Membrane Technology as a Strategy for Improving β-Galactosidase Concentration Processes: The Influence of the pH, Membrane Molecular Weight, Pressure, and Ionic Strength in the Process

Abstract

The enzyme β-galactosidase catalyzes the hydrolysis of lactose into glucose and galactose, although for its effective application it is necessary to establish techniques for purification, concentration, or polishing, such as membrane separation processes, in particular ultrafiltration. The present study aimed to investigate ultrafiltration and diafiltration applied as initial steps for concentration and salt removal, respectively, in the β-galactosidase purification processes. Additionally, the influence levels of the pH (6.5, 7.7, or 7.5), membrane molecular weight cut-off (30, 50, 60, or 100 kDa), operating pressure (1.5, 2.0, or 2.5 kgf/cm²), and ionic strength of the ultrafiltration using NaCL or KCl (0.01–0.1 M) were evaluated considering the enzyme recovery, purification, retention, and concentration factors in relation to the proteins, volume, activity, and protein flux and yield of the processes. The ultrafiltration of the crude enzyme extract at pH 7.5 and 1.5 kgf/cm² with a 50 kDa polyethersulfone membrane resulted in a volume concentration of the β-galactosidase extract up to 7.1-fold greater, a purification factor 1.2-fold greater, and an enzyme recovery rate of 108.9% by eliminating metabolites during the purification process. In addition, the lowest flux variation range (16.0 to 13.1 L/m²·h) was observed under these same conditions, thereby representing a decrease of 18.0%. An increase in the operating pressure and the addition of salts results in reduced enzyme recovery (up to 38% of the process yield (734.1 to 453.7 U/h) and up to 40% of the enzyme recovery rate (108.9 to 60.6%) during the ultrafiltration using NaCl, respectively). The operation in the diafiltration mode allowed salt removal after the purification of β-galactosidase (enzymatic recovery rates above 93.4%) via precipitation and ion-exchange chromatography elution and as part of an aqueous two-phase system using 6 diafiltration cycles, thereby revealing its application potential.

1. Introduction

The enzyme β-galactosidase (E.C 3.2.1.23) catalyzes the hydrolysis of the disaccharide lactose into the monosaccharides glucose and galactose. The enzyme may act in solutions and also in different matrices rich in lactose, being applied in different stages of the processes as a fundamental way to reduce the lactose content of dairy products intended for people with lactose intolerance [1,2]. Around 70% of the adult population worldwide is lactose-intolerant, reaching 95–100% in Asians, for example. This makes it necessary for the food industry to apply technologies to reduce the lactose contents in products intended for these lactose-intolerant people in order to avoid symptoms such as diarrhea, flatulence, and abdominal cramps that appear with the consumption of lactose [3,4]. Moreover, hydrolysis with β-galactosidase contributes to the formation of galactooligosaccharides (GOS) [5,6,7], which are beneficial to humans [8] and improve the technological and sensory characteristics of foods [9,10,11]. However, for effective hydrolysis using enzymes, recovery techniques that do not compromise the enzymatic activity, and consequently the applications, viability, and economic potential, need to be applied [1].

Ultrafiltration (UF) is a technique based on the macromolecule size-exclusion mechanism and can be applied to both the concentration and polishing of biological products [12]. The concentration and separation of the biological material using membranes does not require additives or phase changes and can be performed isothermally at low temperatures and pressures, with lower energy consumption compared with the thermal processes. Furthermore, these conditions ensure minimal denaturation of bioproducts that are highly unstable and are less expensive than other methods [12,13,14]. In ultrafiltration systems, the operating conditions strongly affect the concentration process, particularly the protein size of the bioproducts [15], which should be deeply studied.

Some of the important applications of ultrafiltration techniques include fractionation and the recovery of numerous proteins [16,17,18], polysaccharides [19,20], enzymes [1,21], biosurfactants [22], oils [23,24], and bioactive peptides [25,26]. Most of the studies on β-galactosidases reported the use of only ultrafiltration when modeling the hydrolysis mechanism [27] and galactooligosaccharide synthesis [28,29], and for the removal of impurities such as salts [30] and polymers in an aqueous purification system [31]. The application of these techniques often neglects the characteristic peculiarities of the bioproduct or does not consider the interactions between molecules, which in turn compromise the efficiency of the process and hinder the scale-up procedure, resulting in the infeasibility of the process [1].

