DPP

ED

R

ntbAeprBa0otaroTpioPssr

t

pihst

9

2

Archives of Biochemistry and BiophysicsVol. 365, No. 2, May 15, pp. 262–267, 1999Article ID abbi.1999.1184, available online at http://www.idealibrary.com on

ifferent Sensitivity of Recombinant Aspergillus nigerhytase (r-PhyA) and Escherichia coli pH 2.5 Acidhosphatase (r-AppA) to Trypsin and Pepsin in Vitro

ric Rodriguez, Jesus M. Porres, Yanming Han, and Xin Gen Lei1

epartment of Animal Science, Cornell University, Ithaca, New York 14853

eceived November 20, 1998, and in revised form February 19, 1999

16phciptipgcfiiiilmt(Mwtcpbr

temipaz

Proteolysis of two purified recombinant enzymes,amely, the Aspergillus niger phytase (r-PhyA) andhe Escherichia coli pH 2.5 acid phosphatase (r-AppA),y pepsin and trypsin was investigated in this study.fter r-PhyA and r-AppA were incubated with differ-nt concentrations of pepsin or trypsin, their residualhytase activities and amounts of inorganic phospho-us released from soybean meal were determined.oth enzymes retained more than 85% of their originalctivities at the trypsin/phytase ratios (w/w) 0.001 and.005, while r-AppA and r-PhyA lost 60 and 20% of theriginal activity at the ratio of 0.01 or 0.025, respec-ively. In contrast, there was a 30% increase in phytasectivity after r-AppA was incubated with pepsin at theatios of 0.005 or 0.01. Meanwhile, r-PhyA lost 58 to 77%f its original activity under the same conditions.rypsin and pepsin affected the hydrolysis of phytatehosphorus from soybean meal by r-AppA and r-PhyA

n a similar way to their residual phytase activities. Allf these in vitro proteolyses were confirmed by SDS–AGE analysis. Our results demonstrate different sen-itivities of r-AppA and r-PhyA to trypsin and pepsin,uggesting active trypsin resistant r-PhyA and pepsinesistant r-AppA polypeptides. © 1999 Academic Press

Key Words: phytase; acid phosphatase; proteolysis;rypsin; pepsin; recombinant enzyme.

Phytases are myo-inositol hexakisphosphate phos-hohydrolases that catalyze the stepwise removal ofnorganic orthophosphate from phytate (myo-inositolexakisphosphate) (1). There are two types of phyta-es. One is called 3-phytase (EC 3.1.3.8), which ini-iates the removal of phosphate groups at the positions

1

To whom correspondence should be addressed. Fax: (607) 255-829. E-mail: XL20@cornell.edu.

su

62

and 3 of the myo-inositol ring. The other is called-phytase (EC 3.1.3.26), which first frees the phos-hate at the position 6 of the ring. While no phytaseas been identified from animal tissues, plants usuallyontain 6-phytases, and a broad range of microorgan-sms, including bacteria, filamentous fungi, and yeast,roduce 3-phytases (2–9). Because more than 70% ofhe total phosphorus in foods or feeds of plant origin isn the form of phytate that is poorly available to sim-le-stomached animals and humans, phytases are ofreat uses in improving mineral nutrition of these spe-ies (10–16). Supplemental microbial phytases in dietsor swine and poultry effectively enhance bioavailabil-ty of phytate phosphorus and reduce the need fornorganic phosphorus supplementation (11–15), result-ng less phosphorus pollution in areas of intensive an-mal production (8–15). However, a relatively highevel of phytase supplementation is necessary in ani-

al diets (10–16), because a considerable amount ofhe enzyme is degraded in stomach and small intestine13), probably by proteolysis of pepsin and trypsin.

eanwhile, the proteolytic profiles of various phytasesere not studied. Clearly, a better understanding of

heir sensitivities to trypsin and pepsin hydrolysisould be helpful for improving the nutritional value ofhytases. Aspergillus niger phytase gene (phyA)2 haseen overexpressed in its original host (17) and theecombinant enzyme (r-PhyA, EC 3.1.3.8) has been

