146

Hedgehog Lipoprotein(a) Is a Modulator ofActivation of Plasminogen at the

Fibrin SurfaceAn In Vitro Study

Didier Rouy, P. Michel Laplaud, Michel Saboureau, and Eduardo Angl6s-Cano

Lipoprotein(a) (Lp[a]), a highly atherogenic lipoprotein particle, is the prominent apolipo-proteln B-containing lipoprotein in the hedgehog (Lapland PM et al,J LipidRes 1988^29:1157-1170). In the present work, we studied the consequences of the structural homology between thespecific Lp(a) glycoprotein, apoprotein(a), and plasminogen on the generation of plasmin byfibrin-bound tissue-type plasminogen activator. The activation of plasminogen was initiated byadding either native plasma or Lp(a)-free plasma supplemented with the equivalent of 0.25mg/ml of either purified Lp(a) or albumin to a surface of fibrin prepared on microtitrationplates and to which human tissue-type plasminogen activator was specifically bound. With theLp(a)-free plasma, an increase in the binding and activation of plasminogen as a function oftime was observed. In contrast, in the presence of Lp(a) (i.e., native plasma or the reconstitutedsystem), a significant decrease in the binding of plasmin(ogen) (—60%) was obtained. Thesedata indicate that hedgehog Lp(a) interferes with the binding and activation of plasminogen atthe fibrin surface and may thereby behave as a factor regulating the extent of fibrin deposition.These results support our previous data indicating that high levels of Lp(a) may haveantifibrinolytic effects in humans (Rouy D et al, Arterioscler Thromb 1991;ll:629-638), are inagreement with the observation that Lp(a) is a risk factor for atherosclerotic disease, andprovide further support to the view of Lp(a) as a link between atherosclerosis and thrombosis.(Arteriosclerosis and Thrombosis 1992;12:146-154)

Lipoprotein(a) (Lp[a]) is a genetically deter-mined lipoprotein particle that contains a

f specific apolipoprotein, apoprotein(a)(apo[a]), that is highly homologous to plasminogen,the precursor of the fibrinolytic enzyme plasmin.1-3

However, apo(a), unlike plasminogen, is not a sub-strate for plasminogen activators, and high levels ofthe Lp(a) particle in human plasma are consideredan independent risk factor for atherosclerotic cardio-vascular disease.4-5 A number of studies have there-fore been recently devoted to search for a possiblerole for Lp(a) in both atherosclerosis and thrombo-

From the Institut National de la Santd et de la RechercheM6dicale (INSERM) U. 143, Hdpital de BicStre (D.R., EA-C.),and U. 321, Hdpital de la Pitid (P.M.L.), Paris; the Laboratoire deBiochimie M£dicale (P.M.L.), Faculty de Me'decine, Limoges; andthe Centre d'Etudes Biologiques des Animaux Sauvages (CNRS-UPR 4701) (M.S.), Beauvoir-sur-Niort, France.

Supported by the Institut National de la Sant6 et de la Recher-che Metlicale.

Address for correspondence: Dr. Eduardo Angles-Cano,INSERM U. 143, I.P.C., Hdpital de Bicfitre, F-94275-Cedex,BicStre, France.

Received July 29, 1991; accepted October 1, 1991.

sis. Recent reports have indicated that Lp(a) maybind to and displace plasminogen from the surface ofendothelial cells and monocytes and from immobi-lized fibrinogen and fibrin and, in addition, mayinhibit plasminogen activation in purified systems6-7

and in plasma.8 In subjects with high levels of Lp(a),decreased generation of plasmin at the plasma-fibrininterface was found to be associated with the bindingof Lp(a) to lysine residues of fibrin.8-9

Lp(a) is the prominent apolipoprotein B-con-taining lipoprotein in a typical hibernator, theEuropean hedgehog (Erinaceus europaeus) .10 Asdescribed recently, apo(a) in the hedgehog sharessome physicochemical and immunologic propertieswith its human counterpart.10 The hedgehog maytherefore constitute a useful natural animal modelfor the in vivo investigation of the interactionsbetween this highly atherogenic particle and thefibrinolytic system. In the present study, we havecharacterized the fibrinolytic system of the hedge-hog and have investigated the effect of plasmaLp(a) on the binding and activation of plasminogenat the fibrin surface by tissue-type plasminogen

by guest on April 13, 2017

http://atvb.ahajournals.org/D

ownloaded from

Rouy et al Lp(a) and Fibrinolysis in Hedgehogs 147

activator (t-PA). We show that during ongoingplasminogen activation, both plasminogen andapo(a) (in the form of Lp[a]) bind to the fibrinsurface and that the generation of plasmin dependson the concentration of these proteins. These re-sults clearly indicate that the apo(a)-like glycopro-tein of the hedgehog interacts with fibrin and mayconstitute a regulatory component of the fibrin-olytic system of this hibernator.

MethodsChemicals and Reagents

All the chemicals used were of the best reagentgrade that were commercially available. Other prod-ucts were purchased from the following sources:EDTA, bovine serum albumin, Tween 20, and poly-ethylene glycol (PEG) from Serva, Heidelberg, FRG;Ultrogel AcA 44 and DEAE-Trisacryl from IBF,Villeneuve-La-Garenne, France; chromogenic sub-strate methylmalonylhydroxyprolylarginine-p-nitro-anilide (CBS 1065) from Diagnostica Stago, Asnieres,France; lysine- and gelatin-Sepharose 4B and PD-10Sephadex G-25 M columns from Pharmacia FineChemicals, Uppsala, Sweden; polyvinyl chlorideU-shaped microtitration plates and plate sealers fromDynatech, Marnes-La-Coquette, France; glutaralde-hyde, 25% (vol/vol) aqueous solution from TAABLaboratories, Reading, UK; [^IJNal from AmershamInternational, Amersham, Buckinghamshire, UK; andnitrocellulose from OSI, Paris, France. Benzamidineand 6-aminohexanoic acid were purchased from Aid-rich, Strasbourg, France. Dansyl-L-glutamyl-grycyl-L-arginine chloromethyl ketone (GGACK) was ob-tained from France Biochem, Meudon, France; nonfatdry milk from Gloria, Courbevoie, France; and M-Par-tigen plates and D-phenylalanyl-L-prolyl-L-argininechloromethyl ketone (PPACK) from Behringwerke,Marburg, FRG. Human plasmin and 4-amidinophen-ylmethanesulfonyl fluoride (PMSF) were obtainedfrom Boehringer, Mannheim, FRG. Paranitrophenyl-/j'-guanidinobenzoate was from Sigma, La Verpilliere,France. Protein A was from Cappel (Flobio, Courbe-voie, France). Human t-PA (>95% single-chain,680,000 IU/mg) was purchased from Biopool (Umea,Sweden) and was calibrated to the international t-PAstandard (preparation 517 from the National Institutefor Biological Standards and Control, London, UK).

