platelet interaction with modifiedarticular...

Platelet Interaction with Modified Articular Cartilage

ITS POSSIBLE RELEVANCETO JOINT REPAIR

DOROTHEAZUCKER-FRANKLINand LAWRENCEROSENBERG

From the New York University Medical Center, Departments of Medicine and OrthopedicSurgery, New York 10016

A B S T RA C T During studies concerned with theplatelet-collagen interaction, it was observed thatplatelets did not adhere to bovine or human articularcartilage and that cartilage did not induce plateletaggregation in vivo or in vitro. To study the mech-anism responsible for this observation, the role ofproteoglycans was examined. Purified cartilage colla-gen proved to be fully active as a platelet aggregant.Addition of small amounts of proteoglycan subunit(PGS) blocked platelet aggregation, whereas chon-droitin sulfate, a major glycosaminoglycan com-ponent of cartilage matrix, impaired platelet aggrega-tion only at concentrations which resulted in a markedincrease in viscosity. Moreover, PGS abolishedaggregation of platelets by polylysine but did notprevent aggregation by ADP, suggesting that PGSmay block strategically placed lysine sites on thecollagen molecule. Treatment of fresh articularcartilage with proteolytic enzymes rendered thetissue active as a platelet aggregant. In vivo experi-ments demonstrated that surgical scarification of rab-bit articular cartilage does not result in adhesion ofautologous platelets. Treatment of rabbit knee jointswith intraarticular trypsin 1 wk before the injectionof blood resulted in adhesion and aggregation ofplatelets on the surface of the lesions. Since thereis evidence from other studies that some degree ofcartilage healing may take place after initiation ofan inflammatory response, it is postulated that in-duction of platelet-cartilage interaction may eventuatein cartilage repair.

INTRODUCTION

The role of collagen in platelet adhesion and aggrega-tion has been a subject of intensive research formore than a decade (1-3). In general, these studies

Received for publication 21 October 1976 and in revisedform 7 December 1976.

have been aimed at the elucidation of hemostasisor the pathogenesis of atherosclerosis (for reviewsee 4). The possibility that platelets may also servea crucial function in connective tissue repair has some-times been raised but has never been pursued indepth. It is generally recognized that irreversiblyaggregated platelets form a firm substrate for thedeposition of fibrin as well as for the proliferation offibroblasts. The nature of the intermembranous bondbetween individual platelets or the mechanismwhereby platelets adhere to collagen is therefore ofgreat interest from many points of view. Yet, despiteseveral attractive theories which have been ad-vanced in recent years (5-9), in essence the proc-ess has remained poorly understood. On the basisof currently available evidence, it seems likely thata minimal structural unit consisting of rigidly spacedpolar groups is necessary for platelet adhesion andsubsequent aggregation to occur (10-13). To thisshould be added that, in vivo, the reactivity of agiven connective tissue with platelets may not onlybe a reflection of the spacing and incidence of theactive sites on the collagen molecule, but also ofthe type, concentration, and steric conformation of thesurrounding mucopolysaccharides which may blockthese sites. Thus, we observed that platelets donot adhere to cartilage and that nminced cartilagedoes not induce platelet aggregation in vitro (14).This corroborates clinical experience that intact orlacerated cartilage does not support thrombusformation. On the other hand, after complete extrac-tion of the proteoglycan matrix, cartilage collageninduced platelet aggregation. This process appearedto involve the formation of periodic bridges betweenthe platelet membranes and the dense bands ofthe collagen fibrils. Re-addition of defined moietiesof proteoglycans re-established the inertness ofcartilage collagen vis-a-vis platelets (14). Since itseemed possible that a modification of the cartilagematrix would permit platelets to interact with cartilage

The Journal of Clinical Investigation Volume 59 April 1977- 641 -651 641

collagen in vivo and thereby set the stage for cartilagerepair, a study was undertaken in rabbits to demon-strate that platelets would also adhere to cartilageenzymatically treated in the living animal. The ex-periments involving the action of platelets with na-tive and modified articular cartilage in vitro and invivo form the subject of this report.