2 Abbreviations used: appA, Escherichia coli pH 2.5 acid phospha-ase gene; BAEE, Na-benzoyl-L-arginine ethyl ester; BMGY, buff-red glycerol–complex medium; BMMY, buffered methanol–complexedium; DEAE, diethylaminoethyl; dNTPs, deoxynucleotides; iP,

norganic phosphorus; PCR, polymerase chain reaction; phyA, As-ergillus niger phytase gene; r-AppA, recombinant enzyme of theppA gene expressed in Pichia pastoris; r-PhyA, recombinant en-yme of the phyA gene expressed in A. niger; SDS, sodium dodecyl

ulfate; YPD, yeast extract peptone dextrose medium; PU, phytasenit.

0003-9861/99 $30.00Copyright © 1999 by Academic Press

All rights of reproduction in any form reserved.

uTEholmnreadtv

E

Maimf((tos(vpvTtiYmdcieassTp

geatampdSPttct

ra

(waBmvPrrtd(p

booadppttIrmT

R

nepaaScTpr

bwa1t0aMaedct(id

b

263SENSITIVITY OF RECOMBINANT ACID PHOSPHATASES TO TRYPSIN AND PEPSIN

sed in animal diets as a commercial phytase (13, 14).his enzyme is a glycoprotein of approximately 80 kDa.scherichia coli pH 2.5 acid phosphatase gene (appA)as also been characterized (18, 19). Recently, we haveverexpressed appA in Pichia pastoris as an extracel-ular glycoprotein of approximately 55 kDa. Our ani-

al experiments have demonstrated that the recombi-ant enzyme (r-AppA, EC 3.1.3.2) is as effective as-PhyA in releasing phytate phosphorus in animal di-ts (14). Thus, we consider r-AppA another nutrition-lly useful phytase. These considerations have led us toetermine the effect of trypsin and pepsin on the ac-ivities of the partially purified r-PhyA and r-AppA initro.

XPERIMENTAL PROCEDURES

Expression of r-AppA. The appA gene (Genebank Accession No.58708) was obtained from the E. coli BL21 strain transformed by

n expression vector pAPPA1 (20). A 1.35-kb DNA fragment contain-ng the coding region of appA was amplified by PCR following the

anufacturer instructions (Perkin–Elmer). Primers were derivedrom 59 and 39 regions of the nucleotide sequence (18) and include E2forward, 242–252), 59GGAATTCCAGAGTGAGCCGGA39; and K2reverse, 1468–1490), 59GGGGTACCTTACAAACTGCACG39. Thesewo primers were synthesized by the Cornell University Oligonucle-tide Synthesis Facility (Ithaca, NY). The amplified product wasliced from a 1% agarose gel and eluted with a GENECLEAN II kitBio101). The purified fragment was first cloned into pGEM T-easyector (Promega) and then inserted into the yeast expression vectorPIcZaA (Invitrogen) at the EcoRI site. E. coli strain TOP10F9 (In-itrogen) was used as an initial host to amplify these two constructs.he pPIcZaA vector containing appA was transformed into P. pas-

oris strain X33 by electroporation according to the manufacturer’snstructions (Invitrogen). The transformed cells were plated intoPD-Zeocin agar medium and positive colonies were incubated ininimal medium with glycerol (BMGY) for 24 h. When the yeast cell

ensity reached 2.5 3 108 cells/ml (OD600 5 5), and the cells wereentrifuged and suspended in 0.5% methanol medium (BMMY) tonduce the appA gene expression. Total yeast genomic DNA wasxtracted from the transformed X33 cells after induction and used astemplate to check the presence of the appA gene by PCR using the

