MSU LIBRARIES .—_ RETURNING MATERIALS: P1ace in book drop to remove this checkout from your record. FINES wil] be charged if book is returned after the date stamped be1ow. THE DEVELOPMENT OF A COLORIMETRIC ASSAY TO MEASURE RUMEN DEGRADABILITY OF PROTEIN FEEDS Lg VITRO By Timothy Xavier Terry A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Animal Science 1985 AESTEACT ThE LEVELoPEEET OF A COLORIRETRIC SSAY To MEASUEE EUEEE DEGRAEAEILITY or PROTEIN FEELS Lg VITRo By Timothy Xavier Terry Four protein feed supplements were used in developing an in vitro assay for rumen degradability. Degradability was determined by a synthetic bag procedure using three fistulated bovines. The in zitgg assay was a protease digestion followed by a colorimetric assay for solubilized protein. The fire assays tested were: Lowry, Bio-Rad, A260/A280, ninhydrin, and biuret. Changes were made in the digestion in an attempt to make the digestion and colorimetry compatable. Accuracy of colorimetric methods was determined by correlation of these results with residual nitrogen and in zizg data. The Lowry was too sensitive and suffered from inter- fering substances while Bio-Rad was insensitive to the solubilized peptides. A260/A280 tended to underestimate degradability of some feeds and overestimate others. The ninhydrin overestimated degradabilities or color develop- ment was erratic. Biuret was promising, but while correla- tions with soluble hjeldahl nitrogen within feeds was high (r=.97) correlations considering all feeds were low (r=.53). TABLE OF CONTENTS Page LIST OF TABLES iii LIST OF FIGURES iv INTRODUCTION 1 MATERIALS AND METHODS 18 RESULTS AND DISCUSSION 22 CONCLUSION 45 APPENDIX I - Procedure for Determining Proteolytic Enzyme Degradation of Protein Feeds 46 APPENDIX II - Enymatic Assay of Ficin 48 APPENDIX III - Statistical Analyses 50 APPENDIX IV - 38% Dairy - N.P. - 337 Guaranteed Analysis 53 BIBLIOGRAPHY 54 ii TABLE TABLE TABLE TABLE TABLE TABLE 3. 4. 5. LIST OF TABLES Percent Degradability by Nylon Bag Degradability Values: Nylon Bag vs. Ficin Assay Degradation by Purified Ficin Enzyme Percent Degradation by Ninhydrin Absorbance Before and After Centrifugation Mean Absorbances at Two Working Volumes iii Page 22 24 35 37 4O 41 FIGURE FIGURE FIGURE FIGURE FIGURE FIGURE FIGURE LIST OF FIGURES 1a. Percent N Degradability of SBM by Nylon Bag 1b. Percent N Degradability of CM by Nylon Bag 1c. Percent N Degradability of 38%-D by Nylon Bag 1d. Percent N Degradability of LOM by Nylon Bag 2. Funnel Assembly 3. Absorbance Maxima of Biuret 4. N Determination of Solute: Biuret vs. Kjeldahl iv Page 25 26 27 28 3O 39 43 INTRODUCTION The economic impact of the dairy cow or any ruminant for that matter, has mandated the continued investigation into its nutritional requirements. These requirements are profoundly influenced by the presence of the rumen. Two protein requirements must be considered in ruminant feeding: the needs of the animal itself and the requirements of the microbial population in the forestomach (44). Some species of microorganisms found in the rumen require preformed peptides or free amino acids and if not provided adequately in the diet some microorganisms may disappear from the rumen. There may be a reduction in overall efficiency and protein yield (g protein per Kg fermented organic matter) if the ammonia concentration falls below 80 mg nitrogen per liter of rumen fluid (39). This critical level may be even higher when fermentation rate is rapid. Fermentation in the rumen results in the degradation of some of the dietary protein to volatile fatty acids and ammonia, a process which in some circumstances is wasteful when the overall economy of the animal is considered (1). van't Klooster and Boekholt (46) have noted that in sheep on a low protein diet there was an apparrent net gain of nitrogen as digesta flowed to the lower gastrointestinal tract. This increase was presumably of endogenous origin. When these same sheep were fed high protein diets, nitrogen was wasted via absorption of ammonia from the rumen after the deamination of amino acids. In fact Wright (53) was only able to recover 70% of the feedstuff nitrogen in forms use- ful to the host animal. Early work by Burroughs, et al. (5) and Ganev, et al. (13) suggested that amino acids were released, during deg- radation, in proportions similar to those present in the feedstuff. Consequently, the quality of the undegraded pro- tein was similar to that appearing in the feedbunk. In more recent studies, however, Craig and Broderick (8), and Santos et al. (38), showed that this is not the case. Because all proteins showed characteristic variations in amino acid release, the amino acid pattern of the undegraded protein escaping the rumen differed from that originally fed. Santos (38) has demonstrated that in some instances the residual amino acid pattern may even be improved as the essential amino acid: nonessential amino acid ratios of the duodenal digesta increased relative to the diet. The main nitrogenous components in the digesta passing to the small intestine are cat-amino nitrogen, non-al-amino nitrogen, nucleic acid nitrogen, amide nitrogen, and vari- able proportions of ammonia nitrogen (46). According to Smith and McAllan (41) about 50% of the total protein passing from the stomach to the intestines is of microbial origin, and half of the non-protein nitrogen is in the form of free amino acids (48). It must be recognized that the extent of microbial protein synthesis is also a function of the amount of energy released during fermentation, but even when energy sources are not limiting in the rumen, fermentation and microbial growth may not be sufficient to meet the animal's amino acid requirements for maximal productivity. In such cases the animal requires additional supplementation of protein or amino acids that will by-pass the rumen (1S) and be further digested and absorbed in the intestines (38). For years, feed values and protein requirements have been stated in terms of digestible crude protein (DCP). This regime has been discounted to a certain degree by the studies of Miller (27) who noted that diets of equal DCP values gave different responses in growth. Unfortunately, the DCP concept has persisted, and it has led to the belief that the animal could obtain all of its amino acid requirement from micro- bial protein. Recent research, however, indicate that when amino acid requirements of ruminants are high, insufficient protein is available from microbes. While the rumen micro- organisms are capable of providing sufficient protein for maintenance, slow growth, and early pregnancy they cannot fulfill the needs of fast growth, late pregnancy, or early lactation (15). The contemporary trend is to feed some ruminally unde- gradable or "by-pass" protein as is commonly known. Unfor- tunately, too great a supplementation of undegradable protein may result in a protein deficient microbial fermen- tation, and ultimately decreased rumen degradative effi- ciency. The proper levels of by-pass protein vary according to the level of productive demands of the animal. For example, Sniffen and Chase (55) issued the following recom- mendation for dairy cattle: mid lactation 35-41 total dietary late lactation 34-39 protein early lactation 39-49 3 % by-pass of The quantitative determination of these components, degradable and undegradable protein, and how they vary with different feedstuffs and feeding regimens is central to recent proposals for evaluating the protein contribution of feedstuffs and calculating dietary protein requirements (23). Various methods, both in 33333 and £2.1l12: have been devel- oped to estimate rumen protein degradabilities. Unfortunately, most of these procedures are either complex, laborious, or both, and they vary in degrees of accuracy. Perhaps the most widely used in zigg method is that of Mehrez and Orskov (24) which used bags of dacron parachute material to suspend feed samples in the rumen of sheep. These bags were removed at various times, depending on the type of feed used, and the residual nitrogen determined. The extent of protein degradation then, under normal condi- tions, is assumed to be when 90% of the digestible dry matter has disappeared from the bag (24,32). Like all procedures this one was not without problems. Mehrez and Orskov (24) noted significant variation between sheep and between days of incubation. In the process of accounting for the variation they found that increasing duration of incubation failed to substantially reduce variability, and Stern and Satter have concurred (43). In our lab we have seen an effect of feeding relative to the time of insertion of the bags in that there is an increased rate of degradation shortly