You are on page 1of 7

Separation and Purification Technology 118 (2013) 888–894

Contents lists available at ScienceDirect

Separation and Purification Technology


journal homepage: www.elsevier.com/locate/seppur

Aqueous two-phase poly(ethylene glycol)–sodium polyacrylate system


for amyloglucosidase purification: Equilibrium diagrams and
partitioning studies
Lizzy Ayra Pereira Alcântara a, Kelany Santiago do Nascimento b, Cecília Alves Mourão a,
Valéria Paula Rodrigues Minim a, Luis Antonio Minim a,⇑
a
Department of Food Technology, Federal University of Viçosa, 36570-000 Viçosa, MG, Brazil
b
Biochemistry and Molecular Biology Department, Federal University of Ceará, 60.455-970 Fortaleza, CE, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: Extraction and purification of amyloglucosidase (AMG) using aqueous two phase systems (ATPSs) com-
Received 11 January 2013 posed by polyethylene glycol (PEG, average molar mass of 4000 g/mol) and sodium polyacrylate (NaPA,
Received in revised form 8 August 2013 average molar mass of 15,000 g/mol) has been studied. Phase equilibrium diagrams were obtained at pH
Accepted 26 August 2013
6.0 and 9.0, which were used for partitioning studies. The selectivity (S) and partition coefficient (Ke) of
Available online 4 September 2013
AMG were evaluated as a function of pH, temperature and concentration of PEG and NaPA, and it was
found that these parameters were affected by pH and NaPA concentrations. A face-centered design with
Keywords:
the response surface methodology was applied for selectivity optimization. The highest value found for S
Aqueous two phase systems
Equilibrium diagram
was 1.6 at pH of 6.5, temperature of 25 °C, PEG concentration of 12.5% (w/w) and NaPA concentration of
Amyloglucosidase 11.4% (w/w). The corresponding Ke was 2.0. The purification factor (PF) and the recovery yield (Y) were
Partitioning determined as 1.9% and 86%, respectively.
Optimization Ó 2013 Elsevier B.V. All rights reserved.

1. Introduction and peptides [18,19], lectins [20] and among many other bio-
molecules, due to their advantages over traditional methods.
Amyloglucosidase (AMG, 1,4-a-D-glucan glucohydrolase, E.C. The high water content in each phase of the system ensures
3.2.1.3) is an exo-enzyme that hydrolyzes starches, removing suc- the high stability of biomolecules throughout the extraction pro-
cessive glucose units from the non reducing ends of the starch, cess, which can result in higher yields.
amylose, amylopectin, and amylodextrin chains [1,2]. AMG has It is well-known that the ATPS can be formed by two incompat-
several applications in the food, pharmacy, brewery, and bread ible phases containing two hydrophilic polymers such as dextran
making industries. It is widely used in the production of fructose (DEX) and polyethylene glycol (PEG) or a hydrophilic polymer in
syrup, baking, brewing and distilling [3–5]. Recently it has been the presence of an electrolyte salt at high concentration. ATPSs
incorporated in commercial preparations for simultaneous sac- containing polymer and salt have received more attention than
charification and fermentation of amylacious biomass for produc- polymer–polymer systems because of their low cost, low viscosity,
tion of bioethanol [6]. greater density difference between phases and low interfacial ten-
Various methods have been used to purify AMG including sion [21]. It has been reported that a system constituted of poly-
colloidal gas aphrons [4], ion exchange chromatography [5], ethylene glycol (PEG) and sodium polyacrylate (NaPA) forms two
affinity chromatography [7,8], affinity precipitation [9,10] and phase systems [22–24] and few studies have been conducted for
bioaffinity extraction [11]. Extraction by aqueous two-phase sys- protein partitioning [23–25]. This ATPS separates components rel-
tems (ATPSs) has demonstrated to be an important method for atively fast and has a relatively low viscosity in contrast to other
separating and purifying mixtures of biomolecules [12–14]. traditional two-phase systems [26]. The two polymers PEG and
Applications of aqueous two-phase systems have been success- poly(acrylic acid) are harmless, relatively inexpensive and easily
fully used and reported in literature for the separation and puri- handled [24]. PEG is an uncharged polymer and NaPA is a strongly
fication of several biological active substances such as human negatively charged polymer in pH above 7.0, thus in order for these
antibodies [15], whey proteins [16], amylases [17], amino acids to separate in the useful polymer concentration range of 1–10 wt%,
a minimum salt concentration is needed [25]. Since NaPA has a
⇑ Corresponding author. Tel.: +55 31 3899 1617; fax: +55 31 3899 2208. strong negative charge, it has a strong electrostatic interaction
E-mail address: lminim@ufv.br (L.A. Minim). with charged biomolecules.

