Chemically stabilized trypsin used in dipeptide synthesis

Chemically Stabilized Trypsin Used inDipeptide Synthesis

Ann Murphy, Ciaran O Fagain

School of Biological Sciences, Dublin City University, Dublin 9, Republic ofIreland; telephone: +353 1 704 5288; fax: +353 1 704 5412;e-mail: [email protected]

Received 3 April 1997; accepted 26 September 1997

Abstract: Bovine pancreatic trypsin was treated with eth-ylene glycol bis(succinic acid N-hydroxysuccinimide es-ter). Approximately 8 of 14 lysines per trypsin moleculewere modified. This derivative (EG trypsin) was morestable than native between 30° and 70°C: T50 values were59°C and 46°C, respective. EG trypsin’s half-life of 25 minat 55°C was fivefold greater than native’s. EG trypsin hada decreased rate of autolysis and retained more activityin aqueous mixtures of 1,4-dioxan, dimethylformamide,dimethylsulfoxide, and acetonitrile. EG trypsin had lowerKm values for both amide and ester substrates; its kcatvalues for two amides (benzoyl-L-arginine p-nitroanilideand benzyloxycarbonyl glycyl-glycyl-arginyl-7-amino-4-methyl coumarin) increased, whereas its kcat value for anester (thiobenzoyl benzoyloxycarbonyl-L-lysinate) de-creased slightly. The specific activity (kcat/Km) of EG tryp-sin was increased for both amide and ester substrates.EG trypsin gave higher yields and reaction rates thannative in kinetically controlled synthesis of benzoyl ar-gininyl-leucinamide in acetonitrile and in t-butanol. High-est peptide yields occurred with EG trypsin in 95% ace-tonitrile, where 90% of the substrate was converted toproduct. No peptide synthesis occurred in 95% DMF witheither form of trypsin. © 1998 John Wiley & Sons, Inc. Bio-technol Bioeng 58: 366–373, 1998.Keywords: trypsin; stabilization; peptide synthesis; or-ganic solvents

INTRODUCTION

Proteases are frequently used in peptide synthesis (Ander-sen et al., 1991; Kasche and Haufler, 1984; Morihara,1987). Their many advantages over chemical peptide syn-thesis include stereospecificity, decreased racemization, andfewer side reactions. However, biocatalysts are prone toactivity losses and especially to the effects of high tempera-tures and organic solvents. Enzyme autolysis in solution isalso a risk. One may need to manipulate the enzyme toenhance its stability without losing activity. We have pre-viously described the effects of neutralizing some of thepositive charges on trypsin lysine residues using the mono-functional covalent modifier acetic acidN-hydroxysuccini-

mide ester (AA trypsin; Murphy and O´ Fagain, 1996, 1997).AA trypsin showed reduced autolysis and enhanced ther-mostability in both aqueous and aqueous/organic solventsystems. Increased tolerance of organic cosolvents or dena-turants was not observed at room temperature, however.

Crosslinking agents can increase the stability of proteases(Gleich et al., 1992; Rajput and Gupta, 1987, 1988; Taf-ertshofer and Talsky, 1989; Talsky et al., 1990), probablyby increasing the protein’s structural rigidity. With a bifunc-tional succinimide, one should be able to see the effect ofcrosslinking on trypsin stability and to compare it with co-valent modification of lysine residues. Here we describe theeffects of modification with a bifunctional succinimide, eth-ylene glycol bis(succinic acidN-hydroxysuccinimide ester)(EGNHS), on trypsin’s stability and kinetic properties. Wealso contrast this ‘‘EG trypsin’’ derivative with uncross-linked AA trypsin and report the use of EG trypsin fordipeptide synthesis in high concentrations of water-miscibleorganic solvents.

MATERIALS AND METHODS

Materials

Bovine pancreatic trypsin (EC 3.4.21.4) type III,N-benzoyl-L-arginine (BA),N-benzoyl-L-arginine ethyl ester (BAEE),benzoyl DL-argininep-nitroanilide (DL-BAPNA), benzoyl-L-argininep-nitroanilide (L-BAPNA), p-tosyl-arginyl meth-yl ester (TAME), p-nitrophenyl p8-guanidinobenzoate(NPGB), benzamidine, ethylene glycol bis(succinic acidN-hydroxysuccinimide ester) (EGNHS), suberic acid bis(N-hydroxysuccinimide ester) (SANHS),L-leucinamide HCl,2,4,6-trinitrobenzenesulfonic acid (TNBS), thiobenzoylbenzoyloxycarbonyl-L-lysinate (Z-Lys-SBzl), and 5,58-dithiobis(2-nitrobenzoic acid) (DTNB) were purchasedfrom Sigma. Labscan (Dublin, Ireland) supplied acetone,dimethylformamide (DMF), dimethylsulfoxide (DMSO),1,4-dioxan, methanol, and tetrahydrofuran (THF). Aceticacid, acetonitrile,t-butanol, ethylamine, and trifluoroaceticacid came from BDH. Coomassie blue protein assay wasfrom Bio-Rad and benzyloxycarbonyl glycyl-glycyl-arginyl-7-amino-4-methylcoumarin [(CBz)-Gly-Gly-Arg-AMC] was from Bachem. Benzoyl arginyl-leucinamide

Correspondence to:C. O FagainContract grant sponsor: ForbairtContract grant sponsor: Dublin City UniversityContract grant sponsor: INTAS NetworkContract grant number: 94-4105

© 1998 John Wiley & Sons, Inc. CCC 0006-3592/98/040366-08

(BzlArgLeuNH2) was a generous gift from Dr. B. Walker,Division of Biochemistry, Queen’s University and Dr. P.Harriott, Biosyn (both from Belfast, Northern Ireland, UK).All chemicals were of analytical grade.

