The Journal of Nutrition Vol. 128 No. 3 March 1998, pp. 598-605

Guanidinated Protein Test Meals with Higher Concentration of Soybean Trypsin Inhibitors Increase Ileal Recoveries of Endogenous Amino Acids In Pigs^1,2

William R. Caine^{*, , 3}, Willem C. Sauer, Martin W. A. Verstegen^*, Seerp Tamminga^*, Shaoyan Li, and Hagen Schulze^**

^* Department of Animal Nutrition, Wageningen Institute of Animal Sciences, Wageningen Agricultural University, Wageningen 6709 PG, The Netherlands;  Department of Agricultural, Food and Nutritional Science, University of Alberta, Edmonton T6G 2P5, Canada; and ^** Finnfeeds International Limited, Marlborough, Wiltshire SN8 1NX, United Kingdom

	ABSTRACT

Abstract Introduction Results & Discussion References

The amino acid concentrations of cornstarch-based guanidinated unprocessed (UGM) and autoclaved (AGM) Nutrisoy (defatted soy flour) protein test meals were compared with the respective unguanidinated Nutrisoy diets. Endogenous ileal recoveries and true digestibilities of amino acids were determined in six growing pigs, fitted with a simple T-cannula at the distal ileum, fed the guanidinated protein test meals. The UGM and AGM contained 13.4 (high) and 3.0 (low) g/kg dry matter of soybean trypsin inhibitors (SBTI), respectively. The experiment was a two-period cross-over design with each period lasting 15 d. On d 14 of each period, the pigs were fed the guanidinated test meals followed by 24 h continuous collection of digesta. Concentrations of crude protein and most of the amino acids in the test meals were higher than in the respective diets. Apparent ileal amino acid digestibilities of the test meals did not differ (P > 0.05) from reported values for the respective diets and were higher (P < 0.05) by 22.7 (cysteine) to 61.3 (tyrosine) percentage units for AGM compared with UGM. The ileal recoveries of endogenous amino acids in AGM-fed pigs were lower (P < 0.05) than UGM-fed pigs. Values ranged from 0.10 (arginine) to 0.64 (aspartate + asparagine) and from 0.84 (histidine) to 2.61 (tyrosine) g/kg dry matter intake for AGM- and UGM-fed pigs, respectively. True ileal amino acid digestibilities for AGM were higher (P < 0.05) than UGM with differences ranging from 12.7 (tyrosine) to 38.3 (leucine) percentage units. In conclusion, ileal recoveries of endogenous amino acids were increased in pigs fed guanidinated protein test meals with the higher concentration of SBTI.

	INTRODUCTION

Abstract Introduction Results & Discussion References

Different methods for estimating endogenous ileal recoveries of amino acids in pigs have been reported in the literature. These methods include feeding protein-free diets (e.g., de Lange et al. 1989), peptide alimentation and synthetic amino acid diets (Butts et al. 1993), regression analyses (Furuya and Kaji 1986), ¹⁵N-leucine, ¹⁵N-isoleucine and ¹⁵N-isotope dilution techniques (de Lange et al. 1992). The homoarginine technique described by Hagemeister and Erbersdobler (1985) has been used to estimate the ileal recovery of endogenous lysine in rats (Moughan and Rutherfurd 1990) and pigs (Marty et al. 1994) and endogenous nitrogen in pigs (Barth et al. 1993, Schmitz et al. 1991). Marty et al. (1994) indirectly estimated the flow of endogenous amino acids based on their concentration relative to lysine in endogenous protein collected from pigs fed a protein-free diet as reported by de Lange et al. (1989). The homoarginine technique involves guanidination of dietary protein to chemically convert lysine into the synthetic derivative homoarginine. The guanidinated protein then is fed as test meals. The technique has been used in a number of studies based on the assumption that the chemical transformation does not affect the digestion or absorption of the dietary protein. However, there are no reported comparisons of the amino acid composition of dietary protein before and after guanidination.

Addition of antinutritional factors such as Kunitz trypsin inhibitors (Barth et al. 1993) and lectins (Schulze et al. 1995) to diets increase the amount of undigested endogenous nitrogen leaving the small intestine of pigs. Barth et al. (1993) concluded that the amino acid composition of endogenous ileal nitrogen may be important in terms of maintaining protein homeostasis of pigs fed diets containing protease inhibitors. In this context, Siriwan et al. (1994) adapted the homoarginine technique to estimate recoveries of endogenous amino acids in poultry by determining changes in the ratios of homoarginine to amino acids in diet and digesta after feeding test meals of guanidinated casein and soybean protein. This homoarginine ratios method has not been used in studies with mammals and may be a relatively simple direct approach to determine ileal recoveries of endogenous amino acids.

