General Considerations for Feeding and Diet Formulation
A laboratory animal's nutritional status influences its ability to reach its genetic potential for growth, reproduction, and longevity and to respond to pathogens and other environmental stresses. A nutritionally balanced diet is important both for the welfare of laboratory animals and to ensure that experimental results are not biased by unintended nutritional factors.
Laboratory animals require about 50 nutrients in appropriate dietary concentrations. Tables detailing the estimated minimum nutrient requirements of laboratory animals are presented in this report. It is important to recognize that the estimated requirements in these tables have been determined under specific restrictive conditions. Feed palatability and intake, nutrient absorption and utilization, and excretion can be affected by physicochemical characteristics of feeds such as physical form, sensory properties, naturally occurring refractory or antinutritive compounds, chemical contaminants, and conditions of storage. Many biological factors also affect nutrient requirements.
FACTORS AFFECTING NUTRIENT REQUIREMENTS
Genetic differences among species, breeds, strains, stocks, sexes, and individuals may affect nutrient requirements. For example, the lack of L-gulonolactone oxidase (a key enzyme required for the synthesis of ascorbic acid) in some species is apparently the consequence of genetic mutation (Chatterjee, 1978). L-gulonolactone oxidase activity differs among rodent species, among rat strains, and between sexes within rat strains (Jenness et al., 1980). A mutant rat has even been discovered that, like the guinea pig, lacks L-gulonolactone oxidase and has an obligatory dietary requirement for ascorbic acid (Mizushima et al., 1984; Horio et al., 1985). There is evidence that mouse strains may differ in requirements for riboflavin, pantothenic acid, and other nutrients (Fenton and Cowgill, 1947; Lee et al., 1953; Luecke and Fraker, 1979). Genetic differences in growth potential among species, strains, and sexes may influence the daily requirements for amino acids and other nutrients that are incorporated into tissues (Fenton, 1957; Goodrick, 1973).
STAGE OF LIFE
Nutrient requirements change during stages of the life cycle, especially in response to growth, pregnancy, or lactation. Synthesis of tissues or products requires amino acids, fatty acids, minerals, glucose, or other substrates as well as increased amounts of vitamins and associated cofactors. Research on farm animals demonstrates that rates of growth and of milk production affect nutrient requirements (National Research Council, 1984, 1985, 1988, 1989). The same is probably true for laboratory animals; however, few conclusive studies have been reported. As a result, for most nutrients, it is not currently possible to establish separate requirements for various stages of life for individual laboratory animal species.
Nutrient requirements are usually studied under controlled conditions with minimal diurnal or seasonal variation in temperature, light cycle, or other environmental conditions. Marked modification in these conditions may alter nutrient requirements. For example, exposure to temperatures below the lower threshold of the thermoneutral zone increases energy requirements as animals are obliged to expend energy to maintain a constant body temperature. The consequent increase in food intake may permit the
Housing types can also affect the amounts of nutrients needed in diets. For example, laboratory rodents maintained in either galvanized cages or cages with solid bottoms may have a lower dietary requirement for zinc because of the availability of zinc from the feces and cage materials. Solubilized minerals in drinking water (such as copper from copper water lines) may affect the amounts of these minerals that must be supplied by the diet. If laboratory animals ingest bedding or other "nonfood" materials, these may provide an unintended source of some nutrients or toxins. In studies of the requirements of laboratory animals for constituents that might be needed at extremely low concentrations, even the air supply may be a significant source of contamination.
Under normal rearing conditions, laboratory animals harbor populations of microorganisms in the digestive tract. These microorganisms generate various organic constituents as products or by-products of metabolism, including various water-soluble vitamins and amino acids. The extent to which these nutrients contribute to the nutrition of the host may be substantial but varies according to species, diet composition, and rearing conditions. In the rat and mouse, most of the microbial activity is in the colon, and many of the microbially produced nutrients are not available to the host unless feces are consumed, as is common for rats and other rodents (Stevens, 1988). Prevention of coprophagy may require an increase in the nutrient concentrations that must be supplied by the diet. The loss of some or all microbial symbionts in animals free of specific pathogens and germ-free animals, respectively, may also alter microbial nutrient synthesis and, thereby, influence dietary requirements. Adjustments in nutrient concentrations, the kinds of ingredients, and methods of preparation must be considered when formulating diets for laboratory animals reared in germ-free environments or environments free of specific pathogens (Wostmann, 1975).
Experimental procedures may produce stress or otherwise alter food intake. For example, surgical procedures or test substances in diets may lead to anorexia, necessitating the provision of more palatable diets or diets with elevated nutrient concentrations. Experimental protocols that require restriction of the amount of food offered alter the intakes of all nutrients unless dietary concentrations are altered to account for changes in food consumption.
