- Email: [email protected]
Oriented immobilized anti-LDL antibody carrying poly(hydroxyethyl methacrylate) cryogel for cholesterol removal from human plasma
Materials Science and Engineering C 31 (2011) 1078–1083
Contents lists available at ScienceDirect
Materials Science and Engineering C j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / m s e c
Oriented immobilized anti-LDL antibody carrying poly(hydroxyethyl methacrylate) cryogel for cholesterol removal from human plasma Nilay Bereli a, Gülsu Şener b, Handan Yavuz a,⁎, Adil Denizli a a b
Department of Chemistry, Hacettepe University, Beytepe, Ankara, Turkey Nanotechnology and Nanomedicine Division, Hacettepe University, Ankara, Turkey
a r t i c l e
i n f o
Article history: Received 13 April 2009 Received in revised form 22 October 2010 Accepted 15 March 2011 Available online 29 March 2011 Keywords: Cholesterol removal Cryogels Oriented immobilization Protein A
a b s t r a c t Low density lipoprotein (LDL) cholesterol is a major ingredient of the plaque that collects in the coronary arteries and causes coronary heart diseases. Among the methods used for the extracorporeal elimination of LDL from intravasal volume, immunoaffinity technique using anti-LDL antibody as a ligand offers superior selectivity and specificity. Proper orientation of the immobilized antibody is the main issue in immunoaffinity techniques. In this study, anti-human β-lipoprotein antibody (anti-LDL antibody) molecules were immobilized and oriented through protein A onto poly(2-hydroxyethyl methacrylate) (PHEMA) cryogel in order to remove LDL from hypercholesterolemic human plasma. PHEMA cryogel was prepared by free radical polymerization initiated with N,N,N′,N′-tetramethylene diamine (TEMED). PHEMA cryogel with a swelling degree of 8.89 g H2O/g and 67% macro-porosity was characterized by swelling studies, scanning electron microscope (SEM) and blood compatibility tests. All the clotting times were increased when compared with control plasma. The maximum immobilized anti-LDL antibody amount was 63.2 mg/g in the case of random antibody immobilization and 19.6 mg/g in the case of oriented antibody immobilization (protein A loading was 57.0 mg/g). Random and oriented anti-LDL antibody immobilized PHEMA cryogels adsorbed 111 and 129 mg LDL/g cryogel from hypercholesterolemic human plasma, respectively. Up to 80% of the adsorbed LDL was desorbed. The adsorption– desorption cycle was repeated 6 times using the same cryogel. There was no significant loss of LDL adsorption capacity. © 2011 Elsevier B.V. All rights reserved.
1. Introduction LDL cholesterol is widely recognized as one of the major risk factors in the development of coronary heart diseases because of its ability to build up in the lining of arteries forming atheromas and fatty acid deposits [1–3]. The causative factor in high levels of LDL can be of genetic reasons, such as familial hypercholesterolemia [4,5]. A majority of the hypercholesterolemic patients can be effectively treated by reduced dietary intake and drug therapy to control plasma cholesterol levels, preventing severe heart diseases. However, a more aggressive approach is necessary in severe hypercholesterolemia [6]. Since the first plasmapheresis procedure implemented for homozygous familial hypercholesterolemia; a number of additional methods with different specificity and selectivity have been developed for the extracorporeal removal of LDL from the blood, including cascade filtration, heparin induced extracorporeal LDL precipitation (HELP), thermo-filtration, dextran induced LDL precipitation and direct adsorption of lipoproteins (DALI) [7]. Extracorporeal elimination is an effective and life saving procedure in hypercholesterolemia patients [8]. Of
⁎ Corresponding author. E-mail address: [email protected] (H. Yavuz). 0928-4931/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.msec.2011.03.008
course the most selective binding affinity can be achieved by using immunoaffinity adsorbents containing antibodies as the ligand [9–12]. For instance; in case of high LDL levels in patients, antibodies against apoprotein B100, the main protein component of LDL, are used for the selective extracorporeal removal of harmful LDL without depriving the patients of the useful HDL and other blood components. Antibody immobilization may be a critical step in the design of immunoaffinity adsorbents [13]. When antibodies are covalently immobilized onto adsorbents, their specific binding capacity usually decreases as compared to soluble antibodies [14]. This reduction is attributed to the random orientation of the antibodies on the surface of the solid support used, which disables the accessibility of the antigen binding fragments of the antibodies (Fab) towards their antigens. In normal coupling procedures, the target antibodies are immobilized at many different spots resulting random orientation of the antibodies on the matrix that might prevent the formation of the antibody–antigen complex [15]. More specific immobilization of antibodies through the Fc region of the heavy immunoglobulin chain would be very efficient in terms of providing orderly oriented immobilized antibody so that the antibody binding sites, located on the opposite part of the immobilization sites, become highly accessible to antigen binding. Protein A from Staphylococcus aureus has a unique capability of binding mammalian
N. Bereli et al. / Materials Science and Engineering C 31 (2011) 1078–1083
immunoglobulins through the Fc region [16]. Thus protein A can be employed for the orderly oriented immobilization of antibodies by formation of a protein A immobilized matrix followed by cross-linking of antibody with protein A in the matrix. The advantages of the orderly oriented immobilization process for biologically active proteins can be outlined as good binding accessibility and increased stability [16]. In biological applications, the conventional column chromatography techniques primarily use gel-beads with certain limitations such as high pressure drops and low flow-rates, lowering the efficiency in routine and scale-up applications [17]. In addition, the gel-bead column chromatography is not capable of applying highly viscous fluids such as human blood. Therefore, alternative chromatography techniques such as membranes, monoliths and cryogels [18–21] are applied in cases when gel-bead column chromatography techniques fail. Cryogels provide a potential solution in terms of their low pressure drop and lack of diffusion resistances utilizing macropores as compared to the traditional gel-bead columns [22–24]. Cryogel columns enable high flow-rates allowing voluminous elutions within shorter times. Whole blood can be applied on cryogel columns without any pre-treatment [25–29]. Cryogels are also cheap and thus they can be easily disposed of, eliminating cross-contamination between batches [30]. The PHEMA cryogels were selected for three good reasons: 1) They exhibit a low pressure drop, 2) they lack diffusion resistance and 3) viscous samples such as whole blood can be easily applied on them. We have focused our attention on the development of anti-LDL antibody immobilized immuno-affinity PHEMA cryogels by combining the selectivity of immunoaffinity interaction with the biocompatibility and good flow properties of PHEMA cryogels. LDL removal performance of these immunoaffinity cryogels was reported here. 2. Experimental 2.1. Materials Anti-LDL antibody (anti-human β-lipoprotein; Product No: L-8016) and protein A (from S. aureus, Cowan Strain I; Product No: P-6031) were obtained from Sigma (St. Louis, USA). 2-Hydroxyethyl methacrylate (HEMA) was obtained from Fluka A.G. (Buchs, Switzerland), which was later distilled under reduced pressure in the presence of hydroquinone inhibitor and stored at 4 °C until use. Cyanogen bromide (CNBr), N,N′methylene-bis(acrylamide) (MBAAm), ammonium persulfate (APS) and N,N,N′,N′-tetramethylene diamine (TEMED) were also obtained from Sigma. All other chemicals were of reagent grade and were purchased from Merck AG (Darmstadt, Germany). Water used in adsorption experiments was purified by a Barnstead (Dubuque, IA) ROpure LP® reverse osmosis unit having a high flow cellulose acetate membrane (Barnstead D2731) followed by a Barnstead D3804 NANOpure® organic/colloid removal and ion exchange packed-bed systems. 2.2. Preparation of PHEMA cryogels Preparation of cryogel is described elsewhere [20]. Briefly, monomers (1.3 mL of HEMA, and 10 mL of MBAAm) were dissolved in deionized water and the mixture was degassed under vacuum for about 5 min to remove soluble oxygen, yielding a total monomer concentration of 6% (w/v). The cryogel was prepared by free radical polymerization initiated by firstly addition of 20 mg of APS, 1 % (w/v) of the total monomers, into the solution of the monomers followed by cooling in an ice bath for 2–3 min, and then addition of 25 μL of TEMED, 1% (w/v) of the total monomers, which was stirred for 1 min. Then, the reaction mixture was poured into a plastic syringe (5 mL, internal diameter: 0.8 cm) with closed outlet at the bottom. The polymerization solution in the syringe was frozen at −12 °C for 24 h, which was then thawed at room temperature. After washing with 200 mL of water, the cryogel
1079