BuffersBuffer A was 0.05 M NaH2PO4 buffer, pH 7.4,

containing 0.08 M NaCl. Buffer B was 0.05 MNaH2PO4 buffer, pH 6.8, containing 0.08 M NaCl.Assay buffer was buffer A containing 0.2% (wt/vol)bovine serum albumin and 0.01% (vol/vol) Tween 20.Binding buffer was buffer B containing 0.4% (wt/vol)bovine serum albumin, 0.01% (vol/vol) Tween 20,and 2 mM EDTA. All other buffers were prepared asdescribed in the text.

Animals and DietsMale adult hedgehogs (Erinaceus europaeus) were

bred in the Centre d'Etudes Biologiques des Ani-maux Sauvages located in western central France.They were kept individually in 6-m2 parks undernatural conditions of light, temperature, and rainfall.The animals were fed daily with a mixture of crushedchicken meat and commercial food for dogs (CaninaDuquesne-Purina, Paris, France), containing the fol-lowing proportions by weight of the major constitu-ents: protein, 20%; animal fat, 6%; carbohydrate,5%; vitamin A, 15,000 IU/kg; and vitamin D, 1,500IU/kg. Water was provided ad libitum.

Blood SpecimensBlood was taken from hedgehogs that had been

fasted overnight for approximately 18 hours. Theanimals were slightly anesthetized with halothane(Fluothane, Coopers Veterinaire, Meaux, France) inoxygen, and blood was withdrawn from the jugularvein and collected on 3 mM (final concentration)EDTA. Plasma was then separated by low-speedcentrifugation and brought on ice to the laboratory.

Preparation of Lipoprotein(a)-Free Hedgehog Plasma(d>1.21 glml Density Fraction)

We have previously demonstrated that the Lp(a)particle in the plasma of hedgehogs is contained in the1.040-1.100 g/ml density range.10 Thus, to obtainmaterial containing no Lp(a), aliquots of hedgehogplasma were adjusted to a density of 1.21 g/ml withsolid potassium bromide. Sodium azide (0.02%, vol/vol) was added, and the resulting solution was ultra-centrifuged in an MSE Prepspin 50 ultracentrifuge(MSE Scientific Instruments, Crawley, UK), equippedwith an 8x14-ml aluminum fixed-angle rotor, for 36hours at 100,000^^ at 17°C. The bottom fraction wasthen recovered by aspiration and dialyzed in Spectra-por tubing (Spectrum Medical Industries, Los Ange-les, Calif.; exclusion limit, 6,000-8,000) for three times12 hours at 4°C against a solution containing 0.15 MNaCl, 1 mM EDTA, and 0.02% (vol/vol) NaN3.

Isolation of Hedgehog Lipoprotein(a)To isolate hedgehog Lp(a), the methodology re-

ported in Reference 10 was used. In brief, aliquotsfrom hedgehog plasma pools were first brought to adensity of 1.055 g/ml by addition of potassium bro-mide, layered with sodium chloride (d=1.055 g/ml),and centrifuged for 24 hours at 100,000g^ at 17°C inthe same machine and rotor as reported above. Thetop 3 ml was discarded, and the bottom fraction wascollected, placed in similar new centrifuge tubes, andmixed with sodium chloride (d=1.190 g/ml) to obtaina final density of 1.100 g/ml. Such samples weresubsequently ultracentrifuged for 26 hours at100,000g^ at 17°C. The top fraction (1.5 ml) wasthen collected and dialyzed for three times 12 hoursat 4°C against 0.15 M NaCl, pH 7.0, containing 1 mMEDTA and 0.02% (vol/vol) NaN3. This fraction con-

by guest on April 13, 2017

http://atvb.ahajournals.org/D

ownloaded from

148 Arteriosclerosis and Thrombosis Vol 12, No 2 February 1992

tained a mixture of Lp(a) and LDUike particles. Afterconcentration, these two types of lipoprotein particleswere separated by means of gel-filtration chromatogra-phy on a column of Sepharose CL-4B (12x 1,000 mm)operated at 4°C. Elution was performed with 0.15 MNaCl, pH 7.0, at a rate of 12 ml/hour, and 2-mlfractions were collected. Elution was monitored at 280nm. This type of chromatography led to incompleteseparation of the two lipoprotein populations; there-fore, the fractions corresponding to the first peak werepooled and rechromatographed in the same system.The resulting peak was approximately gaussian. Frac-tions corresponding to the trailing edge of this peakwere discarded, and the remainder of the peak wasshown to contain pure Lp(a).10

Purification of Apoprotein(a)Apo(a) was purified according to a modification of

the methodology of Armstrong et al.11 In brief, nativeLp(a) was brought to a concentration of ~1 mg/mltotal apoprotein in 20 mM tris(hydroxymethyl)ami-nomethane (Tris) pH 7.4, 0.15 M NaCl, 1 mMEDTA, and 0.02% (vol/vol) NaN3. Dithiothreitol wasfreshly added to a final concentration of 10 mM, andthe solution was incubated at 37°C for 3 hours. Thesolution was then adjusted to 1.100 g/ml with solidpotassium bromide and centrifuged in the MSE6xl4-ml titanium swing-out rotor at 150,000?^ for24 hours at 10°C. After completion of centrifugation,apo(a) was a pellet that had sedimented at the baseof the tube. It was solubilized in 20 mM Tris, pH 8.0,1 mM EDTA, and 10 mM dithiothreitol, containing85 mM sodium dodecyl sulfate (SDS).

