Improving the performance of an end-point PCR assay commonly used for the detection of Bacteroidales pertaining to cow feces

Bacteroidales are normal gut flora of warm-blooded animals. Since each host species carries a different diversity of Bacteroidales, the detection of host-associated gene markers of Bacteroidales has emerged as a promising tool for the tracking of the source of fecal pollution in aquatic ecosystems. To detect cow-associated Bacteroidales, a commonly used method has been an end-point PCR assay with the 16S rRNA genes primers CF128F (cow-associated) and Bac708R (all Bacteroidales). The PCR assay has demonstrated high rates of true-positive detection (i.e., high sensitivity) in all previous studies. However, the assay also had high rates of false-positive detection to the samples of non-target hosts in some cases (i.e., low specificity). In opposite to the reason many investigators have proposed, our results suggested that false detection was not necessarily due to the presence of the target sequence of CF128F in the feces of non-target hosts. Instead, we found sequences of non-target hosts having single internal mismatches with CF128F. Those mismatches were well tolerated in PCR, partly due to the universality of Bac708R. To improve the detection performance, we designed a novel primer CF592R (targeting the same clade of sequences as CF128F) to substitute Bac708R. The use of CF529R alleviated false detection and also led to a tenfold reduction in detection limit in the samples tested, compared to the use of Bac708R. Many other end-point PCR assays that detect the 16S rRNA genes in Bacteroidales also use a host-associated primer to couple with Bac708R, and low specificity or sensitivity has been reported. Based on our findings for CF128F, we suggest that the suitability of Bac708R in those PCR assays needs to be revisited.