Np

RÁa

b

c

a

ARRAA

KATPPL

1

pampctpfd

tWmss

Ff

a

h1

Process Biochemistry 50 (2015) 1459–1467

Contents lists available at ScienceDirect

Process Biochemistry

jo ur nal home p age: www.elsev ier .com/ locate /procbio

ovel aqueous two-phase systems based on tetrahydrofuran andotassium phosphate buffer for purification of lipase

anyere L. Souzaa,b, Rafaella A. Limaa, João A.P. Coutinhoc, Cleide M.F. Soaresa,b,lvaro S. Limaa,b,∗

Universidade Tiradentes, Av. Murilo Dantas 300, Farolândia, 49032-490 Aracaju-SE, BrazilITP, Instituto de Tecnologia e Pesquisa, Av. Murilo Dantas, 300-Prédio do ITP, Farolândia, 49032-490 Aracaju-SE, BrazilDepartamento de Química, CICECO, Universidade de Aveiro, 3810-193 Aveiro, Portugal

r t i c l e i n f o

rticle history:eceived 20 November 2014eceived in revised form 14 May 2015ccepted 20 May 2015vailable online 27 May 2015

eywords:

a b s t r a c t

Aqueous two-phase systems (ATPS) based on tetrahydrofuran (THF) + potassium phosphate buffer (pH7) were used in this work for the purification of lipases. Binodal curve, tie lines and critical point wereobtained for the new THF–salt ATPS at 25 ◦C and the binodal curve were successfully correlated withthe Merchuck and Hu equations. To optimize the extraction capability of this ATPS the effects of theconcentration of components and temperature of equilibrium on the partition coefficients and extractionefficiencies were investigated using lipase from Burkholderia cepacia (commercially obtained) as a model

queous biphasic systemsetrahydrofuranhosphate bufferurification

compound. The optimum conditions for the purification of an extracellular lipase obtained by submergedfermentation were established through a surface response analysis by central composite rotational designapplied allowing a purification factor (PF) of 103.9 ± 0.9 and an enzyme recovery of 96.4 ± 1.1 achievedusing this process. Moreover, a commercial lipase by Candida antarctica B recombinant in Aspergillus niger

.24),

ipase was purified (PF = 4.84 ± 0

. Introduction

Lipases, triacylglycerol ester hydrolases (EC 3.1.1.3), occupy alace of prominence among biocatalysts because of their potentialpplications in various industries such as food, dairy, phar-aceutical, detergents, textile, biodiesel, and cosmetic, besides

articipating in the synthesis of fine chemicals and agrochemi-als [1–5]. The lipases especially of microbial origin are producedhrough a fermentation process that, besides the desired com-onents, also generates secondary or intermediate products thatrequently hinder the use of fermented broth in industrial proce-ures [6,7].

Efficient downstream processing techniques are essential forhe success of commercial production of enzymes and proteins.

hen these processes are applied to biological or pharmaceutical

aterials, these purification steps must be delicate enough to pre-

erve the activity of these biomolecules [8]. Aqueous two-phaseystems (ATPS) can be foreseen as an alternative technique for

∗ Corresponding author at: Universidade Tiradentes, Av. Murilo Dantas 300,arolândia, 49032-490 Aracaju-SE, Brazil. Tel.: +55 7932182115;ax: +55 7932182190.

E-mail addresses: alvaro lima@unit.br, alvaro lima@itp.org.br,slima2001@yahoo.com.br (Á.S. Lima).

ttp://dx.doi.org/10.1016/j.procbio.2015.05.015359-5113/© 2015 Elsevier Ltd. All rights reserved.

confirming the potential of this new THF-based ATPS for purifying lipases.© 2015 Elsevier Ltd. All rights reserved.

extraction and/or purification of biocompounds since they have alow cost and lead to a high product purity, while maintaining thebiological activity of the molecules due to their water-rich envi-ronment [9]. Conventional ATPS are formed by two water-solublepolymers or polymer–salt combinations that phase separate abovegiven concentrations [10]. These systems have been used to theseparation and purification of a great number of biological prod-ucts, such as proteins, genetic material, bionanoparticles, cells andorganelles [11,12]. Despite all these advantages, the limited rangeof polarity of the polymer-based systems, limits their applicabilityin the purification of biomolecules [13,14]. Currently, the numberof systems capable of forming two aqueous phases is increasing andsome alternatives include the use of alcohol/salt [15–17], acetoni-trile/sugars [18,19], acetonitrile/polyols [20] and ionic liquids (ILs)[21,22].

ATPS composed of a hydrophilic organic solvent and an inor-ganic salt solution have many advantages, which include rapidphase-separation, high extraction efficiency, low viscosity, highpolarity differences between the phases, a gentle aqueous envi-ronment, and may be formed by inexpensive chemicals easy torecycle [15,23]. These systems, formed by adding a salt solution to

an aqueous solution of an organic compound, have been recentlyproposed and used for the partition of different biomolecules, suchas proteins, amino acids and other natural products [15–17,24,25].

1 chem

isfca(rAaf

bsdwitocottfa(ta

2

2

(cc–aCmbefolo

2

2

pPmifNa7Aw

2

s

460 R.L. Souza et al. / Process Bio

In this sense, the tetrahydrofuran (THF) gained prominencen recent years due your excellent solvent power for numerousubstances, to include in extraction processes of biocompoundsrom plants, such as, commercially valuable compounds (i.e.arotenoids) [26]. Taha et al. [27] have demonstrated thebility of THF to form ATPS with the biological buffer 4-2-hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES). Moreecently Hirayama et al. [28] evaluated the ability of THF to formTPS with the IL [C4mim]Cl. To the best of our knowledge thesere the only works available in the literature that use THF in theormation of ATPS.

In the present work, the phase diagrams of novel ATPS formedy tetrahydrofuran (THF) + potassium phosphate buffer (pH 7) weretudied at 25 ◦C. The binodal curve, tie-lines and critical points wereetermined for the studied system. Additionally, the binodal dataere correlated using the Merchuk [29] or Hu [30] equations. Aim-

ng at exploring the applicability of those ATPS it is here shown, forhe first time, the ability of THF based systems for the purificationf lipase. The enzyme produced by Burkholderia cepacia (commer-ially obtained) was here used as a model to evaluate the profilef the enzymatic partition, namely considering the overall sys-em composition and temperature of equilibrium. Subsequently,o establish the optimum conditions for the purification of lipaserom Bacillus sp. ITP-001 produced by submerged fermentation annalysis of response surface by central composite rotational designCCRD) was used. The ATPS optimum conditions were also usedo evaluating the purification of commercial lipase from Candidantarctica B recombinant in Aspergillus niger.