column was stored in a buffer solution containing 0.02% sodium azide at 4 °C until use. 2.3. CNBr activation As adsorber materials anti-LDL antibody was used for random immobilization while protein A was used for orderly oriented immobilization. The PHEMA cryogel column was activated using cyanogen bromide in order to create reactive sites for protein immobilization, with either anti-LDL antibody or protein A. Prior to the activation process, the PHEMA cryogel column was kept in distilled water for about 24 h, which was then washed by passing a 0.5 M NaCl solution and 50 mL of 0.5 M sodium carbonate buffer solution (pH 10.5). An aqueous solution of CNBr (50 mg CNBr/mL) was carefully prepared in a fume hood and its pH was quickly adjusted to 11.5 using a 4 M NaOH solution and a pH meter (Mettler Toledo Gmbh, Switzerland). The pH of the CNBr solution was maintained between 10.5 and 11.5 during the activation reaction. The CNBr solution was passed from the cryogel column with a flow rate of 1.0 mL/min over 2 h at room temperature. After the activation reaction, the column was washed with cold sodium citrate buffer (0.1 M; pH 6.5) and the filtrate was discarded. 2.4. Random immobilization of anti-LDL antibodies 20 mL of an anti-LDL antibody solution (1.0 mg/mL) dissolved in 0.1 M sodium citrate buffer (pH 6.5) passed through a freshly activated PHEMA cryogel column at a 1.0 mL/min flow rate at 4 °C overnight. After immobilization, 2.0 M ethanol amine solution was recirculated for another 1 h in order to quench side reactions that may form between unreacted sites of the protein and any remaining active groups (e.g. isourea) on the cryogel surface. The amount of anti-LDL antibody immobilization was monitored by measuring the decrease in the protein concentration in the anti-LDL antibody solution by Bradford method. Non-specifically adsorbed anti-LDL antibody molecules on PHEMA were also considered. The amount of immobilized anti-LDL antibody was calculated as: q = ½ ðCi –Ct Þ:V = m
ð1Þ
where, q is the amount of anti-LDL antibody immobilized onto unit mass of the cryogel (mg/g); Ci and Ct are the concentrations of the anti-LDL antibody in the initial solution and in the supernatant after immobilization (mg/mL), respectively, V is the volume of the aqueous phase (mL), and m is the mass of the cryogel column (g). 2.5. Oriented immobilization of anti-LDL antibodies Covalent immobilization of protein A onto freshly activated PHEMA cryogels was performed using the same procedure described in Section 2.4. The CNBr and protein A concentrations were 50 mg/mL and 1.0 mg/mL, respectively. The immobilization (0.1 M sodium citrate) buffer solution used was adjusted at pH 6.5. The amount of protein A immobilized onto the CNBr activated PHEMA cryogel was determined by measuring the decrease in the protein A concentration by the Bradford method, which was calculated with Eq. 1. The protein A-PHEMA cryogel was washed several times using 0.1 M glycine–HCl solution (pH 3.5) and the immobilization buffer in order to remove impurities. The anti-LDL antibody solution was prepared in a borate buffer solution (1.0 mg/mL, pH 7.5). The protein A immobilized PHEMA cryogel column was treated with the anti-LDL antibody solution for 2 h at room temperature to form an antibody protein A complex. Then, protein A was cross-linked to the adsorbed anti-LDL antibody by adding a solution of 0.1% (w/v) cyanamide (CH2N2), prepared in the immobilization buffer solution, into the protein A immobilized column and the reaction was carried out for about 30 min. After the reaction, the column was washed several times with a solution of 0.1 M NaCl and
1080
N. Bereli et al. / Materials Science and Engineering C 31 (2011) 1078–1083
distilled water. The affinity adsorbent was stored in 0.02% (w/v) NaN3 at 4 °C until use. 2.6. Characterization of PHEMA cryogels 2.6.1. Swelling degree The swelling degree (S) of the cryogel was determined as follows: a cryogel sample was taken out of its column and rinsed with ethyl alcohol and distilled water until a clear supernatant was obtained. The cryogel was squeezed to remove unbound water residues as much as possible, which was then transferred to pre-weighed vial and weighed out (mwet gel). The wet cryogel was dried in an oven at 60 °C, which was then weighed out to determine the mass of dried sample (mdry gel). The swelling degree was calculated as: S = mwet gel –mdry gel = mdry gel :
ð2Þ
The total volume of macropores in the swollen cryogel was roughly estimated as follows. The cryogel sample was weighed out after it was swelled in 40 mL of water (mswollen gel) and then the weight of the sample was determined after squeezing (msqueezed gel). The porosity was calculated as: Porosity % = mswollen gel –msqueezed gel = mswollen gel × 100%:
ð3Þ
cholesterol in the plasma after incubation with anti-LDL antibody containing adsorbent. The amounts of the total cholesterol (TC), LDL cholesterol (LDL), HDL cholesterol (HDL), VLDL cholesterol (VLDL) and triglycerides were determined by homogeneous enzymatic colorimetric assay on an automated analyzer instrument (Hitachi 912, Japan). 2.8. Desorption and repeated use Adsorbed LDL was desorbed using a desorption solution of 0.1 M citric acid and 0.02 M Na2HPO4, (pH 3.0). In a typical desorption experiment, 50 mL of the desorption solution was pumped through the cryogel column at a flow rate of 1.0 mL/min for 1 h. The final concentration of LDL in the solution was determined by the same procedure described in Section 2.7. The desorption ratio corresponds to the final LDL concentration in the desorption medium over the amount of LDL adsorbed on the cryogel. In order to test the reusability of anti-LDL antibody containing PHEMA cryogel, the LDL adsorption–desorption procedure was repeated 6 times on a single cryogel column. The PHEMA cryogel was washed with 50 mM NaOH solution after each adsorption–desorption cycle to ensure regeneration and sterilization. 3. Results and discussion 3.1. Characterization of PHEMA cryogel
2.6.2. SEM studies The morphology of a cross section of the dried cryogel was investigated by scanning electron microscopy (SEM). The sample was mounted in a solution of 2.5% glutaraldehyde in 0.15 M sodium cacodylate buffer overnight, which was then kept in a solution of 1% osmium tetroxide for 1 h. Then the sample was dehydrated stepwise in ethanol and transferred to a critical point drier at 10 °C where the ethanol was replaced with liquid carbon dioxide as the transitional fluid. The temperature and pressure of the system were then raised to 40 °C and 100 bar, respectively. Liquid CO2 was transformed directly to gas uniformly throughout the whole sample without heat of vaporization or surface tension forces causing damage. Release of the pressure at a constant temperature of 40 °C resulted in dried cryogel sample. The cryogel was finally coated with gold-palladium (40:60) and examined using a JEOL JSM 5600 scanning electron microscope. 2.6.3. Biocompatibility studies The human blood PHEMA cryogel interaction was monitored by measuring the following parameters; prothrombin time (PT), activated partial thromboplastin time (APTT), fibrinogen time and thrombocyte clotting time. The in vitro hemocompatibility tests were implemented by STA4 Compact Blood Coagulation Analyzer (Diagnostica Stago, France) using the required test kits. 2.7. LDL removal studies The efficiency of the LDL removal procedure was tested in vitro in a continuous experimental setup. The plasma samples with an average cholesterol initial concentration of 431.0 mg/dL were obtained from patients with hypercholesterolemia. Fresh frozen plasma was donated by the Blood Bank at the Mesa Hospital (Ankara). Blood samples were centrifuged at 500 g for 30 min at room temperature to separate plasma. Plasma was filtered using 0.45 μm syringe filters (Model 245-0045 Nalge Co., Rochester, New York), and stored at 4 °C. Sodium azide (0.1% w/v) was added to prevent bacterial growth. Plasma (20 mL) was recirculated from the PHEMA, PHEMA-anti-LDL antibody and PHEMAprotein A-anti-LDL antibody cryogel columns at 20 °C for 4 h. The amount of the removed cholesterol was calculated by Eq. 1, using the initial amount of cholesterol in the plasma and the amount of
A supermacroporous PHEMA cryogel was prepared by copolymerization of HEMA and MBAAm in the frozen state in the presence of APS/TEMED. The SEM image of the pore structure of the cryogel is shown in Fig. 1. The macropores in the PHEMA cryogel bulk structure are open and highly interconnected, forming a network of large channels. The mobile protein phase is forced to flow through these channels, transporting the biomolecules to the active binding sites by convection, which results in an extremely fast mass exchange between the mobile protein phase and the stationary cryogel phase in regards to the traditional packed bed columns [31,32]. The macroporosity and equilibrium swelling degree of the PHEMA cryogels were 67% and 8.89 [g H2O/g cryogel], respectively. The PHEMA cryogel is white, opaque, spongelike and elastic, which can be easily squeezed by hand to remove water accumulated inside the macropores. The squeezed cryogel is able to restore its original size and shape within 1–2 s when soaked in water. Ideally, the pressure drop needed to drive the liquid through any systems should be as low as possible. The pressure drop tests on the PHEMA cryogel columns were performed in water which was used as an equilibration medium. The water phase was passed through the column for 1 min at each incremented linear flow rate ranging from 40 to 382 cm/h. Due to the presence of large and highly interconnected macropores, the PHEMA cryogel column was found to have a very low liquid flow resistance. In addition the PHEMA cryogel column was found to possess a low back pressure and low liquid flow resistance indicating the presence of porous structure with large interconnected macropore structures, as also seen in the SEM picture in Fig. 1. The physicochemical properties determined for the PHEMA cryogel are presented in Table 1. In general cryogels with supermacroporous structures allow direct processing of whole blood containing blood cells [25,26]. In a similar vein, the macropore size of the PHEMA cryogel prepared in our laboratory was tested with whole blood and was found to be very suitable to process blood cells without ever blocking the column. Table 2 briefly outlines the coagulation data obtained in the blood compatibility tests. As seen in Table 2, the clotting times determined for the PHEMA and PHEMA-anti-LDL antibody cryogels are within a comparable range of values for the control plasma. Observed decreases in coagulation times are generally tolerated by the body [33]. Therefore, we conclude that the blood application of PHEMA and anti-LDL antibody
N. Bereli et al. / Materials Science and Engineering C 31 (2011) 1078–1083
1081
Table 2 Coagulation times of human plasma (reported in s)⁎. Material
Control PHEMA PHEMA/anti-LDL antibody
Parameter PT
APTT
Fibrinogen time
Thrombocyte clotting time
12.2 ± 0.8 12.5 ± 0.9 12.4 ± 0.8
27.8 ± 1.5 26.2 ± 1.3 29.3 ± 1.5
14.1 ± 0.9 12.1 ± 0.9 18.3 ± 0.9
15.0 ± 0.8 16.6 ± 1.0 16.4 ± 1.0
⁎ Each result is an average of three determinations.
