2729
J. Sep. Sci. 2015, 38, 2729–2736
Dilara Sac¸lıgil Serap S¸enel Handan Yavuz Adil Denizli Hacettepe University, Department of Chemistry, Beytepe, Ankara, Turkey Received March 9, 2015 Revised April 28, 2015 Accepted April 30, 2015
Research Article
Purification of transferrin by magnetic immunoaffinity beads Immunoaffinity adsorbent for transferrin (Tf) purification was prepared by immobilizing anti-transferrin (Anti-Tf) antibody on magnetic monosizepoly(glycidyl methacrylate) beads, which were synthesized by dispersion polymerization technique in the presence of Fe3 O4 nanopowder and obtained with an average size of 2.0 m. The magnetic poly(glycidyl methacrylate) (mPGMA) beads were characterized by Fourier transform infrared spectroscopy, swelling tests, scanning electron microscopy, electron spin resonance spectroscopy, thermogravimetric analysis and zeta sizing analysis. The density and swelling ratio of the beads were 1.08 g/cm3 and 52%, respectively. Anti-Tf molecules were covalently coupled through epoxy groups of mPGMA. Optimum binding of anti-Tf was 2.0 mg/g. Optimum Tf binding from aqueous Tf solutions was determined as 1.65 mg/g at pH 6.0 and initial Tf concentration of 1.0 mg/mL. There was no remarkable loss in the Tf adsorption capacity of immunoaffinity beads after five adsorption–desorption cycles. Tf adsorption from artificial plasma was also investigated and the purity of the Tf molecules was shown with gel electrophoresis studies. Keywords: Anti-transferrin / Immunoaffinity chromatography / Magnetic beads / Poly(glycidyl methacrylate) / Transferrin DOI 10.1002/jssc.201500216
1 Introduction Transferrin (Tf) is an 80 kDA monomeric glycoprotein with 670–700 amino acids and is responsible for maintaining iron homeostasis through the cellular uptake, storage and transportation of iron [1, 2]. The Tf/Tf receptor mediated cellular iron uptake and transport are critical to avoid cell damage associated with both the iron deficiency and overload in the body [3]. Tf has been widely used to treat a number of diseases such as thalassemia, bacterial infection, diabetes and atransferrinemia [4, 5]. Tf also serves as a biomarker of several clinical conditions such as chronic alcohol abuse [6]. In some cases Tf depletion is required for the detection of low abundant proteins for the proteomic studies [7]. In addition, as cancer cells include lots of Tf receptors on their surface, Tf is widely used as a target ligand for anti-cancer drugs, proteins and genes [8–10]. The conjugation of Tf with proteins, drugs and macromolecule containing hybrid systems makes this situation possible because the Tf is biodegradable, non-toxic and non-immunogenic [11]. Therefore, attempts to develop strategies to obtain pure Tf are crucial.
Correspondence: Dr. Handan Yavuz, PhD, Hacettepe University, Faculty of Science, Department of Chemistry, Biochemistry Division, 06800, Beytepe, Ankara, Turkey E-mail: [email protected] Fax: +90 312 2992163
Abbreviations: Tf, transferrin; mPGMA, magnetic poly (glycidyl methacrylate)
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Immunoaffinity chromatography, which uses polyclonal, monoclonal or anti-idiotypic antibodies as a ligand allows extremely specific product recovery when comparing with other chromatographic techniques [12–15]. Different chemistries, which effect immobilization yield, antigen binding capacity and recovery, could be applied for the preparation of immunoaffinity supports by using monoclonal or polyclonal antibodies and antibody fragments [16]. Due to their specificity for antigens, they are also unique tools in many analytical and preparative systems such as immunoprecipitation [17], immunoaffinity capillary electrophoresis [18], immunofiltration and immunoultrafiltration [19, 20]. Several natural and synthetic materials may be used as stationary phase in immunoaffinity chromatography. Among them, the micron-sized non-porous particles have many advantages, such as good mass transfer characteristics, no diffusion limitation and high resistance to fouling [21]. In addition, affinity ligand immobilized magnetic beads simplifies the process and provides very gentle separation of target proteins [22]. Several types of magnetic affinity adsorbents have been prepared for different types of applications, such as purification of biomolecules [23–25], enzyme immobilization supports [26, 27], removal of pathogenic molecules from blood [28,29], protein depletion for proteomic analysis [30,31], removal of heavy metals [32, 33], molecularly imprinted adsorbents [34, 35], etc. In this study, anti-transferrin antibody (Anti-Tf) immobilized magnetic poly(glycidyl methacrylate) (mPGMA) beads (2 m sized) were synthesized by dispersion polymerization method and the Tf isolation performance from
www.jss-journal.com
2730
J. Sep. Sci. 2015, 38, 2729–2736
D. Sac¸lıgil et al.
Figure 1. FTIR spectrum of magnetic PGMA beads.
one-component aqueous systems and artificial serum samples were reported.
2 Materials and methods 2.1 Materials Human serum Tf, chicken serum anti-Tf antibody (Sigma– Aldrich GW20009F), poly(vinyl pyrrolidone) K-30, Fe(II,III) oxide nanopowder, ethylene glycol dimethacrylate, L-cysteine hydrochloride anhydrate, glycine-hydrochloric acid were obtained from Sigma–Aldrich (St. Louis, MO, USA). Glycidyl methacrylate, ␣,␣ -azobisisobutyronitrile and toluene were supplied from Fluka (Buchs, Switzerland). Other chemicals were of analytical grade and obtained from Merck (Darm stadt, Germany). Barnstead (Dubuque, IA) ROpure LP re verse osmosis unit, Barnstead D3804 Nanopure organic, colloid removal unit and ion exchange column were used for the water purification. The conductivity of deionized water was 18.2 M⍀/cm. R
R
2.2 Synthesis of magnetic PGMA beads The beads were synthesized by dispersion polymerization technique with respect to the procedure given in detailed elsewhere [36]. Briefly, glycidyl methacrylate as a monomer (9.5 mL) and ethylene glycol dimethacrylate as a crosslinker (0.1 mL) were added to 9.0 mL of toluene. 4.0 g poly(vinyl C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
pyrrolidone) K-30 as a stabilizer was added into 84 mL of ethanol. These two phases were mixed and then 0.2 g of ␣,␣ azobisisobutyronitrile as an initiator was added. Following these steps, 50 mg of Fe(II,III) oxide nanopowder was added to the resulting mixture and then this mixture was homogenized. The resulting mixture was taken into a polymerization reactor at 70⬚C for 4 h in a temperature controlled shaking water bath. The resulting beads were washed with aqueous ethanol solutions to remove polymerization residuals.
