Phase behavior and complex coacervation of concentrated pea protein isolate-beet pectin solution

Journal Pre-proofs Phase behavior and complex coacervation of concentrated pea protein isolatebeet pectin solution Yang Lan, Jae-Bom Ohm, Bingcan Chen, Jiajia Rao PII: DOI: Reference:

S0308-8146(19)31655-3 https://doi.org/10.1016/j.foodchem.2019.125536 FOCH 125536

To appear in:

Food Chemistry

Received Date: Revised Date: Accepted Date:

18 June 2019 7 September 2019 14 September 2019

Please cite this article as: Lan, Y., Ohm, J-B., Chen, B., Rao, J., Phase behavior and complex coacervation of concentrated pea protein isolate-beet pectin solution, Food Chemistry (2019), doi: https://doi.org/10.1016/ j.foodchem.2019.125536

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Elsevier Ltd. All rights reserved.

1 2 3 4 5 6

Phase behavior and complex coacervation of concentrated pea protein isolate‒beet pectin solution

7 Yang Lana, Jae-Bom Ohmb, Bingcan Chena, Jiajia Raoa

8 9 10 11

a. Food Ingredients and Biopolymers Laboratory, Department of Plant Sciences, North Dakota State University, Fargo, ND 58102, USA b. Edward T. Schafer Agricultural Research Center, Cereal Crops Research Unit, Hard

12

Spring and Durum Wheat Quality Lab, USDA-ARS, Fargo, North Dakota 58108, United

13

States

14 15 16

Revised Manuscript Submitted to: Food Chemistry

17

Submission Date: May, 2019

To

whom correspondence should be addressed. Tel: 1 (701) 231-6277. Fax: 1 (701) 231-7723. E-mail: [email protected].

1

19

Abstract

20

This study aimed to develop an alternative method named state diagram to identify boundary

21

formation pH for complex coacervates between pea protein isolate (PPI) and sugar beet pectin

22

(SBP) at concentrated solutions (~2.0 wt%). The effects of pH (7–2) and PPI–SBP mixing ratios

23

(1:1–20:1) on coacervates formation were investigated by state diagram, zeta-potential,

24

rheological, and phase composition analysis. Isothermal titration calorimetry, scanning electron

25

microscopy, and Fourier transform infrared spectroscopy were employed to elucidate

26

thermodynamic behaviors, non-covalent bonding of coacervates, and microstructure of

27

coacervates. We demonstrate that state diagram can explicitly identify the three characteristic pH

28

values (pHφ1, pHopt, and pHφ2) at which recognizable transitions take place in concentrated

29

colloids solutions. The mixing ratio dependent of pHφ1 increased to pH 5.5 as PPI–SBP mixing

30

ratio increased to 20:1. The pHopt was recognized at the net charge neutrality or the highest

31

storage modulus of the mixed colloids solutions.

32 33 34

Keywords: Plant protein; Pea protein isolate; Sugar beet pectin; Coacervates;

35

State diagram; Concentrated biopolymer solution

36 37 38 39 2

40

1 Introduction

41

Complex coacervation is a phase separation of a macromolecular solution and composed of two

42

oppositely charged biopolymers (e.g., protein‒polysaccharide, polyelectrolyte‒protein,

43

protein‒protein) based on the physicochemical mechanism of associative separation (Adal,

44

Sadeghpour, Connell, Rappolt, Ibanoglu, & Sarkar, 2017; Gomez-Mascaraque, Llavata-Cabrero,

45

Martinez-Sanz, Fabra, & Lopez-Rubio, 2018). In terms of protein‒polysaccharide system,

46

complex coacervation primarily occurs as a result of attractive electrostatic interactions of

47

charged groups between globular proteins and polysaccharides (Wagoner, Vardhanabhuti, &

48

Foegeding, 2016). The strength of such interactions is typically predominantly determined by the

49

parameters relating to the modulation of electrostatic interactions, including environmental pH,

50

biopolymer ratio, total biopolymer concentration, ionic strength, temperature, and

51

polysaccharides chemical/molecular compositions (e.g., molecular weight, chain branching,

52

charge density and distribution, etc.) (Kayitmazer, Koksal, & Iyilik, 2015; Kizilay, Seeman, Yan,

53

Du, Dubin, Donato-Capel, et al., 2014; Priftis & Tirrell, 2012; Samanta & Ganesan, 2018).

54

Protein‒polysaccharide complex coacervates have great potential applications in the food and

55

pharmaceutical industry, such as encapsulation of bioactive compounds (Gomez-Mascaraque, et

56

al., 2018), fat replacers (Krzeminski, Prell, Busch-Stockfisch, Weiss, & Hinrichs, 2014), films

57

fabrication (Silva, Fonseca, Amado, & Mauro, 2018), stabilization of emulsions (Rodriguez,

58

Binks, & Sekine, 2018), and protein purification (Li, Hua, Chen, Kong, & Zhang, 2016)

59

Many studies have investigated the phase behavior and the possible mechanism of complex

60

coacervates formation in protein‒polysaccharide system at low biopolymer concentrations

61

(typically below 0.3 wt%) (Aryee & Nickerson, 2012; Liu, Shim, Shen, Wang, & Reaney, 2017;

62

Liu, Cao, Ghosh, Rousseau, Low, & Nickerson, 2010). At a lower biopolymer concentration, the 3

63

turbidimetric analysis has long been applied to identify boundary pHs for phase

64

behaviors/transitions based on the optical properties of biopolymer mixtures at the mesoscale

65

regime (Wagoner, et al., 2016). When adjusting the pH of a mixed biopolymer solution from

66

neutral to acidic pH, four different pH regions can be divided with regarding to their interactions

67

by using three critical pH values (pHc, pHφ1, pHφ2), corresponding to four different phase

68

behaviors of biopolymer mixtures (pH > pHc, co-soluble; pHc~ pHφ1, soluble complexes; pHφ1~

69

pHφ2, complex coacervates; pH < pHφ2, dissolution of complexes) (Li, Zhang, Zhao, Ding, &

70

Lin, 2018). Between pHφ1 and pHφ2, a highest turbidity of biopolymer mixture is observed,

71

which is defined as the pH (pHopt), corresponding to the maximum amount of coacervates yield

72

is produced at pHopt.

73

Recently, zeta-potential values have been proposed as an alternative means to determine the

74

phase transition conditions for the formation of complex coacervation at lower biopolymer

75

concentrations. It should be noted that the application of protein–polysaccharide complex

76

coacervates generally requires higher biopolymer concentration in the area of encapsulation, fat-

77

mimetics, and texture modifiers (Klemmer, Waldner, Stone, Low, & Nickerson, 2012; Stenger,

78

Zeeb, Hinrichs, & Weiss, 2017). According to literature reports, a number of researchers have

79

identified the critical formation pHopt using abovementioned turbidimetric method at lower

80

biopolymer concentration. This critical pHopt was then directly applied to prepare complex

81

coacervates with higher biopolymer concentrations under the assumption that pHopt is

82

independent of biopolymer concentration (Stone, Teymurova, Chang, Cheung, & Nickerson,

83

2015). Our recent study has shown that the phase transition pHs were shifted on account of

84

varying biopolymer concentrations (Lan, Chen, & Rao, 2018). Meanwhile the turbidimetric

85

method is not applicable to identify critical formation pHs for high concentration of binary 4

86

protein/polysaccharide system, since the initial turbidity of system is so high. As such, an

87

alternative technique should be developed to properly identify critical pHφ1 and/or pHopt in

88

higher concentration biopolymer systems at which the maximum complexes coacervates are

89

formed (Hosseini, Emam-Djomeh, Razavi, Moosavi-Movahedi, Saboury, Atri, et al., 2013; Ye,

90

Flanagan, & Singh, 2006).

91

Recently, there is growing interest in the utilization of plant proteins, particularly yellow pea

92

protein isolate (PPI) in food application due to health benefits, cheaper price, allergen and

93

gluten-free claim that can be made. The major two fractions of PPI are globulin and albumin,

94

accounting for nearly 70–80% and 10–20% of the total pea protein, respectively (Zha, Dong,

95

Rao, & Chen, 2019). The most important globulins are legumin (11S, S = Svedberg Unit), vicilin

96

(7S) and convicilin (7–8S) (Lam, Can Karaca, Tyler, & Nickerson, 2018). To the best of our

97

knowledge, there is no study on fabricating complex coacervation between pea protein isolate

98

(PPI) and other polysaccharides in concentrated biopolymer systems. It is a challenge and of

99

most importance to be able to precisely identify the pH range for formation of complex

100

coacervates in order to understand their phase transition, structural aspects, and functional

101

properties in concentrated biopolymer systems.

102

In this study, we assembled coacervates-based biopolymer particles using in-house prepared pea

103

protein isolate (PPI) and anionic polysaccharide sugar beet pectin (SBP). The main purpose of

104

this study was to investigate factors (i.e. pH and biopolymer mixing ratios) that influence the

105

phase behavior of PPI‒SBP complex coacervates at concentrated solution (2.0 wt%), and to

106

evaluate the structural properties of these coacervates. We also aimed to develop an alternative

107

method, state diagram, to identify boundary formation pH for complex coacervates under high

108

biopolymer concentration. In addition, isothermal titration calorimetry (ITC), Fourier transform 5

109

infrared spectroscopy (FTIR), and scanning electron microscopy (SEM) were employed to gain

110

structural insights of how subtle pH differences could potentially impact the formation and

111

properties of PPI‒SBP complexes coacervates.

112

2 Materials and methods

113

2.1 Materials

114

Yellow pea flour was manufactured from organic yellow pea seeds and was purchased from

115

Harvest Innovations (Indianola, Iowa, USA). Freeze dried pea protein isolate (PPI) was extracted

116

from yellow pea flour using alkali extraction-isoelectric precipitated method, as stated in our

117

previous report without any modifications (Lan, et al., 2018). The triplicated proximate analyses

118

on final PPI powder were carried out according to AOAC methods (2016): 79.50% protein (%N

119

× 6.25, wet basis), 5.28% moisture, 0.77 % lipid, 4.61% crude ash, and 9.84% carbohydrate (by

120

difference from 100%). Sugar beet pectin (SBP) (Betapec RU301, Lot # 11710767, 45 kDa of

121

Mw, 55% of degree of esterification (DE), 65% of galacturonic acid, as reported by the

122

manufacturer) was kindly donated by Herbstreith & Fox KG (Neuenbürg, Germany) and used

123

without further purification. Analytical grade of sodium hydroxide (NaOH), hydrochloric acid

124

(HCl), and other chemicals and reagents were procured from VWR (Chicago, Illinois, USA). All

125

solutions was prepared using ultrapure water and the concentrations were as weight percentage

126

(wt%) (18.2MΩ·cm, Barnstead GenPure Pro water purification system, Thermo Fisher Scientific

127

Inc., USA).

128 129

2.2 Fabrication of complex coacervates 2.2.1 Preparation of PPI and SBP stock solutions

130

PPI stock solution (2.00 wt%) was prepared by dissolving powdered PPI in water and

131

magnetically stirred at 1000 rpm for 0.5 h at 25 °C. Subsequently, the pH of solutions was 6

132

increased to pH 9.5 with 0.1–2 M NaOH and stirred for 2 h to promote hydration, followed by

133

adjusting pH back to pH 7 using 0.1–2 M HCl and stirred for another 12 h to ensure complete

134

dissolution. SBP stock solution (2.00 wt%) was initially prepared by dispersing SBP in water and

135

stirring mechanically at 500 rpm for 12 h at 25 °C. The SBP solution was then adjusted to pH 7

136

using 0.1–2 M NaOH. Finally, both stock solutions were centrifuged at 5520×g for 15 min

137

(Beckman J2-HS, Beckman Coulter Inc., Indianapolis, IN, USA), and suction filtered through a

138

Whatman #1 filter paper to remove any insoluble matters. Sodium azide (0.02 wt%) was added

139

to both stock solutions to prevent microbial growth.

140

2.2.2 Preparation of PPI-SBP mixture

141

Fixed PPI solution (1.00 wt%) was mixed with SBP solution with concentration varying from

142

0.05 to 1.00 wt% to achieve the initial PPI–SBP ratio ranging from 1:1 to 20:1. The pH values of

143

mixture were then adjusted from 7 to 2 with 0.5 decrement by the addition of HCl with various

144

concentrations (0.1, 0.3, 0.6, 0.9, 2.0 M). The utilization of HCl at various concentrations was to

145

minimize dilution effects as well as conductivity change in mixture solution, as stated by Liu et

146

al (Liu, Low, & Nickerson, 2009). The single PPI and SBP solutions were used as controls. The

147

aqueous mixtures and controls were left to stand static for 24 h at 4 °C to allow phase

148

equilibrium.

149

2.2.3 Construction of state diagram

150

After standing static for 24 h at 4 °C, the state diagram of PPI–SBP mixtures was constructed

151

based on visual observation according to the method described in our recent work (Lan, Chen, &

152

Rao, 2018). Five different symbols were applied to distinguish different phase behaviors of the

153

observed phase separations in the test tubes, namely, ○, □, ▲, ■, ● to represent translucent

7

154

solution, turbid solution, precipitation & cloudy solution, precipitation & clear solution, and

155

precipitation & clear solution with higher volume of precipitation than ■, respectively.

156

2.3 Surface charge analysis

157

Surface charges of single biopolymer solutions (PPI or SBP, 1.00 wt%) and PPI–SBP mixtures

158

were measured using a Zetasizer Nano-ZS 90 (Malvern Instruments, Worcestershire, UK), and

159

reported as zeta-potential (ζ, mV).

160

2.4 Viscoelastic properties of complex coacervates

161

PPI-SBP coacervates were firstly centrifuged at 3220×g for 15 min, and the dynamic viscoelastic

162

properties of precipitates collected after centrifugation were evaluated using a Discovery Hybrid

163

Rheometer-2 rheometer (TA Instruments Ltd., New Castle, DE, USA) described by Liu et al

164

(Liu, et al., 2017) with some modifications. Briefly, coacervates obtained at different mixing

165

ratios (1:1 to 20:1) and pH conditions (pH 4.5–2.5) were placed on the bottom plate and a

166

solvent trap cover was applied to prevent evaporation during the measurement. All rheological

167

tests were carried out at 25 °C (regulated by a circulating water bath of Peltier system), using a

168

cone plate geometry (2°, 40 mm diameter) and a fixed gap (57 µm) between two plates. Samples

169

were equilibrated for 2 min before each measurement.

170

Two types of oscillatory analyses, strain sweep and frequency sweep, were conducted to

171

determine viscoelastic behavior of coacervates. Strain sweep was firstly performed to determine

172

the linear viscoelastic region (LVR), where dynamic G′ and G″ are independent of strain

173

amplitude. Each strain sweep test was performed at a constant frequency of 1 Hz and the strain

174

varying from 0.01 to 100%. Based on LVR, constant strain amplitude of 0.1% was applied for

175

subsequent frequency sweep measurements over an angular frequency range of 0.1–100 rad/s.

8

176

Data analysis was performed with TA TRIOS software (TA Instruments Ltd., New Castle, DE,

177

USA).

178

2.5 Quantification of PPI in coacervate phase

179

All samples were centrifuged (Eppendorf, New York, USA) at 3220×g for 15 min, and the

180

supernatant was recovered to analyze the protein content in accordance to the methods described

181

by Rocha et al. (Rocha, Souza, Magalhaes, Andrade, & Goncalves, 2014) and Warnakulasuriya

182

et al. (Warnakulasuriya, Pillai, Stone, & Nickerson, 2018) with some modifications. Briefly, the

183

percentage of PPI in the supernatant was calculated as: (PPI concentration of supernatant / PPI

184

concentration of non-centrifuged solution) × 100%. The PPI concentration was measured using

185

the Bradford dye-binding method as described in our previous report (Lan, et al., 2018). The PPI

186

in the coacervate phase was calculated as 100% ‒ percentage of PPI in the supernatant.

187

2.6 Isothermal titration calorimetry (ITC)

188

The ITC measurement was performed following the method reported by Bastos et al. (Bastos, de

189

Carvalho, & Garcia-Rojas, 2018) with some modifications. Briefly, the thermodynamic

190

parameters of PPI–SBP interaction at pH 3.5 were investigated using Nano ITC (TA instruments,

191

New Castle, USA) at 25 °C. The titration conditions, including biopolymer concentration and

192

titration volume, were optimized based on our preliminary experimental results. Solutions of PPI

193

(48.31 μM, 10 mg/mL) and SBP (7.41 μM, 0.33 mg/mL) were prepared by dispersing each of

194

them in citrate-phosphate buffer solution (pH 3.5) and mechanical stirring at 500 rpm for 12 h at

195

25 °C to ensure complete hydration. Prior to use, the solution was filtered through a 0.45 μm

196

syringe filter to avoid the pump blocking and degassed for 20 min under vacuum using an ITC

197

degassing apparatus (TA instruments, New Castle, USA). The pH 3.5 buffer was used to

9

198

eliminate a pH mismatch between a titrant (a syringe solution) and a titrand (a cell solution),

199

which composed of 7 mM citric acid and 5 mM dibasic sodium phosphate.

200

SBP, PPI, and buffer solutions were placed in 190 μL of reaction cell, 50 μL of syringe, and

201

reference cell, respectively. The PPI solution was added through 18 successive 2.5 μL injections,

202

with an interval of 200 s between injection, while the first injection was 1 μL and did not collect

203

into result. The stirring speed was set at 350 rpm during titration. The dilution heat from the

204

blank titration was determined by titrating PPI solution into the buffer, then subtracted from raw

205

data to determine the corrected enthalpy changes.

206

Thermograms were recorded by NanoAnalyze software (TA instruments, New Castle, USA).

207

The binding isotherms were obtained by integrating isotherm peaks and subtracting the dilution

208

heat from blank. The binding isotherms were presented as enthalpy changes vs PPI-SBP mole

209

ratios. Thermodynamic parameters including binding stoichiometry (n), binding affinity constant

210

(Ka), reaction enthalpy (ΔH) and change of entropy (TΔS) were calculated by iterative curve

211

fitting of the binding isotherms. The Gibbs free energy change (ΔG) was calculated from the

212

equation (ΔG = ΔH ‒ TΔS), where T is the absolute temperature.

213

2.7 Scanning electron microscopy (SEM)

214

A small volume of coacervates was centrifuged at 3220×g for 15 min and collected precipitate

215

was frozen in slushy nitrogen (‒210 °C), and then lyophilized, according to a method developed

216

by Phadungath (Phadungath, 2005) with slight modifications. The dried coacervates were

217

fractured manually and fragments were attached to cylindrical aluminum mounts with colloidal-

218

silver paste (Structure Probe Inc., West Chester PA, USA) oriented to view the fractured surface.

219

The powdered control samples (PPI and SBP) were attached to adhesive carbon tab on

220

cylindrical aluminum mounts, and the excess was blown off with a stream of nitrogen gas. All

10

221

samples were then sputter coated with gold (Cressington 108auto, Ted Pella, Redding, CA,

222

USA) to make them electrically conductive and finally examined in a scanning electron

223

microscope (JEOL Model JSM-6490LV, Peabody, MA, USA) using an accelerating voltage of

224

15 kV. Micrographs (5000×) were presented as a representative of each sample.

225

2.8 Fourier transform infrared spectroscopy (FTIR)

226

The dried powder of coacervates was obtained by centrifugation (Eppendorf, New York, USA)

227

at 3220×g for 15 min, and the collected precipitate was lyophilized for 48 h (laboratory scale

228

lyophilizer, SP scientific, Gardiner, New York, USA). FTIR spectra of dried coacervates and

229

controls (PPI and SBP powder) were recorded according to our previous published paper without

230

any modification (Lan, Xu, Ohm, Chen, & Rao, 2019).

231

2.9 Statistical analysis

232

A completely randomized design (CRD) with two replicates was applied to arrange all

233

experimental variants. All measurements were performed three times and data was expressed as

234

mean ± SD. One-way analysis of variance (ANOVA) and F-protected LSD by SAS software

235

(Version 9.3, SAS Institute Inc., NC, USA) were conducted, and significant difference was

236

defined at p < 0.05 by Tukey’s test.

237

3 Results and discussion

238

3.1 Effect of pH and mixing ratios on complex coacervation formation

239

3.1.1 State diagram

240

The first question that should be posed here is the boundary pH range of the complex coacervates

241

at concentrated PPI–SBP solutions. As one can see from Fig. 1b, an example of PPI–SBP

242

mixtures prepared at a total biopolymer concentration of 1.20 wt% showed a distinct cloudy

243

appearance in pH range of 6.5–4.5, which represents a solution with undetectable turbidity value. 11

244

This is the major reason that classic turbidimetric method is inappropriate to identify the

245

boundary pH range of complex coacervates, especially the pH with a maximum optical

246

absorbance (pHopt). As such, the phase behaviors of PPI–SBP mixtures after standing at 25 ℃ for

247

24 h as a function of pH and mixing ratios were observed and constructed as a state diagram in

248

Fig. 1. SBP solutions as a function of pH were not included in this state diagram as they

249

presented a translucent appearance throughout pH 7–2. Five distinguishable phase behaviors and

250

appearances were presented in the state diagram (Fig. 1).

251

Figure 1 inserted

252

At neutral pH, two phases of PPI-SBP mixtures or one phase of PPI–SBP solutions were

253

observed depending on the total biopolymer concentrations. For example, when the total

254

biopolymer concentration was > 1.50 wt %, PPI and SBP repelled each other in the ratios of 1:1

255

and 2:1 at pH 7 due to the thermal incompatibility (Kontogiorgos, Tosh, & Wood, 2009). As

256

such, it leads to a phase separation into a PPI-enriched phase and a SBP-enriched phase. In

257

contrast, translucent appearance of one phase solutions were observed in rest of tested mixing

258

ratios (5:1, 10:1, and 20:1) at the same pH, indicating the co-solubility of PPI and SBP (Liu,

259

Shim, Wang, & Reaney, 2015; Zhang, Xu, Li, Wang, Xue, & Xue, 2018). As the pH decreased

260

to acidic conditions, the turbid without precipitation (□), corresponding to the soluble complexes,

261

was observed within a wide pH range of 6–4 in ratios of 1:1 and 2:1 or a narrow pH range of

262

6.5–6 in ratios of 5:1–20:1, suggesting the initiation of electrostatic interactions between PPI and

263

SBP. The further decrease in pH caused the formation of insoluble complexes (coacervates or

264

precipitates) that associated with three various phase behaviors (Fig. 1). According to a

265

prodigious amount of literature on complex coacervation involving oppositely charged protein-

266

polysaccharide, it is generally accepted that the pH regions for complex coacervates starts from

12

267

pHφ1, and then transits to the end of the plateau (pHφ2) using turbidimetric method at low

268

biopolymer concentrations (Ach, Briançon, Dugas, Pelletier, Broze, & Chevalier, 2015;

269

Moschakis & Biliaderis, 2017). In this work, precipitates in three phase behaviors (▲, ■, and ●

270

in Fig 1) were all defined as coacervates with different structural and functional properties. The

271

pH region of coacervates that was labelled using symbol ranging from ▲ to ● in the current

272

study was similar to pHφ1 to pHφ2 at low biopolymer concentration system. The phase behavior

273

of precipitation & clear solution (■) in the state diagram was matches the pHopt that was

274

identified at low biopolymer concentration system. The relationship between the phase behavior

275

of complex coacervates with different visual phase separation and structural, rheological

276

properties were discussed in the following sections. When biopolymer mixture pH was further

277

reduced to the most acidic pH 2, complex coacervates were completely vanished and translucent

278

appearance of one phase solution were observed again, which could be attributed to the pH is

279

below the pKa value of SBP at pH 2 (Fig. 2a). There was no association between PPI and SBP at

280

acidic pH 2.

281

In terms of mixing ratios, the pH region of complex coacervates expanded to that of larger pH

282

regions with the increase of PPI to SBP ratios. For example, boundary pH region 3–2.5 of

283

complex coacervates expanded from pH 3-2.5 to a range of pH 5.5–2.5, when PPI to SBP mixing

284

ratio increased from 1:1 to 20:1 (Fig 1a). This also indicated that higher concentration of protein

285

with lower concentration of sugar beet pectin favors the formation of complex coacervates.

286

Similar phase behaviors and pH-mixing ratio relationship determined via state diagram were also

287

documented in the systems of whey protein isolate (WPI)–gum arabic (GA) (Ach, et al., 2015),

288

soybean protein–chitosan (Yuan, Wan, Yang, & Yin, 2014), and milk protein–tragacanth gum

289

(Azarikia & Abbasi, 2016). 13

290

3.1.2 ζ-potentials of PPI-SBP mixture as a function of pH

291

As formation of complex coacervates are driven by electrostatic interactions, the ζ-potential

292

could provide information regarding the interactions between charged proteins and

293

polysaccharides as well as formation and stability of coacervates. Hence, the ζ-potential of

294

homogeneous PPI, SBP, and PPI–SBP mixture was determined as a function of pH (7–2) and

295

mixing ratios (1:1–20:1) (Fig. 2a).

296

Figure 2 inserted

297

As can be seen from Fig. 2a, the PPI–SBP mixtures displayed intermediate ζ-potentials in

298

between the two individual biopolymers in solution (Fig. 2a). In general, complex coacervation

299

requires opposite charges while this is not the case for soluble complex formation. An overall

300

positively charged pea protein (< pH 4.7) with a negative charge patch could have electrostatic

301

attraction with a cationic SBP to form complex coacervates. Coacervates start to form as the net

302

charge of the biopolymer mixture approaches zero. Therefore, the isoelectric point (IEP) was

303

used to represent the pH at which the zero net charge of a biopolymer mixture was achieved, and

304

the influence of biopolymer mixing ratios on IEP was shown in Fig. 2b. At low mixing ratios of

305

1:1 and 2:1, the presence of SBP shifted the IEP of PPI from 4.7 to nearly 2.5 (Fig. 2b). Mixtures

306

at these two ratios exhibited similar surface charges to that of SBP solution with ζ-potentials

307

remaining negatively over the pH range of 7–2.5 (Fig. 2a). As the mixing ratio increased, IEP

308

values shifted towards higher pH values, reaching an IEP of 4.2 at a mixing ratio of 20:1, which

309

was close to that of PPI alone (Fig. 2b). Similar trend in IEP shift caused by biopolymer

310

interaction was also reported in pea protein-high methoxyl pectin system (Pillai, Stone, Guo,

311

Guo, Wang, & Nickerson, 2019). The IEP results along with state diagram showed that the

312

overall systems were charge neutral at maximum complexation, which is in agreement with

14

313

results reported by Skelte et al. on complex coacervates of lactotransferrin and β-lactoglobulin

314

(Anema & de Kruif, 2014). Charge neutralization can be achieved by altering the charge of one

315

or both biopolymer macroions, or changing the mixing ratios within the protein‒polysaccharide

316

system (Lan, et al., 2018). Above observations testified that PPI–SBP complex coacervates

317

formation occurred largely via electrostatic interactions between the cationic amino groups of

318

pea protein and anionic carboxyl groups of pectin.

319

On the basis of ζ-potential, strength of electrostatic interaction (SEI) was applied to further

320

elucidate the type (attractive or repulsive) and magnitude of forces during complex coacervation

321

(Chang, Gupta, Timilsena, & Adhikari, 2016; Espinosa-Andrews, Enríquez-Ramírez, García-

322

Márquez, Ramírez-Santiago, Lobato-Calleros, & Vernon-Carter, 2013; Timilsena, Wang,

323

Adhikari, & Adhikari, 2016; Yuan, et al., 2014). The SEI of PPI–SBP solutions and PPI at

324

different pHs were calculated as the multiplication of ζ-potential values of PPI and SBP or PPI

325

itself at each pH (Fig. 2a). The positive SEI values indicated that there was an overall repulsive

326

force between biopolymers. In contrast, an attractive force between biopolymers was existed

327

when SEI value was negative. As seen from Fig. 2a, PPI solutions showed an overall repulsive

328

force throughout pH 7–2, with the lowest SEI nearby pH 4.5 (IEP of PPI), and increased SEI as

329

pH was away from IEP. This explains well the phase behaviors of PPI (Fig. 1a) that the highest

330

precipitation occurred at pH 4.5 due to the lowest repulsive forces. By contrast, SEI of PPI–SBP

331

mixtures indicated that the system was dominated by the overall repulsive forces at pH above

332

IEP of PPI, or by attractive forces below the IEP with a highest attractive force at pH 3.5 (Fig.

333

2a). This helps mostly to explain the phase behavior that dense and maximum coacervates were

334

produced around pH 3–4 in PPI–SBP mixtures at a mixing ratio of 5:1 (Fig. 1a).

335

3.1.3 Viscoelastic properties of complex coacervates 15

336

Viscoelastic behaviors (storage modulus G′ and loss modulus G″ versus angular frequency) of

337

coacervates provide knowledge about network structure and play important roles in determining

338

their applications as they affect product texture (Liu, et al., 2017). It is commonly reported that

339

G′ is able to indicate strength of electrostatic interaction between protein and polysaccharide, the

340

higher G′ representing greater electrostatic interaction (Espinosa-Andrews, Sandoval-Castilla,

341

Vázquez-Torres, Vernon-Carter, & Lobato-Calleros, 2010; Liu, et al., 2017; Raei, Rafe, &

342

Shahidi, 2018). Therefore, all coacervates presented in state diagram (Fig. 1) were collected by

343

centrifugation and used in the analysis of viscoelastic properties.

344

Figure 3 inserted

345

First, the effect of pH on viscoelastic properties was evaluated by preparing coacervates at varied

346

pH (pH 2–4.5), but fixed PPI-SBP mixing ratio of 5:1 (Fig. 3a&b). Storage modulus G′ of all

347

PPI–SBP coacervates was higher than loss modulus G″ without a crossover point over the

348

frequency range of 0.01–100 rad/s, implying a highly interconnected gel-like network structure

349

of coacervates (Raei, et al., 2018). Moreover, pH plays a major role in viscoelasticity of

350

coacervates since both dynamic G′ and G″ of coacervates increased as the pH approached pH

351

3.5. The highest G′ and G″ were found at coacervates formed at pH 3.5, close to IEP of PPI–SBP

352

system at mixing ratio of 5:1 (Fig. 2a). This finding indicated that the maximum electrostatic

353

attraction occurred at the pH where PPI–SBP mixture electroneutrality was achieved, and caused

354

the development of strongest entangled gel-like network structures. Similar findings were

355

reported in mixture system of WPI‒flaxseed gum (FG) (J. Liu, et al., 2017) and WPI‒high

356

methoxyl pectin (HMP) (Raei, et al., 2018).

357

Second, in order to address the impact of mixing ratios on the viscoelasticity of PPI–SBP

358

coacervates, the viscoelasticity of coacervates obtained at studied mixing ratios (1:1–20:1) was 16

359

tested, and only the samples at a mixing ratio that had the highest G′ and G″ (maximum

360

electrostatic interaction) were presented in Fig. 3c&d. As one can see, both dynamic moduli (G′

361

and G″) were relatively higher at mixing ratios of 20:1, 10:1, and 5:1, indicating that complex

362

coacervates with stronger gel-like network structures were more easily formed with higher

363

weight percentage of protein. This can be explained by the fact that higher solid mass of protein

364

can lead to higher viscoelasticity.

365

3.1.4 PPI content in complex coacervates

366

Influence of pH and mixing ratio on the content of PPI in coacervates was investigated (Table

367

1). Presumably, the percentage of PPI in the coacervates could be considered as the amount of

368

PPI interacting with SBP after equilibrium. The coacervates were collected at the bottom of the

369

tubes, and the protein contents both in the total sample and the supernatant were measured. The

370

PPI content in coacervation phase was calculated from the difference.

371

Table 1 inserted

372

For instance, at a mixing ratio of 5:1, the PPI content retained in coacervates was increased

373

firstly to a maximum of 96.6% when the pH decreased from pH 4.5 to 3.5, and then reduced to a

374

lower percentage of 57.3% as the pH further dropped to pH 2. This finding, in conjunction with

375

the observations in state diagram, ζ-potential, and rheological tests, proved that maximum

376

associative interaction between PPI and SBP occurred at pH 3.5 (■). The coacervates collected at

377

pH 3.5 had relative denser texture, smaller pores appearance, maximum SEI, and highest storage

378

modulus at a mixing ration of 5:1. (Fig. 1, 2, 3 and 6).

379

Additionally, throughout pH 7–2 at a mixing ratio of 1:1, the highest PPI content in coacervates

380

was around 59% and occurred at pH 2.5. As PPI–SBP mixing ratio increased to 2:1, the protein

381

percentage raised abruptly to 96% and occurred at higher pH value of 3.0. As mixing ratio 17

382

further increased to 20:1, no significant changes in PPI content, except that the occurrence pH

383

continuously shifted towards higher pH was observed. The shift in pH values from pH 2.5 to 4.5

384

was similar to boundary pH shift as seen in state diagram (Fig. 1), which again illustrated that

385

the pH for maximum production of coacervates shifted as the increase of PPI–SBP mixing ratio.

386

3.2 Thermodynamic characterization of interactions between PPI and SBP

387

In order to further understand interactions of PPI–SBP, including binding model and

388

thermodynamics of complex coacervation process, ITC was applied to reveal molecular

389

recognition by titrating PPI with SBP at pH 3.5 and 25 ºC, and to record the released or absorbed

390

heat in a reaction chamber. The typical thermogram of heat rate versus time profile was shown in

391

the top panel of Fig. 4.

392

Figure 4 inserted

393

The titration profile was exothermic, and the exothermic peaks decreased regularly to a steady

394

state after 14th injection of SBP with molar ratio of around 1.6, indicating that the binding

395

between PPI and SBP was saturated. The exothermic reaction was mainly attributed to the

396

nonspecific electrostatic neutralization of the opposite charges carried by two biopolymers

397

demonstrating an enthalpic contribution of complex coacervation (Hadian, Hosseini, Farahnaky,

398

Mesbahi, Yousefi, & Saboury, 2016; Li, et al., 2018). The regular decrease of peaks was

399

associated to a reduction in free PPI remaining in the reaction cell after successive injections,

400

along with a reduction of dilution heat (Li, et al., 2018). Similar exothermic sequences were also

401

found in mixing processes of lactoferrin–sodium alginate (Bastos, et al., 2018), and fish skin

402

gelatin–GA (Li, et al., 2018).

403

The binding isotherm (bottom panel in Fig. 4) was fitted from thermogram using “one

404

independent binding site” model since not two inflection points were observed in binding 18

405

isotherm (Xiong, Ren, Jin, Tian, Wang, Shah, et al., 2016). The related thermodynamic

406

parameters were calculated based on binding isotherm and as follows: binding affinity

407

stoichiometry (n = 1.09±0.09), binding affinity constant (Ka = 4.61 ×106±1.34 ×106 M-1),

408

reaction enthalpy (ΔH = -68.32±0.59 kJ/mol), change of entropy (TΔS = -30.34±1.32 kJ/mol),

409

and change of Gibbs free energy (ΔG = -37.98±0.74 kJ/mol). The value of n indicated that 1.1

410

PPI molecules were bound to one SBP molecule. The binding constant was on the order of 106

411

M−1, indicating strong molecular interactions between PPI and SBP (Bastos, et al., 2018). The

412

negative value of ΔG indicated a spontaneous nature of the interaction between these two

413

biopolymers. Both ΔH and TΔS values were negative which serves as the evidence for the

414

enthalpically favorable and entropically unfavorable of PPI–SBP complexation process. It also

415

implies that hydrogen bonding, electrostatic interactions, and hydrophobic interactions were

416

involved in molecular interactions (Bastos, et al., 2018; Clitor, da Costa, Souza, Simas Tosin, &

417

Garcia-Rojas, 2018; Yang, Liu, Zeng, & Chen, 2018).

418

3.3 Morphological characterization of complex coacervates (Dense colloids vs Soft colloids)

419

3.3.1 Microstructure analysis

420

As above discussed, PPI–SBP coacervates prepared at pH values of 3.5 (■) and 2.5 (●) with

421

PPI–SBP ratio of 5:1 showed significantly different phase behavior and viscoelasticity.

422

Therefore, the microstructure of these two samples were further characterized using SEM to

423

reveal the interior structure to a great part (Fig. 5a-d). PPI prepared at pH 4.5 and SBP powder

424

were applied as controls.

425

Figure 5 inserted

426

PPI collected at its IEP (pH 4.5) presented as clusters with dense three-dimensional networks

427

(Fig. 5a), similar to morphology of isoelectric point precipitated and freeze-dried milk proteins 19

428

(Phadungath, 2005). The highly condensed structure could be attributed to aggregation of pea

429

protein particles at pH where electroneutrality occurred. The SBP powders showed a column/

430

plate-shaped bulky morphology intermingled with small pieces (Fig. 5b).

431

SEM results showed that the coacervates structures were critically dependent on environmental

432

pH, and their evolution could be described in two stages. To begin with, when pH at 3.5, the

433

neutral aggregates of pea protein and the hydrophobic interaction strengthen the complex

434

coacervates between PPI and SBP. Thus, the network structures of coacervates become more

435

condensed and much less organized. At a later stage of coacervation when pH at 2.5(●), no

436

neutral aggregates of pea protein are involved. As such the network structures are highly

437

organized with relatively homogeneous large pore size distribution and smooth inner pore

438

surfaces (Fig. 5d). This interpretation is corroborated by ζ-potentials, state diagram, and

439

viscoelastic measurements that coacervates at pH 3.5 had higher SEI, storage modulus, and

440

denser phase behavior than those at pH 2.5. The relationship between coacervate microstructure

441

and viscoelasticity were similar to previous reports (Anvari, Pan, Yoon, & Chung, 2015;

442

Espinosa-Andrews, et al., 2010; Huang, Du, Xiao, & Wang, 2017). Such findings have led to

443

calls to optimize the application of complex coacervates, particularly in the area of encapsulation

444

since coacervates obtained at pH 2.5 in the current study would have better encapsulation

445

capacity than those prepared at pH 3.5.

446

3.3.2 FTIR analysis

447

Furthermore, FTIR analysis was carried in parallel with SEM image to examine the non-covalent

448

bonding (i.e. electrostatic interactions and hydrogen bonding) between functional groups of PPI

449

and SBP at different pH of coacervates (i.e. pH 3.5 and 2.5).

20

450

The spectrum of PPI was consistent to our previous report (Lan, Xu, Ohm, Chen, & Rao, 2019),

451

with strong –OH contraction vibration band at 3275 cm-1, C–H stretching at 2926 cm-1, and

452

stretching or bending of C=O, N–H and C–N bonds at 1633, 1529, 1392 cm-1 (features of amide

453

I, II and III, respectively), and C–O vibration stretching at 1063 cm-1 (Fig. 5e). The spectrum of

454

SBP exhibited typical absorption bands of polysaccharides at 3388 cm-1 (–OH broad band), at

455

2931 cm-1 (C–H stretching vibration), at 1620 cm-1 (–COO- symmetric stretching vibration), and

456

at 1014 cm-1 (C–O–C stretching vibration) (Fig. 5e). These findings were confirmed from

457

previous reports (Mata, Blázquez, Ballester, González, & Muñoz, 2009; Synytsya, Čopı́ková,

458

Matějka, & Machovič, 2003).

459

Overall, the spectra of PPI–SBP complex coacervates were dominated by the pea protein due to

460

higher protein ratio (5:1), but different from each individual biopolymer (Fig. 5e). For example,

461

a peak of PPI (2926 cm-1) alone also presented in coacervates produced at both pH 3.5 and 2.5,

462

whereas a peak of PPI (1311 cm-1) and two peaks of SBP (1369 and 1323 cm-1) were

463

disappeared in both coacervates, confirming that complex coacervates with a different structure

464

were formed. In addition, the spectra of coacervates changed significantly in the carbonyl-amide

465

region. On one hand, both PPI-SBP coacervates did not show the symmetric –COO- stretching

466

vibration (1620 cm-1) that occurs in SBP. On the other hand, both coacervates showed shift in

467

peaks of amide I, II, and III towards higher wavenumbers as compared to PPI due to the

468

electrostatic interaction between the amino groups of PPI (–NH3+) and carboxylic groups of SBP

469

(–COO-). Similar results were also reported in egg white protein‒xanthan gum and hsian-tsao

470

gum-chitosan system (Souza & Garcia-Rojas, 2017; You, Liu, & Zhao, 2018).The introduction

471

of SBP could also contribute to an increase in hydrogen bonding of complex coacervates (Li, et

472

al., 2018). This result was in agreement with other studies in complex coacervation of chia seed 21

473

protein isolate–chia seed gum (Timilsena, et al., 2016) and lactoferrin–sodium alginate (Bastos,

474

et al., 2018). The spectra of PPI–SBP complex coacervates in “fingerprint region” presented a

475

broad band (approximately 1020–1080 cm-1) which was probably attributed to spectra

476

superposition of PPI (peak at 1063 cm-1, C–O stretching) and SBP (peak 1014 cm-1, C–O–C

477

stretching), implying an electrostatic interaction between PPI and SBP during the formation of

478

complex coacervates (Bastos, et al., 2018). Furthermore, environmental pH (pH 3.5 and 2.5)

479

imposed slightly peak shifts on complex coacervates (Fig. 5e). Compared to PPI, coacervates

480

formed at pH 3.5 had lower amide III peak shift towards higher wavenumbers than those

481

prepared at pH 2.5 (4 vs. 18 cm-1) while both coacervates had same C–O–C peak shift towards

482

higher wavenumbers compared to SBP. These results indicated that the overall intermolecular

483

interaction may be affected by environmental pH conditions (Li, et al., 2018).

484

4 Conclusions

485

This work testifies the feasibility of state diagram for identifying boundary formation pH region

486

(pHφ1 and pHφ2) of complex coacervates in protein-polysaccharide system at a relatively high

487

concentration of biopolymers. The boundary pH region identified by state diagram is in line with

488

zeta-potential, rheological, and phase composition measurements. The pHopt can also be

489

identified using both zeta-potential and rheological measurements. In general, pHφ1 shifts from 3

490

to 5.5 as the PPI–SBP mixing ratio increases from 1:1 to 20:1. PPI–SBP complex coacervation is

491

a spontaneous exothermic process and mainly formed through the electrostatic interaction and

492

hydrogen bonding between nonspecific amino groups of PPI and carboxylic groups of SBP. Both

493

SEM and FTIR results show that the coacervates structures critically depend on the pH.

494

Coacervates possess stronger and denser structure with greater storage modulus at the pHopt

495

where electroneutrality is achieved. The smooth inner pore surfaces with homogeneous large 22

496

pore size distribution of PPI‒SBP microstructure could be formed at the late stage of coacervates

497

when the environmental pH is near the pHφ2. Of particular note in the findings from the current

498

study is that it is the prerequisite to optimize pH and mixing ratio of protein‒polysaccharide

499

complexation for desired microstructure and viscoelastic properties, which can be accomplished

500

using state diagram developed here. Ongoing work will apply these new understandings for the

501

application of PPI–SBP coacervates in the area of microencapsulation.

502

Conflict of Interest

503

There are no conflicts to declare.

504

Acknowledgements

505

This work is supported by USDA National Institute of Food and Agriculture (Hatch ND1594). A

506

scholarship for Mr. Yang Lan is provided by the China Scholarship Council (CSC). We also

507

thank Dr. Sanku Mallik in School of Pharmacy at NDSU for his assistance on ITC experiment.

508 509

References

510

Ach, D., Briançon, S., Dugas, V., Pelletier, J., Broze, G., & Chevalier, Y. (2015). Influence of

511

main whey protein components on the mechanism of complex coacervation with acacia

512

gum. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 481, 367-374.

513

Adal, E., Sadeghpour, A., Connell, S., Rappolt, M., Ibanoglu, E., & Sarkar, A. (2017).

514

Heteroprotein complex formation of bovine lactoferrin and pea protein isolate: A

515

multiscale structural analysis. Biomacromolecules, 18(2), 625-635.

516 517

Anema, S. G., & de Kruif, C. G. (2014). Complex coacervates of lactotransferrin and betalactoglobulin. Journal of Colloid and Interface Science, 430, 214-220.

23

518

Anvari, M., Pan, C.-H., Yoon, W.-B., & Chung, D. (2015). Characterization of fish gelatin–gum

519

arabic complex coacervates as influenced by phase separation temperature. International

520

Journal of Biological Macromolecules, 79, 894-902.

521

Aryee, F. N. A., & Nickerson, M. T. (2012). Formation of electrostatic complexes involving

522

mixtures of lentil protein isolates and gum arabic polysaccharides. Food Research

523

International, 48(2), 520-527.

524 525 526

Azarikia, F., & Abbasi, S. (2016). Mechanism of soluble complex formation of milk proteins with native gums (tragacanth and Persian gum). Food Hydrocolloids, 59, 35-44. Bastos, L. P. H., de Carvalho, C. W. P., & Garcia-Rojas, E. E. (2018). Formation and

527

characterization of the complex coacervates obtained between lactoferrin and sodium

528

alginate. International Journal of Biological Macromolecules, 120(Pt A), 332-338.

529

Chang, P. G., Gupta, R., Timilsena, Y. P., & Adhikari, B. (2016). Optimisation of the complex

530

coacervation between canola protein isolate and chitosan. Journal of Food Engineering,

531

191, 58-66.

532

Espinosa-Andrews, H., Enríquez-Ramírez, K. E., García-Márquez, E., Ramírez-Santiago, C.,

533

Lobato-Calleros, C., & Vernon-Carter, J. (2013). Interrelationship between the zeta

534

potential and viscoelastic properties in coacervates complexes. Carbohydrate Polymers,

535

95(1), 161-166.

536

Espinosa-Andrews, H., Sandoval-Castilla, O., Vázquez-Torres, H., Vernon-Carter, E. J., &

537

Lobato-Calleros, C. (2010). Determination of the gum arabic–chitosan interactions by

538

Fourier Transform Infrared Spectroscopy and characterization of the microstructure and

539

rheological features of their coacervates. Carbohydrate Polymers, 79(3), 541-546.

24

540

Gomez-Mascaraque, L. G., Llavata-Cabrero, B., Martinez-Sanz, M., Fabra, M. J., & Lopez-

541

Rubio, A. (2018). Self-assembled gelatin-iota-carrageenan encapsulation structures for

542

intestinal-targeted release applications. Journal of Colloid and Interface Science, 517,

543

113-123.

544

Hadian, M., Hosseini, S. M. H., Farahnaky, A., Mesbahi, G. R., Yousefi, G. H., & Saboury, A.

545

A. (2016). Isothermal titration calorimetric and spectroscopic studies of β-lactoglobulin-

546

water-soluble fraction of Persian gum interaction in aqueous solution. Food

547

Hydrocolloids, 55, 108-118.

548

Hosseini, S. M. H., Emam-Djomeh, Z., Razavi, S. H., Moosavi-Movahedi, A. A., Saboury, A.

549

A., Atri, M. S., & Van der Meeren, P. (2013). β-Lactoglobulin–sodium alginate

550

interaction as affected by polysaccharide depolymerization using high intensity

551

ultrasound. Food Hydrocolloids, 32(2), 235-244.

552

Huang, G. Q., Du, Y. L., Xiao, J. X., & Wang, G. Y. (2017). Effect of coacervation conditions

553

on the viscoelastic properties of N,O-carboxymethyl chitosan - gum arabic coacervates.

554

Food Chemistry, 228, 236-242.

555

Kayitmazer, A. B., Koksal, A. F., & Iyilik, E. K. (2015). Complex coacervation of hyaluronic

556

acid and chitosan: effects of pH, ionic strength, charge density, chain length and the

557

charge ratio. Soft Matter, 11(44), 8605-8612.

558

Kizilay, E., Seeman, D., Yan, Y. F., Du, X. S., Dubin, P. L., Donato-Capel, L., Bovetto, L., &

559

Schmitt, C. (2014). Structure of bovine beta-lactoglobulin-lactoferrin coacervates. Soft

560

Matter, 10(37), 7262-7268.

25

561

Klemmer, K. J., Waldner, L., Stone, A., Low, N. H., & Nickerson, M. T. (2012). Complex

562

coacervation of pea protein isolate and alginate polysaccharides. Food Chemistry, 130(3),

563

710-715.

564 565 566

Kontogiorgos, V., Tosh, S. M., & Wood, P. J. (2009). Phase behaviour of high molecular weight oat β-glucan/whey protein isolate binary mixtures. Food Hydrocolloids, 23(3), 949-956. Krzeminski, A., Prell, K. A., Busch-Stockfisch, M., Weiss, J., & Hinrichs, J. (2014). Whey

567

protein–pectin complexes as new texturising elements in fat-reduced yoghurt systems.

568

International Dairy Journal, 36(2), 118-127.

569 570 571

Lam, A. C. Y., Can Karaca, A., Tyler, R. T., & Nickerson, M. T. (2018). Pea protein isolates: Structure, extraction, and functionality. Food Reviews International, 34(2), 126-147. Lan, Y., Chen, B., & Rao, J. (2018). Pea protein isolate–high methoxyl pectin soluble complexes

572

for improving pea protein functionality: Effect of pH, biopolymer ratio and

573

concentrations. Food Hydrocolloids, 80, 245-253.

574

Lan, Y., Xu, M., Ohm, J.-B., Chen, B., & Rao, J. (2019). Solid dispersion-based spray-drying

575

improves solubility and mitigates beany flavour of pea protein isolate. Food Chemistry,

576

278, 665-673.

577

Li, X., Hua, Y., Chen, Y., Kong, X., & Zhang, C. (2016). Protein Selectivity Controlled by

578

Polymer Charge Density and Protein Yield: Carboxylated Polysaccharides versus

579

Sulfated Polysaccharides. Journal of Agricultural and Food Chemistry, 64(47), 9054-

580

9062.

581

Li, Y., Zhang, X., Zhao, Y., Ding, J., & Lin, S. (2018). Investigation on complex coacervation

582

between fish skin gelatin from cold-water fish and gum arabic: Phase behavior,

583

thermodynamic, and structural properties. Food Research International, 107, 596-604.

26

584

Liu, J., Shim, Y. Y., Shen, J. H., Wang, Y., & Reaney, M. J. T. (2017). Whey protein isolate and

585

flaxseed (Linum usitatissimum L.) gum electrostatic coacervates: Turbidity and rheology.

586

Food Hydrocolloids, 64, 18-27.

587

Liu, J., Shim, Y. Y., Wang, Y., & Reaney, M. J. T. (2015). Intermolecular interaction and

588

complex coacervation between bovine serum albumin and gum from whole flaxseed

589

(Linum usitatissimum L.). Food Hydrocolloids, 49, 95-103.

590

Liu, S. H., Cao, Y. L., Ghosh, S., Rousseau, D., Low, N. H., & Nickerson, M. T. (2010).

591

Intermolecular interactions during complex coacervation of pea protein isolate and gum

592

arabic. Journal of Agricultural and Food Chemistry, 58(1), 552-556.

593

Liu, S. H., Low, N. H., & Nickerson, M. T. (2009). Effect of pH, salt, and biopolymer ratio on

594

the formation of pea protein isolate-gum arabic complexes. Journal of Agricultural and

595

Food Chemistry, 57(4), 1521-1526.

596

Mata, Y. N., Blázquez, M. L., Ballester, A., González, F., & Muñoz, J. A. (2009). Sugar-beet

597

pulp pectin gels as biosorbent for heavy metals: Preparation and determination of

598

biosorption and desorption characteristics. Chemical Engineering Journal, 150(2), 289-

599

301.

600

Moschakis, T., & Biliaderis, C. G. (2017). Biopolymer-based coacervates: Structures,

601

functionality and applications in food products. Current Opinion in Colloid & Interface

602

Science, 28, 96-109.

603

Phadungath, C. (2005). The mechanism and properties of acid-coagulated milk gels.

604

Songklanakarin Journal of Science and Technology, 27(2), 433-448.

27

605

Pillai, P. K. S., Stone, A. K., Guo, Q., Guo, Q., Wang, Q., & Nickerson, M. T. (2019). Effect of

606

alkaline de-esterified pectin on the complex coacervation with pea protein isolate under

607

different mixing conditions. Food Chemistry, 284, 227-235.

608 609 610 611 612

Priftis, D., & Tirrell, M. (2012). Phase behaviour and complex coacervation of aqueous polypeptide solutions. Soft Matter, 8(36), 9396-9405. Raei, M., Rafe, A., & Shahidi, F. (2018). Rheological and structural characteristics of whey protein-pectin complex coacervates. Journal of Food Engineering, 228, 25-31. Rocha, C. M. R., Souza, H. K. S., Magalhaes, N. F., Andrade, C. T., & Goncalves, M. P. (2014).

613

Rheological and structural characterization of agar/whey proteins insoluble complexes.

614

Carbohydrate Polymers, 110, 345-353.

615

Rodriguez, A. M. B., Binks, B. P., & Sekine, T. (2018). Emulsion stabilisation by complexes of

616

oppositely charged synthetic polyelectrolytes. Soft Matter, 14(2), 239-254.

617

Samanta, R., & Ganesan, V. (2018). Influence of protein charge patches on the structure of

618

protein-polyelectrolyte complexes. Soft Matter, 14(46), 9475-9488.

619

Silva, K. S., Fonseca, T. M. R., Amado, L. R., & Mauro, M. A. (2018). Physicochemical and

620

microstructural properties of whey protein isolate-based films with addition of pectin.

621

Food Packaging and Shelf Life, 16, 122-128.

622

Souza, C. J. F., da Costa, A. R., Souza, C. F., Simas Tosin, F. F., & Garcia-Rojas, E. E. (2018).

623

Complex coacervation between lysozyme and pectin: Effect of pH, salt, and biopolymer

624

ratio. International Journal of Biological Macromolecules, 107, 1253-1260.

625

Souza, C. J. F., & Garcia-Rojas, E. E. (2017). Interpolymeric complexing between egg white

626

proteins and xanthan gum: Effect of salt and protein/polysaccharide ratio. Food

627

Hydrocolloids, 66, 268-275.

28

628

Stenger, C., Zeeb, B., Hinrichs, J., & Weiss, J. (2017). Formation of concentrated biopolymer

629

particles composed of oppositely charged WPI and pectin for food applications. Journal

630

of Dispersion Science and Technology, 38(9), 1258-1265.

631

Stone, A. K., Teymurova, A., Chang, C., Cheung, L., & Nickerson, M. T. (2015). Formation and

632

functionality of canola protein isolate with both high- and low-methoxyl pectin under

633

associative conditions. Food Science and Biotechnology, 24(4), 1209-1218.

634 635 636

Synytsya, A., Čopı́ková, J., Matějka, P., & Machovič, V. (2003). Fourier transform Raman and infrared spectroscopy of pectins. Carbohydrate Polymers, 54(1), 97-106. Timilsena, Y. P., Wang, B., Adhikari, R., & Adhikari, B. (2016). Preparation and

637

characterization of chia seed protein isolate-chia seed gum complex coacervates. Food

638

Hydrocolloids, 52, 554-563.

