SciELO - Scientific Electronic Library Online

vol.17 issue4-7Phenomenological model of the clavulanic acid production process utilizing Streptomyces clavuligerusStudies on the rheology and oxygen mass transfer in the clavulanic acid production by Streptomyces clavuligerus author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Brazilian Journal of Chemical Engineering

Print version ISSN 0104-6632On-line version ISSN 1678-4383

Braz. J. Chem. Eng. vol.17 n.4-7 São Paulo Dec. 2000 



A.L.P.Falco1, L.R.Durrant1, and T.T.Franco2*
1Food Engineering, State University of Campinas (UNICAMP),
2 Chemical Engineering, State University of Campinas (UNICAMP),
P.O. Box 6066, Campinas, 13081-970, Phone: (55) (19) 788-39 66,
Fax: (55) (19) 788-39 65, Brazil,


(Received: January 6, 2000 ; Accepted: April 18, 2000)



Abstract - Alpha-galactosidase taken from a raw extract of Sesbania marginata legume seeds was purified by partitioning in aqueous two-phase systems (ATPS). Initially, galactomannan/dextran 2,000,000 systems were used for the purification, and the partition coefficients of a-galactosidase varied from 1.5 to 4.0. However, mass transport in these systems was poor due to the high viscosity of the employed polymers. Therefore, partitioning in polyethyleneglycol (PEG)/ sodium phosphate systems and the effect of sodium chloride upon the enzyme purification and the yield of a-galactosidase were also investigated. The purification achieved in a single-step was 5.7 with a recovery of 144% of a-galactosidase, possibly due to the removal of materials which inhibited a-galactosidase activity before the purification. The removal of the main protein contaminants and the highest yields were achieved in PEG 4,000/ sodium phosphate + 6% NaCl system at pH 5.0. Further purification by preparative on-exchange chromatography was also developed.
Keywords: alpha-galactosidase; legume seeds; enzyme purification; aqueous two phase systems.




Alpha-galactosidase (EC is widely distributed in the plant kingdom, and it is likely to be related to the depletion of the raffinose family oligosaccharides in leguminous seeds. This enzyme may be employed in the removal of sugars during seed processing and also as a digestive aid when seeds have not been previously treated, as the enzyme is not present in the digestive systems of mammals, included humans. Crystalline a-galactosidase may be used as a reagent to the determination of structures of polysaccharides and glycoproteins (Obara and Hashimoto, 1977). Three different hydrolitic enzymes are known to be involved in the catabolism of galactomannan (Buckeridege and Dietrich, 1996): a-galactosidase, endo-b-mannanase and exo-b-mannanase (or b-mannosidase). Commercially, a-galactosidase would have great potential in the modification of rheological properties of polysaccharides by changing their viscosity and changing their long chains structure. This enzyme from legume seeds has been isolated by Buckeridge and Dietrich (1996) after several chromatographic steps achieving a very low yield. Therefore, the improvement and simplification of the purification procedure of a-galactosidase would be desirable to reduce the cost of enzyme preparation, since extraction, purification and concentration steps are commonly responsible for up to 80% of total production cost (Kula, 1990).

Extraction of biomolecules including enzymes, in aqueous two-phase offers advantages on a large scale, such as the possibility of obtaining a high yield, the possibility of continuous processing and a reduction in operational cost in relation to the costs of conventional processes (Kula, 1990, Silva et al., 2000a). Microbial enzymes as lactate dehydrogenase (Franco et al, 1997), alkaline xylanases from Bacillus pumilus (Duarte et al, 1999), chitosanase from B.cereus (Piza et al, 1999), b-galactosidase from E.coli (Silva et al, 1997) and from yeasts (Silva, et al. 2000a) and vegetal peroxidase from soya beans (Silva et al, 2000b) have recently been successfully extracted and purified in ATPS achieving significatively higher yields than enzymes conventionally extracted.

In this article, polymer-polymer aqueous two-phase systems composed of galactomannan /dextran 2,000,000 and galactomannan/ dextran 500 systems were initially used for the purification of a-galactosidase since the systems and the galactomannan’s molecular masses had been previously characterised by Franco et al., 1996. They were chosen for enzyme partitioning due to the galactose composition, which would potentially act as specific bioligands for a -galactosidase.PEG-phosphate systems and the effect of extra salt in the systems upon the purification of this enzyme were also investigated.