2.3 Characterization of magnetic beads The density of the beads was measured by a pycnometer by using the dispersion of beads in ethanol. The extent of epoxy groups were determined by titration of dispersed sample in 0.1 M tetraethyl ammonium bromide in acetic acid with 0.1 M of perchloric acid. FTIR analysis was carried out using an FTIR-ATR spectrophotometer (Thermo Fisher Scientific, Nicolet IS10, Waltham, MA, USA) over the range of 650–4000 cm−1 . The surface morphology was examined with SEM (JEOL, JEM, 1200 EX, Japan). The dried beads were attached to SEM sample holder with conducting glue and coated with 200 Å thick gold. SEM micrographs were recorded with a magnification ratio of 10.000. The thermogravimetric and differential thermal analysis were conducted under nitrogen atmosphere using TGA–DTA (Shimadzu DTG-60H) with a heating rate of 20⬚C/min. www.jss-journal.com
Other Techniques
J. Sep. Sci. 2015, 38, 2729–2736
2731
2.5 Transferrin adsorption from aqueous solutions Anti-Tf immobilized mPGMA beads were equilibrated in adsorption buffers for 30 min and then centrifuged. The effect of medium conditions on adsorption capacity was examined batch wise. The runs were conducted in a rotator with 250 rpm stirring rate for 2 h. The beads were precipitated with magnetic separator (Dynal MPC-L, Invitrogen Dynal, Norway) when the adsorption process was completed. The amount of Tf adsorbed onto anti-Tf antibody immobilized mPGMA beads was determined by the Bradford method. Tf adsorption from aqueous solutions onto the anti-Tf immobilized mPGMA beads was investigated at different pH (acetate buffer for pH 4.0–5.0 and phosphate buffer for pH 7.0–8.0) and initial Tf concentration (0.05–0.75 mg/mL). Figure 2. SEM images of magnetic PGMA beads.
2.6 Transferrin adsorption from artificial plasma The presence of magnetite in structure was proved with electron spin resonance (ESR) spectroscopy (EL 9, Varian, USA). The particle size and size distribution were found with Nano zeta sizer (Nano S, Malvern Instruments, UK) using suspension of mPGMA beads. The swelling ratio of beads was determined by examining the change in weight following an incubation period of 2 h in 50 mL of water at 25⬚C in a temperature-controlled water bath.
2.4 Anti-Tfimmobilization on magnetic PGMA beads To bind anti-Tf antibodies covalently through the epoxy groups of mPGMA beads, they were first interacted with 0.1 M Na2 CO3 at room temperature for 24 h in a rotator. Covalent attachment was managed by adding anti-Tf solution at different concentrations (i.e., 0.1, 0.3, 0.5 and 1.0 mg/mL) into the suspension of beads and allowed to interact with each other in a rotator for 2 h. After this step, 50 mM L-cysteine hydrochloride anhydrite solution was added and held for 30 min in rotator to inactivate non-bound epoxy groups. Then suspension is centrifuged at 25⬚C for 20 min at 9000 rpm. The protein concentration was determined by the Bradford method using a UV-Vis spectrophotometer (UV-mini 1240, Shimadzu, Japan) recording the absorbance values at 595 nm. The immobilized antibody extent per unit mass of beads, Q (mg/g) was calculated using the following equation;
Artificial plasma (Tokra Medical, Ankara, Turkey) containing lyophilized blood proteins was used as the source of Tf. The plasma sample was dissolved in 1.5 mL physiological serum and interacted with Anti-Tf/mPGMA beads. The amount of Tf was determined turbidimetrically and calculated from the initial and final concentrations. The amount of albumin and IgG was also determined colorimetric (Olympus AU 2700, Japan) and nephelometric (Beckman Array 360, USA) methods and adsorption amounts on the anti-Tf immobilized mPGMA beads were calculated with Eq. 1. The purity of separated Tf was checked with SDS-PAGE. The gels were stained with Coomassie Brillant R250 in acetic acid/methanol/water solution (1:5:5, v/v/v) and washed with ethanol/acetic acid/water (1:4:6, v/v/v). Electrophoresis was conducted at 110 V for 4 h.
2.7 Desorption and reusability To remove adsorbed Tf from magnetic beads, 0.1 M glycine– HCl (pH 2.8) was used as a desorption agent. The Tf adsorbed beads were immersed in elution solution and stirred at 300 rpm at room temperature for 2 h. The adsorption– desorption cycle was repeated for five times by using the same beads to determine the reusability of adsorbent.
3 Results 3.1 Characterization of magnetic PGMA beads
Q = (Co − C ) . V/m,
(1)
where Co and C are anti-Tf concentrations (mg/mL) before and after adsorption; V is the solution volume and m is the mass of polymer. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The GMA beads were preferred because of its epoxy groups, which provide easy ligand binding and its low cost. The beads were synthesized in magnetic form for easy separation from adsorption medium thus preventing the possible denaturation of biomolecule encountered in other separation procedures (i.e. centrifugation). www.jss-journal.com
2732
J. Sep. Sci. 2015, 38, 2729–2736
D. Sac¸lıgil et al.
Figure 3. ESR spectrum of magnetic PGMA beads.
Figure 4. TGA-DTA curve of magnetic PGMA beads.
The presence of epoxy groups are confirmed by FTIR spectrum (Fig. 1) as it involves peaks for epoxy rings at 884 and 906 cm−1 . The ester and methylene group vibration peaks of GMA were observed at 1723 and 3000 cm−1 . The amount of epoxy groups on beads was determined by perchloric acid titration and found as 3.46 mmol/g. The surface morphology of the beads was determined by SEM (Fig. 2). The resulting beads were non-porous and monodisperse with an average size of 2 m. The polydispersity index was calculated as 0.999 by Zeta-Sizer. The non-porous structure is an advantage since it eliminates intraparticle diffusion resistance. The beads were highly hydrophilic and reached the equilibrium-swelling ratio (52%) in 15 min. The densities of non-magnetic PGMA beads, Fe3 O4 nanopowder and mPGMA beads were 1.08, 1.97 and 1.14 g/cm3 at 25⬚C.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The fraction of magnetite nanopowder in beads was calculated by the following equation: = (C − M ) / (C − A) ,
(2)
where A , C and M were the densities of non-magnetic PGMA beads, magnetite and mPGMA beads, respectively. The magnetite nanopowder fraction was calculated as 0.045. The magnetite leaching from beads was also tested in acetic acid (50% v/v, pH 2.0) sodium citrate/NaOH buffer (pH 12.0) media. No leaching was recorded. The magnetite addition into polymer structure was proved by ESR spectroscopy (Fig. 3). The relative intensity of mPGMA beads was 125, indicating the existence of local magnetic field of polymeric structure due to magnetite content.
www.jss-journal.com
Other Techniques
J. Sep. Sci. 2015, 38, 2729–2736
2733
Figure 5. Effect of Anti-Tf concentration on Anti-Tf covalent binding onto mPGMA beads.