The objectives of this study were twofold. The first was to compare the amino acid composition of guanidinated Nutrisoy (defatted soy flour) protein test meals and the respective unguanidinated Nutrisoy diets. The second objective was to measure endogenous ileal recoveries and true digestibilities of amino acids in growing pigs fed guanidinated Nutrisoy protein test meals with high or low concentration of soybean trypsin inhibitors (SBTI⁴) using the homoarginine ratio method.

	EXPERIMENTAL

Animals and diets.  This study was carried out in conjunction with an experiment to determine fecal and ileal apparent digestibilities of energy and amino acids in pigs fed cornstarch-based diets with either unprocessed or autoclaved Nutrisoy (supplied by Archer Daniels Midlands, Decatur, IL; 530 g crude protein/kg dry matter) as the protein source (Li et al. 1997a). Details of animals and management are described in detail by Li et al. (1997a).

The experiment was carried out according to a two-period cross-over design (Petersen 1985). Six barrows, average body weight 53.3 ± 3.7 kg, fitted with a simple T-cannula at the distal ileum were housed in individual metabolism crates in a temperature-controlled (25 ± 1°C) room at the University of Alberta Metabolic Research Facility. Formulation of the experimental diets and guanidinated protein test meals are presented in Table 1. The diets were formulated with either unprocessed or autoclaved Nutrisoy to contain 200 g crude protein/kg with 13.4 (high) and 3.0 (low) g SBTI/kg dry matter, respectively. The guanidinated protein test meals were formulated the same as the respective diets except that dysprosium chloride was included at 0.116 g/kg (providing ~50 mg/kg of dysprosium) as an additional digestibility marker specific to the test meals. Canola oil was included to meet National Research Council (1988) standards for digestible energy. Vitamins, minerals and DL-methionine also were included to meet or exceed NRC (1988) standards. Chromic oxide was included at 3 g/kg of diet as a digestibility marker.

The pigs were fed at a daily rate of 4 g/100 g body weight in two equal meals at 0800 and 2000 h throughout the experiment. The pigs were weighed at the beginning of each experimental period and their feed intake adjusted accordingly. Each period lasted 15 d. From d 1 to 13, the pigs were fed the experimental diets. At 0800 h on d 14, unprocessed (UGM) and autoclaved (AGM) Nutrisoy guanidinated protein test meals were offered to the pigs. A casein enzymatic hydrolysate (CEH) meal of 900 g was given before and after the guanidinated protein test meals (2000 h on d 13 and 14, respectively) to clear the small intestine of any undigested protein and chromic oxide originating from the previous unguanidinated Nutrisoy diets. As for the Nutrisoy diets, the CEH meals were formulated to contain 200 g crude protein/kg (Table 1). The amino acids in the CEH meals were assumed to be absorbed completely based on true digestibilities of cornstarch-based casein diets fed to pigs (Chung and Baker 1992). Particular care was taken to ensure that the guanidinated test meals were consumed completely by the pigs by cleaning the feeders and crates before the test meal and suspending plastic sheets below the metabolism crates to collect spilled feed, which was added back to the feeders. The guanidinated test meals were consumed within 2 h after being offered. Ileal digesta was collected continuously for 24 h starting immediately after the test meal was offered. The digesta was collected into a plastic bag, which was connected to the barrel of the cannula of each pig. The bags contained 10 mL of formic acid (2.5 mol/L) to stop microbial activity. Bags were changed at least once every h and digesta were pooled within pig and period and frozen at 20°C until analyses.

The experimental proposal and surgical procedures were approved by the Animal Care Committees of the Faculty of Agriculture, Forestry and Home Economics at the University of Alberta and the Wageningen Institute of Animal Sciences at the Wageningen Agricultural University. The pigs were cared for in accordance with the guidelines established by the Canadian Council on Animal Care (1980).

