Differentiating enteric Escherichia coli from environmental bacteria through the putative glucosyltransferase gene (ycjM).

This study is to tackle the challenge posed by the "naturalized" Escherichia coli population against the worldwide practice of E. coli-based water quality monitoring. In the literature, the putative glucosyltransferase gene (ycjM) of E. coli has been identified in silico to be one of the 114 genes specific to enteric E. coli. Based on the sequence of E. coli K-12 MG1655, a PCR assay (ycjPCR) targeting ycjM was developed in this study. As demonstrated by the ycjPCR assay using 367 E. coli strains isolated from animal feces, 97.2% of the isolates carried the ycjM with variations from 93.9% to 100% among nine different host sources, but none of the 17 strains of non-E. coli bacteria and only 23.0% of the environment-isolated cryptic Escherichia strains contained the ycjM. These data experimentally confirmed ycjM to be enteric specific. Our study also showed that the ycjPCR assay was superior to the commonly used tuf- or uidA-based PCR methods in differentiating enteric E. coli from ß-D-glucuronidase-positive environmental bacteria. Furthermore, study on 190 E. coli isolates from water samples, using EPA Method 1603 followed by bacterial identification with Biolog MicroStation™ and ycjPCR assay, indicated that the prevalence of ycjM in the E. coli water isolates had a significant (p < 0.05, odds ratio ) spatial variation from 69.6% to 93.8%. These data suggest that E. coli profile using EPA Method 1603 or other ß-D-glucuronidase-activity-based methods may need further analysis using the ycjM profile to accurately determinate fecal pollution in water.