Preparation of Columns of Partially Purified Platelet Extract , and Partially Purified

Univalent antibody fragments prepared from a rabbit antiserum raised against whole human platelets completely inhibited adhesion of platelets to immobilized trimeric collagen in a defined, Mg2+-dependent, adhesion assay. An octylglucoside extract of whole platelets completely neutralized this antibody, and all neutralizing activity bound to immobilized wheat germ agglutinin. Further fractionation on concanavalin A gave rise to subfractions that each neutralized only partially at saturation, when tested against antibody concentrations that inhibit 50% of platelet-collagen adhesion. When tested against higher antibody concentrations that completely inhibited adhesion, each subfraction had no detectable neutralizing effect, although the combined subfractions neutralized completely. This and other evidence suggests that more than one platelet entity participates in platelet-collagen adhesion. Although distinct, they appear to play interdependent roles in a single adhesion process. Considerable evidence is accumulating that a restricted number of surface molecules are involved in cell adhesion. The most extensively studied cell adhesion molecule (CAM), l NCAM (l), is a large glycoprotein originally purified from chick neural retinal cells. It apparently mediates adhesion by direct association of N-CAM molecules on adjacent cells (1). A glycoprotein has also been implicated in adhesion in the cellular slime mold Dictyostelium discoideum, but its role in the adhesion process is unclear (2, 3). The approach used to identify these CAMs has been largely immunological. Rabbits were immunized with a crude membrane fraction, and univalent fragments of the immunoglobulins that were raised were used to inhibit cell-cell adhesion in an in vitro assay. In both D. discoideum and chick neural retina, the antibody preparation blocked adhesion completely; and, as might be expected, crude extracts of the antigenic mixture used for immunization neutralized the antibodies. In both cases, a single glycoprotein was isolated that could completely neutralize the antibody effects. This raised the possibility that in each system there might be only one crucial Abbreviations used in this paper." C~, extract fraction that specifically bound to a concanavalin A column; Cn, extract fraction that flowed through a concanavalin A column; CAM, cell adhesion molecule; W~, extract fraction that specifically bound to a wheat germ lectin column; Wa, extract fraction that flowed through a wheat germ lectin column. CAM, although others might still exist that were not immunogenic in the injected animals. Another experimental system under intensive investigation is the adhesion of human platelets to collagen (4, 5), which is an early step in hemostasis. This system is especially attractive in that platelets are specialized for adhesion. Platelet-collagen adhesion is also of great clinical significance in thrombotic and bleeding disorders. Early methods of quantifying this adhesion by observing the association of platelets with subendothelial collagen in blood vessels (6, 7) have been gradually supplemented by more defined assays (5, 8-11). Immunological techniques have been used in several attempts to identify molecules involved in platelet-platelet aggregation (12, 13) and adhesion (14, 15), but results in various adhesion systems have differed, in part because the assays used did not measure a well-characterized adhesion reaction. Recently, we described a quantitative assay of platelet adhesion to purified type 1 trimeric collagen covalently linked to plastic coverslips (10) under well-defined experimental conditions. Adhesion in this assay required a native collagen structure and was absolutely dependent on Mg 2*. This assay provided an opportunity for identification of the molecular basis of this adhesion. We show here that, as expected, an antiserum raised against human platelets completely inhibited platelet-collagen adhesion in our assay, and that these effects were neutralized by platelet extracts. However, in contrast THE JOURNAL OF CELL BIOLOGY VOLUME 99 DECEMBER 1984 2048-2055 2048 © The Rockefeller University Press 0021-9525/8411112048108 $1.00 on O cber 4, 2017 jcb.rress.org D ow nladed fom with work with D. discoideum (2, 3) or neural retina (1), saturating concentrations of subfractions of the platelet extract could not completely inhibit platelet-collagen adhesion, raising the possibility that more than one entity is directly involved in adhesion in this system. In this article, we describe the results with a rabbit antiserum that led to this conclusion. In the subsequent article (16), we describe work that uses a monoclonal antibody and has identified one of the participating molecules. MATERIALS AND METHODS Platelet Adhesion Assay The adhesion of gel-filtered human platelcts to trimeric collagen immobilized on plastic coverslips was assayed as described previously (10). Briefly, ~Crlabeled platelets were prepared and separated from plasma proteins by gel filtration (17) in modified Tyrode's buffer, pH 7.4, containing 1 mg/ml glucose and 2 mg/ml bovine serum albumin. Type I collagen trimers, derived from lathyritic chick skin (18), were covalently coupled to plastic coverslips using a carbodiimide reagent. The coverslips were placed in Linbro wells (Flow Laboratories, Inc., McLean, VA) and covered with 0.3 ml of Tyrode's buffer containing labeled platelets. The wells were incubated for l0 rain at 37°C on a gyratory shaker and washed. Adhesion, determined as 5tCr bound to coverslips, reached saturation within 20 rain, was Mg 2÷ dependent, and specific for collagen trimers or fibrils (10). Preparation of Adhesion-blocking Antiserum Rabbit antiserum was raised by immunization with whole human platelets that were prepared by gel filtration as described previously (10), except that bovine serum albumin was omitted. Platelets prepared by this method contained <0.1% plasma proteins (flbrinogen or fibronectin) by radioimmunoassay and < 1% lymphocyte contamination. The rabbit was injected with 109 platelets in complete Freund's adjuvant and was initially boosted at~er 28 d with l08 platelets in incomplete Freund's adjuvant. Repeated boosts were made at bimonthly intervals. Sera were collected 5, 7, and 9 d after the third and subsequent boosts and pooled, and IgG was prepared by affinity chromatography on a column of protein A-agarose (Pharmacia Fine Chemicals, Piscataway, NJ). The column was equilibrated in 75 mM Na-K phosphate, 0.15 M NaC1, pH 7.2. IgG was eluted with l M acetic acid, neutralized with solid Tris, and dialyzed into l0 mM Na phosphate, pH 7.3, 0.15 M NaC1, l mM EDTA, and 5 mM ~-mereaptoethanol. To prepare Fab, mercuripapain (Sigma Chemical Co., St. Louis, MO) in 70% ethanol was added to the lgG preparation at 20 ~1 per l0 ml oflgG (10 mg/ml) and digestion was carried out at 37°C for l h (19). Papain was inactivated by addition of iodoacetamide to 30 mM for 15 min at 37°C. After dialysis, Fab fragments were separated from Fc fragments by applying the digest to a second column of protein A-agarose. The flow-through fraction, containing Fab, was collected. Purity was checked by gel electrophoresis in SDS. 1251-Fab Binding to Intact Platelets Fab was labeled with ~2~I, and its binding to platelets was determined at a series of concentrations essentially as described by Pidard ct al. (20) except platelets were prepared by gel filtration as in the adhesion assay. Antisera against Fibronectin, Fibrinogen, and Factor VIII/von Willebrand Factor High-titer rabbit antisera against human fibronectin and human fibrinogen were obtained from Calbiochem-Behring Corp. (La Jolla, CA). The antisera gave strong reactions on immunodiffusion with purified antigens. Rabbit antiserum against Factor VIII/von Willebrand factor was the gift of Dr. C. Hougie, University of California, San Diego. Fab fragments of IgG from these antisera were prepared as described above. Preparation of Outdated Platelet Extracts Outdated human platelet packs were obtained from the Veterans Administration Hospital Blood Bank, La Jolla, and the San Diego Blood Bank. Red blood cell contamination was reduced by repeatedly centrifuging at 500 g for 2 min in a Sorvall GS-4 rotor (DuPont Instruments, Sorvall Biomedical Div., Newtown, CT) equilibrated at 10°C, and discarding the pellet. Platelets were pelleted by centrifugation at 3,000 g for 15 min, resuspended in Tris saline (15 mM Tfis-HCI, pH 7.4, 126 mM NaCI, 5.4 mM KCI, 5 x 10 -5 M CaCI2, I mM MgCI2, 0.1% glucose) and washed three to four times, until the supernatant did not contain appreciable protein, as measured colorimetrically (21). Platelets were then extracted by one of the following methods: F R E E Z E T H A W E X T R A C T I O N O F I N T A C T P L A T E L E T S : P l a t e l e t s were resuspended in Tris saline without EDTA, and the platelct suspension was frozen and thawed three times (22). Solubilized material was collected by centrifugation at 100,000 g for 1 h at 4°C. The supernatant was collected and used in subsequent fractionations. O C T Y L G L U C O S I D E E X T R A C T I O N O F I N T A C T P L A T E L E T S ~ Platelets were resuspended in Tris saline containing 1% octylglucoside and 2 #g/ml phenylmcthylsulfonyl fluoride, and were incubated for 1 h at room temperature on a shaker. Centrifugation was performed as above. To test for neutralizing activity of the supernatant in the adhesion assay, octylglucoside was removed by exhaustive dialysis against Tris saline, pH 7.4. The dialyzate was then clarified by centrifugation at 10,000 g in a Sorvall centrifuge. C H Y M O T R Y P S I N T R E A T M E N T O F I N T A C T P L A T E L E T S : The platelet suspension was incubated with 0.4 U/ml insolubilized a-chymotrypsin (Sigma Chemical Co.) for 2 h at room temperature on a shaker. Platelcts and enzyme-derivatized beads were removed by cent~ifugation at 750 g. Phenylmethylsulfonyl fluoride