CORE Metadata, citation and similar papers at core.ac.uk Provided by Universiteit Twente Repository Effect of fibronectin on the binding of antithrombin 111 to immobilized heparin Youngro Byun,' Harvey A. Jacobs,' Jan Feijen,' and Sung Wan Kim'** 'Department of Pharmaceutics and Pharmaceutical Chemistry, and Center for Controlled Chemical Delivery, The University of Utah, Salt Lake City, Utah; *Department of Chemical Technology, University of Twen te, Enschede, The Netherlands An objective of this research is to verify the mechanism of ence of different fibronectin concentrations. The binding anticoagulant activity of surface-immobilized heparin in the interaction was studied by first binding immobilized hepa- presence of plasma proteins. The competition and binding rin with ATIII, followed by the introduction of fibronectin; interaction between immobilized heparin and antithrombin heparin binding with fibronectin, followed by incubation I11 (ATIII)/thrombin have been described in vituo. However, with ATIII, and simultaneous incubation of surface immo- the strong ionic character of heparin leads to its specific and bilized heparin with ATIII and fibronectin. The extent of nonspecific binding with many other plasma proteins. ATIII binding to heparin in each experiment was assayed Most notably, fibronectin contains six active binding sites using a chromogenic substrate for ATIII, S-2238. for heparin which may interfere with the subsequent bind- The results of this study demonstrate that the displace- ing of heparin with ATIII or thrombin. ment of ATIII from immobilized heparin was proportional Heparin was covalently immobilized through polyethyl- to the fibronectin concentration, and was reversible. Fur- ene oxide (PEO) hydrophilic spacer groups onto a model thermore, the binding sequence did not play a role in the surface synthesized by random copolymerization of styrene final concentration of ATIII bound to immobilized heparin. and p-aminostyrene. The binding interaction of immobi- 0 1996 John Wiley & Sons, Inc. lized heparin with ATIII was then determined in the pres- INTRODUCTION Due to the high molecular weight and ionic char- acter of heparin, it has been studied for its binding Heparin is a polydispersed anionic polysaccharide interaction with other plasma Specifi- molecule with a molecular weight ranging between cally, heparin binding to plasma proteins has been 6000 and 35,000. Heparin binds and catalyzes the in- divided into three classes, depending on the type and teraction of plasma proteins involved in the intrinsic extent of binding interaction. The first class includes and extrinsic clotting cascade, especially antithrom- specific binding of ATIII to heparin, resulting in a bin I11 (ATIII). ATIII is the natural antagonist for conformational change in the protein and catalyzing thrombin, the protein which enzymatically cleaves fi- its anticoagulant activity. The second class of binding brinogen to form the fibrin clot. interaction includes molecules such as heparin cofac- Rosenberg and Damus' described the overall tor 11,4 lip~protein,~histidine-rich glycoprotein,6 and mechanism by which thrombin is inactivated by PF4.7 The binding of this group is dependent on the ATIII. These investigators showed that ATIII neutral- charge density of heparin and protein configuration. izes thrombin by forming a 1:l stoichiometric com- The third class of binding includes proteins such as plex via a reactive site (arginine)-active center (serine) fibronectin,' thrombin, and vitronectin.' This bind- interaction. The complex formation occurs at a rela- ing interaction to heparin depends on the charge den- tively slow rate in the absence of heparin. However, sity and the chain length of heparin. heparin binds to lysine residues on ATIII, thereby In previous research performed in our laboratory, accelerating the inhibition of thrombin. Winterton et al. lo studied the adsorption of fibronec- tin, albumin, and fibrinogen onto surface-immo- *To whom correspondence should be addressed at De- bilized heparin. They showed that both albumin and partment of Pharmaceutics and Pharmaceutical Chemistry, and Center for Controlled Chemical Delivery, Biomedical fibrinogen had no binding affinity to heparin at phys- Polymers Building (#570), Room 205, The University of iological pH. However, human plasma fibronectin Utah, Salt Lake City, LJT 84112. was shown to bind to solution heparin and immobi- Journal of Biomedical Materials Research, Vol. 30, 95-100 (1996) 0 1996 John Wiley & Sons, Inc. CCC 0021-9304/96/010095-06 96 BYUN ET AL. lized heparin. Furthermore, fibronectin was shown to anate-reactive terminal groups. PEO spacer was have six active binding sites for heparin which may grafted onto the polymer substrate through a chem- be sterically blocked in some adsorbed states. ical coupling reaction between the isocyanate group The anticoagulant action of heparin immobilized of modified PEO spacer and the -NH, group of onto polymer surfaces using grafted hydmphilic p-aminostyrene. The surface density of grafted spacer groups has been and the binding spacer was determined to be 16.3 pmol/cm2, as re- mechanisms of immobilized heparin with ATIII and ported in a previous ~tudy.'