Fig. 1. SEM images of the PHEMA cryogel.
immobilized cryogels becomes acceptable with reproducible clotting times in regards to literature values [34]. 3.2. Random immobilization of anti-LDL antibodies Anti-LDL antibody was used as a bioaffinity ligand for the removal of LDL. It should be noted that, immobilization at pH 6.5 is less efficient than higher pH values, but it's less likely to compromise the binding ability of immobilized ligands (especially antibodies) [35]. Non-specific adsorption of an anti-LDL antibody ligand onto plain PHEMA cryogel was 0.17 mg anti-LDL antibody/g, which makes the PHEMA cryogel a potential carrier of the ligand. The amount of the anti-LDL antibody immobilized on the CNBr activated PHEMA cryogel was 63.2 mg antibody/g cryogel as determined in a media containing 50 mg/mL CNBr and 1.0 mg/mL anti-LDL antibody. The performance of the adsorbent is enhanced by the inertness of the cryogel surface and the specificity of the antibody. 3.3. Oriented immobilization of anti-LDL antibodies Some strains of S. aureus biosynthesize protein A, a 40 kDa protein that specifically binds to the Fc region of some antibodies with high
Table 1 Physicochemical properties of the PHEMA cryogel. Specific surface area Pore size diameter Porosity Swelling degree Back pressure
20.2 m2/g 10–200 μm 67% 8.89 g H2O/g cryogel 0.28 MPa
affinity [36,37], except for human IgG3, mouse IgG3, sheep IgG1, and also binds weakly to mouse IgG1. For some species, antibodies do not bind to protein A at all and some monoclonal antibodies show abnormal affinity for the protein A. These properties make the use of protein A advantageous for antibody separation, purification and also oriented immobilization through its Fc region as in this study. Since the IgG-protein A interactions are well known to be strongly pH dependent, the pH effects were mainly examined with three different buffer systems in our previous study [38]. Binding of anti-LDL antibody molecules to protein A occurs through an induced hydrophobic fit and it is promoted by addition of salts. At the center of the Fc binding site as well as on protein A reside histidine residues. At alkaline pH, these residues are uncharged and hydrophobic, strengthening the interaction between protein A and the antibody. As the pH is shifted to acidic values, these residues become charged and repel each other [39]. Ionic strength of buffers was adjusted by adding NaCl (0.15 M). Multiple uses of the immunoadsorbents require the crosslinkage of antibody with protein A through a covalent bond. For this reason, 0.1% (w/v) carbodiimide solution prepared in the same binding buffer was added after the binding reaction has completed. Protein A is covalently immobilized through the active sites on PHEMA cryogel. The amount of immobilized protein A was found as 57 mg/g at 50 mg/mL CNBr and 1.0 mg/mL protein A concentration. The binding solution was 0.1 M sodium citrate buffer (pH 6.5). Anti-LDL antibody binding was achieved with borate buffer at a pH of 7.5 (containing 0.15 M NaCl). The amount of anti-LDL antibody immobilization was 19.6 mg/g. It should be noted that protein A molecules have five Fc binding receptors, three of these receptors are reported to become inactive when the molecule is immobilized. 3.4. Cholesterol removal from human plasma Fig. 2 shows the change of cholesterol concentration in the hypercholesterolemic plasma in the course of incubation time. As seen in these figures, LDL levels were significantly reduced when the PHEMAanti-LDL antibody and PHEMA-protein A-anti-LDL antibody immobilized cryogels were used (i.e., 54% and 59% of the initial LDL, respectively). It should be noted that adsorption equilibrium is reached in nearly half an hour. Adsorption time is especially important for the patient's comfort when the system is thought to be used in extracorporeal treatment. Optimum equilibrium time for the bead based adsorbents is in the range of 3–6 h [1]. Big molecules, such as plasma proteins, diffuse slowly through a particulate support, and therefore longer contact times of the molecule with the adsorbent are required. Because of their favorable mass transfer mechanisms, continuous porous cryogel requires shorter contact times and therefore shorter run times [40]. As seen in these figures, anti-LDL antibody immobilized through protein A (i.e., oriented immobilized) was more efficient than randomly immobilized anti-LDL antibody in removing cholesterol from hypercholesterolemic human plasma, although the immobilized amount of anti-LDL antibody through the protein A was nearly one third of the randomly immobilized anti-LDL antibody onto PHEMA cryogel. This shows the mode of oriented immobilization makes the binding site of antibody always located on the top of its Fab variable regions, well accessible for interaction with LDL molecules. So, we can
1082
N. Bereli et al. / Materials Science and Engineering C 31 (2011) 1078–1083
Fig. 2. Cholesterol removal from hypercholesterolemic human plasma with PHEMA, PHEMA-anti-LDL antibody and PHEMA protein A-anti-LDL antibody cryogels. Total cholesterol concentration: 431 mg/dL; LDL concentration: 295 mg/dL; anti LDL antibody loading: 63.2 mg/g (random), 19.6 mg/g (oriented); T: 20 °C; flow rate: 1.0 mL/min.
say that the binding of antibody molecules through their Fc region was effective. Fig. 3 shows the comparison of LDL and other lipoprotein adsorption amounts of random and oriented anti-LDL antibody immobilized PHEMA cryogels. A support to be used as an affinity adsorbent must be physically inert to prevent false positives and to increase system performance. Our strategy is to develop an affinity adsorbent that selectively adsorbs LDL with low non specific removal of other (lipo) proteins, especially of the protective HDL and other blood proteins. HDL, other lipoproteins and triglyceride adsorption onto these adsorbents was also investigated. As can be seen from this figure, relatively small amounts of HDL, VLDL and triglycerides were adsorbed on both, random and oriented, anti-LDL antibody immobilized adsorbents. Selective adsorption of LDL with low non-specific removal of other (lipo)proteins, especially of the protective HDL; sufficient cholesterol adsorption capacity; and excellent whole blood compatibility allow cholesterol adsorption from whole blood. As can be also seen from the figure, bare PHEMA cryogel adsorbed negligible amount of lipoproteins indicating that the adsorption of lipoproteins onto the immunoaffinity adsorbents occurred due to the existence of the anti-LDL antibody ligand on the PHEMA surface. Different affinity adsorbents with different adsorption capacities were reported in literature for cholesterol removal studies. Denizli
reported 4.5–7.2 mg cholesterol/mL adsorption capacity with antiLDL antibody immobilized poly(HEMA-MAPA) membrane [9]. Yavuz and Denizli have reported 13.3–16.0 mg cholesterol/g adsorption capacity with anti-LDL immobilized poly(HEMA-MAPA) microbeads [10]. Yavuz and Denizli removed 39% and 47% of the total cholesterol with random and oriented anti-LDL antibody immobilized poly (HEMA-EGDMA) beads, respectively [38]. Denizli and Pişkin studied cholesterol removal from hypercholesterolemic human plasma and they reported maximum 4.7 mg cholesterol/g adsorption capacity with low-molecular weight heparin immobilized poly(2-hydroxyethyl methacrylate) microspheres [41]. Tabak et al. used heparin immobilized agarose beads and they obtained in the range of 1.2–3.0 mg/g cholesterol adsorption capacity from healthy human plasma [42]. Ostlund reported 6–8 mg lipoprotein cholesterol per mL column volume with anti-LDL antibody immobilized commercially available Agarose beads from human plasma [43]. Smolik et al. interested in determining the LDL adsorption capacity of dextran sulfate immobilized cellulose affinity beads [44]. They achieved a good selective LDL adsorption from whole blood and presented cholesterol adsorption capacities of 5–10 mg/mL gel. Schmidt used surface modified polyacrylate matrix and he obtained 8–10 mg cholesterol per gram polymer [45]. Lopukhin et al. used macroporous silica beads as the carrier matrix, and immobilized heparin and chytozane sulfate as specific bioligand. They reported cholesterol adsorption capacities around 14.8–15.2 mg/g [46]. Pokrovsky et al. used commercial carrier made of Sepharose carrying monoclonal and polyclonal antibodies as the specific ligands [47]. Their maximum LDL binding capacities were in the range of 0.6–2.5 mg/mL adsorbent. Sinitsyn et al. used Sepharose beads carrying heparin and they reported adsorption values of up to 12 mg cholesterol per gram polymer [48]. Wang et al. prepared L-lysine containing cellulose for electrostatic interaction with LDL and reached up to 60.9% adsorption percentage [49]. Utsumi et al. reduced LDL selectively to sufficient levels in hypercholesterolemia by dextran sulfate columns [50]. Soltys and Etzel prepared immunoaffinity membranes with anti-apo B IgG immobilized through PEG spacer and reached 1.0 mg/mL abo B adsorption capacity [51]. Wang et al prepared dextran adsorbent with amphiphilic ligands for adsorption of low-density lipoprotein and reached 1.916 mg/mL adsorption capacity [52]. Sellergren et al. prepared a series of highly cross-linked molecular imprinted terpolymers in the presence of cholesterol acting as a template molecule [53]. Using a physiological relevant intestinal-mimicking solution of cholesterol, these polymers adsorbed ca. 17 mg cholesterol per gram dry adsorbent. Yu et al. developed a new LDL-adsorbent by using graft polymerization technique [54]. Yavuz et al. prepared cholesterol imprinted poly (HEMA-MAT) particles and obtained 16.23 mg/g cholesterol adsorption capacity [55]. In this study, we demonstrated that random and oriented anti-LDL antibody immobilized PHEMA adsorbents showed excellent LDL adsorption capacities (129 mg/g). 3.5. Desorption and reusability
Fig. 3. Lipoprotein adsorption amounts of random and oriented anti-LDL antibody immobilized PHEMA cryogel. Total cholesterol concentration: 431 mg/dL; anti LDL antibody loading: 63.2 mg/g (random), 19.6 mg/g (oriented); T: 20 °C; flow rate: 1.0 mL/min; adsorption time: 2 h.
