Journal of Chromatography A, 1122 (2006) 28–34
Novel hydrophobic interaction chromatography matrix for specific isolation and simple elution of immunoglobulins (A, G, and M) from porcine serum Gabriela Ramos-Clamont a , Maria del Carmen Candia-Plata b , Roberto Guzman Zamudio c , Luz Vazquez-Moreno a,∗ a
Centro de Investigacion en Alimentacion y Desarrollo A.C. Coordinacion de Ciencia de los Alimentos, Apartado Postal 1735, Hermosillo, Sonora 83000, Mexico b Universidad de Sonora, Departamento de Ciencias Quimico Biologicas, Hermosillo, Sonora 83000, M´ exico c University of Arizona, Department of Chemical & Environmental Engineering, Tucson, AZ 85721-0011, USA Received 4 July 2005; received in revised form 6 April 2006; accepted 11 April 2006 Available online 2 May 2006
Abstract A new, highly acetylated agarose matrix (HA-Sepharose) was synthesized and used as a hydrophobic interaction chromatography (HIC) medium to specifically isolate immunoglobulins (Igs) from porcine serum. Recovery of Igs was in a single step and under mild conditions. HA-Sepharose adsorption was studied in terms of salt, gel acetylation time, flow rate, and protein concentration on the loading buffer. At 0.5 M Na2 SO4 , control with unmodified Sepharose retained a small fraction (0.70 mg/mL of matrix) of serum albumin. On the contrary HA-Sepharose retained primary Igs (IgA, IgG, and 53% of IgM) as revealed by sodium dodecyl sulphate 10% polyacrylamide gel electrophoresis (SDS-PAGE), quantitative radial immunodiffusion and immunodetection. At a flow rate of 1 mL/min, the HA-Sepharose column capacity (3.9 mg/mL of matrix) was similar to the reported capacity for the commercial thiophilic T-gel. However, HA-Sepharose showed higher recovery of IgA and IgM than the T-gel in the same salt conditions, clearly an advantage in terms of immunoglobulin recovery strategies. Acetylation changed the matrix adsorption from albumin to immunoglobulins; thus, the highly acetylated gel rendered recoveries of Igs from unprocessed porcine serum practically free of albumin. © 2006 Elsevier B.V. All rights reserved. Keywords: Blood derivatives; Hydrophobic interaction chromatography; Highly acetylated agarose; Porcine immunoglobulins; T-gel
1. Introduction Hydrophobic interaction chromatography presents an alternative to other chromatographic gels to purify proteins [1–3]. HIC takes advantage of the hydrophobicity of proteins by promoting separation on the basis of hydrophobic interactions between protein non-polar regions and immobilized hydrophobic groups at the stationary phase [4–6]. HIC is encouraged by a high concentration of water structuring salt [7]. In 1990, Porath coined the term salt promoted adsorption chromatography (SPAC) to regroup chromatographic techniques that require a high salt concentration to promote protein adsorption. Protein elution from the matrix is then achieved simply by deletion of the salt from the adsorption buffer [8]. Linear or cyclic alkanes (e.g., octyl- and phenyl-Sepharose) as well as other
∗
Corresponding author. Tel.: +52 662 289 2400; fax: +52 662 280 0058. E-mail address:
[email protected] (L. Vazquez-Moreno).
0021-9673/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2006.04.012
SPAC gels like thiophilic gel (T-gel), mercapthylene pyridinederivatized agarose gel (MP), tricyanoaminopropene–divinyl sulphone agarose gel (DVS-TCP), and bisoxirane-TCP gel were studied in the presence of moderated or high salt concentrations of sodium sulfate. The latter gels displayed differential selectivity for immunoglobulins, while octyl- and phenyl-Sepharose had preference for albumin over immunoglobulins and in minor concentrations for other serum components [9]. Immunoglobulins (Igs) are glycoproteins with antibody activity that constitute 15–20% of total serum proteins. They are important molecules that have many applications in biotechnology as laboratory reagents and for health-promoting purpose [10]. Currently, the separation processes for Igs involve fractionation by cold ethanol, ammonium sulfate or propylene glycol precipitation associated to chromatographic methods based on ion exchange [11,12]. However, these methods suffer from one or several drawbacks that include the poor resolving power of the precipitation methods, and high cost and time-consuming process of the customary chromatographic methods [12].