Reaction of Trypsin with Succinimides

Reaction was performed in 3 mM KH2PO4/K2HPO4/0.1 MKCl, adjusted to pH 8.2 and containing 3 mM benzamidine,at room temperature, for 20 min. Working concentration ofboth enzyme and succinimide was 2 mg mL−1. The reactionmixture contained 5 mg trypsin in 2.375 mL of buffer and125 ml DMSO containing 5 mg of dissolved succinimide.The reaction was terminated by centrifugal gel filtration onSephadex G-25 (Helmerhorst and Stokes, 1980) to removeexcess reagent.

Activity and Other Assays

Amidase activity of trypsin was measured using BAPNA assubstrate in a scaled-down version of the method of Er-langer et al. (1961). Two hundred microliters of BAPNA(1.67 mM; 29 mg dissolved in 1 mL of DMSO and added to39 mL of 0.1M Tris-HCl, pH 8.2, containing 20 mM CaCl2)was equilibrated in a 30°C waterbath. Fifty microliters ofsample was added and incubated at 30°C for 15 min. Thereaction was then terminated with 30% acetic acid (50mL).Absorbances were read at 405 nm on a Titertek Mk IIreader. Esterase activity was determined using TAME(Rick, 1974). Tris buffer, 2.6 mL, 48 mM, pH 8.1, contain-ing 11.5 mM CaCl2, was added to 0.3 mL of TAME stockin H2O (10 mM) and equilibrated at 30°C for 5 min. En-zyme solution (0.1 mL; approx. 0.3 mg mL−1) was added tothe mixture and the absorbance read at 247 nm over 5 min.The number of active sites was determined using NPGB asdescribed by Walsh (1970) but modified slightly. Thirtymicroliters of NPGB (0.01M; 3.37 mg in 1 mL of DMF)was added to 3 mL of sample in buffer (1 mg trypsin mL−1)and absorbance read at 410 nm after 15 min. The number offree amino groups in trypsin was estimated using TNBS,according to Rajput and Gupta (1987). SDS-polyacrylamidegel electrophoresis (10% and 15% acrylamide) followed themethod of Laemmli (1970).

Stability Determinations

The autolysis rate was measured using Coomassie blue dyeas described by Bickerstaff and Zhou (1993). Approx. 0.2mg mL−1 enzyme solution in 0.1M Tris-HCl (pH 8.2) con-taining 20 mM CaCl2 was incubated in a 50°C waterbath. Atregular intervals, 2.6 mL aliquots were taken, cooledquickly to room temperature on ice, and 0.4 mL of Coo-massie blue added. The sample was mixed and after 5 minthe absorbance was read at 595 nm. To obtain a temperatureprofile, trypsin (approx. 0.05 mg mL−1 in buffer containing20 mM CaCl2) was incubated for 10 min at temperaturesbetween 30° and 75°C. Samples were removed and cooled

on ice for 1 min and residual activity was assayed andcompared with the activity of a sample incubated at 30°C.

Thermoinactivation rates were measured by incubation ofapprox. 0.05 mg mL−1 native and modified trypsins inbuffer (3 mM KH2PO4/K2HPO4, 0.1 M KCl, pH 8.2, con-taining 20 mM CaCl2) in a 55°C waterbath. Aliquots weretaken at intervals, cooled quickly to room temperature onice, and the residual trypsin activity assayed. Under theseconditions, autolysis was found to be negligible. Data fittedwell to a first-order exponential decay curve from whichhalf-lives were estimated. To measure tolerance of organicsolvents, trypsin samples (approx. 0.05 mg mL−1) in aque-ous (containing 20 mM CaCl2)/organic mixtures rangingfrom 0% to 90% (v/v) solvent were incubated at 30°C for 1h. Residual activity was assayed and compared with con-trols containing no organic solvent. To test stability in de-naturants, approx. 0.05 mg mL−1 trypsin in buffer contain-ing 20 mM CaCl2 and either guanidine HCl (0 to 1M) orurea (0 to 12.5M) was incubated at 30°C for 1 h, afterwhich time the residual trypsin activity was assayed. Con-trols containing buffer only were used for comparison.

pH Profile

Buffers of 0.1M acetic acid/sodium acetate (pH 4.5 to 6),Tris-HCl (pH 7 to 9) and glycine-NaOH (pH 8 to 10) wereprepared; each contained 20 mM CaCl2. Enzyme solution(approx. 0.05 mg mL−1) was diluted in buffer of each pH.Substrate solutions were prepared using buffer of each pHand activities were measured as described for Tris, pH 8.2.The activity was calculated (%) relative to the maximumactivity of each sample.

