Specific Salt Effects on the Formation and Thermal Transitions Among Β-Lactoglobulin and Pectin Electrostatic Complexes

Purdue University Purdue e-Pubs

Open Access Theses Theses and Dissertations

2013 Specific altS Effects on the Formation and Thermal Transitions Among β-Lactoglobulin and Pectin Electrostatic Complexes Stacey Ann Hirt Purdue University, [email protected]

Follow this and additional works at: https://docs.lib.purdue.edu/open_access_theses Part of the Agriculture Commons, Bioresource and Agricultural Engineering Commons, Food Science Commons, and the Physical Chemistry Commons

Recommended Citation Hirt, Stacey Ann, "Specific altS Effects on the Formation and Thermal Transitions Among β-Lactoglobulin and Pectin Electrostatic Complexes" (2013). Open Access Theses. 34. https://docs.lib.purdue.edu/open_access_theses/34

This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact [email protected] for additional information. Graduate School ETD Form 9 (Revised 12/07) PURDUE UNIVERSITY GRADUATE SCHOOL Thesis/Dissertation Acceptance

This is to certify that the thesis/dissertation prepared

Stacey A. Hirt By

Entitled Speciﬁc Salt Effects on the Formation and Thermal Transitions Among β-Lactoglobulin and Pectin Electrostatic Complexes

Master of Science For the degree of

Is approved by the final examining committee:

Owen Jones

Chair Yuan Yao

Ganesan Narsimhan

To the best of my knowledge and as understood by the student in the Research Integrity and Copyright Disclaimer (Graduate School Form 20), this thesis/dissertation adheres to the provisions of Purdue University’s “Policy on Integrity in Research” and the use of copyrighted material.

Owen Jones Approved by Major Professor(s): ______

Approved by: Rengaswami Chandrasekaran 09/27/2013 Head of the Graduate Program Date

SPECIFIC SALT EFFECTS ON THE FORMATION AND THERMAL TRANSITIONS AMONG β-LACTOGLOBULIN AND PECTIN ELECTROSTATIC COMPLEXES

A Thesis

Submitted to the Faculty

Purdue University

Stacey Ann Hirt

In Partial Fulfillment of the

Requirements for the Degree

of i Master of Science

December 2013

Purdue University

West Lafayette, Indiana

For my parents- Denny and Jackie.

You may not understand what’s written on the following pages, but I owe it all to you

iii

ACKNOWLEDGEMENTS

I would like to thank my parents for absolutely everything. Words cannot describe how grateful I am for the countless opportunities I’ve been given. Thanks for the gift of my education—the years I’ve spent at Villa and Purdue have shaped who I am. Thank you for pushing me hard, never letting me quit, and always expecting the best. I have learned more about myself, who I am, and who I want to be during graduate school than any other point in my life. I am so incredibly fortunate to have such amazing parents. I love you both.

Thanks to my major professor, Owen Jones, for answering my endless barage of questions, teaching and training me especially during the first few months, and being beyond patient with me the past two years. I have learned so much during graduate school and I am repeatedly astounded by your knowledge of our field. Thank you to my iii committee members, Dr. Yao and Dr. Narsimhan, for the new perspectives you’ve brought to my project. Thanks to all the faculty and staff of the Purdue food science department who have helped me build a home and truly find my niche over the past six years.

Thanks to the best lab partner a girl could ever ask for, Laura Zimmerer. Our constant dancing and singing in the lab helped me stay sane during graduate school. You were always there for me whether I needed a bit of motivation, time to vent, or a drink.

I am so happy to have made a lifelong friend during graduate school. And if food science doesn’t pan out it’s nice to know we have our singing career or multiple business ideas to fall back on.

Thanks to my friends, both in graduate school and away. Your words of encouragement and commiseration meant more than you will know. Thanks to Simran

Kaur for being an amazing officemate and Alyssa Beatty for always forcing me to leave my desk during lunch. I couldn’t have made it through my Master’s without the help of my friends. And to my sisters, for listening to me complain, picking up the poor graduate student’s checks on vacations, and giving me a dose of city life outside of Lafayette.

Finally, thank you to Kanye West. His music got me through countless hours of writing and lab work. There will never be a better album than The College Dropout. Ever.

“See I know my destination, but I’m just not there.” –Kanye West

TABLE OF CONTENTS

Page LIST OF TABLES ...... viii LIST OF FIGURES ...... ix LIST OF ABBREVIATIONS ...... xii ABSTRACT ...... xiii CHAPTER 1. INTRODUCTION ...... 1 1.1 Introduction ...... 1 1.2 Objectives ...... 4 CHAPTER 2. LITERATURE REVIEW ...... 5 2.1 Introduction ...... 5 2.2 Protein-Polysaccharide Complex Constituents ...... 5 2.2.1 Proteins ...... 6 2.2.1.1 β-Lactoglobulin ...... 9 2.2.2 Polysaccharides ...... 10 2.2.2.1 Pectin ...... 11 2.3 Interactions Influencing the Formation of Protein-Polysaccharide Complexes 12 2.3.1 Electrostatic Interactions ...... 13

2.3.2 Non-Electrostatic Interactions ...... 14 v

2.3.2.1 Hydrogen Bonds ...... 15 2.3.2.2 Hydrophobic Interactions ...... 15 2.3.2.3 Covalent Bonds ...... 15 2.4 Factors Influencing Complex Formation ...... 16 2.4.1 pH ...... 16 2.4.2 Ionic Strength ...... 17 2.4.3 Protein-Polysaccharide Concentration ...... 17 2.4.4 Charge Density ...... 18 2.5 Specific Ion Effects ...... 18 2.5.1 Stabilization of Proteins ...... 19 2.5.1.1 Chaotropes and Kosmotropes ...... 20

Page 2.5.2 Nature of Electrolyte ...... 21 2.5.2.1 Anion and Cations ...... 21 2.5.3 Direct and Inverse Hofmeister Effects ...... 21 2.6 Applications for Protein-Polysaccharide Complexes ...... 23 2.6.1 Viscosifying and Gelling ...... 23 2.6.2 Foaming and Emulsifying ...... 24 2.6.3 Stabilization and Release of Bioactive Components ...... 25 2.7 Conclusions ...... 26 CHAPTER 3. SPECIFIC ION EFFECTS ON ELECTROSTATIC INTERACTIONS AND COACERVATION OF AQUEOUS β-LACTOGLOBULIN AND PECTIN MIXTURES ...... 27 3.1 Abstract ...... 27 3.2 Introduction ...... 28 3.3 Materials and Methods ...... 31 3.3.1 Materials ...... 31 3.3.2 Biopolymer Solution Preparation ...... 32 3.3.3 Colloidal Charge ...... 33 3.3.4 Solution Turbidity ...... 33 3.3.5 Light Scattering ...... 34 3.3.6 Determination of Critical pH Parameters ...... 34 3.3.7 Statistical Analysis ...... 35 3.4 Results and Discussion ...... 35 3.4.1 Surface Charge of Pure and Mixed Biopolymer in Solution ...... 36 3.4.2 Effect of pH and Salt on Low Concentration β-Lactoglobulin ...... 38

3.4.3 Effect of pH and Salt on Electrostatic Complex Formation ...... 42 vi 3.4.4 Specific Ion Effects on Coacervation Process ...... 48 3.5 Conclusions ...... 55 CHAPTER 4. THERMAL TRANSITIONS AMONG β-LACTOGLOBULIN-PECTIN ELECTROSTATIC COMPLEXES ...... 56 4.1 Abstract ...... 56 4.2 Introduction ...... 57 4.3 Materials and Methods ...... 58 4.3.1 Materials ...... 58 4.3.2 Heated Complex Creation ...... 59 4.3.3 Solution Turbidity ...... 60 4.3.4 Particle Size ...... 60 4.3.5 Turbidity Measurements with Increasing Thermal Treatment ...... 60 4.3.6 Particle Morphology using Atomic Force Microscopy (AFM) ...... 61

vii

Page 4.3.7 Statistical Analysis ...... 62 4.4 Results and Discussion ...... 62 4.4.1 Effect of Ionic Strength on β-Lactoglobulin-Pectin Particle Formation 63 4.4.2 Effect of Ionic Species on β-Lactoglobulin-Pectin Particle Formation . 74 4.5 Conclusion ...... 88 CHAPTER 5. CONCLUSIONS ...... 90 LIST OF REFERENCES ...... 93 APPENDICES Appendix A Chapter 3: Specific Ion Effects on Electrostatic Interactions and Coacervation of Aqueous β-Lactoglobulin and Pectin Mixtures ...... 98 Appendix B Chapter 4: Thermal Transitions Among β-Lactoglobulin-Pectin Electrostatic Complexes ...... 102

vii

viii

LIST OF TABLES

Table ...... Page

Table 3.1 pHc Values ...... 46

Table 3.2 pHΦ values ...... 51

Table 4.1 Z-Average diameter of thermally treated protein-polysaccharide complexes with potassium chloride ...... 69

Table 4.2 Temperature of aggregation of protein-polysaccharide complexes at pH 4.50 with 0-100 mm potassium chloride ...... 72

Table 4.3 Z-Average diameter of heated particles at pH 4.50 ...... 83

Table 4.4 Z-Average diameter of heated particles at pH 4.75 ...... 84 viii

LIST OF FIGURES

Figure ...... Page Figure 3.1 Dependence of zeta-potential on pH of biopolymer solutions ...... 37

Figure 3.2 Effect of salt concentration on turbidity of β-lactoglobulin solutions as a function of pH for (a) KCl and (b) KSCN ...... 39

Figure 3.3 Effect of salt concentration and species on light scattering of β-lactoglobulin and pectin solution with KCl as a function of pH ...... 43

Figure 3.4 Effect of salt concentration and species on light scattering of β-lactoglobulin and pectin solutions as a function of pH (a) 50 mm salt (b) 50 mm ionic strength ...... 44

Figure 3.5 Changes to pHc values as a function of salt concentration and species ...... 46

Figure 3.6 Effect of salt concentration and species on turbidity of β-lactoglobulin and pectin solutions with KCl as a function of pH ...... 49 ix

Figure 3.7 Effect of salt concentration and species on turbidity of β-lactoglobulin and pectin solutions as a function of pH (a) 50 mm salt (b) 50 mm ionic strength ...... 50

Figure 3.8 Changes to pHΦ values as a function of salt concentration and species ...... 52

Figure 4.1 Thermal transitions of a) β-lactoglobulin and b) β-lactoglobulin and pectin solution with KCl as a function of pH ...... 64

Figure 4.2 AFM amplitude images of thermally treated protein-polysaccharide complexes at pH 4.50 with KCl a) 0 mm b) 25 mm c) 50 mm d) 75 mm e) 100 mm ...... 68

Figure Page

Figure 4.3 Changes to the thermal profile of β-Lactoglobulin and pectin with varying concentrations of KCl at pH 4.50 ...... 72

Figure 4.4 Effect of ion species on turbidity of thermally treated β-lactoglobulin and pectin solutions as a function of pH at (a) 50 mm salt (b) 50 mm ionic strength ...... 76

Figure 4.5 Temperature of aggregation as a function of salt species and ionic strength at pH 4.50 for a) β-lactoglobulin solutions and b) β-lactoglobulin and pectin ...... 80

Figure 4.6 Size (Z-average diameter) of particle formed from thermally treated β- lactoglobulin and pectin as a function of salt species and ionic strength at pH 4.50 ...... 84

Figure 4.7 Size (Z-average diameter) of particles formed from thermally treated β- lactoglobulin and pectin as a function of salt species and ionic strength at pH 4.75 ...... 85

Figure 4.8 AFM images of thermally treated protein-polysaccharide complexes at pH

4.50 with 50 mm a) KCl b) NaCl c) KSCN d) NH4SCN e) K2SO4 f) (NH4)2SO4 ...... 87

Appendix Figure

Appendix A Figure 1 Effect of salt concentration on turbidity of β-lactoglobulin solutions x as a function of pH (a) NaCl (b) NH4SCN (c) K2SO4 (d) (NH4)2SO4 ...... 99

Appendix A Figure 2 Effect of salt concentration and species on light scattering of β- lactoglobulin and pectin solutions as a function of pH with (a) K2SO4 (b) (NH4)2SO4 .. 100

Appendix A Figure 3 Effect of salt concentration and species on turbidity of β- lactoglobulin and pectin solutions as a function of pH with (a) K2SO4 (b) (NH4)2SO4 .. 101

Appendix B Figure 1 AFM height images of protein-polysaccharide particles at pH 4.50 with KCl a) 0 mm b) 25 mm c) 50 mm d) 75 mm e) 100 mm ...... 101

Appendix Figure Page

Appendix B Figure 2 AFM height image and apparent size determination of thermally treated BLG and pectin particles with 0 mm KCl using Igor software ...... 101

Appendix B Figure 3 AFM height image and apparent size determination of thermally treated BLG and pectin particles with 50 mm KCl using Igor software ...... 104

Appendix B Figure 4 Thermal transitions β-lactoglobulin and pectin solution with a)

NaCl b) KSCN c) NH4SCN d) K2SO4 and e) (NH4)2SO4 as a function of pH ...... 107

Appendix B Figure 5 Changes to the thermal profile of β-Lactoglobulin with varying concentrations of KCl at pH 4.50 with temperature increase ...... 108

Appendix B AFM amplitude images of thermally treated protein-polysaccharide complexes containing β-Lactoglobulin and pectin at pH 4.50 with 16.67 mm a) K2SO4 b)

(NH4)2SO4 ...... 108

xii

LIST OF ABBREVIATIONS

AFM- atomic force microscopy

BLG- β-Lactoglobulin

DLS- dynamic light scattering

HCl- hydrochloric acid

K2SO4- potassium sulfate

KCl- potassium chloride

KSCN- potassium thiocyanate mm- millimolal

NaCl- sodium chloride

NaOH- sodium hydroxide

NH4SCN- ammonium thiocyanate xii

(NH4)2SO4- ammonium sulfate pI- isoelectric point

UV-Vis- Ultraviolet-Visible Spectroscopy

xiii

ABSTRACT

Hirt, Stacey A. M.S., Purdue University, December 2013. Specific Salt Effects on the Formation and Thermal Transitions among β-Lactoglobulin and Pectin Electrostatic Complexes. Major Professor: Owen Jones.

Factors of ion specificity and ionic strength (I~0-100) were studied in the electrostatic complex formation and protein particle formation by thermal treatment for a

β-lactoglobulin and pectin system. ζ-potential showed β-lactoglobulin and pectin began to interact near pH 5.50 and interactions were strengthened with decrease in pH. Visible light turbidimetry and light scattering at 90° revealed a trend in critical pH transitions for electrostatic complex formation based on both the ionic strength and the anion of the salt species, while effects of the monovalent cation was insignificant. Critical pH values for complex formation and separation (pHc and pHΦ) decreased with increasing ionic xiii

strength, with no significant differences seen between chloride and thiocyanate salts. The ionic strength equivalency of sulfate salts caused significant differences to pHc values, indicating that both ionic strength and specific ion effects influence complex formation.

Heat-treatment at 80°C of the β-lactoglobulin/pectin complexes at pH 4.50 led to the creation of particles with a diameter of 200-400. Light scattering revealed particles at pH

4.50 were significantly smaller than those at pH 4.75. Morphological characterization of particles at pH 4.50 with KCl concentrations of 50 mm and greater revealed disruption in particle structure from rounded to more amorphous shapes and possible flocculation of

xiv particles. Turbidimetry development of complex and heat-treated particles during heat treatment was significantly different from the protein, alone, demonstrating a specific shielding effect on protein-polysaccharide interactions. Ion effects on particle size of heat-treated complexes is a simple means to reliably control particle formation for purposes of controlled release or modified colloidal flow. xiv

CHAPTER 1. INTRODUCTION

1.1 Introduction

The changes on microscopic and macroscopic scales due to interactions of proteins and polysaccharides in solution can be manipulated for utilization in food and pharmaceutical applications. The nature of the proteins and polysaccharides allows them to form complexes with one another as a function of pH due to the reliance on electrostatic interactions for complex formation. Prior to industrial application, a deeper understanding of the factors affecting β-lactoglobulin-pectin complex formation and coacervation including pH, ionic strength, and salt species is necessary.

Specific ion effects have been proposed to explain the effects of salts on protein solutions, including the reoccurring trends of the Hofmeister series. The trends are due more to the anion of the salt than the cation, which follow a kosmotropic to chaotropic

2------series of SO4 , F , Cl , Br , ClO4 , SCN (Zhang & Cremer, 2006). The species listed prior 1

to chloride are known as kosmotropes while the species after chloride are known as chaotropes, with chloride falling in between. Kosmotropes stabilize protein structure leading to decreased solubility of proteins and increased protein aggregation and salting out of solution. Chaotropes destabilize protein structure, which results in greater solubility of proteins and salting in of proteins. The Hofmeister series has seen an increase in interest due to correlations beyond protein solubility including enzyme

2 activity, protein-protein interactions, protein crystallization, and bacterial growth (Zhang

& Cremer, 2006).

Previously, research has focused on the direct Hofmeister series for proteins in solution above the pI. More recently, studies have shown some protein-protein interactions in salt solutions follow an inverse Hofmeister series below the pI in solutions of low ionic strength (M Boström et al., 2005). In lysozyme solutions, charge screening effects that lead to the inverse series below the pI are only dominant below ~200-300 mM salt. Above this concentration, the physical properties of the solution follow a direct series (Zhang & Cremer, 2009).With the inverse Hofmeister series, chaotropes become more effective than kosmotropes at stabilizing protein structure leading to salting out.

Lysozyme has become the quintessential example for the inverse Hofmeister series

(Mathias Boström, Parsons, Salis, Ninham, & Monduzzi, 2011; Grigsby, Blanch, &

Prausnitz, 2001; Zhang & Cremer, 2009).

Complex formation of proteins and polysaccharides is possible due to electrostatic interactions. With a reduction in pH from neutral to more acidic conditions, interactions 2 between protein and anionic polysaccharide in solution increase at just above the isoelectric point (pI) of the protein and weak electrostatic complexes are formed.

Complex formation begins when polysaccharide and protein both have net negative charge, but interactions can occur between the polysaccharide and positive patches on the protein surface. At a critical pH, termed pHc, binding equilibrium between protein and anionic polysaccharide favors the formation of soluble complexes and the increase in colloidal dimensions leads to weak scattering of incident light (detectable by light scattering). The critical value, pHc, reveals the initiation of dominant electrostatic

3 interactions between the protein and polysaccharide and is therefore dependent on ionic strength, the isoelectric point of the protein, and the charge density of the polysaccharide

(Park, Muhoberac, Dubin, & Xia, 1992). Because complexes are held together by weak interactions, they are susceptible to dissociation by high ionic strength or slight increases in pH.

Further reduction in pH below pHc increases the net-cationic charge character of the protein and favors further complex formation with anionic polysaccharide that results in more intense scattering of incident light (detectable by UV-Vis spectroscopy). This critical value, termed pHΦ, results in insoluble complexes termed coacervates. For protein-polyanion systems, reduced pHΦ values indicate a decreased tendency for a protein and anionic polysaccharide to overcome the repulsions and interact to form a charge-neutralized, phase separating complex (F Weinbreck, de Vries, Schrooyen, & de

Kruif, 2003).

Previous research has shown that thermal treatment of BLG-pectin electrostatic complexes can form stable particles by physical cross-linking (Owen G. Jones, Decker, & 3

McClements, 2009). Heating of soluble complexes and coacervates (above the protein’s denaturation temperature) promotes globular protein unfolding and produces more stable particles that have a tendency to more intensely scatter light than their unheated counterparts. Due to the initial electrostatic complexation prior to thermal treatment, these particles are susceptible to changes in pH, ionic strength, and ionic species.

1.2 Objectives

The objective of the following experiments was to understand the effects of low ionic strength of various ion species with and without thermal treatment on dilute solutions of β-lactoglobulin and β-lactoglobulin and pectin and analyze changes to complex formation, coacervation, and thermal transitions of the system.

The goals of the following research were to: i) determine surface charge of individual biopolymer components and of the mixed system and ii) understand effects of pH, ionic strength, and salt type on β-lactoglobulin (BLG) and BLG-pectin electrostatic complex formation and coacervation process and iii) characterize changes to electrostatic complexes with thermal treatment. We hypothesize that both ionic strength and ionic species will affect BLG-pectin complexation, coacervation, and thermal transitions.

CHAPTER 2. LITERATURE REVIEW

2.1 Introduction

Electrostatic complexes and thermally treated particles are of increasing interest to the food and pharmaceutical industries due to their potential applications of stabilization and release of bioactives, viscosifying and gelling, and foaming and emulsifying. When composed of food grade materials, including proteins and polysaccharides, these complexes have the potential to be added to foods and drugs. Prior to application, additional understanding on the factors affecting complex formation and stability are necessary.

The goal of this review is to provide an overview of protein-polysaccharide electrostatic complex composition, interactions between macromolecules, factors affecting complexation, specific ion effects on complex formation, and potential 5 applications.

2.2 Protein-Polysaccharide Complex Constituents

Biopolymer particles can be composed of various food-grade components, notably proteins and polysaccharides. Many proteins and polysaccharides—including whey protein, casein, zein, gelatin, pectin, cellulose, and chitosan—have successfully been used to create biopolymer particles. The materials of composition will determine the functional properties of biopolymer complexes.

2.2.1 Proteins

Proteins are polymers composed of 20 different amino acids linked by substituted amide bonds. Difference in the sequence of amino acids will give rise to thousands of proteins with varying structure and therefore function. Proteins can be classified as globular or fibrous based on their structure. Globular proteins are those in a spherical or ellipsoidal shape due to the folding of the polypeptide chain on itself. Fibrous proteins are elongated chains with twisted linear polypeptide chains (Damodaran, 2008).

Amino acids consist of a central carbon atom covalently attached to a hydrogen atom, an amino group, a carboxyl group and a side chain denoted as the R group. The physicochemical properties of the protein including net charge, solubility, reactivity, and hydrogen bonding potential rely on the R group. Amino acids are ampholytes, meaning they behave both as acids and bases due to the presence of the carboxyl group and amino group, respectively. At around neutral pH, both the carboxyl and amino groups are ionized, and the amino acid is a zwitterion. The pH at which the dipolar ion is electrically neutral is called the isoelectric point (pI). When the pH is decreased below the pI, the 6

COO- group becomes protonated. When the pH is increased above the pI, the NH3+ group becomes deprotonated. In addition, there are seven side groups that have the ability to become ionized (Damodaran, 2008). The overall charge of a protein is a summation of both positive and negative charges in localized ‘patchy’ regions. This allows proteins to bind both positive and negative molecules in these regions even when the overall protein charge is positive or negative.

Proteins provide a wide array of functional properties in the food matrix. Proteins may be responsible for solubility, viscosity, water binding, gelation, elasticity,

7 emulsification, foaming, and fat and flavor binding (Damodaran, 2008). Many factors influence functionality including protein size, shape, and structure; amino acid composition and sequence; net and localized charge; hydrophobicity; and ability to interact with other components and solvents (Damodaran, 2008).

A major factor influencing protein function is the hydrophobicity of the constituent amino acids. The hydrophobic effect works to reduce unfavorable interactions between non-polar molecules and water. The most hydrophobic groups will often be found buried within a protein while the more hydrophilic groups will face outwards when in an aqueous phase. Hydrophobicity will impact a proteins ability to bind small nonpolar molecules (fatty acids, flavors, or surfactants) to adsorb onto non-polar surfaces in order to stabilize foams or emulsions, and to interact with other proteins (Vetri & Militello,

2005).

Amino acid hydrophobicity contributes to protein structure. Primary structure refers to the linear sequence of the amino acids. Secondary structure is the three- dimensional shape of local segments of the protein. The two most common forms of 7 secondary structures are helices or β-sheets. Tertiary structure is the spacial arrangement that arises when a linear protein chain with secondary structures folds into a compact three-dimensional form. Tertiary structure involves multiple forces (hydrophobic, electrostatic, van der Waals, and hydrogen bonding) in order to reduce the free energy.

Covalent disulfide bonds also work to stabilize globular protein structure by coupling thiol groups (Hoffmann & van Mil, 1999). Quaternary structure refers to the arrangement of a protein when it contains more than one polypeptide chain. Various proteins exist as dimers, trimers, tetramers, and so on. For example, β-lactolobulin exists

8 as a dimer in the pH range 5-8, as an octomer in the pH range 3-5, and as a monomer above pH 8 (Damodaran, 2008).

Proteins can serve an array of biological functions as enzyme catalysts, structural proteins, hormones, transfer proteins, antibodies, and storage proteins. In order to exhibit these functions a protein must be in its native state. The native state is thermodynamically the most stable state with lowest feasible free energy. A change in the environment, such as pH, temperature, ionic strength, or solvent characteristics, will force the molecule into a new equilibrium structure. Small changes to the environment allow the protein to adapt, although major changes will force denaturation of the protein. This denaturation will cause drastic changes in secondary, tertiary, and quaternary structure without altering primary structure. Denaturation of proteins will change the structure and therefore alter the functionality (Damodaran, 2008).

Food proteins are often found in their denatured state. While denaturation is sometimes reversible, it is unlikely denaturation of food proteins can be reversed. A common example, the heating of globular protein albumin in egg whites denatures the 8 protein and is irreversible. Denaturation changes the conformation of the protein and exposes the internal hydrophobic amino groups and sulfur-rich groups. Aggregation is then often accompanied due to hydrophobic bonding and disulfide bond formation.

Denaturation of globular can be carefully controlled to produce desired byproducts including fibrils (Arnaudov, Vries, Ippel, & Mierlo, 2003) (Otte, Ipsen, Bauer, Bjerrum,

& Waninge, 2005), and nanoparticles (Giroux, Houde, & Britten, 2010), and nanotubes

(Graveland-Bikker, Fritz, Glatter, & de Kruif, 2006).

Proteins are able to interact with a variety of molecules in close proximity including other proteins, solvents (both hydrophilic and hydrophobic), minerals, lipids, surfactants, and polysaccharides. Interactions include van der Waals, hydrogen bonding, hydrophobic interactions, and electrostatic interactions. In order to create biopolymer complexes and particles, it is important to understand both materials and physical forces that are involved in protein-polysaccharide interactions (Owen Griffith Jones, 2009).

2.2.1.1 β-Lactoglobulin

β-lactoglobulin (BLG) is a globular protein, which represents about 60% of the proteins in bovine whey. It has a radius of about 2 nm and a molecular weight of 18.2 kiloDaltons. It contains two disulfide bridges and one free thiol group (Nicolai, Britten, &

Schmitt, 2011).

BLG is stable in solution due to electrostatic repulsion, except near the isoelectric point (pI) of 5.2, where aggregation at room temperature can occur. This aggregagtion can be reversed once the pH has been increased or decreased away from the pI. The 9 extent and rate at which aggregation occurs near the pI were shown to strongly decrease when NaCl was added to solution at low ionic strengths (Majhi et al., 2006). Close to the pI, the protein has both positive and negative charges and interactions between oppositely charged areas cause protein aggregation, in addition to hydrophobic interactions and hydrogen bonds.

Native BLG is present in solution primarily as monomers and dimers, except in a small pH range near the pI or at high ionic strength (Nicolai et al., 2011). These oligomers dissociate into monomers at elevated temperatures (Verheul, Roefs, & Kruif,

1998). Upon heating above ≈50°C BLG partially unfolds exposing buried hydrophobic groups and the thiol group. The secondary backbone structure is still retained at this temperature, referred to as “the molten-globule state” (Ptitsyn, 1995). With increasing temperature, above the denaturation temperature ≈70°C at neutral pH, secondary structure is lost and intra- and intermolecular disulfide interchange reactions occur, causing irreversible denaturation and aggregation (Hoffmann & van Mil, 1999; Qi et al.,

1997; Verheul et al., 1998). The aggregates formed can have a wide range of shapes— including long strands, short curved strands, and spherical particles—and strongly depends on the pH and presence of added salt (Nicolai et al., 2011).

Due to the different structures and physicochemical properties that can arise from native and denatured BLG, there are many applications for BLG including foams and emulsions, films and coatings, and encapsulation. Encapsulation is of greatest relativity to the following project. Encapsulation using BLG has successfully been done with hydrogels (Gunasekaran, Ko, & Xiao, 2007), aggregated whey protein (Picot & Lacroix,

2004), and coupled with polysaccharide (Guerin, Vuillemard, & Subirade, 2003). 10

2.2.2 Polysaccharides

Polysaccharides are polymers of monosaccharaides composed of ~100-107 units, known as the degree of polymerization (DP). Most polysaccharides have DPs of 200-

3,000. Polysaccharides can ne neutral, cationic (positive), or anionic (negative). The composition, DP, and sequence of the comprising monosaccharides are responsible for the functional properties of polysaccharides. Polysaccharides can be extracted from plants (pectin, starch, cellulose), algae (carrageenan, alginates), or bacteria (xanthan).

They have a wide variety of molecular weights, conformations, branching, electrical characteristics, and hydrophobicity, which lead to differences in their functional properties (solubility, binding properties, viscosity enhancement, and gelation) (Owen

Griffith Jones, 2009). Due to the incredible diversity of physicochemical properties of polysaccharides, the remainder of this review will focus on pectin.

2.2.2.1 Pectin

Pectins are heteropolysaccharides composed of α-(1à4)-D-galactopyranosyl uronic acid with various contents of methyl ester groups (BeMiller & Huber, 2008).

While the linear chain composed of galacturonic acid units are the key feature of pectin, neutral sugars—primarily L-rhamnose—are also present. The native, more complex form of pectin is present in the cell walls and intercellular layers of all land plants. This form is converted into methyl-esterified galacturonoglycans during extraction with acid. Pectins can be isolated from multiple plant sources including citrus peels, apple pomace, and sugar beets. Pectins are often utilized in spreadable gels, formed in the presence of 11 calcium ions or sugar and acid dependent upon the structure of the pectin in use.

The compositions and properties of pectin vary with source, the processing conditions during extraction and preparation, and subsequent treatments (BeMiller &

Huber, 2008). High methoxyl (HM) pectins are those with more than half of the carboxyl groups are in the methyl ester form (-COOCH3) due to preparation. The remainder of carboxyl groups present is a mixture of free acid (-COOH) and salt (e.g. COO-Na+) forms.

Preparations that yield less than half of the carboxyl groups in the methyl ester forms are termed low methoxyl (LM) pectins. The percentage of carboxyl groups esterified with

12 methanol is the degree of esterification (DE) or degree of methylation (DM). Pectin with a DE over 50 is considered a high methoxyl pectin.

High-methoxyl pectin solutions gel when sufficient acid and sugar is present. As the pH of the pectin solution is lowered, the hydrated and charged carboxylate groups are converted to uncharged, slightly hydrated carboxylic acid groups. As a result of losing a portion of their charge and hydration, the polymer molecules can begin to associate over a portion of their length. These associations form junction zones that can entrap the aqueous solution of solute molecules. Junction zone formation is aided by the high concentration of sugar (at least ~55%), which competes with the pectin for the water molecules and reduces hydration leading to increased interaction between the pectin chains (BeMiller & Huber, 2008).