G. Ramos-Clamont et al. / J. Chromatogr. A 1122 (2006) 28–34
Historically, affinity chromatography using Protein A from Staphylococcus aureus has been the preferred method of IgG purification. Protein A is stable over a wide range of pH (pH 2–11) and is able to refold after treatment with denaturating agents such as urea and guanidine salts [10–12]. However, this ligand suffers from several drawbacks, including highproduction cost and its bacterial origin [12–14]. A known problem with Protein A is leaching from the column; this is of potential concern for in vivo use of Igs preparations, because Protein A is a potent immunomodulator and has proven toxicity in clinical trials [13,14]. In addition, derived from ligand leakage, often a subsequent separation step is required to ensure removal of Protein A from the product streams [15,16]. Low pH conditions used for elution are not always optimal for maintaining immunoglobulin biological activity. Therefore, new and efficient methods for antibody purification are always of interest given the importance of this class of molecules. Here, we describe the use of a highly substituted Sepharose (HA-Sepharose) with short hydrophobic chains [6] for separating immunoglobulins from pig serum under mild conditions. HA-Sepharose apparently involves, in addition to their hydrophobic interactions, a recognition site “character” for more specific Igs separations. 2. Experimental 2.1. Chemicals, materials and instrumentation Sepharose CL 6B was purchased from Amersham Pharmacia Biotech (Uppsala, Sweden), and Radial immunodiffusion VETRID kits, and goat anti-pig Igs (IgA, IgG, and IgM) were from Bethyl Labs (Montgomery, TX, USA). T-gel was purchased from Kem-En-Tec (Copenhagen, Denmark). All other reagents were of analytical grade and purchased from Sigma–Aldrich (St. Louis, MO, USA). Dynamic binding capacity experiments were conducted in HR-5 columns (10 cm × 0.5 cm), Amersham Pharmacia Biotech (Uppsala, Sweden) filled with 1 mL of HA Sepharose or T-gel. All other studies of HA-Sepharose, unmodified Sepharose and T-gel were performed using Bio-Rad Econo columns of 2 ml (10 cm × 0.5 cm), 5 ml (10 cm × 1 cm), and 46 ml (20 cm × 2.5 cm) (Bio-Rad, CA, USA). Adsorption, desorption, and regeneration of all matrix were performed in an Econo Chromatographic System from Bio-Rad. 2.2. Sample preparation Porcine blood was collected in sterile containers at the bleeding line of a certified slaughterhouse located in Hermosillo, M´exico. Blood was allowed to coagulate at 25 ◦ C and serum was separated by decanting, followed by centrifugation at 24,000 × g for 15 min at 4 ◦ C. The fat layer was removed at this time. Aliquots of serum were stored at −40 ◦ C until use. 2.3. Matrix synthesis HA-Sepharose was synthesized according to Vazquez Moreno et al. [6]. Briefly, Sepharose CL-6B (50 g) was washed
29
with 100 mL of absolute ethanol on a glass filter followed by an acetone wash (50 mL) and was suction-dried. The gel was suspended in pyridine (50 mL) containing 0.68 mol of acetic anhydride. Reaction was allowed to proceed for 18, 24, and 36 h under continuous shaking at room temperature (25 ◦ C). After reaction, gels were washed on a sintered glass filter with ethanol (100 mL) and washed several times with water to remove unreacted material and stored at 4 ◦ C. 2.4. Elementary analysis Dried gels (1–2 g) were sent to Desert Analytics (Tucson, AZ, USA) for elementary analysis of carbon, oxygen, and hydrogen content. Samples included Sepharose and HA-Sepharose acetylated for 24 and 36 h, respectively. 