This is the accepted version of the article: Alcover, Natàlia; Carceller, Albert; Álvaro Campos, Gregorio; [et al.]. «Zymobac- ter palmae pyruvate decarboxylase production process development : Cloning in Escherichia coli, fed-batch culture and purification». Engineering in life science, Vol. 19, Issue 7 (July 2019), p. 502-512. DOI 10.1002/elsc.201900010 This version is avaible at https://ddd.uab.cat/record/217010 under the terms of the license 1 Zymobacter palmae pyruvate decarboxylase production process 2 development: Cloning in Escherichia coli, fed-batch culture and 3 purification 4 Natàlia Alcover1, Albert Carceller1, Gregorio Álvaro1, Marina Guillén1 1 5 Bioprocess Engineering and Applied Biocatalysis Group, Department of Chemical Biological 6 and Environmental Engineering, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain. 7 8 Corresponding author: 9 Marina Guillén 10 Department of Chemical, Biological and Environmental Engineering, Universitat Autònoma de 11 Barcelona, 08193 Bellaterra, Spain 12 Telephone: +34 93 581 4793 13 Fax: +34 93 581 2013 14 marina.guillen@uab.cat 15 16 Keywords: Pyruvate decarboxylase (PDC), Zymobacter palmae, fed-batch culture, PDC 17 purification, high cell density cultures 18 List of abbreviations: PDC, Pyruvate decarboxylase; DCW, Dry cell weight, HCDC, High cell 19 density culture; MANA-agarose, Amino functionalized agarose; LB, Luria-Bertrani medium; DM 20 Defined medium; TES, Trace elements solution. 1 21 Practical Application 22 Pyruvate decarboxylase (PDC) is a very powerful tool in bioprocesses. PDC has an important role 23 in the ethanol synthesis pathway; therefore its overexpression in target microorganisms is of 24 interest for ethanol production. Moreover, PDC can be found in several multienzymatic systems 25 of interest in biocatalysis such as the synthesis of chiral amines in combination with 26 transaminases, where it drives the equilibrium towards formation of the target chiral amine by 27 producing acetaldehyde and CO2 from pyruvate. Despite its importance in biocatalysis, the use 28 of PDC at large scale is hampered by the lack of an efficient production process. The present 29 study develops a recombinant production and a purification process for PDC from Zymobacter 30 palmae expressed in Escherichia coli, broadening its application of use at high scale. This is the 31 first work reporting the evaluation of the production and downstream processing of PDC on a 32 bioreactor scale. 33 34 35 36 37 38 39 40 41 2 42 Abstract 43 Pyruvate decarboxylase (PDC) is responsible for the decarboxylation of pyruvate, producing 44 acetaldehyde and CO2 and is of high interest for industrial applications. PDC is a very powerful 45 tool in the enzymatic synthesis of chiral amines by combining it with transaminases when 46 alanine is used as amine donor. However, one of the main drawbacks that hampers its use in 47 biocatalysis is its production and the downstream processing on scale. 48 In this paper, a production process of PDC from Zymobacter palmae has been developed. The 49 enzyme has been cloned and overexpressed in Escherichia coli. It is presented, for the first time, 50 the evaluation of the production of recombinant PDC in a bench-scale bioreactor, applying a 51 substrate-limiting fed-batch strategy which led to a volumetric productivity and a final PDC 52 specific activity of 6942 U L-1 h-1 and 3677 U gDCW-1. Finally, PDC was purified in FPLC equipment 53 by ion exchange chromatography. The developed purification process resulted in 100% 54 purification yield and a purification factor of 3.8. 55 56 57 58 59 60 61 62 3 63 1 Introduction 64 Pyruvate decarboxylase (PDC) (E.C. 4.1.1.1) is responsible for the decarboxylation of pyruvate 65 producing acetaldehyde and CO2 using Mg+2 and thiamine pyrophosphate (TPP) as cofactors [1]. 66 This enzyme is a key biocatalyst in the fermentative ethanol pathways since the obtained 67 acetaldehyde is subsequently reduced to ethanol by an alcohol dehydrogenase. PDCs are 68 present in eukaryotes including plants, yeast and fungi. However, this enzyme has been found 69 in a few prokaryotes: Gluconobacter oxydans [2], Zymomonas mobilis [3,4], Sarcina ventriculi 70 [5], Zymobacter palmae [6,7] , Acetobacter pasteurianus [8] and Gluconoacetobacter 71 diazotrophicus [9]. 72 PDC is considered to be an enzyme of high interest for industrial applications. This biocatalyst 73 has been widely studied for the production of bioethanol, especially from cellulosic biomass. In 74 that sense, PDC has been cloned and overexpressed in several microorganisms, such as 75 Escherichia coli, Hansenula polymorpha, Bacillus sp and Klebsiella oxytoca, to improve ethanol 76 production [10–15]. Pyruvate decarboxylase has also been applied to the synthesis of 77 pharmaceutical precursors [16,17], compounds for the cosmetic industry [18], or chiral amines 78 and amino acids which play an important role in the pharmaceutical, agrochemical and chemical 79 industries [19]. However, the use of PDC at industrial scale is still hampered. One of the main 80 drawbacks that hinder the use of enzymes, including PDC, in large-scale industrial processes is 81 the price of the biocatalyst. The production cost of enzymes, including their subsequent 82 purification, is one of the main contributions to the final price of the enzyme [20–22]. 83 Approximately 90% of all industrial enzymes are produced by recombinant microorganisms, 84 especially for high-volume, low-value enzymes [21]. The use of Escherichia coli as a host for 85 recombinant protein production is the most common strategy for process development [23– 86 25]. Thus, Gram negative bacteria remain the preferred system for laboratory investigations and 87 initial development in commercial activities and is a benchmark for comparison with other 4 88 expression platforms [25,26]. This is due to different factors such as ease of genetic 89 manipulation, high specific growth rates and availability of optimized expression plasmids. High 90 cell density culture (HCDC) techniques for culturing E. coli have been developed to improve 91 productivity, and also to provide advantages such as reduced culture volume, enhanced 92 downstream processing, reduced wastewater, lower production costs and reduced investment 93 in equipment [27]. High cell density cultures can be obtained by fed-batch operational 94 strategies, where the use of exponential feeding has been widely used as an efficient strategy 95 to maintain a desired specific growth rate before the induction of protein expression [28,29]. 