Detection and Identification of Mycobacteria by ... - Europe PMC

Vol. 30, No. 8

JOURNAL OF CLINICAL MICROBIOLOGY, Aug. 199,2, p. 2025-2028

0095-1137/92/082025-04$02.00/0

Copyright X 1992, American Society for Microbiology

Detection and Identification of Mycobacteria by Amplification of a Segment of the Gene Coding for the 32-Kilodalton Protein HANNA SOINI,l* MIKAEL SKURNIK,2 KARJ LIIPPO,3 EERO TALA,3 AND MATTI K. VILJANEN' Department of Turku, National Public Health Institute, I Department of Medical Microbiology, University of Turku,2 and Department of Diseases of the Chest, Turku University Central Hospital, 3 Turku, Finland Received 24 March 1992/Accepted 27 May 1992

A polymerase chain reaction (PCR) assay for the rapid detection of mycobacterial DNA is described. Oligonucleotide primers, derived from the sequence of a gene coding for the 32-kDa antigen of Mycobacterium tuberculosis, amplified DNA from all 28 species of mycobacteria tested. All nonmycobacterial species tested were negative. An oligonucleotide probe hybridized to the PCR products of the strains belonging to the M. tuberculosis complex. This method could detect as little as 50 fg, as tested with purified M. tuberculosis DNA. By this amplification method, 127 sputum specimens were tested, with 7.9%o of the specimens proving to be inhibitory in PCR. The sensitivity of detection by PCR compared with that by culture was 55.9%o; when the inhibitory specimens were excluded, the sensitivity was 70.4%. The specificity of PCR combined with hybridization was 1O00. MATERIALS AND METHODS Bacterial strains. The bacterial species studied are listed in Table 1. The mycobacterial species were from the strain collection of the Mycobacterial Reference Laboratory, National Public Health Institute, Turku, Finland. As indicated in the table, many species also included one or more strains originating from either the American Type Culture Collection or the National Collection of Type Cultures (Central Public Health Laboratory, London, United Kingdom). The nonmycobacterial strains were isolated from clinical specimens in the routine microbiological laboratory of the National Public Health Institute. Specimens. Expectorated sputum specimens were obtained from patients with pulmonary tuberculosis at the Department of Chest Diseases, Turku University Central Hospital. The presence of Mycobacterium tuberculosis in these specimens was confirmed by cultivation on Lowenstein-Jensen medium. Control sputa were obtained from patients with bronchiectasia and chronic bronchitis treated at the same department. Treatment of sputum specimens. The sputa were subjected to mucolytic treatment. For this purpose, a solution containing 0.1 M NaOH, 2 M NaCl, and 0.5% sodium dodecyl sulfate (SDS solution) was used. Equal volumes of this solution were added to each sample, the tubes were mixed until mucolysis was complete, and, finally, the specimens were centrifuged. The putative mycobacteria in the pellet were then lysed and DNA was extracted for PCR as described below. Preparation of DNA. Mycobacteria were lysed and their DNA was isolated by the procedure previously described (11). Briefly, bacteria were pelleted by centrifugation, and the pellet was resuspended in 300 ,u of SDS solution. After incubation at 95°C for 15 min with gentle mixing, 200 pul of 0.1 M Tris-HCl (pH 8) was added. The DNA was extracted with phenol-chloroform, precipitated with ethanol, and dissolved in water. Primers and probe. The selection of two primers and one probe was based on the gene sequence published by Borremans et al. (3). A 423-bp region of the gene was amplified by using 27-mer primers MV1 (5'-GGC-CAG-TCA-AGC-TTCTAC-TCC-GAC-TGG-3') and MV2 (5'-GCC-GTT-GCC-

