soybean protein solubility in potassium hydroxide: an ...

Soybean protein solubility in potassium hydroxide: an in vitro test of in vivo protein quality C. M. Parsons, K. Hashimoto, K. J. Wedekind and D. H. Baker J ANIM SCI 1991, 69:2918-2924.

The online version of this article, along with updated information and services, is located on the World Wide Web at: http://jas.fass.org/content/69/7/2918

www.asas.org

Downloaded from jas.fass.org by guest on July 10, 2011

SOYBEAN PROTEIN SOLUBILITY IN POTASSIUM HYDROXIDE: AN IN VITRO TEST OF IN VIVO PROTEIN QUALITY K. J. Wedekind and D. H. Baka

C. M. Parsons', K. Hashimot$,

University of Illinois3, Urbana 61801 ABSTRACT

Experiments were conducted to assess protein solubility in .2% KOH as an indicator of soybean protein quality for chicks and pigs and to assess effects of particle size on protein of soybean meal (SBM) increased, protein solubility solubility. As the particle size (p) (%) decreased (b = -.0206). In two 9-d chick trials, dehulled SBM (48% CP) was subjected to various autoclaving times and then fed as the sole source of dietary protein to young chicks. Increasing autoclaving times from 0 to 40 min at 120°C resulted in a quadratic decrease in protein solubility. A broken-line model was fitted wherein gain:feed of chicks was plotted as a function of protein solubility. The analysis showed no reduction in feed efficiency with solubilities greater than 59 f 1.5% (mean f SEM). When solubility was below 59%,however, gain:feed decreased 1.5%for each 1% decrease in protein solubility. The third trial (13 d) was conducted with 7.5-kg pigs fed autoclaved SBM (44% CP) as the primary source of protein. Feed efficiency was significantly decreased when protein solubility was less than 66%.This study showed that protein solubility in KOH was a good index of in vivo soybean protein quality, and that it is important to standardize SBM particle size when applying the KOH assay. Key Words: Soybean Meal, Protein Solubility, Chicks, Pigs J. Anim. Sci.

1991. 69:2918-2924

lntroduction

to overprocessing (McNaughton et aL, 1981; Dale et al., 1986). An in vitro assay that is a Soybeans are among the more important good indicator of oveIprocessing of SBM is sources of dietary protein for domestic animals needed. in much of the world. Measurement of urease Evans and St. John (1945) reported that as activity, based on change in pH, is one of the raw SBM was autoclaved for increasing most widely used in vitro tests for assessing amounts of time, the proportion of protein quality of soybean meal (SBM) (AOAC, soluble in .2% KOH was reduced. An initial 1980). However, this assay is useN only for study by Araba and Dale (1990) was condetecting undercooking of SBM, because the ducted to evaluate this procedure as an urease activity rapidly falls to zero as the SBM indicator of SBM protein quality for chicks. is heated (Balloun et al., 1953). Moreover, a Their results indicated that this assay could lack of urease activity does not necessarily detect reduced soybean protein quality caused indicate overcooking and thus cannot be used by overprocessing. Thus, the present study was as an indicator of impaired protein quality due conducted to further evaluate the protein solubility assay as an indicator of in vivo soybean protein quality for both chicks and 'Author to whom reprint requests should be addressed. pigs. ~

~~

Carl M. Parsons, 322 Mumford Hall, 1301 W. Gregory Drive, Urbana, IL 61801. 'Present address: Peed Additives Department, Ajinomoto Co., Inc., Shuwa-Higaski-Yayesu Building 9-1, Haccholori 2-Chome. Chuo-Ku, Tokyo, 104, Japan. b p t . of ~ n i m .Sci. Received Augast 29, 1990. Accepted February 15, 1991.