Kinetic Measurements

Rates onL-BAPNA were measured based on the method ofErlanger et al. (1961). A stock solution ofL-BAPNA (4.8mM) was prepared by dissolving 42 mg ofL-BAPNA in 1mL of DMSO and adjusting the volume to 20 mL with 0.1M Tris-HCl (pH 8.2) containing 20 mM CaCl2. A trypsinstock solution (225mg mL−1) was prepared in 3 mMKH2PO4/K2HPO4, adjusted to pH 8.2, containing 0.1MKCl. Substrate (0.93 mL) was added to the cuvette andequilibrated at 30°C for 5 min. Then 0.067 mL of trypsinwas added and the reaction was monitored at 410 nm. Thekinetic assay protocol for (CBz)-Gly-Gly-Arg-AMC wasbased on O’Donnell-Tormey and Quigley (1983). A 4 mMsubstrate stock solution was prepared by adding 25 mg to0.5 mL DMSO. Volume was adjusted to 10 mL by additionof 9.5 mL 0.1 M Tris-HCl (pH 8.2) containing 20 mMCaCl2. Substrate (0.125 mL) was added to a test tube andequilibrated in a 30°C waterbath for 5 min. Ten microlitersof enzyme solution (5mg mL−1) was added; the reactionproceeded at 30°C and was terminated after 5 min by ad-dition of 0.125 mL of acetic acid (30% v/v). The samplewas diluted to 5 mL using distilled water. The release offluorescent AMC from the substrate was measured at 370

MURPHY AND O FAGAIN: STABILIZED TRYPSIN IN DIPEPTIDE SYNTHESIS 367

nm (excitation) and 440 nm (emission) with slitwidths of 5nm. Esterase kinetic experiments used Z-Lys-SBzl. Stocksolutions of 110 mM DTNB in 50 mM Na2HPO4, 20 mMZ-Lys-SBzl in water, and 200 mM Na phosphate–200 mMNaCl (Pi-NaCl, pH 7.5) were prepared. One part each ofDTNB and Z-Lys-SBzl solutions were added to 100 partsPi-NaCl. Nine hundred fifty microliters of this working so-lution was added to a cuvette and equilibrated for 5 min at30°C. Fifty microliters of enzyme solution was added andthe reaction was monitored at 412 nm (Coleman and Green,1981).

To determine catalytic activity in organic solvents, 0.46M Tris-HCl (pH 8.1), containing 0.115M CaCl2 (0.26 mL),was mixed with 0.3 mL of 10 mM TAME stock solution.The volume was made up to 2.9 mL by addition of waterand organic solvent to give the required concentrations. Ap-prox. 0.3 mg mL−1 trypsin solution (0.1 mL) was added tothe reaction mixture and absorbance was read at 247 nmover 3 min. Reference cells without enzyme were includedat each solvent concentration.

Dipeptide Synthesis

A stock reaction mixture of 33.5 mg of LeuNH2 and 34.5mg of BAEE in 950mL of aqueous/organic solvent mixtureand 45mL of triethylamine (to ensure alkaline conditions)was prepared. A portion of this mixture (199mL) was re-moved and equilibrated for 10 min at 4°C. Enzyme solution(10 mL) was added and the entire reaction mixture incu-bated at 4°C. Ten-microliter aliquots were removed at regu-lar intervals, added to 0.5 mL of 50% (v/v) aqueous metha-nol containing 1% (v/v) trifluoroacetic acid, and analyzedby HPLC using a Beckman System Gold apparatusequipped with an Autosampler 507, Diode Array DetectorModule 168, and Beckman C8 column (0.46 mm × 25 cm);with flow rate 1 mL min−1, detection at 204 and 230 nm,mobile phase methanol with 0.05% aqueous trifluoroaceticacid (50/50 by volume). Sample (20mL) was injected ontothe column using an autosampler. Both substrates and prod-ucts were identified and quantified by comparison withstandards of each.

RESULTS

Optimization of Reaction

Modification with SANHS reduced trypsin BAPNA activityby >50% and this reagent was not used further. The suc-cessful EGNHS reaction was optimized with respect tomodifier concentration. (All other reaction conditions hadpreviously been optimized as described by Miland et al.[1996].) Reaction mixtures containing 0.2, 0.4, 2.0, and 10.0mg mL−1 were prepared and the resulting trypsin derivativesanalyzed. All modified fractions had greater BAPNA ac-tivities and were more thermostable than native trypsin (seenext subsection). Active site titration showed an equal num-

ber of active sites in the native and in all EGNHS-modifiedforms; that is, the modification had proceeded without anyloss of enzyme activity. The 2.0 mg mL−1 EGNHS deriva-tive had the highest BAPNA activity (see next subsection).This proved to be the optimal EGNHS concentration and,accordingly, was used in subsequent experiments (‘‘EGtrypsin’’). EGNHS at 10 mg mL−1 gave the highest percent-age of modified lysine residues (85%) and slightly greaterthermostabilization (sevenfold greater half-life at 55°C), butlower activity than the 2.0 mg mL−1 dosage.

Characterization of Trypsin Derivatives

SDS gel electrophoresis showed no higher molecular weightspecies for modified trypsin compared with the native.Thus, no intermolecular crosslinking or aggregation had oc-curred. EG trypsin (modified with 2.0 mg mL−1 EGNHS)showed enhanced amidase activity (203%, BAPNA), butslightly lower esterase activity (90%, TAME), comparedwith the native. (Native and EG trypsin had the same num-ber of active sites.) TNBS determination of free aminogroups indicated that approximately 40% of trypsin lysinesremained, suggesting that 8 of the 14 lysines per trypsinmolecule (Walsh, 1970) had been modified. EG and nativetrypsins gave similar pH-activity profiles in the BAPNAassay over the pH range 4.5 to 10.0.