Figure 8. (A) Pseudo-first-order adsorption kinetics.(B) Pseudosecond-order adsorption kinetics. Table 1. Pseudo-first-order and second-order kinetic constants for mPGMA-Anti-Tf beads
Figure 6. Effect of pH on transferrin adsorption.
[Tf] (mg/mL)
0.5
Experimental
Pseudo 1⬚
Pseudo 2⬚
qeq (mg/g)
k1
qeq
R2
k2
qeq
R2
1.65
0.024
1.35
0.869
1.54
1.01
0.999
The TGA–DTA (Fig. 4) recorded under nitrogen atmosphere showed the complete decomposition of polymeric structure above 450⬚C, the major product being monomer. The other minor products arise from ester decomposition. 3.2 Anti-Tfimmobilization onto mPGMAbeads Figure 7. Effect of initial Tf concentration on Tf adsorption.
The g factor, a numerical characteristic of molecules containing unpaired electrons, was calculated as 2.28 using the following equation: g = h/Hr
(3)
where h is Planck’s constant (6.626 × 10–27 erg.s),  is a universal constant (9.274 × 10–21 erg/Gs), is frequency (9.707×109 Hz) and Hr is the magnetic field resonance (Gs). C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 5 shows anti-Tf binding onto mPGMA beads. As shown in the figure, maximum anti-Tf binding was determined as 2.0 mg/g at an initial anti-Tf concentration of 1.0 mg/mL. These immunoaffinity beads were chosen and used for the Tf adsorption experiments throughout the study. Note that anti-Tf release from the beads was examined and no release was observed at all adsorption–desorption conditions used. It can be seen that the adsorbed amount of anti-Tf is increased with increasing anti-Tf concentration due to high driving force for mass transfer, i.e. the high anti-Tf concentration difference between the solution and the solid bead phases, in the case of high anti-Tf concentration. www.jss-journal.com
2734
J. Sep. Sci. 2015, 38, 2729–2736
D. Sac¸lıgil et al.
3.3 Transferrin adsorption from aqueous solutions 3.3.1 Effect of pH Anti-Tf immobilized mPGMA beads showed highly pH dependent Tf adsorption behavior as shown in Fig. 6. In the studied pH range, highest Tf adsorption was observed at pH 6.0. In more acidic and more alkaline pH regions, adsorption capacity decreased significantly. Note that the pI of Tf is 5.8. At this pH value, probably both the Tf molecules and immobilized anti-Tf molecules are in the most favorable conformation to interact with each other and also the lateral electrostatic repulsions are minimized around isoelectric pH. 3.3.2 Effect of concentration The equilibrium adsorption of Tf first increased with increasing Tf concentration due to the large concentration gradient, the driving force for adsorption, and reached the saturation above 0.5 mg/mL Tf concentration (Fig. 7). When active sites were saturated, the adsorption capacity nearly remained constant and the maximum adsorption amount of Tf was 1.65 mg/g. The Tf adsorption onto mPGMA beads was negligible at all concentrations. There is no functional groups on the unmodified mPGMA beads which bind with Tf molecules, hence, this non-specific binding may be due to Tf diffusion into the swollen polymeric matrix and weak noncovalent interactions between Tf and methacrylate groups of beads. The relatively low adsorption amount of Tf may result from the anti-Tf immobilization procedure in which random binding of antibodies occurs. It may cause inaccessibility of all of the antigen binding regions and results in low adsorption capacity. Figure 9. SDS-PAGE analysis of desorbed transferrin. Band 1: reference protein mixture, Band 2: standard transferrin solution, Band 3: eluted transferrin solution.
Figure 10. Reusability of the magnetic beads.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.3.3 Adsorption kinetics Lagergren’s pseudo-first-order and pseudo-second order rate equations were applied in their linearized forms in equations below, respectively: log q eq − q t = log q eq − (k1 t ) /2.303
(4)
t/q t = 1/k2 q eq 2 + 1/q eq t
(5)
where k1 (min−1 ) and k2 (g/mg.min) were rate constants for pseudo-first-order and pseudo-second-order kinetics, respectively. qeq and qt are denoted the adsorption capacities at equilibrium and at any time in mg/g. The analysis given in Fig. 8 (A) and (B) indicates that the process was controlled chemically, following second-order kinetics, eliminating the diffusion constraints. A comparison of the experimental adsorption capacity with theoretical values is given in Table 1. The correlation coefficient for the linear plot for the pseudo-first-order equation is lower than the correlation coefficient for the pseudo-second-order equation. www.jss-journal.com
Other Techniques
J. Sep. Sci. 2015, 38, 2729–2736
3.4 Transferrin adsorption from artificial plasma Anti-Tf antibody-bound mPGMA beads adsorbed 0.12 mg/g Tf from artificial human plasma. Note that the Tf concentration in artificial plasma used in this study was 2.30 mg/dL, significantly lower than the average Tf concentration in human plasma, which is about 300 mg/dL. The adsorbed Tf concentration by anti-Tf antibody immobilized mPGMA beads was 0.12 mg/g. The adsorption of albumin and IgG on the anti-Tf immobilized mPGMA beads was negligible. To prove the purity of eluted Tf and to show that desorption process did not cause a denaturation, SDS-PAGE analysis was applied (Fig. 9). Lane 1 corresponds to the molecular weight marker. The eluted protein was placed in Lane 3 as a single band, which indicates the purity of Tf. Zhang et al. have used Phenyl Sepharose 6FF hydrophobic interaction chromatography and Q-DEAE Sepharose FF anion-exchange chromatography for Tf purification. They reached 95% purity with a single step anion exchange chromatography and 90% purity with hydrophobic interaction chromatography [1]. For the immunoaffinity purification of serum Tf, Barroso et al. have prepared anti-Tf antibody immobilized using silica-hydrazide columns and CNBr activatedsepharose columns. The adsorption capacities were 51 and 19 mg/mL for the Sepharose and silica columns, respectively [37]. Awade et al. have compared gel-filtration HPLC (Superose 12HR 10/30), reversed-phase HPLC (Supelcosil LC-304) and anion-exchange HPLC (Mono Q HR 5/5) for the ovotransferrin purification from egg white. With respect to their reports, the most promising results were obtained with reversed-phase HPLC [38]. Dye–ligand affinity chromatography has also been used for the serum Tf purification [39]. About 99% purity of Tf has obtained with Cibacron Blue F3GA reactive dye immobilized-Bio-Gel A.
3.5 Desorption and reusability Regeneration of immunoaffinity beads is a crucial step to make these adsorbents preferable for laboratory and commercial applications. It was found that desorption ratios of 86% have been achieved with the 0.1 M glycine–HCl (pH 2.8) as desorption agent for an interaction time of 2 h. The change in capacity was not noticeable after five repeated use (Fig. 10).