after feeding. In the sheep studies of Mehrez and Orskov (24), the animals were fed ad lib and so the feeding effect was not studied. Mehrez and Orskov noted that neither prewashing of the bags to remove soluble feed nor location in the rumen during digestion had any effect on variability. Dry matter entering the bags or microbial attachment made no significant difference as the average increase for blank 5cm by 8cm bags was only .O3g dry matter and only trace amounts of diaminopimelic acid (DAPA) was found on undigested residues. Lindberg, et al. (19) found that pore size seems to have no effect as they saw similar degradabilities of feed samples suspended in bags with pore sizes of either 10um or 40um. A critical consideration in this procedure is the preparation of the samples. Preferable, these samples should be masticated digesta collected from an esophageal canula, but a hammer- mill equipped with a 5mm recutter screen for dry feeds or a mincer for wet feeds will work quite satisfactorily (24). Zinn, et al. (54) cited other possible sources of variation in this procedure including: 1) processing of sample (particle size); 2) lag time between insertion and initiation of normal fermentation; 3) formation of an abnormal microenvironment within the bag as a result of non-physiological moisture content, inadequate removal of end-products, preponderance of feed constituents that poten- tially inhibit digestion, and associative effects of feeds; and 4) lack of microbial adaptation to test feedstuff or roughage level. Weakley (49) noted a decreased rate of nitrogen disappearance with animals fed a high grain diet. Zinn (54) also made some suggestions regarding increas- ing the precision of prediction estimates including: 1) greater constancy of animal, dietary, and analytical factors; and 2) use of a standardized reference feed (i.e.- SBM). In spite of the aforementioned difficulties, the dacron bag technique has the advantage of giving a very rapid, accurate estimate of the digestion of nutrients in a feedstuff (24); provided one has access to surgically prepared animals, a suitable microbial marker, and the time and labor to obtain the necessary measurements. Another class of procedures have been developed to determine feed protein flowing to the duodenum by the difference of total nitrogen flow versus microbial and endogenous nitrogen contributions. All of these methods rely on the identification of microbial nitrogen by the organic or natural labels such as aminoethylene phophonic acid (AEPA), Diaminopimelic acid, and Ribonucleic acid, 358, 15N or ruminally infused isotopic markers such as , or 32P (23); all of which would be incorporated into microbial 7 cells during their growth. The ratio of label to microbial nitrogen is determined in isolates and the ratio can then be used to determine microbial contribution to duodenal flow (23). This value along with the proportion of endog- enous nitrogen is subtracted from total non-ammonia nitrogen in the duodenum and the remainder is assumed to be the contribution of feed protein. Then this value relative to the protein level of the diet yields the percent undegrad- able protein. These procedures have inherent problems. First, like the dacron bags, they requires surgically prepared animals. Secondly, action of pepsin and HCl in the abomasum may bring about the disruption of the cells resulting in a microbial isolate with a disproportionate level of marker (41). Others have circumvented this second problem, especially with 3SS, by measuring other parameters such as the specific activity of cystine or methionine. Unfortunately, this requires extra steps to determine the amino acid bound 35S content of the microbial mass and the whole digesta (23). McAllan and Smith (29) used RNA and DAPA, and found that the two methods showed similar patterns but gave different absolute results. More- over, without an absolute standard it is hard to determine which is more accurate. Thirdly, accuracy of measuring total flow of nitrogen and microbial protein entering the abomasum or small intestine is limited, the measuring techniques are laborious, and, as a consequence, abomasal and duodenal flow measurements are often restricted to a period of 24 hours. Reducing the variation due to non-steady state conditions in the rumen is possible through frequent feed- ing, but this makes the results less meaningful for prac- tical conditions. Therefore, estimates of the proportion of dietary protein escaping microbial degradation in the forestomach are subject to high error (44). Lastly, accuracy may be improved by increasing the number of experimental animals, but according to Miller (28) this procedure would require 10 to 12 animals in order to achieve a level of significance less than 5%. Moreover, certain obvious diffi- culties would arise in terms of handling and disposal if the experimental models chosen were large ruminants such as beef or dairy cattle. Because of the complications seen in xizg many in zitrg methods have been developed which avoid or reduce the prob- lems of time, labor, and animals. These procedures are generally based on either solubility in various solvents (9,10,17,35,43,47,51), ammonia gas production after incuba- tion with viable rumen liquor (21,25,36), or residual nitrogen after incubation with bacterial, fungal, or plant proteases (22,30,34,35,45). The degradation of feed protein appears to be the result of 1) a rapid degradation of readily soluble protein occurring soon after the feed enters the rumen and 2) a slower break down of the less soluble protein extending beyond this initial period (9). or the many factors that affect degradability in the rumen solubility is probably one of the most important (47). Many solubility/insolubility procedures, and as many solvents, have been developed through the years (9,10,17, 35,43,44,47,51). The goal of these techniques is the predic- tion of absorbable or digestible feed protein that escapes rumen degradation and enters the small intestine. There are, however, very few highly predictive results available in the literature (47). Of these few, Poos-Floyd, et al (35) found significant correlations between solubility and in_zizg degradability (growth trials) when the feed sample was solubilized in hot water, 10% Burrough's solution, and bicarbonate-phosphate (BP) buffer. As for the poorly corre- lated results from other studies several factors may account for the descrepancies. These include: solvent type, pH, and ionic strength; extraction time and temperature; degree of agitation; feed patricle size; and different types and properties of soluble protein fractions in the feed (9,10, 17,35,43,44,47,51). Therefore, even though solubility may be an important determinant of rumen degradability the two characteristics should not be considered synonymous (9,17, 44). The solubility of proteins need not be related to susceptibility to enzymatic degradation (22). Thus, an ip gitgg procedure which could duplicate, or at least simulate, the enzymatic proteolysis of the rumen would be of great value. The most obvious way to duplicate rumen proteolysis would be to use rumen fluid or washed bacterial suspensions. 10 This procedure would provide the most meaningful results because, like in the rumen, the system is complex since the bacterial population is mixed, individual organisms are at different stages of growth, and many enzymes are succes- sively involved in the breakdown of proteins to VFA's and ammonia (1). The washing of bacteria, however, may cause a decrease or even loss of activity of some enzymes (1,21). Menke, et al (25) and Raab and coworkers (36) have used the known relationship between the fermentation of carbohy- drates and protein synthesis to determine protein degrad- ability and metabolizable energy. Menke (25) used 150 ml syringes containing 30 ml of a rumen fluid/buffer mix plus 100-200 mg of feed sample, and placed in a rotor located in a drying oven held at 391.500. The syringes were incu- bated for 24 hours and the final reading of gas production was taken at this time. After correcting for the blanks they noted a high correlation (r2=.95) between this in zitgg and in gigg degradability. More recently, Raab, et al. (36) saw similar correlations between gas production and ammonia nitrogen content of the fluid (r=.98), but every feedstuff had its own regression equation. Realizing that they could make use of this high correlation Raab, et al. (36) devel- oped an equation to determine ig_zitgg degradable nitrogen (IVDN). IVDN: (NH3 -N at zero gas production)-(NH3-N of blank) total N of feedstuff incubated A major problem considered by both groups (25,36) was 11 that gas production reflects more the content of