METHODS

Platelet-rich plasma (Pf-P)Y was prepared from normal sub-jects with plastic disposable syringes containing 1 volof 3.8% trisodium citrate for each 9 vol of blood collected.For some experiments, the blood was collected in heparinas described before (15). The specimens were centrifugedat room temperature at 250 g for 7 min, after which the PRPwas removed with a Pasteur pipette. Platelet countsranged between 200,000 and 400,000/mm3. Platelet-poor plasma (PPP) was prepared similarly except that theanticoagulated blood was centrifuged for 20 min at 1,200g.The interaction of various collagen preparations and controlcompounds with platelets was measured turbidimetricallywith the help of an aggregometer (Chrono-Log Corp.,Havertown, Pa.) by a modification of the method of Bornand Cross (16). The instrument was adjusted for 100%light transmission with 0.4 ml PPP and for 0% light trans-mission with 0.4 ml PRP. Generally, the test compoundswere added in volumes of 0.1 ml after equilibration afterthe addition of 0.05 ml 1% calcium chloride. Alternatively,the experiments were performed in the absence of exog-enous calcium chloride. ADP was obtained from SigmaChemical Co., St. Louis, Mo., and D,L-polylysine fromGallard-Schlesinger Chemical Mfg. Corp., Carle Place,N. Y. When native cartilage was used in aggregometerstudies, 12-15 pieces measuring approximately 1 mm3suspended in 0.1 ml Hanks' solution were added to PRP.

Preparation of proteoglycans. Proteoglycans andcollagen were prepared from bovine proximal humeralarticular cartilage. The isolation, chemical, and physicalcharacterization of the proteoglycan aggregate and sub-unit species used in this study were described in detailpreviously (17, 18). Briefly, proteoglycans were extractedfrom bovine articular cartilage in 3 M MgCl2, 0.15 M potas-sium acetate, pH 6.3, at 5°C. Proteoglycan aggregatewas reassociated by dialysis against 0.15 M potassium ace-tate, pH 6.3, at 5°C, then isolated by equilibrium densitygradient centrifugation under associative conditions. Frac-tion 6 from this gradient, recovered at densities greaterthan 1.69 g/ml, contained 16S proteoglycan subunit and 70Sproteoglycan aggregate in approximately equal amounts.Proteoglycan subunit was isolated from the mixture ofaggregate and subunit by equilibrium density gradientcentrifugation under dissociative conditions in 4 M guani-dine hydrochloride, 3.0 M cesium chloride. The proteo-glycan subunit is referred to as PGS.

Preparation of purified articular cartilage collagen (CC).For the complete removal of the extractable proteoglycan,fresh diced bovine articular cartilage was extracted byslow stirring at 5°C, first in 3 M MgCl2 0.01 M MES (2-[N-morpholino] ethane sulfonic acid), pH 6.3, for 48 h,then sequentially with 3 M GuHCl, 0.01 M MES, pH 6.3,

'Abbreviations used in this paper: CC, cartilage collagen;CS, chondroitin sulfate; MES, 2-(N-morpholino) ethanesulfonic acid; PGS, proteoglycan subunit; PPP, platelet-poorplasma; PRP, platelet-rich plasma.

for 48 h as detailed previously (17). The cartilage residuewas suspended in water (1 g/10 ml) and rehydrated by slowstirring at 5°C for 16 h. Aliquots of the residue were thenfrozen under liquid nitrogen and freeze-powdered underliquid nitrogen with a Spex mill (Spex Industries, Inc.,Metuchin, N. J.). When needed, 10 g of the finely powderedcollagen was suspended in 500 ml of water, stirred at5°C for 16 h, then filtered through a fine nylon mesh toremove large particles. The collagen was then again col-lected by centrifugation (5,000 rpm for 30 min) and freezedried.

Preparation of chondroitin sulfate. Proteoglycan dis-sociatively extracted from bovine nasal cartilage was de-graded with 0.2 N NaOH at room temperature for 24 h.The solution was neutralized with HC1 and 1% potassiumacetate was added. The degraded proteoglycan was pre-cipitated with 3 vol of cold ethanol, collected by centrifu-gation, washed with ethanol, followed by ether, and driedin a vacuum. The product was degraded sequentially withpepsin, trypsin, and papain, as previously described (19).Fractions of chondroitin sulfate and keratan sulfate wereseparated by a stepwise 5% ethanol fractionation in 0.8%BaCl2. The chondroitin sulfate fraction isolated in 25%ethanol was converted to the K+ salt and used in the studiesdescribed here.