96 In the present paper the bacterial PDC from Z. palmae was cloned and expressed in E. coli under 97 the lac promoter, inducible by IPTG. The best-performing clone was selected according to results 98 obtained in Erlenmeyer-scale cultures using rich and defined medium. The selected clone was 99 then tested for the production of the enzyme at HCDC, applying a substrate-limiting fed-batch 100 strategy in a 2L-bioreactor. In addition, an efficient purification process was subsequently 101 studied and developed, aiming to obtain a process for PDC production including both, upstream 102 and downstream steps. 103 2 Material and methods 104 2.1 Reagents 105 All reagents were purchased from Sigma Aldrich® (St. Louis, MO, USA) and were of analytical 106 grade if not stated elsewhere. 107 Amino functionalized agarose (MANA-agarose) was purchased from Agarose Bead 108 Technologies® (ABT®, Madrid, Spain). 109 2.2 Bacterial strains and plasmids 110 A pJAM3440 plasmid containing the Zymobacter palmae PDC gene was kindly donated by the 111 Microbiology and Cell Science Department of the University of Florida (Gaisnesville, FL, USA). 5 112 For protein overproduction, the gene was cloned into a pQE-40 vector and transformed into two 113 E. coli strains, M15[pREP4] and SG13009[pREP4]. M15 and SG13009 strains were derived from 114 E. coli K12 and have the phenotype NaIS, StrS, RifS, Thi–, Lac–, Ara+, Gal+, Mtl–, F–, RecA+, Uvr+, 115 Lon+. Both pQE-40 and the strains were obtained from a QIAexpress type IV vector kit from 116 Qiagen® (Venlo, Netherlands). 117 2.3 PDC cloning in E. coli 118 Plasmid pJAM3440 containing the PDC gene was isolated from the donated E. coli strain using a 119 PureYield™ Plasmid Miniprep System from Promega® (Madison, WI, USA) according to the 120 manufacturer's instructions. The PDC gene was then amplified by PCR using a Phusion 121 High-Fidelity DNA Polymerase from Thermo Scientific® (Waltham, MA, USA) and two primers 122 designed with the restriction sequences of BamHI and SalI inserted. Double digestion of the PDC 123 gene and pQE-40 was performed with the mentioned restriction enzymes, followed by an 124 overnight ligation at 16oC. After cleaning with Wizard® SV Gel and PCR Clean-Up System 125 (Promega®), the new plasmid construction was transformed to M15[pREP4] and 126 SG13009[pREP4] electrocompetent cells with a Gene Pulser II electroporator from Bio-Rad® 127 (Hercules, CA, USA) (V= 2500 v; C= 25 µF; R= 200 Ω). Transformed clones were selected using 128 LB-agar plates supplemented with ampicillin 100 mg L-1 and kanamycin 0.05 g L-1. 129 Transformations were confirmed by colony-PCR using GoTaq® Green Master Mix (Promega®). 130 2.4 Media composition 131 Luria-Bertrani (LB) medium, containing 10 g L-1 tryptone, 5 g L-1 yeast extract and 10 g L-1 NaCl, 132 was used for pre-inoculum preparations and preliminary production studies in shake-flasks. 133 A defined medium (DM) with glucose as the main carbon source was used for shake flasks and 134 bioreactor cultivations. For shake flask inoculums, the media contained 5 g L-1 glucose, 13.2 g L- 1 135 K2HPO4, 2.6 g L-1 KH2PO4, 2 g L-1 NaCl, 4.1 g L-1 (NH4)2SO4, 0.5 g L-1 MgSO4·7H2O, 0.026 g L-1 FeCl3, 136 0.005 g L-1 CaCl2·2H2O, 0.1 g L-1 thiamine, 0.033 g L-1 citrate, 0.1 g L-1 ampicillin, 0.05 g L-1 137 kanamycin and 3 mL L-1 of trace elements solution (TES) (TES: 0.04 g L-1 AlCl3, 1.74 g L-1 6 138 ZnSO4·7H2O, 0.16 g L-1 CoCl2·6H2O, 1.55 g L-1 CuSO4·H2O, 0.01 g L-1 H3BO3, 1.42 g L-1 MnCl2·4H2O, 139 0.01 g L-1 NiCl2·6H2O, 0.02 g L-1 Na2MoO4). The same composition was used for the batch phase 140 cultivation in the bioreactor, but with an increased glucose concentration of 25 g L-1. 141 Feeding medium for fed-batch phase contained 425 g L-1 glucose, 11.6 g L-1 MgSO4·7H2O, 0.6 g L- 1 142 FeCl3, 0.35 g L-1 thiamine, 0.09 g L-1 CaCl2·2H2O, 0.1 g L-1 ampicillin, 0.05 g L-1 kanamycin and 143 76.6 mL L-1 of TES. 144 Stock solutions of ampicillin and kanamycin were prepared in a concentration of 10 mg mL-1 and 145 50 mg mL-1 respectively and stored at -20oC. IPTG stock was prepared at 100 mM concentration 146 and stored at -20oC. 147 Vitamins, antibiotics, TES FeCl3, CaCl2·2H2O and IPTG were sterilized by filtration (0.2 μm). 148 Glucose and saline solutions were separately sterilized by autoclaving at 121oC for 30 minutes. 149 2.5 Cultivation conditions 150 Pre-inoculum 151 From cryostocks stored at -80oC, pre-inoculums were prepared in 15 mL of LB medium 152 containing 0.1 mg L-1 ampicillin and 0.05 mg L-1 kanamycin. The growth was performed in Falcon 153 tubes overnight at 37oC and 150 rpm. 154 Shake flask cultures in LB and define medium (DM) 155 Around 5 mL of pre-inoculum was transferred to shake flasks containing 100 mL of LB or defined 156 medium and 0.1 mg L-1 ampicillin and 0.05 mg L-1 kanamycin in order to obtain an initial optical 157 density OD600 of 0.1. Cultures were maintained at 37ºC and 150 rpm until a OD600 of 1 was 158 reached. Then, protein production was induced by the addition of IPTG to a final concentration 159 of 1 mM. 160 Inoculum preparation for bioreactor cultures 161 Inoculums were prepared following the same procedure as described for shake flasks cultures 162 in 100 mL defined medium using 10 g L-1 of glucose. The cultures were kept at 37oC and 150 rpm 163 until an OD600 of around 1.2 was reached. 7 164 Fed-batch cultures in bioreactor 165 Substrate limiting fed-batch cultures were performed in an Applikon ez-Control (Applikon 166 Biotechnology®, Delft, Netherlands) equipped with a 2 L vessel. The first phase of the batch was 167 performed by transferring 80 mL of inoculum to 720 mL defined medium. The temperature was 168 maintained at 37oC and the pH at 7.00 by the addition of 15% NH4OH. Oxygen saturation was 169 maintained at 60% by supplying air at a flow rate of 1.5 L min-1 and with a cascade agitation from 170 450 to 1150 rpm. When the cascade was not enough to maintain that pO2 value, pure oxygen 171 was added with a maximum flow rate of 0.5 L min-1. 