Preparation of the guanidinated test meals.  Lysine residues in batches of unprocessed and autoclaved Nutrisoy were guanidinated as described by Schmitz et al. (1991). Batches (377 g) of the respective Nutrisoy were mixed with 1 L of deionized distilled water in 4 L beakers. Methylisourea (1 L, 0.4 mol/L) was added to each beaker, and the pH adjusted with 4 mol/L NaOH to 10.3, which is optimal for guanidination of soybean protein according to Maga (1981). The methylisourea was prepared previously by dissolving and mixing a solution of 0.4 mol/L o-methylisourea hydrogen sulfate and 0.4 mol/L barium hydroxide octahydrate (Sigma Chemical, St. Louis, MO). The solution then was centrifuged and filtered to remove the barium sulfate precipitate. The Nutrisoy slurries were stirred continuously for 1 h, and the beakers then covered with aluminum foil and stored in a cold room at 4°C for 4 d. The Nutrisoy slurries in the beakers were stirred each day, and the pH adjusted to 10.3. The guanidination reaction was stopped by precipitation of the protein at pH 4.5 (isoelectric point) using 1 mol/L HCl (Schmitz et al. 1991). The contents in each beaker were transferred to four 1-L Beckman polypropylene containers and centrifuged for 10 min at 4000 × g at 4°C. The supernatant was decanted leaving the guanidinated Nutrisoy. The Nutrisoy was washed three times to remove the excess methylisourea by resuspension in distilled water (pH adjusted to 4.5) and centrifuged and decanted, as previously described. The guanidinated Nutrisoy samples were transferred to stainless-steel trays, frozen at 20°C and freeze-dried. After drying, the batches of Nutrisoy were weighed to determine the dry matter recoveries. The resultant guanidinated material was crushed to a particle size similar to the original Nutrisoy.

Chemical analyses.  Samples of the diets, guanidinated test meals and digesta were freeze-dried and ground through a 0.5-mm mesh screen in a Wiley mill (Arther H. Thomas, Philadelphia, PA) before analyses. Dry matter contents were determined according to the Association of Official Analytical Chemists (1990). The gross energy content of the diets and test meals were determined using a Parr 1241 adiabatic oxygen bomb calorimeter (Parr Instrument, Moline, IL). The nitrogen contents were measured with an automated N analyzer (FP-428 Nitrogen Determinator, Leco, St. Joseph, MI). Analyses of chromic oxide and dysprosium in the guanidinated test meals, and ileal digesta were performed by instrument neutron activation analysis according to the procedure described by Kennelly et al. (1980). Samples of the test meals and digesta, ~5 g, were packed into 5 mL irradiation vials and irradiated at the University of Alberta SLOWPOKE II nuclear reactor. After a suitable decay period, samples were counted by measuring the 320.1 and 108.2 keV  rays emitted by the radionuclides ⁵¹Cr (t^1/2 = 27.7 d) and ^165mDy (t^1/2 = 1.258 min), respectively.

Analyses of amino acids in hydrolysates of the test meals and digesta were performed by weighing ~100 mg of sample into 13 × 100 mm screw-capped culture tubes and adding 3 mL of 6 mol/L HCl. The tubes were purged with nitrogen before sealing the screw cap and then incubated in an oven at 100°C for 24 h. Contents of amino acids were determined in duplicate aliquots of the hydrolysates by a fluorometric method involving pre-column derivatization with o-phthaldialdehyde and analysis by high-performace liquid chromatography (HPLC) according to Jones and Gilligan (1983) using a Varian 5000 HPLC system with a Varian 9090 autosampler and a Varian fluorichrom detector (excitation 340 nm, emission 450 nm; Varian Canada, Mississauga, ON). The aliquots were injected on a Supelcosil 3 micron LC-18 reverse phase column (4.6 × 150 mm; Supelco, Sigma-Aldrich Canada, Mississauga, ON) equipped with a Supelco LC-18 reverse phase 20-40 µm guard column (4.6 × 50 mm). The run time for each sample was 38 min with homoarginine eluting off the column at ~22 min, just before alanine and tyrosine. Methionine and cysteine were determined as methionine sulphone and cysteic acid, respectively, after oxidation with performic acid according to AOAC (1990). The oxidized samples were dried and then hydrolyzed and analyzed as described for the other amino acids. Peaks for amino acids were recorded and integrated with the EZchrom chromatography data system (version 2.12, Shimadzu Scientific Instruments, Columbia, MD).

The conversion of lysine to homoarginine (homo) in the guanidinated test meals was calculated as described by Rutherfurd and Moughan (1990), as follows:

(1)

Calculation of flows and digestibilities of amino acids.  The total flow of each amino acid (AA, g/kg dry matter intake) at the distal ileum was calculated using the respective amino acid (aa) concentration and chromic oxide or dysprosium as the indigestible markers (IDM), as follows:

(2)

The endogenous recovery of each amino acid (EndAA) at the distal ileum was calculated from the ratio of homoarginine (homo) to the respective amino acid concentrations in the guanidinated test meals and ileal digesta as suggested by Siriwan et al. (1994), as follows:

(3)

The exogenous (dietary) recovery of each amino acid (ExoAA) at the distal ileum was calculated as follows:

(4)

Apparent and true ileal digestibilities were expressed as percentages. Apparent digestibilities (AD) of crude protein and amino acids (including homoarginine) were calculated as follows:

(5)

The apparent digestibility of homoarginine is assumed to be similar to the true digestibility of lysine as proposed by Hagemeister and Erbersdobler (1985). True ileal digestibility (TD) of each amino acid was calculated from their respective concentration in ileal digesta as follows:

(6)

Estimates of total flow, endogenous and exogenous recoveries and apparent and true digestibilities of lysine were determined using the total of residual lysine plus homoarginine in the test meals and digesta.