Alterations in dietary energy density usually cause a change in feed intake. If high-energy diets are used, it may be necessary to increase nutrient concentrations in the diet to compensate for decreased food consumption. Other interactions occur between nutrients, such as competition for absorption sites among certain minerals that share common active transport systems. Thus in formulating diets containing unusual nutrient concentrations, the potential effects on other nutrients must be considered and adjustments made in nutrient concentrations, if appropriate.
FORMULATION OF DIET TYPES
Diet formulation is the process of selecting the kinds and amounts of ingredients (including vitamin and mineral supplements) to be used in the production of a diet containing planned concentrations of nutrients. Choice of ingredients will be influenced by the species to be fed and the experimental or production objectives. Target nutrient concentrations must take into account estimated nutrient requirements, possible nutrient losses during manufacturing and storage (National Research Council, 1973; Harris and Karmas, 1975), bioavailability of nutrients in the ingredients, and potential nutrient interactions.
Various types of diets are available for use with laboratory animals. Selection of the most appropriate type will depend on the amount of control required over nutrient composition, the need to add test substances, potential effects of feed microbes, diet acceptance by the animals, and cost. Wastage is also a problem with some types of diets, which may be a disadvantage if quantitative intake is to be measured.
The ideal diet for a particular animal colony will depend on production or experimental objectives. The diet must be sufficiently palatable to ensure adequate food consumption and must be nutritionally balanced so that the nutrients essential for the objectives are provided. It should also be free of substances or microorganisms that may be toxic or cause infection. Diets used in research also must be readily reproducible to ensure that the results can be verified by additional studies.
It is common to classify diets for laboratory animals according to the degree of refinement of the ingredients.
Diets formulated with agricultural products and by-products such as whole grains (e.g., ground corn, ground wheat), mill by-products (e.g., wheat bran, wheat middlings, corn gluten meal), high protein meals (e.g., soybean meal, fishmeal), mined or processed mineral sources (e.g., ground limestone, bonemeal), and other livestock feed in-
gredients (e.g., dried molasses, alfalfa meal) are often called natural-ingredient diets. Commercial diets for laboratory animals are the most commonly used natural-ingredient diets, but special diets for research animals may also be of this type. This type of diet is relatively inexpensive to manufacture and, if appropriate attention is given to ingredient selection, is palatable for most laboratory animals. However, variation in the composition of the individual ingredients can produce changes in the nutrient concentrations of natural-ingredient diets (Knapka, 1983). Soil and weather conditions, use of fertilizers and other agricultural chemicals, harvesting and storage procedures, and manufacturing or milling methods can all influence the composition of individual ingredients, with the result that no two production batches of feed are identical. The potential for contamination with pesticide residues, heavy metals, or other agents that might compromise experimental data is another disadvantage (International Council for Laboratory Animal Science, 1987). Natural-ingredient diets are usually unsatisfactory for studies to determine micronutrient requirements, for toxicological studies that are sensitive to low concentrations of contaminants, or for immunological studies that may be influenced by antigens in diets.
The formulation of natural-ingredient diets is complicated by the fact that each ingredient contains many if not most nutrients, so that an adjustment in the amount of any ingredient produces changes in the concentrations of most nutrients in the final product. Hence it is not possible to predetermine the concentration of each nutrient; rather diets are formulated to contain minimal concentrations of particular nutrients (such as crude protein, fiber, fat, calcium, and phosphorus), and other nutrients are added via vitamin and mineral premixes. In feeding domestic animals, cost considerations dictate that natural-ingredient diets be formulated using linear programming techniques that generate diet formulas that conform to set minimal and maximal nutrient concentrations, while minimizing ingredient costs. This has led to the marketing of variable formula products that differ in ingredient composition from batch to batch in response to changing ingredient prices. Such diets may be cost-effective for maintenance and rearing of laboratory animals, but they are too variable to be of use in nutritional, toxicological, or other types of experiments that may be affected by dietary constituents.
An alternative approach has been the development of fixed-formula diets in which the kinds and amounts of ingredients do not vary from batch to batch. These diets are often called open-formula diets when the formula is openly declared, as in specifications used to solicit competitive bids among manufacturers. A fixed-formula diet may contain multiple sources of protein, fat, and carbohydrate, thereby reducing the importance of variation in the composition of any particular ingredient from batch to batch (Knapka et al., 1974). A variety of ingredients also increases the probability that ultra-trace minerals of potential nutritional importance—such as chromium, nickel, and tin—will be provided at appropriate concentrations. Because it is difficult to demonstrate that these minerals are required, and because the amounts in most natural-ingredients are apparently adequate, these minerals are not typically included in mineral premixes for natural-ingredient diets.