2.5. Chromatography conditions For the optimization of chromatographic conditions, HASepharose was equilibrated in different concentrations (0.25, 0.5, and 1 M) of the Na2 SO4 binding buffer, and elution conditions without salt were performed using 10 mM 3-(Nmorpholino) propanesulfonic acid (MOPS) at different pH (6.0, 6.5, 7.2, and 7.6). To assure the removal of proteins from the matrix after chromatographic fractionation, and also to test the stability of the matrix to strong chaotropic salts (guanidine–HCl), HA-Sepharose was washed with two-bed volumes of 4 M guanidine–HCl pH 7.6 and subsequently stored for 7 days in this solution (two-bed volumes). For Igs isolation the following conditions were optimal. HASepharose and its controls (unmodified Sepharose and T-gel) were packed at 1.5 mL/min flow rate and equilibrated with fivebed volumes of 0.5 M Na2 SO4 , 10 mM MOPS, pH 7.6 (Buffer A). Porcine serum molarity was adjusted using an equal volume of 1 M Na2 SO4 , 20 mM MOPS, pH 7.6 and was loaded onto columns. Gels were washed with Buffer A to remove unadsorbed proteins and eluted with 10 mM MOPS buffer, pH 7.2 (Buffer B). Following elution, gels were cleaned with two-bed volumes of 4 M guanidine hydrochloride (guanidine–HCl), pH 7.6, washed with five-bed volumes of distilled water, and then re-equilibrated with buffer A. All chromatographic procedures were monitored by absorbance at 280 nm (Spectronic 21 spectrophotometer, Milton Roy, Rochester, NY, USA), and analyzed for protein content by Bradford [17], using bovine serum albumin (BSA) as a standard. 2.6. Acetylation time effect To establish analytical chromatographic conditions, HASepharose (5 mL bed volume, 10 cm × 1 cm) prepared at three different acetylation times (18, 24, and 36 h) was loaded with 3–5 or 6 mL of porcine serum (80 mg/mL of total protein) and tested according to described in Chromatography conditions. Adsorption capacity was estimated as the total protein (mg) bound divided by bed volume (mL). All determinations were done in triplicate.
30
G. Ramos-Clamont et al. / J. Chromatogr. A 1122 (2006) 28–34
2.7. Dynamic binding capacity Analytical chromatography procedures were done using HASepharose acetylated for 36 h. Dynamic binding capacity (DBC) was determined in columns (10 cm × 0.5 cm) packed with 1 mL of matrix. Columns equilibrated with five volumes of Buffer A were loaded with 120 mg of porcine serum (diluted to 8 mg/mL). Flow rates of 0.2, 0.3, 0.5, and 1 mL/min were tested and the total load of protein was 120 mg. DBC was estimated at 2.5% of breakthrough. 2.8. Chromatographic scale-up By using 46 mL bed (20 cm × 2.5 cm) of HA-Sepharose, the immunoglobulin isolation was scaled up to an injection of 480 or 640 mg of porcine serum protein. The same mobile phases and elution conditions were used as indicated in Chromatography conditions. The solutions (five-bed volumes of each) were passed through the column at 1 mL/min. Regeneration of the bed was carried out as previously described. This process was repeated 10 times to evaluate gel stability. 2.9. Polyacrylamide gel electrophoresis Protein fractions were analyzed by 10% polyacrylamide SDS-PAGE according to Laemmli [18] using a vertical electrophoresis unit Mighty Small 250 (Hoeffer, San Francisco, CA, USA). Gels were stained with 1% Coomassie blue in 30% methanol and 10% acetic acid. 2.10. Immunoblotting assay Immunoblotting assay was performed as described by Towbin et al. [19]. Unadsorbed and elution protein fractions (40 g) were separated on 10% SDS-PAGE polyacrylamide gels under reducing conditions [18] and transferred to nitrocellulose membranes at 2.5 mA/cm2 for 40 min (Semi-dry blotter, Bluchler, Labconco). Serum albumin and the heavy chain of each immunoglobulin type were detected with either anti-pig albumin, anti-pig IgA (␣ chain specific), anti-pig IgG (Fc region specific) or anti-pig IgM ( chain specific) raised in goats, and followed by incubation with anti-goat IgG peroxidase as a secondary antibody. A color reaction was developed by the addition of peroxidase substrate 3,3 -diaminobenzidine (Sigma–Aldrich). 