During the past few decades, various methods have been tested for the rapid diagnosis of mycobacterial infections. So far, however, the demonstration of acid-fast organisms in clinical specimens has proved to be the only practical method in the clinical setting. Even so, the sensitivity of this method is yet unacceptable in that it does not allow identification of the mycobacterial species observed (1). The most recently proposed candidates for rapid diagnosis include labelled DNA probes (9, 14, 21), but they have not been demonstrated to be sensitive enough to detect mycobacteria directly in clinical specimens (19). Investigators have used a logical approach to this problem by amplifying mycobacterial DNA prior to detection with probes. The modem polymerase chain reaction (PCR) technique has been considered to be useful in this regard. The first mycobacterial target used in PCR was a segment of the gene coding for the 65-kDa mycobacterial heat shock protein (11). By using this method, promising results have been obtained (4, 5, 20). The fact that almost all living organisms have heat shock proteins that share large homologous areas (15) nevertheless increases the possibility of cross-reactions. Several other target genes have also been used in PCR for detection of mycobacteria. These include repetitive sequences (10, 23), 16S rRNA (2), antigen b (22), MPB 64 protein (16), and mtp 40 protein (7). As a target for PCR amplification, we have selected the gene coding for the 32-kDa secreted mycobacterial protein. Borremans et al. reported the cloning and sequence determination of this gene (3). The protein evidently is rather specific to mycobacteria and is a constitutive component of different mycobacterial strains (6). The aim of this study was to determine the occurrence of this selected target in different mycobacterial and nonmycobacterial species and, further, to obtain preliminary evidence of the usefulness of this PCR approach for direct determination of mycobacteria in clinical specimens.

*

Corresponding author. 2025

2026

SOINI ET AL.

J. CLIN. MICROBIOL.

TABLE 1. Amplification of a segment of the gene coding for the 32-kDa protein from mycobacterial species with primers MV1 and MV2 and reactivity of probe MV3 with the amplified product No. of strains

Species

Total no.

of strains

PCR

Probe

positive positive

Mycobacterium tuberculosis' Mycobacterium bovis' Mycobacterium bovis BCG Mycobacterium africanuma Mycobacterium microtia Mycobacterium marinuma Mycobacterium kansasiiP Mycobacterium simiae Mycobacterium scrofulaceuma Mycobacterium szulgaib Mycobacterium gordonaea Mycobactenium xenopiP Mycobacterium avium' Mycobacterium intracellularea Mycobacterium gastrib Mycobacterium malmoense" Mycobacterium nonchromogenicuma Mycobacterium terraeb Mycobacterium trivialea Mycobacterium fortuituma Mycobacterium cheloneia Mycobacterium ranaeb Mycobacterium borstelenseb Mycobacterium phleia Mycobacterium smegmatisa Mycobacterium flavescensa Mycobacterium vaccae' Mycobacterium asiaticuma

20 8 5 1 1 3 7 3 5 2 5 5 22 12 2 10 1 4 2 5 6 2 2 3 4 1 1 1

20 8 5 1 1 3 7 3 5 2 4 5 21 12 2 9 1 4 2 5 6 2 2 3 4 1 1 1

20 8 5 1 1 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0

a Prototype strain(s) from the American Type Culture Collection or the National Collection of Type Cultures or clinical isolates. b Clinical isolates only.

GCA-GTA-CAC-CCA-GAC-GCG-3'). The 21-mer probe MV3 (5'-GAC-CCG-CTG-TTG-AAC-GTC-GGG-3') was complementary with a region in the 423-bp sequence. According to the report by Borremans et al., M. tuberculosis and Mycobacterium bovis BCG have the most diverging sequences at this region (between amino acids 234 and 240). The oligonucleotides were synthesized by an automated DNA synthesizer (391 DNA Synthesizer PCR Mate; Applied Biosystems, Inc., Foster City, Calif.) on the basis of phosphoamidite chemistry. After synthesis, the oligonucleotides were cleaved from the support with ammonium (4 times 0.3 ml, 15 min), and the volume was adjusted to 3 ml. Then, each vial was incubated at 55°C for 18 h. After incubation, 1 ml of the oligonucleotide solution was aliquoted. Ammonium was evaporated out and the oligonucleotides were dissolved in water. The concentration of each oligonucleotide was determined spectrophotometrically at 260 nm. Amplification by PCR. Amplification was performed by using a DNA thermal reactor (HB-TR1; Hybaid Ltd., Middlesex, United Kingdom) and a GeneAmp kit with Taq polymerase (AmpliTaq DNA polymerase; Perkin-Elmer Cetus, Emeryville, Calif.). The reaction mixture of 100 ,ul contained 50 mM KCI, 10 mM Tris-HCl (pH 8.3), 1.5 mM MgCl2, 0.01% (vol/vol) gelatin, 200 ,uM deoxyribonucleotides, 20 to 50 pmol of oligonucleotides MV1 and MV2, 2.5 U of Taq polymerase, and various amounts of DNA extracted from the mycobacteria. In the thermal reactor, a total of 30 cycles of denaturation at 94°C for 1 min, annealing at

