EFFECT OF BAKING AND TOASTING ON THE NUTRITIVE VALUE OF BREAD PROTEIN By Nickoias Palamidis A DISSERTATION Submitted to Michigan State University in partial fulfiiiment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Food Science and Human Nutrition 1978 ABSTRACT EFFECT OF BAKING AND TOASTING ON THE NUTRITIVE VALUE OF BREAD PROTEIN By Nickolas Palamidis Changes in the biological value, digestibility and nutritionally available lysine were followed in a series of wheat products representing different degrees of heat treatment of wheat flour. The products used were bread mix, whole bread (crumb and crust together), crumb, crust, light toasted bread (LTB), dark toasted bread (DTB), bread mix containing 4% non fat dry milk solids (NFDM-flour) and bread prepared from the last mix (NFDM-bread). The nutritional value of the above products was evalu- ated by the Protein Efficiency Ratio (PER) and Net Protein Ratio (NPR) methods. In terms of PER values, baking only (whole bread) caused a 38% decrease in the PER value of the unbaked ingredients (bread mix), the crumb showed a 2l% increase in PER, while the crust displayed a negative PER (the rats lost weight). Light toasting of the whole bread caused a 13% decrease in the PER, while dark toasting caused a 66% decrease in the PER of the untoasted bread. Addition of 4% NFDM-solids to the bread mix caused a 56% increase in the PER of the mix and 63% increase in the PER of baked bread (whole bread vs Nickolas Palamidis NFDM-bread). The apparent digestibility of the protein of these products was measured i vivo (rat feeding) and in vitro (using a trypsin-chymotrypsin-peptidase enzyme system). The apparent digestibility decreased, in general, with in- creasing heat treatment. The slight heating of the crumb (protected by the crust) increased its apparent digesti- bility in comparison with that of the unheated ingredients (bread mix). The values for apparent digestibility ob- tained by the in vitro method were in good agreement with the values obtained by the i vivo method. The available lysine of the products was measured by a biological method (rat growth assay) and a chemical method (FDNB-direct). The following values for available lysine in 9 available lysine per 100 g protein, were obtained by the rat growth assay: bread mix 2.14, whole bread 1.99, crumb 2.49, crust l.04, LTB l.88, DTB 1.48, NFDM-flour 2.71 and NFDM-bread 2.18. As percent of the total lysine the available lysine in the bread mix was 79.3%; compared rto the total lysine of the unbaked bread mix, the available lysine in the whole bread was 73.7%, in the crumb 92.2%, in the crust 38.6%, in the LTB 67.5%, in the DTB 54.2%, in the NFDM-flour 83.4% and in the NFDM-bread 67.1%. The values for available lysine obtained by the direct FDNB method were slightly higher than those obtained by the Nickolas Palamidis biological method. Available lysine values with both the 1 vivo and 1 vitro methods were highly correlated with the corresponding PER values (PER vs available lysine i vivo r = 0.989, PER vs available lysine in vitro r = 0.977). When the total lysine value for all products (bread mix, whole bread, crumb etc.) were compared to the corres- ponding available lysine values, high correlation coeffi- cients were obtained: r = 0.903 when the available lysine was determined by the biological method, r = 0.943 when available lysine was determined by the chemical method. To my parents ii ACKNOWLEDGMENTS The author wishes to express his sincere gratitude to his major professor, Dr. P. Markakis for his excellent guidance during the graduate studies at Michigan State Uni- versity and for his valuable assistance in the preparation of the thesis manuscript. Sincere appreciation is also extended to Dr. J.R. Brun- ner and Dr. C. Stine, professors at the Department of Food Science and Human Nutrition and to Dr. L. Bieber and Dr. H. Lillevik, professors at the Department of Biochemis- try for serving on the guidance committee and for their advice in preparation of this manuscript. Thanks are extended to Dr. w. Bergen, professor at the Department of Animal Husbandry, for performing the amino acid analyses. The financial assistance for part of my studies, from the ministry of Coordination of Greece, through the NATO program fellowships, is greatly appreciated. A very special acknowledgment goes to my parents, John and Anastasia Palamidis, for their continuous encour- agement and financial support during the course of my graduate studies in the United States of America. TABLE OF CONTENTS LIST OF TABLES. LIST OF FIGURES INTRODUCTION. LITERATURE REVIEW . . Amino Acid and Protein Requirements. . Evaluation of Nutritional Quality of Proteins. Protein Concentration of Foods Amino Acid Content of Foods. Amino Acid Availability. . . Biological Value of Wheat and Its Products Protein Quality Determination. Biological Methods Chemical Methods . Microbiological Methods. Enzymatic Methods. . . Digestibility of Food Proteins . Methods for Measuring Available Lysine Animal Growth Assays . Chemical Methods . . FDNB- Reactive Lysine by Difference Alternative Reagents to FDNB Protein Damage . . . . . . METHODS AND MATERIALS . . Preparation of the Samples Moisture, Fat and Ash Determination. Nitrogen Determination - Micro Kjeldahl. . . Protein Efficiency Ratio (PER) and Net Protein Ratio (NPR). . . . Apparent Digestibility in vivo Apparent Digestibility in vitro. . . . Available Lysine by Rat Growth Assay . Available Lysine by the FDNB Procedure Amino Acid Analysis. . Acid Hydrolysis. . . Enzymatic Hydrolysis Sulfur Containing Amino.Acids. Determination: iv Page vi viii d NCDNNOKDCDCDVOOUO Page Tryptophan. . . . . . . . . . . . . . . . . 64 RESULTS AND DISCUSSION . . . . . . . . . . . . . . . 67 Nutritional Value of the Heat Treated Products. 70 Digestibility . . . . . . . 77 Available Lysine by Rat Growth Assay. . . . . . 80 Available Lysine by the FDNB Method . . . . . . 84 Total Amino Acid Analysis . . . . . . . . . 89 Enzymatic Hydrolysis Ln vitro . . . . . 91 Destroyed, Inactivated and Available Lysine . . 97 Correlation Between Total and Available Lysine Values. . . 102 Correlation Between Ln vivo and Ln vitro Values for Available Lysine. . . . 105 Correlation Between in vitro Available Lysine and PER and NPR . . . . . . . . . . . . . . . 106 SUMMARY AND CONCLUSIONS. . . . . . . . . . . . . . . 107 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . 110 Table 10 11 12 13 14 LIST OF TABLES Dietary essential and non-essential amino acids for man. Essential amino acid composition of the whole egg protein and reference pattern (FAO/HHO, 1973). . . . Composition of basal diet for the PER. Composition of basal diet for available lysine determination. . Amino acid composition of diets. Color of the air dried and ground samples in the Hunter color difference meter. . Composition of air dried and ground samples. Feed intake, protein intake, weight gain, feed conversion ratio, PER and NPR of the samples fed to rats for 28 days. . Apparent N digestibility (%) of experimental diets, in vivo and in vitro. . . . . . Average food intake, weight gain, weight gain per 100 9 food of rats fed for 28 days synthetic and experimental food. . Available lysine values from rat growth assays Available lysine values by the FDNB method and correction factor from an internal standard. Total amino acid composition of samples (ex- pressed as g residue/100 g protein). . . Total free amino acids in pepsin-pancreatin digest of the samples. . . . . vi Page 20 SO 54 56 68 69 72 75 82 86 88 9O 93 Table Page 15 Percent digestion of the samples by the pepsin- pancreatin-pancreatin enzyme system. . . . . . . 98 16 Total, destroyed, inactivated and available lysine values of the samples . . . . . . . . . . 99 17 Available lysine of samples, percent . . . . . . 101 vii LIST OF FIGURES Figure Page 1a Procedure for determining available lysine by the direct FDNB method (Carpenter's method). . . . . . 34 lb Reaction of bound and free lysine with the l-fluoro-2,4-dinitrobenzene. . . . . . . . . . . . 35 2 Early stages of Maillard reaction. . . . . . . . . 40 3 Average weight gain of weanling rats fed diets containing the bread mix, whole bread, crumb, crust, LTB, DTB, NFDM-flour and NFDM-bread . . . . 71 4 Examples of pH vs time curves obtained by incu- bation of representative samples with trypsin, chymotrypsin and peptidase enzymes mixture . . . . 79 5 Plot of percent available lysine in the diet versus (a) weight gain and (b) versus weight gain per 100 g diet 83 6 Plot of available lysine consumed in g, vs weight gain in 9, during the experimental period. . . . . 85 7 Plot of available lysine by rat growth assay vs PER values . . . . . . . . . . . . . . . . . . . .104 viii INTRODUCTION An area of active current research is the availability of different nutrients in food products, and how the avail- ability is affected by the different processes that the products undergo. One of the nutrients on which consider- able interest has been focused is the essential amino acid lysine, because it is the limiting amino acid in many pro- teins and because its physiological availability is strongly affected by processing and storage practices. Bread and other wheat products are very popular foods in the world. However, these as well as other cereal products are nutritionally inferior as protein sources, because they are deficient in lysine. The different pro- cesses that these products undergo damage their nutritional value by making part of their lysine physiologically unavailable. This happens as a result of the Maillard reaction between lysine and the carbohydrates of cereal products. In this work the loss of nutritional quality of wheat products due to heat treatment (baking and toasting) was evaluated in relation to the loss of available lysine. The protein efficiency ratio (PER) and net protein ratio (NPR) methods were used for the estimation of the biological value of the product proteins. The effect of the heat treatment on the digestibility of the proteins was examined by 1 vivo and in vitro methods. The available lysine of the products was measured with biological and chemical methods and an enzymatic method. The relationships between the results obtained by biological and chemical methods were examined and practical aspects of these relationships discussed, since it is significant at the present time to have relatively simple methods for determining the nutritional value of food products with respect to labelling policy and development of new food products. LITERATURE REVIEW Amino Acid and Protein Requirements People must consume sufficient protein to provide nitro- gen and certain amino acids for the synthesis of new tissue during growth, gestation and lactation, and for the main- tenance of existing tissue. Rose (1949) first found that certain amino acids can be synthesized within the organism while others cannot. The latter must be provided in the diet and are termed "indispensable" or "essential". Those amino acids that can be synthesized in the body are called "dispensable" or "non- essential". Classification of amino acids in this way applies only to dietary needs since all are essential for the synthesis of protein. Table 1 shows the classification of essential and non- essential amino acids. Histidine is considered essential for the infant (Kopple g; 31., 1975). Arginine is dispensable for man, even for the growing infant (Holt, 1967), but is essential for the young rat (Rose 33 11., 1948). Glycine is essential for the chick, but not for mammals (Meister, 1965). Some of the dispensable amino acids can be synthesized only from specific essential amino acids. Cystine can be Table 1. Dietary Essential and Non—Essential Amino Acids Methionine Phenylalanine Threonine Tryptophan Valine for Man Essential Non-Essential Histidine Alanine Isoleucine Aspartic Acid Leucine Arginine Lysine Cystine Glutamic Acid Glycine Hydroxyproline Proline Serine Tyrosine formed only from methionine, and when cystine is present in the diet in adequate amounts, less methionine is re- quired (Rose and Nixon, 1955). The other dispensable amino acids can be synthesized in the body from organic acids that are intermediates in carbohydrate metabolism, e.g., a-ketoglutarate and pyruvate (Steele, 1952) and from the nitrogen of other amino acids or from such compounds as ammonium citrate (Rogers 3; 31., 1970). Nitrogen for adult humans can be supplied in large part by glycine and diammonium citrate (Rose and Nixon, 1955). However, nitrogen balance in adult humans can be adversely affected if only one or two sources of nonspecific nitrogen make up a large part of the nitrogen in an amino acid diet, possibly because the rate of synthesis of the non-essential amino acids is not adequate (Swendseid g£_gl,, 1960). The protein requirements for human diets have been a controversial subject for many years. After many different suggestions on the subject the years before, the Food and Nutrition Board (1945) in its Recommended Daily Allowance adopted, with an allowance for safety, the 1.0 g protein/ kg body wt/day. Hegsted g$_gl, (1946) suggested a value of 0.46 g protein/kg body weight/day on a diet of mixed vegetables and 0.38 g protein/kg body wt/day when the diet supplied one-third of the protein as meat. In 1957 the FAO Committee on Protein Requirements took account of N balance studies in man to obtain an average minimum requirement for adults of 0.35 g of protein per kg of body weight, when the protein consisted of a reference protein of high nutritive value, such as whole egg. Since the nutritive value of most proteins of the diet is less than that of the reference protein, a further correlation based on the chemical scores of the dietary protein was recommended. Thus, in this country with a high standard of living, the allowance of 0.66 g of die- tary protein/kgbody weight was suggested, whereas in a country where the dietary proteins have a lower chemical score, the allowance should be as high as 0.84 g/kg body weight for adults. The committee on Amino Acids of the Food and Nutrition Board (1959) estimated that an intake of 0.31 to 0.34 g/kg body weight/day of a high nutritional value protein, as milk, eggs or meat, would meet the minimal adult requirements. In 1973, it was estimated by the joint FAO/HHO AdHoc Expert Committee that a 65 kg adult man would need 0.57 g of protein/kg body weight/day to meet the nitrogen require- ments. According to the latest edition of the Recommended Dietary Allowances (N.A.S.-N.R.C. 1974), the allowance for the mixed proteins of the United States diet is 0.8 g/kg body weight/day or 56 g protein per day for a 70 kg man. Evaluation of Nutritional Quality of Protein Protein nutritional value of a food product is a reflec- tion of the ability of that food product to meet the protein nutritional needs of the individuals consuming it. The nutritional value of a food protein depends not only on its content in amino acid but also on their physiological availa- bility. This availability varies with the protein source, processing treatments (especially heating), and interaction with other components. All methods used to measure the nutritional value of dietary proteins measure the capacity of proteins to supply essential nutrients for growth and maintenance of the organism. The biological value of protein is the amount of absorbed nitrogen from the gastrointestinal track which is retained by the body to build, repair and maintain tissue protein. The biological value of a protein is determined by its content in essential amino acids. Even more impor- tantly the biological value is determined by the essential amino acid balance, that is, the relative proportions of the amino acids present in a protein. Thus the more closely the essential amino acid pattern of the dietary protein matches the pattern used for protein synthesis, the higher the biological value of the protein will be. Any factor inherent in the protein food itself or caused by processing which changes this pattern has an effect on the biological value of the protein. Such factors are: A. Protein Concentratign_ There is an optimum level of dietary protein for maxi- mum efficiency of utilization (Allison, 1959; Campbell, 1963; National Academy of Sciences, National Research Coun- cil, 1963). The experimental condition identified as "amino acid imbalance“ occurs principally at low levels of dietary protein (Harper, 1959) and the balance or proportions of the amino acids in the protein become more critical. Also, decreasing levels of protein concentration reveal differences in amino acids which do not appear at optimum or higher protein levels (Harper, 1959). B. Amino Acid Content All metabolically active tissues have about the same average amino acid composition, and thus are more or less equal as sources of good quality protein whether they are animal, plant or microorganisms. In contrast, the storage portions of cereals, oilseeds and legumes are very variable in amino acid composition (Bressani, 1962). Since the plant is able to synthesize amino acids from inorganic nitrogen, the storage proteins of seeds need only be suitable sources of nitrogen for the developing seedling. Therefore, the plant is indifferent to the amino acid composition of its storage protein. Some are poor, such as the corn pro~ tein, and some are relatively good such as the soy protein. Another difference between animal and plant proteins is that the ratio of essential to total amino acids is smaller in plant than in animal proteins. This difference is also important in determining the efficiency of protein utilization and may be responsible for the lower utiliza- tion of some plant proteins, even after the essential amino acid pattern is corrected by supplementation (Harper and DeMuelenaere, 1963). C. Amino Acid Availability Amino acid availability of a protein depends on (1) the digestibility coefficient of the protein and (2) the rate of release of amino acids during digestion. A simple chemical examination of the amino acid content of a protein does not always indicate the effective amino acid balance of the nutritive value of the protein. Some of the essential amino acids present in the proteins are not fully released after digestion, and the rate of release of different amino acids varies from protein to protein during digestion (Mauron 33 al., 1955). Gupta g; 31. (1958) found that lysine availability to the weanling rat was only 50 percent for corn, 70 percent for wheat, and 85 percent for a roller dried milk sample. The ten essential amino acids were all highly available from peanut flour and wheat (92-100 percent) whereas in cottonseed meal the