Desorption of LDL molecules from the randomly and orderly oriented anti-LDL antibody immobilized PHEMA cryogels was also studied by recirculating a desorption medium through an LDL adsorbed PHEMA cryogel and measuring the amount of the LDL desorbed in 1 h. The desorption buffer solution consisted of 0.04 M citric acid and 0.02 M Na2HPO4 at pH 3.0. Hypercholesterolemic human plasma was used for repeated cycles of cholesterol adsorption–desorption. It was found that up to 80% of the adsorbed cholesterol molecules were desorbed (Fig. 4). Note that the anti-LDL antibody was not released during the desorption processes, indicating that both the randomly and orderly oriented antibody molecules, as the ligands, were covalently immobilized onto the PHEMA cryogels. It is likely that the citric acid buffer solution alters the tertiary structure of the anti-LDL antibody, weakening the non-covalent bonds at the binding interface, which
N. Bereli et al. / Materials Science and Engineering C 31 (2011) 1078–1083
Fig. 4. Reusability of anti-LDL antibody-PHEMA and anti-LDL antibody-protein A-PHEMA cryogel.
ultimately leads to desorption of LDL. The adsorption–desorption cycle of LDL was repeated 6 times using the same cryogel column. Based on the desorption data presented in Fig. 4, it is concluded that citric acid is a suitable desorption agent for the removal of the LDL from anti-LDL immobilized PHEMA columns, and allows repeated use of the column. 4. Conclusion The results presented in this study reveal that PHEMA and anti-LDL antibody immobilized PHEMA cryogels are biocompatible and that the anti-LDL antibody immobilized PHEMA cryogels developed in our laboratory can be selectively used for the extracorporeal removal of LDL out of hypercholesterolemic human plasma. The maximum anti-LDL antibody immobilization amount was 63.2 mg/g in the case of random immobilization and 19.6 mg/g in the case of oriented immobilization. Random and oriented anti-LDL antibody immobilized PHEMA cryogels adsorbed 111 and 129 mg LDL/g cryogel from hypercholesterolemic human plasma, respectively, which are sufficient values to reduce blood LDL to an acceptable level. Up to 80% of the adsorbed LDL was desorbed and there was no significant loss of LDL adsorption capacity after 6 adsorption–desorption cycles. References [1] [2] [3] [4]
[5] [6] [7] [8] [9] [10]
G.R. Thompson, Atherosclerosis 167 (2003) 1. E. Bruckert, Transfus. Sci. 20 (1999) 43. N. Tani, Ther. Apher. 4 (2) (2000) 135. WHO, Human Genetics Programme. Division of Noncommunicable Diseases, Familial Hyperchoesterolemia (FH): Report of WHO Consultation. WHO/HGN/FH/ CONS/98, 7, 1998, p. 1. M. Bláha, Acta Med. 46 (2003) 3. B. Stegmayr, J.D. Lalau, O. Johnson, Transfus. Apher. Sci. 30 (2004) 213. Gilbert R. Thompson, Atherosclerosis 167 (2003) 1. M. Bláha, M. Cermanova, V. Blaha, P. Jarolim, Atherosclerosis 197 (2008) 264. A. Denizli, J. Chromatogr. B 772 (2002) 357. H. Yavuz, A. Denizli, J. Biomat. Sci. Polym. Edn. 14 (2003) 395.
1083
[11] H. Borberg, LDL-Apheresis and LDL-Elimination Therapies, European School of Haematology, 2nd Seminar of Applied Technology in Haemapheresis (Proceedings), ESH, Paris, 1997, pp. 1–8. [12] S.N. Pokrovsky, A.V. Sussekov, O.I. Afanasieva, I.Y. Adamova, A.A. Lyakishev, V.V. Kukharchuk, Chem. Phys. Lipids 6 (7/68) (1994) 323. [13] P. Batalla, M. Fuentes, V. Grazu, C. Mateo, R. Fernandez-Lafuente, J.M. Guisan, Biomacromol. 9 (2008) 719. [14] D. Josic, W. Hofmann, R. Habermann, J.D. Schlzke, W. Peutter, J. Clin. Chem. Clin. Biochem. 26 (1988) 559. [15] B. Lu, M.R. Smyth, R. O'Kennedy, Analyst 121 (1996) 29R. [16] J. Turkova, J. Chromatogr. B 722 (1999) 11. [17] Z. Ma, S. Ramakrishna, J. Membr Sci. 319 (2008) 23. [18] K. Thommes, M.R. Kula, Biotechnol. Prog. 11 (1995) 357. [19] T.B. Tennikova, B. Reusch, J. Chromatogr. A 1065 (2005) 13. [20] N. Bereli, M. Andaç, G. Baydemir, R. Say, I.Y. Galaev, A. Denizli, J. Chromatogr. A 1190 (2008) 18. [21] L. Uzun, R. Say, A. Denizli, React. Functl. Polym. 64 (2005) 93. [22] V.I. Lozinsky, I.Y. Galaev, F.M. Plieva, I.N. Savina, H. Jungvid, B. Mattiasson, Trends Biotechnol. 21 (2003) 445. [23] P. Arvidsson, F.M. Plieva, V.I. Lozinsky, I.Y. Galaev, B. Mattiasson, J. Chromatogr. A 986 (2003) 275. [24] V.I. Lozinsky, F.M. Plieva, I.Y. Galaev, B. Mattiasson, Bioseparation 10 (2002) 163. [25] W. Noppe, F.M. Plieva, K. Vanhoorelbeke, H. Deckmey, M. Tuncel, A. Tuncel, I.Y. Galaev, B. Mattiasson, J. Biotechnol. 131 (2007) 293. [26] G. Baydemir, N. Bereli, M. Andaç, R. Say, I.Y. Galaev, A. Denizli, Colloid. Surf. B 68 (2009) 33. [27] F. Yilmaz, N. Bereli, H. Yavuz, A. Denizli, Biochem. Eng. J. 43 (2009) 272. [28] İ. Koç, N. Bereli, G. Baydemir, I.Y. Galaev, A. Denizli, Hacettepe J. Biol. Chem. 36 (2008) 61. [29] N. Tüzmen, N. Bereli, M. Andaç, S. Akgöl, A. Denizli, Hacettepe J. Biol. Chem. 36 (2008) 9. [30] P. Arvidsson, F.M. Plieva, I.N. Savina, V.I. Lozinsky, S. Fexby, L. Bülow, I.Y. Galaev, B. Mattiasson, J. Chromatogr. A 977 (2002) 27. [31] M. Barut, A. Podgornik, P. Bryne, A. Strancar, J. Sep. Sci. 28 (2005) 1876. [32] C. Babaç, H. Yavuz, I.Y. Galaev, E. Pişkin, A. Denizli, React. Funct. Polym. 66 (2006) 1263. [33] J. Black, Biological Performance of Materials: Fundamentals of Biocompatibility, 3rd Ed, Marcel Dekker Press, 1999. [34] A. Denizli, J. Appl. Polym. Sci 74 (1999) 655. [35] M. Page, R. Thorpe, in: J.M. Walker (Ed.), The Protein Protocols Handbook, Humana Pres Inc, Totowa, NJ, 1996, p. 739. [36] R. Lindmark, K. Thoren-Tolling, J. Sjöquist, J. Immunol. Methods 62 (1983) 1. [37] G.T. Hermanson, A.K. Mallia, P.K. Smith, Immobilized Affinity Ligand Techniques, Academic Pres, New York, 1992. [38] H. Yavuz, A. Denizli, Macromol. Biosci. 5 (2005) 39. [39] E. Boschetti, D. Judd, W. Schwartz, P. Tunon, Daily Biotech. Updates 20 (2000) 13. [40] A. Tscheliessnig, A. Jungbauer, J. Chromatogr. A 1216 (2009) 2676. [41] A. Denizli, E. Pişkin, J. Chromatography B 670 (1995) 157. [42] A. Tabak, N. Lotan, A. Sideman, A. Tzipiniuk, G. Brook, Life Support Syst. 4 (1986) 355. [43] R.E. Ostlund, Artif. Organs 11 (1987) 366. [44] G. Smolik, B. Schmidt, U. Baurmeister, H.J. Gurland, Artif. Organs 13 (1988) 368. [45] B. Schmidt, Ex-vivo Evaluation of Biocompatibility, Presentation at the Xth International Symposium on Hemoperfusion and Immobilized Reactants, Rome, 1990. [46] Y.M. Lopukhin, V.V. Zuevsky, S.S. Markin, E.S. Nalivaiko, A.N. Rabovsky, Biomat. Artif. Cells Artif. Organs 18 (1990) 571. [47] S.N. Pokrovsky, I.Y. Adamova, G.F. Benelovenskaya, Biomat. Artif. Cells Artif. Organs 18 (1990) 623. [48] V.V. Sinitsyn, G.A. Konovalov, V.V. Kukharchuk, Biomater. Artif. Cells Artif. Organs 18 (1990) 629. [49] S. Wang, C.M. Li, X. Guo, Y. Yu, W. Wang, React. Funct. Polym. 68 (2008) 261. [50] K. Utsumi, M. Kawabe, A. Hirama, K. Ueda, Y. Kamada, K. Arii, Y. Komaba, K. Katsura, Y. Iino, Y. Katayama, Clin. Chim. Acta 377 (2007) 198. [51] P.J. Soltys, M.R. Etzel, Biomaterials 21 (2000) 37. [52] S. Wang, Y. Yu, T. Cui, Y. Cheng, Biomaterials 24 (2003) 2799. [53] B. Sellergren, J. Wieschemeyer, K.S. Boos, D. Seidel, Langmuir 10 (1998) 4037. [54] H. Yu, G. Fu, B. He, Cellulose 14 (2007) 99. [55] H. Yavuz, V. Karakoç, D. Türkmen, R. Say, A. Denizli, Int. J. Biol. Macromol. 41 (2007) 8.
