Food Safety Concerns
SCOPE AND GOAL
This chapter identifies potential food safety concerns for meat or animal products derived from animal biotechnology. The species considered include beef and dairy cattle, sheep, goats, poultry, swine, rabbits, and a wide array of finfishes and shellfishes.
The scope of this chapter encompasses: (1) non-genetically engineered animals that are propagated by nuclear transfer or other cloning techniques, (2) genetically engineered animals developed primarily for meat, milk, or eggs, and (3) genetically engineered animals developed for biomedical or industrial products. This latter category is considered because entry of these animals into the food chain might be proposed at the end of their productive life or sooner, as in the case of unused females and males, which typically are not used to generate the recombinant product (e.g., bulls in which the recombinant protein is expressed in the mammary gland).
The criteria used for identifying important scientific issues were developed considering the hazard (i.e., a compound or agent that has the potential to produce harm), the likelihood of harm resulting from exposure to the hazardous compound or agent, the likelihood that exposure to the hazard would occur, and the severity of any harm that would be realized. In this context, harm ranges from allergic reactions to other forms of illness, including, in the extreme
case, death. Concerns are described on a scale ranging from no concern, to low level of concern, to moderate level of concern, and to high level of concern.
Interest in the quality and safety of food of animal origin began to develop in the United States in the latter part of the nineteenth century, aimed primarily at meat for export to Europe. Regulatory and inspection systems for domestically produced red meat (but not poultry, eggs, or milk) were initiated in 1906 (Wiser, 1986). The U.S. Food and Drug Administration (FDA) prepared a list of food safety hazards (Foster, 1982) about 20 years ago. Food-borne toxigenic and pathogenic microorganisms were named first and considered to be the greatest danger to consumers. Also included were malnutrition, environmental contaminants, toxic natural constituents, reaction products that are formed during processing or preparation and storage for eating, pesticide residues, and finally, food additives. Food safety concerns raised by the use of animal biotechnology add to this list.
As new threats to food safety were recognized, new technologies and regulatory protocols were developed to enhance the safety of food. The occurrence of parasites was managed with slaughterhouse inspections, new husbandry systems, and parasiticidal drugs (Hagstad and Hubbert, 1981). By 1978, only 12 food-borne cases of parasite infection were documented (U.S. Public Health Service, 1978). Residues from drugs used to improve animal health and productivity arose as a food safety concern, but monitoring and inspection protocols largely have been effective in preventing illegal or unsafe levels of residues in food (Meyerholz, 1983; NRC, 1999; FDA, 2000). Microbial pathogens originating in animal fecal material remain the primary concern for the safety of food of animal origin (Tauxe, 1997).
Microbes pathogenic to humans grow in the animal gastrointestinal (GI) tract, and might or might not cause health problems in the animal (Altekruse et al., 1997). Physiologic stress increases the susceptibility of the animal to pathogens, the growth of pathogens in the GI tract, and their shedding into the fecal material of the stressed animal (Salminen et al., 1998). These same pathogens might enter the human food chain when they are transferred to the surface of the meat during slaughter and processing. The role of human food safety related to pathogens from animal fecal material was fully recognized only in the latter part of the twentieth century (Tauxe, 1997).
Secondary concerns for food safety arise from the disposition of carcass remains after removal of the edible meat, and from the disposal of animal fecal material. (Potential environmental concerns related to fecal material from genetically engineered animals are discussed in Chapter 5.) After the edible meat is removed, carcass remains are processed into other products used in a
variety of applications, including food and medical uses (Klinkenborg, 2001). One of the major products is meat and bone meal (MBM), a supplement historically fed to high-production animals. Using MBM from infected cattle in animal feed can transfer bovine spongiform encephalopathy (BSE) to other ruminants, and ultimately, to human consumers, which has occurred in Europe but not in the United States (Bruce et al., 1997). Concern about BSE transmission in the United States has resulted in regulations forbidding the feeding of MBM to ruminants (FDA, 1997). Animal carcasses also are used in a number of other products. Collagen is processed into gelatin for confectionary products such as candies, capsules for pharmaceutical products, and a range of cosmetic products. Bone and connective tissues are used in bone grafts and hernia repair in humans. Therefore, concern for the safety of products derived from animals also must take into account the use to which carcass remnants might be put once the edible portions are removed.