digestible carbohydrates than protein. Futhermore, the results may be additionally affected by atmospheric pressure, pH of the sample, sample size, and the organic acid content of the feedstuff. There is also the difficulty of keeping a stan- dard feedstuff in order to correct for deviations caused by changes in the activity of the rumen liquor, and that feed- stuffs which show significantly slower gas production may actually have a higher in zgzg digestibility. In spite of all this the procedure does have some advantages. There is no separation step which could, depending on the feedstuff, give a bias towards digested or undigested fractions, and it requires a small sample weight-- only 20-30 mg. Mahadevan, et al. (21) also made use of buffered viable rumen fluid (strained), but by a different method. First they linked to SBM with 7-amino-1,3-naphthalene disulfonic acid. The diazotized SBM was then incubated at 37CC for 30,60,120, and 240 minutes. At each interval perchloric acid was added to selected tubes to precipitate the unde- graded SBM and the tubes centrifuged at 15,000xg for 20 minutes. Finally, the pellet was resuspended in .15M NaOH and the absorbance determined at 440nm. This method appears to be superior to the determination of amino acids with ninhydrin because the loss of amino acids by microbial deamination is avoided. Since the pellet is used, rather than supernatant, the complications stemming from reutili- zation of hydrolized protein by the bacteria is eliminated 12 (21). The ninhydrin could be made more compatable with rumen fluid incubations if bacterial deaminases are inacti- vated by adding 0.1mM hydrazine sulfate (6) or destroyed by repeated freezing and thawing (21). This procedure may alter the amino acid's susceptibility to rumen degradation espec- ially the phenolic and imidazole residues. Moreover, in procedures utilizing viable rumen fluid the maintenance of anaerobic conditions throughout the digestion is essential. This is often accomplished by continued gassing with 002 (45), or by the inclusion of cysteine (4) or dithiothreitol (DTT) (6). Rumen fluid activity can vary from day to day and within days due to feeding regimen. In order to provide more controlled conditions in zitgq methods using standard- ized proteases have been developed, and this allows for a greater capability and understanding of comparisons. These proteases are of bacterial (22,30,34,35), fungal (34,35), plant (34,35), or animal origin (34,35,45). Perhaps the most well known of these assays is that of Tilley and Terry (45). In the second stage of their assay they subject the protein to a pepsin digestion which has yielded in vitro values very similar to those found in vivo in sheep. Mahadevan, Erfle, and Sauer (22) used the protease from Bacteriodes amylophilus to simulate rumen proteolysis. The rationale being that B; amylophilus was one of the major proteolytic bacteria of the rumen, plus it does not possess 13 any deaminases which would allow the use of the ninhydrin to determine the extent of degradation. Their research showed that soluble proteins are not degraded at the same rate, and in fact, they noted a six hour lag for some feed samples. They also confirmed the results of earlier workers by noting that solubility is an important factor regarding rumen degradability, but should not be considered a good measure of rumen degradability. Nocek and coworkers (30) chose Streptomyces gresius protease for a similar study because other researchers had shown a close correlation with direct and in_§itu values. Poos-Floyd, et al. (34,35) also examined protease from S; gresius as well as papain (E.C.# 3.4.22.2), bromelain (3.4. 22.4), ficin (3.4.22.3) and neutral fungal protease (NFP) from Aspergillus oryzae. For the most part their results were highly correlated with dacron bag and growth trial data, but the correlations decrease as the periods of incubations increased. NFP, however, did not exhibit this trend and demonstrated correlations as high as .97 at 24 hours in some cases. Ficin and NFP had the lowest coeffi- cients of variation and were therefore selected as the enzymes best suited for in ELEEQ prediction of degradability. Ficin was recommended over NFP because NFP only degraded small amounts of feed sample. The biggest advantages of this procedure is it does not require fistulated animals and the entire digestion can be completed in approximately four hours. 14 A general limitation of all in zitgg methods is that, although they may yield a value for degradability, they do not necessarily yield data representing actual degradation ifl.2112.(44)' While in_xit£g methods have the advantages of being less expensive, less time consuming, and afford the opportunity to maintain more precise conditions than in zixg methods the application of in zitrg data to the pre- diction of results in 1139 is dependent upon how well the in yiyg conditions are known (36). For example, rumen degradability is not only related to the kind of feed and nature of protein, but also a function of residence time in the rumen (31). Most of the aforementioned procedures measure only the fractional rate constant for degradation (Kd). There is, however, one other prime consideration and that is Kr or the fractional rate constant for rumen turn- over. Thus, prediction equations have been developed to calculate rumen degradability based on the interaction of Rd and Kr. Broderick (6) found casein degradation to be first order and from this was able to derive a very simple equa- tion: Percent escape = Kr Kr+Kd Unfortunately when this formula was applied to in vitro values it tended to overestimate the percent escape relative to in vivo data. Orskov and McDonald (31) and Ganev, et al. (13) derived a somewhat more complex equation which took 15 into account different fractions of the feed protein. P = a+b (1-e-Ct ) where: P 2 effective digestibility a = % rapidly disappearing protein (% of total protein) b = % disappearing at constant rate, c, per unit of time, t It must be understood, of course, that this is fitted to a dacron bag incubation which gives values measured under conditions which prevent any passage resulting in an over- estimation of degradation. Therefore, the rate of decrease, K, as estimated by regression analysis of Cr203 passage data can be interpreted as the fractional rate constant. So if 'f' represents the fraction (by weight) of the CrZO3 treated protein which remains in the rumen at 't' hours after feeding then f=e_:Kt (31). Then for a feed free to pass from the rumen cumulative percentage protein degrada- tion up to time 't' can be determined by integration. t .l. f QR dt 0 dt ‘(C+k)t) Pt a+ bc (1-e C+ A problem with this equation is the lack of a lag factor to account for the time required for solubilization and establishment of feed-bacteria contact. Krishnamoorthy and others (18) also recognized the differing fractions or pools of feed protein, a and b, and the fractional passage rate Ki, and were able to derive the following equation: 16 UDN =Zbi Kpi (e'KpiLi ) + (Fe-KP:LLi ) ” Ksi+kpi where: UDN undegraded dietary nitrogen Bi = percent total nitrogen Kpi = fractional turnover rate ksi = fractional degradation rate Li = lag time The correlation of this prediction equation with i situ data, however, was poor (r2=.41). These results are consis- tent with the view that first order in zitrg rate constants for N solubilization do not necessarily reflect rumen degradation. While the data of Broderick (6) supports first order kinetics the lack of fit of the prediction equation to in vivo values may indicate that rumen proteolysis is zero-, first-, or even second order. Moreover, rumen turn- over rates are affected by feeding frequency, type of diet, forage to concentrate ratio, and level of intake. Therefore, it would seem necessary to inelude these terms in a predic- tion model. The solving of such a prediction equation would appear to take a great deal of time and effort, but with the advent of micro computers such an operation would require only seconds to calculate once the data has been entered. It is the lack of relevant data that restricts progress, and in my research I have attempted to reduce the time and labor required to determine the fractional degradation constant, Kd. The two objectives of my research were to: 1) establish in vivo degradabilities for four protein feed supplements 17 using the synthetic bag procedure of Mehrez and Crskov (24); and 2) to evaluate several colorimetric protein assay methods for possible use as a final step in an improved in XEEEQ degradable protein assay which utilized the ficin protease method of Poos-Floyd, et al. (34,35). Finally, the criterion of evaluation for each of the