HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Denkin, S. Articles by Zhang, Y. Search for Related Content PubMed PubMed Citation Articles by Denkin, S. Articles by Zhang, Y. Agricola Articles by Denkin, S. Articles by Zhang, Y.

Microarray-based pncA genotyping of pyrazinamide-resistant strains of Mycobacterium tuberculosis

¹Department of Molecular Microbiology and Immunology, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, Maryland 21205, USA ²Center for Biologics Evaluation and Research, Food and Drug Administration, Kensington, MD 20895, USA

Correspondence Ying Zhang yzhang{at}jhsph.edu

Received 19 April 2005
Accepted 12 August 2005

Drug-resistant Mycobacterium tuberculosis poses a significant threat to the treatment of tuberculosis (TB). The current susceptibility testing for the first-line TB drug pyrazinamide (PZA) is not only time-consuming but also difficult, due to the requirement for acid pH for drug activity. Predominantly, resistance to PZA in M. tuberculosis is caused by mutations in the pncA gene, and the detection of pncA mutations can be an indicator of PZA resistance. In this study, the use of a previously developed microarray method for the rapid detection of PZA-resistant M. tuberculosis based on identifying mutations in the pncA gene was evaluated. Microarray analysis was performed in a blind manner on 33 clinical isolates of M. tuberculosis for which the sequence of the pncA gene had not previously been determined. The results showed that all mutations in PZA-resistant strains identified by DNA sequencing could be unambiguously detected by the microarray method. It is concluded that the microarray method is a valuable tool for the rapid screening and genetic identification of potential PZA-resistant M. tuberculosis strains.

	INTRODUCTION

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Tuberculosis (TB) is a leading infectious disease with about eight million new cases and two million deaths per year worldwide (Raviglione, 2003). The prevalence of TB is increasing in some parts of the world due to HIV infection, which weakens the host immune system and allows latent TB to reactivate (Corbett et al., 2003). Both HIV infection and the emergence of drug-resistant strains pose significant threats to the control of the disease (Raviglione, 2003; Corbett et al., 2003). Rapid identification of drug-resistant strains is important for effective monitoring of drug-resistant Mycobacterium tuberculosis and can provide useful clinical guidance for appropriate treatment of the disease.

Pyrazinamide (PZA) is an important first-line drug that shortens TB therapy (Zhang & Mitchison, 2003). Mutations in the pncA gene encoding nicotinamidase/pyrazinamidase (PZase), which is involved in the conversion of PZA to the active form, pyrazinoic acid (POA), constitute the major mechanism of PZA resistance in M. tuberculosis (Cheng et al., 2000; Scorpio & Zhang, 1996; Scorpio et al., 1997b; Zhang & Mitchison, 2003). The current PZA susceptibility testing method is difficult (Hewlett et al., 1995; Zhang & Mitchison, 2003) due to the requirement of an acidic pH for drug activity (McDermott & Tompsett, 1954). Although the introduction of the BACTEC method (Siddiqi, 1992) has improved PZA susceptibility testing, false resistance can still be a problem in some cases with the resistance cutoff of 100 µg PZA ml^–1 at pH 6.0 currently used (Morlock et al., 2000; Zhang & Mitchison, 2003). In addition, while the BACTEC method has reduced the time taken for drug-susceptibility testing, it is still dependent on the slow growth of M. tuberculosis, which takes about 1–2 weeks.

Molecular testing of mutations in the genes associated with drug resistance has the advantage of being rapid and of eliminating the need for time-consuming phenotype-based susceptibility testing. Although various molecular methods, such as PCR-single-strand conformation polymorphism (SSCP) (Scorpio et al., 1997b), dideoxy fingerprinting (Felmlee et al., 1995), heteroduplex formation (Thomas et al., 2001) and amplification refractory mutation system (ARMS)-PCR (Fan et al., 2003), have been developed for rapid screening of potential drug-resistant mutants, these techniques are still tedious and do not demonstrate the required sensitivity or high-throughput sample-screening capability. A major drawback of the above methods is that they can only be used for detecting known mutations in a defined site or region, and are not sensitive enough to be used for detecting unknown mutations in a target gene. To date, the most accurate and reliable method for mutation detection is DNA sequencing, which can be expensive and challenging if multiple genes are involved in resistance, or if resistance mutations are not clustered in the target gene.