Preparation of Hedgehog Plasminogen and Fibrinogen

Plasminogen was purified from fresh hedgehogplasma by the method of Deutsch and Mertz12 withsome modifications. The plasma was supplementedwith protease inhibitors (100 klU/ml aprotinin, 2 mMPMSF, 10 fiM /7-nitrophenyl-p'-guanidinobenzoate,and 1 iM PPACK), and all steps were performed at4°C. The plasma was passed over a column of lysine-Sepharose 4B. The gel was washed with buffer Acontaining 10 klU/ml aprotinin until the absorbanceat 280 nm equaled that of buffer. The h/sine-adsorbedproteins were eluted with a linear gradient in 6-ami-nohexanoic acid (0-50 mM), concentrated in a dial-ysis bag against solid PEG (Mr~20,000), and appliedto a column of Ultrogel AcA 44 equilibrated withbuffer A. Three peaks were typically observed in theprofile. Fractions corresponding to each peak werepooled separately, concentrated, diaryzed againstbuffer A, and analyzed by SDS-poryacrylamide gelelectrophoresis (SDS-PAGE) and fibrin autography.One of the proteins present in the third peak wassensitive to cleavage by plasminogen activators and isdesignated hereafter as hedgehog plasminogen.

Human and hedgehog fibrinogens were purifiedfrom fresh plasma according to Kazal et al13 andwere depleted in plasminogen and fibronectin asdescribed.12-14

Potyclonal Antibody Against Plasminogen

A poryclonal antibody directed against the proteincontent of the third peak obtained after gel-filtrationchromatography was prepared as follows. Antigen (0.5mg) in Freund's complete adjuvant was injected sub-cutaneously into sheep. The same amount, in Freund'sincomplete adjuvant, was similarly injected at 3 and at6 weeks after the initial immunization. After immuno-logic testing of the antiserum, the sheep was bled, andserum was separated. The immunoglobulin G fractionof the antiserum was then separated by ammoniumsulfate precipitation, ion-exchange chromatographyon DEAE-Trisacryl, and affinity chromatography onprotein A-Sepharose.

Preparation of Solid-Phase Fibrin Surfaces

The solid-phase fibrin surfaces were prepared aspreviously described.15-16 In brief, polyvinyl chloride-bound stable polyglutaraldehyde derivatives were firstproduced by treating U-shaped microtitration polyvi-nyl chloride plates with 2.5% (vol/vol) glutaraldehydein 0.1 M NaHCO3 buffer, pH 9.5, for 2 hours at 22°C.Then fibrinogen (0.3 /JM in 0.1 M NaH2PO4 buffer,pH 7.4, containing 1 mM CaG2) was covalently fixedfor 18 hours at 4°C. After washing, a fibrin networkwas generated by thrombin treatment (1 NationalInstitutes of Health unit/ml in assay buffer containing1 mM CaCl2)- The excess thrombin was eluted bythree washes with a high-ionic-strength solution (0.5M NaCl/8 mM CaCyO.05% [vol/vol] Tween 20). Afinal wash with 5 mM NaH2PO4 buffer (pH 6.8 con-taining 0.05% [vol/vol] Tween 20) was performed, and100 ii\ per well of buffer B, containing 0.02 M lysine,0.2% (wt/vol) bovine serum albumin, and 0.01%NaN3, was added. The plate was sealed and stored inthis state at 4°C until further use. Fibrin surfaces wereprepared with both human and hedgehog fibrinogens.The ability of thrombin to cleave human fibrinogenwas monitored by use of a horseradish peroxidase-labeled mouse monoclonal antibody (Y18-HRP)kindly provided by Organon Teknika (Fresnes,France). The immunoreactivity of this antibody withthe Aa stretch 1-51 of human fibrinogen disappearedon treatment with thrombin.17

Assay MethodsLp(a) in hedgehog plasma was determined by an

immunonephelometric method (Behring nephelom-eter analyzer) with a monospecrfic sheep antiserumto human Lp(a) (Immuno AG, Vienna, Austria) anda reference human standard in which the Lp(a) masscontent had been previously determined by elec-troimmunodiffusion (Immuno AG). The validity ofthis technique for measuring plasma Lp(a) concen-trations in the hedgehog was verified as follows. Thelipoprotein concentration of Lp(a) purified from apool of hedgehog plasmas as indicated above wasascertained by chemical analysis. Immunonephelo-metric measurements of Lp(a) in both serial dilutionsof the purified Lp(a) and in plasmas supplemented

by guest on April 13, 2017

http://atvb.ahajournals.org/D

ownloaded from

Rouy et al Lp(a) and Fibrinolysls in Hedgehogs 149

with increasing concentration of the same Lp(a)preparation were then performed. In both series ofexperiments, results showed that the method in usewas satisfactory for hedgehog Lp(a) measurements.More specifically, the range of results obtained for agiven hedgehog Lp(a) concentration was within 89-106% of those obtained for the same concentrationof human Lp(a); in addition, a linear response tosupplementation in Lp(a) was obtained throughoutthe range of concentrations of this lipoprotein en-countered in our animals according to season, that is,0-100 mg/dl.

Plasminogen antigen concentration in hedgehogplasma was determined by radial immunodiffusionaccording to Mancini et al18 by using the sheepantibody described above and an aliquot of thecontent of the third peak obtained during gel-filtra-tion subfractionation of the tysine-absorbable pro-teins of hedgehog plasma as a standard.

The activatable plasminogen was determined inplasma by spectrophotometric assay consisting offibrin-bound t-PA (see below), a volume dilution(final, 1:80) of plasma in binding buffer, and 1.5 mMCBS 1065. Human plasminogen was used as a refer-ence, and the initial rate of activation under first-order conditions was determined as indicated below.

The antiplasmin activity of hedgehog plasma wasdetermined by titration of added human plasmin in aspectrophotometric assay as described previously,19 ex-cept that residual plasmin activity was detected with 1.5mM of the chromogenic substrate CBS 1065 at adouble-wavelength absorbance ratio (405 nm/490 nm).