colorimetric assays was based upon the degree of correlation with the ip yiyp and standard ficin assay degradabilities. MATERIALS and METHODS Four protein feed supplements-- soybean meal (SBM), canola meal (CM), linseed oil meal (LOM), and 38% Dairy N.P. (38-D) (supplied by Honnegers and Co. Fairbury, IL appendix IV)-- were used as test materials for the develop- ment of the assay. In an effort to conserve time and rea- gents usually only one or two of the feed samples-- SBM and CM-- were digested in the ip xitgg assay. The remaining two feeds were saved in the event a colorimetric assay proved sufficiently accurate to warrant a study of more feeds in order to further characterize the assay. A nylon bag procedure was carried out on the samples utilizing two ruminally fistulated, nonlactating Holstein, cows and one Holstein x Angus crossbred steer. Three replications were made with this procedure. The first two with the Holstein cows and the third with the steer. In the first trial these animals were fed 20 lbs. grass hay fed at 0800 and 20 lbs. at 1600. In the second the animals were on a 40% forage :60% concentrate test diet which was fed at the same amount and frequency as the hay. In the third trial the steer was fed 30 lbs. of corn silage once each day at 0730. The dacron bags measured 5cm x 10 cm and had an mean pore size of 35 um. Approximately 1 g air dried sample was placed in a bag of predetermined weight. The 18 19 bags were then lyophylized for approximately 48 hours and then transferred to a dessicator until the bag with the sample was weighed and the weight recorded. The initial amount of dry matter was determined by difference. The tops of the bags were closed first with a drawstring and second with a number 10 rubber band looped several times around the top. Following this the bags were rinsed in tap water until the effluent was clear and then further rinsed in distilled water. Finally, three bags of each feed were at- tached to each paddle and two empty "blank" bags were also included. Each plexiglass paddle measured 6 cm x cm and had a nylon retaining string of approximately 50cm which was tied off outside of the fistula. The drawstrings were used to tie the bags to the paddles and in a manner to leave about 5 cm of free movement. The bags were evenly distributed along both long sides of the paddles. All paddles were placed in the rumen via the fistula and removed at 4,8, and 12 hours in trials one and two, and at 2,4,8, and 12 hours in trial three. At each designated interval the paddles were taken from the rumen and submerged in an ice water bath to halt the microbial activity. The bags were rewashed in tap water to remove solubilized feed material, microbes, and other rumen particulate matter. This was followed by a final rinse of distilled water and drying in a 100°C forced air oven. Again the bags were retained in a dessicator until weighed and the residual DM determined by difference. The residual was then removed from the bag, 20 weighed, transferred to a flask, and residual nitrogen determined by macro Kjeldahl. The in 11339 assay chosen was that of P005, et al. (35) (appendix I) and is hereinafter referred to as the ficin assay. This particular assay was chosen because the procedure has provided very reliable results. The assay has been altered in an attempt to make it more compatable with the colorimetric final step and these changes will be detailed. At the end of the digestion the solute was sepa- rated from the solid phase by filtration or centrifugation and the liquid phase was saved for colorimetry. The remain- ing solid was analysed for residual nitrogen by macronel- dahl, and this value subtracted from total nitrogen was the gage used to determine the accuracy of each of the colorimetric methods. The five colorimetric assays chosen were the Lowry (20), Bio Rad (3). A260/A280 (14), ninhydrin (2), and biuret (16,37). The absorbance was measured on a Gilford 24008 spectrophotometer using a 10mm light path and quartz cu- vettes. Two ficin sources were used in the assay. The first which was the most frequently used ficin was of a crude form and had an activity of 0.35 units per mg. The second source was of a purified form and had an activity of 1-2 units per mg. These values were determined as per Sigma (11, appendix II) and a unit is defined as a A280 of 1.0 per minute at pH 7 at 3700, when measuring trichloroacetic acid soluble 21 products liberated from casein in a reaction volume of 10.0 ml and a light path of 10 mm. The pH of the ficin and phos- phate buffer solution was about 5.5 and for the complete digestion mixture was 6.5-6.8 which are the optimum pH ranges for initial activation and maximal activity of the ficin (7,50). The enzyme, in powdered form, was stored in an air tight container and kept under refrigeration (5°C until needed. A fresh enzyme mix was prepared for each digestion run. RESULTS and DISCUSSION The results of the nylon bag studies appear in table 1. These results varied over time, diet, and oil meal (coeffi- cient of variation = 5.8%). It was determined that tripli- cate feed samples were adequate to determine degradability due to the lower coefficient of variation (5.8%) than that reported by Mehrez and Orskov (24) (16.5%) utilizing quadru- plicate feed samples. TABLE 1. Percent Degradability by Nylon Bag DIET** 1 2 3 0_IL mm STE 9L4 SEN mi see 0.1: TIME (Hr) 0 13.26 14.00 28.64 30.75 23.44 27.84 4 34.51 36.71 50.85 62.86 32.01 45.38 8 79.24 76.27 63.88 76.07 41.53 59.85 12 63.29 65.91 68.50 75.44 60.16 69.46 ** - 1 = 20 lbs hay; 2 = 40 lbs 40:60; 3 = 30 lbs corn silage * - SBM = soybean meal CM = canola meal There was also an effect between diets. In diet 1 the results show an undulation in degradability values between 4 and 12 hours. Presumanly this is an effect of a second feeding occurring 30 to 45 minutes prior to the 8 hour sampling. The mechanism for this, however, was not deter- mined. Perhaps this problem could have been avoided had the feeding interval been changed to twelve hours and the 22 23 paddles inserted during the morning feeding. There were significant (P <.001) effects of time by diet and diet by oil meal, but not time by oil meal or time by diet by oil meal (appendix III). With diet 2 the degradability of the feeds seemed to be reaching an asymptote after 8 hours. The cause for the two fold difference in the 0 hour values of diet 1 versus diet 2 is unknown, but it may have been due to a technical error such as improperly weighing the feed samples or the bags. The initially greater degradability and the lack of fluctuating values following the pre-8 hour feeding is most likely a result of the high energy 40:60 (forage : concen- trate) buffered ration. The level of readily fermentable energy enabled rapid microbial growth while the buffer, sodium carbonate, maintained a stable rumen environment and fermentation. Also the rumens of the animals on this diet were observed as being considerably wetter (gross obser- vation) and this may have been a factor in the establishment of feed-microbe contact and the removal of the end-products of degradation. The degradability values determined with the steer on diet 3 are somewhat more linear than those of diet 2, but the feeds are still ranked similarly, plus the 0 hour values of diets 2 and 3 are in closer agreement than with the values of diet 1. A 2 hour degradability estimate was also determined on this third diet, but it was not used in the comparison. Because the third degradability trial is more comprehensive, was not influenced by a second 24 feeding, and contains the data of two other protein supple- ments it was the data chosen for evaluation of the colori- metric assays. The results of this third trial are depicted in figures 1a-1d for SBM, CM, 38%-Dairy, and LOM, respec- tively. The ficin assay as per Poos-Floyd, et al. (appendix I) showed similar degradability trends and levels of variation as the composite nylon bag data. The coefficients of vari- ation of two replicates within time periods of the ficin assay for SBM and CM were 6.2, and 5.6, repectively. The absolute values, however, and even the ranking of the feeds is reversed relative to the nylon bag studies (table 2). TABLE 2. Degradability Values*: Nylon Bag vs. Ficin Assay Nylon Bags Ficin hr SBM CM hr SBM CM 0 21.78 24.20 0 34.78 33.11 4 39.12 48.32 .5 61.44 56.26 8 61.55 70.73 1 68.92 63.51 12 63.98 70.72 2 80.40 70.52 *- composite of three trials or assays The CM determinations for water solubility and overall degradability are in reasonable agreement, but the estimates for SBM differ by almost 20 percentage units. One possible explanation for this might be that ficin is capable of cleaving SBM at more or different residues than the enzymes of whole rumen bacteria. The increasing divergence of the ip vivo and in vitro values of SBM as time of digestion increases is consistent with the findings of Poos-Floyd, 25 100— ”- >~ 1: .5 U .0 g ..