172 The substrate limiting fed-batch phase was started when glucose was totally exhausted. An 173 exponential addition of feeding medium took place to maintain a specific growth rate of 0.225 h- 1 174 using an automatic microburete of 2.5 mL capacity. Furthermore, every 30 OD increase, 5 mL 175 of a concentrated phosphate solution (500 g L-1 K2HPO4 and 100 g L-1 KH2PO4) was injected into 176 the bioreactor. When OD600nm reached a value of approximately 70, IPTG was injected to a final 177 concentration of 100 μM in order to induce protein production. Fermentation ended when a 178 glucose accumulation was detected. Fed-batch cultures were carried out in duplicate. The 179 calculated error indicates the confidence interval with a confidential level of 90%. 180 2.6 Analytical methods 181 Monitoring bacterial growth 182 Optical density measurements at 600 nm (OD600) were performed to determine cell 183 concentration using a HACH D3900 (Hach®, Loveland, CO, USA) spectrophotometer. Biomass 184 concentration expressed as Dry Cell Weight (DCW) was calculated considering that 1 OD600 is 185 equivalent to 0.3 gDCW L-1 [30]. Analyses were carried out in duplicate. 186 Glucose analysis 187 Biomass was removed from 1ml of broth sample by centrifugation and filtration (0.45 μm). The 188 obtained supernatant was then used for the glucose analysis using an enzymatic analysis 8 189 performed on an YSI 20170 system (Yellow Spring System). Analyses were carried out in 190 duplicate. 191 Total protein content 192 Culture samples were adjusted to an OD600 of 4 and centrifuged at 10,000 g during 10 minutes 193 at 4oC. The obtained pellet was then re-suspended in lysis buffer (200 mM citrate buffer (pH 194 6.2), 1mM of MgCl2 and 1mM TPP). Cell suspensions were kept on ice and sonicated using a 195 Vibracell ® model VC50 (Sonic & Materials, Newton, CT, USA) with four times 15 second pulses 196 and 2 min intervals on ice between each pulse. The lysate was then centrifuged at 10,000 g 197 during 10 minutes and the clear supernatant was used for protein content analysis. Total 198 intracellular protein content present in cell lysates was determined with the Bradford Method 199 using a Coomassie Protein Assay Reagent Kit (Thermo Scientific®) and Bovine Serum Albumin 200 (BSA) as standard. The assays were performed in 96-microwell plates and Thermo Scentific® 201 Multiskan FC equipment was used for the absorbance reading. Analyses were carried out in 202 duplicate. 203 SDS-Page electrophoresis 204 To determine the percentage of enzyme among the rest of intracellular soluble proteins present 205 in the lysates, a NuPAGE electrophoresis system (Invitrogen®, Carlsbad, CA, USA) was used. 206 Lysates were obtained as previously described. A volume of 10µL sample was mixed with 5µL 207 NuPAGETM LDS Sample Buffer (4X), 2µL NuPAGETM Reducing Agent (10X) and 3µL deionized 208 water. After 10 minutes incubation at 70oC samples were charged to NuPAGE 12% Bis-Tris 209 electrophoresis gel and run using MES-SDS as running buffer at 200 V during 40 minutes. The 210 resultant gel was fixed with a solution containing 40% v v-1 and 10% v v-1 acetic acid in water and 211 stained with Bio-SafeTM Coomassie Stain (BIO-RAD). Pictures were taken with a Gel Doc EZ 212 Imaging System (Bio-Rad®) and analyzed with Image LabTM 6.0 Software (Bio-Rad®) 213 PDC enzymatic activity assay 9 214 PDC activity present in lysates was determined by coupling the pyruvate decarboxylation with 215 alcohol dehydrogenase (ADH) and following NADH oxidation at 340 nm (ƐNADPH = 6.22 mM-1 cm- 1 216 ) and 25oC with a SPECORD® 200 PLUS (Analytik jena®, Jena, Germany) spectrophotometer. 217 Lysates were obtained as described in the previous sections. The reaction mixture contained 218 33 mM sodium pyruvate, 0.11 mM NADH, 3.5 U mL-1 ADH from Saccharomyces cerevisiae 219 (Sigma-Aldrich®), 0.1 mM TPP and 0.1 mM MgCl2 in citrate buffer 200 mM and pH 6. One unit of 220 enzyme activity corresponds to the amount of pyruvate decarboxylase that converts 1 µmole of 221 pyruvate to acetaldehyde per minute. Analyses were carried out in duplicate. 222 2.7 Enzyme purification 223 Cell disruption 224 Biomass samples (3 g) were suspended in lysis buffer (15 mL): 200 mM citrate buffer (pH 6.2), 225 1mM of MgCl2 and 1mM TPP. Cell suspensions were kept on ice and sonicated using Vibracell ® 226 model VC50 (Sonic & Materials®, Newton, CT, USA) with four times 15 second pulses and 2 min 227 intervals on ice between each pulse. The crude cell lysate was then centrifuge at 10,000 g during 228 10 minutes at 4oC and cell debris was rejected. 229 PDC purification by sequential precipitation with ammonium sulphate 230 Ammonium sulphate precipitation was carried out in two steps, using a solution of saturated 231 (NH4)2SO4 (4.1 M at 25oC). In the first step of purification, (NH4)2SO4 was added into a cell lysate 232 in order to promote protein precipitation while keeping the PDC soluble. For that purpose, 233 different concentrations of salt were tested. The mixture was incubated for at least 30 minutes 234 at 25oC or 5oC. The precipitated protein was separated from the supernatant by centrifugation 235 at 10,000 g for 5 minutes. In the second step, additional (NH4)2SO4 was added to the resulting 236 supernatant in order to increase salt concentration that promotes pyruvate decarboxylase 237 precipitation; the mixture was incubated for 30 minutes at 5oC. The precipitated enzyme was 238 separated from the supernatant by centrifugation at 10,000 g for 5 minutes. All fractions were 10 239 sampled to analyse enzyme activity and protein concentration. Samples corresponding to 240 precipitates were re-suspended in lysis buffer before the analysis. 