. Materials and methods

.1. Materials

The chemicals used in the present study, tetrahydrofuranTHF, ≥99.9 wt% purity) and the potassium phosphate salts thatompose the buffer K2HPO4/KH2PO4 (≥99.9 wt% purity) were pur-hased from Sigma-Aldrich. The lipase from Burkholderia cepacia

BCL (≥30,000 U g−1, pH 7.0, 50 ◦C – optimum pH and temper-ture) was purchased from Sigma-Aldrich, lipase produced byandida antarctica B recombinant in Aspergillus niger under com-ercial name Lipozyme CALB L (≥5000 U g−1) was kindly offered

y Novozymes Latin America Ltd and the extracellular lipolyticnzyme from Bacillus sp. ITP-001 was produced by submergedermentation in this work. The ammonium sulphate (P.A.) wasbtained from Synth (Brazil) and coconut oil was purchased at aocal market. The bovine serum albumin (BSA, ≥97 wt% purity) wasbtained from Merck.

.2. Production of lipase by Bacillus sp. ITP-001

.2.1. Fermentation conditionsThe microorganism of Bacillus sp. ITP-001 was isolated from

etroleum-contaminated soil by the Instituto de Tecnologia eesquisa – ITP (Aracaju-Sergipe, Brazil). The bacterial strain wasaintained on nutrient agar slants at 4 ◦C. The strain was cultivated

n 500 mL erlenmeyer flasks containing 200 mL medium with theollowing composition (%, w/v): KH2PO4 (0.1), MgSO4·7H2O (0.05),aNO3 (0.3), yeast extract (0.6), peptone (0.13), and starch (2.0)s the carbon source. The fermentation conditions were: initial pH; incubation temperature of 37 ◦C, and stirring speed of 170 rpm.fter 72 h of cultivation, coconut oil (4%, v/v) and Triton X-100 (1%,/v) were added as inductors as described by Feitosa et al. [31].

.2.2. Pre-purification stepsThe fermented broth was centrifuged at 3000 rpm for 15 min,

o that the bottom phase was discharged (biomass) and the

istry 50 (2015) 1459–1467

supernatant was used to determine the enzymatic activity and thetotal protein. Protein contaminants in the cell-free fermented brothwere precipitated using ammonium sulphate at 80% (w/v) satura-tion. The solution was prepared at room temperature and the brothwas subsequently centrifuged at 3000 rpm for 30 min, separatingthe aqueous solution and precipitate. The aqueous phase was dia-lyzed using MD 25 (cut-off: 12,000–16,000 Da) against ultra-purewater for 24 h at 4 ◦C. The dialyzed solution containing the enzymewas then used to prepare the ATPS. These pre-purification stepswere previously described by our group [7].

2.3. Phase diagrams and tie-lines

The binodal curves of the ATPS were determined through thecloud point titration method at 25 ± 1 ◦C and at atmospheric pres-sure. In a test vial, a THF solution of known concentration wasadded, and a potassium phosphate buffer (K2HPO4/KH2PO4) solu-tion of known mass fraction was then added dropwise until themixture became turbid, then a known mass of water was addedto make the mixture clear again. The potassium phosphate bufferwas a mixture of potassium phosphate monobasic and bibasic ata ratio of 1.087 (w/w) and pH 7. This procedure was repeated toobtain sufficient data for the construction of a liquid–liquid equi-librium binodal curve. The compositions were determined by theweight quantification of all components added within an uncer-tainty of ±10−5 g. The experimental tie-lines (TLs), were measuredwith the procedure outlined in our previous work [20] and theirrespective length (tie line length – TLL) were determined throughthe application of Eq. (1), based on the concentrations of THF andsalt (K2HPO4/KH2PO4, at pH 7) in the two phases.

TLL =√

([THF]T − [THF]B)2 + ([salt]T − [salt]B)2 (1)

where the indexes T and B are of top and bottom phases, respec-tively.

The location of the critical point of the ternary systems was esti-mated by extrapolation of the TLs compositions applying Eq. (2)[21].

[THF] = f + g[salt] (2)

where f and g are fitting parameters.

2.4. Preparation of the ATPS and Lipase partition studies

The biphasic systems were prepared in graduated centrifugetubes (15 mL) by weighing the appropriate amounts of THF(10–20 wt%) and salt (10–25 wt%) All systems contained approx-imately 2 wt% of BCL (≈20 mg mL−1). For the partition studies oflipase from Bacillus sp. ITP-001 the THF and salt (K2HPO4/KH2PO4at pH 7) aqueous solutions were prepared with the dialysate solu-tion obtained from the pre-concentration of the lipase from Bacillussp. ITP-001. Optimization of ATPS conditions for the partition coef-ficient and recovery of enzymes from Bacillus sp. was performedusing a 22 central composite rotational design (CCRD). The CCRDwas used with four axial points and three center points, result-ing in 11 experiments. Version 8.0 of the STATISTICA software wasused to determine regression and for data graphical analysis. Therange and levels of the independent variables and coded values arepresented in Table 1.

Each mixture was prepared gravimetrically within ±10−5 g.After 2 min of gentle stirring, the systems were centrifuged at3000 rpm for 20 min. The tubes were brought to equilibrium in

a thermostatic bath at 25 ± 1 ◦C and local atmospheric pressure(1 atm) overnight (for at least 12 h). The equilibration time usedwas selected after preliminary kinetic tests to ensure that equilib-rium was achieved in the measurements. After this treatment, the

R.L. Souza et al. / Process Biochem

Table 1Experimental design of the central composite rotational, partitioning of lipase andrecovered in the bottom phase from ATPS.

Test set Extraction conditions Parameters of extraction

X1, THF (wt%) X2, Salt (wt%) KE REB

1 −1 (15) −1 (15) 0.25 ± 0.02 91.77 ± 0.712 −1 (15) 1 (25) 0.15 ± 0.01 95.03 ± 0.393 1 (25) −1 (15) 0.18 ± 0.03 89.54 ± 1.714 1 (25) 1 (25) 0.13 ± 0.01 93.13 ± 0.515 −1.41 (12.95) 0 (20) 0.24 ± 0.01 93.63 ± 0.156 1.41 (27.10) 0 (20) 0.25 ± 0.03 91.74 ± 0.997 0 (20) −1.41 (12.95) 0.39 ± 0.05 89.51 ± 1.098 0 (20) 1.41 (27.10) 0.26 ± 0.02 86.22 ± 0.789 0 (20) 0 (20) 0.08 ± 0.02 96.87 ± 0.73

twftakaiaw

tca

K

K

wettc

cvtef

S

R

R

P

TS

2

mps

10 0 (20) 0 (20) 0.09 ± 0.03 96.52 ± 0.5211 0 (20) 0 (20) 0.11 ± 0.01 95.78 ± 0.81

wo phases became clear and transparent and the interface wasell defined. The phases were carefully withdrawn using a pipette

or the top phase and a syringe with a long needle for the bot-om phase. During the equilibration period, to avoid contaminationnd/or evaporation of THF (boiling point at 66 ◦C) the vials wereept closed. The volumes and weights were determined in gradu-ted test tubes (the total mass of the extraction systems prepareds 5.0 g). At least three independent replicates were made and theverage partition coefficients and associated standard deviationsere therefore determined.