4 Conclusion In this study, PGMA beads were prepared magnetically and the Anti-Tf molecules were covalently coupled onto magnetic beads through their epoxy groups for the specific and one step isolation of Tf from plasma. Magnetic beads allowed gentle separation of immunoaffinity adsorbent from adsorption media in minutes. A single band in SDS-PAGE analysis proved the purity of the isolated Tf molecules. The authors have declared no conflict of interest.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2735
5 References [1] Zhang, D., Nandi, S., Bryan, P., Pettit, S., Nguyen, D., Santos, M. A., Huang, N., Protein Expr. Purif. 2010, 74, 69–79. [2] Eid, C., Hemadi, M., Ha-Duong, N., Chahine, J. E. H., Biochim. Biophys. Acta 2014, 1840, 1771–1781. [3] Hirose, M., Biosci. Biotech. Bioch. 2000, 64, 1328–1336. [4] Brandsma, M.E., Jenvikar, A. M., Ma, S., Biotechnol. Adv. 2011, 29, 230–238. [5] Keenan, J., Pearson, D., O’Driscoll, L., Gammell, P., Clynes, M., Cytotechnology 2006, 51, 29–37. [6] Caslavska, J., Joneli, J., Wanzenried, U., Schiess, J., Thormann, W., J. Sep. Sci. 2012, 35, 3521–3528. ¨ ¨ ¨ ¨ [7] Gronwall, C., Sjonberg, A., Ramstrom, M., HoldenGuthenberg, I., Hober, S., Jonasson, P., Stahi, S., Biotechnol. J. 2007, 2, 1389–1398. [8] Qian, Z. M., Li, H., Sun, H., Ho, K., Pharmacol. Rev. 2003, 54, 561–587. [9] Daniels, T. R., Bernabeu, E., Rodriguez, J. A., Patel, S., Kozman, M., Chiapetta, D. A., Holler, E., Ljubimova, J. Y., Helguera, G., Penichet, M. L., Biochim. Biophys. Acta 2012, 1820, 291–317. [10] Soni, V., Jain, S. K., Kohli, D. V., Am. J. Drug Deliv. 2004, 3, 155–170. [11] Li, H., Qian, Z. M., Med. Res. Rev. 2002, 22, 225–250. [12] Hage, D. S., Cazes, J., Handbook of Affinity Chromatography, Taylor & Francis Group, Oxford 2006. [13] Yao, X., Li, J., Deng, N., Wang, S., Meng, Y., Shen, F., J. Sep. Sci. 2011, 34, 3092–3098. [14] Yan, J., Zhao, M. P., Li, Y. Z., J. Sep. Sci. 2005, 28, 1163– 1170. [15] Wen, J., Kong, W., Wang, J., Yang, M., J. Sep. Sci. 2013, 36, 3709–3716. [16] Beyer, N. H., Hansen, M. Z., Schou, C., Horjup, P., Heegaard, N. H. H., J. Sep Sci. 2009, 32, 1592–1604. [17] Heegaard, N. H. H., Hansen, M. Z., Sen, J. W., Christiansen, M., Westermark, P., J. Sep. Sci. 2006, 29, 371– 377. [18] Kalish, H., Phillips, T. M., J. Sep. Sci. 2009, 32, 1605–1612. [19] Ott, S., Niessner, R., Seidel, M., J. Sep. Sci. 2011, 34, 2181–2192. [20] Reiter, E. V., Cichna-Marki, M., Chung, D. H., Zentek, J., Razzazi-Fazeli, E., J. Sep. Sci. 2009, 32, 1729–1739. [21] Burgess, R. R., Thompson, N. E., Curr. Opin. Biotech. 2002, 13, 303–308. [22] Safarik, I., Safarikova, M., Biomagn. Res. Technol. 2004, 2, 7–23. [23] Altıntas¸, E. B., Denizli, A., Mat. Sci. Eng. C 2009, 29, 1627– 1634. ¨ S., Aksoz, ¨ E., Denizli, A., [24] Perc¸in, I., Karakoc¸, V., Akgol, Mat. Sci. Eng. C 2012, 32, 1133–1140. ¨ [25] Turkmen, D., Yavuz, H., Denizli, A., Int. J. Biol. Macromol. 2006, 38, 126–133. [26] Saylan, Y., Uzun, L., Denizli, A., Ind. Eng. Chem. Res. 2015, 54, 454–461.
www.jss-journal.com
2736
D. Sac¸lıgil et al.
¨ S., Karatas¸, M., Denizli, A., Ind. Eng. [27] Sarı, M., Akgol, Chem. Res. 2006, 45, 3036–3043. [28] Odabas¸ı, M., Denizli, A., J. Chromatogr. B 2001, 760, 137– 148. ¨ ¨ ¨ [29] Odabas¸ı, M., Ozkayar, N., Ozkara, S., Unal, S., Denizli, A., J. Chromatogr. B 2005, 826, 50–57. ¨ Karakoc¸, V., Andac¸, M., Turkmen, ¨ ¨ [30] Gokay, O., D., Denizli, A., Artif. Cells Nanomed. Biotechnol. 2015, 43, 62–70. ¨ S., Yavuz, H., Say, R., Denizli, A., Int. [31] Karatas¸, M., Akgol, J. Biol. Macromol. 2007, 40, 254–260. [32] Say, R., Denizli, A., J. Biomater. Sci. Polym. Ed. 2001, 12, 1059–1073. ¨ unes ¨ ¸ , H., Safarikova, M., [33] Yavuz, H., Denizli, A., Gung Safarik, I., Sep. Purif. Technol. 2006, 52, 253–260.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2015, 38, 2729–2736
¨ ¨ [34] Turkmen, D., Bereli, N., C¸orman, M. E., Shaikh, H., Akgol, S., Denizli, A., Artif. Cells Nanomed. Biotechnol. 2014, 42, 316–322. [35] Kartal, F., Denizli, A., J. Sep. Sci. 2014, 37, 2077–2086. ¨ [36] Altıntas¸, E. B., Tuzmen, N., Candan, N., Denizli, A., J. Chromatogr. B 2007, 853, 105–113. [37] Barroso, A., Gimenez, E., Benavente, F., Barbosa, J., Sanz-Nebot, V., Anal. Chim. Acta 2013, 804, 167– 175. [38] Awade, A. C., Efstathiou, T., J. Chromatogr. B 1999, 732, 69–74. [39] VamGelder, W., Huijskes-Heins, M. I. E., Hukshorn, C. J., deJeu-Jaspars, C. M. H., Van Noort, W. L., Van Eijk, H. G., Comp. Biochem. Phys. B 1995, 111, 171–179.
