Stabilized Hemocompatible Coating of Nitinol Devices Based on

Jul 31, 2007 - Stabilized Hemocompatible Coating of Nitinol Devices Based on. Photo-Cross-Linked Alginate/Heparin Multilayer. Meng Liu,† Xiuli Yue,*...
29 downloads 0 Views 361KB Size
Download PDF
9378

Langmuir 2007, 23, 9378-9385

Stabilized Hemocompatible Coating of Nitinol Devices Based on Photo-Cross-Linked Alginate/Heparin Multilayer Meng Liu,† Xiuli Yue,*,‡ Zhifei Dai,†,‡ Lei Xing,† Fang Ma,‡ and Nanqi Ren‡ Nanobiotechnology and Biosensor Lab, Bio-X Center, Harbin Institute of Technology, Harbin 150001, China, and State Key Laboratory of Urban Water Resources and EnVironment (SKLUWRE), Harbin Institute of Technology, Harbin 150090, China ReceiVed February 2, 2007. In Final Form: May 21, 2007 A novel stabilized hemocompatible multicomponent coating was engineered by consecutive alternating adsorption of two polysaccharides, alginate (Alg) and heparin (Hep), onto a Nitinol surface via electrostatic interaction in combination with photoreaction in situ. For this purpose, a photosensitive cross-linker, p-diazonium diphenyl amine polymer (PA), was used as an interlayer between alginate and heparin. The optical intensity of UV/vis spectra increased linearly with the number of layers, indicating the buildup of a multilayer structure and uniform coating. Photo-cross-linking resulted in higher stability without compromising its catalytic capacity to promote antithrombin III (ATIII)-mediated thrombin inactivation. Chromogenic assays for heparin activity proved definitively that anticoagulation activity really comes from surface-bound heparin in multilayer film, not from solution-phase free heparin that has leaked from multilayer film. The activated partial thromboplastin time (aPTT) assay showed that both (PA/Hep)8- and (PA/Alg/PA/Hep)4coated Nitinol were less thrombogenic than the uncoated one. Yet, the latter was found to be more stable under a continuous shaken wash. In addition, (PA/Alg/PA/Hep)4 film exhibited lower suface roughness and higher hydrophilicity than (PA/Hep)8. As a result, hemolysis of (PA/Alg/PA/Hep)4 (0.34 ( 0.064%) was lower than (PA/Hep)8 (0.52 ( 0.241%). The naked Nitinol and (PA/Hep)8-coated Nitinol showed relatively strong platelet adhesion. On the contrary, no sign of any cellular matter was seen on the (PA/Alg/PA/Hep)4 surface. It is believed that the phenomenon of interlayer diffusion resulted in blended structures, hence, the enhanced wettability and antifouling properties after the incorporation of alginate layers. It is likely that the cooperative effect of alginate and heparin led to the excellent blood compatibility of the (PA/Alg/PA/Hep)4 coating. To simplify, there is greater advantage in utilizing cross-linked alginate/heparin surfaces rather than merely the heparin surface for improving blood- and tissue-compatible devices.

Introduction Thrombogenicity and biocompatibility of stents are important factors that influence restenosis and long-term outcome of intravascular stent application. Shape-memory alloys on the basis of nickel-titanium alloys (chemical formula NiTi: Nitinols) possess interesting properties that have stimulated great interest in clinical settings (such as intravascular stents), namely, the shape-memory effect and the superelasticity.1-4 However, the high nickel content of the material may result in acute nickel toxicity5 and allergic reaction.6 In addition, the positive net electrical surface potential7,8 and high free surface energy of metals9 may cause thrombogenicity. To overcome those limitations, some efforts have been made to coat metallic stents with polymers in order to improve their blood compatibility, e.g., * To whom correspondence should be addressed. (X.Y.) State Key Laboratory of Urban Water Resources and Environment (SKLUWRE), Harbin Institute of Technology, Harbin 150090, China. Tel/Fax: 86 451 8628 3008; E-mail: [email protected]. † Nanobiotechnology and Biosensor Lab. ‡ State Key Laboratory of Urban Water Resources and Environment (SKLUWRE). (1) Gisser, K. R. C.; Geselbracht, M. J.; Cappellari, A.; Hunsberger, L.; Ellis, A. B.; Perepezko, J.; Lisensky, G. C. J. Chem. Educ. 1994, 71, 334. (2) Barras, C. D. J.; Myers, K. A. Eur. J. Vasc. EndoVasc. Surg. 2000, 19, 564. (3) Harrison, J. D.; Hodgson, D. E. Using Shape Memory Alloys. Engineering Kit Instruction Manual; Perkins, J., Ed.; Piscataway, NJ, 1988. (4) Michael, B. B. Encyclopedia of Materials: Science and Engineering; Pergamon Press: Oxford, 1986; Vol. 6. (5) Uo, M.; Watari, F.; Yokoyama, A.; Matsuno, H.; Kawasaki, T. Biomaterials 1999, 20, 747. (6) Budinger, L.; Hertl, M. Allergy 2000, 55, 108. (7) DePalma, V. A.; Baier, R. E.; Ford, J. W.; Glott, V. L.; Furuse, A. J. Biomed. Mater. Res. 1972, 6, 37. (8) Ribeiro, P. A.; Gallo, R.; Antonius, J.; Mimish, L.; Sriram, R.; Bianchi, S.; Duran, C. G. Am. Heart J. 1993, 125, 501. (9) Palmaz, J. C. AJR Am. J. Roentgenol. 1993, 160, 613.