1383-5866/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.seppur.2013.08.039
L.A. Pereira Alcântara et al. / Separation and Purification Technology 118 (2013) 888–894 889

A few studies have been found on the partition coefficient of arated and characterized in terms of PEG, NaPA and water
amyloglucosidase in ATPS system, including those with the sys- concentrations.
tems PEG 4000/Dextran 500 [27], PEG (4000, 6000 and 10,000)/
Na2SO4 and KH2PO4[28]. To our knowledge there is no study re- 2.2.1. Components quantification
lated to the partitioning of AMG in systems composed of PEG The concentration of PEG in both phases was determined by li-
and NaPA. Therefore, in this paper partition studies of AMG were quid–liquid extraction using chloroform [22]. From each phase
conducted in these systems. Initially, phase diagrams were deter- (top or bottom), 2 g samples were added to 15 mL tubes together
mined for the system composed of PEG 4000 and NaPA 15,000 at with sodium hydroxide to increase the pH to 11.0. This pH ensures
pH 6.0 and 9.0. The selectivity (S) and partition coefficient (Ke) that NaPA is completely dissociated in the phase. To each tube
were determined at different pH, temperature, and PEG and NaPA 3.0 mL of chloroform was added. All components were thoroughly
concentrations. A face-centered design coupled with the response mixed in a vortex mixer and centrifuged at 4000g for 5 min. The
surface methodology was used for the optimization of S and Ke. bottom phase containing PEG and chloroform was removed and
Purification factor and yield were determined at the optimal stored in tubes previously dried and weighed. This extraction pro-
conditions. cedure was repeated twice. Tubes with the chloroform phase were
then placed in an oven at 105 °C overnight to evaporate the chlo-
roform and the PEG concentration was determined by gravimetry.
2. Materials and methods The standard deviation of PEG concentration measurements was
±0.19 % (w/w). NaPA concentration was determined by refractive
2.1. Chemicals index measurements of the phases at 298.15 K, using a refractom-
eter (Analytic Jena AG Abbe refractometer 09-2001, Germany) with
AMG from Aspergillus niger (molar mass 89 kD, isoeletric point a precision of ±0.0001. The refractive index of the phase samples
3.5) was generously provided by Prozyn Co. (SP, Brazil). Bovine depends on the total concentration of components present, and is
serum albumin (BSA), D+-glucose monohydrate, soluble starch thus considered an additive property. Standard curves for the
and sodium polyacrylate (average molar mass 15,000 g/mol, 35% refraction index of pure NaPA and PEG were prepared in water in
w/w water solution) were purchased from Sigma–Aldrich Chem. the range of 0.5–10.0 % (w/w). The standard deviation for measure-
Co. (St. Louis, USA). Polyethylene glycol (average molar mass ments of NaPA concentration was ±0.03 % (w/w). Water concentra-
4000 g/mol) was purchased from Fluka Chem. Co. (NY, USA). tion was determined by freeze-drying (Christ mod. alpha2-4,
Sodium chloride and sodium hydroxide (analytical grade) were Germany) at 20 °C for 24 h. The standard deviation for measure-
purchased from Vetec (RJ, Brazil). Hydrochloric acid and acetic acid ments of water was ±0.34 % (w/w). All analytical measurements
(analytical grade) were purchased from ISOFAR (RJ, Brazil). were performed in triplicate.
Ultra pure water (Milli-Q Gradient, Millipore, USA) containing
200 mmol/L of NaCl was used throughout the experiments. Stock 2.3. Partition experiments
solutions of PEG (50% w/w) and NaPA (30% w/w) were prepared
prior to performing the experiments. Both solutions contained The systems (10 mL) were prepared in 15 mL centrifuge tubes
200 mmol/L NaCl. The pH of the solutions was adjusted to 6.0 or by adding defined amounts of water, PEG and NaPA stock solu-
9.0 by adding HCl or NaOH and monitoring with a pH meter (HI tions. These tubes were stirred using a vortex mixer and left in
221, HANNA instruments, Brasil). An AMG stock solution was also an incubator (Tecnal TE-184, Brazil) at a desired temperature for
previously prepared in an acetate buffer (0.1 mg/mL). 12 h for reaching equilibrium. This was followed by centrifugation
at 1400g for 10 min to ensure complete phase separation and the
phases were immediately collected using syringes. Partition exper-
2.2. Aqueous two-phase diagrams iments were performed in 15 mL centrifuge tubes, using 2 mL of
the top and bottom pre-equilibrated phases (the phase ratio was
Determination of the binodal curves was carried out using the fixed to avoid effects on the partition coefficient due to possible
turbidimetric titration method [29]. The PEG stock solution changes in the extensive properties of the phases). Volumes of
(1.0 g) was added to a glass tube and titrated with small aliquots 100 lL (292.5 U/mL) of the AMG stock solution were added to
of the NaPA stock solution, approximately (0.01–0.05) g, until the the tubes, which were mixed in a vortex mixer for 2 min and then
solution became turbid, indicating the formation of a two-phase centrifuged at 1400g for 10 min. The tubes were then incubated for
system. Turbidimetric titration was performed in a thermostatic 6 h at a constant temperature to reach equilibrium. Samples from
bath (Quimis, Brasil) with controlled temperature of 298.15 K. the top and bottom phases were immediately collected with syrin-
The total composition of the system was determined by measuring ges and the AMG activity and total protein concentrations were
the amount of NaPA added, using an analytical balance (Ohaus subsequently determined. To avoid interference from phase com-
adventurer™ pro, Ohaus, USA). A known quantity of water was ponents, samples were analyzed against blanks containing the
then added to the tube in order to retrieve the initial one phase same phase composition but without the enzyme.
system. Further titration with the NaPA stock solution resulted in
a new turbid system and this new composition was determined. 2.3.1. Enzyme activity assay
This procedure was repeated several times to obtain the binodal A 5 mL volume of the buffered starch solution (4% w/w, 0.2 mol/
curve. For the determination of the tie lines, different ATPS were L acetate buffer at pH 4.2) was added to a test tube and placed in a
prepared by weighting defined amounts of the components with water bath (60 °C). Samples of 0.2 mL from the upper or lower
composition greater than that of the binodal curve, to a final phase were added to the tube and the reaction proceeded for
weight of 10 g, in 15 mL graduated centrifuge tubes. All system 60 min. The reaction was stopped by adding 0.8 mL of a 4 M NaOH
components were thoroughly mixed with a vortex mixer and left solution. Aliquots of 1 mL were used for glucose determination by
in a thermostatic bath for 12 h (overnight) for phase separation un- the 3,5-dinitrosalicylic acid (DNS) method, according to Miller [30].
til reaching equilibrium. To ensure complete phase separation, the One international unit (U) of AMG was defined as the amount of
systems were centrifuged at 1400g for 10 min at the respective the enzyme that produce 1 lmol of glucose equivalent in 1 min
temperature (Eppendorf 5804, Germany). The volumes of the top at 60 °C, pH 4.2 and 4% of starch solution. Each assay was carried
and bottom phases were measured and then the phases were sep- out in duplicate and the average value was recorded.
890 L.A. Pereira Alcântara et al. / Separation and Purification Technology 118 (2013) 888–894

