Detection of Mycobacterium Tuberculosis in Clinical Samples by Polymerase Chain Reaction

Abstract

The fact that PCR is not dependent on the growth of organisms represents a considerable advantage regarding sensitivity and rapidity of the detection of slow growing organisms as compared to conventional methods. One of the major problems encountered in the application of the PCR for the detection of mycobacteria in clinical samples is the lysis of bacterial cells. We performed PCR on 43 culture positive clinical samples using primers specific for the M. tuberculosis complex developed by Eisenach et al (1990), targeting the multicopy insertion sequence IS6110. DNA suitable for amplification was liberated after a combined treatment which included sonication, incubation in the presence of a detergent and heat. All samples were treated with 0.5 U uracyl-N-glycosylase prior to amplification using dUTP instead of dTTP. The specificity of the PCR was 100%, no amplicons of 123 bp were found when other mycobacteria were detected using conventional methods. In one case we observed a band in a sample from which M. chelonae was isolated. However, this band differed in the molecular size as well as in the restriction fragments generated after Sal I digestion and it hybridized only very weakly with the labeled probe. The sensitivity of our system was 92% for microscopy positive samples and 60% for microscopy negative and culture positive samples. We believe that PCR provides reliable, accurate, and rapid results in the diagnosis of M. tuberculosis suitable for a routine application in a specialized laboratory. However, it should be combined with cultural procedures.