Phosphorylcholine-containing polymers for biomedical ... - Springer Link

2 downloads 0 Views 774KB Size Report
Nov 10, 2004 - polydimethylsiloxane, polytetrafluoroethylene (PTFE), cellulose, and polysulfone (PSf). Thrombus formation is one of the most serious host ...
Anal Bioanal Chem (2005) 381: 534–546 DOI 10.1007/s00216-004-2805-9

R EV IE W

Yasuhiko Iwasaki Æ Kazuhiko Ishihara

Phosphorylcholine-containing polymers for biomedical applications

Received: 1 June 2004 / Revised: 2 August 2004 / Accepted: 6 August 2004 / Published online: 10 November 2004 Ó Springer-Verlag 2004

Introduction

Phosphorylcholine-containing polymers

Numerous polymeric materials have been used in the manufacture of medical devices, especially for application as artificial organs where the polymer is in contact with blood [1]. The polymers currently in use, however, are commodity materials, such as poly(vinyl chloride) (PVC), polyethylene (PE), poly(methyl methacrylate) poly(MMA), segmented polyetherurethane (SPU), polydimethylsiloxane, polytetrafluoroethylene (PTFE), cellulose, and polysulfone (PSf). Thrombus formation is one of the most serious host responses to artificial materials. Because conventional polymer materials do not have sufficient blood compatibility and antithrombogenicity, infusion of an anticoagulant is required during clinical treatment using these medical devices to avoid thrombus formation. As a means of improving blood compatibility, some methods of surface modification using newly designed polymeric materials have been studied. The molecular design of bloodcompatible polymers for making artificial organs is classified into four categories based on the approach to the regulation of blood–material interaction. These categories are listed in Table 1. One of the most effective methods of making a blood-compatible polymer is to modify conventional materials with polymers having a phospholipid polar group mimicking a biomembrane surface.

2-Methacryloyloxyethyl phosphorylcholine (MPC) polymer

Y. Iwasaki (&) Institute of Biomaterials and Bioengineering, Tokyo Medical and Dental University, 2-3-10 Kanda-surugadai, Chiyoda-ku, Tokyo 101-0062, Japan E-mail: [email protected] Tel.: +81-3-52808026 Fax: +81-3-52808027 K. Ishihara Department of Materials Engineering, School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan

The biomembrane surface is regarded as the best surface for smooth interaction with blood components such as proteins and cells. A model of the structure of a biomembrane, which is well-known as the fluid-mosaic model, was proposed by Singer and Nicolson (Fig. 1) [2]. According to this model, amphiphilic phospholipids are arranged in a bilayer structure and proteins are located in or upon it. The distribution of these components is asymmetric. The phospholipids always flow dynamically, and the biomembrane maintains its strength by the supporting proteins. In all cells for which lipid compositional asymmetry has been described, negatively charged phospholipids such as phosphatidylserine are predominantly found on the inner, cytoplasmic side of the membrane, whereas the neutral, zwitterionic phosphorylcholine lipids such as phosphatidylcholines are located in the outer leaflet. The phosphatidylcholine surface provides an inert surface for biological reactions of proteins and glycoproteins to occur smoothly on the membrane. This provides very significant information for the development of novel blood-compatible polymers. Nakabayashi et al. [3] designed a methacrylate monomer with a phospholipid polar group, 2-methacryloyloxyethyl phosphorylcholine (MPC), to obtain new medical polymer materials (Fig. 2). At that time, however, the purity and yield of MPC were insufficient to evaluate their functions. Ishihara et al. [4] then improved the synthetic route to MPC and succeeded in producing MPC as a white powder by recrystallization. MPC, which contains the polymerizable methacrylate group, can readily be co-polymerized and enables the design of numerous polymers having a wide variety of molecular architectures including random [4, 5], block [6–9], graft [10], charged [11, 12], and end functional polymers [10].