polyamide,10 polyurethane,11 or silicone.12 However, all of these coatings are solvent-based and, therefore, may not be applied with precise process control. In addition, the use of pure polymer coatings is also restricted by the limited blood compatibility of these polymer coatings.13,14 The blood compatibility may be further improved using polymer coatings as interfaces for the linkage of bioactive molecules that are able to interact with the surrounding cells in a physiological manner. Nevertheless, a previous step to provide chemically reactive functional groups is necessary. Many of the reactions causing acute clotting and restenosis are thrombin-mediated.15,16 Heparin is the most frequently used agent for thrombin regulation. In order to prevent these reactions, the immobilization of heparin on the surface of the biomaterial has been used to prevent thrombus formation and improve hemocompatibility.17 However, the actual performance of immobilized heparin may vary significantly depending on many factors, including the method of immobilization and the amount and type of heparin.18-21 A mere attachment of heparin to a (10) Yoshioka, T.; Wright, K. C.; Wallace, S.; Lawrence, D. D.; Gianturco, C. Jr. AJR Am. J. Roentgenol. 1988, 151, 673. (11) De Scheerder, I. K.; Wilczek, K. L.; Verbeken, E. V.; Vandorpe, J.; Lan, P. N.; Schacht, E.; De Geest, H.; Piessens, J. Atherosclerosis 1995, 114, 105. (12) Palmaz, J. C.; Garcia, O. J.; Schatz, R. A.; Rees, C. R.; Roeren, T.; Richter, G. M.; Noeldge, G.; Gardiner, G. A.; Becker, G. J.; Walker, C., Jr. et al. Radiology 1990, 174, 969. (13) Verweire, I.; Schacht, E.; Qiang, B. P.; Wang, K.; De Scheerder, I. J. Mater. Sci.: Mater Med. 2000, 11, 207. (14) Rechavia, E.; Litvack, F.; Fishbien, M. C.; Nakamura, M.; Eigler, N. Cathet. CardioVasc. Diagn. 1998, 45, 202. (15) Agnelli, G. CardioVasc. Res. 1996, 31, 232. (16) Seiler, S. M. Semin. Thromb. Hemostasis 1996, 22, 223. (17) Kim, S. W.; Jacobs, H. Blood Purification 1996, 14, 357. (18) Wendel, H. P.; Ziemer, G. European Journal of Cardio-Thoracic Surgery 1999, 16, 342.

10.1021/la7002996 CCC: $37.00 © 2007 American Chemical Society Published on Web 07/31/2007

Stabilized Hemocampatible Coating of Nitinol DeVices

surface may not be optimal if the underlying surface has a significant contact activating capacity. The anticoagulant activity of covalently attached heparin using conventional techniques is often decreased by a chemical activation of heparin or by an excessively tight attachment of heparin molecules to the surface. Instead, the primary noncompatible surface should first be lined with a more compatible and compact layer capable of effectively shielding direct interactions of blood with the primary surface. Previous work by other investigators has shown that increasing the distance between heparin and the surface increased biological activity.22 Nevertheless, monomolecular layers may lose their performance faster due to biodegradation in the blood environment. The layer-by-layer self-assembly of polycations and polyanions into multilayers has emerged as an efficient, versatile, yet simple technique to create biologically active surfaces.23-27 Using this technique, various medical devices were coated with ionically attached assemblies consisting of an eligible number of molecular layers of heparin and cationically charged polyelectrolytes.28-32 However, release of heparin from the surface over time into blood eventually occurs, exposing the cationic polymer surface to the blood, which may cause massive platelet adhesion and aggregation resulting from electrostatic interactions with the exposed cationic charged surface. To overcome this limitation, the whole albuimn-heparin multilayer-assembled coatings were covalently cross-linked with glutaraldehyde to form a stable albumin/heparin coating of the original surface.28-30 Like glycosaminoglycans, alginate is a seaweed-derived acidic polysaccharide composed of β-D-mannuronic acid and R-Lguluronic acid. Alginate gels are known to be biocompatible, degradable, and nontoxic. An additional advantage using alginate in humans could be its nonmammalian source, which eliminates the risk of contamination with pathogenic agents.33 It was reported that the alginate gel sheet was useful as a scaffold for regenerating dermis,34,35 nerve,36-38 and bone.39 Suzuki et al. devised a novel (19) Larm O. L.; Olsson P. Surface-immobilized heparin. Edward Arnold: London, 1989; p 597. (20) Andersson, J.; Sanchez, J.; Ekdahl, K. N.; Elgue, G.; Nilsson, B.; Larsson, R. J. Biomed. Mater. Res., Part A 2003, 67A, 458. (21) Gupta, V.; Aravamuthan, B. R.; Baskerville, S.; Smith, S. K.; Gupta, V.; Lauer, M. A.; Fischell, T. A. J. InVasiVe Cardiol. 2004, 16, 304. (22) Park, K. D.; Okano, T.; Nojiri, C.; Kim, S. W. J. Biomed. Mater. Res. 1988, 22, 977. (23) Donath, E.; Sukhorukov, G. B.; Caruso, F.; Davis, S. A.; Mohwald, H. Angew. Chem., Int. Ed. 1998, 37, 2202. (24) Decher, G. Science 1997, 277, 1232. (25) Dai, Z. F.; Heilig, A.; Zastrow, H.; Donath, E.; Mohwald, H. Chem.sEur. J. 2004, 10, 6369. (26) Dai, Z. F.; Dahne, L.; Mohwald, H.; Tiersch, B. Angew. Chem., Int. Ed. 2002, 41, 4019. (27) Dai, Z. F.; Voigt, A.; Leporatti, S.; Donath, E.; Dahne, L.; Mohwald, H. AdV. Mater. 2001, 13, 1339. (28) Brynda, E.; Houska, M.; Jirouskova, M.; Dyr, J. E. J. Biomed. Mater. Res. 2000, 51, 249. (29) Houska, M.; Brynda, E. J. Colloid Interface Sci. 1997, 188, 243. (30) Brynda, E.; Houska, M. J. Colloid Interface Sci. 1996, 183, 18. (31) Tan, Q. G.; Ji, J.; Barbosa, M. A.; Fonseca, C.; Shen, J. C. Biomaterials 2003, 24, 4699. (32) Thierry, B.; Winnik, F. M.; Merhi, Y.; Silver, J.; Tabrizian, M. Biomacromolecules 2003, 4, 1564. (33) Mourao, P. A. S.; Guimaraes, M. A. M.; Mulloy, B.; Thomas, S.; Gray, E. Br. J. Haematol. 1998, 101, 647. (34) Despang, F.; Borner, A.; Dittrich, R.; Tomandl, G.; Pompe, W.; Gelinsky, M. Materialwissenschaft Und Werkstofftechnik 2005, 36, 761. (35) Suzuki, Y.; Tanihara, M.; Nishimura, Y.; Suzuki, K.; Yamawaki, Y.; Kudo, H.; Kakimaru, Y.; Shimizu, Y. J. Biomed. Mater. Res. 1999, 48, 522. (36) Suzuki, K.; Suzuki, Y.; Tanihara, M.; Ohnishi, K.; Hashimoto, T.; Endo, K.; Nishimura, Y. J. Biomed. Mater. Res. 1999, 49, 528. (37) Suzuki, Y.; Tanihara, M.; Ohnishi, K.; Suzuki, K.; Endo, K.; Nishimura, Y. Neurosci. Lett. 1999, 259, 75. (38) Suzuki, K.; Suzuki, Y.; Ohnishi, K.; Endo, K.; Tanihara, M.; Nishimura, Y. Neuroreport 1999, 10, 2891. (39) Suzuki, Y.; Tanihara, M.; Suzuki, K.; Saitou, A.; Sufan, W.; Nishimura, Y. J. Biomed. Mater. Res. 2000, 50, 405.