Sesbania marginata Benth seeds were obtained from plants cultivated under natural conditions at the Instituto de Botânica, São Paulo, Brazil. Polyethyleneglycol (PEG) Mr 1,500 and 4,000 were from Synth (Campinas, SP, Brazil). Sodium phosphate and sodium chloride were from Ecibra (Campinas, SP, Brazil). Dextran of Mr 2,000,000 was from Amershan-Pharmacia (Upsalla, Sweden). Galactomannan from Dimorphandra mollis was prepared according to Franco et al., 1996. All other chemical and biochemical reagents used were of analytical grade.

Extraction of a-Galactosidase from Sesbania Marginata

Sesbania marginata seeds were scarified and placed in Petri dishes for five days at 30°C. The endosperm plus seed coat were separated from the embryo after the germination of the seeds. The embryo was discharged and the samples were crushed with 20 mM sodium acetate buffer at pH 5.0 and at 5oC, filtered and centrifuged at 10,000 g for 30 min to obtain the a-galactosidase raw enzymatic extract (Buckeridge and Dietrich, 1996).

Determination of a-Galactosidase Activity

Enzyme activity was determined according to Buckeridge and Dietrich, (1996) by using p-nitrophenyl-a-D-galactopyranoside (PNPG) as a substrate. The colour was measured at 405 nm, and the activity unit was defined as the amount of enzyme that hydrolysed 1nmol of PNPG per min at 30oC.

Partitioning of Total Proteins

The total protein content of the isolated-phase samples was determined using the UV spectrophotometric assay (Scopes, 1986). Fifty to 100ml of the top phase were transferred from each system to a cuvette containing 2.4ml of water, mixed well and the absorbance at 280 nm was read against a blank which had 100 ml of a top phase of a system which had been equally prepared without any sample. The procedure was repeated for the bottom phase of each system, and a BSA standard curve was used to calculate protein concentration. The partition coefficient (K) was calculated as the ratio of protein concentration in the top phase to that in the bottom phase at room temperature.

Determination of a-Galactosidase Partition Coefficient

A known volume of each phase of the ATPS was transferred to a tube for the measurement of the enzyme activity. The ratio of the activity in the top phase to the enzyme activity in the bottom phase was calculated.

Aqueous Two-Phase Systems Preparation

Initially aqueous two-phase systems were prepared with 3.0% galactomannan and 20% dextran 2,000,000 (w/w) stock solutions in water containing 0.02% sodium azide. The four chosen compositions of the galactomannan-dextran systems are shown in Figure 1. Stock solutions of 40% (w/w) PEG 1,500 and PEG 4,000 and of 40% (w/w) sodium hydrogen phosphate were weighed in flasks. Solid NaCl was added when necessary, and the remainder was the raw enzymatic extract. The aqueous systems were mixed with a Vortex for 1 min, and the phase separation was achieved by centrifugation for 3min at 3,000g. The compositions of the investigated systems were 16% Peg 1,500 + 16% phosphate and 13% Peg 4,000 + 13% phosphate. The effect of sodium chloride on the partition coefficient was investigated with 3, 6, 9 and 12% NaCl in the aqueous two-phase systems. The volume ratio of the phases was determined in graduated tubes.



Ion Exchange Chromatography

Material recovered from the bottom-phases of ATPS was further purified by ion exchange chromatography (IEC) using a FPLC system with a UV detector and a fraction collector. IEC was developed in an HR5/5 column packed with Q-sepharose FF equilibrated with a 50 mM monobasic sodium phosphate buffer, at pH 6.0. A gradient of 50 mM monobasic sodium phosphate at pH 6.0 (solv. A) and 50 mM monobasic sodium phosphate buffer at pH 6.0 + 1M NaCl (solv.B) were used for elution. The equipment, columns and chromatographic media were from Amershan- Pharmacia.


SDS-PAGE electrophoresis was carried out in a 12% homogeneous gel [5] in the Mini-Protean II system (BioRad). The staining was carried out in Coomassie Brilliant Blue R-250. The standard markers were from Amershan-Pharmacia.