- U) 0 C3 2 o\° .o... 20-4 I l l I I) T O 2 4 b 0 IO 12 t (boon) Figure 1a - Percent N degradability of SBM by nylon bag 26 100- ”-1 >~ C: .5 O» W e .. DD 0) D Z 39.«4 and T fl l 1 T) T o 2 4 o a to n 2 t (hours) Figure 1b - Percent N degradability of CM by nylon bag 27 '0‘)" ”-1 >- 3: .5 0 w 2 60‘ DD 0 C3 2 °\° 40.: . - ao- 1 I I l l. T 0 2 4 6 I IO 12 ' (hows) Figure 1c - Percent N Degradability of 38%-D by nylon bag 28 ‘00— ”- >- 3: .5 0 v E 00- U) 0 D Z e\° ao— 20d T —1 l l I l 0 2 4 6 I 10 12 ' thou") Figure 1d - Percent N Degradability of LCM by nylon bag 29 et al. (35) who noted a decreased correlation between the two determinations as the digestion progressed. Initially the residual and solute were separated by filtering through #54 Whatman paper into a 500 ml Erlenmeyer vacuum flask. This solute plus 27.5 ml of distilled water used to rinse the tube were transferred to a clean tube for storage until assayed. This procedure soon proved to be too cumbersome, and the difficulties associated with a limited number of vacuum flasks and a large number of samples could be forseen. Thus, a new method was devised. A #6 rubber stopper with two 5 mm holes was placed atop a 100 ml (30 mm x 160 mm) centrifuge tube. In one hole a stainless steel elbow was inserted and a vacuum hose connected to the free end. A long stemmed, 60O funnel was inserted into the other hole. This assembly (figure 2) was moved from tube to tube as each digestion was filtered. To minimize any contamina- tion between tubes the funnel was thoroughly rinsed and shaken dry after each filtration. This apparatus proved to be light, rapid, and simple, but one problem arose while filtering digests containing cycsteine hydrochloride. There was often considerable air inclusion during filtration leading to excessive foaming, and from time to time a small amount of foam would be drawn up into the vacuum line. How this affected the results is unknown, but this may have altered the final values if the foam was from the ficin mixture. The foaming did not occur when dithiothreitol was substituted for cysteine 3O Glen Funnel Vacuum Hon 9 Stool Fertile a- 6 rubber Hopper -—--- Figure 2. Funnel Assembly 31 hydrochloride, or it was no longer a factor once the solute volume was reduced. The solute was held in the 100 ml centrifuge tube which was stoppered and refrigerated (5°C) until assayed colori- metrically. A small amount of a whitish-tan substance would precipitate out of solution after 24 hours, but quickly went into solution when the tube was shaken or vortexed. The solute usually remained usable for 14 days. After this time a mold often developed and the solute would have to be discarded. The first colorimetric procedure tested was the folin- phenol method of Lowry, Rosebrough, Farr, and Randall (20). This assay was suggested primarily for its capability of detecting protein concentrations as low as 2 ug per ml. Unfortunately, the concentration of these solutes was uncertain so four dilution rates were tested: 500ul, 100ul, and 50ul all diluted to 1 ml. The standard was composed of bovine serum albumin (BSA) at a concentration of 3.2 mg N/ml diluted to 0.4, 0.8, 1.6, and 2.4 mg N/ml. A solution of the various components of the ficin digestion (triton x-100, buffers, ficin, etc.) was prepared in a ratio similar to the assay but at double the concentration. This was then distributed among labelled tubes at the four rates previ- ously described and diluted to 0.5 ml with distilled water. To this 0.5 ml was added 0.5 ml of the respective BSA solu- tions bringing the final concentrations of the standard to .2, .4, .8, 1.2, and 1.6 mg N/ml. The analysis then 32 proceeded as per Lowry, et al. (20). All tubes were of approximately the same absorbance, but were too dark to read on the spectrophotometer so no quantitative data is available. While there were slight differences between dilutions the absorbance within dilutions was roughly the same. A white particulate matter was noted which could be precipitated by centrifugation, but this did not change the absorbance. Perhaps the MgSO4 or NH4 bicarbonate was the interfering substance, but according to Peterson (33) these, and all other constituents, were at acceptable levels in the final dilution. L-cysteine hydrochloride, however, was very near the upper limit in the final dilution so (3 -0H mercaptoethanol (ME) was substituted at a rate of 0.1 moles per mole of L-cysteine hydrochloride. This switch, unfortu- nately, only served to make the assay unwieldy because ME slowed down the filtration and the acrid odor permeated the entire laboratory. Due to these complications, the procedure of Lowry, et al. (20) was no longer used. The second colorimetric method chosen was the dye binding Bio-Rad assay. This assay is based on the observa- tion that the absorbance maximum for an acidic solution of Coomassie Brilliant Blue G-250 shifts from 465nm to 595nm when binding to a protein (3). The Bio-Rad concentrate was diluted 1 to 4 with distilled water and filtered through #54 Whatman filter paper to remove any coagulated material. Following this, one ml of standard (triplicate) or sample (duplicate) was placed in 16mm x 125mm test tubes with 4ml 33 of the diluted Bio-Rad solution. Each tube was vortexed for for 5 seconds, allowed to react for a minimum of 5 minutes, and the absorbance determined at 595nm. All tubes were of the same deep blue color and it was reasoned that the ficin solution was overloading the dye binding reaction. Therefore, a precipitation using 25% trichloroacetic acid (TCA) diluted 1 to 5 with a sample solution was tested. The solution of TCA and sample was vortexed, centrifuged at 2,000xg for 15 minutes, and then 1ml of supernatant was processed through the Bio-Rad procedure as previously outlined. There was a noticeable difference in color development this time -- all tubes were reddish brown. This is a result of the limited sensitivity of Bio-Rad which can only detect pro- teins of a molecular weight of 3,000 or greater, but TCA precipitates proteins having a molecular weight of greater than 2,500. This is the major reason this assay was discon- tinued, but another factor was the staining of the cuvettes. In order to maintain any reproducibility (even if it was all the same absorbance), the cuvettes had to be rinsed after every other sample with 95% ethyl alcohol to remove the blue tint. The third method of choice was a A260/A28O analysis. This colorimetric procedure is based on an absorbance differential of 260nm versus 280nm and detects aromatic amino acid residues. The protein concentration (mg/ml) is then estimated by the equation: mg protein/ml = (1.45 x A280) - (0.75 x A260) (14). 34 For this assay the samples had to be diluted 1 to 20 in order for the absorbance to fall within the ranges of the spectrophotometer. The degradability values derived from this procedure were quite varied demonstrating a composite coefficient of variation of 43%. Furthermore, the assay failed to correctly estimate the degradabilities as the mean degradability for SBM was only 30% and over 100% for CM. Because of the inaccuracies and variability this assay was no longer used. The last two methods chosen -- ninhydrin and biuret-- were also the most frequently tested. Therefore, there are several changes in the ficin assay which are common between them. First, 5ml of 25% TCA was substituted for the 2ml of t-butyl alcohol to halt the enzyme activity, as Krishnamoorthy, et al (18) showed that t-butyl alcohol had no significant effect on enzyme activity plus the TCA would also precipitate the ficin. This was followed by a centri- fugation at 27,000xg for 20 minutes. It was this centri- fugal force that produced the clearest supernatant (gross observation) in the shortest duration of time. The second procedural change was to use a more purified ficin (1-2 units activity/mg) on an equal activity basis (1.7g ficin /l). Unfortunately, the degradation, as measured by Kjeldahl of the residual material, was severely decreased (table 3). The reason for this decreased degrada- tive activity was not determined, but perhaps there is 35 something that is lost during the purification that aids in contacting the feed sample and/or binding the ficin to the feed. Due to this unfavorable result the purified ficin was no longer used. TABLE 3. Degradation by Purified Ficin Enzyme hour % degradability* CE 0 11.01 18.1% SBM 1 16.63 22.0 2 18.84 3.1 0 24.45 2.3% CM 1 37.00 2.9 2 42.08 4.1 * - means of duplicate samples Increasing sample size