Langmuir, Vol. 23, No. 18, 2007 9379

covalently cross-linked alginate/heparin gel useful for the local delivery of heparin-binding growth factors (HBGFs) in the treatment of ischemic arterial diseases, as well as for regenerating or constructing tissues.40 It was demonstrated by Lee et al. that the use of a biodegradable alginate as a biological sealant was a feasible approach to prepare impervious textile arterial prostheses.41 The general idea followed in this work was the preparation of a compatible interface between a material surface and blood using a defined multilayer assembly of alginate and heparin in order to maximize antithrombin activity. Both alginate and heparin are negatively charged polysaccharides, so a positively charged photosensitive cross-linker, p-diazonium diphenyl amine polymer (PA), is used as the interlayer between alginate and heparin. The procedure requires two steps: “alternative ionic adsorption” of PA and polysaccharides onto the Nitinol surface; and “crosslinking” via photoreaction in situ, which converts the ionic bonding between the neighboring layers to a covalent bonding for cross-linking of the whole alginate/heparin multilayer system, as shown in Scheme 1. These interfaces take advantages: (1) Multilayers of PA and polysaccharides are deposited accurately with respect to coating thickness and precisely follow the shape of the basic device. (2) Immobilization of negatively charged polysaccharides could make surfaces more biocompatible because of the electrostatic repulsion of negatively charged blood components. (3) The use of an alginate as a biological basecoat or interlayer for heparin immobilization instead of proteins such as albumin, collagen, and gelatin may avoid the risk of contamination with pathogenic agents. (4) In contrast to proteincoated endovascular stents, it offers easy handling, storage, and sterilization procedures. Experimental Methods Materials. Heparin (sodium salt, porcine intestinal mucosa, MW 13 500-15 000, 1 mg ) 174 U) was obtained from CalBio Chem. Sodium alginate (also called alginic acid sodium salt; medium viscosity) was purchase from Sigma. The p-diazonium diphenyl amine polymer was synthesized according to the method reported in the literature.42 Human thrombin and ATIII were obtained from Haematologic Technologies, Inc. Chromogenic substrate S-2238 was purchased from American Diagnostica. Nitinol sheets were cut into rectangular pieces (typical dimensions: 10 × 10 × 0.3 mm3) and cleaned with acetone (10 min), ethanol (10 min), and distilled water (rinsing). The sheets were then boiled for 60 min in 30% aqueous H2O2 and rinsed again with distilled water. Etching was performed with 4 M aqueous KOH for 30 min at 120 °C, followed by multiple rinsing with distilled water. The surface of quartz slides were first made hydrophilic by immersion in a hot piranha solution (H2SO4/H2O2, 7:3) for 1 h and then in an H2O/H2O2/NH3 (5:1:1) solution at 60 °C for 30 min. Then, the slides were carefully rinsed in deionized water. Multilayer Construction by Electrostatic LbL Self-Assembly Technique. Multilayer thin films were deposited on the quartz slides and Nitinol sheets using a dip technique from oppositely charged polyelectrolyte solution at room temperature. First, the substrates were immersed in an aqueous PA solution (2 mg/mL) for 5 min, followed by three consecutive washings in deionized water. Subsequently, the substrates were dipped into an aqueous polysaccharide solution (alginate or heparin) (2 mg/mL, 0.5M NaCl) for 5 min, followed by three consecutive washings. After completion of the desired number of dip cycles, the substrates were removed, rinsed (40) Chinen, N.; Tanihara, M.; Nakagawa, M.; Shinozaki, K.; Yamamoto, E.; Mizushima, Y.; Suzuki, Y. J. Biomed. Mater. Res., Part A 2003, 67A, 61. (41) Lee, J. H.; Kim, W. G.; Kim, S. S.; Lee, J. H.; Lee, H. B. J. Biomed. Mater. Res. 1997, 36, 200. (42) Zhao, C.; Chen, J. Y.; Cao, W. X. Angew. Makromol. Chem. 1998, 259, 77.