Materials and Methods

Analytical Methods. Protein solubility was determined by the procedure described previously (Araba and Dale, 1990) as follows. Approximately 1.5 g of SBM was placed in a

291 8


SOYBEAN P R 0 " J

SOLUBILl'IY IN KOH

2919

obtained from a cross of inbred New Hampshire males and inbred Columbian Plymouth Rock females were used in both assays. At hatching, all chicks were fed a 24% CP, comSBM, pretest diet until 8 d of age, fasted overnight, weighed, and allotted to treatment (Sasse and Baker, 1974). The chicks were housed in heated, thermostatically controlled starter batteries with raised wire floors and allowed to consume feed ad libitum. A 24-h constant light schedule was maintained throughout each experiment. Commercial dehulled solventextracted SBM was heated for varying periods of time in a small laboratory autoclave to produce a range of protein solubilities. Samples were autoclaved (120'C and 103.4 kPa) for 0, 5, 10, 20, or 40 min in Exp. 1 and for 0,10,15,20 or 40 min in Exp. 2. The autoclaved SBM samples were then fed as the sole source of protein in diets (Table 1) formulated to be adequate in all essential nutrients (NRC, 1984). Each of the five diets in each experiment was fed to triplicate groups of five male chicks from 8 to 17 d posthatching. Body weight and feed intake of each group were measured at the termination of the experiments. Pig Assay. Forty4ght crossbred pigs of LandraceDuroc-Yorkshire breeding were allotted to experimental groups of three pigs each based on 28-d BW, sex, and litter origin in a randomized complete-block design. All pigs were then fed the basal diet (Table 1) for 2 d, fasted overnight, and assigned to dietary treatments at 31 d of age. Pigs were allowed to consume feed and water ad libitum, and each treatment group was housed in an expanded metal floor pen (1.2 x 1.2 m) equipped with a nipple waterer. Room temperature was main2'C. tained at 26 Commercial hulled, solventextracted SBM was cooked for 0, 10.20 or 40 min in a large industrial autoclave as described above. The autoclaved SBM samples were fed as the primary source of dietary protein (Table 1) for 13 d. Each of the four diets was fed to four replicate groups of three pigs, and BW and feed intake were measured at the end of the experiment. Statistical Analyses. Data from the chick and pig assays were subjected to ANOVA appropriate for complete randomized designs and randomized completeblock designs, r e %del PC-351, Corning Glass Works, Corning, NY. spectively (Steel and Tome, 1980). Treatment 5 ~ & 1 GPR. ~eclrman~ n s m m t palo ~ , p to, CA. responses were evaluated with orthogonal 'central soya, Gibson city, IL.

250-ml beaker and 75 ml of .2% KOH solution (wt/vol; .036 N) was added and the mixture was stirred for 20 min at 22'C at 75% of maximum velocity on a stir plate4 using a magnetic stir bar 3.6 cm in length. Approximately 50 ml of the liquid was then immediately collected and centrifuged5 at 1,250 x g for 10 min. A 15-ml aliquot of the supematant was collected and the nitrogen content of the supematant and the original SBM was determined by the Kjeldahl method (AOAC, 1980). The nitrogen values were multiplied by 6.25 to yield crude protein (CP), and protein solubility was then calculated as a percentage of the total in the original SBM sample. Urease activity in SBM was assessed with the urease index, which is based on change in pH (AOAC, 1980). Evaluation of Particle Size Efects. TWO different samples of heated, solvent-extracted, dehulled SBM (approximately 48% CP) and one sample of heated, solventextracted, hulled SBM (approximately 44% CP) were obtained from a commercial soybean processing plant6. Two assays were conducted. The first assay evaluated the effects of a few markedly different particle sizes on protein solubility of both dehulled and hulled SBM. Protein solubility was determined on dehulled SBM ground with either a coffee grinder or a highspeed analytical mill and on hulled SBM that was either unground or ground with a highspeed analytical mill. A second assay was conducted to better characterize the effects of particle size on protein solubility of another sample of dehulled SBM, thus, the SBM was ground to eight different particle sizes. The various particle sizes were produced using a coffee grinder (varying the setting for fineness of grind) or an analytical mill (varying the size of the screen). Mean particle size was determined after grinding by passing the samples through a series of sieves ranging in opening from 75 to 1,400 pm in diameter and quanwing the amount of sample retained on each sieve (Waldo et al., 1971). Protein solubility was assessed as described above. Chick Assays. Two assays were conducted to determine whether protein solubility is a good index of chick performance. Male chicks

