Identification of an Essential Tryptophan Residue in Alliinase from Garlic (Allium sativum) by Chemical Modification
YoungNamJin,Yong-Hoon Choi, andChul-HakYang*
Department of Chemistry, CollegeofNaturalSciences,Seoul National University, San56-1, Shillim-dong, Kwanak-gu,Seoul151-742, Korea
Received November 22, 2000
Wehave employed chemical modificationto identify amino acids essential for the catalytic activityof alliinase (EC 4.4.1.4) fromgarlic(Allium sativum). AlliinasedegradesS-alkyl-Lcysteinesulfoxides,causing thechar acteristic odor ofgarlic. The activity of alliinase wasrapidlyand completely inactivated by N-bromosuccinim- ide (NBS)andslightly decreasedby succinic anhydrideand N-acetylimidazole. These resultsindicatethat tryptophanyl,lysyl, and tyrosylresiduesplayan importantrole in enzyme catalysis.Thereaction of alliinase with NBA yielded a characteristicdecrease in boththeabsorbanceat 280 nmand theintrinsicfluorescenceat 332 nm with increasingreagentconcentrationof NBS, consistent with the oxidation of tryptophanresidues.
Kineticanalysis,fluorometric titration of tryptophans and correlationtoresidualalliinaseactivityshowed that modification of onlyone residue present on alliinase led tocompleteinhibition of alliinase activity. To identify thisessentialtryptophan residue, weemployedchemicalmodification byNBS in thepresence and absenceof theprotectingsubstrate analogue,S-ethyl-L-cysteine (SEC) and N-terminalsequenceanalysis of peptidefrag ment isolated byreverse phase-HPLC. A fragmentcontainingresidues 179-188 was isolated.Weconclude that Trp182isessential for alliinase activity.
Keywords: Alliinaseactive site, Tryptophan of alliinase.
Introduction
Garlic (Allium sativum Lin花)is a widelydistributed plant and is used in allpartsof the worldnot onlyas a spice and a food, but also as a popular remedy. The odor characteristic of garlic is due to the degradationof S-alkyl-Lcysteinesul foxides byalliinase(EC4.4.1.4). Itis generally believed that thealliinase and theamino acidsubstrateare presentin sepa rated compartments invivo. Uponrupturing or wounding of the cells, the enzyme located in the vacuole and the alkyl cysteine sulfoxides located inthe cytoplasm can react and produce the volatile odorous compounds.1 Allicin (diallyl thiosulfinate or allyl 2-propenethiosulfinate) is the dominant thiosulfinateformedbytherapid condensation of twomole cules of 2-propenesulfenicacid,anintermediate of the alliin (S(+)-allyl-L-cysteinesulfoxide)lysisbyalliinase.
Stoll etal.2 were the first to isolate alliinase from garlic clove. Further investigations by others demonstrated that enzymes which catalyzed similar reactions are not only presentin Alliaceaebut also in Brassicaceae? Fabaceaef broccoli,5 and even in bacteria.6 Since the discovery of the alliinases,therehave beennumerous reports onthe identifi cation, purification and characterization ofalliinases from several Alliaceae species.7-10More recently, alliinasecDNA clones from garlic, shallot (A. ascalonicum L.), and onion (A. cepa L.) were isolated and characterized.11 The holo enzymefromgarlic is aglycoproteincontaining5.5%carbo- hydrate9 and consists of two identical subunits,8each with a
*Corresponding Author. Phone/Fax: +82-2-878-8545; e-mail:
chulyang@plaza. snu.ac.kr
molecular mass of about 55 kDa.11 There are very few reports citingtheimportanceofspecific amino acid residues in the function ofthis enzyme. Human serum was shown recentlyto contain antibodiestothe two major proteinsfrom cloves of garlic.12 These antibodies were directed against alliinase and mannose-specific Allium sativum agglutinin.
Alliinase from garlic has been shown to form a complex with agarlic mannose-specific lectinthrough a glycosylation site atAsn146ofthe enzyme.13 Morerecently, pyridoxal 5’- phosphate (PLP) has been shown to modifyLys251 in allii- nase from onion,14 confirming previous report thatPLP par
ticipates in enzymatic degradation of alliinase.9 There is neither 3-dimensionalstructuraldata nor any informationon the identity and roles ofresidues involved in the catalytic mechanism of theenzyme from any source.
Thegoal of thisstudy was to identify amino acidsimpor
tant for theactivityof alliinase in garlic.Wehave chemically modified theenzymeusingavarietyof reagents to reveal the importance of Trpresidues and to alesser extent Lys and Tyr groups. Bymodifying the enzyme with NBSinthe absence and presence of a substrate analogue, combined with N- terminal sequence analysis of peptides from proteolytic digests, we have identifiedthe essential tryptophan residue as Trp182.