Low-methoxyl pectin gels only in the presence of divalent cations which are required to form crossbridges. Increasing the presence of the divalent cation (calcium is the only divalent cation used to form gels in food applications) increases the gelling temperature and gel strength (BeMiller & Huber, 2008). 12

2.3 Interactions Influencing the Formation of Protein-Polysaccharide Complexes

Formation of protein-polysaccharide complexes is dependent on interactions between the biopolymers. Interactions can be divided in three groups: between the charged macromolecules, between oppositely charged side groups, and between other side groups of polyions available for interaction (Tolstoguzov, 1997). Additionally, they can be classified as weak or strong, specific or nonspecific, and attractive or repulsive.

Electrostatic interactions are the most common forces involved in complex formation

13 between oppositely charged biopolymers (Schmitt, Sanchez, Desobry-Banon, & Hardy,

1998).

2.3.1 Electrostatic Interactions

Electrostatic interactions are the driving force of complex formation between oppositely charged proteins and polysaccharides. Protein molecules have a net positive charge and behave as polycations at pH values below the pI (de Kruif, Weinbreck, & de

Vries, 2004). At mildly acidic and neutral pH values typical of most food systems, carboxyl-containing polysaccharides behave as polyanions. Electectrostatic complex formation between proteins and anionic polysaccharides generally occurs in the pH range between the pKa value of the carboxyl groups on the polysaccharide and the pI of the protein (Tolstoguzov, 1997).

During formation of electrostatic complexes, the overall net charge of anionic polysaccharide decreases with the gradual attachment of each successive protein molecule. The reduction in net opposite charges on the reactants reduces the 13

hydrophilicity and solubility of the resulting complex (Tolstoguzov, 1997). The higher the content of polysaccharide, the pH at which the complex precipitates is decreased.

The process of neutralization of the macromolecules leads to an electrostatically neutral insoluble complex (Doublier, Garnier, Renarda, & Sanchezb, 2000). When the ratio of protein to polysaccharide is equivalent, the maximum amount of insoluble complex is produced (Schmitt et al., 1998). This leads to aggregation and precipitation of complex coacervate.

The formation of electrostatic complexes is usually a reversible process with changes to pH and ionic strength. Generally, complexes dissociate at ionic strengths greater than 0.2-0.3 M or when the pH is raised above the protein’s pI. At pH values above the pI, electrostatic interactions between anionic polysaccharides and positively charged subunits of proteins can still occur (Tolstoguzov, 1997).

Initial reports of electrostatic interactions between proteins and polysaccharides date back to 1911, when complexes of gelatin and acacia gum were formed. The stability of the complexes were dependent on the amount of acid added to the system, so it was surmised there was an electrostatic nature to the interactions (Tiebackx, 1911). Since these initial experiments, complexation between various proteins and polysaccharides has been studied. A wide variety of protein and polysaccharide sources have been studied— including egg albumin, casein, bovine serum albumin, gelatin, collagen, and β- lactoglobulin, with carrageenan, alginate, chitosan, carboxymethylcellulose, acacia gum, and pectin to name a few (Schmitt et al., 1998).

2.3.2 Non-Electrostatic Interactions

When proteins and polysaccharides come into contact, interactional forces other than electrostatic interaction create formation of junction zones between the biopolymers.

These forces depend on the structure and composition of the biopolymers present

(Schmitt et al., 1998).

2.3.2.1 Hydrogen Bonds

Hydrogen bonds refer to interaction of a hydrogen atom attached to an electronegative atom (i.e. nitrogen, oxygen or sulfur) with another electronegative atom

(i.e. oxygen of a carboxyl or carbonyl group). Hydrogen bonds allow weak soluble complex formation at pH values above the protein’s pI (Schmitt et al., 1998). This was demonstrated with pectin and gelatin, where the compatibility of the biopolymers increased with the increased degree of esterification of the pectin. The presence of a greater number of carboxyl groups allowed for greater interaction between the gelatin and pectin (Braudo & Antonov, 1993).

2.3.2.2 Hydrophobic Interactions

Hydrophobic interactions occur between nonpolar molecules in an effort to exclude water. These interactions are entropy driven and promoted by increases in temperature. Hydrophobic interactions result in conformational and structural changes of the biopolymers. These changes cause different macromolecule segments to come into 15 contact and interact. Hydrophobic interactions are critical in the thermal stabilization of protein-polysaccharide complexes (Schmitt et al., 1998).

2.3.2.3 Covalent Bonds

Covalent bonds are highly specific, non-electrostatic interactions involved in complex formation of proteins and polysaccharides. These bonds are often formed by a chemical reaction between the amino groups from the proteins and carboxylic groups from the polysaccharides to create an amide covalent bond. These bonds are favored by

16 the dehydration of the complex (through freeze-drying or heating) or the addition of a cross-linking agent. The covalent bonding is a way to make complexation irreversible and very stable to changes in pH and ionic strength (Schmitt et al., 1998).

2.4 Factors Influencing Complex Formation

Because complexation between proteins and polysaccharides is driven by electrostatic interactions, physicochemical parameters—including pH, ionic strength, protein-polysaccharide concentration, and charge density—will influence the formation of complexes. External factors or processing factors—including shear rate, shear time and pressure—can also impact complex formation although these will not be addressed in depth (Schmitt et al., 1998).

2.4.1 pH

Changes in pH may be used to promote association between proteins and polysaccharides through alteration in electrostatic interactions (Owen Griffith Jones & 16

McClements, 2010a). The pH is critical to the formation of protein-polysaccharide complexes because it affects the ionization degree of the functional side groups of the biopolymers (amino acids for proteins, carboxylic groups for polysaccharides) (Schmitt et al., 1998).

The charge on protein will change from negative to positive when pH is adjusted from above to below the protein’s isoelectric point. When coupled with an anionic polysaccharide, the protein and polysaccharide can begin to associate. Changing pH of solutions from neutral to near and below the pI of the protein is necessary to promote

17 electrostatic interactions which allows for complexation of BLG and pectin (Owen G.

Jones, Lesmes, Dubin, & McClements, 2010).

2.4.2 Ionic Strength

The formation of protein-polysaccharide complexes is dependent on electrostatic interactions between the biopolymers. The addition of microions to the solution can cause charge neutralization and hinder interactions between proteins and polysaccharides. At low ionic strength, there can be small changes to coacervation, although there are sufficient charges on the biopolymers for interactions between proteins and polysaccharides to still occur. At high salt concentrations, the net charge on the biopolymers is reduced by interactions with the ions in solution, resulting in reduced or complete inhibition of interactions, preventing complexation and coacervation. The addition of sodium chloride to many protein-polysaccharide systems has been studied, both at low ionic strengths and higher ionic strengths. In solutions of BLG and pectin at low ionic strengths, pHc and pHϕ values were decreased, while at high ionic strengths, 17

complexation was inhibited (Wang, Lee, Wang, & Huang, 2007; Wang, Wang,

Ruengruglikit, & Huang, 2007). To the author’s knowledge, no research has been published on complexation of BLG and pectin in the presence of salt types besides sodium chloride.

2.4.3 Protein-Polysaccharide Concentration

Both protein-polysaccharide concentration and the ratio of protein to polysaccharide are factors to consider during coacervation. For each protein-

18 polysaccharide system, there is an optimum ratio at which maximum coacervation is obtained. At a biopolymer concentration of 1%, the optimum protein to polysaccharide ratio for coacervation of solutions of BLG and acacia gum has been determined to be 4:1

(Schmitt et al., 1998). For ratios where one of the biopolymers is in excess, soluble complexes are formed because of nonneutralized charges, leading to decreases in turbidity. For solutions at very high biopolymer concentrations, phase separation due to thermodynamic incompatibility occurs because of competition between the macromolecules for solvent.

2.4.4 Charge Density

The charge density of biopolymers is defined as the number of charges present on the biopolymer per unit length. Charge density is important due to the influence of charge on initial complex formation and coacervation can’t occur under a critical charge density.

Complex formation at pH values above the pI, where both the anionic polysaccharide and protein have a net negative charge, can occur due to the presence of patches of protein 18

with increased positive charge density (Schmitt et al., 1998).

2.5 Specific Ion Effects

Specific ion effects have been proposed as an explanation of the changes caused to solutions with the addition of salts. Numerous examples of specific ion effects exist, including changes to thermodynamics, transport properties, and kinetics in simple or complex solutions and near interfaces between macromolecules such as protein and water

(Kunz, 2009).

The influence of salt on protein solutions has been studied since the 1880s, when

Hofmeister demonstrated that salts with a fixed cation, but varying anion, have different capacities in stabilizing protein suspensions. Depending on the anion, different concentrations were required to cause precipitation of the proteins. The salts were ordered based on their effectiveness to precipitate hen egg white, which later was deemed a universal series, known as the Hofmeister series, for a wide set of proteins (M Boström et al., 2005). Beyond protein solubility, the Hofmeister series has been used to explain changes to enzyme activity, protein-protein interactions, protein crystallization, optical rotation of sugar and amino acids, and bacterial growth. More extensive research has been conducted on the effects for macromolecules in aqueous solutions; the molecular level mechanisms of ions are just beginning to be understood (Zhang & Cremer, 2006).

2.5.1 Stabilization of Proteins

While it has been determined that specific ion effects and the Hofmeister series are ubiquitous in chemistry and biology, they were initially proposed to explain changes 19

to stabilization of proteins. Both anions and cations have been divided originally based on an ion’s ability to alter the hydrogen-bonding network of water. Following is a list of anions and cations are in sequence from kosmotropic to chaotropic. Anions to the left of chloride are kosmotropes, while chaotropes fall to the right of chloride. Chloride is a moderate anion, having characteristics of both kosmotropes and chaotropes. Sodium cations and those to the left are kosmotropic and potassium and ions to the right are chaotropic (Zhang & Cremer, 2006). Ions are most effective on the nearby water radius

20 when anions and cations are either both kosmotropes or both chaotropes (i.e. kosmotropic cation/kosmotropic anion or chaotropic cation/chaotropic anion).

In reference to water structure, bulk water—water outside the initial layers of hydration—is unaffected by the presence of low concentration of ions. Ions interact strongly with water over very short distances. Sensitive techniques including Fourier

Transformation Infrared Spectroscopy has shown well-defined radius of hydration with looser bound water molecules (Smiechowski & Stangret, 2006).

THE HOFMEISTER SERIES

KOSMOTROPIC CHAOTROPIC

(STABILIZING/SALTING OUT) (DESTABILIZING/SALTING)

2- 2- 2------Anions: CO3 SO4 S2O3 F- Cl Br NO3 I SCN

Cations: H+ Li+ Na+ K+ Rb+ Cs+

2.5.1.1 Chaotropes and Kosmotropes

Chaotropes are known to destabilize folded proteins and cause salting-in of 20

proteins, meaning greater solubility in solution. They were traditionally called ‘water structure breakers.’ The kosmotropes were believed to be ‘water structure makers,’ causing stabilization to proteins leading to salting out in solution due to decreased solubility. Kosmotropes, including the multivalent anions are strongly hydrated meaning they bind nearby water molecules tightly, immobilizing them whereas large monovalent ions (chaotropes) are able to ‘free up’ nearby water molecules (Collins, Neilson, &

Enderby, 2007).

2.5.2 Nature of Electrolyte

Specific ion effects refer to the relative effectiveness of either anions or cations, both individually and as ion pairs, on a wide range of phenomena. It is generally accepted that anions have a greater effect on protein solubility compared to the cation.

2.5.2.1 Anion and Cations

The hydrating effects of anions are more pronounced than those of cations. For example, fluorine (F-) and potassium (K+) are of approximately similar size and would be expected to have similar accessibility to water molecules in solution, but F- is strongly hydrated whereas K+ is weakly hydrated. The oxygen molecule on water is highly electronegative and readily accepts negative charge transfer from F-; the positive charge transfer from K+ to water is more difficult (Collins et al., 2007).

Another factor contributing to the stronger hydration of anions is the ability for anions to interact with hydrogen atoms of water, allowing hydrogen bonding with the first layer of hydration. Cations compete to interact with oxygen on water, preventing 21 hydrogen bonding (Collins et al., 2007).

2.5.3 Direct and Inverse Hofmeister Effects

The kosmotropic to chaotropic order listed previously for anions and cations has been the traditional (direct) view of the Hofmeister series. More current research has found an inverse in these characteristics at pH points below the pI of the protein when the ionic strength of the system is low. At high ionic strengths, the trends displayed follow a direct Hofmeister series. Above a protein’s pI, the macromolecules bear a net negative

22 charge and the direct Hofmeister series is observed. In this situation, chaotropes

(including SCN-) will lead to protein unfolding and increased solubility in solution while

2- the kosmotropes (including SO4 ) lead to the stabilization of proteins and decreased solubility. If the pH of the solution is below the pI, the net charge of the protein is positive. In this case, an inverse Hofmeister series is observed with chaotropic anions becoming more effective at salting-out proteins from solution than kosmotropic anions

(Zhang & Cremer, 2009).

Lysozyme has been a model protein for inverse Hofmeister effects (Mathias

Boström et al., 2011; Grigsby et al., 2001; Zhang & Cremer, 2009). The addition of anions on the cloud point of lysozyme followed an inverse Hofmeister series at pH values below the pI, but only when ionic strength was below ~0.3M. With an increase in ionic strength to higher values, the direct Hofmeister series was observed even though the pH was below the pI (Zhang & Cremer, 2009).

The cause of the inverse in the Hofmeister series may be due to changes in the aqueous solution volume upon addition of salt. The size of the hydrated ions have been 22

analyzed and it was determined that the larger anions are more effective at screening electrostatic repulsion between protein molecules and promoting salting-out in solutions below the pI. Therefore, an inverse Hofmeister series is followed according to the hydrated volume of the anions. Bigger anions can more easily shed their hydration shells and engage in ion pairing relative to smaller anions. The molar volumes of the ions are

------ClO4 > SCN > I > NO3 > Br > Cl (Zhang & Cremer, 2009).

2.6 Applications for Protein-Polysaccharide Complexes

Protein-polysaccharide complexes have potential application in food systems when composed of food-grade materials. These complexes have the ability to create value-added food products. Beyond food, protein-polysaccharide complexes are also of interest to the cosmetic and pharmaceutical industries. Due to their unique characteristics, protein-polysaccharide complexes can be used for i) viscosifying and gelling, ii) foaming and emulsifying, and iii) stabilization and release of bioactive components (Schmitt &

Turgeon, 2011).

2.6.1 Viscosifying and Gelling

Viscosity is a rheological property characterized by a fluid’s resistance to flow. It is defined as the ratio between pressure or shear stress exerted on a fluid and the resulting shear rate. Rheological properties are related to varying factors including a biopolymer’s molecular weight, shape, flexibility, concentration, and interaction with water. Highly viscous solutions generally result from high molecular weight, greatly flexible, and 23

hydrophilic biopolymers including many polysaccharides (e.g. xanthan, chitosan) and unfolded proteins (e.g. casein-type milk proteins). On the contrary, acacia gum and albumin proteins are compact and will therefore not result in large changes in viscosity

(Schmitt et al., 1998).

Rheological properties of proteins and polysaccharides in associative conditions will exhibit different behaviors due to each individual biopolymer (Schmitt & Turgeon, 2011).

Highly viscous coacervate solutions can be made using whey protein and a globular polysaccharide gum arabic (F Weinbreck et al., 2003) compared to a more gel-like

24 system when the same protein is combined with the linear polysaccharide pectin (Wang,

Lee, et al., 2007). Because electrostatic interactions are the main stabilizing force between proteins and polysaccharides, protein-polysaccharide ratio, pH, and ionic strength will impact the rheological behavior of complexes.

Mixed protein-polysaccharide systems have shown gelling properties, which has been used by the dairy industry to manipulate the rheological behavior of protein-based food. Carrageenan is the most commonly used hydrocolloid in dairy products at neutral pH. Carrageenan can gel at high concentrations as well as stabilize emulsions or suspensions at lower concentrations. It has been used to suspend cocoa in chocolate milk and thicken ice cream (Syrbe, Bauer, & Klostermeyer, 1998). High-methoxyl pectin is also a prominent stabilizer used by the dairy industry in low pH milk beverages. It functions by stabilizing casein against aggregation and precipitation in fruit-flavored milk and yogurt drinks (Syrbe et al., 1998).

2.6.2 Foaming and Emulsifying 24

Proteins-polysaccharide complexes have physicochemical and functional properties contributed by both of the macromolecules. These complexes and coacervates can therefore be used to stabilize the air/water or oil/water interfaces in an array of food applications (Dickinson, 2008).

Dickinson and Izgi investigated the foaming properties of covalent complexes of the nonionic polysaccharide dextran with three proteins (bovine serum albumin (BSA), lysozyme, and β-casein) by monitoring small pressure changes in a closed chamber

(Dickinson & Izgi, 1996). It was determined that complexation of lysozyme with dextran

25 leads to a substantial enhancement in foaming properties, whereas complexation with

BSA leads to only moderate increases, and complexation with β-casein has a negative effect. These results were enhanced by increasing molecular weight of the polysaccharide.

Additionally, research has been conducted to understand the effects of anionic polysaccharides, sodium alginate and lambda-carrageenan on foaming characteristics of milk whey proteins, whey protein concentrate (WPC) and whey protein isolate (WPI)

(Perez, Carrara, Sa, & Rodrı, 2010). The main conclusions were that the foaming characteristics of WPC-polysaccharide and WPI-polysaccharide mixed systems depended on the protein, the chemical structure of PS, and their relative concentrations in the aqueous phase.

Protein/polysaccharide complexes can be used to stabilize emulsions as well as foams. Emulsion stability has received more attention likely based on the practical applications for controlled emulsion formation. Emulsions can be stabilized by complexes or coacervates produced during the emulsification step to yield mixed emulsions or by a layer-by-layer technique (Jourdain, Leser, Schmitt, Michel, & 25

Dickinson, 2008). In this method, a primary emulsion is stabilized by a protein and then dropped into a polysaccharide dispersion to form a bilayer emulsion (Jourdain et al.,

2008).

2.6.3 Stabilization and Release of Bioactive Components

The interfacial properties of protein/polysaccharide complexes and coacervates make them ideal for the encapsulation or emulsions to protect sensitive ingredients.

Encapsulation protects and stabilizes sensitive molecules (flavors, bioactives) against

26 harsh processing and storage conditions. Delivery of materials can be regulated by controlled release in order to free components to the target area (mouth for flavors, gastrointenstinal tract for bioactives). Release can come by mechanical stress, pH changes, or enzymatic action. When composed of food-grade materials, protein- polysaccharide complexes are suitable in many food applications.

Acacia gum-gelatin coacervates have been established for use of encapsulation.

The unique gelling properties of gelatin create interesting characteristics for microencapsulation. Coacervates can be prepared at 50-60°C, higher than the melting point of gelatin, so once cooled there is a rigid shell around the microcapsule. Once consumed, the shell will be disrupted as gelatin is melted in the mouth. Oil, solids, and flavors have been successfully encapsulated (de Kruif et al., 2004) (Yeo, Bellas,

Firestone, Langer, & Kohane, 2005).

2.7 Conclusions

Electrostatic complex creation using proteins and polysaccharides is an area of 26

increasing interest due to potential use in multiple industries. There is still vast potential in identifying the best materials and conditions required to form stable complexes and particles. Prior to application, greater understanding of systems is necessary including basics of formation of complexes with specific functional attributes as well as stability of complexes and particles once they are in a food or drug matrix.

CHAPTER 3. SPECIFIC ION EFFECTS ON ELECTROSTATIC INTERACTIONS AND COACERVATION OF AQUEOUS Β-LACTOGLOBULIN AND PECTIN MIXTURES

3.1 Abstract

The purpose of the study was to understand the effects of pH, ionic strength, and salt type on electrostatic interactions and coacervation of β-lactoglobulin (BLG) and pectin.

Electrical characteristics, light scattering of biopolymer mixtures, and turbidity of solutions was determined in order to understand changes to complexation. With changes in pH, ζ-potential indicated that BLG and pectin begin to interact near pH 5.50 and these interactions were strengthened as pH was reduced. Sulfate salts were able to cause a positive shift in pHc values at the same ionic strength values, indicating a specific ion effect. Sulfate addition at higher ionic strength produced a relatively smaller shift in pHΦ compared to chloride, while thiocyanate anion produced a relatively greater shift. 27

Reverse Hofmeister effects on dilute solutions of BLG occurred at concentration below

100 mm at pH points below the isoelectric point of the protein, with thiocyanate acting as a kosmotrope and sulfate acting as a chaotrope. The extent of interactions between the protein and polysaccharide are proposed to have a significant impact on the formation of supramolecular structure formation during thermal treatment, which will be covered in

Chapter 4.

3.2 Introduction

Interactions of biopolymers in solution, notably proteins and polysaccharides, can lead to interesting phenomena causing changes on both a microscopic and macroscopic scale. The changes due to interactions of proteins and polysaccharides can be manipulated for utilization in food and pharmaceutical applications. The nature of the proteins and polysaccharides allows them to form complexes with one another because of changes in the electrical characteristics of the biopolymers, mainly as a function of pH.

Prior to application, a deeper understanding of the factors affecting β-lactoglobulin-pectin complex formation and coacervation including pH, ionic strength, and salt species is necessary.

During acidification, interactions between protein and anionic polysaccharide in solution increase at pH values just above the isoelectric point of the protein and complexes are formed. At a critical pH, termed pHc, binding equilibrium between protein and anionic polysaccharide favors the formation of soluble complexes and the increase in colloidal dimensions leads to weak scattering of incident light (detectable by light 28

scattering). Complexes begin to form when polysaccharide and protein have net negative charge, but interactions can occur between the polysaccharide and positive patches on the protein surface. Soluble complexes usually resist aggregation with one another due to the high net charge. The critical value, pHc, reveals the initiation of dominant electrostatic interactions between the protein and polysaccharide and is therefore dependent on ionic strength, the isoelectric point of the protein, and the charge density of the polysaccharide

(Park et al., 1992). Because complexes are held together by weak physical interactions, they are susceptible to dissociation by high ionic strength or slight increases in pH.

Further reduction in pH below pHc increases the net-cationic charge character of the protein and favors further complex formation with anionic polysaccharide that results in more intense scattering of incident light (detectable by UV-Vis spectroscopy). This critical value, termed pHΦ, results in insoluble complexes termed coacervates, which have low net-charge and are more prone to coalescence. Coacervates are also reversible structures that are vulnerable to changes in pH and ionic strength. Stable structures result when these complexes are crosslinked by thermal treatment. For protein-polyanion systems, reduced pHΦ values indicate a decreased tendency for a protein and anionic polysaccharide to overcome the repulsions and interact to form a charge-neutralized, phase separating complex (coacervate) (F Weinbreck et al., 2003).

2------series of SO4 , H2PO4 , F , Cl , Br , ClO4 , SCN (Zhang & Cremer, 2006). The species listed prior to chloride are known as kosmotropes while the species after chloride are 29

known as chaotropes, with chloride falling in between. Kosmotropes stabilize protein structure leading to decreased solubility of proteins and increased protein aggregation and salting out of solution. Chaotropes destabilize protein structure, which results in greater solubility of proteins and salting in of proteins. The Hofmeister series has seen an increase in interest due to correlations beyond protein solubility including enzyme activity, protein-protein interactions, protein crystallization, and bacterial growth (Zhang

& Cremer, 2006).

Boström et al., 2005). With the inverse Hofmeister series, chaotropes become more effective than kosmotropes at stabilizing protein structure leading to salting out.

Lysozyme has become the quintessential example for the inverse Hofmeister series

(Mathias Boström et al., 2011; Grigsby et al., 2001; Zhang & Cremer, 2009).

Beyond the solution pH having an effect on the inverse Hofmeister series, salt concentration has also been proven to play a role. In lysozyme solutions, charge screening effects that lead to the inverse series are only dominant below ~200-300 mM salt. Above this concentration, the physical properties of the solution follow a direct series (Zhang & Cremer, 2009).

complex formation and coacervation process. We hypothesize that both ionic strength and ionic species will affect BLG-pectin complexation and coacervation. In particular, we postulate that increasing ionic strength with interfere with BLG-pectin complex formation and coacervation.

3.3 Materials and Methods

3.3.1 Materials

β-lactoglobulin (lot # JE 001-0-415), with a reported composition of 97.9% protein (91.5% β-lactoglobulin), 0.2% fat, 1.8% ash, and 4.4% moisture, was donated by

Davisco Food International (Le Sueur, MN). Further purification of the protein was performed by an adaptation of Jung et al. (Jung, Savin, Pouzot, Schmitt, & Mezzenga,

2008). Briefly, a 10% β-lactoglobulin solution was prepared in 500 mL of water and stirred for 6 hours to fully disperse the protein. Once the protein was solubilized, the pH was adjusted to 4.60. The solution was centrifuged in a Beckman Coulter JA-14 rotor in

250 mL tubes at 10,000 rpm for 30 minutes at 20°C. The supernatant was removed and adjusted to pH 7.00. Dialysis tubing with a molecular weight cutoff of 6-8 kilodaltons

(kDa) was prepared by boiling tubes in 1 mM Ethylenediaminetetraacetic acid (EDTA) for 10 minutes and thoroughly rinsing with ultrapure water. Tubes were filled with protein solution and dialyzed at 40:1 water to protein ratio. Water was replaced after 4, 8, 31

12, 24, 24, and 48 hours for a total of 120 hours. The dialyzed protein solution was then lyophilized to produce a solid, purified β-lactoglobulin powder.

High methoxy pectin (DE 52 sample #3675) was obtained from CPKelco (Atlanta,

GA) and used without further purification. All salts solutions (sodium chloride, potassium chloride, potassium thiocyanate, ammonium thiocyanate, potassium sulfate, ammonium sulfate, and sodium acetate) were prepared from their solid forms (Sigma

Chemical Co., St. Louis, MO). Similarly, sodium hydroxide solutions were prepared from solid sodium hydroxide (Sigma Aldrich Chemical Co., St. Louis, MO).

Hydrochloric acid solutions were prepared from concentrated 12.1 N hydrochloric acid

(Sigma Aldrich Chemical Co., St. Louis, MO). All solutions were prepared with ultrapure deionized water obtained from an in-lab Barnstead E-pure 3 module water purification system (Thermo Scientific, Waltham, MA). Resistivity reached 18 mΩ-cm before water was drawn.

3.3.2 Biopolymer Solution Preparation

Separate solutions of purified 1.0% β-lactoglobulin and 0.5% pectin solutions in

10 millimolal (mm) sodium acetate buffer were prepared from dry powder. Concentrated

500 mm salt solutions (sodium chloride, potassium chloride, potassium thiocyanate, ammonium thiocyanate, potassium sulfate, ammonium sulfate) were similarly prepared in

10 mm sodium acetate buffer. Solutions were stirred on an IKA stirplate (Staufen,

Germany) at 200 rpm for a minimum of 3 hours at ambient temperature and then stored at

4°C overnight for use on the following day. Prior to mixing, the pH value of solutions were measured with a pH meter (Mettler Toledo, Columbus, OH), calibrated daily before 32

use, and adjusted to pH 7.0 using 1.0 N and 0.1 N sodium hydroxide solution. Unless otherwise stated, solutions were mixed to obtain a β-lactoglobulin concentration of 0.1%

(w/w), a pectin concentration of 0.05% (w/w), and salt concentrations between 0-100 mm using 10 mm sodium acetate buffer. After 5 minutes of stirring with a stirbar at 150 rpm, the pH value of the mixtures was reduced using 1.0, 0.1, and/or 0.01 N HCl solutions drop wise using a Pasteur pipet with 2 mL aliquots taken at desired pH value increments.

Additional solutions with the ionic strength equivalency of sulfate salts were prepared by

1 ! using the following equation to determine ionic strength: � = 2 �!�! where �! is the

th molar concentration of the i ion present in solution and �! is its charge.

3.3.3 Colloidal Charge

Average electrophorectic mobility of proteins and polysaccharides in water was determined by laser Doppler micro-electrophoresis using a Zetasizer Nano ZS light scattering instrument (Malvern, Worcestershire, United Kingdom) at a scattering angle of

173°. An electric field is applied to a solution of molecules, which then move with a velocity related to their zeta-potential. The velocity is measured, which enables electrophoretic mobility to be calculated, and from this zeta-potential. All solutions for charge determination were prepared in 10 mm sodium acetate buffer and loaded into a disposable folded capillary cell (Malvern Instruments, Worcestershire, UK). Average colloidal charge is reported as ζ-potential.

3.3.4 Solution Turbidity 33

The transmittance of visible light (λ = 450 nm) through biopolymer solutions was analyzed using an ultraviolet/visible (UV-Vis) spectrophotometer at 25°C (Lambda 25,

PerkinElmer, Waltham, MA). Suspended materials impede transmittance of light through a solution and cause increases in turbidity. Disposable semi-micro cuvettes composed of polystyrene with a path length of 1.0 cm and an internal volume of 1.5 mL were utilized.

Deionized water was used as a reference blank. Results are presented as 100-%T, where

T = I/I0, I is the intensity of light reaching the detector and I0 is the intensity of the light reaching the detector with the reference blank.

3.3.5 Light Scattering

Scattering intensity was determined using a compact light-scattering goniometer

(ALV/CGS-3 Compact Goniometer System, ALV, Langen, Germany) with an effective scattering angle of 12° to 152° and angular resolution of 0.025°. Light from a helium- neon (HeNe) laser at λ=632.8 nm was passed through a sample in the index-matching vat composed of quartz glass containing filtered toluene. Scattered light was detected by

ALV High Q.E. avalanche photodiode (APD) dual detectors in pseudo-cross correlation mode with 50:50 fiber-optical beam splitter. The pseudo-cross correlation of the two detectors reduces background noise and ghosting for increased accuracy.

Scattered light was detected at a scattering angle of 90° with a manually restricted attenuation factor of 3.10%. Count rates of the scattered light were collected over 30 seconds and given as an average count rate frequency (in kHz) between the two detectors.

3.3.6 Determination of Critical pH Parameters

The pH induced structural transitions among proteins and polysaccharides are 34 defined as the pHc when soluble complexes are formed and pHΦ when coacervation occurs (F Weinbreck et al., 2003). Critical pH parameters for the formation of electrostatic complexes between protein and polysaccharide were determined from analyzed light scattering and transmittance data. When the pH is decreased below a certain value, proteins and polysaccharides begin to weakly associate to form soluble complexes (Owen Griffith Jones & McClements, 2010a). This point, known as the pHc, is determined as the point where scattering intensity departs from a constant value

(Mattison, Brittain, & Dubin, 1995). Because of the sensitivity of light scattering

instruments, pHc can be established accurately. To determine values, lines were fit in excel to the data before and after the first significant change in light scattering. The pHc was found by calculating the value of the abscissa at the intersection of these lines.

The point where coacervation occurs and insoluble complexes begin to form is known as the pHϕ. It is marked by phase separation and a dramatic increase in turbidity of solutions (Mattison et al., 1995). Because of the changes in turbidity, UV-Vis spectroscopy can be used to determine when a marked change in transmittance occurs.

The pHϕ value was determined similarly to pHc—fitting lines in excel to the data before and after the first major change in transmittance. The pHϕ was found by calculating the value of the abscissa at the intersection of these lines.

These points were determined for each replicate and reported as an average and standard deviation.

3.3.7 Statistical Analysis

All experiments were performed in triplicate and reported as the mean and standard 35

deviation unless otherwise noted. A Tukey range test with 95% confidence was performed to find significant differences in pHc and pHΦ values.

3.4 Results and Discussion

The objective of the following experiments was to determine changes caused to dilute solutions of β-lactoglobulin and β-lactoglobulin and pectin with the addition of low concentrations of various salts. Initially, ζ-potential was determined for pure and mixed biopolymer solutions to better understand the electrical properties of the materials.

Change in turbidity of dilute BLG solutions with and without pectin and with increasing concentration of salts as a function of pH was examined to understand changes in the electrostatic complex formation and coacervation process of BLG/pectin systems.

Understanding complex formation and coacervation is a critical step prior to thermal treatment due to the influence these processes have on heated particle formation (See chapter 4).

3.4.1 Surface Charge of Pure and Mixed Biopolymer in Solution

The ability for complexation between proteins and polysaccharides is heavily dependent on surface charge of the biopolymers due to electrostatic interactions required in forming complexes. Surface charge for both proteins and polysaccharides can be modulated as an effect of pH. Average surface charge of dilute β-lactoglobulin and pectin in solution was determined by laser Doppler electrophoresis as a function of pH and reported as ζ-potential.

Results in figure 3.1 show the surface charge for β-lactoglobulin (BLG) became 36

more positive as pH was reduced. The point of net neutral charge on the protein was found at pH 4.7, indicating the protein’s isoelectric point (pI). Below this pH, protein gained a net positive charge due to protonation of the amino acids comprising BLG. The isoelectric point of native BLG is approximately 5.0-5.2 (Tokle, Decker, & McClements,

2012). The difference between the value reported in the literature and the present results can be attributed to the inclusion of a portion of denatured protein in the sample. It is possible that a fraction of denatured protein was insufficiently removed during purification or protein became denatured during centrifugation or freeze-drying.

30.00 0.1% B-lac 20.00 0.05% pectin 0.1% B-lac and 0.05% pectin 10.00

0.00 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 -10.00

-20.00 -potential (mV) -potential ζ -30.00

-40.00

-50.00 pH

Figure 3.1 Dependence of zeta-potential on the pH of pure and mixed biopolymer solutions

The ζ-potential for pectin was negative in the entire pH range from 3.0-7.0. This is because pectin is an anionic polysaccharide with carboxyl functional groups and a pKa of 3.5 (BeMiller & Huber, 2008). Carboxylate polysaccharides become deprotonated

(anionic) at pH points above the pKa. Below pH 4.0 an appreciable increase in ζ-potential 37

was seen due to the pH approaching the pKa, although net neutral charge was never achieved.

Solutions containing both BLG and pectin at pH values below 5.5 demonstrated surface charges of the combined colloidal system that were intermediate between those of the pure biopolymer solutions, but relatively closer to pectin. From pH 7.0 to 5.5 the charge of the mixed biopolymer system and pectin are the same followed by a divergence

38 once pH is decreased below 5.5. A point of net zero charge was never reached above pH

3.0.

These results suggest a binding of anionic pectin to the BLG once the surface charge becomes more positive (below pH 5.5) indicating complexation between the protein and polysaccharide. Electrostatic interactions begin between isolated patches of cationic groups on BLG and pectin’s carboxyl groups above the protein’s pI and are strengthened as the pH is decreased below the pI. These results are supported by previous work on complexation of BLG and pectin (Owen G. Jones et al., 2009; Owen Griffith

Jones & McClements, 2010b).

3.4.2 Effect of pH and Salt on Low Concentration β-Lactoglobulin

The effect of salt species and concentration on the turbidity of pure BLG unheated solutions was studied in order to better understand changes in the aggregation and phase behavior of the protein with salts in the absence of pectin. To observe these behaviors of the pure protein, transmission of visible light through 0.1% BLG solutions in the 38 presence of 6 salt species (KCl, NaCl, KSCN, NH4SCN, K2SO4, (NH4)2SO4) at varying concentrations (0, 25, 50, 75, 100 mm) were measured as pH was decreased from 7.0 to

3.0 (figure 3.2 a-b; appendix A, figures 1a-d).

a) 20 18 100 mm 16 75 mm 14 50 mm 12 10 25 mm

100-%T 100-%T 8 0 mm 6 4 2 0 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 pH

b) 40

35 100 mm

30 75 mm

25 50 mm 25 mm 20 39

100-%T 100-%T 0 mm

0 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 pH

Figure 3.2 Effect of salt concentration (in 10-3 mole·kg-1, mm) on turbidity of 0.1% (w/w) β-lactoglobulin solutions as a function of pH for (a) Potassium Chloride (KCl) and (b) Potassium Thiocyanate (KSCN)

With 0 mm salt, a rise in 100-%T values was seen starting near pH 6.0, peaking around pH 4.70, and decreasing until pH 3.5 where a baseline level was reached (figure

3.2a). The peak indicates a rise in turbidity values due to aggregation of protein. These results are in agreement with literature which found maximum aggregation rate of BLG at pH 4.8 (Majhi et al., 2006).

All noticeable aggregation was suppressed with the addition of 25, 50, 75, and

100 mm KCl. Similarly, the addition of NaCl at concentrations greater than 25 mm suppressed all noticeable aggregation (appendix A, figure 1a). The addition of low concentration (<20 mm) sulfate salts, K2SO4 and (NH4)2SO4, reduced aggregation of protein (appendix A, figures 1c-d). Total suppression of aggregation occurred at sulfate salt concentrations greater than 25 mm.

The addition of KSCN to BLG solutions promoted aggregation and there was significant increase in turbidity (figure 3.2b). Turbidity increased with increasing concentration of KSCN. Similar results were seen for the addition of NH4SCN (appendix

A, figure 1b). A shift in the maximum turbidity value occurred from pH 4.75 with 0 mm 40

KSCN to pH 3.75 in the presence of 100 mm KSCN.

It is hypothesized that the results displayed are following an inverse Hofmeister series. The dilute salt concentrations used (≤100 mm) in our research fall within the values in the literature, which state that reverse Hofmeister effects have been observed in solutions of lysozyme with less than 200 mm NaCl (Zhang & Cremer, 2009). With a direct Hofmeister series, we would expect sulfate salts to increase protein stability, leading to protein aggregation and a subsequent rise in turbidity. On the other hand,

41 thiocyanates would decrease protein stability leading to salting in of proteins and a reduction of aggregation and turbidity.

The research shows that with the addition of chloride and sulfate salts at concentrations below 100 mm, aggregation of protein is suppressed and no rise in turbidity is detected. Significant increases in turbidity seem to be completely suppressed with the addition of chloride and sulfate salts, although the sensitivity of transmitted light detection may be inadequate to measure the changes in aggregation behavior for these low concentrations. More sensitive techniques, such as light scattering, could be utilized in future experiments to better measure these changes. The reduction of protein aggregation with the addition of chloride and sulfate salts in solutions below the pI is an indication of reverse Hofmeister series.

With the addition of thiocyanate salts, significant rises in turbidity occurred below pH 5.25, approximately the pI of native BLG. This indicates that while a portion of the

BLG in the sample may be denatured, there is still a significant portion of native BLG.

Aggregation of protein leads to significant rises in turbidity with increasing thiocyanate 41

concentrations. Following typical direct Hofmeister series behavior, one would expect salting in to occur with the addition of thiocyanate salts but instead the thiocyanate is causing aggregation of protein. Since this is not observed, thiocyanate anions are proposed to function according to inverse Homeister series behavior.

With the addition of salts in solutions below the pI of BLG, chloride and sulfate salts reduced protein aggregation and subsequently lowered turbidity values. The addition of thiocyanate salts caused significant rises in turbidity as a function of increasing ionic strength in solutions with the pH below the pI. From these results, we hypothesize an

42 inverse Hofmeister series is occurring with salts ordered from kosmotropic to chaotropic

- - 2- SCN > Cl > SO4 at concentrations below 100 mm at pH values below the pI of BLG.

3.4.3 Effect of pH and Salt on Electrostatic Complex Formation

Detection of scattered light from colloidal objects in liquid dispersions is a commonly used technique to evaluate colloidal dimensions and phase separation (Girard,

Sanchez, Laneuville, Turgeon, & Gauthier, 2004), which are critical parameters in the identification of electrostatic-complexes between polyelectrolytes in solution (Park et al.,

1992). Scattered light from dilute BLG and pectin in solution was obtained as a function of pH and KCl concentration and is given in Figure 3.3. As the pH was decreased below pH 5.5, scattered light intensity increased significantly and indicated the formation of complexes between BLG and pectin. As shown for KCl, increasing addition of ions to the protein-polysaccharide solutions decreased the pH value at which light scattering increased (figure 3.3). 42

180.00

100 mm 160.00 75 mm 140.00 50 mm 25 mm 120.00 0 mm 100.00

Scattering Intensity Scattering (mHz) 80.00

60.00 4.50 4.70 4.90 5.10 5.30 5.50 5.70 5.90 pH

Figure 3.3 Effect of salt concentration (mm) and species on light scattering of 0.1% β- lactoglobulin and 0.05% pectin solution with 0-100 mm Potassium Chloride (KCl) as a function of pH

Changes in scattering intensity with reduction in pH as a function of varying salt species at 50 mm is shown in figure 3.4a. The pH at which scattering intensity increases was similar for chloride and thiocyanate salts but markedly different for sulfate salts at 50 43

mm. The results were similar based on the anion of the salt species.

180.00

KCl 160.00 NaCl

140.00 KSCN NH4SCN 120.00 K2SO4 (NH4)2SO4 100.00 Scattering Intensity Scattering (mHz) 80.00

60.00 4.50 4.70 4.90 5.10 5.30 5.50 5.70 5.90 pH

180.00 KCl 160.00 NaCl KSCN 140.00 NH4SCN 120.00 16.67 mm K2SO4 16.67 mm (NH4)2SO4 44

100.00

Scattering Intensity(mHz) Scattering 80.00

60.00 4.80 4.90 5.00 5.10 5.20 5.30 5.40 5.50 5.60 5.70 5.80 5.90 6.00 pH

Figure 3.4 Effect of salt concentration (mm) and species on light scattering of 0.1% β- lactoglobulin and 0.05% pectin solutions as a function of pH (a) 50 mm of various salt species (b) 50 mm of chloride and thiocyanate salts with 16.67 mm sulfate salts

Sulfate anions are divalent and cause a nonlinear increase in the effective screening of electrostatic interactions that is related to the ionic strength parameter. In order to account for anion differences due to electrostatic screening alone, scattered light intensity from BLG-pectin solution as a function of pH was also determined for sulfate salt concentrations of 8.33, 16.67, 25 and 33.33 mm (appendix A, figures 2a-b), which have ionic strength values corresponding with chloride or thiocyanate concentrations of

25, 50, 75 and 100 mm.

To demonstrate the effect of ionic strength on the formation of electrostatic complexes, scattered light intensity as a function of pH value BLG-pectin solutions with

16.67 mm sulfate is compared against those containing 50 mm thiocyanate and 50 mm chloride in figure 3.4b. The ionic strength equivalency value of 16.67 mm for the sulfate salts had similar light scattering values compared to thiocyanate and chloride salts at 50 mm.

Table 3.1 shows pHc values determined in excel by finding the abscissa before a noticeable increase in scattering intensity. For chloride and thiocyanate salts, pHc values 45

decreased as a function of increasing salt concentration. With the addition of 8.33 mm

K2SO4, there was a slight increase in the pHc value compared to the control. All other sulfate solutions saw decreases in pHc with increasing salt concentrations. This data is presented graphically in figure 3.5. At 50 mm, no statistical significance was seen between the chloride and thiocyanate salts, although both the 16.67 mm and 50 mm sulfate salts were significantly different. For all concentrations, there was no significant difference for pHc values of the chloride and thiocyanate salts (with the exception of

NaCl at 100 mm).

Table 3.1 pHc Values

pHc

KCl NaCl KSCN NH4SCN K2SO4 (NH4)2SO4 4.95 4.87 5.01 4.97 4.68 4.59 100 mm ±0.03A ±0.02B ±0.02A ±0.01A ±0.02C ±0.02D 5.11 5.14 33.33 mm ±0.02E ±0.03E 5.06 5.06 5.06 5.05 4.79 4.76 75 mm ±0.04AB ±0.02AB ±0.03AB ±0.02A ±0.02C ±0.03C 5.22 5.15 25 mm ±0.04D ±0.04BD 5.15 5.13 5.19 5.19 4.86 4.94 50 mm ±0.03A ±0.03A ±0.02A ±0.03A ±0.03B ±0.03B 5.35 5.35 16.67 mm ±0.04C ±0.04C 5.28 5.21 5.26 5.29 5.22 5.15 25 mm ±0.05AC ±0.04AB ±0.04AB ±0.04AC ± 0.04AB ±0.04B 5.43 5.39 8.33 mm ±0.04D ±0.03CD 5.41 5.41 5.41 5.41 5.41 5.41 0 mm ±0.01 ±0.01 ±0.01 ±0.01 ±0.01 ±0.01 5.8 5.7 KCl 5.6 NaCl 5.5 KSCN 5.4 NH4SCN 5.3 K2SO4 46

5.2 (NH4)2SO4 5.1 pHc 5 4.9 4.8 4.7 4.6 4.5 4.4 0 10 20 30 40 50 60 70 80 90 100 Salt Concentration (mm)

Figure 3.5 Changes to pHc values as a function of salt concentration (mm) and species

The effect of ionic strength on formation of protein-polysaccharide complexes has been studied previously, with work primarily focusing on the addition of NaCl to various protein and polysaccharide types. The ions present in solution screen charges on the polymers thereby reducing the range of their associative interactions (Schmitt et al.,

1998). This causes the polymers to interact at a lower pH where the protein carries more positive charges, causing changes to both the pHc and pHΦ. The screening of charges also results in a reduced number of soluble complexes formed, and the scattering intensity is therefore lower (F Weinbreck et al., 2003).

Because the addition of salt tends to weaken electrostatic interactions, these results suggest that there is a reduction in favorable attractive forces between the protein and polysaccharide at higher pH values due to screening effects. As pH is decreased in the presence of salt, net-charge of the protein increases for complexation to occur between protein and polysaccharide, although it will happen at a lower pH as indicated by a decrease in pHc values. At higher concentrations of ions, salt has the ability to inhibit complexation by total screening of charges (F Weinbreck et al., 2003). Previous 47 studies have shown a change to lower pHc values for complexes of bovine serum albumin and heparin with concentrations of NaCl above 10 mm (Seyrek, Dubin, Tribet, & Gamble,

2003).

Due to the lack of significant differences between pHc values of chloride and thiocyanate salts, it is hypothesized that changes to onset of complexation between BLG and pectin in the presence of chloride and thiocyanate salts are due to ionic strength instead of ionic species. The addition of chloride and thiocyanate salts was able to cause a shift to lower pHc values as a function of increasing ionic strength. This indicates that

48 electrostatic forces are screened with increased ionic strength making it less probable for protein and polysaccharide to interact.

The addition of sulfate salts caused changes to the system based both on ionic strength and ion type. If changes occurred based simply on ionic strength, there should not be significant differences in pHc values of solutions with ionic strength equivalencies to those of chloride and thiocyanate. Instead, we see statistically significant differences in the pHc values of the ionic strength equivalencies. At all salt concentrations, the ionic strength equivalencies of sulfate salts have higher pHc values than those of chloride and thiocyanate salts indicating an increase in favorable interactions between BLG and pectin.

At low concentrations, sulfate ions may be reducing water interactions with the surface of the protein and favoring complex formation with pectin, which decreases the exposed surface of the protein molecules. With the addition of low concentrations of sulfate salts to the BLG-pectin system, there is a significant difference in pH values compared to chloride and thiocyanate salts at which onset of complex formation occurs leading to the conclusion that changes are due to both ionic strength and specific ion effects. 48

3.4.4 Specific Ion Effects on Coacervation Process

Figure 3.6 shows changes in turbidity as a function of pH at 0, 25, 50, 75, and 100 mm potassium chloride concentrations. Turbidity values remained constant for all concentrations until the pH was reduced to a critical value where the turbidity abruptly increases. As salt concentration increased, there was a decrease in the critical pH at which a significant rise in turbidity occurred. Maximum turbidity values were also decreased with an increase in KCl concentration. Turbidity of solutions containing 50 mm KCl or

49 greater decreased at pH values below 3.5. Solutions containing 0 and 25 mm KCl induced rapid phase separation of the complexes below pH 3.50 to form a precipitate layer.

Aggregation was significant enough to prevent accurate turbidity determination.

120.00 100 mm 100.00 75 mm 50 mm 80.00 25 mm 60.00 0 mm 100-%T 100-%T 40.00

20.00

0.00 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 pH

Figure 3.6 Effect of salt concentration (mm) and species on turbidity of 0.1% β- lactoglobulin and 0.05% pectin solutions with 0-100 mm Potassium Chloride (KCl) as a function of pH

Turbidity as a function of salt species at 50 mm demonstrates differences due to the anion of the salt species (figure 3.7a). Chloride salts caused the greatest maximum turbidity values followed by thiocyanate and sulfate. With the inclusion of 50 mm sulfate, turbidity values were distinctly lower than chloride and thiocyanate. Turbidity values for chloride and thiocyanate anions remained constant until pH was decreased enough to cause phase separation which resulted in a steady increase in turbidity from pH 4.75 to pH 3.25. When the ionic strength equivalency of 16.67 mm sulfate salts was added to

50 solution, turbidity values were higher than values with 50 mm sulfate and between those of chloride and thiocyanate (figure 3.7b).

120.00 KCl 100.00 NaCl 80.00 KSCN NH4SCN 60.00

100-%T 100-%T K2SO4 40.00 (NH4)2SO4

20.00

0.00 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 pH

120.00 KCl 100.00 NaCl 80.00 KSCN

NH4SCN 50

60.00 16.67 mm K2SO4 100-%T 100-%T 40.00 16.67 mm (NH4)2SO4

20.00

0.00 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 pH

Figure 3.7 Effect of salt concentration (mm) and species on turbidity of 0.1% β- lactoglobulin and 0.05% pectin solutions as a function of pH (a) 50 mm of various salt species (b) 50 mm of chloride and thiocyanate salts with 16.67 mm sulfate salts

Table 3.2 shows pHΦ values determined by finding the abscissa prior to a large rise in turbidity. pHΦ is a critical pH value where phase separation occurs, but does not lead to spontaneous demixing. This leads to a stable colloidal dispersion, so immediate precipitation was not seen. No significant differences in pHΦ values were detected in solutions containing 25 mm chloride and thiocyanate salts. At concentrations above 25 mm, significant differences were seen between the chloride and thiocyanate salts.

Table 3.2 pHΦ values pHΦ

KCl NaCl KSCN NH4SCN K2SO4 (NH4)2SO4 4.56 4.59 4.43 4.43 100 mm - - ±0.01A ±0.01A ±0.02B ±0.01B 4.72 4.72 33.33 mm ± 0.01C ±0.01C 4.71 4.68 4.58 4.63 75 mm - - ±0.01A ±0.01A ±0.02B ±0.02C 4.80 4.79 25 mm ±0.01D ±0.01D 4.82 4.81 4.77 4.78 4.55 4.57 50 mm A AB B B C C

±0.02 ±0.02 ±0.01 ±0.01 ±0.01 ±0.01 51

4.96 4.82 16.67 mm ±0.02D ±0.01A 4.96 4.97 4.91 4.91 4.80 4.79 25 mm ±0.02AB ±0.02AB ±0.02A ±0.02A ±0.01D ±0.01D 5.09 5.02 8.33 mm ±0.02E ±0.02C 4.98 4.98 4.98 4.98 4.98 4.98 0 mm ±0.01 ±0.01 ±0.01 ±0.01 ±0.01 ±0.01

5.2 5.1 KCl NaCl 5 KSCN

4.9 NH4SCN Φ K2SO4

pH 4.8 (NH4)2SO4 4.7 4.6 4.5 4.4 4.3 0 20 40 60 80 100 Salt Concentration (mm)

Figure 3.8 Changes to pHΦ values as a function of salt concentration (mm) and species

Solutions with sulfate salts at the same ionic strength as chloride and thiocyanate had significantly lower pHΦ values. The ionic strength equivalency of sulfate salts had significantly higher pH Φ values compared to chloride and thiocyanate salts with the exception of 16.67 mm (NH4)2SO4. Values for 75 and 100 mm sulfate salts were unable 52 to be determined due to a lack of a definable increase in turbidity (appendix A, figures

3a-b). The addition of 8.33 mm sulfate salts increased pHΦ compared to 0 mm salt. This information is represented graphically in figure 3.8. Sulfate salts caused the greatest decrease in pHΦ, followed by thiocyanate, chloride, and finally the ionic strength equivalency of sulfate.

Changes in turbidity arise from either significant changes in the size or concentration of aggregates in solution (Xia, Dubin, & Dautzenbergt, 1993). Insoluble

53 complexes scatter light more intensely leading to an abrupt change in turbidity that can be detected with UV-Vis spectroscopy. pHΦ is the state boundary between soluble complexes and is dependent upon the ratio of protein to polysaccharide as well as ionic strength (Cooper, Dubin, Kayitmazer, & Turksen, 2005). The phase separation is related to charge-neutralization of the protein-polysaccharide complex. When ionic strength is increased, the pHΦ decreases (Mattison et al., 1995).

Experimentally determined pHΦ critical values were in relatively close proximity to the pI of the protein, which was unexpected. Previous research of coacervation of whey protein and gum arabic (an anionic polysaccharide) at a protein to polysaccharide ratio of 2:1 had an experimentally determined pHΦ value near 4.60, which decreased to under 4.0 with the addition of 50 mm NaCl (Fanny Weinbreck, Wientjes, Nieuwenhuijse,

Robijn, & de Kruif, 2004). It was expected our experimentally determined pHΦ values would be higher, similar to these published values. When pHΦ is reached, there will be a simultaneous increase in light scattering as well as hydrodynamic radius of complexes

(Sperber, Schols, Cohen Stuart, Norde, & Voragen, 2009). Our determination of pHΦ 53

values was performed using UV-Vis spectroscopy and may have introduced error into the determination. Experimentally determined pHΦ values corresponded with the initial increase in turbidity determined at λ=450 nm using UV-Vis spectroscopy; in order to verify that pHΦ values corresponded with phase separation, we would need confirmation of the increase in hydrodynamic radius using more sensitive light scattering techniques or centrifugation of our solutions.

With increasing salt concentration, there were decreased pHΦ values indicating reduction in attractive interactions by electrostatic screening. The changes occurred due

54 to the anion of the salt, which has been confirmed to have a greater effect than the cation

(Zhang & Cremer, 2006).

When only molal concentration was considered, the greatest reductions in pHΦ occurred in the presence of sulfate salts followed by thiocyanate, chloride. At equivalent ionic strength values, pHΦ values were relatively increased for the sulfate anion compared to samples with thiocyanate and chloride anions, indicating a specific ion effect. If only ionic strength equivalency is considered, the trends indicate a reverse

Hofmeister series. At pH below the isoelectric point of the protein, all ions screened charge interactions between the BLG and pectin, thus reducing the pHΦ. In a traditional

Hofmeister series, one would expect thiocyanate to increase solubility and unfolding of the protein, allowing easier interaction with the anionic polysaccharide.

BLG-pectin solutions with sulfate salts at 8.33 mm increases (I = 25 mm) the pHΦ above that of solutions with 0 mm salt. This means with the addition of very low levels of sulfate salts, there is less screening of electrostatic interactions and the protein and polysaccharide can more easily interact allowing phase separation to occur at higher pH 54 values. We saw a specific ion effect on the ability of sulfate salts to increase the pHΦ above the control value.

At ionic strength values of 25 mm, no significant differences were seen between chloride and thiocyanate pHΦ. As ionic strength increased above 25 mm, significant differences in pHΦ values were seen between chloride, thiocyanate, and the ionic strength equivalency of sulfate. These differences in pHΦ values based on salt type, specifically the anion, instead of ionic strength indicate specific ion effects on the BLG-pectin coacervation process.

3.5 Conclusions

These studies have shown that both ionic strength and specific ion effects can influence BLG-pectin complex formation and coacervation. Interactions of proteins and polysaccharides begin near pH 5.5 and are strengthened with decrease in pH as indicated by ζ-potential. Reverse Hofmeister effects occur at pH points below the pI in dilute BLG solutions, with thiocyanate acting as a kosmotrope and sulfate acting as a chaotropes under 100 mm. Shifts in pHc values occurred as a function of increasing ionic strength, with no significant difference between chloride and thiocyanate salts. The ionic strength equivalency of sulfate salts caused significant differences to pHc values, leading us to believe that specific ion effects influence complex formation. Finally, due to the significant differences seen in experimentally determined pHΦ values above 25 mm between chloride, thiocyanate, and sulfate salts we expect specific ion effects can influence the coacervation process of BLG-pectin. Further research into phase behavior of these coacervates would confirm the influence of specific ions on BLG-pectin complexation and coacervation. The results determined in this unheated system will give 55

us greater insight into the changes that occur with the addition of heat.

CHAPTER 4. THERMAL TRANSITIONS AMONG Β-LACTOGLOBULIN-PECTIN ELECTROSTATIC COMPLEXES

4.1 Abstract

The purpose of this study was to characterize changes due to thermal transitions of β-lactoglobulin-pectin electrostatic complexes as a function of pH, ionic species, and ionic strength. Initially, the effect of ionic strength of 0-100 mm of potassium chloride

(KCl) on solutions of 0.1% β-lactoglobulin and 0.05% pectin was determined.

Concentrations of 25-50 mm increased turbidity values compared to the control, while

75-100 mm KCl decreased turbidity. The addition of KCl caused formation of larger particles up to 75 mm KCl as determined by light scattering, which was confirmed visually by atomic force microscopy (AFM). Turbidity measurements with increasing thermal treatment revealed a decrease in aggregation temperatures (Tagg) with the 56

addition of 25-75 mm KCl, while addition of 100 mm caused an increase in Tagg compared to the control.

The effect of ion type was analyzed by addition of chloride, thiocyanate, and sulfate salts to solutions of BLG-pectin. With ionic strength of 50 mm (I=50 mm) in solutions at pH 4.50, sulfate cause the greatest decrease in solution turbidity followed by chloride and thiocyanate. Thermally treated BLG-pectin particles at pH 4.50 had no significant differences in size between all salt species at ionic strength of 50 mm, although size increased as a function of ionic strength. No increase in Tagg occurred for

57 solutions with the addition of thiocyanate salts, and only minimal differences occurred with chloride and sulfate salts at ionic strengths of 100 mm. AFM saw little differences in morphology of particles with 50 mm salt. The study proved changes in size occurred as a function of pH and ionic strength, although fewer changes occurred as a function of salt type.

4.2 Introduction

Formation of protein-polysaccharide electrostatic complexes is possible by adjusting pH of solutions of β-lactoglobulin and pectin, as demonstrated in chapter 3.

These complexes are susceptible to changes in pH, ionic strength, and ionic species.

Previous research has shown that thermal treatment of BLG-pectin electrostatic complexes can form stable particles by physical cross-linking (Owen G. Jones et al.,

2009).

When BLG and pectin are mixed together and pH is reduced, there is initial formation of soluble complexes, followed by coacervation, and finally precipitation. 57

Heating of soluble complexes and coacervates (above the protein’s denaturation temperature) will promote globular protein unfolding and produce more stable particles that have a tendency to more intensely scatter light than their unheated counterparts. Due to the initial electrostatic complexation prior to thermal treatment, these particles will be susceptible to changes in pH, ionic strength, and ionic species.

The goal of the following research was to characterize changes due to thermal treatment of solutions of BLG and BLG-pectin in the presence of 0-100 mm of various salts. This was done by analyzing solution turbidity during and after thermal treatment,

58 determining particle size, and understanding differences in morphology using atomic force microscopy (AFM). It is hypothesized that thermal treatment of electrostatic complexes will be affected by both ionic strength and ionic species and will produce particles with difference physicochemical properties than unheated complexes.

4.3 Materials and Methods

4.3.1 Materials

β-lactoglobulin (lot # JE 001-0-415), with a reported composition of 97.9% protein (91.5% β-lactoglobulin), 0.2% fat, 1.8% ash, and 4.4% moisture, was donated by

Davisco Food International (Le Sueur, MN). Further purification of the protein was performed by centrifugation and dialysis using an adaptation of Jung et al. as described in chapter 3 section 3.3.1.

High methoxy pectin (DE 52 sample #3675) was obtained from CPKelco (Atlanta,

GA) and used without further purification. All salts solutions (sodium chloride, 58

potassium chloride, potassium thiocyanate, ammonium thiocyanate, potassium sulfate, ammonium sulfate, and sodium acetate) were prepared from their solid forms (Sigma

Chemical Co., St. Louis, MO). Similarly, sodium hydroxide solutions were prepared from solid sodium hydroxide (Sigma Aldrich Chemical Co., St. Louis, MO).

Hydrochloric acid solutions were prepared from concentrated 12.1 N hydrochloric acid

(Sigma Aldrich Chemical Co., St. Louis, MO). All solutions were prepared with ultrapure deionized water obtained from an in-lab Barnstead E-pure 3 module water purification

59 system (Thermo Scientific, Waltham, MA). Resistivity reached 18 mΩ-cm before water was drawn.

4.3.2 Heated Complex Creation

Separate solutions of purified 1.0% β-lactoglobulin and 0.5% pectin solutions in

10 millimolal (mm) sodium acetate buffer were prepared from dry powder. Concentrated

0.5 m salt solutions (sodium chloride, potassium chloride, potassium thiocyanate, ammonium thiocyanate, potassium sulfate, ammonium sulfate) were similarly prepared in

10 mm sodium acetate buffer. Solutions were stirred on an IKA stirplate (Staufen,

Germany) at 200 rpm for a minimum of 3 hours at ambient temperature and then stored at

4°C overnight, for use the following day. Prior to mixing, the pH value of solutions were measured with a pH meter (Mettler Toledo, Columbus, OH), calibrated daily before use, and adjusted to pH 7.0 using 1.0 N and 0.1 N sodium hydroxide solution. Unless otherwise stated, solutions were mixed to obtain a β-lactoglobulin concentration of 0.1%

(w/w), a pectin concentration of 0.05% (w/w), and salt concentrations between 0-100 mm 59

using 10 mm sodium acetate buffer. After 5 minutes of stirring with a stirbar at 150 rpm, the pH value of the mixtures was reduced using 1.0, 0.1, and/or 0.01 N HCl solutions drop wise using a Pasteur pipet with 2 mL aliquots taken at desired pH value increments and placed in a glass culture tube. Tubes were heated in a circulating waterbath (Grant,

Cambridgeshire, U.K.) for 15 minutes at 80°C. After heating, tubes were brought to ambient conditions by submersion in a bath at 4°C for 10 minutes.

4.3.3 Solution Turbidity

The transmittance of visible light (λ = 450 nm) through biopolymer solutions was analyzed using an ultraviolet/visible (UV-Vis) spectrophotometer at 25°C (Lambda 25,

Deionized water was used as a reference blank. Results are presented as 100-%T, where

T = I/I0, I is the intensity of light reaching the detector and I0 is the intensity of the light reaching the detector with the reference blank.

4.3.4 Particle Size

Particle size in solution was determined by dynamic light scattering using a

Zetasizer Nano ZS light scattering instrument (Malvern, Worcestershire, United

Kingdom) at a scattering angle of 173°. The diffusion of particles moving under

Brownian motion is measured and converted to particle size and size distribution using 60

the Stokes-Einstein relationship. All solutions for size determination were prepared in 10 mm sodium acetate buffer and loaded into a polystyrene semi-micro disposable cuvette

(Malvern Instruments, Worcestershire, UK). Average particle size is reported as Z-ave

(nm).

4.3.5 Turbidity Measurements with Increasing Thermal Treatment

The turbidity of biopolymer solutions as a function of increasing temperature was determined using a using an ultraviolet/visible (UV-Vis) spectrophotometer (Lambda 25,

PerkinElmer, Waltham, MA) with attached Peltier (PTP-6+6, PerkinElmer, Waltham,

MA) and controlled fluid circulator (PCB 1500, PerkinElmer, Waltham, MA) at 450 nm.

One milliliter of sample was place in a quartz semi-micro cuvette with a path length of 1 cm followed by 4 drops from a Pasteur pipette of mineral oil (Sigma Chemical Co., St.

Louis, MO) to prevent excessive evaporation during heating. Deionized water was used as a reference blank. Temperature scanning was performed from 25°C to 92°C with an increase of 1°C/minute and absorbance readings were taken one time per minute.

Readings were not collected during the cooling period. Aggregation temperatures (Tagg) were determined by finding the inflection point in the turbidity versus temperature curves.

The inflection point was determined from the first derivative of the τ–T curves using

Excel.

4.3.6 Particle Morphology using Atomic Force Microscopy (AFM)

Select samples were chosen for morphological characterization using atomic force microscopy (AFM). 50 µL of heat-treated sample (0.1% BLG, 0.05% pectin, varying salt) 61

at pH 4.50 was deposited on freshly cleaved mica (Ted Pella, Redding, CA) fixed to a glass slide. After adsorption to the mica surface for 3 minutes, excess sample was removed using a Kimwipe, followed by rinsing the mica with 2 mL ultrapure water at pH

4.50, and excess water was again removed with a Kimwipe. Low-pressure air was used to dry the mica and slides were stored overnight in a desiccator for use the following day.

When slides were not in use, they were stored in a covered plastic petri dish in a desiccator.

Topography of deposited samples was measured using a MFP-3D AFM (Asylum

Research, Santa Barbara, CA) mounted with an aluminum coated silicon probe

(Tap150Al-G, 5 N/m force constant, 150 kHz resonant frequency, Ted Pella, Redding,

CA). Environmental vibrations were minimized by use of a vibration table (Herzan TS-

150, Laguna Hills, CA). 3µm x 3µm scans were obtained in intermittent contact mode.

Transferable images were created using Asylum Research software, built upon the IGOR

Pro Platform (Asylum Research, Santa Barbara, CA).

4.3.7 Statistical Analysis

All experiments were performed in triplicate and reported as the mean and standard deviation unless otherwise noted. A Tukey range test with 95% confidence was performed to find significant differences in particle size.

4.4 Results and Discussion

The objective of these experiments was to characterize changes that occur with 62

thermal treatment of electrostatic complexes as a function of pH, ionic strength, and salt type. Initial results show the effect of ionic strength on BLG-pectin particle formation with the inclusion of potassium chloride from 0 to 100 mm. Latter results demonstrate the impact of salt type on BLG-pectin particle formation with chloride, thiocyanate, and sulfate salts from 0 to100 mm. Solution turbidity, particle size, aggregation temperatures, and morphology using AFM were determined in order to understand the effects of ionic strength and ion type.

4.4.1 Effect of Ionic Strength on β-Lactoglobulin-Pectin Particle Formation

In order to determine the effect of ionic strength on the formation of BLG-pectin particles, BLG-pectin complexes were thermally treated (80°C for 15 minutes) at varying pH with potassium chloride, which was chosen for its relatively intermediate kosmotropic/chaotropic character in aqueous solutions. Initially, solutions of 0.1% BLG with KCl were thermally treated to understand the effect of heat on dilute concentrations of BLG (figure 4.1a). When solutions of BLG were heated as a function of ionic strength and pH, aggregation was seen starting at pH 6.0 and continuing through pH 4.0. Protein aggregation led to a very large increase in turbidity values, with maximum aggregation seen from pH 4.75 to 5.25. The addition of KCl slightly decreased maximum turbidity values as a function of increasing ionic strength.

a) 120.000 100 mm 75 mm 50 mm 25 mm 0 mm 100.000

80.000

60.000 100-%T 100-%T 40.000

20.000

0.000 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 pH

b) 120.000 100 mm 75 mm 50 mm 25 mm 0 mm 100.000

80.000

60.000 100-%T 100-%T 64

40.000

20.000

0.000 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 pH

Figure 4.1 Thermal transitions of a) 0.1% (w/w) β-lactoglobulin and b) 0.1% (w/w) β- lactoglobulin and 0.05% (w/w) pectin solution with Potassium Chloride (KCl) (in 10-3 mole·kg-1, mm) as a function of pH

As pH of BLG-pectin solutions with KCl was decreased, turbidity values quickly increased to a maximum value near pH 5.25 before decreasing to a trough with minimum values between pH 4.75 and 4.50 depending on ionic strength (figure 4.1b). For each concentration of KCl there was a pH value between pH 5.0 and 4.0 at which the turbidity of the heated protein solution was relatively reduced and will be referred to as the turbidity minimum. As ionic strength was increased from 0 mm to 100 mm, turbidity minimum values decreased, with 100 mm KCl having the lowest turbidity minimum. The increase in ionic strength caused a shift in the pH at which the minimum turbidity value occurred. For solutions with 0 to 50 mm KCl, turbidity minimums occurred at pH 4.75, while solutions with 75 mm or greater KCl saw minimums at pH 4.50. Following the minimum, turbidity values began to increase again reaching a second peak at pH 3.50 for solutions with greater than 50 mm KCl. Solutions with 0 mm and 25 mm KCl had aggregation of particles below pH 3.50 so accurate turbidity measurements were unable to be determined.

Extensive aggregation in the protein only occurred near the isoelectric point of 65

BLG (pH~5.2), with maximum aggregation seen at pH 4.75-5.25 (figure 4.1a). The isoelectric point of the BLG used was determined by ζ-potential measurements to be closer to pH 4.70 (chapter 3, figure 3.1). In the absence of protein-protein charge repulsion near the isoelectric point and above the protein’s denaturation temperature, the proteins unfold and associate via hydrophobic and disulfide bonds, leading to irreversible formation of aggregates (Hoffmann & van Mil, 1999). There was nearly no increase in turbidity at pH values above 6.50 and below 3.75. At pH values away from the pI (either

66 much higher or lower), protein-protein repulsion inhibits aggregation (Hoffmann & van

Mil, 1999).

Upon the addition of pectin, the turbidity profiles of solutions versus pH (figure

4.1b) were considerably different for solutions of BLG alone, demonstrating that the addition of an anionic polysaccharide can alter the nature of biopolymer aggregates formed during thermal treatment. The initial peak in aggregation near pH 5.25 is likely due to a combination protein aggregation and thermal treatment of soluble complexes able to cause increases in turbidity. At pH 5.25, soluble complexes have begun to form, which would be able to cause changes in turbidity when heated. Also, because complexation is relatively weak at this pH, protein could dissociate from complexes and be denatured leading to protein aggregation.

Due to the reduction in turbidity at pH 4.50-4.75, we hypothesize that the addition of pectin is able to reduce the size and/or concentration of biopolymer aggregates formed, therefore causing decreases in turbidity values. It is expected that anionic polysaccharides formed electrostatic complexes with BLG at ambient temperatures, which inhibited 66

thermal aggregation and denaturation of protein by limiting the ability of protein to interact with itself. Further investigation into particle size was performed on particles at pH 4.50 and pH 4.75 because it was expected that the reduction in turbidity could be correlated to a reduction in particle size. It was predicted that particles created at pH

4.50-4.75 would be suitable and stable for potential applications, which warranted additional characterization of particles by light scattering and AFM.

The differences in turbidity below pH 4.0 between the protein only system and protein-polysaccharide system are due to the addition of pectin. It was seen that below

67 pH 4.0 in the unheated BLG-pectin system, there were still significant changes in turbidity due to formation of insoluble complexes (chapter 3, figure 3.6). It is hypothesized that when these insoluble complexes were heated, increases in turbidity were due to formation of particles. The precipitation of solutions with 0 and 25 mm KCl is due to initial instability of the unheated electrostatic complexes at pH values below 3.5

(chapter 3, figure 3.6). Once thermally treated, these electrostatic complexes precipitated immediately, making it unable to determine accurate turbidity readings.

The addition of KCl led to charge screening and a reduction in turbidity values as

a function of increasing ionic strength. These results are similar to the unheated systems

of BLG and BLG-pectin with KCl. The addition of ions prevents protein-protein

interactions as well as protein-polysaccharide interactions. Interactions between the

BLG and pectin still occur with the addition of KCl at 100 mm, but turbidity values are

reduced in both the protein only and protein-polysaccharide systems. This leads us to

believe that while some charges are screened preventing maximum interaction, there is

not complete inhibition of electrostatic interactions. 67

Morphologies of heated BLG-pectin solutions with 0-100 mm KCl at pH 4.50 were obtained by intermittent contact mode AFM, and amplitude images are compared in figure 4.2 (see also appendix B, figure 1 for height images). Samples were scanned in multiple areas of the mica to obtain a representative image. Amplitude images of particles with 0 and 25 mm KCl show round, uniformly-shaped particles in varying sizes. At ionic strengths of 50 mm and above, morphological differences from the control appeared, particle shape became more irregular and particles aggregated into masses.

68 68

Figure 4.2 AFM amplitude images of thermally treated protein-polysaccharide complexes containing 0.1% β-Lactoglobulin (w/w) and 0.05% pectin (w/w) at pH 4.50 with potassium chloride a) 0 mm b) 25 mm c) 50 mm d) 75 mm e) 100 mm

Table 4.1 Z-Ave (nm) of thermally treated protein-polysaccharide complexes containing 0.1% β-Lactoglobulin (w/w) and 0.05% pectin (w/w) at pH 4.50 with potassium chloride Concentration (mm) Z-Ave (nm) 0 mm 228.0 ± 2.3A 25 mm 265.3 ± 1.3B 50 mm 294.0 ± 6.2CD 75 mm 298.8 ± 3.3C 100 mm 275.9 ± 7.6BD

Clear size and shape differences arise as ionic strength is increased. Round particles with an average diameter of 288 nm (as determined by light scattering) give rise to larger, more irregular shaped particles, which also aggregate with one another.

AFM confirms the size changes as well as gives us a better understanding of changes to structure. As increases in ionic strength caused increases in particle size, there is likely a combination of individual particles becoming larger and particles aggregating with one another (flocculating), creating seemingly larger particles. Particle size increases and become more irregular due to the weakened electrostatic interactions between protein and pectin at higher ionic strength. Particle size increased through 75 mm KCl but then 69

decreased at 100 mm (table 4.1). The AFM image for 100 mm KCl shows aggregates similar in size to those at 75 mm KCl although less flocculation of particles is seen. It is possible that at 100 mm, there was less aggregation of particles, leading to falsely low

Z-ave determination. Greater aggregation of particles would lead to higher Z-ave determinations, and it is likely that there is a distribution in particle size ranging from smaller, individual particles, to larger aggregates of particles. As ionic strength is increased, particles have a tendency to aggregate more readily due to screening of repulsive forces. Further investigation with a more sensitive light scattering technique is

needed to determine if there is truly a particle size difference or if particles are floccing

together.

Table 4.1 shows Z-average diameter values of particles created in solutions of 0.1%

BLG and 0.05% pectin with 0-100 mm KCl at pH 4.50. Particles with no salt were significantly smaller than those with over 25 mm KCl. Size of particles increased with increasing KCl concentration until 100 mm. Particles at 100 mm were similar in size to those at 25 mm and 50 mm but significantly smaller than those at 75 mm KCl.

Apparent size can be determined from the height image AFM as shown in

appendix B, figures 2 and 3. Using Igor software, a line was drawn over the image of the

particle to determine particle size. The base of the particle with 0 mm KCl seems to be

about 175 nm diameter, while the solutions with 50 mm have larger diameters of about

250 nm. These height images provide rough estimations of particle size and it is

expected that the size determined by AFM would be smaller than size determination in

solution due to reduced particle diameter because of drying. These size determinations

also indicated that there is a range of particle size. 70

The influence of thermal treatment on BLG-pectin electrostatic complexes was studied by analyzing turbidity as a function of increasing temperature (Figure 4.3).

Samples containing 0.1% BLG and 0.05% pectin with varying concentrations of KCl at pH 4.50 were heated from 25°C to 92°C at 1°C per minute, in order to ensure solutions were heated beyond the denaturation temperature of BLG (70°C at pH 6.8 (Qi et al.,

1997)). Solutions containing 25 and 50 mm KCl had the highest initial starting turbidities, followed by 0 mm, 75 mm, and 100 mm. These initial values are in agreement with the turbidity values of unheated BLG-pectin mixtures at pH 4.50 from chapter 3, figure 3.6.

As temperature increased in solutions with 0 mm KCl, turbidity values increased gradually and these solutions had the lowest final turbidity value. Solutions with 25 mm

KCl or higher saw initial decreases in turbidity prior to increases. Final turbidity values increased with increasing KCl concentration. Also, the differences in turbidity values between before and after thermal aggregation were dependent on increasing ionic strength, i.e. solutions with 0 mm KCl saw small differences in turbidity near the inflection point, while solutions with 100 mm KCl saw dramatic differences in turbidity near the inflection point.