2.3.2. Protein quantification 2.5. Experimental design and regression analysis


Total protein quantification in the top and bottom phases was
performed according to the Bradford method [31] using a spectro- Equilibrium diagrams were obtained at two pH values (6.0 and
photometer (Biomate 3, ThermoScientific, USA) with wavelength 9.0) and the experiments were conducted in triplicate. The mean
set at 595 nm. Samples of 1 mL from the top and bottom phases values and standard deviations of concentration were determined
were obtained with a syringe and diluted appropriately. To avoid at each condition and 4 tie lines were determined by linear
interference from phase components, samples were analyzed regression.
against blanks containing the same phase composition but without The effects of pH, temperature, and PEG and NaPA concentra-
the enzyme. tions on the partition coefficient and selectivity were studied
according to a factorial design. Initially, a 24 full factorial with 4
2.3.3. Sodium dodecyl sulfate–polyacrylamide gel electrophoresis central points was applied in order to inspect the variables that
(SDS–PAGE) affect the partition coefficient and selectivity of AMG. This type
SDS–PAGE was performed in a vertical gel electrophoresis of design permits evaluation of the principal effects and first order
system (E-C Apparatus Corporation). The acrylamide gel was pre- interaction of each response. Table 1 lists the coded and uncoded
pared as a 15% separating gel and a 5% stacking gel. Protein sam- factors and each level, where x1, x2, x3, and x4 stand for pH, PEG
ples recovered from the phases were concentrated and concentration (% w/w), NaPA concentration (% w/w), and tempera-
precipitated using pure acetone. The pellets were re-suspended ture (°C). Results were analyzed using the SAS software (SAS Insti-
in denaturing buffer (100 mM Tris–HCl pH 6.8, 4% w/v SDS, 4% tute Inc., v. 9.0, Cary, NC, USA). The linear model (Eq. (6)) was fit to
v/v 2-mercaptoethanol, 20% v/v glycerol and bromophenol blue). the experimental results (Table 1), and the statistical significance
The electrophoresis was run at 80 V and 12 mA for 4 h. Gels of the model was evaluated by the analysis of variance (ANOVA)
were fixed overnight in methanol/acetic acid/water (20/10/70). using Fisher’s statistical test (F-test). The significance and magni-
After, the gel was stained with a buffer solution consisting of tude of the effects estimated for each variable and all their possible
0.05% v/v Coomassie1 Brilliant Blue G-250, 30% v/v methanol linear interactions on the response variables were determined. Ef-
and 10% v/v acetic acid, during 18 h. De-staining was done by fects with a p-value higher than 0.05 were discarded.
keeping the gels for about 7 h in methanol/acetic acid/water
(25/7.5/67.5).
X X
ðK e ; SÞ ¼ b0 þ bi xi þ bij xi xj þ e ð6Þ

