Sp

Va

b

a

ARRAA

KANM

1

iayhsdmflshmiinTct

t

T

0d

International Journal of Biological Macromolecules 45 (2009) 188–193

Contents lists available at ScienceDirect

International Journal of Biological Macromolecules

journa l homepage: www.e lsev ier .com/ locate / i jb iomac

elective separation of human serum albumin with copper(II) chelatedoly(hydroxyethyl methacrylate) based nanoparticles

eyis Karakoca, Erkut Yılmaza, Deniz Türkmena, Nevra Öztürkb, Sinan Akgölb, Adil Denizli a,∗

Department of Chemistry, Hacettepe University, Ankara, TurkeyDepartment of Chemistry, Adnan Menderes University, Aydın, Turkey

r t i c l e i n f o

rticle history:eceived 1 March 2009eceived in revised form 29 April 2009ccepted 30 April 2009vailable online 13 May 2009

a b s t r a c t

Poly(hydroxyethyl methacrylate) (PHEMA) nanoparticles with an average size of 300 nm in diameterand with a polydispersity index of 1.156 were produced by surfactant free emulsion polymerization.Specific surface area of the PHEMA nanoparticles was found to be 996 m2/g. Metal-chelating ligand 3-(2-imidazoline-1-yl)propyl(triethoxysilane) (IMEO) was covalently attached to the PHEMA nanoparticles.

eywords:lbumin purificationanoparticlesetal-chelates

IMEO content was 0.97 mmol IEMO/g. The morphology and properties of these nanoparticles were char-acterized with scanning electron microscopy, Fourier transform infrared spectroscopy and atomic forcemicroscopy. The Cu2+-chelated PHEMA–IMEO nanoparticles were used in the adsorption-elution studiesof human serum albumin (HSA) in a batch system. Maximum HSA adsorption amount of the Cu2+ chelatednanoparticles was 680 mg HSA/g. The PHEMA–IMEO–Cu2+ nanoparticles exhibited a quite high adsorp-tion capacity and fast adsorption rate due to their high specific surface area and the absence of internaldiffusion resistance.

. Introduction

Albumin is an important plasma protein with many applicationsn therapeutics and diagnostics. Albumin administration has playedpart in fluid management of acutely ill patients for more than 50ears [1]. Albumin therapy is licensed in the United States includeypovolemia or shock burns, hypoalbunemia or hypoproteinemia;urgery; trauma; cardiopulmonary by-pass; the acute respiratoryistress syndrome; hemodialysis, acute nephrosis, hyperbilirubine-ia, acute liver failure, ascitess and sequestration of protein-rich

uids in acute peritonitis, pancreatitis, mediastinitis and exten-ive cellulitis [2]. Albumin is at present commonly isolated fromuman blood by Cohn’s classical fractionation procedure [3]. Cohn’sethod concerns precipitation of proteins using ethanol with vary-

ng pH, ionic strength and temperature. But this technique, whichs the oldest method of industrial fractionation of blood proteins, isot highly specific and can give partially denaturated proteins [4].he rapid development of biotechnology, biochemistry, pharma-

eutical science, and medicine requires more reliable and efficientechniques for isolation and purification of proteins [5–7].

Affinity chromatography is a well-established method for iden-ification, purification and separation of biomolecules and based on

∗ Corresponding author at: P.K. 1, Samanpazarı, 06242, Ankara, Turkey.el.: +90 312 2977963; fax: +90 312 2992163.

E-mail address: denizli@hacettepe.edu.tr (A. Denizli).

141-8130/$ – see front matter © 2009 Elsevier B.V. All rights reserved.oi:10.1016/j.ijbiomac.2009.04.023

© 2009 Elsevier B.V. All rights reserved.

highly specific molecular recognition. In this method, the moleculepossessing a specific recognition capability is immobilized on asuitable support [8]. The molecule to be isolated is selectively cap-tured by the complementary ligand immobilized on the matrix.Affinity chromatography is an alternative technique to Cohn’s frac-tionation. Ligand stability is becoming an increasingly importantconsideration. The recent trend, therefore, has been to replace highmolecular mass biological ligands with small molecular-mass pseu-dospecific ligands [9–11].

