- Email: [email protected]
A Chick Assay for Availability of Lysine in Wheat
A Chick Assay for Availability of Lysine in Wheat N. A. CAVE1 and C. J. WILLIAMS 2 Research Branch, Agriculture Canada, Ottawa, Ontario, Canada K1A 0C6 (Received for publication April 2, 1979)
1980 Poultry Science 59:799-804 INTRODUCTION
Amino acids represent a considerable proportion of the total cost of a poultry diet. Although 40 to 80% of the amino acids may be provided by cereals, there is little information in the literature on the bioavailabilities of amino acids of cereal grains. As long as there is uncertainty about the available amino acids provided by feedstuffs, practical diets must be formulated to allow considerable margins above actual amino acid requirements. To reduce the size of the margins and achieve optimum utilization of amino acids and protein, knowledge of availability values is essential. A number of growth assays for available lysine have been reported (Hill et al, 1966; Guo et al, 1971; Netke and Scott, 1970). The assays, however, were developed for protein concentrates rather than for cereal feedstuffs. Assays for cereal feedstuffs differ from those for concentrates in that a high level of dietary inclusion of the test material is necessary so that differences in amino acid levels are sufficient to provide marked growth responses. Furthermore, when a single feedstuff may provide up to 80% of dietary amino acids, the ability to detect small differences in the availabilities of those amino acids assumes greater importance than for minor dietary ingredients.
1
Animal Research Institute, Contribution No. 837. Statistical Research Section, Engineering and Statistical Research Institute, Contribution No. 1-80. 2
Jerne and Wood (1949) discussed the requirements for a valid assay. The response of subjects should relate in a consistent manner to increasing input of lysine over the entire range of input used in the assay. The assay diets should be formulated so as to avoid nonspecific growth responses due to addition of feed components other than lysine. This restriction includes both positive responses due to adding limiting nutrients and negative responses due to excessive amounts of a detrimental factor. Uwaegbute and Lewis (1966), Robel and Frobish (1977), and Baker (1978) indicated that excess levels of amino acids may be among factors of the latter type. This paper describes an assay method used in this laboratory for the evaluation of the availability of lysine to chicks together with an assessment of the validity and precision of the method. The availability of lysine is reported for two cultivars of wheat. MATERIALS AND METHODS
The basal assay diet (Table 1) provided adequate amounts of vitamins, minerals (NRC, 1971), and amino acids (Sasse and Baker, 1973). Intact protein and nonessential amino acids in the diet were provided by feather meal and gelatin. L-lysine hydrochloride was added to the basal diet at levels of 0, 1.25, 2.50, 3.75, and 5.00 g/kg to form five standard diets with supplemental lysine of 0, 1.0, 2.0, 3.0, and 4.0 g/kg diet. In addition, six test diets were produced by incorporation into the basal diet
799
Downloaded from http://ps.oxfordjournals.org/ at Uniwersytet Warszawski Biblioteka Uniwersytecka on April 8, 2015
ABSTRACT A chick growth assay was developed to estimate the availability of lysine in cereal grains. Amino acids were supplied in the basal assay diet by hydrolyzed-feather meal, gelatin, and crystalline amino acids. Dietary nitrogen levels were equalized for all assay diets by adjustment of glutamic acid content, and dietary energy levels were equalized by adjustment of the proportions of corn starch, corn oil, and cellulose. The assay method was tested for validity with two wheat cultivars, using dietary lysine level or lysine intake as the independent variable and weight gain or carcass nitrogen gain as the dependent variable. Taking lysine level as the independent variable did not satisfy statistical tests for validity of the assay procedure. Availability of lysine determined by the selected method of regression of weight gain on lysine intake was .79 (with 95% fiducial limits of .94 and .66) and .76 (.87 and .66) for Manitou and Pitic cultivars, respectively.
800
CAVE AND WILLIAMS TABLE 1. Composition of basal diet (g/kg)
Feather meal Gelatin L-arginine-HCl L-histidine-HCl • H, O L-tryptophan L-phenylalanine DL-methionine L-threonine L-leucine L-isoleucine L-glutamic acid Corn starch Corn oil Cellulose Mineral premix a Vitamin premixt> Ethoxyquin
86.4 26.8 2.06 2.79 1.78 1.82 4.49 1.70 2.67 1.88 124.0 549.78 25.0 105.0 53.7 10.0 .13
Mineral premix nutrients per kilogram diet: CaC0 3 , 3 g; Ca3 (P0 4 ) 2 , 28 g; K2 HPo4 , 9 g; MgS0 4 • 7 H 2 0 , 3.5 g; Fe ( C 6 H , 0 , ) 2 • 6 H^O, .5g;ZnCl 2 , .2 g; NaCl, 8.8 g; KI, 40 mg; CuSO, • 5 H 2 0 , 20 mg; H3 B 0 3 , 9 mg; CoS0 4 • 7 H , 0 , 1 mg; MnS0 4 • H2 O, 650 mg; Na2MoO„ • 2 H2 O,9 mg. Vitamin premix nutrients per kilogram diet: thiamine • HC1, 100 mg; nicotinic acid, 100 mg; riboflavin, 16 mg; Ca-pantothenate, 20 mg; pyridoxine • HC1, 6 mg; folic acid, 4 mg; p-aminobenzoic acid, 2 mg ; biotin, .6 mg; cyanocobalamin, .02 mg; menaquinone, 5 mg; a-tocopheryl acetate, 20 mg; inositol, 100 mg; ascorbic acid, 250 mg; cholecalciferol, 600 ICU; retinyl acetate 10,000 IU.