FOOD PRODUCTS FROM NON-GENETICALLY ENGINEERED CLONED ANIMALS
The cloning technologies of embryo splitting (EMS; Willadsen, 1979; Williams et al., 1984) and blastomere nuclear transfer (BNT; Willadsen, 1986; Prather et al., 1987) using embryo cells were introduced into dairy cattle breeding in the 1980s (Chapter 1). Although not widely adopted, a total of 1,472 EMS cloned Holstein females was registered with the American Holstein Association through 2001 (Norman et al., 2002) and evaluated genetically for yield traits, meaning they produced calves and were milked commercially. Yields of female EMS clones were greater than those of the Holstein population by 189 kilograms (kg) milk, 8 kg fat, and 7 kg protein, but slightly less than those of noncloned full siblings. The latter result might indicate an impact of the technology on performance or slightly different management of the two groups. Of 754 EMS cloned bulls registered and 143 evaluated by the U.S. Department of Agriculture (USDA) as sources of donor sperm, only 22 had noncloned full siblings. Results of the evaluations of the sires are not yet available. A total of 187 BNT cloned Holsteins (61 males and 126 females) were registered through 2001 (Norman et al., 2002); 74 had milk yield records, but only 11 had noncloned full siblings. The yields and milk composition of BNT clones exceeded those of the national herd average by 278 kg milk, 10 kg fat, and 10 kg protein, but were similar to those of their noncloned full siblings.
Although existing data for EMS and BNT clones addresses the changes in milk yield and composition, they do not specifically address the food safety of their milk and meat products. Aside from a study on yearling Brangus bulls that compared body measurements and measures of carcass merit obtained from their steer clone-mates (Diles et.al., 1996), there are no published analytical studies of
meat and milk composition comparing the products of cloned animals and full siblings evaluating in detail any unanticipated compositional differences, differences in protein quality, or nutrient bioavailability.
Since the donor nuclei used to produce EMS clones are taken from embryonic cells, there is little if any genomic reprogramming needed to drive embryogenesis. However, blastomeres from embryos of more than eight cells, (i.e., from the stages typically used for BNT), must be reprogrammed upon NT (Van Stekelenberg-Hamers et al., 1995; Kono, 1997; Bordignon et al., 2001), since they express a substantial number of genes, including paternal genes, that are not expressed by the oocyte nucleus. Indeed, the nucleus of the mature donor oocyte is transcriptionally quiescent and is associated with a different set of chromatin proteins (e.g., histones) compared to the recipient oocyte nucleus. A similar array of gestational and postnatal abnormalities seen in somatic cell nuclear transfer also has been observed in BNT (Wilson et al., 1995; Garry et al., 1996; Wilmut et al., 1997; Cibelli et al., 1998) clones. To the degree that inadequate or otherwise different reprogramming relative to that occurring normally in gametic nuclei occurs in BNT (De Sousa et al., 1999; Daniels et al., 2000), the composition of food products from NT animals might differ from that of ordinary animals. Although it is difficult to characterize the level of concern without specific data, it seems unlikely that there are changes in gene expression directly related to EMS and BNT cloning procedures that would raise nutritional or food safety concerns. Food products from BNT clones have been consumed by humans, with no apparent ill effects. Based on current scientific understanding, the committee regards products of EMS and BNT clones as posing a low level of food safety concern. Nevertheless, it would seem appropriate that the FDA use available analytic tests to evaluate the composition of food products from animals that themselves result directly from BNT cloning procedures to verify that they fulfill existing standards for animal-derived food products. The products from the offspring of cloned animals were regarded as posing no food safety concern because the animals are the result of natural matings.
