Purification and partial characterization of draculin, the factor present in the saliva of vampire bats (Desmodus rotundus)

Citation for published version (APA):

Apitz-Castro, R., Beguin, S., Tablante, A., Bartoli, F., Holt, J. C., & Hemker, H. C. (1995). Purification and partial characterization of draculin, the anticoagulant factor present in the saliva of vampire bats (Desmodus rotundus). Thrombosis and Haemostasis, 73(1), 94-100. https://doi.org/10.1055/s-0038- 1653731

Document status and date: Published: 01/01/1995

DOI: 10.1055/s-0038-1653731

Document Version: Publisher's PDF, also known as Version of record

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website. • The final author version and the galley proof are versions of the publication after peer review. • The final published version features the final layout of the paper including the volume, issue and page numbers. Link to publication

General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement: www.umlib.nl/taverne-license

Take down policy If you believe that this document breaches copyright please contact us at: [email protected] providing details and we will investigate your claim.

Download date: 03 Jun. 2020

Thrombosisand Haemostasis @ F.K. SchattauerVerlagsgesellschaft mbH (Stuttgart) 73 (1) 94-100(1995)

Purificotionond PortiolChorocterizolion of Dtoculin, theAnticoogulont Foctor Present in theSolivo of VompireBots (Desmodusrotundus)

RofoelApitz-Costror, Suzette Beguin2, Alfonzo Toblonter, Fulvio Bortolir, John C, Holt3, H,Coenrood Hemker2

Fromthe rlnstituioVenezolono de lnvestigocionesCientificos (lVlC), Corocos, Venezuelo; 2CordiovosculorReseorch lnstutute, Moostricht, The Netherlonds; ond 3ProteinChemistry Dept, Rh6ne-PoulencRorer, Collegeville PA, USA

Summory Here, we report resultson the isolation and partial characterization of the anticoagulantfactor presentin vampirebat saliva and we deter- Fromthe saliva of thevampire bar Desmodus rotundus, we isolated minedits siteof actionin the coagulationmechanism. We havepurified an unknownanticoagulant protein which we havenamed draculin. the anticoagulantfactor to apparenthomogeneity and show that it is an Its molecularmass as determined by non-reducedSDS-PAGE is about acidic(pI = 4.15) 88.5kDa glycoproteinwhich interactswith purified 83kDa Thereduced polypeptide shows a slowermigration. HPLC in factor IXa andXa and inhibits generationin humanplasma. a molecularsieve matrix yields a single,symmetrical peak comespond- ing to 88.5kDa. Isoelectricfocusing shows an acidicprotein with pl = 4.14.2. Aminoacidanalysis is compatiblewith a singiechain Moleriols ond Melhods polypeptideof about80 kDa. Cyanogenbromide cieavage yields a SephacrylS-200 (Pharmacia, Sweden), hydroxyapatite (BioGel HTP), Affi- single16-aminoacid peptide, conesponding to the amino-terminusof Gel-15,and acrylamide(BioRad Labs), Activated Thrombofax, and rabbit thenative molecule. Draculin inhibits the activated form of brainpartial Thromboplastin (Ortho Diagnostic Systems, Raritan, NJ, USA) factorsIX andX It doesnot act on thrombin, , and wereobtained from commercialsources. Coagulation factors were purified doesnot express fibrinolytic activity. The inhibition is immediateand by establishedprocedures by Dr. Rob Wagenvoord(Dept. of Biochemistry, notreadily reversible, with a stoichiometryof abouttwo moleculesof Univ of Limburg,Maastricht, The Netherlands). Chromogenic substrates were ' draculinper molecule of factorIXa or Xa. Surprisingly,the inhibitory obtainedfrom Kabi (Sweden) Thiobenzyl Benzyloxicarbonyl-L-lysinate HCI . activityagainst either factor is notaffected by thepresence of theother. and p'-nitrophenyl-p-guanidobenzoateHCI were from Sigma All other reagentsused were also of thehighest quality available. The buffers used were Draculinbinds quantitatively to eitherimmobilised factor Xa or factor A: 0.05M Tris-HCl,0 I M NaClpH 7.35,with 0.5% of eggalbumin (Sigma) IXa. Our preliminaryinterpretation is that thereare two forms of andB: BufferA atpHT 9 with 20mM EDTA. draculinthat hardly differ in structure.Both bind to factorXa andto factorIXa but oneform inhibits factor Xa andthe other inhibits factor Animals IXa. Whenadded to plasma,draculin increases the lag phase as well as Vampirebats (Devnodus rotundrs,) were regularly captured from wild colo- theheight of thepeak of thrombingeneration. niesliving in a cavein thenorthwest part of Venezuela(State of Falc6n)in a regionwhere rabies is reportedlyabsent. Twenty vampires were curently kept in captivity,in individualcages of themetabolic type (Acme Metal Products, lnlroduclion Chicago,IL, USA),under controlled light and temperature (25o C). The animals Theanticoagulant propefiies of vampirebat saliva have been intui- weremaintained on bovineblood anticoagulated with 3.2%sodium citrate at a ratioof I : Foodwas given every 24 h, alwaysin thelate afternoon. Water tively knownfor manyyears as is witnessedby their name.Vampir, 9, wasgiven ad libitum (6). sincethe earlymiddle ages or evenbefore indicated drinking "semi-dead"in easternEuropean folk1ore. The nameevidently has beenextended to haematophagousbats after these have become known SalivaCollection to Europeans,i. e. after1492. One of thefirst documentedaccounts on For salivacollection, the vampireswere anaesthetized with a mixtureof the anticoagulanteffect of salivafrom an haematophagousbat was 2.5V02-bromo-2-chloro, 1,1,1-trifluorethane (Halothan, Hoechst), 30% nitrous donein 1932byBier, and some years later by Romana(1,2). After oxide,in oxygen(5, 6). Onceanaesthetized, 20 pl of 1%pilocarpine (Isopto these,very few papershave appeared on this subject.At present,the Carpin,Alcon Labs.,Inc, Ft Worth,TX, USA) wasplaced in the mouthin bestcharacterized factor from vampirebat saliva is a recentlyisolated orderto stimulatesalivation. It wasobserved that incidentally wounding of the plasminogenactivator from extracts of vampiresalivary glands (3). It is gingivaewith the plasticpipette tip that was usedto applythe pilocarpine notclear if thiscomesponds to the plasminogen activator from vampire causeda very minor bleedingthat stoppedalmost immediately. Saliva was plasticmicrocentrifuge tubes, placed in ice.The saliva collection batsaliva previously described by Hawkey(4) andCartwdght (5). collectedin took 30 to 40 min, with a regularyield of about0.5 ml/animal. Individual sampleswere kept at -30o C untiluse.

lsolationof theAnticoagulant Factor from VampireBat Saliva Conespondenceto: Dr. H. C Hemker,Department of Biochemistry,Medi- cal Faculty,University of Maastricht,P O Box 616,N-6200MD, Maastricht, The protocolroutinely used was as follows:Saliva (=5 ml, 15mg of TheNetherlands - FAX Number:+3 1 43 670988 protein)was thawed and dialysed for six hoursagainst 10 litres of distilled

94 | .00 waterThe dialysate was centrifuged at 48,000X g during20 min andthe clear supematantwas lyophilized and redissolved in 1.5ml of distilledwater This = wasloaded on a Sephacryl5-200 column (2 5 x 30 cm) andeluted with dis- - tilled waterat a rateof 0 5 ml/min at room temperatureFractions of 2 ml werecollected and the relativeprotein concentration as well as the anti-Xa =^ >c activitywere measured every two fractionsAnti-Xa activity of the fractions =ts was assayedusing a micro-plateadaptation of the FXa assaybased on the hydrolysisof the chromogenicsubstrate 5-2222 as follows: Each microplate dE well contained6 25nM FXa, 50mM Tris pH 7.35 and an aliquotof the pl 20 pl 4 mM fractions,in a totalvolume of I 50 Reactionwas started with of E = 3-2222and after 10 min of incubationat 37" C theabsorbency at 405nm was recordedin a micro-platereader (Bio-Tek, Model EL312e). Fractions with anti-Xaactivity were pooled and lyophilizedThe lyophilizedmaterial was 25 50 75 t00 125 dissolvedin 1 mM NaCl(3-5 ml) andthe pH adjustedto 7 2 with NaOH.The FRACTIONM}I8ER samplewas loadedon a hydroxyapatitecolumn (1 X 7 5 cm, Biogel HTP) Flg.1 Proteinand anti-FXaactivity profile from fractionsobtained after equilibratedwith 1 mM NaCl,pH 7.2 Afterall of thesample was loaded, the molecularsieve of lyophilisedvampire bat saliva through sephacryl S-200 Co- columi was washedwith 6 columnvolumes of 1 mM NaCl (pH7 2) fbl- lumnsize was 2.5 X 30cm; equilibration and elution was done with delonized lowedby 2 columnvolumes of 200mM potassiumphosphate, pH 6 8 At waterand 2 ml fractionswere collected Dashed line comespondsto the inhibi- this point,a phosphategradient from 200mM to 1 M (otal volume= 30ml) tory activityassayed with the microplatesystem described in Methodsand wasinitiated Oneml fractionswere collected at a rate of 0.3ml/min. The expressedas the decrease of absorbencymeasured at 405nm Solidline cor- anti-Xaactivity appeared at about0 3 M phosphate,while most of theprotein respondsto therelative protein concentration determined by Bradford'smethod elutedwith theinitial washing and the 200 mM phosphateThe active fractions werepooled and concentratedby ultrafiltrationthrough an AmiconXM-50 filter. A highdegree of purificationcan also be obtained after cluomatography of crudesaliva (dialysed against 1 mM NaCl)on hydroxyapatite, with omissonof theSephacryl step. Draculin obtained by this singlestep procedure is always slightlycontaminated by a lowermolecular weight polypeptide which elutes 02 14P04 P04 GRADIENT fromthe hydroxyapatite column at thesame phosphate concentration as dracu- I ^^^o.- I lin. o o 94 lqorLJraNrwASHING I

CoagulationAssays F 1 mMNaCl

F a) Humanwhole blood coagulation time (WBCT)was measured in one O ml samplescollected after discardingthe first 2 ml of blood Coagulation 6t I wasallowed to proceedat3'7'C, in glasstubes containing 50 pl of EACA F z (114mg/ml). l) Activatedpartial thromboplastintune (aPTT)was measuredusing a commercialpartial Thromboplastin (Activated Thrombofax, Ortho Diagnostic 05152535 Systems,Raritan, NJ, USA), which contains Elagic acid as activator FRACTIONNUMBER c) Prothrombintime (PT) was measured using a commercialBrain Throm- Flg 2 Antr-FXaactivity profile of materialeluted from a hydroxyapatite boplastin/calciumchloride reagent (Ortho) column.Column (1 X 7 5 cm)was loaded with active material obtained from d) Thrombintime (TT) was determinedwith bovinethrombin (Sigma SephacrylS-200. Fractions volume was I 0 ml.Inhibitory activity expressed as Chem,St Louis,MO, USA) Experimentalconditions were chosen in orderto in Fig. 1 SeeMethods section for details obtaina TT of 30 secondsfor thecontrol. Al1coagulation times were read manually by tiltingthe tubes.

(PS,PC (5 FactorXa Activity (10nM), phospholipid 20/80,I pM), andCa++ mM).The samples containedup to I nM of factorIXa mixedwith up to 2 nM draculin.Factor IXa FactorXa wasassayed by its activityon thechromogenic substrate 5-2337 anddraculin were incubated for at leastone minute before the othercompo- (KabiDiagnostica, Sweden). Usually 10 pl sampleswere put in 0.490ml of nentsof tenasewere added, except for theexperiment of Fig 7 Thereaction bufferB containing4 mM of thesubstrate. The reaction was followed at 37" C, was initiatedby additionof purifiedfactorx (200nM) The productionof at 405nm in a homemade fixed dual-wavelengthphotometer (Biochemistry factorXa wasfollowed on subsamplestaken into EDTA buffer,containing the andInstrumental Dept., Univ of Limburg),and all relevantparameters were chromogenicsubstrate 5-2337, as described above. Typically 1 nM of factor calculatedby a dedicatedsoftware developed for thatinstrument (Laudy, P., IXa will causea reactionvelocity of 200nM qf factorXa beinglbrmed per InstrumentalDepartment, Univ. of Limburg) FactorXa concentrationwas minutein thissystem Because never more than 2 nM of draculinwas added per determinedby titrationwith p'-N-p-guanidobenzoateasdescribed by Jameson nM of factorIXa, the inihibitoryaction of draculinon factorXa would not et al.(7). significantlyinterfere with theoutcome of thefactor IXa determination Additionally,FIXa amidolytic activity was tested on thiobenzyl benzylocar- bonyl-LJysinateas described by Greenand Shaw (l 0) FactorIXa Activity

FactorIXa was assayedaccording to the principlesworked out by van Determinationof OtherProteases Dieyenet al (8) asfurther elaborated by Wagenvoordet al. (9) We measured theenzymatic activity on factorX activationin a systemwhere the non-factor The followingproteases were tested on the chromogenicsubstrates indi- IXa componentsof the tenasecomplex are presentin excess:factorVllla cated:thrombin-S2238; trpysin-S2160; chymotrypsin-S2586; plasmin-S2160; 95 MWS(kD) Effectof ImmobilizedFlXa or FXaon the Activityof Draculin Draculinwas incubated for twominutes with 150pl of inactivatedAffiGel- 15 (control)and after centrifugation for oneminute at 12,000X g an aliquot of the supernatantwas assayed for inhibitionof FIXa or FXa in thestandard assayThe amount of draculinwas chosen to obtain95-100% inhibition of the 94 enzymaticactivity of thecoresponding factors. Identical amounts of draculin w wereincubated with a) FXa-AffiGel,or b) FIXa-AffiGeland similarly tested

Thrombingeneration experiments (TGT) were executedas previously 6T described(15).

AminoacidComposition and Partial Aminoacid Sequence

43 Draculinwas further purified, prior to proteinchemistry, by reversephase HPLCon a C-1 8 column(Vydac 201 TP, The Separations Group, Hesperia CA, or AsahipakODP-50, from AnspecCo., Ann Arbor,MI) equilibratedin 0.1% trifluoroaceticacid and developed with a gradient(0.87o per minute) of aceto- nitrile.Each column yielded a singlepeak. Amino acid composition was deter- s 30 minedby derivatisationwith phenylisothiocyanate (PITC) after vapour phase acidhydrolysis for 16h at 110"C ("Picotag"methodology, Milligen Corp, WatersDivision).

{r 20.I Resullsond Discussion

Theanticoagulant factor present in vampirebat saliva was pudfied to homogeneityas described in theMethods section. Fig. 1 showsthe 14.4 proteinseparation patterrl and anti-Xa activity obtained after molecular sieveof dialysed,lyophilized saliva, through Sephacryl 5-200. Fig.2 showsresults of thechromatography of the pooled, active fractions, on a columnof hydroxyapatiteas describedin Methods.Most of the proteincomes through with the washing (1 mM NaCl)and 0.2 M phos- phate(pattern not shownin the figure),while anti-Xaactivity only appeaNafter establishment of the gradient from 0.2M to 1.0M phos- phate,at about0.3 M phosphate. fig.3 SDS-Polyacrylamidegel electrophoresisof draculin.Lane 1: crude SDS-polyacrylamidegel electrophoresisof the active material (10 saliva;lane 2: non-reduceddraculin (10 pg), lane3: reduceddraculin pg); elutedfrom the hydroxyapatitecolumn, showed a singleband cone- Lane4 conespondsto the low molecularweight markers from Pharmacia. spondingto a molecularmass in therange of 80 to 85kDa. Under re- Proteinstainine with CoomassieBlue ducingconditions, this band migrates as an apparentlyslightly higher molecularmass component, suggesting the presenceof intra-chain disulphidebonds (Fig. 3, lanes 2 and3). The protein band gives a strong positivereaction for carbohydrateswith theperiodic acid-Schiff stain plasmakallikein-S2302. Their activity was tested in thepresence and in the (notshown). of dracu- absenceof roughlyequimolar as well asup to tenfoldexcess amounts HPLCof thepurified native draculin showed a singie,symmetrical lin. proteinpeak eluting at a positioncoffesponding to a molecularmass of Proteindetermination was done using the methoddescribed by Bradford (11). 88.5kDa (Fig. 4). This result was obtained using 0.25 M phosphatefor lowerionic strengtheluent SDS-Polyacrylamidegelelectrophoresis was done as described by Shagger elutionof thecolumn. Several trials using andvon Jagow (12). Periodic acid-Schiff staining for giycoproteinswas done resultedin apparentbinding of theprotein to thecolumn matrix, which asdescribed (13). markedlyretarded elution of theprotein. Fig. 5 showsthat isolelectric Isoelectricfocusing was performedon aSarosein a Multiphor LKB focusingof draculinin anagarose based gel resulted in a singlepoly- (Sweden)electrophoresis chamber as described by Vesterbergand Gramstrup- peptideband which immobilizes at pI = 4.14.2. Christensen(14), with minor modifications.Ampholine (pH 3 5-10, LKB, The aminoacid composition of non-reduced(native) and reduced, Sweden)was added to obtaina finalconcentratton of 5Vo. alkylated(S-pyridilethyl) draculin is shownin Table1. In SPEdra- HPLCof purifieddraculin was performed in a ShimadzuHPLC equipment culin,seven moles of SPEcysteine per moleof proteinwas also re- (Shrmadzu (10 pm Corp.,Japan) using a silicabased molecular sieve column covered,Form the aminoacid compositiona molecularweight of ProteinPak 300 sw,Waters,Millipore Cory., USA). Equilibrationof the co- approximately80 Da canbe calculated, which is in agreementwith the lumnand elution of thesample was done with 250mM potassiumphosphate, molecularmass obtained from SDS-PAGE. The contribution of thecar- pH 6.8.Calibration of this columnwas done with molecularweight markers thepolypeptide fromPharmacia (Sweden). bohydratemoiety seems to amountto aboutI\Vo of Immobilizationof FIXaand FXa on AffiGel-L5 was done as indicated in the weight,as indicated by themolecular mass obtained from HPLC mole- manufacturerbrochure, using 1.2 mg of eachpurified factor for couplingto cularsieve of nativedraculin (88.5 kDa). Cleavage of draculinwith 0 5 ml of gel (bedvolume). cyanogenbromide yields a singlepeptide with apparentmolecular

96 weightaround 7000 Da, with thesequence NH2-ARRRGVRWCTISK- PEA.This sequencecontains the 16-aminoacidsequence from intact draculin,indicating that it representsthe amino-terminus of themole- cule. C Drac ulin Preliminaryexperiments on theeffect of draculinon bloodcoagula- E I tion indicatedthat draculin prolonged the coagulationtime of whole humanblood (WBCT), as well asthe prothrombin time (PT, extrinsic F pathway)or theactivated partial thromboplastin time (aPTT, instrinsic z. O pathway)of citratedhuman plasma. About 50 nM final concentration F of draculindoubled the aPTT and the WBCT. The same concentration z. F prolongedthe PT from 14(control) to 18s. Under the same conditions, L]l t draculindid not affectthe thrombintime or the activityof purified thrombinon thechromogenic substrate 5-2238. Neither did it inhibit plasmin trypsin,chymotrypsin, or plasmakallikein. These results 48 53 58 suggesteda relatively specific anticoagulant effect located in the co- LOG NlOLECULARMASS aguiationcascade at the levelof factorX, or factorX activation.To furtherdefine the mechanism(s) of the anticoagulant effect of draculin Fig.4 HPLCof purifieddraculin. A ProteinPak 300sw column (7 8 x 300 we investigatedits effecton thecatalytic activity of purifiedfactors. mm),was equilibrated and eluted with 250mM potassiumphosphate, pH 6 8 (kDa) to molecularweight markers from Pharmaciaused Fig.6 showsthe progress of thehydrolysis of thechromogenic sub- Figures conespond for columncalibration (n = 5, t SEM).No differencein retentiontime of the strateS-2337 by factorXa,expressed as the change in absorbencyat markerswas observed with Tris or HEPESbuffers The insert conesponds to 405nm. Additionof draculinto the systemresults in an immediate a typicalrun of 10pg of draculin(S = start,E = endof therun) inhibitionof thecatalytic activity of theenzyme. iuggesting a veryfast interactionbetween factor Xa anddraculin. We repeatedthis experi- mentat a seriesof factorXa anddraculin concentrations but inhibition alwayswas apparently complete within the 5-10 s thatare required for additionand mixing of draculinin the solutionNo stoppedflow ex- perimentscould be caniedout in our laboratories.To our surprise draculinalso inlibits theactivation of factorX catalysedby factorIXa. If 0.3,0.5or 1.0nMof factorlXais incubatedwith anequimolar or excessamount of draculin,then, after addition of theother components z of thetenase system in excess,i.e. factorVllla(10 or 20nM) and o t phospholipid(2 pM), O or 20 nofactor X convertingactivity is observed. - Thisis illustratedin Fig.7. Thisfigure also shows that the order of ad- o 4f-pl=41-42 ditionof thereaction components influences the velocity of inhibition. Thismeans that factor IXa in thetenase complex system is to a certain ril degreeprotected from draculinaction and that fast inhibition is only i o'-v- r -vt(D I achievedwhen free factor IXa interactswith draculin.This sussests I ! 2345618 f4IGRATI0N(cm)

Flg.5 Isoelectricfocusing of purifieddraculin. Fifteen pr"g of proteinwas Table I Amino acid compositionof non-reduced(native) and reduced, applied.Insert shows the actual agarose gel after staining with CoomassieBlue alkylated(s-pyridylethyl, SPE) draculin (0 indicatesoriginal position of the sample)Isolelectric point markersused were:1, Trypsinogen (9 3);2, Lentil Lectin (8.15); 3, Horse myoglobin (6.85); Aminoacid SPE Native 4, A (5.2);5, Soybean trypsin inhibitor (a.55); 6, Amylogluco- 't8.9 BJactoglobulin ASX 78.I sidase(3.5) 't9 GLX 75.0 6 SER 51.1 53.2 GLY 70.5 69.4 HIS 8.3 9.0 thatthe inhibition observed is basicallydue to theeffect of draculinon ARG 46.2 45.4 protection THR 310 34.4 free factorIXa. Comparable of the amidolyticactivity of '75.9 ALA 7r8 factorXa on 5-2337is observedwhen the componentsof the pro- PRO 36.8 34.6 thrombinasesystem (factor Va andphospholipids) are added (results TYR 25.6 22.5 notshown). VAL 45.4 43.1 The dose-responsefor the interactionbetween sub-equimolar MET 5.9 6.8 amountsof draculinand factor IXa or Xa is depictedin Fig. 8. As can ILE 17.9 16.5 be seen,the inhibitory effect is dose-dependentand shows a tendency LEU 722 69.7 towardscomplete inhibition when about two moles of draculin(MW = PHE 32.7 33'l 88.5kDa) are added to onemole of eitherfactor. A similarconclusion LYS 4'7.l 464 canbe drawnfrom theresults shown in theinsert, which describe the Thefigures indicate residues per mole effectof draculinon theactivity of factorIXa on thechromogenic sub-

91 of thrombin-antithrombininteraction being 1.49min-t/pM,which C0NTR0L ^ooooo accountsfor a halflife timeof thrombinof over25 s in a 1 Lr,Msolution of antithrombin(16), the interactionof antithrombinwith other ^oo- ^oo' activatedclotting factors is evenconsiderably slower]. Furthermore the ADDITI0N completelyunexpected observation that factorXa- and factorIXa E -ooo- c I o" *oo AAAA AAA AA AA A A AA 6A A AA A inhibitionare not mutuallyexclusive is, to out knowledge,not found ;AA AAA ^ DRACULIN with anyof theknown serpins. d Thislatter fact is for themoment the most difficult to explain.One ^A hasto reconcilethe followingfour observations:a) apparenthomo- geneityof our preparationby the usualcriteria, inclusive a single aminoterminalsequence, b) a 2:1 bindingstoichiometry of factor Xa o or IXa inhibitionthat is variablebetween 1.8 and2.7. c) lackof com-

0L 0 10 20 30 40 50 60 TII4E(sec) Flg.6 Inhibitionof factorXa amidolytic activity by draculin. Progress curve forthe hydrolysis of thechromogenic substrate S-2337 by purified coagulation factorXa (25 nM), expressed asthe change in absorbencyat405 nm After22 seconds,buffer (Control) or draculin (50 nM) was added and the reaction was =

allowedto proceed for another 30 s t o E o (t strateThiobenzyl Benzyloxycarbonyl-L-lysinate (Z-Lys-SB21 . HCI). x

Sincethis is not a goodsubstrate for factorIXa, relativelyhigh con- o t! centrationsof enzymemust be usedin orderto havea reasonablerate IL to hydrolysis.Due to thishandicap, only two molecular ratios 0f dra- culin:factorIXa were examined,however, their inhibitory effect clearlyfalls within the expocted range assuming a molar ratio of 2:1 anda mo1ecu1armass for draculinof 88.5kDa (dashedline). This 0306090 Time (sec) inhibitionof theactivity of FIXaon an artificial substrate confirms that the inhibitoryactivity 0f draculinobserved in the tenasesystem is Fig 7 Effeclof draculinon factorX activationby factorIXa. Thecomplete indeeddue to a directeffect of draculinon factorIXa. tenasesystem was used, as described in MethodsThis figurealso shows the effectof varyingthe orderof additionof draculin(DRAC) to the reaction Resultsfrom dose-responseexperiments done with six batchesof system purifieddraculin, show a variabilityin the stoichiometrythat ranges I Controlwithout draculin; + sequenceof addition:factor IXa, phospholipids, betweenL8 and2.l This variabilitycould be relatedto a variable factor VIIIa, draculin;! sequenceof addition:factor IXa, phospholipids, degreeof microheterogeneity,e.g. causedby differentdegrees of gly- draculin,factor VIIIa A Sequenceof addition:factor IXa, draculin,phospho- cosylation,ofnative draculin At presentwe do not have experimental lipids,factor VIIIa. The substrateof the reaction,i e. factorX, was always evidenceon the possibledependence of the biologicalactivity of addedlast to startthe reaction draculinand the level of glycosylation. Table2 showsthat the effect of draculinis specificfor theactivated formof factorX. As canbe seen, the inhibitory effect of draculinon the activityof FXa is not affectedby thepresence of FX or a chemically modifiedFXa (Dansyl-Glu-Gly-Arg-chloromethylketone-FXa) which E is catalyticallyinactive, Table 2 alsoshows a suryrisinglack of com- \ petitionbetween FIXa and FXa in theirinteraction with draculin.Incu- = E bationof draculinwith anexcess of FIXadoes not affect the inhibition o= of FXaby draculin.Table 3 indicatesthat interaction of draculinwith <6 JO 15 2t 25 eitherfactor IXa or Xa which havebeen immobilized to a matrix M MllO (ldlniFtu) (AffiGel-15),aiso immobilizes the inhibitory activity of theprotein ;x againstthe otherfactor. We attributethe weak residualinhibition d observedafter incubation with FIXa-AffiGelto a lessthan optimal couplingof thisfactor to thematrix. At themoment we cannotpropose a detailedmechanism of action v 05 1,0 1.5 2,0 25 of draculin0n factorXa and/orfactor IXa. From its sizeone might I'10LARRATIO (DracuJ.in:FXa or IXa) thinkdraculin to be relatedto oneof themany known serpins but the Fig 8 Stoichiometryof draculin-clottingfactor interaction. Stoichiometry N{erminalamino acid sequencedoes not showhomology with this of theinhibition of purifiedfactor IXa (circles)and purified factor Xa (squares) groupof proteins.Also the inhibition is muchfaster than that obtained by draculin(standard assay) Insert shows results obtained for theeffect of dra- with mostserpins. Compared to the typicalanticoagulant serpin anti- culinon purified factor IXa usingZ-Lys-SBzl . HCIas substrate (filled circles) thrombinwe seethat the specificity is higher(no inhibition of throm- Thedashed line corespondsto the theoreticalinhibition that will be obtained bin)and the action is muchfaster [the pseudo first orderdecay constant assuminga 2: I stoichiometry

98 100 shownin Fig.9. As canbe seen,thrombin generation is bothretarded and inhibitedby draculin.The increaseof the lag phaseis brought c0NTR0L ,o o aboutby factorXa inhibition,while the decrease of thepeak value is BO o o dueto factorIX inhibition(17). To summarize,in thispaper we report the isolation from vampire bat c o OU salivaof a newanticoagulant glycoprotein with a molecularmass of z m 885 kDaand isoelectric point of 4.14.2.I1appears as a singlechain o4e' polypeptide.However, from the functionaland bindingstudies we a I haveto acceptthe possibility of a microheterogeneity.This polypep- F tide, which we nameddraculin, seems [o be highly specificfor the 20 activatedforms of coagulationfactors IX and X and its inhibitory f effectis immediate The stoichiometry (2: l) of inhibitionand the inter- actionwith immobilised activated factor Xa andIXa as well as the lack 051015 of competitionbetween FIXa and FXa for draculin,suggests the TIME (min) existenceof two specificinhibrtory sites, possibly present in two Fig 9 Effect of draculinon intrinsicthrombin generatron in human,defi- hardlydifferent polypeptide chains. brinatedplasma Experimental protocol and analysis of thedata, as described in Draculinis differentfrom other anti-Xa molecules that have been re- Methodsand Ref. 13. Open cilcles, control; closed circles, 1 8 nM draculin centlyisolated from other animal sources, such as Antistasin (18), TAP (19),the anti-FXaisolated from blackfly salivaryglands (20) and anticoagulantsfrom leeches(21) All of themare polypeptides with molecularweights below 20 kDaand their effect rs characterizedby a Table2 Effectof theaddition of FX, modifiedFXa and FIXa on theanti-Xa slowonset that requires from l0 to l5 minof preincubationfor maxi- activityof draculin mal inhibition.AntilX activityhas been observed in crudeextracts Addition* FXaactivity (mA/min) from the tick Ornithodorusmoubata, however, it is not clearif the FXa 610 effectof theanticoagulant is on FIX or FIXa (22).Draculin is also FXa+ draculin 330 differentfrom Ecotin the serine protease inhibitor isolated homE. coli FXa+FX+draculin 327 (23). FXa+mFXa+draculin 339 To ourknowledge, draculin is thefirst natural polypeptide for which FXa+FIXa+draculin 332 immediateanti-Ixa and anti-Xa activities have been described. The *FXa, mFXa and FIXa were20 nM Draculinconcentration was chosen to discoveryof thisnew anticoagulant provides a newtool for studiesof produceabout 5070 inhibition (= 20 nM basedon a MW = 805 kDa).mFXa is thecoagulation mechanism(s), and the possibility of designof sub- factorXa of whichthe active site is blocked with dansyl-Glu-Gly-Arg-Chloro- stratesfor theassay of factorIXa. Further,it opensa newapproach to methylketone thestrategy in thesearch for newanticoagulants with therapeuticpo- tential.The observation that vampires are immune to theanticoagulant actionof theirown saliva suggests that there are interesting differences betweenthe clotting system of thesemammals and others, that remain TabLe3 Effectof immobilizedfactor IXa andXa on theactivity of draculin tobe investigated. Anti-Xa Anti-lXa Draculin+ blockedAffiGel 100 r00 AcknowLedgements Draculin+ IX-AffiGel 20 25 Supportedin part by ContractsCI1*/0575 and CT920062 (lnternational Draculin+ Xa-AffiGel 4 2 ScientificCooperation European Community-Andean Pact Countries) The An aliquotof a draculinsolution was incubated with the AffiGel preparation authorsare also deeply indebted to themembers of the Centro de Investigacion Thenthe AffiGel was removed by centrifugation.The inhibitory capacity of dela EcologiadelaZona Arida (CIEZA) of theUniv. Francisco de Miranda thesupernatant is expressed asa percentageof that of theoriginal draculin (Venezuela)lbr their invaluable collaboration in the capture of vampirebats, solution andto Dr. R. Wagenvoord(Dept. of Biochemisrry,U of Limburg)for the supplyof purifiedcoagulation factors

petitionbetween the two substratesfactor Xa andfactor IXa andd) bindingof bothactivities to immobilisedfactor Xa or factorIXa. The minimalhypothesis seems to bethat two very similar protein-chains are References presentin ourpreparation, one (drac-Xa) that inhibits factor Xa andthe otherthat inhibits factor IXa (drac-IXa).To explainthe retention on gel 1.Bier OG Actionanticoagulante et fibrinolytiquede l'extraitdes glandes we needthe extra assumption of eitheran interaction between the two salrvairesd'une chauve-souris h6matophage (De smodus ruy'rs). Comptes RendusSoc Biol (Paris)1932;1.10:129-31. formsof draculin(heterodimer) or a noninhibitory interaction between 2 RomanaC Action anticoagulantede la salivedu vampire(Desmodus factorXa and drac-IXaand vice versa.More complexinteractions, rotundus).Soc Pathol Exot 1939; 32:399403. suchas binding of drac-IXato thefactor Xa-drac-Xa complex remain 3,Gardell SJ, Duong LT, DiehlRE, York JD, Hare TR, Register RB, Jacobs p0ssible. problem This is atpresent under investigation. JW, Dixon RAF, FriedmanPA Isolation,charactenzation, and cDNA Theeffect of draculinon the generation of thrombinmediated by the cloningof a vampirebat salivary plasminogen activator J BiolChem 1989; componentsof the tenasecomplex (intrinsic thrombin generation) is 264:1'1941-52.

