of rpoB mutant and wild-type (strain H37Rv) bacteria. Mixtures of mutant and wild-type target DNAs were as follows: 531TGG and wild-type DNAs (A), 526 TAC and wild-type DNAs (B), 516 TAC and wild-type DNAs (C), 513 CGA and wild-type DNAs (D). Mixtures of mutant and wild-type target DNAs were 50% (lane 1), 20% (lane 2), 10% (lane 3), 4% (lane 4), 2% (lane 5), and 1% (lane 6). The amount of DNA was kept at 2 ng per reaction.

The PMCPC assay successfully identified a variety of mutations in 40 M. tuberculosis strains (Fig. 1). With the addition of a PNA probe to the PCR mixture, the corresponding polymerase readthrough of the wild-type allele was hindered (absence of amplicon 1) whereas its amplification proceeded for the mutant allele (presence of amplicon 1). The 100-bp universally obtained fragments in all PCRs confirmed the success of the PCR (Fig. 1). The 517-bp fragments obtained by PCR in the absence of PNA probes indicated the presence of M. tuberculosis in the sample (Fig. 1A), where the specificity of this assay was confirmed with 20 species of nontuberculosis mycobacteria and 8 of nonmycobacteria (Table 3). The detection limit of the PMCPC assay was determined to be 100 fg of DNA (data not shown).

FIG. 1.

Profiles generated by the PMCPC assay targeting four rpoB codons: codon 531 with the PNA 531 probe (B), codon 526 with the PNA 526 probe (C), and codons 516 and 513 with the PNA 513_516 probe (D). (A) PMCPC assay in the absence of PNA probes. Lanes: 1, strain H37Rv; 2 and 3, strains with rpoB 531 mutant alleles (TTG and TGG); 4 to 7, strains with rpoB 526 mutant alleles (CGC, TAC, GAC, and GGC); 8, strain with an rpoB 516 mutant allele (TAC); 9, strain with an rpoB 513 mutant allele (CGA); 10, no template DNA; M, 100-bp DNA ladder.

TABLE 3.

Mycobacteria and nonmycobacteria used to determine the specificity of the PMCPC assay

Species	Straina
Mycobacteria
M. abscessus	ATCC 19977
M. avium	ATCC 25291
M. chelonae	ATCC 19237
M. diernhoferi	ATCC 19340
M. flavescens	ATCC 14474
M. fortuitum	ATCC 6841
M. gordonae	ATCC 14470
M. intracellulare	ATCC 13950
M. kansasii	ATCC 12478
M. lentiflavum	Clinical isolate
M. marinum	ATCC 927
M. nonchromogenicum	ATCC 19530
M. phlei	ATCC 11758
M. scrofulaceum	ATTC 19981
M. simiae	ATCC 25275
M. smegmatis	ATCC 14468
M. szulgai	ATCC 35799
M. triviale	ATCC 23292
M. tuberculosis	H37Rv
M. vaccae	ATCC 15483
M. xenopi	ATCC 19250
Nonmycobacteria
Corynebacterium flavescens	NBRC 14136
Corynebacterium xerosis	NBRC 12684
Klebsiella pneumoniae	NBRC 3318
Nocardia farcinica	NBRC 15532
Pseudomonas aeruginosa	NBRC 3080
Rhodococcus equi	NBRC 3730
Streptococcus equinus	NBRC 12553
Streptococcus salivarius	NBRC 13956

^aNBRC; National Institute of Technology and Evaluation Biological Resource Center.

It is noteworthy that complete suppression of PCR amplification of wild-type alleles could be achieved even when an excess amount of purified DNA (100 ng per reaction) was used. This proved that amplicon 1 is never obtained from wild-type alleles and there is no interference by predominantly existing wild-type DNA with the amplification of mutant allele-specific 517-bp fragments (amplicon 1). The mutant subpopulations, at a level of 2% of the total DNA in the sample, could be detected as a faint band representing amplicon 1 (Fig. 2). Sequence analysis confirmed that these faint bands were derived from the mutant and did not represent a false positive derived from an excess amount of wild-type DNA (data not shown). This potential of the PMCPC assay is favorable given that a drug-resistant subpopulation that exceeds 1% of the total is considered clinically important and represents one advantage of this assay over other genotypic methods of detecting drug-resistant M. tuberculosis. Further quantitative analysis, such as the use of densitometric data to provide physicians with more detailed information relating to the amount of mutant DNA in the wild-type DNA, would be valuable but requires more equipment and a heavier workload. Further investigations in this area are required.

The Line Probe assay (LiPA) is probably the most frequently used method at this moment (5, 7, 10). Its superiority over our assay is its ability to identify four specific mutations by R probes, which covers about 75% of the Rif^r M. tuberculosis strains. However, a drawback of the LiPA is that it failed to distinguish the TTC insertion mutation at position 514 from the wild-type sequence (10). Failure to detect insertional and deletional mutations in the region covered by PNA probes is not likely to happen with the PMCPC assay since the PCR clamping can be seen only when perfectly matched sequences (wild type) exist. In addition, the LiPA cannot distinguish mutants coexisting with the wild type, which are not targeted by R probes. PMCPC has a limitation with regard to its specificity; i.e., mutations in regions other than those covered by the three PNA probes, which are estimated at about 5% of Rif^r M. tuberculosis strains, fail to be detected. However, this is specific not only to the PMCPC assay but to all other molecular methods. A PNA probe costs about 50 times more than a DNA probe. The use of a PNA approach represents a rather new technology, and only a limited number of manufacturers supply PNA probes. When the use of PNA probes becomes more commonplace, it is anticipated that the associated cost will comparable to that of methods using DNA probes. Bockstahler et al. (2) first described the effectiveness of PNA probes in identifying katG and rpoB mutations in M. tuberculosis by the PNA probe hybridization-enzyme-linked immunosorbent assay method, which differs quite significantly from the approach taken in our assay.

The newly developed PMCPC technique is rapid and easy to perform with conventional PCR and agarose gel electrophoresis equipment, and the results are easy to interpret. Moreover, this assay can detect drug-resistant subpopulations with reasonable sensitivity. Our preliminary study also demonstrated the potential use of this assay for direct analysis of sputum specimens. Finally, we intend to evaluate the assay for direct analysis of clinical specimens with a large number of samples.