of Manitou wheat, at levels of 190, 380, and 570 g/kg and Pitic wheat at levels of 217, 433, and 650 g/kg of final diet. The incorporation of different levels of L-lysine hydrochloride and wheat was made at the expense of corn starch, corn oil, and glutamic acid in the basal diet. All assay diets were made isocaloric (3.27 kcal/g) by adjusting the quantities of corn starch, corn oil, or cellulose and were made isonitrogenous (3.52 g N/kg) by adjusting the quantity of glutamic acid. The diets contained not less than 10 g/kg of corn oil and 10 g/kg of cellulose. The two samples of wheat assayed represent cultivars of high (Manitou) and low (Pitic) nitrogen (N) content. Their determined metabolizable energy values and chemical composition are described in Table 2. Day-old male chicks of a commercial meattype strain (Shaver Starbro) were housed in electrically heated, wire floored, tier brooders
TABLE 2. Metabolizable energy, chemical composition, and 1000-kemel weight of Manitou and Pitic •wheat3-'0
Lysine (g/kg) Nitrogen (g/kg) Metabolizable energy (kcal/g) Ether extract (g/kg) Starch + sugar (g/kg) Crude fiber (g/kg) Ash (g/kg) Dry matter (g/kg) 1,000-kernel weight (g)
Manitou
Pitic
4.8 32.0 3.52 11.6 663.7 31.2 18.8 896 33.2
4.3 22.2 3.37 13.9 733.4 28.8 19.2 878 31.6
All data except dry matter and 1,000-kernel weight are expressed on a dry matter basis. All data except lysine are taken from Sibbald and Price (1976).
Downloaded from http://ps.oxfordjournals.org/ at Uniwersytet Warszawski Biblioteka Uniwersytecka on April 8, 2015
Ingredient
(14 chicks per pen) and given free access to water and chick starter diet (Table 3). At 8 days of age chicks were weighed individually and divided into groups with a 5 g range in body weight. Chicks of extreme body weights were discarded and the remainder were assigned to four replicate groups according to body weight. Each replicate group was divided among 12 pens of 10 chicks, equalizing pen weights within replicates. One pen of chicks from each replicate was killed immediately, frozen, and stored for subsequent analysis of total carcass nitrogen. The carcass N was used in computing nitrogen gain during the feeding period. Each diet was fed ad libitum to one pen of 8-day-old chicks in each replicate for 7 days. Pen weights were obtained at 8 and 15 days of age and feed intake determined for the period 8 to 15 days of age. At 15 days of age chicks were killed by carbon dioxide inhalation, frozen, and ground preparatory to carcass analysis. Samples of ground carcass from 8- and 15-day-old chicks were freeze-dried and analyzed for water and Kjeldahl nitrogen by AOAC (1970) methods. Grain and diet samples were hydrolyzed with 6N HC1 (sample weight: acid weight = 1:200) for 22 hr at 120 C (Cremer and Menden, 1970) in the presence of mercaptoethanol (Keutmann and Potts, 1969). After hydrolysis samples were filtered, evaporated to dryness, and dissolved in .2N sodium citrate buffer (pH 2.2) for estimation of amino acid contents by ion exchange chromatography
LYSINE AVAILABILITY IN WHEAT TABLE 3. Composition of chick starter diet (g/kg)
Corn Soybean meal (49% protein) Corn oil Ground limestone Dical. phosphate Salt, iodized Vitamin premix a Mineral premix^ DL-methionine
543.25 390 30 13.5 15 2.5 3.0 2.0 .75 1000.00
Vitamin premix ingredients per kilogram diet: vitamin A, 9000 IU; vitamin D 3 , 1650 ICU; riboflavin, 4.72 mg; Ca-pantothenate, 6.6 mg; nicotinic acid, 24.2 mg; pteroyl monoglutamic acid, .55 mg; cyanocobalamin, .009 mg. Mineral premix nutrients per kilogram diet: manganese, 60 mg; zinc, 50 mg; iron, 10 mg; copper, 5 mg.