Partitioning of a-galactosidase was initially developed in 5.0 g systems of galactomannan /dextran 2,000,000. These systems were easily mixed and handled at this size, but when they were scaled up to 50.0 g, a larger variation on the reproducibility of the data was observed. The viscosity of the systems increased with galactomannan concentration, possibly due to the interaction of both polymers. However, in the smaller-scale systems, the partition coefficient of a-galactosidase remained from 1.0 to 1.5 in the systems with concentrations of 4.0 and 5.0% dextran, respectively. The partition coefficient increased to 3.5 in a 1% dextran + 1.6% galactomannan system, and conversely, fell bellow 1 in a system with 10% dextran, indicating that the higher the dextran concentration, the lower the partition coefficient of the enzyme (Table 1). An effect of the increase of galactomannan concentration was also observed in the systems upon the K of a-galactosidase.



When polymer-polymer systems were replaced by PEG-phosphate, the reproducibility of the partition coefficients of the enzyme and of the total proteins were improved in larger-scale experiments (50.0 g). The partition coefficient of a-galactosidase was 0.03 in 16% PEG 1,500 - 16 % phosphate and in 13% PEG 4,000 - 13 % phosphate (both at a pH 5.0) in the absence of any extra NaCl. Conversely, the partition coefficients of the other contaminant proteins (total proteins) were 4.4 and 0.7, and the purification factors were 6.3 and 2.6, respectively. These results indicate that PEG 1,500 - phosphate is an useful system in the separation of a-galactosidase from its natural contaminants, although the recovery of the enzyme was only 89% in the bottom phosphate-rich phase. The effect of sodium chloride concentration upon the purification of a-galactosidase from the contaminants was only significant in PEG 4,000-phosphate systems. Table 2 shows that the purification factor (PF) increased from 2.6 to 5.4 in PEG 4,000 - phosphate systems with increasing NaCl concentrations up to 6% and then decreased to 4.8 with NaCl concentrations of 12% (Figure 2). However, amongst all the PEG-salt systems, the highest purification factor was achieved in a PEG 1,500 - phosphate system without any NaCl (Table 2). Conversely, the highest recovery of the enzyme was obtained in PEG 4,000-phosphate systems + 6% NaCl rather than other NaCl concentrations, possibly due to the removal of materials which inhibited a-galactosidase activity before the purification. It was possible to recover more than 100% of the enzyme in a single-step partitioning and it was verified that a further partitioning in a fresh ATPS did not improve the purification factor. The improvement in the separation of the target protein and the main contaminants in PEG/salt systems is usually attributed to the differences of surface hydrophobicity. Franco et al (1996b,c) proposed a resolution factor based on hydrophobicity, R, to measure the systems separation capacity for proteins of increasing hydrophobicity. It was observed that the R value of the PEG1500/phosphate system increased 2.5 fold with the addition of 6% NaCl but that this value raised only 1.3 fold for the PEG4000/phospahte system with the same addition of NaCl. An extensive work showed that the effect of NaCl on the improvement of the resolution of ATPS toward R was clearly related to hydrophobicity: The increase in hydrophobic differences between the phases, the increase in hydrophobicity of the PEG-rich phase promoted by the NaCl and the increase in the strength of hydrophobic interactions between the proteins and the PEG molecules were found to probably be the main reason for the improvement in R.



Diluted material isolated from the bottom phase was further purified in an anion exchanger media (Q-Sepharose) with a gradient of NaCl at pH 6.0.

Figures 3 and 4 show that an almost pure a-galactosidase was eluted mainly at 1.0M NaCl. Table 3 reports that the purification factor and yield which were achieved after partitioning in a 13% PEG 4,000 + 13% phosphate + 6% NaCl system were 5.4 and 144%, and after IEC the purification was 12.9 fold. The enzyme yield was 20%. A successful purification procedure for a-galactosidase was established by using just two-steps: partitioning in PEG4.000/phosphate + 6%NaCl system and followed by IEC. Alpha-galactosidase from soya beans was also extracted and purified by partitioning in a PEG/phosphate system (data not shown) achieving a purification factor of 86 fold in a single-step procedure with 42% yield.