The steps in formulating a natural-ingredient diet are reviewed by Knapka (1985) and the International Council for Laboratory Animal Science (1987). It is important to recognize that the bioavailability of nutrients may be lower in natural-ingredient diets than in purified diets. Factors that may affect bioavailability include the chemical form of nutrients, constituents that may bind nutrients (such as phytate, tannins, and lignin), nutrient interactions, and effects of processing. Thus it is prudent to include nutrients at concentrations higher than the minimal requirements but within the safe range. Information on the vitamin and mineral tolerances of animals has been summarized in prior reports (National Research Council, 1980, 1987). Common practice is to use greater margins of safety for particularly labile vitamins and for trace minerals.
PURIFIED AND CHEMICALLY DEFINED DIETS
Diets that are formulated with a more refined and restricted set of ingredients are designated purified diets. Only relatively pure and invariant ingredients should be used in these formulations. Examples of such ingredients are casein and soybean protein isolate (as sources of protein), sugar and starch (as sources of carbohydrate), vegetable oil and lard (as sources of fat and essential fatty acids), a chemically extracted form of cellulose (as a source of fiber), and chemically pure inorganic salts and vitamins. The nutrient concentrations in a purified diet are less variable and more easily controlled via formulation than in a natural-ingredient diet. However, even these ingredients may contain variable amounts of trace nutrients, and experimental diets intended to produce specific deficiencies may need to be even more restrictive as to ingredient specifications (International Council for Laboratory Animal Science, 1987). The potential for chemical contamination of these diets is also low. Purified diets are often used in studies of specific nutritional deficiencies and excesses. Unfortunately, they are not readily consumed by all species and are more expensive to produce than natural-ingredient diets.
Chemically Defined Diets
For studies in which strict control over nutrient concentrations and specific constituents is essential, diets have been made with the most elemental ingredients available, such as individual amino acids, specific sugars, chemically defined triglycerides, essential fatty acids, inorganic salts, and vitamins. Such diets are called chemically defined diets; they represent the highest degree of control over nutrient concentrations. Unfortunately, chemically defined diets are not readily consumed by most species of laboratory animals and are usually too expensive for general use. Although the nutrient concentrations in these diets are theoretically fixed at the time they are manufactured, the bioavailability of nutrients may be altered by oxidation or nutrient interactions during storage. Chemically defined diets that can be sterilized by filtration have been developed for use in germ-free and low-antigen studies (Pleasants, 1984; Pleasants et al., 1986).
The ingredients used in purified and chemically defined diets have the advantage that each is essentially the source of a single nutrient or nutrient class, which greatly simplifies the task of formulation. Each ingredient must be carefully selected on the basis of purity, consistency of supply and composition, and physicochemical properties, but the decision of how much to use is primarily a function of the planned nutrient concentration. Attention must be paid to providing sources of all essential nutrients because inadvertent omission of trace and ultra-trace nutrients in purified and chemically defined diets is more likely than with natural-ingredient diets. Margins of safety above requirement concentrations should be modest and relate to potential losses caused by oxidative degradation or other reactions that may occur during and after manufacture. The ingredients and formulation of purified diets have been described by Navia (1977).
Impurities remain a major concern with purified diets (International Council for Laboratory Animal Science, 1987). Protein sources may supply variable but unknown amounts of vitamins, minerals, and essential fatty acids; starch may contain traces of lipid and essential fatty acids; and oils may contain fat-soluble vitamins. Thus it is necessary to select specific ingredients if strict control of a particular nutrient is required. Protein sources used to produce trace mineral deficiencies include Torula yeast for chromium and selenium; lactalbumin for cobalt; casein for copper, iron, and manganese; and dried egg white for zinc (International Council for Laboratory Animal Science, 1987). Casein contains phosphorus; soybean protein contains phytate, which binds minerals (Wise, 1982). Casein is often extracted to reduce vitamin content, but even "vitamin-free casein" may have significant residual amounts of vitamin B6 (Quinn and Chan, 1979).
Chemically defined diets are formulated using chemically pure (analytical grade) nutrients such as amino acids, fatty acid esters, glucose, vitamins, and mineral salts. In selecting ingredients one must consider such factors as chemical stability and solubility (in liquid diets); obviously all essential nutrients must be added individually. The availability of the different chemical forms of nutrients is a primary concern in the formulation of chemically defined diets. For example, the l-isomeric forms of amino acids occur in natural food protein. However, the d-isomers of several of the essential amino acids will support growth in the rat. Of these, methionine alone appears to be as well utilized in either form (Wretlind and Rose, 1950). Details about the composition and use of chemically defined diets are provided by Pleasants and colleagues (Pleasants et al., 1970; Pleasants, 1984; Pleasants et al., 1986).