Activation of Plasma Plasminogen by Fibrin-BoundTissue-Type Plasminogen Activator

The activation of hedgehog plasminogen by humant-PA on a fibrin surface was performed as indicatedelsewhere.20 In brief, a solid-phase fibrin plate waswashed three times with binding buffer, and 50 pd perwell of the same buffer containing 20 IU/ml humant-PA was incubated for 1 hour at 37°C. The plate wasthen extensively washed with binding buffer to elim-inate unbound proteins, and the reaction was startedby adding 50 p\ per weH of the activation mixture.This mixture was either native plasma or Lp(a)-freeplasma supplemented with the equivalent of 0.25mg/ml of either Lp(a) or albumin. The activation wasperformed in a series of three wells and was moni-tored for 4 hours. Every 5-15 minutes, the reactionwas stopped by removing the fluid phase and washingthe wells with assay buffer. When indicated, a traceamount (final concentration, 0.3-0.7 /tg/ml) ofhedgehog mI-plasminogen was added. Products pres-ent in the solid phase were analyzed by spectropho-tometry, radioisotopic counting, SDS-PAGE autora-diography, and specific immunoblotting as follows.Fibrin-bound plasmin activity was detected by adding50 fil 1.5 mM CBS 1065 in assay buffer per well andmeasuring the change in absorbance (AA^/min) ata double-wavelength absorbance ratio (A^^/A^with a microtitration plate counter (MR 610, Dyna-

tech) equipped with a thermostatic device to main-tain temperature at a constant 37°C. Initial velocitiesof each reaction were calculated by using an appro-priate statistical analysis of data in its original coor-dinate system, that is, AA^ versus time, as describedpreviously.15 Because CBS 1065 is a plasmin-selectivechromogenic substrate and fibrin-bound plasmin wasmeasured at the end of the activation in the absenceof plasminogen, it is very unlikely that fibrin-boundt-PA was interfering with the assay. Moreover, exper-iments were performed to detect any possible hydrol-ysis of the chromogenic substrate by Lp(a).

To quantify the amount of fibrin-bound plasmino-gen and to identify the molecular derivatives of plas-minogen bound to fibrin, the experiments were per-formed in the presence of ^I-hedgehog plasminogen,and the reaction was stopped with assay buffersupplemented with 1 mM benzamidine and 10 /tMGGACK. At the end of activation, the radioactivityin the wells was counted in a gamma radiationcounter. In parallel experiments, the molecular spe-cies bound to the fibrin surface were eluted byincubating 50 yX 0.2 M 6-aminohexanoic acid for 18hours at 22°C. The eluates were used to identify theplasminogen-derived products by SDS-PAGE andimmunoblotting.

Electrophoretic MethodsGradient gel electrophoresis. Continuous-gradient

slab-gel electrophoresis was performed in a Pharma-cia electrophoresis apparatus GE 2/4 loaded withgradient gels PAA 2/16 (Pharmacia), according to theconditions of Nichols et al.21 For particle-size cali-bration, we used a series of markers ranging inhydrated diameters from 71 to 170 A (High Molec-ular Weight electrophoresis calibration kit, Pharma-cia) and a solution of latex particles (Dow Chemical,380-A diameter). The Stokes' diameters of the par-ticles were calculated by the Stokes-Einstein equa-tion described by Anderson et al.22

Sodium dodecyl sulfate-pofyaaylamide gel electropho-resis and immunoblotting. SDS-PAGE and immunoblot-ting were performed essentially as described in Refer-ences 23 and 24, with minor modifications. Afterblotting the proteins from the poh/acrylamide gel, thenitrocellulose paper was incubated with 5% (wt/vol)nonfat dry milk in 10 mM Tris buffer, pH 7.4, to blockfree binding sites. Blotted proteins were probed withthe antiserum against hedgehog plasminogen diluted1:10 with assay buffer. After washing, the membranewas soaked in a solution of radiolabeled protein A(—500,000 cpm/ml). The nitrocellulose was washed andautoradiographed on Kodak XS-1 films for 8-12 hoursat -70°C, using a Kodak X-Omatic intensifying screen.

Radioiodination of ProteinsHedgehog plasminogen and protein A were radio-

iodinated with 125I using the lodogen method ofFraker and Speck25 with the following modifications:an iodination time of 4 minutes at 4CC and removal offree 125I by molecular sieving on a PD-10 Sephadex

by guest on April 13, 2017

http://atvb.ahajournals.org/D

ownloaded from

150 Arteriosclerosis and Thrombosis Vol 12, No 2 February 1992

G-25 column. The specific activity obtained was 9-11nCi/ng protein for plasminogen and 40 nCi/ng pro-tein for protein A.

Fibrin Autography

Slab-gel/fibrin-agar autography was performed asfollows. Samples were electrophoresed under nonre-ducing conditions. A portion of the gel was stainedwith 2.5% (wt/vol) Coomassie blue. The remainingpart of the gel was soaked for 1 hour in 2.5% (vol/vol)Triton X-100 to remove SDS and then in distilledwater for 5 minutes. The gel was then overlaid on afibrin-agar indicator gel prepared in a moist chamberas described previously.26 Autographs were allowedto develop at 37°C for 6-48 hours and then photo-graphed with dark-ground illumination.

ResultsHedgehog Lysine-Adsorbable Proteins

Proteins eluted from the lysine-Sepharose 4B col-umn were concentrated and then rechromato-graphed on a column of Ultrogel AcA 44. Typically,three peaks were obtained. In each series of experi-ments, proteins in each fraction were analyzed bySDS-PAGE. In one experiment, proteins in eachpeak were examined equally by gradient-gel electro-phoresis, thus demonstrating that the first peak con-tained a protein of apparent Mr of —620,000. Such avalue was comparable to that already reported by usfor certain forms of hedgehog apo(a).10 This mole-cule exhibited cross-immunoreactivity with the sheepantibody against hedgehog plasminogen, as deter-mined by immunoblotting (see below). Proteins inthe third peak were resolved into three protein bandsby SDS-PAGE. Plasminogen was identified by fibrinautography in the protein band migrating with an M,of —93,000 (Figure 1), a molecular weight similar tothat of human plasminogen. Two other bands migrat-ing in a more anodic position and exhibiting relativemolecular weights of 79,000 and 72,000 did notproduce fibrinolytic activity, either spontaneously orin the presence of activators. Although a trace ofplasminogen was also detected in the last fractions ofthe second peak, only the third peak was consideredto contain plasminogen for the purpose of the exper-iments described here. Although proteins in thesecond chromatographic peak were eluted with anapparent molecular weight corresponding to that of7S gamma globulins, they were resolved by SDS-PAGE into two protein bands identical to those withA/r 79,000 and 72,000 found as contaminants ofplasminogen in the third peak. Fibrinogen purifiedfrom hedgehog plasma migrated as human fibrinogenpurified and electrophoresed under similar condi-tions (not shown).