Immobilized metal affinity chromatography (IMAC) is primarilybased on affinity adsorption and therefore possesses advantagesas well as disadvantages associated with this type of separa-tion technology [12–15]. Using this technology, large amounts ofrecombinant protein can be recovered, most often in a singlechromatographic step. The advantages of IMAC, which include lig-and stability, high protein loading, mild elution conditions, simpleregeneration and low cost, are very improtant when developingprotein purification procedures [16–19].

The goal of this paper is to report the synthesis ofpoly(hydroxyethyl methacrylate) (PHEMA) nanoparticles carry-ing 3-(2-imidazoline-1-yl)propyl(triethoxysilane) (IMEO) and theiruse in the purification of human serum albumin (HSA) from human

serum by metal-chelation. As indicated by several studies, theimidazole moiety in histidine contains a pyridine like nitrogenatom, which is a good ligand for metal ions [20,21,12]. PHEMAnanoparticles (300 nm in diameter) were produced by a surfactantfree emulsion polymerization. These nanoparticles were chemically

Biological Macromolecules 45 (2009) 188–193 189

mPFbct

2

2

d(ataMpwc(wepahbe

2

irPr0amatmbmwma3pn

2

1otrariuPo

V. Karakoc et al. / International Journal of

odified to introduce IMEO groups on the nanoparticle surface. TheHEMA–IMEO nanoparticles were characterized by SEM, AFM andT-IR. Albumin adsorption studies were performed to evaluate theinding capacity of HSA onto the PHEMA–IMEO–Cu2+ nanoparti-les. Elution of HSA and regeneration of the nanoparticles were alsoested.

. Experimental

.1. Materials

Hydroxyethyl methacrylate (HEMA) and ethylene glycolimethacrylate (EGDMA) were purchased from Sigma Chem. Co.St. Louis, USA). Commercial HEMA contains residual methacryliccid and crosslinkers due to fabrication process. The polymeriza-ion inhibitor 4-methoxyphenol also needs to be removed. HEMAnd EGDMA were distilled under reduce pressure (0.01 mBar, 70 ◦C.onomers were stored at 4 ◦C until use. 3-(2-imidazoline-1-yl)

ropyl(triethoxysilane) (IMEO, molecular weight: 274.43 g/mol)as purchased from Sigma. Potassium persulphate (KPS) was pur-

hased from Aldrich (Munich, Germany). Human serum albuminHSA, lyophilized, Fraction V) was purchased from Sigma and usedithout further purification. All other chemicals were of the high-

st purity commercially available and were used without furtherurification. All water used in the experiments was purified usingBarnstead (Dubuque, IA) ROpure LP® reverse osmosis unit with aigh flow cellulose acetate membrane (Barnstead D2731) followedy a Barnstead D3804 NANOpure® organic/colloid removal and ionxchange packed bed system.

.2. Synthesis of PHEMA nanoparticles

Surfactant free emulsion polymerization was carried out accord-ng to the literature procedure with minor modifications aseported elsewhere [22]. Briefly, the stabilizer, polyvinyl alcoholVAL (0.5 g), was dissolved in 50 ml deionized water for the prepa-ation of the continuous phase. Then, the HEMA/EGDMA mixture.6 ml/0.01 ml was added to the dispersion which was mixing inn ultrasonic bath for 30 min. KPS (initiator) concentration in theonomer phase was 0.44 mg/ml. Prior to polymerization, KPS was

dded to the monomer phase and nitrogen gas blown throughhe medium for about 1–2 min to remove dissolved oxygen. Poly-

erization was carried out in a constant temperature shakingath at 70 ◦C, under nitrogen atmosphere for 24 h. After the poly-erization, the PHEMA nanoparticles were cleaned by washingith methanol and water several times to remove the unreactedonomers. For this purpose, the nanoparticles were precipitated

nd collected with the help of a centrifuge (Zentrifugen, Universal2 R, Germany) which is the rate of 25,000 g for 1 h and resus-ended in methanol and water several times. After that, the PHEMAanoparticles were further washed with deionized water.