(Beckman 121M, Beckman Instruments, Palo Alto, CA). The availability of lysine in each of the two wheat samples was estimated by fitting the multiple regression model: Y ij = b 0 + r i + b s X s + b m X r n + b p X p + e ij [1] where Yjj is the observed response (gain in weight or carcass nitrogen); X s , X m , and X p are the lysine intakes or lysine levels (g/kg final diet) from the standard rations, Manitou wheat and Pitic wheat, respectively; b s , b m and b p are the corresponding regression coefficients; r; is the effect of the ith replicate (2r; = 0); b 0 is the intercept; and ejj is the random error assumed =N(0,ff 2 ). This regression model was tested for blank, curvature, and validity by the methods described by Campbell (1966). The availabilities of lysine in Manitou and Pitic wheats were estimated from the ratios b m / b s and b p / b s , respectively. Fiducial limits of these estimates were calculated by Fieller's theorem (Finney, 1964). RESULTS AND DISCUSSION Mean weight gains obtained by chicks fed
the standard diets ranged from 1.7 to 8.2 g/chick/day over the period 8 to 15 days of age (Table 4). These gains may be compared with weight gains of 11.4 and 11.6 g/chick/day obtained in a preliminary experiment by chicks fed the basal ration (Table 1) supplemented with 7.5 g L-lysine hydrochloride/kg and a 22% protein control ration (Table 3), respectively. It is evident that dietary lysine supported growth throughout the range of concentrations used in this assay but even at the highest level tested it did not support maximum growth. When dietary lysine level was taken as the independent variable the results of the statistical tests indicated significant (P<.05) fundamental invalidity. The regression coefficients obtained from this analysis were not used to estimate lysine availability. When lysine intake was the independent variable, neither of the response variables showed significant departures from the model (Table 5). More particularly, the tests for blank and curvature indicated that the responses to the diets were not significantly nonlinear over the entire range of lysine levels used and that the response was in the region of constant efficiency of lysine utilization for gain. Nonsignificance of the test for validity indicated that the basal ration had been formulated in such a way (by the provision of essential nutrients and of minimum levels of plant oil and fiber) that the test diets behaved as simple dilutions of the standard for lysine. The upper level of lysine intake was 1.25 times higher for the standard diets than for the Manitou wheat. Since Manitou wheat was of unusually high lysine content (4.8 g/kg DM) the range of dietary lysine contents provided by the set of standard diets used here should be sufficient for assaying most if not all cereal grains. The lysine availabilities of the two wheats were estimated from the regression coefficients (Table 6) obtained by fitting the data to model [1] using lysine intake as the independent variable. The lysine availability of Manitou wheat, estimated from weight gain, was .79 with 95% fiducial limits of .94 and .66; the corresponding values for Pitic wheat were .76, and .87 and .66. Estimates of lysine availabilities from carcass N were .43 with 95% fiducial limits of .75 and .13 for Manitou wheat and .58 with fiducial limits of .82 and .36 for Pitic wheat. Thus the estimates of lysine availability using carcass N as a measure of response were smaller and the fiducial interval
Downloaded from http://ps.oxfordjournals.org/ at Uniwersytet Warszawski Biblioteka Uniwersytecka on April 8, 2015
Ingredient
801
CAVE AND WILLIAMS
802
TABLE 4. Lysine intake and average gains in body weight and carcass N of chicks in the assay of wheat cultivars Gain
Lysine supplement Lysine intake Source
Level (g/kg)
(g/chick/day)
(mgN/chick/day)
41.6 49.9 76.1 105.6 127.2
1.7 2.1 4.4 6.5 8.2
.02 .04 .06 .16 .20
Standard diets L-lysine • HCI L-lysine • HCl L-lysine • HCl L-lysine • HCl L-lysine • HCl Test diets Manitou wheat Manitou wheat Manitou wheat
190 380 570
58.2 76.0 92.9
2.4 3.8 5.2
.04 .05 .09
Pitic wheat Pitic wheat Pitic wheat
217 433 650
58.0 80.3 103.8
3.0 4.1 5.5
.04 .07 .11
0 1.25 2.50 3.75 5.00
was much wider than the values obtained using weight gain. The large amount of variability associated with the estimation of carcass N gain makes this measure of response unsuitable for the estimation of available lysine. Weight gain appears to be a satisfactory response to use for estimating lysine availability by the present method. Nonetheless, the widths of the fiducial intervals were quite large (.28 for Manitou and .21 for Pitic wheat). One reason for the size of these intervals can be seen in the narrow range of dietary lysine levels between the unsupplemented basal diet and the test diets with highest levels of wheat (Table 4). Because of the low contents of energy, protein, and lysine in Pitic wheat, it was not possible to include a greater quantity of Pitic-wheat lysine in the test diet while ensuring provision of other essential nutri-
ents. The widths of the fiducial intervals can be reduced by increasing the number of observations. For example, if the 11-point design used in this study had been replicated 8 times instead of 4, then (assuming the same results in the additional 4 replicates as observed in 4 replicates used) the width of the fiducial intervals would have been .19 and .15 for Manitou and Pitic wheats, respectively. The precision of the estimates of lysine availability can also be improved by reducing the number of intermediate levels of lysine or wheat. For example, if only two lysine levels (basal and basal plus 5 g/kg) were used for the standard and only one level of wheat (the highest) for the test rations, giving a 4-point design with 11 replicates, then the width of the fiducial intervals are expected to be .21 and .16 for Manitou and Pitic wheat, respectively. Doubling
TABLE 5. Analyses of variance for the assay of two cultivars of wheat using weight gain or carcass N gain as dependent variable and lysine intake as independent variable Mean squares
Source of variation
Degrees of freedom
Weight gain
Carcass N gain
Regression Blank Nonvalidity Curvature Replication Residual
3 1 4 2 3 30
254,571** 115 1,418 851 262 575
194.28** 2.10 1.63 2.60 1.58 2.55
* 'Statistically significant (P<.001).