avail- ability ranged from 64.5 to 93.4 percent. The utilization of lysine from nineteen food proteins ranged from 49 to 90 percent and of methionine from 48 to 83 percent (Guthneck et al., 1953; Schweigert g£__l., 1954). It is well known that heat processing causes reduction of the availability of certain amino acids. But it is not 10 known why certain amino acids of native, raw proteins are not 100% available. Biological Value of Wheat and Its Products Osborne and Mendel (1914) showed that the wheat pro- teins are generally of poor biological value as compared with animal proteins. The studies of Muraver and Harper (1959), Howard 31 31. (1958) and Bender (1958) suggest that the protein content of wheat flour is inferior to animal protein for growth and maintenance of rats. These studies have shown that lysine is the primary deficient amino acid as far as growth of rats is concerned (Mitchell and Block, 1946; Block and Weiss, 1956). Jonick and Kawalizyk (1965) showed in several samples of wheat bread that lysine and methionine levels control the biological value of wheat. When lysine was added to the wheat diet, there was a large increase in the growth promoting value of wheat (Hutchinson g; al., 1958; Rosen- berg and Rohdenburg, 1952; Flodin, 1956). While lysine is the most limiting amino acid in wheat and wheat products, additional amino acids have been shown to be limited. The addition of threonine along with lysine further increased the quality of wheat protein when the level of the protein in the rat ration was 9.5% or less. Above this level, the addition of threonine had no effect (Rosenberg g; 11., 1954). Bender (1957, 1958) showed that when a diet of bread fed to the rats for 10 days at 1.5% 11 nitrogen level, the first and second limiting amino acids were lysine and threonine, but methionine proved to be the third limiting amino acid. There is general agreement between investigators that the first limiting amino acid in wheat and wheat products is lysine. However, depending on the type of product, whole grain, wheat flour or bread, the order of the second and the third limiting amino acid will vary between threonine and methionine. During baking, certain amino acids are lost or des- troyed. Horn 2; al. (1958) found no destruction of amino acids during fermentation but losses of lysine, methionine and cystine during baking were significant. Most of them occured in the crust. Jansen 2£.il- (1964) measured the nutritional losses of added lysine during baking by rat assay. They showed that 30% of lysine became unavailable nutritionally when the time of baking was increased from zero to 50 minutes. At 20 min. baking time, no loss was observed at 425°F. Larsen (1966) suggested that protein bound and free lysine can be made nutritionally unavailable by the browning reaction in which the e-group of lysine would react with carbohydrates or other compounds containing carbonyl groups. PROTEIN QUALITY DETERMINATION Biological Methods The Biological Value (BV) method of Thomas Mitchell utilizes the principle of nitrogen balance in estimating the quality of proteins. Biological Value is calculated as the ratio of the retained nitrogen over the absorbed nitrogen and measures the efficiency of utilization of nitrogen absorbed by the test organism (McLaughlan, 1972). This method is laborious, because of the extreme care which is necessary in collecting feces and urine free of feed contamination and difficulties in accurate measure of endogenous urinary and fecal nitro- gen. N retained Nin-(Nf-Nmf)-(Nur.-Neur) x 100 BV= = N absorbed Nin-(Nf-Nmf) where: Nin = nitrogen intake, Nf = fecal nitrogen, Nmf = metabolic fecal nitrogen, Nur. = urine nitrogen, Neur. = endogenous urinary nitrogen The metabolic fecal N and endogenous urinary N are measured in practice for each animal during a period of non-protein feeding prior or subsequent to the test period. 12 13 The assumption is made that the amount of metabolic N in urine and feces is the same in animals consuming diets containing protein or no protein. It was shown by Mitchell (1924) that this is related linearly to the body weight and also that it is directly proportional to the roughage content of the diet. BV does not take into account the digestibility of the foods tested, since it concerns only the utilization of absorbed Nitrogen. The Net Protein Utilization (NPU) developed by Bender and Miller (1953) eliminates some of the laborious aspects of BV and takes into account the digestibility of the pro- teins. NPU = N retained = N£_;_flg x 100 N 1ntake Nc where: Nt = body nitrogen of test group, Ne = body nitrogen of the group on protein free diet, Nc = nitrogen consumed by the test group. The Nitrogen retained in this method is measured as the difference in carcas nitrogen between the rats fed the test diet and rats fed the non-protein diet . This requires a very careful pairing of the animals according to the body weight before the start of the experiment. The nitrogen lost by the group fed the non-protein diet represents the nitrogen required for maintenance. The Net Protein Ratio (NPR) method described by Bender and Doell (1957) is based on the easier measurement of body 14 weight change instead of the laborious measurement of body nitrogen. NPR = wtgt + wtle Wtct where" wtgt = weight gain of the group on test protein, wt]e = weight loss of non-protein group, wtct = wt of protein consumed. This method takes into account the maintenance require- ments of the rats. It also assumes that the body composition is constant, assumption which may not be valid when widely different test and protein deficient diets are fed to the animals. Hegsted and co-workers (1965, 1968) proposed a slope- ratio assay for the assessment of protein quality. In this procedure the protein is fed at different levels and the slopes of the test proteins are compared with that of a standard protein, such as lactalbumin. The use of a standard protein under the same conditions as the test proteins aims at eliminating the causes of variability which cannot be clearly identified and controlled. The slope of the test protein is expressed as a percent of the lepe of the standard protein. A valid slope-ratio assay requires that the response curves for sample and standard be linear and that they meet at the zero dosage level. This does not always happen. Only values falling on the linear portion of the curve are used in computation of the slope assay value. However, the slope may not always be a valid index of protein quality. For example, lysine 15 deficient proteins yield a lower slope; while threonine deficient proteins tend to yield higher slope values (Hack- ler, 1977). The Protein Efficiency Ratio method (PER) (Osborn and Mendel, 1919) is defined as the ratio of weight gained over the protein consumed by weaning rats for a 28-day test period. It is the most simple bioassay and it is the offi- cial method of AOAC for measuring protein quality. However, there are fundamental problems associated with PER. No allowance is made for amino acid maintenance requirements of the test animals. Thus, a protein that meets the main- tenance requirements but does not promote growth will be scored with zero PER value. It also assumes constant compo- sition of body weight under different diets as the NPR does. As with all bioassay methods, it is difficult to measure complementary effects of two or more proteins, by using PER as an indicator of their protein quality. The single level of protein, 10%, used in PER tests, is considered too high for good quality proteins (Hegsted, 1977). Since PER has become the major protein quality test and basis for nutritional labeling, extensive research on the factors that affect the variability of results obtained has been done. The effect of some factors reviewed by Stenke (1977) and some of them are: 1. animal quality; 2. strains of rats; 3. acclimation period; 4. diet instability; 5. nutritional balancing of diets; 6. diet hydration; 7. high moisture samples. The results of other studies (Morrison & 16 Campbell, 1960; Hurt g; 31., 1975; Hegarty, 1975) indicated that l) younger rats (21-23 day old vs 29 day old) yield higher PER values; 2) female rats grow slower than males, thus producing different PER values; 3) PER values are con- siderably higher at 2 weeks vs 4 weeks. However, adjusted PER's calculated by Hegarty's data indicated that 2 weeks assay only would give accurately enough results. These data suggested that under standardized conditions, the PER assay could be shortened to two weeks only (Hackler, 1977). PER also varies with the amount of protein in the diet; for example, high quality proteins, such as the egg protein, show higher PER value at about 8% protein in the diet, whereas poor quality proteins, such as wheat gluten, show maximum PER at 15% protein level (McLaughlan, 1972). Mineral content and dietary fat (saturated vs unsatu- rated) do not affect PER. But the level of fat and fiber was found to influence the results significantly. Fiber and fat were found to affect the PER of casein to the same extent as they affected the test proteins. Therefore, if both the test diets and casein are balanced with respect to these ingredients, the results will not be affected (Hurt _g _l., 1975). The carbohydrates in the test samples must also be con- sidered in the diet composition. Rats prefer a diet with some sweet taste and may consume higher quantities of the diet (Steinke, 1977). This problem may occur when foods containing sucrose are compared with casein control diet 17 using starch as the carbohydrate source. Because PER values are affected by so many factors, strict adherence to certain conditions is necessary for obtaining comparable and meaningful results. PER shows good correlation with the other biological methods. Hackler (1977) computed the correlation coeffi- cients for the BV, Digestibility (D), NPU and PER bioassays from data on plant and animal proteins published by FAO. With this particular set of data the correlation coefficient (r) for NPU and PER was the highest (.973). For BV:NPU r = 0.899, BV:PER r = 0.885, BV:D r = 0.647, DzNPU r = 0.576, DzPER r = 0.479. McLaughlan and Keith (1975) introduced a modified PER procedure that takes into account maintenance requirements. In this method, Nitrogen Utilization (NU) is defined as weight gain plus 0.1 times the initial weight plus final weight divided by the protein consumption. When NU of a test protein is compared to a standard protein, usually lactalbumin, it is termed Relative Nitrogen Utilization (RNU). NU for test protein NU for lactalbumin RNU = x 100 The factor 0.1 x (initial + final weight) is similar to the weight loss of the non-protein group in the NPR procedure. McLaughlan and Keith (1975) compared the PER, RNU slope RPV and NPR values for twelve different proteins of varying protein quality. They found similar values for 18 each protein, except that the PER values tended to underes- timate the quality of lower quality proteins. In a similar comparison of twelve different food mixtures common to Latin America, all the dietary proteins used were similarly ranked by RPV, RNV, PER and NPR (Chavez and Pellet, 1976). In all bioassay methods, the rat is the most common animal used in assessing the nutritional quality of food proteins for human consumption. Rats and humans, however, do not have exactly the same nutritional requirements. Also the rate growth, maintenance needs and the consumption of a variety of foods containing various proteins by humans at each meal are very different from the procedures used in rat bioassays. The essential amino acid requirements of man for maintenance are much lower than those for growth in the rat. Also, the protein requirements for growth or mainte- nance are different in the rat than in humans at all stages of life (Bodwell, 1977). Since there is no bioassay method with which absolute accuracy can be achieved in determining the nutritional value of proteins for humans, and each one of the existing methods has its limitations, the selection of the method to be used is based on criteria such as sim- plicity of the test, economics, labor and reproducibility within and between laboratories. Chemical Methods The chemical methods for nutritional evaluation of Pr0teins are based on the chemical analysis of the amino 19 acid content of the proteins. Mitchell and Block (1946) used the chemical composition of food proteins and related it to their nutritive value. The basis for their method was the fact that the biological value of a protein is dictated by the limiting amino acid in that protein (the essential amino acid with the greatest percentage deficit compared to the amino acid needs of the rat or compared to its content in whole egg protein). The "chemical score” value method is defined as the lowest ratio of an essential amino acid of the test protein to the same amino acid in whole egg protein. The whole egg protein was chosen as reference protein because it is highly digestible and almost perfectly utilizable in rodent metabolism, being better than milk protein in this respect (Mitchell, 1946). This was shown later to be true also for the dog (Alison g; 1., 1949) and the adult man (Hawley g;_al., 1948). The essential amino acid composition of whole egg protein is shown in Table 2. Later studies (Bender, 1961) showed that whole egg proteins contain excessive amounts (relative to rat requirement) of tryptophan; also sulfur containing amino acids (methionine and cystine) are in slight excess, but lysine is present at just about the required level. There- fore, the chemical score of a test protein will be good for lysine deficient proteins, while it will underestimate the value of proteins deficient in sulfur-containing amino acids (McLaughlan t 1., 1959). 20 Table 2. Essential Amino Acid Composition of the Whole Eg Protein and the Reference Pattern (FAO/WHO, 197 T Amino Acids Whole Egg FAO/WHO (g/l6 gr N) Reference Pattern (9/16 9" N) Lysine .8 5.4 Methionine + Cystine 5.3 3.5 Phenylalanine & Tyrosine 9.3 6.1 Leucine 8.8 7.0 Isoleucine 5.9 4.0 Valine 7.1 4.9 Threonine 4.9 4.0 Tryptophan 1.4 1.0 Histidine 2.6 - TOTAL 53.1 36.0 21 Because of the excess of essential amino acids in the whole egg protein, the FAO/WHO in 1973 proposed a provi- sional amino acid pattern (Table 2) based on the amino acid requirements of pre-school infants. The index calculated according to this pattern of amino acids is referred to as Amino Acid Score. Oser (1951) proposed the Essential Amino Acid Index (EAAI) as a method for protein quality evaluation. This method of rating protein quality is based not only on one essential amino acid, as the chemical score, but on the contribution of all the essential amino acids. Oser felt that each essential amino acid is specific in its own right and all are equally essential. Therefore, in his method he included all the essential amino acids. The EAAI is calcu- lated by computing the average of the logariths of the ten "egg ratios" (the ratio of the essential amino acids in a protein relative to their respective amounts in whole egg protein) and then taking the antilogarithm of this average value. The EAAI method modified by Mitchell (1954) by adjusting to 100 the egg ratios that happen to have values higher than that. The antilogarithm of the average of the egg ratio logarithms that included adjusted values is called Modified Essential Amino Acid Index (MEAAI). The major problem with the chemical methods for protein quality evaluation is the assumption that all the amino 22 acids are biologically available. This assumption is not always valid and especially with proteins that have been heat treated or processed. 0n the other hand, the advan- tages of these indexes are a) the small size of the sample required; b) the short time of analysis required compared to bioassays and c) that they give information concerning the identity of the limiting amino acid. Microbiological Methods These methods utilize the proteolytic activity of certain microorganisms for the estimation of the nutritional value of proteins. Different microorganisms have been used through the years. The production of lactic acid was first used as a measure of the growth of bacteria in media in which one amino acid was limiting. The growth response of Streptococcus zymogenes (Ford E.J., 1960) and the protozoa Tetrahymena pyriformis (Rosen 31 al., 1958; Pilcher t al., 1954) were used for the deter- mination of the biological value of proteins. With a vigorous proteolytic strain of Streptococcus zymogenes, Ford (1960) obtained protein quality values for several meat meals that were closely correlated with the available lysine content as it was found by others with a combination of an enzymatic-microbialogical technique (Bunyan, 1960). Using §treptococcus zymogenes he also found that heating diminished the nutritive value of skim milk powder. It seems though 23 that foods rich in carbohydrates are not suitable techni- cally for microbiological assays for nutritional value estimation. Studies showed poor correlation between micro- biologically obtained values for the biological value of cereal products and results obtained with animal experiments (Menden & Cremer, 1970). Animal protein foods assayed for methionine and lysine showed fair agreement with those for animal experiments (Rao 35 al., 1963; Scott _5 _l., 1966). The microbiological tests are of shorter duration and lower cost than the animal experiments. Enzymatic Methods Enzymatic methods are an improvement over chemical methods because they determine only amino acids that can be freed from the protein molecules and thus considered as available under physiological conditions. They can be used to measure amino acid availability or to estimate the biological value of a protein. Melnick e; al., 1946, first reported that different proteins, or proteins that had been processed differently, were hydrolyzed by pancreatin at different rates. In his procedure food proteins were digested with pancreatin only. Sheffner _£._l., 1956, used a peptic (100:2.5 Substrate: Enzyme, S:E) digest which was analyzed microbiologically for available essential amino acids to formulate his Pepsin Digest Ratio (PDR). Akeson and Stahman (1964) used Pepsin (100:1.5 S:E) at 37°C for 3 hours, followed by Pancreatin (25:1 S:E) at 24 37°C for 24 hours, and analyzed the digest with an amino acid analyzer. Enzyme blanks were prepared in which the test protein was omitted. They calculated the Pepsin Pan- creatin Digest Index (PPDI) for ten food proteins. The indices were correlated highly with biological values (r = 0.99) and the chemical scores (r = 0.940). Menden and Cramer (1966) used