9380 Langmuir, Vol. 23, No. 18, 2007

Dai et al.

Scheme 1. Photoreaction among PA and Alginate and Heparin under UV Irradiation

with Millipore water, and air-dried. The deposition process of the films was conducted in the dark to avoid the decomposition of the PA. The multilayer growth was monitored by UV-visible absorption spectra executed on a Varian Cary 4000 spectrophotometer. Photoreaction Induced by Visible Light. Irradiation was performed using UV lamps (365 nm). The multilayer thin films on the substrates were irradiated for a given time under ambient conditions from above at a distance of approximately 3 cm (relative intensity ∼2 mW/cm2). Static Water Contact Angles. Contact angles on the Nitinol sheet samples were detected by employing drops of pure deionized water using the contact angle and surface tension meter (KAM101, KSV). Measurements are reported as average advancing/receding degree ( standard deviation of ten data points. For each sample, at least five measurements on different surface sites were averaged. Infrared Spectroscopy. Spectra of the coated Nitinol sheet samples were acquired using a Varian resolution Fourier transform infrared spectrometer (Varian FTS 3100) equipped with a wideband MCT detector, collected with 512 background scans and 4 cm-1 resolution. Variable angle specular reflectance mode (VASRM) spectra were acquired with the 86° incidence angle. The singlebeam spectrum of the VASRM accessory was used as a background. Scanning was conducted in the mid-IR range from 400 to 4000 cm-1. Spectra manipulations performed on the data, such as baseline correction, and CO2 peak removal (from 2250 to 2405 cm-1) were performed using the Varian Resolution Pro software package. Infrared band assignments were obtained from reference values previously reported in the literature.43,44 Shaken Wash Experiments. To study the binding force of heparin onto the substrate, a shaken wash model was designed to allow for exposure of the heparinized substrate to fluid-flow stress. The heparinized quartz plates were secured on a customized holder and subjected to a tangential shaken wash at 74 to 80 rpm (selected to mimic resting heart rate) in 20 mM phosphate buffer saline (PBS; pH 7.4) for a given time at 37 °C. Measurements of Bioactivity of Heparin in Vitro. The activated partial thrombin time (aPTT) was determined using a semiautomatic blood coagulation analyzer (Steellex LG-PAPER-I, China). The grafts were cut into sheets (1.0 × 1.0 cm2) and were placed carefully in vials containing 1 mL of coagulation control normal plasma (Pacific (43) Hou, X. L.; Wu, L. X.; Sun, L.; Zhang, H.; Yang, B.; Shen, J. C. Polym. Bull. 2002, 47, 445. (44) Colthup, N. B.; Daly, L. H.; Wiberley, S. E, Introduction to Infrared and Raman Spectroscopy, 3rd ed.; Academic Press: New York, 1990.

Hemostasis) with shaking at 37 °C. 0.1 mL plasma was taken out of the vial for subsequent measurement at different interval (0 min, 0.5 min, 1.5 min, 5 min, etc.). For aPTT assay, the test method and dosages of the agents were performed according to the manuals. 0.1 mL of tested plasma was mixed well with 0.1 mL of KONTACT reagent (1.2% rabbit brain phospholipids, 0.03% magnesium aluminum silica, 0.4% phenol, 0.8% buffers, salt, and stabilizers) at 37 °C for 5 min. A stopwatch was started simultaneously with the addition of 0.025 M CaCl2, and the solution was mixed well. The time of clot formation was recorded with a chronometer. Chromogenic Assays for Heparin Activity. The heparinized sample (Nitinol sheet: 1.0 × 1.0 cm2) was incubated in Hepes buffer (20 mM Hepes, 190 mM NaCl, 0.5 mg/mL BSA, 0.02% NaN3) at 37 °C with 10 nM AT III for 5 min, followed by adding 10 nM human thrombin. At timed intervals, aliquots are removed and added to Tris buffer (50 mM Tris, 175 mM NaCl, 0.5 mg/mL BSA) containing 20 mM EDTA, followed by adding chromogenic substrate S-2238 (0.2 mM final concentration) for thrombin detection at 405 nm by a Varian Cary 4000 spectrophotometer. Percent Hemolysis. Hemolysis assay experiments were performed on Nitinol sheets. In brief, three sample sheets (1.0 × 1.0 cm2) were equilibrated in normal sodium chloride saline water (0.9% w/v NaCl) for 24 h. Then, 10 mL of 0.9% saline was added to each sample for 30 min at 37 °C. Afterward, 0.2 mL dilute blood (anticoagulated blood was diluted with normal saline water according to volume ratio 1/1.25) was added, and the samples were incubated in different test tubes for 60 min at 37 °C. Positive and negative controls were obtained by adding 0.2 mL dilute blood to 10 mL saline solution and bidistilled water, respectively. All test tubes were incubated at 37 °C for 60 min. After incubation, the tubes were centrifuged at 2000 rpm for 5 min. The percentage hemolysis was calculated by measuring the optical density of the supernatant solution at 545 nm in a UV-visible spectrophotometer (Cary 4000, Varian) as per the following formula: Hemolysis (%) ) (X1 - X3)/(X2 - X3) × 100% X1, X2, and X3 were absorbance of the sample, positive control, and negative control. Platelet Adhesion in Vitro. The assay of platelet adhesion could be used to examine the activation of platelets, fibrin clots, and so forth. The Nitinol sheet samples were allowed to contact with 1 mL of fresh platelet-enriched plasma for 3 h at 37 °C under the static conditions. After washing with PBS, the sheets were fixed using 2%