99 4. HawkeyC. Plasminogenactivator in salivaof the vampirebat Desmodus 15.Hemker HC, Willems GM, B6guinS. A computerassisted method to obtain rotundus.Nature 1966; 2ll:434-5. theprothrombin activationvelocity in wholeplasma independent ofthnom- 5. CartwrightT.Theplasminogen activator ofvampire bat saliva. Blood 1974; bin decayprocesses. Thomb Haemost1986; 56:9-17. 43:318-26. 16.B6guin S, KesselsH, Dol F, HemkerHC. The consumption of antithrom- 6. DicksonJM, GreenDG. The vampire bat (Desmoilus rotundus)'.Improved bin III during coagulation,its consequencesfor the calculationof pro- methodsof laboratorycare. Laboratory Animals 1970; 4: 3T44. tlrombinaseactivity and the standardisationof heparinactivity. Thromb 7. JamesonGW, RobertsDV, AdamsRW, Kyle WSA, ElmoreDT, Thioben- Haemost1992; 68: 13642. zyl benzylocarbonyl-L-lisinate.Substrate for a sensitiveassay for trypsin- 17.B6guinS, Dol F,HemkerHC. FactorXa inhibitionconhibutes to the lite enzymes.Biochem J 1973;131l0T-lT. heparineffect. Thromb Haemost l99I; 66 306-9. 18.Tuszynski GP, Gasic TB, GasicGJ. Isolation and characterization of Anti- 8. VanDiejen G, Tans G, Rosing J, HemkerHC. The role of phospholipidand stasin.I Biol Chem1987 262: 97 18-23. factorVIIIa in theactivation 6f bovinefactor X. J Biol Chem1981: 256: ; 19.Waxman L, SmithDE, ArrcuriKE, VlasukGP. Tick anticoagulantpeptide 343342. (TAP)is a novelinhibitor of bloodcoagulation factor Xa. Science1990; 9. WagenvoordR, HendrixH, TransT, HemkerHC. Developmentof a rapid 248:593-6. and sensitivefactorX assayfor clinical use. Haemostasis1990;20: 20. JacobsJW, CruppEW, SardanaM, FriedmanPA. Isolationand character- 2',t6-88. izationof a coagulationfactor Xa inhibitor from black fly salivaryglands. 10.Green DJ, ShawE. ThiobenzylBenzyloxycarbonyl-L-lysinate, a substate ThrombHaemostas 1990; @:235-8. for a sensitivecolorimetric assay for TrypsinJikeenzymes. Anal Biochem 2l.BrankampRG, BlankenshipDT, SunkaraPS, CardinAD. Ghilantens. 19'19;93:2234. Anticoagulant-antimetastaticproteins from the South American leech, 11. BradfordMM. A rapid andsensitive method for the quantitationof micro- Haementeriaghilianii.JLab Clin Med 1990;115: 89-97. gram quantitiesof protein utilizing the principle of protein-dyebinding. 22.HellmanK, HawkinsRL The actionof tick extractson blood coagulation Anal BiochemI9'7 6; 72: 248-54. andfibrinolysis. Thrombosis Diath Haemonh1967 ; 18: 617-25. 12.Shagger H,,von Jagow G. Tricine-SodiumDedecyl Sulfate-Polyacrylanide 23.Seymur JL, LindquistRN, DennisMS, Moffat B, YansuraD, Reilly D, gel electrophoresisfor the separationof proteinsin the rangefrom 1 to WessingerME, LazarusRA. Ecotinis a potentanticoagulant and reversible 100kDa, Anal Biochem 1987 : 166:368J9. tightbindinginhibitor of factorXa Biochemistry1994;33:3949-58. 13.Fairbanks G, StackTL, WallackDFH. Electrophoreticanalysis of the ma- jor polypeptidesof ttrehuman ery;throcyte membrane. Biochemisfy 1971; l0:261742. [4. VesterbergO, Gramstrump-ChristensenB. Sensitive silver stainingof proteinsafter isoelectric focusing in agarosegels. Electrophoresis 1984; 5: 282-5. ReceivedJune 29, 1994 Acceptedafter resubmission September 14,1994

100