Hybridization of DNA samples to miniature glass microchips containing oligonucleotide probes has been used in a variety of genomic studies (Chizhikov et al., 2001; Volokhov et al., 2002). This technique allows for the analysis of several genetic markers in one simple hybridization. Microarrays composed of short oligonucleotide probes have been demonstrated to be a valuable tool for detecting minor genetic changes in the microbial population (Cherkasova et al., 2003). Such an approach has previously been used for the successful detection of rpoB mutations in a small region of 81 bp responsible for rifampicin resistance in M. tuberculosis (Gingeras et al., 1998; Troesch et al., 1999). However, use of this technology for the detection of other, more challenging drug resistances in TB has not been reported. We have recently developed a sliding-frame oligonucleotide microarray method for the rapid detection of PZA-resistant M. tuberculosis strains, based on mapping pncA mutations that cause PZA resistance (Wade et al., 2004). This microarray method requires a much smaller number of oligoprobes (80 versus over 5000 in the resequencing method) and is more practical and economical. In a recent preliminary study, using strains that had previously been characterized for pncA mutations, we showed that this microarray method has promise for the detection of pncA mutations that lead to PZA resistance (Wade et al., 2004). In the present study, we further evaluated this method in parallel with DNA sequencing for uncharacterized PZA-resistant M. tuberculosis strains. The results indicate that this microarray method shows 100 % concordance with DNA sequencing for the detection of pncA mutations involved in PZA resistance.

	METHODS

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Bacterial strains, PZA susceptibility testing, and PZase activity assay.

PZA-resistant M. tuberculosis clinical isolates were obtained from the New York State Department of Health, Albany, NY, and from the Veterans Administration Medical Center, West Haven, CT. Nine strains were obtained in the form of live cultures from the Veterans Administration Medical Center, whereas 24 strains from the New York State Department of Health were obtained in the form of heat-killed bacterial cells. PZA-resistant M. tuberculosis strains were identified by using the pH 6.0 liquid medium in the BACTEC radiometric method with a PZA concentration of 100 µg ml^–1 as a cutoff for resistance (Siddiqi, 1992). PZA resistance was defined as resistance to at least 100 µg PZA ml^–1 for the BACTEC method at pH 6.0 and as 50 µg PZA ml^–1 for the 7H11 agar method at pH 5.5. For PZA susceptibility testing, the susceptible strain M. tuberculosis H37Rv and PZA-resistant M. bovis BCG Pasteur were included as susceptible and resistant controls, respectively. PZase activity was assayed using the radioactive [¹⁴C]PZA method, as described by Zhang et al. (1999). A PZase-positive culture (PZA-susceptible H37Rv) and a PZase-negative culture (BCG Pasteur) were included as controls for the PZase assay.

Bacterial genomic DNA isolation, PCR, and DNA sequencing.

Genomic DNA was isolated as previously described (Zhang et al., 1992). Two primers, Myc-Forw (CTGCCGCGTCGGTAGGCAAACTGC) and T7-Myc-Rev (CGTTAATACGACTCACTATAGGGCCAACAGTTCA TCCCGGTTCGGC), positioned 31 bp upstream and 7 bp downstream of the M. tuberculosis pncA gene, were used for PCR amplification of the entire pncA gene. The reverse PCR primer contained the T7 RNA polymerase promoter sequence (underlined) at the 5' end for the sequential synthesis of the single-stranded RNA (ssRNA) probe. The standard PCR mixture (30 µl) contained 1.5 units HotStarTaq DNA polymerase, 1x the recommended buffer supplemented with 2.5 mM MgCl₂ (Qiagen), 500 nM each of forward and reverse primers, 200 µM each dNTP (dATP, dGTP, dCTP and dTTP) and 1 µl DNA template (0.1 µg DNA template). PCR was performed using a Gene Amp PCR System 9700 thermocycler (Applied Biosystems) with the following cycle conditions: initial denaturing at 95 °C for 10 min followed by 40 cycles at 94 °C for 40 s, 55 °C for 40 s, 72 °C extension for 40 s, and the final extension at 72 °C for 10 min. The presence of amplified PCR products was detected by 1 % agarose gel electrophoresis followed by ultraviolet detection after ethidium bromide staining. To determine the pncA sequence, PCR products containing the pncA gene were purified from the agarose gel after electrophoresis using a gel-purification kit (Qiagen) according to the manufacturer's instructions. The PCR products were directly sequenced using an ABI 377 automatic DNA sequencer (Applied Biosystems) and forward and reverse primers as described by Scorpio et al. (1997a).

Microarray detection of PZA-resistant M. tuberculosis.

