The in¯uence of hydrochloric acid concentration
and measurement method on the determination
of amino acid levels in soya bean products
David M. Albin
*, Jennifer E. Wubben, Vince M. Gabert
Department of Animal Sciences, University of Illinois at Urbana-Champaign, 1207 W. Gregory Drive, Urbana, IL 61801, USA
Received 14 September 1999; received in revised form 10 March 2000; accepted 16 August 2000
Abstract
Accurate determination of amino acid levels in soya products facilitates optimum diet formulation and amino acid supplementation. A study was carried out to investigate the effect of hydrochloric acid (HCl) concentration and method of amino acid measurement on the measurement of amino acid levels. Six different soya bean products were evaluated. Hydrolysis was carried out with ®ve different acid concentrations (1, 3, 6, 9 and 12 M HCl). Amino acids were analysed by both ion-exchange chromatography with ninhydrin detection and pre-column derivatisation with
phenyl isothiocyanate. Both acid concentration and measurement method affected (P<0:05)
measurements of most amino acids. Standard hydrolysis conditions (hydrolysis in 6 M HCl at
1108C for 24 h) rarely provided the maximal amino acid values, and thus, correction factors were
calculated to standardise amino acid levels by dividing the maximum value by the value obtained with 6 M HCl. Therefore, when determining amino acid concentrations, the patterns of release and degradation of amino acids, as well as correction factors, should be considered to obtain the most
accurate values.#2000 Elsevier Science B.V. All rights reserved.
Keywords:Amino acids; Soya bean products; Measurement methods; Acid concentration
1. Introduction
The amino acid levels in feedstuffs used in animal diets must be determined to understand the nutritional value of these ingredients and to optimise diet formulations. Commonly, an acid hydrolysis of the samples is performed. Then, amino acids are
87 (2000) 173±186
*Corresponding author. Tel.:1-217-333-9749; fax:1-217-333-7088.
E-mail address: [email protected] (D.M. Albin).
measured with high pressure liquid chromatography using pre-column derivatisation with phenyl isothiocyanate (PITC), or ion-exchange chromatography (IEC) with post-column ninhydrin reaction (Bidlingmeyer et al., 1984; Elkin and Wasynczuk, 1987; Fountoulakis and Lahm, 1998).
The release of amino acids from proteins with acid hydrolysis is the most important step in the determination of amino acid concentrations. Standard hydrolysis procedures involve 24 h acid hydrolysis in 6 M HCl at 1108C (Fountoulakis and Lahm, 1998; Gehrke et al., 1985; Heinrikson and Meredith, 1984; Rowan et al., 1992). Hydrolysis time (16± 72 h) has been shown to affect the release and degradation of amino acids in a diet for growing pigs, and standard hydrolysis conditions rarely provided the maximum amino acid concentration (Rowan et al., 1992). Other acids, such as methanesulfonic acid andp -toluenesulfonic acid, have been used to hydrolyse proteins, and the effects of time and temperature on hydrolysis have been investigated (Fountoulakis and Lahm, 1998).
Soya bean products are used widely in animal nutrition (Emmert and Baker, 1995). However, the effect of acid concentration on amino acid measurements in commonly used soya protein products has not been investigated. The purposes of the present study were threefold. The ®rst purpose was to determine the effect of acid concentration on amino acid measurements in soya bean products. Secondly, the effect of method (PITC versus IEC) on amino acid measurements was examined. Finally, for amino acids levels that were not maximised with hydrolysis in 6 M HCl, correction factors were calculated.
2. Materials and methods
2.1. Procedure
Representative samples of soya bean meal from two cropping years (SBM96, SBM97), soya protein concentrate (SPC; Arcon F 65±301, Archer-Daniels-Midland, Decatur, IL), soya protein isolate (SPI; Ardex 66±960, Archer-Daniels-Midland, Decatur, IL), whole soya beans (WholeSB; Williams 82 variety) and soya bean hulls (Soya hulls) were obtained. Homogenous samples (approximately 10 g) of each were ®nely ground in a coffee bean grinder (Mr. Coffee, Model # IDS-50, Bedford Heights, OH) with approximately 10 ml liquid nitrogen for about 20 s, mixed and stored frozen atÿ108C.