Stability of EG Trypsin

Modified trypsin was less prone to autolysis than native (seeFig. 1). The absorbance of native trypsin decreased to 40%of its initial value after 5 h, whereas the EG trypsin re-mained stable. During thermal inactivation at 55°C, data forboth enzymes fitted well to a first-order decay curve (Fig.2). EG trypsin had a half-life of 25 min (k 4 0.028 ± 0.004)compared with 5 min (k 4 0.140 ± 0.018) for native. Re-sidual activities of EG and native trypsins were comparedover the temperature range 30° to 75°C (see Fig. 3). EGtrypsin retained 100% activity up to 50°C, whereas the na-tive progressively lost activity above 40°C.T50 values fornative and EGNHS trypsin were 46°C and 59°C, respec-tively. Native and EG trypsins’ tolerance of acetonitrile,methanol, DMF, and THF were compared. In each case, EGtrypsin showed enhanced organotolerance (see Fig. 4 for anexample). The threshold concentrations,C50 (defined byMozhaev et al. [1989] as the percentage of solvent yieldinghalf of the initial enzyme activity), were calculated. EGtrypsin had higherC50 values for amidase activity in each ofthe solvents tested (Table I). For both trypsins, esteraseactivity was enhanced at low solvent concentrations andwas retained even at high solvent concentrations. No sig-nificant difference in esteraseC50 values was observed be-tween native and EG trypsin (Table I). The effects of gua-nidine hydrochloride and urea on both trypsins were inves-tigated. No stability difference was observed in guanidinehydrochloride, but EG trypsin was more tolerant of urea,retaining higher activity over the concentration range tested

368 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 58, NO. 4, MAY 20, 1998

(see Fig. 5). This was especially noticeable at high concen-trations: EG trypsin retained >50% activity (native <10%)in 10M urea. In these experiments, trypsin was incubated invarying concentrations of urea and then diluted into aBAPNA reaction mixture for assay under optimal condi-tions. Activities were also measured in assay mixtures con-taining increasing concentrations of urea. Under these con-ditions, EG trypsin was slightly more active than native inthe range 4 to 8M.

Kinetics of EG Trypsin in Aqueous Systems

Km andkcat values were calculated by the direct linear plot(Tipton, 1992) and are summarized in Table II.Km for bothamide and ester substrates decreased upon modification. EGtrypsin’s kcat values were 2 and 1.5-fold greater than na-

Figure 1. Autolysis of native and modified trypsin at 50°C. The absor-bance is shown in percentages relative to initial absorbance of samples.Circles: native trypsin; squares: EG trypsin.

Figure 2. Thermoinactivation of native and modified trypsin at 55°C.The activity is shown as percentage value relative to initial activity ofsamples. Circles: native trypsin; squares: EG trypsin.

Figure 3. Effect of 10-min incubations at increasing temperature on na-tive and modified trypsin. Activities were measured at various tempera-tures and are percentage values relative to activity at 30°C. Circles: nativetrypsin; squares: EG trypsin.

Figure 4. Effect of various concentrations of acetonitrile on native andmodified trypsin at 30°C. The activity is shown as percentage values rela-tive to activity of solvent-free samples (i.e., in buffer only). Circles: nativetrypsin; squares: EG trypsin.


tive’s for L-BAPNA and (CBz)-Gly-Gly-Arg-AMC, respec-tively. Thekcat for the ester substrate Z-Lys-SBzl decreasedto 80% of native’s. Thekcat/Km ratios for EG trypsin in-creased slightly for all substrates tested: values were 3.8-,2.8-, and 1.4-fold greater than those of native trypsin forL-BAPNA, (CBz)-Gly-Gly-Arg-AMC, and Z-Lys-SBzl, re-spectively.

Dipeptide Synthesis

Conditions for dipeptide synthesis by native trypsin wereinvestigated. The optimum enzyme concentration was 1 mgmL−1 whereas 0.2M leucinamide gave the highest rate ofpeptide synthesis. All subsequent reactions used these con-centrations at 4°C, the optimal temperature. (While the re-action rate was slowest at 4°C, hydrolysis of the Bzl-Arg-LeuNH2 product was minimal, resulting in maximal productyield.) Hydrolysis of both substrate and the dipeptide prod-uct decreased with increasing concentration of acetonitrile

for both native and EG trypsins. At low acetonitrile con-centrations, little difference was observed between nativeand EG trypsins. In 95% acetonitrile, however, EG trypsinhad the greater synthetic rate (see Fig. 6). After 1 day, EGtrypsin had utilized all the BAEE substrate but nativetrypsin did not utilize all the BAEE even after 3 days.The synthesis/hydrolysis ratio was approximately 9:1, withEG trypsin converting 90% of the substrate to Bzl-Arg-LeuNH2. In 95% t-butanol, both trypsins were capable ofpeptide synthesis. Again, EG trypsin showed greater syn-thetic rates than native (see Fig. 7). EG trypsin reachedmaximum product yield after 1 day, whereas native did notattain maximum product yield even after 6 days. Over 60%of the substrate was converted to Bzl-Arg-LeuNH2 and theremainder to Bzl-Arg (the hydrolysis product).