Viscosity measurements. Solutions were prepared by dis-solving samples dried to constant weight in 0.15 M NaCl,0.02 M N-Tris hydroxymethyl-2-aminoethane sulfonicacid, pH 7. Addition of solvent to dry samples and all dilu-tions were made by weight. Viscosity measurements weremade with an Ubberlohde type viscometer (Cannon Instru-ment Co., State College, Pa.) in a constant temperaturewater bath maintained at 20+0.01°C with a mercury in glasstype thermoregulator (Arthur H. Thomas Co., Philadelphia,Pa., model 9655G), and a Beckman differential thermometer(Beckman Instruments, Inc., Fullerton, Calif.).

Preparation of whole fresh wet cartilage. Bovinearticular cartilage was collected and added immediately toice-cold 0.15 M sodium chloride. Specimens of humanarticular cartilage were parepared from limbs amputatedfor a variety of reasons. The cartilage was diced in 1-mmcubes, washed and used as such, or aliquots were incu-bated at 37°C overnight in the following solutions: (a)Trypsin: 1 mg/ml (0.1%) in 0.15 M NaCl, 0.01 M CaCl2,0.05 M Tris buffer at pH 8.0; (b) Papain: 1 mg/ml in0.15 M NaCl, 0.005 M cysteine, 0.005 M EDTA, 0.05 MMES at pH 6.0; (c) Pepsin: 1 mg/ml in 0.15 M NaCl,0.05 MMESat pH 3.8.

Electron microscopy. For ultrastructural studies, 2-5-mlaliquots of PRP were incubated with fresh cartilage, puri-fied CC proteoglycans, or other agents in the same pro-portions as they were used for the turbidimetric assays.Fixation was accomplished in suspension by the additionof equal amounts of 3% phosphate-buffered glutaralde-hyde at 37°C when gross aggregation had begun in amanually agitated test tube. When no aggregation ensuedthe specimens were fixed after a minimum of 15 min ofincubation with the compound tested. Such specimens werecentrifuged at 1,200 g for 5 min to pellet the formed elementsbefore fixation. To improve contrast of the platelet surfacecoat and the inter-platelet bridges, ruthenium red at a con-centration of 1 mg/ml was added to the fixatives essen-tially as described by Luft (20). Postfixation was achievedin 2%osmium tetroxide for 2 h. Dehydration and embeddingin Epon (Shell Chemical Co., New York) was carriedout as in previous reports (21). Thin sections were cutwith an LKB ultrotome equipped with a diamond knife(LKB Instruments, Inc., Rockville, Md.). A Siemens Elmiskop

642 D. Zucker-Franklin and L. Rosenberg

100T Control cartilage

C

CaCI2

I I I1 2 3

Time (minutes)

7 s,,0SS; ; ,

.J F , t,M

lc

Cartiloge Collagen (CC) - 0.2 mg/mI

Time (minutes)

FIGURE 1 (a) Electron micrograph of platelets sedimented onto fresh articular cartilage.Although the platelets have been deformed by centrifugation, they are not degranulatedand show no adherence to the cartilage. Rbc, erythrocytes; M, cartilage matrix; Arrowsindicate collagen fibrils. Magnification x8,000. (b) Aggregometer curve obtained with freshcontrol cartilage shows that no aggregation is taking place. (c) Curve obtained with colla-gen extracted from same cartilage. Arrows indicate when substances were added.

I electron microscope (Siemens Corp., S. Iselin, N. J.)was the instrument used for all studies.

In vivo experiments. The procedure for creation of thejoint cartilage lesions has become standard in a numberof laboratories. In brief, adult NewZealand rabbits weighingabout 4 kg underwent arthrotomy of both knees underether anesthesia. The patella was retracted medially ex-

posing the articular surface of the femur. Longitudinal

scarifications were made into the articular cartilage with a

no. 15 scapel blade. The wound was irrigated with normalsaline, closed with interrupted sutures of no. 4 chromiccatgut or nylon. The wounds were allowed to heal up to2 wk after which the experimental joint of each animal was

injected with 5 mg trypsin in saline whereas the controljoint of the same animal was only given saline. 1 wk later,0.5 ml of freshly collected heparinized autologous blood

Platelet Reaction with Modified Cartilage 643

80

c6O

,-. _0

20

lb

I I4 5

la

FIGURE 2 High resolution of platelets aggregated with CC as shown in Fig. lc. Degranula-tion is evident. Microtubules (T) and actin filaments (F) (21) remain. Arrows indicateperiodic "bridges" between platelets (short arrows) and between platelets and collagen(long arrows). C, collagen. Magnification x72,000.

