Appl Biochem Biotechnol (2015) 175:454–468 DOI 10.1007/s12010-014-1273-8
A Novel Affinity Disks for Bovine Serum Albumin Purification Nalan Tuzmen & Tülden Kalburcu & Deniz Aktaş Uygun & Sinan Akgol & Adil Denizli
Received: 6 May 2014 / Accepted: 23 September 2014 / Published online: 12 October 2014 # Springer Science+Business Media New York 2014
Abstract The adsorption characteristics of bovine serum albumin (BSA) onto the supermacroporous poly(hydroxyethylmethacrylate)-Reactive Green 19 [p(HEMA)-RG] cryogel disks have been investigated in this paper. p(HEMA) cryogel disks were prepared by radical polymerization initiated by N,N,N′,N′-tetramethylene diamine (TEMED) and ammonium persulfate (APS) pair in an ice bath. Reactive Green (RG) 19 was covalently attached to the p(HEMA) cryogel disks. These disks were used in BSA adsorption studies to interrogate the effects of pH, initial protein concentration, ionic strength, and temperature. BSA adsorption capacity of the p(HEMA)-RG cryogel disk was significantly improved after the incorporation of RG. Adsorption capacity reached a plateau value at about 0.8 mg/mL at pH 4.0. The amount of adsorbed BSA decreased from 37.7 to 13.9 mg/g with increasing NaCl concentration. The enthalpy of BSA adsorption onto the p(HEMA)-RG cryogel disk was calculated as −58.4 kJ/ mol. The adsorption equilibrium isotherm was fitted well by the Freundlich model. BSA was desorbed from cryogel disks (over 90 %) using 0.5 M NaSCN, and the purity of desorbed BSA was confirmed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The experimental results showed that the p(HEMA)-RG cryogel disks have potential for the quick protein separation and purification process. Keywords Bovine serum albumin . Affinity adsorption . Cryogels . Dye ligand . Blood
N. Tuzmen (*) Department of Chemistry, Science Faculty, Dokuz Eylül University, 35600 Buca, Izmir, Turkey e-mail: [email protected] T. Kalburcu Department of Chemistry, Arts and Science Faculty, Aksaray University, Aksaray, Turkey D. A. Uygun Department of Chemistry, Arts and Science Faculty, Adnan Menderes University, Aydın, Turkey S. Akgol Department of Biochemistry, Science Faculty, Ege University, İzmir, Turkey A. Denizli Department of Chemistry, Science Faculty, Hacettepe University, Ankara, Turkey
Appl Biochem Biotechnol (2015) 175:454–468
455
Introduction Serum albumins are the most abundant of the proteins in blood plasma, and albumin is one of the most extensively studied proteins in the serum albumin family. It has many functions throughout the body [1]. In addition to its use in cell culture, blocking agents, and protein binding, serum albumin is also used in drug delivery to transport pharmaceutical drugs like antibiotics and anti-inflammatories [2]. It is frequently used as an excipient in pharmaceuticals because of its ability to stabilize other proteins in solution [3]. Serum albumin has several direct applications in human health including clinical applications for burn victims, hemorrhages, malnourishment, and liver or kidney failure [4]. Its availability, inherent stability, and ligand binding ability make albumin an essential component in the production of many therapeutic drugs and diagnostic tests [1, 5]. Nowadays, it has been pointed out in the literature that bovine serum albumin (BSA) has aroused the interest of researchers greatly due to its great abundance in blood plasma proteins, as well as its wide applications in many fields [6]. It has been widely studied because of its stability and low cost, unusual ligand-binding properties, and particularly its structural homology with human serum albumin (HSA) [7]. The rapid development of biotechnology and biomedicine requires more reliable and efficient separation technologies for isolation and purification of the biomolecules, such as proteins, enzymes, and nucleic acids. Because of its high specific selectivity, affinity chromatography has been extensively used to isolate various biomolecules [8–10]. Recently, dye-ligand affinity chromatography has been used extensively in laboratory- and large-scale protein purification [11–15]. Dye ligands are commercially available, inexpensive, and can easily be immobilized, especially on matrices bearing hydroxyl groups. Although dyes are all synthetic in nature, they are still classified as affinity ligands because they interact with the active sites of many proteins mimicking the structure of the substrates, cofactors, or binding agents for those proteins. A number of textile dyes, known as reactive dyes, have been used for protein purification. Most of these reactive dyes consist of a chromophore (either azo dyes, anthraquinone, or phathalocyanine), linked to a reactive group (often a mono- or dichlorotriazine ring). The interaction between the dye ligand and proteins can be by complex combination of electrostatic, hydrophobic, and hydrogen bonding [16]. Recently, cryogels have been considered as a novel generation of stationary phases in the separation science and used as a carrier for various biomedical and bioengineering applications [17–22]. Their supermacroporous structures with interconnected pores offer a unique combination of high interconnected porosity, high diffusivity, and high mechanical strength [23]. The interconnected system of large pores makes various cryogels promising materials for the production of new chromatographic matrices tailormade for the separation of biological nanoparticles and microparticles (plasmids, viruses, cell organelles, and even intact cells), and also for the implementation as carriers for immobilization of molecules and cells. In addition, cryogels are cheap materials and they can be used as disposable. Cryogels can be synthesized in different formats like monoliths, diskshaped, thin sheets, and beads. Thus, owing to these properties, cryogels present a very interesting chromatographic material allowing the direct separation of proteins from unprocessed crude extracts or even from fermentation broth in the case of extracellulary expressed proteins [17]. The main objective of this work was to investigate the effect of BSA adsorption onto the poly(hydroxyethylmethacrylate) [p(HEMA)] cryogel disks and to purify BSA from bovine plasma by dye affinity chromatography using these cryogel disks. p(HEMA) cryogel was prepared by radical polymerization which proceeds in aqueous solution of monomers frozen between two glass plates (cryopolymerization). Reactive Green (RG) 19 was covalently attached to the p(HEMA) cryogel disks. p(HEMA)-RG cryogel disk was characterized by Fourier transform infrared (FTIR), scanning electron microscopy (SEM), and swelling ratio analysis. Then, BSA adsorption onto the dye-ligand affinity cryogel disks was optimized by varying different parameters such as pH, initial BSA concentration, ionic strength, and
456
Appl Biochem Biotechnol (2015) 175:454–468
temperature. In the last part, desorption of BSA, purity of desorbed BSA, and stability of these materials were tested.