The design of oligonucleotide probes for scanning-frame microarray, microchip design and fabrication, hybridization conditions, and microarray scanning and image analysis were as described in detail previously (Wade et al., 2004). Briefly, overlapping oligonucleotide probes were designed on the basis of the nucleotide sequence of the complete pncA gene of M. tuberculosis (GenBank accession no. NC_000962). In total, 79 oligoprobes were designed for microarray analysis of mutations in the pncA gene with appropriate physico-chemical properties (oligoprobe size from 14 to 20 nucleotides, GC content 47–86 %, and predicted melting temperature 49–55 °C). An additional oligoprobe complementary to the forward primer was used to monitor the synthesis of full-length ssRNA. The microarray containing 79 pncA-specific oligoprobes was spotted in quadruplicate using a contact microspotting robotic system PIXSYS 5500 (Cartesian Technologies) equipped with a microspotting pin CMP7 (ArrayIt). Quadruplication of each oligoprobe on the microchip ensured the quality of microchip fabrication and enabled the application of statistical methods to the analysis of the microarray data. ssRNA samples for microarray analysis were synthesized by T7 polymerase-driven transcription of the amplicons using the MEGAscript T7 High Yield Transcription kit (Ambion). RNA transcription was conducted in a 30 µl reaction mixture containing 2 µl MEGAscript T7 Enzyme Mix, 1x reaction buffer, 5 mM of ATP, UTP, CTP and GTP, and 0.1–0.5 µg of the pncA PCR product as template. After 2 h incubation at 37 °C, the unincorporated NTPs were removed by purification through Centrisep spin columns (Princeton Separations) according to the manufacturer's protocol. The MICROMAX ASAP RNA Labelling kit (Perkin Elmer) was used to incorporate Cy3 fluorophore into the ssRNA. Fluorescently labelled ssRNA samples were purified using the Centrisep spin columns, dried under vacuum, and solubilized in the supplemented MICROMAX Hybridization Buffer III at a final concentration of 0.5–1.0 µM. Hybridization between fluorescently labelled ssRNA samples and microarray oligoprobes was performed as described previously (Wade et al., 2004). The fluorescent images of processed microarray slides were generated using a ScanArray 5000 (Perkin Elmer) equipped with two lasers operating at 632 nm (for excitation of the Cy5 dye) and 543 nm (for excitation of the Cy3 dye). The position(s) of mutation(s) in pncA were identified by comparing the intensities of fluorescent signals from each array element measured for the reference wild-type pncA gene and for the potentially mutated gene, where the ratios of the signals were normalized using a linear regression model. Thus, an alteration of more than twofold in the hybridization signal ratio between the wild-type (sensitive) control strain and the mutant strain indicates a mutation in pncA.

	RESULTS AND DISCUSSION

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Since our previous study demonstrated that the sliding-frame microarray method that we developed was able to identify PZA-resistant strains that had previously been characterized and were known to contain pncA mutations (Wade et al., 2004), in the present study we tested if the microarray method can be used to detect PZA resistance using 33 PZA-resistant clinical isolates that have not been characterized for pncA mutations. Coded DNA samples from the 33 M. tuberculosis isolates (Table 1) were analysed by microarray analysis and by DNA sequencing simultaneously. Microarray analysis identified 19 out of the 33 PZA-resistant strains as having alterations in different positions of the pncA gene, as judged by changes in the array hybridization signal ratio (Table 1). As expected, the sensitive control strains H37Ra and H37Rv had no alterations in pncA.

DNA sequence analysis of the 33 PZA-resistant strains indicated that 19 of the 33 strains (58 %) had various types of mutations in the pncA gene, including substitutions, deletions and insertions (Table 1). Novel types of mutations were identified, when compared with the pncA mutation profiles of previously characterized PZA-resistant strains, that have not previously been observed (Wade et al., 2004). For example, strain 11 had a 9 bp deletion at nucleotide position 383, which removes three amino acids, Val-Asp-Val, from its PncA. Other novel mutations include a G insertion at nucleotide 288 in strain 2, and a G to A change at nucleotide position 485 causing the amino acid substitution Gly162Asp in strain 16. Strains 3, 19 and 28 were M. bovis strains that have the characteristic C to G change at nucleotide position 169, causing the amino acid substitution His57Asp. It is worth noting that the 19 strains that had pncA mutations were exactly the 19 strains that were identified by the microarray analysis, indicating a 100 % correlation between the microarray analysis and DNA sequencing.