2.2. Chemical analysis
Triplicate samples of each soya bean product were used to determine dry matter and crude protein (N6:25) according to procedures outlined by AOAC (1990). The
amino acids; phenyl isothiocyanate (protein sequencing grade, cat. # P-1034); and triethylamine (cat. # 13,206±3), were obtained from Sigma±Aldrich, St. Louis, MO.
Triplicate samples of SBM96, SBM97, WholeSB, and Soya hulls (at least 200 mg), and triplicate samples of SPC and SPI (at least 100 mg) were accurately weighed into screw-capped test tubes (15 ml, Pyrex, cat. # 9826±16x, Corning, NY) with te¯on-lined caps. HCl (12 ml) were added to the tubes. The acid concentrations used in this study were the following: 1, 3, 6, 9 and 12 M. The tubes were purged with N2for 10 s, mixed
and hydrolysed in an oven (Fisher Scienti®c, Model # 500 Series, Pittsburgh, PA) at 1108C for 24 h. After removal from the oven, the samples were allowed to cool. Once the samples had cooled, 0.5 ml of a-amino butyric acid (AABA, 50mM) and 0.5 ml of norleucine (Nor, 50mM) were accurately weighed on a balance and added to each tube. The tubes were inverted 200 times and centrifuged at 1100g for 10 min to pellet debris. Then, the hydrolysates were diluted in a 1:2.5 ratio by adding 300ml of distilled, deionised water to a 200ml aliquot of hydrolysate in a 1.5 ml microfuge tube (Fisher Scienti®c, cat. # 05±408±10, Pittsburgh, PA). Amino acid standards (2.5mM) were prepared by weighing individual amino acids into a 250 ml volumetric ¯ask and dissolving them in 0.1 M HCl. Three standards were used. Standard 1 consisted of Asp, Ser, Gln, Citrulline, Arg, AABA, Val, Ile, Nor and Trp. Standard 2 consisted of Glu, Asn, Taurine, Thr, Pro, AABA, Met, Leu, Nor and Ornithine. Standard 3 consisted of Hydroxyproline, Gly, His, Ala, AABA, Tyr, Cys, Nor, Phe, Lys and Homoarginine. Standards were prepared to accommodate analyses of hydrolysates and physiological samples. Samples and standards were prepared for PITC and IEC analysis.
The procedures for amino acid determination with PITC were the following: 20ml of diluted hydrolysates or 10ml of either standard 1, 2 or 3 were pipetted into polypropylene tubes (Fisher Scienti®c, cat. # 1495910AA, St. Louis, MO). They were allowed to dry under vacuum overnight in a freeze-drier (Labconco, Model # 77500, Kansas City, MO). The samples were re-dried by adding 20ml of 1:1:1 (v/v/v) methanol:water:triethylamine to each sample. The tubes were held at a 458angle and spun several times to resolubilise amino acids. The tubes were then allowed to vacuum-dry for 4 to 6 h. Finally, the samples were derivatised by adding 20ml of 7:1:1:1 (v/v/v/v) methanol:water:triethylamine:phe-nyl isothiocyanate, and mixed by holding the tubes at a 458 angle and spinning them several times. The tubes were capped and derivatisation was allowed to occur for 35 min at room temperature (228C). The samples were vacuum-dried for 4±6 h following derivatisation. The samples were then reconstituted in 200ml of sample diluent, which was composed of a mixture of 95:5 (v/v) phosphate buffer (5 mM sodium phosphate dibasic, pH 7.4, adjusted witho-phosphoric acid):acetonitrile. The samples were vortexed (Barnstead/Thermolyne, Type 16700 Mixer, Dubuque, IA), allowed to stand for approximately 15 min, and then vortexed again. A pipette was used to transfer most of the liquid to a polypropylene HPLC vial (Bio-Rad Laboratories, cat. # 223±9471, Hercules, CA), while leaving any debris at the bottom of the microfuge tube. The injection volume used ranged from 30 to 60ml depending on the protein content of the sample.
