Michael W. Pariza
5.2 Underlying Considerations
5.3 Safety Evaluation
5.4 Evaluating Protein Safety .. References
Microbial enzymes used in food processing are not pure substances. Rather, they are complex mixtures that include the desired enzyme as well as other metabolites generated by the production strain, in addition to intentionally added materials such as preservatives and stabilizers. Accordingly, safety evaluation of food enzyme preparations poses special challenges that are not typically encountered with other food ingredients. To address these challenges we developed a scientific framework1,2 that focuses on the safety of the production organism and its metabolites rather than simply on the desired enzyme. This framework may also serve as a model for evaluating the safety of other complex food matrices that contain intentionally modified proteins.
In the United States, the U.S. Food and Drug Administration (FDA) has the primary regulatory jurisdiction over the use of food ingredients, including, of course, enzymes used in food processing. The uses of most food ingredients are regulated under FDA's GRAS (Generally Regarded As Safe) provisions, which are available online at http:// www.fda.gov/. With regard to microbial enzymes used in food processing, the FDA considers the safety of the producing organism to be of paramount importance. For example, the regulation (21CFR184.1685) dealing with the enzyme chymosin, which is produced via microbial fermentation and used to make cheese, reads in part as follows:
"Chymosin preparation is a clear solution containing the active enzyme chymosin (E.C. 126.96.36.199). It is derived, via fermentation, from nonpathogenic and nontoxigenic strains of Escherichia coli K-12 containing the prochymosin gene. The prochymosin is isolated as an insoluble aggregate that is acid-treated to destroy residual cellular material and, after solubilization, is acid-treated to form chymosin. It must be processed with materials that are generally recognized as safe, or are food additives that have been approved by the Food and Drug Administration for this use."
The FDA also lists in 21CFR184.1685 other acceptable microorganisms for chy-mosin manufacture, including nonpathogenic and nontoxigenic strains of Kluyvero-myces marxianus and Aspergillus niger. The important elements are that the production strain must be safe (i.e., nontoxigenic and nonpathogenic) and processed with materials that are either GRAS for use in food enzyme manufacture, or regulated food additives that have been approved by the FDA for this use. The Enzyme Technical Association (ETA) maintains a current listing of production microorganisms and enzymes in commercial use, including enzymes used in food processing. The listing can be accessed at http://www.enzymetechnicalassoc.org/.
It bears repeating that food enzymes are not manufactured and sold as pure substances but, rather, as complex mixtures that include the desired enzyme, other metabolites generated by the production strain, and intentionally added materials such as preservatives and stabilizers. The intentionally added materials (preservatives, stabilizers, etc.) should be GRAS for use in food enzyme manufacture, and used in accordance with current Good Manufacturing Practice (cGMP) as defined by the FDA. These, of course, are FDA requirements and must be strictly adhered to.
It is important to recognize that enzymes likely to be used in foods (carbohy-drases, lipases, proteases) are already present in the human digestive tract in far larger amounts than one would typically encounter in a processed food. This "natural enzyme background" consists of enzymes that are synthesized endogenously and secreted into the gut; enzymes synthesized by microbes that inhabit the gut; and enzymes that occur naturally in the foods we eat, particularly uncooked foods. There are, of course, a few rare enzymes with known toxic properties (for example, toxic enzymes found in venom or associated with microbial pathogens such as Coryne-bacterium diphtheriae) but these would never be considered for food processing use because, among other things, they would serve no useful purpose in functionally modifying any component in a food matrix. In addition, although allergies to certain food proteins are a serious matter for some individuals, it is worth noting that there is no documented case of an allergic reaction to an ingested enzyme from a commercially processed food. There are rare instances of allergic reactions to inhaled enzymes, but these did not involve commercially processed foods.3
Given these considerations, it follows that safety evaluation should focus on the production strain and its metabolites, including but not limited to the desired enzyme protein, that comprise the enzyme preparation. In this regard it is critically important that the production strain not produce toxins that are active via the oral route.
Evaluating the toxigenic potential of a microorganism and its metabolites would certainly be a daunting task were it not for the extensive scientific literature base that is available concerning toxigenic and pathogenic microorganisms.124 Because of this literature base we know that very few microorganisms that grow in food will produce illness via either intoxication or infection. Moreover, the most prominent of the foodborne toxigenic/pathogenic microorganisms have long been recognized and are well characterized in terms of their capacity to produce human illness.
Microbial foodborne intoxication requires the presence of a toxic agent but not necessarily viable cells of the toxin-producing strain. For example, staphylococcal enterotoxins and botulinal neurotoxins can be produced in food by, respectively, toxigenic strains of Staphylococcus aureus and Clostridium botulinum. The toxins may then persist even under conditions where the respective producing organisms have been inactivated or inhibited from growing. By contrast, foodborne pathogens such as Salmonella species, which induce illness via an infectious process rather than intoxication, present a risk to consumers only if viable organisms are present in the final product. There is also a third category, represented by Clostridium perfrin-gens, where ingestion of the of viable organisms leads not to infection but to intoxication from an enterotoxin that is synthesized by the organisms in situ. C. perfringens enterotoxin is a spore coat protein that is produced during sporulation. Intoxication results from ingesting viable vegetative cells which then sporulate in the gastrointestinal tract, thereby producing and releasing the enterotoxin.