-423

1 2 3 4 5 6 7 8 9 10 FIG. 1. Ethidium bromide-stained 1.5% agarose gel containing mycobacterial DNA amplified with primers MV1 and MV2. Lanes: 1, control without mycobacterial DNA; 2, M. tuberculosis H 37 Rv; 3, M. tuberculosis ATCC 35825; 4, M. tuberculosis ATCC 35827; 5, M. bovis NCTC 10772; 6, M. avium ATCC 15769; 7, M. intracellulare ATCC 13950; 8, Mycobacterium kansasii ATCC 25100; 9, M. malmoense clinical isolate; 10, 100-bp ladder (GIBCO-Bethesda Research Laboratories).

55°C for 1 min, and synthesis at 72°C for 2 min were carried out. The tubes were allowed to cool at ambient temperature, and the reaction products were stored at 4°C. All samples were tested for inhibition. Five nanograms of M. tuberculosis DNA was added to the sample DNA, and these were amplified together as described above. The sample was considered to be inhibitory if no amplification product was observed. Analysis of PCR products. A 20-,ul volume of the reaction mixture was run in a 1.5% agarose gel. After being stained with ethidium bromide, PCR products were visualized and photographed under UV light. DNA from the gel was transferred to nylon membranes (GeneScreen Plus; DuPont, Boston, Mass.). Hybridization was carried out at 65°C for 18 h with the probe labelled at the 5' end with [,y-32P]ATP (5,000 Ci/mmol; Amersham, Amersham, United Kingdom). Membranes were washed twice at 20°C for 2 h and once at 65°C for 10 min. Finally, membranes were held for 24 h with radiographic films in cassettes with amplifying screens. The films were developed with an automated developer (Curix 60; Agfa). Commercial radioactive probes. Commercial radioactive probes (GeneProbe, San Diego, Calif.) were used to identify PCR-negative mycobacterial strains. RESULTS

Sensitivity of the PCR. When purified M. tuberculosis DNA was tested, a distinctive band of the expected size was seen by ethidium bromide staining of the gel with as little as 50 fg of DNA (data not shown). Reactivity of the primers with mycobacterial species. The 20 M. tuberculosis strains tested, including 4 strains from the American Type Culture Collection, gave a clearly positive reaction. Except for 1 strain of M. avium, one strain of M. gordonae, and 1 strain of M. malmoense, all 143 tested mycobacterial strains, from 28 species, were PCR positive (Table 1 and Fig. 1). All nonmycobacterial species tested were PCR negative (one Nocardia asteroides, one "Nocardia paratuberculosis" (Nocardia farcinica), one Bordetella parapertussis, one Bordetella pertussis, three Branhamella catarrhalis, two Enterococcus faecalis, five Escherichia

DETECTION OF MYCOBACTERIA BY PCR

VOL. 30, 1992

TABLE 2. Comparison of cultivation and PCR results in the

detection of M. tuberculosis in sputum specimens Cultivation results PCR Result

Positive Negative Inhibitory Total

Positive (n)

Negative (n)

19 8 7 34

84

Total (n)

25 92 31b0 93 127 6a

aThe amplified product obtained from these specimens did not hybridize with the oligonucleotide probe specific for M. tuberculosis complex. b These specimens were considered PCR negative.