MaterialsandMethods
Materials.Bulbsofgarlic (A. sativum L.) werepurchased locally.Thestandard substrate, S-ethyl-L-cysteinesulfoxide (SECS), was synthesized from S-ethyl-L-cysteine (SEC) purchased from Sigma, by oxidation with acid H2O2 as
described previously.15 Thefollowing chemical modification reagents were purchased from Sigma; dithiothreitol (DTT), succinic anhydride (SA), N-acetylimidazole (NAI), maleic anhydride, 1 -ethy l-3 -(3 -dimethy laminopropyl)carbodiimide (EDC), iodoacetic acid, phenylmethylsulfonylfluoride (PMSF), diethylpyrocarbonate (DEPC), 1-cyclohexyl-3-(2-morpholino- ethyl)carbodiimide (CMC). N-bromosuccinimide (NBS) was obtainedfrom WAKO. Polyvinylpolypyrrolidine, con- canavalin A Sepharose 4B, methyl a-D-mannopyranoside, pyridoxal 5’-phosphate(PLP),EDTA(disodiumsalt), pyru vic acid (sodium salt), 2,4-dinitrophenylhydrazine, guani dine hydrochloride (Gdn-HCl), and TPCK-treated trypsin were purchased from Sigma. Trifluoroacetic acid (TFA)for protein sequencing was obtained fromAldrich. Hydroxyapa tite, acetonitrile (HPLC grade), and trichloroaceticacid(TCA) were purchasedfrom Bio-Rad, Fisher, and Lancaster, re
spectively. Potassium dibasicphosphate, sodiummonobasic phosphate, and allotherreagentswereanalytical grade.
Enzyme Purification and Activity Assay. Alliinasewas purified from garlic as described previously except for buffer composition.9,16 The buffer A consisted of 50 mM NaKphosphate(pH 6.5), 5 mM EDTA 1 mM DTT, and10%
glycerol. NaKphosphate buffers were obtained by mixing varyingproportions of equimolar solutions ofNaH2PO4and K2HPO4 to obtain the desiredpH. The enzymepreparation wasover95%pure as judgedby resolving proteinbands by SDS-PAGE (10% gels) and by staining with Coomassie Brilliant Blue.17 Alliinase activity was measuring using a modificationof the method of Nock et al..9 The final reac
tion volume (0.5 mL) contained 100 mM NaK phosphate buffer(pH6.5), 0.25mMPLP, 4 mM SECS, and appropriate aliquots of enzyme. The reactionmixture wasincubated for 5 minat room temperature and the reaction was terminated bythe addition of 1 mLof 10% (w/v) TCA.Protein precipi
tates, if present, were removedbycentrifugation. The pyru vate produced by the alliinase was measured by the total keto-acid method ofFriedemann et al..18 Protein concentra tions were determined quantitatively by the method of Bradford19 using BSA as a standard. Forallinactivation and spectroscopic studies, the amount of protein used was expressed as monomer concentration (Mr =55,000).
Gener지 Chemic지 Modification Procedures. 10 juM of the enzyme protein was incubated with various concentra tions of chemical modification reagents in 1 mLof buffers under the following conditions: 10 mM phenylglyoxal, 2 mM SA, 10mMmaleicanhydride, and 10mMiodoacetate for 15 min and 20 mM NAI for 30 min in 0.1 M borate buffer (pH 7.0); 10 mM EDC and 10mM CMC in 0.1 M Mes buffer (pH 6.5) for 15 min; 5 mM PMSF in 0.1 M HEPESbuffer(pH6.5) for 15 min; 10 mM DEPC in 50 mM NaKphosphatebuffer(pH 6.5) for 30 min; 10 mMDTT and 5 mM NBS (in the dark)in50 mMNaK phosphate buffer (pH 5.5) for 5 min. The stock solutions of the modifying reagents used above were prepared with distilled water exceptCMC,DEPC,phenylglyoxal, and NBS that were dis
solved in absolute EtOH. The concentrations of the stock solutions of these reagents dissolved in ethanol were
adjustedat 2% (v/v), orless, inthefinalreaction mixture to ensure that the amountofaddedethanoldoes not disturbthe enzymatic activity. All these reactions were carried out at room temperature. Each appropriate aliquot out of the respective reaction mixturewas subjectedtoassay for allii- nase activity. Inthe NBS modificationreaction, the enzyme solution to react withNBS had to bedialyzedagainstbuffer A without DTT, which willbereferredtobufferA’ through
out this report, because NBS reacts with DTT faster than tryptophan. Therefore, DTT wasused for terminationofthe modification reaction with NBS. Since the absorbance at 450 nm was slightlyincreased by the product made by the reaction of NBS and DTT, this product andexcess ofNBS and DTT had toberemoved from the reaction tubesby dial
ysis, or gel filtration prior to enzyme assay when excess DTT (about more than300molarexcess) was used.In the experiments for acquiringthekinetic data ofinactivation by modifying reagents, after alliinase was treated with varying concentration ofSA, NAI, and NBS with changingreaction time,theresidualactivityofalliinasewasrecorded.
Effectsof NBSModification onAlliinase SpectralPro perties.
UV Absorption Spectra: The samples for measuring of UV absorption consisted ofthe following conditions. In all cases ofmodification reaction withNBS, the final concen
tration ofprotein was maintained at 5 uM. A Control not modified with NBS, “Modified 1” oxidized by 0.5 mM NBS, and“Modified 2” oxidizedby 1 mM NBS were pre
pared in bufferA’ (pH 5.5) at room temperature. The reac
tion mixture wasincubated for 1 minin thedark to minimize unspecificreactions and quenched by additionof a threefold molar excess of DTT. After reaction, these samples were dialyzed against buffer A’ (pH 5.5), and centrifuged to remove, if present, insolubleparticles. The sample concen
tration was adjusted to 0.2 mg/mL and added to a quartz cuvette. Ultraviolet spectra were scannedover the rangeof 240-320nm using a KontronUvikon930spectrophotometer at room temperature.
Fluorescence Emission Spectra: Fluorescence measure
ments were carried outwith a KontronInstruments (model SFM 25) spectrofluorometer. Fluorescence intensities were determinedusingan excitation wavelength of 295 nm.20The emission spectrawere obtained overthe range of 310-380 nm atroomtemperature. Themodificationprocedureswere identical with the above cases. It is unnecessary todialyze reaction mixtures, because only protein couldemitfluores cence by excitationat 295 nm. Thefinalprotein concentra tion in thecuvettewasadjusted to 1uM.
Titration of Unmodified Tryptophan Residues: The number of tryptophanresidues unmodified was determined by the Pajot method20 slightly modifiedfor this experiment.
600uLof 5 uMprotein, 2393 uLof6 M Gdn-HCl in 0.1%
NH4HCO3 (pH 8.0), and7 uL of14 M 2-mercaptoethanol wasdirectlysubmitted to a fluorescencecuvette.Afterincu bation for 1 h for protein denaturation, its fluorescence was measured at 354 nm asdescribed in thefluorescenceemis sion spectra procedures. The standard curve was obtained
with knownconcentrations of freeL-tryptophan solution.
Identificationof anEssentialTryptophan Residue. Determination of Modification ConditionofAlliinase: We have employed chemical modification with different concentrations of NBSfor 20 sec inbuffer A’(pH 5.0) in order to findthespecificcondition inwhich theactive site of alliinase could be protectedby SECfrommodificationwith NBS. The reaction had to bestopped quicklyby addition of a threefold molar excess ofDTTafter 20 sec. The residual activity of the respective samples was analyzed and the unmodified tryptophan residues were titrated as described previously.
Trypsin Digestion: The native alliinase and NBS-modi- fiedderivativesweredialyzedextensivelyovernightagainst 0.1% NH4HCO3 (pH 8.0) and lyophilized. The respective aliquots of10 nmolweredissolvedin 100
卩
L of8Murea, and then 100卩
L of0.4 M NH4HCO3 and2卩
L of0.5 M DTTwere added. These samples wereincubated ina water bath adjusted to 50oC for 30min,cooled to roomtempera ture, and then 20卩
L of0.1 M iodoacetamide was added.After 15 min, sample solutions were diluted by additionof 250
卩
L oftriply distilled water that will be referred to as tdH2O throughout the rest of this paper. Tryptic cleavage wasthencarried out with 22卩
L of TPCK-treatedtrypsin (1 mg/mL) atanenzyme/substrate ration of 1 : 25 (w/w)at 37 oC for 24 h. All ofthe abovereagents had to be dissolved in tdH2O. The digestionwas terminated by acidification with 20% TFA to pH 2-2.5.21 The insoluble particles were removed by centrifugation prior to separation by reverse phase-HPLC.Sampling for Reverse Phase-HPLC: Samples for sequencingwere diversely prepared bythefollowingproce dures. All protein samples had to be dialyzedagainstbuffer A’ (pH 5.0) prior tomodification with NBS. The concentra tion of proteinswas controlled at 5 ^Min all experiments.
“Control” was not treated with NBS. “Modified 1, 2,and 3” were modified with 1.5 mM NBS for 8, 20, and 30 sec, respectively,and quenched with a threefoldmolarexcess of DTT. “Protected 1” was incubated with 2.5 mM SEC for 20 min for protection ofthe activesite, dialyzed against buffer A’ (pH5. 0), and concentrated by a centricon.After the con
centration of “Protected 1” was adjusted to 5 p,M, it was modified with 5 mM NBSfor 20 sec, and then quenched.
“Protected 2” was prepared by modification with 10 mM NBS for 20 sec after incubation with2.5 mM SEC for 20 min. Allsamplesweredigested with TPCK- treated trypsin bythe method described previously.
Reverse Phase-HPLC Separation Condition: Separa tion ofpeptides generated by trypsin digestion of control, modified, and protectedsamples wasachieved usinga Beck man HPLC system (model: system GOLD). Separations were carried out usingaVydac 218TP54 C18 reverse phase column(300 A, 5pm,4.6x 250 mm). Inallcases, solvent A was 0.1%TFA in tdH?O : acetonitrile (99 : 1)and solvent B was 0.1% TFA in tdH?O : acetonitrile (2 : 8). The flowrate wasfixed to 1 mL/min. After sample injection, the solvent mixture waskeptconstantat0% sol. B for 10min and then
raisedto20%sol. Bover 30 min, 50% sol. Bover 60min, and 80% sol. Bover20 min. Absorbance was monitoredat 280 nm to have the advantage of recording absorbance of tryptophanylresidues.
Sequence Determination: N-terminal sequence analyses of alliinase were performed by automatic Edmandegrada
tion using a model Procise 491 (Applied Biosystems)auto matic gas phase sequencer at the Korea Basic Science Institute (KBSI).
Results
Inactivation Kinetics of AlliinasebyS이ected Chemic지 Modifying Reagents. It was revealed that tryptophan, lysine, and tyrosineresidues,specifically modifiedbyNBS, SA,and NAI, are critical for theenzymatic activity.Modifi
cation by the other reagents (DTT, EDC, PMSF, DEPC, CMC,maleicanhydride, idoaceticacid) seems not toinhibit theenzymaticactivity (notshownhere). SA has proved use ful in themodification oflysine.22Alliinaselost60% of its activity in 2 mM SAby the method described previously.
(>
--A=
윦
%)
2.5 2 1.5 1
0 5 10 15
2mM 3mM 5mM 0.5 10mM
0
-
■-Mo 96)ro
2
20 40 60
I* 5mM
◊ 10mM
■ 20mM A 40mM
(>
■->
Joe 一 汶
)
1
Time (min)
Figure 1. Time course of inactivation of alliinase by chemical reagents; Plots of Log (% activity) against time to determine the pseudo- first order rate constants (而).[A] succinic anhydride (SA) and [B] N-acetylimidazole (NAI) in 0.1 M borate buffer (pH 7.0), [C] N-bromosuccinimide (NBS) in buffer A (50 mM NaK phosphate buffer (pH 6.5), 10% glycerol, and 5 mM EDTA).
Figure 2. A plot of ki against the concentration of inhibitors to determine the second order rate constant k2 of inactivation of the enzyme by inhibitors. [A] SA, [B] NAI, and [C] NBS.
Lys251 is already known to participate in binding PLP,14 an important cofactor and lysine-modifyingreagentitself. NAI was used as a reagent for the selective modification of tyrosyl residues.23 The treatment of enzyme with 20 mM NAI led to inactivation to about 50%, suggesting that the tyrosyl residuesplay a roleincatalysisby alliinase. Alliinase was completely inactivated by addition of 5 mM NBS, selective for tryptophan,suggestingthatthetryptophan resi-
Table 1. Kinetic data for the reaction of alliinase with succinic anhydride, N-acetylimidazole, and N-bromosuccinimidea
modifying reagents modified residue (s)
k2
(M-1min-1)
reaction order (n) Succinic anhydride" Lys, Tyr 17.8 1.51
N-acetylimidazole" Tyr 0.4 0.883
N-bromosuccinimidec Trp 1398 1.14
aFor modification conditions, see Experimental Procedures. "These reagents were used in 0.1M borate buffer (pH 7.0). cThe reactionwith this reagent wasperformed in buffer A’ (pH 5.5).
dueis the mostprobableamino acid thatparticipates in the substrate binding or catalysisofalliinase.
To calculate the kinetics of inactivation by modifying reagents, the logarithm of enzyme activity remaining was plotted as a function oftime at different concentrations of reagents (Figure 1). Theseplots give a seriesofstraight lines at differentconcentrations ofreagents. The values of k1, the first order rate constants, can be obtained from the slopes of the straight lines. Theresults showthat with theincrease of reagent concentration, the value of k1 increased. The rela tionship betweenk1 and theinhibitor concentration [I] can be written24:
k1= k2[I]n or logk1 = logk2 + nlog[I]
where k1 and k2 arethe first and secondorder rate constants ofinactivation, respectively. [I] is the concentration ofthe inhibitor. A plot of k1 against[I]gives a straightline which passesthrough theorigin of thecoordinate,asshowninFig ure 2, wherethe slopes of the straightlinesgive the second orderrateconstants k2 (Table 1). Under thesereaction condi tions, the rate on inactivation using NBS as a reagent is much faster than observed with SA or NIA. The n is the reactionorder, i.e.the numberof essential residuesmodified.
Plot oflogk1 againstlog[I] gives astraight line, where the slopes of thestraight linesgivethevaluesof n (Table 1).The data imply that the loss of alliinase activity is due to the modification ofapproximately 1 Trp, 1 Lys, and1 Tyr resi
due. The stoichiometry of inactivation is slightly higher for SA and consistent with the fact that this lysine specific reagent can also modify Tyr groups, albeitnot effectively.
Likewise, NBSyielded areaction orderof1.14, indicating thatthis reagent can also modify Tyrgroups, although at a much slowerrate than Trp residues. Out of 9 Trpresidues present per monomer in the primary sequence, only the modification of a single Trp results in a loss of enzyme activity.
EffectofNBS Modification on Alliinase Spectr지 Pro perties. Incubation of alliinase with NBS produced charac teristic changes of its spectral properties depending on the NBS/alliinase molar ratio(Figure3).
UV Absorption Spectra: It hasbeen reported thatsome proteins with tryptophan residuesmodifiedby NBS have a decreaseofabsorbanceat280nm.21,25-27Increasingamounts of NBSled to a decrease in theabsorbanceat 280 nm, but an increase at 250 nm (Figure 3A), showing that tryptophan residuesweretransformed to their oxindolederivative. SDS- PAGE confirmed that NBS treatment did notresult in cleav
ageof thepolypeptidechains(notshown here).
Fluorescence Emission Spectra: The intrinsic fluores cenceemission spectrum wasalsostrongly affected(Figure 3B). As an excitation wavelengthof 295 nm was used, the decrease in fluorescence shows the modification of the tryptophan residues. With the increase of the concentration of NBS, the fluorescence emission intensity at 332 nm decreasedrapidly;moreover, when alliinasewastreated with 300-fold molar excess ofNBS for 5 min, its intrinsic fluore
scence emission couldbe hardly detected at 332 nm (data
튿(
。。
으 ①
°
드 은
。
g v
(汶 ) ① 은①
°
s은
。
근
으
一프
은
305 325 345 365
Wavelength (nm)
Figure 3. Effects of NBS modification on alliinase spectral properties. [A] UV absorption spectra of native alliinase (thin line) and 100- (thick line) or 200-fold (——)molar excess NBS-oxidized alliinase in buffer A (pH 5.5). The protein concentration in the cuvette was 0.2 mg/mL. [B] Fluorescence emission spectra of native alliinase (1) and alliinase oxidized with 10-, 30-, 50-, 100-, 200-, 300-fold molar excesses of NBS (2-7). Excitation wavelength was 295 nm. The protein concentration in the cuvette was 1 ^M.
not shown). This result indicates that all nine tryptophan residues present in alliinasecouldbeoxidizedbyNBS. For theemission spectra of the modifiedalliinase with different modified extents, no marked red shift was obtained. This suggests that the tryptophan residuesrelatively close to the surfaceof theenzymehavebeen modified and denaturation, or gross conformational changes, has not occurred in the modifiedalliinase.
IdentificationofEssentialTryptophan Residue.
Determination of Modification ConditionofAlliinase: Wetried to protectthe active site ofalliinase with SEC, a substrate analogue. Alliinasewaspreincubated with 500 fold molar excessofSEC and thenmodified with 300 fold molar excess of NBS in buffer A (pH 6.5) with 1 mM, 200 fold molarexcess DTT. But wefound that themodifiedalliinase did not showthedecreaseof the fluorescence emission,sug gestingthattryptophanresidues of alliinasewere not modi
fied with NBS. This result indicatesthatSEC and DTT react with NBS faster thantryptophan inpH 6.5. Thereforeallii- nase in buffer A had tobedialyzedagainstbufferA (with out DTT) prior tomodification with NBS. Since tryptophan residuesmaybe rapidly28 and specificallymodifiedatmildly acidic pH, which increases both reactivity29 and selectivity,30 we had to find theappropriate pH. This is thepH in which alliinasehastoberelatively stable and the specificreactivity
o I--- --- ---Lf-d 0 0 50 100 150 200 250 300
잉( ra드 - -
™ - 을 - 을 ) su Llds
d
」'
p 으
느
POEU3
Figure 4. [A] Concentration-dependent effects of NBS modifi
cation on the enzymatic activity of alliinase. 5 ^M alliinase was treated with variable amounts of NBS for 20 sec in buffer A (pH 5.0). Measurement of residual activity ( • ) and titration of unmodified tryptophan residues ( △) was performed by experimental procedures described previously. [B] Fraction of alliinase activity remaining (a) plotted against the fraction of residual Trp residues (b).
of NBS for tryptophan must be improved as comparedwith pH 6.5. The optimal pH was controlled to pH 5.0 inwhich alliinase shows75%residualactivity,relatively stable.
It is noted that the oxidation ofmethionine is a possible side reaction ofthe treatment of proteins with NBS.31 In ordertofind outwhat is produced bythereaction with NBS and SEC, a simple test was performed on the assumption that SEC would be converted toSECS, a substrate of allii- nase, by oxidationwith NBS, because SEC is similar with methionine in structure. SEC was preincubated with NBS for 5 min, quenched, and added to the alliinase solution.
After5 min, we checked enzymeactivityasdescribedinthe experimental procedures. The product ofthe reaction with NBS and SECshowed a similarcharacteristic with SECSas a substrate (not shown), showing clearly that SECmustbe converted to SECSby oxidation with NBS. SinceSECScan be located in theactive site, both SEC and SECSproduced byNBS can sufficiently protect the active site of alliinase.
However, ifthe incubationtime ofthe modification is long enough for the whole SEC to react with NBS, protection cannot beeffective,becauseall the SECSproduced was very rapidlydegradedbyalliinase. Reacting SEC-protectedallii- nase with NBS for 5 min resulted in completemodification, asevidencedby theabsenceof absorbingspecies (280 nm) normally resolved byRP-HPLC(data not shown). Therefore we also had to determine the modification time. Alliinase was modified with different concentrations of NBS for 20
0.007
( 튿
。'°
으 잉 듀 은
。sq<
0.003 0.001 0.005
30
10
25 75
응、
s 으
B
50
(wuo
ro으 8 듀 은
osq<
-0.001 -
0 100
0.007
(wuo
ro으 8 듀 은
osq<
-0.001 0 0.007 0.003 0.001
50
Time (min)
0.005
30
10 5?g) B
25 50 75
Time (min)
50
( 튿
。
覺°
° rou은
。导
100 -0.001
( 튿
。'°
으 잉 듀 은
。sq<
-0.001 - 0 0.005 0.003 0.001
25 50 75
Time (min)
50
%5으8
o
o 3
1
100
( 튿
。'°
으 잉 듀 은
。sq<
0 25 50 75
Time (min)
0.007 0.005 0.003 0.001 0.007 0.005 0.003 0.001 -0.001
汶s
으
B
o
o 3
1 50
O
0 25 50 75 100
次S
으
B
o
o 3
1
-0.001 - 0
Time (min)
0.005 0.003 0.001
so_
10 B
25 50 75
Time (min)
50
50
-10
100
Figure 5. RP-HPLC elution profile of tryptic peptides of Control [A], Protected 1 [B], 2 [C], Modified 1 [D], 2 [E], and 3 [F]; Sol. A was 0.1% TFA in tdH?O : acetonitrile (99 : 1) and Sol. B was 0.1 % TFA in tdH?O : acetonitrile (2 : 8). The flow rate was fixed to 1 ml/min.
sec. Modification with 300 fold molar excess of NBSledto complete inactivation. On the basis of these data, all the modifiedsampleswereprepared under theaboveconditions, i.e. 300 foldmolar excessof NBS for 20secinbuffer A’ (pH 5.0), except for modification of protected samples. In this condition, the number of unmodified tryptophan residues was 3.2 (Trp mol/alliinase mol), indicating that six of the nine Trpresidues are morereactive on NBS than the other three. Figure 5B shows the relationship between the frac tional activity remaining (a) and the fractional Trp residue remaining (b) ofthe enzyme asa Tsou plot.32 It can be seen that among the six reactive tryptophan residues modified only one is essential for the alliinase activity and three less reactive residues do not have an effect on the enzymatic activity.This fact wasevidently proven in the HPLCelution profile(Figure 5).
Tryptic Peptide An지ysis forActive Site Tryptophan Identification: On oxidation, the indole chromophore of tryptophan with a strongabsorbanceat280nm is converted to oxindole with a considerably lower absorbance at this wavelength. This decrease inthe absorbance at280 nm was
used to distinguish the tryptophanresiduesinthetryptic pep
tides of the control, modified, and SEC-protected peptides (Figure 5). Althoughtryptophan residues out of the active site in the protected alliinase were not totally modified by NBS (Figure 5B, C), thetryptophan residue in theactivesite could be efficiently protected from modification by NBS and distinguished from others since it is obviousthat trypto phan in the active center is much lessreactive for NBS by protection than thoseoutoftheactive site.
Nine main peaks, denoted 1, 2, 3, a, b, c, d, W73, and W182, weremonitoredfrom native alliinase (control) (Fig ure5A). Peak 1, 2,and 3 showed a littledecreaseat280nm in the absence of SEC (Figure 5D, E, F), indicating that these residues are buried in the relatively hydrophobic regions of the protein and accordingly slowly modified by NBS in comparison with otherresidues. Therefore it turns outthatthese three residues do not have an effect onallii- nase activity. This result is in accordance with data above (Figure 4). The absorbance ofpeak a, b, c, and dwas mark
edly reduced in SEC-protected sampleswhile the peaks of W73 and W182were slightly decreased in absorbance (Fig-
Table 2. N-Terminal sequence of tryptic alliinase fragments un
modified or less modified from NBS, a little decreased in absor- bance at 280 nma
Fragment Sequence
fragment denoted W73 predicted fragment 54-74 fragment denoted W182 predicted fragment 179-213
I QGCSADVAS...
I QGCSADVASGDGLFLEEYWK GYVCAGNAAN...
GYVWAGNAANYVNV . .PEGLLR aThe peak W73 and W182 obtainedin Figure 6 from native alliinase (control) was collected, lyophilized, and submitted to sequencing by automatic Edman degradation. The twoobtained N-terminal sequences were identified by comparison with predicted fragments [11]. Even though in the fragment W73tryptophan wasnot identified, we could determine that tryptophan is situated at 73 by comparison with the predicted fragment.X was notdetected bysequencing, butmust be W.
Figure 6. Changes of fluorescence intensity (a) and the number of unmodified tryptophan residues (b) upon each sample in Figure 6.
ure 5B, C). As compared with theControl,peaks a, b, c, and d showed about 57%, 86%, 78%, and 70% oxidation, respectively on the average, indicating that these residues were not protected by SECfrom modification. In the cases of W73 and W182, these two residuesappeartobeprotected by SEC from modification,suggesting that at least one of the two residues is located in the active center of alliinase. In addition, we prepared samples protectedby various proce- dures. In these samples, the peaks of W73 and W182 were similar in Figure 5B, C (not shown). Therefore these two fragments were collected, lyophilized, and submitted to 10 cycles of protein N-terminal sequencing by automatic Edmandegradation. The two obtained N-terminalsequences were identified by comparison with predicted fragments (Table 2).
In order to determine which residueofthe twois directly involved in the active site, alliinasewasmodified with NBS for3, 7, and 20 sec. The activity of alliinase showed 60%
residualactivity at 3 sec and wasperfectly inactivatedat 7 sec (result not shown). Therefore, 8 sec is enoughtime to inactivatecompletely theactivity of alliinase. InFigure5D, RP-HPLCprofileof samplemodified for 8 sec, peaks a, b, c could be observed and especially W73, which could be detectedevenin Figure 5E, was a little oxidized. This de monstrates that these residuescouldbe excludedfromprob abilitythatamong these residuesthere might be anessential residue. Consequently, we come to the conclusion that Trp182is situated intheactive, or binding, siteofalliinase and plays animportantrole in the activity, sincethe modifi
cationofW182 alone is sufficienttoinduce a complete loss of activity.