2.4. Extraction parameters determination


where b0 is the model intercept, xi and xj are the levels the of inde-
In this work, pre-equilibrated phases with equal volumes were pendent variable, e is the error, and bi and bij are the linear, and
used throughout the experiments, so the volume phase ratio (R) interaction coefficients, respectively.
was always 1. The enzyme was added to the bottom phases and From the results of the ANOVA and after discharging the non
after the partition experiments it was verified that the R-value significant factors (p > 0.05), the selectivity and partitioning of
had not changed. AMG were optimized using response surface methodology (RSM),
The AMG partition coefficient (Ke) is defined as the ratio of the which includes factorial designs and regression analysis, and this
volumetric activity (U/mL) in the top phase (AT) to that in the bot- methodology is also used to evaluate the relative significance of
tom phase (AB): several factors even in the presence of complex interactions [32].
A face-centered design (FCD) with four central points, three vari-
AT
Ke ¼ ð1Þ ables and three levels was performed using the variables pH (x1),
AB PEG concentration (x2) and NaPA concentration (x3), with temper-
The protein partition coefficient (Kp) is the ratio of the equilib- ature maintained constant at 25 °C. The FCD was applied using the
rium concentration (mg/mL) of the total protein in the top phase SAS software (SAS Institute Inc., v. 9.0, Cary, NC, USA).
(CT) to that in the bottom phase (CB):
CT Table 1
Kp ¼ ð2Þ Partition coefficient (Ke) and selectivity (S) for each experimental condition according
CB
to the 24 full factorial design.
Selectivity (S) was calculated as the ratio of the enzyme parti-
Assay Uncoded variables Coded variables Ke S
tion coefficient (Ke) to the protein partition coefficient (Kp):
x1 x2 x3 x4 x1 x2 x3 x4
Ke 1 6.0 8.0 4.0 20.0 1 1 1 1 0.722 0.452
S¼ ð3Þ
Kp 2 9.0 8.0 4.0 20.0 +1 1 1 1 1.417 0.712
3 6.0 8.0 4.0 30.0 1 1 1 +1 1.007 0.962
The purification factor (PF) was calculated as the ratio of the 4 9.0 8.0 4.0 30.0 +1 1 1 +1 0.487 0.401
5 6.0 8.0 8.0 20.0 1 1 +1 1 0.743 0.433
specific activity (U/mg) in the top phase to that added to the sys-
6 9.0 8.0 8.0 20.0 +1 1 +1 1 1.762 1.192
tem, before partitioning: 7 6.0 8.0 8.0 30.0 1 1 +1 +1 1.234 1.243
8 9.0 8.0 8.0 30.0 +1 1 +1 +1 1.303 0.979
AT =C T 9 6.0 12.0 4.0 20.0 1 +1 1 1 1.206 1.146
PF ¼ ð4Þ
AF =C F 10 9.0 12.0 4.0 20.0 +1 +1 1 1 0.786 0.282
11 6.0 12.0 4.0 30.0 1 +1 1 +1 1.047 1.099
where CT and CF are the total protein concentrations, expressed as 12 9.0 12.0 4.0 30.0 +1 +1 1 +1 0.879 0.671
13 6.0 12.0 8.0 20.0 1 +1 +1 1 1.505 1.593
mg/mL, in the top phase and that added to the system, respectively.
14 9.0 12.0 8.0 20.0 +1 +1 +1 1 0.791 0.412
The activity yield was determined as the ratio of the volumetric 15 6.0 12.0 8.0 30.0 1 +1 +1 +1 1.568 1.833
activity (U/mL) in the top phase to that added to the system, and 16 9.0 12.0 8.0 30.0 +1 +1 +1 +1 1.056 0.765
expressed as percentage: 17 7.5 10.0 6.0 25.0 0 0 0 0 1.323 0.799
  18 7.5 10.0 6.0 25.0 0 0 0 0 1.321 0.613
AT 19 7.5 10.0 6.0 25.0 0 0 0 0 1.295 1.127
Y¼ R  100 ð5Þ 20 7.5 10.0 6.0 25.0 0 0 0 0 1.344 0.891
AF
L.A. Pereira Alcântara et al. / Separation and Purification Technology 118 (2013) 888–894 891

The experimental data obtained from the FCD (Table 2) was


analyzed by the response surface regression procedure (RSREG,
SAS Institute Inc., v. 9.0, Cary, NC, USA) using the second order
polynomial equation (Eq. (7)):
X X XX
K e ; S ¼ b0 þ b i xi þ bii x2i þ bij xi xj ð7Þ
i i i j

where b0, bi, bii and bij are regression coefficients for the intercept,
linear, quadratic and interaction effects, respectively, and xi and xj
are the coded factors.
An analysis of variance (ANOVA) for the model was performed
and the model significance was examined by Fisher’s statistical test
(F-test) by testing for significant differences between sources of
variation in experimental results, i.e. the significance of the regres-
sion, the lack of fit, and the coefficient of multiple determination
(R2). Parameters with less than 95% significance (p > 0.05) were ex-
Fig. 1. ATPS equilibrium diagram for PEG 4000 + NaPA 15,000 at 298.15 K: ( – pH
cluded and added to the error term. 6.0; s – pH 9.0).

3. Results and discussion


Table 3
3.1. Phase diagram Equilibrium data for PEG 4000 + NaPA 15,000 system at 278.15 K and pH of 6.0 and
9.0.