from .25g to .30g was the next alteration in procedure. The rationale being that if the enzyme could be saturated with substrate then only the feed sample would be undergoing degradation. This would eliminate self-degradation of the ficin and further degradation of liberated peptides. The majority of the liberated proteins would thus be of one fraction -- small peptides -- which would allow the use of the biuret while reducing the varia- tion caused by protein existing in other undetectable forms such as amino acids or large precipitated peptides. In order to minimize variability and interference an attempt was made to reduce the background nitrogen in the assay. There are only three reagents in the digestion that contain any nitrogen: ammonium bicarbonate, ficin, and 36 cysteine hydrochloride. The ammonium bicarbonate is in a very low concentration (‘<.8g/l), and, as demonstrated with the Lowry, the activity of the ficin was reduced when ammo- nium bicarbonate was omitted. Ficin is the enzyme that simulates microbial proteases, and without it the procedure becomes a solubility trial. The cysteine hydrochloride cannot be deleted as it is necessary to maintain a reduced state in the digestion under aerobic conditions (4). More- over, cysteine acts not only as a reducing agent, but also as a coenzyme and an activator of ficin which involves the formation of dissociable compounds between the activator and the enzyme (50). Therefore cysteine cannot be deleted, but only substituted. The most logical substitute at this time appeared to be dithiothreitol (DTT). This dithioanologue of the reduced sugar threitol is a white powder with little, if any, odor, and the oxidation thereof results in an inactive cyclic disulfide which does not react with protein disulfides as p-OH mercaptoethanol has been known to do (40). The most effective concentration appears to be 15mM (2.313g/ l phOSphate buffer) as determined by the Sigma procedure (11). The enzyme and phosphate buffer solution seem to foam much less than when cysteine was used and it remained stable under refrigeration (500) for periods of up to a week whereas the cysteine solution was usable for only 48 hours, maximum. The fifth and final change was the substitution of 2ml of 50% sulfasalicylic acid (SSA) for TCA or PCA. SSA does not interfere with the Kjeldahl as does TCA, and it 37 precipitates large peptides ( 2,000 m.w.) more thoroughly than PCA resulting in a clearer and more uniform supernatant (gross observation). The ninhydrin procedure was first tested with the solute from the digestion using the purified ficin enzyme. The assay was carried out as per AOAC (2) with a sample of the 27,000xg supernatant. The results showed a difference only between the 0 hour and 1 hour samples for canola meal, and even then the values severely underestimated the degree of degradation (table 4). TABLE 4. Percent Degradation by Ninhydrin oil meal SBM QM hour 0 1.04 1.29 1 2.97 3.11 2 3.88 3.00 The ninhydrin method was also tested with the solute from the digestion using DTT instead of L-cysteine hydro- chloride primarily because of increased hydrolysis seen with other assays and now that cysteine was no longer a part of the assay perhaps the overloading previously encountered would be eliminated. Unfortunately, this was not the case. The assay was attempted twice and both times were unsuccessful. The color development was erratic and very unstable even though all reagents were made up fresh before each attempt. The only explanation to offer for this occurance is that possibly the 1,2,3 Triketohydrindene was 38 no longer stable as it was the only organic compound in the assay other than the digestion supernatant itself. The final test of the ninhydrin procedure was made using the supernatant from the precipitation using 50% SSA. Unfortunately, similar problems were encountered with the ninhydrin so the results are still inconclusive. The fifth and most promising method tested was the biuret assay of Koch and Putman (16) and Robinson and Hogden (37). Biuret appears to have two absorbance maxima at 275nm and 600nm (fig. 3). Koch and Putman (16) used this assay at 330nm for assaying whole bacteria with success, and I found similar success with standards and samples using their procedure. Regression coefficients for standards were often 0.999 or greater, and there was a noticeable difference between time periods of digestion. Similar results were seen with the method of Robinson and Hogden (37) who deter- mined the absorbance at 560nm, and it was this wavelength that was chosen. At 560nm the area under the absorbance curve is greater which should allow one to discern differ- ences between samples with greater accuracy, and the slepe of the line at this point in the graph is low so any inaccu- racies in wavelength will affect the results only moderately,_ Plus, at this wavelength there is much less variability due to electronic interferences. There were, however, discrepancies within time periods. Within each time period the absorbance of the smaller sample was consistently greater than that of the larger sample. 39 popsflm Mo msflxmz oommnmomp< I m omzmam .52. .. fmco_o>o>> a o _ L )—-1 111/ )t)\\.1 I N. — aouquosqv .Iv.u 40 It was presumed that this was the result of the larger feed sample being able to bind more ficin thereby reducing the amount in the filtrate. Proceeding on the assumption that the ficin molecule was of a greater molecular weight than any of the digestion products an aliquot of filtrate was centrifuged at 2,000xg for 15 minutes without the aid of a precipitating agent. The gross appearance of the solute was not changed, but a dark brown precipitate was noted. This additional step served to sufficiently alter the absorbance such that the relationship with the original sample weights fell within expected values (table 5). TABLE 5. Absorbance* Before and After Centrifugation Hour Sample weight (g) Before After 0 .2502 5.575 4.508 .2519 5.240 4.720 1/2 .2502 5.308 4.500 .2506 5.243 4.717 1 .2556 5.272 5.000 .2546 5.161 4.899 * - Absorbance per gram SBM, n=2 In an attempt to further reduce variation the working volume was reduced in order to concentrate the soluble pro- tein so that the aliquot used in the biuret assay would be of a greater proportion of the whole digestion solute and the ratio of soluble protein to background nitrogen (ficin, NH bicarb, etc.) would be increased. The ficin solution 4 was concentrated two fold so that only half of the original 41 amount need be used, and the volume of the rinse was reduced to 15 ml. Thus the final volume was now 38.5ml as originally used. This did not significantly alter the variation between samples within time periods, but did increase the differ- ences between time periods (table 6). TABLE 6 - Mean Absorbances* at Two Working Volumes .22 El 3§a2.El hour Absorbance 91. Absorbance Cl 0 1.281 5.2% 1.131 4.0% 1.377 1.199 1/2 1.582 4.8 1.308 5.2 1.481 1.241 1 1.528 5.6 1.535 2.1 1.523 1.581 FAbsorbance: n=2, Coefficient of Variation: n=4 The biuret was also tested with the samples from the purified ficin digestion. It showed very favorable results with SBM in that trends with time were clearly visible, but the degradability was always ten percentage units below that which was expected. The canola meal, however, proved to be different. The only difference lay between 0 and 1 hour, plus the degradability was significantly overestimated -- 186% degradable. Because of the poor results demonstrated with the purified ficin the procedure returned to using the crude grade ficin and the results of the Kjeldahls were more normal. When the biuret was tested with the two different 42 sample sizes there were no differences between sets of samples sizes in terms of degradability. The supernatant (27,000xg, 20 minutes) also underwent a Kjeldahl analysis in an attempt to correlate, or at least establish a ratio between, mg nitrogen per ml as determined by Kjeldahl versus mg nitrogen per ml as per biuret. The results of this com- parison were varied, showed no trends, and in some cases were quite extreme. Part of the discrepancy may be due to TCA changing to chloroform during distillation of the Kjeldahl leading to a shift in the titration point (26). Substitution of 3ml of 3M perchloric acid (PCA) for TCA followed by centrifugation and neutralization with 3M K2 CO3 seemed to remedy this problem. The correlations of N content by Kjeldahl and biuret were .9794 and .9748 for SBM and CM, respectively. The graphs of the regression equations are depicted in figure 4. The slopes were not significantly different, but the combined regression of the biuret nitrogen determination on the values determined by Kjeldahl of SBM and CM was only .5841. The