Activation of Plasminogen by Fibrin-BoundTissue-Type Plasminogen Activator

Activatable plasminogen in hedgehog plasmas(from eight animals) was equivalent to 226 ±97 nM

92 —

66 —

Pg Pn M III B I Pg Pn

FIGURE 1. Identification of plasminogen by electrophoreticfibrin autography. Hedgehog plasma was chromatographed onlysine-Sepharose. Affinity-adsorbed proteins were eluted with6-aminohexanoic acid, fractionated by gel filtration, andfurther separated by sodium dodecyl sulfate (SDS) -potyacryl-amide (10%) gel electrophoresis under nonreducing condi-tions. Panel A: Section of gel stained with Coomassie brilliantblue. Panel B: Fibrin autography. SDS was exchanged withTriton X-100, and the gel was then overlaid on a urokinaselfibrin/agar indicator film, incubated for 18 hours at 37°C, andphotographed under dark-ground illumination. I, II, and IIIrefer to peaks observed in a typical chromatographic profile. Pgand Pn represent the electrophoretic mobility and activity ofpurified human plasminogen (Mr 93,000) and plasmin (Mr

85,000), respectively. The Mr values of protein standards(phosphorylase b, 92,000; bovine serum albumin, 66,000) areshown at left.

human plasminogen. The corresponding value inhumans is 721 ±98 nM (10 donors). The difference ismost probably related to the higher level of antiplas-min activity observed in hedgehog plasma (145-200% compared with 80-125% for normal humanplasmas) and to the presence of a substantial amountof nonactivatable lysine-binding proteins. The higherplasminogen antigen value (350-500 mg/1) found inthese animal plasmas relative to the value in humans(150-200 mg/1) is in agreement with the latter ob-servation and with the existence of a cross-reactionbetween the fysine-adsorbable proteins and the anti-bodies raised against plasminogen present in thethird chromatographic peak.

Plasminogen activation experiments were per-formed using surfaces of fibrin prepared from mono-layers of human or hedgehog fibrinogen covalentlybound through polyglutaraldehyde to a solid support.The validity for use of such a surface of fibrin forplasminogen activation experiments at the plasma-fibrin interface has been previously demonstrated.20

In these experiments, human t-PA was first boundto the fibrin surface, unbound proteins were elimi-nated by washing, and hedgehog plasma was addedto initiate the activation. The extent of plasmingeneration at the surface of fibrin was measured witha plasmin-selective chromogenic substrate. The pos-sible hydrolysis of the chromogenic substrate byLp(a) was investigated before these experimentswere performed; even at high Lp(a) concentrationsand after 3-4 hours of incubation at 37°C, we wereunable to obtain an increase in the absorbance at 405nm relative to that of the substrate alone. The results

by guest on April 13, 2017

http://atvb.ahajournals.org/D

ownloaded from

Rouy et al Lp(a) and Fibrinolysis in Hedgehogs 151

10 . v (A A 406 /mln)

2-4 r-

1.8

1.2

0.6

% of maximum binding

100 r-

B

50 100 150 200Activation time (mln)

250 50 100 150 200Activation time (mln)

250

FIGURE 2. Line plots of activation of hedgehog plasma plasminogcn by fibrin-bound tissue-type plasminogen activator (t-PA). Aconstant amount of t-PA (20 IU/ml) was first bound to the fibrin surface. Plasma (n=4), with or without a trace amount ofhedgehog 125I-plasminogen, was then added, and the extent of plasminogen activation was studied as a function of time. Reactionwas stopped at regular intervals by removing the activation mixture. Panel A: Activation was performed in plasma, the fibrin surf acewas washed, and fibrin-bound plasmin was detected spectrophotometricalfy by adding 1.5 mM methylmalonylhydroxyprolylargi-nine-p-nitroanilide. Absence of hydrolysis of the chromogenic substrate by lipoprotein(a) was verified before these experiments wereperformed. Amount of plasmin generated as a function of time is represented by the initial velocities (v) expressed as103 •v(^A^05/min), where ^A405lmin is the change in absorbance per minute at 405 nm. Panel B: Activation in plasmasupplemented with a trace amount of125I-plasminogen. At the end of activation, the wells were washed, dried, and cut out, and theirrespective radioactivities were counted in a gamma radiation counter. Plasmin(ogen) derivatives bound as a function of time arerepresented as percentages relative to the maximum bound at 4 hours.

of plasminogen activation (Figure 2A) show thatplasmin activity was progressively detected at thesurface of fibrin as a function of the activation time.The mass of protein bound to fibrin as a function oftime was detected by counting the radioactivity in thewells when hedgehog 125I-plasminogen was added toplasma as a tracer and is represented in Figure 2B asa percentage with reference to the maximum boundat 4 hours. It is noteworthy that identical results wereobtained with human and hedgehog fibrin surfaces.Proteins present on the surface of fibrin were elutedwith 6-aminohexanoic acid and analyzed by SDS-PAGE and immunoblotting. Results after 2 and 4hours of activation are represented in Figure 3 (lanesA-C). As shown, one protein band of high molecularweight and at least three other protein bands of lowmolecular weight were detected with a sheep immu-noglobulin G directed against hedgehog plasminogenand radiolabeled protein A for tagging. In the low-molecular-weight zone, fibrin-bound proteins wererepresented mostly by a band of M, 56,000, corre-sponding to that of plasmin heavy chain, and twobands with a more cathodic migration correspondingto an M, of 65,000 and 72,000, respectively. The latterband was similar to that of the third protein bandpresent in peak III of the AcA 44 chromatogram(Figure 3, lane D), whereas the band with an Mt of65,000 had a migration identical to that of a proteinband present in the second chromatographic peak(not shown). A faint band with a relative molecularmass (M, 79,000) similar to that of the second proteinband of peak III was also present. In no case wasplasminogen detected at the surface of fibrin, indi-cating that all fibrin-bound zymogen molecules weretransformed into plasmin. At 4 hours of activation, a

band of high molecular weight (equally observed as atrace at 2 hours, lane C), probably related to thematerial present in the first chromatographic peak, wasclearly detected Indeed, according to the particularexperiment considered, this latter material migrated inSDS-PAGE gels at a position indicative of an Mr closeor identical to that of purified hedgehog apo(a).