The Waters HPLC system consisted of either a 712 WISP or a 700 Satellite WISP autosampler, two 510 pumps, a column heater (468C) and a 484 tunable absorbance detector set at 254 nm. Peaks were identi®ed and integrated with Waters Maxima 820
software. The HPLC column was a Waters Pico-Tag1
3:9 mm30 cm reverse-phase column (Waters, cat. # WAT010950, Milford, MA). The packing consisted of 4mm Silica/ C18 beads. A 4:6 mm5 cm Supelcosil reverse-phase C18 guard column with 40mm Pellicular packing (Sigma±Aldrich, Supelco, cat. # 5±8232, Bellefonte, PA) was used. Two eluents were used. Eluent A consisted of 2.5:97.5 (v/v) 70 mM sodium acetate, pH 6.55:acetonitrile. Eluent B consisted of 50:35:15 (v/v/v) acetonitrile:water:methanol. Both eluents were vacuum ®ltered through a 0.45mm nylon ®lter before use. The ¯ow rate began at 1.0 ml/min. At 75 min, the ¯ow rate was increased to 1.3 ml/min. The ¯ow rate returned to 1.0 ml/min at 76 min. The gradient which was run for the separation consisted of 100% eluent A until 13.5 min, at which point the level of eluent A decreased to 97%, eluent B increased to 3% (vertical change, Waters No. 11). The level of eluent A continually decreased while eluent B increased, and this pattern is indicated in the following: 24 min, concave curve, Waters No. 9/(A, 95%: B, 5%); 30 min, convex curve, Waters No. 5/(A, 91%; B, 9%); 50 min, linear change, Waters No. 6/(A, 66%; B, 34%); 65 min, linear change, Waters No. 6/(A, 0%; B, 100%). The column was reequilibrated with 100% eluent A (linear change, Waters No. 6) at 76 min for 89 min.
The procedures for amino acid determination with IEC were the following: duplicate samples of SBM97 and SPC were utilised. Samples of hydrolysate were taken from the same tubes that were used for the PITC procedure. Following hydrolysis, addition of internal standards and centrifugation at 1100gfor 10 min, as described above, 200ml of each sample was diluted 1:5 (v/v) hydrolysate: distilled, deionised water. The diluted samples were then pH adjusted with a buffer solution of NaOH in 10 ml sodium citrate (2%, Beckman Instruments, Inc., Palo Alto, CA) in a ratio of 1:5 diluted sample:buffer solution. For samples that were hydrolysed with 6 M HCl, 0.1 g of NaOH was used. For samples that were hydrolysed with 1 and 3 M HCl, 0.01 g of NaOH was used. For samples that were hydrolysed with 9 and 12 M HCl, 0.15 g of NaOH was used. The pH of the samples was measured with litmus paper and found to be approximately 2. The samples were then analysed using post-column detection with ninhydrin using procedures that have been described previously (Spitz, 1973). It should be noted that the internal standards were not used for calculations of amino acid concentrations with IEC, but were used for PITC calculations only. External standards were used for the IEC procedure.
2.3. Data analysis
1988) was used for all calculations. Analevel of 0.05 was used to determine statistically
signi®cant differences.
