Fibronectin-Based Masking Molecule Blocks Platelet Adhesion

Synopsis. Vessel wall extracellular matrix, which underlies the endothelium, is a potemt stimulator of platelet adhesion and activation. Exposure of t...
0 downloads 0 Views 191KB Size
JULY/AUGUST 2003 Volume 14, Number 4 © Copyright 2003 by the American Chemical Society

COMMUNICATIONS Fibronectin-Based Masking Molecule Blocks Platelet Adhesion David H. Geho,*,† William I. Smith, Jr.,§ Lance A. Liotta,†,‡ and David D. Roberts†,‡ Laboratory of Pathology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892, and Department of Pathology, Suburban Hospital, Bethesda, Maryland 20814. Received March 12, 2003

Vessel wall extracellular matrix, which underlies the endothelium, is a potent stimulator of platelet adhesion and activation. Exposure of this matrix can result from damage incurred by vascular interventions, such as saphenous vein bypass grafting and angioplasty. Fibrillar collagens are an important component of the thrombogenic extracellular matrix. Herein we describe a means of targeting poly(ethylene glycol) (PEG)-mediated blockade directly to platelet-binding ECM molecules, such as type I collagen, thereby selectively blocking platelet adhesion to vascular matrix. Purified fibronectin (FN), a matrix protein that interacts with fibrillar collagens and platelets, was selectively pegylated to generate a targeted molecular shielding reagent that masked ECM ligands from platelet recognition and adhesion. This approach protects the functions of other vascular proteins, including surface proteins on intact endothelium. To mask the platelet-binding site of FN, PEG-propyl moieties (5000 Da) were covalently appended to lysine residues on the surface of FN, generating FNPEG-5K. To preserve the collagen-binding function of FN, it was pegylated while bound to a gelatin agarose matrix. We demonstrate that FNPEG-5K blocks platelet adhesion to purified type I collagen. Moreover, the same preparation blocks platelet adhesion to vascular wall components, including collagens.

INTRODUCTION

Platelet adhesion to extracellular matrix (ECM)1 proteins exposed by damage incurred during vascular pro* Correspondence: David Geho, Laboratory of Pathology, NCI, Building 10-2N212, 10 Center Drive, Bethesda, MD 20892. Telephone: 301-594-2944. Fax: 301-480-9488. E-mail: gehod@ mail.nih.gov. † National Institutes of Health. § Suburban Hospital. ‡ Both of these senior authors contributed equally to this investigation. 1 Extracellular matrix (ECM), fibronectin (FN), phosphate buffer saline (PBS), poly(ethylene glycol) (PEG), bovine serum albumin (BSA), Tyrode’s calcium-free buffer (TCFB), prostaglandin E1 (PGE1).

cedures can lead to thrombotic occlusions at the intervention sites (1). A number of vascular ECM proteins contribute to platelet adhesion, including collagens and fibronectin (FN) (2, 3). Herein we describe a means of targeting poly(ethylene glycol) (PEG)-mediated blockade directly to platelet-binding ECM molecules, thereby selectively blocking platelet adhesion to vascular matrix. Purified FN, a matrix protein that interacts with fibrillar collagens and platelets, was selectively pegylated to generate a targeted molecular shielding reagent that masked ECM ligands from platelet recognition and adhesion2 (3-5). This approach protects the functions of other vascular proteins, including surface proteins on intact endothelium.

10.1021/bc034037v CCC: $25.00 © 2003 American Chemical Society Published on Web 06/27/2003

704 Bioconjugate Chem., Vol. 14, No. 4, 2003

Geho et al.