The partition coefficient was defined as the protein concentra-ion (KP) or enzyme activity (KE) in the top phase, divided by theorresponding value in the bottom phase, as describe of by Eqs. (3)nd (4).

P = CT

CB(3)

E = EAT

EAB(4)

here C is the total protein concentration (mg mL−1), EA is thenzyme activity (U mL−1), and the subscript T and B are top and bot-om phases, respectively. Selectivity (S) was defined as the ratio ofhe lipase enzyme partition coefficient (KE) to the protein partitionoefficient (KP).

In order to evaluate the purification process, the enzyme spe-ific activity (SA, U mg−1 protein) was calculated using Eq. (5), theolume ratio between the volumes of top and bottom phases (RV),he contaminant proteins recovered in the top phase (RPT, %), thenzyme recovered in the bottom phase (REB, %), and the purificationactor (PF – fold) were calculated according to Eqs. (6)–(8).

A = EA

C(5)

PT = 1001 + (1\RVKP)

(6)

EB = 1001 + RVKE

(7)

F = SA

SAi(8)

he purification factor (PF) was calculated as the ratio between theA in the bottom phase and the initial specific activity (SAi).

.5. Enzymatic activity

Lipolytic activity was assayed using the modified oil emulsionethod proposed by Soares et al. [32]. The substrate was pre-

ared by mixing 50 mL of olive oil with 50 mL of Arabic gumolution (7%, w/v). The reaction mixture containing 5 mL of the

istry 50 (2015) 1459–1467 1461

oil emulsion, 2 mL of 100 mM sodium phosphate buffer (pH 7)and enzyme extract (1 mL) was incubated in a thermostated batchreactor for 5 min at 37 ◦C. A blank titration was done on a samplewhere the enzyme was replaced with distilled water. The reactionwas stopped by the addition of approximately 0.33 g of sample in2 mL of acetone–ethanol–water solution (1:1:1). The liberated fattyacids were titrated with 40 mM potassium hydroxide solution inpresence of phenolphthalein as indicator. One unit (U) of enzymeactivity was defined as the amount of enzyme that liberated 1 �molof free fatty acid per min (�mol min−1) under the assay conditions(37 ◦C, pH 7, 120 rpm).

2.6. Protein assay

The total concentration of protein in each aqueous phase wasquantified through Bradford’s method [33], using a Varian Cary 50Bio UV–Vis Spectrophotometer at 595 nm, and a calibration curvepreviously established for the standard protein bovine serum albu-min (BSA).

2.7. SDS-PAGE electrophoresis

Electrophoresis was performed with the Mini-PROTEAN TetraSystem (BioRad, Brazil) using 12% resolving gels and 5% stack-ing gels as described by Laemmli [34]. Proteins were visualizedby staining with silver stain [7,35]. Protein markers used weretrypsin inhibitor (21.5 kDa), carbonic anhydrase (31 kDa), oval-bumin (45 kDa), bovinealbumin (66.2 kDa), and phosphorylase(97.4 kDa) purchased from BioRad (Brazil).

3. Results and discussion

3.1. Binodal curve and correlation

The development of novel, more performant, and economicATPS, with low viscosity and high polarity difference between thetwo phases to recovery or purify enzymes is a priority issue in thiswork, the choice of salt (potassium phosphate buffer, pH 7) wasdue their ability to form ATPS with less polar compounds, due toits hydration capacity [36] and because it was previously used withsuccess for the purification of lipases by ATPS [7,8,35,37,38]. Aque-ous solutions of THF (80 wt%) and of salt (40 wt%) were initiallyprepared and used for the determination of the binodal curve at25 ± 1 ◦C and atmospheric pressure. Using the cloud point titrationmethod [13], it was possible to observe the ability of THF to formtwo phases with the salt. This is a direct consequence the forma-tion of hydration complexes between the water and salt, reducingthe ability to hydrogen bond between the salting water and THF.The binodal curve, shown in Fig. 1, was fitted using the Merchukequation (Eq. (9)) [29].

[THF] = A × exp{(B × [salt]0.5) − (C × [salt]3)} (9)

In order to get a more accurate fitting, we also used a non-linerempirical expression (Eq. (10)) [30], with four fitting parameters tocorrelate the binodal data. This equation has been successfully usedfor the correlation of binodal data of alcohols + salt ATPS [39,40].

[THF] = exp(A + B × [salt]0.5 + C × [salt] + D × [salt2] (10)

The regression parameters were estimated by least-squaresregression using Eqs. (9) and (10), and their values with the respec-tive standard deviations (std) and correlation factors (R2) along with

the weight fraction experimental data (w) for the system are givenin Table 2. On the basis of the obtained R2 and std in Table 2, in gen-eral, good correlation were obtained for the two equations used,indicating that these fittings can be used to predict data in a given

1462 R.L. Souza et al. / Process Biochemistry 50 (2015) 1459–1467

Fig. 1. Binodal curve for the ternary system composed of THF +[K2HPO4/KH2PO4] + water, at 25 ± 1 ◦C and atmospheric pressure. , calcu-

lated binodal from Eq. (9); , experimental solubility data; , tie-line data; ,

tie-lines; , auxiliary curve data; , critical point.

Table 2Parameters obtained through Eqs. (9) and (10) with the respective standarddeviations (std) and correlation factors (R2) along with the weight fraction experi-mental data (w) for the systems composed of THF (1) + potassium phosphate buffer(2) + H2O, at 25 ± 1 ◦C.

