Abstract: Effects of PEG molecular weight, system polymer concentrations (tie-line lengths), pH and sodium chloride concentration, on the partition coefficient of trypsin in PEG-purified Anacardium occidentale L. (cashew tree) exudate gum polysaccharide aqueous two-phase systems have been investigated. Changes in PEG molecular weight, tie-line length and pH, had relatively little effect on trypsin partitioning, with partition coefficients (Κ) < 0.3, i.e., trypsin partitioned preferentially into the A. occidentale L. gum (lower) phase. However, addition of sodium chloride (0.1 M) resulted in dramatic increases in Κ values with increasing pH using the PEG 4000 (9% w/w)-purified A. occidentale L. gum (18% w/w) system, increasing to 3.70 at pH 7.0 and 9.77 at pH 8.0, i.e., trypsin partitioned preferentially into the PEG 4000 (upper) phase. Relative trypsin activities in the phases were investigated for these systems (with different sodium chloride levels) and high activities were obtained at pH 7.0 (70.9%) and pH 8.0 (90.9%) in 0.1 M sodium chloride, which in conjunction with their Κ values, clearly demonstrates their suitability as lower cost enzyme purification procedures (compared with PEG-dextran).
INTRODUCTION
Aqueous liquid-liquid two-phase extraction has been utilised for many years as a separation/ purification methodology with wide application in areas of biochemistry, cell biology and biotechnology (Walter et al., 1985; Albertsson, 1986; Walter and Johansson, 1986; Zaslavsky, 1995). Compared with other commonly used techniques, partitioning in aqueous two-phase systems has several advantages, such as easy scale-up and the ability to handle particulate materials and process streams continuously (Carlson, 1988).
Aqueous two-phase systems are composed of two substances dissolved in water above a certain critical phase separating concentration and the first kind of such systems, used for protein partitioning, were composed of water, a water-soluble organic solvent and a salt. The second kind of aqueous two-phase systems consist of a polymeric substance and a salt in water. The most commonly used systems are based on poly (ethylene glycol) (PEG) and potassium phosphate or ammonium sulphate and for economical reasons these systems have been used for large-scale processes, particularly enzyme purification (Kula et al., 1989; Veide et al., 1983; Kroner et al., 1982, 1984; Tjerneld et al., 1987a, b; Tjerneld and Johansson, 1990; Cascone et al., 1991; Schmidt et al., 1994; Queiroz et al., 1995; Sinha et al., 1996). However, their disadvantages are that they can damage fragile proteins and also that they result in waste disposal problems (Vernau and Kula, 1990). The third kind of aqueous two-phase systems are composed of two polymers in water. The most commonly used are PEG-dextran systems, which offer good flexibility in terms of water content and salt concentration and the localisation of the respective polymers in different phases offers the possibility to restrict chemical groups on the partitioned substrate to one phase by direct interaction with the corresponding polymer (Tjerneld and Johansson, 1990).
Partitioning of bioactive materials in aqueous two-phase systems is affected by the surface properties of bioactive materials such as surface net charge, molecular weight, shape, surface hydrophobicity and the existence of specific binding sites and the partition coefficient is affected by the concentration and phase-forming polymer properties, incorporated ions, pH and the affinity of the macromolecule for the phase-forming polymer (Albertsson, 1986). The electrical potential between the upper and the lower phases of aqueous two-phase systems are created by the non-homogeneous distribution of ions in the upper and lower phases (Han and Lee, 1997).
By using polymer-polymer aqueous two-phase systems it is possible to have phases with high water contents (80-90% w/v) and low interfacial tension, with good solubility for proteins and other biomacromolecules. PEG-dextran systems have been utilised for the separation of many target substances on a laboratory scale (Albertsson et al., 1990; Park and Wang, 1991; Schmidt et al., 1994; Han and Lee, 1997; Gündüz and Korkmaz, 2000; Lin et al., 2003). However, larger scale biotechnological applications have been slow to develop due to the inherently high costs of fractionated dextrans of high purity and well defined molecular weight profiles (with relatively low degrees of polydispersity), which are necessary for good reproducibility (Tjerneld and Johansson, 1990). Cheaper crude dextrans have been investigated, however they contain high molecular weight fractions that result in high viscosity and therefore associated handling difficulties (Kroner et al., 1982; Tjerneld et al., 1985). Other sources of purified dextran (Ghosh et al., 2004) and dextran derivatives (including hydroxypropyl-, carboxymethyl-, diethylaminoethyl- and sulphated dextran) have also been used for two-phase systems (Albertsson, 1986).