Role of Lipoprotein(a) in the Activationof Plasminogen

To determine whether Lp(a) bound to the fibrinsurface was interfering with the binding of plasmino-gen, comparative activation experiments were per-formed with native hedgehog plasma, Lp(a)-freeplasma, and Lp(a)-free plasma supplemented withthe equivalent of 0.25 g/1 protein mass of eitherLp(a) or albumin. Typical activation profiles ob-tained under these conditions are shown in Figure 4.The amount of plasminogen derivatives bound tofibrin as a function of time was markedly decreasedwhen the activation was performed with nativeplasma (80% decrease at 4 hours; Figure 4A) andwith Lp(a)-free plasma supplemented with 0.25 g/1Lp(a) (60% decrease at 4 hours; Figure 4B) com-pared with the activation obtained in the absence ofLp(a). It is of note that the respective antiplasminactivities of Lp(a)-free and native plasmas weresimilar, thus indicating that the increased generationof plasmin was actually related to the content in Lp(a).

DiscussionIn the human species, the generation of plasmin at

the plasma-fibrin interface is finely regulated at thelevel of both t-PA and plasminogen.27 Plasminogeninteractions with two plasma proteins, histidine-rich

by guest on April 13, 2017

http://atvb.ahajournals.org/D

ownloaded from

152 Arteriosclerosis and Thrombosis Vol 12, No 2 February 1992

- Apo(a)

8 2 .

66- m - Pn

TIME (h)

FIGURE 3. Identification by immunoblotting of activationproducts eluted from the fibrin surface. Activation of hedgehognative-plasma plasminogen was performed as indicated in Figure2 After the wells were washed, bound proteins were eluted fromthe fibrin surface with 0.2 M 6-aminohexanoic add and electro-phoresed under reducing conditions on a 5 -15%potyacrylamidegel Proteins were identified by immunoblotting with antibodiesdirected against peak III from gel filtration of fysine-adsorbableproteins and autoradiography with 125I-protein A. Lanes A, B,and C represent samples obtained after 2 and 4 hours ofactivation with three different hedgehog plasmas. Lane D showsCoomassie blue staining of proteins in peak III electrophoresedexactly under the same conditions as in lanes A-C. Mr values ofprotein standards (phosphorylase b, 92,000; bovine serum albu-min, 66,000) are shown at left. Apo(a), Pg, and Pn indicatemobility of hedgehog apoprotein(a), plasminogen, andplasmin,respectively.

glycoprotein and a2-antiplasmin, and with fibrin aremediated through lysine-binding sites present in thekringle domains of plasminogen.20-28-29 As a conse-quence, the amount of plasminogen available forfibrin and its transformation into plasmin are low-ered by 50% in plasma (activatable plasminogen was721±98 nM in 10 normal volunteers). Exposure ofnew plasminogen-binding sites on plasmin-degradedfibrin is considered to be a potent effector functionthat increases the rate of plasmin formation byt-PA.30-31 A mechanism that would modulate theextent of fibrinorysis by decreasing the number ofsuch new binding sites has not yet been described. Inparticular, under physiological conditions, no mole-cule or circulating protein has been shown to behaveas plasminogen vis-a-vis its fibrin-binding properties.The demonstration that apo(a) is composed mainlyof structural domains homologous to kringles 4 and 5of human plasminogen suggests that this protein maybe a likely ligand for plasminogen-binding sites infibrin. Indeed, we have recently shown that in humansubjects with high levels of Lp(a), a decreased gen-eration of plasmin at the plasma-fibrin interface wasassociated with the binding of Lp(a) to lysine resi-dues of fibrin.8'9 In the present work, we have furtherinvestigated the possible role of apo(a) as a modula-

tor of plasminogen activation. For this purpose, wehave studied an animal model, the hedgehog, whichhas been shown to possess Lp(a) as its prominentapolipoprotein B-containing plasma lipoprotein.10

Except for the human source of t-PA, these studieswere performed with hedgehog plasma or its derivedproducts, including Lp(a), apo(a), fibrinogen, andplasminogen. From the results reported in the pres-ent study, two main aspects should be stressed. First,a group of lysine-adsorbable proteins was separatedfrom plasma by lysine affinity chromatography andsubsequently fractionated by gel filtration. One ofthese proteins had an unusually high Mr (—620,000),resembling that of purified hedgehog apo(a) andreacting with antibodies against hedgehog plasmino-gen and against human Lp(a) on immunoblotting.Another protein had the same molecular mass ashuman plasminogen (MT 93,000) and was efficientlytransformed into a fibrinolytic enzyme by plasmino-gen activators; it was therefore considered to beplasminogen. Two other protein bands migratingwith a molecular mass (Mr 79,000 and 72,000) lowerthan that of plasminogen had no proteoh/tic activityand were not activated by plasminogen activators.These proteins were the main components of thesecond chromatographic peak of the lysine-adsorb-able proteins, and their presence in the peak ofplasminogen was due to incomplete chromatographicresolution. It is of note that their relative molecularweights determined by SDS-PAGE were lower thanexpected from their chromatographic behavior, sug-gesting the existence in plasma of a molecular formof relatively higher molecular weight. Because theisolation of these proteins was performed underconditions that avoided proteolytic degradation im-mediately after sampling and during the purificationprocedure, we suggest that such forms are probablypresent in circulating plasma. Furthermore, electro-phoresis of the purified products under nondenatur-ing conditions also showed the presence of suchmolecular forms. Further biochemical and functionalcharacterization of these lysine-adsorbable proteinsis in progress in our laboratory.