Statistical analysis.  The results were subjected to analysis of variance using the general linear model procedure of the SAS Institute (1990). The statistical model included experimental periods (P) and dietary treatments (D) as main effects and their interaction with pigs within group as the source of variation: where i = 2 and j = 2. Treatment means were compared using the Student's t test according to Steel and Torrie (1980). Means were considered to be different when P < 0.05.

	RESULTS AND DISCUSSION

Abstract Introduction Results & Discussion References

The crude protein and amino acid contents of the Nutrisoy diets and corresponding guanidinated protein test meals are presented in Table 2. Recoveries of dry matter after guanidination of the Nutrisoy batches used for the preparation of UGM and AGM were 620 and 760 g/kg, respectively. The apparent loss of material from the Nutrisoy batches were assumed to consist mainly of soluble carbohydrates and some amino acids, lost during the multiple washing to remove methylisourea after guanidination. As the protein test meals were formulated using the same quantities of guanidinated Nutrisoy as with the diets, this resulted in an increase in their content of crude protein and most of the amino acids. The crude protein contents of UGM and AGM were 25 and 15 g/100 g higher, respectively, than in the unprocessed and autoclaved Nutrisoy diets. The total sum of the amino acids presented in Table 2 accounted for 91 and 90 g/100 g of the crude protein content of the unprocessed and autoclaved Nutrisoy diets, respectively. Corresponding values were 81 and 87 g/100 g for UGM and AGM. These differences suggest that methylisourea may not have been removed entirely by washing the batches of guanidinated Nutrisoy, resulting in non-amino acid nitrogen enrichment in the test meals. However, the indispensable to dispensable amino acid ratio in the diets was 0.93 and higher than values of 0.88 and 0.87 for UGM and AGM, respectively. In this respect, the concentration of most of the dispensable amino acids, particularly aspartate + asparagine and glutamate + glutamine, were increased in the protein test meals. On the other hand, the concentration of the indispensable amino acids were usually similar between the protein test meals and the respective diets. This suggests a disproportionate loss of more soluble indispensable than dispensable amino acids from the batches of guanidinated Nutrisoy. Differences in the proportion of amino acids lost after guanidination probably were related to their content in the different protein components of plant material.

The conversions of lysine to homoarginine according to Equation 1 were 46 and 30 g/100 g for UGM and AGM , respectively. The sum of the concentrations of homoarginine and residual lysine of UGM was 0.9 g/kg higher than the lysine concentration of the unprocessed Nutrisoy diet, whereas the sum of the concentrations of homoarginine and residual lysine of AGM was 1.0 g/kg lower than the lysine concentration of the autoclaved Nutrisoy diet. The concentration of arginine was also 0.9 g/kg lower in AGM compared with the respective diet. The lower concentration of residual lysine plus homoarginine in AGM compared to the lysine concentration of the autoclaved Nutrisoy is difficult to explain. Perhaps autoclaving the Nutrisoy before guanidination may have denatured the protein in a manner that caused a disproportionate loss of lysine and arginine during washing. The Nutrisoy was autoclaved using steam at 120°C, at 221 kPa, for 15 min. With the exception of lower content of lysine and cysteine, these conditions caused only minor increases in the amino acid concentration of autoclaved compared with unprocessed Nutrisoy in the diets. Nevertheless, decrease in the content of lysine in the autoclaved Nutrisoy indicates temperature sensitivity that in part would explain the lower concentration of homoarginine and residual lysine in AGM.

