​

A TENNESSEE FOOD DISTRIBUTOR had to recall an entire batch of Jack cheese late in 2007 after pieces were found to be contaminated with Listeria monocytogenes, a bacteria that can cause serious and sometimes fatal infections in people with weakened immune systems. The good news in this case was that the recall resulted from a routine sampling program by the state’s Department of Agriculture, rather than a report of human infection. The bad news: The product was distributed between August 28 and November 19, 2007. By the time of the recall notification on November 21, consumers had already been exposed to a life-threatening pathogen from which some of them potentially could have died.

The case, a minor incident among many on the U.S. Food and Drug Administration’s Web site, shows both the promise of food-safety testing and its current limitations.

For testing to be effective, it must keep pace with today’s automated, ever-accelerating, RFID-tracked supply chains. In the 24 hours it could take to grow Salmonella culture in a laboratory Petri dish, that chicken potpie—mistakenly undercooked in a microwave—could be on someone’s dinner plate.

Both government regulators and industry see the need to accelerate testing while making it better at detecting even the smallest quantity of bacteria or harmful substance that could have been introduced accidentally or intentionally. Together researchers and entrepreneurs are pushing the scientific envelope to answer the need. They are enhancing traditional testing methods, such as those that rely on laboratory cultures, while also taking advantage of technology from other fields, including DNA splitting, fiber optics, and spectroscopy.

Out of the Petri Dish

Antigen-antibody testing is currently the most common method of assessing the presence of bacteria-based foodborne pathogens like E. coli, Salmonella, or Clostridium. Also called enzyme-linked immunosorbent assay (ELISA), the method uses a pathogen’s natural enemy, the antibody, to track it down and demonstrate its presence.

In ELISA testing, mass-produced antibodies for a given pathogen, like E. coli, are harvested from animals such as rabbits. Those antigens are then treated with fluorescent dye and mixed with a test sample. If the pathogen is present, individual antibodies will bind to the bacteria.

The combined elements are then rinsed with a chemical solution that eliminates loose antibodies. What’s left is exposed to fluorescent light. If the material glows, the test technician knows that the suspect pathogen is present because the fluorescent antibodies have bound to it.

A similar method relies on phages instead of antibodies. Phages are essentially unique viruses that can only bond to the outside of specific bacteria. If dyed, they can prove the presence of bacteria in the same way.

ELISA’s drawbacks, traditionally, have included varying degrees of sensitivity to low concentrations of pathogens. The test is more sensitive if a sample is “grown up” to increase the amount of pathogens, so results can take as long as 24 to 72 hours.

A newer means of conducting ELISA testing solves that problem. In this method, developed by Invitrogen of Carlsbad, California, microscopic metal beads, roughly 100 micrometers (1/10 mm) in diameter, are coated with the same E. coli antibodies used in normal ELISA tests. They are now “immunomagnetic” beads, or as the company calls them, Dynabeads.

For a test using Dynabeads, the sample’s test solution is dyed to fluoresce, and it is placed into a test container along with the Dynabeads. If a pathogen—in this case E. coli—is present, it will find and bind to the antibodies on the surface of the Dynabeads.

A magnet is then placed next to the wall of the test container, attracting the Dynabeads. A special rinse flushes out the container, leaving behind the Dynabeads, which are tested for flourescence. If they glow, the test is positive for E. coli.

The entire test takes less than 20 minutes, says Invitrogen spokeswoman Revelle Anderson. This technology has been selected for E. coli screening of food served to athletes at this year’s Beijing Olympic Games.

DNA revolution. By the 1990s, the revolutionary DNA-based polymerase chain-reaction (PCR) test, used most famously in crime forensics, came to the world of food-safety testing. PCR testing is often referred to as “genetic photocopying.” While cultural and ELISA tests multiply existing bacteria by growing them, in PCR tests, heat is used to break bacteria cells open and to split their internal DNA strands in two. As the cellular material is repeatedly heated and cooled, the DNA is multiplied by the millions. It is then heated in a final step so that its strands end up separate.