241 Enzyme purification by ionic adsorption using MANA-agarose 242 Ionic adsorption was carried out first in batch mode, where 1.07 grams of MANA-agarose 243 support was mixed with 1 mL of cell lysate and adjusted to 20 U mL of support-1 using 8 mL of 244 phosphate or acetate buffer (at pH 6.2 and 5, respectively) containing 1 mM MgCl2 and 1 mM 245 TPP (Thiamine pyrophosphate). The suspension was incubated at 25oC for 30 minutes with mild 246 agitation and 100 µL aliquots of the supernatant and the suspension were taken every 15 247 minutes. After incubation, the supernatant was separated from the support by filtration, the 248 support was washed with 9 mL of the buffer as described above and a sample was taken from 249 the combined supernatants and analysed as above. Finally, elution of the adsorbed proteins was 250 done by adding 9 mL of the buffer containing 1 M NaCl and incubating at 25oC for 1 hour with 251 mild agitation. Supernatant was recovered by filtration and a sample of 100 µL was taken. In 252 parallel, a blank was performed under the same conditions, except for the MANA-agarose 253 support that was substituted by distilled water. All the samples were used to analyse enzyme 254 activity, protein concentration and PDC content by SDS-Page electrophoresis. 255 Using Fast Protein Liquid Chromatography (FPLC) performed in an ÄKTA Pure (GE Healthcare®, 256 Chicago, IL, USA), 0.5 mL of cell lysate was applied at 0.1 mL min-1 to a MANA-agarose column 257 (packed with 1.16 mL of support), equilibrated with buffer A (50mM acetate buffer at pH 5). The 258 column was washed at 0.8 mL min-1 with buffer A during 5-15 column volumes. In the elution 259 step, a linear gradient of buffer B (50 mM acetate buffer, 1 M NaCl, at pH 5) from 0 to 100% 260 during 10 column volumes or a step gradient (first step at 31% of buffer B and second step at 261 100% buffer B during 10 column volumes each step) was applied at 0.8 mL min-1. Within each 262 chromatographic procedure, fractions of 1 mL were collected and analysed for enzyme activity, 263 protein concentration and SDS-Page electrophoresis. 11 264 Desalting 265 Purified PDC fractions, obtained after MANA-agarose chromatography with step gradient, were 266 pooled and applied at a rate of 10 mL min-1 to a XK 26/20 (GE Healthcare®) column packed with 267 60 mL of Sephadex 10 and equilibrated with lysis buffer (200 mM citrate at pH 6.2 with 1 mM 268 MgCl2 and 1 mM TPP). After sample application, lysis buffer was applied to the column at a rate 269 of 10 mL min-1 during 2 column volumes. Fractions of 10 mL each were obtained and analysed 270 for enzyme activity and protein concentration. 271 3 Results and discussion 272 3.1 Shake-flask preliminary experiments 273 Firstly, clones of E. coli strains M15[pREP4] and SG13009[pREP4] expressing PDC were prepared 274 following the procedure described in the Methods section. Several transformed clones were 275 obtained and two of them, one from each E. coli strain, were selected. The colony PCR confirmed 276 that both selected clones contained the PDC gene. 277 Aiming to compare the capability of both clones to grow and to produce PDC, preliminary 278 cultures were performed in shake flasks with 100 mL LB medium supplemented with the 279 corresponding antibiotics (Figure 1). Both in the case of M15[pREP4][pQE-PDC] and 280 SG13009[pREP4][pQE-PDC], growth took place with a maximum specific growth rate of around 281 0.9 h-1 and a final biomass concentration of about 0.65 gDCW L-1 was achieved. During the 282 exponential growth phase an early pulse of IPTG, to a final concentration of 1 mM, was applied 283 at approximately 0.3 gDCW L-1 in order to induce PDC expression. In both cases, 2 hours after 284 induction, a maximum activity of around 5000 U gDCW-1 was achieved and the culture contained 285 a total protein concentration of about 230 mg protein gDCW-1. No PDC activity or PDC protein 286 were detected in the samples taken before induction (Figure 2), thus indicating that the PDC 287 promoter is strongly repressed by the pREP4 plasmid in both strains. Therefore, using LB rich 12 288 media no significant differences were detected between M15[pREP4][pQE-PDC] and 289 SG13009[pREP4][pQE-PDC] during cell growth or PDC expression. 290 One of the main objectives of this work was to prove that PDC can be produced in E. coli in a 291 bioreactor using a substrate-limiting fed-batch strategy. Therefore, the performance of each 292 strain was also studied in a defined medium which allows economical process scale-up. Under 293 these conditions, M15[pREP4][pQE-PDC] and SG13009[pREP4][pQE-PDC] showed similar 294 maximum specific growth rate and final biomass concentration in shake flask cultures (Table 1). 295 Nevertheless, PDC specific activity obtained with the SG13009[pREP4][pQE-PDC] strain was 1.7- 296 fold higher compared to that of M15[pREP4][pQE-PDC] (Table 1). According to these results, 297 SG13009[pREP4][pQE-PDC] was selected for further scale-up experiments. 298 3.2 PDC production in bioreactor cultures using a substrate limiting fed-batch strategy 299 The selected strain SG13009[pREP4][pQE-PDC] was used for PDC production using a substrate 300 limiting fed-batch strategy (see Methods section). Figure 3 shows glucose, DCW, and specific 301 PDC activity profile against time. Glucose was completely depleted during the batch phase after 302 11 hours of culture, reaching a final biomass concentration of 9.32 gDCW L-1. During this batch 303 phase the yield was 0.37 gDCW g glucose-1 and the maximum specific growth rate reached 0.51 304 h-1. The substrate limiting fed-batch phase was then started to maintain a constant growth rate 305 of 0.22 h-1. Once biomass reached 20 g L-1 (3.5 hours after initiation of the fed-batch phase), 306 induction was performed adding a pulse of IPTG to a final concentration in the bioreactor of 100 307 µM. The culture was stopped after 4 hours post-induction since the glucose started to 308 accumulate in the media, probably due to cellular metabolic stress [31–34]. The maximum 309 biomass concentration and PDC specific activity were 34 gDCW L-1 and 3677 U gDCW-1, 310 respectively. Enzyme volumetric productivity was also calculated, achieving 6942 U L-1 h-1. 