Materials and Method Materials 2-Hydroxyethylmethacrylate (HEMA), N,N,N′,N′-tetramethylene diamine (TEMED), and ammonium persulfate (APS) were supplied by Fluka A.G. (Buchs, Switzerland). N,N′-N,N′methylene-bis(acrylamide) (MBAAm) and BSA were purchased from Sigma (St. Louis, USA). All other chemicals were of the highest purity and used without further purification. Ultrapure water filtered by Millipore S.A.S 67120 Molsheim, France, was used for all experiments unless otherwise stated. Before use, the laboratory glassware was rinsed with water and dried in a dust-free environment. Preparation of p(HEMA) Cryogel HEMA was polymerized by free radical polymerization by using APS and TEMED as the initiator. Water and MBAAm were included in the polymerization recipe as the pore-former and cross-linker, respectively. A typical preparation procedure was as follows: HEMA (2.60 mL) was dissolved in 10.0 mL water. MBAAm (0.566 g) was dissolved in 20 mL water. Second solution was mixed with previous one. The cryogel was then produced by free radical polymerization initiated by TEMED (50 μL) and APS (40 mg). After adding APS, the solution was cooled in an ice bath for 2–3 min. TEMED was added, and the reaction mixture was stirred for 1 min. Then, the reaction mixture was poured between two glass plates separated with 1.5-mm-thick spacers. The polymerization mixture was frozen at −16 °C for 24 h and then thawed at room temperature. After washing with 200 mL of water, the cryogel was cut into circular disks (0.8 cm in diameter) and mass of cryogel disks was determined over a range of 20–25 mg. Cryogelic disks stored in buffer containing 0.02 % sodium azide at 4 °C until use. Immobilization of Dye Ligands onto the p(HEMA) Cryogel RG 19 immobilization was performed as outlined in [24]. Briefly, 100 mg RG 19 dye was dissolved in 30 mL water. Immobilization was performed at room temperature with shaking (200 rpm). Cryogel disks were incubated at 60 °C for 1 h with shaking, followed by the addition of 0.15 g NaCl and 30-min incubation with shaking. The temperature was then raised to 70 °C, and 0.015 g Na2CO3 was added to the dye solution and incubated at 70 °C for 2 h with shaking. After the dye immobilization progress, the cryogels were removed from the dye solution and washed with water until it ran clear (Fig. 1). RG 19 concentrations in the initial solution and at equilibrium were calculated from a calibration curve based on absorbance measured with a UV spectrophotometer (Schimadzu 1601, Japan) at 630 nm. Finally, cryogel disks were washed with first with 1.0 M NaCl and then with ethanol. RG 19 immobilized cryogel disks were stored in 0.02 % sodium azide at 4 °C. Characterization of the p(HEMA)-RG Cryogel To determine the swelling ratio of the cryogel disks (S), they were dried in a vacuum oven at 55 °C, 100 mbar, and the masses of the dried cryogel disks were then determined (mdried gel).
Appl Biochem Biotechnol (2015) 175:454–468
457
Fig. 1 a Chemical structure of RG 19. b Optical photograph of p(HEMA) and p(HEMA)-RG cryogel disks
Masses of swollen cryogel disks were determined regularly for a 24-h period (mwet swelling ratio was calculated as follows: S ¼ mwet gel –mdried gel =mdried gel
gel).
The ð1Þ
The percentages of porosity and porosity for macropores were calculated according to the following equation: %Porosity ¼ mswollen gel −mwater bound =mswollen gel 100 ð2Þ Porosity for macropores ¼
mswollen gel −msqueezed gel =mswollen gel 100
ð3Þ
The morphology of a cross section of the dried cryogel disk was investigated by SEM. The sample was fixed in 2.5 % glutaraldehyde in 0.15 M sodium cacodylate buffer overnight and postfixed in 1 % osmium tetroxide for 1 h. Then, the sample was dehydrated stepwise in ethanol and transferred to a critical point drier set to +10 °C where the ethanol was exchanged for liquid carbon dioxide as a transitional fluid. The temperature was then raised to +40 °C and the pressure to about 100 bar. 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 the final dried cryogel sample. Finally, it was coated with gold–palladium (40:60) and examined using a JEOL JSM 5600 SEM (Tokyo, Japan). FTIR spectra of the p(HEMA) and p(HEMA)-RG cryogel disks were obtained by using a FTIR spectrophotometer (FTIR 8000 Series, Shimadzu, Japan). The dried disk (about 0.01 g) was thoroughly mixed with KBr (0.1 g, IR Grade, Merck, Germany) and pressed into a pellet, and the spectrum was then recorded. BSA Adsorption from Aqueous Solutions In the adsorption experiments, the effects of contact time, initial BSA concentration, pH, temperature, and ionic strength on the adsorption capacity of p(HEMA)-RG cryogel disks were determined. BSA adsorption on the p(HEMA)-RG cryogel disk was studied over a range of pHs (3.5–8.0), temperatures (15–45 °C), BSA concentrations (0.1–1.5 mg/mL), and ionic strengths (0.0–1.0 M NaCl). Cryogel disks weighing 20–25 mg were swelled up and conditioned with sample buffer, and then 10-mL BSA solution was added to swelled cryogel disks. Adsorption experiments were conducted in a thermostated shaker (150 rpm) for 60 min, the equilibrium period for the adsorption of BSA at room temperature. Initial and final protein concentrations were determined by UV spectrophotometry (Schimadzu 1601, Japan) at 280 nm. BSA adsorption experiments were performed in three replicates. For each set of data, standard statistical methods were used to determine the mean values and standard deviations. Confidence intervals of 95 % were calculated for each set of samples in order to determine the margin of error.