Phase diagrams for aqueous two-phase systems containing PEG Total composition Top phase Bottom phase TLL
4000-NaPA 15,000-NaCl at 298.15 K and at pH 6.0 and 9.0 were PEG NaPA PEG NaPA PEG NaPA
experimentally determined. These diagrams are constituted by (w/w) (w/w) (w/w) (w/w) (w/w) (w/w)
tie lines that describe the compositions of the two phases in equi- pH 6.0
librium and a binodal curve that delimits the biphasic to the mon- 0.0719 0.0711 0.1308 0.0038 0.0125 0.1262 0.170
ophasic region (Fig. 1). The experimental liquid–liquid equilibrium 0.0957 0.0946 0.1862 0.0012 0.0046 0.1702 0.248
results for these aqueous two-phase systems at different pH values 0.1274 0.126 0.2587 0.0013 0.0027 0.2272 0.341
0.1677 0.1696 0.3591 0.0012 0.0003 0.2918 0.462
along with the tie line length (TLL) are presented in Table 3. For
each diagram four distinct total compositions were selected and pH 9.0
0.0720 0.0703 0.1367 0.0249 0.0063 0.1528 0.183
the tie lines were determined according to:
0.0948 0.0935 0.1888 0.0167 0.0021 0.2012 0.262
 2  2 0:5 0.1268 0.1252 0.2537 0.0180 0.0010 0.2512 0.343
TLL ¼ wt1  wb1 þ wt2  wb2 ð8Þ 0.1677 0.1690 0.3642 0.0205 0.0003 0.3101 0.465

where w1 and w2 are the mass fractions of PEG and NaPA, respec-
tively, and the superscripts t and b refer to the top and bottom
phases. The tie lines were determined by linear regression of each
The parameters for this equation were determined by least-squares
corresponding set of total, bottom phase, and top phase concentra-
regression and are presented in Table 4 along with the root of mean
tions and good linear fits were obtained.
square error (RMSE).
Each experimental binodal curve was adjusted to the empirical
For each biphasic system, Eq. (9) was successfully adjusted to
equation suggested by Merchuk and co-workers [33]
the experimental data obtained for the binodal curves. The effect
w1 ¼ a expðb w0:5 0:5
2  c w2 Þ ð9Þ of pH on phase separation was evaluated in PEG 4000-NaPA
15,000 as illustrated in Fig. 1. Tie-line slope and TLL was slightly
modified by variations in pH, although a discrete change in the
Table 2 composition of the phases was observed. According to the phase
Partition coefficient, selectivity, purification factor, and yield for each experimental
condition according to the FCD design.
diagram, an increase in pH leads to an increase in the NaPA con-
centration in the top phase, decreasing the size of the two-phase
Assay Uncoded variables Coded variables Ke S region. PEG is an uncharged polymer and NaPA is dissociated only
x1 x2 x3 x1 x2 x3 in pH above 5.0. Consequently, it is more charged at pH 9.0 than 6.0
1 6.0 11.0 10.4 1 1 1 0.830 0.808 and the more the NaPA is charged the more the incompatibility be-
2 6.0 11.0 11.4 1 1 +1 1.258 1.597 tween the polymers and smaller amount of polymers is needed for
3 6.0 12.0 10.4 1 +1 1 1.057 1.168 phase separation.
4 6.0 12.0 11.4 1 +1 +1 0.925 1.238
5 7.0 11.0 10.4 +1 1 1 1.522 0.991
6 7.0 11.0 11.4 +1 1 +1 1.570 0.877 3.2. Analysis of the factors affecting AMG partitioning
7 7.0 12.0 10.4 +1 +1 1 1.326 0.954
8 7.0 12.0 11.4 +1 +1 +1 1.483 0.820
A full factorial design was applied for screening the variables
9 6.0 11.5 10.9 1 0 0 1.022 1.173
10 7.0 11.5 10.9 +1 0 0 1.076 0.741 that significantly affect (p < 0.05) the partition coefficient and
11 6.5 11.0 10.9 0 1 0 1.732 1.425
12 6.5 12.0 10.9 0 +1 0 1.704 1.548
13 6.5 11.5 10.4 0 0 1 1.865 1.488 Table 4
14 6.5 11.5 11.4 0 0 +1 2.018 1.576 Values of parameters of Eq. (9) for the PEG–NaPA systems at 298.15 K.
15 6.5 11.5 10.9 0 0 0 1.707 1.401
pH a b c RMSE
16 6.5 11.5 10.9 0 0 0 1.765 1.538
17 6.5 11.5 10.9 0 0 0 1.653 1.515 6.0 0.6941 27.1012 5.1396 0.0389
18 6.5 11.5 10.9 0 0 0 1.700 1.507 9.0 165.140 40.219 904.946 0.0169
892 L.A. Pereira Alcântara et al. / Separation and Purification Technology 118 (2013) 888–894