Hagemeister and Erbersdobler (1985) first proposed guanidination of protein to convert dietary lysine to the amino acid derivative homoarginine to directly quantify the ileal recovery of endogenous lysine in pigs. The use of guanidinated test meals involves several assumptions that have been discussed elsewhere (e.g., Marty et al. 1994, Moughan and Rutherfurd 1990, Schmitz et al. 1991). The determination of endogenous amino acid recoveries using the homoarginine ratio method assumes that guanidination does not affect the digestion and absorption of the test protein and that the absorption of homoarginine is similar lysine. It also is assumed that homoarginine is not incorporated into endogenous secretions. However, guanidinated protein test meals usually have been fed to rats (Moughan and Rutherfurd 1990), poultry (Siriwan et al. 1994) and small pigs (e.g., Barth et al. 1993, Butts et al. 1993, Schmitz et al. 1991) that require only small quantities of guanidinated protein. Rutherfurd and Moughan (1990) used dialysis against distilled water instead of multiple washing and centrifugation to remove excess methylisourea. This approach is suitable for preparation of small, but not for larger, quantities of guanidinated protein. Imbeah et al. (1996) determined the optimum conditions for guanidination of casein and soybean protein and reported a maximum conversion of 78 g/100 g in soybean protein (incubated in 0.4 mol/L methylisourea solution, at pH 10.5 and 20°C for 24 h) when the molar ratio of lysine to methylisourea was 1:10. In the present study, batches of Nutrisoy were guanidinated with a 1:6 ratio of lysine to methylisourea in 0.2 mol/L methylisourea solution (1 L of distilled water plus 1 L of 0.4 mol/L of methylisourea solution), at pH 10.3 and 4°C for 96 h. Under these conditions, the conversion of lysine residues to homoarginine was lower than reported by Imbeah et al. (1996). Interestingly, Imbeah et al. (1996) also reported 8-11 g/100 g higher conversion rates in large-scale (5 kg) compared with laboratory-scale (20 g) batches of soybean protein. They did not explain these differences in conversion, but it is likely, as seems to be the case in the present study, that soluble protein and carbohydrate were removed from soybean protein during the multiple washing to remove methylisourea. A proportionally greater loss of lysine than homoarginine residues during the procedure for guanidination of the large batches of soybean protein would account for their greater conversion rate. An alternative method of guanidinating large quantities of protein needs to be developed for studies with larger animals.

A large proportion of the lysine residues in the test protein need to be guanidinated for random distribution of homoarginine (e.g., Moughan and Rutherfurd 1990, Siriwan et al. 1994). Whether the conversion of lysine to homoarginine occurred in a uniform manner with respect to the composition of the guanidinated Nutrisoy protein is difficult to measure. Siriwan et al. (1994) assumed random distribution of homoarginine residues in casein and soybean protein based on a constant ratio between homoarginine and the other amino acids after sequential in vitro enzymatic digestion. However, Schmitz et al. (1991) reported that the in vitro rate of proteolysis of guanidinated casein by trypsin was slower (P < 0.05) than for unguanidinated casein, whereas the rate of in vitro proteolysis by chymotrypsin was not affected (P > 0.05) by guanidination. This suggests that replacing lysine with homoarginine in guanidinated protein actually may impede enzymatic digestion. Consequently, constant ratios of homoarginine to amino acids after sequential enzymatic digestion would not necessarily indicate a random distribution of the lysine derivative.

Endogenous amino acid recoveries were higher (P < 0.05) in pigs fed UGM (24.80 g/kg dry matter intake) than AGM (4.00 g/kg dry matter intake) with all individual amino acids following a similar pattern (Table 3). The endogenous amino acid recoveries ranged from 0.84 (histidine) to 2.61 (tyrosine) g/kg dry matter intake for UGM and from 0.10 (arginine) to 0.64 (aspartate + asparagine) g/kg dry matter intake for AGM. Differences in endogenous recoveries between UGM and AGM were greatest for the aromatic, branched-chain and hydroxy amino acids. The exogenous (dietary) recoveries of all amino acids were higher (P < 0.05) in pigs fed UGM compared with AGM (Table 3). The exogenous recoveries ranged from 1.30 (tyrosine) to 20.42 (glutamate + glutamine) g/kg dry matter intake for UGM and from 0.80 (histidine and tyrosine) to 9.47 (glutamate + glutamine) g/kg dry matter intake for AGM.

In the present study, it was intended that flows of amino acids would be determined using dysprosium as a specific marker in the guanidinated protein test meals. However, flows of most of the amino acids in the UGM-fed pigs, based on dysprosium (data not shown), were higher than estimates based on chromic oxide. In the AGM-fed pigs, flows of amino acids determined using dysprosium were not different (P > 0.05) from estimates based on chromic oxide. These results suggest an incomplete recovery of dysprosium from the test meal in UGM-fed pigs. Imbeah et al. (1995) discussed the possibility of dysprosium migrating between particulate matter or forming complexes with endogenous organic compounds in the digestive tract of pigs, thus affecting estimates of marker passage.