458
Appl Biochem Biotechnol (2015) 175:454–468
Desorption Studies To determine the desorption ratio, 0.5 M NaSCN was used as desorption agent. BSA adsorbed p(HEMA)-RG cryogel disks were incubated in the desorption medium at room temperature for 1 h with shaking (150 rpm). The desorption of the protein from the cryogel disk was evaluated by comparing the amounts of protein eluted and adsorbed. Desorption ratioð%Þ ¼
Amount of BSA desorbed 100 Amount of BSA adsorbed
ð4Þ
To test the repeated use of the p(HEMA)-RG cryogel disks, BSA adsorption–desorption cycle was repeated for five times using the same cryogel. The cryogel disks were washed with deionized water for 1 h and reequilibrated in the adsorption buffer and were then reused for protein adsorption. BSA Purification from Bovine Plasma Bovine blood was obtained thoroughly from a slaughterhouse of Izmir Municipality. No preservatives were added to the samples. Bovine blood was collected into EDTA-containing vacutainers, and red blood cells were separated from plasma by centrifugation at 4,000g for 30 min at room temperature, then filtered (3 mm Sartorius filter) and frozen at −20 °C. Before use, the plasma was thawed for 1 h at 37 °C. Before application, the viscous sample was diluted with 50 mM phosphate buffer containing 0.1 M NaCl (pH 7.4). Dilution ratios were 1/ 50 and 1/100. Five milliliters of the diluted bovine plasma was incubated with p(HEMA)-RG cryogel disks weighing about 20–25 mg at a shaking rate of 150 rpm for 1 h. The initial and final concentrations of BSA in serum before and after adsorption processes were determined by bromocresol green (BCG) dye method at 628 nm [25]. Adsorbed BSA was desorbed from p(HEMA)-RG cryogel disks by using 0.5 M NaSCN with shaking at 150 rpm for 1 h. In order to test the purity, gel electrophoresis was carried out as described in details previously [26]. The procedure could be summarized as follows: The desorbed BSA samples was applied to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) using 10 % separating gel (9 cm×7.5 cm), and 5 % stacking minigels were stained with 0.25 % (w/v) Coomassie Brillant R 250 in acetic acid–methanol–water (1:5:5, v/v/v) and destained in ethanol–acetic acid–water (1:4:6, v/v/v). Electrophoresis was run for 2 h with a voltage of 110 V.
Results and Discussion Characterization of p(HEMA)-RG Cryogel p(HEMA) cryogel disk was produced by polymerization in the frozen state of the HEMA in the presence of APS/TEMED as an initiator/activator pair. The functional hydroxyl groups on the surface of the pores in the cryogel disks allowed their modification with the ligand RG 19. RG 19 is a dichlorotriazine dye (Fig. 1), and it contains six sulfonic acid groups, one primary and four secondary amino groups. RG 19 was covalently attached to the p(HEMA) cryogel disk via the reaction between the chloride groups of the dye molecules and the hydroxyl groups of the HEMA. The incorporation rate of RG was found to be 0.150± 0.004 μmol RG/disk. The visual observations (the color of the cryogel) ensured attachment
Appl Biochem Biotechnol (2015) 175:454–468
459
of dye molecules (Fig. 1). The p(HEMA)-RG cryogel disk was extensively washed with methanol until to ensure that there was no dye leakage in any medium used throughout this study. p(HEMA)-RG cryogel disks produced in this manner have a porous structure and thin polymer walls, with large continuous interconnected pores (10–100 μm in diameter) that provide channels through which the mobile phase can flow (Fig. 2). The pore size of the matrix was much larger than the size of the protein molecules, allowing them to easily pass through. BSA is 14 nm×4 nm×4 nm, and its molecular weight is 66.5 kDa. As a result of the convective flow of the solution through the pores, the mass transfer resistance is practically negligible. The equilibrium swelling degree of the p(HEMA) cryogel disk was 4.8±0.4 g H2O/ g cryogel. The p(HEMA)-RG cryogel disk is opaque and sponge-like. This cryogel disk could be easily compressed by hand to remove the water that had accumulated inside the pores. When the compressed piece of cryogel was submerged in water, it soaked up the water and was restored to its original size and shape within 1–2 s. The FTIR spectra of the p(HEMA) and p(HEMA)-RG cryogel disks are shown in Fig. 3. The FTIR bands observed around 1,160 cm−1 were assigned to symmetric stretching of S–O, as also pointed out on the chemical structure of the RG. The band observed at 3,500 cm−1 was assigned to the –OH functional group. The intensity of the –OH band increased after RG attachment. Also, the split of the band at 3,400– 3,600 cm−1 indicates SO3H and NH2 groups. These bands show the attachment of RG within the p(HEMA) cryogel. Visual observation (the color of the cryogel) confirmed the attachment of dye molecules. Optimization of BSA Adsorption Effect of Contact Time To determine the effect of contact time on BSA adsorption onto p(HEMA)-RG cryogel disk, adsorption studies were performed from 0 to 120 min. As shown in Fig. 4, BSA adsorption
Fig. 2 SEM micrographs of p(HEMA) cryogel disk
460
Appl Biochem Biotechnol (2015) 175:454–468
Fig. 3 FTIR spectrum of a p(HEMA) and b p(HEMA)-RG cryogel disks
increased with time and reached a plateau of saturation at 60 min. Therefore, all adsorption studies were performed at 60 min. Effect of pH The optimal pH values for the adsorption of BSA (0.8 mg/mL) onto p(HEMA)-RG cryogel disk were investigated in the pH range of 3.5–8.0 at 25 °C. As shown in Fig. 5, the maximum adsorption of BSA was observed at pH 4.0. In some studies reported in the literature, higher adsorption capacities for BSA was observed at pH 5 which is the isoelectric point of BSA [22, 27, 28]; however, in our study adsorption capacity was higher at pH 4. This observation may be explained with the effects of buffer conditions used, ion species (anion–cation effects), and ionic strengths. Significantly lower adsorption capacities were obtained in more acidic and in
Fig. 4 Effect of contact time on BSA adsorption onto the p(HEMA)-RG cryogel disks (T 25 °C; C 0.8 mg/mL; mdisk:20–25 mg)
Appl Biochem Biotechnol (2015) 175:454–468
461
Fig. 5 Effect of pH on BSA adsorption onto p(HEMA)-RG cryogel disks (contact time 60 min; T 25 °C; C 0.8 mg/mL; mdisk 20–25 mg)