The cloning of animals from somatic cells is more recent. Limited sample size and health and production data, as well as rapidly changing cloning protocols, make it difficult to draw conclusions regarding the safety of milk, meat, or other products from somatic cell clones and their offspring. The key scientific issue is whether and to what degree the genomic reprogramming that occurs when a differentiated nucleus is placed into an enucleated egg and forced to drive the development of a clone might result in gene expression that raises food safety concerns. Differences in patterns of developmental gene expression in non-engineered individuals and somatic cell clones would be greatest during early development when reprogramming is incomplete. A number of datasets suggest that the health and wellbeing of neonatal and young somatic cell clones often are impaired relative to those of normal individuals (see Chapter 6
regarding animal wellbeing). Direct effects of any abnormalities in patterns of gene expression on food safety are unknown. However, because stress from these developmental problems might result in shedding of pathogens in fecal material, resulting in a higher load of undesirable microbes on the carcass, the food safety of products, such as veal, from young somatic cell cloned animals, might indirectly present a food safety concern. As a somatic cell clone develops and nuclear reprogramming is completed, patterns of gene expression would approach those of a non-engineered individual. Indeed, the health and wellbeing of somatic cell clones approximated those of normal individuals as they advance into the juvenile stage. For example, somatic cell cloned cattle reportedly were physiologically, immunologically, and behaviorally normal, and exhibited puberty at the expected age, with high rates of conception upon artificial insemination (Lanza et al., 2001). Two of these individuals have given birth to calves that seem phenotypically normal. There are to date no published comparative analytical data assessing the composition of meat and milk products of somatic cell clones, their offspring, and conventionally bred individuals (although several studies are in progress; Bishop, personal communication, 2002). However, the committee found it difficult to characterize the level of concern without further supporting evidence regarding food product composition. Currently, there is no evidence that food products derived from adult somatic cell clones or their progeny pose a hazard (i.e., there is no evidence that they present a food safety concern).
GENETICALLY ENGINEERED ANIMALS
A number of types of genetically engineered animals will be developed primarily for food, and others will be developed primarily for producing non-food materials such as pharmaceuticals, vaccines, fibers, and other high value products. The principles for assessing the safety of food from genetically engineered animals are qualitatively the same as for non-engineered animals, but animals genetically engineered for non-food products might present additional concerns relating to the nature of the products that they generate. As for all foods or food products, those from genetically engineered animals should be evaluated for agents—chemical or biologic—which affect the safety of the food for the human consumer.
Animals used for xenotransplantation are not considered safe for human consumption and are excluded from the food chain by current regulations (see Chapter 7 for information on food animal regulations). Their exclusion is based primarily on concerns about persistent tissue residues of agents used to anaesthetize the animal prior to harvesting the tissues and organs. If there were any possibility that such animals might be rendered and considered for further processing into useful human food or medical products, concerns about
anesthetic residues would remain pertinent. If animals genetically engineered for xenotransplantation, but not used for that purpose, were presented for entry into the food chain, the food safety of such animals also would have to be evaluated based on protocols developed for evaluating other genetically engineered animals.
Animals might be genetically engineered to produce non-food products in their milk or eggs. Half of the genetically engineered population will be male, and will not be directly useful for production of heterologous proteins in, for example, milk or eggs. It is likely that companies producing such animals will seek early entry of males that are transgenic, but incapable of producing milk or eggs, into the food chain. In addition, companies might want to enter females that are “no takes”, which do not express high levels of the product of interest, or that have reached the end of their productive lives, into the human food supply. The safety of food products from such animals that were culled from transgenic lines might present concerns.
Numerous experiments have shown that the level and specificity of transgene expression in an animal is predictable only to a limited extent, probably because all the factors affecting gene expression have not yet been identified (Houdebine, 2000). Transgenes might be expressed at a low level in various tissues in which the promoter is not expected to be active. Such ectopic expression might be due to genomic position effects attributable to the action of neighboring enhancer elements. In addition, ectopic expression might result from basal-level transcription at the site of integration (Ashe et al., 1997; Travers, 1999). Recombinant proteins whose expression is driven by regulatory elements directing expression in mammary glands have been observed in the blood of transgenic animals during lactation (Bishoff et al., 1992; Devinoy et al., 1994; Thepot et al., 1995). The presence of transgene products in blood might result from leakage of the mammary epithelium or from secretion at the apical side of mammary cells. For example, although the promoter from the whey acidic protein (WAP) gene has been used to direct expression of a transgene in mammary tissue, and some concentration of WAP normally is found in the blood of lactating animals (Grabowski et al., 1991). Hence, through bioactivity, allergenicity, or toxicity pathways, ectopic gene expression might directly affect the safety of food products derived from tissues, sexes, or life stages of transgenic animals where transgene expression is not expected. In some cases, recombinant proteins produced in milk have deleterious effects on mammary gland function (Bishoff et al., 1992; Shamay et al., 1992; Bleck et al., 1995; Ebert et al., 1994) or on the transgenic animals more generally (Burdon et al., 1991; Reddy et al., 1991; Jhappan et al., 1993; Devinoy et al., 1994; Hennighausen et al., 1994; Thepot et al., 1995; Massoud et al., 1996; Litscher et al., 1999). These effects might stem from ectopic expression of the transgene or from transfer of the recombinant proteins from mammary gland to blood. Animals with variable levels or ectopic expression of the transgene presumably
will be identified in the development of the transgenic lines. Should products from such individuals be released to commercialization channels, they could pose a food safety concern unless the protein of concern is screened for and found absent. It is expected that well-established transgenic lines to be used in routine production will have been subjected to selection, and that concerns posed by unstable or ectopic gene expression will have been addressed to a large degree. Should pharmaceuticals or other biologically active proteins enter the food supply through products of such animals, associated food safety concern could be high. Additionally, the effects of transgene expression on animal wellbeing might indirectly affect the safety of food products derived from their tissues through stress-mediated mechanisms.