Downloaded from http://ps.oxfordjournals.org/ at Uniwersytet Warszawski Biblioteka Uniwersytecka on April 8, 2015
(mg/day)
LYSINE AVAILABILITY IN WHEAT
803
TABLE 6. Regression analysis of weight gain and carcass nitrogen gain on lysine intake Weight gain Source of supplementary lysine L-lysine • HCl Manitou wheat Pitic wheat
Carcass N gain
ba
Relative availability^
ba
Relative availability 0
.0760 ± .0023 .0604 + .0055 .0579 ± .0042
1.0 .79 ± .14 .76 ± .11
.0021 + .00014 .0009 ±.00033 .0012 ± .00025
1.0 .43 ± .31 .58 ± .23
Lysine availability relative t o L-lysine • HCl ± approximate fiducial interval half-widths.
the size of this experiment (4 points and 22 replicates) would yield fiducial interval widths of .15 and .11 for Manitou and Pitic wheat, respectively. It is noted that in a 4-point design it is not possible to test the validity of the model used. In this study it was necessary to test for validity and for the appropriateness of the basal diet; in future assays of similar feedstuffs, however, it would be possible to change the numbers of assay points and replicates in this way, to increase assay precision without increasing the size of experiment. Lysine availability in wheat has been estimated previously from digestibility determinations. The availabilities estimated in this assay are higher than the .68 lysine availability in wheat determined with chicks of 10 to 40 days old by fecal analysis technique (Kristen et al, 1966) but lower than .81 estimated with colostomized hens (McNab and Shannon, 1974); however these differences are within the limits of chance variation. The lysine availabilities obtained by the weight gain method in the present study are also lower than estimates of wheat protein digestibility (Hainan, 1928). It is characteristic of lysine of feedstuffs, however, that digestibility or absorbability is lower than that of most other amino acids (Poppe and Meier, 1971; Sarwar and Bowland, 1975).
ACKNOWLEDGMENTS The authors wish to thank S. Cadieux for his able technical assistance. Analysis of amino acids was conducted by the Chemical and Biological Research Institute, Agriculture Canada.
REFERENCES Association of Official Analytical Chemists, 1970. Official methods of analysis. 11th ed. AOAC, Washington, DC. Baker, D. H., 1978. Nutrient bioavailability of feedstuffs: methodology for determining amino acid and B-vitamin availability in cereal grains and soybean meal. Pages 1-12 in Proc. 1978 Georgia Nutr. Conf. for Feed Ind. Campbell, R. C , 1966. The chick assay of lysine. Biometrics 2 2 : 5 8 - 7 3 . Cremer, H. D., and E. Menden, 1970. Laboratory methods for the evaluation of changes in protein quality. Pages 123-161 in Newer methods for the evaluation of changes in protein quality. A. A. Albanese, ed. Academic Press, New York. Finney, D. J., 1964. Statistical method in biological assay. 2nd ed. Charles Griffin and Co. Ltd., London. Guo, L. S., J. D. Summers, and E. T. Moran, 1971. Assaying feedstuffs for available lysine content using a feather meal basal diet. Can. J. Anim. Sci. 51:161-168. Hainan, E. T., 1928. Digestibility trials with poultry. II. Digestibility of weak and strong wheats and their value for poultry feeding. J. Agr. Sci. 18:421-431. Hill, D. C , J. Singh, and G. S. Ashton, 1966. A chick bioassay for lysine. Poultry Sci. 45:554—560. Jerne, N. K., and E. C. Wood, 1949. The validity and meaning of the results of biological assays. Biometrics 5:273-299. Keutmann, H. T., and J. T. Potts, 1969. Improved recovery of methionine after acid hydrolysis usingmercaptoethanol. Anal. Bioch. 29:175—185. Kristen, H., S. Poppe, and W. Wiesemuller, 1966. Untersuchungen zum Eiweisstoffwechsel bei kiiken. 1. Lysinbilanzen. Archiv. Geflugelzucht Kleintierkunde 15:317-323. McNab, J. M., and D.W.F. Shannon, 1974. The nutritive value of barley, maize, oats and wheat for poultry. Brit. Poultry Sci. 15:561-567. National Research Council, 1971. Nutrient requirements of poultry. Natl. Acad. Sci., Washington, DC. Netke, S. P., and H. M. Scott, 1970. Estimates on the availability of amino acids in soybean meal as determined by chick growth assay, methodology as applied to lysine. J . Nutr. 100:281-288.
Downloaded from http://ps.oxfordjournals.org/ at Uniwersytet Warszawski Biblioteka Uniwersytecka on April 8, 2015
Regression coefficient (g gain/mg lysine intake) ± standard error.
804
CAVE AND WILLIAMS
Poppe, S., and H. Meier, 1971. Untersuchungen iiber die Aminosaurenresorption aus verschiedenen Proteintragern beirn Gefliigel. Arch. Tierernahrung 21:515-529. Robel, E. J., and L. T. Frobish, 1977. Evaluation of the chick bioassay for estimating sulfur amino acid, lysine and trytophan availability in soybean meal. Poultry Sci. 56:1399-1404. Sarwar, G., and J. P. Bowland, 1975. Availability of amino acids in wheat cultivars used in diets for weanling rats. Can. J. Anim. Sci. 55:579—586.