Figure 1. FNPEG-5K binds collagen, lacks platelet binding properties, and binds to blood vessels. (A) FN (lane 1, 2 µg) and FNPEG5K (lane 2, 2 µg) were run on a reducing 7.5% SDS-PAGE and stained with Coomassie blue. FN migrated at Mr 220 000. (B) Binding of iodinated native (0) or modified FN (b) to type I collagen coated wells. (C and D) A human vein was incubated with FNPEG-5K that contained biotinylated PEG moieties, and biotin reactivity resulting from FNPEG-5K adherence was probed. (C) No FNPEG5K added to tissue (negative control). (D) Biotinylated FNPEG-5K (10 µg/mL). (E) Binding of fluorescent labeled platelets to plate bound FN and FNPEG-5K. The results shown for experiments B and E are representative of three similar experiments for each, mean ( SD; C and D are representative of two similar experiments.

To derive an ECM-targeted masking molecule from FN, we neutralized its ability to bind platelets while protecting its collagen-binding properties (2, 3). Covalent addition of PEG residues to protein surfaces can inhibit undesirable interactions, such as stimulating an immune response (6). Accordingly, modification of lysines present on the surface of plasma FN with mPEG-propionaldehyde via reductive amination (7, 8) provided a means to neutralize its platelet adhesive function. In order for FNPEG-5K to be a targeted masking molecule, it needed 2 Human FN was purified from human plasma using a previously described protocol (see ref 5). Briefly, human plasma was applied to a sepharose CL-4B column and then directly applied to a gelatin agarose matrix that had been pretreated by washing with 4 M urea, 10 mM Tris-HCl, pH 7, followed by equilibration with phosphate-buffered saline (PBS). The column was then washed with 1 M NaCl in PBS followed by washing with plain PBS. FN was then eluted by using an acidic buffer (0.1 M NaCl, 0.05 M citric acid, pH 5.5) and neutralized by adding 1.0 M Na2HPO4 at 1/20th volume. Protein concentrations were measured using a bicinchoninic acid kit. To make FNPEG5K, purified FN was bound to gelatin agarose followed by the addition of 8 mg/mL sodium cyanoborohydride and 32 mg/mL of monomethyl poly(ethylene glycol) (PEG)-propionaldehyde 5 kDa (Shearwater, Huntsville, AL) in 0.1 M sodium phosphate buffer (pH 8.0). After at least 16 h, the protein was eluted. For biotinylated FNPEG-5K, biotin-PEG-NHS (3400 Da) (Shearwater) was incubated with gelatin-immobilized FN for 1 h followed by pegylation as above. To assess the extent of pegylation, FN and FNPEG-5K were compared by 7.5% TrisHCl SDS-PAGE and a fluorescamine assay using a standard protocol (see ref 10).

to retain its binding activity for fibrillar collagen in the subendothelial matrix. An immobilized susbstrate was used previously to protect enzyme active sites from modification during pegylation (9). Gelatin, which is a denatured form of type I collagen, binds FN with a higher affinity than native collagen (4). Purified plasma FN, therefore, was bound to a gelatin agarose affinity matrix during pegylation (5). SDS-PAGE assessment showed that FNPEG-5K prepared in this manner migrated more slowly than native FN, consistent with the addition of PEG moieties (Figure 1A). Conditions for the pegylation of FN were optimized to minimize platelet adhesion and preserve collagen binding by varying the duration of derivatization and the concentrations of both mPEG-propionaldehyde and sodium cyanoborohydride. A fluorescamine assay indicated that 25-30% of available lysines were pegylated using optimal reaction conditions (10). In binding studies, iodinated FNPEG-5K bound plate-immobilized collagen in a manner similar to native FN (Figure 1B).3 Thus, the key collagen-binding properties of FN were selectively protected by this targeted modification approach. A biotinylated form of FNPEG-5K was incubated with a human blood vessel that had been stretched to induce injury to confirm that FNPEG-5K retained binding to exposed blood vessel wall components (Figure 1C,D).4 Having FNPEG-5K in hand, we next tested whether the PEG moieties inhibited platelet adhesion (Figure 1E)5 (11, 12). Unlike native FN, immobilized FNPEG-5K did not mediate adhesion of platelets. The cell-binding domain of FN contains an RGD sequence as well as a