more alkaline pH regions. The decrease in the BSA adsorption capacity in more acidic and more alkaline pH regions may be due to the bilateral electrostatic repulsion of protein molecules since they are in the same electric charge (positive at acidic pH and negative at basic pH). Effect of the Initial BSA Concentration Figure 6 shows both non-specific and specific binding of BSA onto the p(HEMA) and p(HEMA)-RG cryogel disks. It is well known that one of the main requirements in dye affinity chromatography is the specificity of the affinity adsorbent for the target molecule. The non-specific interaction between the support, which is the p(HEMA) cryogel disk in the present case, and the molecules to be adsorbed, which are the BSA molecules here should be minimum to consider the interaction as specific. As shown in this figure, negligible amount of BSA (0.9±0.1 mg/g) was adsorbed non-specifically on the p(HEMA) cryogel disks, while dye immobilization significantly increased the BSA coupling capacity of the cryogel (up to 37.7±4.2 mg/g). We observed that the amount of adsorbed protein increased first with the initial concentration of BSA then reached a plateau value (about 0.8 mg/mL) which represents saturation of the active adsorption sites (which are available and accessible for BSA) on the
Fig. 6 Effect of initial BSA concentration on BSA adsorption onto p(HEMA)-RG cryogel disks (contact time 60 min; T 25 °C; pH 4.0; mdisk 20–25 mg)
462
Appl Biochem Biotechnol (2015) 175:454–468
cryogel. This increase in the BSA binding capacity may have resulted from cooperative effect of different interaction mechanisms such as hydrophobic, electrostatic, and hydrogen bonding. Adsorption Isotherms In order to obtain information about the properties and mechanism of the sorption process, the experimental results of BSA adsorption onto the p(HEMA)-RG cryogel disks were represented by adsorption isotherms and fitted with four model equations (Langmuir, Freundlich, Temkin, and Dubinin–Radushkevich). The Langmuir isotherm model [29] is based on the assumption that the adsorption process takes place on a homogeneous surface, assuming monolayer adsorption onto a surface with a finite number of identical sites, so a monolayer of adsorbate is formed at saturation on the adsorbent surface. The Langmuir isotherm model is given by the following expression: Qeq ¼
Qmax bC eq 1 þ bC eq
ð5Þ
where Qmax is the maximum adsorption capacity (mg/g), Ceq is the equilibrium concentration of BSA at the equilibrium time (mg/L), and b is the Langmuir constant (L/mg). Freundlich [30] presented a fairly satisfactory empirical model to describe non-ideal adsorption on heterogeneous surfaces, as well as multiplayer adsorption. It also assumes that as the adsorbate concentration in solution increases, so too does the concentration of adsorbate on the adsorbent surface and, therefore, has an exponential expression. Qeq ¼ K f C eq 1=n
ð6Þ
where Ceq represents the equilibrium concentration of BSA at the equilibrium time (mg/L), n is the Freundlich constant related to adsorption intensity (dimensionless), and Kf, the Freundlich constant, represents the relative adsorption capacity (mg/g). On the other hand, the Temkin isotherm was derived assuming that the fall in the heat of sorption is linear rather than logarithmic, as implied in the Freundlich equation. It is expressed as follows [31]: Qeq ¼ K T lnC eq þ K T ln f
ð7Þ
where Qeq is the equilibrium mass adsorbed per unit mass of cryogel, Ceq is the equilibrium concentration of protein, and f and KT are constants. In this work, the correlation obtained from the fitting of the Freundlich model (Table 1) was better than the fit using either the Langmuir or Temkin models. The Freundlich model is an empirical model, which allows for molecular interactions between the surface and adsorbate, as well as adsorbate–adsorbate interactions in solution. In contrast, the Langmuir model is based on a reversible adsorption process involving the exchange of adsorbed and free proteins in the absence of intermolecular interactions and the Temkin model, which is a chemical adsorption model based on strong adsorbate–substrate interactions. The results of the modeling therefore indicated that the adsorption of BSA to the p(HEMA)-RG cryogel disks was more complicated than the ideal process described
Appl Biochem Biotechnol (2015) 175:454–468
463
Table 1 Langmuir, Freundlich, Temkin, and Dubinin–Radushkevich isotherm constants for BSA adsorption Langmuir
Freundlich
Temkin
Dubinin–Radushkevich
Qmax 119.0 mg/g
Kf 66.5 mg/g
KT 14.5 L/mol
Qmax 44.6 mg/g
b 0.8 L/mg
n 1.2
f 19.9 J/mol
B 4×10–8 mol2/J2
R2 0.8481
R2 0.9911
R2 0.9433
R2 0.9741
by the Langmuir model and that both protein–protein interactions and protein–surface interactions were important in the adsorption process. The Freundlich constants, Kf and n were calculated to be 66.5 and 1.2, respectively. Kf and n are indicators of adsorption capacity and adsorption intensity [32]. A value of n>1 for the RG ligand indicates positive cooperativity in bonding while a Kf value of 66.5 suggests easy adsorption of BSA from the adsorption medium onto the p(HEMA)-RG cryogel disks. The Dubinin–Radushkevich equation [33] was used in order to gain insight on whether BSA adsorption on the p(HEMA)-RG cryogel disk is due to physical or chemical interactions. The Dubinin–Radushkevich isotherm is expressed by the following equation: ln Qeq ¼ lnQmax −B½ε2
ð8Þ
where Qmax represents the maximum amount of protein adsorbed onto the p(HEMA)-RG cryogel (mg/g), B is a constant related to the sorption energy (mol2/J2), and ε is the Polanyi potential, given by the following equation: ε ¼ RT ln 1 þ 1=C eq ð9Þ where Ceq is the equilibrium concentration of the sorbate (mg/L), R is the ideal gas constant (8.314 J/mol K), and T is the temperature (Kelvin). The calculated parameters are shown in Table 1. Energy related to adsorption can be found from the following equation [34]: E ¼ B−1=2 =√2
ð10Þ
Energy values ranging between 1 and 8 kJ/mol indicate that the sorption is due to physical interactions between adsorbent and adsorbate [35]. The values of E (3.6 kJ/mol) calculated from Eq. 10 indicate that physisorption due to weak van der Waals forces plays a significant role in the adsorption process. Effect of Ionic Strength To determine the effect of ionic strength on BSA adsorption, adsorption studies were carried out in the range of 0.0–1.0 M NaCl. As shown in Fig. 7, the amount of BSA adsorbed onto the p(HEMA)-RG cryogel disks decreased significantly (63.07 %, p
A Novel Affinity Disks for Bovine Serum Albumin Purification Nalan Tuzmen & Tülden Kalburcu & Deniz Aktaş Uygun & Sinan Akgol & Adil Denizli
Received: 6 May 2014 / Accepted: 23 September 2014 / Published online: 12 October 2014 # Springer Science+Business Media New York 2014