relatively high concen- trations of pancreatin (2:1 S:E) and avoided using pepsin beforehand to limit possible hydrolysis of the pepsin by the pancreatin. They digested the sample for 15 hours, a time they felt was short enough to avoid autolysis of the pancreatin enzymes. Ford and Salter (1966) reported that prolonged digestion with pepsin (10:1 S:E), pancreatin (10:1 S:E) and erepsin (3:1 S:E) released between 80-85% of the amino nitrogen as free amino acids when assayed microbiologically. The above in 11333 digestions were conducted in a flask under static environmental conditions, in which undigested protein material was removed only at the end of the diges- tion period. Products, which 1 vivo would be rapidly absorbed, accumulate in the digestion mixture and may inhibit the digestive reactions. Indeed, end product in- hibition studies indicated that amino acids do competitively inhibit the action of pancreatin on milk proteins (Dimler, 1975). Mauron (1955) also used pepsin and pancreatin but he digested the samples in a dialysis sac simulating, in that 25 way, closely the digestion under physiological conditions and avoiding end product inhibition. They calculated the Pepsin Pancreatin Dialysis Digest (PPDD) index, in a way similar to the PPD index, for several proteins and found quite close agreement to the rats biological value. A discrepancy occurred when a heat-damaged milk protein sample was assayed. Ford and Salter (1966) also were concerned about end product inhibition. They digested samples on a column of sephadex G-lO to provide for chromatographic removal of small products of digestion. After digestion and depro- teinization the samples were fractionated on a Sephadex G-25 column. Amino acid analysis was done on the fraction with molecular weight of less than 250. An enzyme score, calculated like the chemical score, comparing the essential amino acids released from the test protein and those released from egg protein, was determined on the heat processed casein (Stahman & Woldegiorgis,1975). Chemical score and enzyme score were compared well with the PPDI. However, the chemical score tended to underestimate the quality of the more severely heated proteins. The major problem with the use of 13 113:3 digestion methods for protein quality evaluation is the incomplete digestion of the test protein. Therefore, a reference protein is always assayed in conjunction with the test pro- tein. The enzymatic methods have the considerable advan- tages of 1) lower cost and shorter analysis time; 2) less 26 variation in the results than found with rat bioassays; 3) generation of information on amino acid availability and relative nutritional adequacy (Stahman and Woldegiorgis, 1975). Recently Hsu _£ _1. (1977) developed a model for rapid prediction of PER of a protein by using amino acid profiles of the sample protein and a reference casein, along with their in vitro digestibilities. The in vitro digestibility is measured by the drop in pH that the enzyme system tryp- sin-chymotrypsin-peptidase cause in 10 minutes, and from a regression equation that correlates the drop in pH in 10 minutes by the enzyme system, to the apparent digestibility as it is measured by rat assays. By a combination of the digestibility and the essential amino acids profile ex- pressed as percentage of the FAD/WHO - 1973 provisional pattern a computed PER (C-PER) is derived. The average differences between the C-PER and the rat based PER for 45 foods that they used in their experiments, was 0.12. This method gives a fast (72 hour) estimate of protein quality and also gives data as to why that quality is high or low, via either high or low protein digestibility or the deficiency or abundance of one or several essential amino acids (Satterlee 53 al., 1977). Digestibility of Food Proteins A measure of the digestibility of a food protein can be obtained by subtracting the amount of Nitrogen excreted in the feces (F) from the amount ingested (I) and expressing 27 this value as a percentage of the intake. This is termed as apparent digestibility (A D). I-F A D = x 100 As fecal Nitrogen includes nitrogen from bacterial and sloughed intestinal mucosal cells (fecal metabolic nitrogen), as well as that of the undigested food proteins, apparent digestibility is not an accurate measure of the digestibility of the protein. If correction is made for the amount of fecal nitrogen of no-dietary origin, the true digestibility is obtained. (F - Fm) True digestibility = I - ———E————-x 100 where Fm = fecal metabolic nitrogen. The fecal metabolic nitrogen can be estimated when the subject is consuming either a protein-free diet or just enough of a highly diges- tive protein to prevent excessive loss of body protein (Harper, 1974). In vivo conditions of the digestive tract include a pH range of 2 to 6 in the stomach and 4.5 to 6.5 in the small intestine (Nasset, 1957). Digestive enzymes exhibit definite pH values for opti- mum activity. For pepsin (3.4.4.1), the proteinase in the gastric juice, the optimum pH is 1.8 (Bovey §§.1L, 1960). The pancreatic enzymes released by the pancreas to the 28 small intestine consist primarily of trypsin (3.4.4.4) and chymotrypsin (3.4.4.5), with optimum activities of pH 7 to 8. Each of the digestive enzymes cleave proteins dif- ferently (Hill, 1965). Pepsin, the least specific, cleaves peptide bonds in which phenylalanine, tyrosine, glutamic acid, cystine and cysteine contribute either the amino or carboxyl group of the bond. Trypsin has a specificity limited to cleaving peptide bonds with lysine or arginine in the carboxyl position. Chymotrypsin cleaves the peptide bonds with tyrosine, phenylalanine and tryptophan in the carboxyl position. The L-isomeric form of the amino acid residue is preferred for each of the enzymes (Rupp t al., 1966). Methods for Measuring Available Lysine Animal Growth Assays These methods are based on the weight gain response of the animals as a measure of available lysine. A stan- dard curve that relates the growth response to the amount of available lysine in the diet is necessary for these methods. Diets in which lysine is the only limiting factor may be prepared either with a natural feeding stuff defi- cient in lysine, such as wheat gluten (Gupta $3.11., 1957), sesame meal (S.C. Weigert 33 al., 1953), zein and feather meal (Carpenter, 1957; Guo t 1., 1971), or with a mixture of synthetic amino acids (Tsien t 1., 1957; Calhoun g; 31., 29 1960; de Muelenaere 25 al., 1967; Netke e£_al., 1970). The tested food protein added to the basal diet pro- vides not only lysine but also other amino acids, thus changing the balance of amino acids. Schweigert 3; al., (1953) showed that in lysine deficient diets nitrogen addi- tional to the requirements of rat did not affect their response to the limiting levels of lysine. The rate of weight gain of the rats was almost identical for two groups, one of which was fed with basal diet containing 50% more of all amino acids. Diets of both groups had the same degree of lysine deficiency. Excess of amino acids in the diets did not impair the utilization of lysine, isoleucine or leucine, which were the most limited amino acids, by chicks used in the assay (Netke g£_al., 1970), even though excess of amino acids depressed feed intake and thus weight gain (Netke _5 _l., 1969). Concerning the effect of type of carbohydrates on the results obtained, Gupta g£_al. (1957) showed that the results were influenced by the dietary carbohydrate if weight gains were related to amino acid concentration in the diet, but they were independent of the nature of dietary carbohydrate when the weight gains were plotted against actual amino acid intake. Schweigert _£ 11. (1952) used adult rats that had been deprived of protein during a 12 to 15 day period on a nitro- gen free diet and for a test period of 11 to 12 days. After the test period the animals were depleted again of protein 30 and used for a second test. The same researchers in sub- sequent work (Schweigert g; 1., 1954) with lysine and other amino acids, used young rats or chicks. The live weight gain is usually taken as measure of response. Calhoun g; 31. (1960) measured "empty weight" (body weight minus intes- tinal contents) or carcas nitrogen as a refinement. This method improved the precision slightly, but it has not been adapted generally. Dose response relations used for calculation of results are: a) % lysine in diets vs weight gain; b) % lysine in diet vs weight gain per 100 9 food eaten or; c) g lysine eaten vs 9 of weight gain. The two last ways have been preferred because they give values less affected by amino acid imbalance (Netke g; 31., 1969) and by other compounds influencing food consumption (Gupta t 1., 1958; Calhoun g: _l., 1960). Animal growth assays are expensive, slow and imprecise because of the variation in growth response between indi- vidual animals. Despite the above limitations, they are the only direct way to test for the biological availability of nutrients and the method against which the validity of values obtained by other procedures are tested. Kuiken E: El. (1948) determined the availability of all ten essential amino acids in several foodstuffs by measuring the amino acids in the food and in the feces. Gupta £3 1. (1957) used the same method along with the animal growth method for several proteins of animal and 31 plant origin. The results he obtained from the lysine con- tent of the feces method agreed reasonably well with the values obtained using the animal growth data. Lysine con- tent of the protein and the feces was measured microbio- logically with Leuconostoc mesenteroides as test organism. Proteolytic microorganisms such as Streptococcus zymo- genes (Ford, 1962), Streptococcus faecalis (Ford, 1965), the protozoa Tetrahymena pyriformis (Stott g3 31., 1966; Boyne g£_gl,, 1967) or enzymatic digestion in vitro (Evans _£__l_-, 19.48; Mauron _;c__l., 1955; Ford, 1965) were also used for measuring available lysine, the same way as these methods are used for overall protein quality determination. Chemical Methods These methods are based on the reaction of the free or non-bound e-NH2 groups of lysine with different chemical reagents. The most often used chemical is the l-fluoro- 2,4 dinitrobenzene (FDNB). Carpenter and Ellinger (1955) first used FDNB to measure the available lysine in feed- stuffs (Carpenter's method). Subsequently (Carpenter _£__l, (1957); Carpenter, 1960), the method was improved by avoiding interfering substances in the final measurement. More recently, Booth (1971) improved the method by using a more accurate recovery factor to correct for the loss of derivatized lysine during the acid hydrolysis of the sample. The procedure for measuring available lysine by the direct FDNB method and the reaction of bound and free lysine 32 with the l-fluoro-2,4-dinitrobenzene reagent, are shown in Figure l. a-DNP-amino acids (including a, e-di-DNP lysine) that arise from the presence of free amino acids and N-terminal amino acids in the test material and also much of the dini- trophenol that arises from hydrolysis of FDNB, are removed by simple extraction of the hydrolyzate with diethyl ether. e-DNP-lysine is separated from 0-DNP-tyrosine im-DNP-histi- dine, a-DNP arginine, 6-DNP-ornithine and e-hydroxylysine that may interfere, with the use of methoxy carbonyl chloride (MCC), also called chloroformate. The ether-washed sample is made alkaline and shaked with MCC. e-DNP-lysine reacts at its o-NH2 group to form the methoxy carbonyl derivative. The solution is then reacidified and the lysine compound and im-DNP-histidine are extracted with diethyl ether. Thus, interferences are overcome by measuring the absorbance of the ether-washed hydrolyzate before and after treatment with MCC and after extraction with ether (Carpenter, 1960). The residual color after the second extraction serves as a blank. The humin color which is formed during the acid hydroly- sis of products rich in carbohydrates is another source of interference. It is appreciably reduced by the treatment with MCC (Matheson, 1968b; Milner and Carpenter, 1969). Booth (1971), using a modified Carpenter technique employing a large sample:acid ratio, suggested that in practice the amount of humin artifact under these conditions was small 33 Figure 1. Procedure for determining available lysine by the direct FDNB method (Carpenter's method) (la) and reaction of bound and free lysine with the 1-fluoro-2,4 dinitrobenzene (1b). 34 Protein amino acids + +e-DNP-lys-protein+hydrolysis + + + FDNB e-DNP-lys Extraction and measurement of g-DNP—lys (absorption at 435 nm) l.a. Procedure for determining available lysine in the d1rect FDNB method (Carpenter's method) m HN-CH-COW 1 CH2 I CH2 CH ' 2 CH2 I NH2 ‘K-e-amino group lysine bound to protein molecule HzN-CH-COOH 1 CH2 1 CH2 + l 1'2 CH ' 2 NH2 free‘lysine 35 W HN-CH-CO M“ F CH2 1 N0 CH 2 + 1 2 + HF :“2 N02 CH2 NH N02 N02 1-flu0r0-2,4 dini- e-dinitrophenyl tro benzene lysine __ N02 0 N ' HN-CH-COOH 2 Q / . F CH2 1 2 A 02 CH2 '1’ 2H1: 1 CH2 1 H NO2 f 2 NH Cl l-fluoro-2,4 dini- trobenzene N02 a,e-dinitrophenyl lysine l.b. Reaction of bound and free lysine molecule with the l-fluoro-2,4 dinitrobenzene. 36 enough to be neglected. Under these conditions of hydroly- sis most of the humin formed is insoluble and is removed by filtration. Part of the e-DNP-lysine of the sample is lost during the acid digestion step. Thus a recovery factor has to be calculated. Using e-DNP-lysine, Carpenter (1960) found that the recovery is 90% or more with animal materials, but with vegetable foods rich in carbohydrates, the recovery of added free e-DNP-lysine is only 60-85%. When DNP-lysine is refluxed with carbohydrates in hydrochloric acid, the N02 groups on e-DNP-lysine are reduced to NH2 (Handwerck, Bujard and Mauron, 1966). As the amino compound is less yellow than the nitro compound, the mixture is unsuitable for colorimetry. Booth (1971) showed that protein-bound DNP-lysine is less sensitive to destruction than the free molecule which is only gradually formed during refluxing in acid. He used dinitrophenylated gluten and dinitrophenylated casein to estimate the recoveries and he suggested that the true loss of DNP-lysine from the test material can be roughly estimated as one half the percentage loss of added free DNP-lysine. EQNB-reactive Lysine by Difference This method consists in calculating the lysine with a free e-NHZ group as a difference between the total lysine after acid hydrolysis and lysine after treatment with FDNB 37 and subsequent hydrolysis. The hypothesis behind this ap- proach is that the lysine units in processed foods, in which the e-NH2 is blocked, are unable to react with FDNB, but still yield lysine which appears as part of the "total" lysine value obtained on ordinary analysis after acid hy- drolysis. Also, it is assumed that the FDNB-unreactive lysine will still be recovered as lysine after treatment of the sample with FDNB and then acid hydrolysis. This method overcomes the problem of partial reduction of DNP-lysine, during acid hydrolysis, of the direct FDNB method and the need for estimation of recovery factors. ' Couch (1974) in a collaborative study of 3 samples of 44% protein soybean meal, found the differential FDNB method to be satisfactory for determining available lysine content of foods and feeds and recommended it for adoption as offi- cial first action AOAC method. Alternative Reagents to FDNB Trinitrobenzene-sulfonic acid (TNBS) was used by Kakade and Liener (1969) for the direct determination of lysine having a free c-NH2 group. The method they developed is similar to the direct FDNB method. TNBS is water soluble and does not require the presence of ethanol for the reac- tion with the proteins in alkaline solution. O-methylisourea reacts with e-NH2 groups of lysine in Proteins to form a quinidine derivative (Hurrel and Carpen- ter, 1974) which, on hydrolysis, yields homoarginine and can be analyzed by the method of Spackman, Stein and 38 Moore (1958), on the amino acid analyzer. No homoarginine is formed when lysine units are combined with sugar in a Maillard reaction (Finot & Mauron, 1972). The disadvantage is that O-methylisourea reacts slowly (4 days at room tem- perature) which is perhaps the reason why this procedure has not been used extensively. Acrylonitrile also reacts, though slowly, with the e-NH2 groups of lysine units to form a cyanoethylated deriva- tive that is stable under ordinary conditions for the acid hydrolysis of the proteins (Riehn 8 Scheraga, 1966). The use of methyl acrylate as an alternative to acrylonitrile has been also proposed (Finley and Friedman, 1973). Thomas (1970) has used sodium borohydride to reduce the reaction products between aldehydes and lysine units to forms that are stable to acid hydrolysis and has taken the lysine that is released from the test material by acid hydrolysis after this treatment as a direct measure of available lysine. Protein Damage Reaction with Carbohydrates The reaction of proteins with carbohydrates, the Mail- lard reaction or non-enzymatic browning, has been studied extensively (Hannan and Lea, 1949, 1952; Bender, 1972; An- drian, 1974). It is the reaction between amino groups of amino acids or proteins and aldehyde groups from reducing 39 sugars or carbonyl compounds present in foods. It is actu- ally a series of complex reactions which result in destruc- tion of certain amino acids and in the formation of brown products (Fig. 2). Although free a-NH2 N-terminal groups are at least as reactive as the e-NH2 groups of lysine (Schwartz and Lea, 1952), the latter provides the great majority of the reactive groups in most proteins. The rate of Maillard reaction is maximal at a water activity (aw) of 0.6 - 0.8 and decreases at higher concen- trations of water; it is still appreciable at a aw as low as 0.2 (Chefter, 1977). Low pH inhibits browning since the first step of Maillard reaction depends on the condensation of carbonyl groups with nonionized e-amino groups of lysine. Alkaline pH generally enhances Maillard reactions. For example, the reaction rate between sodium caseinate with 11% glucose at 37°C and 70% relative humidity at pH 3 was only one-tenth of the rate of pH 7, but it_was only slightly accelerated by more alkaline conditions (Lea and Hannan, 1949). Different reducing sugars also react with different rates under the same conditions. The order of activity was xylose > arabinose > glucose > lactose, maltose > fructose (Lewis and Lea, 1950). In general, pentoses are more active than hexoses in promoting Maillard reactions. The non-redu- cing disaccharide, sucrose will induce browning only after it is inverted. When protein foods containing reducing sugars are heated moderately, Maillard reaction occurs in preference 40 9”? CH2 R ' LYSIVE 92 NHZ * 1111)le H-C=0 R-N FEARRNVEVENTS Rm ( ) a momma 1“" .. = a... - a —- .. CHZOH ' 014 (611011) POWERS CH2 1 3 Fonmnou CH20H GLUC(BE SCHIFF BASE 1-DE0xv-2- KETOSYL comm LATE STAGES OF 111111111111 REACTIO‘l Figure 2. Early stages of Maillard reaction. 