Expression of transgenes also might be intended to change the nutritional attributes or improve the safety of food products. For example, expression of transgenes in milk might optimize milk composition, add neutraceuticals to milk, or reduce the incidence of infectious disease (Zuelke, 1998; Houdebine, 2000). Several systems are being developed to reduce lactose concentration in milk (Alton et al., 1998; Whitelaw, 1999). Secretion of bovine -lactalbumin (an enzyme) in pig milk increased piglet growth (Bleck et al., 1998; Wheeler, 1994), showing the potential for changing the nutritive value of milk. Immunoglobulin A directed against viruses infecting the digestive tract might be expressed in milk (Saif and Wheeler, 1998; Castilla et al., 1998; Sola et al., 1998), and viral antigens activated by oral administration might be used to vaccinate humans and animals against viral disease (Houdebine, 2000). Changes of these types raise a moderate level of food safety concern. Claims of nutritional attributes, safety, and efficacy of milk or other food products from transgenic animals must be demonstrated.
Animals might be developed to produce food products designed to fit special human dietary needs. Possible future products might include milk that lacks the most common allergenic protein, eggs that are lower in cholesterol, meat with enhanced vitamin content, or fat content modified in quality or quantity (Young, 2002). The nutrient profiles of meat and animal products are well documented, and changes in this profile raise concerns. Changes might be unwanted by some consumers, and might add value for others. If these changed products were labeled in order to appeal to targeted consumers, and identifiable to those who have medical or other reasons to avoid such foods, they would be of low concern. Novel proteins also can be produced by genetic engineering. Although proteins are necessary components of the human diet, they can exert undesirable effects, including: (1) allergenicity and hypersensitivity, (2) bioactivity, and (3) toxicity.
Allergenicity and Hypersensitivity
Food allergies are adverse reactions to a protein or glycoprotein in food that elicits a heightened response of the immune system in some people. Among several types of immunologic responses causing food allergies, the most common type of reaction is mediated by allergen-specific immunoglobulin E (IgE) antibodies. IgE-mediated reactions are known as immediate or acute hypersensitivity reactions because symptoms occur within minutes to several hours after ingestion of the allergenic food. Food allergies also include delayed hypersensitivity reactions whose mechanisms are less clear. These include cell-mediated reactions where the onset of symptoms occurs more than eight hours after ingestion of the allergenic food. In the United States, the prevalence of food allergies is 1.5 percent of the general population, and 5 percent of children under three years of age (Sampson, 1997). The prevalence of these types of reactions in infants remains uncertain, but cases have been well documented (FAO, 2001). Many children outgrow their food allergies (Sampson, 1997; Taylor et al., 1999). There are eight foods or food groups that account for more than 90 percent of the food allergies in the United States. These include cow’s milk, eggs, fish, crustaceans, peanuts, soybeans, tree nuts, and wheat (Taylor et al., 1999). However, more than 160 other foods have been identified as causing food allergies (Hefle et al., 1996).