We found that 14 of the 33 PZA-resistant strains (42 %) had no mutations in the pncA gene. This is somewhat surprising, as our previous studies have indicated that over 90 % of PZA-resistant strains have mutations in pncA (Scorpio et al., 1997b; Cheng et al., 2000). Of the 14 strains without pncA mutations, only four, strains 1, 5, 6 and 7, were obtained as live cultures, whereas the other 10 strains were obtained in the form of heat-killed bacteria. Since it is well known that PZA susceptibility testing is often problematic and since only four of the 14 strains were available as live cultures, we therefore evaluated the four PZA-resistant strains on 7H11 agar plates containing PZA (50 µg ml^–1) at acid pH (pH 5.5). Interestingly, all four strains turned out to be susceptible to 50 µg PZA ml^–1, as tested by the 7H11 agar method. In addition, we also performed a PZase assay for the four strains without pncA mutations, and these strains had apparently normal PZase activity (Fig. 1, lanes 3–6) compared to the susceptible control strain H37Rv (Fig. 1, lane 1) in converting [¹⁴C]PZA to POA.

View larger version (127K): [in this window] [in a new window]   Fig. 1. Detection of PZA conversion in M. tuberculosis clinical isolates. [14C]PZA was added to cultures of PZase-positive control M. tuberculosis H37Rv (lane 1), PZase-negative control M. bovis BCG (lane 2), and strains 1, 5, 6 and 7 (lanes 3–6, respectively) and allowed to incubate at 37 °C for 16 h. [14C]PZA conversion to [14C]POA in the culture supernatant was analysed by TLC. The radioactive PZA and POA are indicated by arrows.

In our previous studies, we have found that the few PZA-resistant strains without pncA mutations that are PZase-positive are usually low-level resistant or of intermediate susceptibility (Scorpio et al., 1997b; Cheng et al., 2000). Such strains may have enhanced efflux of POA, but the mechanism has not been identified. Given that few such strains are found among clinical isolates, the clinical significance of such strains is unclear, because of their low-level resistance (Zhang & Mitchison, 2003). It is quite likely that only strains that contain pncA mutations and are thus highly resistant to PZA may present clinically relevant resistance. If so, identifying the mutation in pncA by the microarray method would be a useful tool for rapid detection of PZA resistance. It is worth noting that our microarray method correctly identified all PZA-resistant strains containing pncA mutations and correlated well with the susceptibility testing by the 7H11 agar method. The molecular detection of pncA mutations using microarrays offers a rapid alternative approach to PZA-resistance detection.

Since interpretation of our microarray assay is based on the recognition of hybridization patterns, it perhaps suffers by the presence of sequence polymorphism in pncA, i.e. single silent point mutations. Two strains, 15 and 24, have the same single silent point mutation, 195CT, which probably represents the natural polymorphism of the gene but is not relevant to the resistant phenotype. While microarrays have the potential to significantly improve and simplify the detection and identification of mutation(s) in viral and bacterial gene(s), the broad application of this technology is still likely to be hampered by the relatively high cost of the assays. The current cost of microarray analysis is relatively high, but there are several potential approaches to reduce significantly the expense of microarrays and to make this technology affordable for the majority of clinical and public health laboratories. The use of simplified portable scanners and the development of new protocols to prepare fluorescently labelled samples for microarray analysis using less-expensive dyes are the main ways to achieve this goal. Moreover, the worldwide presence and ongoing development of DNA microarray facilities, which allow researchers at remote sites to submit biological samples and receive information by rapid customized analysis, could also potentially make the array detection of drug resistance more accessible and affordable.

We noted that this study analysed only a relatively small number of strains. Nevertheless, this study is encouraging, as we found that the microarray method is as reliable as DNA sequencing in identifying pncA mutations in PZA-resistant strains that have not previously been characterized for alterations in pncA (Wade et al., 2004); thus it may serve as a means for the rapid detection of PZA-resistant strains. Future studies are required to further evaluate the use of microarrays for the rapid detection of PZA resistance in a large of number of PZA-resistant strains, and to develop a single microarray chip for the detection of all clinically relevant drug resistances in M. tuberculosis.

	ACKNOWLEDGEMENTS

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
This work was supported by NIH grants AI44063 and AI/HL49485, and by the Natural Science Foundation of China (30328031). The receipt of clinical isolates from Linda Parsons, Bereneice Madison and Wendy Gross is gratefully acknowledged.

	REFERENCES

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES