Expression, purification and thermal stability evaluation of an engineered amaranth protein expressed in Escherichia coli Academic research paper on "Biological sciences"

ARTICLE IN PRESS

EJBT-00163; No of Pages 8

Electronic Journal of Biotechnology xxx (2015) xxx-xxx

HOSTED BY

ELSEVIER

Contents lists available at ScienceDirect

Electronic Journal of Biotechnology

Research Article

Expression, purification and thermal stability evaluation of an engineered amaranth protein expressed in Escherichia coli

Jocksan I. Morales-Camacho a, Octavio Paredes-López b, Edgar Espinosa-Hernández a, Daniel Alejandro Fernández Velascoc, Silvia Luna-Suárez a *

a Instituto Politécnico Nacional-CIBA-IPN, Ex-Hacienda San Juan Molino Carretera estatal Tecuexcomac-Tepetitla Km 1.5, 90700 Tepetitla, Tlaxcala, Mexico

b Departamento de Biotecnología y Bioquímica, Centro de Investigación y de Estudios Avanzados de IPN, Libramiento Norte Carretera Irapuato León Kilómetro 9.6, Carretera Irapuato León, 36821 Irapuato, Guanajuato, Mexico

c Laboratorio de Fisicoquímica e Ingeniería de Proteínas, Departamento de Bioquímica, Facultad de Medicina, Universidad Nacional Autónoma de México, D.F., Mexico, Mexico

ARTICLE INFO

ABSTRACT

Article history: Received 3 December 2015 Accepted 14 March 2016 Available online xxxx

Keywords: Globulin 11S Protein expression Protein engineering Thermal stability

Background: The acidic subunit of amarantin (AAC)—the predominant amaranth seed storage protein—has functional potential and its third variable region (VR) has been modified with antihypertensive peptides to 22 improve this potential. Here, we modified the C-terminal in the fourth VR of AAC by inserting four VY 23 antihypertensive peptides. This modified protein (AACM.4) was expressed in Escherichia coli. In addition, we 24 also recombinantly expressed other derivatives of the amarantin protein. These include: unmodified 25 amarantin acidic subunit (AAC); amarantin acidic subunit modified at the third VR with four VY peptides 26 (AACM.3); and amarantin acidic subunit doubly modified, in the third VR with four VY peptides and in the 27 fourth VR with the RIPP peptide (AACM.3.4). 28

Results: E. coli BL21-CodonPlus (DE3)-RIL was the most favorable strain for the expression of proteins. After 6 h of 29 induction, it showed the best recombinant protein titer. The AAC and AACM.4 were obtained at higher titers 30 (0.56 g/L) while proteins modified in the third VR showed lower titers: 0.44 g/L and 0.33 g/L for AACM.3 and 31 AACM.3.4, respectively. As these AAC variants were mostly expressed in an insoluble form, we applied a 32 refolding protocol. This made it possible to obtain all proteins in soluble form. Modification of the VR 4 33 improves the thermal stability of amarantin acidic subunit; AAC manifested melting temperature (Tm) at 34°C 34 and AACM.4 at 37.2°C. The AACM.3 and AACM.3.4 did not show transition curves. 35

Conclusions: Modifications to the third VR affect the thermal stability of amarantin acidic subunit. 36

©2015 Pontificia Universidad Católica de Valparaíso. Production and hosting by Elsevier B.V. All rights reserved. 38

1. Introduction

Hypertension is a worldwide health problem and is a major risk factor for cardiovascular diseases such as heart failure, stroke, peripheral arterial disease and myocardial infarction. Some drugs developed for hypertension treatment cause side effects including the angiotensin converting enzyme (ACE) inhibitors. They inhibit the conversion of angiotensin I to angiotensin II within the renin-angiotensin system. Because of this, several researchers have focused on the study of new alternatives including nutritive compounds that contribute to the treatment of hypertension. These include bioactive peptides that are usually composed of 2-20 amino acid residues but may be larger [1,2,3]. These structures have been isolated from different foods and are usually inactive within the

* Corresponding author. E-mail addresses: ssilvials2004@yahoo.com.mx, sluna@ipn.mx (S. Luna-Suárez). Peer review under responsibility of Pontificia Universidad Católica de Valparaíso.

sequences of parent proteins but can be released by enzymatic 61 hydrolysis during gastrointestinal digestion or during food preparation 62 to subsequently acquire bioactivity [4,5]. Other strategies to obtain 63 these products include the design of antihypertensive peptide 64 multimers or protein engineering. Biopeptides can be introduced into 65 proteins with nutritional or functional importance, and these new 66 structures may be used as components in foods or additives in the 67 formulation of new products [6,7]. An alternative to synthesizing 68 these proteins is through recombinant systems such as Escherichia coli 69 because the genetic manipulation is easy, culture is inexpensive, and 70 expression is fast to quickly produce large amounts of protein [8]. 71 In our group, we purified and characterized the amarantin acidic 72 subunit (AAC) by expressing it in E. coli. Amarantin is important 73 because of its nutritional value, it has an excellent essential amino acid 74 balance; the AAC is a good candidate for modification in terms of its 75 structure [9], it harbors four variable regions on the molecular surface 76 (Fig. 1). The variable regions have been suggested as suitable targets 77 for modification because they are not important in the structure 78

http://dx.doi.org/mi 016/j.ejbt.2016.04.001

0717-3458/© 2015 Pontificia Universidad Católica de Valparaíso. Production and hosting by Elsevier B.V. All rights reserved.

ARTICLE IN PRESS

J.!. Morales-Camacho et al. / Electronic Journal of Biotechnology xxx (2015) xxx-xxx

a Schematic representation Protein name Reference

VR1 VR2 VR3 VR4

1-10 86-131 185-209 24M90 AAC [9]

N-B H-c

199-(4)VY-200

N-B H-c AACM.3 [10]

199-(4)VY-200 290-RIPP

N-B M-c AACM.3.4 [11]

272-(4)VY-273

N-B M-c AACM.4 (this research)

Fig. 1. Schematic representation of AAC variants and putative 3D model. (a) AAC variants previously reported and variant generated in this research (AACM.4). Fourvariable regions are indicated in AAC scheme by VR1, VR2, VR3 and VR4; residue numbering is based on the PDB structure of amaranth 11S proglobulin seed storage protein from A. hypochondriacus L. (3QAC); white and black areas represent conserved and variable regions, respectively. Numbers in the second to fourth schemes indicate insertion in each protein. (b), putative 3D model using as template 11S globulin from amaranth (PDB 3QAC). Variable regions are indicated with VR1, VR2, VR3, and VR4 that correspond with unordered structure exposed in surface.

formation [10]. Thereafter, we generated the AACM.3 protein inserting four Val-Tyr (VY) antihypertensive peptides into the AAC third variable region (VR) to improve its nutraceutical properties [11]. We also generated and overexpressed the doubly modified acidic subunit (AACM.3.4) into the third VR with four VY and a RIPP peptide at the C-terminal [12]. This last protein showed antihypertensive effect both in in vitro and in vivo trials [13]. The objective of this research was to modify AAC by inserting four VY antihypertensive peptides into the C-terminal to generate the protein AACM.4; and to compare expression levels and thermal stability of the new protein and the above mentioned variants of AAC. This might lead to the identification of a good candidate or model to establish the basis for the production of a functional food additive or to improve the nutraceutical properties in some crops in the near future.

2. Materials and methods

Fig. 1a is a schematic diagram that summarizes the proteins used in this work.

2.1. Homology modeling of AAC

In order to evaluate the spatial positioning of the variable regions, a 3D model was generated using amaranth 11S globulin (PDB 3QAC) as template [14] and RaptorX server [15] which predict secondary and tertiary structures together with disordered regions. Moreover, as VRs in 3QAC are disordered regions, which are not visible in the crystal structure, the model of AAC generated by RaptorX was optimized using ModLoop server (https://modbase.compbio.ucsf.edu/modloop/) which is a software tool used to predict disordered regions by satisfaction of spatial restraints. Then, the resulting homology 3D model suggests that the VRs are exposed on the surface (Fig. 1b).

2.2. Construction of the modified amarantin acidic subunit expression 107

plasmid to express AACM.4 108

The plasmid pET-ACID-R4-6His that codes for AACM.4 was 109

constructed from pET-AC-6His [9] which contains the cDNA that 110

codifies amarantin acidic subunit. This plasmid is derived from 111

pSPORT-11S, which contains the cDNA that codifies amaranth 11S 112

globulin (GenBank accession no. CAA57633.1) [16]. pET-AC-6His was 113

used as PCR template for carrying out site-directed mutagenesis to 114

insert four VY biopeptides in tandem into the fourth variable region 115

corresponding to the C-terminal region in the AAC. Platinum® Taq 116

DNA Polymerase High Fidelity (Invitrogen) was used to obtain the 117

PCR products of AAC mutated into its fourth variable region. The 118

primers for amplification were: forward 5'-TGGGTGATTAATGGAAGG 119

AAGG-3' (underlined corresponding to the Aseirestriction site that 120

enabled cloning into the pET-32b( + ) vector). Reverse: 5'-CTCGAG 121

GTAAACGTAAACGTATACGTAAACCCTATTGGGAAGG-3' (the encoding 122

sequence of the VYVYVYVY peptide inserted in bold letters, with Xho i 123

restriction site underlined). This enabled cloning into the pET-32b( + ) 124

vector. 125

Following amplification, the PCR product was cloned into 126

pPCR®2.1-TOPO®. E. coli cells harboring recombinant plasmid were 127

selected on LB plates containing 100 ^g/mL ampicillin and X-gal. The 128

DNA fragment encoding AACM.4 was released from pPCR®2.1-TOPO 129

vector using VspI and XhoI restriction enzymes. The VspI restriction 130

site enabled cloning the region of AACM.4 into the ribosome binding 131

site in the pET-32b(+) vector. The split by XhoI enabled cloning of 132

AACM.4 onto the frame with the His tag region of the vector. The 133

VspI/XhoI fragment was ligated into plasmid pET-32b( + ) and 134

transformed into the TOP 10 cloning host as described by Sambrook 135

et al. [17]. The E. coli transformants were selected on LB plates 136

containing 100 |ag/mL carbenicillin. The positive clones were 137

confirmed with PCR, restriction analysis and DNA sequencing using 138

both T7 forward and reverse sequencing primers. 139

ARTICLE IN PRESS

J.l. Morales-Camacho et al. / Electronic Journal of Biotechnology xxx (2015) xxx-xxx

140 2.3. Transformation of expression cells

141 E. coli, Rosetta 2 and BL21-CodonPlus(DE3)-RIL strains were

142 transformed with pET-AC-6His codifying AAC [9], pET-AC-Ml

143 codifying AACM.3 [11], pET-ACID-R4-6His codifying AACM.4, and

144 pET-AC-M3-6His codifying AACM.3.4 [12] by applying heat shock

145 following the manufacturer's instructions. All used plasmids contained

146 an ampicillin resistance gene (AmpR) it makes the E. coli strain

147 ampicillin resistant. Subsequently, successfully transformed cells were

148 selected on LB (Sigma) agar plates containing 100 |ag/mL ampicillin

149 and 34 |ag/mL chloramphenicol.

150 2.4. Inoculum preparation and shake flask fermentation

151 Transformants expressing AAC and its variants were used to

152 inoculate 5 mL LB broth (Sigma) supplemented with appropriate

153 antibiotics as mentioned previously. Pre-cultures were grown

154 overnight at 37°C on an orbital shaker at 200 rpm. Shake flask

155 fermentations were performed in 250 mL Erlenmeyer flasks

156 containing 42 mL of LB broth. Cultivations were inoculated with

157 2.5% (v/v) pre-culture. Cultures were shaken at 200 rpm in an orbital

158 incubator at 37°C. Expression of different variants of AAC (in both

159 strains) were initiated when cultures reached 0.5 OD at 600 nm

160 by adding lactose as an inducer at 0.5% (w/v) according to

Morales-Camacho et al. [18]. This process was continued after 161 induction and samples were taken at 0, 1.5, 3, 6, and 24 h to 162 determine the best strain to express all variants of AAC; 1 mL from 163 each culture was transferred to 1.5 mL centrifuge tube. All samples 164 were harvested by centrifugation at 13,300 x g for 5 min at room 165 temperature. Supernatants were discarded and cell pellets were 166 stored at -20°C for further analysis. 167

2.5. Bioreactor fermentation 168

To increase the production yield of AAC and its variants, the 169 fermenter cultures were performed in a 5 L bioreactor Biostat 170 A (Sartorius) with an 80% working volume as described in 171 Morales-Camacho et al. [18] with some modifications. For each 172 expression experiment, (AAC, AACM.3, AACM.4, and AACM.3.4) batch 173 culture conditions were as follows: culture media, 4 L (200 g/L potato 174 waste, 12 g/L tryptone, 4 g/L glycerol, 17 mM KH2PO4 and 72 mM 175 K2HPO4); inoculum, 2.5% (v/v); agitation, 340 rpm; aeration, 1 vvm; 176 pH was maintained at 7 by addition of 2 M NaOH or 5 M H2SO4; initial 177 temperature, 37°C and subsequent induction temperature was 178 adjusted to 30°C. Protein expression was induced when the culture 179 had an OD60o of 0.5 with 0.5% (w/v) lactose. The cells were harvested 180 6 h after induction by centrifugation 13,300 x g for 5 min at room 181 temperature. 182

Fig. 2. SDS-PAGE and Western blot of recombinant proteins. (a) AAC protein and schematic representation; (b) AACM.3 protein and schematic representation; (c) AACM.4 and schematic representation; (d) AACM.3.4 protein and schematic representation. The fourvariable regions are shown with labels I, II, III, and IV. The modification is indicated. Lanes: MW: molecular weight; 1:0 h of induction; 2:1.5 h of induction; 3:3 h of induction; 4:6 h of induction; 5:24 h of induction in E.coli Rosetta 2; 6:1.5 h of induction; 7:3 h of induction; 8:6 h of induction; 9:24 h of induction in E. coli BL21-CodonPlus (DE3 )-RIL. The image below corresponds to Western blot of each recombinant protein. Arrows indicate the presence of AAC and their variants in each SDS-PAGE gels. 5 |jg of protein was loaded into each lane.

ARTICLE IN PRESS

J.!. Morales-Camacho et al. / Electronic Journal of Biotechnology xxx (2015) xxx-xxx

Time expression (h)

i-AAC -»-AACM.3 AACM.4 AACM.3.4

_J 0.08

a- 0.06

0 0.04

2 0.02

Time expression (h)

Fig. 3. Expression of recombinant proteins by E. coli BL21-CodonPlus(DE3)-RIL (a) Yields achieved after induction; (b) productivity achieved in recombinant proteins after induction. Each value is the mean of triplicate experiments. ANOVA analysis evaluated significant differences in expression (P < 0.05).

183 2.6. Detection ofAAC and its different variants

184 Cell pellets were resuspended in 0.25 mL of distilled water and

185 0.25 mL of loading buffer. Samples were analyzed by 12% SDS-PAGE

186 [19] and stained with Coomassie brilliant blue G-250. Western

187 blotting was used to detect the proteins, which were transferred to

188 PVDF membrane using a Mini Trans-Blot cell (Bio-Rad). Membrane

189 was incubated with a polyclonal antiamarantin as first antibody, and

190 anti-rabbit conjugated to alkaline phosphatase as secondary antibody

191 in a dilution 1:60,000 and 1:3,000, respectively [9].

192 2.7. Protein assay

193 Protein concentration was determined using BCA (bicinchoninic

194 acid) assay using bovine serum albumin (Sigma) as a protein

195 standard. The concentration of recombinant proteins was determined

196 in cell pellets disrupted in alkaline buffer (0.2 M NaOH, 1% SDS) and

197 by SDS-PAGE (12%) under denaturing and reducing conditions. Gels

were scanned, and the area and intensity of the bands were quantified 198

by densitometric analysis using Image Lab 4.0 (Bio-Rad). 199

2.8. Experimental design 200

A simple factorial design for three factors was used to carry out 201

experiments for the expression of four recombinant proteins (AAC, 202

AACM.3, AACM.4, and AACM.3.4) in two E. coli strains (Rosetta 2 and 203

BL21-CodonPlus(DE3)-RIL) at different induction times (0, 3, 6, and 204

24 h). All trials were performed in triplicate. 205

The statistical software package Statgraphics 16.2.04 (Statistical 206

Graphics Corp) was used to analyze the data. An analysis of variance 207

(ANOVA) was used to assess the best strain to express AAC and the 208

modified variants according to protein titers. 209

2.9. Analysis of soluble and insoluble fraction of all variants ofAAC 210

Cell disruption was used to determine the fraction in which each 211

variant of AAC was accumulated inside cells. The samples were 212

I j I I

AACM.4 AACM.3.4

Fig. 4. SDS-PAGE and Western blot of soluble and insoluble protein fraction in AAC, AACM.3, AACM.4, and AACM.3.4. (a) Extraction of AAC and AACM.3 expressed by BL21-CodonPlus(DE3)-RIL and Rosetta 2. (b) Extraction of AACM.4 and AACM.3.4 expressed by BL21-CodonPlus(DE3)-RIL and Rosetta 2. Odd numbers in lanes represent the soluble protein fraction and even lanes are the insoluble protein fraction. Arrows indicate recombinant proteins.

ARTICLE IN PRESS

J.!. Morales-Camacho et al. / Electronic Journal of Biotechnology xxx (2015) xxx-xxx

213 harvested 6 h after induction and were used to induce cell disruption.

214 This treatment was carried out in 5 mL extraction buffer (EB) (0.2 M

215 NaCl, 20 mM phosphates, pH 7.5) applying sonication for 1 min at an

216 amplitude of 60% and then, kept on ice for 3 min. This process was

217 repeated six times. Subsequently, the disrupted cells were fractionated

218 by centrifugation at 15,600 x g for 30 min at 4°C. The supernatants

219 were discarded and the cell debris pellets were resuspended in 10 mL

220 solubilization buffer (SB) (6 M urea, 0.2 M NaCl, 20 mM phosphates,

221 pH 7.5) and shaken at 200 rpm and room temperature for 4 h.

222 Subsequently, the samples were centrifuged at 15,600 x g for 30 min

223 at 4°C to remove insoluble material. The supernatants were

224 recuperated and analyzed by 12% SDS-PAGE and stained with

225 Coomassie brilliant blue G-250 to detect recombinant proteins.

226 2.10. Purification ofAAC and its modified variants

227 The insoluble crude extract of each variant of AAC was applied to an

228 AP-2 20 x 300 mm column (Waters) packed with Protino Ni-TED resin

229 (Macherey-Nagel) coupled to BioLogic DuoFlow™ Chromatography

230 System (Bio-Rad) and eluted with buffer SB plus imidazole at a

231 flow rate of 2.5 mL/min and at room temperature. Increasing

232 concentrations of imidazole were applied during purification process

233 to evaluate which imidazole concentration eluted each variant of AAC.

Five column volumes of SB plus 5, 25, 50, 125, 250, and 500 mM 234 imidazole were passed through the column. All fractions collected 235 during the purification process were analyzed by SDS-PAGE and 236 stored at 4°C. 237

2.11. Protein refolding ofAAC and its modified variants 238

The refolding protocol was established to eliminate urea and allow 239 refolding of AAC, ACM.3, ACM.4, and ACM.3.4. First, 0.5 mL at 0.5 and 240 1 mg/mL for each sample of protein was dialyzed against 50 mL of 241 refolding buffer (4 M urea, 0.2 M NaCl, 20 mM phosphates, pH 7.5) for 242 1.5 h in an analogue tube roller (Bibby Scientific) at4°C. This process 243 was repeated to diminish the urea concentration to 3,1.5, and 0.5 M. 244 Finally, three cycles of dialysis were applied against 50 mL of EB. 245

Before and after refolding, the samples were assayed by fluorescence 246 (Supplementary Fig. 1). 247

2.12. Thermal stability using Thermofluor assay

The proteins AAC, AACM.3, AACM.4, and AACM.3.4 were subjected 249 to thermofluor assay [20], implemented using an ABI 7900 Real-Time 250 PCR (Applied Biosystems) system. Sypro Orange (Invitrogen) with 251 emission wavelength at 625 nm was used as a fluorescent dye. Briefly, 252

Fig. 5. SDS-PAGE and Western blot of purification and refolding process of recombinant proteins. (a) AAC and AACM.3 proteins; (b) AACM.4 and AACM.3.4 proteins. Lanes: MW: molecular weight; 1 and 4: soluble extract of each protein; 2 and 5: insoluble extract of each protein; 3 and 6: pure proteins after refolding.

ARTICLE IN PRESS

J.!. Morales-Camacho et al. / Electronic Journal of Biotechnology xxx (2015) xxx-xxx

253 20 |aL of the sample (final volume) containing 1 mg/mL of the protein

254 and 5 x SYPRO orange dye in EB buffer were used. The temperature

255 was increased from 25 to 95°C at a rate of 1°C/min. Equilibrium time

256 was 5 s for each temperature. All experiments were in triplicate,

257 melting curve analysis and determination of melting temperature (Tm)

258 were undertaken using Origin software (version 6.0) (Northampton).

259 2.13.2-D electrophoresis

260 AAC and all modified variants were subjected to isoelectric focusing

261 using 7 cm immobilized strips of 3-10 pH gradient (Bio-Rad) in a

262 Protean i12 IEF system (Bio-Rad) to determine the isoelectric point

263 (pi) of the purified proteins [11].

264 3. Results and discussion

265 3.1. Expression and detection of AAC and its variants

266 Rosetta 2 (Novagen) and BL21-CodonPlus(DE3)-RIL (Stratagene) E.

267 coli strains were used to evaluate the expression of the AAC and their

268 modified variants at flask level. SDS-PAGE data showed that the titers

269 of all variants of AAC were different in both strains. There was an

270 accumulative effect in all recombinant proteins for up to 6 h of

271 expression; the presence of these proteins was confirmed by Western

272 blot (Fig. 2). The AAC and AACM.4 were expressed at high titers in

273 both strains at 24 h following induction in both strains (Fig. 2a and c).

274 Meanwhile AACM.3 and AACM.3.4 were expressed at lower titers

275 (Fig. 2b and d). The AACM.3 was not detected at 24 h in Rosetta 2

276 (Fig. 2b lane 5). The ANOVA data on densitometric values revealed

that the E. coli BL21-CodonPlus(DE3)-RIL was better for the expression 277 of all recombinant proteins. 278

The expression levels of AAC and AACM.4 expressed by 279 BL21-CodonPlus(DE3)-RIL did not show statistically significant 280 differences (SSD) (P< 0.05) after 6 h. In contrast, proteins modified in 281 the third VR (AACM.3 and AACM.3.4) did not show SSD after 3 h of 282 expression; at 6 h SSD was clear. At 24 h after induction, all variants 283 showed SSD (P < 0.05) (Fig. 3a). According to the results, the level of 284 expression for recombinant proteins can be summarized as: 285

AAC>AACM.4 >AACM.3 >AACM.3.4

[Equation 1]

According to titers and productivity (Fig. 3b), AAC and AACM.4 are proteins that manifest greater in vivo or kinetic stability than AACM.3 288 and AACM.3.4. This might be because of modifications in the third VR. 289 The cultures were fractionated to define where the recombinant 290 proteins were accumulated at 6 h after induction. Soluble and 291 insoluble fractions are presented in Fig. 4. The AACM.4 protein 292 expressed by E. coli BL21-CodonPlus(DE3)-RlL manifested a higher 293 proportion in soluble form (approximately 5% of total protein titer) 294 than the other proteins. Most recombinant proteins were expressed 295 in insoluble form; the results were similar to those in the literature [9, 296 11,12]. 297

3.2. Bioreactor fermentation, purification and refolding of AAC and their 298 modified variants 299

Expression in E. coli BL21-CodonPlus(DE3)-RlL was scaled-up in a 300 bioreactor to increase the production yield for all the recombinant 301

3 26500 <

£ 26000

£ 25500 O J3 LL

25 29 33 37 41 45 49 53 Temperature (°C)

<D U C <11 u

25 29 33 37 41 45 49 53 Temperature (°C)

a> o in <u

22500 -

29 33 37 41 Temperature (°C)

a> o c 0) o

u- 45000

25 29 33 37 41 45 Temperature (°C)

Fig. 6. Different melting curve profiles obtained for each recombinant protein by thermofluor assay. (a) Transition curve of the AAC protein; (b) transition curve of AACM.3 protein; (c) and (d): transition curves of AACM.4 and AACM.3.4 proteins, respectively. The insets in (a) and (c) correspond to the plot of Tm values for AAC (34°C) and AACM.4 (37.2°C). Each value is the mean of triplicate experiments.

ARTICLE IN PRESS

].¡. Morales-Camacho et al. / Electronic Journal of Biotechnology xxx (2015) xxx-xxx

302 proteins. The titers at 6 h after induction were AAC, 1.6 g/L; AACM.4,

303 1.55 g/L; AACM.3, 1.37 g/L and AACM.3.4, 1.32 g/L. These results are

304 3-fold higher than at the flask level because the available oxygen was

305 increased by agitation and aeration. The cells harvested 6 h after

306 induction were disrupted to extract proteins. SDS-PAGE and Western

307 blot data are shown in Fig. 5. Recombinant proteins were expressed as

308 insoluble aggregates, which may be due to an increased translation

309 rate leading to protein misfolding and poor solubility [21].

310 Furthermore, the insoluble fractions were used to purify the

311 recombinant proteins by IMAC. The AAC eluted at 5 mM imidazole.

312 The AACM.3.4, AACM.3 and AACM.4 eluted at 50 and 125 mM

313 imidazole, respectively. Next, purified proteins were refolded by

314 applying dialysis to eliminate urea. The percentages of refolded

315 proteins according to initial concentration were 20% AAC, 18%

316 AACM.3, 36% AACM.4, and 30% AACM.3.4. These differences may be

317 due to different conformations occurring in recombinant proteins

318 because of modifications with the biopeptides. Fig. 5 shows the

319 SDS-PAGE and Western blot of all refolded pure proteins.

320 3.3. Thermal stability and 2-D electrophoresis

321 The Tm provides a quantitative indication of protein stability. A more

322 stable protein will require higher temperatures to unfold [22]. All

323 proteins were subjected to a thermofluor assay in EB, and Tm values

324 were 34.0 ± 0.11°C for AAC and 37.2 ± 0.2°C for AACM.4. This may

325 mean that AAC exposes its hydrophobic regions at lower temperatures

326 than AACM.4. Thus, AACM.4 is a more thermostable protein. By other

327 hand the absence of thermal transition curves in the VR3 mutants

328 may indicate that these proteins are unfolded, but the fluorescence

329 spectra (Supplementary Fig. 1) indicated that they don't have the

330 same maximum emission wavelength after the urea were eliminated,

331 so they are partially folded. Thus, modifications in the third VR may

332 generate conformational changes that only enable solubilization of

333 these proteins by the refolding protocol such that AACM.3 and

334 AACM.3.4 remained as molten globule (Fig. 6). The fact that the

335 molten globules (VR3 mutants) were the lowest expressed by E. coli,

336 it's probably related to kinetic or in vivo stability.

337 Tandang-Silvas et al. [14] reported that thermal stability increases by

338 diminishing the loop length, but for AACM.4 the insertion of bioactive

339 peptides into the fourth VR increases the loop. Here, thermal stability

340 is greater than that of the AAC. Perhaps the insertion of bioactive

341 peptides in AACM.4 enhances thermal stability because it generates

342 interactions that stabilize the protein.

343 The 2-D electrophoresis demonstrated that the pl of AAC is 6,

344 whereas AACM.3, AACM.4, and AACM.3.4 are 6.1, 5.93, and 6.3,

345 respectively. These results are close to those calculated via the

346 ProtParam tool.

347 In conclusion, there are differences in the expression and recovered

348 percentage based on the refolding protocol of recombinant proteins.

349 This may be due to structural modifications due to the insertions done

350 in the molecule. They are also reflected in properties such as thermal

351 stability.

352 It is well known that mutations by amino acid insertions in

353 secondary structure elements tend to be more susceptible to

354 destabilization of proteins [23]. Nevertheless, in this research variable

355 regions (unstructured regions) were chosen to carry out mutations to

356 decrease this possibility; even so, insertions in third variable region Q3 (VR3) destabilize the structure.

358 Financial support

359 This work was supported by Secretaría de Investigación y Q4 Posgrado-IPN and Consejo Nacional de Ciencia y Tecnología

361 (CONACYT).

Acknowledgments 362

We thank the Secretaría de Investigación y Posgrado-IPN and 363 Consejo Nacional de Ciencia y Tecnología (CONACYT) for the 364 scholarship to JIMC to obtain his PhD degree. We also thank 365 Bueno-Sánchez C., López y López V.E. and Llamas-García M.L. for their 366 technical assistance. 367

Appendix A. Supplementary data 368

Supplementary data to this article can be found online at http://dx. 369 doi.org/10.1016/j.ejbt.2016.04.001. 370

References 371

[1] Fitzgerald RJ, Murray BA, Walsh DJ. Hypotensive peptides from milk proteins. [Cited 372 April 1,2014]. J Nutr 2004;134:980-8 [Available from Internet at: http://jn.nutrition. 373 org/content/134/4/980S.long]. 374

[2] Madureira AR, Tavares T, Gomes AMP, Pintado ME, Malcata FX. Invited review: 375 Physiological properties of bioactive peptides obtained from whey proteins. J Dairy 376 Sci 2010;93:437-55. http://dx.doi.org/10.3168/jds.2009-2566. 377

[3] Palmer BF. Managing hyperkalemia caused by inhibitors of the renin-angiotensin- 378 aldosterone system. N Engl J Med 2004;351:585-92. http://dx.doi.org/10.1056/ 379 NEJMra035279. 380

[4] Hernández-Ledesma M, Del Mar Contreras M, Recio I. Antihypertensive peptides: 381 Production, bioavailability and incorporation into foods. Adv Colloid Interface Sci 382 2011;165:23-35. http://dx.doi.org/10.1016/jxis.2010.11.001. 383

[5] Kawasaki T, Seki E, Osajima K, Yoshida M, Asada K, Matsui T, et al. Antihypertensive 384 effect of valyl-tyrosine, a short chain peptide derived from sardine muscle hydroly- 385 zate, on mild hypertensive subjects. J Hum Hypertens 2000;14:519-23. http://dx. 386 doi.org/10.1038/sj.jhh.1001065. 387

[6] Onishi K, Matoba N, Yamada Y, Doyama N, Maruyama N, Utsumi S. Optimal design- 388 ing of ß-conglycinin to genetically incorporate RPLKPW, a potent anti-hypertensive 389 peptide. Peptides 2004;25:37-43. http://dx.doi.org/10.1016/j.peptides.2003.11.006. 390

[7] Rao S, Su Y, Li J, Xu Z, Yang Y. Design and expression of recombinant antihyperten- 391 sive peptide multimer gene in Escherichia coli BL21. J Microbiol Biotechnol 2009;19: 392 1620-7. http://dx.doi.org/10.4014/jmb.0905.05055. 393

[8] Peti W, Page R. Strategies to maximize heterologous protein expression in 394 Escherichia coli with minimal cost. Protein Expr Purif 2007;51:1-10. http://dx.doi. 395 org/10.1016/j.pep.2006.06.024. 396

[9] Luna-Suárez S, Medina-Godoy S, Cruz-Hernández A, Paredes-López O. Expression 397 and characterization of the acidic subunit from 11S amaranth seed protein. 398 Biotechnol J 2008;3:209-19. http://dx.doi.org/10.1002/biot.200700146. 399