ame primers as described above. The amplified DNA fragment wasequenced at the Cornell University DNA Services-Facility usingaq Cycle automated sequencing with Dye Deoxy terminators (Ap-lied Biosystems, Forster City, CA).Purification of r-PhyA and r-AppA. We obtained r-PhyA as a

enerous gift from BASF (Mt Olive, NJ). Both r-PhyA and r-AppAnzymes were initially suspended into 50 mM Tris–HCl, pH 7, andmmonium sulfate was added to 25% of saturation. After the mix-ure was centrifuged (25,000g, 20 min), the supernatant was savednd ammonium sulfate was added to 75% of saturation. Then, theixture was centrifuged (25,000g, 20 min) and the pellet was sus-

ended into 10 mL of 25 mM Tris–HCl, pH 7. The suspension wasialyzed overnight against the same buffer and loaded onto a DEAE–epharose column (Sigma) equilibrated with 25 mM Tris–HCl, pH 7.roteins were eluted with 0.2 M NaCl, 25 mM Tris–HCl, pH 7, afterhe column was washed with 200 mL of 25 mM Tris–HCl, pH 7. Allhe collected fractions were assayed for phytase activity and proteinoncentration (21). The whole purification was conducted at 4°C, andhe fractions were stored at 220°C before analysis.

Proteolysis and protein electrophoresis. The purified r-AppA and

-PhyA (2 mg/mL) were incubated with different amounts of pepsinnd trypsin following the manufacturer instructions (Sigma). Pepsin u

800 units/mg protein) and trypsin (1500 BAEE units/mg protein)ere dissolved into 10 mM HCl, pH 2 (0.1 mg/mL) and 80 mMmmonium bicarbonate, pH 7.5 (0.1 mg/mL), respectively. OneAEE unit was defined as 0.001 absorbance change at 253 nm perinute at pH 7.6 and 25°C, with BAEE as a substrate. In a final

olume of 100 mL, 10 mg of purified r-PhyA (0.1 PU) or r-AppA (0.08U) was incubated with trypsin or pepsin at protease/phytase (w/w)atios ranging from 0.001 to 0.01, at 37°C for 1 to 120 min. Theeaction was stopped on ice and the pH of the mixture was adjustedo 8 for protein electrophoresis and phytase activity assay. Theigested protein mixtures were analyzed by sodium dodecyl sulfateSDS)–polyacrylamide or urea–SDS–polyacrylamide gel electro-horesis as previously described (22, 23).Phytase activity and hydrolysis of phytate phosphorus from soy-

ean meal. Phytase activities of both r-PhyA and r-AppA, prior tor at various time points of proteolysis, were determined as previ-usly described (24). The released inorganic phosphorus (iP) wasssayed by the method of Chen et al. (25). One phytase unit (PU) wasefined as the activity that releases 1 mmol of iP from sodium phytateer minute at 37°C. To confirm the proteolytic effects of trypsin andepsin on the residual activities of r-PhyA and r-AppA, we monitoredhe hydrolysis of phytate phosphorus from soybean meal by thesewo enzymes incubated with different amounts of trypsin or pepsin.n a 5-mL total reaction, 0.5 mg of the purified r-PhyA (5 PU) or-AppA (4 PU) was incubated with 1 g soybean meal and pepsin in 10M HCl, pH 2.5 or trypsin in 0.2 M citrate, pH 6.8 at 37°C for 2 h.he released iP was determined as described above.