535

Fig. 2 Chemical structure of MPC

Fig. 1 Fluid-mosaic model of biomembrane

The homopolymer of MPC is soluble in water and the solubility of MPC polymers can be easily controlled by changing the structure and fraction of the co-monomer. Figure 3 shows the chemical structure of poly(MPC-coBMA) (PMB). Using this polymer, we have investigated the blood compatibility of MPC polymers with special focus given to each blood component, e.g. cells, plasma protein, phospholipids, and water [13–20]. Figure 4 shows experimental columns containing polymer-coated poly(methyl methacrylate) beads and micrographs of the bead surface after contact with human blood without used of any anticoagulant. On the hydrophobic poly(n-butyl methacrylate) (poly(BMA)), many adherent cells are observed. Moreover, activation and aggregation of the cells and clot formation were observed on the poly(BMA). In contrast, poly(MPC-coBMA) (PMB) with 30 mol% MPC PMB30) effectively Table 1 blood-compatible polymer materials

Fig. 3 Chemical structure of PMB

suppressed cell adhesion. The adherent platelets readily detach from the surface, but are strongly activated in contact with the polymer surface, which induces embolization. It is, therefore, necessary to evaluate the activation of platelets just coming into contact with the polymer surface and those adhering to the surface to understand true blood-compatibility. Activation of platelets after contact with PMB was evaluated by measuring the concentration of cytoplasmic calcium ions ([Ca2+]i) in the platelets [18]. Figure 5 shows fluorescence micrographs of platelets after contact with glass and PMB30 surfaces. The brightness of the color is

536

Fig. 4 Microsphere columns after passage of human whole blood for 15 min and SEM micrographs of polymer coated-beads packed in the columns Fig. 6 Amount of protein adsorbed on polymer surfaces from human plasma

Fig. 5 Fluorescence micrographs of platelets after contact with glass and PMB30 surfaces

attributed to the [Ca2+]i level of the platelets. On the glass surface, a remarkable increase in the [Ca2+]i of the platelets was observed. In contrast, the [Ca2+]i level of the platelets in contact with PMB did not increase. Protein adsorption on the material surfaces causes serious biological reactions such as thrombus formation, immune response, complement activation, capsulation, etc. [21, 22]. To understand the blood-compatibility of surfaces, it is necessary to determine not only the amount of adsorbed protein but also the species of the protein. Therefore, the effects of MPC units on protein adsorption were investigated. Figure 6 shows the amount of protein adsorbed on PMB, poly(HEMA), and poly(BMA) after contact with plasma for 60 min [15]. On poly(BMA), many more proteins were adsorbed than on poly(HEMA) and PMB. The amount of proteins adsorbed on PMB decreased with an increase in the ratio of the MPC composition in the polymer. The species and distribution of the protein adsorbed on PMB were also determined by gold-colloid and radiolabeled immunoassay [14]. From these experiments it was clarified that the PMB could reduce plasma protein adsorption nonspecifically. Thrombus formation on conventional polymeric materials occurred through the multilayers of plasma proteins denatured by contact with the surfaces. The secondary structures of bovine

serum albumin (BSA) and bovine plasma fibrinogen (BPF) adsorbed on the PMB were evaluated by circular dichroism (CD) spectroscopy [19, 20]. Figure 7 shows CD spectra of BSA in PBS and that adsorbed on the polymer surface. In the BSA in PBS, the mean molecular residual ellipticity, (h), had a large negative value at 222 nm. The CD spectrum of BSA adsorbed on PMB was almost the same at that in PBS. The negative ellipticity at 222 nm of BSA adsorbed on the MPC polymers increased as the ratio of MPC decreased, then became almost zero for BSA adsorbed on poly(HEMA). We found the same tendency for BPF. Calculation of the a-helix content of BSA and BPF revealed that the PMB could effectively suppress the conformational change of proteins even when the proteins were adsorbed on the surface [19, 20]. In contrast, the a-helix content of both proteins adsorbed on poly(HEMA) was observed to decrease significantly. The protein adsorption-resistant

Fig. 7 Circular dichroism spectra of BSA in PBS and of that adsorbed on polymer surfaces: a poly(HEMA), b PMB10, c PMB30, d BSA/PBS