Ultrafiltration can be used as an alternative to other concentration techniques; therefore, it is considered an important process for applications in the food and pharmaceutics industries. This technique can also be applied for the removal of salts and polyethylene glycol used in the purification process when utilizated in the diafiltration mode. In this case, ultrafiltration is achieved after precipitation or separation in an aqueous two-phase system of β-galactosidase from Kluyveromyces marxianus used in the production of GOS [5] or other products of industrial interest [32].

Thus, we evaluated the influence of the pH, membrane molecular weight cut-off (MWCO), operating pressure, and ionic strength of the ultrafiltration process of β-galactosidase from K. marxianus CCT 7082. Furthermore, the ultrafiltration process was applied in the diafiltration mode as an alternative for salt removal from the β-galactosidase extract after the purification process. To the best of our knowledge, no previous study has been performed in this regard.

2. Materials and Methods

2.1. Materials

The malt extract, yeast extract, peptone, glucose, agar, ammonium sulfate ((NH₄)₂SO₄), potassium phosphate monobasic (KH₂PO₄), bovine serum albumin (BSA), and magnesium sulfate heptahydrate (MgSO₄·7H₂O) were purchased from Merck (Darmstadt, Germany). The lactose, potassium chloride (KCl), and sodium chloride (NaCl) were purchased from Synth (São Paulo, Brazil). Membranes with molecular weight cut-offs (MWCO) of 30 kDa (polyethersulfone—PBTK Millipore^® 15005), 50 kDa (polyethersulfone—PBQK Millipore^® 15005), 60 kDa (polysulfone—Osmonics^® YMEWSP3001), and 100 kDa (polyethersulfone—PBHK Millipore^® 15005) were purchased from Merck-Millipore (Darmstadt, Germany). The o-nitrophenyl-galactopyranoside used as the substrate was purchased from Sigma Aldrich (Darmstadt, Germany). All other chemicals used were of analytical grade and were used as received, without any further purification.

2.2. Microorganism, Inoculum, and Submerged Cultivation

The microorganism K. marxianus CCT 7082 was maintained at 4 °C in yeast malt (YM) agar slants containing malt extract (3 g/L), yeast extract (3 g/L), peptone (5 g/L), glucose (10 g/L), and agar (20 g/L) [33]. The inoculum was prepared in 500 mL Erlenmeyer flasks with 150 mL of culture medium containing yeast extract (1 g/L), (NH₄)₂SO₄ (1.2 g·L⁻¹), KH₂PO₄ (5 g/L), MgSO₄·7H₂O (0.4 g/L), and lactose (10 g/L) in 0.2 mol/L of potassium phosphate at pH 5.5. The culture conditions were 30 °C and 180 rpm for 24 h [34].

The inoculum was added to the cultivation medium in a proportion of 10 % (v/v). The medium contained yeast extract (17 g/L), (NH₄)₂SO₄ (8.8 g/L), KH₂PO₄ (5 g/L), MgSO₄.7H₂O (0.4 g/L), and lactose (28.2 g/L) in 0.2 M of potassium phosphate buffer at pH 6.0. The microorganism was cultivated at 30 °C and 180 rpm for 96 h [33].

2.3. Enzyme Extraction

The enzyme was extracted by adding 1.1 g of glass beads (0.6–0.8 mm diameter) for each mL of cells resuspended (40 mg dry cell/mL) in a 50 mmol/L potassium phosphate buffer at pH 6.6 [35]. The suspension was stirred in a vortex mixer for 40 min; thereafter, the cell extract was clarified via centrifugation at 4700× g and 4 °C for 10 min [36].

2.4. β-Galactosidase Concentration Using Ultrafiltration Step

2.4.1. Assessment of Menbrane Molecular Weight Cut-Off and Operating pH and Pressure

The ultrafiltration runs were carried out in a stirred dead-end cell system (designed and produced by our research group and not commercially available) as illustrated in Figure 1. The system was kept under refrigeration at 4 °C to minimize losses due to denaturation. Membranes with molecular weight cut-offs of 30, 50, 60, and 100 kDa were tested. The operating pressure and pH values varied from 1.5 to 2.5 kgf/cm² and from 6.5 to 7.5, respectively. All assays were performed in triplicate. The enzymatic activity and total protein content values in the feed, retentate, and permeate were measured after each test.