Binodal parameters

Eq. (9) Eq. (10)

A 97.7 ± 0.8 4.48 ± 0.06B −0.545 ± 0.004 −0.456 ± 0.053C 9.9 × 10−6 ± 1.1 × 10−6 −0.0198 ± 1.3 × 10−2

D 1.9 × 10−13 ± 2.010−4

R2 0.9997 0.9995

Experimental data

100 w1 100 w2 100 w1 100 w2 100 w1 100 w2

39.84 2.66 23.42 6.63 9.39 17.8738.44 2.89 22.72 7.07 9.16 18.1936.91 3.36 21.97 7.29 8.56 18.5835.11 3.50 21.56 7.46 8.35 19.0934.15 3.72 21.15 7.61 8.06 19.4933.41 3.89 19.82 8.66 7.50 20.3832.51 4.21 19.49 8.77 5.78 24.6431.33 4.45 19.06 8.99 5.51 25.1930.36 4.60 18.69 9.24 5.22 26.2329.78 4.75 18.29 9.47 4.83 26.6928.78 4.96 17.92 9.70 4.52 27.3228.08 5.16 17.35 9.97 4.22 28.0527.50 5.40 17.02 10.17 3.99 29.3227.01 5.54 16.68 10.34 3.70 30.6526.52 5.80 16.36 10.54 2.75 33.5125.55 6.04 16.03 10.79 2.20 35.5425.11 6.17 15.74 10.98

ra

w

Fig. 2. Effect of concentration of THF on the stability of lipase from Burkholderia cepa-cia. The crude lipase feedstock was incubated at room temperature up to 24 h. The

TE

24.71 6.25 14.25 12.6824.16 6.45 13.54 13.1323.79 6.53 9.80 17.49

egion of the phase diagram where no experimental results arevailable.

A series of tie-lines in the two-phase region of the binodal curveere investigated and are reported in Table 3 and shown in Fig. 1,

able 3xperimental data of TLs, TLLs and critical point values of the coexisting phases for the TH

Weight fraction/(wt%)

[THF]M [salt]M [THF]T [salt]T [THF]B

13.22 19.98 53.34 1.23 6.23

20.14 19.98 64.76 0.57 4.87

29.99 17.97 75.44 0.23 4.41

19.99 23.96 83.53 0.08 3.76

a In the critical point: f = 52.84; g = 4:53; and R2 = 0.9877.

relative activity was measured using a lipase assay. The lipase activity of phosphatebuffer (pH 7.0) was used as the control. The THF concentrations were expressed as(wt%).

together with the overall composition, TLL, and critical point. Thetie-lines are approximately parallel to each other, thus allowingeasily to estimate the coexisting phase compositions for any sys-tem. The critical point (cp) for the studied systems was estimatedby extrapolation from the TLs compositions applying Eq. (2).

After the complete characterization of the studied ATPS by thedetermination of the phase diagram, TLs, TLLs and critical point, theeffect of THF on the stability of the lipase and the ability of the sys-tem composed of THF + salt to purify the enzyme was investigated.

3.2. Effect of THF in lipase activity

In the traditional purification processes of proteins, several stepsof manipulation are required and the enzyme activity unavoidablydecreases in each step of the purification. In order to examine theeffect of composition in THF-rich phase on the lipase-stability, solu-tions were prepared at different concentrations by dissolving thepure THF in distilled water, and then crude BCL was mixed withTHF solutions. For this study a wide range of concentrations of theTHF (10–80 wt%), and times of the incubation (up to 24 h) wereconsidered. The study was carried at 25 ± (0.1) ◦C.

The results are shown in Fig. 2, from which it is surprising to findthat the activity of lipase was not decreased up to 18 h in contactwith THF. In general, there is an increase of enzymatic activity withthe use of THF, regardless of its concentration. The THF is an organicsolvent of hydrophilic character (log P = 0.46) [41]. Although it isknown that hydrophobic organic solvents (log P > 4) may improvethe stability of enzymes by stimulating the open conformation ofthe active site of the lipase [42], some studies have also reportedincreased stability of lipases using hydrophilic organic solvents.Activity of lipase from Streptomyces sp. CS133 was significantlyincreased in presence of organic solvents with log P = 0.87 (diethylether), log P = 1.25 (dichloromethane) and log P = 2.0 (benzene)

while the relative activities were 123, 129 and 161%, respectively[43]. In addition, previous studies have shown that lipase fromBurkholderia can maintain highly stable catalytic activity in organicsolvents, such as ethanol, 1-propanol and 2-propanol, all leading to

F + potassium phosphate buffer (pH 7) system at 25 ± 1 ◦C.

[salt]B TLL [Y]Criticala [X]Critical

a

23.24 51.99 11.23 14.1526.62 65.3227.95 76.5230.06 85.22

chem

a(ibsbTAh

3

butottsi

w2eb

rtata

Fpcaw

R.L. Souza et al. / Process Bio

relative activity above 99.5% in organic solvent solutions at 40%v/v) [15,44]. Lipase from Aspergillus carneus was also investigatedn various organic solvents, and it was found stable in iso-octane,enzene, toluene and xylene [4]. It is believed that the organicolvent contributes to maintain the enzymes open conformationy exposing its active site and thereby stimulating lipase activity.herefore, at the present stage it is appropriate to admit that theTPS formed with THF are promising systems for the efficient andigh activity extraction and purification of lipases.

.3. Partition of the model lipase

To optimize the lipase partition in a THF + potassium phosphateuffer (pH 7), eight systems (each in triplicate) were evaluatedsing the BCL as model lipase. These systems systematically variedhe THF and potassium phosphate concentrations and temperaturef formation of the system. It should be remarked that for the sys-ems tested, the bottom phase is the salt-rich phase whereas theop phase corresponds to the THF-rich phase. These systems wereelected so that the liquid–liquid systems could be formed takingnto account their phase diagram.

According to Fig. 3, the optimal condition for partitioning lipaseas observed in the THF 20 wt% and potassium phosphate buffer

0 wt% system, which has a KE of 0.42 ± 0.01. For the extractionfficiency it was observed for THF 10 wt% and potassium phosphateuffer 20 wt% (REB = 91.8 ± 0.3%).

In Fig. 3 (i), the effect of potassium phosphate at concentrationsanging between 10 and 20 wt% was investigated. The concentra-

ion of THF was fixed at 20 wt% and the system was operated at 25 ◦Cnd pH 7. A gradual increase in salt concentration, favored the par-itioning and the enzyme recovery for the bottom phase (salt-rich)s the value of KE decreased from 0.73 to 0.42, and the value of REB

ig. 3. Enzyme recovered in the bottom phase (%, REB – represented by bars) andartition coefficients (KE – represented by symbols) of lipase from Burkholderiaepacia, for systems based in THF + [K2HPO4/KH2PO4] + water, at 25 ± (0.1) ◦C andtmospheric pressure, as a function of concentration: (i), 20 wt% THF + wt% salt; (ii),t% THF + 20 wt% salt.

istry 50 (2015) 1459–1467 1463

increase from 56.8 to 65.8%. This trend is observed up to concentra-tions of 20 wt%, after this optimal concentration other effects leadto the migration of lipase for the opposite phase, THF-rich (KE > 1).According to Babu et al. [45], increasing the concentration of saltdecreases the solubility of biomolecules in the salt-rich phase (bot-tom), which results in increased partitioning of biomolecules to thetop phase, due to a salting out effect. Souza et al. [8] observed thesame trend, where the partition efficiency of porcine pancreaticlipase is negatively affected, for phosphate concentrations above of18 wt% in ATPS with PEG 1500 g mol−1 and above 20 wt% in ATPSwith PEG 8000 g mol−1. In all these works, the lipase preferentiallymigrated to the salt-rich phase. This fact is due to the pH of thesaline phase (pH = 7.0) being above the isoelectric point (pI = 6.0) ofthe enzyme, resulting in increased affinity for the more hydrophilicsalt-rich phase [7,46].