Stabilized Hemocampatible Coating of Nitinol DeVices

Langmuir, Vol. 23, No. 18, 2007 9381

Figure 1. Absorption spectra of (PA/polysaccharide)n multilayer films: n is 0, 1, 2, 3, 4, and 5 from bottom to top on quartz slides. Polysaccharide with odd numbers is alginate, whereas even numbers represent heparin. gluteraldehyde solution for 30 min, then washed with PBS again, then immersed into 55%, 70%, 80%, 90%, 95%, and 100% ethanol solution in sequence and finally dried in a desiccator. The sheets were then gold-sputtered before being imaged by SEM (X-650 Scanning Electron Micro Analyzer).

Results and Discussion Fabrication PA/Polysaccharide Multilayer via Electrostatic Layer-by-Layer Self-Assembly. In general, the fabrication procedure is based on the adsorption of subsequent alternating layers caused by the electrostatic attraction between the opposite charges on the top of the previously adsorbed layer and by molecules adsorbing from solution. PA bears a net positive charge, while both alginate and heparin are negatively charged polysaccharides due to the presence of ionized sulfate and carboxylate groups. Thus, the first PA monolayer is adsorbed on a primary surface that serves in the next steps as an anchor for the alternating, charge-driven adsorption of polysaccharides. In the next step, polysaccharide is adsorbed on the first PA layer. The interface with the solution becomes overcharged so that onto the first polysaccharide layer the second layer of PA can be adsorbed, and the sequence of these steps can be repeated until the required number of layers is deposited. The gradual increases are seen in the intensity of UV/vis spectra of PA/polysaccharide multilayer films assembled on a quartz slide as a function of deposition cycles (Figure 1). The peak at 376 nm is associated with the contribution of π-π* transition of the diazonium group of PA and increases linearly with the number of layers, indicating the successful multilayer growth and uniform coating. The pronounced optical absorbance of PA (bright yellow films) provides additional insight into the film growth and degree of PA incorporation. The contact angle measurements provide additional evidence for the alternative layer-by-layer deposition of PA and polysaccharides onto the Nitinol surface (Figure 2). The pristine Nitinol sheet gave an advancing contact angle of 57.56 ( 2.24°. To eliminate surface contaminants and to produce a consistent and reproducible titanium oxide surface layer for binding, the Nitinol sheets were etched using 4 M aqueous KOH for 30 min at 120 °C. After etching, the advancing contact angle decreased to 34.95 ( 3.89°, indicating a noticeable increase in hydrophilicity due to the formation of metal oxides on the Nitinol surface. An intermediary PA molecule was then deposited to the oxide layer, followed by the binding of either alginate or heparin to the PA

Figure 2. Contact angle of PA/polysacaride multilayer films with different number of layers on Nitinol sheet: (a) Alg/PA/Hep multilayer films with alginate layer and (b) Hep/PA/Hep without alginate layer. Odd numbers represent films with PA as the outermost layer, whereas even-numbered films have polysaccharide (alginate or heparin) as the outermost layer. Measurements are reported as the average value of advancing or receding contact angles of at least five data points.

layer via electrostatic interaction. For comparative study, two systems, Alg/PA/Hep multilayer films with alginate layer and Hep/PA/Hep multilayer films without alginate layer, were fabricated. The advancing contact angles of films with PA as the outmost layer were determined to be in the range from 32° to 55°, while films with polysaccharides as the outmost layer give the advancing contact angles from 19° to 40°. After the first two bilayers, the contact angles of the film fluctuated periodically following the PA and polysaccharide deposition cycles. Such behavior has often been observed, with a few polyelectrolyte layers required prior to regular multilayer growth for films assembled by the layer-by-layer technique.45,46 The actual number of layers assembled prior to regular growth depends on the nature of the surface, the polyelectrolytes, and the adsorption conditions. The obvious alternate change of contact angle data verified the progressive buildup of the film by alternate deposition of the polyelectrolytes. An apparent increase of advancing and receding contact angles for both Alg/PA/Hep and Hep/PA/Hep systems (45) Lvov, Y.; Decher, G.; Mo¨hwald, H. Langmuir. 1993, 9, 5. (46) Decher, G.; Schimitt, J. Prog. Colloid Polym. Sci. 1992, 89, 160.

9382 Langmuir, Vol. 23, No. 18, 2007

Figure 3. Absorption spectra of multilayers of (PA/Alg/PA/Hep)4 on quartz slide under irradiation at different times: 0, 5, 15, 30, 45, 60, 75, 90, and 150 s.

after UV irradiation, indicating an increase in hydrophobicity, can be measured. In comparison with system Hep/PA/Hep, system Alg/PA/Hep showed lower contact angles and smaller hysteresis, suggesting an increase in wettability and a decrease in surface roughness. The contact angle hysteresis is simply calculated by substracting the measured advancing (maximum) contact angle with the measured receding (minimum) contact angle, mainly due to surface roughness of the solid substrate. The addition of an alginate layer into the films resulted in the decrease in solid surface roughness, which reduced advancing contact angle and enlarged receding contact angle. Hence, decreased hysteresis was observed for the whole Alg/PA/Hep system. Covalent Attachment of Multilayer Films via Photoreaction. The (PA/Alg/PA/Hep)4 multilayer film was exposed under light for a given time. A dramatic change was shown in UV-vis spectra against irradiation time (Figure 3). As irradiation time increased, the absorbance at 376 nm decreased. Simultaneously, a new peak appeared at 290 nm. These changes tended to be limited within 3 min. We considered here that PA covalently attached to polysaccharides through the photoreaction of phenyldiazonium in PA with sulfate in heparin or carboxylate groups in heparin and alginate due to the decomposition of the diazonium group. Under irradiation of UV light, PA is converted into its phenyl cationic form after releasing N2; then, an SN 1 type of nuclear displacement by sulfate and carboxylate groups occurs, as shown in Scheme 1. Ti hydrogel was formed on the Nitinol surfaces which had been etched using 4 M aqueous KOH.47-49 PA was adsorbed onto the surface mainly via hydrogen bonding interaction between the hydroxyl groups of the Nitinol substrate and diazonium of PA.50,51 Under UV irradiation, the hydrogen bonds converted to covalent bonds following the decomposition of the diazonium group.50,51 As a result, the whole system is covalently fixed via photoreaction between polysaccharides and PA under UV irradiation, and heparin is retained in the alginate network by covalent bond. Actually, there was a possibility that the diazonium moiety reacted with other sites of heparin and (47) Yan, W. Q.; Nakamura, T.; Kawanabe, K.; Nishigochi, S.; Oka, M.; Kokubo, T. Biomaterials 1997, 18, 1185. (48) Kujala, S.; Ryhanen, J.; Danilov, A.; Tuukkanen, J. Biomaterials 2003, 24, 4691. (49) Prusi, A.; Arsov, L. J. Solid State Electrochem. 2007, 11, 355. (50) Yang, Z. H.; Cao, W. X. J. Polym. Sci., Part A: Polym. Chem. 2003, 41, 3103. (51) Cao, T. B.; Yang, S. M.; Yang, Y. L.; Huang, C. H.; Cao, W. X. Langmuir 2001, 17, 6034.

Dai et al.

Figure 4. FTIR spectra of multilayer film of (PA/Alg/PA/Hep)4 on Nitinol sheet before and after UV irradiation.

alginate such as OH. But in our case, it should mainly react with sulfate and COO- moieties. The reason was that, on one hand, the multilayer film was constructed by employing electrostatic interaction as driving force, which associated cationic phenyldiazonium moiety and anionic moieties between adjacent layers more tightly. FT-IR was employed to further confirm the change taking place in the film before and after irradiation. FTIR spectra of a multilayer film of (PA/Alg/PA/Hep)4 were shown on Nitinol sheet (Figure 4). For the specimen before irradiation, two quite strong vibration bands at 2227 and 2173 cm-1 could be assigned to the asymmetric and symmetric stretching modes of diazonium, respectively. The appearance of the two distinct absorption bands at 1586 cm-1 and 1119 cm-1, which were attributed to the stretching mode of the phenyl group containing a substituted diazonium group in PA and the NfO stretching mode between diazonium and sulfate/carboxylate groups due to the strong ion pair effect (-NtN+ f SO3- and -NtN+ f COO-),44 respectively. However, after UV irradiation, the former was separated into two bands at 1606 and 1645 cm-1 and the latter disappeared completely, indicating the decomposition of the diazonium groups in the film. The symmetric stretching vibration of the highly polar -SO3- group is observed at 1033 cm-1 before and after UV irradiation. Thus, we could conclude that heparin and alginate were photopolymerized to PA in solid multilayer film, leading to a three-dimensional cross-linking network. Stability of Multilayer Films. We compared the stability of the multilayer films before and after UV irradiation by a vigorous tangential shaken wash of the corresponding films designed to approximate resting heart rate and circulating blood volume. Through UV-vis absorption spectra, we found that, before irradiation treatment, the absorbance at 376 nm for the noncross-linked multilayer coating decreased much faster than that at 290 nm for the photo-cross-linked multilayer coating after irradiation treatment during a continuous shaken wash (Figure 5). After a 4 h shaken wash, the PA content dropped to 60% and 80% of the preshaken level for non-cross-linked (PA/Hep)8 and (PA/Alg/PA/Hep)4, respectively. However, no apparent drop of PA content was seen for cross-linked (PA/Hep)8 and (PA/Alg/ PA/Hep)4 films. With additional exposure to shaken wash, significantly more PA was lost continuously with ultimately only 25% and 13% retained PA after a 48 h shaken wash for noncross-linked (PA/Hep)8 and (PA/Alg/PA/Hep)4, respectively.

Stabilized Hemocampatible Coating of Nitinol DeVices

Figure 5. Effect of shaken wash on the (PA/Hep)8 and (PA/Alg/ PA/Hep)4 multilayer films before and after UV irradiation treatment. The absorbance was obtained at 376 and 290 nm for the cross-linked and non-cross-linked multilayer films, respectively, through UVvis absorption spectra.

Nevertheless, after irradiation treatment, the reducing percent based on the total PA was estimated to be less than 40% and 15% for (PA/Hep)8 and (PA/Alg/PA/Hep)4 after a 48 h shaken wash, respectively. It is obvious that covalently cross-linking the whole system significantly improved immobilization of heparin onto the Nitinol surface quantitatively and durably. It is easily understood that both (PA/Hep)8 and (PA/Alg/PA/Hep)4 films became more stable after photo-cross-linking. However, interestingly, (PA/Alg/PA/Hep)4 film showed higher stability than (PA/ Hep)8 film before and after irradiation treatment due to the addition of alginate layers into the film. Since Nitinol is widely used as a shape-memory alloy, it is required to undergo a shape change during use. The experiments showed that the (PA/Alg/PA/Hep)4 multilayer is able to withstand the strong bending that is characteristic for a superelastic shapememory alloy. This is probably due to the high elasticity of the polymeric coating that is able to accommodate extension and compression. There is no difficulty for the coating to withstand a thermal shock and also the dilatation encountered during the martensite-austenite phase transformation and vice versa. Figure 6 showed (PA/Alg/PA/Hep)4 multilayer-coated Nitinol band that was cooled to 77 K with liquid nitrogen and left to warm up again to room temperature (during about 120 s). Neither a rupture of the coating nor a loss of adhesive strength was observed. The results indicated that the alloys could maintain good shape though layer-by-layer coating. Evaluation of Bioactivity of Heparin in Vitro. The aPTT is a measure of the activity of the intrinsic pathway of coagulation (VIII, IX, XI, XII). In recent times, it has been applied in the evaluation of in Vitro antithrombogenicity of biomaterials. The normal reference range of aPTT for a healthy blood plasma was regarded as 28-40 s. The poor platelet plasma which was in contact with the Nitinol sheets was evaluated by aPTT assay. The aPTT values were 98.1 and 90.8 s for (PA/Hep)8 and (PA/ Alg/PA/Hep)4-coated Nitinol sheets, respectively. Both of them were much longer than that of the pristine Nitinol sheet (39.5 s). These in Vitro results indicate that the bioactively coated stents are less thrombogenic than virgin metallic stents. The clotting time of incubated plasma in contact with pristine Nitinol and heparinized Nitinol was studied by varying the contact time. The results indicated that the aPTT values of both (PA/ Hep)8 and (PA/Alg/PA/Hep)4-coated Nitinol sheets increased

Langmuir, Vol. 23, No. 18, 2007 9383

with increasing contact time, while the uncoating Nitinol changed little (Figure 7). It was known that the immobilized heparin could catalyzed antithrombin III (AT-III), then the thrombin would bond to AT-III, thus preventing blood coagulation. Under shaking conditions, the thrombin/AT-III complex could dissociate from immobilized heparin, and the “recovered” heparin was available for subsequent binding of more AT-III and thrombin, resulting in the sustained increased clotting time (aPTT) in the plasma. The similar mechanism of the AT-III and thrombin binding to surface-immobilized heparin was previously elucidated under flow conditions by Byun et al.52 Heparin surface concentrations were determined to be 33.3 pmol/cm2 (0.5 µg/cm2) and 16.7 pmol/cm2 (0.25 µg/cm2) for (PA/Hep)8 and (PA/Alg/PA/Hep)4, respectively, according to the method reported by Park et al.53 Heparin surface concentration of (PA/Alg/PA/Hep)4 is 50% lower than that of (PA/Hep)8, but (PA/Alg/PA/Hep)4 exhibited the APTT, similar to (PA/Hep)8. These data define a mass transfer limited regime in which the heparin-potentiated thrombin inactivation is not enhanced by a further increase in the number of heparin layers which is in proportion to surface heparin density. It is proposed that the outmost heparin plays a key role for the activity of the intrinsic pathway of coagulation. Chromogenic Assays for Heparin Activity. To prove definitively that anticoagulation activity really comes from surface-bound heparin in the layer-by-layer film, the capacity of PA/Hep multilayer on the Nitinol surface to inactivate thrombin by antithrombin III (ATIII) was analyzed. As we know, heparinaccelerated thrombin inactivation is indeed a two-step mechanism, which includes the rapid formation of a heparin-ATIII inhibitory complex that subsequently reacts with free thrombin. Although it is anticipated that thrombin inactivation would be higher in the presence of physiologic concentrations of antithrombin (1.52.5 µM,54 the current system at concentrations of 20 nM ATIII) and 10 nM thrombin illustrates that PA/Hep multilayer on the Nitinol surface is catalytically active before and after photocross-linking. In the absence of surface heparin, approximately 4 nM of thrombin was inactivated by ATIII after 30 min, which represents 40% of initial thrombin concentration (10 nM). In stark contrast, in the presence of surface heparin, more thrombin was inactivated, as shown in Figure 8. The (PA/Hep)5 multilayer coating has a better effect on thrombin inactivation before and after photo-cross-linking, but the stability of heparin activity is much different. Before photo-cross-linking, heparin is released from the surface into solution in a short time; as a result, we can see that an apparent thrombin decay and time to reach a steadystate level of thrombin decay was 20 min, shorter than crosslinked (PA/Hep)5 (more than 30 min). After being stored in PBS at 37 °C for 10 days, the non-cross-linked (PA/Hep)5 system performed close to control sample in the absence of heparin, possibly due to the loss of heparin. Nevertheless, there was no significant change in thrombin inactivation for cross-linked (PA/ Hep)5 even after being stored in PBS for 10 days, indicating that photo-cross-linking resulted in higher stability of heparin activity. It provided a strong evidence that for the photo-cross-linked PA/Hep multilayer coating on Nitinol the observed anti-protease activities are not due to solution-phase free heparin that has leaked from the layer-by-layer deposited film. Hemolysis. No big difference in hemolysis was observed between uncoated and coated Nitinol bands, and all values were (52) Byun, Y.; Jacobs, H. A.; Kim, S. W. Biotechnol. Prog. 1996, 12, 217. (53) Park, K. D.; Piao, A. Z.; Jacobs, H.; Okano, T.; Kim, S. W. J. Polym. Sci., Part A: Polym. Chem. 1991, 29, 1725. (54) Murano, G.; Williams, L.; Miller-Andersson, M.; Aronson, D. L.; King, C. Thromb. Res. 1980, 18, 259.

9384 Langmuir, Vol. 23, No. 18, 2007

Dai et al.

Figure 6. Mechanical stability of (PA/Alg/PA/Hep)4 multilayer coating on Nitinol: shape-memory effect from 77 K (liquid nitrogen temperature) to room temperature (left to right).