In 50% DMF, EG trypsin had a slightly higher reaction

Table I. C50 values for amidase (BAPNA) and esterase (TAME) activi-ties of native and EG trypsin in water-miscible solvents.

Solvent

C50 (%): amidase C50 (%): esterase

Native EG trypsin Native EG trypsin

Acetonitrile 39 ± 0 60 ± 2 69 ± 4 64 ± 6Methanol 63 ± 3 73 ± 3 63 ± 2 62 ± 2Dimethylformamide 63 ± 2 67 ± 2 ND NDTetrahydrofuran 35 ± 4 49 ± 9 71 ± 3 71 ± 11,4-Dioxan 46 ± 1 54 ± 3 52 ± 2 52 ± 1

ND 4 not determined.

Table II. Kinetic parameters of native and EG-trypsin.

Substrate Native EG trypsin

L-BAPNAKm 2.5 ± 0.2 1.31 ± 0.3kcat 2.4 ± 0.2 4.77 ± 1.0kcat/Km 0.96 ± 0.16 3.67 ± 0.2

(CBz)-G-G-R-AMCKm 0.48 ± 0.02 0.25 ± 0.01kcat 0.93 ± 0.04 1.36 ± 0.04kcat/Km 1.9 ± 0.1 5.4 ± 0.2

Z-Lys-SBzlKm 0.041 ± 0.003 0.027 ± 0.01kcat 29 ± 3 23 ± 1.4kcat/Km 710 ± 130 975 ± 37

Units: Km 4 mM, kcat 4 s−1, kcat/Km 4 s−1 mM−1.

Figure 5. Effect of various concentrations of urea on native and modifiedtrypsin at 30°C. The activity is shown as percentage values relative toactivity of urea-free samples. Circles: native trypsin; squares: EG trypsin.

Figure 6. Rate of dipeptide synthesis by native and modified trypsin in95% acetonitrile (v/v) at 4°C. Circles: native trypsin; squares: EG trypsin.


rate than native. Maximum product yield was achieved after20 min, after which time product hydrolysis occurred. After24 h, Bzl-Arg-LeuNH2 product concentrations had droppedto 60% and 70% for EG trypsin and native, respectively. In95% DMF, used in an attempt to prevent hydrolysis, neithertrypsin yielded dipeptide.

DISCUSSION

Crosslinking of Trypsin

Bifunctional reagents have successfully been used to stabi-lize proteins (Wong and Wong 1992). Bis-N-hydroxy-succinimide esters are homobifunctional crosslinking re-agents. Modifications with EGNHS, which spans a distanceof 16.1 Å, was found to enhance activity (BAPNA assay),but treatment with SANHS (11 Å long; Ji, 1983)decreasedactivity. Thus, the crosslinking distance has a significanteffect on trypsin activity. Torchilin et al. noted that thelength of intramolecular crosslinks introduced intoa-chy-motrypsin (Torchilin et al., 1978) and glyceraldehyde-3-phosphate dehydrogenase (Torchilin et al., 1983) influencedthermostability. Gleich et al. (1992) showed that trypsincrosslinked withN-hydroxysuccinimide esters of dicarbox-ylic acids, and dianhydrides and bisimidoesters had en-hanced thermostability, but confined their report to heattolerance. EG trypsin’s BAPNA activity was 203% of na-tive’s. This compares with the uncrosslinked AA trypsin,which had 130% of the native enzyme’s BAPNA activity(Murphy and OFagain, 1996).

Stability of EG Trypsin

Crosslinking of trypsin decreases autolysis (Gleich et al.,1992; Rajput and Gupta, 1987, 1988) by increasing the ri-

gidity of the enzyme. The reduced autolysis observed heremay be due to the modification of lysine residues rather thanto crosslinking; a similarly reduced autolysis was observedin AA trypsin (Murphy and OFagain, 1996). EG trypsinhad a fivefold greater half-life at 55°C than native trypsin,and a 13°C higherT50 value. These values compare with adoubled 55°C half-life, and a 5°C increase inT50 for un-crosslinked AA trypsin (Murphy and O´ Fagain, 1996). (Theextent of lysine modification was the same in both deriva-tives.)

Tolerance of organic solvents (acetonitrile, DMF, THF,and methanol; these have denaturation capacity values of64.3, 63.3, 100.0, and 30.5, respectively [Khmelnitsky et al.,1991]) was enhanced in EG trypsin and itsC50 values werehigher than native’s for all solvents tested (see Table I). At30°C, in contrast, AA trypsin’s tolerance of methanol,DMF, and THF resembled that of the native (Murphy and O´

Fagain, 1996). Gorman and Dordick (1992) showed thatenzymes lose their bound water when suspended in organicsolvents, with the highest degree of desorption resultingfrom polar solvents. The stabilization of EG trypsin is likelydue to increased conformational rigidity resulting from in-tramolecular crosslinks. St. Clair and Navia (1992) showedthat chemically crosslinked thermolysin crystals (T-CLECs)had enhanced stability to organic solvents.