Following the study of optimization, the concentration of thepotassium phosphate was fixed in 20 wt% and the concentrationof THF ranged from 10 to 20 wt% (Fig. 3 (ii)). The increase of theTHF concentration leads to lower partition coefficients of lipase,that is, to a higher ability of lipase to migrate for the salt-richphase. The values of KE decrease from 0.64 (with 10 wt% THF) to0.42 (with 20 wt% THF). This effect is due to the enrichment of themost hydrophobic region (THF-rich phase, with log P 0.53), whichfavors the migration of lipases for the salt-rich phase. However,the enzyme recovery in the bottom phase has a tendency oppo-site to the partition coefficient. The maximum value of recoveryis achieved with 10 wt% THF (REB = 91.8 ± 0.3%). At low concentra-tions of THF in the top phase, the volume ratio of the phases ismuch smaller (RV = 0.14) than in higher concentrations, for exam-ple with 20 wt% of THF (Rv = 1.25). This is a result of the gradualincrease in the ability of THF to conduct interactions with watervia hydrogen bonds [47]. For the following steps of this work, dataof enzyme partition (KE) are taken into account due to the higherselectivity of lipase for the salt-rich phase, which may allow theincrease of the purification factor (PF) when the objective is to applythese conditions for the purification of lipase from Bacillus sp. ITP-001. Data of enzyme recovery in the bottom phase (REB), partitioncoefficients (KE) and volumetric ratio (RV) for ATPS with differentcompositions of THF and potassium phosphate buffer in (wt%) areshown in Supplementary Table A.1.

To assess the effect of temperature on lipase partitioning, thesystem composed of THF at 20 wt% and potassium phosphate bufferat 20 wt%, was chosen because it represents the best condition ofpartitioning of the lipase. The thermodynamic functions calculatedfor the transfer of lipase, namely the molar Gibbs energy (�G◦

m),the molar enthalpy (�H◦

m) and the molar entropy of transfer (�S◦m),

Eqs. (11)–(13) were used according to,

ln K = −�H◦m

R× 1

T+ �S◦

m

R(11)

�G◦m = �H◦

m − T�S◦m (12)

�G◦m = −RT ln(K) (13)

Fig. 4 shows the profile of KE as a function of equilibrium temper-ature, adjusted from 5 to 35 ◦C. An increase in temperature slightlyfavors the extraction of lipase for the salt-rich phase. The parti-tioning of lipase from the THF-rich phase to the salt-rich phaseis spontaneous, as shown by the calculated negative value for�G◦

m (−1.67 KJ/mol) and endothermic (�H◦m = 10.80 KJ/mol). The

main forces that govern the migration of biomolecules are entropic(�S◦

m = 43.51 J/mol K), since T × �S◦m > �H◦

m.In summary, the optimal condition for partitioning of the BCL

was obtained in ATPS composed of THF at 20 wt% and potassiumphosphate buffer at 20 wt%, at an equilibrium temperature of 25 ◦C.Therefore, these conditions were chosen for the purification oflipase from fermented broth.

1464 R.L. Souza et al. / Process Biochemistry 50 (2015) 1459–1467

Fc

3

ca1twpTcpT

3

lesodvepmic[Tt

ieaa

ta

Fig. 5. Pareto chart for the effects of THF concentration (X1) and [K2HPO4/KH2PO4]

TPs

ig. 4. Effect of temperature on partition coefficient (KE) of lipase from Burkholderiaepacia for the ATPS based on THF + [K2HPO4/KH2PO4].

.4. Production and pre-purification of lipase

Before assessing the ability of this new ATPS in purifying extra-ellular lipase from Bacillus sp. ITP-001, the steps of productionnd pre-purification must be considered. The fermentation was44 h long, then was applied the salt (NH4)2SO4 for the precipita-ion of lipase, followed by a dialysis step to remove low moleculareight compounds, including inorganic salts of the precipitationrocess, all this process is described in detail elsewhere [7,35].able 4 reports the enzymatic activity (EA – U mL−1), total proteinoncentration (C – mg mL−1), specific activity (SA – U mg−1) andurification factor (PF – fold) in the fermented broth and dialyzed.he purification factor found on the dialysate was 12.7 ± 0.2 fold.

.5. Purification of lipases using ATPS

After the production and pre-purification step of extracellu-ar lipase from Bacillus sp. ITP-001, the purification was assessedmploying representative conditions of this new ATPS previouslytudied. The extraction systems were prepared by adding 20 wt%f THF + 20 wt% of potassium phosphate buffer (pH 7) + 60 wt% ofialysate solution containing the lipolytic enzyme produced. In pre-ious studies performed by some of us the optimum pH for thextracellular lipase from Bacillus sp. ITP-001 was established to beH 7 [48]. A buffer was then used in all experiments, because itade possible to keep the pH constant during the entire exper-

ments, allowing to keep the enzyme stable and at its optimalonditions, thus preventing its denaturation or loss of activity13,49]. The systems were held at equilibrium for 18 h at 25 ◦C.hese optimized conditions were based on the partitioning of BCLo the bottom phase (rich in salt).

Additionally, a surface response analysis by central compos-te rotational design (CCRD) was carried to evaluate the optimalxtraction conditions. The interactions between the biomoleculesnd the phases of the system, discussed in Section 3.3, will not be

gain here discussed.

The Pareto charts were used to evaluate the weight of contribu-ion of each factor to the response and the possible cross-effectmong these variables. The bar lengths in these charts are

able 4urification factor, enzymatic activity, specific activity, and protein concentration at the ep. ITP-001.

Steps Process EA (U mL−1)

Production Fermentation 6167.3

Pre-purification Dialyse 6135.4

Purification ATPS 36,210.4

concentration (X2) in the partitioning of lipase (i) and recovered in the bottom phase(ii) by ATPS.

proportional to the absolute value of the estimated effects and thevertical line represents the 95% confidence interval. Therefore, thefactors having a significant effect on the response are those beyondthis line. The Pareto charts of two parameters of extraction (parti-tion coefficient – KE [Fig. 5 (i)] and recovery of enzyme in bottomphase – REB [Fig. 5 (ii)]) were similar and showed that only of con-centration of salt solution (wt%, KH2PO4/K2HPO4) was found to besignificant to the control of the analyzed parameters followed bythe concentration of THF (Fig. 5).