PHYSICAL FORM OF DIETS
Diets for laboratory animals can be provided in different physical forms. The most common form in use for laboratory animals is the pelleted diet, which is typically formed by adding water to the mixture of ground ingredients and then forcing it through a die. The size and shape of the holes in the die determine pellet shape and rotating blades control the length; the diet is then dried to firmness. Binders are sometimes used to improve pellet quality. Pelleted diets are easy to handle, store, and use; reduce dust in animal facilities; prevent animals from selecting choice ingredients; and tend to minimize wastage. It is not easy, however, to add test compounds or otherwise alter pelleted diets after manufacture.
Extruded diets are similar to pelleted diets except the meal is forced through a die under pressure and at high temperature after steam has been injected, so the product expands as it emerges from the die. Extruded diets are less dense than pelleted diets and are preferred by some animals (e.g., dogs, cats, and nonhuman primates). Extruded diets are not commonly used for laboratory rodents because of the increased wastage during feeding and higher production costs.
Diets in meal form are sometimes used because they permit incorporation of additives and test compounds after the diet has been manufactured. These diets are often inefficient, however, because large amounts may be wasted unless specially designed feeders are available. Also, meals cake under certain storage conditions. An additional problem is that dust generated from the feed may be hazardous if toxic compounds have been added. One solution to this problem is to add jelling agents and water to the meal to form a jelled mass that can be cut into cubes for feeding; however, the jelling agents may contain carbohydrate,
amino acids, or minerals that must be accounted for in diet formulations. The gel diet requires refrigeration to retard microbial growth and must be fed daily or more frequently to maintain moisture content and thus food intake.
Crumbled diets are prepared by crushing pelleted or extruded diets and screening particles to the most appropriate size for a particular age or size of laboratory animal, including fish and birds. Crumbled diets offer a method of presenting small particles of diet that, theoretically, contain all dietary ingredients present in pelleted diets. Crumbled diets offer the convenience, without the problems, of diets in meal form; they are not frequently used for rodents, however.
Liquid diets have been developed to accommodate specific requirements such as filter sterilization. Liquid diets are often used in studies of the effects of alcohol on nutrient utilization and requirements. In some cases purified diets will take the form of a stable emulsion when blended with water (Navia, 1977). Neonatal animals are also fed liquid diets that are derived primarily from milk products. As with gel diets, care must be taken to store liquid diets properly to avoid microbial growth.
MANUFACTURE AND STORAGE PROCEDURES AND OTHER CONSIDERATIONS
The efficient manufacture of natural-ingredient diets requires a large capital investment for facilities, milling apparatus, and inventories of ingredients that are least expensive when purchased in bulk. Therefore, these laboratory animal diets are usually commercially manufactured. Laboratory animal diets should not be manufactured or stored in facilities used for farm feeds or any products containing additives such as rodenticides, insecticides, hormones, antibiotics, growth factors, or fumigants. Areas where ingredients and diets are stored and processed should be kept clean and enclosed to prevent entry of feral rodents, birds, and insects. Routine pest control is essential.
The initial step in manufacturing natural-ingredient diets is to grind all ingredients to a similar particle size so they can be uniformly blended into a homogeneous mixture. Particle size depends on the pore size of the screen used in a hammer mill or other grinder. The optimal particle size of ground ingredients depends on the kind of ingredients involved and the planned physical form of the final product. Grinding may improve the digestibility of the ingredient by increasing the surface area that is exposed to digestive enzymes; however, grinding can also increase subsequent rates of destruction of nutrients by increasing exposure to atmospheric oxygen and by releasing enzymes responsible for autocatalytic processes.
Ingredients used in large amounts are added directly, while those used in small amounts, such as vitamins and minerals, are added via premixes. Separate vitamin and mineral premixes should be used to minimize destruction of vitamins by oxidation reactions catalyzed by minerals. Premixes should be prepared with a carrier such that a sufficient amount is added (e.g., 1 percent of the diet) to avoid weighing errors and to ensure homogeneous distribution of these micronutrients. Errors such as omitting ingredients or adding incorrect amounts can be minimized by verifying on a check sheet each ingredient as it is added.
The length of time a particular combination of ingredients should be mixed for maximal homogeneity depends on a number of factors including particle size, particle density, mixer speed, and mixer size. Overmixing can occur, resulting in particle separation associated with differences in density, physical form, and susceptibility to static electrical charges that can develop in mixers (Pfast, 1976).