The proportion of total amino acids, from either endogenous or exogenous origin, recovered at the distal ileum of pigs, depends on the true digestibility of the protein source being investigated and the content of antinutritional factors such as SBTI. Previous studies have reported greater increases in endogenous than exogenous nitrogen recoveries in pigs fed diets with higher levels of SBTI. For instance, Barth et al. (1993) found that exogenous nitrogen represented only 7.9 and 9.2 g/100 g of total nitrogen recovered at the distal ileum of miniature pigs fed guanidinated casein meals supplemented without or with 3.0 g of purified Kunitz trypsin inhibitors, respectively. This relatively small change in exogenous nitrogen recovery is surprising given the high rate of formation of trypsin inhibitor-enzyme complexes that would be expected with the level of Kunitz trypsin inhibitors supplemented to the diet. In the present study, endogenous and exogenous nitrogen recoveries, calculated from the respective sum of amino acids, assuming 16 g nitrogen/ 100 g protein, were 22 and 78 g/100 g, respectively, for UGM-fed pigs. Corresponding recoveries for AGM-fed pigs were 10 and 90 g/100 g. These results suggest that SBTI have an inhibitory effect on the activity of pancreatic enzymes but do not cause an increase in their secretion.

Soybean trypsin inhibitors cause hypersecretion of pancreatic enzymes in rats and chicks, but this effect has not been demonstrated in pigs (Li et al. 1997b). The high content of SBTI in soybean products such as Nutrisoy has a detrimental effect on enzyme activities by forming complexes with pancreatic enzymes that result in lower apparent ileal digestibilities of amino acids in pigs (Li et al. 1997a). If Nutrisoy is autoclaved, the content of SBTI is reduced substantially, and apparent amino acid digestibilities are increased. The lower recoveries of endogenous amino acids in AGM-fed compared with UGM-fed pigs suggest that the SBTI formed complexes with pancreatic enzymes such as trypsin and chymotrypsin, thereby, increasing protein recoveries at the distal ileum. These results support the study by Li et al. (1997b), who showed that total activities of trypsin and chymotrypsin were not affected (P > 0.05) in pancreatic juice collected from pigs fed unprocessed compared with autoclaved Nutrisoy diets.

It is remarkable that other studies have not reported comparisons for amino acid content of guanidinated protein test meals and their respective diets. For this reason, it is uncertain if the differences in amino acid content between the test meals and diets are specific to Nutrisoy or to all protein sources. In this context, to ascertain whether the test meals were representative of the diets a comparison was made between the corresponding apparent amino acid digestibilities. Apparent ileal digestibilities of dry matter, crude protein and amino acids for the guanidinated test meals are presented in Table 4. The digestibility of crude protein was 33.7 percentage units higher (P < 0.05) for AGM compared with UGM. Corresponding amino acid digestibilities also were higher (P < 0.05) with differences ranging from 24.5 (aspartate + asparagine) to 61.3 (tyrosine) percentage units. With the exception of a lower (P < 0.05) methionine digestibility for UGM, the apparent digestibilities of all amino acids were similar (P > 0.05) between the guanidinated test meals and their respective diets. The apparent digestibilities of the respective diets were reported previously by Li et al. (1997a). The digestibilities of residual unguanidinated lysine in UGM and AGM were 24.8 and 65.0% (SEM = 3.3), respectively, and lower (P < 0.05) than corresponding digestibilities of lysine in their respective diets. This was expected as there would be a disproportionately greater contribution of endogenous to total lysine recovered at the distal ileum of pigs fed the guanidinated protein test meals compared with the respective diets. The similarity of apparent ileal amino acid digestibilities indicates that UGM and AGM were still representative of their respective diets even though there were some differences in amino acid content.

True ileal digestibilities of amino acids for the guanidinated test meals are presented in Table 4. True digestibilities were higher (P < 0.05) for AGM compared with UGM with differences ranging from 11.7% (tyrosine) to 38.3% (leucine). True amino acid digestibilities for UGM were higher than the corresponding apparent digestibilities. These differences ranged from 3.5% (glutamate + glutamine) to 49.9% (tyrosine). True amino acid digestibilities for AGM were also usually higher than their apparent values, although the differences were not as large and ranged from 0.6% (phenylalanine) to 6.0% (threonine). The small difference between apparent and true amino acid digestibilities of AGM-fed compared with UGM-fed pigs suggests that SBTI mainly increased the recoveries of exogenous amino acids. However, true digestibilities of lysine were not different (P > 0.05) from the corresponding apparent digestibilities of homoarginine.