2.11. Comparison between HA-Sepharose and T-gel HA-Sepharose was compared with T-gel, a commercially available matrix for immunoglobulin separation where protein adsorption is also promoted by salt [8,9] was assayed under the same chromatographic conditions used for HASepharose. T-gel DBC was estimated, as described before for HA-Sepharose. The relative amount of porcine immunoglobulins M (IgA, IgG, and IgM) in HA-Sepharose and T-gel unadsorbed and adsorbed fractions, as well as in serum samples, were determined by radial immunodiffusion (RID) according
to Fahey and McKelvey [20]. Quantitative RID kits containing goat anti-pig IgA, goat anti-pig IgG or goat anti-pig IgM (Bethyl Labs) were used according to vendor. Control plate wells were loaded with reference standards of appropriate immunoglobulin or with 160 g of protein from each chromatographic fraction. Immunoglobulin concentration of unknown samples was determined by locating their precipitation-ring diameters on a semi-log plot. Pig IgA (60–480 mg/dL), IgG (625–5000 mg/dL) or IgM (75–600 mg/dL) were used as standard solutions. Ring precipitation diameters were measured with a VET-RID reader and an ocular device Finescale (Horscale, Labconco, USA). Duplicate samples were each analyzed for immunoglobulin content at three independent times. 3. Results and discussion Proteins adsorb to a variety of solid phases. Major interactions involved in adsorption can be classified as hydrophobic or electrostatic interactions [21]. Frequently chemical (ionic strength, pH) or physical changes (temperature) lead to desorption and many times to separation and purification of proteins [22,23]. At low salt concentrations, agarose is typically non-adsorbent and inert to react with proteins; this is one of the main reasons it is widely used as a support chromatographic medium. In this work, it is shown that a modest modification of agarose (acetylation) and 0.5 M Na2 SO4 stimulates the interaction of immunoglobulins. 3.1. Elementary analysis The elementary analysis of unmodified and modified Sepharose is given in Table 1. As expected the carbon content increased with modification from Sepharose to HA-Sepharose. Acetylation raised the carbon content from 21.95 to 50.15% in HA-Sepharose acetylated for 24 h and to 50.47% in HASepharose acetylated for 36 h. These values indicate that at 24 h of reaction with acetic anhydride the acetylation probably has reached saturation. During column packing, 1 g of Sepharose hydrated to 1.35 mL, while HA-Sepharose acetylated for 36 h only to 1.11 mL. Composition and hydratation behavior indicate a difference in hydrophobicity on both gels. This is in agreement with Garcia et al. [24] where agarose acetylation caused a reduction in the solubility of agarose in boiling water and an increase in the matrix hydrophobic character.
Table 1 Elementary analysis results on unmodified and modified Sepharose Gel
Carbon (% weight)
Hydrogen (% weight)
Oxygen (% weight)
Sepharose CL-6B HA Sepharose 24 ha HA Sepharose 36 ha
21.95 50.16 50.47
7.34 6.12 5.95
57.00 45.49 44.63
a
24 and 36 h correspond to acetylation time (in hours).
G. Ramos-Clamont et al. / J. Chromatogr. A 1122 (2006) 28–34
31
Fig. 1. Elution profile of proteins from porcine serum by hydrophobic interaction chromatography. (A) HA-Sepharose, (B) unmodified Sepharose (5 mL bed volume, 10 cm × 1 cm). Gels were equilibrated with 0.5 M Na2SO4, 10 mM MOPS, pH 7.6 (buffer A), and porcine serum (8 mL) applied. Non-adsorbed proteins remained in the flow through (a). Adsorbed proteins eluted with 10 mM MOPS, pH 7.6 (b). Columns (10 cm × 1 cm) were washed with buffer A until the absorbance was