.3. IMEO binding on PHEMA nanoparticles

For the IMEO binding, PHEMA nanoparticles and IMEO (mol ratio:10) were mixed and stirred at 25 ◦C for about 4 days. At the endf this period, stirring was stopped. The IMEO binding reactionakes place at 25 ◦C without any catalyst as shown in Fig. 1. Theesulting IMEO modified PHEMA nanoparticles were centrifuged

nd washed with dichloromethane. Then, the nanoparticles wereesuspend in distilled water. To evaluate the degree of IMEO bind-ng, the PHEMA nanoparticles were subjected to silisium analysis bysing flame atomizer atomic absorption spectrometer (Analyst 800,erkinElmer, Bodenseewerk, USA). Fig. 1 shows chemical structuref the PHEMA–IMEO nanoparticles.

Fig. 1. Structure of the PHEMA–IMEO nanoparticles.

2.4. Characterization studies

The average nanoparticle size and size distribution were deter-mined by Zeta Sizer (Malvern Instruments, Model 3000 HSA,England).

The morphology and size distribution of the nanoparticles wereexamined by scanning electron microscopy (SEM, Philis, XL-30SFEG, Germany). The nanoparticle sample was initially dried in airat 25 ◦C before being analyzed. The sample was mounted on aSEM sample mount and was sputter coated for 2 min. The samplewas then mounted in a SEM. The surface of the sample was thenscanned at the desired magnification to study the morphology ofthe nanoparticles.

The size of the PHEMA nanoparticles was analyzed by atomicforce microscopy (AFM) (Digital Instruments, MMafm-2/1700 EXL).Scanning was performed at a scan rate of 1.001 Hz and scan size of5.000 �m. The tip loading force was minimized to avoid structuralchanges of the sample.

FTIR spectra of the samples were obtained using a FTIR spec-trophotometer (Varian FTS 7000, USA). The dry nanoparticles(about 0.1 g) were thoroughly mixed with KBr (0.1 g, IR Grade,Merck, Germany), and pressed into a tablet form and the spectrumwas recorded. To prepare a liquid sample (e.g., IMEO) to FTIR analy-sis, firstly place a drop of the liquid on the face of a highly polishedKBr (0.1 g, IR Grade, Merck, Germany) plate, then place a secondplate on top of the first plate so as to spread the liquid in a thinlayer between the plates and clamps the plates together. Finallywipe off the liquid out of the edge of plate and the spectrum wasthen recorded.

The surface area of the PHEMA nanoparticles was calculatedusing the following expression:

N = 6.1010S

��sd3(1)

Here, N is the number of nanoparticles per mililiter; S is the% of solids; �s is the density of bulk polymer (g/ml); d is thenanoparticle diameter (nm). The number of nanoparticles in mlsuspension was determined by utilizing from mass-volume graphof nanoparticles. From all these data, specific surface area of thePHEMA nanoparticles was calculated by multiplying N and surfacearea of 1 nanoparticle.

2.5. Cu2+ Chelation

The PHEMA–IMEO–Cu2+ nanoparticles were prepared as fol-lows: PHEMA–IMEO nanoparticles were mixed with aqueoussolutions containing 30 ppm Cu2+ ion, at a constant pH of 5.0(adjusted with HCl and NaOH), which was the optimum pH for Cu2+

chelate formation and at room temperature. A 1000 ppm atomicabsorption standard solution (containing 10% HNO3) was used asthe source of Cu2+ ions. The flasks were stirred magnetically at600 rpm for 1 h (sufficient to attain equilibrium). The concentra-

tion of the Cu2+ ions in the resulting solution was determined witha graphite furnace atomic absorption spectrophotometer. The Cu2+

chelation step was shown in Fig. 2. The amount of chelated Cu2+

ions was calculated by using the concentrations of the Cu2+ ions inthe initial solution and in the equilibrium.

190 V. Karakoc et al. / International Journal of Biological Macromolecules 45 (2009) 188–193

F

2

niItTa(aa

Q

nipp

smTbeuH

2

PHtabEsa3wwPdni

pai(ai

Fig. 3. Scanning electron microscopy image of PHEMA nanoparticles.

ig. 2. Schematic diagram for the chelation of Cu2+ ions through the nanoparticles.