ESULTS

Preparation of r-AppA and r-PhyA. Over 30 colo-ies of X33 transformed with the appA gene expressedxtracellular phytase activity that hydrolyzes sodiumhytate. Colony 26 had the highest activity (88 U/mL)nd was chosen for further studies. After the r-PhyAnd the r-AppA samples were eluted from the DEAE–epharose column, 45 fractions of 4 mL each wereollected for both enzymes to assay for phytase activity.he fractions used for proteolysis had a specifichytase activity of 9.6 and 8.1 U/mg of protein for the-PhyA and r-AppA, respectively.Effects of trypsin digestion on the phytase activities of

oth enzymes. After 2 h of trypsin digestion, thereere significant differences in the residual phytasectivities between the r-PhyA and the r-AppA (Fig.A). Although both enzymes retained more than 85% ofheir original activities at the trypsin/phytase ratios of.001 and 0.005, r-AppA lost 64 and 74% of its originalctivity at the ratio of 0.01 and 0.025, respectively.eanwhile, r-PhyA lost only 14 and 23% of its original

ctivity, respectively. Because of the apparent differ-nce in sensitivities of these two enzymes to trypsinigestion at the ratio of 0.01, a time course study wasonducted with this ratio. Up to 2 h of trypsin diges-ion, r-PhyA still retained 90% of its original activityFig. 2). In contrast, r-AppA lost 64, 77, 87, and 95% ofts original activity after 1, 5, 30, and 120 min ofigestion, respectively.Effect of pepsin digestion on the phytase activities of

oth enzymes. After 2 h of pepsin digestion, the resid-

al phytase activity of r-AppA was totally unexpected.

AeteHacatpifra

Awaptodraovuw

ardf

lbtf0Mcpmicpr

D

oprma

Far0 he0

264 RODRIGUEZ ET AL.

t the ratios of 0.001 and 0.002, the phytase activityither remained unchanged or slightly increased. Athe ratios of 0.005 and 0.01, the phytase activity wasnhanced by 30% compared with the initial value.owever, r-PhyA lost 58 and 77% of its original activityt these two high ratios (Fig. 1B). Because of signifi-antly different responses between r-PhyA and r-AppAt the ratio of 0.005, this ratio was used for a follow-upime course study. There was a stepwise increase inhytase activity along with time when the r-AppA wasncubated with pepsin from 0 to 30 min. Thereafter, nourther increase was observed (Fig. 2). However,-PhyA lost 42, 73, 82, and 92% of its original activityfter 1, 5, 30, and 120 min of incubation, respectively.SDS–polyacrylamide gel electrophoresis. When r-ppA was incubated with trypsin, the enzyme proteinas degraded at the ratios above 0.01 and was invisiblet the ratio of 0.025. There was a major band of ap-roximately 28 kDa, with several other bands betweenhis band and the intact protein in the three low ratiosf trypsin. However, that major band was clearly re-uced and the other bands disappeared at the highestatio of trypsin (Fig. 3A). There were many intermedi-ry bands when the r-PhyA was incubated with vari-us amounts of trypsin and there were at least threeisible bands at the highest ratio of trypsin (Fig. 3B). A

IG. 1. (A) Phytase activity changes of r-PhyA and r-AppA incubatnd 0.025). Symbols: r-PhyA (■) and r-AppA (F). The results are mean-PhyA or r-AppA control. (B) Phytase activity changes of r-PhyA a.001, 0.002, 0.005, and 0.01). Symbols: r-PhyA (h) and r-AppA (E). T.01 versus untreated r-PhyA or r-AppA control.

nique band of approximately 8.4 kDa was shownhen r-AppA was incubated with pepsin at the ratio p

bove 0.002 (Fig. 4A). On the other hand, proteolysis of-PhyA by various amounts of pepsin resulted in manyiffused and smearing bands, in addition to a majorragment of approximately 14 kDa (Fig. 4B).

Effects of proteolysis on phytate–phosphorus hydro-ysis by r-PhyA and r-AppA. When r-AppA was incu-ated with soybean meal and different amounts ofrypsin for 2 h at 37°C, the reduction in iP releasedrom soybean meal was 3, 13, 34, and 52%, at ratios of.001, 0.005, 0.01, and 0.025, respectively (Fig. 5A).eanwhile, the reduction for r-PhyA under the same

onditions was 3, 6, 13, and 28%, respectively. Addingepsin to r-AppA (ratio $ 0.005) and the soybean mealixture resulted in an approximately 30% increase in

P released from soybean meal, compared with theontrol (Fig. 5B). In contrast, the same treatmentsroduced more than 50% reduction in iP release by-PhyA.

