When CDC is working on an outbreak of foodborne illness, it’s our goal to pinpoint the source as quickly and precisely as possible so we can prevent any further threat to public health. For several years now, the PulseNet Methods Development and Reference Unit (PMDRU) has been developing and validating new methods to complement pulsed-field gel electrophoresis (PFGE), which is currently the standard typing method used in outbreak detection. The method that we have found most helpful in detecting subtle differences between closely related bacterial strains is a technique known as multiple-locus variable-number tandem repeat analysis (MLVA).
Since MLVA is still a fairly new technique, most state health laboratories don’t yet have the capability to perform it themselves. They have to send their samples to CDC for testing. As a microbiologist in the PMDRU, my job is to perform the hands-on testing on the E. coli and Salmonella strains we receive from the states.
The recent outbreak of E. coli 0157:H7 related to eating raw cookie dough is a great example of how we’ve put this new technique to work. A week after our PulseNet colleagues identified the potential cluster of illness, we requested samples from around the country that fit the PFGE pattern we were investigating. Within two weeks it was clear that most of the state samples shared the same MLVA pattern as well.
Over the next few weeks, samples continued to come in and MLVA was run as soon as they arrived. Meanwhile, the epidemiologists at the local state health departments and CDC were busy conducting interviews on the case patients. Interestingly, a very unusual vehicle, raw cookie dough, emerged as a likely source for this outbreak.
On June 29, the Food and Drug Administration (FDA) notified us that they had been able to isolate E. coli O157:H7 from cookie dough and the request was made to send the strain to the CDC for immediate testing. We tested the isolate immediately upon arrival, only to discover that the MLVA pattern was quite different from the main outbreak pattern.
Even though we haven’t found our “smoking dough” yet, the information gathered by interviewing patients strongly supports the cookie dough theory. Unlike other older DNA fingerprinting techniques, MLVA helped us see more accurately which patients shared a common source for their illness.
More information about PulseNet and its role in foodborne outbreak detection can be found at http://www.cdc.gov/pulsenet/.