As Food Microbiology Division Lecturer of the American Society of Microbiology (ASM) at the organization’s national meeting in Washington, DC, in 1995, Daniel Y.C. Fung, Ph.D. was asked to give a lecture reviewing rapid methods and automation in microbiology and provide predictions of the future of this important field in food and beverage analysis. Nearly five years later, Food Safety Magazine (then Food Testing & Analysis) published the first update of those predictions to see whether the noted food microbiologist’s forecasts had achieved “real-world use” status. In that 1999 article, entitled “Predictions of the Future of Rapid Methods in Microbiology,” Dr. Fung wrote, “Predicting the future is risky business at best. Several of the predictions I made in 1995 actually became reality. Hope- frilly, a few of the 1999 predictions I’ve forecast here will become a reality in the next decade. And, while my prognostications are speculative in nature, I foresee a bright future for rapid and automated microbiological methods and instrumentation in the food industry as we move out of the Information Age and into the era some observers are already dubbing the ‘Age of Hyper-Technology.’ I can’t wait to see how this set of predictions measures up!” In this FSM interview, Dr. Fung once again takes that risk and analyzes his earlier predictions to see what has transpired in the field.
Food Safety Magazine: In 1999, you wrote about a few of the basic drivers in the development of rapid and automated detection capabilities for the food industry. Have these drivers proved to be catalysts in advances?
Daniel Y.C. Fung: As I stated in my 1999 article, food microbiology has been rapidly catching up with clinical microbiology in terms of sophistication of rapid methods and automated techniques for quite a few years now. My statement that the drivers for this acceleration are largely the increase in the requirements for the monitoring of the safety of food supplies in national and international trade and the important role that food microbiologists play in the entire field of food safety as microbial challenges to foods emerge has proved true and will continue to prove true, especially in light of new bioterrorism concerns.
In that article I also cited a 1998 Strategic Consulting Inc. market survey which reported that the volume of microbiological tests was estimated to be a worldwide market of more than 750 million tests conducted annually, of which 420 million were run in the food industry. The report stated that, of these, just over 60 million tests, or roughly 15°/o of all microbiology tests performed in the food industry, were for specific organisms such as Salmonella, Escherichia coli O157:H7, and so on, with the remaining tests falling into the routine testing category. A 2002 report from the same surveyor [see, “New Report Analyzes Global Market for Food Microbiology Testing”] finds not only did the testing volumes increase globally to 501 million conducted by the food processing industry worldwide, but that overall growth of food microbiological testing has increased about 6% annually since 1999, and that pathogen testing has out- paced (13.5%) routine testing (4.5%) in annual growth rate comparisons. I think this is an indicator that we are making progress in better, more usable rapid microbiological detection and analytical methods and technologies that can provide the speed to result that many food companies require.
I also noted that there was no doubt that the Hazard Analysis & Critical Control Points (HACCP) system would play a key role in the present and future of food safety, and that food microbiology testing would also play an important role in this regard. HACCP has become widely accepted and used as part of many food safety programs, and of course, we must verify whether the HACCP program is working or not. I like to use canned goods as an example of this, because with every possible combination of heating curve and processing technique used one would assume that the canning will be perfect, right? But you still have to test the canned product at the tail end of the process to see whether or not you have Clostridium botulinum spores. Thus, the beginning, middle and end-point testing of the target pathogens are still very important tasks, because even though you think you have the best HACCP program, you still have to know at the beginning of the process what potential microbial contaminants are coming in to your plant, as well as ensure that product going out of the plant is not contaminated. I am from the school of thought that end-point testing is still an important step to validate that your HACCP program did the job.
Food Safety Magazine: Will you review your 10 predictions from 1999 and talk about whether you think these forecasts on food microbiology advances in rapid and automated detection have been realized?
Fung: Many of my predictions have now been realized, while some are still a few steps away from realization. However, I think that they all remain valid for future improvements and development.
Prediction 1. Viable Cell Counts Will Still Be Used. The ability to tell whether cells are alive or dead in a sample is important, because live cells will grow into millions of cells in just a few hours while a dead cell will not. Certainly, if we cook our food properly, we have eaten Salmonella that is dead, which will not harm us. So, we are concerned only with living cells, and viable cell count enables us to identify the spoilage potential of a food and, of course, the potential of pathogens to grow in large numbers that will make us sick. Viable cell count also is an important tool for processing segments, such as the dairy industry in which total count standards must be met for raw and unpasteurized milk, as well as for some countries in which milk suppliers actually receive a higher price for milk with a lower viable cell count.
