This is the accepted version of the article: García Bofill, Miquel; Sutton, Peter W.; Guillén Montalbán, Marina; [et al.]. «En- zymatic synthesis of vanillin catalysed by an eugenol oxidase». Applied catalysis A: general, Vol. 582 (July 2019), art. 117117. DOI 10.1016/j.apcata.2019.117117 This version is avaible at https://ddd.uab.cat/record/217011 under the terms of the license 1 Enzymatic synthesis of vanillin catalyzed by an eugenol oxidase 2 Miquel Garcia-Bofill, Marina Guillén1, Peter W. Sutton1, Gregorio Álvaro1, 1 3 Bioprocess Engineering and Applied Biocatalysis Group, Department of Chemical Biological and 4 Environmental Engineering, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain. 5 miquel.garcia.bofill@uab.cat; marina.guillen@uab.cat; peter.sutton@uab.cat; 6 gregorio.alvaro@uab.cat; 7 8 Corresponding author: 9 Gregorio Álvaro 10 Department of Chemical, Biological and Environmental Engineering, Universitat Autònoma de 11 Barcelona, 08193 Bellaterra, Spain 12 Telephone: +34 93 581 1806 13 Fax: +34 93 581 2013 14 gregorio.alvaro@uab.cat 15 16 17 18 19 20 21 1 22 Abstract 23 24 Vanillin is one of the most important flavours produced in the world. Due to its increasing value 25 and its demand, mainly in the food industry, several ways to obtain it at lower prices are under 26 study. One of the routes is based on the oxidation of vanillyl alcohol by Eugenol oxidase (EUGO), 27 which has high potential to be used at industrial scale owing to the high space time yields that 28 can be obtained at lab scale (2.9 g L-1 h-1 of vanillin). Additionally, EUGO can be immobilised 29 efficiently onto different supports (MANA-agarose, Epoxy-agarose and Purolite 8204F) which 30 can be reused several times to perform the oxidation preserving good stability and improving 31 more than 3-fold the biocatalyst yield. 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 2 50 1. Introduction 51 Natural vanilla (Vanilla planifolia; V. pompona) is one of the most important flavours harvested 52 in the world [1], with a current volume of almost 8 thousand tonnes per annum. 77% of crop 53 production is concentred in Madagascar, Indonesia and China and its production price over the 54 last 10 years is around 5700 USD T-1 [2]. There are around 180 aromatic compounds in vanilla, 55 and, the one that gives its characteristic flavour is vanillin (4-hydroxy-3-methoxybenzaldehyde) 56 which only represents 2% (w/w) of vanilla [3]. Even though agricultural production is increasing, 57 it is not sufficient to supply demand, which has doubled over the past 20 years. Its high price 58 and widespread use in the food, cosmetic, and pharmaceutical industries triggered the necessity 59 to produce it by other means. It is estimated that less than 1% of total world production of 60 vanillin now comes from the natural beans [4]. 61 Various different ways to produce vanillin using chemical synthesis [5–9], microbial [10– 62 18]/plant cells [19] and enzymatic biotransformation have been described in the literature. 63 [20,21],[22] However, the productivities of the biotransformation approaches are still far from 64 being industrially implementable. Chemical synthesis is used to produce the majority of vanillin 65 production, but it is declining in interest in many markets (chiefly in the food industry) in favour 66 of product produced by microorganisms, plants cells and enzymes, that can be labelled as 67 “natural” [23]. Furthermore, consumers are increasingly willing to pay a premium for products 68 produced using natural means [24]. 69 Biotransformation processes using isolated enzymes can often be more expensive than those 70 using whole cells due to the extra processing required during biocatalyst production, but this 71 can be considerably offset by the reduction in secondary reactions and reduced product 72 purification costs. One of the ways to improve the industrial viability of a bioprocess is to reuse 73 the biocatalyst by immobilisation. Immobilisation does not only permit the reuse and easier 74 recovery of the enzyme, but also allows the utilization of different reactor configurations such 3 75 as continuous or cross flow in addition to traditional batch formats. It can also improve 76 biocatalyst stability towards various reaction parameters such as pH, temperature, organic 77 solvent or inhibitors. Immobilisation can sometimes also enhance biocatalyst activity, specificity 78 and selectivity [25],[26],[27]. 79 In the present paper, Eugenol oxidase (EUGO) was immobilised covalently onto different 80 supports and used to perform the conversion of vanillyl alcohol to vanillin (Figure 1Figure 1), Con formato: Fuente: Sin Negrita 81 reusing the biocatalyst, in order to improve the process metrics compared to reaction using 82 soluble enzyme. 83 84 Figure 1: Oxidation of vanillyl alcohol to vanillin catalysed by Eugenol oxidase (EUGO) and catalase. 85 86 2. Materials and methods 87 2.1 Materials 88 Eugenol oxidase from Rhodococcus jostii (EUGO) was provided by InnoSyn B.V. (The 89 Netherlands) as Escherichia coli lysates. Amino functionalized agarose (AMINO-agarose) and 90 non-functionalised agarose supports, used to obtain epoxy-agarose-UAB, were obtained from 91 Agarose Beads Technology® (ABT®) brands. Purolite® supports, methacrylate matrix activated 92 with amino (ECR8409 and ECR8415, with a pore size of 600-1200Å and 1200-1800 Å, 93 respectively) and epoxy groups (Praesto epoxy 45), were generously donated by Purolite® Life 4 94 Sciences (Bala Cynwyd, PA, USA). Catalase and all other reagents were supplied by Sigma-Aldrich 95 Chem. Co. 96 97 2.2 Methods 98 2.2.1 Protein determination and enzyme content 99 Total protein content of lysates containing EUGO was analysed using the Bradford method with 100 bovine serum albumin as standard [28]. 101 Enzyme content was determined using the sodium dodecyl sulphate polyacrylamide gel 102 electrophoresis (SDS-PAGE) (NuPage 12%, Invitrogen, USA) run in a Mini-PROTEAN II apparatus 103 (BioRad, USA) following the protocol of Laemmli et al. [29]. Low range protein markers (Biorad, 104 USA) were used for molecular weight determination. Gels were stained using Comassie G250 105 colloidal stain solution [34% (v/v) ethanol, 2% (v/v) H3PO4, 17% (w/v) NH4SO4 and 0.066% 106 Comassie G250] and the Image LABTM software (BioRad, USA) was used for image processing. 107 108 2.2.2 Enzymatic activity assays 109 The activity of EUGO was measured spectrophotometrically at 340 nm following the production 110 of vanillin (ɛ= 27 mM-1cm-1) from vanillyl alcohol. A 50 µl aliquot of enzyme sample was added 111 to a cuvette with 950 µl of substrate dissolved in buffer to a final concentration of 0.5 mM vanillyl 112 alcohol in 50 mM glycine-NaOH, pH 9.5 at 30°C. One unit of EUGO activity (U) is defined as the 113 amount of enzyme required to produce 1 µmol of vanillin alcohol per minute at the conditions 114 described above. 115 Catalase activity was measured spectrophotometrically at 240 nm following the consumption of 116 the substrate, H2O2 (ɛ= 0.0383 mM-1cm-1). A 50 µl aliquot of enzyme sample was combined with 5 117 950 µl of 20 mM H2O2 in 50 mM potassium phosphate buffer, pH 7.0 at 25°C. One unit of catalase 118 activity (U) is defined as the enzyme required to convert 1 µmol of H2O2 per minute at the 119 conditions defined previously. 120 Activity assays were carried out using a Cary 50 Bio UV-visible spectrophotometer (Palo Alto, 121 USA). 122 Activity assays were carried out using 1.5 mL cuvettes suitable for UV. When the activity of EUGO 123 immobilised derivatives was measured, double of each volume was added and 3 mL cuvettes 124 were used with magnetic stirring to maintain a proper suspension of the derivatives during the 125 measurement. 126 127 2.2.3 Stability of EUGO under different pH conditions 128 The effect of pH on the stability of EUGO was studied by incubating the enzyme at different pHs 129 in the range 5-10. 1 U mL-1 of EUGO was added to in 100 mM buffer solution with a final volume 130 of 10 mL. Samples were incubated on a roller (Movil-Rod Selecta S.A.) for 24 hours at 25°C. The 131 buffers used (100mM) were: sodium acetate (pH 5-5.5), potassium phosphate (pH 6.0-8.0), Tris- 132 HCl (pH 8.5-9), and sodium bicarbonate (pH 10). The activity of the samples was analysed, as 133 described above, at 0, 0.5, 1, 2, 4, 8 and 24 hours. Experiments were carried out in duplicate. 134 135 2.2.4 Effect of hydrogen peroxide on EUGO stability 136 In order to evaluate the effect of the hydrogen peroxide (formed as a by-product in vanillyl 137 alcohol oxidation) on EUGO stability, experiments were performed in presence of hydrogen 138 peroxide. 1 U EUGO mL-1 was incubated under reaction conditions (30% acetone, 50 mM 139 potassium phosphate pH 7.5, mild agitation conditions, 25°C), with different concentrations of 6 140 hydrogen peroxide (400, 200, 100, 50, 25, 10 and 0 mM) for 24 hours. EUGO activity was 141 analysed, as described above, at different times. Experiments were performed in duplicate. 142 143 2.2.5 EUGO immobilisation 144 All the immobilisation studies shown below were carried out as follows: support was previously 145 hydrated, washed with distilled water several times and two times with immobilisation buffer. 146 Supports were mixed with enzyme preparation in immobilisation buffer in a 1:10 (v/v) 147 support/solution ratio with a final volume of 10 mL. The mixtures were incubated under mild 148 agitation conditions at 25°C for all the immobilisation processes. Experiments were performed 149 in duplicate. 150 All the experiments were performed using the minimum units of activity per mL of support 151 where there are no diffusional limitations. To determine the limit without diffusional limitations, 152 different low loadings of enzyme (30, 10, 5 and 2.5 U mL-1 support) were immobilised to the 153 supports. When two different loadings show the same immobilisation yields and retained 154 activities, it can be affirmed that there are not diffusional limitations at these loadings. 155 During the immobilisation process, samples of supernatant and suspension were taken at 156 different times to test EUGO activity during immobilisation course. The blank was prepared by 157 using enzyme preparation in immobilisation buffer with water instead of the support for 158 knowing the behaviour of soluble EUGO under immobilisation conditions. 159 Once the enzyme was immobilised onto the support, the immobilised derivative was washed 160 and filtered to remove residual water, leaving moist beads. 161 In order to compare the immobilisations, immobilisation yield and retained activity values were 162 determined, measuring the activity as described above, as: 7 !"!# $ 163 % = & 100 !"!# !% !"!# !% !"# $ 164 ) * + % = & 100 165 Final supernatant activity: activity of the supernatant when the immobilisation finished. 166 Final suspension activity: activity of the suspension (supernatant and immobilised derivative) 167 when the immobilisation finished. 168 169 2.2.5.1 Immobilisation of EUGO onto amino functionalized supports 170 EUGO immobilisation onto amino functionalized support was carried out as follow: 10% (v/v) 171 support/solution mixture was incubated during 30 minutes to adsorb ionically the enzyme to 172 the support. Once the enzyme was ionically adsorbed, 25 mM (N-(3-dimethylaminopropyl)-N- 173 ethyl)carbodiimide (CDI) (final concentration) was then added and the mixture stirred for a 174 further 2 h. Then, 1 h incubation with 1 M NaCl was performed to ensure that the retained 175 activity is attached covalently. 176 177 2.2.5.2 Immobilisation of EUGO onto epoxy functionalised supports 178 Immobilisation of EUGO onto supports with epoxy groups was performed by incubating the 179 enzyme with the support in a high strength buffer (1 M potassium phosphate buffer at pH 7.5 180 and 8) for 4 hours at 25˚C. 0.2 M β-Mercaptoethanol was then added and the mixture incubated 181 for a further 4h at 4˚C in order to block epoxy groups that have not formed covalent bounds 182 with the enzyme. Two different epoxy functionalised supports were used: Functionalisation 183 method 1, containing 30 µmol of epoxy groups per g of support (F.M. 1) described by Axarli et 184 al. [30] and functionalisation method 2, containing 80 µmol of epoxy groups per g of support 185 (F.M. 2) described by Sunberg et al. [31]. 8 186 The immobilised derivatives were washed with the buffer needed and used immediately. 187 2.2.6 Stability of soluble and immobilised EUGO under reaction and storage 188 conditions 189 Stability of biocatalyst under reaction conditions (30% acetone in 50 mM potassium phosphate 190 buffer pH 7.5, 25°C, mild agitation conditions) and storage conditions (25 mM potassium 191 phosphate buffer pH 6, 4°C) was studied for free and immobilised EUGO onto the selected 192 supports (10% v/v). The activity of the biocatalyst was analysed over time until the enzyme had 193 lost half of its initial activity. Experiments were performed in duplicate. 194 195 2.2.7 Determination of the maximum enzyme loading on the selected supports 196 The experiments were performed by increasing the offered activity per mL of support following 197 the immobilisation procedures described above for each functionalization type until the activity 198 observed in the supernatant was higher than 10% of initial offered activity. Experiments were 199 performed in duplicate. 200 201 2.2.8 Preparative Scale Vanillin Synthesis Reactions with soluble and immobilised 202 EUGO 203 Reactions with immobilised EUGO were performed using 10% (v/v) of the immobilised derivative 204 in order to ensure efficient mixing. To compare catalyst performance in reactions, both soluble 205 and immobilised EUGO were used with the same units per mL of reaction. 206 Reactions were carried out at 10 mL scale using the following conditions: 400 mM of vanillyl 207 alcohol, 30% (v/v) acetone in 50 mM potassium phosphate buffer at pH 7.5, 25˚C, 1 vvm air 208 (hydrated with a 30% acetone solution in water), magnetic stirring (500 rpm), and 10 µl 9 209 antifoam. 9 mg mL-1 (35847 U mL-1 reaction) of catalase were added to the reaction to control 210 the peroxide formed by EUGO in the oxidation. Units of EUGO employed were determined as 211 explained above. Reaction completion time was determined as the moment that no substrate 212 consumption or product production was observed, monitored by GC as described in section 213 2.2.10. 214 215 2.2.9 Reusability of the immobilised derivative 216 Several cycles of reactions were performed using the same conditions as in section 2.2.8 using 217 immobilised EUGO as catalyst. At the end of each cycle, the immobilised catalyst was washed 218 three times with 20 mL of water and three times with 20 mL of reaction buffer (30% acetone, 50 219 mM potassium phosphate, pH 7.5). Immediately, the same quantities of fresh reaction 220 components used in the previous cycle were added to the recycled immobilised enzyme to start 221 a new reaction cycle. All the cycles were performed during the same time. The first cycle was 222 ended to reach completion (using fresh biocatalyst) and the conversion was determined by GC 223 as shown in section 2.2.10. The reaction time for the rest cycles was the same as the first one. 224 Error bars correspond to standard deviation. 225 226 2.2.10 GC analysis of vanillyl alcohol and vanillin 227 Vanillyl alcohol and vanillin quantification in preparative scale reactions shown above was 228 carried out by gas chromatography. Samples were extracted with ethyl acetate (1:20 v/v) 229 containing methylbenzoate (5 mM) as internal standard. The organic phase was analysed using 230 a 7890A Agilent gas chromatograph equipped with an Innowax 19095N-123 (30m x 530 µm x 1 231 µm) column. The column temperature was held at 100 °C for 5 min, then increased to 240 °C at 232 10°C per minute and maintained at this temperature for a further 2 minutes. The injector was 10 233 kept at 225 °C; for the flame ionization detector, the temperature was 250 °C. Helium was used 234 as a carrier gas at constant pressure of 10 psi. Retention times were 6.9, 17.7 and 19.6 for 235 internal standard, vanillin and vanillyl alcohol, respectively. 236 237 3. Results 238 3.1 Characterization of EUGO 239 Cell free extracts (CFE) containing EUGO activity were provided by InnoSyn B.V. These CFE were 240 characterised regarding protein content (42.9 mg mL-1 CFE) and enzymatic activity (180.0 U mL- 1 241 CFE). Aiming to choose the most suitable and appropriate conditions and supports for the 242 immobilisation of EUGO, the effect of pH on stability was studied (Figure 2Figure 2). EUGO was Con formato: Fuente: Sin Negrita 243 found to be stable over a wide pH range, preserving more than 90% of its initial activity after 24 244 hours between pH 5 and 8 (Figure 2Figure 2). However, at pH’s higher than 8, the stability Con formato: Fuente: Sin Negrita 245 decreases rapidly. These results are well aligned with a previous publication by Q. Nguyen, et al. 246 [32]. 247 248 Figure 2: Influence of pH on EUGO stability. Stability was carried out by incubating 1 U mL-1 EUGO in 100 mM buffers: 249 sodium acetate, pH 5 (●); potassium phosphate, pH 6.0 (○), 7.0 ▼ , 8.0 (△); Tris-HCl, 9.0 (■), and carbonate- 250 bicarbonate, pH 10.0 (□), 25°C. 11 251 252 3.2 Effect of hydrogen peroxide onto EUGO stability 253 Oxidations catalysed by EUGO produce H2O2 (Figure 1Figure 1) in stoichiometric amounts which Con formato: Fuente: Sin Negrita 254 could negatively affect the stability of the enzyme [33]. Therefore, the stability of EUGO was 255 studied under reaction conditions at different concentrations of hydrogen peroxide (0-400mM) 256 to analyse the effect of H2O2 in EUGO performance. The rank of concentrations was selected 257 taking into account that the initial concentration of substrate used in the target reactions is set 258 at 400mM. 259 The obtained results are depicted in Figure 3Figure 3 A. At 10 mM concentration, H2O2 had no Con formato: Fuente: Sin Negrita 260 impact on EUGO stability, whereas, at successively higher concentrations an increasingly 261 negative impact was observed. After 24 hours, less than 50% of initial activity was detected at 262 concentrations higher than 50 mM. Therefore, due to the strong effect of H2O2 on EUGO 263 stability, catalase was used in preparative reactions to convert it to water in-situ. Stability 264 studied in the presence of up to 400 mM of H2O2 and catalase (9 mg mL-1 (35847 U mL-1)) showed 265 no loss in EUGO activity compared to background (Figure 3 B). Therefore, it was decided to 266 perform the oxidation of vanillyl alcohol in the presence of catalase. 12 A 100 B 120 80 100 Activity (%) 60 Activity (%) 80 60 40 40 20 20 0 0 0 5 10 15 20 25 0 5 10 15 20 25 Time (h) Time (h) 267 268 Figure 3: Stability of EUGO under reaction media (30% acetone, 50 mM potassium phosphate pH 7.5, mild 269 agitation conditions, 25°C) at different concentrations of H2O2 of 0 mM (●), 10 mM (○), 25 mM (▼),50 270 mM (Δ), 100 mM (■),200 mM (□) and 400 mM (♦) (A) and with different concentrations of H2O2 (0, 50, 271 100, 200 and 400 mM) and catalase (9 mg ml-1 (35847 U mL-1)) (B) 272 273 3.3 EUGO immobilisation 274 The extensive pH range where EUGO is stable allows the use of different immobilisation 275 methods to attach the enzyme covalently onto different supports. Different supports with 276 different matrices (agarose and methacrylate) and different functional groups (amino and 277 epoxy) were selected for the immobilisation screening (Table 1Table 1). Con formato: Fuente: Sin Negrita 278 Firstly, in order to ensure that the loading of enzyme offered to the supports did not show 279 diffusional limitations, enzyme concentrations of 30, 10, 5 and 2.5 U mL-1 support were selected. 280 Loadings equal or lower than 10 U mL-1 support led to the retained activities showed in table 1. 281 When 10 U mL-1 were used the obtained retained activity were similar:18% (Epoxy-agarose 282 (UAB) F.M.1.; ph 7.5), 35% (Epoxy-agarose (UAB) F.M.1.; pH 8), 38% (Epoxy-agarose (UAB) 283 F.M.2.; pH 8), 29% (Praesto epoxy 45), 44% (Purolite 8204F), 65% (MANA-agarose), 35% (Amino 284 ECR8409) and 21% (Amino ECR8415). Similar values were obtained when 2.5 U mL-1 were 285 immobilised. When 30 U mL-1 were immobilised, retained activity values were lower showing 13 286 the diffusional limitations. Therefore, loadings equal or lower than 10 U/mL did not show 287 diffusional limitations for all the supports studied. This behave could be because enzyme 288 molecules are being placed forming in a monolayer. Then, 10 U mL-1 support was the loading 289 used in all the experiments to characterise the immobilisations. 290 Supports functionalised with amino groups (rows 6, 7, 8) behave as expected for ionic exchange 291 resins. The enzyme initially becomes attached by ionic adsorption and in order to provide 292 covalent linkages between carboxyl groups of the enzyme and amino groups of the support, CDI 293 is used. The best result from this support type was obtained using MANA-agarose, with 100% 294 immobilisation yield and 63% retained activity. 295 Epoxy groups from supports react mainly with amino groups from the enzyme, but also with 296 hydroxyl and thiol groups [34]. Alkaline pH favours this reaction because non protonated NH2- 297 groups are reactive. Firstly, the effect of pH on immobilization was tested with Epoxy-agarose 298 (UAB) F.M.1. Two pHs (7.5 and 8 (rows 1 and 2)) were selected taking into account the stability 299 decrease suffered by the EUGO at higher pHs (Figure 2Figure 2). The best results were obtained Con formato: Fuente: Sin Negrita 300 at pH 8, as expected, reaching 97% and 37% immobilization yield and retained activity 301 respectively. These results represent an increase of 1.5-fold and 2.3-fold, respectively, 302 compared to the experiment performed under pH 7.5. -1 303 Regarding the activation grade, 9, 30 and 80 µmols of epoxy groups g support were tested 304 (row 4, 2 and 3) and the results show that increasing the activation grade from 9 to 30 µmols of 305 epoxy groups g-1 support increase the immobilization yield from 51% to 97% (rows 4 and 2). 306 However, an increase from 30 to 80 did not significantly improve immobilization yield or 307 retained activity (rows 2 and 3). It was also noticed that using a support with similar activation 308 grade but a different matrix did not lead to a variation of the immobilization performance (rows 309 3 and 5). 14 310 Taking into account all the immobilization studies, the best results were obtained with MANA- 311 agarose, Epoxy-agarose (UAB) F.M.2 and Purolite 8204F. They showed 100% immobilisation 312 yield and 63%, 38% and 43% retained activity respectively. 313 These three supports were pre-selected to perform further experiments, in order to select the 314 best immobilised derivative to perform the vanillyl alcohol oxidation reaction. 315 316 Table 1: EUGO immobilisation screening. Nº support U of activation EUGO Support Retained Immobilisation Immobilisation Support grade offered Matrix functional activity pH yield (%) (µmols g-1 mL-1 group (%) support) support 1 Epoxy-agarose ≈ 30 10 agarose epoxy 7.5 64 16 (UAB) F.M. 1 2 Epoxy-agarose ≈ 30 10 agarose epoxy 8 97 37 (UAB) F.M.1 3 Epoxy-agarose ≈ 80 10 agarose epoxy 8 100 38 (UAB) F.M.2 4 Praesto epoxy ≈9 10 agarose epoxy 8 51 35 45 (Purolite) 5 Purolite 8204F ≈ 73 10 Methacrylate epoxy 8 100 43 (purolite) 6 MANA-Agarose 40-60 10 agarose amino 6 100 63 (ABT) 7 Amino ECR8409 unknown 10 Methacrylate amino 6 100 33 (Purolite) 8 Amino ECR8415 unknown 10 Methacrylate amino 6 100 22 (Purolite) 317 318 319 3.4 Stability of soluble and immobilised EUGO under reaction and storage conditions 320 The selected biocatalysts were incubated in reaction media (30% acetone, 50 mM potassium 321 phosphate pH 7.5, mild agitation conditions, 25°C), in order to analyse if immobilisation 322 improved enzyme stability under these conditions compared to the soluble enzyme. Stability 15 323 under storage conditions was also studied (25 mM potassium phosphate pH 6, 4°C). Under both 324 conditions tested, soluble EUGO was found to be very stable, with a half-life of about 11.5 days 325 under simulated reaction conditions. Similar profiles were observed for EUGO immobilised on 326 MANA-agarose (t1/2: 8.3 days) and Purolite 8204F (t1/2: 17.3 days). In contrast, EUGO immobilised 327 on Epoxy-agarose-UAB showed a 6-fold improvement, with a half-life of 77.5 days (Figure Con formato: Fuente: Sin Negrita 328 4Figure 4 A). Under storage conditions (Figure 4 B), both soluble and immobilised forms 329 preserved more than 70% initial activity after 14 days of storage. A 100 B 100 80 80 Activity (%) Activity (%) 60 60 40 40 20 20 0 0 0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16 330 Time (d) Time (d) 331 Figure 4: Stability of EUGO, soluble (○) and immobilised onto MANA-agarose (●), Epoxy-agarose (▲) and Purolite 332 8204F (■) in reaction media (30% acetone, 50 mM Potassium phosphate pH 7.5, 25°C) (A) and in storage conditions 333 (25 mM potassium phosphate pH 6, 4°C) (B). 334 335 3.5 Maximum loading of enzyme that can be immobilised onto the supports 336 The maximum quantity of enzyme that can be attached to each selected support was analysed 337 in order to optimise the concentration of enzyme offered to the support without loosening 338 unnecessary amounts of biocatalyst in the supernatant. Different amounts of enzyme were 339 immobilised onto the selected supports in order to determine this value. Results shown in Table 340 2 determined that EUGO can be immobilised efficiently at high loadings onto MANA-agarose, 341 Epoxy-agarose-UAB and Purolite 8204F. MANA-agarose and Epoxy-agarose-UAB can be loaded 16 342 up to 75 mg of protein mL-1 support. That equates to 442 and 479 U immobilised mL-1 support 343 respectively. At high loadings of enzyme the activity of immobilised derivatives cannot be 344 analysed due to the presence of diffusional limitations. Then, in order to determine the retained 345 activity, it is supposed that the enzyme retains the same percentage of activity as determined 346 during immobilisation at low enzyme loading (without diffusional limitations). Taking into 347 account the retained activity percentages determined in the characterization (i.e. 63% for 348 MANA-agarose and 38% for Epoxy-agarose) the supports can retain up to 279 and 179 U mL-1 of 349 support. Regarding the Purolite 8204F, the methacrylate-based support could be loaded with up 350 to 50 mg of protein mL-1 support, which corresponds to a retained activity of 130 U mL-1 of 351 support. 352 As expected, measured activities of the immobilised derivatives at these loadings (39, 32 and 29 353 U ml-1 support for MANA-agarose, Epoxy-agarose and Purolite 8204F) were lower than the 354 theoretical values, probably due to diffusional limitations. 355 Table 2: Results of maximum loading enzyme study. Measuredactivity Theoretical* Offered mg Theoretical U of the U offered mL- % retained Support protein mL-1 1 immobilised -1 immobilised support Immobilisation activity (U mL support mL-1 support derivative (U mL- support) 1 ) MANA-agarose 75 456 97 442 279 39 Epoxy-agarose 75 509 94 479 179 32 Purolite 8204F 50 340 88 300 131 29 356 *Thereoretical retained activity in the absence of diffusional limitations. 357 3.6 Vanillyl alcohol oxidation using soluble and immobilised EUGO 358 Preparative scale vanillin synthesis reactions were performed on a 10 mL scale with 400 mM 359 vanillyl alcohol, as shown in the experimental section, using 10% v/v of immobilised enzyme 360 onto MANA-agarose (279 U mL-1 of support), Epoxy-agarose (179 U mL-1 of support) and Purolite 361 8204F (131 U mL-1 of support). For comparative purposes the same overall number of units of 362 soluble enzyme to compare the obtained process metrics. All reactions also contained 35847 U 17 363 of catalase mL-1 of reaction media to eliminate the peroxide formed by EUGO. Reactions with 364 28, 18 and 13 U mL-1 of soluble EUGO gave high conversions of 86.2, 85.7 and 83.8% after 30, 41 365 and 37 hours respectively. 366 Reactions with immobilised EUGO gave comparable conversions to that of the free enzyme 367 when using similar overall number of units activity under similar reaction conditions. Similar 368 space-time yields were obtained for both forms of Epoxy-agarose (immobilised and soluble) (1.8 369 and 1.6 g P L-1h-1, respectively) and Purolite 8204F (1.2 and 1.4 g P L-1h-1, respectively) were used 370 as immobilisation support, but a 50% increase was observed with EUGO immobilised into 371 MANA-agarose (2.9 g P L-1h-1). This variation was produced because the reaction time was lower 372 in the reaction with immobilised derivative. This behave maybe could be produced because the 373 immobilised enzyme is protected from the gas interphase which has been demonstrated that 374 can deactivate enzymes [35]. Soluble EUGO gave a space-time yield of 1.9 g P L-1h-1. These high 375 values, compared with other biotechnological ways to produce vanillin [10,12,17,18,22,36], 376 make EUGO a promising enzyme to perform the vanillyl alcohol oxidation at industrial scale. 377 3.7 Reusability of immobilised EUGO for vanillyl alcohol oxidation 378 Immobilised derivatives were each reused over 5 reaction cycles and the results are shown in 379 Table 3Table 3 and Figure 5Figure 5. Purolite 8204F displayed lowest recyclability, retaining less Con formato: Fuente: Sin Negrita Con formato: Fuente: Sin Negrita 380 than 65% conversion and less than 15% yield by the last cycle. EUGO immobilised onto Epoxy- 381 agarose-UAB and MANA-agarose showed better operational stability with conversions higher 382 than 80%, for both derivatives and yields higher than 50% and 30% in the last cycle, respectively. 383 The biocatalyst yield improved more than 2 fold compared to soluble enzyme, with a maximum 384 obtained using the Epoxy-agarose immobilised derivative (7.3 mg vanillin U-1 EUGO). It should 385 be mentioned that the particles were abraded by magnetic stirring during the re-cycles and, 386 therefore, the particle size was reduced during the experiments. This effect can produce a 18 387 decrease on diffusional limitations which would increase the initial reaction rate. However, this 388 effect would be counteracted by the activity decrease due to enzyme stability. 389 Only one study was found in the literature that reported vanillin production with an immobilised 390 enzyme (oxygenase Cso2) [22]. Although a different pathway (from isoeugenol) was used, they 391 obtained 0.68 mg of vanillin per mL over ten cycles. We could produce 45 mg of vanillin mL-1 in 392 5 cycles, when MANA-agarose was used as immobilisation carrier, which represents an increase 393 of 66-fold over the alternative literature procedure. 394 395 Table 3: Process metrics obtained in the vanillyl oxidation using soluble and immobilised EUGO. 400 mM of vanillyl 396 alcohol, 30% (v/v) acetone in 50 mM potassium phosphate buffer at pH 7.5, 25˚C, 1 vvm air (hydrated with a 30% 397 acetone solution), magnetic stirring (500 rpm), and 10 µl antifoam. 9 mg mL-1 reaction (35847 U mL-1 reaction) of 398 catalase. Biocatalyst* -1 yield (mg P U mL Reaction % Total Biocatalyst Form U-1 of reaction time (h) conversion vanillin (g) biocatalyst offered) Soluble 30 86 0.4 1.4 MANA-agarose 28 Immobilised 18 87** 1.1 4.1 Soluble 41 86 0.4 2.3 Epoxy-agarose 18 Immobilised 30 90** 1.3 7.3 Soluble 37 84 0.3 2.7 Purolite 8204F 13 Immobilised 36 76** 0.7 5.6 399 *Determined when the reaction finished. 400 ** Overall conversion. 401 19 402 403 404 Figure 5: Conversion (bars) and Yield (●) of the reacFon cycles MANA-agarose 283.5 U mL-1 support (A), Epoxy agarose 405 182.8 U mL-1 support (B) and Purolite 8204F 124 U mL-1 support (C), and Catalase (35847 U mL-1 reaction). Reaction 406 conditions: 10 mL, 25˚C, magnetic stirring (500 rpm), 1 vvm, 400 mM Vanillyl alcohol/50 mM Potassium phosphate 407 pH 7.5. 408 409 Conclusions 410 Eugenol oxidase is an active and stable enzyme over a wide range of pH. It can be immobilised 411 efficiently into different supports (MANA-agarose, Epoxy-agarose and Purolite 8204F) at high 412 enzyme loadings (279, 179 and 131 U EUGO mL-1 support, respectively). Immobilised derivatives 413 showed high stability under the reaction conditions used with a 6-fold improvement in half-time 20 414 for EUGO immobilised onto Epoxy-agarose compared with the soluble form. High conversions 415 (more than 90%) and space-time yield of 2.9 g vanillin L-1 h-1 could be obtained in the vanillyl 416 alcohol oxidation with soluble and immobilised forms. The immobilised derivatives could be 417 reused satisfactorily for more than 5 cycles, with EUGO immobilised into Epoxy-agarose being 418 the operationally most stable biocatalyst tested, retaining more than 80% of conversion in the 419 last cycle, which represents 90% of retained activity compared to fresh biocatalyst. It presented 420 the highest biocatalyst yield of 7.3 mg vanillin U-1 EUGO which is 3-fold more than the soluble 421 form. 422 423 Acknowledgements 424 The research for this work has received funding from the European Union project ROBOX (grant 425 agreement n° 635734) under EU’s Horizon 2020 Programme Research and Innovation actions 426 H2020-LEIT BIO-2014-1. This document reflects only the author's view and the Agency is not 427 responsible for any use that may be made of the information it contains. The Department of 428 Chemical, Biological and Environmental Engineering of Universitat Autònoma de Barcelona 429 constitutes the Biochemical Engineering Unit of the Reference Network in Biotechnology and 430 the research group 2017 SGR 1462, Generalitat de Catalunya. Authors also thank COST Action 431 CM 1303–Systems Biocatalysis for financial support and Purolite Life Sciences® for the supports 432 given. Miquel García acknowledges also UAB for funding his Ph.D. grant. 433 PS would like to acknowledge the the European Union for his current funding: “This project has 434 received funding from the European Union’s Horizon 2020 research and innovation programme 435 under Marie Skłodowska-Curie grant agreement No 665919”. 436 21 437 4. Bibliography 438 [1] A. Lomascolo, C. Stentelaire, M. Asther, L. Lesage-Meessen, Trends Biotechnol. 17 439 (1999) 282–289. 440 [2] Food and Agriculure Organization-FAO site, (2018). 441 [3] S. Ramachandra Rao, G. Ravishankar, J. Sci. Food Agric. 80 (2000) 289–304. 442 [4] M.M. Bomgardner, Chem. Eng. News (2016) 38–42. 443 [5] A.L. Mathias, A.E. Rodrigues, Holzforschung 49 (1995) 273–278. 444 [6] C. Qu, M. Kaneko, K. Kashimura, K. Tanaka, S. Ozawa, T. Watanabe, ACS Sustain. Chem. 445 Eng. 5 (2017) 11551–11557. 446 [7] J.D.P. Araújo, C.A. Grande, A.E. Rodrigues, Chem. Eng. Res. Des. 88 (2010) 1024–1032. 447 [8] J.D.P. Araújo, C.A. Grande, A.E. Rodrigues, Catal. Today 147 (2009) 330–335. 448 [9] E.A.B. da Silva, M. Zabkova, J.D. Araújo, C.A. Cateto, M.F. Barreiro, M.N. Belgacem, A.E. 449 Rodrigues, Chem. Eng. Res. Des. 87 (2009) 1276–1292. 450 [10] A. Muheim, K. Lerch, Appl. Microbiol. Biotechnol. 51 (1999) 456–461. 451 [11] S.-H. Yoon, C. Li, J.-E. Kim, S.-H. Lee, J.-Y. Yoon, M.-S. Choi, W.-T. Seo, J.-K. Yang, J.-Y. 452 Kim, S.-W. Kim, Biotechnol. Lett. 27 (2005) 1829–1832. 453 [12] S.H. Yoon, C. Li, Y.M. Lee, S.H. Lee, S.H. Kim, M.S. Choi, W.T. Seo, J.K. Yang, J.Y. Kim, 454 S.W. Kim, Biotechnol. Bioprocess Eng. 10 (2005) 378–384. 455 [13] D. Di Gioia, L. Sciubba, L. Setti, F. Luziatelli, M. Ruzzi, D. Zanichelli, F. Fava, Enzyme 456 Microb. Technol. 41 (2007) 498–505. 457 [14] J. Ni, F. Tao, H. Du, P. Xu, Sci. Rep. 5 (2015) 1–12. 458 [15] A. Paz, D. Outeiriño, R. Pinheiro de Souza Oliveira, J.M. Domínguez, N. Biotechnol. 40 22 459 (2018) 186–191. 460 [16] E.H. Hansen, B.L. Møller, G.R. Kock, C.M. Bünner, C. Kristensen, O.R. Jensen, F.T. Okkels, 461 C.E. Olsen, M.S. Motawia, J. Hansen, Appl. Environ. Microbiol. 75 (2009) 2765–2774. 462 [17] L. Zheng, P. Zheng, Z. Sun, Y. Bai, J. Wang, X. Guo, Bioresour. Technol. 98 (2007) 1115– 463 1119. 464 [18] W. Yang, H. Tang, J. Ni, Q. Wu, D. Hua, F. Tao, P. Xu, PLoS One 8 (2013). 465 [19] B. Suresh, G.A. Ravishankar, Plant Physiol. Biochem. 43 (2005) 125–131. 466 [20] J. Jin, H. Mazon, R.H.H. Van Den Heuvel, D.B. Janssen, M.W. Fraaije, FEBS J. 274 (2007) 467 2311–2321. 468 [21] R.H.H. Van Den Heuvel, M.W. Fraaije, C. Laane, W.J.H. van Berkel, J. Agric. Food Chem. 469 49 (2001) 2954–2958. 470 [22] T. Furuya, M. Kuroiwa, K. Kino, J. Biotechnol. 243 (2016) 25–28. 471 [23] E.L.P. Europeo, E.L. Consejo, D.E.L.A. Uni, P. Europeo, P. Europeo, (2008) 34–50. 472 [24] S. Román, L.M. Sánchez-Siles, M. Siegrist, Trends Food Sci. Technol. 67 (2017) 44–57. 473 [25] C. Mateo, J.M. Palomo, G. Fernandez-Lorente, J.M. Guisan, R. Fernandez-Lafuente, 474 Enzyme Microb. Technol. 40 (2007) 1451–1463. 475 [26] U. Guzik, K. Hupert-Kocurek, D. Wojcieszynska, Molecules 19 (2014) 8995–9018. 476 [27] R. Di Cosimo, J. Mc Auliffe, A.J. Poulose, G. Bohlmann, Chem. Soc. Rev. 42 (2013) 6437– 477 6474. 478 [28] M.M. Bradford, Anal. Biochem. 72 (1976) 248–254. 479 [29] U.K. Laemmli, Nature 227 (1970) 680–685. 480 [30] I. Axarli, A. Prigipaki, N.E. Labrou, Biomol. Eng. 22 (2005) 81–88. 23 481 [31] L. Sunberg, J. Porath, J. Chromatogr. 90 (1974) 87–98. 482 [32] Q.T. Nguyen, G. de Gonzalo, C. Binda, A. Rioz-Martínez, A. Mattevi, M.W. Fraaije, 483 ChemBioChem 17 (2016) 1359–1366. 484 [33] K. Hernandez, A. Berenguer-murcia, R.C. Rodrigues, R. Fernandez-lafuente, Curr. Org. 485 Chem. 16 (2012) 2652–2672. 486 [34] E. Katchalski-katzir, D.M. Kraemer, J. Mol. Catal. B Enzym. 10 (2000) 157–176. 487 [35] A.S. Bommarius, A. Karau, Biotechnol. Prog. 21 (2005) 1663–1672. 488 [36] T. Furuya, M. Miura, M. Kuroiwa, K. Kino, N. Biotechnol. 32 (2015) 335–339. 489 24