Contents lists available at ScienceDirect
Materials Science and Engineering C j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / m s e c
Oriented immobilized anti-LDL antibody carrying poly(hydroxyethyl methacrylate) cryogel for cholesterol removal from human plasma Nilay Bereli a, Gülsu Şener b, Handan Yavuz a,⁎, Adil Denizli a a b
Department of Chemistry, Hacettepe University, Beytepe, Ankara, Turkey Nanotechnology and Nanomedicine Division, Hacettepe University, Ankara, Turkey
a r t i c l e
i n f o
Article history: Received 13 April 2009 Received in revised form 22 October 2010 Accepted 15 March 2011 Available online 29 March 2011 Keywords: Cholesterol removal Cryogels Oriented immobilization Protein A
a b s t r a c t Low density lipoprotein (LDL) cholesterol is a major ingredient of the plaque that collects in the coronary arteries and causes coronary heart diseases. Among the methods used for the extracorporeal elimination of LDL from intravasal volume, immunoaffinity technique using anti-LDL antibody as a ligand offers superior selectivity and specificity. Proper orientation of the immobilized antibody is the main issue in immunoaffinity techniques. In this study, anti-human β-lipoprotein antibody (anti-LDL antibody) molecules were immobilized and oriented through protein A onto poly(2-hydroxyethyl methacrylate) (PHEMA) cryogel in order to remove LDL from hypercholesterolemic human plasma. PHEMA cryogel was prepared by free radical polymerization initiated with N,N,N′,N′-tetramethylene diamine (TEMED). PHEMA cryogel with a swelling degree of 8.89 g H2O/g and 67% macro-porosity was characterized by swelling studies, scanning electron microscope (SEM) and blood compatibility tests. All the clotting times were increased when compared with control plasma. The maximum immobilized anti-LDL antibody amount was 63.2 mg/g in the case of random antibody immobilization and 19.6 mg/g in the case of oriented antibody immobilization (protein A loading was 57.0 mg/g). Random and oriented anti-LDL antibody immobilized PHEMA cryogels adsorbed 111 and 129 mg LDL/g cryogel from hypercholesterolemic human plasma, respectively. Up to 80% of the adsorbed LDL was desorbed. The adsorption– desorption cycle was repeated 6 times using the same cryogel. There was no significant loss of LDL adsorption capacity. © 2011 Elsevier B.V. All rights reserved.
1. Introduction LDL cholesterol is widely recognized as one of the major risk factors in the development of coronary heart diseases because of its ability to build up in the lining of arteries forming atheromas and fatty acid deposits [1–3]. The causative factor in high levels of LDL can be of genetic reasons, such as familial hypercholesterolemia [4,5]. A majority of the hypercholesterolemic patients can be effectively treated by reduced dietary intake and drug therapy to control plasma cholesterol levels, preventing severe heart diseases. However, a more aggressive approach is necessary in severe hypercholesterolemia [6]. Since the first plasmapheresis procedure implemented for homozygous familial hypercholesterolemia; a number of additional methods with different specificity and selectivity have been developed for the extracorporeal removal of LDL from the blood, including cascade filtration, heparin induced extracorporeal LDL precipitation (HELP), thermo-filtration, dextran induced LDL precipitation and direct adsorption of lipoproteins (DALI) [7]. Extracorporeal elimination is an effective and life saving procedure in hypercholesterolemia patients [8]. Of
⁎ Corresponding author. E-mail address: [email protected] (H. Yavuz). 0928-4931/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.msec.2011.03.008
course the most selective binding affinity can be achieved by using immunoaffinity adsorbents containing antibodies as the ligand [9–12]. For instance; in case of high LDL levels in patients, antibodies against apoprotein B100, the main protein component of LDL, are used for the selective extracorporeal removal of harmful LDL without depriving the patients of the useful HDL and other blood components. Antibody immobilization may be a critical step in the design of immunoaffinity adsorbents [13]. When antibodies are covalently immobilized onto adsorbents, their specific binding capacity usually decreases as compared to soluble antibodies [14]. This reduction is attributed to the random orientation of the antibodies on the surface of the solid support used, which disables the accessibility of the antigen binding fragments of the antibodies (Fab) towards their antigens. In normal coupling procedures, the target antibodies are immobilized at many different spots resulting random orientation of the antibodies on the matrix that might prevent the formation of the antibody–antigen complex [15]. More specific immobilization of antibodies through the Fc region of the heavy immunoglobulin chain would be very efficient in terms of providing orderly oriented immobilized antibody so that the antibody binding sites, located on the opposite part of the immobilization sites, become highly accessible to antigen binding. Protein A from Staphylococcus aureus has a unique capability of binding mammalian
N. Bereli et al. / Materials Science and Engineering C 31 (2011) 1078–1083
immunoglobulins through the Fc region [16]. Thus protein A can be employed for the orderly oriented immobilization of antibodies by formation of a protein A immobilized matrix followed by cross-linking of antibody with protein A in the matrix. The advantages of the orderly oriented immobilization process for biologically active proteins can be outlined as good binding accessibility and increased stability [16]. In biological applications, the conventional column chromatography techniques primarily use gel-beads with certain limitations such as high pressure drops and low flow-rates, lowering the efficiency in routine and scale-up applications [17]. In addition, the gel-bead column chromatography is not capable of applying highly viscous fluids such as human blood. Therefore, alternative chromatography techniques such as membranes, monoliths and cryogels [18–21] are applied in cases when gel-bead column chromatography techniques fail. Cryogels provide a potential solution in terms of their low pressure drop and lack of diffusion resistances utilizing macropores as compared to the traditional gel-bead columns [22–24]. Cryogel columns enable high flow-rates allowing voluminous elutions within shorter times. Whole blood can be applied on cryogel columns without any pre-treatment [25–29]. Cryogels are also cheap and thus they can be easily disposed of, eliminating cross-contamination between batches [30]. The PHEMA cryogels were selected for three good reasons: 1) They exhibit a low pressure drop, 2) they lack diffusion resistance and 3) viscous samples such as whole blood can be easily applied on them. We have focused our attention on the development of anti-LDL antibody immobilized immuno-affinity PHEMA cryogels by combining the selectivity of immunoaffinity interaction with the biocompatibility and good flow properties of PHEMA cryogels. LDL removal performance of these immunoaffinity cryogels was reported here. 2. Experimental 2.1. Materials Anti-LDL antibody (anti-human β-lipoprotein; Product No: L-8016) and protein A (from S. aureus, Cowan Strain I; Product No: P-6031) were obtained from Sigma (St. Louis, USA). 2-Hydroxyethyl methacrylate (HEMA) was obtained from Fluka A.G. (Buchs, Switzerland), which was later distilled under reduced pressure in the presence of hydroquinone inhibitor and stored at 4 °C until use. Cyanogen bromide (CNBr), N,N′methylene-bis(acrylamide) (MBAAm), ammonium persulfate (APS) and N,N,N′,N′-tetramethylene diamine (TEMED) were also obtained from Sigma. All other chemicals were of reagent grade and were purchased from Merck AG (Darmstadt, Germany). Water used in adsorption experiments was purified by a Barnstead (Dubuque, IA) ROpure LP® reverse osmosis unit having a high flow cellulose acetate membrane (Barnstead D2731) followed by a Barnstead D3804 NANOpure® organic/colloid removal and ion exchange packed-bed systems. 2.2. Preparation of PHEMA cryogels Preparation of cryogel is described elsewhere [20]. Briefly, monomers (1.3 mL of HEMA, and 10 mL of MBAAm) were dissolved in deionized water and the mixture was degassed under vacuum for about 5 min to remove soluble oxygen, yielding a total monomer concentration of 6% (w/v). The cryogel was prepared by free radical polymerization initiated by firstly addition of 20 mg of APS, 1 % (w/v) of the total monomers, into the solution of the monomers followed by cooling in an ice bath for 2–3 min, and then addition of 25 μL of TEMED, 1% (w/v) of the total monomers, which was stirred for 1 min. Then, the reaction mixture was poured into a plastic syringe (5 mL, internal diameter: 0.8 cm) with closed outlet at the bottom. The polymerization solution in the syringe was frozen at −12 °C for 24 h, which was then thawed at room temperature. After washing with 200 mL of water, the cryogel
1079
column was stored in a buffer solution containing 0.02% sodium azide at 4 °C until use. 2.3. CNBr activation As adsorber materials anti-LDL antibody was used for random immobilization while protein A was used for orderly oriented immobilization. The PHEMA cryogel column was activated using cyanogen bromide in order to create reactive sites for protein immobilization, with either anti-LDL antibody or protein A. Prior to the activation process, the PHEMA cryogel column was kept in distilled water for about 24 h, which was then washed by passing a 0.5 M NaCl solution and 50 mL of 0.5 M sodium carbonate buffer solution (pH 10.5). An aqueous solution of CNBr (50 mg CNBr/mL) was carefully prepared in a fume hood and its pH was quickly adjusted to 11.5 using a 4 M NaOH solution and a pH meter (Mettler Toledo Gmbh, Switzerland). The pH of the CNBr solution was maintained between 10.5 and 11.5 during the activation reaction. The CNBr solution was passed from the cryogel column with a flow rate of 1.0 mL/min over 2 h at room temperature. After the activation reaction, the column was washed with cold sodium citrate buffer (0.1 M; pH 6.5) and the filtrate was discarded. 2.4. Random immobilization of anti-LDL antibodies 20 mL of an anti-LDL antibody solution (1.0 mg/mL) dissolved in 0.1 M sodium citrate buffer (pH 6.5) passed through a freshly activated PHEMA cryogel column at a 1.0 mL/min flow rate at 4 °C overnight. After immobilization, 2.0 M ethanol amine solution was recirculated for another 1 h in order to quench side reactions that may form between unreacted sites of the protein and any remaining active groups (e.g. isourea) on the cryogel surface. The amount of anti-LDL antibody immobilization was monitored by measuring the decrease in the protein concentration in the anti-LDL antibody solution by Bradford method. Non-specifically adsorbed anti-LDL antibody molecules on PHEMA were also considered. The amount of immobilized anti-LDL antibody was calculated as: q = ½ ðCi –Ct Þ:V = m
