Detection and identification of mycobacteria in sputum from suspected tuberculosis patients

Background Detection of Tuberculosis agent like nontuberculous mycobacteria (NTM) species by culture and microscopic methods remains difficult and time consuming. A fast and reliable diagnosis of tuberculosis would greatly improve the control of the disease. The purpose of this study is to compare the conventional multiplex PCR and multiplex PCR reverse cross blot hybridization assay to culture method in terms of mycobacteria species detection. Findings Among the 117 positively cultured samples, nontuberculous mycobacteria (NTM) species were found in 9 samples of multiplex PCR reverse cross blot hybridization assay; compared to only 3 NTM species found in our conventional multiplex PCR, and 13 NTM species were successfully identified among 162 negatively cultured samples compared to only 5 NTM species identification in conventional multiplex PCR results. Conclusions The sensitivity of the multiplex PCR reverse cross blot hybridization assay comparing to culture method was 86.03%, the specificity is 35.46%, the positive predictive value was 41.94% and the negative predictive value was 82.41%. For conventional multiplex PCR these values are 81.62%, 38.65%, 41.89%, 79.51%, respectively. Furthermore, in terms of mycobacteria species detection, the conventional multiplex PCR was relatively equal compared to the multiplex PCR reverse cross blot hybridization assay, and to be particularly having no significant discrepant results on the identification of Mycobacteria tuberculosis in both methods.


Background
Tuberculosis is caused by Mycobacterium tuberculosis, Mycobacterium bovis or Mycobacterium africanum. Together with Mycobacterium microti and the vaccine strain M. bovis BCG they belong to the Mycobacterium tuberculosis complex. Some other cases caused by nontuberculous mycobacteria (NTM), are mostly mycobacteria belonging to the Mycobacterium avium-Mycobacterium intracellulare complex [1]. Opportunistic mycobacteria commonly associated with the human immunodeficiency virus (HIV) are Mycobacterium kansasii, Mycobacterium xenopi, Mycobacterium fortuitum and Mycobacterium scrofulaceum [1]. An estimated 1.7 billion individuals are infected with Mycobacterium tuberculosis [2]. Mortality is highest in developing countries, where over three-quarters of cases occur [3].
Early detection is of major importance in the control of tuberculosis [4]. The emergence of multidrug resistant strains and its association with outbreaks on community in endemic areas illustrates that rapid diagnosis is essential [5,6]. A fast and reliable diagnosis of tuberculosis would greatly improve the control of the Tuberculosis [7]. Current conventional diagnosis of tuberculosis or other mycobacteria could be time-consuming, because the culture of mycobacteria may take 4 to 8 weeks. Some mycobacteria are very difficult or almost impossible to grow in vitro such as M. genavense and M. leprae [4]. Direct staining and microscopy of clinical samples lack sensitivity and specifity [1]. In principle, these drawbacks could be solved by an application of PCR, which allows in vitro amplification of target DNA to a detectable level within a matter of hours [8].
Various researchers have recently described the rapid detection of M. tuberculosis by PCR, and many have reported a high degree of sensitivity in detecting M. tuberculosis in clinical samples by means of DNA amplifications [8]. Recently a nested PCR has been developed in detecting Salmonella typhi in blood, feces and urine from suspect typhoid fever and multiplex PCR to detect M. tuberculosis complex bacteria and other mycobacteria which this technique is based on the amplification of the specific insertion sequence IS6110 and 16S rDNA respectively [9,10]. This study uses the multiplex PCR reverse cross blot hybridization assay and the conventional multiplex PCR to detect and identify the Mycobacterium species from clinical samples of patients suspected of mycobacterial diseases in comparison with the Conventional methods.

Methods
Three hundred and eighty-seven samples of sputum from patients suspected of mycobacterial disease were obtained from the lung hospital in Makassar, Indonesia. Microscopy and culture were performed according to the standard methods at the Department of Medical Microbiology, Molecular Biology and Immunology Laboratory. Ziehl Neelsen staining with some modifications was used for microscopic detection [11]. Sputum samples were decontaminated and cultured on Lowenstein Jensen medium, which is locally produced [12,13], after being extracted with Boom Method, the PCR assays were performed.

Ethical Approval
This study was reviewed and approved by Hasanuddin University and informed consent was obtained from all participants or their parents or their guardians.

