experimental design
Catherine Teixeira de CARVALHOa, Sérgio Dantas de OLIVEIRA JÚNIORa, Wildson Bernardino de Brito LIMAa, Fábio Gonçalves Macêdo de MEDEIROSb, Ana Laura Oliveira de Sá LEITÃO a, Everaldo Silvino dos SANTOSa, Gorete Ribeiro de MACÊDOa, Francisco
Canindé de SOUSA JÚNIORc *
a Laboratory of Biochemical Engineering, Chemical Engineering Department, Federal
University of Rio Grande do Norte (UFRN), 59078-970, Natal-RN, Brazil
b Bioprocess Laboratory, Chemical Engineering Department, Federal University of Rio Grande
do Norte (UFRN), 59078-970, Natal-RN, Brazil
c Laboratory of Bromatology, Department of Pharmacy, Health Sciences Center, Federal
University of Rio Grande do Norte (UFRN), 59012-970, Natal-RN, Brazil
*Corresponding author:
Francisco Canindé de Sousa Júnior
Laboratory of Bromatology, Department of Pharmacy, Health Sciences Center, Federal University of Rio Grande do Norte (UFRN), 59012-970, Natal, Brazil. E-mail: fcfarma@yahoo.com.br
Abstract
The use of β-galactosidase in food products has been a major focus of the industry. Therefore, the development of efficient and inexpensive methodologies to purificate this product is extremely important. Thus, the aim of this study was to partially purified the enzyme β- galactosidase (β-gal) by ion exchange chromatography, in a fixed bed column, evaluating the influence of the parameter’s pH and ionic strength, in order to obtain a high purification factor. To determine the initial adsorption conditions of the β-gal, four types of adsorbents resins were tested in rotary incubators: Capto MMC multimodal, Streamline DEAE, Streamline SP and Amberlite XAD polymeric-XDA 7. The pH influence was also studied, so the systems were exposed to different pHs 6-8. Then, after determined the best adsorption conditions, an experimental design 22 with three central points was applied to optimize the purification using a FPLC, evaluating the ionic strength (50 and 200 mM) and the pH (6.0 and 7.0). Subsequently, the β-gal enzyme fractions of the best purification assay were submitted to a qualitative analysis by electrophoresis. The enzyme adsorption capacity in the rotary incubator was favored by the high pH and the Streamline DEAE resin had the best selectivity to the target molecule, with a β-gal retention capacity of 5.79 ± 0.79 U/g at pH 7.0. The design results showed the passage from the lower level (-1) to the superior (+ 1) of the ionic strength and pH increases the purification fold (PF) of the β-gal enzyme to 2.00 ± 0.42, however the ionic strength exerted greater influence on the PF response The fraction of the β-gal enzyme in the best elution range using between 0.1 and 0.4 M of NaCl in the and purification condition with a buffer with 200mM and pH 7.0 was submitted to electrophoresis, where an evident band with molecular weight in the range between 60 and 120 kDa was observed, configuring the enzyme of interest. These results point to the recovery of a stable β-galactosidase of K. lactis with potential industrial application from low-cost residue.
1. Introduction
β-galactosidase (β-gal) is an enzyme widely used in the food industry, it acts as a catalyst in the hydrolysis of lactose in milk and its derivatives, and in the synthesis of Galacto- Oligosaccharides (GOS) with prebiotic properties (Zhao et al., 2018).
Lactose, the carbohydrate found in milk and derivatives, is a disaccharide formed by glucose and galactose. In humans, lactose intolerance occurs in 75% of the population, being caused by the insufficiency of β-gal in the body, which results in a decrease of the enzyme activity in the membrane of the mucosa edge located in the small intestine of adults. Enzyme formulations of β-gal in dairy products can reduce lactose and promote a beneficial effect for lactose intolerant (Braga et al., 2014; Zhao et al., 2018)
The lactases or β-gal of biotechnological interest that are produced by yeasts of the genus Kluyveromyces, are generally intracellular and obtained by submerged fermentation. This fermentation, as in most biotechnological process, involves the purification of proteins and peptides from different strategies (Machado et al., 2015). In this context, one of the promising techniques used for this purpose is chromatography, where the protein binds to an adsorbent by ionic or hydrophobic interactions being subsequently eluted for recovery.
The chromatography using an ion exchange matrix is widely used for protein adsorption, in which the concept is based on the attraction between the proteins molecules and the resin that presents opposite charges. One of the advantages of this process is that it provides smooth separation conditions, allowing proteins to maintain their conformation. (Braga et al., 2014; Medeiros et al, 2012).
Thus, the aim of this study was to evaluate the best condition of purification using a single chromatographic step varying the adsorbent, pH and ionic strength, using purification fold and yield as reponses.
2. Materials and methods
2.1 Microorganism and inoculum production
Kluyveromyces lactis NRRL Y-8279 used in this study was provided by the ARS
Collection culture (Peoria, Illinois, USA). The strain was maintained in a 30% (v/v) solution of glycerol at -20 °C. For pre-inoculum preparation, K. lactis was cultivated in YEPD medium containing 10 g/L of yeast extract, 20 g/L of dextrose, 20 g/L of peptone and 20 g/L of agar for 24 h, at 30 °C. Then, three isolated colonies (inoculum) were transferred into 50 mL of the culture medium containing 10 g/L of lactose, 5.0 g/L of potassium phosphate, 1.2 g/L of ammonium sulfate, 1.0 g/L of yeast extract and 0.4 g/L of magnesium sulfate, prepared in potassium phosphate buffer (0.2 M, pH 5.5) according to (Lima et al., 2013). The Lactose used in the growth medium was exclusively from the cheese whey partially deproteinized (CWD). The inoculums were cultivated in Erlenmeyers under agitation of 180 rpm, at 30 °C, for 16 h in a rotating incubator.
2.2 Production of enzymatic extract
The enzyme β-galactosidase (β-gal) was produced through submerged cultivation por
Kluyveromyces lactis NRRL Y-8279 using rotary incubators (Tecnal, model TE-241), with 250
mL Erlenmeyers flasks containing 50 mL of culture medium (20 g/L of QWD, 1.3 g/L of (NH4)2SO4, 12 g/L of yeast extract, 5.0 g/L of KH2PO4 and 0.4 g/L MgSO4.7H2O, in 0.2 M
potassium phosphate buffer at pH 5.5, according to the adapted methodology of (Braga et al., 2014). The fermentation was performed with 10% of inoculum (v/v) prepared as described in section 2.1. The culture medium was previously sterilized and the lactose from the QWD was filtered in a 0.22 μm membrane. The cultivations in rotary incubators were performed at 30 °C, 180 rpm. Samples were removed after 20 h of fermentation.
The enzymatic extract was obtained from the yeast cell disruption after the fermentation, according to (Braga et al., 2014). The procedure was performed in centrifuge tubes of 50 mL, containing 25 mL of the cell suspension and 27.5 g of glass pearls (diameter ranging from 0.95 to 1.05 mm) under vortex agitation for 5 min, followed by 1 minute in an ice bath. This procedure was repeated 4 times. Then, the suspension was centrifuged (Cientec, CT-5000R model, Brazil) at 5200 × g during 10 min at 4 ºC. The enzyme extract was collected and stored at -20 °C.
2.3. Determination of adsorption conditions for -gal purification
In order to determine the initial adsorption conditions of the β-Gal enzyme, four types of adsorbents (Multimodal capto MMC, Streamline DEAE, Streamline SP and Amberlite XAD polymeric-XDA 7) and three pHs (6.0, 7.0 and 8.0) were tested. The assays were performed in batch, at 30 ºC and 150 rpm, in a rotary incubator for 60 minutes. In 25 mL Erlenmeyer vials, 0.2 g of resin, 2 mL of sodium phosphate buffer 200 mM in the respective pH, and 2 mL of enzymatic extract were added (Padilha et al., 2017). The enzymatic activity retained in the solid phase calculated according to equation 1:
𝑞 (𝑈/𝑔) =𝑉 (𝑐0−𝑐)
𝑉𝑎𝑑𝑠 (1)
Being equation 1: q the amount of enzyme adsorbed on the resin, V the volume of the enzyme extract, C0 the value of the initial activity, C the value of the activity in the equilibrium, Vads
the volume of the adsorbent.
2.4. -gal purification
The resin that presented the best adsorption capacity to the target molecule was selected for the purification assays using a fast protein liquid chromatography (FPLC) system (AKTA
start, GE healthcare Bio-Science, New Jersey, USA) using a fixed bed column HR 16/5 with a bed volume of 6 mL (GE Healthcare Bio-Sciences AB, Uppsala, Sweden).
The process was performed with a superficial velocity of the mobile phase of 100 cm.h-
1. The column containing the DEAE streamline resin was initially equilibrated for 30 minutes
with buffer A (sodium phosphate). Then, 4.5 mL of the crude extract was injected into the system. After the load, the non-bonded or weakly bonded molecules were removed from the system in the washing step, in which a volume of 7.5 mL of buffer A was used. And later, during the elution stage, a mobile phase ionic strength was linearly increased by adding the buffer B (sodium phosphate 50 mM, 125 mM and 200 mM at pH 6, 6.5 and 7 and NaCl 1 M) from 0 to 100%, aiming to separate the proteins according to the strength with which they were adsorbed in the resin. Samples were collected at each 1.5 mL. Two runs were performed with the same parameters to atone to the reproducibility of the system.
The influence of the factors (independent variables) ‘ionic strength’ and ‘pH’ on the dependent variables ‘fold purification’ (FP) and the ‘yield’ (Y) of β-Gal was analyzed through an experimental design 22 with 3 central points, resulting in 7 assays. The factors and levels
chosen for the design are presented in Table 1.
Table 1: Factors and levels used in the experimental design 22
Factor Levels -1 0 +1 Ionic strength 50 mM 125 mM 200 mM pH 6.0 6.5 7.0
The analysis of the experimental design data was performed by of Analysis of Variance (ANOVA), at a significance level of p ≤ 0.05, and the extent of Variance explained by each
model was given by the coefficient of determination (R ). The Software Statistica® 7.0, was used for the graphical analysis of the data.
2.5 Analytical methods
To determine the hydrolytic activity of the β-gal enzyme, ortho-nitrophenyl-β-D- galactoside (ONPG) was used according to (Braga et al., 2014). The β-Gal hydrolytic activity was determined at 37 ºC and pH 6.6 for 10 minutes and expressed in U/mL. A unit (U) of enzymatic activity corresponds to the amount of enzyme that catalyzes a reaction with the formation velocity of 1.0 µmol of ortho-nitrophenol (o-nitrophenol) for 1 minute, in the conditions aforementioned. The total protein concentration was determined using the method proposed by (Bradford, 1976). Bovine serum albumin (BSA) was used as a standard protein from a calibration curve that ranged from 0 to 1.5 mg/ml of protein, in which the measurements were performed in triplicate. The protein concentration was expressed in mg/mL and used to calculate the specific activity of the enzyme β-Gal (U/mg).
From the enzymatic activity of the crude and partially purified extract, the purification yield was obtained according to equation 2 and the fold purification calculated from the specific activities of the crude and purified extract according to equation 3 (Leitão et al., 2018).
Yield (%) = 𝐸𝑛𝑧𝑦𝑚𝑎𝑡𝑖𝑐 𝑎𝑐𝑡𝑖𝑣𝑖𝑡𝑦 𝑜𝑓 𝑝𝑢𝑟𝑖𝑓𝑖𝑒𝑑 𝑒𝑥𝑡𝑟𝑎𝑐𝑡
𝐸𝑛𝑧𝑦𝑚𝑎𝑡𝑖𝑐 𝑎𝑐𝑡𝑖𝑣𝑖𝑡𝑦 𝑜𝑓 𝑐𝑟𝑢𝑑𝑒 𝑒𝑥𝑡𝑟𝑎𝑐𝑡 x 100 (2)
Fold purification =𝑆𝑝𝑒𝑐𝑖𝑓𝑖𝑐 𝑎𝑐𝑡𝑖𝑣𝑖𝑡𝑦 𝑜𝑓 𝑝𝑢𝑟𝑖𝑓𝑖𝑒𝑑 𝑒𝑥𝑡𝑟𝑎𝑐𝑡
𝑆𝑝𝑒𝑐𝑖𝑓𝑖𝑐 𝑎𝑐𝑡𝑖𝑣𝑖𝑡𝑦 𝑜𝑓 𝑐𝑟𝑢𝑑𝑒 𝑒𝑥𝑡𝑟𝑎𝑐𝑡 (3)
The β-gal enzyme fractions of the best purification range were subjected to qualitative analysis using polyacrylamide gel electrophoresis (PAGE) at 9%, according to the methodology adapted from (General Eletric Conpany, 2000)
3. Results and discussion
The results obtained in the adsorption experiments of the β-gal enzyme in different adsorbents and pHs using the sodium phosphate buffer are shown in Figure 1.
In Figure 1, it was observed that the retention capacity of the enzyme rises in all the basic regions tested, and the Streamline DEAE resin presents the best selectivity to the target molecule, which promotes a better performance of the chromatographic system with a β-gal enzyme retention capacity of 18.77 ± 0.14 U/g for pH 6.0, 5.79 ± 0.79 U/g at pH 7.0 and 13.50 ± 1.14 U/g at pH 8.0.
Figure 1 – Adsorption of β-gal on 4 resins in the pH range of 6.0-8.0, 30° C, 150rpm and 0,2g
support. Different lowercase letters (a, b, c, d) indicate statistical differences (p < 0.05) for different pH with the same adsorbent. Different uppercase letters (A, B, C) indicate statistical differences (p < 0.05) for different adsorbents considering the same pH.
For proteins, purification in its active form was dependent on the solubility, load and size of the molecule (Thadathil and Velappan, 2014). In most cases, the number of steps in the
entire purification process will increase according to the required degree of purity (Wilken and Nikolov, 2012). Each process requires resins with specific characteristics, they are usually polymers of high chemical, mechanical and biological stability such as: cellulose, agarose, dextran, polyacrylamide and polystyrene. In addition to the type of polymer, for the ion- exchange chromatography, another important factor is the ionized groups on the surface of the enzymes. These groups are generated from amino acid residues and their load balance (between negative and positive groups) resulting in the enzyme load. The presence of these groups varies with the pH of the medium and when they are present in equal number it is called isoelectric point (IP). Therefore, there are matrices loaded with positive groups, such as the DEAE type (diethylaminoethyl) called anionics, and those with negative groups, such as those of type CM (Carboxy-methyl), called cationic (Lovato et al., 2017).
3.2 Central composite design using FLPC
After identification of the best adsorbent and pH for the β-gal adsorption, a experimental design 22, with three central points, was carried out to optimize the enzyme purification. The DEAE streamline resin was tested in the automated chromatography system FPLC (AKTA Start in the pH Ranges 6.0, 6.5 and 7.0.). It was investigated the influence of the factors (independent variables) pH and ionic strength (IS) on the Fold Purification (FP) of the enzyme and the Yield (Y) of the process (dependent variables). The experimental conditions and the responses for linear experimental design are presented in Table 2. One can observe that the results related to FP ranged from 0.56 (assay 2) to 2.00 (assay 3), while those obtained for the Y ranged from 49.08% (assay 1) to 91.07% (assay 2).
Table 2: Experimental design 2 with coding and levels of variables used to purify the β-gal enzyme as a function of pH and ionic strength.
Run pH IS (mM) FP Y (%) 1 -1 (6.0) +1 (200) 1.29 ± 0.29 49.08 ± 0.75 2 -1 (6.0) -1 (50) 0.56 ± 0.52 91.07 ± 0.12 3 +1 (7.0) +1 (200) 2.00 ± 0.42 51.64 ± 0.35 4 +1 (7.0) -1 (50) 0.64 ± 0.13 79.95± 0.45 5* 0 (6.5) 0 (125) 1.21± 0.01 65.73± 0.45 6* 0 (6.5) 0 (125) 1.05 ± 0.85 66.47± 0.77 7* 0 (6.5) 0 (125) 1.08 ± 0.01 68.11± 0.41 *Central points
Caption: IS - Ionic strength; FP- Fold Purification e Y- Yield of activity purified enzyme
The use of statistical methods and mathematical models contributed to the development and optimization of processes over time. The response surface methodology (RSM) is a tool to evaluate and model experimental data in order to identify independent and combined influences of independent variables in the output variables of the process (Câmara junior et al., 2016).
Figure 2: Pareto Charts of the standardized effects for the responses yield (Y) of activity
purified enzyme and fold purification of β-gal (FP), respectively
Based on the adjusted regression coefficients obtained by the software Statistica 7.0, a statistical model that correlates the responses yield (Y%) and fold purification (FP) with the variables ionic strength and pH were constructed, as observed in equations 2 and 3, in which X1 is the pH and X2 the ionic strength.
A)
Y= 67.427 – 2.142.X1 – 17.576.X2 + 3.422.X1.X2 (2)
FP = 1.114 + 0.186.X1 + 0.512.X2 + 0.148.X1.X2 (3)
One can observe on the Pareto is charts obtained that the Y and FP responses were influenced by both linear effects of the selected variables. The concept of interaction terms indicates that changes in a combination of process variables had a synergistic effect on the experimental response (Zhao et al., 2018). The passage from the lower level (-1) to the higher (+ 1) of the ionic strength and pH increases the purification degree of the β-gal enzyme to 2.00, but the ionic strength exerts greater influence on the FP. The contributions of ionic strength and pH for enzyme recovery were -2,142 and -17,576, respectively.
The analysis of variance (ANOVA) performed in the obtained models confirmed the significance (p< 0.05) evidenced in Table 3 and the Lack-of-Fit test (LoF) was significant (p> 0.05), suggesting the capacity to predict the models. High correlation coefficients (R²; 0.98858 and 0.99594, for FP and Y respectively, indicate a strong concordance of the adjusted data with the empirical results, responsible for robust empirical models (Câmara Junior et al., 2016).
Table 3: Analysis of variance (ANOVA) of the adjusted models to the experimental reponses FP and Y. Y VARIANCE FACTOR gl SS MQ F p-valor Model 3 1300.497 433.499 245.192 0.000 Residual 3 5.305 1.768 Lack of fit 1 2.327 2.327 1.563 0.033 Pure error 2 2.978 1.489 Total 6 1305.802 R2 0.98858
Caption: Y- β-gal yield; FP- fold purification; gl – degrees of freedom; SS – Square sum; MQ – Mean square; R² – correlation coefficient.
Through the F test, it can be observed that the models proposed in equations 2 and 3 are statistically significant and useful for predictive purposes. For the FP, the Fcal value was 15.26
times higher than the Ftab (9.28), and for the Y the calculated value was 26.42 times higher than
the tabulated. FP VARIANCE FACTOR gl SS MQ F p-valor Model 3 1.274 0.425 141.666 0.042 Residual 3 0.015 0.003 Lack of fit 1 0 0 0 0.000 Pure error 2 0.015 0.007 Total 6 1.289 R2 0.99594
The 3D representations of the adjusted models (equations 2 and 3) are shown in the Figures 3 A e 3B. The highest region found on the response surfaces indicates the ideal conditions for maximizing the responses of the purification process of β-galactosidase.
Figure 3: Response surface (RS) of Y and FP. (A) RS of the yield and (B) of the fold
purification of β-gal as function of pH x IS.
It is observed in Figure 3(A) that lower values of ionic strength (IS) and pH favor the degree of recovery of the enzyme, reaching 91.07% ± 0.12 of yield. In contrast, the purity
A)
degree of the biomolecule of interest is not significant, achieving a 0.56 ± 0.52 fold purification. It is also evident that higher values of IS and pH increase the degree of purity for the enzyme reaching a FP of 2.00 ± 0.43 with a significant Y of 51.64% ± 0.35.
The results of Y and FP obtained in this work are promising and superior to those reported by (Lima et al., 2016) who used the technique of multimodal chromatography for recovery and purification of β-gal and reached around 48% of recovery with a FP of 1.17. In the case of (Medeiros et al, 2012), they used sepharose Q resin and obtained recovery values close to 88%, however it is important to highlight that in this study only commercial enzyme solutions were used, which favors the process, given the higher degree of purity of samples. In this way, Figure 4 shows the variation of protein concentration and enzymatic activity through stages of chromatography (load, elution and washing) of the run with the best performance.
Figure 4: Purification chromatogram of β-gal produced by K. Lactis in a fixed bed, submitted
The variation of the two responses analyzed during the entire run was observed. At the first moment, during the load, there was a slight concentration of protein, with little significant enzymatic activity. In the course of washing, it can be identified a slight increase in the amount of enzyme present in the mobile phase, considering that the function of this stage is to carry out the molecules not adsorber on to the matrix of the column and due to the high hydrophilic character of biomolecule of interest.
During the elution phase, it was evidenced that the total protein and enzyme concentrations did not increase proportionally to the ionic strength of the mobile phase, showing a purification range using between 0.1 and 0.4 M NaCl presenting similar results to (Lima et
al., 2016). The enzyme recovery from the crude broth is considered significant, even with the
presence of contaminants after cell lysis.
The chromatography using as matrix an ion exchange resin is well used in protein adsorption, in which the concept is based on the attraction between the molecules of proteins that are charged electrostatically and the resin that presents opposite loads. One of the advantages of this process is that it provides smooth separation conditions allowing proteins to maintain their conformation(Braga et al., 2014; Medeiros et al, 2012).
3.3 PAGE gel
The fraction of the β-gal enzyme in the best elution range using 0.1 to 0.4 M NaCl in the optimum purification condition with IS of 200mM and pH 7.0 were subjected to qualitative analysis by means of a 9% polyacrylamide PAGE gel under non-denatured conditions as shown in Figure 5.
Figure 5: Native PAGE gel of polyacrylamide 9%. Lane M – Maker protein (MM), Lane 1–
crude extract (EB), Lane 2 – elution sample of 0,1 to 0,4 M of NaCl.
The wide-band marker protein used contained molecular mass ranging from 66 to 669 kDa, serving as a standard to identify the molar mass of the β-gal enzyme used in this research. The β-gal enzyme may vary its properties according to its source. Molecular mass may range from 850 kDa of enzyme produced by Escheriria coli to 201 kDa for those produced by
Kluyveromyces Marxianus (Gekas, V, and Lopez-Leiva, 1985). The literature reports that β-gal
molecules originating from Kluyveromyces lactis have a molecular mass varying between 120 and 140 kDa, values calculated by particle size exclusion chromatography (Boeris et al., 2012; Carminatti, 2001). It is observed in Figure 4 an evident band with molecular weight in the range between 60 and 120 kDa was observed, configuring the enzyme of interest.
4. Conclusion
In the present study, the enzyme adsorption capacity of the resind in the rotary incubator rises in all the basic regions tested, and the Streamline DEAE resin presented the best selectivity to the target molecule, with a β-gal retention capacity of 18.77 U /g ± 0.14 for pH 6.0, 5.79 U/g ± 0.79 at pH 7.0 and 13.50 U/g ± 1.14 for pH 8.0. The results of the experimental design showed
the passage from the lower (-1) to the higher (+1) level of Ionic strength and pH increases the