was injected into both joints and the animals were sacri-ficed within 30 min. The joint cartilages were preparedfor light and electron microscopy. To ensure satisfactoryfixation the entire distal ends of the femora were removedand submerged in 3% glutaraldehyde where dissectionof the articular cartilage was accomplished with the helpof a dissecting microscope. In one instance, Safranin 0,at a concentration of 0.6%, was added to the fixatives toassess the degree of proteoglycan extraction which hadresulted from the treatment with proteolytic enzymes (22,23).The specimens were washed in phosphate buffer ovemightand postfixed for 2 h in 2% osmium tetroxide. Fixationof specimens in situ with fixatives containing Safranin0 proved to be disadvantageous since the dye obliteratedthe lesions making proper orientation of the tissue during theembedding procedure impossible.

RESULTS

Finely minced fresh articular cartilage did not aggre-gate platelets, nor did platelets adhere to this tissueeven when the cartilage and platelets were pelletedtogether by centrifugation (Figs. la and b). On theother hand, when proteoglycans were completelyextracted, the collagen residue (CC) which had re-tained its tertiary fibrillar structure supported plateletaggregation (Fig. 1c). Moreover, on electron micros-copy, it became evident that the "bridges" whichform between the membranes of platelets aggregatedby other agents (15) could also be resolved between


platelets and collagen, and that the periodicity of the"bridges" corresponded to the dense banding of thecollagen fibrils (arrows, Fig. 2). The addition of smallamounts of PGS to the CC 5 min before incubationwith platelets inhibited the platelet collagen interac-tion as evidenced by the turbidimetric as well asmorphologic studies (Figs. 3 and 4). It is noteworthythat purified proteoglycans alone did not cause plate-let aggregation even when high concentrationsamounting to 2 mg of PGS were used. Since PGSmarkedly increased the viscosity of the suspendingmedium, the possibility existed that platelets wereprevented from aggregating on this basis alone. Twostudies were performed to exclude this possibility.Platelets were incubated with PGS after which anamount of ADPwas added known to induce aggrega-tion in the absence of PGS. As is seen in Fig. 5,a 20-fold increase in the minimum concentration ofPGS which had impaired aggregation of plateletswith collagen did not prevent aggregation of plate-lets with ADP. Secondly, chondroitin sulfate (CS),the major glycosaminoglycan component of PGS, wastested for its effect on the platelet collagen interac-tion. Fig. 6 shows the relative viscosity curves ofPGS and CS. Only when the calculated viscosityof CS was 10 times as high as that of PGSand whenits concentration was almost 200-fold that of PGS

PGS- 0.1 mg/mi + CC- 0.2 mg/ml

I

CaCd2

I I1 2 3

Time (minutes)

.4 5

- a .

FIGURE 3 Aggregometer curve obtained when proteoglycan was added to CC. No aggrega-tion ensued.

FIGuRE 4 Appearance of platelets sedimented with CC and PGS used in the experimentillustrated in Fig. 3. Platelets are not aggregated or degranulated. Black material be-tween platelets represents collagen fibrils and PGS fixed in the presence of ruthenium red.Magnification x 14,000.

O.01 mMADP

so

I*1 -

0A4

20

a

I g . I

0 1 2 4Time (minues)

bII

1 2 3Time (minutes)

4

FIGURE 5 (a) Control platelets aggregated with ADP. (b) An aggregometer curve demonstrat-ing that aggregation of platelets by ADP is not impaired in the presence of a concen-tration of PGS (1 mg/ml) five times higher than that required to inhibit aggregation withcollagen.