3. Results
The dry matter and crude protein (g kgÿ1, dry matter basis) contents, respectively, of the soya bean samples were as follows: SBM96 (888, 536 g kgÿ1); SBM97 (885, 512 g kgÿ1
3.1. Acid concentration and amino acid composition
Mean amino acid values for the soya bean samples obtained using the various acid concentrations during hydrolysis are shown in Figs. 1±4. Most amino acid measurements were affected by acid concentration (P<0:05) regardless of soya bean
sample. Acid concentration did not affect (P>0:05) the release of the following
amino acids: lysine in SBM97; aspartic acid in SPC, SPI, WholeSB and Soya hulls; threonine in Soya hulls; and phenylalanine in Soya hulls. Most amino acids exhibited more release with acid hydrolysis in 3 M than in 1 M (P<0:05), regardless of soya
bean sample. Aspartic acid was maximised with 1 M in SBM96 and SBM97, but it was only signi®cant (P<0:05) for SBM96. Maximum valine measurements were obtained
with acid concentrations greater than 6 M, but the differences were only signi®cant (P<0:05) for SBM96, SPC and SPI. Isoleucine was also maximised at concentrations
greater than 6 M, except in SBM97 and SPC, and the differences were only signi®cant (P<0:05) for SBM96 and SPI. There was a trend for serine concentrations to be
maximised (P>0:05) with 3 M for SBM96 and Soya hulls, while threonine
concentrations were maximised with 9 M for SBM96, SBM97 and SPI. Threonine release was greater (P<0:05) with 9 M acid for SBM96 only. There was a trend
for tyrosine to be maximised (P>0:05) with 3 M in SBM96, SBM97, SPC and
WholeSB. Acid hydrolysis with 9 and 12 M degraded (P<0:05) tyrosine in all soya
bean samples. Glycine, histidine, alanine, arginine, proline, leucine, phenylalanine and lysine tended (P>0:05) to increase with increasing acid concentration for most soya
bean samples until 6 M HCl. There was a tendency for concentrations of isoleucine, leucine, phenylalanine and lysine in SBM97 to decrease (P>0:05) with hydrolysis
in acid concentrations greater than 6 M, while the concentrations of the same amino acids increased (P>0:05; P<0:05 for isoleucine) in SBM96 with hydrolysis in
acid concentrations greater than 6 M. Glutamic acid increased (P<0:05) from 1 to 3 M
in all soya bean samples, and then tended to increase or remain constant (P>0:05) from
3 to 12 M.
3.2. Measurement method and amino acid composition
The in¯uence of measurement method on amino acid measurements of the soya bean samples is shown in Table 1. With the exception of aspartic acid, most amino acid
concentrations in SBM97 were higher (P>0:05) when determined with PITC than when
determined with IEC. With the exception of aspartic acid and histidine, most amino acid concentrations in SPC were numerically higher (P<0:05) when determined with PITC
than when determined with IEC.
Fig. 1. Effect of acid concentration on the mean yield of amino acids (Yaxis, g kgÿ1dry matter basis) from soya
3.3. Correction factors
Correction factors calculated for the soya bean samples are shown in Table 2. The correction factors were determined by dividing the maximum amino acid measurement by the value obtained with hydrolysis in 6 M HCl. Correction factors were only calculated for measurements conducted with PITC. Most of the correction factors ranged from 1.00 to 1.32. However, the correction factors for isoleucine in SBM96 and lysine in Soya hulls were greater than 1.32.
Fig. 2. Effect of acid concentration on the mean yield of amino acids (Yaxis, g kgÿ1dry matter basis) from soya
bean meal 1996 (full line, shaded diamond), soya bean meal 1997 (long-dashed line, open square) and soya bean hulls (short-dashed line, shaded circle). Error bars indicate sample standard error (n3).
4. Discussion
4.1. Acid concentration and amino acid composition
It is well known that valine and isoleucine are released slowly, while serine and threonine are continually degraded, over time during acid hydrolysis (Gehrke et al., 1985; Rees, 1946; Rowan et al., 1992). Increasing acid concentration with ®xed hydrolysis time
Fig. 3. Effect of acid concentration on the mean yield of amino acids (Yaxis, g kgÿ1dry matter basis) from soya
also resulted in the slow release of valine and isoleucine, as measurements of the two amino acids increased with increasing acid concentration in the present study (Figs. 1±4). Serine was progressively destroyed from hydrolysis with 3 M to greater acid concentrations in only SBM96 and Soya hulls. However, threonine was maximised with 9 M HCl for SBM96, SBM97 and SPI. There was no indication of threonine degradation with hydrolysis using 9 M, but threonine was degraded with 12 M. This study indicates that the differences in the release and degradation of valine, isoleucine, serine and
Fig. 4. Effect of acid concentration on the mean yield of amino acids (Yaxis, g kgÿ1dry matter basis) from soya
protein isolate (full line, shaded triangle), soya protein concentrate (long-dashed line, cross) and whole soya beans (short-dashed line, open circle). Error bars indicate sample standard error (n3).