537

properties of the MPC polymer were also determined by other researchers [23–25]. We considered why the MPC polymer suppressed any significant interaction between proteins and cells. Is it possible to construct a biomembrane-like structure on the MPC polymer? Figure 8 shows the amount of phospholipids adsorbed on the PMB, poly(HEMA), and poly(BMA) after contact with plasma for 60 min [15]. In contrast with protein adsorption, the amount of phospholipids adsorbed on PMB was greater than that on poly(HEMA) and poly(BMA), and it increased with an increase in the ratio of the MPC composition in the polymer. The MPC unit in the polymer has affinity for the phospholipids in blood. We also evaluated the organization and distribution of phospholipids adsorbed on the MPC polymer and poly(HEMA) surfaces using differential scanning calorimetry (DSC) [6, 17], X-ray photoelectron spectrometry (XPS) [6], a quartz crystal microbalance (QCM) [26], and atomic force microscopy (AFM) [27]. Figure 9 shows the AFM images of the polymer surface and that treated with a dipalmitoyl phosphatidylcholine (DPPC) liposomal suspension. The surface of the polymer-coated mica film was quite flat,

Fig. 8 Amount of phospholipid adsorbed on polymer surfaces from human plasma, determined with a clinical kit (Phospholipid C-Test Wako)

Fig. 9 Atomic force microscopy images of polymer surfaces before and after contact with DPPC liposomal suspension

Fig. 10 Fibrinogen adsorption pattern on polymer surfaces from fibrinogen/PBS with and without DMPC liposomes: a Contact with fibrinogen/PBS. b Contact with fibrinogen/PBS containing DMPC liposomes

but this changed after the film was soaked in the DPPC liposomal suspension. The change in the roughness of the poly(HEMA) surface was slight, but it changed drastically on the MPC polymer surface, because of the adsorbed DPPC liposomes without deformation. From these measurements it was concluded that the phospholipid adsorbed on the MPC polymer surface formed a highly organized structure. The competitive adsorption of proteins and phospholipids on the MPC polymer was also evaluated using albumin, fibrinogen, and dimyristoylphosphatidylcholine (DMPC) as model components [28]. Figure 10 shows the fibrinogen adsorption pattern on the polymer surfaces from fibrinogen/PBS, with and without the DMPC liposome, as determined by gold-colloid-labeled immunoassay. On poly(BMA), many white dots could be observed with no relationship with the existence of the DMPC liposome. When PMB with 10 mol% MPC unit (PMB10) was treated with the fibrinogen solution, the number of white dots on the surface was the same as that on poly(BMA). However, the number of white dots was drastically reduced by addition of the DMPC liposome to the fibrinogen solution. On the PMB30, no effect of the DMPC liposome on fibrinogen adsorption was observed, because the amount of fibrinogen adsorbed on the original PMB30 was quite small even without treatment with the DMPC liposomal suspension. These results indicate that the DMPC liposomes were preferentially adsorbed on the MPC polymer surface compared with the protein, and that the adsorbed DMPC played an important role in reducing protein adsorption on the MPC polymer. Phospholipid adsorption is considered one of the reasons for the excellent blood compatibility of the MPC polymer. It has recently been reported that the structure of water absorbed in the polymer materials influences protein adsorption on their surfaces. Lu et al. [29] proposed a significant model for protein adsorption on the surface of a polymer. When proteins are adsorbed on the surface of a polymer by hydrophobic interactions, exchange of bound water between the protein and the surface is necessary [29]. The amount of bound water