Fig. 6 (i) and (ii) shows the response surface charts for theextraction parameters (partition coefficient and recovery of theenzyme, respectively). Through of analysis these graphs, it is pos-sible to relate them with the data previously presented in Section3.3, by which a commercial lipase was used as a model for study-ing the partitioning in ATPS. The data shows a good agreementbetween the two approaches, the optimal partitioning of the

lipase being obtained for 20 wt% of THF and 20 wt% of salt ATPSwith an KE = 0.10 ± 0.02 (for the salt-rich phase), as well as theREB = 96.4 ± 0.5.

nd of each step of the production and pre-purification of lipase produced by Bacillus

C (mg mL−1) SA (U mg−1) PF (fold)

1.15 5,365.7 –0.09 68,171.1 12.7 ± 0.20.06 557,846.9 103.9 ± 0.9

R.L. Souza et al. / Process Biochemistry 50 (2015) 1459–1467 1465

F 2HPOi

mBtttapoa(bt(iooiip

i2stApAtotp

ppspbl(o

ep

EBwhich migrate for the THF-rich phase (RPT = 53.2 ± 1.3) by follow-ing the partition coefficients of the enzyme (KE = 0.045 ± 0.004) andprotein contaminants (KP = 4.06 ± 0.24), thus confirming the same

ig. 6. Response surface plot showing the effect of THF concentration (wt%) and [Kn the bottom phase (REB).

The proposed application of this ATPS revealed a good perfor-ance in the purification of the lipolytic lipase produced from

acillus sp. ITP-001. The data suggest that the purification factor ofhe enzyme was increased from 12.7 to 103.9 ± 0.9 fold, comparinghe steps of pre-purification (by use of dialysis) with the purifica-ion step using ATPS (Table 4). The increase of the purification factorchieved by the use of the ATPS is related to the selectivity of thehases constituting the system, resulting mainly from the removalf the contaminants that act as inhibitors [50]. The results show

lower selectivity of contaminating proteins to the bottom phaseS = 1.69), compared with the enzyme (S = 20.0). This is probablyecause the enzymes are almost completely recovered in the bot-om phase (REB = 96.4 ± 0.5), due to their very low isoelectric pointpI = 3.0) [7] and negatively charged at pH 7.0 [35,37], which resultsn the increase of its hydrophilic character, creating a higher affinityf the enzyme for the salt-rich phase. Furthermore, the migrationf enzymes to bottom phase (KE = 0.10 ± 0.02) and the contaminat-ng proteins to the top phase (KP = 1.45 ± 0.18) indicate an increasen the specific activity of the enzyme in the salt-rich phase (bottomhase), increasing the purification factor.

Previous studies by us, focused on the purification of enzymes,ncluding lipase from Bacillus sp. ITP-001 using IL/salt ATPS,5 wt% of [C8mim]Cl and 30 wt% of phosphate buffer potas-ium (pH 7), showed lower purification values (PF = 51 ± 2 fold)han those obtained here [35]. However, the use of conventionalTPS (polyethylene glycol – PEG 8000 g mol−1 + phosphate bufferotassium) show higher values purification (PF = 201.53 fold) [7].lthough the system here reported shows a lower performance

han the ATPS of PEG/salt, it is less expensive and the viscosityf the phases is reduced when compared with the PEG/salt sys-em, enhancing the mass transfer, and simplifying the fluid flowroblems when one considers the scaling up of the process.

To support our interpretation of the results concerning theurification capacity of the ATPS based in THF/potassium phos-hate buffer, an electrophoresis analysis was performed usingamples of the bottom phase (system 20 wt% of THF + 20 wt% ofotassium phosphate buffer at pH 7 – considered in this work theest purification system) and crude fermentation broth. The three

anes shown in Fig. 7 correspond to the molecular mass standardlane P), the crude fermentation broth (Lane 1) and bottom phase

btained from the THF-based ATPS (Lane 2).

The presence of multiple light bands in Lane 1 confirms the pres-nce of some contaminant proteins. In Lane 2, it is possible to see theresence of the target enzyme with a molecular weight of around

4/KH2PO4] concentration (wt%): (i) partitioning of lipase (KE); (ii) lipase recovered

54 kDa [7,35] (here abbreviated as Enz) and the presence of fewother protein bands (contaminants) which were not completelyremoved. The results from the electrophoresis are consistent withpurification factor values reported.

Furthermore, in order to prove the potential of this new ATPSwas also evaluated the purification of a commercial lipase fromCandida antarctica B recombinant in Aspergillus niger (CALB L)using the optimized conditions. The results demonstrate that therecovery of the CALB L was to the bottom phase, the salt-richphase (R = 98.8 ± 0.1), in opposition to the contaminating proteins

Fig. 7. SDS-PAGE analysis of purified lipase from Bacillus sp. ITP-001. The purity ofpartitioned lipase was assessed by 12% acrylamide gel stained with silver nitratesolution. The molecular weights of the standard protein marker ranged between21.5 and 97.4 kDa. Lane P: protein molecular markers; Lane 1: fermented broth;Lane 2: bottom phase obtained from the THF-based ATPS.

1 chem

pTrmbansn

4

psbatdbcaaebtsRfmB9ttsccf

A

CtfCM

A

t

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

466 R.L. Souza et al. / Process Bio

artitioning profile shown for the other lipases tested in this work.he purification factor for CALB L was found to 4.84 ± 0.24. Theesults seem to corroborate the data previously found in the opti-ization step by using the commercial BCL and lipase produced

y Bacillus sp. ITP-001, despite the fact that the purification resultschieved for the lipase produced by Bacillus sp. are much more sig-ificant due to the completely different level of purity of the lipaseamples (the fermentation broth has a higher content of contami-ant proteins).