Table 1
Amino acid levels (g kgÿ1, dry matter basis) in soya bean meal (SBM97) and soya protein concentrate (SPC)
determined using 6 M HCl for acid hydrolysis and using pre-column derivitisation with phenyl isothiocyanate (PITC) or post-column detection with ninhydrin (IEC)
Amino acid SBM97 SPC
PITC IEC Sig.a PITC IEC Sig.
Aspartic acid 48.82.30b 51.50.20 NS 71.03.10 71.10.20 NS
Correction factorsafor amino acids in soya bean meal (SBM96, SBM97), soya protein concentrate (SPC), soya
protein isolate (SPI), whole soya beans (WholeSB) and soya bean hulls (Soya hulls)
Amino acid SBM96 SBM97 SPC SPI WholeSB Soya hulls
Aspartic acid 1.19 1.08 1.02 1.03 1.02 1.14
Glutamic acid 1.08 1.00 1.01 1.00 1.00 1.17
Serine 1.05 1.00 1.00 1.02 1.00 1.02
Glycine 1.10 1.06 1.00 1.00 1.17 1.08
Histidine 1.10 1.02 1.07 1.07 1.00 1.16
Threonine 1.05 1.00 1.00 1.02 1.00 1.00
Alanine 1.11 1.02 1.00 1.03 1.03 1.29
Arginine 1.08 1.02 1.01 1.00 1.00 1.16
Proline 1.07 1.08 1.00 1.02 1.12 1.00
Tyrosine 1.03 1.04 1.00 1.00 1.02 1.00
Valine 1.17 1.05 1.19 1.11 1.04 1.32
Isoleucine 1.45 1.00 1.00 1.11 1.09 1.15
Leucine 1.06 1.00 1.01 1.03 1.00 1.01
Phenylalanine 1.11 1.00 1.00 1.02 1.00 1.12
Lysine 1.21 1.00 1.00 1.04 1.02 1.65
aCorrection factors determined by expressing the maximum amino acid yield as a proportion of the yield
threonine are, at least in part, dependent on protein source. Other researchers (Glazer et al., 1976; Rowan et al., 1992) have attributed differences in degradation and release of amino acids to protein source. Tyrosine, reported to be susceptible to oxidation (Finley, 1985; Gehrke et al., 1985), was degraded with the use of 9 and 12 M HCl. However, similar concentrations of tyrosine were obtained with 6 M HCl relative to those obtained with 3 M HCl, although some of the values were slightly lower at 6 M compared to 3 M HCl. Rowan et al. (1992) reported similar observations concerning tyrosine, and attributed the stability to the addition of phenol. In the present study, however, phenol was not added prior to hydrolysis, and tyrosine was essentially not degraded with acid concentrations as strong as 6 M HCl. Aspartic acid tended to decrease as acid concentration increased from 1 M for samples of SBM96, SBM97 and Soya hulls, indicating mild sensitivity of aspartic acid to acid degradation in these samples. Similar patterns were not observed for SPC, SPI and WholeSB.
The relative stability or gradual increase in measurements of aspartic acid, glycine, histidine, alanine, arginine, proline, leucine, phenylalanine and lysine with increasing acid concentration used during this study re¯ected the lack of sensitivity of these amino acids to increasing acid concentration, even using 12 M HCl.