Figure 6 shows changes of fluorescence intensity (a) and the number of unmodifiedtryptophan residues (b) in each samplein Figure 5. The totalintensity of peaksinRP-HPLC is well in accord with thetotal fluorescence intensity in Fig
ure 6. In particular, it shows that the decrease ofpeak inten
sity in modified samples (Figure 5D, E, F) is modification time-dependent.The number ofunmodified tryptophanresi duesbetweenB and Dwasapproximatelytwo. This demon
strates that theessentialTrpresidue(s) should beoneinstead of tworesidues, for the fluorescence effect of non-essential residues inB that were not modifiedon account of interfer
enceof SEC must beexcluded.
Discussion
Chemic지 Modification of Alliinase with NBS. Oxida tion of thetryptophanindolemoietyintooxindole by NBS is particularly interesting since the additional sterichindrance of the substituted group is very limited, contrary to other chemical modifiers,30 and since the modified residues become totally nonfluorescent.33 No subunit dissociation is observed in polyacrylamide gel electrophoresis under non- denaturating conditions(not shown), consistent with thelack of red shiftofthe fluorescence emissionspectra. No peptide bond cleavage is apparent under the modifying conditions used, as monitored by subsequent analysis of alliinase on SDS-PAGE (notshownhere). Thisagrees with theobserva tion that proteins aremore susceptiblethan model peptides to NBS-dependent tryptophan modification and cleavage, which additionally requiredenaturatingconditions, strongly acidic pH and a long incubation time.34,35 It is known that NBS can react with not onlytheindolegroup of tryptophan but also the SH group of cysteine in a number of pro- teins.25,36 Our results have shown that the modification of SHgroupsdid not haveaneffecton theinactivation of cata lytic activity ofthe enzyme by chemicalmodification with iodoacetate and DEPC, which modify cysteine residues.
Therefore, theresultspresentedinthe above sections show clearlythat themodification of tryptophanresiduesofallii- nase by NBS leads to the complete inactivation of this enzyme. Eventhoughthe modificationoftyrosine residues, which might beinvolved in theactive site of alliinase,may have occurred, it turns out that the tryptophan residue is much moreessential for activityofthe enzyme than tyrosine through thechemical modificationstudy. Thelinearityofthe plot and the analysis for inactivation rate and Trp residue modification rate show that among the nine tryptophanyl residuesonlyone is anessentialTrpresidue.
PropertiesofTryptophans in Alliinase. Alliinase modi
fied with NBS for about 5 min becomes nonfluorescent, sug gesting that all tryptophan residues become completely oxidized by NBS. In general, NBS oxidizesexposed but not buried indole side chains in folded proteins. For example,
Table 3. Comparison of the amino acid sequences surrounding the specific tryptophan residue in alliinase from garlic (A. sativum L.), shallot (A. ascalonicum L.), and onion (A. cepa L.)
Species Alliinase Sequence
65 182 191
Garlic YPVFREQTKYFNKKGYVWAGNAANYVN
Shallot YPVFREQTKYFDKKGYEWKGNAADYVN
Onion YPVFREQTKYFDKKGYEWKGNAADYVN
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
Program is gratefullyacknowledged.
References
Lancaster, J. E.; Collin, H. A. Plant Sci. Lett. 1981, 22, 169.
Stoll, A.; Seebeck, E. H이v. Chim. Acta 1949, 32, 197.
Mazelis, M. Phytochemistry 1963, 2, 15.
Mazelis, M.; Creveling, R. K. Biochem. J. 1975, 147, 485.
Hamamoto, A.; Mazelis, M. Plant Physiol. 1986, 80, 702.
Nomura, J.; Nishizuka, Y.; Hayaishi, O. J. Biol. Chem.
1963, 238, 1441.
Tobkin, Jr, H. E.; Mazelis, M. Arch. Biochem. Biophys.
1979 193 150,,.
Nock, L. P.; Mazelis, M. Arch. Biochem. Biophys. 1986, 249, 27.
Nock, L. P; Mazelis, M. Plant Physiol. 1987, 85, 1079.
Fujita, M.; Endo, M.; Sano, M. Agric. Biol. Chem. 1990, 54, 1077.
Van Damme, E. J. M.; Smeets, K.; Torrekens, S.; Van Leu
ven, F.; Peumans, W. Eur. J. Biochem. 1992, 209, 751.
Tchernychev, B.; Rabinkov, A.; Mirelman, D.; Wilchek, M. Immunol. Lett. 1995, 47, 53.
Rabinkov, A.; Wilchek, M.; Mirelman, D. Glycoconj. J.
1995, 12, 690.
Kitamura, N.; Shimomura, N.; Iseki, N.; Honma, M.;
Chiba, S.; Tahara, S.; Mizutani, J. Biosci. Biotech. Bio- chem. 1997, 61, 1327.
Stoll, A.; Seebeck, E. Adv. Enzymol. 1951, 11, 377.
Won, T.; Mazelis, M. Physiol. Plantarum 1989, 77, 87.
Laemmli, U. K. Nature 1970, 227, 680.
Friedemann, T. E.; Haugen, P. E. J. Biol. Chem. 1943, 147, 415.