coli, five Haemophilus influenzae, three Kiebsiella oxytoca, five Kiebsiella pneumoniae, one Proteus vulgaris, five Pseudomonas aeruginosa, one Pseudomonas fluorescens, five Staphylococcus aureus, five Staphylococcus epidermidis, five Streptococcus pneumoniae, eight Streptococcus group A, five Streptococcus group B, six Streptococcus group C, and five Streptococcus group G strains). Probe MV3 hybridized with all M. tuberculosis strains and, unexpectedly, also with the BCG strains and with other strains belonging to M. tuberculosis complex (Mycobacterium bovis, Mycobacterium africanum, and Mycobacterium microti). Moreover, one strain of Mycobacterium marinum, one strain of Mycobacterium intracellulare, and one strain of Mycobacterum chelonei gave a faintly positive result with the probe (Table 1). Clinical specimens. A preliminary assessment of the clinical sensitivity of the PCR was carried out by retrospectively testing 127 sputum specimens taken from 112 patients. M. tuberculosis had been cultivated from 34 of these specimens (Table 2). Of all specimens, 7.9% (n = 10) proved to be inhibitory in PCR. Of the 34 culture-positive specimens, 21% (n = 7) proved to be inhibitory. When these seven specimens were included, the sensitivity of the PCR method compared with that of the culture method was 55.9% (19 of 34); when they were excluded, the sensitivity was 70.4% (19 of 27, Table 2). Six specimens from five patients were positive by PCR detection but not by cultivation. None of these proved to be positive when the PCR product was studied with the probe recognizing bacteria belonging to the M. tuberculosis complex. M. intracellulare had been isolated earlier from one of these five patients. None of the six specimens had yielded atypical mycobacteria in cultivation. Therefore, the specificity of the PCR, combined with hybridization, in detection of M. tuberculosis was 100% (93 of 93, Table 2). The corresponding positive and negative predictive values were 100% (19 of 19) and 86.1% (93 of 108, three cultivation negative; inhibitory specimens were considered negative), respectively (Table 2). DISCUSSION The gene coding for the 32-kDa protein proved to be a promising target for PCR for detection of mycobacteria. The 423-bp segment was amplified from all mycobacterial species tested, strongly suggesting that the 32-kDa-protein gene is present in all mycobacterial species. Thus, our results significantly complement the earlier phenotypical studies of the prevalence of this protein among mycobacteria (6). This high prevalence, together with its antigenic dominance, renders the 32-kDa protein a very likely candidate as a diagnostic

antigen.

2027

The three strains for which no amplification was observed were definitely mycobacteria; they were clearly acid-fast organisms. Furthermore, the identity of two of them (Mycobacterium avium and Mycobacterium gordonae) was confirmed by commercial radioactive probes. The third strain, Mycobacterium malmoense, was identified not only by routine biochemical tests but also by the fatty acid profile analysis. These strains were also retested by PCR several times, and negative results were observed. A plausible explanation for the negative PCR results is that one or the other or both of the sequences to which the primers attach may have been altered by mutation in these strains. It is known that a single-base replacement in the target can abolish the binding of an oligonucleotide to that target. The selected primers showed excellent specificity when studied with a large panel of nonmycobacterial strains. The panel contained several respiratory pathogens and bacteria which can contaminate specimens taken for mycobacterial tests. Clinically, it is important to determine whether the patient is infected with M. tuberculosis or with other mycobacteria. For this purpose, an oligonucleotide probe recognizing a sequence in the amplified product was synthesized. On the basis of the sequence published by Borremans et al. (3), the probe should theoretically have been able to discriminate M. tuberculosis from BCG and from other mycobacteria as well. However, in our experiments, the probe only partially met these requirements. The probe could not differentiate M. tuberculosis from other species in the M. tuberculosis complex. Nevertheless, what is clinically more important is that the probe did not react with atypical mycobacteria except for three isolates, and these reactions were much weaker than those caused by M. tuberculosis complex. Except for M. tuberculosis, the other members of the complex are rarely of clinical significance. For a clinician, establishing that a mycobacterium from a patient specimen belongs to this complex essentially implies that the patient should be treated as having M. tuberculosis. The probe reacted consistently with our five BCG strains. Unfortunately, we could not test the BCG strain used in the work of Borremans et al. However, De Wit et al. later showed that the genes coding for the 32-kDa proteins are strongly conserved in M. tuberculosis and BCG. Strains compared by these investigators were identical except for a single nucleotide change (8). To determine the possible variations between the different mycobacterial species, the amplified DNA segments can be sequenced. On the basis of that information, specific probes can then be synthesized for clinically important mycobacteria. This can also be accomplished by using specific nested primers or restriction enzymes. TIhe sensitivity of the PCR was very good when tested with purified DNA ofM. tuberculosis. The level detected, 50 fg per test, corresponds to the DNA mass of some tens of microbes. The sensitivity of the PCR as a clinical test was preliminarily assessed with retrospective sputum material. The unacceptably lower sensitivity could have been due to several factors. The best portions of the specimens had already been used for cultivation and staining, leaving for this PCR analysis less-than-optimum test samples. Furthermore, DNA degradation may have occurred during the rather long storage of the specimens (1 to 12 months in -20°C) prior to DNA extraction. The most serious problem negatively affecting sensitivity was the presence of inhibitory activity in some specimens. Of all specimens, 7.9% proved to have inhibitory elements.