Sasse, C. E., and D. H. Baker, 1973. Modification of the Illinois reference standard amino acid mixture. Poultry Sci. 52:1970-1972. Sibbald, I. R., and K. Price, 1976. Relationships between metabolizable energy values for poultry and some physical and chemical data describing Canadian wheats, oats and barleys. Can. J. Anim. Sci. 56:255-268. Uwaegbute, H. O., and D. Lewis, 1966. Chick bioassay of lysine. 1. Development of the assay procedure. Brit. Poultry Sci. 7:249-257. Downloaded from http://ps.oxfordjournals.org/ at Uniwersytet Warszawski Biblioteka Uniwersytecka on April 8, 2015
1980 Poultry Science 59:799-804 INTRODUCTION
Amino acids represent a considerable proportion of the total cost of a poultry diet. Although 40 to 80% of the amino acids may be provided by cereals, there is little information in the literature on the bioavailabilities of amino acids of cereal grains. As long as there is uncertainty about the available amino acids provided by feedstuffs, practical diets must be formulated to allow considerable margins above actual amino acid requirements. To reduce the size of the margins and achieve optimum utilization of amino acids and protein, knowledge of availability values is essential. A number of growth assays for available lysine have been reported (Hill et al, 1966; Guo et al, 1971; Netke and Scott, 1970). The assays, however, were developed for protein concentrates rather than for cereal feedstuffs. Assays for cereal feedstuffs differ from those for concentrates in that a high level of dietary inclusion of the test material is necessary so that differences in amino acid levels are sufficient to provide marked growth responses. Furthermore, when a single feedstuff may provide up to 80% of dietary amino acids, the ability to detect small differences in the availabilities of those amino acids assumes greater importance than for minor dietary ingredients.
1
Animal Research Institute, Contribution No. 837. Statistical Research Section, Engineering and Statistical Research Institute, Contribution No. 1-80. 2
Jerne and Wood (1949) discussed the requirements for a valid assay. The response of subjects should relate in a consistent manner to increasing input of lysine over the entire range of input used in the assay. The assay diets should be formulated so as to avoid nonspecific growth responses due to addition of feed components other than lysine. This restriction includes both positive responses due to adding limiting nutrients and negative responses due to excessive amounts of a detrimental factor. Uwaegbute and Lewis (1966), Robel and Frobish (1977), and Baker (1978) indicated that excess levels of amino acids may be among factors of the latter type. This paper describes an assay method used in this laboratory for the evaluation of the availability of lysine to chicks together with an assessment of the validity and precision of the method. The availability of lysine is reported for two cultivars of wheat. MATERIALS AND METHODS
The basal assay diet (Table 1) provided adequate amounts of vitamins, minerals (NRC, 1971), and amino acids (Sasse and Baker, 1973). Intact protein and nonessential amino acids in the diet were provided by feather meal and gelatin. L-lysine hydrochloride was added to the basal diet at levels of 0, 1.25, 2.50, 3.75, and 5.00 g/kg to form five standard diets with supplemental lysine of 0, 1.0, 2.0, 3.0, and 4.0 g/kg diet. In addition, six test diets were produced by incorporation into the basal diet
799
Downloaded from http://ps.oxfordjournals.org/ at Uniwersytet Warszawski Biblioteka Uniwersytecka on April 8, 2015
ABSTRACT A chick growth assay was developed to estimate the availability of lysine in cereal grains. Amino acids were supplied in the basal assay diet by hydrolyzed-feather meal, gelatin, and crystalline amino acids. Dietary nitrogen levels were equalized for all assay diets by adjustment of glutamic acid content, and dietary energy levels were equalized by adjustment of the proportions of corn starch, corn oil, and cellulose. The assay method was tested for validity with two wheat cultivars, using dietary lysine level or lysine intake as the independent variable and weight gain or carcass nitrogen gain as the dependent variable. Taking lysine level as the independent variable did not satisfy statistical tests for validity of the assay procedure. Availability of lysine determined by the selected method of regression of weight gain on lysine intake was .79 (with 95% fiducial limits of .94 and .66) and .76 (.87 and .66) for Manitou and Pitic cultivars, respectively.
800
CAVE AND WILLIAMS TABLE 1. Composition of basal diet (g/kg)
Feather meal Gelatin L-arginine-HCl L-histidine-HCl • H, O L-tryptophan L-phenylalanine DL-methionine L-threonine L-leucine L-isoleucine L-glutamic acid Corn starch Corn oil Cellulose Mineral premix a Vitamin premixt> Ethoxyquin
86.4 26.8 2.06 2.79 1.78 1.82 4.49 1.70 2.67 1.88 124.0 549.78 25.0 105.0 53.7 10.0 .13
Mineral premix nutrients per kilogram diet: CaC0 3 , 3 g; Ca3 (P0 4 ) 2 , 28 g; K2 HPo4 , 9 g; MgS0 4 • 7 H 2 0 , 3.5 g; Fe ( C 6 H , 0 , ) 2 • 6 H^O, .5g;ZnCl 2 , .2 g; NaCl, 8.8 g; KI, 40 mg; CuSO, • 5 H 2 0 , 20 mg; H3 B 0 3 , 9 mg; CoS0 4 • 7 H , 0 , 1 mg; MnS0 4 • H2 O, 650 mg; Na2MoO„ • 2 H2 O,9 mg. Vitamin premix nutrients per kilogram diet: thiamine • HC1, 100 mg; nicotinic acid, 100 mg; riboflavin, 16 mg; Ca-pantothenate, 20 mg; pyridoxine • HC1, 6 mg; folic acid, 4 mg; p-aminobenzoic acid, 2 mg ; biotin, .6 mg; cyanocobalamin, .02 mg; menaquinone, 5 mg; a-tocopheryl acetate, 20 mg; inositol, 100 mg; ascorbic acid, 250 mg; cholecalciferol, 600 ICU; retinyl acetate 10,000 IU.