Bradford, M. M. Anal. Biochem. 1976, 72, 248.
Pajot, P. Eur. J. Biochem. 1976, 63, 263.
Puri, K. D.; Surolia, A. J. Biol. Chem. 1994, 269, 30917.
Klotz, I. M. Methods Enzymol. 1967, 11, 576.
Riordan, J. F.; Wacker, W. E. C.; Vallee, B. L. Biochemis
try 1965, 4, 1758.
Hollenberg, P. F.; Flashner, M.; Coon, M. J. J. Biol. Chem.
1971, 246, 946.
Freisheim, J. H.; Huennekens, F. M. Biochemistry 1969, 8, 2271.
Divita, G.; Jault, J. M.; Gautheron, D. C.; Di Pietro, A.
Biochemistry 1993, 32, 1017.
Zheng, W. Z.; Chen, Q. X.; Zhao, H.; Zhang, Z.; Zhang, W.; Zhou, H. M. Biochem. Mol. Biol. Int. 1997, 41, 951.
Ohnishi, M.; Kawagishi, T.; Hiromo, K. Arch. Biochem.
Biophys. 1989, 272, 46.
Spande, T. F.; Witkop, B. Methods Enzymol. 1967, 11, 528.
Lundblad, R. L.; Noyes, C. M. In Chemical Reagents for Protein Modification; Lundblad, R. L., Noyes, C. M., Eds.; CRC Press: Boca Raton, FL, U. S. A., 1984; p 47.
Spande, T. F.; Witkop, B. Methods Enzymol. 1967, 11, 498.
Tsou, C. L. Sci. Sinica. 1962, 11, 1535.
Imoto, T.; Forster, L. S.; Rupley, J. A.; Tanaka, F. Proc.
Natl. Acad. Sci. U.S.A. 1972, 69, 1151.
Ramachandran, L. K.; Witkop, B. Methods Enzymol.
1967, 1 1, 283.
Shechter, Y.; Patchornik, A.; Burstein, Y. Biochemistry
The essentialtryptophanresidue (W) is indicated in boldface type.
only 4 ofthe 8tryptophans in xylanase A fromSchizophyl- lum commune3 and3of the 11 tryptophans inEGIIfrom T.
reesei3 are oxidized byNBS. Therefore, itis unusual thatall nine tryptophans in alliinase are susceptible to oxidation.
Threeresidues (peaks 1,2, and 3)are locatedinarelatively hydrophobic region of alliinase and are less reactive for NBS thanothers. However, theyare not completely seques teredwithin the interior of alliinase and may still be accessi bletoNBS. Theothersareexpected to lie nearthesurface of theprotein.
Sequence Comparison of Alliinasewith Others. Amino acid sequences of the alliinase have been deduced from
cDNAs isolatedfrom garlic, shallot, andonion.* * * * * * 11 * * * * * * * * * * * * * * * * * * * * * * * * Compari son of thealignedsequencesof thesealliinase (Table 3) indi
cate that Trp182 in alliinase is perfectly conserved among these species and amino acid sequence surrounding this tryptophan is alsohighly conserved.Thus,thisregion, espe
ciallyW182,may constitute a partof the enzyme activesite.
This paper is the first evidence for the localization ofan essential tryptophan residue (W182) in alliinase. Another significantvalue is anew approach of a chemical modifica
tionmethod. By chemical modification, there have been a number ofproblems inidentifyingessentialresidue(s)among the probable residues. In the case oftryptophan modifica
tion, there can be three problems. First, it is difficult to obtainthe site-specificchemicalmodificationof tryptophan residues in protein without any sidereactionsofother amino acid residues. Second, since the tryptophan residues in the hydrophobic regions in the enzyme are unreactive, or less reactive, there may be a possibility that essential Trp resi dues in these regions cannot be identified. Third, in many cases, it is hard to obtain complete protectionof the active site ofthe protein with substrate analogues, or active site binding molecules from the modification with NBS, for NBSis veryreactive.Forexample,Brayet al.39and Clottes et al.40 could not identify essential tryptophan residue(s), because they could not protect the active, or binding, site effectively. In our study, weovercame the protection prob lem byregulatingthe reaction conditionsusing kineticinfor mation.
Acknowledgment. We are grateful to Hae-Young Kwon and Hoong-Sun Park, Seoul National University, for valu able advice and helpful discussion and Myoung-Hee Nam, KBSI, for performing N-terminal sequencing. This study was supportedby a Grant from the Ministry ofEducation (2000). Financial support in part from the Brain Korea 21
1976, 15, 5071.
36. O’Gorman, R. B.; Matthews, K. S. J. Biol. Chem. 1977, 252, 3565.
37. Bray, M. R.; Clarke, A. J. Biochemistry 1995, 34, 2006.
38. Macarron, R.; Henrissat, B.; Claeyssens, M. Biochim. Bio
phys. Acta 1995, 1245, 187.
39. Bray, M. R.; Johnson, P. E.; Gilkes, N. R.; Mcintosh, L. P;
Kilburn, D. G.; Warren, R. A. J. Protein Sci. 1996, 5, 2311.
40. Clottes, E.; Vial, C. Arch. Biochem. Biophys. 1996, 329, 97.