Ground, mixed feeds are often pelleted. Ingredient composition, amount of moisture and heat, die size, operating conditions, and other factors influence the size, hardness, and nutrient concentrations of pellets. Some loss of labile vitamins may occur during pelleting, especially if it is done at high temperatures; however, the heat of the pelleting process may also inactivate enzymes, reduce bacterial populations in the diet, and, in some cases, improve digestibility (Slinger, 1973; International Council for Laboratory Animal Science, 1987). Many species prefer pelleted products and, therefore, increase voluntary food intake. Pelleting also allows for a reduction in wastage.
PURIFIED AND CHEMICALLY DEFINED DIETS
Purified or chemically defined diets can be efficiently prepared in laboratories or diet kitchens with a minimum amount of special apparatus. All diets for laboratory animals should be prepared in facilities used only for this purpose and under strict rules to prevent contamination or errors in the kinds and amounts of ingredients used. Navia (1977) presents a detailed discussion regarding the preparation of purified diets. Purified diets may be pressed into tablets, pelleted, or fed as a powder, paste, or gel. Great care must be exercised to ensure homogeneous mixing when test substances are added to purified diets (International Council for Laboratory Animal Science, 1987).
ENVIRONMENTAL CONDITIONS OF STORAGE AREAS
Nutrient stability of feeds generally increases as temperature and humidity decrease. The shelf-life of any particular lot of feed depends on the environmental conditions of the storage area. Feed stored where temperature and
humidity are high can deteriorate within several weeks. Natural-ingredient diets stored in air-conditioned areas should be used within 180 days of manufacture; diets containing vitamin C should be used within 90 days of manufacture (National Institutes of Health, 1985). Vitamins C and A are especially labile. Diets stored for long periods or under unusual environmental conditions should be assayed for nutrients prior to use. Diets formulated without antioxidants or with large amounts of highly perishable ingredients, such as fat, may require special handling or storage procedures. Sterilization of diets is essential for germ-free and specific-pathogen-free animals and is often advisable for conventionally reared animals. Autoclaving at temperatures greater than 100° C can be effective in achieving complete sterility so long as steam penetrates the entire load for a sufficient amount of time, but excessive exposure should be avoided as this exacerbates vitamin losses and affects protein quality (Zimmerman and Wostmann, 1963; Coates, 1984). Some autoclaves permit rapid heating to high temperatures under vacuum, with consequent reduction of exposure time and nutrient losses. Diets can be sterilized by ionizing radiation with less damage to nutrients than is caused by heat sterilization as long as diets are packed under vacuum or nitrogen and little moisture is present (Ley et al., 1969; Coates, 1984). It has been suggested that supplements of heat-labile vitamins be increased two to fourfold in diets to be sterilized to compensate for potential losses during sterilization (International Council for Laboratory Animal Science, 1987).
High-lipid diets require several formulation and storage precautions. Unsaturated lipid in the diet is susceptible to oxidation, which reduces the amount of available essential fatty acids (EFA). Rancid characteristics of oxidized lipid may reduce diet acceptability. An antioxidant (butylated hydroxytoluene or ethoxyquin at 0.01 to 0.02 percent of oil) should be added to the oil (American Institute of Nutrition, 1980). As an additional precaution to reduce decomposition, diets should be stored at temperatures £4° C in a container that has been flushed with argon or nitrogen before sealing (Fullerton et al., 1982). When very highly unsaturated oils are fed (e.g., fish oils), the diet should be changed every 24 to 48 hours (Johnston and Fritsche, 1989). In addition, extra DL-α-tocopherol (e.g., 5 to 10 times the concentration in low-lipid diets) may need to be included in the diets to prevent in vivo peroxidation (Garrido et al., 1989; Johnston and Fritsche, 1989). (The fatty acid composition of several common dietary oils is shown in Appendix Table 1 to help researchers choose the dietary lipid source most appropriate for a particular experimental protocol.)
QUALITY ASSURANCE AND POTENTIAL CONTAMINANTS
Given the potential importance of diet quality to consistent experimental results, a routine program to assay nutrients should be implemented to verify the composition of diets fed to laboratory animals. Although accidental omission or inadvertent inclusion of ingredients is uncommon, when it does occur it can have disastrous consequences. Discrepancies between expected and actual nutrient concentrations in laboratory animal feeds can occur as a result of errors in formulation, losses of labile nutrients during manufacture and storage, and variation of the nutrient content of ingredients from average values presented in tables (e.g., National Research Council, 1976, 1982).