EG trypsin was much more tolerant of urea than native.This strongly suggests the existence of crosslink(s) in EGtrypsin. Simple modification of the protein surface would beunlikely to oppose the unfolding effects of urea. The stabi-lization against urea most likely results from crosslinkingand rigidification of the protein backbone following bis-succinimide modification. Note that the modified but un-crosslinked AA trypsin was no more tolerant of urea thannative (Murphy and O´ Fagain, 1996). It would be interestingto confirm the existence of a crosslink by proteolytic diges-tion and electrophoresis of fragments, and to ascertain itsexact position from peptide maps and sequence information.EG trypsin’s urea tolerance and decreased autolysis sug-gests its possible application as a catalyst for the trypticdigestion of resistant or insoluble proteins. However, whenurea was included in the BAPNA assay mix, EG trypsin’sactivity was only slightly greater than native’s in the range4 to 8 M. Curiously, no difference between the native andEG trypsin was observed in guanidine HCl. Gabel (1973)found that trypsin covalently coupled to Sephadex resisteddenaturation by urea but not by guanidinium chloride. Hesuggested that the guanidinium chloride unfolding proceedsvia a different activated state from that in urea, and that thisstate may be reached in the same way whether the enzymeis soluble or bound to a carrier.

Kinetics

Table II shows thatKm for both amide and ester substratesdecreases upon modification. Values ofkcat increased foramide substrates but decreased for the ester substrate. Simi-lar amidase effects were noted for AA trypsin but itskcat for

Figure 7. Rate of dipeptide synthesis by native and modified trypsin in95% t-butanol (v/v) at 4°C. Circles: native trypsin; squares: EG trypsin.


Z-Lys-SBzl increased by∼50% (Murphy and O´ Fagain,1997). For both substrate types, thekcat/Km ratio was higherfor EG trypsin. Succinimide modification neutralizes thepositive charge of some trypsin lysine residues. The posi-tively charged scissile residues (Lys, Arg) binding to tryp-sin’s active site will be repelled by like charges in the activesite’s vicinity. Neutralization of some trypsin positivecharges may enhance substrate binding to the active site.

Serine proteases hydrolyze amides and esters by the acyl-enzyme mechanism (Fersht, 1985). For trypsin, the rate-limiting step of ester hydrolysis is the breakdown of theacyl-enzyme intermediate (Guinn et al., 1991), whereas therate-determining step in amide hydrolysis is acylation of theactive site Ser. Nucleophilic attack by this Ser requiresproper alignment between enzyme and bound substrate. Es-ter hydrolysis may depend less on the active site structure,as it is governed by hydrolytic attack of water on the acyl-enzyme intermediate. Amidase activity is, therefore, ex-pected to be more sensitive than esterase activity to active-site conformational changes and is more strongly affectedby immobilization (Sears and Clark, 1993). Enhanced es-terase activity was observed at low solvent concentrationsfor both native and EG trypsin, as reported by Barbas et al.(1988) and Guinn et al. (1991). Low concentrations of or-ganic solvents are thought slightly to accelerate deacylationof the acyl enzyme, the rate-limiting step of ester hydroly-sis. C50 values for esterase activity of EG trypsin differedlittle from native in the solvents tested.

Dipeptide Synthesis

The dipeptide synthesis protocol had been optimized (forAA trypsin; Murphy and OFagain, 1997) prior to compari-son of native and EG trypsin. The highest aminolysis rateoccurred with a twofold molar excess of acyl acceptor (leu-cinamide) over the acyl donor (BAEE). Higher leucinamideconcentrations gave no further increase in aminolysis. Tryp-sin-catalyzed synthetic yields increased with decreasingtemperature. Similarly, lower temperatures led to higheryields in the chymotrypsin-catalyzed synthesis of X-Phe-Leu-NH2 (Calvet et al., 1992). The hydrolysis rate for theacyl enzyme (and the peptide product) increased with tem-perature much more than did the synthetic rate. Blanco et al.(1991) suggested that the absorption constant of the nucleo-phile to the enzyme/complex decreases with increasing tem-perature: the proportion of trypsin molecules containing leu-cinamide in their active centers is much lower at highertemperatures.

The effect of organic solvents on peptide synthesis wasinvestigated using acetonitrile,t-butanol, and DMF. Sol-vents were used to reduce the unwanted secondary hydro-lysis of product, to improve the esterase/amidase ratio, andto exploit EG trypsin’s greater tolerance of water-miscibleorganic solvents. (In this regard, we should point out thatthe ratio ofkcat/Km values for Z-Lys-SBzl/L-BAPNA is 740for native, 513 for AA trypsin [Murphy and O´ Fagain, 1997]and 266 for EG trypsin [Table II]; thus native appears to

have the greater esterase/amidase ratio in aqueous solution.Nevertheless, EG trypsin accumulated higher dipeptideyields than native in solvent systems.) Peptide synthesis wasaccomplished in varying concentrations of acetonitrile,which was chosen because it has successfully been used forpeptide synthesis and because of the high stability of EGtrypsin in its presence. Cerovsky (1990) showed that trypsincould synthesize peptides in acetonitrile with low watercontent (5% v/v). Hydrolysis of BAEE and secondary pep-tide hydrolysis both were suppressed and the reaction wasslower. His findings agree with the effects of varying ace-tonitrile concentrations on the rate of peptide synthesis bynative and EG trypsin. The aminolysis/hydrolysis ratio in-creased with acetonitrile concentration, and the highest ac-cumulation of peptide product occurred at 95% acetonitrile.EG trypsin also had a higher peptide synthesis rate thannative in 95%t-butanol, with suppressed product hydroly-sis, as in acetonitrile. In 50% DMF, 80% of BAEE wasconverted to product after 20 min, but no peptide productwas formed in 95% DMF, which likely inactivated both EGand native trypsins: neither trypsin showed amidase activityin 95% DMF. This coincides with Cerovsky (1990), whoobserved little or no synthesis by trypsin in 95% DMF.Calvet et al. (1992) noted higher reaction yields at low DMFconcentrations in chymotrypsin-catalyzed peptide synthesis.