100 :

8.Q .

r60.0

._L2 -,0

40

20L

0

3

*1*1 -

i-4

20

4Q

4r

0

* IC

PGS(@) 1 2 3

CS (0) 10 20 30

CONCENTRATION,mg/ml

FIGURE 6 Relative viscosity curves of proteogate and chondroitin sulfate.

did inhibition of aggregation by CS becable (Fig. 7). It is thus likely that Ptspecific than CS in blocking the criticalcollagen molecule which interact with plait had been suggested that free lysine gcollagen molecule may play an imporltriggering the platelet reaction (5) andalso been demonstrated that polylysineplatelet aggregant (24, 25), it seemed o

CS - 18mg/ml + CC- 0.2 mg/ml

so

60.2.A

.T0

0-4O

I I I I I I I I0 1 2 3 4

Time (minutes)

FIGURE 7 Aggregometer curve showing thatcentration of CS causes delayed and incompletThis effect may be due to slow release of endc

test whether PGScould block the effect of polylysineon platelets. Indeed, small amounts of PGS wereable to completely impair platelet aggregation byD,L-polylysine (Figs. 8 and 9). Low concentrationsof CS were ineffective in this regard. The likelihoodthat the proteoglycan in the cartilage matrix preventedadhesion of platelets to the collagen fibrils present inthis tissue, and the knowledge that proteoglycanspossesses a core protein (17) made it seem reason-able to prepare cartilage for platelet adhesion in vivoby means of proteolytic enzymes. Indeed trypsin orpapain treatment of diced articular cartilage overnightrendered the tissue active as regards platelet adhesionand aggregation (Figs. lOa-c). On electron micros-copy, platelets had formed aggregates in association

4- with the collagen fibrils and showed the morpho-4 logical counterpart of the release reaction, i.e., loss

40 so of granules and other cytoplasmic structures. En-couraged by these results, studies were initiatedto examine whether platelet adhesion to cartilage

lycan aggre- could be achieved in rabbit joints by analogousmethods. As reported in detail elsewhere (26, 27)scarifications restricted to the knee joint cartilage

ome notice- of rabbits and not entering subchondral bone (28)GS is more showed no evidence of inflammation or repair nosites on the matter how long after arthrotomy such lesions were

ttelets. Since examined.2 In the present study, no platelets orroups on the fibrin were seen to adhere to the cartilage surfacetant role in 48 h after the intraarticular injection of 0.5 ml ofsince it had heparinized whole blood, although a few morphologi-

is a potent cally unaltered erythrocytes and leukocytes weref interest to present in the space created by the lesion (Figs. Ila-

c). On the other hand, joints which had been treatedwith intraarticular trypsin 8-12 days before the in-jection of blood showed not only adhesion of numerousplatelets to the surface of the cartilage lesion, butin many areas platelets had aggregated and under-gone degranulation (Fig. 12). Leukocytes whichwere probably derived from the injected blood weresometimes seen in association with such plateletaggregates.

DISCUSSION

The observations recorded here have theoretical aswell as practical implications. As regards the plate-let collagen interaction, it has become generallyrecognized that a multimeric or tertiary structure ofcollagen is a prerequisite for platelet aggregation(10-13) even though such polymerization may takeplace in reaction mixtures which, at the outset, con-tain only glycopeptide subfragments (29). The corolary,that this requirement is not met by collagen prepara-

200-fold con- tions or tissues which fail to aggregate platelets, has:e aggregation.vgenous ADP. 2 Rosenberg, L. Unpublished observations.


_ Poly-O -L- Iysine- 0.05 mg/ ml

_~~~~~~~~~~~~~~~~~-__+_B

60

40

20

COC12I I I I I I I

9% . %q . _

.2.0Ec9

4

I £ 3Time (minutes)

6Cr

.2.E -

` 4

20

4 5

FiGuRE 8 Aggregation of platelets with D,L-polylysine.

been tacitly assumed. Since all phases of the presentstudy have been carried out with fibrillar collagenconfirmed by electron microscopy, a discussion ofinhibitors which may interfere with collagen polym-erization and therefore preclude platelet aggrega-tion is not relevant. On the other hand, few investi-gators have given consideration to other moleculespresent in connective tissue matrices which even atlow concentration could block collagen sites whichinteract with the platelet membrane. As has beenshown in this study, small amounts of some proteinpolysaccharide moieties can block collagen aggrega-tion of platelets whereas others, such as CS do not.Therefore, the reason why some connective tissues orbasement membranes are more reactive with plateletsthan others could be attributable to the type and(or)location of the protein polysaccharides rather than

U

COC12 a

3I,Ti_ -__-

4 5

PGS- 0.1lmg/mlI Poly-iD - L- lysine-QO.0 mg/mi

I

CaC12. . I JI

1 2 3Ti me (minutes)

I I4 5

FIGURE 9 Complete inhibition of platelet aggregation byD,L-polylysine in the presence of 0.1 mg/ml PGS whichis reminiscent of the inhibitory effect of PGS on plateletaggregation by collagen.