The genetic engineering of animals intended for use as food will involve the expression of new proteins in animals; hence the safety, including the potential allergenicity of the newly introduced proteins, will have to be assessed. While most known allergens are proteins, only a few of the innumerable proteins found in foods are allergenic under typical circumstances of exposure (Taylor and Hefle, 2001). While the common sources of food allergens have been identified and characterized, many others are less known and poorly understood. If the new protein originates from a known allergenic source or its amino acid sequence is similar to that of a known allergen, the protein can be tested to determine whether it causes a reaction with sera from individuals with known food allergies. However, the potential allergenicity of a protein can be reasonably assessed only when the protein is known to trigger an immune response in sensitive subjects. By contrast, the potential allergenicity of a protein of unknown allergenicity cannot easily be predicted, as no immunosera of allergic subjects are available (Mendieta et al., 1997). A more difficult issue arises when a new protein comes from a source that historically is not a human food. Assessing the potential allergenicity of transferred proteins remains one of the most difficult aspects in the overall safety assessment of transgenic foods. An adequate allergenicity assessment will require an understanding of several factors, including the source of the transferred protein, its level of expression, the physical and chemical properties of the protein, and any structural similarities to known allergens. No single factor can be considered definitive,
but consideration of all these factors together might provide some indication of potential allergenicity (Gendel, 1998 a,b; Taylor and Hefle, 2001). Concerns regarding the potential allergenicity of these new compounds in food are the lack of predictive and testing methodologies, and the feasibility of performing adequate assessments for an increasing number of transgenic products. The possibility that particular novel gene products might trigger allergenicity or hypersensitivity responses will vary with the gene product at issue, and poses a moderate level of food safety concern (i.e., the likelihood of a reaction is of moderate concern, but when it occurs, it could be a severe). The committee notes that the World Health Organization and other bodies are working to develop and standardize protocols for testing allergenicity.
Other Bioactive Compounds
In some cases, the aim of genetic engineering is to enhance expression of an economically important trait (e.g., growth rate; Pursel et al., 1990; Devlin et al., 2001), or to improve resistance to disease (e.g., mastitis; Kerr et al., 2001). In others, animals will be engineered to express proteins of pharmaceutical interest (Wright et al., 1991). These applications involve the expression of biologically active proteins or polypeptides encoded by a transgene. The possibility exists that such molecules could retain their bioactivity after consumption, raising a food safety concern.
The bioactive product of a transgene, in most cases, will be a protein or, in some cases, a polypeptide. During digestion, proteins and larger polypeptides largely are broken down into small peptide fragments and amino acids by proteolytic enzymes in the digestive tract. Di- and tripeptides that are absorbed into digestive epithelial cells are broken down into amino acids by intracellular enzymes. Few intact, small peptides are absorbed into the bloodstream during digestion. Substantial degradation of the intact protein effectively destroys its original bioactivity, since that bioactivity depends on the integrity of at least a portion of the protein or peptide. Allergenicity might, of course, remain a problem for sensitive individuals. Many food allergens are absorbed into the bloodstream. While most intact proteins generally are not absorbed into the bloodstream of healthy adults with an intact, properly functioning digestive system, absorption might occur in individuals whose digestive epithelium has been compromised by disease or injury (Simon, 1985), possibly posing allergenic response. In such cases, the normal array of digestive enzymes might be absent, or the integrity of the epithelium as a barrier might be compromised. Gastroenteritis, for example, can reduce the secretion of digestive enzymes and cause the breakdown of digestive epithelium, resulting in the passage of intact proteins and peptides into the bloodstream. A food safety concern thus arises
when individuals whose digestive system has been compromised by disease, injury, or advanced age ingest foods containing bioactive proteins or peptides.
The digestive epithelium of newborn infants permits the transient absorption of whole proteins or large protein fragments until closure of the gut epithelium occurs. Closure is facilitated by breastfeeding and delayed in infants that are formula fed. The timing of the closure might range from weeks to months, depending on dietary factors. Prior to closure, a wide variety of intact proteins might cross the digestive epithelium by a non-selective mechanism and enter the bloodstream. Thus, consumption of food (especially milk) containing bioactive proteins or peptides could result in the transfer of such molecules into the bloodstream of newborn infants. This possibility raises a concern regarding recombinant bioactive molecules present in milk used in infant formulas.