Second, the preparation of a hedgehog fibrin surfacewas feasible and proved to be useful for the binding ofhuman t-PA and the activation of hedgehog plasmino-gen either in a reconstituted system or in native plasma.The generation of plasmin by fibrin-bound t-PA wasmarkedly decreased when the activation was performedwith Lp(a)-free plasma supplemented with purifiedLp(a). In addition, we have positively documented thebinding of the Mr —620,000 apo(a)-Uke protein and, toa lesser extent, of other rysine-absorbable proteins tofibrin during ongoing plasminogen activation. Consid-ered together, these data indicate that in the hedgehogthe activation of plasminogen at the plasma-fibrininterface is modulated by proteins containing kringlestructures, mainly apo(a), that bind to fibrin. Becausethis phenomenon was documented in intact animalsbred under natural conditions, our data strongly sug-

by guest on April 13, 2017

http://atvb.ahajournals.org/D

ownloaded from

Rouy et al Lp(a) and Fibrinolysis in Hedgehogs 153

% of maximum binding100 r-

75

50

25

% of maximum binding

100 ,-

B

50 100 150 200Activation time (mln)

250 25050 100 150 200Activation time (mln)

FIGURE 4. Line plots of effect of hedgehog lipoprotein(a) (Lp[a]) on activation of plasminogen. Activation experiments wereperformed as indicated in Figure 2 with activation mixtures containing a trace amount of hedgehog123I-plasminogen. Fibrin-boundactivation products were detected by counting the radioactivity in the wells with a gamma radiation counter. Amount of plasminogenderivatives bound to fibrin as a function of activation time is expressed as a percentage relative to the maximum bound at 4 hours.Panel A: Activation with hedgehog native plasma containing 0.3 g/l Lp(a) (o) and with the d>1.21 g/ml fraction ( • ) of the sameplasma, that is, Lp(a)-free plasma. Final plasminogen concentration was similar in the two samples. Panel B: Activation withLp(a)-free plasma supplemented either with the equivalent concentration of purified hedgehog Lp(a) (o) similar to that of originalplasma (0.25 g/l) or with a solution of albumin (•). Concentration of plasminogen was kept constant by adjusting the final volumesof the plasmas.

gest that, at least in the hedgehog, Lp(a) may constitutean important regulatory factor of fibrinolysis.

The European hedgehog exhibits naturally occurringseasonal variations in the plasma concentration ofLp(a).32 These variations are reproducible from oneyear to another and are, at least in part, under endo-crine control. In the human species in contrast, theplasma level of Lp(a) appears to be genetically deter-mined33 and is relatively constant, being insensitive tomost dietary changes and hypocholesterolemic drugs.34

The modulation of plasminogen activation by Lp(a) inthe hedgehog, as demonstrated in the present report, isobviously a seasonal physiological phenomenon. There-fore, such seasonal variations may help to provide aclue to the observation that high levels of Lp(a) inhumans are a risk factor for atherosclerotic cardiovas-cular disease and to support the view of Lp(a) as a linkbetween atherosclerosis and thrombosis.

AcknowledgmentsM J. Chapman is gratefully acknowledged for crit-

ical reading of the manuscript and for most valuablesuggestions. We are indebted to D. Durand for thepreparation of the sheep poh/clonal antibody tohedgehog plasminogen.

References1. Mac Lean JW, Tomlinson JE, Kuang WJ, Eaton DL, Oien

EY, Fless GM, Scanu AM, Lawn RM; DNA sequence ofhuman apolipoprotein(a) is homologous to plasminogen.Nature 1987;330:132-137

2. Eaton DL, Fless GM, Kohr WJ, Mac Lean JW, Xu QT, MillerCG, Lawn RM, Scanu AM: Partial amino acid sequence ofapolipoprotein(a) shows that it is homologous to plasminogen.Pwc NatlAcad Set U S A 1987;84:3224-3228

3. Scanu AM: Lipoprotein(a): A genetically determined lipopro-tein containing a gh/coprotein of the plasminogen family.Semin Thromb Haanost 1988;14:266-270

4. Sandkamp M, Funke H, Schulte H, Kohler E, Assmann G:Lipoprotein(a) is an independent risk factor for myocardialinfarction at a young age. Clin Chem 1990^6:20-23

5. Dahlen G, Guyton JR, Attar M, Fanner JA, Gautz JA, GottoAM: Association of levels of lipoprotein Lp{a), plasma lipidsand other lipoproteins with coronary artery disease docu-mented by angiography. Circulation 1986;74:758-765

6. Harpel PC, Gordon BR, Parker TS: Plasmin catalyzes bindingof lipoprotein(a) to immobilized fibrinogen and fibrin. ProcNatlAcad Sci USA 1989;86:3847-3851

7. Edelberg JM, Gonzales-Gronow M, Pizzo SV: Iipoprotein(a)inhibition of plasminogen activation by tissue-type plasmino-gen activator. Thromb Res 1990^7:155-162

8. Rouy D, Grailhe P, Nigon F, Chapman J, Angles-Cano E:Lipoprotein(a) impairs the generation of plasmin by fibrin-bound t-PA: In vitro studies in a plasma milieu. ArteriosclerThromb 1991;ll:629-638

9. Angles-Cano E, Fleury V, Rouy D, Chapman J: The compet-itive binding of lipoprotein(a) and plasminogen to plasmin-degraded fibrin (abstract). Circulation 1990;82(suppl III):m-92

10. Laplaud M, Beaubatie L, Rail SC, Luc G, Saboureau M:Lipoprotein(a) is the major apoB-containing lipoprotein in theplasma of a hibernator, the hedgehog (Erinaceus europaeus). JLipid Res 1988;29:1157-1170

11. Armstrong VW, Walli AK, Seidel D: Isolation, characteriza-tion and uptake in human fibroblasts of an apo(a)-free lipo-protein obtained on reduction of lipoprotein(a). / Lipid Res1985;26:1314-1323

12. Deutsch DG, Mertz ET: Plasminogen purification fromhuman plasma by affinity chromatography. Science 1970; 170:1095-1096

13. Kazal LA, Amsel S, Miller OP, Tocantins LM: The prepara-tion and some properties of fibrinogen precipitated fromhuman plasma by glycine. Pwc Soc Exp Biol Med 1963;113:989-994