ISCUSSION

To our knowledge, there have been no specific datan sensitivities of microbial phytases to trypsin andepsin. In this study, we exposed two partially purifiedecombinant phytases to single protease digestions andeasured the effects of proteolysis on their residual

ctivities and their capacity of releasing phytate phos-

with different ratios of trypsin/protein (w/w) (r 5 0.001, 0.005, 0.01,SE from five independent experiments. *P , 0.01 versus untreated

r-AppA incubated with different ratios of pepsin/protein (w/w) (r 5results are means 6 SE from seven independent experiments. *P ,

eds 6

nd

horus from soybean meal. Our results have demon-

slp

atrafa1owRXpt

swsPsistasuaett

Fsmd(ri

F(Tf

265SENSITIVITY OF RECOMBINANT ACID PHOSPHATASES TO TRYPSIN AND PEPSIN

trated that r-PhyA is more resistant to trypsin andess resistant to pepsin than r-AppA. The proteolyticatterns of these two phytases, shown by SDS–PAGE

IG. 2. Residual phytase activity of r-PhyA and r-AppA after tryp-in or pepsin hydrolysis during a time course (0, 1, 5, 30, and 120in). Symbols: trypsin-digested r-PhyA (■) or r-AppA (F); pepsin-

igested r-PhyA (h) and r-AppA (E). The ratio of trypsin/phytasew/w) used was r 5 0.01 (w/w). The ratio of pepsin/phytase used was5 0.005. The results are means 6 SE from six independent exper-

ments. *P , 0.01 versus untreated r-PhyA or r-AppA control.

IG. 3. SDS–polyacrylamide gel electrophoresis of r-AppA (12%, Ar 5 0.001, 0.005, 0.01, and 0.025 (w/w)). Proteins were stained using

he protein marker (M) is a 10-kDa ladder (10, 20, 30, 40, 50, 60, 70, 80, 9

rom four independent experiments.

nalysis, are also distinctly different. Presumably,hese different susceptibilities to proteases between-PhyA and r-AppA may be associated with their char-cteristics of primary amino acid sequence and peptideolding, because there is a low homology (;15%) ofmino acid sequences between these two enzymes (17,8). However, caution should be given in considerationf the molecular mechanism of phytase proteolysis,hich is beyond the original scope of the present study.ecent progress in crystallization and (or) preliminary-ray analysis of the phyA phytase (26) and an E. colihytase (27) would help us in understanding the struc-ural basis for their proteolytic responses.

Unexpectedly, r-AppA showed a 30% increase in re-idual phytase activity after pepsin digestion. Like-ise, this enzyme also released 30% more iP from

oybean meal in the presence of pepsin. From the SDS–AGE analysis, r-AppA was clearly degraded intomall peptides by pepsin along different periods ofncubation. Likely, there may be potential pepsin-re-istant polypeptides with higher phytase activity thanhe intact r-AppA protein. Although the SDS–PAGEnalysis did not offer us any specific information onuch peptides, pepsin has been shown to convert nat-ral or synthetic proteins in active polypeptides, suchs converting porcine endothelin to active 21-residuendothelin (28). Pepsin may also modulate the struc-ure and functions of certain proteins (29, 30). As men-ioned above, the availability of the recent crystalliza-

r-PhyA (20%, B)-digested products by different amounts of trypsinomasie blue. T, trypsin control; C, purified r-AppA (A) or r-PhyA (B).