reason for the signifi- cantly different y-intercept of SBM cannot be explained. The results of the biuret analysis on the solute with DTT showed that while the activity of the ficin was still the same when DTT was used, it appeared to be more efficient at hydrolyzing liberated peptides. The values, as indicated by the biuret, reached a maximum at the half hour interval, but the residual N values (Kjeldahl) continued to decrease 43 I — Canola Maal / --$aybaan Meal mg N lml "a...” r T I .2 3 ,4 mg N lml (Kialdahl) '0—1) Figure 4 - N Determination of Solute: Biuret vs Kjeldahl 44 through two hours of digestion. The final test of the biuret procedure was made using the supernatant from the precipitation using 50% SSA. Unfortunately, similar problems were encountered with this procedure so the results are still inconclusive. CONCLUSION The digestion and colorimetric assays are incompatable as they exist at the present time. Further changes will need to be made in either the digestion, colorimetry, or both in order for the procedure to be successful. Perhaps some of these changes include replacing ficin with a strict endopeptidase or increasing sample size even more so that the net effect would be more peptides and less free amino acids, and a greater compatability with the biuret assay. Perhaps using a different colorimetric procedure, but the biuret does seem to have the most promise of the five assays tested. This assay is simple, rapid, does not seem to be as adversely affected by interfering substances, and, other than the spectrophotometer, it requires no highly specialized equipment. 45 APPENDI CES APPENDI x _I_ Procedure for Determining Proteolytic Enzyme Degradation of Protein Feeds Mary Poos-Floyd, Terry Klopfenstein, and R.A. Britten Materials: 50 ml test tubes, #6 rubbers stoppers w/ small holes Ficin Incubator at 39-400 0 Reagents: A. Stock phosphate buffer 0.5M 10.54g KZHPO4 / litre 2 4 B. Working phosphate buffer 0.1M 59.66g'KHm / litre 200ml stock phosphate buffer/litre 6.1g cysteine hydrochloride/litre Ficin enzyme is dissolved in working phosphate buffer at a concentration of 4. 3g/litre C. 1% sodium azide solution (w/v) D. 1% triton X-100 solution (v/v) E. 80% t-butyl alcohol F. In vitro rumen buffer and Macromineral solutions In vitro rumen buffer gllitre Ammonium —bicar50nate Sodium bicarbonate 35 Macromineral solution a2 4 (afihy.) 5.7 KH 2P04 (anhy. ) 6.2 MgSO4 7H2 0 0.6 Procedure: A. Weigh out enough sample to provide 15mg N per tube. 1. Prepare duplicate tubes for each sampling time 2. Use 0, .5, 1, 2 hour samples 46 47 B. Prewet samples with 5ml distilled water 12 hours before next step. 0. Add 10ml of 1:1 solution of in vitro rumen buffer and macromineral solution to each tube. D. Add 1ml sodium azide solution and .5ml Triton X-100 solution to each tube. Stopper and swirl and incubate at 39°C for two hours. E. After two hours, remove tubes from the incubator and add 10ml of prewarmed enzyme solution to all except 0 hour test tubes. Stopper, swirl, and return to incubator. - 1. 0 hr tubes serve as an estimate of protein solubility. 2. 0 hr samples should be filtered and washed through Whatman #541 filter paper. F. At the appropriate time, remove tubes and halt incubation be adding 2ml of 80% t-butyl alcohol. Filter and wash samples on Whatman #541 filter paper and perform Kjeldahl analysis of the filter paper plus residue. Calculations: % Residual N = %’N Remaining % Total N Residual N as % = % N Remaining of insoluble N % N in 0 hr sample I. II. III. APPENDIX ;I_ ENZYMATIC ASSAY g_ FICIN f (Sigma Chemical Co., February 1973) Principle icin hydrolyzes peptides, amides, and esters; its specificity is similar to that of Papain Reagents and Procedure Into two suitable vials labelled "Blank" and "Test" pipette the following reagents: Rea ents BLANK TEST a. 2% casein in 0.1M KH2P04 buffer 2.0ml 2.0ml g. k. pH 7.6, heat to dissolve, but DO NOT BOIL, cool to 37°C for assay .25M L-cysteine, adjust to pH 7 0.2 0.2 with solid sodium bicarbonate .25M EDTA 0.2 0.2 25 N NaOH 0.2 0.2 1.0M KH2P04 buffer, adjust to pH 0.6 0.6 7.0 with NaOH distilled water 0.8 0.4 Equilibrate at 37°C then, at zero time, add; enzyme solution --- 0.4 for a crude powder prepare a sol- ution in reagent 'e' containing 0.1 mg enzyme/ml after adding enzyme solution mix well and incubate at 37°C for exactly 20 minutes. add 6.0 ml of 5% TCA and let stand at 37°C for 60 minutes filter through Whatman #50 or equivalent filter paper read A280 of "Test" versus "Blank" Calculation A280 = Units/mg 20 x mg enzyme/ 10 ml reaction mix 48 49 IV. Unit Definition one unit is equivalent to a AIA280 of 1.0 per minute at pH 7 at 37°C when measuring TCA soluble products liberated from casein in a reaction volume of 10 ml and 1 cm lightpath. APPENDIX III STATISTICAL ANALYSES ANALYSIS OF VARIANCE: SOYLEAN MEAL (ficin assay) TABLES pg MEANS TIEE (LIN) o 30 60 120 % DEGRADABILITY 34.78 61.44 68.92 80.40 REPS 6 6 6 4 SE 3.672 ANALYSIS OF VARIANCE: CANOLA MEAL (ficin assay) TABLES pg MEANS TIME (MIN) 0 30 60 120 % DEGRADABILITY 33.11 56.26 63.52 70.52 REPLICATICNS 6 6 6 4 SE 3.042 50 51 AEALYSIS OF VARIANCE: PERCE§T NITROGEN DEGRADABILITY (ficin assay TABLES pg AEANS TIME (LIN) 0 30 60 120 % DEGRADAEILITI 33.94 58.85 66.22 75.46 REPLICATES 12 12 12 s OIL MEAL SBM M 59.66 54.52 TIME SBM CM O 34.78 33.11 REPS 6 6 30 61.44 56.26 REPS 6 6 60 68.92 63.51 REPS 6 6 120 80.40 70.52 REPS 4 4 SE 3.372 ANALYSIS OF VARIANCE: SOYBEAN MEAL (nylon bags) TABLES pg MEANS TIRE (HR) % DEGRADABILITY DIET % DEGRADABILITY DIET TIRE (HR) O 4 8 12 SE 0 21.78 1 47.58 13a26 34.51 79.24 63.29 4 39.12 3.082 2 52.97 28.64 50.85 63.88 68.50 8 61.55 12 63.98 39.28 23.44 32.01 41.53 60.16 52 AEALYSIS OF VARIANCE: CANOLA MEAL (nylon bag) TABLES pg MEANS TIME (HR) 0 4 8 % DEGRADABILITY 24.20 48.32 70.73 DIET 1 2 % DEGRADABILITY 48.22 61.28 DIET 1 2 TIME 0 14.00 30.75 4 36.71 62.86 8 76.27 76.07 12 65.91 75.44 SE 2.736 12 70.27 50.63 27.84 45.38 59.85 69.46 APPENDIX I_V 38% Dairy - N.P. - 337 GUARANTEED ANALYSIS Crude Protein, not less than 8.0% Crude fat, not less than 0.5% Crude fiber, not more than 9.0% Calcium, maximum 3.0% Calcium, minimum 2.0% Phosphorus, minimum 0.7% Iodine, minimum 0.00009% Selenium, maximum 0.0001% NaCl, maximum 2.5% NaCl, minimum 1.5% Vitamin A 20,000 USP units per lb INGREDIENTS Animal protein products, plant protein products, processed grain by-products, grain screenings, cane molasses, Vitamin A supplement, D-Activated animal sterol, Vitamin E supple- ment, salt, calcium carbonate dicalcium phosphate, iron carbonate, manganous oxide, copper oxide, cobalt carbonate, calcium iodate, zinc oxide, zinc sulfate, magnesium oxide, natural and artificial flavors, sodium selenite 53 BI BLI O GRAPHY 10. 11. 12. BIBLIOGRAPHY Annison, E.F., Nitrogen metabolism in the sheep: Protein digestion in the rumen, Biochem. i; 64:705-14 (1956) Association Of Official Analytical Chemists, Official Methods of Analysis, 12th ed. Washington 1975 Bio-Rad protein assay (research manual), Bio-Rad Labora- tories, Richmond, California Blackburn, T.H. and P.N. Nelson, Proteolysis in the sheep rumen by whole and fractionated rumen contents, q; Gen. Microbiol. 22:272-81 (1960) Burroughs, W., D.K. Nelson, and D.R. Mertens, Protein physiology and its application in the lactating cow: The metabolizable protein feeding standard, g; Anim. Sci, 41:933 (1975) Broderick, G.A., IQ Vitro procedures for estimating rates Of ruminal protein degradation and proportions of protein escaping the rumen undegraded, gleutr. 108:181- 90 (1978) Cohens W., Characterization of Ficin, Nature 182:650-60 1958 Craig, W.M. and G.A. Broderick, Amino acids released during protein de radation by rumen microbes, {y Anim. Sci. 58: 436-443 T1984) Crawford, R.J., W.H. Hoover, C.J. Sniffen, and B.A. Crooker, Degradation of feedstuff nitrogen in the rumen vs. nitrogen solubility in three solvents, J; Anim. Sci. 46:1768-75 (1978) Crooker, B.A., C.J. Sniffen, W.H. Hoover, and L.L. Johnson, Solvents for soluble nitrogen measurements, 1; Dairy Sci. 61:437-47 (1978) Enzymatic Assay of Ficin, Sigma Chemical Co., Inc., St. Louis, MO Evans, E., An evaluation of the relationships between dietary parameters and rumen solid turnover rate, Can. i; Anim. Sci. 61:97-103 (1981) 54 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 55 Ganev, G., E.R. Orskov, and R. Smart, Effect of rough- age or concentrate feeding and rumen retention time on total degradation of protein in the rumen, g; Agri. Sci., Camb. 93:651-56 (1979) Kalckar, H.M., Differential spectrophotometry of purine compounds by means of specific enzymes: III. studies of the enzymes of purine metabolism, g; Biol. Chem. 167: 461-86 (1947) Kempton, T.J., J.V. Nolan, and R.A. Leng, Principles for the use of non-protein nitrogen and by-pass protein in thg diets of ruminants, World Anim. Rev. 22:2-10 1979 Koch, A.L., S.L. Putman, Sensitive biuret method for the determination of protein in an impure system such as whole bacteria, Anal. Biochem. 