.6. HSA adsorption studies from aqueous solution

HSA adsorption of the PHEMA and the PHEMA–MEO–Cu2+

anoparticles were performed in a batch experimental set-up. Thenitial HSA concentration was changed between 0.05 and 2.0 mg/ml.n a typical adsorption experiment, HSA was dissolved in 10 ml ofhe buffer solution containing NaCl and nanoparticles were added.he adsorption experiments were carried out for 1 h at 25 ◦C atstirring rate of 250 rpm. At the end of the equilibrium period

60 min), the nanoparticles were separated from the solution. HSAdsorption capacity was determined by using Bradford method. Themount of adsorbed HSA was calculated as:

= [(C0 − C)V ]m

(2)

Here, Q is the amount of HSA adsorbed onto unit mass ofanoparticles (mg/g); C0 and C are the concentration of HSA in the

nitial solution and in the aqueous phase after treatment for certaineriod of time, respectively (mg/ml); V is the volume of the aqueoushase (ml); and m is the mass of the nanoparticles used (g).

HSA elution experiments were performed in 50 mM imidazoleolution. HSA-adsorbed nanoparticles were placed in the elutionedium and stirred for 1 h at 25 ◦C, at a stirring rate of 250 rpm.

he final HSA concentration in the elution medium was determinedy Bradford method. In the case of Cu2+-carrying nanoparticles, thelution of Cu2+ ions were also measured in the elution mediumsing AAS. The elution ratio was calculated from the amount ofSA adsorbed on the nanoparticles and the amount of HSA eluted.

.7. HSA adsorption from human plasma

Adsorption of HSA from human plasma on theHEMA–IMEO–Cu2+ nanoparticles was studied in a batch wise.uman plasma was collected from voluntary blood donors. All

he blood samples tested negative for HBS antigen and HIV I–IInd hepatitis C antibodies. No preservative was added into thelood samples. The human blood samples were collected intoDTA-containing vacutainers in which the red blood cells wereeparated out of plasma by centrifugation at 4000 g for 30 mint room temperature. The plasma was then filtered out through�m Sartorius filter and frozen at −20 ◦C. The plasma sampleas thawed for 1 h at 37 ◦C. The PHEMA–IMEO nanoparticlesere incubated at 20 ◦C for 30 min with 20 ml of human plasma.

hosphate buffered saline (PBS, pH: 7.4, NaCl: 0.9%) was used forilution of human plasma. The amount of HSA adsorbed throughanoparticles was determined by a solid phase enzyme linked

mmunosorbent assay method (ELISA) [23].The purity of HSA was assayed by sodium dodecylsulfate-

olyacrylamide gel electrophoresis (SDS-PAGE). All SDS-PAGE

nalysis of the serum samples was performed on 10% separat-ng minigels (9 cm × 7.5 cm) for 120 min at 200 V. Stacking gels5%) were stained with 0.25% (w/v) Coomassie Brillant R 250 incetic acid–methanol–water mixture (1:5:5, v/v/v) and destainedn ethanol–acetic acid–water mixture (1:4:6, v/v/v). The SDS-PAGE

Fig. 4. AFM image of PHEMA nanoparticles.

gels were scanned using a Shimadzu dual-wavelength flying spotscanning densitometer (Shimadzu, Tokyo, Japan).

3. Result and discussion

3.1. Characterization of nanoparticles

Nanoparticles can produce larger specific surface area and there-fore may result in high metal-chelating ligand loading and proteinadsorption. Therefore, it may be useful to synthesize nanoparticleswith large surface area and utilize them as suitable carriers for theadsorption of metal ions. The specific surface area was calculatedas 996 m2/g. Surfactant free emulsion polymerization producedPHEMA nanoparticles with an average size of 300 nm in diameterwith a polydispersity index of 1.156. It is apparent that the PHEMAnanoparticles are perfectly spherical with a relatively smooth

surface and uniform as shown by the scanning electron microscopy(SEM) images (Fig. 3). The small polydispersity index suggest thatnucleation is fast compared to particle growth and also the absenceof a secondary nucleation step. In addition, the total monomer