2028

SOINI ET AL.

Inhibition is an acknowledged phenomenon in PCR methodology. The prevalence of inhibition reported by BrissonNoel et al. (4) was about the same as that seen in our material. Unfortunately, in our study the majority (7 of 10) of the inhibitory specimens was among the culture-positive specimens. This suggests that infection due to M. tuberculosis may lead to the introduction of inhibitory substances into the sputum. It is known that, e.g., some blood components, such as heme, can inhibit the PCR (17, 18). Elimination of inhibitory factors continues to be one of the major challenges for the widespread acceptance and use of PCR methodology in clinical diagnostics. Several approaches have been reported to abrogate such inhibition (13). For example, specimens have been dialyzed or absorbed onto a column. These methods, however, usually require a considerable quantity of DNA or specimen. Therefore, they were not applicable in the present study. Six specimens which had been negative by cultivation were found to be PCR positive. Hybridization with the probe specific for M. tuberculosis complex indicated that these reactions were obviously due to atypical mycobacteria. M. intracellulare had been detected in an earlier specimen from one of the five patients. The cultivation system used was optimized for M. tuberculosis. Studies are currently under way to improve cultivation efficacy. Conventional medium will, of course, be tested. As well, specimens will also be cultivated on a medium with low pH, the conditions of which should be more conducive to the recovery of atypical mycobacteria (12). ACKNOWLEDGMENTS We thank Marja Katila for the fatty acid profile analysis of M. malmoense strains and Marita Kiijonen for excellent technical assistance. The language of this manuscript was revised by Jeri L. Hill. This study was supported by the Finnish Antituberculosis Foundation, the Tampere Tuberculosis Foundation, and the Foundation of Vaino and Laina Kivi. REFERENCES 1. Bates, J. H. 1979. Diagnosis of tuberculosis. Chest 76:757-763. 2. Boddinghaus, B., T. Rogall, T. Flohr, H. Blocker, and E. C. Bottger. 1990. Detection and identification of mycobacteria by amplification of rRNA. J. Clin. Microbiol. 28:1751-1759. 3. Borremans, M., L. De Wit, G. Volckaert, J. Ooms, J. De Bruyn, K. Huygen, J.-P. Van Vooren, M. Stelandre, R. Verhofstadt, and J. Content. 1989. Cloning, sequence determination, and expression of a 32-kilodalton-protein gene of Mycobacterium tuberculosis. Infect. Immun. 57:3123-3130. 4. Brisson-Noel, A., C. Aznar, C. Chureau, S. Nguyen, C. Pierre, M. Bartoli, R. Bonete, G. Pialoux, B. Gicquel, and G. Garrigue. 1991. Diagnosis of tuberculosis by DNA amplification in clinical practice evaluation. Lancet 338:364-366. 5. Brisson-Noel, A., B. Gicquel, D. Lecossier, V. Levy-Frebault, X. Nassif, and A. J. Hance. 1989. Rapid diagnosis of tuberculosis by amplification of mycobacterial DNA in clinical samples. Lancet i:1069-1071.