44:239-45 (1971) Krishnamoorthy, U., T.V. Muscato, C.J. Sniffen, and P.J. Van Soest, Nitrogen fractions in selected feed- stuffs, g; Dairy Sci. 65:217-25 (1982) Krishnamoorthy, U., C.J. Sniffen, M.D. Stern, and P. J. Van Soest, Evaluation of a mathematical model of rumen digestion and an ip vitro simulation of rumen proteol- ysis to estimate the rumen undegraded nitrogen content of feedstuffs, Brit. g; Nutr. 50:555-68 (1983) Lindberg, J.E., H.S. Soliman, and S. Sanne, A study of the rumen degradability of.untreated and heat treated rapeseed meal and of whole rapeseed, including a comparison of two n lon bag techniques, Swed. g; Agric. Res. 12:83-88 (1982 Lowry, 0.H., N.J. Rosebrough, A.L. Farr, and R.J. Randall, Protein measurement with the folin-phenol reagent, g; Biol. Chem. 193:265-275 (1951) Mahadevan, S., J.D. Erfle, and F.D. Sauer, A colori- metric method for the determination of proteolytic degradation of feed proteins by rumen microorganisms, g; Anim. Sci. 48:947-53 (1979) Mahadevan, S., J.D. Erfle, and F.D. Sauer, Degradation of soluble protein by Bacteriodes amylophilgg protease and b rumen microorganisms. g;_gpggy Sci. 50:723-28 (1980) Mathers, J.C. and E.L. Miller, A simple procedure using 3SS incorporation for the measurement of microbial and undegraded food protein in ruminant digesta, Brit. 13. 14. 15. 16. 17. 18. 19. 20. 22. 23. 56 Ganev, G., E.R. Orskov, and R. Smart, Effect of rough- age or concentrate feeding and rumen retention time on total degradation of protein in the rumen, J; Agri. Sci.. Camb. 93:651-56 (1979) Kalckar, H.M., Differential spectrophotometry Of purine compounds by means of specific enzymes: III. studies of the enzymes of purine metabolism, g; Biol. Chem. 167: 461-86 (1947) Kempton, T.J., J.V. Nolan, and R.A. Leng, Principles for the use of non-protein nitrogen and by-pass protein in the diets Of ruminants, World Anim. Rev. 22:2-10 (1979) Koch, A.L., S.L. Putman, Sensitive biuret method for the determination of protein in an impure system such as whole bacteria, Anal. Biochem. 44:239-45 (1971) Krishnamoorthy, U., T.V. Muscato, C.J. Sniffen, and P.J. Van Soest, Nitrogen fractions in selected feed- stuffs, 1; Dairy Sci. 65:217-25 (1982) Krishnamoorthy, U., C.J. Sniffen, M.D. Stern, and P. J. Van Soest, Evaluation of a mathematical model Of rumen digestion and an £3 vitro simulation of rumen proteol- ysis to estimate the rumen undegraded nitrogen content of feedstuffs, Brit. J; Nutr. 50:555-68 (1983) Lindberg, J.E., H.S. Soliman, and S. Sanne, A study of the rumen degradability of untreated and heat treated rapeseed meal and of whole rapeseed, including a comparison of two n lon bag techniques, Swed. l; Agric. Res. 12:83-88 (1982 Lowry, 0.H., N.J. Rosebrough, A.L. Farr, and R.J. Randall, Protein measurement with the folin-phenol reagent, 1; Biol. Chem. 193:265-275 (1951) Mahadevan, S., J.D. Erfle, and F.D. Sauer, A colori- metric method for the determination of proteolytic degradation of feed proteins by rumen microorganisms, g; Anim. Sci. 48:947-53 (1979) Mahadevan, S., J.D. Erfle, and F.D. Saue , Degradation of soluble protein by Bacteriodes amylop ilus protease and b rumen microorganisms. J; Anim. Sci. 50:723-28 (1980) Mathers, J.C. and E.L. Miller, A simple procedure using 35 S incorporation for the measurement of microbial and undegraded food protein in ruminant digesta, Brit. 24. 25. 26. 27. 28. 29. 30. 32. 33. 34. 57 ;; Nutr. 45:505-13 (1980) Mehrez, A.Z. and E.R. Orskov, A study of the artificial bag technique for determining the digestibility of feeds in the rumen, Jy_Agric. Sci., Camb. 88:645-50 (1977) Menke, K.H., L. Raab, A. Salewski, H. Steingrass, D. Fritz, and W. Schneider, The estimation of the digest- ibility and metabolizable energy Of ruminant feeding- stuffs from the gas production when they are incubated with rumen liquor gp vitro, 1; Agric. Sci., Camb. 93: 217-22 (1979) Merck Index, 8th ed. Merck and Co., Inc., Rahway, NJ 1978 Miller, E.L., Symposium on nitrogen utilization by the ruminant, Proc. Nutr. Soc. 32:79-84 (1973) Miller, E.L., Research into end-products Of digestion in relation to the proposed ARC protein system, pp. 15.1-15.7 in D.F. Osbourn, D.E. Beers, and D.J. Thompson (ed), Ruminant Digestion and Feed Evaluation, Agric. Res. Council, London McAllan, A.B. and R.H. Smith, The efficiency of micro- bial protein synthesis in the rumen and the degradation of feed nitrogen between the mouth and the abomasum of gteerg given different diets, Brit. g; Nutr. 51:77-83 1984 Nocek, J.E., J.H. Herbein, and C.E. Polan, Total amino acid release rates of soluble and insoluble protein fractions of concentrate feedstuffs by Streptomyces griseus, J; Dairy Sci. 66:1663-67 (1983) Orskov, E.R. and I. McDonald, The estimation of protein degradability in the rumen from incubation measurements weighted accordin to rate Of passage, J; Agric. Sci., Camb. 92:499-503 I1979) Orskov, E.R. and A.Z. Mehrez, Estimation of extent of protein degradation from basal feeds in the rumen of sheep, Proc. Nutr. Soc. 36:78A (1977) (abstr.) Peterson, G.L., Review of the folin-phenol protein quantitation method of Lowry, Rosebrough Farr, and Randall, Anal. Biochem. 100:201-20 (1979) Poos, M., T. Klopfenstein, R.A. Britton, and D.G. Olson, An enzymic technique for determining rumen protein degradation, J; Dairy Sci. 63: (suppl. 1) 142 (abstr.) 36. 37. 38. 59. 40. 41. 42. 43. 44. 45. 46. 58 Poos-Floyd, M., T. Klopfenstein, and R.A. Britton, Evaluation of Laboratory techniques for predicting fuminal protein degradation, J; Dairy Sci. 68:829-39 1985 Raab, L., B. Cafantaris, T. Jilg, and K.H. Menke, Rumen protein degradation and biosynthesis 1. a new method for the determination of protein degradation in rumen fluid lg vitro, Brit. {y_Nutr. 50:569-82 (1983) Robinson, H.W. and C.G. Hogden, The biuret reaction in the determination of serum proteins, J; Biol. Chem. 155:727 (1940) Santos, K.A., M.D. Stern, and L.D. Satter, Protein degradation in the rumen and amino acid absorption in the small intestine of lactating dairy cattle fed various protein sources, J; Anim. Sci. 58:244-55 (1984) Satter, L.D. and L.D. Slyter, Effect of ammonia concen- tration on ruminal microbes in vitro, g; Anim. Sci. 55:273 (1972) (abstr.) Scopes, R., Protein Purification, 1st ed. Springer- Verlag, New York 1982, pp. 195-97 Smith, R.H. and A.B. McAllan, Nucleic acid metabolism in the ruminant: Amounts of nucleic acids and ammonia nitrogen in digesta from the rumen, duodenum, and ileum of calves, Brit. 1; Nutr. 25:181-89 (1971) Stern, M.D., K.A. Santos, and L.D. Satter, Protein degradation in the rumen and amino acid absorption in small intestine of lactating dairy cattle fed heat treated whole soybeans, 1; Dairy Sci. 68:45-56 (1985) Stern, M.D. and L.D. Satter, Evaluation of nitrogen solubility and the dacron bag technique as methods for estimating protein degradation in the rumen, gy_Anim. Sci. 58:714-24 (1984) Tamminga, 8., Protein degradation in the forestomaehs Of ruminants, J; Anim. Sci. 49:1615-30 (1979) Tilley, J.M.A. and R.A. Terry, A two stage technique for the in vitro digestion Of forage crops, J; Brit. Grassland Soc. 18:104-11 (1963) van't Klooster, A. Th. and R.A. Boekholt, Protein digestion in the stomach and intestines of the cow, Neth. §;_Agric. Sci. 20:272-84 (1972) 47. 48. 49. 50. 51. 52. 54. 55. 59 Waldo, D.R. and H.K. Goering, Insolubility of proteins in ruminant feeds by four methods, J; Anim. Sci. 49: 1560-68 (1979) Warner, A.C.I., Proteolysis by rumen microorganisms, J; Gen. Microbiol. 14:749-62 (1956) Weakley, D.G., M.D. Stern, and L.D. Satter, Factors affecting disappearance of feedstuffs from bags suspended in the rumen, J; Anim. Sci. 56:493-507 (1983) Winnick, T., W.H. Cone, and D.M. Greenberg, Experiments pn thI activation of Ficin, J; Biol. Chem. 153:465-70 1944 Wohlt, J.E., C.J. Sniffen, and W.H. Hoover, Measurement Of protein solubility in common feedstuffs, 1; Dairy Sci. 56:1052-57 (1973) Wohlt, J.E., C.J. Sniffen, W.H. Hoover, L.L. Johnson, and C.K. Walker, Nitrogen metabolism in wethers as affected by dietary protein solubility and amino acid profile, J; Anim. Sci. 42:1280-89 (1976) Wright, P.L., Protein and carbohydrate degradation by rumen microorganisms, J; Anim. Sci. 31:259 (1971) (abstr.) Zinn, R.A., L.S. Bull, and R.W. Hemken, Degradation of supplemental proteins in the rumen, J; Anim. Sci. 52: 857-66 (1981) Sniffen, C.J. and L.E. Chase, Formulating Diets for soluble and degradable protein. Proc. 1982 Feed Dealer Seminars, pp. 58-61d.