to the nature or degree of multimerization of thecollagen contained therein. In any case, proteoglycanblockage may be effective even when the plateletaggregant is a single chain polypeptide as has beenexemplified by the observation reported here thatsmall amounts of PGS can prevent platelet aggrega-tion by polylysine (Fig. 9). One wonders whethercyanogen bromide peptides of collagen which pur-portedly did not reassociate (30) would also beinhibited by PGS. At any rate, the observations makeit seem unnecessary to postulate a platelet receptorwhich is specific for collagen. The architecture of theintact collagen fibril may merely provide suitablyspaced polar groups (be they positive or negative)

.0 1 ,A Tig_04"4T, "%

FIGURES lOa-c Comparison of curves obtained with (a) untreated control cartilage showingno platelet aggregation, (b) cartilage treated with papain showing complete aggregation aftera 2-min lag phase, and (c) curve obtained with trypsin-treated cartilage.


100 log

I

;

I . , .*.I%

if.-1 .1I 7

I (o::

S.'

FIGURES lla-c Illustrations of surgically induced articular cartilage lesion. (a) Conventionalphotomicrograph of a lesion 8 mo after surgery stained with Safranin 0. The grooves re-mained patent and no repair is evident. (b) Phase photomicrograph of a thick sectionthrough the lesion from a joint which had been injected with autologous blood. Erythrocytesare seen in the groove, but no thrombus has formed. (c) Electron micrograph of the lesionillustrated in Fig. llb verified the absence of adhering platelets. R, Erythrocytes; C, Cartilage.Magnification x2,400.

able to trigger platelet release. In this context theidentical spacing of bridges between aggregated plate-lets and those between platelet membranes and colla-gen fibrils (Fig. 2) may be significant. It should alsobe noted that the location of the bridges correspondsto the dense bands of the collagen fibril, the pre-sumed site of polar residues in most models of colla-gen structure (31). It will be of interest to examinewhether the number of bridges corresponds to thenumber of sites on the platelet membrane which needto be cross-linked covalently or ionically, as the casemay be, for the release phenomenon to occur. Inthe final analysis, the mechanisms that induce plate-let aggregation may be analogous to those which takeplace when various ligands or multivalent anti-bodies cross-link sites on the lymphocyte membranecausing this cell to undergo transformation. It ispossible that in the case of platelets, the number of

sites which need to be cross-linked for release tooccur may be found by an analysis of the precisenature and binding site of PGS to collagen. Forthat purpose, ultrastructural studies will have to becarried out utilizing smaller tagged subunits ofproteoglycan or its core protein rather than the aggre-gate used in the present experiments.

The practical aspects of the studies described inthis communication were designed to explore condi-tions which may promote healing of articular carti-lage. After injury to most other tissues an inflam-matory response ensues which, broadly speaking,consists of adhesion and aggregation of platelets, con-version of fibrinogen to fibrin, and the invasion offibroblasts and other cells. When the injury is limitedto articular cartilage, no inflammatory responsetakes place. The reason for the lack of such responsehas long been attributed to the absence of blood ves-

FIGURES 12a-c Several areas of the surface of the articular cartilage lesions obtained froma rabbit knee joint which had been treated with intra-articular trypsin 1 wk before theinjection of autologous blood. Platelets are seen adherent in many areas. The majority ofplatelets have undergone degranulation (arrows). Occasionally, fairly large thrombi haveformed in intimate association with the cartilage (12b). The cartilage matrix seems to havebeen partially extracted (12c). Periodicity of fibrils (F) is seen at higher magnification (12c).Magnification: 12a, x6,000; 12b, x4,000; 12c, x11,000.


.q

'ti7 ' I.' . ;t'

I'2 '..:

t2a >1'0* a*N > ,2 % st Ri ' t rS~~~~~q

t"''' "A 'u -e

v.j'' -s ' ' *;

f~,K,>* t l,Fj

*}s tis \ s ~~~~~~~~~'

4.

\

V..K

N 4

6 . . *, r > > t wgorl,

- ,* t ,Bi4/w Et* t

13c1 NN% \..I N%Se%

gr

p -.

II h

k , V.-"Y..A" ". ...