~,'~ thrombin were also delineated in vitr~.'~,~~However, Low-molecular-weight heparin (Hepar Industries, since heparin (both immobilized and solution) is able Inc., Franklin, OH), with a molecular weight of 6000 to bind other plasma proteins, especially in vivo, stud- and an anticlotting activity of 95 USP Ulmg (Factor ies defining the competitive binding between ATIII Xa), was used in these experiments. This heparin was and plasma proteins is necessary. As mentioned, fi- fractionated on an ATIII-affinity column to obtain bronectin is a prominent plasma protein with specific high-ATIII-affinity heparin. Approximately 12% of binding sites for heparin, and plays a vital role in the initial heparin concentration was isolated in the hemostasis. Therefore, the ability of immobilized high-affinity fraction, resulting in an absolute activity heparin to bind to ATIII in the presence of fibronectin of 323 USP Uimg (FXa), nearly 3.4 times the activity of in vitro may help explain the anticoagulant mecha- the unfractionated nism of immobilized heparin in vim. This high ATIII affinity heparin fraction was immo- bilized onto the PEO spacer by a coupling reaction between isocyanate groups of the PEO spacer and the amine or hydroxyl groups of heparin. The surface EXPERIMENTAL PROCEDURES density of immobilized heparin was 13.7 pmol/cm2, as cited in a previous report.I4 The binding mole ratio Heparinized surfaces of immobilized heparin to PEO spacer was 0.91:1, implying that nearly one PEO group was coupled with one heparin m~lecule.'~ Heparin was immobilized onto a surface via a hy- drophilic spacer, as shown in Figure 1. Random co- polymers of styrene (Aldrich Chemicals, Milwaukee, WI) and p-aminostyrene (Polysciences, Warrington, Purification of fibronectin PA) were synthesized as the polymer substrate. The optimum composition ratio of styrene and p-ami- Fibronectin was isolated from human plasma by nostyrene was 74126, as defined in previous re- affinity chromatography on a gelatin-Sepharose col- ~earch.'~,'~The polymer substrate was then coated umn at room temperature. Phenylmethysulfonylflu- onto glass beads (100 ? 10 pm diameter, Ferro- oride (0.1 mM) was added as a protease inhibitor to Cataphote Co., Jackson, MS) for subsequent polyeth- plasma, as well as to all buffers. Tris buffer (pH 7.4) ylene oxide (PEO) and heparin coupling. included 10 mM EDTA, 0.02% NaN,, 5 mM EACA PEO was used as a hydrophilic spacer group to (€-amino capric acid), and 1 mM benzamidine. couple heparin to the polymer surface. Hydroxyl Plasma (250 mL, Bloodbank Twente-Achterhoek, En- groups of PEO (Sigma Chemical Co., St. Louis, MO) schede, The Netherlands) was centrifuged at 300 g for (MW 3400) were modified with tolylene diisocyanate 30 min. This plasma was eluted over a Sepharose 4B (TDI, Kodak, Co., Rochester, NY) to generate isocy- column (500 mL), and subsequently applied to a 125 mL column of gelatin-Sepharose 4B. The gelatin- Sepharose 4B column was washed with 50 mM Tris- HC1 buffer for 3 h at a flow rate of 75 mL/h. After HEPARIN washing with 50 mM Tris-HC1 buffer overnight, the column was washed with 1M NaCl in 50 mM Tris- HC1 buffer (250 mL) to eliminate any nonspecifically bound fibronectin. The flow rate was 90 mL/h and the washing volume was 250 mL. After washing again with buffer for 3 hours (75 mLih), the column was washed with 1M urea in 50 mM Tris-HC1 buffer (250 mL) to elute the weak-affinity fibronectin. Finally, 4h4 urea in 50 mM Tris-HC1 buffer was eluted with 20 mL/h flow rate. High-affinity fibronectin was eluted Figure 1. Heparin immobilization onto the polymer sur- in the sixteenth through twenty-first volume (42 mL), face. based on UV absorbance at 280 nm. The fibronectin FIBRONECTIN IN BINDING OF ANTITHROMBIN I11 AND HEPARIN 97 solution was diluted to 0.5 mgimL, dialyzed in PBS 100 ~J.Lconcentrated acetic acid. The reaction between (pH 7.4) at 4"C, and stored at -30°C. S-2238 and free thrombin was monitored by measur- ing the enzymatic cleavage of S-2238 to release p-ni- troanilide fragments (pNA, A,,, = 405 nm). Glass treatment Elgue et al.I5reported that ATIII binding to heparin Binding of ATIII to immobilized heparin decreased by nearly 15% when the kinetic experi- ments were performed in plastic tubes, rather than The interaction of ATIII with this heparinized sur- siliconized glass tubes. Therefore, the glass vials used face has been reported.14 In these studies, 250 mg of in subsequent kinetic experiments were treated with polymer-PEO-heparin beads (-60 cm2) were incu- dichlorodimethylsiloxane (Kodak Co., Rochester, bated with various concentrations (0-1 pM) of ATIII NY) to prevent nonspecific protein adsorption. Glass solution. After 10 minutes (the time to obtain maxi- vials were soaked in 5% dichlorodimethylsiloxane in mum binding interaction), the concentration of ATIII toluene at 25°C for 1 h, and then washed three times was assayed and a Scatchard plot was constructed.