41 to any other type of protein damage (Chefter, 1977). Pro- tein foods containing small amounts of reducing sugars and submitted to a more severe heating may undergo both late Maillard-type and protein-protein interaction, resulting not only in loss of lysine but also of other amino acids (particularly arginine, aspartic acid, and glutanic acid in the case of casein with glucose) (Carpenter, 1973). Rao and McLaughlan (1967) autoclaved mixture of equal parts of casein and glucose at 120°C for 10 minutes. The values for lysine they obtained as percentages of the cor- responding values for the unheated control were: FDNB- reactive lysine 57, lysine (by rat growth) 54. When Boctor and Harper (1958) autoclaved egg white with 2% glucose at 121°C for 12 hours, the total lysine value fell by 30% and the FDNB value by 40%, but the rat growth value fell by 80%. The latter was in accordance with the fall in the digestibility of nitrogen. When egg protein was autoclaved in the absence of glucose, the damage did not reach the same degree of severity. Valle-Riesta and Barnes (1970) sugges- ted that with limited supply of reducing sugars in the system and severe heat treatment, cross-linkages may be formed, many of which still contain FDNB-reactive lysine. In these cases, the FDNB values are overestimates of the amount of available lysine. From the nutritional point of view, initial and late Maillard reactions do not induce the same type of damage (Chefter, 1977). Although all the reaction products of 42 proteins with carbohydrates are not known, the metabolism of two of them have been studied in model systems. e-N- (l-deoxy-lactulosyl)-lysine (e-DL-L) represents 70-75% of all the Maillard compounds in samples of overheated milk (Finot, 1973). This compound results from the reaction of the e-NH2 group of lysine with lactose and reduces markedly the availability of this amino acid. e-N-(l-deoxy-D- fructosyl)-L-lysine, (e-DF-L), corresponds to the reaction of glucose with lysine. Both of these compounds, if protein bound, are digested poorly (Erbersdobler, 1976). Synthetic e-DF-L is not utilized by the body; it is absorbed through the intestinal wall by diffusion and is rapidly excreted through the urine as an intact molecule. The e-DF-lysine which is not absorbed reaches the lower intestine where it is presumably destroyed by microorganisms (Finot, 1973; Erbersdobler, 1976). Deoxy-lactulosyl-lysine residues from milk protein are found in the feces rather than in the urine. In advanced stages of non-enzymatic browning, the reac- tive unsaturated polycarbonyl compounds which are formed polymerize and bind to the terminal, 5 and other amino groups of polypeptide chains. This results in formation of colored, highmolecular weight, highly cross-linked protein-carbohydrate polymers, with low solubility, diges- tibility and nutritional value (Clark and Tannenbaum, 1974). The initial product e-DF-lysine, upon acid hydrolysis, yields lysine, furosine and pyridosine; the freed lysine 43 accounts for only 44-50% of the lysine of the derivative (Finot, 1973). Severely damaged proteins regenerate even less lysine upon acid hydrolysis (Chefter, 1977). Protein - Protein Interactions Severe heating of pure proteins or protein foods low in carbohydrate content (fish and meat), may not cause con- siderable change in the amino acid profile obtained by acid hydrolysis but may cause decrease in protein digestibility and loss of protein nutritional value (Donoso gi al., 1962; Miller _5 _l., 1965a, c). It was also observed, with rats fed fish proteins, that abnormally high levels of peptides remained undigested and free amino acids and small peptides were found in their gut (Ford gi al., 1966; Ford, 1973). They suggested that inefficient digestion resulted from the resistance of heated proteins to proteases (Ford et al., 1971). The finding of peptides containing lysine, aspartic and glutamic acid, in the urine of the rats, led to thoughts that acylation of e-NH2 group of lysine by glu- tamic or aspartic residues, would make this amino acid biologically unavailable, although hydrolysis with hydro- chloric acid was able to regenerate lysine. Bjarnison and Carpenter (1970) demonstrated the formation of e-N-(y-glutamyl)-lysyl peptide from lysyl and glutamine residues, with release of ammonia. Later, however, it was found that free (not combined) e-N-(v-glutamic)-L-1ysine can be fully utilized as source of lysine by the rat and the 44 chick (Warbel _5 _l., 1972). Thus, simple acylation of e-NHZ groups of lysyl residues cannot explain the effects of severe heating on proteins. Mauron (1975) suggested that these effects may be explained by the formation of intra and intermolecular cross-links of e-N-(y-glutamyl)- lysine or of e-N-(y-aspartyl) lysine, since these cross- links do not appear to break in the gastrointestinal track. The presence of many of these cross-links may prevent easy access of proteolytic enzymes by steric hindrance and slow down or even prevent proteolytic attack.' Such mechanisms may be responsible for the decrease in availability of several amino acids in severely heated proteins. The time delay in the release of some amino acids and peptides that can not be utilized could explain why in some cases the protein digestibility appears higher than the utilization of essential amino acids (Chefter, 1977). Recently, strong interest was shown on the formation of lysinoalanine (LAL) cross-links. These links occur in alkali-treated proteins but they were also found in a variety of foods that had not been exposed to pH above 7. The presence of LAL in the structure of proteins has been reported to reduce the digestibility of proteins and net protein utilization (DeGroot, 1969). It was also suspected for causing renal lesions characterized by cytoplasmic and nuclear enlargement of the tubular epithelium in the rats. These changes, however, seemed to be species specific to the rat. Mice, hamsters, dogs and monkeys failed to exhibit 45 renal cytomegalic effects when fed synthetic LAL (DeGroot, 1969). The finding of LAL in a variety of foods (Sternberg _£ 31., 1975) heat-treated at pH values considerably lower than those obtained during alkali treatment suggest the universal occurence of LAL in cooked foods and that humans have long been exposed to proteins containing LAL. METHODS AND MATERIALS Preparation of the Samples Commercial bakery flour (unbleached) containing wheat flour, malted barley flour and potassium bromate, was ob- tained from International Multifoods, Minneapolis, Minne- sota. The bread was prepared according to the following formula (bread mix): Flour 100 parts Water 65 parts. Salt 1 part Sugar 6 parts Yeast (dry granular) 3 parts The yeast was allowed to hydrate in the water for 5 minutes in the stainless steel bowl of the mixer and the sugar, salt and approximately half of the flour were added. The above ingredients were mixed for one minute, and half of the rest of the flour was added and mixed for another minute. The additional flour was added gradually with con- tinuous mixing and the dough was kneaded for 7 minutes. Then the dough was shaped into loaves of approximately 37.5 x 11.5 cm and 1,700 gr wt., put into loaf pans and left to ferment for about 30 minutes in a proofing cabinet 46 47 at about 38°C. The dough was punched down and left to ferment for another 40 minutes or until it filled up the pan. The loaves were baked in a preheated oven at 190°C for 55 minutes, until crust was golden brown. The loaves were removed from the pan and allowed to cool. Part of the bread was separated into crumb and crust and part of it was sliced into slices of 1.1 cm. thickness. The slices were toasted in a household toaster either for 1.5 minutes or for 2.5 minutes. The highest temperature reached during toasting was 270°C for the 1.5 minutes and 350°C for the 2.5 minutes period. The 1.5 minute toasting period produced toasted bread of light yellow-brown color, the light toasted bread (LTB), the 2.5 minutes toasting period resulted in a dark brown product, the dark toasted bread (DTB). In the BDOVL ;read formula, 4% non-fat dry milk solids were added (NFDM-bread mix) and bread was prepared from this mixture as described above (NFDM-bread). All the above products were dried overnight, in a cabinet dryer with circulating hot air at 37°C to a con- stant moisture content. They were ground in a mill to pass through a 0.50 mm screen and were kept refrigerated until they were used. Moisture, Fat and Ash Determination Moisture, fat and ash of the samples were determined according to AOAC (1975) procedures. 48 Nitrogen Determination - Micro Kjeldahl Samples containing approximately 7.0 mg dried protein were digested for 1 hour in duplicate according to AOAC. Sulfuric acid of 1.84 specific gravity was used for diges- tion and potassium sulfate and mercuric oxide were added as catalysts. After cooling the flasks, the sides were rinsed with deionized water and digestion was continued for an additional hour. The digests transferred into the distillation apparatus by using approximately 10 ml deio- nized water. The digestion mixture was neutralized with 15 m1 of a 50% NaOH solution, containing 5% of sodium thio- sulfate. The released ammonia was steam-distilled into 5 m1 of 5% boric acid solution, containing 4 drops of methyl red-methylene blue indicator containing 2 parts of 0.2% in alcohol Methyl red with 1 part of 0.2% in alcohol methylene blue. The distillation was continued until the volume in the receiving flask reached 25 ml. The ammonium borate complex was titrated with 0.02 N HCl which had been accurately standardized against tris-hydroxy-amino methane as primary standard. Nitrogen was calculated from the following formula: % N (ml HCl - ml blank) (Normality of HCl) (14.007) x 100 mg of sample Protein Efficiency Ratio (PER) and Net Protein Ratio (NPR) PER was performed according to AOAC (1975). Male rats of Sprague-Dawley strain, 21 days old, were used, ten 49 for each assay. After an acclimation period of 4 days, during which the rats were fed a standard rat diet, they were randomly divided into groups. Individual rat weights differed by less than 10 g and the average weight of groups varied from 71.2 g to 76.5 g. Rats were housed in indivi- dual cages; diet and H20 were provided ad libitum. One group was fed the reference casein diet. Another group was fed a non—protein diet and the results of this group used in the calculation of the Net Protein Ratio. The body weight of each rat was recorded on the first day of the assay period and again weekly up to the 28th day. The food intake was also determined weekly. The composition of the basal diet is shown in Table 3. The diets were adjusted according to the proximate analysis of test materials so that all diets, samples and reference had the above composition. The test materials were fed as sole source of protein at the 10% protein level. Corn oil and corn starch were used to make up for the oil content and the carbohydrates of the diet. The vitamin mix was obtained from Teklad Test Diets and the salt mixture was prepared in our lab with composition similar to the specified USP XIX. The average 28 day weight gain and protein (N x 5.75 for bread, N x 6.25 for casein) intake per rat for each group were calculated. The PER (g wt gain/g protein intake) was calculated for each group and the values were adjusted to 2.5 for the reference casein. 50 Table 3. Composition of basal diet for the PER f Ingredients Amount % Protein 10 Corn oil 8 Salt mixture1 5 Vitamin mixture2 1 Non-nutritive fiber 1 Corn starch and sugar to complete 100 1USP XIX Salt mixture composition (%): Sodium chloride (NaCl) 13.93; Potassium iodide (KI) 0.079; Potassium phosphate monobasic (KH2P04) 38.90; Magnesium sulfate (MgSO4) 5.73; Calcium carbonate (CaCO3) 38.14; Ferrous sulfate (FeSO47H20 0.039; Manganese sulfate (MnSO4H20) 0.401; Zinc sulfate (ZnSO 7H 0) 0.548; Cupric sulfate (CuSO4 5H20) 0.0477; Coba1t c loride (CoClz 6H20) 0.0023. 2Vitamin mix AOAC (Teklad Test Diets) contained (Inter- national Units or mg/100 g diet) Vitamin A 2000 (IU); Vitamin D 200 (IU); Vitamin E 10 (IU), Medadione 0.5; Choline 200; P-Amino benzoic acid 10; Inositol 10; Nia- cin 4; Ca-D-Pantothenate 4; Riboflavin 0.8; Thiamin HCl 0.5; Pyridoxine HCl 0.5; Folic acid 0.2; Biotin 0.04; Vitamin 812 0.0003; 51 The NPR was calculated from the formula: wt gain + wt loss of non N diet group protein intake NPR = The feces from each group were collected during the period of 16th and 26th day of the experiment and the nitro- gen content used in calculation of the protein digestibili- ties. Feces were separated from adhering particles of diet before they were ground for nitrogen determination. Apparent Digestibility in Vivo It was calculated from data on nitrogen consumption from the PER experiment and the nitrogen content of the feces for the period between the 16th and 26th day of the experiment. The following formula was used: N in the diet (g)—N in feces (9) N in diet (9) Apparent digestibility = Apparent Digestibility in Vitro The enzymes used in this experiment were: Trypsin from Hog Pancreas, Type IX Crystalized, dialyzed and Lyo- philized powder, 14,350 BAEE units activity per mg protein. a-Chymotrypsin from Bovine Pancreas, Type 11, 3x crystalized and Lyophilized with 50 units activity per mg protein. 52 Peptidase from Hog Intestinal Mucosa, Grade III, with 30 units activity per g . solid. All enzymes were obtained from Sigma Chemical Company. The method utilizing a combination of trypsin-chymo- trypsin-peptidase (Hsu g3 al., 1977) was used. Amount of sample providing 6.25 mg protein/ml was suspended in 50 ml of distilled deionized water and the pH was adjusted to pH 8.0 with 0.1 N H01 and/or NaOH, while stirring in a 37°C water bath. The multienzyme solution containing 1.6 mg trypsin, 3.1 mg chymotrypsin and 1.3 mg peptidase/ml, was maintained in an ice bath and adjusted to pH 8 with 0.1 N HCl and/or NaOH. Five milliliters of the multienzyme solution were added to the protein suspen- sion, which was being stirred at 37°C. A rapid decline of pH occurred immediately, caused by the freeing of amih: acid carboxyl groups from the protein chain by the proteo- lytic enzymes. The pH drop was followed in a Corning Scientific Instruments pH meter model 10 and recorded every minute over a 10 minute period. The multienzyme solution was freshly prepared before each series of tests, and its activity was determined using a casein of known in 1119 apparent digestibility. The drop of pH in 10 minutes was used to estimate the apparent digestibility from the equation y = 210.464 - 18.108 x, 53 where “x" was the pH of sample suspension immediately after the 10 minute digestion with the multienzyme solu- tion. The buffering capacities of various samples were determined as follows: 50 ml of-a sample suspension (6.25 mg protein/ml) was adjusted to pH 8, with 0.1 N NaOH or HCl at 37°C. The protein suspension was then slowly titra- ted with 0.1 N HCl to pH 6.45 over a 10 min. period. The amount of acid consumed was used as an index for the buffer capacity of the protein source. Available Lysine by Rat Growth Assay Male weanling rats, 21 days old, of the Sprague-Dawley strain were used. After the adjustment period of 4 days, during which they were fed a standard rat diet, the rats were closely matched in weight and divided into groups of six. The average weight of a rat was 73 i 2 g. The groups were assigned at random to the various diets. Rats were kept in individual cages and diet and water were provided ad libitum for 27 days. The basal diet was composed, based on the nutrient requirements for optimum growth of the rat, as they are defined in the edition of National Academy of Sciences #10, "Nutrient requirements for laboratory animals". The compo- sition of basal diet is showing in Table 4. The amino acid mixture contained all the 17 amino acids (omitting lysine), dispensible and indispensible, 54 Table 4. Composition of basal diet for available lysine determination Amino acid mixture 11.29% Salts 4% Vitamin mix 1% Oil 5.5% Sucrose 30.0% Starch 32.39% necessary for optimum rat growth. The amino acid mixture composition is showing in Table 5. Each amino acid in this diet is represented by an amount equal to that suggested by the National Academy of Science for rat growth, plus 80% of the quantity of that amino acid present in 100 g of wheat flour. Itis assumed that the average availability of wheat flour amino acids is 80% (Kuiken, A.K. t al., 1948; Calhoun, K.W. _3 _l., 1960). Lysine hydrochloride was added to these diets in amounts to give lysine content 0.0, 0.1, 0.2, 0.3 and 0.4% of the diet. The test materials were incorporated into the basal diet at the 10% protein level at the expense of carbohy- drates. The prepared diets were refrigerated during the dura- tion of the experiments. Available Lysine by FDNB Procedure Carpenter's direct FDNB method as modified by Booth (1971) was used. The samples were fine enough to pass through a 0.125 mm sieve. Quantity of sample 1.0-1.2 gr was placed in a 500 ml round bottom flask, with four anti-bump glass beads. Ten ml of NaHCO3 solution (8% w/v) was added, the flask was gently shaken by hand and left to stand for 7-10 min until the sample was wetted. Care was taken so that the sample would not be widely scattered in the flask. 