*


2920

PARSONS ET AL. TABLE 1. COMPOSITION OF BASAL DIET (46) ~

~~

compomt

Chickassaysa

Dehulled soybean meal (48% CP)b H U Usoybean ~ meal (44% C P ) ~

48.00

Dex@OSeC SUroSeC

43.90

-

4.00 2.20 1.00

21.05 21.05 9.95 2.00 1.55 1.00

*a=# %bYwt

-

-

42.50

-

Cornstarchc

Dried whey (12% CP) corn oil Dicalcium phosphate Limestone Iodized sodium chloride DL-methionine choline cbloride (50%) Vitaminmixd Tracemineralmix& BamLmmycins (.44%)f carbadox (1.M)B

20 -

.40 20 .10

.10 .35

.10

.05

-

.M

-

25

achick and pig diets calculated to contain 23 and 20% CP, respectively. bcentra~soya, Gibson city, IL.

'A. E. Staky Manufactming Co., Decatur, E. dSee Southern and Baker (1982) for composition of chick vitamin mix and Edreonds and Bakm (1987) for composition of pig vitamin and trace mineral mixes. 'Chick mineral mix contained (46 by wt): Mn, 15.0; Fe, 15.0; Zn, 15.0; Cu, 1.0, I. .15; Se, .02. fFhvomycin: Hoechst-Rowel Agri. Vet. Co., SomrviUe, NJ. SPfizer, Inc., New York, NY.

polynomial comparisons and linear regression analysis (Steel and Tome, 1980). The feed efficiency data from the two chick assays were also calculated as a percentage of the control (nonautoclaved SBM, treatment 1) within experiment and the combined results were analyzed with a broken-line model fitted by

the method of least squares (Robbins et al., 1979; Robbins, 1986) to objectively estimate the breakpoint or Critical level of protein solubility associated with optimal feed utilization. Chick feed efficiency was the dependent variable and protein solubility was the in&pendent variable in this model.

TABLE 2. J3FFECT OF PARTICLE SIZE ON F'ROTEIN SOLUBIUTY IN POTASSIUM HYDROXIDE (ASSAY 1)

m of

soybeanmeal

Dehulleda Dehulled Hulledb Hulled Pooled SEI&

Griading rnetbod

coffee gcindep Analyticalmill' ungroundf Analyticalmill=

Mean particle size, pm 386 207 1,025 253 2.4

Protem solubility, 46

77 84

67 89 .8

'%khd1ed solvent-extracted. bSolvent-extracted with hulls. cPooled SEM calculated from duplicate determinations on the same sample. dModel G3; Bum-0-Matic Corp., Springfield, IL. %del A-10 Analytical Mill; Tekmar, Inc., C i n c h & OH. fParticle size of sample BS received from c o m m processing ~ plant with w a d d i t i ~grinding.


2921

SOYBEAN PROTEIN SOLUBILITY IN KOH TABLE 3. EFFJXT OF PARTICLE SIZE ON PROTEIN SOLUBlHTY OF DEHULLED SOYBEAN MEAL IN POTASSIUM HYDROXIDE (ASSAY 2) Grinding

method Analyticalmillb Analyticalmill Analytical mill Analyticalmill Analyticalmill

Coffee grindef Coffee grinder Coffee grinder Pooled S E I d ~~~~~

Mean particle size, pm

Rotein solubili@, %

184 25 1 299 556 599 707 831 939 3.5

89.6 83.3 81.6 79.2 77.4 76.3 73.9 70.2 .7

~

%near decrease as a function of increased particle size (P < .OO01). Equation obtained solubility on mean particle size was Y = 91.6 - .(EO6 (* .0017)X; r = .94.