Abstract The adsorption characteristics of bovine serum albumin (BSA) onto the supermacroporous poly(hydroxyethylmethacrylate)-Reactive Green 19 [p(HEMA)-RG] cryogel disks have been investigated in this paper. p(HEMA) cryogel disks were prepared by radical polymerization initiated by N,N,N′,N′-tetramethylene diamine (TEMED) and ammonium persulfate (APS) pair in an ice bath. Reactive Green (RG) 19 was covalently attached to the p(HEMA) cryogel disks. These disks were used in BSA adsorption studies to interrogate the effects of pH, initial protein concentration, ionic strength, and temperature. BSA adsorption capacity of the p(HEMA)-RG cryogel disk was significantly improved after the incorporation of RG. Adsorption capacity reached a plateau value at about 0.8 mg/mL at pH 4.0. The amount of adsorbed BSA decreased from 37.7 to 13.9 mg/g with increasing NaCl concentration. The enthalpy of BSA adsorption onto the p(HEMA)-RG cryogel disk was calculated as −58.4 kJ/ mol. The adsorption equilibrium isotherm was fitted well by the Freundlich model. BSA was desorbed from cryogel disks (over 90 %) using 0.5 M NaSCN, and the purity of desorbed BSA was confirmed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The experimental results showed that the p(HEMA)-RG cryogel disks have potential for the quick protein separation and purification process. Keywords Bovine serum albumin . Affinity adsorption . Cryogels . Dye ligand . Blood
N. Tuzmen (*) Department of Chemistry, Science Faculty, Dokuz Eylül University, 35600 Buca, Izmir, Turkey e-mail: [email protected] T. Kalburcu Department of Chemistry, Arts and Science Faculty, Aksaray University, Aksaray, Turkey D. A. Uygun Department of Chemistry, Arts and Science Faculty, Adnan Menderes University, Aydın, Turkey S. Akgol Department of Biochemistry, Science Faculty, Ege University, İzmir, Turkey A. Denizli Department of Chemistry, Science Faculty, Hacettepe University, Ankara, Turkey
Appl Biochem Biotechnol (2015) 175:454–468
455
Introduction Serum albumins are the most abundant of the proteins in blood plasma, and albumin is one of the most extensively studied proteins in the serum albumin family. It has many functions throughout the body [1]. In addition to its use in cell culture, blocking agents, and protein binding, serum albumin is also used in drug delivery to transport pharmaceutical drugs like antibiotics and anti-inflammatories [2]. It is frequently used as an excipient in pharmaceuticals because of its ability to stabilize other proteins in solution [3]. Serum albumin has several direct applications in human health including clinical applications for burn victims, hemorrhages, malnourishment, and liver or kidney failure [4]. Its availability, inherent stability, and ligand binding ability make albumin an essential component in the production of many therapeutic drugs and diagnostic tests [1, 5]. Nowadays, it has been pointed out in the literature that bovine serum albumin (BSA) has aroused the interest of researchers greatly due to its great abundance in blood plasma proteins, as well as its wide applications in many fields [6]. It has been widely studied because of its stability and low cost, unusual ligand-binding properties, and particularly its structural homology with human serum albumin (HSA) [7]. The rapid development of biotechnology and biomedicine requires more reliable and efficient separation technologies for isolation and purification of the biomolecules, such as proteins, enzymes, and nucleic acids. Because of its high specific selectivity, affinity chromatography has been extensively used to isolate various biomolecules [8–10]. Recently, dye-ligand affinity chromatography has been used extensively in laboratory- and large-scale protein purification [11–15]. Dye ligands are commercially available, inexpensive, and can easily be immobilized, especially on matrices bearing hydroxyl groups. Although dyes are all synthetic in nature, they are still classified as affinity ligands because they interact with the active sites of many proteins mimicking the structure of the substrates, cofactors, or binding agents for those proteins. A number of textile dyes, known as reactive dyes, have been used for protein purification. Most of these reactive dyes consist of a chromophore (either azo dyes, anthraquinone, or phathalocyanine), linked to a reactive group (often a mono- or dichlorotriazine ring). The interaction between the dye ligand and proteins can be by complex combination of electrostatic, hydrophobic, and hydrogen bonding [16]. Recently, cryogels have been considered as a novel generation of stationary phases in the separation science and used as a carrier for various biomedical and bioengineering applications [17–22]. Their supermacroporous structures with interconnected pores offer a unique combination of high interconnected porosity, high diffusivity, and high mechanical strength [23]. The interconnected system of large pores makes various cryogels promising materials for the production of new chromatographic matrices tailormade for the separation of biological nanoparticles and microparticles (plasmids, viruses, cell organelles, and even intact cells), and also for the implementation as carriers for immobilization of molecules and cells. In addition, cryogels are cheap materials and they can be used as disposable. Cryogels can be synthesized in different formats like monoliths, diskshaped, thin sheets, and beads. Thus, owing to these properties, cryogels present a very interesting chromatographic material allowing the direct separation of proteins from unprocessed crude extracts or even from fermentation broth in the case of extracellulary expressed proteins [17]. The main objective of this work was to investigate the effect of BSA adsorption onto the poly(hydroxyethylmethacrylate) [p(HEMA)] cryogel disks and to purify BSA from bovine plasma by dye affinity chromatography using these cryogel disks. p(HEMA) cryogel was prepared by radical polymerization which proceeds in aqueous solution of monomers frozen between two glass plates (cryopolymerization). Reactive Green (RG) 19 was covalently attached to the p(HEMA) cryogel disks. p(HEMA)-RG cryogel disk was characterized by Fourier transform infrared (FTIR), scanning electron microscopy (SEM), and swelling ratio analysis. Then, BSA adsorption onto the dye-ligand affinity cryogel disks was optimized by varying different parameters such as pH, initial BSA concentration, ionic strength, and
456
Appl Biochem Biotechnol (2015) 175:454–468
temperature. In the last part, desorption of BSA, purity of desorbed BSA, and stability of these materials were tested.