As I predicted in 1995 and 1999, viable cell count (total aerobic count, anaerobic count, differential count, and pathogenic count) is still an important parameter to assess the potential safety and hygienic quality of our food supply. During the last few years, improvements in alternative methods, including rapid culture methods, spiral plating systems, direct epifluoroscent technique (DEFT), and other innovations in more efficient dilution instruments and sample preparation instruments, have made viable cell count procedures more efficient and rapid.
Some of the specific advances I predicted in this area are closer to reality, most recently the improvement of vital stains using very specific dyes that differentiate living versus non-living cells in less than one hour. As noted in 1999, I stated that these vital stains can greatly reduce detection time of viable cells. We now are moving to detection times of less than one hour. If this can be perfected, we could have a viable cell count in about a half an hour.
At this moment, we do have the DEFT test, which takes an hour, but the problem with this system is that you need a very sophisticated microscope to read the test. There are some companies now designing hand-held units in which you can actually take a drop of milk, for example, which will react with a vital dye. After staining the drop, a biochip will suck the liquid into a film and you stick the slide stain into this hand-held unit, which will automatically scan the entire film with an increase in sensitivity. If you use an all-immersion microscope, which provides a very small field of vision, you need to have 106 cells to get a reliable reading. However, these newer systems offer a greater field of vision and can scan the entire film to get down to 104 or even 103 cells/mL. The reality is that there are some systems where, if they find the right dye and use correct staining and optics, can give a viable cell count in less than an hour.
Another of my predictions related to viable cell counts was the development of technologies for early sensing of viable colonies on agar. I stated that I believed that in the future it would be possible to use heat-sensing or colony-sensing devices to detect the development of colonies on the agar without having to “see” the colonies with one’s eyes. This development would shorten the procedure to about three to four hours instead of the traditional 48 hours to count a visible colony. Well, I just saw something while on the East Coast that really is like science fiction. A person showed me a series of pictures of agar plates showing from one, individual cell on a plate all the way to an actual colony (which could be seen with the naked eye) by a procedure involving an antibody reaction and the amplification of specific dyes. The ability to identify one E. coli O157:H7 on the surface of an agar is really amazing. This type of development, which shows our ability to see cells much sooner now, supports my earlier predictions.
Prediction 2. Real-Time Monitoring of Hygiene Will Be in Place. In my earlier predictions, I focused on the increased use of ATP bioluminescence and catalase methods, which are technologies that have been proven as useful to the industry. ATP bioluminescence- based diagnostic test kits in food sanitation monitoring now enjoy widespread use in the food industry worldwide. Again, these systems have proved very useful in improving hygiene monitoring and promoting good cleaning practices in food plants, and they likely will continue to be used extensively in the near- and long-term future. Research and development in the use of the catalase enzyme to test the hygienic status of food processing or preparation surfaces has continued, as well. I think that the catalase test is still a viable method, and a good way to help people tell whether the surface is dirty or not, so that they have the ability to properly clean those surfaces. And many other developments in the use of surface contact plates, paddle methods, swabs, the 3M Petriflim contact method, impedance instrumentation and optical method automated systems have proved to help users obtain very rapid indicators of hygiene.
I also forecast in 1999 that other online monitoring systems, such as measuring the presence of protein, fat, carbohydrate, and so on, would also be developed for hygiene monitoring in the near future and provide instantaneous results. The first generation of these tests, which detect protein residues in five to 10 seconds are now available, and this is an exciting development. Industry is now using five-second protein tests, where you can take a swab, activate it, swab the target surface, and if it changes to a green color, you know that too much protein is present, which indicates that the surface is not clean. These types of hygiene monitoring techniques, which can be done in about a minute or less, are truly defined as rapid methods. There are several companies developing tests that give almost instant detection of residues, including carbohydrates and lipids.
Prediction 3. Polymerase Chain Reaction, Ribotyping, and Genetic Tests Will Become Reality in Food Laboratories. This was a bold prediction in 1995, and a not-so-bold one in 1999. Today, I am very comfortable saying that this prediction was a good one, because genetic engineering procedures have become more and more user-friendly such that that one does not have to be a genetic engineer to run polymerase chain reaction (PCR) and ribotyping tests in food laboratories. This is due to the enhanced automation of such systems, which has made it efficient for large food laboratories to adopt and use. These systems will likely filter down to smaller food labs to use as the market develops.
Interest in bioterrorism prevention is spurring a lot of companies to perfect PCR and ribotyping systems for increased speed. Two years ago, we were talking about genetic-based rapid methods as “maybe” being useful for rapid testing and analyses of microorganisms, most specifically E. coli O157:H7. Now we are talking about these rapid methods as being more useful, more developed and more able to detect many other organisms in addition to E. coli organisms.
Prediction 4. Enzyme-Linked Immunosorbent Assay and Immunological Tests Will Be Completely Automated and Widely Used. This 1995 prediction certainly had come true by my 1999 prediction, specifically in that automated systems for immunological analysis had become widely used in the food industry, government and consultant laboratories around the world at that time. An analyst only needs to pre-enrich a food sample in the appropriate broth for a period of time (usually overnight) and then a small aliquot is introduced to the automated instrument and the analysis is completed automatically in about an hour or an hour-and-a-half.
The push right now is in the area of lateral flow test system diagnostic kits. I’m really amazed, because I thought this detection technology was stabilized given the amount of ELISA systems commercially available. But right now, the battleground is to make these tests more sensitive. ELISA tests usually requires 106 organisms to be effective, but the new generation of lateral flow tests make particles more sensitive, and as such, are getting down to 105 and even 104, which is incredible. But this raises another question: How fast can we get results from a meat sample? Most lateral flow tests require overnight incubation and growth. The push is to have it done in eight hours, after enrichment in the specially designed enrichment broth. The refinement of lateral migration to more sensitive units is now the reality.
Prediction 5. Dipstick Technology Will Provide Rapid Answers. This prediction, again, became reality in 1999, and many forms of dipsticks have been available for the rapid screening of pathogens, including those based on lateral flow, as described above, and those in which the user directly dips the kit into the enriched liquid sample and by capillary action the target cells are “sucked” into the kit for subsequent reactions or those are based on colorimetric assay (direct-labeled probe) formats. Dipstick technology has been well-received by the food industry and likely will continue to be improved even more as time goes on.
Prediction 6. Biosensors Will Be in Place in HACCP Programs. In 1999, I said that, “This prediction is for the future!” ‘Well, the future is much closer, although I still believe that we are a few steps away from the ability to place biosensors in food processing systems for the instantaneous detection of pathogens in the food matrix. And again, there is no lack of sensors available, particularly if one considers monitoring of ATP, catalase, enzymes, pH, oxygen gas, hydrogen gas, other metabolic products, conductance, impedance, capacitance, microcalorimetry, and so on, as “biosensors” for microbial activities. These methods can be used successfully now, but they will not enable the user to detect specific pathogens in “real time,” and thus are not fully efficient for use in a HACCP program at this time.
Having said this, biosensors are getting better and better, and there are many fiber optics that can sense analytes better than two years ago. The problem with the biosensor is discovering how to get the target organism to interact with the sensor. While there is no problem when using a biosensor with a pure culture, there is a problem in determining how to get one Salmonella in 25 grams to react with the sensor. The problem is the size of the microorganisms we want to react with the sensor. For example, one E. coli is 10-12 gram, and the ability to pull that size of a cell to interact with the fiber optic biosensor is a major challenge. We still need to amplify or concentrate the cells and that means continued efforts to develop innovative filtration, concentration, electric attraction and other sample preparation techniques to advance the usefulness of the biosensor.
Also, it appears that microarray chip technology is becoming very inexpensive to use. I have spoken to several people recently who are saying that microarray chips costs are as low as two cents per DNA hybridization per spot on a chip. This means that if you have a chip that can detect 100 different pathogens on a slide, the cost is only about $20. In addition, there are some microarray biochips that enable you to spot 50,000 DNA on one slide, which makes the area of bioinformatics, or how to interpret the data, an increasingly important field of study.
For the quality assurance and food safety personnel in the food processing sector, the ability to spot 100 pathogens on a slide is good, but what is better is the ability to use a microarray chip to differentiate between pathogens. For example, if you have a multipathogen testing chip on which you can accommodate 10 different major E. coli, 10 major Salmonella, or 10 major Listeria—all on that same chip—then you can take a sample of liquid from your meat, do some sample cleanup, put it across this slide, slip the slide into a reader, and you can actually identify not only whether this food has Listeria monocytogenes, but whether it is pathogenic in nature. This type of multiplexing on a biochip potentially can give you a reading of all the major pathogens’ occurrence (or not) in your food, and very inexpensively at two cents per dot.
Prediction 7. Instant Detection of Target Pathogens Will Be Possible By Computer-Generated Matrix in Response to Particular Characteristics of Pathogens. In 1995, my thought was that we should find some way to find the toxin gene or matrix for virulence, which is why I placed this prediction on my list. The exciting news is that this is exactly what researchers are doing right now in the areas of proteomics, genomics and phenomics. Proteomics is the study of finding the function of the proteins; genomics is the study of how to express the gene; and phenomics is the study of how to use the expressed gene characteristic to identify the organisms.
For example, carbohydrate fermentation can be described as a phenotypic expression of the genotypic characteristic of the cell. The genotypic is more stable, but it may not be expressed. Even though you might know the entire genome of the E. coli O157:H7, the question remains, how are the cells expressed? A good example is the Biolog system, which is similar to my 1967 research in which I was the first to put biochemicals into the microtiter plate for systematic diagnostic microbiology. Biolog used the same philosophy and improved on it with computerization of the system. They can provide you four microtiter plates, each one with about 100 wells, so that you have approximately 400 characteristics available, making the combination/permutation opportunities almost endless for phenotypic microarray. This is a powerful tool, and while the method is not new, the trend toward finding how to use the information is new.
Prediction 7 is based on developments in technology with regard to how to find the virulence factor, but at the same time, how to detect expression of the cell’s gene. When I first predicted this type of development in 1995, it was philosophical, but as it now migrates into the realm of practicality, I think it is a valid prediction that this area of development will continue to be actively researched and advanced.
Prediction 8. Effective Separation, Concentration of Target Cells Will Greatly Assist in Rapid Identification. Again, advanced separation technology saves at least one day in the enrichment procedure of many pathogen detection protocols. As I’ve said, concentration of cells by effective filtration, differential centrifugation, electric precipitations, and so on, will greatly assist rapid identification of target cells, but we still need fuller development of a lot of ingenious methods. Without these advances, a lot of the new and improved technologies like immunomagnetic separation, filtration and centrifugation, enzyme systems and particle attraction cannot move forward. Efficient separation of the pathogens from the background flora which enables us to stimulate the growth of target pathogens increases their chances of being detected by secondary tests such as PCR, gene probes, ELISA and other methods. This is moving along, but we need to move it further.
Prediction 9. A Microbiological Alert System Will Be in Food Packages. As we know, microbial cells during growth and spoilage will generate a variety of compounds that can be detected by devices such as gas and pH indicators. My original prediction was that a series of reagents could be embedded in a unit within the food package and when enough of a certain gas (e.g., ammonia, hydrogen sulfide, CO2. or pH changes) is generated, a color change will occur similar to temperature-sensitive papers to indicate a potential spoilage problem exists. It appears that this is getting closer and closer to reality now; specifically, there are many sensors in food packages that can sense CO2. pH, ammonia, and different compounds. There are several researchers who are trying to make antibody on packaging film for automatic detection of target pathogens.
In the future, I predict that we will have different barcodes encoded with information about target analytes such that if the microorganisms in the food generate some kind of a combination of acid, gas or temperature abuse, these will be detected through a hand-held barcode scanner/reader via a color change. For example, if five of the 15 barcode lines change color, the food is suspect. This may be total science fiction, but I do know some researchers who are putting antibodies from Salmonella on the packaging film and if Salmonella is present, it reacts with this antibody and then further reacts with another antibody inside this unit that automatically makes a sandwich ELISA test on the packaging film. They can actually design it so it will spell out the word Salmonella!
Prediction 10. Consumers Will Have Rapid Alert Kits for Pathogens at Home. It remains conceivable that rapid alert kits for food spoilage and even potential food pathogen detectors available for home use will be developed. I know of some individuals who are trying to make CO2 detection technology (CO2 is a universal byproduct of anaerobic and aerobic growth) into solid state technology so that people could actually check food packages for themselves at home.
As I stated earlier, we would need a lot of consumer education and the building of consumer knowledge does take awhile. For example, when the microwave oven was introduced, it took about 10 years before people became very knowledgeable about how to cook with the new appliance and how to treat microwave- able food products differently from conventional oven cooking.
Food Safety Magazine: You’ve done a lot of interesting research using dried plum extracts to counter E. coli in foods. Can you tell us a bit more about that?
Fung: The reason we are excited about the dried plum extracts is two-fold. They can suppress the growth of many pathogens in foods, but more exciting is a recent finding by Kansas State University researchers that dried plum extracts retard spoilage organisms, and therefore, preserve food longer. We’ve put the seven different types of dried plum extracts in ground beef and placed the samples in 7C, which is abuse refrigeration temperature. We found that in the samples without the extracts, the cells start growing and the meat spoils in about a week, whereas the samples with the seven different types of dried plum extracts preserved the ground beef much longer. From a microbiological standpoint, these dried plum extracts are a very interesting compound, because they have two significant antioxidants in them— chlorogenic acid and neochlorogenic acid— that provide the antimicrobial effect.
But even more exciting is that these extracts contain a lot of sorbitol, and this carbohydrate acts as a humectant that helps to keep the moisture in the ground beef. If you cook the ground beef, freeze it and reheat it, the hamburger becomes like a piece of rock. Now, you say, so what? In many high schools, school cooks no longer cook hamburgers on-site from raw ground beef patties. The U.S. Department of Agriculture (USDA) school lunch program now produces millions of pre-cooked hamburgers and sends them to the schools where kitchen staff reheat and serve them to students. Often, students do not like the the meat patties that are so hard in texture, and as a result, they don’t eat them. Now, USDA is trying to use this dried plum extract to help retain the moisture of these hamburger patties. Dried plum extract also adds very little of its own taste or color to the meat product, so the functionality is a good selling point for its use.
Also, people are interested in natural compounds for use in foods. Consumers will have positive feelings about a label listing “dried plum extract” on the product, regardless of the active ingredients. Natural compound research like this has a very good return on investment (ROI) in terms of usage, consumer-friendly attributes, and food safety and quality assurance for the food company.
Food Safety Magazine: In your experience, Dr. Fung, what are the most significant foodborne illness-causing pathogens of concern to the food and beverage industry today?
Fung: Historically, the three most significant pathogens have been Salmonella, Clostridium pefringens and Staphylococcal enterotoxin in terms of total impact on society in the past 50 years. I still feel Salmonella is the most important pathogen of concern, internationally. You have pathogens come and go, but Salmonella is the key pathogen of concern, in my opinion. Now, Campylobacter may be popping up as an important concern, as well, and in fact, there are some consumer groups trying to get USDA to add this pathogen to the HACCP rule standards. But historically and in the foreseeable future, I still feel Salmonella is the base organism that continues to pose a most pervasive problem all around the world, and thus it must be monitored carefully all the time.
If I had to give a “top five” list of microorganisms of concern, I’d include Salmonella, Clostridium, Staphylococcus, E. coli, Listeria monocytogenes, and variations of all these, because even these have antibiotic-resistant strains. For example, E. coli O157:H7 is a pathogen that is very important to monitor, but we know that there are many other strains of E. coli that are important to monitor, as well. In terms of toxins, Staphylococcal enterotoxins are associated with a large number of outbreaks, and while they may not kill you, you’ll wish you were dead if you become ill from this toxin. So, we should not jump on the bandwagon of the “top five” list; we must monitor all potentially harmful pathogens in foods and beverages.
Food Safety Magazine: Recently released statistics from USDA and CDC show an overall reduction of 23% in the incidence of most of these pathogens in foods. Have rapid and automated methods contributed to that?
Fung: The information from the USDA on the reduction of Salmonella in ground beef since the implementation of HACCP is good news. This whole field of rapid and automated microbiological methods is helping us to better monitor pathogens and encourages us to move toward better, safer food products. People are more concerned than ever about how to best handle food and take hygiene measures, so I think the whole field is a catalyst for moving to a better society.
Don’t forget that we are talking about our high-tech society that is using these technologies at this moment. Other places—Asia, Africa, South America, and so on—are not well exposed to these technologies. Global development will be even more important in the next 10 years, and as I travel extensively, I can see that there is more and more excitement in non-industrialized countries about the promises of these technologies. As standards increase, there will be more testing to ensure that those standards are met. For example, although regular diarrhea is a normal occurrence in the lives of many in developing countries, if people find out that they can control and monitor the bacteria that causes the illness, they will take the opportunity to eat safer, better quality foods.
Daniel V.C. Fung, PhD., is professor of food science at Kansas State University (KSU), Manhattan, KS, and director of KSU’s annual International Workshop on Rapid Methods and Automation in Microbiology, which will be held this year July 12-19. Fung is an internationally recognized authority in the field of rapid methods, authoring more than 600 papers on the subject Among the numerous honors awarded to him for outstanding achievements in the field of food microbiology the most recent include his participation as an invited lecturer at the 100-year commemoration of the death of Louis Pasteur in Paris, the International Award given by the Institute of Food Technologists (IFT) in 1997, his election to the first group of Fellows by the International Academy of Food Science and Technology in 1999, and his receipt of the Waksman Outstanding Educator Award given by the Society for Industrial Microbiology in 2001. Fung can be contacted at Dfung@oznet.ksu.edu.
According to a new market report released in April 2002, the global food processing industry conducted more than 501 million microbiological tests last year at a cost of approximately $1.2 billion. These testing volumes are leveling off in North America and Europe, but are poised for significant growth in Asia and developing countries.
“Food Diagnostics: Global Review of Microbiological and Residue Testing in the Food Processing Industry,” from Stra