V. Karakoc et al. / International Journal of Biological Macromolecules 45 (2009) 188–193 191

O; (B

cwgntbga

a

Fig. 5. FTIR spectra of (A) IME

onversion was determined as 98.5% (w/w). PHEMA nanoparticlesere highly dispersive in water by ultrasonication due to hydroxyl

roups on the surface of nanoparticles. The dispersion state of theanoparticles was confirmed visually by the observed white color ofhe suspension. The aqueous dispersion of nanoparticles were sta-

le for several days. The size of the PHEMA nanoparticles were alsoiven by AFM in Fig. 4. AFM image confirms that the particle size isbout 300 nm.

In the FTIR spectrum of the IMEO, the strong absorption bandst 1605 cm−1, assigned to the characteristic n(C N) vibrations and

) PHEMA; (C) PHEMA–IMEO.

indicated a strong band at 2928 and 2974 cm−1 n(C–H) (Fig. 5A).The n(O H) stretching vibration in PHEMA is observed in the3600–3410 cm−1 range as broad absorptions, indicated a strongband at 1730 cm−1 due to n(C O) group and the 2954 cm−1 n(C H)stretching of CH3, the 1268 cm−1 n(C O) stretching vibration

(Fig. 5B). The characteristic n(C O), n(C N) and n(C H) stretchingvibration bands of the PHEMA–IMEO nanoparticles are observedat 1728 cm−1, 1660 cm−1, 2954 cm−1, respectively (Fig. 5C). Inaddition to, the n(Si O C) vibration band is observed in the1264 cm−1. As a result, the peak position of 1264 cm−1 is related to

192 V. Karakoc et al. / International Journal of Biological Macromolecules 45 (2009) 188–193

F0

nao

3

3

tt0noacto

FC

ig. 6. Effect of time on HSA adsorption; IMEO content: 0.97 mmol/g; Cu2+ loading:.95 mmol/g; HSA concentration: 1.0 mg/ml; pH: 5.0.

(Si O C) and the observance of C N bands of the PHEMA–IMEOt 1660 cm−1, the shifts of the C N vibration to higher frequenciesf 1660 cm−1 due to the IMEO binding of nanoparticles.

.2. HSA adsorption-elution studies

.2.1. Effect of timeFig. 6 shows the effects of time on HSA adsorption amount on

he PHEMA–IMEO–Cu2+ nanoparticles. As seen from this figure,he adsorption of HSA on the PHEMA nanoparticles was low, about.25 mg/g. There are no chelating functional groups on the PHEMAanoparticles for HSA binding, therefore this negligible amountf adsorbed HSA may be due to weak interactions between HSA

nd hydroxyl groups on the surface of the PHEMA nanoparti-les. But high adsorption capacity (680 mg/g) was achieved withhe PHEMA–IMEO–Cu2+ nanoparticles due to the incorporationf IMEO–Cu2+ chelating groups onto the polymer structure.

ig. 7. Effects of HSA concentration on HSA adsorption; IMEO content: 0.97 mmol/g;u2+ loading: 0.95 mmol/g; pH: 5.0; Incubation time: 2 h.

Fig. 8. Repeated use of PHEMA–IMEO–Cu2+ nanoparticles; IMEO content:0.97 mmol/g; Cu2+ loading: 0.95 mmol/g; HSA concentration: 1.0 mg/mL; pH: 5.0;Incubation time: 2 h.

Additionally, the long chain of IMEO–Cu2+ complex acted as aspacer arm that could catch HSA easily. Short adsorption equilib-rium time is observed with the PHEMA–IMEO–Cu2+ nanoparticlesat the beginning of adsorption, plateau value (showing adsorp-tion equilibrium) is completely achieved within 10 min. A majoradvantage of nonporous nanoparticles is the absence of significantintraparticle diffusion resistance, making it particularly useful forrapid separation of proteins [24]. This very fast adsorption equilib-rium time is probably due to the high binding rate on the surfaceof nanoparticles between PHEMA–IMEO nanoparticles and Cu2+

ions.