Materials and Method Materials 2-Hydroxyethylmethacrylate (HEMA), N,N,N′,N′-tetramethylene diamine (TEMED), and ammonium persulfate (APS) were supplied by Fluka A.G. (Buchs, Switzerland). N,N′-N,N′methylene-bis(acrylamide) (MBAAm) and BSA were purchased from Sigma (St. Louis, USA). All other chemicals were of the highest purity and used without further purification. Ultrapure water filtered by Millipore S.A.S 67120 Molsheim, France, was used for all experiments unless otherwise stated. Before use, the laboratory glassware was rinsed with water and dried in a dust-free environment. Preparation of p(HEMA) Cryogel HEMA was polymerized by free radical polymerization by using APS and TEMED as the initiator. Water and MBAAm were included in the polymerization recipe as the pore-former and cross-linker, respectively. A typical preparation procedure was as follows: HEMA (2.60 mL) was dissolved in 10.0 mL water. MBAAm (0.566 g) was dissolved in 20 mL water. Second solution was mixed with previous one. The cryogel was then produced by free radical polymerization initiated by TEMED (50 μL) and APS (40 mg). After adding APS, the solution was cooled in an ice bath for 2–3 min. TEMED was added, and the reaction mixture was stirred for 1 min. Then, the reaction mixture was poured between two glass plates separated with 1.5-mm-thick spacers. The polymerization mixture was frozen at −16 °C for 24 h and then thawed at room temperature. After washing with 200 mL of water, the cryogel was cut into circular disks (0.8 cm in diameter) and mass of cryogel disks was determined over a range of 20–25 mg. Cryogelic disks stored in buffer containing 0.02 % sodium azide at 4 °C until use. Immobilization of Dye Ligands onto the p(HEMA) Cryogel RG 19 immobilization was performed as outlined in [24]. Briefly, 100 mg RG 19 dye was dissolved in 30 mL water. Immobilization was performed at room temperature with shaking (200 rpm). Cryogel disks were incubated at 60 °C for 1 h with shaking, followed by the addition of 0.15 g NaCl and 30-min incubation with shaking. The temperature was then raised to 70 °C, and 0.015 g Na2CO3 was added to the dye solution and incubated at 70 °C for 2 h with shaking. After the dye immobilization progress, the cryogels were removed from the dye solution and washed with water until it ran clear (Fig. 1). RG 19 concentrations in the initial solution and at equilibrium were calculated from a calibration curve based on absorbance measured with a UV spectrophotometer (Schimadzu 1601, Japan) at 630 nm. Finally, cryogel disks were washed with first with 1.0 M NaCl and then with ethanol. RG 19 immobilized cryogel disks were stored in 0.02 % sodium azide at 4 °C. Characterization of the p(HEMA)-RG Cryogel To determine the swelling ratio of the cryogel disks (S), they were dried in a vacuum oven at 55 °C, 100 mbar, and the masses of the dried cryogel disks were then determined (mdried gel).
Appl Biochem Biotechnol (2015) 175:454–468
457
Fig. 1 a Chemical structure of RG 19. b Optical photograph of p(HEMA) and p(HEMA)-RG cryogel disks
Masses of swollen cryogel disks were determined regularly for a 24-h period (mwet swelling ratio was calculated as follows: S ¼ mwet gel –mdried gel =mdried gel
gel).
The ð1Þ
The percentages of porosity and porosity for macropores were calculated according to the following equation: %Porosity ¼ mswollen gel −mwater bound =mswollen gel 100 ð2Þ Porosity for macropores ¼
mswollen gel −msqueezed gel =mswollen gel 100
ð3Þ
The morphology of a cross section of the dried cryogel disk was investigated by SEM. The sample was fixed in 2.5 % glutaraldehyde in 0.15 M sodium cacodylate buffer overnight and postfixed in 1 % osmium tetroxide for 1 h. Then, the sample was dehydrated stepwise in ethanol and transferred to a critical point drier set to +10 °C where the ethanol was exchanged for liquid carbon dioxide as a transitional fluid. The temperature was then raised to +40 °C and the pressure to about 100 bar. 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 the final dried cryogel sample. Finally, it was coated with gold–palladium (40:60) and examined using a JEOL JSM 5600 SEM (Tokyo, Japan). FTIR spectra of the p(HEMA) and p(HEMA)-RG cryogel disks were obtained by using a FTIR spectrophotometer (FTIR 8000 Series, Shimadzu, Japan). The dried disk (about 0.01 g) was thoroughly mixed with KBr (0.1 g, IR Grade, Merck, Germany) and pressed into a pellet, and the spectrum was then recorded. BSA Adsorption from Aqueous Solutions In the adsorption experiments, the effects of contact time, initial BSA concentration, pH, temperature, and ionic strength on the adsorption capacity of p(HEMA)-RG cryogel disks were determined. BSA adsorption on the p(HEMA)-RG cryogel disk was studied over a range of pHs (3.5–8.0), temperatures (15–45 °C), BSA concentrations (0.1–1.5 mg/mL), and ionic strengths (0.0–1.0 M NaCl). Cryogel disks weighing 20–25 mg were swelled up and conditioned with sample buffer, and then 10-mL BSA solution was added to swelled cryogel disks. Adsorption experiments were conducted in a thermostated shaker (150 rpm) for 60 min, the equilibrium period for the adsorption of BSA at room temperature. Initial and final protein concentrations were determined by UV spectrophotometry (Schimadzu 1601, Japan) at 280 nm. BSA adsorption experiments were performed in three replicates. For each set of data, standard statistical methods were used to determine the mean values and standard deviations. Confidence intervals of 95 % were calculated for each set of samples in order to determine the margin of error.