of Manitou wheat, at levels of 190, 380, and 570 g/kg and Pitic wheat at levels of 217, 433, and 650 g/kg of final diet. The incorporation of different levels of L-lysine hydrochloride and wheat was made at the expense of corn starch, corn oil, and glutamic acid in the basal diet. All assay diets were made isocaloric (3.27 kcal/g) by adjusting the quantities of corn starch, corn oil, or cellulose and were made isonitrogenous (3.52 g N/kg) by adjusting the quantity of glutamic acid. The diets contained not less than 10 g/kg of corn oil and 10 g/kg of cellulose. The two samples of wheat assayed represent cultivars of high (Manitou) and low (Pitic) nitrogen (N) content. Their determined metabolizable energy values and chemical composition are described in Table 2. Day-old male chicks of a commercial meattype strain (Shaver Starbro) were housed in electrically heated, wire floored, tier brooders
TABLE 2. Metabolizable energy, chemical composition, and 1000-kemel weight of Manitou and Pitic •wheat3-'0
Lysine (g/kg) Nitrogen (g/kg) Metabolizable energy (kcal/g) Ether extract (g/kg) Starch + sugar (g/kg) Crude fiber (g/kg) Ash (g/kg) Dry matter (g/kg) 1,000-kernel weight (g)
Manitou
Pitic
4.8 32.0 3.52 11.6 663.7 31.2 18.8 896 33.2
4.3 22.2 3.37 13.9 733.4 28.8 19.2 878 31.6
All data except dry matter and 1,000-kernel weight are expressed on a dry matter basis. All data except lysine are taken from Sibbald and Price (1976).
Downloaded from http://ps.oxfordjournals.org/ at Uniwersytet Warszawski Biblioteka Uniwersytecka on April 8, 2015
Ingredient
(14 chicks per pen) and given free access to water and chick starter diet (Table 3). At 8 days of age chicks were weighed individually and divided into groups with a 5 g range in body weight. Chicks of extreme body weights were discarded and the remainder were assigned to four replicate groups according to body weight. Each replicate group was divided among 12 pens of 10 chicks, equalizing pen weights within replicates. One pen of chicks from each replicate was killed immediately, frozen, and stored for subsequent analysis of total carcass nitrogen. The carcass N was used in computing nitrogen gain during the feeding period. Each diet was fed ad libitum to one pen of 8-day-old chicks in each replicate for 7 days. Pen weights were obtained at 8 and 15 days of age and feed intake determined for the period 8 to 15 days of age. At 15 days of age chicks were killed by carbon dioxide inhalation, frozen, and ground preparatory to carcass analysis. Samples of ground carcass from 8- and 15-day-old chicks were freeze-dried and analyzed for water and Kjeldahl nitrogen by AOAC (1970) methods. Grain and diet samples were hydrolyzed with 6N HC1 (sample weight: acid weight = 1:200) for 22 hr at 120 C (Cremer and Menden, 1970) in the presence of mercaptoethanol (Keutmann and Potts, 1969). After hydrolysis samples were filtered, evaporated to dryness, and dissolved in .2N sodium citrate buffer (pH 2.2) for estimation of amino acid contents by ion exchange chromatography
LYSINE AVAILABILITY IN WHEAT TABLE 3. Composition of chick starter diet (g/kg)
Corn Soybean meal (49% protein) Corn oil Ground limestone Dical. phosphate Salt, iodized Vitamin premix a Mineral premix^ DL-methionine
543.25 390 30 13.5 15 2.5 3.0 2.0 .75 1000.00
Vitamin premix ingredients per kilogram diet: vitamin A, 9000 IU; vitamin D 3 , 1650 ICU; riboflavin, 4.72 mg; Ca-pantothenate, 6.6 mg; nicotinic acid, 24.2 mg; pteroyl monoglutamic acid, .55 mg; cyanocobalamin, .009 mg. Mineral premix nutrients per kilogram diet: manganese, 60 mg; zinc, 50 mg; iron, 10 mg; copper, 5 mg.