ð1Þ
where, q is the amount of anti-LDL antibody immobilized onto unit mass of the cryogel (mg/g); Ci and Ct are the concentrations of the anti-LDL antibody in the initial solution and in the supernatant after immobilization (mg/mL), respectively, V is the volume of the aqueous phase (mL), and m is the mass of the cryogel column (g). 2.5. Oriented immobilization of anti-LDL antibodies Covalent immobilization of protein A onto freshly activated PHEMA cryogels was performed using the same procedure described in Section 2.4. The CNBr and protein A concentrations were 50 mg/mL and 1.0 mg/mL, respectively. The immobilization (0.1 M sodium citrate) buffer solution used was adjusted at pH 6.5. The amount of protein A immobilized onto the CNBr activated PHEMA cryogel was determined by measuring the decrease in the protein A concentration by the Bradford method, which was calculated with Eq. 1. The protein A-PHEMA cryogel was washed several times using 0.1 M glycine–HCl solution (pH 3.5) and the immobilization buffer in order to remove impurities. The anti-LDL antibody solution was prepared in a borate buffer solution (1.0 mg/mL, pH 7.5). The protein A immobilized PHEMA cryogel column was treated with the anti-LDL antibody solution for 2 h at room temperature to form an antibody protein A complex. Then, protein A was cross-linked to the adsorbed anti-LDL antibody by adding a solution of 0.1% (w/v) cyanamide (CH2N2), prepared in the immobilization buffer solution, into the protein A immobilized column and the reaction was carried out for about 30 min. After the reaction, the column was washed several times with a solution of 0.1 M NaCl and
1080
N. Bereli et al. / Materials Science and Engineering C 31 (2011) 1078–1083
distilled water. The affinity adsorbent was stored in 0.02% (w/v) NaN3 at 4 °C until use. 2.6. Characterization of PHEMA cryogels 2.6.1. Swelling degree The swelling degree (S) of the cryogel was determined as follows: a cryogel sample was taken out of its column and rinsed with ethyl alcohol and distilled water until a clear supernatant was obtained. The cryogel was squeezed to remove unbound water residues as much as possible, which was then transferred to pre-weighed vial and weighed out (mwet gel). The wet cryogel was dried in an oven at 60 °C, which was then weighed out to determine the mass of dried sample (mdry gel). The swelling degree was calculated as: S = mwet gel –mdry gel = mdry gel :
ð2Þ
The total volume of macropores in the swollen cryogel was roughly estimated as follows. The cryogel sample was weighed out after it was swelled in 40 mL of water (mswollen gel) and then the weight of the sample was determined after squeezing (msqueezed gel). The porosity was calculated as: Porosity % = mswollen gel –msqueezed gel = mswollen gel × 100%:
ð3Þ
cholesterol in the plasma after incubation with anti-LDL antibody containing adsorbent. The amounts of the total cholesterol (TC), LDL cholesterol (LDL), HDL cholesterol (HDL), VLDL cholesterol (VLDL) and triglycerides were determined by homogeneous enzymatic colorimetric assay on an automated analyzer instrument (Hitachi 912, Japan). 2.8. Desorption and repeated use Adsorbed LDL was desorbed using a desorption solution of 0.1 M citric acid and 0.02 M Na2HPO4, (pH 3.0). In a typical desorption experiment, 50 mL of the desorption solution was pumped through the cryogel column at a flow rate of 1.0 mL/min for 1 h. The final concentration of LDL in the solution was determined by the same procedure described in Section 2.7. The desorption ratio corresponds to the final LDL concentration in the desorption medium over the amount of LDL adsorbed on the cryogel. In order to test the reusability of anti-LDL antibody containing PHEMA cryogel, the LDL adsorption–desorption procedure was repeated 6 times on a single cryogel column. The PHEMA cryogel was washed with 50 mM NaOH solution after each adsorption–desorption cycle to ensure regeneration and sterilization. 3. Results and discussion 3.1. Characterization of PHEMA cryogel
2.6.2. SEM studies The morphology of a cross section of the dried cryogel was investigated by scanning electron microscopy (SEM). The sample was mounted in a solution of 2.5% glutaraldehyde in 0.15 M sodium cacodylate buffer overnight, which was then kept in a solution of 1% osmium tetroxide for 1 h. Then the sample was dehydrated stepwise in ethanol and transferred to a critical point drier at 10 °C where the ethanol was replaced with liquid carbon dioxide as the transitional fluid. The temperature and pressure of the system were then raised to 40 °C and 100 bar, respectively. Liquid CO2 was transformed directly to gas uniformly throughout the whole sample without heat of vaporization or surface tension forces causing damage. Release of the pressure at a constant temperature of 40 °C resulted in dried cryogel sample. The cryogel was finally coated with gold-palladium (40:60) and examined using a JEOL JSM 5600 scanning electron microscope. 2.6.3. Biocompatibility studies The human blood PHEMA cryogel interaction was monitored by measuring the following parameters; prothrombin time (PT), activated partial thromboplastin time (APTT), fibrinogen time and thrombocyte clotting time. The in vitro hemocompatibility tests were implemented by STA4 Compact Blood Coagulation Analyzer (Diagnostica Stago, France) using the required test kits. 2.7. LDL removal studies The efficiency of the LDL removal procedure was tested in vitro in a continuous experimental setup. The plasma samples with an average cholesterol initial concentration of 431.0 mg/dL were obtained from patients with hypercholesterolemia. Fresh frozen plasma was donated by the Blood Bank at the Mesa Hospital (Ankara). Blood samples were centrifuged at 500 g for 30 min at room temperature to separate plasma. Plasma was filtered using 0.45 μm syringe filters (Model 245-0045 Nalge Co., Rochester, New York), and stored at 4 °C. Sodium azide (0.1% w/v) was added to prevent bacterial growth. Plasma (20 mL) was recirculated from the PHEMA, PHEMA-anti-LDL antibody and PHEMAprotein A-anti-LDL antibody cryogel columns at 20 °C for 4 h. The amount of the removed cholesterol was calculated by Eq. 1, using the initial amount of cholesterol in the plasma and the amount of
A supermacroporous PHEMA cryogel was prepared by copolymerization of HEMA and MBAAm in the frozen state in the presence of APS/TEMED. The SEM image of the pore structure of the cryogel is shown in Fig. 1. The macropores in the PHEMA cryogel bulk structure are open and highly interconnected, forming a network of large channels. The mobile protein phase is forced to flow through these channels, transporting the biomolecules to the active binding sites by convection, which results in an extremely fast mass exchange between the mobile protein phase and the stationary cryogel phase in regards to the traditional packed bed columns [31,32]. The macroporosity and equilibrium swelling degree of the PHEMA cryogels were 67% and 8.89 [g H2O/g cryogel], respectively. The PHEMA cryogel is white, opaque, spongelike and elastic, which can be easily squeezed by hand to remove water accumulated inside the macropores. The squeezed cryogel is able to restore its original size and shape within 1–2 s when soaked in water. Ideally, the pressure drop needed to drive the liquid through any systems should be as low as possible. The pressure drop tests on the PHEMA cryogel columns were performed in water which was used as an equilibration medium. The water phase was passed through the column for 1 min at each incremented linear flow rate ranging from 40 to 382 cm/h. Due to the presence of large and highly interconnected macropores, the PHEMA cryogel column was found to have a very low liquid flow resistance. In addition the PHEMA cryogel column was found to possess a low back pressure and low liquid flow resistance indicating the presence of porous structure with large interconnected macropore structures, as also seen in the SEM picture in Fig. 1. The physicochemical properties determined for the PHEMA cryogel are presented in Table 1. In general cryogels with supermacroporous structures allow direct processing of whole blood containing blood cells [25,26]. In a similar vein, the macropore size of the PHEMA cryogel prepared in our laboratory was tested with whole blood and was found to be very suitable to process blood cells without ever blocking the column. Table 2 briefly outlines the coagulation data obtained in the blood compatibility tests. As seen in Table 2, the clotting times determined for the PHEMA and PHEMA-anti-LDL antibody cryogels are within a comparable range of values for the control plasma. Observed decreases in coagulation times are generally tolerated by the body [33]. Therefore, we conclude that the blood application of PHEMA and anti-LDL antibody
N. Bereli et al. / Materials Science and Engineering C 31 (2011) 1078–1083
1081
Table 2 Coagulation times of human plasma (reported in s)⁎. Material
Control PHEMA PHEMA/anti-LDL antibody
Parameter PT
APTT
Fibrinogen time
Thrombocyte clotting time
12.2 ± 0.8 12.5 ± 0.9 12.4 ± 0.8
27.8 ± 1.5 26.2 ± 1.3 29.3 ± 1.5
14.1 ± 0.9 12.1 ± 0.9 18.3 ± 0.9
15.0 ± 0.8 16.6 ± 1.0 16.4 ± 1.0
⁎ Each result is an average of three determinations.