458
Appl Biochem Biotechnol (2015) 175:454–468
Desorption Studies To determine the desorption ratio, 0.5 M NaSCN was used as desorption agent. BSA adsorbed p(HEMA)-RG cryogel disks were incubated in the desorption medium at room temperature for 1 h with shaking (150 rpm). The desorption of the protein from the cryogel disk was evaluated by comparing the amounts of protein eluted and adsorbed. Desorption ratioð%Þ ¼
Amount of BSA desorbed 100 Amount of BSA adsorbed
ð4Þ
To test the repeated use of the p(HEMA)-RG cryogel disks, BSA adsorption–desorption cycle was repeated for five times using the same cryogel. The cryogel disks were washed with deionized water for 1 h and reequilibrated in the adsorption buffer and were then reused for protein adsorption. BSA Purification from Bovine Plasma Bovine blood was obtained thoroughly from a slaughterhouse of Izmir Municipality. No preservatives were added to the samples. Bovine blood was collected into EDTA-containing vacutainers, and red blood cells were separated from plasma by centrifugation at 4,000g for 30 min at room temperature, then filtered (3 mm Sartorius filter) and frozen at −20 °C. Before use, the plasma was thawed for 1 h at 37 °C. Before application, the viscous sample was diluted with 50 mM phosphate buffer containing 0.1 M NaCl (pH 7.4). Dilution ratios were 1/ 50 and 1/100. Five milliliters of the diluted bovine plasma was incubated with p(HEMA)-RG cryogel disks weighing about 20–25 mg at a shaking rate of 150 rpm for 1 h. The initial and final concentrations of BSA in serum before and after adsorption processes were determined by bromocresol green (BCG) dye method at 628 nm [25]. Adsorbed BSA was desorbed from p(HEMA)-RG cryogel disks by using 0.5 M NaSCN with shaking at 150 rpm for 1 h. In order to test the purity, gel electrophoresis was carried out as described in details previously [26]. The procedure could be summarized as follows: The desorbed BSA samples was applied to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) using 10 % separating gel (9 cm×7.5 cm), and 5 % stacking minigels were stained with 0.25 % (w/v) Coomassie Brillant R 250 in acetic acid–methanol–water (1:5:5, v/v/v) and destained in ethanol–acetic acid–water (1:4:6, v/v/v). Electrophoresis was run for 2 h with a voltage of 110 V.
Results and Discussion Characterization of p(HEMA)-RG Cryogel p(HEMA) cryogel disk was produced by polymerization in the frozen state of the HEMA in the presence of APS/TEMED as an initiator/activator pair. The functional hydroxyl groups on the surface of the pores in the cryogel disks allowed their modification with the ligand RG 19. RG 19 is a dichlorotriazine dye (Fig. 1), and it contains six sulfonic acid groups, one primary and four secondary amino groups. RG 19 was covalently attached to the p(HEMA) cryogel disk via the reaction between the chloride groups of the dye molecules and the hydroxyl groups of the HEMA. The incorporation rate of RG was found to be 0.150± 0.004 μmol RG/disk. The visual observations (the color of the cryogel) ensured attachment
Appl Biochem Biotechnol (2015) 175:454–468
459
of dye molecules (Fig. 1). The p(HEMA)-RG cryogel disk was extensively washed with methanol until to ensure that there was no dye leakage in any medium used throughout this study. p(HEMA)-RG cryogel disks produced in this manner have a porous structure and thin polymer walls, with large continuous interconnected pores (10–100 μm in diameter) that provide channels through which the mobile phase can flow (Fig. 2). The pore size of the matrix was much larger than the size of the protein molecules, allowing them to easily pass through. BSA is 14 nm×4 nm×4 nm, and its molecular weight is 66.5 kDa. As a result of the convective flow of the solution through the pores, the mass transfer resistance is practically negligible. The equilibrium swelling degree of the p(HEMA) cryogel disk was 4.8±0.4 g H2O/ g cryogel. The p(HEMA)-RG cryogel disk is opaque and sponge-like. This cryogel disk could be easily compressed by hand to remove the water that had accumulated inside the pores. When the compressed piece of cryogel was submerged in water, it soaked up the water and was restored to its original size and shape within 1–2 s. The FTIR spectra of the p(HEMA) and p(HEMA)-RG cryogel disks are shown in Fig. 3. The FTIR bands observed around 1,160 cm−1 were assigned to symmetric stretching of S–O, as also pointed out on the chemical structure of the RG. The band observed at 3,500 cm−1 was assigned to the –OH functional group. The intensity of the –OH band increased after RG attachment. Also, the split of the band at 3,400– 3,600 cm−1 indicates SO3H and NH2 groups. These bands show the attachment of RG within the p(HEMA) cryogel. Visual observation (the color of the cryogel) confirmed the attachment of dye molecules. Optimization of BSA Adsorption Effect of Contact Time To determine the effect of contact time on BSA adsorption onto p(HEMA)-RG cryogel disk, adsorption studies were performed from 0 to 120 min. As shown in Fig. 4, BSA adsorption
Fig. 2 SEM micrographs of p(HEMA) cryogel disk
460
Appl Biochem Biotechnol (2015) 175:454–468
Fig. 3 FTIR spectrum of a p(HEMA) and b p(HEMA)-RG cryogel disks
increased with time and reached a plateau of saturation at 60 min. Therefore, all adsorption studies were performed at 60 min. Effect of pH The optimal pH values for the adsorption of BSA (0.8 mg/mL) onto p(HEMA)-RG cryogel disk were investigated in the pH range of 3.5–8.0 at 25 °C. As shown in Fig. 5, the maximum adsorption of BSA was observed at pH 4.0. In some studies reported in the literature, higher adsorption capacities for BSA was observed at pH 5 which is the isoelectric point of BSA [22, 27, 28]; however, in our study adsorption capacity was higher at pH 4. This observation may be explained with the effects of buffer conditions used, ion species (anion–cation effects), and ionic strengths. Significantly lower adsorption capacities were obtained in more acidic and in
Fig. 4 Effect of contact time on BSA adsorption onto the p(HEMA)-RG cryogel disks (T 25 °C; C 0.8 mg/mL; mdisk:20–25 mg)
Appl Biochem Biotechnol (2015) 175:454–468
461
Fig. 5 Effect of pH on BSA adsorption onto p(HEMA)-RG cryogel disks (contact time 60 min; T 25 °C; C 0.8 mg/mL; mdisk 20–25 mg)