. Conclusions

The formation of ATPS composed of THF and potassium phos-hate buffer (pH 7) was studied to analyze the ability of this newystem to purify the lipase produced in a submerged fermentationy the Bacillus sp. ITP-001, by starting with an optimization study on

model lipase produced by Burkholderia cepacia, and then applyinghe best systems to the purification of an extracellular lipase pro-uced by submerged fermentation by Bacillus sp. The experimentalinodal data was successfully correlated with the empirical Mer-huck and Hu equations, in addition, the binodal curve, tie-linesnd critical point were obtained and serve as support for futurepplications of this system. The optimal extraction conditions werestablished to be 20 wt% of THF and 20 wt% of potassium phosphateuffer (pH 7) through an 22 analysis of response surface by cen-ral composite rotational design (CCRD). The enzyme partitioning ispontaneous and governed by entropic effects. A PF of 103.9 ± 0.9,EB of 96.4 ± 0.5 and KE = 0.10 ± 0.2 were achieved for the lipaserom Bacillus sp. ITP-001 using this recovery process from a fer-

entation broth. For the commercial lipase by Candida antarctica recombinant in Aspergillus niger a PF of 4.84 ± 0.24 and a REB of8.8 ± 0.1 were found. The THF/salt ATPS proved thus to be effec-ive for the purification of solvent tolerant lipase and it was shownhat the lipase’s enzymatic activity was not affected by the organicolvent. Because of the ease of organic solvent recovery, lower vis-osity and the cost-effectiveness of the process, the THF/salt ATPSould be potentially developed as a commercial recovery processor lipase derived from microbial sources.

cknowledgments

The authors acknowledge the financial support of FAPITEC/SE,APES and CNPq. The authors also thank financial support throughhe doctoral grant R.L. Souza and PROBIC/UNIT for the scholarshipor R.A. Lima. J.A.P. Coutinho acknowledges the Fundac ão para aiência e a Tecnologia for the funding of CICECO-Aveiro Institute ofaterials (Ref. FCT UID/CTM/50011/2013).

ppendix A. Supplementary data

Supplementary data associated with this article can be found, inhe online version, at doi:10.1016/j.procbio.2015.05.015

eferences

[1] R. Sharma, Y. Chisti, U.C. Banerjee, Production, purification, characterization,and applications of lipases, Biotechnol Adv 19 (2001) 627–662.

[2] K.-E. Jaeger, T. Eggert, Lipases for biotechnology, Curr Opin Biotechnol 13(2002) 390–397.

[3] F. Hasan, A.A. Shah, A. Hameed, Industrial applications of microbial lipases,Enzyme Microb Technol 39 (2006) 235–251.

[4] R.K. Saxena, W.S. Davidson, A. Sheoran, B. Giri, Purification andcharacterization of an alkaline thermostable lipase from Aspergillus carneus,Process Biochem 39 (2003) 239–247.

[5] A. Houde, A. Kademi, D. Leblanc, Lipases and their industrial applications,Appl Biochem Biotechnol 118 (2004) 155–170.

[6] A.S. Lima, R.M. Alegre, A.J.A. Meirelles, Partitioning of pectinolytic enzymes inpolyethylene glycol/potassium phosphate aqueous two-phase systems,Carbohydr Polym 50 (2002) 63–68.

[

istry 50 (2015) 1459–1467

[7] J.M.P. Barbosa, R.L. Souza, A.T. Fricks, G.M. Zanin, C.M.F. Soares, A.S. Lima,Purification of lipase produced by a new source of Bacillus in submergedfermentation using an aqueous two-phase system, J Chromatogr B 879 (2011)3853–3858.

[8] R.L. Souza, J.M.P. Barbosa, G.M. Zanin, M.W.N. Lobao, C.M.F. Soares, A.S. Lima,Partitioning of porcine pancreatic lipase in a two-phase systems ofpolyethylene glycol/potassium phosphate aqueous, Appl Biochem Biotechnol161 (2010) 288–300.

[9] P.A.J. Rosa, A.M. Azevedo, S. Sommerfeld, W. Backer, M.R. Aires-Barros,Aqueous two-phase extraction as a platform in the biomanufacturingindustry: economical and environmental sustainability, Biotechnol Adv 29(2011) 559–567.

10] B.Y. Zaslavsky, Aqueous two-phase partitioning, physical chemistry andbioanalytical application, Marcell Dekker, New York, 1995.

11] M. Rito-Palomares, Practical application of aqueous two-phase partition toprocess development for the recovery of biological products, J Chromatogr B807 (2004) 3–11.

12] J.A. Asenjo, B.A. Andrews, Aqueous two-phase systems for protein separation:phase separation and applications, J Chromatogr A 1238 (2012) 1–10.

13] R.L. Souza, V.C. Campos, S.P.M. Ventura, C.M.F. Soares, J.A.P. Coutinho, A.S.Lima, Effect of ionic liquids as adjuvants on PEG-based ABS formation and theextraction of two probe dyes, Fluid Phase Equilib 375 (2014) 30–36.

14] R.L. Souza, S.P.M. Ventura, C.M.F. Soares, J.A.P. Coutinho, A.S. Lima, Lipasepurification using ionic liquids as adjuvants in aqueous two-phase systems,Green Chem 17 (2015) 3026–3034.

15] C.W. Ooi, B.T. Tey, S.L. Hii, S. Mazlina, M. Kamal, J.C.W. Lan, A. Ariff, T.C. Ling,Purification of lipase derived from Burkholderia pseudomallei withalcohol/salt-based aqueous two-phase systems, Process Biochem 44 (2009)1083–1087.

16] Tan Z-j, Li F-f, X-l. Xu, Extraction and purification of anthraquinonesderivatives from Aloe vera L. using alcohol/salt aqueous two-phase system,Bioproc Biosyst Eng 36 (2013) 1105–1113.

17] I.A.O. Reis, S.B. Santos, F.D.S. Pereira, C.R.S. Sobral, M.G. Freire, L.S. Freitas,C.M.F. Soares, Á.S. Lima, Extraction and recovery of rutin from acerola wasteusing alcohol-salt-based aqueous two-phase systems, Sep Sci Technol 49(2013) 656–663.

18] G.B. Cardoso, T. Mourão, F.M. Pereira, M.G. Freire, A.T. Fricks, C.M.F. Soares,Á.S. Lima, Aqueous two-phase systems based on acetonitrile andcarbohydrates and their application to the extraction of vanillin, Sep PurifTechnol 104 (2013) 106–113.

19] P.B. Dhamole, P. Mahajan, H. Feng, Phase separation conditions forsugaring-out in acetonitrile–water systems, J Chem Eng Data 55 (2010)3803–3806.

20] G.B. Cardoso, I.N. Souza, T. Mourão, M.G. Freire, C.M.F. Soares, A.S. Lima, Novelaqueous two-phase systems composed of acetonitrile and polyols: phasediagrams and extractive performance, Sep Purif Technol 124 (2014) 54–60.

21] M.G. Freire, A.F.M. Claudio, J.M.M. Araujo, J.A.P. Coutinho, I.M. Marrucho, J.N.C.Lopes, L.P.N. Rebelo, Aqueous biphasic systems: a boost brought about byusing ionic liquids, Chem Soc Rev 41 (2012) 4966–4995.

22] J.H. Santos, F.A. Silva, S.P.M. Ventura, J.A.P. Coutinho, R.L. Souza, C.M.F. Soares,A.S. Lima, Ionic liquid-based aqueous biphasic systems as a versatile tool forthe recovery of antioxidant compounds, Biotechnol Progr 31 (2015) 70–77.