56 Table 5. Amino Acid Composition of Diets (g . of amino acid/100 g . diet) Amino acid A B C L-arginine .67 .36 1.03 L-asparagine .44 .20 .64 L-glutamic acid 4.40 2.20 6.60 L-histidine .33 .15 .48 L-isoleucine .61 .33 .94 L-leucine .83 .50 1.33 L-methionine .34 .ll .45 L-cystine .34 .17 .51 L-phenylalanine .44 .37 .81 L-tyrosine .44 .28 .72 L-proline .44 .20 .64 L-threonine .56 .22 .78 L-tryptophane .17 .10” .27 L-valine .67 .33 1.00 Glysine .20 .10 .30 L-alanine .20 .10 .30 L-serine .20 .10 .30 Total 11.28 17.10 A: Amino acid content of diet for optimum rat growth (Nu- trient Requirements for Laboratory Animals, ed. National Academy of Sciences #10). B: Eighty percent of the amino acid content of wheat flour (by Swaminathan, Ch. 4, Vol. III in Newer Methods in Biochemical Nutrition). C: Amino acid composition of the diets used for construction of the standard curve, relating response of animals and different amounts of available lysine in the diet. 57 Fifteen m1 of ethanol into which 0.5 ml of FDNB was dissolved (Eastman Kodak Co.) was added to each flask. The flask was shaken, gently at first, for 4 hrs. at room tem- perature in a Junior Orbit Shaker (of the Lab-Line Instru- ments, Inc.) at 200 rpm. Halfway through the shaking period, the flask was twirled to disperse the sample and to make sure that all particles were wetted by the FDNB solution. After the shaking period, the ethanol was evapo- rated in a boiling water bath (it was checked by making sure the flasks lost 12.5 g of their weight). The mixture was cooled - mixed with 30 ml of 8.1 N HCl and refluxed for 16 hr. The heat was turned off, the condenser was washed with a little water and the flask disconnected. The contents were filtered, while still hot, through a Whatman filter paper #2 and rinsed repeatedly with hot water until the total filtrate was almost 250 m1. When the filtrate was cooled, it was made to volume and mixed. The hydrolyzates were stored at 4°C, in the dark, until used for further analysis. Two ml of the filtrate was pipetted into each of two stoppered test tubes, A and B. The contents of tube B were extracted with about 5 ml of peroxide-free diethyl ether. Most of the ether was sucked out with a water aspirator. The dissolved ether was removed by placing the tube in a steam bath until effervescence from the residual ether ceased. A drop of phenolphthalein solution was added 58 followed by the addition of NaOH solution (120 g/litter) until the first pink appeared. Two ml of carbonate buffer pH 8.5 (19.5 g NaHC03, and l g Na2C03 dissolved in 250 ml of water and pH adjusted to 8.5) were added. Under the fume hood 5-6 drops (about 0.01 ml each) of methyl chloro- formate (Eastman Kodak Co.) was added. The tube firmly stoppered and shaken vigorously. The pressure was released cautflnwly. After about 8 min., 0.75 ml concentrated HCl was added dropwise and very cautiously at first, with agi- tation to prevent frothing. The tube was shaken occasional- ly. The solution was extracted with ether 4 times as described above, and the residual ether was removed by warming the tube in a steam bath. The tube was cooled and the contents made up to 10 ml with water. In the meantime, the tube A was extracted three times with peroxide-free diethyl ether. Residual ether was removed and the contents made up to 10 ml with 1 N HCl. The absorbance of both A and B were read at 435 nm against deionized water. Reading A minus reading B (the blank) is the net absorbance attributable to DNP-lysine. With each set of samples analyzed, the absorbance of the standard e-DNP-lysine solution, carried through the same procedure as the samples, was determined. The standard solution was prepared by dissolving 254 mg of mono-e-N-dinitrophenyl-lysine hydrochloride (e-DNP- lysine, mol. wt. 348.8, lysine content 41.9%) (Sigma 59 Chemical Company), in 200 m1 of 8.1 N HCl. Two tenths of 1 ml of this solution was used in each run for tubes A and B. This amount contained 0.106 mg lysine when it was diluted to 10 ml and gave a net absorbance of .400 at 435 nm. The available lysine content was calculated from the following formula: w c = ixflxfl‘. As ah WuxP x 100 0 11 mg available lysine/100 gr protein Ws = mg of lysine in standard solution used in each run (usually 0.2 ml of standard solution used containing 0.106 mg lysine) As = net absorption (absorption of tube A - absorption Tube B) of standard solution Au = net absorption of unknown ah = aliquot of filtrate (usually 2 ml used) Vh = total volume of filtrate (usually 250 ml) W = weight of sample hydrolyzed in mg 60 P = protein content of sample Because part of the DNP-lysine of the sample was lost during the reflux digestion, a recovery factor was calcu- lated by using e-DNP-lysine as internal standard. Dupli- cate samples of each material were treated as described above as far as the addition of 8.1 N HCl, but only part of the last reagent was added. Standard e-DNP-lysine solution was then added containing lysine approximately equal to the amount of available lysine in the sample. More 8.1 N HCl was added to a total volume of 30 m1, and the procedure was continued as previously described. From the difference in available lysine between the two samples and the known amount of lysine added into the second sample, the percent recovery factor was calculated. The values of samples for available lysine were corrected only for half of the loss estimated by the internal standard since it has been shown by Booth (1971) that the loss of protein-bound dinitrophenyl lysine (as the lysine of the sample) is only half of the loss of free e-DNP-lysine (as that of the internal standard). It is important in the above procedure that the stan- dard e-DNP-lysine be added to the sample after it has been acidified with the addition of part of the 30 ml of 8.1 N HCl. Otherwise, the added e-DNP-lysine in the alkaline solution of the sample and with excess of FDNB, will be converted to a, e, di-DNP-lysine and will be removed during 61 the following step of washing with diethyl ether. Amino Acid Analysis Acid Hydrolysis Amino acid analyses were performed on HCl hydro- lyzates of protein using a Technikon Instruments Co. amino acid analyzer, according to the procedure of Moore 33 l. (1958). Samples consisting of approximately 10 mg f C protein were weighed into 10 ml ampoules. Five ml of 6N HCl were added to the ampoules. The contents were frozen in a dry-ice acetone bath. The ampoules were evacuated with a high vacuum pump. As the contents were slowly melted, the gases were removed. The contents were then refrozen and the ampoules were sealed using an air-propane flame. The sealed ampoules were placed in an oil bath in a forced draft, recirculating oven regulated at 110 :2C2for 24 hours. After hydrolysis the ampoules were opened, and 1 ml solution containing 1 um/ml nor-leucine and l um/ml S-B- (4-pyridy1ethyl)-L-cysteine (PEC) was added as internal standard. The hydrolyzate was then quantitatively trans- ferred from the ampoule to a 25 ml pear-shaped flask. The hydrolyzate was evaporated to dryness on a rotary evaporator. The dried sample was washed with a small amount of deionized water and again taken to dryness. In all, three washings were performed to remove residual HCl. The washed and 62 dried hydrolyzate was dissolved in lithium citrate buffer (pH 1.9) and diluted to a volume of 4 ml. The solution was then filtered through a 0.22 u Mil- lipore Filter and 0.2 m1 aliquots were used for analysis. The chromatograms were quantitated by the peak area method. Standard amino acid mixtures were analyzed using the same ninhydrin solution. Enzymatic Hydrolysis The following enzymes were used (Sigma Chemical Com- pany). Pepsin from Hog stomach mucosa, 2x crystalized and lyophilized powder with 3400 units activity per mg solid. Pancreatin from Porcine pancreas, grade VI, with acti- vity 4xNF Grade. In a 125 ml flask, the sample was dispersed in 25 ml deionized water and the pH was adjusted to 1.8 by slowly adding HCl 1N. The volume was adjusted to 30 ml with de- ionized water. The amount of sample used was such as to give 1% protein concentration at the volume of 30 m1. To the deionized water 0.02% Na azide (Sigma Chemical Co.) was added to prevent microbial growth. Fifteen miligrams pepsin were added and after the pepsin was dissolved, the flasks were stoppered and placed in a 37°C water bath for 3 hrs. under continuous stirring. Peptic digestion was terminated by adding 1 N NaOH until pH 8.3 was reached. Deionized water was added to 63 make the total volume of added liquid 5 ml. Then 17.3 mg pancreatin was added to each flask and contents were main- tained at 37°C. After 3 hours of pancreatin digestion, a second 17.3 mg pancreatin was added and the digestion con- tinued for an additional 21 hrs. at 37°C and with continuous stirring. Enzyme blanks were treated identically except for the absence of the protein sample. The digestion of the samples by pepsin and pancreatin was stopped by adding sufficient tichloroacetic acid (TCA) to obtain a final concentration of 15%. After sitting overnight at 4°C, the samples were centrifuged in a clinical centrifuge at lOOxg for 5 min. to remove the precipitated material. The supernatant was filtered through Whatman filter paper #1 and the filtrate was analyzed for nitrogen and free amino acids. Three ml of the TCA supernatant plus 1 ml of nor-leucine solution (111mole/ ml) were washed with 30 m1 diethyl ether to remove the dissolved TCA. Residual ether was removed by heating the sample at 80°C for 3 min. Two ml of the washed sample was diluted to 5 ml with lithium citrate buffer, pH 1.9. The diluted sample was filtered through 0.22 millipore filter. Amino acid analyses were performed as previously described. Since acid hydrolysis was not employed, the sulfur amino acids were not destroyed and could be quantitated directly. 64 Sulfur Containing Amino Acids Since methionine and cystine are oxidized during acid hydrolysis, they are analyzed separately. The methods of Schram gt _1. (1964) and Lewis (1966) are used. These methods involve performic acid oxidation of methionine and cystine to methionine sulfore and cysteic acid, respecti- vely. The performic acid solution was prepared by mixing one volume of 30 percent (w/w) hydrogen peroxide with nine volumes of 8.8 percent (w/w) formic acid. This mixture was allowed to stand for one hour at room temperature. Amount of samples representing 5-8 mg protein was weighed into a 25 m1 pear-shaped flask. Ten ml of performic acid solution was added and oxidation was carried out at 4°C for 24 hrs. After the oxidation, 1 ml of norleucine ( 2 umole/'ml) solution was added. The performic acid was removed in a rotary evaporator. The dried sample was quan- titatively transferred to a 10 ml ampoule with 5 ml of 6N redistilled HCl. Hydrolysis and amino acid analysis were performed as previously described. Igyptophan Tryptophan is very labile during acid hydrolysis, and after prolonged hydrolysis, little or none of the amino acid is left. Therefore, it was determined colorimetrically after hydrolysis with the enzyme pronase as described by Spies (1967). 65 a. Thirty mg sample was weighed directly into a 2.0 ml glass vial with a screw cup. b. To each vial, 0.1 m1 of pronase hydrolytic solution and a drop of toluene, as preservative, was added. Pronase hydrolytic solution was prepared fresh by adding 100 mg pronase to 10 ml of 0.1 M phosphate buffer, pH 7.5. The suspension was shaken gently for 15 min., then clarified by centrifugation for 15 min. at 10,000 rpm. c. The vials were closed and incubated for 24 hrs. at 40°C. After incubation, 0.9 m1 of 0.1 M phos- phate buffer, pH 7.5 were added to each vial. The uncapped vials were placed into 50 ml Erlen- meyer flasks containing 9.0 m1 of 21.2 N sulfuric acid and 30 mg of dimethylaminobenzaldehyde (DAB). The vials were tipped over and the con- tents were quickly mixed by rotating the Erlen- meyer flasks. Samples were cooled to room temperature and kept in the dark at 25°C for six hours. d. 0.1 ml of 0.045 percent sodium nitrate solution was added to each Erlenmeyer flask. After gentle shaking, the flasks were left standing for 30 min. for the development of the color. The absorbance was measured at 590 nm, using a Beck- man DU spectrophotometer. 66 Duplicate blanks of the pronase hydrolytic solution were treated similarly and the tryptophan content of pronase was subtracted from the total tryptophan content. A standard curve from zero to 120 mg tryptophan was prepared according to the Procedure E described by Spies and Chambers (1948). D, Lvtryptophan (2.4 mg) was dissolved in 20 m1 21.2 N sulfuric acid containing 60 mg of dimethylaminobenzalde- hyde (DAB). 0, 1, 2, 4, 6, 8 and 10 m1 of this solution were made up to 10 ml with solutions of 21.2 N sulfuric acid containing 600 mg DAB/200 ml, and placed in 50 ml Erlenmeyer flasks. The mixtures were kept in the dark at 25°C for six hours, then 0.1 ml of 0.045 percent sodium nitrate was added to each flask. The flasks were allowed to stand 30 minutes for color development and absorbance was measured at 590 nm, using a Beckman DU spectrophoto- meter. A straight line relationship was obtained between absorbance and tryptophan content. RESULTS AND DISCUSSION The loaves of bread were weighed and their dimensions were measured after they were cooled to room temperature. The average weight was 1422.3 9 and their average specific volume was 3.06 cm3/g . The crumb made the 74.5% and the crust the 25.5% of the whole loaf by weight. The color of the dried and ground samples was measured with a Hunter Color Difference Meter D-25 model. The instru- ment was adjusted with the white standard and then with the yellow brownish standard #2814. The powdered samples were held in a Petri dish and 4 readings of each sample were taken, rotating the dish by 900 after each reading. Table 6 shows the average L, a, b readings for each sample. The L. a, b values reflect the changes in color due to baking and toasting of the samples. The L values measure lightness of the color from 100 for white to zero for black; a, measures redness, when plus, gray when zero and greenness when minus, and b values measure yellowness when plus, gray when zero and blueness when minus. The powdered samples were analyzed for moisture, oil content, total protein and ash. The results of these analyses are shown in Table 7. 67 68 m.m_ mm.a _.~P e.gp 0.0P om.~_ m.¢. N.m a k.m mm. _.¢ N.¢ 0.0 mk. m.m o.P m o.~o m.,m m.~¢ F.0m P._m m.e~ o.Po m.Pm 4 women xFE women czopm mho awn; Ezcu . zouz -zoaz mks u a apes: umacm .mmzpo> Leave mucogmmmwu Lopou Lopez: .mmfinsmm nczogm ccm umwcu me mnu we Lopou .o mpawh 69 Table 7. Composition of air—dried and ground samples i vi Product Moisture Protein, % Ash Crude % (Nx5.75) % fat, % Bread mix 11.48 13.13 1.25 .57 Whole bread 4.73 13.33 1.27 .54 Crumb 4.86 13.44 1.25 .56 Crust 4.48 13.09 1.43 .50 LTB 4.59 12.99 1.34 .57 DTB 4.32 13.60 1.33 .55 NFDM-flour 10.10 13.60 1.05 .56 NFDM-bread 4.84 13.52 1.33 .77 70 Nutritional Value of the Heat Treated Products The growth curves for the 8 groups of the rats fed the experimental diets are shown in Figure 3. Rats on the NFDM-flour diet grew at the highest rate, 1.6 g/rat/day, while the rats on the DTB diet had the slowest rate of growth 0.13 g/rat/day. The crust diet could not even support the maintenance of rats, thus resulting in weight loss during the experimental period of 28 days. The crumb diet resulted in a growth rate of 1.0 g/rat/day, while the Bread mix and the NFDM-bread gave almost the same growth rate, 0.75 g/rat/day. LTB gave a growth rate of 0.35 g/ rat/day, which is between those of DTB 0.13 g/rat/day and whole bread 0.44 g/rat/day. The average values on food intake, protein intake, weight gain, food conversion ratio, PER and NPR are shown in Table 8. Heat treatment decreased, in general, the nutritive value of the products (except for the case of the crumb), resulting in increasing food conversion ratio =ggfood intake 9 weight gain as the heat treatment increased. Food conversion ratio was 52.6 for the DTB, 20.0 for the LTB, 19.2 for the whole bread, 12.2 for the Bread mix, 11.7 for the NFDM-bread, 9.0 for the crumb and 7.0 for the NFDM—flour. For the crust no food conversion ratio could be calculated, since the 71 .ummgnuzouz vcm Assopwuzodzv esopm xpms xcv gamucoc .Ampov ummgn umwmmou xgmc .AmHAV umwca vmummou ucmwp .umaco .nszgu .ummgn wpozz .xve vmmgm asp mcwcwmucou mumwu um» mum; mcwpcomz do swam pgmmmz mmmcm>< .m mczmwu mast .mewh om mm mm . mP up op o N _ _ A mko 411. 4 - cm 9 0m names mFosz 0 \\\\\\\¥ \\. ummcplzouz o st ummcm a cop nascu \\\\\\\\\ opp _ omp xzopmuzouz 6 ‘1u619M Kpog 72 .mxmucw :wmpoca\umwu zucoc co mum; we mmop p: + :Pmm p: n .m.N «we cwammu toe umpmswu< Auv Auv mmz .Amo.o v av apocau uwmwcmwm gmwwwu coEEou cw Lmupmp m paoguwz mcmmz .ummu wmcmc m—nwppsz m.:mu::o Any .m .:?mm pgmwmz\m axons? uwmu Amv mm.m om.~ cm. A ww.~ m.m m.mp n m.P~_ m.m A P.~¢ mm A N.No¢ :wmmmu Pm.P mn. one. A um. N.P_ m.~ H m.FN ~.N u m.e~ m.o~ A o.mm~ ummgnuzouz am.~ n_.~ mom. A mm.P 0.5 «.mp H m.m¢ ~.¢ H F.mm m.F¢ n o.o~m Lsopm-zodz mm. mp. cou now» .cmmm p3 .mxmucw :wmpoga .mxmucv ummu .w mpnmh 73 consumption of food resulted in weight loss rather than in gain. PER values calculated from the formula weight gain/pro- tein consumed. One group of rats was fed a diet containing 10% casein protein and it was used as control to calculate the corrected PER values. The approximate weight gain for this group was 4.36 g/rat/day with an average feed intake of 14.33 g/rat/day. The PER of this group was taken as 2.5 and the PER of the other groups were adjusted to the casein 2.5 value. The adjusted PER value for the bread mix was 0.75, rep- resenting 30% of the casein PER. The whole bread PER, 0.46, represents 18% of the casein PER. The diets containing the crumb showed higher PER, 0.91, representing 36% of the casein PER, LTB and DTB gave PER values of 0.40 and 0.16, corresponding to 16% and 6.2% of the casein PER,respectively. Fortification of the bread formula with 4% non-fat dry milk solids increased the PER of the mix to 1.17, repre- senting 47% of the casein PER, and the PER of the NFDM- bread to 0.75 or 30% of the casein PER. The inferior quality of the wheat protein has been observed by many researchers. Osborn and Mendel (1914) showed that wheat proteins are generally of poor biological value when they are compared with animal proteins and studies by Muramer and Harper (1959), Howard t 1. (1958) and Bender (1958) showed that the protein quality of wheat 74 flour is inferior to that of animal proteins for the growth and maintenance of rats. Non-fat dry milk solids are known to improve the nutritional value of bread. Saviston and Kennedy (1957) showed that addition of non-fat dry milk to whole wheat bread at the levels of 3% and 6%, effectively improved the protein quality. In this study, 4% NFDM-solids in the unbaked ingre- dients (a commonly used level) caused 56% increase in the PER of the unbaked ingredients and 63% increase in the baked bread. The protein of the crust was damaged most severely. A negative PER, -0.23, was observed as a result of the fact that rats on this diet lost about 4.7 grams weight during the experimental period of 28 days. Next most severely damaged nutritionally was the protein of DTB, the PER of which decreased 79% as compared to that of the unbaked bread mix. The PER of LTB decreased by 48% and that of the bread by 38%. Compared to the whole bread, light toasting caused a 13% decrease in PER, while dark toasting caused a 66% decrease. Crumb showed an 18% increase in the PER compared to the bread mix, which was not heated at all. The mild heating of crumb (protected by the crust) probably caused slight denaturation of the proteins and thus increased their utilization by the rat. This hypothesis is supported by the fact that the in 1119 apparent digestibility of crumb was higher (88%) than that of the bread mix (86%) (Table 9). 