more alkaline pH regions. The decrease in the BSA adsorption capacity in more acidic and more alkaline pH regions may be due to the bilateral electrostatic repulsion of protein molecules since they are in the same electric charge (positive at acidic pH and negative at basic pH). Effect of the Initial BSA Concentration Figure 6 shows both non-specific and specific binding of BSA onto the p(HEMA) and p(HEMA)-RG cryogel disks. It is well known that one of the main requirements in dye affinity chromatography is the specificity of the affinity adsorbent for the target molecule. The non-specific interaction between the support, which is the p(HEMA) cryogel disk in the present case, and the molecules to be adsorbed, which are the BSA molecules here should be minimum to consider the interaction as specific. As shown in this figure, negligible amount of BSA (0.9±0.1 mg/g) was adsorbed non-specifically on the p(HEMA) cryogel disks, while dye immobilization significantly increased the BSA coupling capacity of the cryogel (up to 37.7±4.2 mg/g). We observed that the amount of adsorbed protein increased first with the initial concentration of BSA then reached a plateau value (about 0.8 mg/mL) which represents saturation of the active adsorption sites (which are available and accessible for BSA) on the
Fig. 6 Effect of initial BSA concentration on BSA adsorption onto p(HEMA)-RG cryogel disks (contact time 60 min; T 25 °C; pH 4.0; mdisk 20–25 mg)
462
Appl Biochem Biotechnol (2015) 175:454–468
cryogel. This increase in the BSA binding capacity may have resulted from cooperative effect of different interaction mechanisms such as hydrophobic, electrostatic, and hydrogen bonding. Adsorption Isotherms In order to obtain information about the properties and mechanism of the sorption process, the experimental results of BSA adsorption onto the p(HEMA)-RG cryogel disks were represented by adsorption isotherms and fitted with four model equations (Langmuir, Freundlich, Temkin, and Dubinin–Radushkevich). The Langmuir isotherm model [29] is based on the assumption that the adsorption process takes place on a homogeneous surface, assuming monolayer adsorption onto a surface with a finite number of identical sites, so a monolayer of adsorbate is formed at saturation on the adsorbent surface. The Langmuir isotherm model is given by the following expression: Qeq ¼
Qmax bC eq 1 þ bC eq
ð5Þ
where Qmax is the maximum adsorption capacity (mg/g), Ceq is the equilibrium concentration of BSA at the equilibrium time (mg/L), and b is the Langmuir constant (L/mg). Freundlich [30] presented a fairly satisfactory empirical model to describe non-ideal adsorption on heterogeneous surfaces, as well as multiplayer adsorption. It also assumes that as the adsorbate concentration in solution increases, so too does the concentration of adsorbate on the adsorbent surface and, therefore, has an exponential expression. Qeq ¼ K f C eq 1=n
ð6Þ
where Ceq represents the equilibrium concentration of BSA at the equilibrium time (mg/L), n is the Freundlich constant related to adsorption intensity (dimensionless), and Kf, the Freundlich constant, represents the relative adsorption capacity (mg/g). On the other hand, the Temkin isotherm was derived assuming that the fall in the heat of sorption is linear rather than logarithmic, as implied in the Freundlich equation. It is expressed as follows [31]: Qeq ¼ K T lnC eq þ K T ln f
ð7Þ
where Qeq is the equilibrium mass adsorbed per unit mass of cryogel, Ceq is the equilibrium concentration of protein, and f and KT are constants. In this work, the correlation obtained from the fitting of the Freundlich model (Table 1) was better than the fit using either the Langmuir or Temkin models. The Freundlich model is an empirical model, which allows for molecular interactions between the surface and adsorbate, as well as adsorbate–adsorbate interactions in solution. In contrast, the Langmuir model is based on a reversible adsorption process involving the exchange of adsorbed and free proteins in the absence of intermolecular interactions and the Temkin model, which is a chemical adsorption model based on strong adsorbate–substrate interactions. The results of the modeling therefore indicated that the adsorption of BSA to the p(HEMA)-RG cryogel disks was more complicated than the ideal process described
Appl Biochem Biotechnol (2015) 175:454–468
463
Table 1 Langmuir, Freundlich, Temkin, and Dubinin–Radushkevich isotherm constants for BSA adsorption Langmuir
Freundlich
Temkin
Dubinin–Radushkevich
Qmax 119.0 mg/g
Kf 66.5 mg/g
KT 14.5 L/mol
Qmax 44.6 mg/g
b 0.8 L/mg
n 1.2
f 19.9 J/mol
B 4×10–8 mol2/J2
R2 0.8481
R2 0.9911
R2 0.9433
R2 0.9741
by the Langmuir model and that both protein–protein interactions and protein–surface interactions were important in the adsorption process. The Freundlich constants, Kf and n were calculated to be 66.5 and 1.2, respectively. Kf and n are indicators of adsorption capacity and adsorption intensity [32]. A value of n>1 for the RG ligand indicates positive cooperativity in bonding while a Kf value of 66.5 suggests easy adsorption of BSA from the adsorption medium onto the p(HEMA)-RG cryogel disks. The Dubinin–Radushkevich equation [33] was used in order to gain insight on whether BSA adsorption on the p(HEMA)-RG cryogel disk is due to physical or chemical interactions. The Dubinin–Radushkevich isotherm is expressed by the following equation: ln Qeq ¼ lnQmax −B½ε2
ð8Þ
where Qmax represents the maximum amount of protein adsorbed onto the p(HEMA)-RG cryogel (mg/g), B is a constant related to the sorption energy (mol2/J2), and ε is the Polanyi potential, given by the following equation: ε ¼ RT ln 1 þ 1=C eq ð9Þ where Ceq is the equilibrium concentration of the sorbate (mg/L), R is the ideal gas constant (8.314 J/mol K), and T is the temperature (Kelvin). The calculated parameters are shown in Table 1. Energy related to adsorption can be found from the following equation [34]: E ¼ B−1=2 =√2
ð10Þ
Energy values ranging between 1 and 8 kJ/mol indicate that the sorption is due to physical interactions between adsorbent and adsorbate [35]. The values of E (3.6 kJ/mol) calculated from Eq. 10 indicate that physisorption due to weak van der Waals forces plays a significant role in the adsorption process. Effect of Ionic Strength To determine the effect of ionic strength on BSA adsorption, adsorption studies were carried out in the range of 0.0–1.0 M NaCl. As shown in Fig. 7, the amount of BSA adsorbed onto the p(HEMA)-RG cryogel disks decreased significantly (63.07 %, p