selectivity of AMG in ATPS composed of PEG and NaPA, and the re- for both response variables by regression analysis using only the
sults are presented in Table 1. It was observed that AMG has great- significant factors listed previously. Selectivity and the partition
er affinity for the PEG rich-phase, indicated by the partition coefficient could be explained by the following second-order poly-
coefficient and selectivity which presented values greater than nomial equations (Eqs. (10) and (11)).
one. The values of the partition coefficient found in this work were
greater than those found by Furuya et al. [27] (Ke of 0.31–0.69) S ¼ 1:4998  0:160x1 þ 0:003x2 þ 0:0699x3  0:463x21
with the system composed of PEG 4000/Dextran 500. Conse-
 0:138x1 x3  0:0924x2 x3 ð10Þ
quently, these preliminary results indicate the applicability of this
system for the selective extraction of AMG. Considering that the
objective of this study was to purify AMG, selectivity was selected K e ¼ 1:7145 þ 0:1885x1 þ 0:0654x3  0:6787x21 þ 0:2138x23 ð11Þ
for optimization so it is necessary to determine the factors that sig-
nificantly affect this variable.
The ANOVA performed on the reduced models indicated that both
The results obtained from the 24 full factorial experiments were
models are statistically valid, with significant p-values (p < 0.05).
tested by the analysis of variance (ANOVA) and the results are pre-
The coefficients of determination (R2) were found to be 0.931 and
sented in Table 5. It was verified that only the factors pH (x1), NaPA
0.916, for selectivity and partition coefficient, respectively, besides
concentration (x3), and the interaction of pH with PEG concentra-
a non-significant lack of fit (p > 0.05). These results indicated the
tion (x1 * x2) significantly affected (p < 0.05) AMG selectivity. Only
suitability of the model for an appropriate representation of the real
the factors x1 and x3 affected the partition coefficient of AMG; con-
relationship among the studied factors. The conditions that lead to
sequently, the factors x1, x2, and x3 were chosen to be evaluated in
the best selectivity and the partition coefficient of AMG were pH of
the FCD for partitioning optimization.
6.5, PEG concentration of 11% and NaPA concentration of 11.4%. The
optimum selectivity and partition coefficient obtained were 1.6 and
3.3. Optimization of the AMG selectivity 2.0, respectively.
The effects of pH and NaPA concentration on the partition coef-
The experimental results of the FCD are presented in Table 2. ficient of total protein and AMG, and selectivity are visualized in
Statistical analysis of selectivity and the partition coefficient for the plots presented in Fig. 3, with PEG concentration fixed at its
AMG using the full quadratic model was tested by ANOVA using central level (11.0% (w/w)). It is observed in Fig. 3A that in the
a Fisher statistical test. The outcome of the ANOVA can be visual- studied conditions, the effect of pH on the partition coefficient of
ized in a Pareto chart (Fig. 2), in which the standardized estimated the total protein was very strong and the effect of NaPA concentra-
effects of each factor are plotted in decreasing order and compared tion was minimal. When observing the same figure at conditions of
to the minimum magnitude of a statistically significant factor with lower pH (6.0) and higher NaPA concentration (11.4% (w/w)), it is
95% confidence (p = 0.05), represented by the vertical dashed line. observed that the partition coefficient was lower than one, which
As can be observed in Fig. 2A, the factors x1 and x21 , and the inter- means that the protein prefers to transfer to the bottom phase. Un-
actions x1 * x3 and x2 * x3 significantly affected (p < 0.05) AMG der these conditions, NaPA is only partially charged and although
selectivity. On the other hand, the partition coefficient was affected electrostatic repulsion occur between NaPA and the proteins, it is
only by the factors x1 and x3, as well as their quadratic effects (see reasonable to assume that partitioning to the bottom phase is
Fig. 2B). The factors that were not statistically significant were due to the excluded volume effect since the repulsive forces be-
pooled into the error term and new reduced models were obtained tween the NaPA-rich bottom phase and the negatively charged
protein are not effective. It is important to observe in Fig. 3A that
as the concentration of NaPA increased, the partition coefficient
Table 5 Kp also increased, which means that repulsive forces start to play
Analysis of variance performed on the results of AMG selectivity. a role on protein partitioning to the upper phase. As the pH is in-
Source SS DF MS F-values p-values creased, repulsive forces are intensified because the NaPA becomes
x1 0.6999 1 0.6999 15.31 0.0297
more negatively charged and this compensates the opposite effect
x3 0.4648 1 0.4648 10.17 0.0498 of excluded volume in such a way that the partition coefficient is
x1 * x2 0.8714 1 0.8714 19.06 0.0222 greater than one. It can be assumed that the spontaneity of parti-
Lack of fit 0.5722 6 0.0954 2.09 0.2916 tioning is resulted of the increase in the system entropy as the
Pure error 0.1371 3 0.0457
chemical species (H2O or NaPA macro-ions) that were interacting
Total 3.1947 19
with PEG are released into the solution due to the partition of

Fig. 2. Pareto chart for the standardized effects of the variables pH (x1), PEG (x2) and NaPA (x3) concentrations on the AMG partition coefficient (A) and selectivity (B).
L.A. Pereira Alcântara et al. / Separation and Purification Technology 118 (2013) 888–894 893

Fig. 4. SDS–PAGE analysis on the partition of amyloglucosidase. Lane 1: protein


molecular markers; lanes 2 and 3: bottom and top phase of the ATPS, respectively;
lane 4: amyloglucosidase stock solution.

contained the enzyme (89 kDa) pointing towards the increase in


purity of amyloglucosidase by eliminating many other bands pres-
ent in the crude extract (Fig. 4).