www.jss-journal.com
J. Sep. Sci. 2015, 38, 2729–2736
Dilara Sac¸lıgil Serap S¸enel Handan Yavuz Adil Denizli Hacettepe University, Department of Chemistry, Beytepe, Ankara, Turkey Received March 9, 2015 Revised April 28, 2015 Accepted April 30, 2015
Research Article
Purification of transferrin by magnetic immunoaffinity beads Immunoaffinity adsorbent for transferrin (Tf) purification was prepared by immobilizing anti-transferrin (Anti-Tf) antibody on magnetic monosizepoly(glycidyl methacrylate) beads, which were synthesized by dispersion polymerization technique in the presence of Fe3 O4 nanopowder and obtained with an average size of 2.0 m. The magnetic poly(glycidyl methacrylate) (mPGMA) beads were characterized by Fourier transform infrared spectroscopy, swelling tests, scanning electron microscopy, electron spin resonance spectroscopy, thermogravimetric analysis and zeta sizing analysis. The density and swelling ratio of the beads were 1.08 g/cm3 and 52%, respectively. Anti-Tf molecules were covalently coupled through epoxy groups of mPGMA. Optimum binding of anti-Tf was 2.0 mg/g. Optimum Tf binding from aqueous Tf solutions was determined as 1.65 mg/g at pH 6.0 and initial Tf concentration of 1.0 mg/mL. There was no remarkable loss in the Tf adsorption capacity of immunoaffinity beads after five adsorption–desorption cycles. Tf adsorption from artificial plasma was also investigated and the purity of the Tf molecules was shown with gel electrophoresis studies. Keywords: Anti-transferrin / Immunoaffinity chromatography / Magnetic beads / Poly(glycidyl methacrylate) / Transferrin DOI 10.1002/jssc.201500216
1 Introduction Transferrin (Tf) is an 80 kDA monomeric glycoprotein with 670–700 amino acids and is responsible for maintaining iron homeostasis through the cellular uptake, storage and transportation of iron [1, 2]. The Tf/Tf receptor mediated cellular iron uptake and transport are critical to avoid cell damage associated with both the iron deficiency and overload in the body [3]. Tf has been widely used to treat a number of diseases such as thalassemia, bacterial infection, diabetes and atransferrinemia [4, 5]. Tf also serves as a biomarker of several clinical conditions such as chronic alcohol abuse [6]. In some cases Tf depletion is required for the detection of low abundant proteins for the proteomic studies [7]. In addition, as cancer cells include lots of Tf receptors on their surface, Tf is widely used as a target ligand for anti-cancer drugs, proteins and genes [8–10]. The conjugation of Tf with proteins, drugs and macromolecule containing hybrid systems makes this situation possible because the Tf is biodegradable, non-toxic and non-immunogenic [11]. Therefore, attempts to develop strategies to obtain pure Tf are crucial.
Correspondence: Dr. Handan Yavuz, PhD, Hacettepe University, Faculty of Science, Department of Chemistry, Biochemistry Division, 06800, Beytepe, Ankara, Turkey E-mail: [email protected] Fax: +90 312 2992163
Abbreviations: Tf, transferrin; mPGMA, magnetic poly (glycidyl methacrylate)
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Immunoaffinity chromatography, which uses polyclonal, monoclonal or anti-idiotypic antibodies as a ligand allows extremely specific product recovery when comparing with other chromatographic techniques [12–15]. Different chemistries, which effect immobilization yield, antigen binding capacity and recovery, could be applied for the preparation of immunoaffinity supports by using monoclonal or polyclonal antibodies and antibody fragments [16]. Due to their specificity for antigens, they are also unique tools in many analytical and preparative systems such as immunoprecipitation [17], immunoaffinity capillary electrophoresis [18], immunofiltration and immunoultrafiltration [19, 20]. Several natural and synthetic materials may be used as stationary phase in immunoaffinity chromatography. Among them, the micron-sized non-porous particles have many advantages, such as good mass transfer characteristics, no diffusion limitation and high resistance to fouling [21]. In addition, affinity ligand immobilized magnetic beads simplifies the process and provides very gentle separation of target proteins [22]. Several types of magnetic affinity adsorbents have been prepared for different types of applications, such as purification of biomolecules [23–25], enzyme immobilization supports [26, 27], removal of pathogenic molecules from blood [28,29], protein depletion for proteomic analysis [30,31], removal of heavy metals [32, 33], molecularly imprinted adsorbents [34, 35], etc. In this study, anti-transferrin antibody (Anti-Tf) immobilized magnetic poly(glycidyl methacrylate) (mPGMA) beads (2 m sized) were synthesized by dispersion polymerization method and the Tf isolation performance from
www.jss-journal.com
2730
J. Sep. Sci. 2015, 38, 2729–2736
D. Sac¸lıgil et al.
Figure 1. FTIR spectrum of magnetic PGMA beads.
one-component aqueous systems and artificial serum samples were reported.
2 Materials and methods 2.1 Materials Human serum Tf, chicken serum anti-Tf antibody (Sigma– Aldrich GW20009F), poly(vinyl pyrrolidone) K-30, Fe(II,III) oxide nanopowder, ethylene glycol dimethacrylate, L-cysteine hydrochloride anhydrate, glycine-hydrochloric acid were obtained from Sigma–Aldrich (St. Louis, MO, USA). Glycidyl methacrylate, ␣,␣ -azobisisobutyronitrile and toluene were supplied from Fluka (Buchs, Switzerland). Other chemicals were of analytical grade and obtained from Merck (Darm stadt, Germany). Barnstead (Dubuque, IA) ROpure LP re verse osmosis unit, Barnstead D3804 Nanopure organic, colloid removal unit and ion exchange column were used for the water purification. The conductivity of deionized water was 18.2 M⍀/cm. R
R
2.2 Synthesis of magnetic PGMA beads The beads were synthesized by dispersion polymerization technique with respect to the procedure given in detailed elsewhere [36]. Briefly, glycidyl methacrylate as a monomer (9.5 mL) and ethylene glycol dimethacrylate as a crosslinker (0.1 mL) were added to 9.0 mL of toluene. 4.0 g poly(vinyl C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
pyrrolidone) K-30 as a stabilizer was added into 84 mL of ethanol. These two phases were mixed and then 0.2 g of ␣,␣ azobisisobutyronitrile as an initiator was added. Following these steps, 50 mg of Fe(II,III) oxide nanopowder was added to the resulting mixture and then this mixture was homogenized. The resulting mixture was taken into a polymerization reactor at 70⬚C for 4 h in a temperature controlled shaking water bath. The resulting beads were washed with aqueous ethanol solutions to remove polymerization residuals.
