Abstract
Gut colonization with enterotoxigenic Bacteroides fragilis (ETBF) appears to be associated with the development of colorectal cancer. However, differences in carriage rates are seen with various testing methods and sampling sites. We compared standard PCR, SYBR green and TaqMan quantitative PCR (qPCR) and digital PCR (dPCR) in detecting the B. fragilis toxin (bft) gene from cultured ETBF, and from matched luminal and faecal stool samples from 19 colorectal cancer patients. Bland-Altman analysis found that all three quantitative methods performed comparably in detecting bft from purified bacterial DNA, with the same limits of detection (<1 copy/mu l). However, SYBR qPCR under-performed compared to TaqMan qPCR and dPCR in detecting bft in clinical stool samples; 13/38 samples were reported positive by SYBR, compared to 35 and 36 samples by TaqMan and dPCR, respectively. TaqMan qPCR and dPCR gave bft copy numbers that were 48-fold and 75-fold higher for the same samples than SYBR qPCR, respectively (p < 0.001). For samples that were bft-positive in both fecal and luminal stools, there was no difference in relative abundance between the sites, by any method tested. From our findings, we recommend the use of TaqMan qPCR as the preferred method to detect ETBF from clinical stool samples.