75 .ces oe cw we.o on o eoee meanm «so ea In men mmomeuoo on ooeeacme em: 2 Fo.o ea memoeeeeeeeeon on auwoonou e.m 0.x ~.w 0.5 m.m N.m m.m m.n ¢.~_ unvemwwam n.mm m.~m o.om n.—w w.Pm m.mn w.mm m.~m P.om oeuw> :H m.Fm w.mw m.mm N.m~ “.mm m.m~ m.ww w.mm m.om o>w> =H women esopw women st common -zomz -znmz mpg men umzeu nsseu mponz womem oepw> cw wco o>w> cw .mumpw Foucmswemaxm we Afiv xuwpwnvpmmmew z ucmeoaq< .m mpnoh 76 Other researchers (Hankes, Riesen _t _l., 1948) noticed that when dry casein was subjected to mild heat treatment, the in vitro availability of amino acids increased, although severe processing lowered their ifl_vitrg digestibility. Osner and Johnson (1975) also reported that susceptibility of heated caseins to enzymatic digestion increases just before the onset of denaturation and before any cross-linking takes place. The severe heat damage of the crust and the decrease of nutritional value of the bread was subject of earlier studies. Stout and Drosten (1933) studied the time tem— perature relationships in the baking of bread, and found that interior temperature of loaves did not exceed 1000 to 102°C during the baking process. Barackman and Bell (1938) showed that the temperature of the crust rises to 100°C and then to 150°C as the dehydration of the dough progresses and the browning sets on. Studies in 1931 by Morgan and by Kon and Markuse have shown that, while the protein quality of the crust is lower than that of the crumb, the protein quality of the crumb and of the whole loaf is the same. The results of this work do not agree with the above con- clusion. The PER of whole bread is 50% that of the crumb. Although the exact PER of the mixture crumb (74.5%) and crust (25.5%) that make up the whole bread cannot be calculated from the PER values of these two portions, because of the non-proportionality of the PER, the effect 77 of the very low PER value of the crust on the nutritional value of the whole bread is obvious, as the PER of the mixture is 50% that of the crumb. Baking only caused a decrease of 38% in the PER of the unbaked ingredients and 35% of the ingredients con- taining 4% NFDM-solids. The NPR method grades the proteins of the samples in the same order as the PER method, but the NPR values are higher than the corresponding PER values for all samples. This is understandable as NPR takes into account the main- tenance requirements of the rats, thus scoring the proteins for their potential to support maintenance as well as growth of the rats. For the same reason the percent differences in NPR values between different samples are smaller than the differences in PER of the same samples; for example the percent difference in PER values between crumb and DTB is 88% while this difference is only 41.3% for the NPR values of the same samples. The same comparison holds true for other samples. Digestibility The apparent nitrogen (N) digestibility of the experi- mental diets were estimated 1 vivo and 1 vitro. The in vivo N digestibility was calculated from data for N consumption and N excreted in feces, collected during the PER tests. The in vitro N digestibility was calculated 78 from the drop in the pH of the sample at 10 minutes, using the multienzyme system trypsin-chymotrypsin-peptidase, and by the equation y = 210.464 - 18.103 x where y = apparent digestibility, x = pH of the sample after 10 min. The values obtained by both methods are shown in Table 9 along with the buffering capacities of the samples. Figure 4 shows the typical curve of decreasing pH during the 10 min period, by the action of the enzymes on the samples. The samples show a variation in their apparent diges- tibility values 1 vivo which are, in general, parallel to the intensity of heat treatment. For example, both breads, whole bread and NFDM-bread, had apparent digestibility values lower than the mix of ingredients from which they were made. Crust, LTB & DTB had lower values than the less heated whole bread, crumb and bread mix. However, the digestibility values i vivo for crust (79.9%) and DTB (77.2%) do not follow the pattern of decreasing digestibility with increasing heat treatment, observed for the rest of the samples. The higher N digestibility of the crumb over that of the bread mix can be explained by the hypothesis presented earlier, that mild heat treatment of the crumb, protected by the crust, caused slight denaturation of the crumb pro- teins and increased their susceptibility to the digestive enzymes. The addition of NFDM-solids to the ingredients increased slightly the apparent 1 vivo digestibilities 79 .meaust mmstcm mmowwuoma wco :_queuosanu .cwmnaeu nu_3 mmpasom m>wuoucmmmeome mo cowuonaocw xn wmcwouno mm>e=u mew» m> :Q mo mm_quxm .o mezmwm :_E .mawh op m m n m m w m N P o.o commouo it'll/C XPE ll. womemu o.~ o.m of the non-heated mixes (88.8 vs 86.3%), while it did not have any apparent effect on the baked breads (83.8% for both whole bread and NFDM-bread). The in vitro apparent digestibility values were in close agreement with their corresponding values 1 vivo except for the case of crumb for which it showed a lower value than that for the bread mix. This is contrary to the results obtained by the 1 vivo method. The buffering capacities of the protein samples did not differ much from each other and were not expected to have affected the measurement of the in vitro digestibili- ties. Available Lysine by Rat Growth Assay The choices for a basal diet for such experiments are either a natural feeding stuff low in lysine content, or a mixture of synthetic amino acids. The synthetic amino acids are expensive and result in a relatively unpalatable and unphysiological diet for the rat, due to their high osmotic pressure. 0n the other hand, a basal diet con- taining natural feeding stuffs of low lysine content has the disadvantage that an assumption has to be made as to the biological availability of lysine it contains. An all synthetic amino acid diet was used in this work in order to have an accurate control of the amount of available lysine in the diets used for the construction of the 81 standard curve. The diets prepared as it is described in the Methods and Materials section, had an amino acid con- tent in excess to the requirements for optimum growth of the rats, creating that way an amino acid imbalance in the diet. This imbalance was expected not to have any adverse effect on the response of the rats to the limiting levels of lysine which was tested in these experiments. It was shown by Schweigert _t _l. (1965) that a 50% increase in the amounts of all amino acids of a similar basal diet did not affect the response of rats to the limiting levels of lysine. The same was shown to be true for the chick, which is as susceptible to an amino acid imbalance as the rat. Fisher et a1, (1960) showed that such amino acid imbalances resulted in reduced feed consumption, but there was no evidence that the limiting amino acid was utilized less well than in the case where no imbalance was created. Table 10 shows the food intake, weight gain and weight gain per 100 9 food for the rats under the differ- ent diets used in this experiment. The rat performance data obtained were subjected to three methods of calcula- ting available lysine. A standard curve was drawn by plotting the weight gain versus percentage of added lysine to the diet (Figure 5a). As it is shown in this figure, the weight gain is influenced strongly by the differences in food consumption; food consumption increases as the percent lysine in the diet increases being 0.72, 0.92, 82 Table 10. Average food intake, weight gain, weight gain per 100 9 food of rats fed for 28 days synthetic and experimental diets. Supplement to Food Height Weight gain per basal diet intake gain 100 9 food None 129.3 -12.2 -9.4 .1% Lys 164.8 -3.0 -l.8 .2% Lys 192.2 16.4 8.5 .3% Lys 250.4 40.8 16.3 .4% Lys 330.2 80.7 24.5 Bread mix 200.2 16.3 8.1 whole bread 196.2 16.7 8.5 crumb 239.7 32.0 13.3 crust 150.9 -2.0 -1.3 LTB 186.8 14.5 7.8 DTB 175.0 7.2 4.1 NFDM—flour 257.1 38.5 15.0 NFDM-bread 232.3 26.8 11.5 woo» a co. emu zoom unmemz mamem> Any wco coon unmwm: mamem> on umww mnu no m:_me mpnopwo>o unmuema .m me=m_e neeo e. oeemee a “mew ee oeemee a e. o. N. _. o e. m. N. e. 83 1% _ _ a l _ q u _ op- O O F p00; 6 001/6 U196 1H513M .1 O N O N l o m l o m J o e ‘u196 zqfigaM l o LO 6 I o \o 86 1.07, 1.39, 1.83 gr/day/rat for the groups under the diets containing 0.0%, 0.1%, 0.2%, 0.3% and 0.4% lysine respec- tively. In the second method, the standard curve was obtained by plotting weight gain per 100 gr of food con- sumed (Figure 5b). Gupta _£._l- (1957) stated that when weight gains were adjusted in this manner, greater unifor- mity among experiments was observed regarding the availa— bilities of lysine from various food items. In the third method, the weight gains were plotted against the grams of available lysine consumed (Figure 6). Because this method takes into account the differences in actual lysine con- sumption, it represents a refinement over adjustment for food intake. The available lysine content of the samples was calculated from this last curve and from the average weight gain of groups under the test diets. The values obtained are shown on Table 11. Available Lysine by the FDNB Method The FDNB-method was applied to the individual sam- ples which were previously tested for available lysine by the rat growth assay. Carpenter's method was used as it was modified by Booth (1971). A large sample/acid ratio (3.25 - 3.9 mg N/ml acid) was used in the hydrolysis of the samples which were treated with FDNB and the hydroly- zates were filtered while still hot and washed with hot water. Both of these practices helped to recover as much as possible DNP-lysine from the hydrolyzed sample. Because 85 80 '- 70 l— 60 __ 40 #— Neight gain. 9 20 10 «— -]0 1' _ 1 1 1 L 1 1 l 0.0 .2 .4 .6 .8 1.0 1.2 1.4 Available lysine consumed, 9 Figure 6. Available lysine consumed in 9, versus weight gain in g, during 28 days. 86 o_.N ee.N we.e mo._ eo.e me.N mm.e ee.N enema, meno_eo>< women esopm women xwe -znez -znez men men “ween asaeu mean: women mommo npzoem uoe seem Acemuoea a oo_\mv mmapo> mwwmxp mpnopwo>< .pp mpnoh 87 during acid hydrolysis part of the DNP-lysine of the sample was destroyed, an internal standard, free e-dinitrophenyl lysine (e-DNP-lys), was taken through the same procedure as the sample, and a recovery factor was calculated for each sample. The amount of s-DNP-lys added as internal standard in each sample was approximately equal to the amount of available lysine present in the sample. However, the values for available lysine were corrected for only one half of the losses of the internal standard, since the losses of e-DNP—lys still bound to the protein are half of the free e-DNP-lys (Booth, 1971). Table 12 shows the uncorrected values, percent recovery of internal standard, correction factor, and corrected values of available lysine obtained by the FDNB method. The average recovery value for the internal standard was 72.9%, although there was some variation in the reco- very, from product to product. Booth (1971) extensively studied the effect of sample composition on the recovery of the internal standard and found that it is greatly affected by the presence of carbohydrates in the sample. He reported the following percentages of recovered e-DNP- lys: 92% for fried chicken meat, 92% for blood meal, 86% for wheat gluten, 80% for bean flour, 69% for oat-feed and 68% for wheat. The recovery with the products of this experiment, 72.9%, is in good agreement with that of Booth for wheat. The loss of added, as well as the indigenous 88 mo.~ mp. em om.m mw.m mmp. mo om.m m_.~ mm—. mu ~m.~ .m ooe\mee m oop\map mmapo> wmuumeeoucn .uoea .Fwo>o a mmapo> wmuumeeou eouuoe comummeeou weowcoam Focemucw we >em>oume ucmuema .poeo .Pwo>o m women zouz eaoPe 12amz women mean: weowcoum Pocemucw co soee meouooe cowpmmeeou wco wonums mzom mnu an mm=Po> mcwmap mpnopwo>< .NP mpnoh 89 e-DNP-lys. is attributed to the reduction of the N02 groups of e-DNP-lys. to NH2 groups, by the carbohydrates in the sample, during acid hydrolysis. Table 12 shows that pro- cessed products gave higher recoveries of e-DNP-lys. than the products that had not been processed i.e. whole bread 73%, crust 74%, LTB 77%, DTB 80%, NFDM-bread 74% versus bread mix 70%, crumb 70%, NFDM-flour 69%. This variation in recoveries might be due to the decrease of the reducing power of sample carbohydrates during processing. Possibly, the carbohydrates lost part of their reducing power during the heat treatment and were not able to reduce as many N02 groups to NH2 as the carbohydrates of the non-heated products. Total Amino Acid Analysis Table 13 shows the amino acid content of the products, in grams of amino acid per 100 grams protein. In the same table the FAD/WHO reference pattern of essential amino acids is also shown. Comparison of the amounts of essential amino acids of the product with those of the FAD/WHO pat- tern shows that lysine is the first limiting amino acid in all products, with threonine as the second limiting amino acid. The chemical scores of the products (9 lys of product/g lys in FAO/NHO pattern) are 49 for the Bread mix, 46 for the whole bread, 50 for the crumb, 39 for the crust, 45 for the LTB, 40 for the DTB, 61 for the NFDM- flour and 55 for the NFDM-bread. 90 Table 13. Total amino acid composition of samples (expressed as g residue/100 gr protein). __L Amino FAO/ Bread Whole Crumb Crust LTB DTB NFDM- ”NFDM acid WHO mix bread flour bread Lys 5.5 2.70 2.49 2.69 2.10 2.45 2.15 3.25 2.97 Thr 4.0 2.05 2.29 2.37 3.03 2.71 2.43 2.94 2.30 Cyst. 2.00 1.94 1.90 1.80 1.95 1.86 2.05 2.00 Met 3'5 1.70 1.85 1.75 1.70 1.77 1.80 1.72 1.73 Val 5.0 3.85 4.13 3.91 4.89 4.11 4.07 4.35 4.41 Ile 4.0 3.59 3.66 3.41 3.93 3.79 3.67 3.50 4.03 Leu 7.0 6.15 7.19 7.66 7.70 7.56 6.39 6.35 8.53 Tyr 2.74 3.02 2.60 3.62 3.26 2.94 3.06 3.17 Phe 6'0 3.96 4.29 3.95 5.14 4.78 4.51 4.44 4.77 Trp 1.0 0.99 .98 1.00 .97 .98 .95 .98 .96 His 1.98 2.59 2.11 3.53 2.11 1.64 2.04 2.17 Arg 3.56 4.93 4.93 4.28 4.53 3.77 4.78 4.69 Asp 3.39 4.11 3.67 4.91 4.27 3.89 4.82 4.62 Ser 3.64 3.51 3.81 4.89 4.20 3.72 3.68 3.99 Glu 43.74 39.40 42.40 29.81 36.46 41.17 37.79 36.66 Pro 9.45 9.03 6.28 11.80 8.34 9.25 9.18 7.83 Gly 3.45 3.49 3.62 3.80 3.73 3.75 3.23 3.07 Ala 2.71 2.93 2.55 3.42 3.76 2.75 2.79 2.89 91 Total lysine content of the products decreases with increasing heat treatment, the values varying from 2.7 g lys/100 g protein for the Bread mix and crumb to 2.10 g lys/100 g protein for the crust. The NFDM fortification increased the lysine content of NFDM-flour to 3.25 g/ 100 g protein and that for NFDM-bread to 2.97 g lys/100 g protein. Wheat and its products are known to be deficient in lysine. When lysine was added to the wheat diet, there was an increase in the growth promoting value of wheat E (Rosenberg and Rohdenburg, 1952; Flodin, 1956; Saviston and Kennedy, 1957). While there is general agreement in the literature that lysine is the most limiting amino acid in wheat and wheat products, the order of the second and third limiting amino acid varies between valine, methio- nine and threonine. According to studies by Sure (1952, 1954) the sequence of most limiting amino acids in milled wheat flour varies as follows; lysine, threonine and valine. Bender (1957, 1958) showed that the first and second limi- ting amino acids in bread were lysine and threonine, but methionine was the third limiting amino acid. Enzymatic Hydrolysis in vitro Enzymatic hydrolysis of the samples by the pepsin- pancreatin-pancreatin system was employed as a biological method for looking at changes occurring to individual amino 92 acids, during the heat processing of the samples. It has been suggested by many researchers (Melnick gt gt., 1946; Mauron gt gt., 1955; Sheffner gt _t., 1956; Akeson and Stahman, 1964) that enzymatic methods 19.11212 give a more realistic estimation of the nutritional value of proteins, because they take into account the possibility of the amino acids to be liberated from the intact protein by the action of the digestive enzymes. Also these methods provide an indication of the site of damage of the protein, during the processing. The total amino acids in the TCA soluble fraction, liberated by the pepsin-pancreatin-pancreatin proteolysis are shown in Table 14. Unexpectedly the enzyme blank gave values of amino acids that are larger than most of the amino acid values in the TCA soluble fraction of the sam- ples making impossible the calculation of amino acids liberated by the enzymatic action of the sample. It is possible that the enzymes during the 27 hour period of incubation and in the absence of substrate suffered exces- sive autolysis, thus giving high blank amino acid values, while the autolysis of the enzymes in the presence of substrate (samples) was very small. Indeed, digestion of the enzyme blank, as estimated by N analysis, showed a 83% digestion of the enzyme mixture. However, if it is assumed that the contribution of the amino acids from the enzymes is the same in all samples, then comparison of the 93 Table 14. Total free amino acids in pepsin-pancreatin digest of the samples. 94 Free Bread mix Whole bread Crumb Crust 22130 Digest1 (%)? Digest1 (%)2 Digest1 (%)2 Digest1 (%)2 Lys 1.83 69 1.16 47.0 1.83 68 .81 38.0 Thr .61 29 .71 31.0 .99 42 1.55 51.0 Cys .09 5 .12 6.0 .16 8.5 .10 5.5 Met 1.18 69 1.09 59.0 1.50 86 1.09 64.0 Val 1.63 42 1.87 45.0 .51 13 1.58 32.0 I1e 2.05 59 1.90 52.0 1.26 37 1.87 47.0 Leu 9.92 161 11.39 158.0 9.68 126 5.35 69.0 Tyr 2.14 78 2.89 93.0 2.18 89 2.98 82.0 Phe 5.49 138 3.11 73.0 4.09 104 3.66 72.0 His 1.54 78 .86 33.0 1.24 58 .91 26.0 Arg 3.25 91 2.56 52.0 3.58 73 2.38 55.0 Asp .87 26 .91 22.0 1.12 30 .99 20.0 Ser .54 15 .86 25.0 .27 7.0 1.02 21.0 Clu 3.38 8 8.23 21.0 5.35 13.0 4.39 15.0 Pro .20 2. .26 3.0 .29 4.7 .27 2.3 Cly .39 11. .55 16.0 .77 21.0 .58 11.0 A1 .98 36 1.37 47.0 2.08 81.0 1.64 48.0 1 Values expressed as g/100 gr protein of the digest zValues expressed as percent of total amino acid as deter- mined by acid hydrolysis. 