Assaying is particularly important when commercial diets of undeclared formula are used because nutrient concentrations may deviate from those published by the manufacturer. For example, because commercial cornstarch can contain significant quantities of linoleic acid (Holman, 1968), diets designed to induce essential fatty acid deficiency are more effective when sucrose, rather than starch, is used. Batch-to-batch variation in nutrient composition may be substantial even in fixed-formula diets made from natural-ingredients. For example, in 94 batches of a fixed-formula diet assayed, concentrations varied about sixfold for vitamin A, nearly fourfold for thiamin, and twofold for calcium (Rao and Knapka, 1987). However, some of this variation may have been the result of sampling or analytical error. Variation in purified diets, although of lesser magnitude, may be important if nutrients are provided at requirement concentrations.
Samples for assaying should be taken from multiple bags or containers of feed. Care must be exercised to obtain a representative well-mixed subsample, especially if any settling or segregation of diet particles has occurred. Nutrient analyses should be conducted by a reputable laboratory and in accordance with Association of Official Analytical Chemists methods of analysis (Association of Official Analytical Chemists, 1990). Analyses should at least include proximate constituents (moisture, crude protein, ether extract, ash, and crude or acid-detergent fiber) and any nutrients of particular interest. Some vitamins and other nutrients are difficult to assay because of low concentrations or interfering compounds or both.
Potential chemical and biological contaminants of feeds are a major source of concern for toxicological and immunological research but may impact on other types of experiments as well. The International Council for Laboratory Animal Science noted seven unwanted substances in laboratory animal diets (International Council for Laboratory Animal Science, 1987):
pests (especially insects and mites);
bacteria, bacterial toxins, and mycotoxins;
natural plant toxins;
breakdown products of nutrients;
nitrates, nitrites, and nitrosamines; and
In addition, errors in formulation or manufacture can result in hazardous amounts of those nutrients, such as vitamins A and D, and copper, that can be toxic at concentrations not greatly in excess of requirements. The greater potential for contaminants and other unwanted substances in natural-ingredient diets may make these diets unsuitable for certain types of research. However, fixed-formula diets can omit ingredients that tend to be particularly variable (such as some fish and meat meals) and rigorous pretesting of raw ingredients for specific contaminants may eliminate most potential problems. For example, in the manufacture of a fixed-formula rodent diet, it was necessary to restrict the fish meal to batches that had been demonstrated to be low in nitrosamine concentrations (Rao and Knapka, 1987).
Recommended maximum acceptable concentrations of chemical contaminants have been published by various agencies (e.g., Food and Drug Administration, 1978; Environmental Protection Agency, 1979; International Council for Laboratory Animal Science, 1987). Based on observed contaminant amounts and potential toxic effects, Rao and Knapka (1987) provide a list of recommended limits for about 40 contaminants, including aflatoxins, nitrosamines, heavy metals, chlorinated hydrocarbons, organophosphates, polychlorinated biphenyls, nitrates and nitrites, preservatives, and estrogenic activity. They also proposed a scoring system for diets to be used in chemical toxicology studies that permits separation of tested diets into those acceptable for long-term use, those acceptable only for short-term (transitory) use, and those that should be rejected. Testing for contaminants should be routine in toxicological research and may be valuable on at least an occasional basis in other studies.
Good manufacturing technique, appropriate storage conditions, and feeders that prevent fecal and urinary contamination of diets will minimize, but not eliminate, bacterial and other biological agents in diets. Diet is a potential source of pathogens for laboratory animals (Williams et al., 1969). Clarke et al. (1977) described procedures for sampling and assaying feeds for various pathogenic organisms as well as standards regarding the number and kinds of organisms acceptable in diets. As mentioned previously, sterilization procedures are employed for diets fed to germ-free and specific-pathogen-free animal colonies. Because microbial residues may be unacceptable in the low-antigen diets required for immunologic studies, the use of chemically defined diets may be necessary.
Traditionally, maximal growth and reproduction have been used as criteria for the evaluation of laboratory animal diets. However, evidence from a number of studies indicate that restricting the caloric intake of laboratory animals may have beneficial effects on life span, the incidence and severity of degenerative diseases, and the onset and incidence of neoplasia (Weindruch and Walford, 1988; Snyder and Towne, 1989; Yu, 1990; Bucci, 1992). Based on these results, allowing animals to eat ad libitum to produce maximum growth and reproduction may not be consistent with objectives of long-term toxicological and aging studies.
It is important to achieve caloric restriction of test animals without producing unintended nutrient deficiencies. Elevation of nutrient concentrations in the diet may be necessary to ensure that the nutrient intake of animals whose eating is restricted is comparable to that of animals allowed to eat ad libitum. Unfortunately, relatively little information is available about the extent to which caloric restriction affects nutrient requirements.