3.2.2. Effect of HSA concentrationFig. 7 shows the effects of HSA concentration on adsorption.

The pH of the adsorption medium was 5.0 (i.e., the isoelectricpoint of HSA), adjusted with CH3COONa–CH3COOH buffer system.As expected, HSA adsorption first increased with increasing con-centration of HSA in the incubation medium and then reached asaturation value at an HSA concentration of 0.75 mg/ml. The non-specific HSA adsorption was very low (0.25 mg/g). Cu2+ chelationsignificantly increased the HSA adsorption capacity of the particles(up to 680 mg/g), because of the specific interactions between HSAand chelated Cu2+ ions.

3.3. Elution studies

The elution of the adsorbed HSA from the PHEMA–IMEO–Cu2+

nanoparticles was studied in a batch experimental set-up. The HSAadsorbed nanoparticles were placed within the 50 mM imidazoleelution medium and the amount of HSA and Cu2+ released in 1 hwas determined. It was then repeatedly used in adsorption of HSA.Five successive cycles of protein elution and adsorption were car-ried out without noticeable decrease in the adsorption capacityof the PHEMA–IMEO–Cu2+ nanoparticles toward HSA (Fig. 8). Thisillustrates the reversibility of this concept of chelated Cu2+-protein

interactions. It should be noted that there was only negligible Cu2+

release in this case which shows that Cu2+ ions are attached tohistidine molecules on the nanoparticle surface by strong chelateformation.

V. Karakoc et al. / International Journal of Biolog

Fig. 9. SDS-PAGE of serum fractions. The fractions were assayed by SDS-PAGE using10% separating gel (9 cm × 7.5 cm), and 5% stacking minigels were stained with 0.25%(dL(

3

bHhPePdninr

Piews

[

[[[[[

[

[[[

w/v) Coomassie Brillant R 250 in acetic acid–methanol–water (1:5:5, v/v/v) andestained in ethanol–acetic acid–water (1:4:6, v/v/v). Lane 1, 1/10 diluted serum;ane 2, 1/10 diluted serum after adsorption; Lane 3, eluted sample; Lane 4 BiomarkerSigma). Equal amounts of samples were applied to each line.

.4. HSA purification from human plasma

The adsorption of HSA from human plasma was performedatch wise. There was a very low non-specific adsorption ofSA (2.4 mg/g) on the plain PHEMA nanoparticles, while muchigher adsorption values (795 mg/g) were obtained when theHEMA–IMEO–Cu2+ nanoparticles were used. The purity of HSAluted from PHEMA–IMEO–Cu2+ nanoparticles was assayed by SDS-AGE. As clearly seen in Fig. 9, HSA in serum (Lane 1) was almostisappeared in Lane 2 after adsorption onto PHEMA–IMEO–Cu2+

anoparticles. Furthermore, the presence of only bands at Lane 3ndicate the purity of HSA after desorption of PHEMA–IMEO–Cu2+

anoparticles. The purity of HSA obtained was found to be in theange of 88.3%.

It is also worth to note that adsorption of HSA onto the2+

HEMA–IMEO–Cu nanoparticles was higher than those obtained

n the studies in which aqueous solutions were used. This may bexplained as follows; the conformational structure of HSA moleculeithin their native environment (i.e. human plasma) is much more

uitable for specific interaction with Cu2+ ions. The high HSA

[[[

[

ical Macromolecules 45 (2009) 188–193 193

concentration may also contribute to this adsorption capacity dueto the high driving force between the aqueous and solid phases.