95 LTB DTB NFDM-flour NFDM-bread Enzyme Digest1 (%)2 Digest](%)2 Digest1 (%)2 Digest1 (%)2 Blank 1.23 50.0 .98 33.0 2.27 59 1.32 44.0 3.26 .46 17.0 .73 30.0 .37 13.0 .76 33.0 1.13 .12 6.0 .09 5.0 .07 3.5 .06 3.0 .13 1.26 70.0 .84 46.0 .96 56.0 .94 54.0 .54 1.46 35.0 1 06 26.0 1.44 33.0 1.54 35.0 1.97 3.46 90.0 1 72 47.0 1.92 55.0 1.79 44.0 .28 10.30 136.0 8.89 139.0 8.39 132.0 9.56 112.0 5.19 3.58 109.0 4.10 139.0 2.81 92.0 2.86 90.0 1.74 3.84 80.0 2.89 64.0 2.97 67.0 3.19 67.0 1.34 .98 46.0 .90 55.0 .82 40.0 .82 38.0 .75 2.73 60.0 2.43 65.0 3.78 79.0 2.65 56.0 .99 .49 11.5 .74 19.0 .62 13.0 .66 14.0 2.54 .64 15.0 .88 24.0 .39 10.6 .93 23.0 1.66 2.66 7.3 3.64 8.8 6.84 18.11 3.45 9.4 4.35 .20 2.4 .20 2.2 .20 2.3 .18 2.3 .27 .39 10.6 .42 11.2 .15 4.5 .38 12.5 .89 1.32 35.0 1.39 50.0 1.96 70.0 1.14 39.0 1.56 96 amino acid profiles of the samples would be possible; in that case the differences between the samples would reflect differences in the amino acids liberated from the samples. The Table 14 also shows the amounts of amino acids released enzymatically as percentages of the total amino acids recovered from the samples by acid hydrolysis. Lysine shows a progressive decrease in the percentage released, as the heat treatment of the products increases from bread mix (68%), to the whole bread (47%), LTB (50%), DTB (33%). This pattern is not seen in any other of the essential amino acids. During the heat treatment the lysine with its e-amino group reacts with reducing sugars to form Maillard type products. In view of the specifi- ties of the proteolytic enzymes it is assumed that the sugar-amino linkages are not split by these enzymes, and the lysine units involved in these linkages are not measured by the amino acid analysis. The total amino acids released by the enzymatic digestion did not exceed the 40% of the nitrogen content of the TCA soluble fraction. They varied from 31.2% for the crust to 40% for the crumb. Acid hydrolysis of the TCA soluble fraction revealed that the rest of the nitrogen in the TCA soluble fraction was in the form of peptides not being analyzed by the amino acid analyzer. As was expected, according to the specificities of the enzymes used, high percentages of arginine, tyrosine, 97 leucine, phenylalanine, lysine and methionine were present in the TCA soluble fraction of all samples. The percent digestion of the samples, based on the Kjeldahl analysis of nitrogen in the TCA soluble fraction was as high as 89.5% for the crumb and 77.3% for the DTB with the rest products ranging between these two values (Table 15). It is interesting to note that the above calculated percentage digestion values are in good agreement with the 1 vivo digestion values showing in Table 9. Destroyed, Inactivated and Available Lysine Table 16 shows the values for total lysine obtained after HCl acid hydrolysis and the values for available lysine obtained by rat growth assay and by FDNB method. The difference between the total and available lysine by rat growth assay gives the inactivated lysine, i.e. the portion of lysine which is unavailable to the organism but susceptible to hydrolysis by HCl acid. The difference in total lysine between the different products and that of the non-heated (bread mix) sample gives the "destroyed" lysine, i.e. the portion of the lysine that is not reco- vered with the HCl acid hydrolysis. The sum of inactivated and destroyed lysine constitutes the "total deterioration" of the lysine in the sample, and the difference between the total deterioration and the total lysine gives the 98 oéw nKm m.: mom méw m.mw Now Tum cowummmww mcmuemm women ewe—m women xPE lzomz nzouz moo mhn pmseu nsweu mponz womem mstcm cwuomeuconucwuomemcoalcwmnma mnu an mm_anm mnu mo Emumxm :o_ummm_w acmuemm .mp mpnoh- 99 .pcmpcou momm»_ esoFelznmz mo memon mnp co wmpopzupou F mw.~ mw.m mm._ om.F mo._ mm.m wF.N Fo.m wonumelmzom xn . mcpmxp mpnopwo>< _m~.o Fand No.0 mm.o 00.? o~.o om.o mm.o mcwmxp wmuo>wuuocm mF.N _N.N m¢.F mm.m wo._ mo.m mm.P w~.N mommo npzoem poe mcwmxp .Fwo>< Fw~.o Fo mm.o mm.o on.o Fo.o Fm.o o mcwmxp wmxoeummo mm.m mm.m m~.N m¢.~ oP.N on.~ m¢.~ o~.~ .Poewzn wwmo nu: mn mcemap Pooch women esope women st -mez -zmmz men men omsem masem oeomz oooem :_mpoem a oop\m .mcwmx— mpnopwo>o wco wmuo>muoocw .oomoeomoo .emmoe .me oenoe 100 available lysine, which is the nutritionally important portion of the lysine. The terms "inactivated" and "des- troyed" lysine come from methodological considerations only. Nutritionally both these portions of lysine being tied-up with their e-amino group in Maillard reaction pro- ducts are not utilized by the rat organism. It was shown recently (Erbersdobler, 1976) that fructoselysine and lactuloselysine, the products formed by heating proteins in the presence of glucose or lactose, if protein bound, are digested poorly. Free fructoselysine is not actively transported out of the intestine but absorbed by diffusion; the adsorbed part is not utilized by the organism and it is excreted rapidly as an intact molecule. Fructoselysine which is not absorbed reaches the lower intestine, where it is presumably utilized by microorganisms living there. Table 17 shows the portions of destroyed, inactivated and available lysine in the samples, as percentages of the total lysine after acid hydrolysis of the unheated samples. It is interesting to note that only 79.2% of the lysine in the raw bread mix (not heated) is available. This value for available lysine is in good agreement with the 83% availability of wheat flour lysine reported by Calhoun _t _l,, 1970. Mild heating resulted in increased lysine availability as shown by the fact that the crumb value for available lysine was 92.2% vs 79.3% for the bread mix. 101 Table 17. Available lysine of samples, percent. Destruction1 $211561 Detlg1gration] Available1 Bread mixa 0 20.74 20.74 79,3 Whole breada 7.77 18.52 25.29 73,7 Crumba 0-37 7.40 7.77 92.2 Crusta 22-22 39.23 61.45 38.5 LTBa 9-25 23.26 32.52 67.5 0133 20 37 24.81 45.18 54.8 NFDM-flourb 0.0 16.6] 16.5] 83,4 NFDM-breadb 8.6 24.30 32.90 57,} 1Calculated as percent of total lysine of bread mix (a) or NFDM-flour (b). 102 The destructiaiof lysine, due to the Maillard reac- tion, can be extended to a considerable portion of the total lysine. Tanaka gt gt. (1977) reported that due to browning, during 10 days of storage of egg albumin with glucose at 15% moisture and 37°C, 52% of the originally available lysine was lost. Correlation Between Total and Available tysine Values The total lysine values for the products analyzed are highly correlated with the available lysine values obtained by either the rat assay method or the FDNB-method, E as shown by the following regression equations and correla- tion coefficients. .< ll 1.239x - 1.233 r = 0.903 where v available lysine by rat growth assay x = total lysine, HCl acid hydrolysis Y = 1.204x - 0.943 r = 0.957 where Y = available lysine by FDNB method x = total lysine, HCL acid hydrolysis This good correlation can be explained, in view of the findings that Maillard reaction products on acid hydrolysis release a portion of the lysine being tied-up with its e-amino group. The mechanism responsible for the des- truction and inactivation of lysine is the same, i.e. the 103 tying-up of the e-amino group during the Maillard reaction. The differences in values between total lysine content and lysine availability arises from the fact that acid hydro- lysis used in total lysine determination regenerates part of the inactivated lysine. Although this hypothesis has not been tested with the samples used in this work, it was shown to be true for the Maillard type products in other materials (Lea and Hannan, 1950; Andrian, 1961; Finot, 1973). However, the proportions of regenerated lysine are not well-known. Lea and Hannan, 1950, found that in their casein-glucose system 70% of lysine combined with glucose could be recovered by acid hydrolysis. Adrian (1961) reported that he did not recover more than 10% of the lysine from heated lysine-glucose mixture after HCL acid hydrolysis. Finot (1973) reported that from the synthetic derivative fructoselysine, which is one of the main Mail- lard type products in heated milk, 44-50% of the lysine of the derivative was recovered after acid hydrolysis. A comparison between the values for available lysine by rat growth assay and PER values, shows that these two sets of values vary almost in parallel. The relationship between these two sets of values is essentially linear (Figure 7) and it is expressed by the regression equation y = 0.825x - 1.094 and r = 0.989 where Y PER and x = available lysine by rat growth. Also 104 o.m mm=_o> «we m> zommo nuzoem woe an momma, mpnopwo>o on Home Acwmuoea a oop\mv mcmmxp mpno—wo>< m.~ o.~ m.~ o.— .m me:m_m 83d 105 the relationship between available lysine by rat growth and NPR is excellent, as it is shown by the regression equation Y = 0.700x - 0.086 and r = 0.987 where Y = NPR and x = available lysine by rat growth. These two high correlations could be used in predicting the PER and NPR of samples of this series from the avail- able lysine values. Comparison Between In Vivo and In Vitro Values for Available Lysine The values for available lysine obtained by the chemi- cal method were in good agreement with the values obtained by the biological method for all samples except for the crust (Table 16). For the crust the chemical method gave a value 41% higher than the value obtained by the biologi- cal method. This high value could probably be explained in view of the fact that during the severe heat treatment of proteins cross linkages are formed many of which still contain FDNB-reactive lysine. The chemical method is based on the assumption that lysine with its c-amino group only reacts with FDNB, while lysine with durable bonds on the e-amino group does not. The similarity of the results between FDNB method and rat growth assay, support the hypothesis that indeed the 6- amino group controls the availability of lysine by 106 transforming this essential amino acid into a form not utilizable by the rat organism. Correlation Between in vitro Available Lysine and PER and N33 The relationships between the values for available lysine by the FDNB method and the PER and NPR values are given by the equations: Y = 0.888x - 1.395 and r = 0.977 where Y = PER and x = available lysine by FDNB method and Y = 0.767x - 0.371 and r = 0.986 where Y = NPR and x = available lysine by the FDNB method. The equations and coefficients of correlation show excellent correlations between available lysine values and PER and NPR values. These excellent relationships, espe- cially with PER, could probably be utilized in predicting the PER values of samples of the series used in this work, from the available lysine values obtained with the simple and easy to perform FDNB method. SUMMARY AND CONCLUSIONS Changes in the biological value, digestibility and nutritionally available lysine were studied in a series of wheat flour products representing different degrees of heat treatment of the wheat flour. The products used were bread mix, whole bread (crumb and crust together), crumb, crust, light toasted bread (LTB), dark toasted bread (DTB), bread mix containing 4% non fat dry milk solids (NFDM-flour) and bread prepared from the last mix (NFDM-bread). The nutritional value of the above products was evaluated by the Protein Efficiency Ratio (PER) and Net Protein Ratio (NPR) methods. The nutritional value of the products decreased, with the exception of crumb, as the heat treat- ment increased, with the PER values for the products (after being adjusted to 2.5 for casein) being: 0.75 for the bread mix, 0.46 for the whole bread, 0.91 for the crumb, —0.23 for the crust, 0.40 for the LTB, 0.16 for the DTB, 1.17 for the NFDM-flour and 0.75 for the NFDM-bread. Compared to the unbaked ingredients (bread mix) baking caused a 38% decrease in the PER value, crumb alone showed a 21% increase in the PER, while the crust displayed a negative PER (-0.23). Light toasting of the whole bread caused a 13% decrease in the PER of the bread, while dark 107 108 toasting caused a 66% decrease. The addition of 4% NFDM- solids to the bread mix caused a 56% increase in the PER of the mix and 63% increase in the PER of the baked bread (whole bread vs NFDM-bread). The apparent digestibility of the protein was measured by in vivo and lg vitro (using a trypsin-chymotrypsin- peptidase enzyme system). The apparent digestibility decreased, in general, with increasing heat treatment, being 88.8% for the NFDM-flour, 88.3% for the crumb, 86.3% for the bread mix, 83.8% for the NFDM-bread, 82.7% for the LTB, 79.9% for the crust and 77.2% for the DTB. The mild heating of the crumb (protected by the crust) increased the apparent digestibility in comparison with that of the unheated ingredients (bread mix). The values for apparent digestibility obtained by the 11 gtttg method were in good agreement with the values obtained by the 1 vivo method. The available lysine of the products was measured by a biological method (rat growth assay) and a chemical method (FDNB-method). The values for available lysine, in 9 available lysine per 100 g protein, obtained by the rat growth assay were: bread mix 2.14, whole bread 1.99, crumb 2.49, crust l.04, LTB l.88, DTB 1.48, NFDM-flour 2.71 and NFDM-bread 2.18. As percent of the total lysine, the available lysine in the bread mix was 79.3%; compared to the total lysine of the unbaked bread mix, the available lysine in the whole bread was 73.7% in the crumb 92.2%, 109 in the crust 38.5%, in the LTB 67.5%, in the DTB 54.8%. Compared to the total lysine in the NFDM-flour the avail- able lysine was 83.4% for the NFDM-flour and 67.1% for the NFDM-bread. Available lysine values obtained by both 1 vivo and 1g vitro methods were in good correlation with the corres- ponding PER values of the products as indicated by the regression equations and correlation coefficients: Y = 0.825x - 1.094 and r = 0.989 where Y = PER value and x = available lysine by rat growth Y = 0.888x - 1.395 and r = 0.977 where Y = PER and x = available lysine by FDNB method. The good agreement of the available lysine values obtained by the FDNB method with the values obtained by the biological method and the excellent correlation with the PER values, indicate that the simple and less time consuming FDNB method could be used for estimating the biologically available lysine and predicting the PER values of wheat flour products such as those used in this work. BIBLIOGRAPHY BIBLIOGRAPHY Akeson, R.W. and Stahman, A.M. 1964. A pepsin pancreatin digest index of protein quality evaluation. J. Nutri- tion 83:257. Akeson, R.W. and Stahmann, A.M. 1965. Nutritive value of leaf protein concentrate, an in vitro digestion study. J. Agr. Food Chem. V01? 13, No. 2, p. 145. Allison, J.B. 1959. Calories and protein nutrition. In: Protein Nutrition. Ann. N.Y. Acad. Sci. 69 (1009). Allison, J.B., Anderson, J.H. and White, J.I. 1949. The nutritive value of dietary proteins. Trans. Am. Assoc. Cereal Chemists 7:24. Andrian, J. 1974. Nutritional and physiological conse- quences of Maillard reaction. World Review of Nutri- tion and Dietetics 19:71. Association of official analytical chemists. 1975. Methods of analysis, AOAC, Washington, D.C. Baldwin, R.R., Lowry, R.J., and Thiessen, R. 1951. Some effects of processing on the nutritive properties of proteins. Food Research 16, 107. Bariga, P.B., Bayllis, E., and Lyman, C. 1959. Determina- tion of free-Lys e—Amino groups in cottonseed meals & preliminary studies on relation to protein quality. Archives of Bioch. & Biophys. 84, 1-6. Barackman, R.A. and Bell, R.N. 4938. Temperatures attained in baking products. Cereal Chemistry 15:841. Bender, A.E. and Miller, 5.0. 1953. A new brief method of estimating net protein value. Bioch. J. 53:vii. Bender, A.E. and B.H. Doell. 1957. Biological evaluation of proteins. A new aspect. Br. J. of Nutrition 11:140. Bender, A.E. 1958. Nutritive value of bread protein forti- fied with amino acids. Proc. Nutr. Soc. 16; viii. Science 127:87. 110 111 Bender, A.E. 1961. National Research Council. National Academy of Science, Publ. 843, 407. Bender, A.E. 1972. Processing damage to protein food. A review. J. Food Technology 7:239. Bjarnason, J. & Carpenter, J.K. 1969. Mechanisms of heat damage in proteins. 1. Models with acylated lysine units. Br. J. Nutr. 23:859. Block, R.J. and Mitchell, H.H. 1946. The correlation of the amino acid composition of proteins with their nutritive value. Nutrition Abstracts & Reviews, Vol. 16, No. 2, p. 249. Block, R.J. and K.W. Weiss. 1956. Amino acid handbook; methods and results of protein analysis. C.C. Thomas, Springfield, Illinois. Blom, L., Hendricks, J. and Caris, J. 1967. Determination of available lysine in foods. Analytical Bioch. 21: 328. Boctor, A.M. and Harper, A.E. 1968. Measurement of avail— able lysine in heated and unheated foodstuffs by chemical and biological methods. Journal of Nutrition 94:289. Bodwell, E.C. 1977. Application of animal data to human protein nutrition:A review. Cereal Chem. 54(4):958. Booth, H.V. 1971. Problems in the determination of FDNB- available lysine. J. Sci. Fd. Agric. 1971. Vol. 22: 658. Bovey, F.A. and Yanaris, S. 1960. Pepsin. In: The Enzymes Vol. 4, Ed. Lardy, H. and Myrbach, K. Academic Press, N.Y. Bressani, R. and Valliente, A.T. 1962. All vegetable pro- tein mixtures for human feeding. J. Food Sci. 27:394. Bricker, M.L. and Mitchell, H.H. 1947. The protein require- ments of the adult rat interns of the protein contained in egg. milk and soy flour. J. Nutrition 34:491. Calhoun, W.K., Hepburn, F.N. and Bradley, W.B. 1960. The availability of lysine in wheat, flour, bread and gluten. Journal of Nutrition 70:337. 112 Carpenter, J.K. and Ellinger, M.G. 1955. The estimation of "available lysine" in protein concentrates. Bio- chemical Journal 61, xi. Carpenter, J.K., Jones, W.L. and Mason, E.L. 1959. Coloured reaction products from (im-DNP) histidine and methoxy- carbonyl chloride. Biochemical Journal 73, ll Carpenter, J.K., Ellinger, G.M., Munro, M.I. and Rolfe, E.J. 1957. Fish products as protein supplements to cereals. British Journal of Nutrition 11:162. Carpenter, J.K. 1973. Damage to lysine in food processing. its measurement and its significance. Nutrition Ab- stracts & Reviews. Vol. 43, No. 6, page 431. - Chavez, J.F. and Pellet, P.L. 1975. Protein quality of some representative Latin American diets by rat bio- assay. J. of Nutrition 106:792. Cheftel, C. 1977. Chemical and nutritional modifications of food proteins due to processing and storage. In: Food Proteins. Ed. Whitaker, J. and S. Tannenbaum. AVI, Westport, Conn. Clark, A.V. and Tannenbaum, S.R. 1974. Isolation and characterization of pigments from protein-carbonyl browning systems. Models for two insulin-glucose pig- ments. J. Agr. Food Chem. 22:1089. Conkenton, E.J. and Frampton, V.L. 1959. Reaction of Gossypol with free e-amino groups of lysine in pro- teins. Arch. Bioch. Bioph. 