We know a great deal about the chemical nature of microbial toxins and their physiological affects on humans and animal models. Microbial foodborne toxins range in size from relatively large-molecular-weight toxic proteins produced by toxigenic bacteria to small-molecular-weight toxic organic compounds produced by toxigenic molds, algae and (rarely) certain bacterial species. These toxins induce a range of toxin-specific adverse effects that include vomiting and diarrhea (e.g., staphylococcal enterotoxins), paralysis and death (e.g., botulinal neurotoxins), and acute hepatic necrosis and cirrhosis and ultimately hepatocarcinoma (e.g., aflatoxin produced by toxigenic species of the mold genus Aspergillus). Notably, all of these foodborne toxins induce acute toxic effects that are evident within a few hours to a few days after exposure. In some cases chronic toxicity may also occur (e.g., long-term paralysis from exposure to a botulinal neurotoxin or liver cancer from aflatoxin ingestion), but in every instance, at sufficient exposure levels, all known foodborne microbial toxins will first induce symptoms of acute toxicity in susceptible animal species. This realization is critical to developing effective safety evaluation strategies for microbial products, including enzymes.
The scientific framework for evaluating the safety of microorganisms for use in enzyme manufacture that we developed12 begins with a thorough characterization of the organism using molecular classification technology, for example, 16S rRNA gene alignment. This is necessary to ensure that the organism has been correctly classified with regard to genus and species, and to identify relatedness with other microbial species. The next step is to conduct a thorough literature review to determine whether the species to which organism belongs, as well as other closely related species, have been associated with human illness. It is particularly important to determine whether the organism or closely related species have been associated with the production of toxins that are active via the oral route. It is also common practice to at least partially sequence the genome and to utilize this information to determine whether the genome contains any known toxin genes. If the production organism is genetically modified, then additional considerations come into play regarding the nature of the modification, the characteristics of the donor organism with regard to potential toxigenicity, the presence of transmissible antibiotic resistance markers, and so forth.
The enzyme preparation, which contains not only the desired enzyme activity but also other metabolites produced by the production strain, is then evaluated using appropriate chemical and biochemical tests for potentially "adverse" agents. This includes the FDA requirement that molds be screened to ensure that they do not produce antibiotics or mycotoxins for which appropriate chemical or biochemical tests are available. European regulators often require mutagenicity testing using the Ames test, although it should be noted that this requirement has never generated any useful information with regard to the evaluation of enzyme safety; for this reason, we have not included it among our recommendations.12
After the various molecular, chemical, and biochemical screening tests for known adverse agents are completed, the enzyme preparation is evaluated with appropriate animal feeding tests to ensure the absence of any previously unknown substances that might induce adverse health effects. Typically this involves standard subchronic (91-day) feeding trials in rats.
The forgoing is focused on ensuring the absence of toxins that act via the oral route. Pathogenic potential is, of course, also important — not so much with regard to consumer safety, because enzyme preparations rarely contain viable production organisms, but, rather, with regard to worker safety and the feasibility of safely growing the organism in a fermentation plant. In assessing pathogenic potential it is important to distinguish between true pathogens and opportunistic pathogens. True pathogens, which are relatively rare among microbial species, are able to overcome host barriers that have not been compromised, and to induce infection. By contrast, opportunistic pathogens will produce infections only in compromised hosts (e.g., individuals with suppressed immune systems). Accordingly, although only a relatively small number of microbial species are true human pathogens, many microorganisms are associated with occasional opportunistic infections. Hence, occasional reports of opportunistic pathogenicity should not by itself exclude an organism from consideration for enzyme manufacture.1,2
It should also be noted that although appropriate animal models are available for assessing toxic potential, microbial pathogensis is a far more complex process and microbial pathogens often exhibit host specificity, which greatly limits the ability to use animal models to screen for potential pathogenesis in humans.
Ensuring the safety of bacterial enzyme preparations necessitates ensuring that enterotoxins and other toxic proteins active via the oral route are not present. Hence, food enzyme testing protocols must be designed so that the detection of such toxic proteins is assured. In this regard it is again worth noting that all known toxic proteins induce acute toxic responses. Moreover, the toxic responses induced by toxic proteins do not include long-term chronic conditions, such as cancer. Said another way, there are no known proteins that, when ingested, will induce cancer or related chronic illness.5 (In rare instances, nonprotein prosthetic chromophores with DNA-damaging activity have been reported in association with unusual proteins, but the DNA-damaging properties reside solely with the nonprotein chromophores, not with the associated apoprotein structure.)6 Accordingly, there is no justification whatever for conducting chronic toxicity feeding tests on proteins. Subchronic (91-day) feeding trials are fully sufficient for assessing the safety of proteins, irrespective of whether the proteins are naturally occurring or intentionally modified.
1. Pariza, M.W. and Foster, E.M., Determining the safety of enzymes used in food processing, J. Food Prot., 46, 453, 1983.
2. Pariza, M.W. and Johnson, E.A., Evaluating the safety of microbial enzyme preparations used in food processing: Update for a new century, Regul. Toxicol. Pharmacol., 33, 173, 2001.
3. Quirce, S., et al., Respiratory allergy to Aspergillus-derived enzymes in Baker's asthma, J. Allergy Clin. Immunol., 90, 970, 1992.
4. D'Mello, J.P.F., Ed., Food Safety, Contaminants and Toxins, CABI Publishing, Cambridge, MA, 2003.
5. International Food Biotechnology Council (IFBC), Assuring the safety of foods produced by genetic modification, Regul. Toxicol. Pharmacol., 12, S1, 1990.
6. Povrik, L.F. and Goldberg, I.H., Covalent adducts of DNA and the nonprotein chromo-phore of neocarzinostatin contain a modified deoxyribose, Proc. Natl. Acad. Sci. USA, 79, 369, 1982.
Was this article helpful?
What Is The First Essential Step For Going Green With Food? Get Everything You Need To Know To Get Started With Helping The Earth And Going Green With Food. This Book Is One Of The Most Valuable Resources In The World When It Comes To Everything You Need To Know About Green Agriculture.