Figure 7. The aPTT of human blood plasma incubated with pristine Nitinol and heparinized Nitinol sheets at different contact times.

Figure 8. Thrombin inactivation by antithrombin III in the presence of Nitinol sheets coated with multilayer of (PA/Hep)5 before and after photo-cross-linking. Results are presented for the (PA/Hep)5coated Nitinol sheets after being stored in PBS (pH 7.4) at 37 °C for different time periods: 0 and 10 days. Each data point represents a mean value of four measurements at each reaction time point.

much lower than the permissible hemolysis level of 5% (Figure 9). Acute cytotoxicity of the uncoated and coated Nitinol bands was excluded by negligible hemolysis. The hemolysis of (PA/ Hep)8 was determined to be 0.52 ( 0.241%, while this value decreased to 0.34 ( 0.064% for (PA/Alg/PA/Hep)4-coated Nitinol. An increased wettability as a result of the incorporation of negatively charged alginate into the multilayer coating of Nitinol may lead to decreased hemolysis due to a diminished or more reversible protein adsorption. Platelet Adhesion. The thromboresistance of an uncoated and coated metallic substrate was investigated using a platelet adhesion assay. The SEM analysis after the incubation in fresh plateletenriched plasma showed clear differences among the samples (Figure 10). The naked Nitinol surface displayed a slightly porous

Figure 9. Hemolysis of pristine Nitinol and heparinized Nitinol. Error bars indicate standard deviations (n ) 3).

topography caused by the etching of concentrated KOH solution. These surface features were completely masked by a multilayer coating consisting of 16 layers and replaced by nanosized clusters in the range of hundreds of nanometera for (PA/Hep)8 and tens of nanometers for (PA/Alg/PA/Hep)4. It may be an indication of the formation of complex coacervates, as reported previously in the case of the LbL self-assembly of poly(L-lysine) and HA.55 Note that both (PA/Hep)8 and (PA/Alg/PA/Hep)4 coatings have heparin as the outmost layer, but they exhibit different appearances. The much smaller clusters and higher uniformity were observed on the (PA/Alg/PA/Hep)4 film, which is in agreement with the contact angle measurements. The phenomenon of interlayer diffusion resulted in blended structures, hence, the enhanced wettability after the incorporation of alginate layers.56 (PA/Hep)8 and (PA/Alg/PA/Hep)4 coatings exhibited a similar anticoagulant effect, which was demonstrated by the above aPTT assay, but altered in the platelet function due to the incorporation of alginate. Both uncoated Nitinol and (PA/Hep)8-coated Nitinol exhibits a relatively strong platelet adhesion, but the shapes of attached platelets are different (Figure 10a,b,c). On the surface of the naked Nitinol, the adhered platelets were distorted with pseudopodia, while no deformation and no pseudopodia were seen for the platelets on the (PA/Hep)8 coating. Indeed, the (PA/ Hep)8 coating increased platelet binding on the devices and led to platelet activation, an effect which is related to the release of platelet adenosine diphosphate.57 In comparison, the (PA/Alg/ PA/Hep)4 coating results in a distinct reduction of platelet adhesion and almost no sign of any cellular matter on it, indicating that the incorporation of the alginate can effectively reduce the platelet adhesion and activation. Platelet adhesion is known to be initiated by adsorption of plasma proteins, such as fibrinogen. Concerning the lower platelet binding onto our (PA/Alg/PA/Hep)4 coating, one can speculate that the coating procedure we used led to a (55) Picart, C.; Lavalle, P.; Hubert, P.; Cuisinier, F. J. G.; Decher, G.; Schaaf, P.; Voegel, J. C. Langmuir 2001, 17, 7414. (56) Picart, C.; Mutterer, J.; Richert, L.; Luo, Y.; Prestwich, G. D.; Schaaf, P.; Voegel, J. C.; Lavalle, P. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 12531. (57) Xiao, Z. H.; Theroux, P. Circulation 1998, 97, 251.

Stabilized Hemocampatible Coating of Nitinol DeVices

Langmuir, Vol. 23, No. 18, 2007 9385

Figure 10. SEM photomicrographs of Nitinol surface after being contacted with fresh platelet-enriched plasma: (a) pristine Nitinol, (b) (PA/Hep)8-coated Nitinol, (c) (PA/Alg/PA/Hep)4-coated Nitinol. The magnification is ×10.0 k.

significant loss of platelet adhesive sites. The antifouling properties of alginate immobilized onto surfaces have been welldocumented. It is believed that alginate may keep the heparin active. It is likely that the cooperative effect of alginate and heparin resulted in the excellent blood compatibility of such a (PA/Alg/PA/Hep)4 coating.

Conclusions The layer-by-layer technique in combination with photoreaction in situ provides a simple and efficient way to cover biomedical devices with a stabilized multicomponent coating of a required and uniform thickness, irrespective of the surface shape. In comparison to the coating with mere heparin, the alginate/heparin coating becomes more efficient. The addition of alginate into the multilayer film resulted in an increase in hydrophilicity and stability and a decrease in surface roughness. The photo-cross-

linked alginate/heparin multilayer was shown to improve the blood compatibility by decreased thrombogenicity and strongly reduced platelet adhesion. This surface modification of implants with both alginate and heparin could be used to advantageously influence the initial stages of biomaterial implantation by activating and/or controlling the early biological events at the implant/host interface. Detailed and extensive animal studies must be carried out to utilize the alginate/heparin surfaces for improving blood- and tissue-compatible devices as well as tissueengineered devices. Acknowledgment. This research was financially supported by National Natural Science Foundation of China (NSFC50573015) and Program for New Centry Excellent in University. LA7002996