81:130. Couch, Russell J. 1975. Collaborative study of the deter- mination of available lysine in proteins and feeds. Journal of the AOAC Vol. 58, No. 3, p. 559. DeMuelenaere, H.J.H., Chen, L.M. and Harper, E.A. 1967. Assessment of factors influencing estimation of lysine availability in cereal products. J. Agr. Food Chemis- try 15(2):310. Dimler, R.S. 1975. tg vitro enzymatic digestion of milk proteins. M.S. thesis, Michigan State University. Donoso, 6., Lewis, 0.A., Miller, D.S. and Payne, P.R. 1962. Effect of heat treatment on the nutritive value of proteins: Chemical and balance studies. J. Sci. Food Agr. 13:192. 113 Dryden, J.H., J.G. Kendrick, L.D. Satterlee, L.J. Schroeder and R.S. Block. 1977. Predicting protein digestibility and quality using an enzyme-tetrahymena pyriformis W Bioassay. Journal of Food Biochemistry 1(1) Ellis, R.G. 1969. The Maillard reaction. In: Advances in Carbohydrate Chemistry 14:63. Erbersdobler, F.H. 1977. The biological significance of carbohydrate-lysine crosslinking during heat-treatment of food proteins. In: Protein crosslinking-B, nutri- tional and medical consequences. Ed. by Mendel Friedman. Plenum Publ. Co., N.Y. FAO. 1957. Protein requirements. Nutr. Studies No. 16, FAO/UN. FAO/WHO. 1965. Protein requirements report of joint FAO/WHO expert group. WHO Technical Report Series No. 301. FAQ. 1972. Amino-acid content of foods and biological data on proteins. Rome, Italy. FAO/WHO. 1973. Energy and protein requirements WHO, Geneva: Technical reports series No. 522. Finley, W.J. & M.F. Friedman. 1973. Chemical methods for available lysine. Cereal Chemistry. Vol. 40, 101. Finot, A.P. 1973. Non-enzymatic browning Chapt. 34. In: Proteins in Human Nutrition. Ed. by Porter, J.W.G., Rolls, B.A., Academic Press. London. Fisher, H., Grimminger, P. Leveille, G.A. and Shapiro, R. 1960. Quantitative aspects of lysine deficiency and amino acid imbalance. J. Nutrition 71:213. Flodin, N.W. 1956. The philosophy of amino acid fortifi- cation of foods. Cereal Sci. Today 1:165. Ford, J.E. 1960. A microbiological method for assessing the nutritional value of proteins. Brit. J. Nutrition 14:485. Ford, E.J. 1962. A microbiological method for assessing the nutritional value of proteins. 2. The measurement of ”available" methionine, leucine, isoleucine, argi- nine, histidine, tryptophane and valine. Brit. J. Nutr. 1962, 16, 409. 114 Ford, J.E. and Salter, D.N. 1966. Analysis of enzymati- cally digested food proteins by Sephadex-gel filtration. Brit. J. Nutrition 20:843. Ford, J.E. 1972. Some effects of processing on nutritive value. In: Proteins in human nutrition. Ed. J.W.G. Porter and B.A. Polls. Academic Press, N.Y. Guo, L.S., Summers, J.D. and Moran, E.T. 1971. Assaying feedstuffs for available lysine content using a feather meal basal diet. Canadian Journal of Animal Science 51:161. Gupta, D.J., Dakroury, M.A., Harper, E.A. and Elvehjem, A.C. 1957. Biological availability of lysine. J. Nutrition 64:259. Guthneck, T.B., Bennett, A.B. and Schweigert, S.B. 1953. Utilization of amino acids from foods by the rat. II. Lysine. J. of Nutrition 46, p. 289. Hackler, R.L. 1977. Methods for measuring protein quality: A Review of Bioassay Procedures. Cereal Chemistry 54 (4):984. Handwerck, V., Bugard, E. and Mauron, J. 1966. Helvetica Chim. Acta, 49:419. Hankes, L.V., Riesen, W.H., Henderson, L.H. and Elvehjem, C.A. 1948. Liberation of amino acids from raw and heated casein by acid and enzyme hydrolysis. J. Bio- logical Chem. 176:467. Hannam, R.S. and Lea, C.H. 1952. Studies on the reaction between proteins and reducing sugars in the "dry" state. VI. The reactivity of the terminal amino groups of lysine in model systems. Biochimica and Biophysica Acta 9:293. Harper, A.E. 1959. Amino acid balance and imbalance. 1. Dietary level of protein and amino acid imbalance. J. Nutrition 68:405. Harper, A.E. and de Muelenaere, H.J.H. 1961. Nutrition value of cereal proteins with Special reference to the availability of amino acids. 5th International Bioch. Congress, Moscow, 1961. Harper, E.A. 1974. Basic concepts. In: Improvement of protein nutriture.Committee on amino acids food and nutrition, National Research Council, National Academy of Sciences, Washington, D.C. 115 Hawley, E.E., Mulvin, J.R., Nasset, E.S. and Saymanaki, T.A. 1948. Biological value of six partially puri- fied proteins. J. Nutr. 36:153. Hegarty, P.V.J. 1975. Some biological considerations in the nutritional evaluation of foods. Food Techno- logy 29:52. Hegsted, D.M., R. Neff and J. Worcester. 1968. Determina- tion of the relative nutritive value of proteins. Factors affecting precision and validity. J. Agr. Food Chem. 16:190. Hegsted, D.N. 1977. Protein quality and its determina- tion. p. 347. In: Food Proteins. Ed. by J. Whitaker and S. Tannenbaum, AVI Publ. Co., Westport, Connec- ticut. Hill, R.L. 1965. Hydrolysis of proteins. In: Advances in Protein Chemistry. Ed. C.B. Anfinsen, M.L. Anson, J.T. Edsall and F.M. Richards. Vol. 20, Academic Press, N.Y. Hodge, E. John. 1953. Chemistry of browning reactions in model systems. J. Agricultural and Food Chemistry Vol. 1, No. 15. Holm, H. 1971. Micro-method for determination of non-N- terminal lysine in milk and similar products. J. Sci. Fd. Agr. 22:378. Holt, L.E. Jr. 1967. Amino acid requirements of infants. Current Ther. Res. 9:149. Horn, M.J., C.C. Fitied, A.E. Blum and H.W. Warren. 1958. The distribution of amino acids in wheat and certain wheat products. Cereal Chem. 35:411. Howart, H.W., W.J. Monson, C.D. Bauer and R.J. Block. 1958. The nutritive value of bread flour proteins as affected by practical supplementation with lactalbumin, nonfat dry milk solids, soybean proteins, wheat gluten, and lysine. J. Nutr. 64:151. Hsu, W.H., Vavak, L.D., Satterlee, D.L. et a1. 1977. A multienzyme technique for estimating protein digesti- bility. J. of Food Science, Vol. 42, No. 5. ‘ Hurrell, F.R. and Carpenter, J.K. 1974. Mechanisms of heat damage in proteins. 4. The reactive lysine content of heat-damaged material as measured by different ways. Brit. J. Nutrition 32, 589. 116 Hurt, H.D., R.H. Forsythe, and C.H. Krieger. 1975. Factors which influence the biological evaluation of protein quality by the protein efficiency ratio method. In: Protein Nutritional Quality of Food and Feeds. Part I. Assay Methods Biological, Biochemical and Chemical, Ed. M. Friedman. Marcel Dekker, Inc., New York. Hutchinson, J.B., T. Moran and J. Pace. 1958. Bread and the growth of weanling rat: the lysine threonine balance. Nature 181:1733. Jansen, G.R., S.R. Ehle and N.L. House. 1964. Studies of the nutritive loss of supplemented lysine in baking. 1. Loss in a standard white bread containing 4% non-fat dry milk. Food Techn. 18:367. Kakede, L.M. and Irvin E. Liener. 1969. Determination of available lysine in proteins. Analytical Biochemistry 27:273. Kon, S.K. and Markuze, Z. 1931. The biological values of the proteins of breads baked from rye and wheat flours alone or combined with yeast or soya bean flour. Bioch. J. 25:1476. Kopple, J.D. and Surendseid, M.E. 1975. Evidence that histidine is an EAA in normal and chronically uremic man. J. Clin. Invest. 55:881. Kuiken, A.K. and Lyman, M.G. 1948. Availability of amino acids in some foods. J. of Nutrition 36:359. Lea, H.C. and Hannan, S.P. 1949. Studies of the reaction between proteins and reducing sugars in the "dry" state. I. Effect of activity of water, pH and temperature on the primary reaction between casein and glucose. Bio- chemica and Biophysica Acta 3:313. Lea, H.C. and Hannan, S.P. 1950. Studies of the reaction between proteins and reducing sugars in the "dry" state. 11. Further observations on the formation of the casein- glucose complex. Bioch. and Biophy. Acta 4:518. Lea, H.C. and Hannan, S.P. 1950. Studies of the reaction between proteins and reducing sugars in the "dry" state. 111. Nature of the protein groups reacting. Bioch. and Biophy. Acta 5:433. Lewis, V.M. and Lea, C.H. 1950. A note on the relative rates of reaction of several reducing sugars and sugar derivatives with casein. Bioch. and Bioph. Acta 4:532. 117 Lewis, 0.A.M. 1966. Short ion-exchange column method for the estimation of cystine and methionine. Nature 209: 1239. Mauron, J., Mottu, F., Bujard, E. and Egli, R. 1955. The availability of lysine, meth., & Tryptophan in con- densed milk & milk powder. In vitro digestion studies. Archives of Bioch. & Bioph. 59:433. Mauron, J. 1970. Nutritional evaluation of proteins by enzymatic methods. In: Evaluation of novel protein products; Ed. A.E. Bender, R. Kihlberg, B. Lofquist and L. Munch. Pergamon Press, N.Y. Mauron, J. 1972. Influence of industrial and household handling on food protein quality. In: Protein and amino acid functions. Ed. E.J. Bigwood, Pergamon Press, Oxford. McLaughlan, J.M., Rogers, G.C., Chapman, 6.0. and Campbell, A.J. 1959. Canadian J. of Physiology 37:1293. McLaughlan, J.M. 1972. Effects of protein quality and quantity of protein utilization. In: Newer Methods of Nutritional Biochemistry. Vol. 5:33, Academic Press, New York. McLaughlan,J.M.and Keith, N.0. 1975. Bioassays for protein quality. In: Protein Nutritional Quality of foods and feeds. Part I. Assay Methods - biological, biochemical and chemical. Ed. by M. Friedman, Marcel-Dekker, Inc. N.Y. Meisten A. 1965. The role of amino acids in nutrition. In: Biochemistry of the amino acids. Vol. 1. pp. 201-230. Academic Press, New York. Melnick, D., Oser, L.B. and Weiss, S. 1946. Rate of enzymic digestion of proteins as a factor in nutrition. Science p. 326, Vol. 103. Menden, E. and H.D. Cremer. 1970. Laboratory methods for the evaluation of changes in protein quality. In: Newer Methods in Nutritional biochemistry 4:123. Academic Press, New York. Miller, E.L., Carpenter, K.J. and Milner, C.K. 1965A. Availability of amino acids in protein foods. Chemical and nutritional changes in heated cod muscle. Br. J. Nutr. 19:547. Miller, E.L., Hartley, A.W. and Thomas, D.C. 1965C. Availa- bility of sulfur amino acids in protein foods. Effect of heat treatment upon the total amino acid content of cod muscle. Br. J. Nutrition 19:565. 118 Mitchell, H.H. 1924. A method of determining the biological value of proteins. J. Biological Chemistry 58:873. Mitchell, H.H. and Carman, G.G. 1926. The biological value of the nitrogen of mixtures of patent white flour and animal foods. J. Biol. Chem. 68:183. Mitchell, H.H. and R.J. Block. 1946. Some relationships between the amino acid contents of proteins and their nutritive values for the rat. J. Biolog. Chem. 163: 599. Mitchell, H.H. 1954. In "Symposium on Methods for the evaluation of nutritional adequacy and status" eds. Spector H., Peterson, M.S. and T.E. Friedmann, p. 13, National Research Council, Washington, D.C. Moore, 5., D.H. Spackman and W.H. Stein. 1958. Chromatogra- phy of amino acids on sulfonated polystyrene resin. Anal. Chem. 30:1185. Morgan, A.F. 1931. The effect of heat upon the biological value of cereal proteins and casein. J. Biol. Chem. 90: 771. Morrison, A.B. and J.A. Campbell. 1960. Evaluation of protein in foods. V. Factors influencing the protein efficiency ratio in foods. J. of Nutr. 70:112. Muramer, S.M. and A.E. Harper. 1959. Amino acid balance and imbalance. 11. Dietary level of protein and lysine requirement. J. Nutr. 69:58. Nasset, E.S. 1957. Role of the digestive track in the utilization of proteins and amino acids. J. Amer. Med. Assoc. 164:172. Netke, S.P., Scott, M.H. and Allee, L.G. 1969. Effect of excess amino acids on the utilization of the first limiting amino acid in chick diets. J. Nutrition 99:75. Netke, S.P. and Scott, H.M. 1970. Estimates on the availa- bility of amino acids in soybean oil meal as determined by chick growth assay. Methodology as applied to lysine. Journal of Nutrition 100:281. Olley, J. and Watson, H. 1961. The "available lysine" content of fish meals. Journal of the Science of Food and Agriculture 12:316. Osborne, 1.8. and Mendel, L.B. 1914. Amino acids in nutri- tion and growth. J. Biol. Chem. 17:325. 119 Osborn, T.B., Mendel, L.B. and E.L. Ferry. 1919. A method of expressing numerically the growth promoting value of proteins. J. Biol. Chem. 37:223. Oser, B.L. 1951. Method for integrating essential amino acid content in the nutritional evaluation of proteins. J. Am. Dietet. Assoc. 27:396. Osner, R.C. and Johnson, R.M. 1974. Nutritional and chemi- cal changes in heated casein. I. Preliminary study of solubility, gel filtration pattern and amino acid pattern. J. Food Technology 9:301. Pilcher, L.H. and Williams, H.H. 1954. Microbiological evaluation of protein quality. 11. Studies of the responses of tetrahymena pyridiformis W to intact pro- teins. J. Nutrition 53:589. Rao, S.R., Carter, F.L. and Frampton, V.L. 1963. Determi- nation of available lysine in oilseed meal proteins. Analytical Chemistry 35:1927. Rao, N.M., Sreenivas, H. Swaminathan, M., Carpenter, J.K., and Morgan, C.B. 1963. The nutritionally available lysine and methionine of heated casein-glucose mixtures. J. Sci. Food Agr. 14:544. Rao, N.M. and McLaughlan, J.M. 1967. Lysine and methionine availability in heated casein-glucose mixtures. J. of the Association of Official Analytical Chemists 50:704. Reynolds, M.T. 1963. Chemistry of non-enzymatic browning. I. The reaction between aldoses and amines. In: Advances in Food Research 12:1. Reynolds, M.T. 1965. Chemistry of non-enzymatic browning. In: Advances in Food Research 14:168. Roach, G.A., S. Anderson and Williams. 1967. Comparison of methods for the determination of available lysine value in animal and vegetable protein sources. J. Sci. Fd. Agric., Vol. 18, 1967. Rogers, G.R., Chen, D.M.Y. and Harper, A.E. 1970. Importance of dispensable amino acids for maximal growth in the rat. Proc. Soc. Exp. Biol. Med. 13:517. Rose, W.C., Oesterling, M.J. and Womack, M. 1948. Compara- tive growth on diets containing ten and nineteen amino acids with further observations upon the role of glutamic and aspartic acid. J. Biol. Chem. 176:753. 120 Rose, W.C. 1949. Amino acid requirements of man. Fed. Proc. 8:546. Rose, W.C. and Nixon, R.L. 1955. The amino acid require- ments of man. XIV. The sparing effect of tyrosine on the phenylalanine requirement. J. Biol. Chem. 217:95. Rose, W.C. 1957. The amino acid requirements of adult man. Nutr. Abstr. Rev. 27:631. Rosen, 0.6. and Fernell, R.W. 1956. Microbiological evalu- ation of protein quality with Igtrahymena gyriformis. 2. Relative nutritive value of proteins and foodstuffs. Brit. J. Nutr. 10:156. Rosenberg, H-R. and Rohdenburg, L.E. 1952. The fortifica- tion of bread with lysine. II. The nutritional value of fortified bread. Arch. Bioch. & Biophy. 37:461. Rosenberg, H.R, E.L. Rohdenburg and J.T. Baldini. 1954. The fortification of bread with lysine. III. Supple- mentation with essential amino acids. Arch. Bioch. Biophys. 49:263. Ruderus, H. and Kihlberg, R. 1970. Determination of availa- ble lysine by column chromatography. In: Evaluation of novel protein products. Eds: Bender, A.E., B. Lofquist, R. Kihlberg, L. Munch. Pergamon Press, Oxford, N.Y. Rupp, J.R., Neimann, C. and Hein, G.E. 1966. The primary specificity of chymotrypsin. Further evidence for "wrong-way" binding. Biochemistry 5:4100. Sanger, F. 1945. The free amino groups of insulin. Bio- ch. J. 39:507. Satterlee, D.L., Kendrick, 6.0. and Miller, A.C. 1977. Rapid ifl vitro assays for estimating protein quality. Food Technology 31(7):69. Saviston, R.A., Kennedy, M.B. 1957. Effect of baking of the nutritive value of proteins in wheat bread with and without supplements of non-fat dry milk and lysine. Cereal Chemistry 34:94. Schram, E., Moore, S. and Bigwood, R.J. 1954. Chromato- graphic determination of cystine as cysteic acid. Bioch. J. 57:33. Schwartz, H.M. and Lea, C.H. 1952. The reaction between proteins and reducing sugars in the "dry" state. Rela- tive reactivity of the a- and c-amino groups of insulin. Biochemical Journal 50:713. 121 Schweigert, S.B. and Guthneck, T.B. 1953. Utilization of amino acids from foods by the rat. 1. Methods of testing for lysine. J. of Nutrition 46:277. Sheffner, A.L., Eckfeldt, G.A. and Spector, H. 1956. The pepsin—digest-residue (PDR) amino acid index of the net protein utilization. J. of Nutrition 60:105. Spies, J.R. and Chambers, D.C. 1948. Chemical determina- tion of tryptophan. Analytical Chem. 20:30. Spies, J.R. 1967. Determination of tryptophan in proteins. Anal. Chemistry 39:1412. Stahmann, M.A. and Woldegiorgis, G. 1975. Enzymatic methods for protein quality determination. In: Protein nutri- tional quality of foods and feeds. Part I. Assay methods - Biological, Biochemical and Chemical, Ed. M. Friedman. Marcel Dekker, New York. Steinke, F.H. 1977. Protein efficiency ratio pitfalls and causes of variability: A Review. Cereal Chemistry 54 (4) 949-957. Sternberg, M., C.Y. Kim and F.J. Shwende. 1975. Lysino- alanine determination in proteins. J. of Food Science 40:1168. Sternberg, M., C.Y. Kim and F.J. Schwende. 1975. Lysino- alanine: Presence in Foods and Food Ingredients. Science 190_992, 1975. Stout, L.E. and Drosten, F. 1973. Heat flow through bakery products. II. Time temperature relationship existing during the baking of bread. Ind. Eng. Chem. 25:428. Swaminathan, M. 1967. Availability of plant proteins. Chapter 4. In: Newer Methods of Nutritional Biochemis- try. Vol. 111. Ed. by Anthony A. Albanese. Academic Press, N.Y., London. Swenseid, M.D., Harris, G.L. and Tuttle, S.G. 1960. The effect of sources of non-essential nitrogen on N balance in young adults. J. Nutr. 71:105. Tanaka, M., Kimiagar, M., Tung-Ching, Lee and C.L. Chichester. 1977. Effect of Maillard reaction on nutritional quality of protein. In: Protein crosslinking-B. Nutritional and Medical Consequences. Ed. by Mendel Friedman, Plenum Publishing Co. 122 Thomas, M.C. 1972. A chemical method for the determina- tion of available lysine. Ph.D. Thesis, Texas A & M University Tsen, C.C. and Reddy, K.R.P. 1977. Effect of toasting on the nutritive value of bread. J. Food Sci. 42(5):1370. Tsien, W.S., Johnson, E.L. and Liener, I.E. 1957. The availability of lysine from torula yeast. Archives of Biochemistry and Biophysics 71:414. Valle-Riesta, J. and Barnes, R.H. 1970. Digestion of heat-damaged egg albumen by the rat. Journal of Nutri- tion 100:873. Waibel, P.E. and Carpenter, K.J. 1972. Mechanisms of heat damage in proteins. 3. Studies with e-(y-L—glutamyl)- L-lysine. Br. J. Nutrition 27:509. Williams, H.H., Harper, A.E., Hegsted, M.D., Arroyave, G. and Holt, E.L., Jr. 1974. Nitrogen and amino acid requirements. In: Improvement of protein nutriture. Committee on Amino Acids, Food and Nutrition Board, National Research Council, National Academy of Sciences, Washington, D.C.