2.4.2. Ionic Strength Influence

The increases in ionic strength in the enzyme concentration and prepurification processes were evaluated by adding potassium chloride (KCl) and sodium chloride (NaCl) to the initial enzyme extract. The runs were performed in triplicate with the following concentrations: 0.01, 0.05, and 0.1 mol/L. The system was operated at 1.5 kgf/cm², pH 7.5, and 4 °C, using a UF membrane of 50 kDa.

2.4.3. Diafiltration

The diafiltration was carried out using the same ultrafiltration apparatus to remove the ammonium sulfate, sodium chloride, and potassium phosphate salts homogenized with the enzyme extract obtained from published and recommended protocols for precipitation and ion-exchange chromatography elution and as part of the composition of the aqueous two-phase system (ATPS), respectively [1,37,38]. The enzymatic extracts were obtained after reproducing the protocols and the purified enzymatic extract, which was rich in salt and used in the diafiltration process. The complete removal of the salts was performed using six diafiltration cycles with two volumes of diafiltration solution per cycle (50 mmol/L potassium phosphate buffer) at a suitable pH for each case.

2.5. Ultrafiltration Performance

The performance in the enzyme recovery and concentration processes was evaluated using the recovery (Equation (1)), purification (Equation (2)), retention and concentration factors (retained proteins (Equation (3)), volume concentration factor (Equation (4)), activity (Equation (5)) and protein concentration factors (Equation (6)), flux (Equation (7)), and yield of the processes (Equation (8)). The uncertainties of the measured values were, on average, below 5%:

$$\mathrm{PF}=\frac{{\mathrm{A}}_{\mathrm{sf}}}{{\mathrm{A}}_{\mathrm{si}}}$$

(1)

where A_sf and A_si are the respective specific activities (U/mg) of the final and initial (crude extract) fractions. The specific activity is defined as the ratio between the activity (U/mg) and the total protein concentration (mg/mg) in the same sample:

$$\mathrm{REC} \left(\%\right)=\frac{{\mathrm{A}}_{\mathrm{f}}{.\mathrm{V}}_{\mathrm{f}}}{{\mathrm{A}}_{\mathrm{i}}{.\mathrm{V}}_{\mathrm{i}}}. 100$$

(2)

where A_f and A_i are the respective enzymatic activities (U/mL) of the final and initial (crude extract) fractions; V_f and V_i are the volumes (mL) of the final and initial (crude extract) fractions, respectively:

$$\mathrm{RP}(\%)=\left(1-\frac{{\mathrm{P}}_{\mathrm{cp}}}{{\mathrm{P}}_{\mathrm{cf}}}\right). 100$$

(3)

where P_cp and P_cf are the protein concentrations (mg/mL) in the permeate and feed, respectively:

$$\mathrm{Fc} \left(\mathrm{V}\right)=\frac{{\mathrm{V}}_{\mathrm{f}}}{{\mathrm{V}}_{\mathrm{r}}}. 100$$

(4)

where V_f and V_r are the volume (mL) fed into the ultrafiltration system and the volume remaining after permeation:

$$\mathrm{Fc} \left(\mathrm{A}\right)=\frac{{\mathrm{A}}_{\mathrm{f}}}{{\mathrm{A}}_{\mathrm{r}}}. 100$$

(5)

where A_f and A_r are the activity levels(U/mL) at the beginning of the process and the end of the process:

$$\mathrm{Fc} \left(\mathrm{P}\right)=\frac{{\mathrm{P}}_{\mathrm{f}}}{{\mathrm{P}}_{\mathrm{r}}}. 100$$

(6)

where P_f and P_r are the protein concentrations (mg/mL) at the beginning of the process and the end of the process:

$$\mathrm{Flux}=\frac{{\mathrm{V}}_{\mathrm{p}}}{\mathrm{A} . t}$$

(7)

where V_p is the permeate volume (L), A is the membrane area (m²), and t is the time (h):

$$\mathrm{Yield}=\frac{{\mathrm{A}}_{\mathrm{t}}}{t}$$

(8)

where A_t is the total activity (U) obtained in the final process and t is the total time of the process (h).

2.6. Analytical Methods

Enzyme Activity and Total Protein

The enzyme activity was determined using o-nitrophenyl-galactopyranoside as the substrate (ONPG, Aldrich), according to the method proposed by Inchaurrondo, et al. [39]. One unit of enzyme activity (U) is defined as the amount of enzyme necessary to yield 1 μmol of o-nitrophenol per minute at 37 °C and pH 6.6. The total protein content was assayed using the method proposed by Bradford [40], using bovine serum albumin (BSA) as the standard.

2.7. Statistical Analysis

All analyses were performed in triplicate. The results are presented as means ± standard deviations from the mean. The results were analyzed using a unilateral analysis of variance (ANOVA), and Tukey’s test (p < 0.05) was used to determine differences between means using Statistica 7.0 for Windows (Statsoft Inc., Tulsa, OK, USA).

3. Results and Discussion

3.1. β-Galactosidase Recovery and Concentration Using Ultrafiltration

The use of enzymatic extracts at various pH values and membranes with different molecular weight cut-offs (MWCO) showed no significant changes (p < 0.05) in the parameters evaluated during the β-galactosidase ultrafiltration process (

Table 1. Effects of membrane molecular weight cut-offs on β-galactosidase concentrations at pH 6.5, 7.0, and 7.5.

pH	MWCO (kDa)	%R A	PF B	RP(%) C	Fc (V) D	Fc (A) E	Fc(P) F
6.5	30	93.0 ± 0.01 b	0.93 ± 0.03 cd	98.5 ± 1.16 ab	6.7 ± 0.0 ab	5.79 ± 0.68 ab	6.31 ± 0.84 a
	50	79.5 ± 4.8 b	0.89 ± 0.06 d	99.4 ± 0.66 ab	6.4 ± 0.26 b	5.0 ± 0.3 b	5.67 ± 0.05 ab
	60	89.3 ± 2.9 ab	0.95 ± 0.09 cd	99.0 ± 0.96 ab	6.4 ± 0.5 b	5.7 ± 0.30 ab	5.93 ± 0.58 ab
	100	96.2 ± 2.2 ab	1.08 ± 0.08 abcd	99.2 ± 0.96 ab	7.2 ± 0.5 ab	6.9 ± 0.66 a	6.16 ± 0.53 ab
7.0	30	98.1 ± 1.4 ab	1.09 ± 0.02 abcd	99.9 ± 0.16 a	7.6 ± 0.64 ab	7.03 ± 0.64 a	6.44 ± 0.59 a
	50	96.4 ± 2.7 ab	1.2 ± 0.08 ab	99.7 ± 0.12 ab	6.8 ± 0.67 ab	6.54 ± 0.73 a	5.42 ± 0.54 ab
	60	96.3 ± 2.7 ab	1.03 ± 0.11 bcd	99.6 ± 0.13 ab	6.3 ± 0.36 b	6.03 ± 0.36 ab	5.83 ± 0.7 ab
	100	99.5 ± 0.3 ab	1.19 ± 0.06 ab	99.0 ± 0.42 ab	6.4 ± 0.36 b	6.42 ± 0.46 ab	5.38 ± 0.24 ab
7.5	30	86.0 ± 6.2 ab	1.14 ± 0.03 abc	99.1 ± 0.64 b	8.1 ± 0.25 a	6.97 ± 0.36 a	6.12 ± 0.46 ab
	50	108.9 ± 4.2 a	1.22 ± 0.06 ab	96.9 ± 2.7 ab	6.6 ± 0.18 ab	7.15 ± 0.41 a	5.88 ± 0.06 ab
	60	96.4 ± 2.3 ab	1.28 ± 0.14 ab	98.9 ± 0.3 ab	6.8 ± 0.78 ab	6.17 ± 0.39 ab	4.82 ± 0.38 b
	100	97.9 ± 1.76 ab	1.05 ± 0.04 bcd	98.9 ± 0.45 ab	7.1 ± 1.0 ab	6.14 ± 0.39 ab	5.75 ± 0.27 ab

^A β-Galactosidase recovery; ^B purification factor; ^C retained proteins; ^D volume concentration factor; ^E activity concentration factor; ^F protein concentration factor. Data followed by the same lowercase letters in an individual column are not significantly different (p > 0.05) from the same parameter.

). All runs, except for the 30 and 50 kDa membranes at pH 6.5, presented purification factors >1. These low purification factors might be due to the similar sizes of the contaminating proteins and the target enzyme. Another hypothesis for the low purification factor found in the present work is the retention caused by the dynamic membrane formed by the proteins retained on the membrane surface layer (gel layer) [41], which contributes to an increase in filtration resistance, rapid flow reduction, and the shortening of the operational cycles, affecting the process viability [42].

The runs using the 50 kDa membrane at pH 7.5 exhibited the highest enzyme recovery rate (108.9%). This behavior can be explained by the presence of metabolites during the purification process, which can inhibit the enzyme activity. When the contaminants are removed from the medium, the enzyme may be activated [43,44,45].

All membranes tested, regardless of the pH, were able to fully retain the enzyme, while the enzymatic activity was not verified in the permeate flux. The high recovery rates of enzymatic activity (in relation to the total activity at the beginning of the process) ranged from 79.5 to 108.9%. The volume, activity, and protein concentration factor values were up to 8.1, 7.1, and 6.4-fold greater, respectively.

These results demonstrated the effectiveness of the ultrafiltration steps for the recovery and concentration of β-galactosidase. Furthermore, the ultrafiltration reduced the working volume, which is a crucial factor in the purification processes [1].

Moreover, the permeate flux was evaluated during the ultrafiltration process, since a low flux can lead to a lengthy and expensive process, making it one of the factors limiting deployment [12]. Figure 2A–C shows the permeate flux variation during ultrafiltration at pH 6.5, 7.0, and 7.5, respectively.

In the present work, a decrease in flux during permeation was observed in all conditions, which is considered a typical behavior of concentration polarization phenomena and fouling on the membrane surface. When the solution is permeated, the protein concentration at the membrane surface increases, thereby elevating the polarized layer. Depending on the situation, a gel layer may be formed, which in turn leads to a further decrease in the permeate flux [15].

At pH 7.5, low variation in the initial permeate flux as a function of the molar mass cut-off was observed, with the values ranging from 16.0 to 13.1 L/m²·h, representing a decrease of 18.0%. Throughout the process, the highest variation observed was from 16.7 to 12.1 L/m².h, representing a reduction of 27.5%, with a maximum operation time of 100 min.

The lowest initial permeate flux, except for the 100 kDa membrane, occurred at pH 6.5 (range of 12.2 to 11.1 L/m²·h). The largest decrease in permeate flux was observed in the 60 kDa membrane (11.0 to 5.0 L/m²·h), representing a decrease of 54.5% over the process and a maximum operating time of 130 min. The low permeate flux at this pH may have been due to protein agglomeration and the production of a deposit on the membrane surface, thereby reducing the membrane permeability. Furthermore, the protein adsorption on the membrane surface may result in a second membrane that alters the function of the original membrane [46,47,48].

Although the 60 kDa membrane presented similar results to the other membranes regarding the purification factor, this membrane should be analyzed and used with caution due to the process presenting a longer operating time and a lower final permeate flux when compared to the other membranes, affecting the yield and viability of the process. This reduction in the yield of the process can lead to significant increases in the concentration and purification costs of this enzyme and can even make it unfeasible at the industrial scale. This effect may have occurred due to greater fouling of the membrane, as a larger pore diameter favors the permeation of compounds similar in size to the membrane pores, leading to a greater occurrence of blockage of the pores [12]. In addition, the differences in membrane material can also affect the separation performance during ultrafiltration, since the chemical composition can affect the functionality, charge, and interactions with the material, among other factors [49].

At pH 7.0, the average flux values ranged from 21.2 to 12.0 L/m²·h, with a reduction of 43.4% at a maximum operating time of 80 min. Although the application of this pH led to the highest initial flow, a high flux reduction rate was observed during the ultrafiltration process. The permeate flux variation is a phenomenon inherent in the process and is observed in several processes using ultrafiltration. Cassini, et al. [50] studied three ultrafiltration membranes used for recovering soy protein isolates. These authors reported reductions in the permeate flux by 30%, 69%, and 30% for the 5, 20, and 50 kDa membranes, respectively. Similarly, Chollangi and Hossain [51] reported fluxes of 10 and 125 L/m²·h for the 5 and 10 kDa membranes, respectively, using a pressure of 3.0 kgf/cm² in the ultrafiltration of proteins and lactose from dairy effluents. These authors observed a decrease in the permeate flux, indicating that the protein increased fouling on the membrane and thereby affected the permeate flux.

The evaluation of all process factors—including the flow and its variations throughout the process, the enzymatic recovery and purification values, as well as the concentration and retention parameters for the protein and enzymatic activity—is necessary to define the conditions for carrying out ultrafiltration [1,21].

Based on the previous results, the effect of the operating pressure was assessed using the 50 kDa membrane at pH 7.5, which presented higher purification factor and enzyme recovery values and a slight decrease in the permeate flux during β-galactosidase ultrafiltration. Figure 3 presents these results.

After comparing the initial flux with the final flux in the process at different pressures (Figure 3), decreases of approximately 3.5%, 38.4%, and 47.6% were observed, respectively, indicating that the increase in pressure causes a significant decrease in permeate flux during the process. In the runs operated at low pressures, the concentration was gradual, leading to a slight drop in the permeate flux, a consequent reduction in the operating time, and an increase in the process yield (734.1 U/h) when compared to those at higher pressures (484.0 and 453.7 U/h for pressures of 2.0 and 2.5 kgf/cm², respectively). Additionally, a comparison between pressures of 2.0 and 2.5 kgf/cm² also showed the effect of higher pressures on the membrane concentration process, where an inversion of permeation efficiency was observed over time, especially at 25 min. Despite initially presenting a higher flux, the pressure of 2.5 kgf/cm² resulted in the faster concentration of the proteins (see the description of the effects below), altering the permeation capacity of the membrane, which made the process less efficient over time when compared to the process using a pressure of 2.0 kgf/cm², for example.

After 35 min, no difference was observed between the pressures tested, and from this point onwards the processes seem equivalent. However, it is important to consider the advantages that reduce the operating costs and make the process viable, such as the constant filtration rate and low flow variation, the permeation of smaller particles present in the material, the membrane deformation, as well as consumption of gas used in the membrane permeation process [12,52,53].

The reduced recovery and purification factors of the enzyme in the systems operating with higher pressures may be associated with greater flux reductions caused by the membrane deformation (related to the compaction that occurs depending on the time of use, the constant conditions of high pressure, and the material used, which influences the permeability and increases the filtration resistance), concentration polarization, and fouling due to the applied pressure, leading to the inefficiency of the process; however, the occurrence of these phenomena is inevitable, can harm the performance of the membrane, and influence the process viability [54,55].

The decrease in resistance with increasing pressure is generally not observed in permeation experiments. In general, the opposite contradictory trend is observed; that is, the higher pressure results in the increased formation of the second layer on the membrane, thereby reducing the permeability [56,57].

Although working at reduced pressures with lower flux may seem inevitable, the results indicate that under less polarized conditions, the fouling is lower and the permeate flux is less affected during the process.

Sodium and potassium chloride were added to the enzymatic extract of β-galactosidase, as these salts can change the ionic strength of the solution [58], act as protective agents of the enzyme [59], and modify the permeability during the separation of some compounds [58,60,61]. However, the addition of NaCl and KCl in the β-galactosidase extract revealed no positive effect compared to the control test (without added salt) (

Table 2. Effects of ionic strength on the enzyme concentration and purification using a 50 kDa membrane at pH 7.5 and 1.5 kgf/cm² of operating pressure.

Salt	Concentration (mol.L)	%R	FC(A)	PF
Control	-	108.9 ± 4.2 a	7.15 ± 0.4 a	1.22 ± 0.06 a
KCl	0.01	92.4 ± 4.1 b	4.6 ± 0.7 b	0.9 ± 0.05 b
	0.05	88.3 ± 7.3 bc	5.9 ± 0.1 abc	0.8 ± 0.1 b
	0.1	73.9 ± 4.7 cd	3.2 ± 0.6 c	0.7 ± 0.09 b
NaCl	0.01	85.9 ± 3.9 bc	4.3 ± 0.4 c	0.8 ± 0.08 b
	0.05	63.7 ± 6.1 d	3.1 ± 0.7 c	0.7 ± 0.1 b
	0.1	60.2 ± 5.7 d	3.4 ± 0.3 c	0.7 ± 0.06 b

Data followed by the same lowercase letters in one individual column are not significantly different (p > 0.05) to the same parameter.

) and reduced the activity, purity, concentration, and enzyme recovery. The retention of the activity and protein levels did not change when compared to the control assay.

The addition of salt to the extract did not cause migration of the enzyme or increased migration of contaminants to permeate proteins (data not shown). Although the protective effect of KCl was not observed, the denaturation of the enzyme was lower when compared to the NaCl recovery. The same behavior was observed by Braga, Lemes and Kalil [37] in β-galactosidase purification via ion-exchange chromatography. These authors found that the purification factor and enzyme recovery were higher when the KCl in the same concentration used in the present work (1 M) was used in the purification when compared to NaCl and ammonium sulfate.

The purification factor and enzyme recovery during the ultrafiltration decreased when the sodium and potassium chloride concentrations were increased from 0.01 M to 0.1 M. The prolonged contact of the enzyme with the salts seemed to have caused enzymatic denaturation. In addition, several changes were observed in the structure of the molecules and the electrostatic interactions between proteins, thereby resulting in protein agglomeration and consequent deposition and adsorption in the membrane surface to a larger extent [48,62].

3.2. Diafiltration

The removal of salts used in the purification processes is essential for maintaining the enzyme catalytic power, since it is known that large salt concentrations in certain cases can affect and even denature the target enzyme. This evidence reveals the importance of this process [63,64].

Thus, the ultrafiltration process was applied in diafiltration mode as an alternative to remove salts and other remaining contaminants at the end of the previously established purification procedures (aqueous two-phase system, precipitation, and ion-exchange chromatography) [1]. The fractions containing the purified enzyme derived from various purification procedures were diafiltered by adding a solute-free solvent (potassium phosphate buffer 0.05 M) and the filtration was stopped when the target salts were not detected in the enzyme extract. The number of cycles (6 cycles), as previously established for ammonium sulfate, was considered equivalent for other salts [65]. Moreover, our research group previously defined that at least five cycles were required for the total removal of the remaining PEG from the aqueous two-phase system in purification steps [21].

The combination of ultrafiltration and diafiltration processes was effective, particularly for concentrating and desalting the enzyme extract; furthermore, the total recovery of the sample was possible. The β-galactosidase recovery rate via diafiltration, regardless of the salt used, was always greater than 90% (

Table 3. Purification factor and β-galactosidase recovery in the diafiltration process after different techniques.

Purification Steps	Salts	Diafiltration Cicles *	% R	PF
Aqueous two-phase system	Potassium phosphate	6	93.4 ± 2.7 a	1.1 ± 0.1 a
Precipitation	Ammonium sulfate		97.2 ± 3.4 a	1.0 ± 0.1 a
Ion exchange chromatography	Potassium phosphate		93.4 ± 2.1 a	1.0 ± 0.12 a

* Two diavolumes of diafiltration solution per cycle. Data followed by the same lowercase letters in an individual column do not differ significantly (p > 0.05) in relation to the same parameter.

), thereby indicating that this process is feasible for removing the salts present in the enzyme extract without significantly affecting the enzyme recovery.

Another advantage of the diafiltration mode is associated with the possibility of reducing the concentration polarization and membrane fouling effects, in certain cases [66]. Thus, the choice of operating parameters is crucial in achieving good recovery of the enzyme. The diafiltration mode did not provide high purification factors; however, the process was effective for the complete removal of the salts present in the extract.

4. Conclusions

The ultrafiltration and concentration of β-galactosidase from Kluyveromyces marxianus CCT 7082 were efficiently applied, despite not promoting high purification factors (=1.0-fold). In this case, the use of ultrafiltration at pH 7.5 and 1.5 kgf/cm² with a 50 kDa membrane resulted in the maximum concentration of beta-galactosidase extract (7.1-fold) and 100% enzyme recovery. The increase in operating pressure caused losses in the enzyme recovery of up to 38% of the process yield due to the greater flux reduction caused by the membrane deformation, concentration polarization, and fouling due to the applied pressure, leading to inefficiency in the process. The use of salts in the feed solution (KCl and NaCl) caused decreases in enzyme purification and recovery of up to 40% during the process using NaCl, probably due to protein denaturation. The diafiltration mode was efficient for salt removal from the purified enzymatic extract using different techniques, including an aqueous biphasic system, ion exchange chromatography, and ammonium sulfate precipitation, thereby indicating its potential application (recovery >93.4% and purification factor of 1.0-fold) compared to other more expensive and laborious techniques.