Our results for trypsin-catalyzed peptide synthesis com-pare well with the effects of solvents on chymotrypsin-mediated peptide synthesis (Nagashima et al., 1992). Na-gashima’s group found that acetonitrile,t-butanol, and DMFgave yields of 80%, 74%, and 0% after 24 h in approxi-mately 95% solvent systems. Although many of the pep-tides directly synthesizable by trypsin will be water soluble,the coupling of short defined peptides to yield a larger prod-uct molecule is a potentially important application of enzy-matic synthesis. Not all peptides for coupling will be watersoluble, which explains our interest in EG trypsin’s organicsolvent performance.

CONCLUSION

Modification with a bis-succinimide increased trypsin’s sta-bility toward autolysis, thermoinactivation, hydrophilic or-ganic solvents, and urea. This modification leads toKm

increases for both amide and ester substrates, increasedkcat

values for amide substrates, and a slightly decreasedkcat forthe ester substrate. EG trypsin synthesized the dipeptideBzl-Arg-Leu-NH2 at higher rates than native in 95% aceto-nitrile (90% BAEE converted to product and only 10% hy-drolysis observed) ort-butanol. EG trypsin may prove tohave further application as a catalyst in presence of urea andas a tool for peptide synthesis.

We thank Dr. Brian Walker, Division of Biochemistry, Queen’sUniversity, and Dr. Pat Harriott, Biosyn Ltd. (both of Belfast,Northern Ireland, UK) for Bzl-Arg-Leu-NH2 and much helpfuladvice. A reviewer of this paper also made helpful suggestions.This laboratory is a participant in an INTAS network.


References

Andersen, A. J., Fomsgaard, J., Thorbek, P., Aasmul-Olsen, A. 1991. Cur-rent enzymatic possibilities in the technology of peptide synthesis II.Process development in enzymatic peptide synthesis. Chimicaoggi4:17–23.

Barbas, C. F., Matos, M., Blair West, J. R., Wong, C.-H. 1988. A search forpeptide ligase: Cosolvent-mediated conversion of proteases to ester-ases for irreversible synthesis of peptides. J. Am. Chem. Soc.110:5162–5166.

Bickerstaff, G. F., Zhou, H. 1993. Protease activity and autodigestion as-says using Coomassie Blue dye binding. Anal. Biochem.210: 155–158.

Blanco, R. M., Alvaro, G., Guisan, J. M. 1991. Enzyme reaction engineer-ing: Design of peptide synthesis by stabilised trypsin. Enzyme Microb.Technol.13: 573–583.

Calvet, S., Clapes, P., Vigo, J. P., Xano, N., Jorba, X., Mas, R. M., Torres,J. L., Valencia, G. 1992. Application of empirical design methodolo-gies to the study of the influence of reaction conditions and N-alpha-protecting group structure on the enzymatic x-Phe-Leu-NH2 dipeptidesynthesis in buffer/DMF solvent systems. Biotechnol. Bioeng.39:539–549.

Cerovsky, V. 1990. Free trypsin-catalysed peptide synthesis in acetonitridewith low water content. Biotechnol. Lett.12: 899–904.

Coleman, P. L., Green, G. D. J. 1981. A coupled photometric assay forplasminogen activator, pp. 408–414. In: L. Lorand (ed.), Methods inenzymology. Academic Press, New York.

Erlanger, B. F., Kokowsky, N., Cohen, W. 1961. The preparation andproperties of two new chromogenic substrates of trypsin. Arch. Bio-chem. Biophys.95: 271–278.

Fersht, A. 1985. Enzyme structure and mechanism. W.H. Freeman, NewYork.

Gabel, D. 1973. The denaturation by urea and guanidium chloride of tryp-sin and N-acetylated-trypsin derivatives bound to Sephadex and aga-rose. Eur. J. Biochem.33: 348–356.

Gleich, M., Talsky, G., Spannagl, R. 1992. Stabilization of trypsin bymodification with bifunctional reagents. Dechema BiotechnologyConference (vol. 5A), pp. 121–124.

Gorman, L. S., Dordick, J. S. 1992. Organic solvents strip water off en-zymes. Biotechnol. Bioeng.39: 392–397.

Guinn, R. M., Blanch, H. W., Clark, D. S. 1991. Effect of a water-miscibleorganic solvent on the kinetic and structural properties of trypsin.Enzyme Microb. Technol.13: 320–326.

Helmerhorst, E., Stokes, G. B. 1980. Microcentrifuge desalting: A rapid,quantitative method for desalting small amounts of protein. Anal. Bio-chem.104: 130–135.