Communications

Bioconjugate Chem., Vol. 14, No. 4, 2003 705

Figure 2. FNPEG-5K inhibits platelet adhesion. (A) Soluble FN or FNPEG were bound to immobilized type I collagen. After being washed and blocked, platelets were incubated for 30 min. After repeated washings, adherent platelets were fixed with glutaraldehyde, stained, and counted. The level of background binding of platelets to blocked plastic was below 5% in the experiment shown. The result shown is representative of three similar experiments, mean ( SD. (B, C, D, E) Human ileac artery autopsy specimens were coated with FNPEG-5K and then incubated with biotinylated platelets. After unbound platelets were washed away, frozen sections of the tissues were made, and biotin reactivity resulting from platelet adherence was probed. (B) No FNPEG-5K and no platelets added to tissue (negative control). (C) Only biotinylated platelets added (positive control). (D) FNPEG-5K (30 µg/mL) and platelets added. (E) FNPEG-5K (60 µg/mL) and platelets added. The results shown in B, C, D, and E are representative of two similar experiments.

number of residues that contribute to an electrostatic interaction with the platelet integrins (13). PEG residues attached to the cell-binding domain may sterically hinder 3 Fibronectin-125I Binding Assay. Iodinated FN was added to gelatin agarose, equilibrated as above, and incubated, rotating for 1 h at room temperature. Following this incubation, the supernatant was removed and a reaction mixture containing 4 mg/mL NaCNBH3 and 8 mg/mL of mPEG-propionaldehyde 5kDa diluted in 0.1 M Na2HPO4 buffer, pH 8.0, was added and rotated overnight for 16 h at room temperature. Following this incubation, the protein was eluted as before. Equivalent counts of native Fn-I125 and Fn-I125 PEG-5K were added to Immulon 2 wells that had been precoated with type I collagen (Vitrogen 100, Collagen Corporation, Palo Alto, CA) overnight at 4 °C followed by 1 h blocking with 1% BSA in DPBS at room temperature. After an incubation at 4 °C for 3 h, the wells were washed and aspirated twice with 100 µL of DPBS. The wells were then collected and counted in a gamma counter. 4 Binding Experiments with Blood Vessels. Blood vessels were obtained from either autopsy specimens or anonymized samples from excess surgical material. The National Institutes of Health Office of Human Subjects Research deems these types of specimens exempt. The vessels were dissected, damaged by lightly stretching the vessel during handling, and incubated in 0.5% BSA. Depending on the experiment, FNPEG-5K or a biotinylated form of FNPEG-5K was added to the vessels. After an incubation of at least 30 min, the vessel pieces were washed three times with PBS. For vessels treated with nonbiotinylated FNPEG-5K, platelets that had been labeled with sulfo-succinimidyl-6-(biotinamido)hexanoate were then added to the vessels. After 30 minutes, the nonadhered platelets were removed by three washes. The vessels were fixed with glutaraldehyde and processed as frozen sections. The slides were rinsed and probed for biotin reactivity using Vectalabs Elite ABC and DAB kits (Vector Laboratories, Burlingame, CA).