Trypsin preferentially hydrolyzes peptide linkages next to the basic amino acids. If homoarginine impedes the rate of proteolysis by trypsin, then true amino acid digestibilities of the guanidinated test meals would be lower than that of the respective diets. Furthermore endogenous amino acid recoveries determined as ratios to homoarginine probably are underestimated and therefore provide minimum estimates. Exogenous recoveries measured as the difference between total and endogenous flows of amino acids would be overestimated. This limitation of the ratio method is indicated by the negative values for endogenous recoveries of arginine and phenylalanine in AGM-fed pigs (Table 3), although the large differences in endogenous amino acid recoveries between unprocessed and autoclaved Nutrisoy suggest that the ratio method can be used to determine qualitative estimates. The effect of SBTI were shown clearly by large differences in endogenous amino acid recoveries between UGM- and AGM-fed pigs. These differences followed the pattern of appearance of amino acids from protein after enzymatic hydrolysis based on the specificities of the proteases and peptidases in the intestinal tract of the pig (Low 1980). The lower content of SBTI in AGM- compared with UGM-fed pigs decreased the endogenous recoveries of the basic, aromatic, branched-chain and hydroxy amino acids to a greater extent than of the other amino acids.

In general, amino acids that have a relatively higher content in intestinal and pancreatic secretions contributed a greater proportion to the sum of endogenous amino acids recovered from UGM-fed pigs. Pancreatic secretions have a high content of the branched-chain amino acids, glycine, aspartate and glutamate (Corring and Jung 1972, Gabert et al. 1996). The same is true for serine and threonine, which have a high content in intestinal mucins where they serve as attachment sites for oligosaccharide chains, and for cysteine, which is necessary to maintain conformation of the protein core (Dekker 1990). In the AGM-fed pigs, there was not a particularly high endogenous recovery of any amino acid. Presumably, digestion of protein of pancreatic and intestinal origin was not inhibited in the small intestine of the AGM-fed pigs to the same extent as the UGM-fed pigs because of the lower content of SBTI in AGM.

In conclusion, guanidination of Nutrisoy changed the amino acid composition of the test meals with respect to the diets. The homoarginine ratio method was an effective approach that provided qualitative differences of endogenous and exogenous recoveries of amino acids from the distal ileum of pigs fed Nutrisoy diets differing in the content of SBTI. Further studies to find methods that provide quantitative estimates are warranted.

	ACKNOWLEDGMENTS

The authors acknowledge the assistance of Rick Allan, Steve Melnyk and Brenda Tchir during animal surgery and Gary Sedgwick for assistance with chemical analyses. The authors are indebted to J. Duke at the University of Alberta, SLOWPOKE II nuclear reactor facility for the analyses of dysprosium and chromic oxide in the guanidinated test meals and digesta samples.

	FOOTNOTES

Manuscript received 21 April 1997. Initial reviews completed 6 June 1997. Revision accepted 30 October 1997.