4.2. Measurement method and amino acid composition
Different investigators have evaluated the two amino acid determination methods used in this study, and they have reported that the two procedures provide very similar results (Bidlingmeyer et al., 1984; Elkin and Wasynczuk, 1987; Heinrikson and Meredith, 1984). However, in the current study, method affected amino acid measurements. For many amino acids, in both SBM97 and SPC using hydrolysis in 6 M HCl, the PITC procedure provided signi®cantly higher measurements. Also, it should be noted that the hydrolysis curves for SBM97 and SPC determined with IEC followed the same patterns of release as those shown in Figs. 1±4 (data not shown). Comparisons with published amino acid values have indicated that the measurements determined with PITC and IEC were comparable for amino acid levels in soya bean samples. For example, using PITC and IEC in this study, the threonine content of SBM97 was 25.8 and 18.5 g kgÿ1
, respectively (dry matter basis). Other publications (Cavins et al., 1972; Emmert and Baker, 1995; NRC, 1998; Rudolph et al., 1983) have reported threonine concentrations in soya bean meal of 21.5, 20.6 and 18.6 g kgÿ1, respectively (dry matter basis). The threonine content of SPC using PITC and IEC was found to be 28.1 and 25.2 g kgÿ1, respectively (dry matter basis). Emmert and Baker (1995) and NRC (1998) reported threonine concentrations in SPC of 27.3 and 31.1 g kgÿ1, respectively (dry matter basis). There are several possible explanations for the differences in amino acid measurements determined with the two methods in the current study. Many steps of the IEC procedure demanded accurate volumetric measurements. Before amino acids were determined with the IEC procedure, the hydrolysate needed to be diluted and buffered. This was done in two different steps. As part of this procedure, external standards were used and, therefore, dilution of standards and injection volumes needed to be accurate. The PITC procedure did not require accurate volumetric measurements because internal standards and tube weighing were used. Therefore, more opportunities for error existed with the IEC
procedure than when the PITC procedure was used. Despite this, it is interesting to note that less variation due to method existed for amino acids in SPC compared to SBM97. During soya bean processing, SPC is the result of further processing than soya bean meal. Thus, SPC is more puri®ed and re®ned than soya bean meal, and, therefore, may contain fewer factors than soya bean meal that could alter the determination of amino acids. Even though both procedures provided values for most amino acids that are within the range found in the literature, further research is necessary to increase the accuracy and decrease the variation from procedure-to-procedure.
4.3. Correction factors
Garnett (1985) suggested the use of generalised correction factors as a simple approach to correcting amino acid concentrations. Correction factors have been calculated by others (Kohler and Palter, 1967; Rowan et al., 1992; Slump, 1980; Tkachuk and Irvine, 1969) to correct amino acid measurements determined with 24 h hydrolysis to the maximum values obtained with different hydrolysis times. Recently, research projects have begun to use correction factors to obtain amino acid measurements that are more accurate (Lenis et al., 1990; Mroz et al., 1994). However, correction factors have not been determined to correct the values obtained with hydrolysis in 6 M HCl to the maximum values obtained with other acid concentrations. For hydrolysis time, correction factors for serine, isoleucine and threonine in foods have been reported (Kohler and Palter, 1967; Rowan et al., 1992; Slump, 1980; Tkachuk and Irvine, 1969) and ranged from 1.04 to 1.14, 1.02 to 1.21 and 1.02 to 1.08, respectively. In the present study, correction factors for serine, isoleucine and threonine ranged from 1.02 to 1.05, 1.09 to 1.15 and 1.02 to 1.05, respectively. Rowan et al. (1992) and Slump (1980) have reported correction factors for 24 h hydrolysis for valine of 1.20 and 1.08, respectively. In the present study, valine correction factors ranged from 1.04 to 1.32. Most of the correction factors in this study were similar to those determined with different hydrolysis times. For isoleucine in SBM96 and lysine in Soya hulls, the correction factors were 1.45 and 1.65, respectively. These should be used with caution, as comparison correction factors would suggest that these factors are abnormally high (Rowan et al., 1992). However, there is nothing to indicate that analytical error produced abnormal measurements of these amino acids.
Acknowledgements
The authors would like to thank Zoran Magas of LC Tech Services, Burlington, Ont., Canada for technical advice and maintenance of the high pressure liquid chromatography equipment used in this study.
References
Association of Of®cial Analytical Chemists, 1990. Of®cial Methods of Analysis, 15th Edition. AOAC, Arlington, VA, USA, 1298 pp.
Bidlingmeyer, B.A., Cohen, S.A., Tarvin, T.L., 1984. Rapid analysis of amino acids using pre-column derivatization. J. Chromatogr. 336, 93±104.
Cavins, J.F., Kwolek, W.F., Inglett, G.E., Cowan, J.C., 1972. Amino acid analysis of soybean meal: interlaboratory study. J. Assoc. Off. Anal. Chem. 55, 686±691.