Investigations have been performed into the formation and application of aqueous two-phase systems using less expensive carbohydrate polymers, such as pullulan (Nguyen et al., 1988), cellulose derivatives (Albertsson, 1986; Tjerneld, 1989; Skuse et al., 1992), starch derivatives (Tjerneld et al., 1986, 1987b; Sturesson et al., 1990; Venâncio et al., 1993; Almeida et al., 1998; Wu et al., 2001), maltodextrin (starch hydrolysate) (Szlag and Guiliano, 1988; Atkinson and Johns, 1994; da Silva and Meirelles, 2000a, b), agarose (Medin and Jansson, 1993), guar gum (Venâncio et al., 1995), locust bean gum (Venâncio et al., 1996) and larch arabinogalactan (Christian et al., 1998). Information on the structural characteristics of these different carbohydrate-based polymeric biomaterials and their derivatives can be obtained from many sources (Kennedy, 1974; Kennedy and White, 1983; Kennedy, 1988; de Belder, 1993; Kennedy et al., 1995; Carioca et al., 1996; Heinze and Glasser, 1998; Barsby et al., 2001; Yuryev et al., 2002; Gotlieb and Capelle, 2005; Knill and Kennedy, 2005; Collins, 2006).
The potential utilisation of the exudate gum from Anacardium occidentale L. (cashew) as an aqueous-phase-forming polymer and its capacity for protein (bovine serum albumin) separation/partitioning in aqueous two-phase systems with PEG (effect of PEG molecular weight, pH, temperature and relative polymer concentrations (tie-line length)), has been assessed previously (Sarubbo et al., 2000; Oliveira et al., 2001). A. occidentale L. is found in many tropical and subtropical countries, especially Brazil, Venezuela, India, Malaysia and Papua New Guinea. Its exudate gum, produced in epithelial cells which border the gum ducts (Nair et al., 1983), is part of the plants biochemical defences (Marques and Xavier-Filho, 1991) and has found local use as a substitute for gum Arabic (de Paula and Rodrigues, 1995), since it has some similar physicochemical characteristics (Mothé and Rao, 1999, 2000).
| Fig. 1: | Structural representation of a fragment of purified A. occidentale L. exudate gum polysaccharide (Anderson and Bell, 1975; Sarubbo et al., 2000) {A = primary backbone; B and C = side-chains; R = D-Gal, D-Glc, L-Ara, L-Rha, D-Man, D-GlcA or 4-O-Me-D-GlcA non-reducing end groups} |
The primary polysaccharide component of A. occidentale L. exudate gum is a highly branched acidic hetero-galactan (Fig. 1) (a complex mixture of Ca, Mg, K and Na salts, MW ~ 1.0-1.5 x 104), which has a (1→3)-linked β-D-Galp backbone (A chain in Fig. 1), with (1→6)-linked …→3)-β-D-Galp-(1→… galactan side chains (B and C chains in Fig. 1) and D-Gal, D-Glc, L-Ara, L-Rha, D-Man, D-GlcA and 4-O-Me-D-GlcA non-reducing end groups and D-Xyl residues have also been detected in some samples (Bose and Biswas, 1970; Anderson et al., 1974; Anderson and Bell, 1975; de Paula and Rodrigues, 1995; de Pinto et al., 1995; Zakaria and Rahman, 1996; de Paula et al., 1998; Menestrina et al., 1998; Mothé and Rao, 1999; Silva et al., 2004). Brazilian A. occidentale L. exudate gum (utilised in this study) is composed (w/w) of Gal (~ 72-73%), Glc (~ 11-14%), Ara (~ 4-5%), Rha (~3-4 %), Man (~ 0-1%), GlcA/4-O-Me-GlcA (~ 5-6%) (Zakaria and Rahman, 1996; de Paula et al., 1998; Menestrina et al., 1998; Mothé and Rao, 1999; Silva et al., 2006).