(Beckman 121M, Beckman Instruments, Palo Alto, CA). The availability of lysine in each of the two wheat samples was estimated by fitting the multiple regression model: Y ij = b 0 + r i + b s X s + b m X r n + b p X p + e ij [1] where Yjj is the observed response (gain in weight or carcass nitrogen); X s , X m , and X p are the lysine intakes or lysine levels (g/kg final diet) from the standard rations, Manitou wheat and Pitic wheat, respectively; b s , b m and b p are the corresponding regression coefficients; r; is the effect of the ith replicate (2r; = 0); b 0 is the intercept; and ejj is the random error assumed =N(0,ff 2 ). This regression model was tested for blank, curvature, and validity by the methods described by Campbell (1966). The availabilities of lysine in Manitou and Pitic wheats were estimated from the ratios b m / b s and b p / b s , respectively. Fiducial limits of these estimates were calculated by Fieller's theorem (Finney, 1964). RESULTS AND DISCUSSION Mean weight gains obtained by chicks fed
the standard diets ranged from 1.7 to 8.2 g/chick/day over the period 8 to 15 days of age (Table 4). These gains may be compared with weight gains of 11.4 and 11.6 g/chick/day obtained in a preliminary experiment by chicks fed the basal ration (Table 1) supplemented with 7.5 g L-lysine hydrochloride/kg and a 22% protein control ration (Table 3), respectively. It is evident that dietary lysine supported growth throughout the range of concentrations used in this assay but even at the highest level tested it did not support maximum growth. When dietary lysine level was taken as the independent variable the results of the statistical tests indicated significant (P<.05) fundamental invalidity. The regression coefficients obtained from this analysis were not used to estimate lysine availability. When lysine intake was the independent variable, neither of the response variables showed significant departures from the model (Table 5). More particularly, the tests for blank and curvature indicated that the responses to the diets were not significantly nonlinear over the entire range of lysine levels used and that the response was in the region of constant efficiency of lysine utilization for gain. Nonsignificance of the test for validity indicated that the basal ration had been formulated in such a way (by the provision of essential nutrients and of minimum levels of plant oil and fiber) that the test diets behaved as simple dilutions of the standard for lysine. The upper level of lysine intake was 1.25 times higher for the standard diets than for the Manitou wheat. Since Manitou wheat was of unusually high lysine content (4.8 g/kg DM) the range of dietary lysine contents provided by the set of standard diets used here should be sufficient for assaying most if not all cereal grains. The lysine availabilities of the two wheats were estimated from the regression coefficients (Table 6) obtained by fitting the data to model [1] using lysine intake as the independent variable. The lysine availability of Manitou wheat, estimated from weight gain, was .79 with 95% fiducial limits of .94 and .66; the corresponding values for Pitic wheat were .76, and .87 and .66. Estimates of lysine availabilities from carcass N were .43 with 95% fiducial limits of .75 and .13 for Manitou wheat and .58 with fiducial limits of .82 and .36 for Pitic wheat. Thus the estimates of lysine availability using carcass N as a measure of response were smaller and the fiducial interval
Downloaded from http://ps.oxfordjournals.org/ at Uniwersytet Warszawski Biblioteka Uniwersytecka on April 8, 2015
Ingredient
801
CAVE AND WILLIAMS
802
TABLE 4. Lysine intake and average gains in body weight and carcass N of chicks in the assay of wheat cultivars Gain
Lysine supplement Lysine intake Source
Level (g/kg)
(g/chick/day)
(mgN/chick/day)
41.6 49.9 76.1 105.6 127.2
1.7 2.1 4.4 6.5 8.2
.02 .04 .06 .16 .20
Standard diets L-lysine • HCI L-lysine • HCl L-lysine • HCl L-lysine • HCl L-lysine • HCl Test diets Manitou wheat Manitou wheat Manitou wheat
190 380 570
58.2 76.0 92.9
2.4 3.8 5.2
.04 .05 .09
Pitic wheat Pitic wheat Pitic wheat
217 433 650
58.0 80.3 103.8
3.0 4.1 5.5
.04 .07 .11
0 1.25 2.50 3.75 5.00
was much wider than the values obtained using weight gain. The large amount of variability associated with the estimation of carcass N gain makes this measure of response unsuitable for the estimation of available lysine. Weight gain appears to be a satisfactory response to use for estimating lysine availability by the present method. Nonetheless, the widths of the fiducial intervals were quite large (.28 for Manitou and .21 for Pitic wheat). One reason for the size of these intervals can be seen in the narrow range of dietary lysine levels between the unsupplemented basal diet and the test diets with highest levels of wheat (Table 4). Because of the low contents of energy, protein, and lysine in Pitic wheat, it was not possible to include a greater quantity of Pitic-wheat lysine in the test diet while ensuring provision of other essential nutri-
ents. The widths of the fiducial intervals can be reduced by increasing the number of observations. For example, if the 11-point design used in this study had been replicated 8 times instead of 4, then (assuming the same results in the additional 4 replicates as observed in 4 replicates used) the width of the fiducial intervals would have been .19 and .15 for Manitou and Pitic wheats, respectively. The precision of the estimates of lysine availability can also be improved by reducing the number of intermediate levels of lysine or wheat. For example, if only two lysine levels (basal and basal plus 5 g/kg) were used for the standard and only one level of wheat (the highest) for the test rations, giving a 4-point design with 11 replicates, then the width of the fiducial intervals are expected to be .21 and .16 for Manitou and Pitic wheat, respectively. Doubling
TABLE 5. Analyses of variance for the assay of two cultivars of wheat using weight gain or carcass N gain as dependent variable and lysine intake as independent variable Mean squares
Source of variation
Degrees of freedom
Weight gain
Carcass N gain
Regression Blank Nonvalidity Curvature Replication Residual
3 1 4 2 3 30
254,571** 115 1,418 851 262 575
194.28** 2.10 1.63 2.60 1.58 2.55
* 'Statistically significant (P<.001).