.,t

sels and the resulting inability of blood-bome ele-ments to reach the injured site. However, in thepresent as well as in many previous studies (27),it has been clearly demonstrated that cartilage lacera-tions (Figs. lla and b) do not support platelet orfibrin adhesion even when blood is present in thejoint as a result of arthrotomy or intraarticular in-jection. In the initial phases of this work, we believedthat inhibition of platelet adhesion to cartilage wasdue to a physical effect, i.e., contact between plate-lets and collagen fibrils was prevented by the denseconcentration of proteoglycan aggregates cover-ing the collagen fibrils. The finding that proteoglycansubunit solutions at extremely low concentrationsand low viscosities completely inhibited collagen-induced platelet aggregation while CS solutions atmuch higher concentrations and viscosities did not,came as a surprise. This observation raises the possi-bility that the mechanism of inhibition may involvethe specific binding of proteoglycan subunit to colla-gen fibrils at sites close to or identical to the sitesrequired for platelet adhesion. As suggested in theforegoing, treatment with proteolytic enzymes mayaffect the core protein of the proteoglycan subunitwhich may interact with the collagen molecule, therebymaking crucial sites accessible to platelets. Oneadditional observation deserves mention in thecontext of this discussion. During platelet aggregationa host of active substances are released, known toplay a role in the inflammatory response. Recentlya new platelet-derived factor has been reported (32)that is released after platelet aggregation and thatappears to enhance the proliferation of fibroblasts andother connective tissue cells (33, 34). Therefore, it ispossible that similar factors released by plateletswhich have been induced to adhere and aggregateon the surface of enzyme-modified cartilage couldpromote the proliferation of chondrocytes and(or)the synthesis of proteoglycans by these cells. Thefirst step, namely the induction of platelet adhesionand aggregation on articular cartilage in the livinganimal, has been achieved. Whether this willeventuate into repair of the cartilage lesions mustawait long-term observations which are currently inprogress.

ACKNOWLEDGMENTS

The excellent assistance of George Grusky and SusanDittmar is gratefully acknowledged.

This investigation was supported by U. S. Public HealthService grants AM012274 and AM01431.

REFERENCES

1. Bounameau, Y. 1959. L'accolement des plaquettes auxfibres sous-endotheliales. C. R. Sednces Soc. Biol. Fil.153: 865-867.

2. Hugues, J. 1960. Accolement de plaquettes au colla-gene. C. R. Seances Soc. Biol. Fil. 154: 866-868.

3. Zucker, M. B., and J. Borrelli. 1962. Platelet clumpingproduced by connective tissue suspensions and bycollagen. Proc. Soc. Exp. Biol. Med. 109: 779-787.

4. Mustard, J. F. 1968. Hemostasis and thrombosis. Semin.Hematol. 5: 91-106.

5. Wilner, G. D., H. L. Nossel, and E. C. LeRoy. 1968.Aggregation of platelets by collagen. J. Clin. Invest. 47:2616-2621.

6. Wilner, G. D., H. L. Nossel, and T. L. Procupez. 1971.Aggregation of platelets by collagen: polar activesites of insoluble human collagen. Am. J. Physiol. 220:1074- 1079.

7. Barber, A. J., and G. A. Jamieson. 1971. Platelet colla-gen adhesion characterization of collagen glucosyl-transferase of plasma membranes of human bloodplatelets. Biochim. Biophys. Acta. 252: 533-545.

8. Chesney, C. McI., E. Harper, and R. W. Colman. 1972.Critical role of the carbohydrate side chains of collagenin platelet aggregation. J. Clin. Invest. 51: 2693-2701.

9. Puett, D., B. K. Wasserman, J. D. Ford, and L. W.Cunningham. 1973. Collagen-mediated plateletaggregation. Effects of collagen modification involvingthe protein and carbohydrate moieties. J. Clin. Invest.52: 2495-2506.

10. Muggli, R., and H. R. Baumgartner. 1973. Collagen in-duced platelet aggregation: requirement for tropo-collagen multimers. Thromb. Res. 3: 715-728.

11. Jaffe, R., and D. Deykin. 1974. Evidence for a structuralrequirement for the aggregation of platelets by collagen.

J. Clin. Invest. 53: 875-883.12. Brass, L. F., and H. B. Bensusan. 1974. The role of

collagen quaternary structure in the platelet: collageninteraction. J. Clin. Invest. 54: 1480-1487.