In a step similar to the ELISA test, the separated or “denatured” DNA strands are dyed. The sample DNA material is poured over the slide coated with split DNA strands from the subject pathogen. If pathogen DNA is present in the sample material, it will automatically bond to the corresponding material on the slide, and that spot on the slide will fluoresce.

Devices such as the GeneChip, pioneered by Affymetrix of Santa Clara, California, offer tests for multiple substances on one slide using PCR technology. The chip’s “microarrays” may contain thousands of different sample spots to test for different varieties of DNA. After PCR processing and application, the slide is placed in a machine that reads the fluorescence of each spot, determining the materials present in a sample.

Kevin King, a principal scientist at the Midwest Research Institute in Kansas City, Missouri, says that the method can currently produce results within eight hours. Eventually, they hope to see results within four or five hours, which would be well within the duration of a single work shift, King says.

PCR holds great promise for field-deployable real-time pathogen testing, which may come in about five years, says Jerry Kelly, a senior technical consultant with testing vendor Shuster Labs in Canton, Massachusetts.

Yet PCR technology has its limitations as well, chief among them, false positives. Even if a pathogen bacteria that was present within a food item has been destroyed, such as by pasteurization, its split DNA strands remain, and would produce a false positive in a PCR test.

To address that shortfall, as well as problems detecting pathogens quickly, food suppliers and testing vendors seek to bundle multiple technologies into a single test.

Marlene Janes, an associate professor of food safety and microbiology at Louisiana State University and a researcher with the school’s AgCenter, works closely with her state’s seafood industry toward the goal of rapid testing of raw shellfish for Vibrio parahaemolyticus and Vibrio vulnificus. The former can make a victim violently ill, while the latter kills half those it infects.

Janes’ goal is a combination of ELISA, phage test, and PCR, which, if effective, could determine whether pathogens are both present in a sample and still alive.

The “3-in-1” test would use immunomagnetic beads like Dynabeads to trap any bacteria in a sample. Next, phages would be introduced. If they find their counterpart bacteria, they would break open the cells as they entered the organism, releasing its internal mechanisms, including DNA strands. The freed DNA strands could then be denatured and duplicated using normal PCR procedures, and they could be identified using established methods, such as a slide or microarray.

Janes hopes that a 3-in-1 test will be available within two years. For now, she is focused on developing an effective, portable PCR testing device that can provide shellfishers results out on a boat within a few hours, she says.

Fiber optics. Researchers at the U.S. Naval Research Laboratory in Washington, D.C., have adopted a technology traditionally associated with telephony—fiber optics—to detect food pathogens, incorporating principles also at play in ELISA and PCR tests.

When used in communications, fiberoptic lines are essentially light pipes. The sound of a voice, for example, is translated into digital data, transmitted as light, then translated back into sound when it arrives at a faraway telephone.

Chris Taitt, a research biochemist at the Naval Research Laboratory, explains that the food-test aspect of fiber optics takes advantage of the fact that the tiniest bit of light leaks out of what is otherwise a closed environment—the pipe. The bit that escapes is called an evanescent wave. That wave can “excite” fluorescence in molecules coating the exterior of a fiberoptic device.

The fiber optics used in the Naval Research Laboratory’s fiber-optic biosensors are not so much pipes as tapered segments fitted to a test device. For the test, each segment is coated with a specific pathogen’s antibody. A food sample is then exposed to the biosensor’s optical fibrils. Next, a laser is shot into the fiberoptic array, eliciting fluorescence in the molecules alongside it. Sensors then measure characteristics of the light generated by that fluorescence.

Each pathogen, when bound together with its antibody, fluoresces light at a specific wavelength and frequency. If, for example, an E. coli bacteria has bound to an antibody on the fiber-optic array, the sensor will detect it. The technology has been proven to detect one cell in a 10-gram sample, Taitt says.

An attractive aspect of this technology, which is currently applied in the Raptor brand device manufactured by Research International, is its speed. The 13-pound machine, roughly the size of a pair of toasters, can conduct tests for up to four pathogens within 15 minutes. All the user has to do is insert a sample and push one button.

The machine’s optical fibrils can be reused until a positive test occurs, at which point they must be replaced because of the inherent contamination.

Next-generation fiber-optic biosensor arrays, expected to be available for mass-market use in the coming years, will take a page from Affymetrix with a “lab on a chip” concept. Instead of a needle shaped fibril bearing one antibody, a glass slide-like array will bear a field of different antibodies, Taitt says.

Instead of being shot down a narrow fibril, laser light would be introduced into the slide laterally, exciting fluorescence in the molecules on the face of the slide. Dozens of different pathogens could be assayed on a single slide, Taitt says.

The device is dubbed the Leopard due to the slides’ antibody spots, and it has been licensed to three firms, one focused on food safety. Hanson Technologies, Inc., of Carlisle, Pennsylvania, is using the Leopard to screen leafy greens like those blamed for the 2006 E. coli outbreak that killed three people.

A traditional test for E. coli in leafy greens calls for a 25-gram sample from a 5,000-pound lot, explains Chairman and CEO William Hanson. That sample is cultured in a warm, nutrient-rich broth for one day, then subjected to a test, most commonly an ELISA test using immunomagnetic beads. Because of the size and randomness of the sampling, Hanson calls the value of the method “almost zero.”

Scientists at Hanson Technologies had a better idea, based on their knowledge that companies processing bagged leafy greens typically submit them to massive washing measures in an effort to remove contaminants. The idea was simple: Don’t test the product, test the used wash water. The result of this breakthrough is the OmniFresh 1000 System.

The washers used by large-scale leafy green processors clean around 5,000 to 7,000 pounds of product per hour, removing roughly 90 percent of contaminants, Hanson says. To use the OmniFresh system, processors divert 10 to 15 percent of each batch’s used wash water through a proprietary filtration system, which Hanson says isolates and extracts potential pathogens.

Those culled materials are tested using fiber-optic-array biosensor technology licensed from the Naval Research Laboratory. The process, which takes two hours, increases the effective sample size 30,000 to 50,000 times over standard random testing methods, Hanson says.

Results of formal, large-scale testing conducted with a major produce firm in California during October and November of 2007 were encouraging, he says. The manufacturer allowed a program in which Hanson Technologies tested a series of 2,000 to 3,000 batches of fresh vegetables. For the test, small batches of the product were purposefully contaminated at random with 11 to 20 pound samples of fouled greens. The OmniFresh system detected the contamination in each case, Hanson says.

A separate pilot program was conducted with Carlisle, Pennsylvania’s Verdelli Farms, where the OmniFresh system was used to test 40,000 pounds of produce over the course of a month, Hanson says.

While fiber-optic devices like the Raptor and Leopard currently deliver the test results through a standard notebook computer plugged into the array, this kind of technology holds the greatest promise of handheld miniaturization, states King, of the Midwest Research Institute.

Star Gazing

Some technologies are currently far less developed but hold enough potential to merit more research.

Spectroscopy. We all learned in high school science class that astronomers could determine the elements present in a faraway star by passing its light through a spectroscope—a device that breaks down light into its constituent color elements. Each element burning in a faraway star has a specific spectral fingerprint that can be read by a measuring device called a spectrometer.

Spectroscopy works most easily with pure elements. The process becomes more complicated when used to detect molecules, in particular complex ones like the proteins that offer telltale signs of foodborne disease bacteria. The proteins themselves may vary while hiding among thousands of other chemicals on the outside of a bacterium.

The technique, however, holds the promise of effective food-safety testing. Light from the infrared spectrum is analyzed using highly sensitive detection equipment; that data is analyzed with software, King says.

W. Fred McClure, professor emeritus with the department of biological and agricultural engineering at North Carolina State University, explains that infrared spectroscopy for food testing would typically involve a small sample placed on a glass slide similar to one used under a traditional microscope.

Light is shone either onto the sample from above, or through the slide from below, intermittently and incrementally, at different wavelengths. At each different wavelength the spectrometer registers whether the light is reflected by the test material or absorbed.

Using analytical software, the results are presented on a graph, with wavelength represented on the “X” axis and absorption on the “Y” axis. Different compounds have distinct fingerprints on the graph, McClure explains.

Effective field use of infrared spectroscopy in food testing remains years away due to the number and complexity of compounds associated with pathogens, and the challenge of pinpointing them amid others, King says.

A more advanced iteration of spectroscopy, and one still further from practical application in this field, is referred to as Raman spectroscopy. The method is based on the physical effect discovered by Chandrasekhara Venkata Raman in 1928, earning him the Nobel Prize for Physics. Raman discovered minute variations in the frequency and wavelength of light reflected by specific atoms and molecules.

If future advances in spectrometry and software analysis allow the practical application of Raman spectroscopy in detection of foodborne pathogens, its precision would allow what scientists have referred to as “spectroscopy with a tweezer.”

Smart labels. It’s a heck of an idea: A tiny sticker or label on the inside of clear, plastic food packaging that changes color, or even better, displays a “sad” face when food has gone bad. It’s the brainchild of Mitch Sanders, founder and executive vice president of ECI Biotech in Worcester, Massachusetts.

The company’s concept relies on proprietary enzymes, engineered to react when exposed to chemicals or organic compounds produced by stale or outright rotten food. The same technology could theoretically be applied to many foodborne pathogens, Sanders says. His company touts speed, low cost, and, of course, ease of interpretation.

The idea may sound great to a wide-eyed consumer, but food manufacturers didn’t bite, Sanders says. They saw the little stickers not as smart labels, but liability labels. What if a consumer got sick after eating a product contaminated with Listeria, but the sticker only indicated Salmonella? What if a consumer ate some packaged cold cuts, and bore no ill effects, but looked at the package two hours later and saw an ominous frown?

Further, tests demonstrated 83 percent accuracy in samples carrying pathogens, and 14 percent false positives in clean samples. The food industry said “no thanks” for now, Sanders says.

ECI’s technology is not dead in the water, however. It may soon turn up at your local drugstore, featured in traditional adhesive bandages. If the cut on your finger becomes infected, the bandage would turn a telltale color. Further, it may be marketed to industry for testing of products during processing, Sanders says.

Focus on Process

All food testing methodologies are further complicated by the mass of the food supply, and its inherent dynamism. For example, a single sample taken from a processing plant hopper may miss a deadly pathogen tucked in another corner of the vat. Alternatively, a pristine lot that passed a safety screening test may then undergo a slight change in temperature or pH. Even this minor shift could foster growth of a harmful pathogen.

Food suppliers further face the question of cost. At processing plants, a fast, cheap, disposable, reaction based dipstick test, resembling a litmus strip or home pregnancy test, is useless unless a pathogen is present in massive and, therefore, unlikely quantities, King says. Meanwhile, 100 percent product sample testing is inherently impossible.

In comparison, the future may offer handheld Raman spectrometers for quality control and testing. The devices could cost $50,000 to $100,000 apiece, King estimates, but the additional cost for each sample would be nil, and the results would be immediate.

Lee Johnson, director of process and technical support for poultry manufacturer Butterball, LLC, and former head of food safety for ConAgra, says manufacturers like his employer determine the size and frequency of testing samples based on the pathogen, level of risk, cost, safety, test accuracy, and supply chain volume. Most major tests at Butterball rely on PCR and ELISA technology, and each can take 24 to 48 hours.

Johnson’s top concerns include traditional poultry pathogens and more recently avian flu, which his company tests for in each flock, working in close collaboration with the U.S. Centers for Disease Control and Prevention.

For now, the company is sticking with tried and true technology. It is leery of investing money in novel ways to test the product. “We don’t have a lot of time to mess around with theoretical tests or failure,” says Johnson.

Food-safety experts like Johnson place the value of safety testing in perspective. They say it’s a critical, yet relatively minor element of an overall process management regimen that serves as the real defense against foodborne illness.

“You can’t test food safety into a product,” says Gary Ades, president of G&L Consulting Group, which oversees food safety for several manufacturers and retailers. “You have to build it in.”

Joseph Straw is an assistant editor at Security Management.