The effects of PEG molecular weight, relative polymer concentrations (tie-line length), pH and sodium chloride concentration, on the partition coefficient and enzyme activity of trypsin in PEG-Anacardium occidentale L. exudate gum polysaccharide aqueous two-phase systems are presented in this research.
MATERIALS AND METHODS
A. occidentale L. exudate gum polysaccharide isolation: Crude natural exudate gum (~ 1 kg) was collected from cultivated A. occidentale L. trees (common-type yellow cashew producers, ~ 20 years old) in various locations within the sate of Pernambuco (PE), Brazil. Clear exudate gum nodules (~ 1-10 g), free of bark, were selected for purification as Na salts using a method based on those of Villain-Simonnet et al. (1999 and 2000), as detailed below. Crude gum was dissolved in sodium chloride solution (NaCl, 0.5 M, 25±1→C) to give a gum concentration of ~5 g L-1. The resultant crude gum solution was centrifuged (25 min, 30000 G, 25±1→C) and the supernatant filtered (0.45 μm). The polysaccharide was precipitated from the filtered supernatant using ethanol (giving a final ethanol concentration of 55% v/v), isolated by suction filtration, washed with ethanol: sodium chloride solution of increasing ethanol concentration (75-100% v/v ethanol: 25-0% v/v 0.5 M NaCl) and dried under vacuum for 48 h at 30→C. Purified A. occidentale L. exudate gum polysaccharide recovery was ~ 90% w/w (which is in close agreement with Anderson et al. (1974) and Marques and Xavier-Filho (1991).
Determination of phase diagrams: Phase diagrams were determined according to the method of Albertsson (1986), as detailed in previous investigations (Sarubbo et al., 2000; Oliveira et al., 2001) (Fig. 2). The binodial line, the demarcation between PEG (4000 and 8000)-purified A. occidentale L. exudate gum polysaccharide compositions showing monophasic and biphasic behaviour, was obtained by direct observation of two-phase formation for triplicate solutions of varying polymer concentrations (Table 1) in phosphate buffer (15 mM, pH 6.0), which were mechanically stirred for several minutes to ensure equilibrium conditions and the phases generated were allowed to separate over a 24 hour period in beakers placed in a constant temperature bath (27±2→C). PEG (4000 and 8000)-purified A. occidentale L. exudate gum polysaccharide two-phase systems (Table 1) were prepared from stock solutions of PEG 4000 and 8000 (50% w/w) and purified A. occidentale L. exudate gum polysaccharide (30% w/w) in 15 mM phosphate buffer (pH 6.0). Systems that displayed a distinct phase-phase interface were considered biphasic.
| Fig. 2: | Phase diagrams for PEG-purified A. occidentale L. exudate gum polysaccharide systems (27±2→C, pH 6.0): (a) PEG 4000; (b) PEG 8000 (Sarubbo et al., 2000; Oliveira et al., 2001) |
| Table 1: | Initial system and final phase compositions (% w/w) of PEG-purified A. occidentale L. exudate gum polysaccharide systems (27±2→C, pH 6.0): (a) PEG 4000; (b) PEG 8000; (Sarubbo et al., 2000; Oliveira et al., 2001) |
| * All quoted values (%w/w) are derived from the mean of triplicate analyses (all with%variation values<5%) (% variation = 100 x (standard deviation/mean), # Vr = system volume ratio = volume of upper phase/volume of lower phase | |
Polymer concentrations in the upper and lower phases were determined (Table 1): PEG concentration using the method of Skoog (1979) and A. occidentale L. exudate gum polysaccharide concentration using the 3,5-dinitrosalicylic acid assay for reducing sugars (Miller, 1959; White and Kennedy, 1981; Chaplin, 1994), using purified A. occidentale L. exudate gum polysaccharide and D-galactose as standards. PEG 4000 and PEG 8000 were obtained from Sigma-Aldrich Company Ltd., Poole, Dorset, UK.