2.3 Characterization of magnetic beads The density of the beads was measured by a pycnometer by using the dispersion of beads in ethanol. The extent of epoxy groups were determined by titration of dispersed sample in 0.1 M tetraethyl ammonium bromide in acetic acid with 0.1 M of perchloric acid. FTIR analysis was carried out using an FTIR-ATR spectrophotometer (Thermo Fisher Scientific, Nicolet IS10, Waltham, MA, USA) over the range of 650–4000 cm−1 . The surface morphology was examined with SEM (JEOL, JEM, 1200 EX, Japan). The dried beads were attached to SEM sample holder with conducting glue and coated with 200 Å thick gold. SEM micrographs were recorded with a magnification ratio of 10.000. The thermogravimetric and differential thermal analysis were conducted under nitrogen atmosphere using TGA–DTA (Shimadzu DTG-60H) with a heating rate of 20⬚C/min. www.jss-journal.com
Other Techniques
J. Sep. Sci. 2015, 38, 2729–2736
2731
2.5 Transferrin adsorption from aqueous solutions Anti-Tf immobilized mPGMA beads were equilibrated in adsorption buffers for 30 min and then centrifuged. The effect of medium conditions on adsorption capacity was examined batch wise. The runs were conducted in a rotator with 250 rpm stirring rate for 2 h. The beads were precipitated with magnetic separator (Dynal MPC-L, Invitrogen Dynal, Norway) when the adsorption process was completed. The amount of Tf adsorbed onto anti-Tf antibody immobilized mPGMA beads was determined by the Bradford method. Tf adsorption from aqueous solutions onto the anti-Tf immobilized mPGMA beads was investigated at different pH (acetate buffer for pH 4.0–5.0 and phosphate buffer for pH 7.0–8.0) and initial Tf concentration (0.05–0.75 mg/mL). Figure 2. SEM images of magnetic PGMA beads.
2.6 Transferrin adsorption from artificial plasma The presence of magnetite in structure was proved with electron spin resonance (ESR) spectroscopy (EL 9, Varian, USA). The particle size and size distribution were found with Nano zeta sizer (Nano S, Malvern Instruments, UK) using suspension of mPGMA beads. The swelling ratio of beads was determined by examining the change in weight following an incubation period of 2 h in 50 mL of water at 25⬚C in a temperature-controlled water bath.
2.4 Anti-Tfimmobilization on magnetic PGMA beads To bind anti-Tf antibodies covalently through the epoxy groups of mPGMA beads, they were first interacted with 0.1 M Na2 CO3 at room temperature for 24 h in a rotator. Covalent attachment was managed by adding anti-Tf solution at different concentrations (i.e., 0.1, 0.3, 0.5 and 1.0 mg/mL) into the suspension of beads and allowed to interact with each other in a rotator for 2 h. After this step, 50 mM L-cysteine hydrochloride anhydrite solution was added and held for 30 min in rotator to inactivate non-bound epoxy groups. Then suspension is centrifuged at 25⬚C for 20 min at 9000 rpm. The protein concentration was determined by the Bradford method using a UV-Vis spectrophotometer (UV-mini 1240, Shimadzu, Japan) recording the absorbance values at 595 nm. The immobilized antibody extent per unit mass of beads, Q (mg/g) was calculated using the following equation;
Artificial plasma (Tokra Medical, Ankara, Turkey) containing lyophilized blood proteins was used as the source of Tf. The plasma sample was dissolved in 1.5 mL physiological serum and interacted with Anti-Tf/mPGMA beads. The amount of Tf was determined turbidimetrically and calculated from the initial and final concentrations. The amount of albumin and IgG was also determined colorimetric (Olympus AU 2700, Japan) and nephelometric (Beckman Array 360, USA) methods and adsorption amounts on the anti-Tf immobilized mPGMA beads were calculated with Eq. 1. The purity of separated Tf was checked with SDS-PAGE. The gels were stained with Coomassie Brillant R250 in acetic acid/methanol/water solution (1:5:5, v/v/v) and washed with ethanol/acetic acid/water (1:4:6, v/v/v). Electrophoresis was conducted at 110 V for 4 h.
2.7 Desorption and reusability To remove adsorbed Tf from magnetic beads, 0.1 M glycine– HCl (pH 2.8) was used as a desorption agent. The Tf adsorbed beads were immersed in elution solution and stirred at 300 rpm at room temperature for 2 h. The adsorption– desorption cycle was repeated for five times by using the same beads to determine the reusability of adsorbent.
3 Results 3.1 Characterization of magnetic PGMA beads
Q = (Co − C ) . V/m,
(1)
where Co and C are anti-Tf concentrations (mg/mL) before and after adsorption; V is the solution volume and m is the mass of polymer. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The GMA beads were preferred because of its epoxy groups, which provide easy ligand binding and its low cost. The beads were synthesized in magnetic form for easy separation from adsorption medium thus preventing the possible denaturation of biomolecule encountered in other separation procedures (i.e. centrifugation). www.jss-journal.com
2732
J. Sep. Sci. 2015, 38, 2729–2736
D. Sac¸lıgil et al.
Figure 3. ESR spectrum of magnetic PGMA beads.
Figure 4. TGA-DTA curve of magnetic PGMA beads.
The presence of epoxy groups are confirmed by FTIR spectrum (Fig. 1) as it involves peaks for epoxy rings at 884 and 906 cm−1 . The ester and methylene group vibration peaks of GMA were observed at 1723 and 3000 cm−1 . The amount of epoxy groups on beads was determined by perchloric acid titration and found as 3.46 mmol/g. The surface morphology of the beads was determined by SEM (Fig. 2). The resulting beads were non-porous and monodisperse with an average size of 2 m. The polydispersity index was calculated as 0.999 by Zeta-Sizer. The non-porous structure is an advantage since it eliminates intraparticle diffusion resistance. The beads were highly hydrophilic and reached the equilibrium-swelling ratio (52%) in 15 min. The densities of non-magnetic PGMA beads, Fe3 O4 nanopowder and mPGMA beads were 1.08, 1.97 and 1.14 g/cm3 at 25⬚C.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The fraction of magnetite nanopowder in beads was calculated by the following equation: = (C − M ) / (C − A) ,
(2)
where A , C and M were the densities of non-magnetic PGMA beads, magnetite and mPGMA beads, respectively. The magnetite nanopowder fraction was calculated as 0.045. The magnetite leaching from beads was also tested in acetic acid (50% v/v, pH 2.0) sodium citrate/NaOH buffer (pH 12.0) media. No leaching was recorded. The magnetite addition into polymer structure was proved by ESR spectroscopy (Fig. 3). The relative intensity of mPGMA beads was 125, indicating the existence of local magnetic field of polymeric structure due to magnetite content.
www.jss-journal.com
Other Techniques
J. Sep. Sci. 2015, 38, 2729–2736
2733
Figure 5. Effect of Anti-Tf concentration on Anti-Tf covalent binding onto mPGMA beads.