4. Conclusion

Recent breakthroughs in nanotechnology have made variousnanostructured materials more affordable for a broader range ofapplications. Various nanostructures have been examined as hostsfor protein adsorption. Nano-structures, generally providing a largesurface area for the adsorption of proteins, have been actively devel-oped for selective separation. Only limited work has been publishedon the application of nanosized particles in the separation of pro-teins. Nanosized particles can produce larger specific surface areaand therefore, may result in high adsorption capacity for proteins.Therefore, it may be useful to synthesize nanosized particles withlarge surface area and utilize them as suitable carriers for the selec-tive separation of proteins. Metal chelate affinity chromatographyis a sensitive and selective method for protein separation. The num-ber of location of surface exposed electron-donating imidazole andthiol groups and their ability to coordinate with chelated metal ionsdictate the adsorption of proteins on metal-immobilized nanopar-ticles. The PHEMA–IMEO–Cu2+ nanoparticles were used for theadsorption of HSA in a batch system. The adsorption behaviourof HSA onto the nanoparticles was investigated using variousexperimental conditions. The properties of the PHEMA–IMEO–Cu2+

nanoparticles seem to provide an adequate approach to selectiveHSA adsorption based on their chelation properties. The IMEO–Cu2+

containing PHEMA nanoparticles revealed good adsorption proper-ties as a nano-support and will be useful in the protein separationtechnology.

References

[1] W. Norbert, Fundamental of Clinical Chemistry, W.B. Saunders, London, 1976.[2] M.M. Wilkes, R.J. Navickis, Ann. Intern. Med. 135 (2001) 149.[3] E.J. Cohn, L.E. Strong, W.L. Hughes, D.J. Mulford, J.N. Ashworth, M. Melin, H.L.

Taylor, J. Am. Chem. Soc. 68 (1946) 459.[4] J.F. Stotz, C. Rivat, C. Geschier, P. Colosett, F. Streiff, Swiss Biotech. 8 (1990) 7.[5] A. Kassab, H. Yavuz, M. Odabasi, A. Denizli, J Chromatogr. B 746 (2000) 123.[6] E.B. Altıntas, H. Yavuz, R. Say, A. Denizli, J. Biomater. Sci. Polym. Ed. 17 (2006)

213.[7] N. Öztürk, S. Akgöl, M. Arısoy, A. Denizli, Sep. Purif. Technol. 58 (2007) 83.[8] A. Denizli, E. Piskin, J. Biochem. Biophys. Methods 49 (2001) 391.[9] P.Y. Huang, R.G. Carbonell, Biotechnol. Bioeng. 63 (1999) 633.10] C.G. Gomez, M.C. Strumia, J. Polym. Sci. Polym. Chem. 46 (2008) 2557.

[11] S. Ozkara, B. Garipcan, E. Piskin, A. Denizli, J. Biomater. Sci. Polym. Ed. 14 (2003)761.

12] E.K.M. Ueda, P.W. Gout, L. Morganti, J. Chromatogr. A 988 (2003) 1.13] S. Emir, R. Say, H. Yavuz, A. Denizli, A. Biotechnol. Prog. 20 (2004) 223.14] R. Gutierrez, E.M. Martin del Valle, M.A. Galan, Sep. Purif. Rev. 35 (2007) 71.15] E.B. Altintas, N. Tuzmen, L. Uzun, A. Denizli, Ind. Eng. Chem. Res. 46 (2007) 7802.16] M.B. Ribeiro, M. Vijayalakshmi, D.T. Balvay, S.M.A. Bueno, J. Chromatogr. B. 86

(2008) 64.17] P. Jain, L. Sun, J. Dai, G.L. Baker, M.L. Bruening, Biomacromolecules 8 (2007)

3102.18] S. Akgöl, D. Türkmen, A. Denizli, J. Appl. Polym. Sci. 93 (2004) 2669.19] Z.Y. Ma, Y.P. Guan, X.Q. Liu, H.Z. Liu, Langmuir 21 (2005) 6987.20] D. Türkmen, N. Öztürk, S. Akgöl, A. Elkak, A. Denizli, Biotechnol. Prog. 24 (2008)

1297.21] D. Türkmen, H. Yavuz, A. Denizli, Int. J. Biol. Macromol. 38 (2006) 126.22] N. Öztürk, N. Bereli, S. Akgöl, A. Denizli, Coll. Surf. B 67 (2008) 14.23] E. Harlow, D. Lane, Antiibodies: A Laboratory Manual, Cold Spring Harbor Lab-

oratory, New York, 1988.24] G.Y. Lee, C.H. Chen, T.H. Wang, W.C. Lee, Anal. Biochem. 312 (2003) 235.