Elkin, R.G., Wasynczuk, A.M., 1987. Amino acid analysis of feedstuff hydrolysates by precolumn derivatization with phenylisothiocyanate and reversed-phase high-performance liquid chromatography. Cereal Chem. 64, 226±229.
Emmert, J.L., Baker, D.H., 1995. Protein quality assessment of soy products. Nutr. Res. 15, 1647±1656. Finley, J.W., 1985. Reducing variability in amino acid analysis. In: Finley, J.W., Hopkins, D.T. (Eds.),
Digestibility and Amino Acid Availability in Cereals and Oilseeds. American Association of Cereal Chemists, St. Paul, MN, USA, pp. 15±30.
Fountoulakis, M., Lahm, H.-W., 1998. Hydrolysis and amino acid composition analysis of protein. J. Chromatogr. 826, 109±134.
Garnett, N.K., 1985. The in¯uence of amino acid hydrolysis times on the determination of amino acid requirements for poultry. In: Ideris, A., Hamid, R. (Eds.), Proceedings of the Regional Seminar on Future Developments in the Poultry Industry. World Poultry Science Association, Malaysia, pp. 14±17. Gehrke, C.W., Wall Sr, L.L., Absheer, J.S., Kaiser, F.E., Zumwalt, R.W., 1985. Sample preparation for
chromatography of amino acids: acid hydrolysis of proteins. J. Assoc. Off. Anal. Chem. 68, 811±821. Glazer, A.N., Delange, R.J., Sigman, D.S., 1976. Chemical modi®cation of proteins. In: Work, T.S., Work, E.
(Eds.), Laboratory Techniques in Biochemistry and Molecular Biology. North-Holland, Amsterdam, pp. 1±205.
Heinrikson, R.L., Meredith, S.C., 1984. Amino acid analysis by reverse-phase high-performance liquid chromatography: precolumn derivatization with phenylisothiocyanate. Anal. Biochem. 136, 65±74. Kohler, G.O., Palter, R., 1967. Studies on methods for amino acid analysis of wheat products. Cereal Chem. 44,
512±520.
Lenis, N.P., Van Diepen, J.Th.M., Goedhart, P.W., 1990. Amino acid requirements of pigs. Part 1. Requirements for methioninecystine, threonine and tryptophan of fast-growing boars and gilts, fed ad libitum. Neth. J. Agric. Sci. 38, 577±595.
Mroz, Z., Jongbloed, A.W., Kemme, P.A., 1994. Apparent digestibility and retention of nutrients bound to phytate complexes as in¯uenced by microbial phytase and feeding regimen in pigs. J. Anim. Sci. 72, 126± 132.
National Research Council, 1998. Nutrient Requirements of Swine, 10th Edition. National Academy Press, Washington, DC, USA, 198 pp.
Ott, L., 1993. Multiple comparisons. In: Lozyniak, K., Krikorian, S. (Eds.), An Introduction to Statistical Methods and Data Analysis, 4th Edition. Wadsworth, Belmont, CA, pp. 807±841.
Rees, M.W., 1946. The estimation of threonine and serine in proteins. Biochem. J. 40, 632±640.
Rowan, A.M., Moughan, P.J., Wilson, M.N., 1992. Effect of hydrolysis time on the determination of amino acid composition of diet, ileal digesta, and feces samples and on the determination of dietary amino acid digestibility coef®cients. J. Agric. Food Chem. 40, 981±985.
Rudolph, B.C., Boggs, L.S., Knabe, D.A., Tanksley Jr, T.D., Anderson, S.A., 1983. Digestibility of nitrogen and amino acids in soybean products for pigs. J. Anim. Sci. 57, 373±386.
Slump, P., 1980. Symposium on protein metabolism and nutrition. In: Proceedings of the 3rd European Association for Animal Production. European Association for Animal Production, Wageningen, The Netherlands, p. 14.
Spitz, H.D., 1973. A new approach for sample preparation of protein hydrolyzates for amino acid analysis. Anal. Biochem. 56, 66±73.
Statistical Analysis Systems Institute Inc., 1988. SAS/STAT User's Guide, Release 6.03. SAS Institute Inc., Cary, NC, USA, 956 pp.