538

might, therefore, be the key property enabling understanding of protein adsorption. Tsuruta [30] reported that the random networks of water molecules on the material surface were very important in explaining the protein adsorption. Protein adsorption processes are considered to start with protein trapping by the random networks of water molecules on the material surface. The material surface, which cannot undergo hydrogen bonding with water, will then reduce protein adsorption. Table 2 lists the free water concentration in the hydrated polymer membrane, as determined with differential scanning calorimetry [20]. When the degree of hydration of the polymers was adjusted at 0.36, the free water fraction of water absorbed in the MPC polymer was 0.65, which was found to be significantly higher than that in poly(HEMA), which was 0.28. In addition, the structure and hydrogen bonding of water in the vicinity of PMB were analyzed in their aqueous solutions and in thin films with contours of O–H stretching of the Raman and attenuated total reflection infrared (ATR–IR) spectra, respectively [31, 32]. The relative intensity of the collective band (C value) corresponding to a long-range coupling of O–H stretching of the Raman spectra for the aqueous solution of PMB was very close to that for pure water, which is in contrast with the smaller C value in the aqueous solution of ordinary polyelectrolytes. A similar tendency was also observed on hydrated thin polymer films. These results suggest that the PMB do not significantly disturb the hydrogen bonding between water molecules in either the aqueous solution or the thin film systems. The equilibrium amount of proteins, BSA, and BPF adsorbed on the polymer surface was measured and related to the free water fraction in the hydrated polymers, as shown in Fig. 11 [33]. The amounts of both proteins adsorbed on poly(HEMA), poly(acryl amide (AAm)-coBMA), and poly(N-vinylpyrrolidone(VPy)-co-BMA) were larger than those on the poly(MPC-co-dodecyl methacrylate) (PMD) and PMB. It was reported that the theoretical amount of BSA and BPF adsorbed on the surface in a monolayer state are 0.9 and 1.7 lg cm 2,

Table 2 Characteristics of the hydration states of the polymers

Fig. 12 Schematic representation of the mechanism of nonthrombogenicity observed on MPC polymer

respectively. On the surface of the MPC polymers, the amount of adsorbed proteins was less than these theoretical values. From an evaluation of the interaction between blood components and PMB, the mechanism of the blood compatibility of PMB can be represented as shown in Fig. 12. When the PMB was exposed to blood, the phospholipids were adsorbed on the PMB surface in preference to adsorption of proteins and to adhesion of platelets. The adsorbed phospholipids on the PMB surface form a biomimetic surface. The highly water-free content of hydrated PMB also contributes to the demonstrably excellent blood compatibility. Recently, the biocompatibility of MPC polymers has also been determined by measuring the amount of IL-1b mRNA secreted from adherent macrophage-like cells on polymer surfaces [34, 35]. The expression of mRNA from the adherent cells on the MPC polymer surfaces was significantly lower than from those on conventional polymeric biomaterials. This property was the basis of studies of the MPC polymer as a material for tissue engineering [36, 37]. Other phosphorylcholine-containing polymers

Fig. 11 Relationship between free water fraction in hydrated polymer membrane and amount of proteins adsorbed on the polymer: squares BSA, [BSA] in PBS=0.45 g dL 1, circles BPF, [BPF]=0.30 g dL 1

Numerous studies have been performed to create phospholipid-assembled surfaces that suppress any biological response with blood. Although black membrane or a phospholipid liposome has interesting properties, because of its biomimetic structure, its stability is quite poor. Ringsdorf and Schlarb [39] and Bonte et al. [38]

539

Fig. 13 Formation of a polyconjugated, phospholipid polymer from the diacetylenic monomer

Fig. 14 Chemical structures of phosphorylcholine ethanolamine and phosphorylcholine ethylene glycol dimethylsilyl chloride

investigated the polymerization of phospholipids with a polymerizable group and found that polymerized liposomes did not induce platelet aggregation in either plasma or blood. Hence, the adsorption of liposomes on the polymer support was a good method for preparing a phospholipid-assembled surface. They were, however, unable to make a biomimetic membrane, because the polymerization ability and mobility of the phospholipid polymer were quite low. Chapman et al. also started a study of blood-compatible surfaces in the 1980s from a biological perspective. Phospholipid molecules containing diacetylenic groups in their acyl chain (Fig. 13) were Fig. 15 Various designs of phosphorylcholine-assembled monolayers on solid surfaces