[10] Tandang-Silvas MRG, Tecson-Mendoza EM, Mikami B, Utsumi S, Maruyama N. 400 Molecular design of seed storage proteins for enhanced food fhysicochemical prop- 401 erties. Annu Rev Food Sci Technol 2011;2:59-73. http://dx.doi.org/10.1146/ 402 annurev-food-022510-133718. 403

[11] Luna-Suárez S, Medina-Godoy S, Cruz-Hernández A, Paredes-López O. Modification 404 of the amaranth 11S globulin storage protein to produce an inhibitory peptide 405 of the angiotensin I converting enzyme, and its expression in Escherichia coli. 406 J Biotechnol 2010;148:240-7. http://dx.doi.org/10.1016/jobiotec.2010.06.009. 407

[12] Castro-Martínez C, Luna-Suárez S, Paredes-López O. Overexpression of a modified 408 protein from amaranth seed in Escherichia coli and effect of environmental condi- 409 tions on the protein expression. J Biotechnol 2012;158:59-67. http://dx.doi.org/10. 410 1016/j.jbiotec.2011.12.012. 411

[13] Medina-Godoy S, Rodríguez-Yáñez SK, Bobadilla NA, Pérez-Villalva R, Valdez-Ortiz 412 R, Hong H, et al. Antihypertensive activity of AMC3, an engineered 11S amaranth 413 globulin expressed in Escherichia coli, in spontaneously hypertensive rats. J Funct 414 Foods 2013;5:1441-9. http://dx.doi.org/10.1016/jjff.2013.06.001. 415

[14] Tandang-Silvas MR, Cabanos CS, Carrazco Peña LD, De la Rosa AP, Osuna-Castro JA, 416 Utsumi S, et al. Crystal structure of a major seed storage protein, 11S proglobulin, 417 from Amaranthus hypochondriacus: Insight into its physico-chemical properties. 418 Food Chem 2012;135:819-26. http://dx.doi.org/10.1016/j.foodchem.2012.04.135. 419

[15] Källberg M, Wang H, Wang S, Peng J, Wang Z, Lu H, et al. Template-based protein 420 structure modeling using the RaptorX web server. Nat Protoc 2012;8:1511 -22. 421 http: //dx.doi.org/10.1038/nprot2012.085. 422

[16] Barba de la Rosa AP, Herrera-Estrella A, Utsumi S, Paredes-López O. Molecular charac- 423 terization, cloning and structural analysis of a cDNA encoding an amaranth globulin. 424 J Plant Physiol 1996;149:527-32. http://dx.doi.org/10.1016/S0176-1617(96)80329-4. 425

[17] SambrookJE, Fritsch EF, Maniatis T. Molecular cloning: A laboratory manual. 3rd ed. 426 New York: Cold Spring Harbor Laboratory Press; 1989. 427

[18] Morales-Camacho JI, Domínguez-Domínguez J, Paredes-López O. Overexpression of 428 a modified amaranth protein in Escherichia coli with minimal media and lactose as 429 inducer. Recent Pat Biotechnol 2013;7:61-70. http://dx.doi.org/10.2174/ 430 1872208311307010006. 431

[19] Laemmli UK. Cleavage of structural proteins during the assembly of the head of bac- 432 teriophage T4. Nature 1970;227:680-5. http://dx.doi.org/10.1038/227680a0. 433

[20] Pantoliano MW, Petrella EC, Kwasnoski JD, Lobanov VS, Myslik J, Graf E, et al. High- 434 density miniaturized thermal shift assays as a general strategy for drug discovery. 435 J Biomol Screen 2001;6:429-40. http://dx.doi.org/10.1177/108705710100600609. 436

ARTICLE IN PRESS

J.¡. Morales-Camacho et al. / Electronic Journal of Biotechnology xxx (2015) xxx-xxx

437 [21 ] Chen D, Duggan C, Ganley JP, Kooragayala LM, Reden TM, Texada DE, et al.

438 Expression of enterovirus 70 capsid protein VP1 in Escherichia coli. Protein Expr

439 Purif 2004;37:426-33. http://dx.doi.org/10.1016/j.pep.2004.06.027.

440 [22] Phillips K, De la Peña AH. The combined use of thermofluor assay and ThermoQ

441 analytical software for the determination of protein stability and buffer optimization

as an aid in protein crystallization. Curr Protoc Mol Biol 2011;10. http://dx.doi.org/ 442

10.1002/0471142727.mb1028s94. 443

[23] Vetter IR, Baase WA, Heinz DW, Xiong JP, Snow S, Matthews BW. Protein structural 444

plasticity exemplified by insertion and deletion mutants in T4 lysozyme. Protein Sci 445

1996;5:2399-415. http://dx.doi.org/10.1002/pro.5560051203. 446