New technique reveals distinct E. coli O157:H7 populations.

A new genetic fingerprinting technique developed by University of Nebraska scientists (Department of Food Science and Technology, 330 Food Industry Complex, Lincoln, NE 68583) is revealing important insights about potentially deadly E. coli O157:H7 bacteria. The new technique shows that there are two genetically-distinct E. coli O157:H7 populations found in cattle: one that sometimes causes fatal food poisoning in people and a second that is not commonly isolated from food-poisoning cases. Previously, relationships of different E. coli O157:H7 populations weren't well-understood. The technique, called octamer-based genome scanning (OBGS), allows researchers to pinpoint where genetic differences exist on E. coli's DNA. It offers a way to rapidly clone and identify the genes at those DNA sites. Genetic results suggest that the population most commonly found in cattle either is nonvirulent and incapable of causing disease, or it is not easily transmitted to people. The new method gives a very high-resolution snapshot of the genome and allows scientists to see things they couldn't see before.

The researchers are fine-tuning and automating the OBGS method, making it faster and easier to use. They hope to develop a quick test for industry in the next few years. They've applied for patent protection. Perhaps the most practical application of the OBGS method will be in developing a more sensitive test of E. coli O157:H7 isolates. Scientists caution that researchers must be careful not to over-interpret the data at this point. The newly discovered population must be tested to determine if it is non-virulent.

The OBGS technique will be broadly applicable to many organisms. After spending eight months optimizing their innovation, scientists tested the technique on E. coli isolates that were drawn from humans and cattle. All the isolates originated in a three-county region of Wisconsin. Researchers expected that the isolates, when illustrated on an evolutionary tree-type structure called a dendrogram, would be dispersed throughout the structure. But most of the human isolates were clustered together in one part of the dendrogram, and most of the cattle isolates were clustered together, indicating that there were significant genetic differences between the two groups. So either their data were biased because all were collected from the same region, or the E. coli population infecting cattle in Wisconsin was distinct from the bacterial population making people sick.

To test for bias, investigators used the method on a large group of isolates gathered from 16 states. Again, two distinct populations-human and bovine-appeared on the dendrograms. The hypothesis: The cattle bacterial population either is very inefficient at causing human illness, or it is a weak population that does not survive the necessary hurdles to infect people and cause E. coli-related illness.

Data recently released by USDA/ARS officials found E. coli O157:H7 to be present in more than half of cattle tested. Only about 20,000 cases of human infection with E. coli O157:H7 are reported each year. One might expect a much higher number, given the number of cattle infected. In comparison, an estimated 2 million cases of Salmonella poisoning occur annually, many caused by eating contaminated poultry and beef.

The scientists' numbers suggest that two-thirds of the isolates from cattle are genetically related to the apparently nonvirulent population. They are undertaking a much larger sample to get a better idea of exactly what that number is. Scientists analyzed 78 E. coli O157:H7 samples to reach the current results. They plan to test hundreds to get a good fix on the numbers. Further information. Andrew Benson; phone: [phone]; fax: [phone]; email: [e-mail].