Trypsin partitioning: PEG (4000 and 8000)-purified A. occidentale L. exudate gum polysaccharide two-phase systems (Table 1) were prepared from stock solutions of PEG 4000 and 8000 (50% w/w) and purified A. occidentale L. exudate gum polysaccharide (30% w/w) in 15 mM phosphate buffer (pH 6.0, 7.0 and 8.0). Sodium chloride (NaCl) was directly dissolved into replicate systems to afford salt concentrations of 0.1 and 1.0 M as required. The total weights of the final systems used were 4 g. Enzyme partitioning experiments were performed in triplicate at 27±2→C (constant temperature bath) by mixing the systems with trypsin solution (2 mg mL-1, 400 μL). The systems were vortex mixed (5 min) and centrifuged (5 min, 236 G, 25±1→C) to obtain two clear phases. Thus, triplicate system variables included 3 different tie-line lengths (relative polymer concentrations/compositions), 2 PEG molecular weights, 3 pH’s and 3 salt concentrations. Trypsin (from bovine pancreas, EC 3.4.21.4) was obtained from Sigma-Aldrich Company Ltd., Poole, Dorset, UK and had an activity of > 10,000 BAEE U/mg protein. [Unit definition: 1 U corresponds to the amount of enzyme which increases the absorbance at 253 nm (ΔA253) by 0.001 per minute at pH 7.6 and 25→C using N-benzoyl-L-arginine ethyl ester (BAEE) as substrate].
Determination of trypsin partition coefficients (K): Protein (trypsin) concentrations were determined (in triplicate) in the upper phases according to the method of Bradford (1976) and in the lower phases calculated by mass balance, due to high viscosity, according to the methods of Venâncio et al. (1993) and Almeida et al. (1998). Partition coefficient (K) was defined as the ratio between trypsin concentration in the upper (PEG) and lower (A. occidentale) phases. Effects of system composition (PEG molecular weight, tie-line length, pH and sodium chloride) on trypsin partition coefficients are presented Table 2 and data of specific interest presented graphically in Fig. 3.
Determination of trypsin activity: Trypsin activity in the upper and lower phases of selected systems (chosen based upon their partition coefficients, determined using the methodology detailed in the previous section) was measured in triplicate according to the method of Ginther (1979) using azocasein (Sigma-Aldrich Company Ltd., Poole, Dorset, UK, 1.0% w/w in 0.2 M Tris-HCl buffer, pH 7.2, containing 1.0 mM CaCl2). [Unit definition: 1 U corresponds to the amount of enzyme which increases the absorbance at 440 nm (ΔA440) by 1.0 per hour at pH 7.2 and 25→C using azocasein as substrate]. Determined trypsin activities for the analysed systems are presented in Table 3 (quoted values are relative trypsin activity (%), with 100% activity being that measured for trypsin in 15 mM sodium phosphate buffer (pH 6.0, 7.0 or 8.0), i.e., the appropriate polymer-free control solutions).
RESULTS AND DISCUSSION
Effect of system composition on trypsin partition coefficient (K): Effects of system composition (PEG molecular weight, tie-line length, pH and sodium chloride) on trypsin partition coefficients are presented in Table 2. In the absence of sodium chloride, the investigated systems were not significantly influenced by PEG molecular weight, pH or tie-line length (all determined K-values < 0.3, i.e., trypsin partitioned preferentially into the A. occidentale L. gum (lower) phase). Partition behaviour is more sensitive to changes in polymer molecular weight for proteins > 50 kDa. The relatively low molecular weight of trypsin (25 kDa) may account for this, the same behaviour being observed for cutinase (22 kDa) partitioning in PEG-starch systems (Almeida et al., 1998). Increases in polymer concentrations (tie-line length) either resulted in a reduction in K values (or had relatively little effect). Increasing PEG concentration causes molecular exclusion of trypsin from the upper to the lower phase (Sturesson et al., 1990; Almeida et al., 1998).