	LITERATURE CITED

Abstract Introduction Results & Discussion References

• Association of Official Chemists (1990) Official Methods of Analysis, 14th ed. Assoc. Official Anal. Chemists. Washington, DC.
• Barth C. A., Lunding B., Schmitz M., Hagemeister H. Soybean trypsin inhibitor(s) reduce absorption of exogenous and increase loss of endogenous protein in miniature pigs. J. Nutr. 1993; 123:2195-2200
• Butts C. A., Moughan P. J., Smith W. C., Carr D. H. Endogenous lysine and other amino acid flows at the terminal ileum of the growing pig (20 kg bodyweight): the effect of protein-free, synthetic amino acid, peptide and protein alimentation. J. Sci. Food Agric. 1993; 61:31-40
• Canadian Council on Animal Care (1980) Guide to the Care and Use of Experimental Animals, vol. 1 (with addendum). Ottawa, Ontario.
• Chung T. K., Baker D. H. Apparent and true amino acid digestibility of a crystalline amino acid mixture and of casein: comparison of values obtained with ileal-cannulated pigs and cecectomized cockerels. J. Anim. Sci. 1992; 70:3781-3790 [Abstract]
• Corring T., Jung J. The amino acid composition of pig pancreatic juice. Nutr. Rep. Int. 1972; 6:187-190
• Dekker, J. (1990) Gastric Mucin, Structure and Biosynthesis. Doctoral thesis, Rijksuniversiteit Utrecht, Utrecht, The Netherlands.
• de Lange C.F.M., Sauer W. C., Souffrant W. B. The effect of protein status of the pig on the recovery and amino acid composition of endogenous protein in digesta collected from the distal ileum. J. Anim. Sci. 1989; 67:755-762
• de Lange C.F.M., Sauer W. C., Souffrant W. B., Lien K. A. ¹⁵N-Leucine and ¹⁵N-isoleucine isotope dilution techniques versus the ¹⁵N-isotope dilution technique for determining the recovery of endogenous protein and amino acids in digesta collected from the distal ileum in pigs. J. Anim. Sci. 1992; 70:1848-1856 ^[Abstract]
• Gabert V. M., Sauer W. C., Li S., Fan M. Z. Exocrine pancreatic secretions in young pigs fed diets containing fababeans (Vicia faba) and peas (Pisum sativum): concentrations and flows of total, protein-bound and free amino acids. J. Sci. Food Agric. 1996; 70:256-262
• Furuya S., Kaji Y. True ileal digestibility of crude protein and amino acids in protein sources as determined by a regression method for growing pigs. Jpn. J. Zootech. Sci. 1986; 57:859-870
• Hagemeister, H. & Erbersdobler, H. F. (1985) Chemical labelling of dietary protein by transformation of lysine to homoarginine: a new technique to follow intestinal digestion and absorption. Proc. Nutr. Soc. 44: 133A (abs.).
• Imbeah M., Angkanaporn K., Ravindran V., Bryden W. L. Investigation on the guanidination of lysine in proteins. J. Sci. Food Agric. 1996; 72:213-218
• Imbeah M., Sauer W. C., Caine W. R. Comparison of the single dose and withdrawal method for measuring the rate of passage of two digestibility markers in digesta collected from the distal ileum and feces in growing pigs. Anim. Feed Sci. Technol. 1995; 52:41-50
• Jones B. N., Gilligan J. P. O-Phthaldialdehyde precolumn derivitization and reverse-phase high performance liquid chromatography of polypeptide hydrolysates and physiological fluids. J. Chromatogr. 1983; 266:471-482 [Medline]
• Kennelly J. J., Aherne F. X., Apps M. J. Dysprosium as an inert marker for swine digestibility studies. Can. J. Anim. Sci. 1980; 60:441-446
• Li, S., Sauer, W. C. & Caine, W. R. (1997a) Response of nutrient digestibilities to feeding diets of low and high levels of soybean trypsin inhibitors in growing pigs. J. Sci. Food Agric. (in press).
• Li, S., Sauer, W. C., Huang, S. & Hardin, R. T. (1997b) Response of pancreatic secretions to feeding diets with low and high levels of soybean trypsin inhibitors in growing pigs. J. Sci. Food Agric. (in press).
• Low A. G. Nutrient absorption in pigs. J. Sci. Food Agric. 1980; 31:1087-1130 [Medline]
• Maga J. A. Measurement of available lysine using the guanidination reaction. J. Food Sci. 1981; 46:132-134
• Marty B. J., Chavez E. R., de Lange C.F.M. Recovery of amino acids at the distal ileum for determining apparent and true ileal amino acid digestibilities in growing pigs fed various heat-processed full-fat soybean products. J. Anim. Sci. 1994; 72:2029-2037 [Abstract]
• Moughan P. J., Rutherfurd S. M. Endogenous flow of total lysine and other amino acids at the terminal ileum of protein- or peptide-fed rats: the chemical labelling of gelatin protein by transformation of lysine to homoarginine. J. Sci. Food Agric. 1990; 52:179-192
• National Research Council (1988) Nutrient Requirements of Swine, 9th ed. National Academy of Sciences-National Research Council, Washington, DC.
• Petersen, R. G. (1985) Design and Analysis of Experiments, 2nd ed. pp. 305-306. Marcel Dekker, New York.
• Rutherfurd S. M., Moughan P. J. Guanidination of lysine in selected dietary proteins. J. Agric. Food Chem. 1990; 38:209-211
• SAS Institute Inc. (1990) SAS/STAT User's Guide, Version 6.08. SAS Institute, Cary, NC.
• Schmitz M., Hagemeister H., Erbersdobler H. F. Homoarginine labelling is suitable for determination of protein absorption in miniature pigs. J. Nutr. 1991; 121:1575-1580
• Schulze H., Saini H. P., Huisman J., Hessing M., van den Berg W., Verstegen M.W.A. Increased nitrogen secretion by inclusion of soya lectin in the diets of pigs. J. Sci. Food. Agric. 1995; 69:501-510
• Siriwan P., Bryden W. L., Annison E. F. Use of guanidinated dietary protein to measure losses of endogenous amino acids in poultry. Br. J. Nutr. 1994; 71:515-529 [Medline]
• Steel, R.G.D. & Torrie, J. H. (1980) Principles and Procedures of Statistics: A Biometrical Approach, 2nd ed. McGraw-Hill Publishing, New York.

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences