Immobilization of Aspergillus oryzae β-galactosidase in an agarose matrix functionalized by four different methods and application to the synthesis of lactulose

Accepted Manuscript Immobilization of Aspergillus oryzae β-galactosidase in an agarose matrix functionalized by four different methods and application to the synthesis of lactulose Cecilia Guerrero, Carlos Vera, Nestor Serna, Andrés Illanes PII: DOI: Reference:

S0960-8524(17)30108-6 http://dx.doi.org/10.1016/j.biortech.2017.02.003 BITE 17566

To appear in:

Bioresource Technology

Received Date: Revised Date: Accepted Date:

2 December 2016 1 February 2017 2 February 2017

Please cite this article as: Guerrero, C., Vera, C., Serna, N., Illanes, A., Immobilization of Aspergillus oryzae βgalactosidase in an agarose matrix functionalized by four different methods and application to the synthesis of lactulose, Bioresource Technology (2017), doi: http://dx.doi.org/10.1016/j.biortech.2017.02.003

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Immobilization of Aspergillus oryzae β-galactosidase in an agarose matrix functionalized by four different methods and application to the synthesis of lactulose. Cecilia Guerrero1*, Carlos Vera2, Nestor Serna1, Andrés Illanes1. 1. School of Biochemical Engineering, Pontificia Universidad Católica de Valparaíso (PUCV), Valparaíso, Chile. 2. Department of Biology, Faculty of Chemistry and Biology, Universidad de Santiago de Chile (USACH), Santiago, Chile. *: corresponding author. Tel. 56 32- 2272035; E-mail address: [email protected]

Abstract Aspergillus oryzae β-galactosidase was immobilized in monofunctional glyoxyl-agarose and heterofunctional supports (amino-glyoxyl, carboxy-glyoxyl and chelate-glyoxyl agarose), for obtaining highly active and stable catalysts for lactulose synthesis. Specific activities of the amino-glyoxyl agarose, carboxy-glyoxyl agarose and chelate-glyoxyl agarose derivatives were 3676, 430 and 454 IU/g biocatalyst with half-life values at 50°C of 247, 100 and 100 h respectively. Specific activities of 3490, 2559 and 1060 IU/g were obtained for fine, standard and macro agarose respectively. High immobilization yield (39.4%) and specific activity of 7700 IU/g was obtained with amino-glyoxyl-agarose as support. The highest yields of lactulose synthesis were obtained with monofunctional glyoxyl-agarose. Selectivity of lactulose synthesis was influenced by the support functionalization: glyoxyl-agarose and amino-glyoxyl-agarose derivatives retained the selectivity of the free enzyme, while selectivity with the carboxy-glyoxyl-agarose and chelate-glyoxyl-agarose derivatives was reduced, favoring the synthesis of transgalactosylated oligosaccharides over lactulose.

Keywords: agarose matrix, β-galactosidase, Aspergillus oryzae, immobilization, lactulose.

1

1. Introduction Lactulose (4-O-β-D-galactopyranosyl-β-D-fructofuranose) is a synthetic non-digestible disaccharide formed by one galactose and one fructose moiety linked by a β-glycosidic bond (Schumann, 2002). Considerable attention has been given to the synthesis of lactulose due to its outstanding therapeutic and nutritional properties, being a highly valuable lactose derivative (Wang et al., 2013). Most significant applications refer to its medical use in the treatment of chronic constipation and hepatic encephalopathy and to a lesser extent to its use as a prebiotic additive in functional foods and as a sucrose replacer (Schuster-WolffBühring et al., 2010; Panesar and Kumari, 2011; Song et al., 2013; Wang et al., 2013; Nahla and Musa 2015). Lactulose can be produced by chemical isomerization with alkaline catalysts, by enzyme-catalyzed synthesis from fructose and lactose with β-galactosidases, and more recently by epimerization of lactose with cellobiose 2-epimerase (Hicks and Parrish, 1980; Schuster-Wolff-Bühring et al., 2010; Kim and Oh, 2012; Song et al., 2013; Aït-Aissa and Aïder, 2014), but industrial production of lactulose is currently done by chemical synthesis only (Aider and Halleux, 2007; Panesar and Kumari, 2011; Wang et al., 2013). Chemical synthesis has several drawbacks: high concentrations of catalyst are required and the reaction is poorly specific generating undesirable side-products like epilactose, galactose, tagatose, isosaccharic acids and colored compounds, so that several purification steps are required to remove them (Zokaee et al., 2002; Aider and de Halleux, 2007; Panesar and Kumari, 2011; Nooshkam and Madadlou, 2016). Taking this into consideration, the enzymatic synthesis of lactulose appears as a promising technological option in compliance with the principles of green chemistry. It has the advantage that can be produced from byproducts of the dairy industry (whey or whey permeate), while the chemical synthesis requires high-purity lactose to reduce secondary reactions (SchusterWolff-Bühring et al., 2010); however, the main obstacle is the poor stability of the enzyme under reaction conditions (Illanes, 1999; Mateo et al., 2007a). Several strategies have been proposed to overcome this limitation going from the redesign of the enzyme molecule by protein engineering techniques to the rational design of matrices and procedures for enzyme immobilization (Fernández-Lafuente et al., 1995; Illanes, 1999; López-Gallego et al., 2005; Mateo et al., 2007a). Enzyme immobilization allows not only an increase in stability but also a better control of the reaction, a higher flexibility for reactor design, easiness of