from regressing protein

184, 251, 299, 556, and 599 pm mean particle sizes were produced using screen sizes of 250, 375, 550, 850, pa, respectively. Source of analytical mill was Thomas-Wiley Intermediate Mill, Thomas Scientific, Swedesboro, NJ (Model 3383-L10). %e 707,831, and 939 jun particle sizes were produced using d a m 1 fmeness of grind s e w s .Coffee m e r was same as that used in Assay 1. dPooled SEM calculated from duplicate and triplicate determinations on the same samples for mean particle size and protein sohbility, respectively. %e

and 1,700

Results

The CP contents of the SBM used in Exp. 1 to 3 were 47.7, 49.3, and 43.3%, respectively, and were unaffected by autoclaving. Mean particle sizes varied markedly by grinding method in the fnst assay. The largest particle size was obtained for unground hulled SBM and the smallest for the dehulled SBM ground with an analytical mill (Table 2). As the particle size decreased, protein solubility increased. In the second particle size assay. protein solubility decreased linearly (P < .OO01) as mean particle size of dehulled SBM increased from 184 to 939 pm (Table 3). Because of the effects of particle size, all protein solubility determinations reported for

the animal trials were performed on samples ground to approximately 210 vm with an analytical mill. No urease activity was detected in the nonautoclaved SBM evaluated in Ekp. 1 (Table 4). Increasing the autoclaving time resulted in a quadratic decrease (P c .005) in protein solubility. Rate and efficiency of weight gain of chicks fed the autoclaved SBM also decreased quadratically (P c .005) as autoclaving time increased. The nonautoclaved SBM evaluated in Exp. 2 had a urease index of .14 units of pH change (Table 5). The index decreased to zero within 10 min of autoclaving. As observed in Exp. 1, protein solubility decreased quadratically (P < .005) as autoclaving time increased. Increasing

TABLE 4. EFFECT OF AUTOCLAVING TIME ON QUALITY OF DEHULLED SOYBEAN MEAL FOR CHICKS @XP. 1) Autoclaving

time,min. 0 5 10 20 40 Pooled SEM

Urease index, units of pH increase

Rotein solubiitya, 46

Weigt P *g

fcedab,Bkg

0 0 0 0 0

84 72 63 52 36 .6

166 164 164 138 87 3

689 705 689 641 485 7

Gain/

-dratic decrease as a function of iucreased autoclaving time (P < .W). %dues are means of three groups of five male chicks from 8 to 17 d posthatching. Average initial weight was 77 g.


2922

PARSONS ET AL.

L S U 0

M

R * = .94

Protein s o l . o f SBM ( % ) Figure 1. Effect of protein solubility of soybean meal (SBM) on efficiency of feed utilization by chicks from the. combined results of Exp. 1 and 2. The results of each experiment are expressed as a percentage of the control (chicks fed SBM that was not autoclaved) and were analyzed with a broken-line regression model (Robbins. 1986) to estimate the critical level of protein solubility associated with optimal feed utilization. The breakpoint represents the point of intersection of the horizontal line @ I = 0) and the line with decreasing slope 0.The slope (b2) is the change in feed efficiency per unit decrease in protein solubility (Exp. 1, 0; Exp. 2, 0).

the autoclaving time resulted in a quadratic decrease (P < .005) in chick weight gain and feed efficiency. The results of analyzing the feed efficiency data from the two chick assays with a brokenline model are shown in Figure 1. From this analysis, it was estimated that protein solubilities of 59 f 1.5% (mean f SEM) or greater had no effect on feed utilization. The 95% confidence limit for this mean was 55 to 63%. Feed efficiency decreased linearly by 1.5% for each 1%reduction in protein solubility below 59%. The urease index of the nonautoclaved SBM used in the pig assay (Exp. 3) was .17 units of pH change and decreased to approximately zero by 10 min of autoclaving (Table 6). Protein solubility was quadratically reduced (P < .001) from 89 to 56% as autoclaving time increased. Weight gain and feed efficiency decreased linearly (P < .05) as autoclaving time increased. Dlscusslon