Dextran-galactomannan systems were able to purify a-galactosidase from its main protein contaminants, but as they became extremely viscous when scaled-up to a 50g system, phase separation and data reproducibility were difficult.

PEG/salt systems were used to replace polymer-polymer systems for the extraction and purification of a-galactosidase. The highest purification achieved was 6.3 in a 16% PEG 1,500 + 16% phosphate (pH 5.0), but the recovery of the target enzyme in the bottom phosphate-rich phase was lower than in PEG 4,000 + phosphate systems. The highest recovery of a-galactosidase (144%) was obtained in a 13% PEG4,000 + 13% phosphate + 6% NaCl, also at a pH 5.0. Here a-galactosidase was mostly collected in the bottom phase, and the main protein contaminants were partitioned mostly in the top phase, and the purification achieved was 5.4 in a single-step partitioning. A further step for the purification of the isolated material from the bottom phase of the above system by anion-exchange chromatography was developed, leading to a nearly pure enzyme. The results show that a reduced number of unity operations was easily achieved by introducing a single-step partitioning procedure in a 13% PEG 4,000 + 13% phosphate + 6% NaCl for the removal of the main contaminants before performing ion-exchange chromatography in a profitable purification procedure.



The financial assistance received from Capes (in the form of a scholarship for A.L.P.), from FAPESP and from CNPq (Brazil) for support of the research project are gratefully acknowledged.



Blakesley,R.W. and Boezi,J.A. Anal. Biochem. 82: 580-582 (1977).        [ Links ]

Buckeridge,M.S. and Dietrich, S.M.C. Plant Sci. 117:33-43 (1996).        [ Links ]

Duarte,M.C, Portugal,E.P,Ponezi,A., Franco,T.T. Alkalophilic xylanase production. Bioresource Technology 68: 49-53, 1999.        [ Links ]

Franco, T.T.; Rodrigues, N.R.; Serra, G.E.; Panegassi, V.R. and Buckeridge, M.S. J. Chromatogr. B 680: 255-261 (1996a).        [ Links ]

Franco, T.T.; Andrews, A.T. and Asenjo, J.A. Use of chemically modified proteins to study the effect of a single protein characteristics in aqueous two-phase systems. Conservative modifications. Biotech. Bioeng.,vol.49: 290-299, 1996b.        [ Links ]

Franco, T.T.; Andrews, A.T. and Asenjo, J.A. Use of chemically modified proteins to study the effect of a single protein characteristics in aqueous two-phase systems. Effect of surface hydrophobicity. Biotech. Bioeng., vol.49:300-308, 1996c.        [ Links ]

Franco, TT;Galaev,I, Kaul,R; Mattiasson,B. Purification of lactate dehidrogenase by aqueous two phase systems using a thermosensitive polymer. Biotech. Letters. 4(11) 1997.        [ Links ]

Goldfarb, AR; Saidel,J. and Mosovich,E. J. Biol. Chem.193: 397-404 (1951).        [ Links ]

Kula, M-R. Trends and future of aqueous two-phase extraction. Bioseparation, vol.1,pp. 181-189(1990).         [ Links ]

Laemmli. V.K. Nature 227: 680-685 (1970).        [ Links ]

Obara,K. and Hashimoto,S. Sugar Technol. Rev. 3: 209-258 (1977).        [ Links ]

Piza, F. A , Siloto, A, Carvalho,CV; Franco,TT Production, characterization, purification of chitosanase from Bacillus cereus. Braz.J.Chem.Eng.16(2): 185-192, 1999        [ Links ]

Silva, M.E, Pellogia, C., Piza, F.A.T. e Franco, T.T. Purification of three different microbian b-galactosidases by partitioning in aqueous two-phase systems. Ciência e Tecnologia de Alimentos, 17(3):219-223, 1997        [ Links ]

Silva, M.E. and Franco, T.T. Purification of microbial b-galactosidase from Kluyveromyces fragilis by bioaffinity partitioning. J.Microbiology, press        [ Links ]

Silva, M.E. & Franco, T.T. Purification of soybean peroxidase (Glycine max) by metal affinity partitioning in aqueous two-phase systems. Journal of Chromatography B, 743: 287-294, 2000.        [ Links ]



*To whom correspondence should be addressed

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License