14. Engvall E, Ruoslahti E: Binding of soluble forms of fibroblastsurface protein, fibronectin, to collagen. Int J Cancer 1977^20:1-5

15. Masson C, Angles-Cano E: Kinetic analysis of the interactionbetween plasminogen activator inhibitor-1 and tissue-typeplasminogen activator. Biochem J 1988^256:237-244

16. Angles-Cano E: A spectrophotometric solid-phase-fibrin-tissue plasminogen activator activity assay (SOFIA-tPA) for

by guest on April 13, 2017

http://atvb.ahajournals.org/D

ownloaded from

154 Arteriosclerosis and Thrombosis Vol 12, No 2 February 1992

high-fibrin-affinity tissue plasminogen activators. Anal Bio-chem 1986;153:201-210

17. Koppert PW, Huijsmans CMG, Nieuwenhuizen W: A mono-clonal antibody specific for human flbrinogen, fibrinopeptideA-containing fragments and not reacting with free fibrinopep-tide A. Blood 1985;66:503-507

18. Mancini G, Carbonara AO, Heremans JF: Immunochemicalquantitation of antigens by single radial immunodiffusion.Immunochemistry 1965;2:235-254

19. Teger-Nilsson AC, Friberger P, Gyzander E: Determinationof a new rapid plasmin inhibitor in human blood by means ofa plasmin specific tripeptide substrate. ScandJ Clin Lab Invest1977;37:403-409

20. Rouy D, Angles-Cano E: The mechanism of activation ofplasminogen at the fibrin surface by t-PA in a plasma milieu invitro: Role of alpha-2 antiplasmin. Biochan J 1990;271:51-57

21. Nichols AV, Krauss RM, Musliner TA: Nondenaturing poly-acrylamide gradient gel electrophoresis. Methods Enzymol1986;128:417-431

22. Anderson DW, Nichols AV, Forte TM, Lindgren FT: Particledistribution of human serum high density lipoproteins. Bio-chim Biophys Ada 1977;493:55-68

23. Laemmli UK: Qeavage of structural proteins during theassembly of the head of bacteriophage T4. Nature 1970;227:680-685

24. Towbin H, Staehelin T, Gordon J: Electrophoretic transfer ofproteins from polyacrylamide gels to nitrocellulose sheets:Procedure and some applications. Proc Natl Acad Sci USA1979;6:4350-4354

25. Fraker PJ, Speck JC Jr Protein cell membrane iodinationswith a sparingly soluble chloroamide, 1,3,4,6-tetrachloro-3a,6a-diphenylgrycoluril. Biochan Biophys Res Commun 1978;80:849-857

26. Angles-Cano E, Boyer B, Gisselbrecht S, Debre P: Heteroge-neity of plasminogen activator expression in various moloneyvirus-induced tumor cell lines. Int J Cancer 1984^3:277-280

27. Bachmann F: Fibrinotysis, in Verstrate M, Vermylen J, LijnenHR, Arnout J (eds): Thrombosis Haemostasis 1987. Leuven,Belgium, International Society of Thrombosis and Haemosta-sis and Leuven University Press, 1987, pp 227-265

28. Lijnen HR, Hoylaerts M, Collen D: Isolation and character-ization of a human plasma protein with affinity for the lysinebinding sites in plasminogen. / Biol Chem 1980;255:10214-10222

29. Lijnen HR, Van Hoef B, Collen D: On the role of thecarbohydrate side chains of human plasminogen in its inter-action with alpha2-antiplasmin and fibrin. Eur J Biochem1981;120:149-154

30. Suenson E, Lutzen O, Thorsen S: Initial plasmin-degradationof fibrin as the basis of a positive feed-back mechanism infibrinolysis. Eur J Biochem 1984;140:513-522

31. Fleury V, Angles-Cano E: Characterization of the binding ofplasminogen to fibrin-surfaces: The role of carboxy-terminallysines. Biochemistry 1991^30:7630-7636

32. Laplaud M, Saboureau M, Beaubatie L, El-Omari B: Seasonalvariations of plasma lipids and lipoproteins in the hedgehog,an animal model for lipoprotein(a) metabolism: Relation toplasma thyroxine and testosterone levels. Biochim BiophysAda 1989;1005:143-156

33. Utermann G, Hoppichler F, Dieplinger H, Seed M, ThompsonG, Boerwinkle E: Defects in the low density lipoproteinreceptor gene affect lipoprotein(a) levels: Multiplicative inter-actions of two gene loci associated with premature atheroscle-rosis. Proc Natl Acad Sci U S A 1989;86:4171-4174

34. Scanu AM: Lipoprotein(a): A potential bridge between thefields of atherosclerosis and thrombosis. Arch Pathol Lab Med1988;112:1045-1047

KEY WORDS • solid-phase fibrin • plasminogen activationfibrinolysis • hedgehogs • lipoprotein(a)

by guest on April 13, 2017

http://atvb.ahajournals.org/D

ownloaded from

D Rouy, P M Laplaud, M Saboureau and E Anglés-CanoAn in vitro study.

Hedgehog lipoprotein(a) is a modulator of activation of plasminogen at the fibrin surface.

Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 1992 American Heart Association, Inc. All rights reserved.

Avenue, Dallas, TX 75231is published by the American Heart Association, 7272 GreenvilleArteriosclerosis, Thrombosis, and Vascular Biology

doi: 10.1161/01.ATV.12.2.1461992;12:146-154Arterioscler Thromb Vasc Biol. 

http://atvb.ahajournals.org/content/12/2/146World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://atvb.ahajournals.org//subscriptions/

at: is onlineArteriosclerosis, Thrombosis, and Vascular Biology Information about subscribing to Subscriptions:

  http://www.lww.com/reprints

Information about reprints can be found online at: Reprints: 

document.Permissions and Rights Question and AnswerFurther information about this process is available in theis being requested is located, click Request Permissions in the middle column of the Web page under Services.Clearance Center, not the Editorial Office. Once the online version of the published article for which permission

can be obtained via RightsLink, a service of the CopyrightArteriosclerosis, Thrombosis, and Vascular Biology Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on April 13, 2017

http://atvb.ahajournals.org/D

ownloaded from