311 Protein analysis revealed a total protein concentration and of 432 mg gDCW-1 with 30% being 312 PDC (128 mg PDC gDCW-1) which means a quality of 29 U mg PDC-1. Similar values have been 313 obtained previously by the research group applying the substrate limiting fed-batch strategy 13 314 with IPTG pulse induction. The production of rhamnulose-1-phosphate alsolase (RhuA) 315 expressed in E. coli M15∆GlyA [pREP4] following the aforementioned strategy led to 160 316 mgRhuA gDCW-1 and 25.5 gDCW L-1 [32]. 317 3.3 PDC purification by sequential purification with ammonium sulphate 318 Fractional purification with ammonium sulphate is based on the different solubilities of the 319 proteins present in the lysate [35,36]. The process is based on the salting out effect produced 320 by the (NH4)2SO4. The strategy is based on a two-stage process. In the first step, the salt 321 concentration is increased up to a value that leads to the maximum precipitation of lysate 322 proteins whilst maintaining PDC soluble. Thus, contaminant proteins can be removed by simple 323 filtration or centrifugation. In the second stage, the salt concentration is increased until PDC 324 precipitates. 325 Firstly, different preliminary experiments were carried out at two different temperatures (5oC 326 and 25oC) to determine the best condition to carry out the purification. The results showed that 327 enzymatic deactivation occurred under at temperatures. However, at 25ºC the total recovered 328 activity (supernatant and precipitate) was 58%, whilst at 5ºC it was 15% higher (73 %). . Thus, 329 5oC was selected as the temperature to perform the purification study. 330 Regarding the first precipitation step, saturated concentrations of 10, 20, 40 and 50% w/v of 331 (NH4)2SO4 were studied. The optimum in terms of final specific activity was obtained using 40% 332 (NH4)2SO4 concentration, containing 85% of the total final recovered activity (supernatant and 333 precipitate) and representing a 61% of the total initial activity of the lysate. However, the specific 334 activity was not increased compared to the lysate due to the activity loss suffered during the 335 process. Aiming to reduce the enzyme deactivation during the first precipitation step at 40% w/v 336 of (NH4)2SO4 concentration, the incubation time was reduced from 2 h to 0.5 h. Results showed 337 that the final specific activity reached 12.0 U mg protein-1; corresponding to a purification factor 338 of 1.5 (Table 2). The resultant supernatant contained 91% of the total final recovered activity 14 339 (supernatant and precipitate) and represented 82% of the total initial activity of the lysate (Table 340 2). Therefore, the obtained results demonstrate that the precipitation process is improved when 341 incubation time is reduced from 2 to 0.5 hours due to a reduction in enzyme deactivation and 342 treatment with a40% w/v of (NH4)2SO4 concentration at 5oC for 0.5 h was optimum for the first 343 precipitation step. 344 Regarding the second purification step (Table 2), (NH4)2SO4 concentrations of 65% w/v and 70% 345 w/v were studied. The highest specific activity was obtained with 65% w/v of salt, reaching 15.7 346 U mg protein-1, corresponding to a purification factor of 2. It represented a 97% of the total final 347 recovered activity (supernatant and precipitate) and 79% of the total initial activity of the lysate. 348 The use of 70% w/v salt concentration led to a 25% decrease in the purification factor. Therefore, 349 65% w/v salt concentration was selected as the best option for the second precipitation step. 350 In summary, sequential purification with the (NH4)2SO4 strategy, applying a two-step process of 351 40% w/v and 65% w/v of (NH4)2SO4 leads to a final purification factor of 2 and a purification yield 352 of 79%. 353 3.4 PDC purification by ionic adsorption using MANA-agarose 354 Ion exchange purification is based on the interaction that takes place due to the difference in 355 charge of protein and functional groups in the chromatography matrix. Ion exchange is one of 356 the most used chromatographic techniques for the separation and purification of proteins, 357 nucleic acids, polypeptides, polynucleotides, among others. The reasons for the success of ion 358 exchange are its widespread applicability, its high resolving power, its high capacity, and the 359 simplicity and controllability of the method [37]. 360 Monoaminoethyl-N-aminoethyl agarose (MANA-agarose) is a widely used matrix for protein 361 purification by ionic adsorption. This matrix is positively charged at acidic pH leading to an ionic 362 adsorption of those proteins that are negatively charged under the process conditions. PDC 363 purification was performed using MANA-agarose under two different pH’s (6.2 and 5) in a batch 15 364 strategy. Even though at pH 6.2 the support is positively charged, according to the pKa of the 365 matrix (6.7) and the PDC is negatively charged, according to the pI (4.9), it was hypothesized that 366 a lower pH could increase the amount of positively charged proteins in the lysate, thus reducing 367 the competition among PDC and other contaminant proteins to be adsorbed on the support. 368 The results showed a higher final specific activity when the process was performed under the 369 most acidic pH, reaching 23.5 U mg protein-1, corresponding to a purification factor of 4 which 370 represents a 56% increase, compared with the process performed at pH 6.2 (Table 3). It should 371 be mentioned that in both cases, 100% of total initial activity was recovered after desorption 372 treatment (1M NaCl), indicating that no enzyme deactivation had occurred during the processes, 373 giving a 100% purification yield. 374 The results obtained with the ion exchange strategy showed a 2-fold improvement in the 375 purification factor and a 21% improvement of the purification yield compared to the sequential 376 ammonium sulphate precipitation strategy. Therefore, it was decided to optimize and 377 automatize the process using an FPLC. The objective of this study was to find the optimum salt 378 concentration for the elution process. Firstly, the elution step (see Methods section) was 379 performed by using a linear gradient from 0 to 100% buffer B (NaCl 1 M) during 10 column 380 volumes and the results are shown in Figure 4. Two main protein fractions were eluted whose 381 maximum absorbance was detected at 31.03 mS cm-1 (33% buffer B; 0.33 M NaCl) and 57.8 mS 382 cm-1 (68% buffer B; 0.68 M NaCl) eluent conductivity. The samples recovered in the first peak 383 mainly contained contaminant proteins since the activity was only a 1.6 % of the total initial 384 loaded activity with a final specific activity of 0.94 U mg protein-1. The activity assay of the sample 385 from the second peak revealed that almost 100% of the total initial loaded activity was present 386 with a specific activity of 39.22 U mg protein-1. Protein electrophoresis also corroborated the 387 results obtained by the activity assays (Figure 5), thus indicating that PDC was eluted during the 388 second main fraction corresponding to an eluent conductivity of 57.8 mS cm-1 (68% buffer B; 389 0.68 M NaCl). Using this strategy, a purification factor of 3.4 and a purification yield of 100% was 16 390 obtained. The decrease in purification factor from 4 to 3.4 compared to the batch strategy could 391 be due to a slight overlap between the two main fractions obtained in the elution step. 392 Aiming to improve the purification process by eliminating peak overlap, a new approach was 393 applied based on a two-stage elution step. In the first elution stage, a 33% buffer B (0.33 M 394 NaCl), corresponding to a conductivity value of 31.03 mS cm-1 was maintained constant until the 395 contaminant proteins were eluted. In the second stage, once all contaminant proteins had been 396 eluted, the composition of the eluent was changed to 68% of buffer B (0.68 M NaCl), 397 corresponding to a conductivity value of 57.8 mS cm-1, keeping it constant until all PDC was 398 eluted. In figure 6 the enzymatic activity and protein concentrations achieved during the 399 purification process are depicted. As can be seen, during the first stage of the elution (31.03 400 mS·cm-1), no PDC activity was detected, thus indicating that only contaminant proteins were 401 eluted. In the second stage it can be seen that PDC activity was detected, corresponding to PDC 402 elution. The obtained purified protein presented a specific activity of 42.95 U mg protein-1, 403 corresponding to a purification factor of 3.8 and a 100% yield during this ionic exchange 404 purification. The purification factor obtained with the optimized FPLC strategy using a two-stage 405 elution represented a 2-fold improvement compared to the sequential purification with 406 (NH4)2SO4. Other authors have also reported a purification factor of 4 for the Z. palmae PDC 407 expressed in E. coli using a two- step downstream process based on ion exchange 408 chromatography and size-exclusion chromatography [6]. Other bacterial PDCs expressed in E. 409 coli have also been purified by other authors. PDCs from Acetobacter pasteurianus, Zymomonas 410 mobilis and Sarcina ventriculi were purified by ion exchange and size-exclusion chromatography, 411 reaching a purification factor of 14, 12 and 6, respectively [6]. Studies on native bacterial PDC 412 purification have also been reported by other authors. Lowe et al. reached a purification factor 413 of 8 for native PDC from S. ventriculi by applying a three step process based on (NH4)2SO4 414 precipitation, ion exchange chromatography and size exclusion chromatography [5]. Native PDC 415 from Acetobacter pasteurianus was purified up to 110-fold by ionic exchange and 17 416 hydroxyapatite chromatography while native PDC from Z. mobilis has also been purified up to 417 447-fold by dye-ligand interaction and ion exchange chromatography [6, 38]. 418 3.5 PDC Purification: desalting step 419 In order to get a PDC preparation ready for storage before utilization, a last step of desalting 420 was performed by using the FPLC. After the PDC purification by ionic exchange with MANA- 421 agarose using a two-stage elution step process, the eluted peak 2 containing the PDC (52,3 U) 422 was desalted in order to remove NaCl. After sample application to the column, packed with 423 Sephadex 10, desalting was performed with lysis buffer (200 mM citrate, pH 6.2, containing 1 424 mM MgCL2 and 1 mM TPP) and the eluted peak corresponding to desalted PDC contained 54 U 425 of PDC activity. Thus, a 100% yield was obtained for the desalting step. 426 4 Concluding remarks 427 A production process for recombinant PDC from Zymobacter palmae expressed in E. coli has 428 been developed. The PDC was cloned and overexpressed reaching up to 30% of the total protein. 429 In this work it is presented for the first time the production of recombinant PDC in a bench-scale 430 bioreactor applying a substrate-limited fed-batch strategy which lead to volumetric productivity 431 and a final PDC specific activity of 6942 U L-1 h-1 and 3677 U gDCW-1 , respectively. 432 A purification process has also been developed by comparing two strategies: sequential 433 purification by (NH4)2SO4 and ion exchange purification using MANA-agarose. The last one 434 resulted in the best purification factor and purification yield, reaching values of 3.8 and 100% 435 respectively, after process optimization by FPLC. Finally, the eluted PDC from ion exchange 436 chromatography was desalted in order to remove NaCl and a 100% yield was obtained. 437 Thus, a complete PDC production process, including cloning in E. coli, production in a bench- 438 scale bioreactor (upstream) and purification by ion exchange and desalting (downstream), was 439 developed to achieve a PDC preparation ready for storage. 18 440 Acknowledgments 441 This work was supported by the Spanish MINECO (project No. CTQ2014-53114R), co-financed 442 by European Regional Development Fund. The Department of Chemical, Biological and 443 Environmental Engineering of UAB constitutes the Biochemical Engineering Unit of the 444 Reference Network in Biotechnology, and the research group 2017-SGR-1462, Generalitat de 445 Catalunya. Financial support from Spanish MINECO for the PhD scholarship (FPI: BES-2015- 446 072542) of N.A. Zymobacter palmae PDC gene was kindly donated by the Microbiology and Cell 447 Science Department of the University of Florida (USA). 