Figure 8. (A) Pseudo-first-order adsorption kinetics.(B) Pseudosecond-order adsorption kinetics. Table 1. Pseudo-first-order and second-order kinetic constants for mPGMA-Anti-Tf beads
Figure 6. Effect of pH on transferrin adsorption.
[Tf] (mg/mL)
0.5
Experimental
Pseudo 1⬚
Pseudo 2⬚
qeq (mg/g)
k1
qeq
R2
k2
qeq
R2
1.65
0.024
1.35
0.869
1.54
1.01
0.999
The TGA–DTA (Fig. 4) recorded under nitrogen atmosphere showed the complete decomposition of polymeric structure above 450⬚C, the major product being monomer. The other minor products arise from ester decomposition. 3.2 Anti-Tfimmobilization onto mPGMAbeads Figure 7. Effect of initial Tf concentration on Tf adsorption.
The g factor, a numerical characteristic of molecules containing unpaired electrons, was calculated as 2.28 using the following equation: g = h/Hr
(3)
where h is Planck’s constant (6.626 × 10–27 erg.s),  is a universal constant (9.274 × 10–21 erg/Gs), is frequency (9.707×109 Hz) and Hr is the magnetic field resonance (Gs). C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 5 shows anti-Tf binding onto mPGMA beads. As shown in the figure, maximum anti-Tf binding was determined as 2.0 mg/g at an initial anti-Tf concentration of 1.0 mg/mL. These immunoaffinity beads were chosen and used for the Tf adsorption experiments throughout the study. Note that anti-Tf release from the beads was examined and no release was observed at all adsorption–desorption conditions used. It can be seen that the adsorbed amount of anti-Tf is increased with increasing anti-Tf concentration due to high driving force for mass transfer, i.e. the high anti-Tf concentration difference between the solution and the solid bead phases, in the case of high anti-Tf concentration. www.jss-journal.com
2734
J. Sep. Sci. 2015, 38, 2729–2736
D. Sac¸lıgil et al.
3.3 Transferrin adsorption from aqueous solutions 3.3.1 Effect of pH Anti-Tf immobilized mPGMA beads showed highly pH dependent Tf adsorption behavior as shown in Fig. 6. In the studied pH range, highest Tf adsorption was observed at pH 6.0. In more acidic and more alkaline pH regions, adsorption capacity decreased significantly. Note that the pI of Tf is 5.8. At this pH value, probably both the Tf molecules and immobilized anti-Tf molecules are in the most favorable conformation to interact with each other and also the lateral electrostatic repulsions are minimized around isoelectric pH. 3.3.2 Effect of concentration The equilibrium adsorption of Tf first increased with increasing Tf concentration due to the large concentration gradient, the driving force for adsorption, and reached the saturation above 0.5 mg/mL Tf concentration (Fig. 7). When active sites were saturated, the adsorption capacity nearly remained constant and the maximum adsorption amount of Tf was 1.65 mg/g. The Tf adsorption onto mPGMA beads was negligible at all concentrations. There is no functional groups on the unmodified mPGMA beads which bind with Tf molecules, hence, this non-specific binding may be due to Tf diffusion into the swollen polymeric matrix and weak noncovalent interactions between Tf and methacrylate groups of beads. The relatively low adsorption amount of Tf may result from the anti-Tf immobilization procedure in which random binding of antibodies occurs. It may cause inaccessibility of all of the antigen binding regions and results in low adsorption capacity. Figure 9. SDS-PAGE analysis of desorbed transferrin. Band 1: reference protein mixture, Band 2: standard transferrin solution, Band 3: eluted transferrin solution.
Figure 10. Reusability of the magnetic beads.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.3.3 Adsorption kinetics Lagergren’s pseudo-first-order and pseudo-second order rate equations were applied in their linearized forms in equations below, respectively: log q eq − q t = log q eq − (k1 t ) /2.303
(4)
t/q t = 1/k2 q eq 2 + 1/q eq t
(5)
where k1 (min−1 ) and k2 (g/mg.min) were rate constants for pseudo-first-order and pseudo-second-order kinetics, respectively. qeq and qt are denoted the adsorption capacities at equilibrium and at any time in mg/g. The analysis given in Fig. 8 (A) and (B) indicates that the process was controlled chemically, following second-order kinetics, eliminating the diffusion constraints. A comparison of the experimental adsorption capacity with theoretical values is given in Table 1. The correlation coefficient for the linear plot for the pseudo-first-order equation is lower than the correlation coefficient for the pseudo-second-order equation. www.jss-journal.com
Other Techniques
J. Sep. Sci. 2015, 38, 2729–2736
3.4 Transferrin adsorption from artificial plasma Anti-Tf antibody-bound mPGMA beads adsorbed 0.12 mg/g Tf from artificial human plasma. Note that the Tf concentration in artificial plasma used in this study was 2.30 mg/dL, significantly lower than the average Tf concentration in human plasma, which is about 300 mg/dL. The adsorbed Tf concentration by anti-Tf antibody immobilized mPGMA beads was 0.12 mg/g. The adsorption of albumin and IgG on the anti-Tf immobilized mPGMA beads was negligible. To prove the purity of eluted Tf and to show that desorption process did not cause a denaturation, SDS-PAGE analysis was applied (Fig. 9). Lane 1 corresponds to the molecular weight marker. The eluted protein was placed in Lane 3 as a single band, which indicates the purity of Tf. Zhang et al. have used Phenyl Sepharose 6FF hydrophobic interaction chromatography and Q-DEAE Sepharose FF anion-exchange chromatography for Tf purification. They reached 95% purity with a single step anion exchange chromatography and 90% purity with hydrophobic interaction chromatography [1]. For the immunoaffinity purification of serum Tf, Barroso et al. have prepared anti-Tf antibody immobilized using silica-hydrazide columns and CNBr activatedsepharose columns. The adsorption capacities were 51 and 19 mg/mL for the Sepharose and silica columns, respectively [37]. Awade et al. have compared gel-filtration HPLC (Superose 12HR 10/30), reversed-phase HPLC (Supelcosil LC-304) and anion-exchange HPLC (Mono Q HR 5/5) for the ovotransferrin purification from egg white. With respect to their reports, the most promising results were obtained with reversed-phase HPLC [38]. Dye–ligand affinity chromatography has also been used for the serum Tf purification [39]. About 99% purity of Tf has obtained with Cibacron Blue F3GA reactive dye immobilized-Bio-Gel A.
3.5 Desorption and reusability Regeneration of immunoaffinity beads is a crucial step to make these adsorbents preferable for laboratory and commercial applications. It was found that desorption ratios of 86% have been achieved with the 0.1 M glycine–HCl (pH 2.8) as desorption agent for an interaction time of 2 h. The change in capacity was not noticeable after five repeated use (Fig. 10).