The results of this study show that protein solubility in KOH is a sensitive index of in vivo SBM protein quality. Under the conditions herein, the critical level of protein

solubility associated with maintaining optimal feed efficiency in chicks was estimated to be 59 f 1.5%. Protein solubilities below 59% resulted in a linear decrease in feed efficiency when expressed relative to the control. The effect of protein solubility on efficiency of feed utilization by pigs was similar to that observed with chicks, although an insufficient number of autoclaving time periods prevented precise quantification of the critical protein solubility level in pigs. However, weight gain of pigs was adversely affected when protein solubility decreased to between 66 and 71% (Table 6), whereas chick growth was unaffected at these levels. The reduction in growth rate of pigs within this protein solubility range was almost entirely due to decreased feed intake. Araba and Dale (1990) recently conducted initial experiments with chicks fed diets deficient in lysine to evaluate the protein solubility test for overprocessing of SBM. Their results indicated that growth perfoimance was significantly reduced when protein solubility of autoclaved SBM decreased to between 65 and 74% and to between 61 and 67%, respectively, in two different expenments. The authors subsequently concluded that protein solubility values below 70% are suggestive of overprocessing and decreased SBM protein quality. The higher critical protein solubility level suggested by the Araba and Dale (1990) study compared to ours (70% vs 59 f 1.5%, respectively) may be associated with the type of diet in which the overprocessed SBM was evaluated. In the Araba and Dale (1990) study, chicks were fed lowprotein, lysinedeficient diets containing the autoclaved SBM as the primary source of dietary protein. In contrast, the diets in our study were formulated to meet or exceed the NRC (1984) requirements for protein and all amino acids. Further work in our laboratory (parsons et al., 1988) has indicated that digestibilities of most amino acids (determined with cecectomized chickens) in SBM were reduced by autoclaving, whereas ME was unaffected. The effect on amino acid digestibility was much greater for lysine than for other amino acids, wherein the concentration of digestible lysine was reduced from 3.00 f .05 to 2.25 f .03 g/100 g SBM by autoclaving for 40 min. Moreover, the concentration of analytically determined lysine (ionexchange chromatography after acid hydrolysis) in the


2923

SOYBEAN PROTEIN SOLuBJLlTY IN KOH TABLE 5. EFFECT OF AUTOCLAVING TIME ON QUALITY OF DEHULLED SOYBEAN MEML FOR CHICKS @XP. 2) ~~

Gainl

Autoclaving time, min.

Urease index,units of pH i n m e

Protein

solubility",%

WeigJt gain * g

feedab, &

0 IO 15 20

.14 0 0

85 66 57 49 36 .5

157 146 147 124 68 4

706 686 665 6 18 420 12

40 Pooled SEM

0 0

BQuadratic decrease as a function of increased autoclaviug time (P < .005). bvalues are means of three groups of five male chicks from 8 to 17 d posthatching. Average initial weight was 70 g.

SBM was decreased 15% by autoclaving for 40 min, suggesting that substantial quantities of Maillard products were produced. Our study clearly shows that the protein solubility test is markedly affected by particle size of the SBM. Protein solubility increased as particle size decreased. This relationship was linear within the range of 184 to 939 pm particle sizes for dehulled SBM, although the results suggested that the effects on protein solubility may be somewhat greater at the smallest particle sizes (i.e., 184 to 251 pn). Thus, quality control laboratories must grind SBM samples to a consistent particle size to obtain repeatable results with the KOH protein solubility test. Length of stirring time and stirring speed should also be kept constant because increasing the stirring time from 20 to 40 min for a few samples of SBM increased protein solubility by 2.7% in our laboratory. The effects of autoclaving on the urease index support previous work (Balloun et al., 1953; Dale et al., 1986) showing that this test is of little or no value in detecting overprocessed SBM. As autoclaving time increased, urease activity rapidly dropped to zero

before protein quality was reduced enough to impair animal performance. Hence, a urease index value of zero does not necessarily indicate overprocessing of SBM. Urease index values substantially greater than zero, however, may indicate underprocessing of SBM (McNaughton et al., 1981). The effect of autoclaving on protein solubility differed somewhat between the chick and pig assays. Protein solubility values for d e hulled SBM were very consistent between chick experiments but were lower than those obtained in the pig study in which hulled SBM was used. The differences may have resulted from the type of SBM used or the method of autoclaving. A large industrial autoclave was used in the pig study to facilitate processing of larger quantities of SBM, whereas a small laboratory autoclave was used for the chick assays. lmpllcatlons