American Institute of Nutrition. 1980. Second report of the ad hoc committee on standards for nutritional studies. J. Nutr. 110:17-26.
Association of Official Analytical Chemists. 1990. Official Methods of Analyses of the Association of Official Analytical Chemists, 15th. Ed., K. Helrich, ed. Arlington, Va.: Association of Official Analytical Chemists.
Bucci, T. J. 1992. Dietary restriction: Why all the interest? An overview. Lab Anim. 21:29–34.
Chatterjee, I. B. 1978. Ascorbic acid metabolism. World Rev. Nutr. Dietet. 30:69–87.
Clarke, H. E., M. E. Coates, J. K. Eva, D. J. Ford, C. K. Milner, P. N. O'Donoghue, P. P. Scott, and R. J. Ward. 1977. Dietary standards for laboratory animals: Report of the Laboratory Animals Centre Diets Advisory Committee. Lab. Anim. 11:1–28.
Coates, M. E. 1984. Sterilization of diets. Pp. 85–90 in The germ-free animal in biomedical research, M. E. Coates, and B. E. Gustafsson, eds. Laboratory Animals Handbooks 9. London: Laboratory Animals Ltd.
Environmental Protection Agency. 1979. Proposed health effects test standards for toxic substances control act. Test rules. Good laboratory practice standards for health effects. Federal Register, Part 2, 27334–27375; Part 4, 44054–44093.
Fenton, P. F. 1957. Hereditary factors in protein nutrition. Am. J. Clin. Nutr. 5:663–665.
Fenton, P. F., and G. R. Cowgill. 1947. The nutrition of the mouse. I. A difference in the riboflavin requirements of two highly inbred strains. J. Nutr. 34:273–283.
Food and Drug Administration. 1978. Nonclinical laboratory studies. Good laboratory practice regulations. Federal Register, Part 2, 59986–60025.
Fullerton, F. R., D. L. Greenman, and D. C. Kendall. 1982. Effects of storage conditions on nutritional qualities of semipurified (AIN-76) and natural-ingredient (NIH-07) diets. J. Nutr. 112:567–573.
Garrido, A., F. Garrido, R. Guerro, and A. Valenzuela. 1989. Ingestion of high doses of fish oil increases the susceptibility of cellular membranes to the induction of oxidative stress. Lipids 24:833–835.
Goodrick, C. L. 1973. The effects of dietary protein upon growth of inbred and hybrid mice. Growth 37:355–367.
Harris, R. S., and E. Karmas, eds. 1975. Nutritional Evaluation of Food Processing, 2nd Ed. Westport, Conn.: AVI Publishing.
Holman, R. T. 1968. Essential fatty acid deficiency. Prog. Chem. Fats Other Lipids 9:275–348.
Horio, F., K. Ozaki, A. Yoshida, S. Makino, and Y. Hayashi. 1985. Requirement for ascrobic acid in a rat mutant unable to synthesize ascorbic acid. J. Nutr. 115:1630–1640.
International Council for Laboratory Animal Science. 1987. ICLAS Guidelines on the Selection and Formulation of Diets for Animals in Biomedical Research, M. E. Coates, ed. London: International Council for Laboratory Animal Science.
Jenness, R., E. C. Birney, and K. L. Ayaz. 1980. Variation of L-gulonolactone oxidase activity in placental mammals. Comp. Biochem. Physiol. B 67:195–204.
Johnston, P. V., and K. L. Fritsche. 1989. Nutritional methodology in dietary fat and cancer research. Pp. 9–25 in Carcinogenesis and Dietary Fat, S. Abraham, ed. Boston: Kluwer Academic.
Knapka, J. J. 1983. Nutrition. Pp. 51–67 in The Mouse in Biomedical Research . Vol. 3, H. L. Foster, J. D. Small, and J. G. Fox, eds. New York: Academic Press.
Knapka, J. J. 1985. Formulation of diets. Pp. 45–59 in Methods for Nutritional Assessment of Fats, J. Beare-Rogers, ed. Champaign, Ill.: American Oil Chemists Society.
Knapka, J. J., K. P. Smith, and R. J. Judge. 1974. Effect of open and closed formula rations on the performance of three strains of laboratory mice. Lab. Anim. Sci. 24:480–487.
Lee, Y. C. P., J. T. King, and M. B. Visscher. 1953. Strain difference in vitamin E and B12 and certain mineral trace-element requirements for reproduction in A and Z mice. Am. J. Physiol. 173:456–458.
Ley, F. J., J. Bleby, M. E. Coates, and S. J. Patterson. 1969. Sterilization of laboratory animal diets using gamma radiation. Lab. Anim. 3:221–254.