4 Conclusion In this study, PGMA beads were prepared magnetically and the Anti-Tf molecules were covalently coupled onto magnetic beads through their epoxy groups for the specific and one step isolation of Tf from plasma. Magnetic beads allowed gentle separation of immunoaffinity adsorbent from adsorption media in minutes. A single band in SDS-PAGE analysis proved the purity of the isolated Tf molecules. The authors have declared no conflict of interest.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2735
5 References [1] Zhang, D., Nandi, S., Bryan, P., Pettit, S., Nguyen, D., Santos, M. A., Huang, N., Protein Expr. Purif. 2010, 74, 69–79. [2] Eid, C., Hemadi, M., Ha-Duong, N., Chahine, J. E. H., Biochim. Biophys. Acta 2014, 1840, 1771–1781. [3] Hirose, M., Biosci. Biotech. Bioch. 2000, 64, 1328–1336. [4] Brandsma, M.E., Jenvikar, A. M., Ma, S., Biotechnol. Adv. 2011, 29, 230–238. [5] Keenan, J., Pearson, D., O’Driscoll, L., Gammell, P., Clynes, M., Cytotechnology 2006, 51, 29–37. [6] Caslavska, J., Joneli, J., Wanzenried, U., Schiess, J., Thormann, W., J. Sep. Sci. 2012, 35, 3521–3528. ¨ ¨ ¨ ¨ [7] Gronwall, C., Sjonberg, A., Ramstrom, M., HoldenGuthenberg, I., Hober, S., Jonasson, P., Stahi, S., Biotechnol. J. 2007, 2, 1389–1398. [8] Qian, Z. M., Li, H., Sun, H., Ho, K., Pharmacol. Rev. 2003, 54, 561–587. [9] Daniels, T. R., Bernabeu, E., Rodriguez, J. A., Patel, S., Kozman, M., Chiapetta, D. A., Holler, E., Ljubimova, J. Y., Helguera, G., Penichet, M. L., Biochim. Biophys. Acta 2012, 1820, 291–317. [10] Soni, V., Jain, S. K., Kohli, D. V., Am. J. Drug Deliv. 2004, 3, 155–170. [11] Li, H., Qian, Z. M., Med. Res. Rev. 2002, 22, 225–250. [12] Hage, D. S., Cazes, J., Handbook of Affinity Chromatography, Taylor & Francis Group, Oxford 2006. [13] Yao, X., Li, J., Deng, N., Wang, S., Meng, Y., Shen, F., J. Sep. Sci. 2011, 34, 3092–3098. [14] Yan, J., Zhao, M. P., Li, Y. Z., J. Sep. Sci. 2005, 28, 1163– 1170. [15] Wen, J., Kong, W., Wang, J., Yang, M., J. Sep. Sci. 2013, 36, 3709–3716. [16] Beyer, N. H., Hansen, M. Z., Schou, C., Horjup, P., Heegaard, N. H. H., J. Sep Sci. 2009, 32, 1592–1604. [17] Heegaard, N. H. H., Hansen, M. Z., Sen, J. W., Christiansen, M., Westermark, P., J. Sep. Sci. 2006, 29, 371– 377. [18] Kalish, H., Phillips, T. M., J. Sep. Sci. 2009, 32, 1605–1612. [19] Ott, S., Niessner, R., Seidel, M., J. Sep. Sci. 2011, 34, 2181–2192. [20] Reiter, E. V., Cichna-Marki, M., Chung, D. H., Zentek, J., Razzazi-Fazeli, E., J. Sep. Sci. 2009, 32, 1729–1739. [21] Burgess, R. R., Thompson, N. E., Curr. Opin. Biotech. 2002, 13, 303–308. [22] Safarik, I., Safarikova, M., Biomagn. Res. Technol. 2004, 2, 7–23. [23] Altıntas¸, E. B., Denizli, A., Mat. Sci. Eng. C 2009, 29, 1627– 1634. ¨ S., Aksoz, ¨ E., Denizli, A., [24] Perc¸in, I., Karakoc¸, V., Akgol, Mat. Sci. Eng. C 2012, 32, 1133–1140. ¨ [25] Turkmen, D., Yavuz, H., Denizli, A., Int. J. Biol. Macromol. 2006, 38, 126–133. [26] Saylan, Y., Uzun, L., Denizli, A., Ind. Eng. Chem. Res. 2015, 54, 454–461.
www.jss-journal.com
2736
D. Sac¸lıgil et al.
¨ S., Karatas¸, M., Denizli, A., Ind. Eng. [27] Sarı, M., Akgol, Chem. Res. 2006, 45, 3036–3043. [28] Odabas¸ı, M., Denizli, A., J. Chromatogr. B 2001, 760, 137– 148. ¨ ¨ ¨ [29] Odabas¸ı, M., Ozkayar, N., Ozkara, S., Unal, S., Denizli, A., J. Chromatogr. B 2005, 826, 50–57. ¨ Karakoc¸, V., Andac¸, M., Turkmen, ¨ ¨ [30] Gokay, O., D., Denizli, A., Artif. Cells Nanomed. Biotechnol. 2015, 43, 62–70. ¨ S., Yavuz, H., Say, R., Denizli, A., Int. [31] Karatas¸, M., Akgol, J. Biol. Macromol. 2007, 40, 254–260. [32] Say, R., Denizli, A., J. Biomater. Sci. Polym. Ed. 2001, 12, 1059–1073. ¨ unes ¨ ¸ , H., Safarikova, M., [33] Yavuz, H., Denizli, A., Gung Safarik, I., Sep. Purif. Technol. 2006, 52, 253–260.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2015, 38, 2729–2736
¨ ¨ [34] Turkmen, D., Bereli, N., C¸orman, M. E., Shaikh, H., Akgol, S., Denizli, A., Artif. Cells Nanomed. Biotechnol. 2014, 42, 316–322. [35] Kartal, F., Denizli, A., J. Sep. Sci. 2014, 37, 2077–2086. ¨ [36] Altıntas¸, E. B., Tuzmen, N., Candan, N., Denizli, A., J. Chromatogr. B 2007, 853, 105–113. [37] Barroso, A., Gimenez, E., Benavente, F., Barbosa, J., Sanz-Nebot, V., Anal. Chim. Acta 2013, 804, 167– 175. [38] Awade, A. C., Efstathiou, T., J. Chromatogr. B 1999, 732, 69–74. [39] VamGelder, W., Huijskes-Heins, M. I. E., Hukshorn, C. J., deJeu-Jaspars, C. M. H., Van Noort, W. L., Van Eijk, H. G., Comp. Biochem. Phys. B 1995, 111, 171–179.
www.jss-journal.com
