Detection of Shigellae from Stools of Dysentery Patients by Culture and Polymerase Chain Reaction Techniques

Abstract

In Bangladesh, the isolation rates of Shigella spp. range from 11% to 12% by the conventional culture technique. Since the sensitivity of this technique is low, the polymerase chain reaction (PCR) technique was used for detecting small number of Shigellae from patients’ stools. Sensitivity and specificity of the two techniques were also compared. Stool samples were collected from 41 patients with dysentery who attended the Clinical Research and Service Centre of the ICDDR,B: Centre for Health and Population Research. All stool specimens were directly plated onto MacConkey, Salmonella-Shigella, Xylose lysin deoxycholate and Hectoen enteric agar media, and Shigellae were detected following standard procedures. DNA was extracted from the stool samples, and the target sequence of invasive plasmid antigen (ipa)H locus was amplified by PCR with 130 ng each of two primers (primer H8 [5'-GTTCCTTGACCGCCTTTCCGATAC-3'] and primer H15 [5'-GCCGGTCAGCCACCCTA-3']) following standard procedures. The amplified product was hybridized using an ipaH probe. The isolation rates of Shigella dysenteriae type 1, S. flexneri, S. sonnei, and S. boydii were, respectively, 17.1%, 19.5%, 4.9% and 2.4% by the conventional method. The results of the PCR technique showed that 700 bp fragment was generated in 18 of the 18 culture-positive and in 7 of the 23 culture-negative stools. One hundred twenty-three strains of Escherichia coli were also tested by PCR for identifying the enteroinvasive E. coli, but none of them yielded any positive result. This study showed that the sensitivity of the culture technique is 72% and specificity is 100%, when the PCR technique was considered as gold standard. Therefore, the PCR may be considered a more sensitive and specific technique than the conventional culture technique and has the potential to be employed in routine diagnosis of dysentery in clinical centres as well as in epidemiologic studies.