these interactions between FNPEG-5K and platelet integrins, leading to a loss of platelet binding. Since modified FNPEG-5K retained the collagen-binding properties of native FN while lacking its intrinsic platelet adhesive properties, the ability of FNPEG-5K 5 Plastic-based Platelet Adhesion Assays. On the basis of a previously published protocol (see ref 11), Immulon 1B 96well plates were coated with FN or FNPEG-5K and blocked with 100 µL of Tyrode’s calcium-free buffer (140 mM NaCl, 3 mM KCl, 12 mM NaHCO3, 0.4 mM NaH2PO4‚H2O, 1 mM MgCl2‚6 H2O, 0.35% BSA, 0.1% glucose, pH 7.33) (TCFB) with 3% BSA and 20 ng/mL prostaglandin E1 (PGE1). The wells were then washed three times with TCFB. Platelets, approved for use under the National Institutes of Health IRB protocol 99CC0168, were retrieved from the blood bank and resuspended in TCFB (1 × 109/ml) with 24 µg/mL of 5-(and-6)-carboxyfluorescein diacetate succinimidyl ester (Molecular Probes, Eugene, OR), 20 ng/mL PGE1, and 130 µg/mL apyrase and incubated for at least 30 min at 37 °C (see ref 12). Another incubation followed for 10 min at 37 °C in TCFB with 20 ng/mL PGE1. After being washed, the platelets were incubated in FN-coated wells and then washed multiple times, and the plate was read using a spectrofluorimeter plate reader with excitation of 485 nm and emission of 535 nm. For platelet/collagen binding assays, Falcon 1008 plastic dishes were coated with 6 µL drops of type I collagen. The plates were washed, and then FN or FNPEG-5K was added for at least 1 h at room temperature followed by aspiration and blocking with TCFB (1% BSA, 20ng/mL PGE1). Two mL of platelets (prepared as before except for fluorescent label) in TCFB (20 million/mL) were added to the dish and incubated for 30 to 40 min at room temperature. Unbound platelets were then washed away and the bound platelets were fixed with 1% glutaraldehyde. The bound platelets were stained with DiffQuick (Dade Behring AG, Dudengen, Switzerland) and counted.

706 Bioconjugate Chem., Vol. 14, No. 4, 2003

to inhibit platelet adherence to type I collagen, a key ECM component, was investigated (Figure 2A). Although native FN enhanced platelet adhesion to a collagencoated substrate, FNPEG-5K potently and dramatically protected type I collagen from platelet binding. In a similar manner, the adventitia of blood vessels, a rich source of type I collagen, were incubated with FNPEG5K, followed by incubation with biotinylated platelets (14). The platelets bound readily to adventitial vascular wall components (Figure 2C). In contrast, preincubation of FNPEG-5K with the blood vessel markedly decreased, in a dose-dependent manner, the number of platelets that adhered to the blood vessel wall (Figure 2D,E). Thus, coating with FNPEG-5K protects vessel wall components from platelet adhesion. In this communication, we describe a method for transforming native FN from a platelet-binding protein into FNPEG-5K, a targeted molecular shielding reagent that masks ECM collagens from platelet recognition and adhesion. Such a reagent could potentially be used in vascular interventions. For example, during vessel harvesting for bypass procedures, the vessel could be bathed in FNPEG-5K prior to anastomosis. In so doing, sites of platelet adhesion could be masked with FNPEG-5K prior to reestablishment of blood flow through the vessel. This would leave nondamaged vascular cell surface proteins free to function normally. This represents a significant advance over a previously described method wherein function-attenuating PEG was added indiscriminately to vascular graft proteins (15). Moreover, platelet function would not be compromised, a noted side effect of integrin IIb/IIIa antagonists (16). Because PEG is not immunogenic, molecules such as FNPEG-5K should not be expected to generate antibody responses, even after repeated administrations. Thus, a multifunctional adhesion protein can be engineered through artificial posttranslational modifications into a targeted masking molecule. Moreover, this provides an example of how PEG-mediated steric blocking can be selectively targeted to binding sites for specific ligands of therapeutic relevance. The next step in characterizing FNPEG-5K will involve testing whether this bioconjugate will block platelet binding in an animal-based vasculature intervention model. In a broader sense, this technique represents a proteinbased alternative to antisense oligonucleotides that have been envisioned as potential vessel wall therapeutics (17). By selectively binding and coating targeted molecules, one can neutralize undesired intermolecular interactions. Similar chimeras have the potential to be useful tools in cellular engineering and protein-based therapies (18). ACKNOWLEDGMENT

We thank Dr. M. Pendrak for preparing iodinated FN and Ms. J. Cashel for preparing purified FN. We also thank Dr. William Pritchard for his helpful advice. LITERATURE CITED (1) Motwani, J. G., and E. J. Topol. (1998) Aortocoronary saphenous vein graft disease: pathogenesis, predisposition, and prevention. Circulation 97, 916-931.