synthesized and formed into polymers upon irradiation with ultraviolet light [40]. Because diacetylene polymerization requires the order of solid-state reactivity, it is a function of the preservation of the structural order displayed by the monomeric lattice. The dynamic properties of the monomeric, diacetylenic phospholipids were found to resemble those of naturally occurring lipids. Diacetylenic phospholipids have two significant advantages for the utilization of multilayered films. First, polymerization of the diacetylene group forms a perfectly regular polymer with a stable polymeric phospholipid. This should render the surface more biocompatible, especially when the polymer is engineered to mimic the surface of the host cells. Diacetylenic phospholipid can be coated on the surfaces of a variety of materials to keep out the phosphorylcholine group. Kono et al. [41] reported that polyamide microcapsules treated with phosphatidylcholines suppressed platelet adhesion. Hall et al. [42] performed a thromboelastographic study of a variety of surfaces treated with phosphatidylcholines and observed prolongation of clotting time compared with untreated surfaces. Durrani et al. [43] and Hayward et al. [44, 45] synthesized phosphorylcholine derivatives with dimethylsilylchloride and ethanolamine groups (Fig. 14). These phosphorylcholine derivatives were coated on glass or a variety of polymer surfaces by covalent bonding. Tegoulia and Cooper [46] synthesized alkanethiols, with a variety of functional groups and which reacted with a gold surface. It was clear that the alkanethiols fixed on the gold surface create a self-assembled monolayer (SAM). The water contact angle of the SAM surface prepared with the alkanethiol having a hydrophilic polar group was less than that of the gold surface. On the SAM surface prepared with the phosphorylcholine group, adhesion of the neutrophils was effectively reduced.

540 Fig. 16 Chemical structures of MPC derivatives

Fig. 17 Resonance frequency change of polymer-coated resonators after addition of DPPC liposomal suspension

Marra et al. [47] synthesized the phospholipid monomer 1-palmitoyl-2-[12-(acryloyloxy)dodecanoyl]sn-glycero-3-phosphorylcholine, as unilamellar vesicles, and fused it on to alkylated glass. Free-radical polymerization was performed in an aqueous solution at 70°C. It was clear from XPS analysis that the phospholipid assembly had a close-packed monolayer formation. This formation is very stable under static conditions in water and air, and in an environment having a high shear flow. Blood compatibility was assessed in a baboon arteriovenous shunt model, which revealed minimal platelet deposition over observation for 2 h. Kohler et al. [48] prepared a glass surface that reacted with 3-aminopropyltrimethoxysilane. The carboxylated phosphatidylcholine was grafted with a coupling agent. The number of adherent platelets on the glass surface

modified with the carboxylated phosphatidylcholine was suppressed. As shown in Fig. 15, various types of phosphorylcholine-assembled layers on solid surfaces have been explored, and the ‘‘non-biofouling’’ properties of these surfaces have been reported [49–52]. Matsuda et al. [53] prepared a phosphorylcholineendcapped oligo(N,N-dimethylacrylamide) [oligo(DMEAA)] and a block co-oligomer with oligo(styrene) by a photoiniferter-based quasiliving polymerization technique. The oligomer has amphiphilic properties and chemisorbs on a gold surface with hydrophobic anchoring. The surface coated with the oligomers reduced plasma protein adsorption and cell adhesion. The authors also explored a surface design for producing one or two phosphorylcholine groups capped at the terminal end of a graft chain of poly(DMEAA) [54]. Recently, syntheses of a variety of MPC derivatives and their polymers have been conducted, as is shown in Fig. 16 [55–59]. Iwasaki et al. [55, 60, 61] synthesized xmethacryloyloxyalkyl phosphorylcholine, with a variety of methylene chain lengths between the phospholipid polar group and the backbone. This enhances the affinity of the MPC polymer for phospholipids in blood. Figure 17 shows the change in the frequency of resonators coated with the MAPC polymers after the DPPC liposomal suspension had been soaked in PBS at 25°C for 60 min [27]. A decrease in frequency was observed, which corresponded to the adsorption of the DPPC liposomes on the surface. The magnitude of the change in frequency observed on the MAPC resonators increased in the order PMB30