Addition of sodium chloride can be utilised to significantly increase protein partition coefficients in aqueous two-phase systems, an effect that has been attributed to possible hydrophobic interactions (Cascone et al., 1991; Schmidt et al., 1994). In the investigated systems, addition of sodium chloride (0.1 M) resulted in dramatic increases in Κ values with increasing pH using the PEG 4000 (9% w/w)-purified A. occidentale L. gum (18% w/w) system, increasing to 3.70 at pH 7.0 and 9.77 at pH 8.0 (Table 2, Fig. 3), i.e., trypsin partitioning preferentially into the PEG 4000 (upper) phase.
| Table 2: | Effect of system composition on trypsin partition coefficients (K) in PEG-purified A. occidentale L. exudate gum polysaccharide aqueous two-phase systems (27±2oC): (a) PEG 4000; (b) PEG 8000 |
| * All quoted K values are derived from the mean of triplicate analyses (all with % variation values < 10%) [% variation = 100 x (standard deviation/mean)] | |
| Table 3: | Effect of sodium chloride on relative trypsin activities (%) for selected PEG-purified A. occidentale L. exudate gum polysaccharide systems (25±2→C) |
| * All quoted values (%) are derived from the mean of triplicate analyses (all with % variation values < 5%) [% variation = 100 x (standard deviation/mean)] | |
Effect of system composition on relative trypsin activity: The upper and lower phases from the first tie-line lengths for each of the PEG 4000 and PEG 8000-purified A. occidentale L. gum polysaccharide systems were selected for determination of relative trypsin activity (Table 3), since aspects of these systems demonstrated the greatest effect on partition coefficients (Table 2, Fig. 3). In general, higher relative trypsin activities were obtained in the pH 7.0 and 8.0 systems and higher molecular weight systems (PEG 8000) resulted in lower relative trypsin activities than their lower molecular weight (PEG 4000) counterparts (Table 3). Similar behaviour was observed for lipolytic activity recovery in PEG-phosphate systems (Queiroz et al., 1995). The maximum relative trypsin activity was obtained with PEG 8000 at pH 7.0 in 1.0 M NaCl (93.1%), in the lower phase. High relative trypsin activities were obtained for the lower phases of the systems with the highest partition coefficients in Table 2 (i.e., the most enzyme material in the upper phases), namely 70.9% at pH 7.0 and 90.9% at pH 8.0 (in 0.1 M sodium chloride), clearly demonstrating the suitability of these systems for trypsin partitioning/ purification.
Certain systems (Table 3) have combined upper and lower phase relative trypsin activities of > 100%, which can be due to a number of distinct factors that affect enzyme activity by altering access to active sites, such as microviscosity (Lindeström-Lang and Schellman, 1959). Increasing solvent viscosity has been shown to decrease enzyme catalytic rate constants for hyaluronate lyase, lactate dehydrogenase and trypsin (Laurent, 1971; Demchenko et al., 1989), however contradictory results, i.e., increases in activity with increasing viscosity, have been observed for selected enzymes, including trypsin, with polyvinylpyrrolidone (Trón et al., 1976; Van der Sanden et al., 1995) and PEG 400 and 6000 (Van der Sanden et al., 1995; Sinha et al., 1996).
| Fig. 3: | Effect of sodium chloride addition on trypsin partition coefficient in selected PEG-purified A. occidentale L. exudate gum polysaccharide systems (27±2→C) |
Charged molecules show a strong viscosity dependence on ionic strength. A. occidentale L. gum polysaccharide is composed of ~ 5-6% w/w GlcA/4-O-Me-GlcA residues (Zakaria and Rahman, 1996; de Paula and Rodrigues, 1995; de Paula et al., 1998; Menestrina et al., 1998; Mothé and Rao, 1999; Silva et al., 2006). Ionisation of the GlcA/4-O-Me-GlcA carboxyl groups above pH 7.0 could result in changes in viscosity, as observed for mucilage gum containing similar uronic acid levels (Trachtenberg and Mayer, 1982; Medina-Torres et al., 2000).
CONCLUSIONS
Manipulation of the PEG-purified A. occidentale L. gum system composition can be utilised for successful trypsin purification/fractionation, since bulk inactive
material has been successfully separated from small quantities of material with high relative trypsin activity. Such systems therefore have great potential as lower cost enzyme purification procedures (compared with PEG-dextran).
ACKNOWLEDGMENTS
The authors would like to express their thanks to FAPESP, CAPES and the British Council, for providing the financial support for these investigations.