Test finds E. coli in beef faster, could better trace outbreaks

August 30, 2010

WEST LAFAYETTE, Ind. — Infrared spectroscopy can detect E. coli faster than current testing methods and can cut days off investigations of outbreaks, according to a study at Purdue University.

Lisa Mauer, an associate professor of food science, detected E. coli in ground beef in one hour using Fourier transform infrared spectroscopy, much less than the 48 hours required for conventional plating technology, which requires culturing cells in a laboratory. Mauer said spectroscopy could be done in the same laboratories, just in much less time.

The spectroscopy method also differentiates between strains of E. coli 0157:H7, meaning outbreaks could be tracked more effectively and quickly. Current tests are multistep and take almost one week to get results.

"Even with all the other bacteria present in ground beef, we could still detect E. coli and recognize different strains," said Mauer, whose findings were reported in the August issue of the Journal of Food Science.

Mauer demonstrated two methods for separating bacteria from ground beef for testing. An antibody-capture method, which binds bacteria to antibodies attached to magnetic beads, gave results in four hours. A filtration method achieved results in about an hour.

Infrared spectroscopy could detect as little as one E. coli cell if the bacteria was cultured for six hours. Conventional plating techniques used for E. coli detection require culturing cells for 48 hours.

E. coli has a specific infrared spectrum that can be read with a Fourier transform infrared spectrometer. Infrared light is passed over a sample. The spectrometer reads the spectrum created by the combination of energy that has been absorbed and energy that has been reflected back.

"Energy is only absorbed by certain components of a sample," Mauer said. "If that component or bacteria isn't there, the energy is reflected back."

About 70,000 Americans are sickened by E. coil each year, according to the Centers for Disease Control and Prevention. People become infected after ingesting food contaminated with the bacteria, which comes from human or animal feces. Symptoms include severe stomach cramps, diarrhea and vomiting, and in rare occasions the infection can be life-threatening.

Mauer's testing methods also can differentiate between living and dead E. coli cells, something current testing methods cannot.

"If the cells are dead, they're not harmful. But the presence of that dead population could tell you something about the quality of the product," Mauer said.

Mauer believes the ground beef tests show promise for using the technology to find other pathogens in additional types of foods. She has already shown that spectroscopy can detect melamine -- which sickened about 300,000 infants in China and killed at least six in 2008 -- down to one part per million in powdered baby formula.

Mauer next plans to assess spectroscopy for detection of more pathogens in different food products. The U.S. Department of Agriculture Agricultural Research Service and the Purdue Center for Food Safety Engineering funded the study.

Writer: Brian Wallheimer, 765-496-2050, bwallhei@purdue.edu

Source: Lisa Mauer, 765-494-9111, mauerl@purdue.edu

Ag Communications: (765) 494-2722;
Keith Robinson, robins89@purdue.edu
Agriculture News Page



Detection of Escherichia coli 0157:H7 from Ground Beef Using Fourier Transform Infrared (FT-IR) Spectroscopy and Chemometrics

Reeta Davis, Joseph Irudayaraj, Bradley L. Reuhs, and Lisa J. Mauer

FT-IR spectroscopy methods for detection, differentiation, and quantification of Escherichia coli 0157:H7 strains separated from ground beef were developed. Filtration and immunomagnetic separation (IMS) were used to extract live and dead E. coli 0157:H7 cells from contaminated ground beef prior to spectral acquisition. Spectra were analyzed using chemometric techniques in OPUS, TQ Analyst, and WinDAS software programs. Standard plate counts were used for development and validation of spectral analyses. The detection limit based on a selectivity value using the OPUS ident test was 105 CFU/g for both Filtration-FT-IR and IMS-FT-IR methods. Experiments using ground beef inoculated with fewer cells (101- 102 CFU/g) reached the detection limit at 6h incubation. Partial least squares (PLS) models with cross validation were used to establish relationships between plate counts and FT-IR spectra. Better PLS predictions were obtained for quantifying live E. coli 0157:H7 strains (R2 ≥0.9955, RMSEE = ≤0.17, RPD ≥ 14) and different ratios of live and dead E. coli 0157:H7 cells (R2 =0.9945, RMSEE =2.75, RPD=13.41) from ground beef using Filtration FT-IR than IMS-FT-IR methods. Discriminant analysis and canonical variate analysis (CVA) of the spectra differentiated various strains of E. coli 0157:H7 from an apathogenic control strain. CVA also separated spectra of 100% dead cells separated from ground beef from spectra of 0.5% live cells in the presence of 99.5% dead cells of E. coli 0157:H7. These combined separation and FT-IR methods could be useful for rapid detection and differentiation of pathogens in complex foods.