4. Conclusions

The present study evaluated the AMG partitioning behavior in


ATPS composed of PEG and NaPA in order to optimize the system
conditions that provide high values of partition coefficient and
selectivity. Phase equilibrium diagrams were determined at pH
values of 6.0 and 9.0, and it was observed that a small reduction
in the two-phase region occurred as the pH was increased. Several
parameters affecting selectivity and partitioning of AMG were
Fig. 3. Effect of pH and NaPA concentration on the partition coefficient of the total studied, and the FCD associated with the response surface method-
protein (A), AMG partition coefficient (B) and selectivity (C) in PEG/NaPA systems. ology was applied in order to optimize enzyme extraction. The re-
sults presented in this work confirmed the feasibility of AMG
extraction by the system proposed, considering the high selectivity
amyloglucosidase and diffuse to the bottom phase, increasing the and partition coefficient that led to a high recovery (86%) and a
configurational entropy of the system. purification factor of 1.9.
On the other hand, when only the AMG partition coefficient is
take in account it is important to consider the effect of phase com- Acknowledgements
ponents on enzyme activity. It is observed in Fig. 3B that the par-
tition coefficient of AMG decreases when the pH is higher than The authors acknowledge the financial support provided for
6.5 and the effect of increasing NaPA concentration is unimportant. this work by the CNPq (National Counsel of Technological and Sci-
According to Hernandez et al. [34], the optimal amylase activity entific Development), FAPEMIG (Foundation for Research Support
produced from A. niger was found at pH of 4.95 and its activity of the State of Minas Gerais) and Prozyn – Brazil.
was visibly decreased (around 30% of the initial activity) as the
pH was increased to 8.0. Thus, the lower partition coefficient of References
AMG with the increasing pH may be a result of stability loss.
Fig. 3C depicts the behavior of AMG selectivity as affected by [1] M. Li, J.-W. Kim, T.L. Peeples, Amylase partitioning and extractive
the pH and NaPA concentration. A region of optimal selectivity bioconversion of starch using thermoseparating aqueous two-phase systems,
J. Biotechnol. 93 (2002) 15–26.
(S = 1.6) was observed at pH 6.5 and 11.4% (w/w) of NaPA. This [2] J.A. James, B.H. Lee, Glucoamylases: microbial sources, industrial applications
condition leads to an optimal partition coefficient for AMG of 2.0. and molecular biology – a review, J. Food Biochem. 21 (1997) 1–52.
Under such conditions, a purification factor of 1.9 and yield of [3] H. Guzmán-Maldonado, O. Paredes-López, Amylolytic enzymes and products
derived from starch: a review, Crit. Rev. Food Sci. Nutr. 35 (1995) 373–403.
86% were obtained as calculated according to Eqs. (4) and (5), thus [4] A.Z. Zidehsaraei, M. Moshkelani, M.C. Amiri, An innovative simultaneous
indicating the usefulness of the system. glucoamylase extraction and recovery using colloidal gas aphrons, Sep. Purif.
SDS–PAGE profile indicating the molecular marker (lane 1), Technol. 67 (2009) 8–13.
[5] A.P. Manera, E.S. Kamimura, L.M. Brites, S.J. Kalil, Adsorption of
amyloglucosidase crude extract (lane 4), and top and bottom phase amyloglucosidase from Aspergillus niger NRRL 3122 using ion exchange resin,
of ATPS (lane 2 and 3) evidenced that the top phase of ATPE mainly Braz. Arch. Biol. Technol. 51 (2008) 1015–1024.
894 L.A. Pereira Alcântara et al. / Separation and Purification Technology 118 (2013) 888–894