Ji, T. H. 1983. Bifunctional reagents. pp. 580–609. In: C. H. W. Hirs, S. N.Timasheff (eds.), Methods in enzymology. Academic Press, NewYork.

Kasche, V., Haufler, U. 1984. Enzyme catalyzed semisynthesis of conden-sation products—kinetic vs. equilibrium control. Third Meeting of theEuropean Congress on Biotechnology (vol. 1), pp. 425–430.

Khmelnitsky, Y. L., Mozhaev, V. V., Belova, A. B., Sergeeva, M. V., Mar-tinek, K. 1991. Denaturation capacity: A new quantitative criterion forselection of organic solvents as reaction media in biocatalysis. Eur. J.Biochem.198: 31–41.

Laemmli, U. K. 1970. Cleavage of structural proteins during the assemblyof the head of bacteriophage T4. Nature277: 680–685.

Miland, E., Smyth, M. R., O´ Fagain, C. 1996. Modification of horseradishperioxidase with bifunctional N-hydroxysuccinimide esters: Effects onmolecular stability. Enzyme Microb. Technol.19: 242–249.

Morihara, K. 1987. Using proteases in peptide synthesis. Trends Biotech-nol. 5: 164–170.

Mozhaev, V. V. 1993. Mechanism based strategies for protein thermosta-bilization. Trends Biotechnol.11: 88–95.

Mozhaev, V. V., Khmelnitsky, Y. L., Sergeeva, M. V., Belova, A. B., Kly-achko, N. L., Levashov, A. V., Martinek, K. 1989. Catalytic activityand denaturation of enzymes in water/organic cosolvent mixtures. Al-pha-chymotrypsin and laccase in mixed water/alcohol, water/glycoland water/formamide solvents. Eur. J. Biochem.184: 597–602.

Murphy, A., O Fagain, C. 1996. Stability characteristics of chemically-modified soluble trypsin. J. Biotechnol.49: 163–171.

Murphy, A., OFagain, C. 1997. Dipeptide synthesis using acetylated tryp-sin derivative. Biotechnol. Techniq.11: 13–16.

Nagashima, T., Watanabe, A., Kise, H. 1992. Peptide synthesis by prote-ases in organic solvents: Medium effect on substrate specificity. En-zyme Microb. Technol.14: 842–847.

O’Donnell-Tormey, J., Quigley, J. P. 1983. Detection and partial charac-terization of a chymostatin-sensitive endopeptidase in transformed fi-broblasts. Proc. Natl. Acad. Sci. USA88: 344–348.

Rajput, Y. S., Gupta, M. N. 1987. Reaction of trypsin with some crosslink-ing reagents. Enzyme Microb. Technol.9: 161–163.

Rajput, Y. S., Gupta, M. N. 1988. Reaction of trypsin with dimethyl adipi-midate: Purification and characterization of a trypsin derivative withdecreased autolysis. Enzyme Microb. Technol.10: 143–150.

Rick, W. 1974. Trypsin, pp. 1013–1024. In: Bergmeyer, H. U. (ed.), Meth-ods of enzymatic analysis. Academic Press, New York.

Sears, P. S., Clark, D. S. 1993. Comparison of soluble and immobilizedtrypsin kinetics: Implications for peptide synthesis. Biotechnol. Bio-eng.42: 118–124.

St. Clair, N. L., Navia, M. A. 1992. Cross-linked enzyme crystals as robustbiocatalysts. J. Am. Chem. Soc.114: 7314–7316.

Tafertshofer, G., Talsky, G. 1989. Intra- and intermolecular crosslinking ofenzymes by dianhydrides and N-hydroxysuccinimide esters. Dechema-Biotechnology Conference (vol 3A), pp. 171–174.

Talsky, G., Tafertshofer, G., Gleich, M. 1990. Influence of intra- andintermolecular crosslinking on activity, stability and kinetic of en-zymes. Dechema Biotechnology Conference (vol. 4A), pp. 121–243.

Tipton, K. F. 1992. Principles of enzyme assay and kinetic studies, pp.1–58. In: R. Eisenthal and M. J. Danson (eds.), Enzyme assays: Apractical approach. IRL Press, Oxford.

Torchilin, V. P., Maksimenko, A. V., Smirnov, V. N., Berezin, I. V., Klib-anov, A. M., Martinek, K. 1978. The principles of enzyme stabiliza-tion. III. The effect of the length of intra-molecular cross-linkages onthermostability of enzymes. Biochim. Biophys. Acta.522: 277–283.

Torchilin, V. P., Trubetskoy, V. S., Omel’Yanenko, V. G. 1983. Stabiliza-tion of subunit enzyme by intersubunit crosslinking with bifunctionalreagents: Studies with glyceraldehyde-3-phosphate dehydrogenase. J.Molec. Catal.19: 291–301.

Walsh, K. A. 1970. Trypsinogen and trypsins of various species, pp.41–63. In: G. E. Perlmann and L. Lorand (eds.), Methods in enzymol-ogy. Academic Press, New York.

Wong, S. S., Wong, L-J. C. 1992. Chemical crosslinking and the stabili-zation of proteins and enzymes. Enzyme Microb. Technol.14:866–874.


Chemically stabilized trypsin used in dipeptide synthesis

Documents

Transcript of Chemically stabilized trypsin used in dipeptide synthesis