J. CLIN. MICROBIOL. 6. De Bruyn, J., R. Bosmans, J. Nyabenda, and J. P. Van Vooren. 1989. Effect of zinc deficiency on the appearance of two immunodominant protein antigens (32-kDa and 65-kDa) in culture filtrates of mycobacteria. J. Gen. Microbiol. 135:79-84. 7. Del Portillo, P., L. A. Murillo, and M. E. Patarroyo. 1991. Amplification of a species-specific DNA fragment of Mycobactenum tuberculosis and its possible use in diagnosis. J. Clin. Microbiol. 29:2163-2168. 8. De Wit, L., A. de la Cuvellerie, J. Ooms, and J. Content. 1990. Nucleotide sequence of the 32 kDa-protein gene (antigen 85 A)

of Mycobactenum bovis BCG. Nucleic Acids Res. 18:3995. 9. Drake, T. A., J. A. Hindler, G. W. Berlin, and D. A. Bruckner. 1987. Rapid identification of Mycobactenum avium complex in culture using DNA probes. J. Clin. Microbiol. 25:1442-1445. 10. 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. 11. Hance, A. J., B. Grandchamp, V. Levy-Frebault, D. Lecossier, J. Rauzier, D. Bocart, and B. Gisquel. 1989. Detection and identification of mycobacteria by amplification of mycobacterial DNA. Mol. Microbiol. 3:843-849. 12. Katila, M. L., and J. Mattila. 1991. Enhanced isolation of MOTT on egg media of low pH. APMIS 99:803-807. 13. Khan, G., H. 0. Kangro, P. J. Coates, and R. B. Heath. 1991. Inhibitory effects of urine on the polymerase chain reaction for cytomegalovirus DNA. J. Clin. Pathol. 44:360-365. 14. Kiehn, T. E., and F. F. Edwards. 1987. Rapid identification using a specific DNA probe of Mycobacterium avium complex from patients with acquired immunodeficiency syndrome. J. Clin. Microbiol. 25:1551-1552. 15. Lindquist, S., and E. A. Craig. 1988. The heat-shock proteins. Annu. Rev. Genet. 22:631-677. 16. Manjunath, N., P. Shankar, L. Rajan, A. Bhargava, S. Saluja, and Shriniwas. 1991. Evaluation of a polymerase chain reaction for the diagnosis of tuberculosis. Tubercle 72:21-27. 17. Mercier, B., C. Gaucher, 0. Feugeas, and C. Mazurier. 1990. Direct PCR from whole blood, without DNA extraction. Nucleic Acids Res. 18:5908. 18. Panaccio, M., and A. Lew. 1991. PCR based diagnosis in the presence of 8% (v/v) blood. Nucleic Acids Res. 19:1151. 19. Pao, C. C., S.-S. Lin, S.-Y. Wu, and W.-M. Juang. 1988. The detection of mycobacterial DNA sequences in uncultured clinical specimens with cloned Mycobacterium tuberculosis DNA as probes. Tubercle 69:27-36. 20. Pao, C. C., B. Yen, J..B. You, J.-S. Maa, E. H. Fiss, and C. H. Chang. 1990. Detection and identification of Mycobacterium tuberculosis by DNA amplification. J. Clin. Microbiol. 28:18771880. 21. Roberts, M. C., C. McMillan, and M. B. Coyle. 1987. Whole chromosomal DNA probes for rapid identification of Mycobacterium tuberculosis and Mycobactenium avium complex. J. Clin. Microbiol. 25:1239-1243. 22. 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. 23. Thierry, D., A. Brisson-Noel, V. Vincent-.Levy-Fr6bault, S. Nguyen, J.-L. Guesdon, and B. Gicquel. 1990. Characterization of a Mycobacterium tuberculosis insertion sequence, IS6110, and its application in diagnosis. J. Clin. Microbiol. 28:26682673.