448 Conflicts of Interest 449 The authors declare no financial or commercial conflict of interest. 450 References 451 [1] Pohl, M., Grötzinger, J., Wollmer, A., Kul,a M -R., Reversible Dissociation and Unfolding 452 of Pyruvate Decarboxylase from Zymomonas mobilis. Eur. J. Biochem. 1994,61,224:651. 453 [2] Van Zyl, L.J., Taylor, M.P., Eley, K., Tuffin, M., Cowan, D. a., Engineering pyruvate 454 decarboxylase-mediated ethanol production in the thermophilic host Geobacillus 455 thermoglucosidasius. Appl. Microbiol. Biotechnol. 2014,98,1247–59. 456 [3] Dobritzsch, D., KÃ\Pnig, S., Schneider, G., Lu, G., High resolution crystal structure of 457 pyruvate decarboxylase from Zymomonas mobilis. J. Biol. Chem. 1998,273,20196–204. 458 [4] Candy, J.M., Duggleby, R.G., Structure and properties of pyruvate decarboxylase and site- 459 directed mutagenesis of the Zymomonas mobilis enzyme. Biochim. Biophys. Acta - 460 Protein Struct. Mol. Enzymol. 1998,1385,323–38. 461 [5] Lowe, S.E., Zeikus, J.G., Purification and characterization of pyruvate decarboxylase from 462 Sarcina ventriculi. J. Gen. Microbiol. 1992,138,803–807. 463 [6] Raj, K.C., Talarico, L. a., Ingram, L.O., Maupin-furlow, J. a., Cloning and Characterization 464 of the Zymobacter palmae pyruvate decarboxylase gene (pdc) and comparison to 465 bacterial homologues. Microbiology. 2002,68,2869–2876. 466 [7] Buddrus, L., Andrews, E.S.V., Leak, D.J., Danson, M.J., Arcus, V.L., Crennell, S.J., Crystal 467 structure of pyruvate decarboxylase from Zymobacter palmae. Acta Crystallogr. Sect. 468 Struct. Biol. Commun. 2016,72,700–706. 19 469 [8] Chandra Raj, K., Ingram, L.O., Maupin-Furlow J. a., Pyruvate decarboxylase: A key enzyme 470 for the oxidative metabolism of lactic acid by Acetobacter pasteurianus. Arch. Microbiol. 471 2001,176,443–51. 472 [9] Van Zyl, L.J., Schubert, W.D., Tuffin, M.I., Cowan, D. a., Structure and functional 473 characterization of pyruvate decarboxylase from Gluconacetobacter diazotrophicus. 474 BMC Struct. Biol. 2014,14,1–13. 475 [10] Ohta, K., Beall, D.S., Mejia, J.P., Shanmugam, K.T., Ingram, L., Genetic Improvement of 476 Escherichia coli for Ethanol Production : Chromosomal Integration of Zymomonas mobilis 477 Genes Encoding Pyruvate Decarboxylase and Alcohol Dehydrogenase I. Appl. Environ. 478 Microb. 1991,57,893–900. 479 [11] Lewicka, A.J., Lyczakowski, J.J., Blackhurst, G., Pashkuleva, C., Rothschild-Mancinelli, K., 480 Tautvaišas, D., et al., Fusion of pyruvate decarboxylase and alcohol dehydrogenase 481 increases ethanol production in escherichia coli. ACS Synth. Biol. 2014,3,976–978. 482 [12] Yang, M., Li, X., Bu, C., Wang, H., Shi, G., Yang, X., et al., Pyruvate decarboxylase and 483 alcohol dehydrogenase overexpression in Escherichia coli resulted in high ethanol 484 production and rewired metabolic enzyme networks. World J. Microbiol. Biotechnol. 485 2014,30,2871–2883. 486 [13] Ishchuk, O.P., Voronovsky, A.Y., Stasyk, O. V., Gayda, G.Z., Gonchar, M. V., Abbas, C .a., 487 et al., Overexpression of pyruvate decarboxylase in the yeast Hansenula polymorpha 488 results in increased ethanol yield in high-temperature fermentation of xylose. FEMS 489 Yeast. Res. 2008, 8,1164–74. 490 [14] Ingram, L.O., Aldrich, H.C., Borges, a. C.C., Causey,T.B., Fernando, M.A.M., Saleh, A., et 491 al., Enteric Bacterial Catalysts for Ethanol Production. Biotechnol. Prog. 1999,15,855–66. 492 [15] Barbosa, M. de FSBO., Ingram,L.O., Expression of theZymomonas mobilis alcohol 493 dehydrogenase II (adhB) and pyruvate decarboxylase (pdc) genes in Bacillus. Curr. 494 Microbiol. 1994,28,279–82. 495 [16] Yun., H., Kim, B.-G., Enzymatic Production of ( R ) -Phenylacetylcarbinol by Pyruvate 496 Decarboxylase from Zymomonas mobilis. Biotech. Bioprocess Eng. 2008,13,372–6. 497 [17] Breuer, M., Bernhard, H., Friedrich, T., Production and use or pyruvate decarboxylase. 498 United States Patent ,7189545,2005. 499 [18] Krutsakorn, B., Imagawa, T., Honda, K., Okano, K., Ohtake, H., Construction of an in vitro 500 bypassed pyruvate decarboxylation pathway using thermostable enzyme modules and 501 its application to N-acetylglutamate production. Microb. Cell Fact. 2013,12,1–9. 502 [19] Höhne, M., Kühl, S., Robins, K., Bornscheuer, U.T., Efficient asymmetric synthesis of Chiral 503 amines by combining transaminase and pyruvate decarboxylase. ChemBioChem. 504 2008,9,363–5. 505 [20] Klein-Marcuschamer, D., Oleskowicz-Popiel, P., Simmons, B. a., Blanch, H.W., The 506 challenge of enzyme cost in the production of lignocellulosic biofuels. Biotechnol. Bioeng. 507 2012,109,1083–7. 20 508 [21] Ferreira, R.D.G., Azzoni, A.R., Freitas, S., Techno-economic analysis of the industrial 509 production of a low-cost enzyme using E. coli: The case of recombinant β-glucosidase. 510 Biotechnol. Biofuels. 2018,11,1–13. 511 [22] Lima-Ramos, J., Nordblad, M., Woodley, J.M., Guidelines and Cost Analysis for Catalyst 512 Production in Biocatalytic Processes. Org. Process Res. Dev. 2011,15,266–74. 513 [23] Schumann, W., Ferreira, L.C.S., Production of recombinant proteins in Escherichia coli. 514 Genet. Mol. Biol. 2004,27,442–53. 515 [24] Tripathi, N.K., Production and Purification of Recombinant Proteins from Escherichia coli. 516 ChemBioEng. Rev. 2016,3,116–133. 517 [25] Chen, R., Bacterial expression systems for recombinant protein production: E. coli and 518 beyond. Biotechnol. Adv. 2012,30,1102–1107. 519 [26] Wingfield, P.T., Overview of the Purification of Recombinant Proteins. Curr. Protoc. 520 Protein Sci., 2016,80,1-50. 521 [27] Lee, S.Y., High cell-density culture of Escherichia coli. Trends Biotechnol. 1996,14,98–105. 522 [28] Huang, C.-J., Lin, H., Yang, X., Industrial production of recombinant therapeutics in 523 Escherichia coli and its recent advancements. J. Ind. Microbiol. Biotechnol. 2012,39,383– 524 99. 525 [29] Yee, L .BH., Recombinant Protein Expression in High Cell Density Fed-Batch Cultures of 526 Escherichia Coli. Biotechnology. 1992,10,1550–1556. 527 [30] Vidal, L., Pinsach, J., Striedner, G., Caminal, G., Ferrer, P., Development of an antibiotic- 528 free plasmid selection system based on glycine auxotrophy for recombinant protein 529 overproduction in Escherichia coli. J. Biotechnol. 2008,134,127–36. 530 [31] Rosano GL, Ceccarelli E a. Recombinant protein expression in Escherichia coli: Advances 531 and challenges. Front Microbiol 2014;5:1–17. doi:10.3389/fmicb.2014.00172. 532 [32] Ruiz, J., Fernández-Castané, A., De Mas, C., González, G., López-Santín, J., From 533 laboratory to pilot plant E. coli fed-batch cultures: Optimizing the cellular environment 534 for protein maximization. J. Ind. Microbiol. Biotechnol. 2013,40,335–343. 535 [33] Schweder, T., Lin, H., Jürgen, B., Breitenstein, A., Riemschneider, S., Khalameyzer, V., et 536 al., Role of the general stress response during strong overexpression of a heterologous 537 gene in Escherichia coli. Appl. Microbiol. Biotechnol. 2002,58,330–337. 538 [34] Mühlmann, M., Forsten, E., Noack, S., Büchs, J., Optimizing recombinant protein 539 expression via automated induction profiling in microtiter plates at different 540 temperatures. Microb. Cell Fact. 2017,16,1–12. 541 [35] Jan-ChristerJanson, (Ed.). Protein purification : Principles, high resolution methods and 542 Applications. 3rd ed. Wiley 2011. 543 [36] Bonner, L.R. P., Protein purification. Garland Science 2007. 21 544 [37] Ullah, H., Ion Exchange Technology. The Role of Ion Exchange Chromatography in 545 Purification and Characterization of Molecules. 2012. 546 [38] Hoppner, T.C., Doelle, H.W., Purification and kinetic characteristics of pyruvate 547 decarboxylase and ethanol dehydrogenase from Zymomonas mobilis in relation to 548 ethanol production. Eur. J. Appl. Microbiol. 1983, 17, 152-157 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 22 570 Table 1. Maximum specific growth rates (h-1), specific PDC activities (U mg-1) and final biomass 571 concentrations (gDCW L-1) obtained in shake flask cultures of M15[pREP4][pQE-PDC] and 572 SG13009[pREP4][pQE-PDC] in defined medium. Culture was performed at 37oC and 150 rpm. Maximum Final biomass Specific activity Strain specific growth concentration (U mg-1) rate (h-1) (gDCW L-1 ) M15[pREP4][pQE-PDC] 0.43 3.1 0.45 SG13009[pREP4][pQE-PDC] 0.41 5.2 0.41 573 574 575 576 577 578 579 580 581 582 583 584 585 23 586 Table 2. PDC purification by sequential precipitation with ammonium sulphate. Enzymatic 587 activity obtained after the two-step sequential precipitation with 40% w/v of (NH4)2SO4 588 saturation in the first step and different percentages of (NH4)2SO4 in the second step. Incubations 589 with (NH4)2SO4 were performed at 5oC during 0.5 h. The initial activity of the second step 590 corresponds to the activity of the supernatant after the first step (40% w/v (NH4)2SO4). (NH4)2SO4 Total Activity in Activity in Recovered Specific Purification (% Saturation) initial the the activity activity factor activity precipitate supernatant (%) (U mg-1) (U) (U) (U) Lysate - - - 7.9 - 40 34.1 2.6 27.8 81.6a 12.0 a 1.5 (1st step) 65 27.1 0.8 79.3b 15.7 b 2 (2nd step) 27.8 70 26.4 0.7 77 b 11.8 b 1.5 (2nd step) 591 a) In the supernatant 592 b) In the precipitate 593 594 595 596 597 598 599 24 600 Table 3. PDC purification with MANA-agarose by using a batch-strategy. Firstly, PDC (20 U mL 601 support-1) was absorbed in MANA-agarose at 25oC during 0.5 h at different pH 6.2 (phosphate 602 buffer 50 mM) and pH 5.0 (acetate buffer 50 mM). Finally PDC was desorbed from MANA- 603 agarose with 1 M NaCl. pH Total Total Recovered Recovered Recovered Specific Purification initial initial activity protein activity activity factor activity protein after NaCl after NaCl (%) (U mg-1) (U) (mg) desorption desorption (U) (U) lysate - - - 5.7 - 6.2 20 3.49 20.3 1.4 100 15.1 2.6 5 20.3 0.9 100 23.5 4.1 604 605 606 607 608 609 610 611 612 613 614 25 615 Figure legends 616 Figure 1. Shake flask cultures of M15[pREP4][pQE-PDC] and SG13009[pREP4][pQE-PDC] grown 617 with LB-medium. Biomass (filled diamonds) and PDC activity (filled circles) for M15[pREP4][pQE- 618 PDC] strain; biomass (empty diamonds) and PDC activity (empty circles) for 619 SG13009[pREP4][pQE-PDC] strain. Cultures were performed at 37oC and 150 rpm. The arrow 620 indicates the IPTG pulse for the induction. 621 Figure 2. SDS-Page electrophoresis of M15[pREP4][pQE-PDC] and SG13009[pREP4][pQE-PDC] 622 samples from shake flask cultures in LB-medium. M: molecular weight marker. PI: pre-induction 623 sample. 1 h, 2 h, 4 h: time after induction with IPTG. Pyruvate decarboxylase subunits 624 correspond to the 60 kDa bands. Culture was performed at 37oC and 150 rpm. 625 Figure 3. PDC production in fed-batch culture in a 2L-bioreactor using the strain 626 SG13009[pREP4][pQE-PDC]. Biomass (black diamonds), glucose concentration (white circles) 627 and PDC activity (black triangles). The start of the fed-batch phase is represented with the 628 vertical dotted line. The arrow indicates the IPTG pulse for induction. 629 Figure 4. FPLC chromatogram profile of PDC purification using MANA-agarose and a linear 630 gradient elution. The wash step was performed using buffer A (50 mM acetate buffer pH 5) and 631 elution step was performed mixing buffer B (1 M NaCl, 50 mM acetate buffer pH 5) with buffer 632 A linearly increasing (0-100% buffer B) over 10 column volumes. 633 Figure 5. SDS-Page electrophoresis of samples obtained in the linear gradient elution 634 chromatography. Each line corresponds to a 1 mL fraction from the FPLC after the 635 chromatographic process. 636 Figure 6. Enzyme activity and protein concentration profiles with FPLC using a two-stage elution 637 step. The elution first step is performed at 33% of buffer B (during 10 column volumes). The 638 second step is performed at 100% of buffer B (during 10 column volumes). 26 639 Figure 1 640 641 642 643 644 645 646 647 648 649 650 27 651 Figure 2 652 653 654 655 656 657 658 659 28 660 Figure 3 661 662 663 664 665 666 667 668 669 670 671 29 672 Figure 4 673 674 675 676 677 678 679 680 681 682 683 30 684 Figure 5 685 686 687 688 689 690 31 691 Figure 6 692 32