Soybean meal is the primary protein source used in poultry and swine diets throughout much of the world Although the urease index

TABLE 6. EPPECT OF AUTOCLAVING TIME ON QUALITY OF H I J T L J D SOYBEAN MEAL FOR PIGS (EXE'. 3) Autoclaving time, min. 0 10 20

40 Pooled SEM

Urease index, units of pH increase

Rotein solubiliv, %

Weip gain .ks

Gain/ feedb, &

.17

89

.02

71 66 56 .7

4.87 4.54 3.83 3.75 .285

680 658 65 1 599 22

0 0

decrease as a function of increased autoclaving time (P c .001). b e a r decrease as a function of increased autoclaving time (P < .05). -dratic

'Values are means of four groups of three crossbred pigs from 33 to 46 d of age. Average initial weight was 7.5 kg.


2924

PARSONS ET AL.

is widely used for detecting underprocessed soybean meal, a good in vitro test for detecting overprocessed soybean meal has not been available. The results herein indicate that protein solubility in KOH is a good in vitro test of in vivo protein quality for overprocessed soybean meal. The test is very simple, rapid, and inexpensive and should have broad application for monitoring quality of soybean products used in animal nutrition. However, it is very important that laboratory assay conditions such as sample particle size be standardized when this test is used. Llterature Cfted Arab4 M. and N. M. Dale. 1990. Evaluation of protein solubility as an indicator of overprocessingof soybean meal. Poult. Sci. 69:76. AOAC. 1980. official Methods of Analysis. (13th Ed.). Associationof official AnalyticalChemists,Washington, Dc. Balloun, S. L., E. Johnston and L. K. Arnold. 1953. Laboratory estimationof the nutritive value of soybean meals. Poult. Sci. 32517. Dale, N., 0. W. Charles and S. Duke. 1986. Reliability of urease activity as an indicator of overprocessing of soybean meal. Poult. Sci. 65(S11ppl. 1):164 (Abstr.). Edmonds, M. S. and D. H. Baker. 1987. Failure of excess dietary lysine to antagonize arginine in young pigs. J. Nutr. 11713%.

Evans, R. J. and J. L. St. John. 1945. EFtimation of the relative nutritive value of vegetable proteins by two chemical methods. J. Nu@. 30209. McNaughton, J. M., F. N. Reem and J. W. Deaton. 1981. Relationships between color, trypsin inhibitor contents, urease index of soybean meal and effects on broiler performance. Poult. Sci. 60:393. NRC. 1984. Nuhient Requirements of Poultry. (8th Ed.). National Academy Ress, Washington, DC. Parsons, C. M..K. Hashimoto, K. 1.Wedekind and D. H. Baker. 1988. Soybean protein solubility in KOH as an in vim test of in vivo proteinquality: Effects of particle size and overprocessing.Poult. Sci. 68 (Suppl. 1): 110 (Abstr.). Robbins, K. R. 1986. A method, SAS program and example for fitting the broken line to growth data. Univ. of Tennessee Res. Rep. 86-09. Univ. of Tennessee Agric. Exp. Sta., Knoxville. Robbins. K. R., H. W. Norton and D. H.Baker. 1979. Estimation of nutrient requirementsfrom growth data J. Nutr. 109:1710. Sasse, C.E.and D.H.Baker. 1974. Factors affecting sulfatesalfur utilization by the young chick. Poult. Sci. 53: 652.

Southem, L. L. and D. H. Baker. 1982. Eimeria acervulina infection in chicks fed cobalt in the presence or absence of excess dietary methionine. J. Nu*. 112: 1220.

Steel, RGD. and J. H. Tome. 1980. Principles and Proceduresof Statistics:A BiometricalApproach (2nd Ed.). McGraw-Hill Book Co., New York. Waldo, D. R., L. W.Smith, E.L. Cox, B. T. Weinland and H. L. Lucss, Jr. 1971. Logarithmicnormal distributionfor descriptionof sieved forage materials. J. Dairy Sci. 5 4 1465.


Citations

This article has been cited by 8 HighWire-hosted articles: http://jas.fass.org/content/69/7/2918#otherarticles