[6] S. Srichuwong, T. Orikasa, J. Matsuki, T. Shiina, T. Kobayashi, K. Tokuyasu, [21] B. Mattiasson, Applications of aqueous two-phase systems in
Sweet potato having a low temperature-gelatinizing starch as a promising biotechnology, Trends Biotechnol. 1 (1983) 16–20.
feedstock for bioethanol production, Biomass Bioenergy 39 (2012) 120–127. [22] V.S. Sampaio, R.C.F. Bonomo, E.S. Monteiro Filho, V.P.R. Minim, L.A. Minim,
[7] S. Harsa, S. Furusaki, Chromatographic separation of amyloglucosidase from Physical properties and liquid–liquid equilibrium of aqueous two-phase
the mixtures of enzymes, Biochem. Eng. J. 8 (2001) 257–261. systems containing poly(ethylene glycol) ± potassium chloride ± sodium
[8] I. Paszczynski, J. Miedziuk, J. Lobazewski, J. Kochmanska, Trojanowski, A polyacrylate, J. Chem. Eng. Data 57 (2012) 3651–3657.
simple method of affinity chromatography for the purification of glucoamylase [23] S. Saravanan, J.A. Reena, J.R. Rao, T. Murugesan, B.U. Nair, Phase equilibrium
obtained from Aspergillus niger C, FEBS Lett. 149 (1982) 63–66. compositions, densities, and viscosities of aqueous two-phase poly(ethylene
[9] A. Sharma, S. Sharma, M.N. Gupta, Purification of wheat germ amylase by glycol) + poly(acrylic acid) system at various temperatures, J. Chem. Eng. Data
precipitation, Protein Expr. Purif. 18 (2000) 111–114. 51 (2006) 1246–1249.
[10] S. Teotia, R. Lata, S.K. Khare, M.N. Gupta, One-step purification of glucoamylase [24] H.-O. Johansson, F.M. Magaldi, E. Feitosa, A.P. Junior, Protein partitioning in
by affinity precipitation with alginate, J. Mol. Recognit. 14 (2001) 295–299. poly(ethylene glycol)/sodium polyacrylate aqueous two-phase systems, J.
[11] T. Gouveia, B.V. Kilikian, Bioaffinity extraction of glucoamylase in aqueous Chromatogr., A 1178 (2008) 145–153.
two-phase systems using starch as free bioligand, J. Chromatogr., B 743 (2000) [25] H.-O. Johansson, M. Ishii, M. Minaguti, E. Feitosa, T.C.V. Penna, A. Pessoa
241–246. Jr., Separation and partitioning of green fluorescent protein from
[12] S. Tanuja, N.D. Srinivas, K.S.M.S. Raghava Rao, M.K. Gowthaman, Aqueous two- escherichia coli homogenate in poly(ethylene glycol)/sodium-
phase extraction for downstream processing of amyloglucosidase, Process poly(acrylate) aqueous two-phase systems, Sep. Purif. Technol. 62 (2008)
Biochem. 32 (1997) 635–641. 166–174.
[13] M. Li, T.L. Peeples, Purification of hyperthermophilic archaeal amylolytic [26] L.A. Minim, R.C.F. Bonomo, I.V. Amaral, M.F.T. Reis, A.A.A. Oliveira, V.P.R.
enzyme (MJA1) using thermoseparating aqueous two-phase systems, J. Minim, Density and viscosity of binary and ternary mixtures of poly(ethylene
Chromatogr., B 807 (2004) 69–74. glycol) and polyacrylic acid (sodium salt) at temperatures of (288.15 to
[14] S. Shahriari, M. Manouchehr Vossoughi, V. Taghikhani, A.A. Safekordi, I. 318.15) K, J. Chem. Eng. Data 55 (2010) 2328–2332.
Alemzadeh, Experimental study and mathematical modeling of partitioning of [27] T. Furuya, Y. Iwai, Y. Tanaka, H. Uchida, S. Yamada, Y. Arai, Measurement and
b-amylase and amyloglucosidase in PEGsalt aqueous two-phase systems, J. correlation of partition coefficients of hydrolytic enzymes for
Chem. Eng. Data 55 (2010) 4968–4975. dextran+poly(ethylene glycol) + water aqueous two-phase systems at 20 °C,
[15] A.M. Azevedo, A.G. Gomes, P.A.J. Rosa, I.F. Ferreira, A.M.M.O. Pisco, M.R. Aires- Fluid Phase Equilib. 110 (1995) 115–128.
Barros, Partitioning of human antibodies in polyethylene glycol–sodium [28] P.A. Albertsson, Partitioning of Cell Particles and Macromolecules, third ed.,
citrate aqueous two-phase systems, Sep. Purif. Technol. 65 (2009) 14–21. John Wiley & Sons, New York, 1986.
[16] L.A. Alcântara, L.A. Minim, V.P. Minim, R.C. Bonomo, L.H. da Silva, Application [29] R. Hatti-Kaul, Aqueous Two-Phase Systems: Methods and Protocols, first ed.,
of the response surface methodology for optimization of whey protein Humana Press, New Jersey, 2000.
partitioning in PEG/phosphate aqueous two-phase system, J. Chromatogr., B: [30] G.L. Miller, Use of dinitrosalicylic acid reagent for determination of reducing
Anal. Technol. Biomed. Life Sci. 879 (2011) 1881–1885. sugar, Anal. Chem. 31 (1959) 426–428.
[17] R. Kammoun, H. Chouayekh, H. Abid, B. Naili, S. Bejar, Purification of CBS [31] M.M. Bradford, A rapid and sensitive method for the quantitation of
819.72 [alpha]-amylase by aqueous two-phase systems: modelling using microgram quantities of protein utilizing the principle of protein-dye
response surface methodology, Biochem. Eng. J. 46 (2009) 306–312. binding, Anal. Biochem. 72 (1976) 248–254.
[18] K. Berggren, A. Wolf, J.A. Asenjo, B.A. Andrews, F. Tjerneld, The surface exposed [32] A.M. Azevedo, P.A.J. Rosa, I.F. Ferreira, M.R. Aires-Barros, Optimisation of
amino acid residues of monomeric proteins determine the partitioning in aqueous two-phase extraction of human antibodies, J. Biotechnol. 132 (2007)
aqueous two-phase systems, Biochim. Biophys. Acta 1596 (2002) 253–268. 209–217.
[19] A. Zaslavsky, N. Gulyaeva, B. Zaslavsky, Peptides partitioning in an aqueous [33] J. Merchuk, B. Andrews, J. Asenjo, Aqueous two-phase systems for protein
dextran-polyethylene glycol two-phase system, J. Chromatogr., B 743 (2000) separation-studies on phase inversion, J. Chromatogr., B 711 (1998) 285–
271–279. 293.
[20] K.S. Nascimento, A.M. Azevedo, B.S. Cavada, M.R. Aires-Barros, Partitioning of [34] M.S. Hernandez, R.R. Marilu, P.G. Nelson, P.R. Renato, Amylase production by
Canavalia brasiliensis lectin in polyethylene glycol–sodium citrate aqueous Aspergillus niger in submerged cultivation on two wastes from food industries,
two-phase systems, Sep. Sci. Technol. 45 (2010) 2180–2186. J. Food Eng. 73 (2006) 93–100.

You might also like