23] Z. Li, H. Teng, Z. Xiu, Extraction of 1,3-propanediol from glycerol-basedfermentation broths with methanol/phosphate aqueous two-phase system,Process Biochem 46 (2011) 586–591.

24] T.W. Tan, Q. Huo, Q. Ling, Purification of glycyrrhizin from Glycyrrhizauralensis Fisch with ethanol/phosphate aqueous two phase system, BiotechnolLett 24 (2002) 1417–1420.

25] W. Zhi, Q. Deng, Purification of salvianolic acid B from the crude extract ofSalvia miltiorrhiza with hydrophilic organic/salt-containing aqueoustwo-phase system by counter-current chromatography, J Chromatogr A 1116(2006) 149–152.

26] Q. Su, K.G. Rowley, N.D.H. Balazs, Carotenoids: separation methods applicableto biological samples, J Chromatogr B 781 (2002) 393–418.

27] M. Taha, L. Khoiroh, M. Lee, Phase behavior and molecular dynamicssimulation studies of new aqueous two-phase separation systems induced byHEPES buffer, J Phys Chem B 117 (2012) 563–582.

28] N. Hirayama, T. Higo, H. Imura, Salting-out phase separation system ofwater–tetrahydrofuran with co-using 1-butyl-3-methylimidazolium chlorideand sodium chloride for possible extraction separation of chloro-complexes,Solvent Extr Res Dev 21 (2014) 71–76.

29] J.C. Merchuk, B.A. Andrews, J.A. Asenjo, Aqueous two-phase systems forprotein separation: studies on phase inversion, J Chromatogr B 711 (1998)285–293.

30] M. Hu, Q. Zhai, Z. Liu, S. Xia, Liquid–liquid and solid–liquid equilibrium of theternary system ethanol + cesium sulfate + water at (10, 30, and 50) ◦C, J ChemEng Data 48 (2003) 1561–1564.

31] I.C. Feitosa, J.M.D. Barbosa, S.C. Orellana, A.S. Lima, C.M.F. Soares, Lipaseproduction by bacterial isolates from petroleum contaminated soil, ActaSci-Technol 32 (2010) 27–31.

32] C.M.F. Soares, H.F. De Castro, F.F. De Moraes, G.M. Zanin, Characterization and

utilization of Candida rugosa lipase immobilized on controlled pore silica,Appl Biochem Biotechnol 77–79 (1999) 745–757.

33] M.M. Bradford, Rapid and sensitive method for quantitation of microgramquantities of protein utilizing principle of protein–dye binding, Anal Biochem72 (1976) 248–254.

chem

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

R.L. Souza et al. / Process Bio

34] U.K. Laemmli, Cleavage of structural proteins during assembly of head ofbacteriophage-T4, Nature 227 (1970) 680–685.

35] S.P.M. Ventura, R.L.F. de Barros, J.M.D. Barbosa, C.M.F. Soares, A.S. Lima, J.A.P.Coutinho, Production and purification of an extracellular lipolytic enzymeusing ionic liquid-based aqueous two-phase systems, Green Chem 14 (2012)734–740.

36] F.M. Hofmeister, M. Yalon, S. Iida, J. Stacholy, E.P. Goldberg, Evaluation of thetissue-protective properties of hydrophilic surface modified intraocular-lensimplants, Abs Pap Am Chem Soc 196 (1988) 22.

37] G. Bassani, B. Farruggia, B. Nerli, D. Romanini, G. Picó, Porcine pancreaticlipase partition in potassium phosphate–polyethylene glycol aqueoustwo-phase systems, J Chromatogr B 859 (2007) 222–228.

38] Y.J. Zhou, C.L. Hu, N. Wang, W.W. Zhang, X.Q. Yu, Purification of porcinepancreatic lipase by aqueous two-phase systems of polyethylene glycol andpotassium phosphate, J Chromatogr B 926 (2013) 77–82.

39] H. Katayama, K. Sugahara, Liquid–liquid phase equilibria of the systemethanol (1) + water (2) + tripotassium citrate (3), J Chem Eng Data 53 (2008)1940–1943.

40] Y. Wang, S. Hu, J. Han, Y. Yan, Measurement and correlation of phase diagramdata for several hydrophilic alcohol + citrate aqueous two-phase systems at298.15 K, J Chem Eng Data 55 (2010) 4574–4579.

41] C. Hansch, A. Leo, D. Hoekman, Exploring QSAR – hydrophobic, electronic, andsteric constants, American Chemical Society, Washington, DC, 1995.

42] M.L. Rúa, T. Díaz-Maurino, V.M. Fernández, C. Otero, A. Ballesteros,Purification and characterization of two distinct lipases from Candidacylindracea, Biochim Biophys Acta 1156 (1993) 181–189.

[

istry 50 (2015) 1459–1467 1467

43] P. Mander, S.S. Cho, J.R. Simkhada, Y.H. Choi, D.J. Park, J.C. Yoo, An organicsolvent–tolerant lipase from Streptomyces sp. CS133 for enzymatictransesterification of vegetable oils in organic media, Process Biochem 47(2012) 635–642.

44] J. Yang, D. Guo, Y. Yan, Cloning, expression and characterization of a novelthermal stable and short-chain alcohol tolerant lipase from Burkholderiacepacia strain G63, J Mol Catal B-Enzym 45 (2007) 91–96.

45] B.R. Babu, N.K. Rastogi, K.S.M.S. Raghavarao, Liquid–liquid extraction ofbromelain and polyphenol oxidase using aqueous two-phase system, ChemEng Process 47 (2008) 83–89.

46] D. Forciniti, C.K. Hall, M.R. Kula, Protein partitioning at the isoelectric point –influence of polymer molecular-weight and concentration and protein size,Biotechnol Bioeng 38 (1991) 986–994.

47] D.D. Purkayastha, V. Madhurima, Interactions in water–THF binarymixture by contact angle. FTIR and dielectric studies, J Mol Liq 187 (2013)54–57.

48] J.M.P. Barbosa, R.L. Souza, C.M. de Melo, A.T. Fricks, C.M.F. Soares, A.S. Lima,Biochemical characterisation of lipase from a new strain of Bacillus spITP-001, Quim Nova 35 (2012) 1173–1178.

49] J.F.B. Pereira, F. Vicente, V.C. Santos-Ebinuma, J.M. Araujo, A. Pessoa, M.G.Freire, J.A.P. Coutinho, Extraction of tetracycline from fermentation broth

using aqueous two-phase systems composed of polyethylene glycol andcholinium-based salts, Process Biochem 48 (2013) 716–722.

50] S.P.M. Ventura, S.G. Sousa, M.G. Freire, L.S. Serafim, A.S. Lima, J.A.P. Coutinho,Design of ionic liquids for lipase purification, J Chromatogr B 879 (2011)2679–2687.