639

Wagoner, T., Vardhanabhuti, B., & Foegeding, E. A. (2016). Designing whey protein-

640

polysaccharide particles for colloidal stability. In M. P. Doyle & T. R. Klaenhammer

641

(Eds.), Annual Review of Food Science and Technology, Vol 7, vol. 7 (pp. 93-116). Palo

642

Alto: Annual Reviews.

643

Warnakulasuriya, S., Pillai, P. K. S., Stone, A. K., & Nickerson, M. T. (2018). Effect of the

644

degree of esterification and blockiness on the complex coacervation of pea protein isolate

645

and commercial pectic polysaccharides. Food Chemistry, 264, 180-188.

646

Xiong, W. F., Ren, C., Jin, W. P., Tian, J., Wang, Y. T., Shah, B. R., Li, J., & Li, B. (2016).

647

Ovalbumin-chitosan complex coacervation: Phase behavior, thermodynamic and

648

rheological properties. Food Hydrocolloids, 61, 895-902.

29

649

Yang, J., Liu, G., Zeng, H., & Chen, L. (2018). Effects of high pressure homogenization on faba

650

bean protein aggregation in relation to solubility and interfacial properties. Food

651

Hydrocolloids, 83, 275-286.

652

Ye, A. Q., Flanagan, J., & Singh, H. (2006). Formation of stable nanoparticles via electrostatic

653

complexation between sodium caseinate and gum arabic. Biopolymers, 82(2), 121-133.

654 655 656

You, G., Liu, X. L., & Zhao, M. M. (2018). Preparation and characterization of hsian-tsao gum and chitosan complex coacervates. Food Hydrocolloids, 74, 255-266. Yuan, Y., Wan, Z.-L., Yang, X.-Q., & Yin, S.-W. (2014). Associative interactions between

657

chitosan and soy protein fractions: Effects of pH, mixing ratio, heat treatment and ionic

658

strength. Food Research International, 55, 207-214.

659

Zha, F., Dong, S., Rao, J., & Chen, B. (2019). Pea protein isolate-gum arabic Maillard

660

conjugates improves physical and oxidative stability of oil-in-water emulsions. Food

661

Chemistry, 285, 130-138.

662

Zhang, T., Xu, X., Li, Z., Wang, Y., Xue, Y., & Xue, C. (2018). Interactions and phase behaviors

663

in mixed solutions of kappa-carrageenan and myofibrillar protein extracted from Alaska

664

Pollock surimi. Food Research International, 105, 821-827.

665 666 667 668 669 670 671 672 673 674

Figure Captions Fig. 1 (a) State diagram of PPI, and PPI-SBP mixtures at different mixing ratios during acid titration. (b) The appearance of PPI-SBP mixture (5:1 mixing ratio) as function of pH. The PPI concentration was fixed at 1 wt%. The appearance was observed after 24h standing. ○ represents translucent solution; □ represents turbid solution; ▲ represents precipitation and cloudy solution; ■ represents precipitation and clear solution; ● represents precipitation and clear solution that has higher volume of precipitation than ■.

30

675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693

Fig. 2 (a) Zeta (ζ)-potential changes of PPI, SBP, PPI-SBP mixtures at different mixing ratios and the strength of electrostatic interaction (SEI) of PPI-PPI and PPI-SBP during acid titration. (b) Plot of isoelectric point (IEP) against PPI-SBP mixing ratio (including data points of pure SBP and PPI). The PPI concentration was fixed at 1 wt%. Fig. 3 The (a) storage modulus G′ and (b) loss modulus G″ versus angular frequency for PPISBP coacervates prepared at different pH values (the PPI-SBP mixing ratio was fixed at 5:1). The (c) G′ and (d) G″ of coacervates prepared at different mixing ratios and at the pH that the highest G′ occurred within each mixing ratio. The PPI concentration was fixed at 1 wt%. Fig. 4 Thermogram (top panel) and binding isotherm (bottom panel) corresponding to the titration of PPI (48.31 μM) with SBP (7.41 μM) dispersion in citrate-phosphate buffer (pH 3.5) at 25 °C. Fig. 5 SEM micrographs (5000×) of (a) PPI collected at pH 4.5, (b) SBP powder, and PPI-SBP coacervates collected at (c) pH 3.5 and (d) pH 2.5; (e) FTIR spectra of the coacervates prepared at pH 3.5 and 2.5. The PPI-SBP mixing ratio was fixed at 5:1, and PPI concentration was fixed at 1 wt%. Bar in SEM images represents 5 µm.

31

694 695 696 697 698 699 700

Fig. 1 (a) State diagram of PPI, and PPI-SBP mixtures at different mixing ratios during acid titration. (b) The appearance of PPI-SBP mixture (5:1 mixing ratio) as function of pH. The PPI concentration was fixed at 1 wt%. The appearance was observed after 24 h standing. ○ represents translucent solution; □ represents turbid solution; ▲ represents precipitation and cloudy solution; ■ represents precipitation and clear solution; ● represents precipitation and clear solution that has higher volume of precipitation than ■.

32

701

702 703 704 705 706 707

Fig. 2 (a) Zeta (ζ)-potential changes of PPI, SBP, PPI-SBP mixtures at different mixing ratios and the strength of electrostatic interaction (SEI) of PPI-PPI and PPI-SBP during acid titration. (b) Plot of isoelectric point (IEP) against PPI-SBP mixing ratio (including data points of pure SBP and PPI). The PPI concentration was fixed at 1 wt%.

33

708 709 710 711 712

Fig. 3 The (a) storage modulus G′ and (b) loss modulus G″ versus angular frequency for PPISBP coacervates prepared at different pH values (the PPI-SBP mixing ratio was fixed at 5:1). The (c) G′ and (d) G″ of coacervates prepared at different mixing ratios and at the pH that the highest G′ occurred within each mixing ratio. The PPI concentration was fixed at 1 wt%.

34

713

714 715 716 717 718

Fig. 4 Thermogram (top panel) and binding isotherm (bottom panel) corresponding to the titration of PPI (48.31 μM) with SBP (7.41 μM) dispersion in citrate-phosphate buffer (pH 3.5) at 25 °C.

35

719

720 721 722 723 724 725

Fig. 5 SEM micrographs (5000×) of (a) PPI collected at pH 4.5, (b) SBP powder, and PPI-SBP coacervates collected at (c) pH 3.5 and (d) pH 2.5; (e) FTIR spectra of the coacervates prepared at pH 3.5 and 2.5. The PPI-SBP mixing ratio was fixed at 5:1, and PPI concentration was fixed at 1 wt%. Bar in SEM images represents 5 µm.

36

726 727 728

Graphical abstract

729 730 731 732 733 734 735 736 737

Table Table 1. Percentage (%) of PPI in PPI‒SBP coacervate phases collected at different mixing ratios during acid titration†. pH 4.5 4.0 3.5

1:1 – – –

PPI‒SBP mixing ratio 2:1 5:1 10:1 – 57.2±6.6a 97.0±0.7a – 91.8±0b 97.4±0.3a 65.8±1.0a 96.6±0.1b 93.3±1.7a

20:1 97.6±0.7a 96.8±0.3a 85.1±2.0b 37

738 739 740 741

3.0 33.1±4.0a 96.3±0.2b 87.5±1.7b 75.7±0b 67.8±5.6c 2.5 58.7±6.8b 71.2±2.9a 68.6±3.1a 56.8±3.2c 50.9±1.9d 2.0 – – 57.3±10.1a 43.5±2.8d – †The concentration of PPI was fixed at 1 wt%. Values are expressed as mean ± standard deviation of at least three different samples. Values with different superscript letters in the same column are significantly different at p<0.05.

742 743 744 745 746 747

Highlights    

Concentrated pea protein isolate -pectin mixture was characterized by state diagram The maximum coacervation yield was recognized at the net charge neutrality The microstructure of coacervates was influenced by pH The electrostatic interaction and hydrogen bonding were involved in complexation

748 749 750 751 752 753 754

Declaration of interest The authors are not aware of any affiliations, memberships, funding, or financial holdings that might be affecting the objectivity of this article.

38