Bioactive peptides and proteins also might exert their effects in the digestive system, prior to absorption. For example, recombinant human bile-salt stimulated lipase (BSSL) has been expressed in the milk of transgenic sheep; this protein is intended for oral administration as a therapeutic agent for treating patients suffering from pancreatitis (PPL Therapeutics, 2001). Consumption of food products (i.e., milk and meat) from animals expressing bioactive molecules such as BSSL could alter digestion in otherwise healthy individuals, and presents a food safety concern. Lysostaphin, a bactericidal protein expressed by certain bacteria, has been expressed in murine milk, where it reduced mastitis caused by Staphylococcus aureus (Kerr et al., 2001). Transgenic cattle expressing lysostaphin in milk have been generated with the intent of reducing mastitis in that species (Suszkiw, 2001). Similarly, Jia et al. (2000) proposed production of transgenic fish expressing the hybrid antimicrobial peptide cecropin-melittin for control of fish pathogens. Preliminary studies using injections demonstrated the effectiveness of the antimicrobial peptide to protect fish against infections and suggested that the strategy of overexpressing the peptides in transgenic fish might provide a method of decreasing bacterial disease problems in fish. Milk containing lysostaphin or fish expressing cecropin-melittin could alter the balance of digestive tract flora of consumers of these products; in addition, widespread use of such antimicrobial agents also could foster the emergence of lysostaphin-resistant strains of pathogenic S. aureus or Vibrio anguillarum. Thus, food products containing antimicrobial proteins might present a food safety concern in view of their potential to alter the balance of consumers’ intestinal flora, and might foster the evolution of microbial strains resistant to specific agents.
Many genetically engineered fish and shellfish express an introduced growth hormone (GH) gene—most often a fish GH gene—in order to promote rapid growth. Hence, it is particularly important to make sure that such a transgene product has no biologic activity in humans or animals that consume fish or shellfish expressing such a transgene. The food safety of GH proteins was evaluated when administration of recombinant bovine GH (rbGH, also
called somatotropin) to dairy cattle was considered by the FDA in the late 1980s (Juskevich and Guyer, 1990). The FDA cited data showing that non-primate GH proteins are not biologically active in humans; nor are fragments of the GH molecule, nor insulin-like growth factors secreted by the host in response to GH administration. Neither bovine GH nor bovine insulin-like growth factor I (IGF-I) were orally active in rats, a species responsive to parenterally administered bovine GH. The FDA also cited studies showing that bovine, ovine, whale, and porcine growth hormones are not biologically active in humans, which suggests that piscine growth hormones are unlikely to be biologically active in humans. The degree to which full-term human infants absorb intact proteins is equivocal; FDA cited studies showing that concentrations of IGF-I in milk of rbGH-treated cows was within the physiologic range in human breast milk, and IGF-I is denatured under conditions used to process cow’s milk for infant formula. In the unlikely case that products of GH-transgenic fish or shellfish would be fed to human infants, cooking would denature active GH and IGF-I molecules.
Against the background of the discussion above, the committee regards the likelihood that a bioactive product poses a hazard will vary among gene products, food products, and consumers, in various cases posing a low to moderate level of food safety concern. For a susceptible individual, however, such a hazard could have severe consequences.
Many toxins are well studied and genes for known toxins would not be transferred purposefully into a food animal. As noted earlier, genetic engineering will have the potential of introducing novel proteins expressed in food animals. Because proteins generally are broken down in the digestive system into common amino acids, the direct toxicity of proteins (beyond the possibilities of allergenicity and bioactivity discussed above) is unusual and generally of low food safety concern. Purposefully expressed proteins that remain intact or otherwise pose a potential safety concern presumably will be fully evaluated in the pre-market review and approval process, and thus pose a relatively low concern.
Of greater concern to the committee with respect to possible toxicity are the unintended and unanticipated effects and byproducts of the genetic engineering of a food animal, including but not limited to these novel proteins. For example, the engineering could alter a metabolic process that then results in a toxic metabolite being present in edible tissue. The question, as suggested earlier in the chapter, is whether edible products of genetically engineered animals have been screened adequately to detect the presence of unanticipated compositional changes that might introduce toxicity. Assuming that adequate analytic methods and screening protocols exist (an issue that the committee did