Fig. 1. SEM images of the PHEMA cryogel.
immobilized cryogels becomes acceptable with reproducible clotting times in regards to literature values [34]. 3.2. Random immobilization of anti-LDL antibodies Anti-LDL antibody was used as a bioaffinity ligand for the removal of LDL. It should be noted that, immobilization at pH 6.5 is less efficient than higher pH values, but it's less likely to compromise the binding ability of immobilized ligands (especially antibodies) [35]. Non-specific adsorption of an anti-LDL antibody ligand onto plain PHEMA cryogel was 0.17 mg anti-LDL antibody/g, which makes the PHEMA cryogel a potential carrier of the ligand. The amount of the anti-LDL antibody immobilized on the CNBr activated PHEMA cryogel was 63.2 mg antibody/g cryogel as determined in a media containing 50 mg/mL CNBr and 1.0 mg/mL anti-LDL antibody. The performance of the adsorbent is enhanced by the inertness of the cryogel surface and the specificity of the antibody. 3.3. Oriented immobilization of anti-LDL antibodies Some strains of S. aureus biosynthesize protein A, a 40 kDa protein that specifically binds to the Fc region of some antibodies with high
Table 1 Physicochemical properties of the PHEMA cryogel. Specific surface area Pore size diameter Porosity Swelling degree Back pressure
20.2 m2/g 10–200 μm 67% 8.89 g H2O/g cryogel 0.28 MPa
affinity [36,37], except for human IgG3, mouse IgG3, sheep IgG1, and also binds weakly to mouse IgG1. For some species, antibodies do not bind to protein A at all and some monoclonal antibodies show abnormal affinity for the protein A. These properties make the use of protein A advantageous for antibody separation, purification and also oriented immobilization through its Fc region as in this study. Since the IgG-protein A interactions are well known to be strongly pH dependent, the pH effects were mainly examined with three different buffer systems in our previous study [38]. Binding of anti-LDL antibody molecules to protein A occurs through an induced hydrophobic fit and it is promoted by addition of salts. At the center of the Fc binding site as well as on protein A reside histidine residues. At alkaline pH, these residues are uncharged and hydrophobic, strengthening the interaction between protein A and the antibody. As the pH is shifted to acidic values, these residues become charged and repel each other [39]. Ionic strength of buffers was adjusted by adding NaCl (0.15 M). Multiple uses of the immunoadsorbents require the crosslinkage of antibody with protein A through a covalent bond. For this reason, 0.1% (w/v) carbodiimide solution prepared in the same binding buffer was added after the binding reaction has completed. Protein A is covalently immobilized through the active sites on PHEMA cryogel. The amount of immobilized protein A was found as 57 mg/g at 50 mg/mL CNBr and 1.0 mg/mL protein A concentration. The binding solution was 0.1 M sodium citrate buffer (pH 6.5). Anti-LDL antibody binding was achieved with borate buffer at a pH of 7.5 (containing 0.15 M NaCl). The amount of anti-LDL antibody immobilization was 19.6 mg/g. It should be noted that protein A molecules have five Fc binding receptors, three of these receptors are reported to become inactive when the molecule is immobilized. 3.4. Cholesterol removal from human plasma Fig. 2 shows the change of cholesterol concentration in the hypercholesterolemic plasma in the course of incubation time. As seen in these figures, LDL levels were significantly reduced when the PHEMAanti-LDL antibody and PHEMA-protein A-anti-LDL antibody immobilized cryogels were used (i.e., 54% and 59% of the initial LDL, respectively). It should be noted that adsorption equilibrium is reached in nearly half an hour. Adsorption time is especially important for the patient's comfort when the system is thought to be used in extracorporeal treatment. Optimum equilibrium time for the bead based adsorbents is in the range of 3–6 h [1]. Big molecules, such as plasma proteins, diffuse slowly through a particulate support, and therefore longer contact times of the molecule with the adsorbent are required. Because of their favorable mass transfer mechanisms, continuous porous cryogel requires shorter contact times and therefore shorter run times [40]. As seen in these figures, anti-LDL antibody immobilized through protein A (i.e., oriented immobilized) was more efficient than randomly immobilized anti-LDL antibody in removing cholesterol from hypercholesterolemic human plasma, although the immobilized amount of anti-LDL antibody through the protein A was nearly one third of the randomly immobilized anti-LDL antibody onto PHEMA cryogel. This shows the mode of oriented immobilization makes the binding site of antibody always located on the top of its Fab variable regions, well accessible for interaction with LDL molecules. So, we can
1082
N. Bereli et al. / Materials Science and Engineering C 31 (2011) 1078–1083
Fig. 2. Cholesterol removal from hypercholesterolemic human plasma with PHEMA, PHEMA-anti-LDL antibody and PHEMA protein A-anti-LDL antibody cryogels. Total cholesterol concentration: 431 mg/dL; LDL concentration: 295 mg/dL; anti LDL antibody loading: 63.2 mg/g (random), 19.6 mg/g (oriented); T: 20 °C; flow rate: 1.0 mL/min.
say that the binding of antibody molecules through their Fc region was effective. Fig. 3 shows the comparison of LDL and other lipoprotein adsorption amounts of random and oriented anti-LDL antibody immobilized PHEMA cryogels. A support to be used as an affinity adsorbent must be physically inert to prevent false positives and to increase system performance. Our strategy is to develop an affinity adsorbent that selectively adsorbs LDL with low non specific removal of other (lipo) proteins, especially of the protective HDL and other blood proteins. HDL, other lipoproteins and triglyceride adsorption onto these adsorbents was also investigated. As can be seen from this figure, relatively small amounts of HDL, VLDL and triglycerides were adsorbed on both, random and oriented, anti-LDL antibody immobilized adsorbents. Selective adsorption of LDL with low non-specific removal of other (lipo)proteins, especially of the protective HDL; sufficient cholesterol adsorption capacity; and excellent whole blood compatibility allow cholesterol adsorption from whole blood. As can be also seen from the figure, bare PHEMA cryogel adsorbed negligible amount of lipoproteins indicating that the adsorption of lipoproteins onto the immunoaffinity adsorbents occurred due to the existence of the anti-LDL antibody ligand on the PHEMA surface. Different affinity adsorbents with different adsorption capacities were reported in literature for cholesterol removal studies. Denizli
reported 4.5–7.2 mg cholesterol/mL adsorption capacity with antiLDL antibody immobilized poly(HEMA-MAPA) membrane [9]. Yavuz and Denizli have reported 13.3–16.0 mg cholesterol/g adsorption capacity with anti-LDL immobilized poly(HEMA-MAPA) microbeads [10]. Yavuz and Denizli removed 39% and 47% of the total cholesterol with random and oriented anti-LDL antibody immobilized poly (HEMA-EGDMA) beads, respectively [38]. Denizli and Pişkin studied cholesterol removal from hypercholesterolemic human plasma and they reported maximum 4.7 mg cholesterol/g adsorption capacity with low-molecular weight heparin immobilized poly(2-hydroxyethyl methacrylate) microspheres [41]. Tabak et al. used heparin immobilized agarose beads and they obtained in the range of 1.2–3.0 mg/g cholesterol adsorption capacity from healthy human plasma [42]. Ostlund reported 6–8 mg lipoprotein cholesterol per mL column volume with anti-LDL antibody immobilized commercially available Agarose beads from human plasma [43]. Smolik et al. interested in determining the LDL adsorption capacity of dextran sulfate immobilized cellulose affinity beads [44]. They achieved a good selective LDL adsorption from whole blood and presented cholesterol adsorption capacities of 5–10 mg/mL gel. Schmidt used surface modified polyacrylate matrix and he obtained 8–10 mg cholesterol per gram polymer [45]. Lopukhin et al. used macroporous silica beads as the carrier matrix, and immobilized heparin and chytozane sulfate as specific bioligand. They reported cholesterol adsorption capacities around 14.8–15.2 mg/g [46]. Pokrovsky et al. used commercial carrier made of Sepharose carrying monoclonal and polyclonal antibodies as the specific ligands [47]. Their maximum LDL binding capacities were in the range of 0.6–2.5 mg/mL adsorbent. Sinitsyn et al. used Sepharose beads carrying heparin and they reported adsorption values of up to 12 mg cholesterol per gram polymer [48]. Wang et al. prepared L-lysine containing cellulose for electrostatic interaction with LDL and reached up to 60.9% adsorption percentage [49]. Utsumi et al. reduced LDL selectively to sufficient levels in hypercholesterolemia by dextran sulfate columns [50]. Soltys and Etzel prepared immunoaffinity membranes with anti-apo B IgG immobilized through PEG spacer and reached 1.0 mg/mL abo B adsorption capacity [51]. Wang et al prepared dextran adsorbent with amphiphilic ligands for adsorption of low-density lipoprotein and reached 1.916 mg/mL adsorption capacity [52]. Sellergren et al. prepared a series of highly cross-linked molecular imprinted terpolymers in the presence of cholesterol acting as a template molecule [53]. Using a physiological relevant intestinal-mimicking solution of cholesterol, these polymers adsorbed ca. 17 mg cholesterol per gram dry adsorbent. Yu et al. developed a new LDL-adsorbent by using graft polymerization technique [54]. Yavuz et al. prepared cholesterol imprinted poly (HEMA-MAT) particles and obtained 16.23 mg/g cholesterol adsorption capacity [55]. In this study, we demonstrated that random and oriented anti-LDL antibody immobilized PHEMA adsorbents showed excellent LDL adsorption capacities (129 mg/g). 3.5. Desorption and reusability
Fig. 3. Lipoprotein adsorption amounts of random and oriented anti-LDL antibody immobilized PHEMA cryogel. Total cholesterol concentration: 431 mg/dL; anti LDL antibody loading: 63.2 mg/g (random), 19.6 mg/g (oriented); T: 20 °C; flow rate: 1.0 mL/min; adsorption time: 2 h.