was added to neutralize any residual protease activity, and the extract was clarified at 100,000 g as described above. Assay of Antibody Effects on Adhesion and Their Neutralization Antibody effects on platelct-collagen adhesion were determined by preincubating the platelets in assay medium containing a series of concentrations of immune Fab for l0 min at 37°C and then performing the adhesion assay over a 10-min period. To test platelet extracts for neutralization of the effects of immune Fab on platelet-collagen adhesion, a series of concentrations were reacted with antibody that was at a concentration that inhibited -50% of platelet-collagen adhesion. In general, antiplatelet Fab, at 15 ug/ml, was preincubated with serial dilutions of platelet extract for 30 min at 37°C. 5~Crlabeled human platelets were then added and the tubes incubated for l0 min in a 37°C water bath. Aliquots of 0.3 ml were added to triplicate Linbro wells containing collagen-derivatized coverslips. Platelet adhesion was assayed over a 10-min period as described and quantified by measuring 5~Cr bound to the coverslip. Neutralization of adhesion-blocking antibodies was calculated as described by Brackenbury et al. (23) for each dilution of an extract: % neutralization = % adhesion (Fab + extract) % adhesion (Fab alone)/% adhesion (extract alone) % adhesion (Fab alone) x 100. One unit of neutralizing activity was defined as that amount that gave 50% neutralization of the adhesion-blocking antibody under defined conditions. It was estimated by plotting the log of extract protein concentration against the percent neutralization for serial assay points and fitting a line by the least-squares method (24). Specific activity was obtained by dividing activity, in units per milliliter, by the extract protein concentration (21), in milligrams per milliliter. Estimation of Platelet Lysis Because apparent inhibition of adhesion could result from platelet lysis during the assay, with release of 5JCr, we evaluated lysis at the end of the assay incubation. The supernatant containing unbound platelets and any free 5~Cr was collected and lightly fixed with 0.5% paraformaldehyde in 0.03 M Na phosphate buffer, pH 7.4 (17) for l0 min. An aliquot was then filtered over a 0.2-#m filter (Millipore Corp., Medford, MA), and free ~Cr in the filtrate was determined. This was compared to total 5~Cr in an unfiltered aliquot. Under standard assay conditions, lysis estimated by this method was 0.3% or less, and was not affected by addition of immune Fab. Fractionation of Platelet Extracts on Lectin Columns The platelet extract was applied to a 3-ml column of wheat germ agglutininagarose (Vector Labs, Burlingame, CA), equilibrated in 0.05 M Tris-HCl, pH 7.4, containing 0.15 M NaCl. After washing in this buffer, specific elution was carried out with 0.1 M N-acetyl-o-glucosamine in this buffer. The fraction eluted from the wheat germ column W~, was dialyzed into 0.05 M Tris-HCl, pH 8, containing 0.15 M NaCl, 1 mM CaCl2, I mM MgCl2, and was chromatographed on a 2-ml column ofconcanavalin A-agarose (E-Y Laboratories, San Mateo, CA). The flow-through (W~Cn) was saved, and this column was eluted with 0. l M a-methyl mannoside in the same buffer to give rise to the fraction W~C~. All fractions were dialyzed, lyophilized, and resuspended in Tris saline to assay neutralizing activity. SHADLE AND BARONDES Distinct Entities in Platelet-Collagen Adhesion 2049 on O cber 4, 2017 jcb.rress.org D ow nladed fom Preparation of Columns of Partially Purified Platelet Extract, and Partially Purified Antiplatelet IgG To partially purify the components of antiplatelet IgG that bound to entities that influenced platelet-collagen adhesion, we prepared a fraction enriched in platelet proteins that did not neutralize the adhesion-blocking activity of the antiserum and used this to adsorb irrelevant antiplatelet antibodies. To this end, an aliquot of an octylglucoside extract of whole platelets was repeatedly passed through columns of wheat germ agglutinin-agarose, and the final flowthrough fraction, Wa, was collected. An aliquot of Wa containing ~20 mg of protein was coupled to cyanogen bromide-activated Sepharose (Pharmacia Fine Chemicals) at 4"C (25). The gel was postincubated with 1 M ethanolamine, pH 9, to block unreacted residues. Antiplatelet IgG was then reacted with this column by circulating l0 mg IgG in phosphate-buffered saline, pH 7.2, over Wa agarose for 24 h at 4"C (26). The flow-through fraction, enriched in antibodies that blocked platelet-collagen adhesion, was collected, coupled to cyanogen bromide-activated Sepharose, and used to purify the platelet antigens with which it reacted. To re~over the antibodies that bound to the Wa agarose column, it was washed (26) and specifically eluted with 0.2 M diethylamine, p H I 1.5. The IgG fractions that flowed through and bound to the Wft agarose were digested to Fab fragments, as above, to assess their binding to platelets and their effect on platelet-collagen adhesion.