Luecke, R. W., and P. J. Fraker. 1979. The effect of varying zinc levels on growth and antibody mediated response in two strains of mice. J. Nutr. 109:1373–1376.
Mizushima, Y., T. Harauchi, T. Yoshizaki, and S. Makino. 1984. A rat mutant unable to synthesize vitamin C. Experientia 40:359–361.
National Institutes of Health. 1985. Guide for the Care and Use of Laboratory Animals, Publication No. 86–23. Bethesda, Md.: National Institutes of Health.
National Research Council. 1973. Effect of Processing on the Nutritional Value of Feeds. Washington, D.C.: National Academy Press.
National Research Council. 1976. Atlas of United States and Canadian Feeds. Washington, D.C.: National Academy Press.
National Research Council. 1980. Mineral Tolerance of Domestic Animals. Washington, D.C.: National Academy Press.
National Research Council. 1982. United States—Canadian Tables of Feed Composition. Washington, D.C.: National Academy Press.
National Research Council. 1984. Nutrient Requirements of Beef Cattle, Sixth Revised Edition. Washington, D.C.: National Academy Press.
National Research Council. 1985. Nutrient Requirements of Sheep, Sixth Revised Edition. Washington, D.C.: National Academy Press.
National Research Council. 1987. Vitamin Tolerance of Animals. Washington, D.C.: National Academy Press.
National Research Council. 1988. Nutrient Requirements of Swine, Ninth Revised Edition. Washington, D.C.: National Academy Press.
National Research Council. 1989. Nutrient Requirements of Dairy Cattle, Sixth Revised Edition. Washington, D.C.: National Academy Press.
Navia, J. M. 1977. Animal Models in Dental Research. Tuscaloosa, Ala.: University of Alabama Press.
Pfast, H. B., ed. 1976. Feed Manufacturing Technology. North Arlington, Va.: Feed Production Council, American Feed Manufacturing Association.
Pleasants, J. R. 1984. The germ-free animal fed chemically defined ultrafiltered diet. Pp. 91–109 in The Germ-Free Animal in Biomedical Research. Laboratory Animals Handbook 9, M. E. Coates, and B. E. Gustafsson, eds. London: Laboratory Animals Ltd.
Pleasants, J. R., M. H. Johnson, and B. S. Wostmann. 1986. Adequacy of chemically defined, water-soluble diet for germ-free BALB/c mice through successive generations and litters. J. Nutr. 116:1949–1964.
Pleasants, J. R., B. S. Reddy, and B. S. Wostmann. 1970. Qualitative adequacy of chemically defined liquid diet for reproducing germ-free mice. J. Nutr. 100:498–508.
Quinn, M. R., and M. M. Chan. 1979. Effect of vitamin B6 deficiency on glutamic acid decarboxylase activity in rat olfactory bulb and brain. J. Nutr. 109:1694–1702.
Rao, G. N., and J. J. Knapka. 1987. Contaminant and nutrient concentrations of natural-ingredient rat and mouse diet used in chemical toxicology studies. Fund. Appl. Toxicol. 9:329–338.
Snyder, D. L., and B. Towne. 1989. The effect of dietary restriction on serum hormone and blood chemistry changes in aging Lobund-Wistar rats. Prog. Clin. Biol. Res. 287:135–146.
Slinger, S. J. 1973. Effect on pelleting and crumbling methods on the nutritional value of feeds. Pp. 48–66 in Effect of Processing on the Nutritional Value of Feeds: Proceedings. Washington, D.C.: National Academy of Sciences.
Stevens, C. E. 1988. Comparative physiology of the vertebrate digestive system. New York: Cambridge University Press.
Weindruch, R., and R. L. Walford. 1988. The Retardation of Aging and Disease by Dietary Restriction. Springfield, Ill.: Charles C Thomas.
Williams, L. P., J. B. Vaughn, A. Scott, and V. Blanton. 1969. A ten-month study of salmonella contamination in animal protein meals. J. Am. Vet. Med. Assoc. 155:167–174.
Wise, A. 1982. Interaction of diet and toxicity—The future role of purified diet in toxicological research. Arch. Toxicol. 50:287–299.
Wostmann, B. S. 1975. Nutrition and metabolism of the germ-free animal. World Rev. Nutr. Dietet. 22:40–92.
Wretlind, K. A. J., and W. C. Rose. 1950. Methionine requirement for growth and utilization of its optical isomers. J. Biol. Chem. 187:697–703.
Yu, B. P. 1990. Food restriction: Past and present status. Rev. Biol. Res. Aging 4:349–371.
Zimmerman, D. R., and B. S. Wostmann. 1963. Vitamin stability in diets sterilized for germ-free animals. J. Nutr. 79:318–322.