2

product recovery and enzyme recovery and reutilization (Fernández-Lafuente et al., 1998; Illanes et al., 2006; Rodrigues et al., 2013). Many different supports and protocols for βgalactosidase immobilization have been developed (Albayrak and Yang, 2002; Panesar et al., 2006; Freitas et al., 2011; Urrutia et al., 2013). Among them, covalent immobilization to functionalized agarose outstands, allowing the linkage of the enzyme by different zones of its surface which is quite versatile in terms of the enzyme-support interactions (Mateo et al., 2010; Rodrigues et al., 2012). Heterofunctional immobilization allows orienting the immobilization of the enzyme by different zones in its surface, which is a sound alternative to the immobilization in glyoxyl-agarose that will only occurred by the region with the high density of lysine residues. Even though immobilization by other zones may reduce multipoint attachment, it may produce an increase in stability if that region is prone to inactivation. This new orientation may also allow a higher recovery of enzyme activity after immobilization and could eventually modify the enzyme selectivity (Mateo et al., 2007a; Mateo et al., 2010; Rodrigues et al., 2012). A strategy used for modifying the bonding pattern of the enzyme is the use of supports with more than one functional group (heterofunctional supports). The enzyme is first immobilized by physical or chemical interaction with complementary groups in the support and then immobilization conditions are changed to allow for covalent intramolecular bonding of the enzyme to the support using the functional groups of the latter (Mateo et al., 2010). With the purpose of favoring different orientations of the immobilized enzyme functional groups are incorporated to the support, for instance, amino groups to orient the immobilization of the enzyme by its region having the higher density of anionic amino acids (i.e. amino-glyoxyl-agarose), or carboxyl groups to orient the immobilization of the enzyme by its region having the higher density of cationic amino acids (i.e. carboxy-glyoxyl-agarose). It is also possible to incorporate metal chelates to orient immobilization by its region having the higher density histidine (i.e. chelate-glyoxyl-agarose) (Mateo et al., 2007b; Mateo et al., 2010; Rodrigues et al., 2012; Urrutia et al., 2013). This type of immobilization may modify the enzyme selectivity, which is particularly interesting in this case where a mixture of lactulose and transgalactosylated oligosaccharides (TOS) are synthesized (Guerrero et al., 2015); therefore, immobilization may allow an increase in lactulose yield by favoring its synthesis over the synthesis of

3

TOS. Four catalyst were prepared by covalent multi-point immobilization of the enzyme to glyoxyl-agarose (GA) and the heterofunctional supports amino-glyoxyl-agarose (Am-GA), carboxyl-glyoxyl agarose (Cx-GA) and chelate-glyoxyl-agarose (Che-GA). Since the morphology of the support will affect the biocatalyst properties, agarose of three different sizes (fine, standard and macro) and five degrees of crosslinking (2, 4, 6, 8 and 10%) were used as immobilization supports. In this way, the objective of the work is to evaluate the biocatalysts obtained in the synthesis of lactulose, determining the effect of the type of immobilization on the specific activity and immobilization yield of expressed activity, and on the yield and specificity of lactulose synthesis from lactose and fructose.

2. Materials and Methods 2.1. Materials Agarose Bead standard (2, 4, 6, 8, and 10 % cross-linked with epichlorohydrin) and Agarose Bead Fine and Macro (6 % cross-linked with epichlorohydrin) were purchased from Agarose Bead Technologies (Madrid, Spain). Lactulose was provided by Discovery Fine Chemicals (Wimborne, UK). D(+) galactose, D(+) glucose and D (+) lactose monohydrate, o-nitrophenol (o-NP), o-nitrophenyl-β-D-galactopyranoside (o-NPG) and galacto-oligosaccharides (GOS) standards were supplied by Sigma (St Louis, MO, USA). All other reagents were analytical grade and provided either by Sigma or Merck (Darmstadt, Germany). The commercial preparation of Aspergillus oryzae β-galactosidase traded under the name Enzeco™ Fungal Lactase Concentrate (AOG) was kindly donated by Enzyme Development Corporation, New York, USA and used in all experiments.

2.2. HPLC analysis of the reaction products Substrates and products of synthesis were analyzed in a Jasco RI 2031 HPLC equipment, provided with refractive index detector, isocratic pump (Jasco PU2080) and autosampler (Jasco AS 2055), using BP-100 Ca++ columns (300 mm x 7.8 mm) for carbohydrate analysis (Benson Polymeric, Reno, USA). Samples were eluted with milli-Q water at a flow-rate of 0.5 mL
min-1. Column and detector temperatures were 80 °C and 40 °C respectively. Chromatograms were integrated using the software ChromPass. Composition 4

of samples was determined by assuming that the area of each peak is proportional to the weight percentage of the respective sugar; such assumption was checked by a material balance according to Boon et al. (1999).

2.3. Determination of hydrolytic activity In order to determine the maximum hydrolytic potential of the enzyme, one international unit of hydrolytic activity (IUH) was defined as the amount of β-galactosidase that hydrolyzes 1 µmol of o-NPG per minute at 45mM o-NPG, 40 ºC and 4.5 pH (Guerrero et al., 2011). Hydrolysis activity was determined by measuring o-NP formation during incubation time at the above conditions as determined by absorbance at 420 nm in a Jenway 6715 spectrophotometer. The free enzyme preparation had a specific activity (asp) of 186,000 IUH⋅g-1 at the above assay conditions. The assays were carried out in triplicate, with standard deviations lower than 5%.

2.4. Determination of protein Protein concentration of the commercial preparation Enzeco™ Fungal Lactase Concentrate was 0.45 g⋅g-1 determined by the Bradford assay (Bradford, 1976). The assays were carried out in triplicate, with standard deviations lower than 5%.

2.5. Immobilization of Aspergillus oryzae β-galactosidase Immobilization of AOG in GA was done following the procedure described by Guisan (1988); immobilization on the heterofunctional supports was done as reported by Mateo et al. (2010). Agarose of three different sizes: fine (20-50 µm) standard (50-150 µm) and macro (150-350 µm) and five levels of crosslinking (2, 4, 6, 8 and 10%) were used. For all immobilization evaluated, agarose with standard size and a degree of crosslinking of 6% will be used as the baseline of comparison. When changing the size and/or degree of crosslinking, it will be indicated in the text.

5

Immobilization was monitored by measuring the enzymatic activity of samples of the suspension containing the catalyst and of the catalyst-free filtrate at regular time intervals. Immobilization was evaluated in terms of the following parameters, as previously described by Guerrero et al., 2015: - Protein immobilization yield (YIP), representing the mass (M) percentage of the contacted protein which is immobilized:

YIP =

M immobilized protein Mcontacted

⋅100

(Eq.1)

protein

The mass of immobilized protein is calculated as the difference between the mass of contacted protein and the mass of protein in the filtrate.

- Immobilization yield of expressed activity (YIA), representing the percentage of contacted activity that is expressed in the biocatalyst after immobilization:

YIA =

A expressed A contacted

⋅100

(Eq.2)

A is the catalytic activity expressed in IUH.

2.5.1. Glyoxyl agarose support Immobilization of β-galactosidase in GA proceeds in two steps. The first one is the activation of agarose with glycidol followed by oxidation with periodate according to the procedure described by Guisan (1988). The second step refers to the linkage of the enzyme to the activated support. In order to produce biocatalysts with different enzyme loads, four different concentrations of enzyme preparation were used for the immobilization in standard agarose support: 5, 10, 20 and 30 mgprotein⋅g-1support). Agarose standard size with

6

6% of crosslinking was the one used in the case of the monofunctional supports as baseline support. Immobilization was done at pH 10 using 0.1 M bicarbonate buffer. Due to the instability of A. oryzae β-galactosidase at such pH, immobilization was conducted in 20% (v/v) glycerol at 4°C. Then the Schiff bases formed are reduced with 0.5 M sodium borohydride (NaBH4) during 40 min at 4 ºC to produce a stable secondary amine (Guisan, 1988; Urrutia et al., 2013).

2.5.2. Heterofunctional agarose supports Epoxide-agarose, used as basis for all heterofunctional supports, was prepared as described by Mateo et al. (2010). Once prepared, different procedures were used for introducing new functional groups. In the case of amino-glyoxyl-agarose (Am-GA) activation was done with triethylamine; in the case of carboxy-glyoxyl-agarose (Cx-GA) activation was done with iminodiacetic acid; in the case of chelate-glyoxyl-agarose (Che-GA) activation was done with CuSO4 solution (Mateo et al., 2007b; Mateo et al., 2010; Urrutia et al., 2013). Then, the previously activated heterofunctional supports were oxidized with 10 mM NaIO4. Agarose standard size with 6% of crosslinking was the one used in the case of the heterofunctional supports as baseline support. For the immobilization in Am-GA, Che-GA and Cx-GA supports, the enzyme solution (30 mgprotein⋅gsupport-1) was firstly contacted with 1 g of support and kept under agitation at 25°C for 2 h. In the case of Am-GA and Cx-GA, the enzyme solution was prepared in 5 mM phosphate buffer pH 7 to favor the ionic interaction of the enzyme with the support; then multi-point covalent immobilization proceeded by filtering the suspension and incubating the retained catalyst in 5 mM bicarbonate buffer containing 20% (v/v) glycerol (Urrutia et al., 2013). For the immobilization in Che-GA, the retained catalyst was washed with 50 mM for promoting enzyme adsorption only by the histidine residues of the enzyme and the metal chelates (Urrutia et al., 2013).

7

2.6. Thermal stability of β-galactosidases under non-reactive conditions. Thermal stability of the free enzyme and the immobilized biocatalysts were compared at 50°C under non-reactive conditions, for assessing the potential stabilization due to immobilization, which may allow the reutilization of the biocatalysts. Inactivation profiles for the free enzyme and all biocatalysts were well described by one-stage first-order kinetics with no residual activity (Henley and Sadana, 1986): e = exp(− k ⋅ t ) e0

(Eq.3)

where e0 and e are the initial and residual enzyme activity after time t respectively, and k is the first-order kinetic rate constant of inactivation. Half-life (t1/2) is defined as the time when the remaining enzyme activity is 50% of the initial. Stabilization factor (SF) is defined as the of the half-life ratio of the immobilized and the free enzyme.

2.7. Synthesis of lactulose with free and immobilized enzyme. Lactulose synthesis was performed in 100 mL Erlenmeyer flasks at 50°C, pH 4.5, 50% (w/w) total initial sugars concentration, 100 IUH⋅g-1 lactose and fructose/lactose molar ratios of 1, 4, 8, 12 and 16. Lactose and fructose were dissolved in 100 mM citratephosphate buffer pH 4.5 at 95 °C and then the solution was cooled down to the reaction temperature. Reaction was conducted until reaching the point of maximum lactulose concentration. The assays were carried out in duplicate, with standard deviations never exceeding 5%. Quantification of sugars was done as described in section 2.2. Lactulose synthesis was evaluated in terms of the following parameters, as previously described by Guerrero et al., 2015: - Lactulose yield (YLu), representing the mass of lactulose synthesized (MLu) per unit mass of limiting substrate (lactose) (MLac), evaluated at the maximum lactulose concentration attained during synthesis: 8

YLu =

M Lu M Lac

(Eq.4)

- TOS yield (YTOS), representing the mass of TOS synthesized (MTOS) per unit mass of limiting substrate (lactose) (MLac), evaluated at the maximum lactulose concentration attained during synthesis:

M TOS M La

YTOS =

(Eq.5)

- Specific productivity of lactulose synthesis (πLu), representing the mass of lactulose produced per unit mass of protein in the enzyme preparation (MP) and unit time (t), evaluated at the maximum lactulose concentration attained during synthesis: π Lu =

M Lu MP ⋅t

(Eq.6)

- Specific productivity of TOS synthesis (πTOS), representing the mass of TOS produced per unit mass of protein in the enzyme preparation and unit of time, evaluated at the maximum lactulose concentration attained during synthesis:

πTOS =

M TOS MP ⋅ t

(Eq.7)

- Lactose conversion (XLac), representing the fraction of the initial lactose mass that is transformed during the reaction of synthesis: X Lac =

M Lac,0 − M Lac , t M Lac,0

(Eq.8)

where MLac,0 and MLac,t represent the initial lactose mass and remaining lactose mass after reaction time t respectively

- Selectivity of lactulose synthesis (SLu), representing the molar ratio of lactulose to TOS in the reaction medium, evaluated at the maximum lactulose concentration attained during synthesis:

9

SLu =

(Eq.9)

N Lu N TOS

where NLu and NTOS represent the moles of lactulose and TOS respectively.

3. Results and Discussion 3.1. Immobilization in glyoxyl-agarose support. Protein loading capacity of glyoxyl-agarose was tested with AOG at four different levels of offered protein. Figure 1 shows that asp increased with the increase in protein offered within the range evaluated but YIA decreased. YIP varied only slightly in the rage from 5 to 20 mg protein
g-1support, the highest value being obtained at 30 mg protein
g-1support, which was selected for further experiments. Similar results have been reported for the covalent immobilization of Kluyveromyces lactis β-galactosidase in glyoxyl-Sepharose support (Bernal et al., 2013a) and Bacillus circulans β-galactosidase in activated chitosan (Urrutia et al., 2014) and glyoxyl-agarose (Urrutia et al., 2013; Bernal et al., 2013b). Figure 1 Since the morphology of the support in an important variable in the formation of multipoint linkages (Mateo et al., 2006), the effect of the particle size and degree of crosslinking of agarose was studied on YIA, YIP and asp (Figure 1). It is well known that the support structure will affect its loading capacity and the level of its interaction with the enzyme. Physical characteristics of the support, like its particle size and pore size, determine its surface area affecting the loading capacity and determining the impact of diffusional restrictions (Brena and Batista-Viera, 2006; Klein et al., 2012). Porous supports are usually preferred because of the high surface of contact they offer for enzyme immobilization and the protection they provide against inactivation by the reaction conditions. However, diffusional restrictions are likely to be severe, limiting their catalytic potential. On the other hand, if the support is composed by fibers of smaller diameter than the enzyme intense interactions will be precluded due to a low geometric congruence (Mateo et al., 2006). Figure 1d and 1e shows that asp and YIA are reduced by

10

the increase in particle size, while YIP is not affected (Figure 1f). Thus, biocatalysts with asp of 3490 IUH
g-1, 2659 IUH
g-1 and 1060 IUH
g-1, and YIA of 39.4%, 31% and 19.9% were obtained with fine, standard and macro agarose respectively. Similar results were obtained by Klein et al. (2012) in the immobilization of K. lactis β-galactosidase in activated chitosan. Figure 1 also shows that values of YIA, YIP and asp are similar for all degrees of crosslinking except at 8% where values are slightly higher. All glyoxyl-agarose biocatalysts were then evaluated in the synthesis of lactulose and TOS from lactose with the purpose of establishing if reaction yield and selectivity of lactulose synthesis are affected by the biocatalyst particle size and degree of crosslinking.

3.2. Immobilization in heterofunctional supports. Figure 2 shows the effect of protein load on the immobilization of AOG in the heterofunctional supports Am-GA and Che-GA. Independent of the support functionalization, an increase in the offered protein produced an increase in the asp of the biocatalysts which is consistent with the results obtained with AOG immobilized in glyoxyl-agarose (Figure 1). An increase in protein load produced insignificant changes in YIA in the case of the Am-GA biocatalyst (Am-GA-AOG) but a definite increase in the case of the Che-GA biocatalyst (Che-GA-AOG). Thus, this load of 30 mgprotein
g-1support was selected to carry out the immobilization in the heterofunctional support Cx-GA; at such load YIA (4.9%) and asp (431 IUH
g-1) of the biocatalyst (Cx-GA-AOG) were obtained. These values are similar than obtained for Che-GA-AOG (YIA of 5.13 % and asp of 455 IUH
g-1); however, they are quite lower than obtained with Am-GA-AOG (YIA of 29.1% and asp of 3676 IUH
g-1). Figure 2 Since, according to Figure 1, agarose particle size was the variable most affecting the asp of the GA biocatalysts, its effect on the biocatalysts immobilized in the heterofunctional supports Am-GA and Che-GA was evaluated. As shown in Figure 2, the highest values of asp and YIA were obtained with fine agarose as support, being 7700 IUH
g-1 and 57 % for Am-GA-AOG and 780 IUH
g-1 and 13.4% for Cx-GA-AOG.

11

Results in Figures 1 and 2 show that the immobilization to small-sized GA and to Am-GA resulted in the biocatalysts with the highest asp while the lower values of asp and YIA were obtained with Cx-GA-AOG and Che-GA-AOG. The values obtained with GA-AOG and Am-GA-AOG were higher than reported for the immobilization of AOG by other methods; for instance, immobilization in silica produced a biocatalyst with asp of 1800 IUH
g-1 and YIA of 17.2 %, while immobilization in GA produced one with asp of 570 IUH
g-1 and YIA of 31.7 % (Bernal et al., 2013b). In the case of AOG immobilized by aggregation and crosslinking (CLAGs), YIA of 30 % was obtained and a very high asp of 15,000 IUH
g-1 was reported, which is due to the absence of an inert support (Guerrero et al., 2015). Values of YIA of 38.5, 37.5 and 21.5% were reported for the immobilization of B. circulans βgalactosidase in GA, Am-GA and Che-GA respectively (Urrutia et al, 2013).

3.3. Thermal stability of β-galactosidases in non-reactive conditions. In order to select the best support for AOG immobilization, stability of the four biocatalysts produced (see sections 3.1 and 3.2) and the free enzyme was determined under non-reactive conditions at 50°C and pH 4.5 (100 mM citrate-phosphate buffer) for free AOG, GA-AOG, Am-Ga-AOG, Cx-GA-AOG and Che-GA-AOG using in each case fine, standard and macro agarose. Immobilized enzymes in all the particle size supports and at all the protein loads were more stable than the free enzyme, protein load not affecting the stability of the immobilized biocatalysts. Thermal inactivation of free AOG, GA-AOG, Am-Ga-AOG, Cx-GA-AOG and Che-GAAOG was well described by a one-stage first-order mechanism with no residual activity (Table 1). Table 1 As shown in Table 1, 40% increase in stability factor with respect to free AOG was obtained with GA-AOG which is due to the rigidification of the enzyme structure that makes it more resistant to configurational changes. The same occurred with Am-GAAOG with more than 110% increase in stability with respect to AOG and 55 % with respect to GA-AOG. Slight increase in stability was observed for Cx-GA-AOG and Che-

12

GA-AOG; similar results were reported for CLAGs of AOG (Guerrero et al., 2015). Particle size of the support had no significant effect on the inactivation kinetics of the biocatalysts.

3.4. Synthesis of lactulose. Biocatalysts produced as described in sections 3.1 and 3.2 were evaluated in the synthesis of lactulose from lactose and fructose, for selecting the best one in terms of lactulose yield and selectivity. Figure 3 shows the values of YLu, YTOS, πLu, πTOS and SLu obtained. Figure 3 As shown, YLu and SLu increased with the increase in the fructose/lactose molar ratio, while YTOS significantly decreased, which occurred with the free enzyme and all immobilized enzymes. These results are in agreement with the results reported for CLAGs of AOG (Guerrero et al., 2015). It can also be seen that the distribution of products varied significantly among biocatalysts being the YLu obtained with the enzyme immobilized in the heterofunctional supports (Am-GA-AOG, Cx-GA-AOG and Che-GA-AOG) lower than obtained with the free AOG (Figure 3a); however YTOS higher than with the free enzyme were obtained with Cx-GA-AOG and Che-GA-AOG at some of the fructose/lactose molar ratios analyzed so that their SLu were lower (Figure 3e). Actually, SLu for all immobilized enzymes was lower than for the free enzyme, being the one obtained with GA-AOG the one more closely resembling it. With respect to πLu, the highest value was obtained with GA-AOG and the lowest with Cx-GA-AOG and CheGA-AOG at all the fructose-lactose ratios tested.

Figure 4 shows the effect of crosslinking degree of GA and size distribution of GA, AmGA and Che-GA on YLu, YTOS and SLu, obtained at different fructose/lactose molar ratios. Particle size of GA, Am-GA and Che-GA has a slight effect increasing YLu and reducing YTOS and SLu, which is attributed to the increase of internal diffusional restrictions with

13

catalyst particle size. The increase in crosslinking degree produced no significant effects on the parameters of lactulose synthesis obtained at different fructose/lactose molar ratios; values of such parameters were in all cases lower than the ones of the free enzyme due to the effect of internal diffusional restrictions (Guerrero et al., 2011; Guerrero et al., 2015). Figure 4 Profiles of lactulose and TOS obtained with the four biocatalysts are compared in Figure 5 with the free enzyme at a fructose/lactose molar ratio of 8. As seen, lactulose concentration obtained with GA-AOG was similar than obtained with AOG, but TOS concentration was slightly lower. Compared to AOG, concentrations of lactulose and TOS were significantly lower with Am-GA-AOG (Figures 5c and 5d), while in the case of CxGA-AOG and Che-GA-AOG lactulose concentrations were lower but TOS were higher (Figures 5f and 5h). Figure 5 Likewise, profiles of lactulose and TOS synthesis are compared in Figure 6 for the immobilized biocatalysts of different particle size and degree of crosslinking. Figure 6 The different particle size (Figure 6a and 6c) and degree of crosslinking of GA (Figure 6d) did not affect the behavior of Am-GA-AOG and Che-GA-AOG, but the concentrations of lactulose and TOS were lower than obtained with free AOG (Figure 5). Using the smallest particle size (fine) for GA, similar results were obtained than with free AOG, while results worsen in the bigger size catalysts (standard and macro) as a consequence of diffusional restrictions. This effect was stronger on the concentration of lactulose than TOS (Figures 6b).

4. Conclusions Active and stable biocatalysts were prepared by immobilizing AOG in functionalized agarose supports and evaluated in the synthesis of lactulose. Particle size had a strong effect on YA and asp of all biocatalysts. Highest YLu and SLu were obtained with GA-AOG.

14

Specific activity and stability of Am-GA-AOG were high, but YLu and SLu were lower than obtained with respect GA-AOG, though the highest among heterofunctional supports. In summary, to better define the best biocatalyst for lactulose synthesis a compromise has to be solved between the activity and stability of the biocatalysts and their performance in lactulose synthesis.

Acknowledgements Work financed by Chilean Fondecyt Grant 1160216 and 3130339. We acknowledge generous donation of β-galactosidase from Enzyme Development Corporation, DSM Food Specialties and Novozymes.

References 1. Aider, M., de Halleux, D., 2007. Isomerization of lactose and lactulose production: review. Trends Food Sci. Tech. 187: 356-364. 2. Aït-Aissa, A., Aïder, M., 2014. Lactulose: production and use in functional food, medical and pharmaceutical applications. Practical and critical review. Food Sci.Technol. 49: 1245-1253. 3. Albayrak, N., Yang, S.T. 2002. Production of galacto-oligosaccharides from lactose by Aspergillus oryzae β-galactosidase immobilized on cotton cloth. Biotechnol. Bioeng. 77: 819. 4. Bernal, C., Marciello, M., Mesa, M., Sierra, L., Fernandez-Lorente, G., Mateo, C., Guisan, J.M. 2013a. Immobilisation and

stabilization of β-galactosidase from

Kluyveromyces lactis using a glyoxyl support. Int. Dairy J. 28: 76-82. 5. Bernal, C., Urrutia, P., Illanes, A., Wilson, L. 2013b. Hierarchical meso-macroporous silica grafted with glyoxyl groups: opportunities for covalent immobilization of enzymes. New Biotechnol. 30: 500-506.

15

6. Boon, M.A., Janssen, A.E.M., van der Padt, A. 1999. Modeling and parameter estimation of the enzymatic synthesis of oligosaccharides by β-galactosidase from Bacillus circulans. Biotechnol. Bioeng. 64: 558-567. 7. Bradford, M. 1976. A rapid and sensitive method for the quantitation of microgram quantites of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: 248254. 8. Brena, B.M., Batista-Viera, F,. 2006. Immobilization of Enzymes, in: Guisan, M.J., (Eds.), Methods in Biotechnology: Immobilization of Enzymes and Cells. Humana Press Inc., Totowa, NJ, USA, pp. 15-30. 9. Fernández-Lafuente, R., Rosell, C.M., Guisán, J.M. 1995. The use of stabilised penicillin acylase derivatives improves the design of kinetically controlled synthesis. J. Mol. Catal. A-Chem. 101: 91-97. 10. Fernández-Lafuente, R., Armisén, P., Sabuquillo, P., Fernández-Lorente, G., Guisán, J.M. 1998. Immobilization of lipases by selective adsorption on hydrophobic supports. Chem. Phys. Lipids 93: 185-197. 11. Freitas, F.F., Marquez, L.D.S., Ribeiro, G.P., Bradao, G.C. 2011. A comparison of the kinetic properties of free and immobilized Aspergillus oryzae β-galactosidase. Biochem. Eng. J. 58-59: 33-38. 12. Guerrero, C., Vera, C., Plou, F., Illanes, A. 2011. Influence of reaction conditions on the selectivity of the synthesis of lactulose with microbial β-galactosidase. J. Mol. Catal. BEnzym. 72: 206-212. 13. Guerrero, C., Vera, C., Conejeros, R., Illanes, A. 2015. Repeated-batch operation for the synthesis of lactulose with β-galactosidase immobilized by aggregation and crosslinking. Bioresour. Technol. 190: 122-131. 14. Guisán, J.M. 1988. Aldehyde-agarose gels as activated supports for immobilization stabilization of enzymes. Enzyme Microb. Technol. 10: 375-382. 15. Henley, J., Sadana, A. 1986. Deactivation theory. Biotechnol. Bioeng. 28:1270-1282.

16

16. Hicks, K.B., Parrish, F.W. 1980. A new method for the preparation of lactulose from lactose. Carbohydr. Res. 82: 393-397. 17. Illanes, A. 1999. Stability of biocatalysts. Electron. J. Biotechnol. 2:15-30. 18. Illanes, A., Wilson, L., Caballero, E., Fernández-Lafuente, R., Guisán, J.M. 2006. Crosslinked penicillin acylase aggregates for synthesis of β-lactam antibiotics in organic medium. Appl. Biochem. Biotech. 133: 189-202. 19. Kim, Y-S, Oh, D-K. 2012. Lactulose production from lactose as a single substrate by a thermostable cellobiose 2-epimerase from Caldicellulosiruptor saccharolyticus. Bioresour. Technol. 104: 668-672. 20. Klein, M.P., Nunes, M.R., Rodrigues, R.C., Benvenutti, E.V., Costa, T.M.H., Hertz, P.F., Ninow, J.L. 2012. Effect of the support size on the properties of β-galactosidase immobilized on chitosan: Advantages and disadvantages of macro and nanoparticles. Biomacromolecules 13: 2456-2464. 21. López-Gallego, F., Montes, T., Fuentes, M., Alonso, N., Grazu, V., Betancor, L., Guisán, J., Fernández-Lafuente, R. 2005. Improved stabilization of chemically aminated enzymes via multipoint covalent attachment on glyoxyl supports. J. Biotechnol. 116:1-10. 22. Mateo, C., Palomo, J.M., Fuentes, M., Betancor, L., Grazu, V., López-Gallego, F., Pessela, B.C.C., Hidalgo, A., Fernández-Lorente, G., Fernández-Lafuente, R., Guisán, J.M. 2006. Glyoxyl agarose: A fully inert and hydrophilic support for immobilization and high stabilization of proteins. Enzyme Microb. Technol. 39: 274-280. 23. Mateo, C., Palomo, J.M., Fernández-Lorente, G., Guisán, J.M., Fernández-Lafuente, R. 2007a. Improvement of enzyme activity, stability and selectivity via immobilization techniques. Enzyme Microb. Technol. 40:1451-1463. 24. Mateo, C., Grazu, V., Palomo, J.M., Lopez-Gallego, F., Fernandez-Lafuente, R., Guisan, J.M. 2007b. Immobilization of enzyme on heterofunctional epoxy supports. Nat. Protocols 2: 1022-1033. 25. Mateo, C., Bolivar, J.M., Godoy, C.A., Rocha-Martin, J., Pessela, B.C., Ciriel, J.A., Muñoz, R., Guisan, J.M., Fernández-Lorente, G. 2010. Improvement of enzyme properties

17

with

a

two-step

immobilization

process

on

novel

heterofunctional

supports.

Biomacromolecules. 11: 3112-3117. 26. Nahla, T. K., Musa, T. N., 2015. Chemical isomerization of whey lactose to lactulose by using batch reaction. Pakistan J. Nutr. 14: 255-258. 27. Nooshkam, M., Madadlou, A., 2016. Maillard conjugation of lactulose with potentially bioactive peptides. Food Chem. 192: 831-836. 28. Panesar, P.S., Panesar, R., Singh, R.S., Kennedy, J.F., Kumar, H. 2006. Microbial production, immobilization and applications of β-galactosidase. J. Chem. Technol. Biotechnol. 81: 530-543. 29. Panesar, P. S., Kumari, S. 2011. Lactulose: production, purification and potential applications. Biotechnol. Adv. 29: 940-948. 30. Rodrigues, R.C., Ortiz, C., Berenguer-Murcia, A., Torres, R., Fernández-Lafuente, R. 2013. Modifying enzyme activity and selectivity by immobilization. Chem. Soc. Rev. 42: 6290-6307. 31. Schumann, C., 2002. Medical, nutritional and technological properties of lactulose. An update. Eur. J. Nutr. 41: 17-25. 32. Schuster-Wolff-Bühring, R., Fischer, L., Hinrichs, J. 2010. Production and physiological action of the disaccharide lactulose. Int. Dairy J. 20: 731-741. 33. Song, Y. S., Suh, Y. J., Park, C., Kim, S. W. 2013. Improvement of lactulose synthesis through optimization of reaction conditions with immobilized β-galactosidase. Korean J. Chem. Eng. 30: 160-165. 34. Urrutia, P., Mateo, C., Guisan, J.M., Wilson, L., Illanes, A. 2013. Immobilization of Bacillus circulans β-galactosidase and its application in the synthesis of galactooligosaccharides under repeated-batch operation. Biochem. Eng. J. 77: 41-48. 35. Urrutia, P., Bernal, C., Wilson, L., Illanes, A. 2014. Improvement of chitosan derivatization for the immobilization of Bacillus circulans β-galactosidase and its further application in galacto-oligosacharide synthesis. J. Agr. Food Chem. 62: 10126-10135.

18

36. Wang, H., Yang, R., Hua, X., Zhao, W., Zhang, W. 2013. Enzymatic production of lactulose and 1-lactulose: current state and perspectives. Appl. Microbiol. Biotechnol. 97, 6167-6180. 37. Zokaee, F., Kaghazchi, T., Zare, A., Soleimani, M., 2002. Isomerization of lactose to lactulose. Study and comparison of three catalytic systems. Process Biochem. 37, 629635.

19

Figure Captions

Figure 1: Effect of support load, size distribution and crosslinking degree of glyoxyl-agarose on the specific activity asp (a, d and g), immobilization yield of expressed activity YIA (b, e and h) and protein immobilization yield YIP (c, f and i) in the covalent immobilization of A. oryzae β-galactosidase in glyoxylagarose.

Figure 2: Effect of support load and size distribution on the specific activity asp and immobilization yield of expressed activity YIA in the covalent immobilization of A. oryzae β-galactosidase in heterofunctional amino-glyoxyl-agarose (a, b, e and f) and chelate-glyoxyl-agarose (c, d, g and h). Agarose standard size with 6% of crosslinking was used as baseline support.

Figure 3: Lactulose yield YLu (a), transgalactosylated oligosaccharides yield YTOS (b), productivity of lactulose synthesis πLu (c), productivity of transgalactosylated oligosaccharides πTOS (d) and selectivity of lactulose synthesis SLu (e) obtained in the synthesis of lactulose at 50°C, 50% w/w sugars, 100 IUH
g-1 and pH 4.5 with A. oryzae β-galactosidase free (
), immobilized in glyoxyl-agarose (), immobilized in amino-glyoxyl-agarose (), immobilized in carboxy-glyoxyl-agarose (), and immobilized in chelateglyoxil-agarose (). Agarose standard size with 6% of crosslinking was used as baseline support.

Figure 4: Effect of size distribution of glyoxyl-agarose (a, b and c), amino-glyoxyl-agarose (g, h and i) and chelate-glyoxyl-agarose (j, k and l) and crosslinking degree of glyoxyl-agarose on lactulose yield (YLu), on transgalactosylated oligosaccharides yield (YTOS) and on the selectivity of lactulose synthesis (SLu) obtained in the synthesis of lactulose at 50°C, pH 4.5, 50% w/w sugars and 100 IUH
g-1 with A. oryzae β-galactosidase immobilized in glyoxyl-agarose, at fructose /lactose molar ratio of 4 (
), 8 () and 12 ().

Figure 5: Profiles of lactulose (a,c,e,g,i) and transgalactosylated oligosaccharides (b,d,f,h,j) obtained in the synthesis of lactulose at 50 °C, pH 4.5, 50 % w/w sugars, 100 UIH
g-1 and fructose /lactose molar ratio of 8 with A. oryzae β-galactosidase free () and immobilized (
) in glyoxyl agarose (a,b), amino-glyoxyl agarose (c,d), carboxyl-glyoxyl agarose (e,f) and chelate-glyoxyl agarose (g,h). Agarose standard size with 6% of crosslinking was used as baseline support.

Figure 6: Effect of size distribution (fine: black, standard: grey and macro: white) and crosslinking degree (agarose standard size with 2 %: black, 4 %: dark grey, 8 %: light grey and 10 %: white) on the profiles of lactulose (diamonds) and transgalactosylated oligosaccharides (circles) obtained in the synthesis of lactulose at 50 °C, pH 4.5, 50 % w/w sugars, 100 IUH
g-1 and fructose /lactose molar ratio of 8 with A. oryzae βgalactosidase immobilized in amino-glyoxyl agarose (a), glyoxyl agarose (b and d) and chelate-glyoxyl agarose (c).

20

3000

100

a)

asp (IUH
g-1)

c)

b)

75

YIP (%)

2500

60

YIA (%)

2000 1500 1000

50

40

20 25

500 0

0 5

10

20

0

30

5

Support load(mgprotein
gsupport-1 )

10

20

30

4000

45

10

20

30

Support load(mgprotein
gsupport-1 ) 100

d)

e)

f) 80

YIP (%)

3000 30

YIA (%)

asp (IUH
g-1)

5

Support load(mgprotein
gsupport-1 )

2000

60

40 15

1000

20

0

0

0

Fine

Standard

Macro

4000

Fine

Standard

Fine

Macro

50

Standard

Macro

100

g)

h) 40

YIP (%)

30

2000

i)

80

YIA (%)

asp (IUH
g-1)

3000

60

20

40

10

20

1000

0

0 2

4

6

8

Crosslinking degree (%)

10

0

2

4

6

8

Crosslinking degree (%)

10

2

4

6

8

10

Crosslinking degree (%)

Figure 1

21

5000

40

b)

a)

30

YIA (%)

asp (IUH
g-1)

4000

3000

2000

20

10 1000 0

0 5

10

20

5

30

Support load(mgprotein
gsupport-1 ) 500

c)

30

d)

8

YIA (%)

asp (IUH
g-1)

20

10

400

300

200

6 4 2

100

0

0

5

10

20

5

30

Support load(mgprotein
gsupport-1 )

10

20

30

Support load(mgprotein
gsupport-1 )

8000

80

e)

4000

f)

60

YIA (%)

asp (IUH
g-1)

6000

2000

40

20

0

0

Fine

Standard

Macro

Fine

1000

Standard

Macro

16

g)

600

400

h)

12

YIA (%)

800

asp (IUH
g-1)

10

Support load(mgprotein
gsupport-1 )

8

4

200

0

0

Fine

Standard

Macro

Fine

Standard

Macro

Figure 2

22

0.25

a)

0.3

YTOS (g

g-1)

YLu (g

g-1)

0.4

0.2

b)

0.2 0.15 0.1

0.1

0.05

0

0 1

4 8 12 16 Fructose/Lactose (mol

mol-1)

1

0.6

4 8 12 16 Fructose/Lactose (mol

mol-1)

0.4

d) πTOS (g

mg-1
h-1)

π Lu (g

mg-1
h-1)

c) 0.4

0.2

0.2

0

0 1

4

8

12

1

16

Fructose/Lactose (mol

mol-1) 16

8

12

16

e)

12

S Lu/TOS (mol

mol-1)

4

Fructose/Lactose (mol

mol-1)

8

4

0 1

4 8 12 16 Fructose/Lactose (mol

mol-1)

Figure 3

23

0.35

0.12

a)

b)

0.1

0.25

0.15 0.1

0.06 0.04 0.02

0.05 0

0

Fine

Standard

Fine

Standard

0.06

YTOS (g

g-1)

YLu (g

g-1)

0.25

0.04

0.1

0.02

0.05 0

0

2

4

6

8

2

6

8

h)

0.1 0.2

0.15 0.1 0.05

0.08 0.06 0.04

0.15

k)

0.08 0.06 0.04

0.1

10

i)

4 3 2

Fine

Standard

Macro

8

0.1

YTOS (g

g-1)

0.2

8

5

Macro

SLu/TOS (mol

mol-1)

j)

6

0

Standard

0.12

0.25

4

1

Fine

Macro

0.35 0.3

2

Crosslinking degree (%)

0

0

Standard

4

2

0.02

Fine

6

10

SLu/TOS (mol

mol-1)

g)

f)

6

0.12

YTOS (g

g-1)

YLu (g

g-1)

4

0.14

0.25

Macro

8

Crosslinking degree (%)

0.3

Standard

0

10

Crosslinking degree (%)

YLu (g

g-1)

Fine

SLu/TOS (mol

mol-1)

e)

0.08

0.15

2

10

d)

0.2

4

Macro

0.1

0.3

6

0

Macro

0.35

c)

8

0.08

0.2

YTOS (g

g-1)

YLu (g

g-1)

10

SLu/TOS (mol

mol-1)

0.3

l)

6

4

2

0.02

0.05 0

0

Fine

Standard

Macro

0

Fine

Standard

Macro

Fine

Standard

Macro

Figure 4

24

2

4

TOS (% w/w)

Lactulose (% w/w)

6

a) 2

1

b) 0

0 0

0.5

0

1

0.5

X Lactose 2

4

TOS (% w/w)

Lactulose (% w/w)

6

c)

2

1

d) 0

0 0

0.2

0.4

0.6

0.8

0

1

0.2

XLactose

0.6

0.8

1

XLactose

TOS (% w/w)

Lactulose (% w/w)

0.4

2

6

4

e) 2

0

1

f) 0

0

0.5

1

0

X Lactose

0.2

0.4

0.6

0.8

1

X Lactose 2

4

g) 2

0

TOS (% w/w)

6

Lactulose (% w/w)

1

X Lactose

1

h) 0

0

0.2

0.4

X Lactose

0.6

0.8

0

0.5

1

X Lactose

Figure 5

25

Sugars concentrations (% w/w) 4

0

6

5

c)

4

3

2

1

0

Sugars concentrations (% w/w)

Sugars concentrations (% w/w)

a)

3

2

1

Sugars concentrations (% w/w)

5

0 0.5

0 0.5

X Lactose 1

6

5

1

b)

4

3

2

1

0

X Lactose 0

X Lactose

0.5

5

0 0.5

1

6

d)

4

3

2

1

0

X Lactose 1

Figure 6

26

Table 1: Parameters of inactivation at 50°C and pH 4.5, determined according to a one-stage first-order mechanism of inactivation, for A. oryzae β-galactosidase immobilized in glyoxyl-agarose (GA-AOG) in amino-glyoxyl-agarose (Am-GA-AOG), carboxy-glyoxyl-agarose (Cx-GA-AOG) and chelate-glyoxylagarose (Che-GA-AOG).

Free enzyme GA 6% BCL Standard Am-GA 6% BCL Standard Che-GA 6% BCL Standard Cx-GA 6% BCL Standard GA 6% BCL Fine GA 6% BCL Macro Am-GA 6% BCL Fine Am-GA 6% BCL Macro Che-GA 6% BCL Fine Che-GA 6% BCL Macro

k (h-1)

t1/2 (h)

SF

0.0088 0.00645 0.00413 0.0069 0.0067 0.00693 0.00607 0.00424 0.00452 0.00710 0.00641

78 109 169 100 103 100 114 166 178 97 107

1.00 1.40 2.17 1.28 1.32 1.28 1.46 2.13 2.28 1.24 1.37

k: first-order kinetic rate constant of inactivation.; t1/2: half-life; SF: stabilization factor (half-life ratio of the immobilized and the free enzyme).

27

- β-galactosidase was immobilized in mono- and heterofunctional agarose supports

- Higher specific activity (3676 IU/g) was obtained in amino-glyoxyl agarose as support - Such catalyst had a very high thermal stability at reaction temperature of 50°C

- Highest product yield was obtained using monofunctional glyoxyl agarose as support - Lactulose yield and selectivity were influenced by the support functionalization

- Selectivity was the highest when using the monofunctional glyoxyl agarose as support

28