13. Simons, E. R., C. McI. Chesney, R. W. Colman, E.Harper, and E. Samberg. 1975. The effect of theconformation of collagen on its ability to aggregateplatelets. Thromb. Res. 7: 123-139.

14. Zucker-Franklin, D., and L. Rosenberg. 1972. Studieson platelet-collagen interaction. III. Congress onThrombosis and Haemostasis Abstracts. 268. (Abstr.)

15. Zucker-Franklin, D. 1969. Microfibrils of blood plate-lets: Their relationship to microtubules and thecontractile protein. J. Clin. Invest. 48: 165-175.

16. Bom, G. V. R., and M. J. Cross. 1963. The aggrega-tion of blood platelets. J. Physiol. (Lond.). 168: 178-195.

17. Rosenberg, L. C., S. Pal, and R. J. Beale. 1973. Pro-teoglyeans from bovine proximal humeral articularcartilage. J. Biol. Chem. 248: 3681-3690.

18. Rosenberg, L., W. Hellman, and A. K. Kleinschmidt.1975. Electron microscopic studies of proteoglycanaggregates from bovine articular cartilage. J. Biol.Chem. 250: 1877-1883.

19. Rosenberg, L., M. Schubert, and J. Sandson. 1967.The protein-polysaccharides of bovine nucleus pulposus.

J. Biol. Chem. 242: 4691-4701.20. Luft, J. H. 1971. Ruthenium red and violet. II. Fine

structural localization in animal tissues. Anat. Rec. 171:369-416.

21. Zucker-Franklin, D., and G. Grusky. 1972. The actinand myosin filaments of human and bovine bloodplatelets. J. Clin. Invest. 51: 419-430.

22. Rosenberg, L. 1971. Chemical basis for the histologicaluse of safranin 0 in the study of articular cartilage.

J. Bone Jt. Surg. 53A: 69-82.23. Shepard, N., and N. Mitchell. 1976. The localization of


proteoglyean by light and electron microscopy usingSafranin 0. A study of epiphyseal cartilage. J. Ultra-struct. Res. 54: 451-460.

24. Schneider, W., W. Kubler, and R. Gross. 1968. Induc-tion of blood platelet aggregation by cationic polypep-tides. Thromb. Diath. Haemorrh. 19: 307.

25. Massini, P., and E. F. Luscher. 1972. On the mech-anism by which cell contact induces the release reac-tion of blood platelets; the effect of cationic polymers.Thromb. Diath. Haemorrh. 27: 121-133.

26. Fuller, J. A., and F. N. Ghadially. 1972. Ultrastructuralobservations on surgically produced partial-thicknessdefects in articular cartilage. Clin. Orthop. Relat. Res.86: 193-205.

27. Mankin, H. J. 1974. The reaction of articular cartilageto injury and osteoarthritis (First of two parts). N.Engl. J. Med. 291: 1285-1292.

28. Mitchell, N., and N. Shepard. 1976. The resurfacing ofadult rabbit articular cartilage by multiple perfora-tions through the subchondral bone. J. Bone Jt. Surg.58A: 230-233.

29. Kang, A. H., E. M. Beachey, and R. L. Katzman. 1974.

Interaction of an active glycopeptide from chick skincollagen (al-CB5) with human platelets. J. Biol. Chem.249: 1054-1059.

30. Chiang, T. M., E. H. Beachey, and A. H. Kang. 1975.Interaction of chick skin collagen fragment (al-CB5)with human platelets. Biochemical studies during theaggregation and release reaction. J. Biol. Chem. 250:6916-6922.

31. Hodge, A. J., and F. 0. Schmitt. 1960. The chargeprofile of the tropocollagen macromolecule and thepacking arrangement in native-type collagen fibrils.Proc. Natl. Acad. Sci. U. S. A. 46: 186-197.

32. Kohler, N., and A. Lipton. 1974. Platelets as a sourceof fibroblast growth-promoting activity. Exp. Cell Res.87: 297-301.

33. Ross, R. 1975. Connective tissue cells, cell prolifera-tion and synthesis of extracellular matrix-A review.Philos. Trans. R. Soc. Lond. B. Biol. Sci. 271: 247-259.

34. Rutherford, R. B., and R. Ross. 1976. Platelet factorsstimulate fibroblasts and smooth muscle cells quiescentin plasma serum to proliferate.J. Cell Biol. 69: 196-203.


platelet interaction with modifiedarticular...

Documents