Tailing of oligonucleotide probes with dTTP
The tailing reactions were performed with 200 pmol of the oligonucleotide probe. The probes were fixed to the membrane in a hybridization oven for 10 minutes. The membrane was washed twice with 10× SSC. The probes chosen for the identification of 16S rDNA and IS6110 PCR products (Table 1). On the rotary shaker for at least 5 minutes, the membrane was put in the cross blotter on the rubber mat, and a different mould with 34 slots [2 × 50 mm (numbered 0-33)] or a mould with three blocks of 34 slots (each 2 × 15 mm) was placed on top of it. The hybridized PCR product on the membrane was detected by incubation with streptavidin-alkaline phosphatase and a color substrate (4-nitroblue tetrazolium chloride and 5-bromo-4-chloro-3-indolylphosphate) according to the instruction of the manufacturer (Boehringer Mannheim, Germany).

Statistical analysis
Difference in the results between positive and negative groups for culture, microscopy, and conventional multiplex PCR in the same sputum samples and for multiplex PCR reverse cross blot hybridization assay results were analyzed using the SPSS (SPSS Inc., Chicago, Il) computer package.

Findings
The electrophoresis of conventional multiplex PCR and pattern of multiplex PCR reverse cross blot hybridization in nitrocellulose membrane The conventional multiplex PCR was set by specific primers to determine the insertion sequences of IS6110 and IS1245. Figure 1. shows representative of the DNA amplified products by conventional multiplex PCR and these amplicons of PCR product which were analyzed by electrophoresis on 1.8% agarose gels stained with ethidium bromide ( 15, 16, 18, 19, 23, 25, 27, 29), pGen1 (line 17), pSme3 (line 21), respectively. In positive control, the amplicons of mixed PCR product of mycobacteria was used (line 34).

The culture, microscopy and multiplex PCR reverse cross blot hybridization assay
The culture positive samples were divided into 117 positive and 19 negative multiplex PCR reverse cross blot hybridization results (table 2). From the 117 positive in culture samples, 108 were also positive in PCR reverse cross blot hybridization assay (for the probes pMyc5a, pt3 and pTub1 9 samples were found negative for the probe pMyc5a and pPt3) and 3 were positive pAvi7; 2 samples were positive pChe3 and 1 sample was positive in the probe pInt5, pInt7, pKan7, pFor1, pSme3, respectively (table 1). From the 251 negative culture results 162 were positive in multiplex PCR reverse cross blot hybridization assay (table 2). In 162 positive in multiplex PCR reverse cross blot hybridization assay, 149 samples gave positive results with the probes pMyc5a, pt3 and pTub1 except 13 samples were found negative for the probe pMyc5a and pPt3 and the other 5 samples were positive with the probe pAvi7; 3 samples gave positive result in pChe3 and 1 sample was positive in the probe pInt5, pInt7, pKan7, pFor1, pSme3, respectively (table  1). Sputum samples with microscopy positive were found also positive in sputum cultures and no significant statistical difference exists between microscopy and culture tests (p > 0.05) (table 2). The sensitivity of the Multiplex PCR reverse cross blot hybridization assay compared to culture method was 86.03%, the specificity was 35.46%, the positive predictive value was 41.94% and the negative predictive value was 82.41%.

Comparison between conventional multiplex PCR and multiplex PCR reverse cross blot hybridization assay results
From 117 positive cultured samples which then amplified with multiplex PCR reverse cross blot hybridization assay, it was found that 108 samples were M. . The sensitivity of the conventional Multiplex PCR comparing to culture method was 81.62%, the specificity was 38.65%, the positive predictive value was 41.89% and the negative predictive value was 79.51%. No significant difference was found in identification of M. tuberculosis by conventional multiplex PCR and multiplex PCR reverse cross blot hybridization assay results in both culture positive and negative samples.

Discussion
The results of this study show that multiplex PCR and reverse cross blot hybridization significantly are more sensitive than culture and microscopic methods to detect mycobacteria strain (p < 0.05) (  (table 4). Furthermore, no significant difference was found on the identification of M. tuberculosis by conventional multiplex PCR and multiplex PCR reverse cross blot hybridization assay results in both positive and negative samples of culture results. The low specificity of both PCR assays was presumably due to the higher positive result among negative culture of samples. This could be resulting from a high number of samples with fastidious or non cultivable mycobacteria content such as M. genavense and M. leprae, and also there was an evidence that some of the samples were positive in culture but negative on both of our PCR methods, this probably due to contamination of other bacteria that easily happened in culture methods.
In terms of mycobacteria species detection, the conventional multiplex PCR was relatively equal compared to the multiplex PCR reverse cross blot hybridization.  The conventional multiplex PCR and Multiplex PCR reverse cross blot hybridization assay should be suitable for a rapid and correct diagnosis of patients suspected of having mycobacterial disease. These two methods will help the clinicians significantly in deciding the suitable antimicrobial treatment for their patients within shorter period of time.