Downloaded from http://ps.oxfordjournals.org/ at Uniwersytet Warszawski Biblioteka Uniwersytecka on April 8, 2015
(mg/day)
LYSINE AVAILABILITY IN WHEAT
803
TABLE 6. Regression analysis of weight gain and carcass nitrogen gain on lysine intake Weight gain Source of supplementary lysine L-lysine • HCl Manitou wheat Pitic wheat
Carcass N gain
ba
Relative availability^
ba
Relative availability 0
.0760 ± .0023 .0604 + .0055 .0579 ± .0042
1.0 .79 ± .14 .76 ± .11
.0021 + .00014 .0009 ±.00033 .0012 ± .00025
1.0 .43 ± .31 .58 ± .23
Lysine availability relative t o L-lysine • HCl ± approximate fiducial interval half-widths.
the size of this experiment (4 points and 22 replicates) would yield fiducial interval widths of .15 and .11 for Manitou and Pitic wheat, respectively. It is noted that in a 4-point design it is not possible to test the validity of the model used. In this study it was necessary to test for validity and for the appropriateness of the basal diet; in future assays of similar feedstuffs, however, it would be possible to change the numbers of assay points and replicates in this way, to increase assay precision without increasing the size of experiment. Lysine availability in wheat has been estimated previously from digestibility determinations. The availabilities estimated in this assay are higher than the .68 lysine availability in wheat determined with chicks of 10 to 40 days old by fecal analysis technique (Kristen et al, 1966) but lower than .81 estimated with colostomized hens (McNab and Shannon, 1974); however these differences are within the limits of chance variation. The lysine availabilities obtained by the weight gain method in the present study are also lower than estimates of wheat protein digestibility (Hainan, 1928). It is characteristic of lysine of feedstuffs, however, that digestibility or absorbability is lower than that of most other amino acids (Poppe and Meier, 1971; Sarwar and Bowland, 1975).
ACKNOWLEDGMENTS The authors wish to thank S. Cadieux for his able technical assistance. Analysis of amino acids was conducted by the Chemical and Biological Research Institute, Agriculture Canada.
REFERENCES Association of Official Analytical Chemists, 1970. Official methods of analysis. 11th ed. AOAC, Washington, DC. Baker, D. H., 1978. Nutrient bioavailability of feedstuffs: methodology for determining amino acid and B-vitamin availability in cereal grains and soybean meal. Pages 1-12 in Proc. 1978 Georgia Nutr. Conf. for Feed Ind. Campbell, R. C , 1966. The chick assay of lysine. Biometrics 2 2 : 5 8 - 7 3 . Cremer, H. D., and E. Menden, 1970. Laboratory methods for the evaluation of changes in protein quality. Pages 123-161 in Newer methods for the evaluation of changes in protein quality. A. A. Albanese, ed. Academic Press, New York. Finney, D. J., 1964. Statistical method in biological assay. 2nd ed. Charles Griffin and Co. Ltd., London. Guo, L. S., J. D. Summers, and E. T. Moran, 1971. Assaying feedstuffs for available lysine content using a feather meal basal diet. Can. J. Anim. Sci. 51:161-168. Hainan, E. T., 1928. Digestibility trials with poultry. II. Digestibility of weak and strong wheats and their value for poultry feeding. J. Agr. Sci. 18:421-431. Hill, D. C , J. Singh, and G. S. Ashton, 1966. A chick bioassay for lysine. Poultry Sci. 45:554—560. Jerne, N. K., and E. C. Wood, 1949. The validity and meaning of the results of biological assays. Biometrics 5:273-299. Keutmann, H. T., and J. T. Potts, 1969. Improved recovery of methionine after acid hydrolysis usingmercaptoethanol. Anal. Bioch. 29:175—185. Kristen, H., S. Poppe, and W. Wiesemuller, 1966. Untersuchungen zum Eiweisstoffwechsel bei kiiken. 1. Lysinbilanzen. Archiv. Geflugelzucht Kleintierkunde 15:317-323. McNab, J. M., and D.W.F. Shannon, 1974. The nutritive value of barley, maize, oats and wheat for poultry. Brit. Poultry Sci. 15:561-567. National Research Council, 1971. Nutrient requirements of poultry. Natl. Acad. Sci., Washington, DC. Netke, S. P., and H. M. Scott, 1970. Estimates on the availability of amino acids in soybean meal as determined by chick growth assay, methodology as applied to lysine. J . Nutr. 100:281-288.
Downloaded from http://ps.oxfordjournals.org/ at Uniwersytet Warszawski Biblioteka Uniwersytecka on April 8, 2015
Regression coefficient (g gain/mg lysine intake) ± standard error.
804
CAVE AND WILLIAMS
Poppe, S., and H. Meier, 1971. Untersuchungen iiber die Aminosaurenresorption aus verschiedenen Proteintragern beirn Gefliigel. Arch. Tierernahrung 21:515-529. Robel, E. J., and L. T. Frobish, 1977. Evaluation of the chick bioassay for estimating sulfur amino acid, lysine and trytophan availability in soybean meal. Poultry Sci. 56:1399-1404. Sarwar, G., and J. P. Bowland, 1975. Availability of amino acids in wheat cultivars used in diets for weanling rats. Can. J. Anim. Sci. 55:579—586.
Sasse, C. E., and D. H. Baker, 1973. Modification of the Illinois reference standard amino acid mixture. Poultry Sci. 52:1970-1972. Sibbald, I. R., and K. Price, 1976. Relationships between metabolizable energy values for poultry and some physical and chemical data describing Canadian wheats, oats and barleys. Can. J. Anim. Sci. 56:255-268. Uwaegbute, H. O., and D. Lewis, 1966. Chick bioassay of lysine. 1. Development of the assay procedure. Brit. Poultry Sci. 7:249-257. Downloaded from http://ps.oxfordjournals.org/ at Uniwersytet Warszawski Biblioteka Uniwersytecka on April 8, 2015