) orCo

0, 100, 110, 120, and 200 kDa; Gibco). The results are representative

twdupdpgdttt

ur

tptspwHsl

F0p 90f

For

266 RODRIGUEZ ET AL.

ion data on the phyA (26) and the E. coli phytases (27)ould facilitate targeting site-directed mutageneses oreletions of the appA gene. Thereby, we may be able tonveil the molecular mechanism for the increase ofhytase activity of r-AppA associated with pepsin hy-rolysis. Despite the biochemical uncertainty of theepsin-resistant r-AppA peptides, our finding has areat nutritional implication. Because pepsin, a well-escribed aspartic protease, is the major protease inhe stomach (31), a pepsin-resistant phytase polypep-ide could allow us to supplement a low level of enzymeo the diets with sufficient activity. Thus, expense for

IG. 4. SDS–polyacrylamide gel (20%) electrophoresis of r-AppA (A.001, 0.002, 0.005, and 0.01 (w/w)). Proteins were stained using Coorotein marker (M) is a 10-kDa ladder (10, 20, 30, 40, 50, 60, 70, 80,rom six independent experiments.

IG. 5. Amounts of inorganic phosphorus (iP) released from soybea

f trypsin (r 5 0.001, 0.005, 0.01, and 0.025 (w/w)) (A) or pepsin (r 5 0.00esults are means 6 SE from three independent experiments. *P , 0.0

se of dietary phytase in animal production will beeduced.It is difficult to compare the activity levels of pro-

eases used in the present study with those under thehysiological conditions, because the in vivo concentra-ions of pepsin and trypsin have not been well de-cribed. An average trypsin activity of 20 to 25 U/mg ofrotein has been reported in the intestine of pig (32),hich is much higher than the doses we used herein.owever, we used multiple levels of trypsin and pep-

in, with 10- to 25-fold range differences between theowest and highest levels. In addition, we measured

r r-PhyA (B) digested products by different amounts of pepsin (r 5sie blue. P, pepsin control; C, purified r-AppA (A) or r-PhyA (B). The, 100, 110, 120, and 200 kDa; Gibco). The results are representative

eal by r-PhyA and r-AppA incubated with different concentrations

) oma

n m

1, 0.002, 0.005, and 0.01) (B). Symbols: r-AppA (■), r-PhyA (h). The1 versus untreated r-AppA or r-PhyA control.

tidwtbvd

A

n

R

1

1

1

1

1

1

1

1

1

1

2

2

22

2

2

2

2

2

23

3

3

267SENSITIVITY OF RECOMBINANT ACID PHOSPHATASES TO TRYPSIN AND PEPSIN

he iP release from soybean meal by r-AppA or r-PhyAn the presence of pepsin or trypsin, a simulated in vivoigestive condition. Although both r-AppA and r-PhyAere partially purified, all the data consistently point

oward distinct responsive patterns of these two recom-inant enzymes to pepsin and trypsin. Thus, our initro observation could be relevant to physiological con-itions.

CKNOWLEDGMENTS

This research was supported by a grant from the Cornell biotech-ology program, New York State Science and Technology.

EFERENCES

1. Pallauf, J., and Rimbach, G. (1997) Arch. Tierernahr. 50, 301–319.

2. Greiner, R., Konietzny, V., and Jany, K. D. (1993) Arch. Biochem.Biophys. 303, 107–113.

3. Kim, Y. O., Kim, H. K., Bae, K. S., Yu, J. H., and Oh, T. K. (1998)Enzymol. Microbiol. Technol. 22, 2–7.

4. Shieh, T. R., and Ware, J. H. (1968) Appl. Microbiol. 16, 1348–1351.

5. Mitchell, D. B., Vogel, K., Weimann, B. J., Pasamontes, L., andvan Loon, A. P. (1997) Microbiology 143, 245–252.

6. Pasamontes, L., Haiker, M., Wyss, M., Tessier, M., and VanLoon, A. P. G. M. (1997) Appl. Environ. Microbiol. 63, 1696–1700.