Geho et al. (2) Sixma, J. J., G. H. van Zanten, E. U. Saelman, M. Verkleij, H. Lankhof, H. K. Nieuwenhuis, and P. G. de Groot. (1995) Platelet adhesion to collagen. Thromb. Haemost. 74, 454-459. (3) Houdijk, W. P., and J. J. Sixma. (1985) Fibronectin in artery subendothelium is important for platelet adhesion. Blood 65, 598-604. (4) Yamada, K. M. (1989) Fibronectin domains and receptors. Fibronectin (D. F. Mosher, Ed.) pp 47-121, Academic Press, San Diego. (5) Miekka, S. I., Ingham, K. C., and Menache, D. (1982) Rapid methods for isolation of human plasma fibronectin. Thromb. Res. 27, 1-14. (6) Abuchowski, A., van Es, T., Palczuk, N. C., and Davis, F. F. (1977) Alteration of immunological properties of bovine serum albumin by covalent attachment of poly(ethylene glycol). J. Biol. Chem. 252, 3578-3581. (7) Jentoft, N., and Dearborn, D. G. (1979) Labeling of proteins by reductive methylation using sodium cyanoborohydride. J. Biol. Chem. 254, 4359-4365. (8) Chamow, S. M., Kogan, T. P., Venuti, M., Gadek, T., Harris, R. J., Peers, D. H., Mordenti, J., Shak, S., and Ashkenazi, A. (1994) Modification of CD4 immunoadhesin with monomethoxypoly(ethylene glycol) aldehyde via reductive alkylation. Bioconjugate Chem. 5, 133-140. (9) Caliceti, P., Schiavon, O., Sartore, L., Monfardini, C., and Veronese, F. (1993) Active site protection of proteolytic enzymes by poly(ethylene glycol) surface modification. J. Bioact. Compat. Polym. 8, 41-50. (10) Stocks, S. J., Jones, A. J., Ramey, C. W., and Brooks, D. E. (1986) A fluorometric assay of the degree of modification of protein primary amines with poly(ethylene glycol). Anal. Biochem. 154, 232-234. (11) McGregor, J., Gayet, O., Mercier, N., McGregor, L., and Chignier, E. (1996) Identification, isolation, and characterization of platelet glycoproteins mediating platelet adhesion. Platelets: a practical approach (S. Watson, and K. Authi, Eds.) pp 132-133, Oxford University Press, New York. (12) Baker, G. R., Sullam, P. M., and Levin, J. (1997) A simple, fluorescent method to internally label platelets suitable for physiological measurements. Am. J. Hematol. 56, 17-25. (13) Kauf, A. C., Hough, S. M., and Bowditch, R. D. (2001) Recognition of fibronectin by the platelet integrin alpha IIb beta 3 involves an extended interface with multiple electrostatic interactions. Biochemistry 40, 9159-9166. (14) Rhodes, R. K. Blood Vessels. (1982) Collagen in Health and Disease ( J. B. Weiss, and M. I. V. Jayson, Eds.) pp 376-387, Churchill Livingston, New York. (15) Deible, C. R., Beckman, E. J., Russell, A. J., and Wagner, W. R. (1998) Creating molecular barriers to acute platelet deposition on damaged arteries with reactive poly(ethylene glycol). J. Biomed. Mater. Res. 41, 251-256. (16) Waters, J. H., Anthony, D. G., Gottlieb, A., and Sprung, J. (2001) Bleeding in a patient receiving platelet aggregation inhibitors. Anesth. Analg. 93, 878-882. (17) Simons, M., Edelman, E. R., DeKeyser, J. L., Langer, R., and Rosenberg, R. D. (1992) Antisense c-myb oligonucleotides inhibit intimal arterial smooth muscle cell accumulation in vivo. Nature 359, 67-70. (18) Tykocinski, M. L., Kaplan, D. R., and Medof, M. E. (1996) Antigen-presenting cell engineering. The molecular toolbox. Am. J. Pathol. 148, 1-16.

BC034037V