Comparison of Detection Methods for Non-O157 Shiga Toxin-Producing Escherichia coli in 375 g Beef Trim

Introduction: Shiga toxin-producing Escherichia coli (STEC) strains of serotypes other than O157:H7 have emerged as an increasing concern to public health and the food industry. While O157:H7 remains the primary concern, six other serogroups (O26, O45, O103, O111, O121, and O145, referred to as the Big 6) account for approximately 70% of non-O157 STEC infections in the US from 1983-2002 (CDC). Current testing methods target two virulence factors (stx and eae) found in the Big 6.  Available commercial methods were developed allowing for industry to meet the requirements specified by USDA even though these virulence factors can be present in other E. coli serogroups.

Purpose: The objective was to compare commercial detection methods to the reference method (USDA MLG 5B.01) for detection of non-O157 STEC in 375 g beef trim using limit of detection (LOD). In addition, the goal was to evaluate isolation and confirmation protocols used in Europe (EU) compared to strategies provided by USDA MLG.

Methods: Each O group was analyzed separately in a series of inoculated studies. Beef trim intended for 80% lean ground beef production was sourced from a local supplier. Twenty-five (25) g samples were cut, weighed out, inoculated and stored at 4°C to allow adaptation to the product and to cold stress the cells. Three replicate samples at each of 7 dilution levels were inoculated.  Each 25 g inoculated sample was combined with 350 g un-inoculated trim and 1.5 l of appropriate pre-warmed enrichment broth and homogenized. Samples were incubated and tested according to the manufacturer’s protocol. A Most Probable Number (MPN, 3-tube) was calculated along with 95% confidence intervals. 

Results: Overall, the commercial platforms performed well, showing similar levels of sensitivity for detection of presumptive positives for O45, O26, O103 and O121 (screen results only). For O111, one method that utilizes an integrated immunomagnetic separation (IMS) and PCR approach was more sensitive than a PCR-only screen approach. Additionally, one commercial method showed more confirmed positives. There was difficulty detecting O111 with the USDA MLG 5B.01 method, likely due to the high novobiocin concentration in the enrichment media; this is consistent with industry feedback that O111 does not grow to the same levels as the other Top 7 STEC in media containing antibiotic in the same amount of time. The USDA method also was less sensitive for O103 compared to the three commercial methods.

Significance: Use of an IMS tool, such as antibody-coated beads, aided considerably with the confirmation protocol and is an important step when confirming suspect samples. Generally, results from the modified Rainbow plates were better than the EU confirmation plates (MacConkey with or without rhamnose). In summary, detection of non-O157 STEC in 375 g beef trim can be performed by any of the three commercial methods evaluated in this study.