7. Barberic, S., Kozulic, B., Reis, B., and Mildner, P. (1984) J. Biol.Chem. 259, 878–883.

8. Wodzinski, R. J., and Ullah, A. H. J. (1996) Adv. Appl. Microbiol.42, 263–302.

9. Liu, B. L., Rafiq, A., Tzeng, Y. M., and Rob, A. (1998) Enz.Microbiol. Technol. 22, 415–424.

0. Harper, A. F., Kornegay, E. T., and Schell, T. C. (1997) J. Anim.Sci. 75, 3174–3186.

1. Lei, X. G., Ku, P. K., Miller, E. R., and Yokoyama, M. T. (1993)J. Anim. Sci. 71, 3359–3367.

2. Lei, X. G., Ku, P. K., Miller, E. R., Yokoyama, M. T., and Ullrey,D. E. (1993) J. Anim. Sci. 71, 3368–3375.

3. Jongloed, A. W., Mroz, Z., and Kemme, P. A. (1992) J. Anim. Sci.70, 1159–1168.

4. Stahl, C. H., Han, Y. M., Roneker, K. R., House, W. A., and Lei,X. G. (1999) J. Anim. Sci., in press.

5. Jongbloed, A. W., Lenis, N. P., and Mroz, Z. (1997) Vet. Q. 19,130–134.

6. Skoglund, E., Nasi, M., and Sandberg, A.-S. (1998) Can. J. Anim.Sci. 78, 175–180.

7. Van Hartingsveldt, W., van Zeijl, C. M. J., Harteveld, G. M.,Gouka, R. J., Suykerbuyk, M. E. G., Luiten, R. G. M., VanParidon, P. A., Selten, G. C. M., Veenstra, A. E., van GorcomCess, R. F., and Van Hondel, C. A. (1993) Gene 127, 87–94.

8. Dassa, J., Marck, C., and Boquet, P. L. (1990) J. Bacteriol. 172,5497–5500.

9. Pradel, E., and Boquet, P. L. (1988) J. Bacteriol. 170, 4916–4923.

0. Ostanin, K., Harms, E. H., Stevis, P. E., Kuciel, R., Zhou, M. M.,and Van Etten, R. L. (1992) J. Biol. Chem. 267, 22830–22836.

1. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.(1951) J. Biol. Chem. 193, 265–275.

2. Laemmli, U. K. (1970) Nature 227, 680–685.3. Sherton, C. C., and Wool, I. G. (1974) J. Biol. Chem. 249, 2258–

2262.4. Piddington, C. S., Houston, C. S., Paloheimo, M., Cantrell, M.,

Miettinen-Oinonen, A., Nevalainen, H., and Rambosek, J. (1993)Gene 133, 55–62.

5. Chen, P. S., Toribara, T. Y., and Warner, H. (1956) Anal. Chem.28, 1756–1758.

6. Kostrewa, D., Gruninger-Leitch, F., D’Arcy, A., Broger, C.,Mitchell, D., and van Loon, A. P. G. N. (1997) Nature Struct. Biol.4, 185–189.

7. Jia, Z., Golovan, S., Ye, Q., and Forsberg, C. W. (1998) ActaCrystallogr. D. Biol. Crystallogr. 54, 647–649.

8. Takaoka, M., Takenobu, Y., Miyata, Y., Ikegawa, R., Mat-sumura, Y., and Morimoto, S. (1990) Biochem. Biophys. Res.Commun. 166, 436–442.

9. Fruton, J. S. (1977) Adv. Exp. Med. Biol. 95, 131–140.0. Savoie, L., and Charbonneau, R. (1990) Plant Foods Hum. Nutr.

40, 233–242.1. Wiesemuller, W., and Poppe, S. (1990) Arch. Tierenahr. 40, 689–

693.2. Tivey, D. R., Strullu, F., Blum, J. W., Le Jean, C., and Le

Dividich, J. (1997) in Digestive Physiology in Pigs (Laplace, C.,

Ed.), pp. 140–143, EAAP No. 88, INRA, Lescure Theol. Press,Paris.