Desorption of LDL molecules from the randomly and orderly oriented anti-LDL antibody immobilized PHEMA cryogels was also studied by recirculating a desorption medium through an LDL adsorbed PHEMA cryogel and measuring the amount of the LDL desorbed in 1 h. The desorption buffer solution consisted of 0.04 M citric acid and 0.02 M Na2HPO4 at pH 3.0. Hypercholesterolemic human plasma was used for repeated cycles of cholesterol adsorption–desorption. It was found that up to 80% of the adsorbed cholesterol molecules were desorbed (Fig. 4). Note that the anti-LDL antibody was not released during the desorption processes, indicating that both the randomly and orderly oriented antibody molecules, as the ligands, were covalently immobilized onto the PHEMA cryogels. It is likely that the citric acid buffer solution alters the tertiary structure of the anti-LDL antibody, weakening the non-covalent bonds at the binding interface, which
N. Bereli et al. / Materials Science and Engineering C 31 (2011) 1078–1083
Fig. 4. Reusability of anti-LDL antibody-PHEMA and anti-LDL antibody-protein A-PHEMA cryogel.
ultimately leads to desorption of LDL. The adsorption–desorption cycle of LDL was repeated 6 times using the same cryogel column. Based on the desorption data presented in Fig. 4, it is concluded that citric acid is a suitable desorption agent for the removal of the LDL from anti-LDL immobilized PHEMA columns, and allows repeated use of the column. 4. Conclusion The results presented in this study reveal that PHEMA and anti-LDL antibody immobilized PHEMA cryogels are biocompatible and that the anti-LDL antibody immobilized PHEMA cryogels developed in our laboratory can be selectively used for the extracorporeal removal of LDL out of hypercholesterolemic human plasma. The maximum anti-LDL antibody immobilization amount was 63.2 mg/g in the case of random immobilization and 19.6 mg/g in the case of oriented immobilization. Random and oriented anti-LDL antibody immobilized PHEMA cryogels adsorbed 111 and 129 mg LDL/g cryogel from hypercholesterolemic human plasma, respectively, which are sufficient values to reduce blood LDL to an acceptable level. Up to 80% of the adsorbed LDL was desorbed and there was no significant loss of LDL adsorption capacity after 6 adsorption–desorption cycles. References [1] [2] [3] [4]
[5] [6] [7] [8] [9] [10]
G.R. Thompson, Atherosclerosis 167 (2003) 1. E. Bruckert, Transfus. Sci. 20 (1999) 43. N. Tani, Ther. Apher. 4 (2) (2000) 135. WHO, Human Genetics Programme. Division of Noncommunicable Diseases, Familial Hyperchoesterolemia (FH): Report of WHO Consultation. WHO/HGN/FH/ CONS/98, 7, 1998, p. 1. M. Bláha, Acta Med. 46 (2003) 3. B. Stegmayr, J.D. Lalau, O. Johnson, Transfus. Apher. Sci. 30 (2004) 213. Gilbert R. Thompson, Atherosclerosis 167 (2003) 1. M. Bláha, M. Cermanova, V. Blaha, P. Jarolim, Atherosclerosis 197 (2008) 264. A. Denizli, J. Chromatogr. B 772 (2002) 357. H. Yavuz, A. Denizli, J. Biomat. Sci. Polym. Edn. 14 (2003) 395.
1083
[11] H. Borberg, LDL-Apheresis and LDL-Elimination Therapies, European School of Haematology, 2nd Seminar of Applied Technology in Haemapheresis (Proceedings), ESH, Paris, 1997, pp. 1–8. [12] S.N. Pokrovsky, A.V. Sussekov, O.I. Afanasieva, I.Y. Adamova, A.A. Lyakishev, V.V. Kukharchuk, Chem. Phys. Lipids 6 (7/68) (1994) 323. [13] P. Batalla, M. Fuentes, V. Grazu, C. Mateo, R. Fernandez-Lafuente, J.M. Guisan, Biomacromol. 9 (2008) 719. [14] D. Josic, W. Hofmann, R. Habermann, J.D. Schlzke, W. Peutter, J. Clin. Chem. Clin. Biochem. 26 (1988) 559. [15] B. Lu, M.R. Smyth, R. O'Kennedy, Analyst 121 (1996) 29R. [16] J. Turkova, J. Chromatogr. B 722 (1999) 11. [17] Z. Ma, S. Ramakrishna, J. Membr Sci. 319 (2008) 23. [18] K. Thommes, M.R. Kula, Biotechnol. Prog. 11 (1995) 357. [19] T.B. Tennikova, B. Reusch, J. Chromatogr. A 1065 (2005) 13. [20] N. Bereli, M. Andaç, G. Baydemir, R. Say, I.Y. Galaev, A. Denizli, J. Chromatogr. A 1190 (2008) 18. [21] L. Uzun, R. Say, A. Denizli, React. Functl. Polym. 64 (2005) 93. [22] V.I. Lozinsky, I.Y. Galaev, F.M. Plieva, I.N. Savina, H. Jungvid, B. Mattiasson, Trends Biotechnol. 21 (2003) 445. [23] P. Arvidsson, F.M. Plieva, V.I. Lozinsky, I.Y. Galaev, B. Mattiasson, J. Chromatogr. A 986 (2003) 275. [24] V.I. Lozinsky, F.M. Plieva, I.Y. Galaev, B. Mattiasson, Bioseparation 10 (2002) 163. [25] W. Noppe, F.M. Plieva, K. Vanhoorelbeke, H. Deckmey, M. Tuncel, A. Tuncel, I.Y. Galaev, B. Mattiasson, J. Biotechnol. 131 (2007) 293. [26] G. Baydemir, N. Bereli, M. Andaç, R. Say, I.Y. Galaev, A. Denizli, Colloid. Surf. B 68 (2009) 33. [27] F. Yilmaz, N. Bereli, H. Yavuz, A. Denizli, Biochem. Eng. J. 43 (2009) 272. [28] İ. Koç, N. Bereli, G. Baydemir, I.Y. Galaev, A. Denizli, Hacettepe J. Biol. Chem. 36 (2008) 61. [29] N. Tüzmen, N. Bereli, M. Andaç, S. Akgöl, A. Denizli, Hacettepe J. Biol. Chem. 36 (2008) 9. [30] P. Arvidsson, F.M. Plieva, I.N. Savina, V.I. Lozinsky, S. Fexby, L. Bülow, I.Y. Galaev, B. Mattiasson, J. Chromatogr. A 977 (2002) 27. [31] M. Barut, A. Podgornik, P. Bryne, A. Strancar, J. Sep. Sci. 28 (2005) 1876. [32] C. Babaç, H. Yavuz, I.Y. Galaev, E. Pişkin, A. Denizli, React. Funct. Polym. 66 (2006) 1263. [33] J. Black, Biological Performance of Materials: Fundamentals of Biocompatibility, 3rd Ed, Marcel Dekker Press, 1999. [34] A. Denizli, J. Appl. Polym. Sci 74 (1999) 655. [35] M. Page, R. Thorpe, in: J.M. Walker (Ed.), The Protein Protocols Handbook, Humana Pres Inc, Totowa, NJ, 1996, p. 739. [36] R. Lindmark, K. Thoren-Tolling, J. Sjöquist, J. Immunol. Methods 62 (1983) 1. [37] G.T. Hermanson, A.K. Mallia, P.K. Smith, Immobilized Affinity Ligand Techniques, Academic Pres, New York, 1992. [38] H. Yavuz, A. Denizli, Macromol. Biosci. 5 (2005) 39. [39] E. Boschetti, D. Judd, W. Schwartz, P. Tunon, Daily Biotech. Updates 20 (2000) 13. [40] A. Tscheliessnig, A. Jungbauer, J. Chromatogr. A 1216 (2009) 2676. [41] A. Denizli, E. Pişkin, J. Chromatography B 670 (1995) 157. [42] A. Tabak, N. Lotan, A. Sideman, A. Tzipiniuk, G. Brook, Life Support Syst. 4 (1986) 355. [43] R.E. Ostlund, Artif. Organs 11 (1987) 366. [44] G. Smolik, B. Schmidt, U. Baurmeister, H.J. Gurland, Artif. Organs 13 (1988) 368. [45] B. Schmidt, Ex-vivo Evaluation of Biocompatibility, Presentation at the Xth International Symposium on Hemoperfusion and Immobilized Reactants, Rome, 1990. [46] Y.M. Lopukhin, V.V. Zuevsky, S.S. Markin, E.S. Nalivaiko, A.N. Rabovsky, Biomat. Artif. Cells Artif. Organs 18 (1990) 571. [47] S.N. Pokrovsky, I.Y. Adamova, G.F. Benelovenskaya, Biomat. Artif. Cells Artif. Organs 18 (1990) 623. [48] V.V. Sinitsyn, G.A. Konovalov, V.V. Kukharchuk, Biomater. Artif. Cells Artif. Organs 18 (1990) 629. [49] S. Wang, C.M. Li, X. Guo, Y. Yu, W. Wang, React. Funct. Polym. 68 (2008) 261. [50] K. Utsumi, M. Kawabe, A. Hirama, K. Ueda, Y. Kamada, K. Arii, Y. Komaba, K. Katsura, Y. Iino, Y. Katayama, Clin. Chim. Acta 377 (2007) 198. [51] P.J. Soltys, M.R. Etzel, Biomaterials 21 (2000) 37. [52] S. Wang, Y. Yu, T. Cui, Y. Cheng, Biomaterials 24 (2003) 2799. [53] B. Sellergren, J. Wieschemeyer, K.S. Boos, D. Seidel, Langmuir 10 (1998) 4037. [54] H. Yu, G. Fu, B. He, Cellulose 14 (2007) 99. [55] H. Yavuz, V. Karakoç, D. Türkmen, R. Say, A. Denizli, Int. J. Biol. Macromol. 41 (2007) 8.