An aggregation temperature coinciding with large increases in turbidity near the expected denaturation temperature was calculated by finding the inflection point in the turbidity versus temperature curves (table 4.2). The Tagg for solutions without KCl was determined to be 80.5°C. Tagg values for solutions with the addition of salt were all below that of the control with the exception of 100 mm KCl, which increased the Tagg value to

82.0°C. 71

120 100 mm 75 mm 50 mm 25 mm 0 mm 100

60 100-%T 100-%T 40

0 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Temperature (°C)

Figure 4.3 Changes to the thermal profile of 0.1% β-Lactoglobulin (w/w) and 0.05% pectin (w/w) with varying concentrations of Potassium Chloride at pH 4.50 with temperature increase of 1°C per minute

Table 4.2 Temperature of aggregation of protein-polysaccharide complexes containing 0.1% β-Lactoglobulin (w/w) and 0.05% pectin (w/w) at pH 4.50 with varying concentrations of potassium chloride Concentration (mm) Tagg (°C) 0 80.5 ± 0.7 72

25 79.0 ± 0.0 50 79.0 ± 1.4 75 80.0 ± 0.7 100 82.0 ± 0.0

The initial turbidity values for solutions with 0-50 mm KCl can be attributed to the formation of electrostatic complexes at ambient conditions. Solutions with 0 and 25 mm saw gradual increases in turbidity at temperatures above 40°C, which indicates protein started aggregating prior to the denaturation temperature. Once above the denaturation temperature the rate of aggregation increased significantly, similarly to

73 solutions with greater than 50 mm. Solutions with 75 and 100 mm KCl possessed low initial turbidity values compared to those with 25 and 50 mm due to electrostatic screening and prevention of interactions between the protein and polysaccharide at ambient temperatures. The initial decrease in turbidity of solutions containing 50-100 mm is hypothesized to be caused by promotion of diffusivity of biopolymers promoting interactions of proteins and polysaccharides prior to denaturation and aggregation. Below

75°C there was diminished rates of aggregation with increasing ionic strength. Once above 75°C, there was a significant increase in turbidity due to protein aggregation above the thermal denaturation temperature. These results indicate that increasing ionic strength is able to screen charges at low temperatures but unable to inhibit protein aggregation above the thermal denaturation temperature. Solutions with 0 and 25 mm KCl had lower final turbidity values compared to 50-100 mm, which had similar final turbidity values.

These results are supported by AFM, where there was an apparent difference in particle size above 50 mm KCl that could cause an increase in scattering and final turbidity values. 73

The decrease in Tagg in solutions with 25 to 75 mm KCl indicates a destabilizing effect to the protein-polysaccharide complexes. The complexes become more susceptible to changes due to thermal treatment and therefore the Tagg is reduced. At ion concentrations up to 75 mm, ions shield protein-polysaccharide and any remaining protein-protein interactions and promote easier protein aggregation. At 100 mm KCl, aggregation temperature increased to 82.0°C, which is above the value of the control.

Higher ion concentration seems to have a chaotropic-like behavior because the formation of insoluble components is shifted to higher temperatures. Onset temperatures of

aggregation were similar for all solutions from 0-100 mm but the Tagg was different.

Follow-up studies with differential scanning calorimetry could be performed to determine denaturation temperature and cooperativity of unfolding in order to provide more information on thermal transitions.

4.4.2 Effect of Ionic Species on β-Lactoglobulin-Pectin Particle Formation

The effect of the specific cation and anion on formation of BLG-pectin particles was studied by thermally treating solutions with various chloride, thiocyanate, and sulfate salts at different pH values and analyzing changes in their solution turbidity. When no salt was added to the system, there was an increase in turbidity from pH 5.75 to 5.25, followed by a minimum turbidity value of 4.75, and then an increase in turbidity until pH

3.50 where aggregation and sedimentation occurred (figure 4.1b). These results were discussed previously in regards to figure 4.1b.

When 25 mm of chloride and thiocyanate salts were added to the BLG-pectin system, there was an increase in the turbidity minimum value (figure 4.1b and appendix b, 74

figures 4a-c). With sulfate salts at 25 mm, turbidity minimum values were lower than those of the control although at the ionic strength equivalency of 8.33 mm there was an increase in turbidity minimum values, similar to chloride and thiocyanate salts (appendix b, figures d-e).

With the addition of 50 mm of each salt type, a significant increase was initially seen with a peak maximum at pH 5.25, followed by a decrease leading to a minimum value at pH 4.75 for chloride salts, 4.50 for thiocyanate salts and K2SO4, and 4.25 for

(NH4)2SO4 (figure 4.4a). After reaching minimum turbidity values, turbidities for all

75 solutions increased, with chloride and thiocyanate salts coming to a plateau and sulfate salts reaching another peak before descending. Turbidity values for solutions containing

50 mm sulfate salts were noticeably decreased compared to chloride and thiocyanate solutions. Taking into consideration the ionic strength equivalency of the 16.67 mm sulfate salts (I=50 mm), solution turbidity values at each pH value displayed fewer deviations among the different salt species (figure 4.4b). Peak maximums were seen at pH 5.25 followed by minimums at pH 4.75. After the minimum, solution turbidy values increased prior to a gradual decrease below pH 3.50.

120 110 KCl NaCl KSCN NH4SCN K2SO4 NH42SO4 100 90 80 70 60

100-%T 50 40 30 20 10 0 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 pH

120 110 100 90 80 70 60 76

100-%T 50 40 30 KCl NaCl 20 KSCN NH4SCN 10 16.67 mm K2SO4 16.67 mm NH42SO4 0 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 pH

Figure 4.4 Effect of specific ion species on the turbidity of thermally treated 0.1% β- lactoglobulin and 0.05% pectin solutions as a function of pH considered at (a) 50 mm salt concentration (b) 50 mm ionic strength ([sulfate salts] = 16.67 mm)

For solutions with 75 mm or more of chloride salts, there was a noticeable decrease in minimum turbidity values, making the changes in turbidity due to pH more defined (figure 4.1b and appendix b, figure 4a). Also, turbidity minimums occurred at pH

4.50 instead of at pH 4.75 like that of the control. The addition of thiocyanate salts caused an increase in turbidity minimums and there were less noticeable differences in turbidity as a function of pH (appendix B, figures 4b-c). With sulfate salts at greater than

75 mm, turbidities were significantly decreased at all pH values below 5.50 compared to chloride and thiocyanate salts (appendix B, figures 4d-e).

The dependence of turbidity on solution pH is caused by the thermal denaturation and aggregation of free protein as well as changes to light scattering of heated electrostatic complexes, which relates to particle size and concentration The initial peak in turbidity at pH 5.25 is due to unassociated protein in the system that has been thermally denatured leading to aggregation. All the salt solutions at 50 mm have a pHc value lower than 5.25 (chapter 3, table 3.1), indicating that complexation through electrostatic interactions has not yet occurred, leading us to conclude that the peak at pH 77

5.25 is due to free protein in solution. Following the peak, turbidity values of the heated solution decreased and approached a minimum between pH 4.25 and pH 4.75. At the minimums of 4.75 for chloride salts, 4.50 for thiocyanate salts and K2SO4, and 4.25 for

(NH4)2SO4, the pHc value has been reached in the unheated system. This indicates the soluble complexes have been created in the unheated system and once thermally treated these particles can scatter light more intensely than their unheated counterparts.

Reductions in turbidity were attributed to the electrostatic complex formation between protein and polysaccharide, which prevented full aggregation phenomena and led to the

78 formation of particles instead of large aggregates of protein as was observed without the presence of pectin (figure 4.1a). This indicates a protective effect of the polysaccharide on the protein in solution in its ability to reduce thermal aggregation. Following the minimum turbidity value, an increase in turbidity occurred below pH 4.25 for all solutions. In solutions without pectin (figure 4.1a) there was a sharp drop in turbidity values below pH 4.0 leading us to believe that the changes in turbidity compared to solutions of only BLG, are due to the presence of pectin. Below pH 4.25, insoluble complexes would have been formed in the unheated system and once thermally treated, these particles could cause large increases in turbidity.

Similar to the unheated system, the inclusion of sulfate salts at 50 mm significantly reduced turbidity values in solutions with 0.1% BLG and 0.05% BLG compared to 50 mm chloride and thiocyanate salts. If the ionic strength equivalency of the 16.67 mm sulfate salts was considered (I=50 mm), turbidity values were similar to those of 50 mm chloride and thiocyanate salts. Minimums were seen at pH 4.75, below the pHΦ values for the sulfate salts at 16.67 mm. Once again, the reduction in turbidity is 78

due to complexation between the protein and polysaccharide. The minimum of the sulfate salts at 16.67 mm saw the lowest turbidity values, indicating that most likely fewer and/or smaller aggregates are formed in the presence of sulfate salts compare to the other salts.

The aggregation temperatures (Tagg) of protein solutions, with or without pectin, were determined by observing their solution turbidity as a function of increasing temperature in order to determine the effects of pectin and specific ions on the protein interactivity during thermal treatment. For solutions of 0.1% BLG without pectin, the aggregation temperature was 78°C (figure 4.5a). With increasing ionic strength,

79 aggregation temperatures were increased. Sulfate salts at the ionic strength equivalency of 100 mm (33.33 mm sulfate salts) increased Tagg to 83.0°C with K2SO4 and 83.5°C with

(NH4)2SO4. The addition of chloride salts increased Tagg to 82.5°C with 100 mm NaCl and 82.0°C with 100 mm KCl. Thiocyanate only minimally increased Tagg to 79.5°C with

100 mm NH4SCN and 78.5°C with 100 KSCN. Upon addition of 0.05% pectin to solutions of 0.1% BLG, the aggregation temperature without added salts was increased to

80.5°C (figure 4.5b). Addition of up to 75 mm chloride and thiocyanate salts to BLG- pectin solutions decreased the observed Tagg value relative to the 0 mm solution, although the Tagg value increased consistently as the salt concentration increased from 25 to 75 mm.

With the addition of 100 mm of chloride salts and the ionic strength equivalency of 100 mm sulfate salts, Tagg values were increased relative to 0 mm salt while the Tagg values of solutions containing 100 mm of thiocyanate salts were lower than the control. If the ionic strength of sulfate salts is taken into consideration, the change in Tagg value of the BLG- pectin solutions with increasing ions is comparable among the different salt species, indicating that the shift in aggregation temperature is largely a charge-screening effect 79

and more heavily dependent upon ionic strength.

90 KCl NaCl KSCN NH4SCN K2SO4 NH42SO4 88

agg 82 T 80

74 0 10 20 30 40 50 60 70 80 90 100 Ionic Strength (I)

88 KCl NaCl KSCN NH4SCN K2SO4 NH42SO4

agg 82 T

80 80

74 0 20 40 60 80 100 Ionic Strength (I)

Figure 4.5 Temperature of aggregation as a function of salt species and ionic strength at pH 4.50 for a) 0.1% β-lactoglobulin solutions and b) 0.1% β-lactoglobulin and 0.05% pectin

Aggregation temperatures were determined for solutions of 0.1% BLG and 0.1%

BLG with 0.05% pectin at pH 4.50. Initial Tagg for solutions of 0.1% BLG with no salt was determined to be 78.0°C. The addition of each salt species caused increases in Tagg, with the greatest changes due to sulfate salts, then chloride and thiocyanate salts.

Previous research suggests that NaCl concentration greater than 20 mM can increase the aggregation temperature of whey proteins, with higher ionic strengths increasing protein stability by increased the hydration and solubility of the proteins (Xiong, 1992). The trend here shows sulfate salts having the greatest stabilizing effect on the protein only solution at pH 4.50. Because this pH falls below that of the pI of BLG, it is expected that there is a specific ion effect following the reverse Hofmeister series. Sulfate salts seems to be protecting the BLG from thermal aggregation by the increase in Tagg. Thiocyanate salts only slightly increased the Tagg even in solutions up to 100 mm.

The addition of 0.05% pectin to 0.1% BLG at pH 4.50 with 0 mm salt increased

Tagg from 78.0°C to 80.5°C (figure 4.5). This increase in Tagg is consistent with the literature (Owen G. Jones et al., 2009). With the addition of pectin, electrostatic 81

complexes form below thermal denaturation temperature, preventing protein self- association. Once the temperature is increased above the thermal denaturation temperature, the protein is able to unfold and aggregate to form nano- and micro-particles that scatter light causing a sharp increase in turbidity. It is expected that the protein and pectin stay attached and aggregate together to form a particle, although it is possible that the protein and pectin detach and protein aggregation leads to an increase in turbidity.

The addition of chloride and thiocyanate salts through 75 mm caused little changes to the Tagg of solutions of 0.1% BLG and 0.05% pectin at pH 4.50. This same

82 information is presented graphically in figures 4.6 and 4.7. A slight increase was initially seen, followed by an increase once 100 mm chloride salts were added. Thiocyanate caused decreases in Tagg, causing destabilization of protein and decreased Tagg. Chloride was able to make an impact at 100 mm where it seems to be stabilizing the protein, leading to greater solubility and increased Tagg. Finally, the initial Tagg in the presence of sulfate salts decreased with the addition at low ionic strength (<16.67 mm) and then significantly increased Tagg with increasing ionic strength. Again, the influence of sulfate, chloride, and thiocyanate of protein solubility follows the reverse Hofmeister series for solutions at pH 4.50 although greater effects occurred due to ionic strength. While the influence of pectin changed the profile of the heating curves, the effect of specific ions caused more prevalent changes. The addition of pectin was unable to suppress the effect the ions had on the nature of protein in solution.

Table 4.3 and 4.4 give observed particle size (Z-average diameter, in nm) of thermally treated particles of 0.1% BLG and 0.05% pectin as determined by light scattering. Determination of particle size was critical in order to know whether or not 82

controlling parameters, including pH and addition of ions, would affect size. Control of particle size is crucial for texture and optical properties, which will help to determine potential applications. In all solutions compared by salt type and concentration, particles had a smaller average particle diameter at pH 4.50 than in solution at pH 4.75 with the exception of the control containing 0 mm salt and 75 mm (NH4)2SO4 (table 4.3 and 4.4).

Due to the smaller particle size at pH 4.50, it was decided that AFM would be performed on solutions at this pH. At pH 4.50, there were no significant differences between all solutions containing 50 mm salts or the ionic strength equivalency of the sulfate salts at

16.67 mm (table 4.3). At pH 4.75, chloride and thiocyanate salts at 50 mm were statistically similar, as were sulfate salts at 50 mm, and sulfate salts at 16.67 mm (table

4.4).

Table 4.3 Z-Ave (nm) of heated particles at pH 4.50 Z-Ave (nm) at pH 4.50

KCl NaCl KSCN NH SCN K SO (NH ) SO 4 2 4 4 2 4 275.9 ± 296.8 ± 437.1 ± 413.0 ± 450.7 ± 431.0 ± 100 mm 7.6A 13.5A 17.7B 3.3B 37.5B 58.3B 277.8 ± 278.8 ± 33.33 mm 2.6A 12.2A 298.8 ± 300.9 ± 349.2 ± 326.6 ± 402.9 ± 418.0 ± 75 mm 3.3AC 30.0AC 44.3ABD 5.4ACD 8.3BD 12.4B 280.2 ± 280.0 ± 25 mm 0.4C 2.1AC 294.0 ± 311.2 ± 309.8 ± 300.5 ± 284.1 ± 306.8 ± 50 mm 6.2A 19.4A 22.4A 20.5A 2.3A 6.7A 275.9 ± 280.0 ± 16.67 mm 2.4A 4.5A 265.3 ± 282.4 ± 264.3 ± 264.2 ± 280.2 ± 280.0 ± 25 mm 1.3ACDE 9.4B 1.1ACD 2.8ACD 0.4BE 2.1ABE 250.5 ± 273.5 ± 8.33 mm 0.8C 4.8ABDE

228.0 ± 228.0 ± 228.0 ± 228.0 ± 228.0 ± 83 0 mm 228.0 ± 2.3 2.3 2.3 2.3 2.3 2.3

550.0 KCl NaCl KSCN 500.0 NH4SCN K2SO4 (NH4)2SO4 450.0

400.0

350.0 Siza (nm) Siza 300.0

250.0

200.0 0 10 20 30 40 50 60 70 80 90 100 Ionic Strength (I)

Figure 4.6 Size (Z-average diameter) of particle formed from thermally treated 0.1% β- lactoglobulin and 0.05% pectin as a function of salt species and ionic strength at pH 4.50

Table 4.4 Z-Average diameter (nm) of heated particles at pH 4.75 Z-Ave (nm) at pH 4.75

KCl NaCl KSCN NH SCN K SO (NH ) SO 4 2 4 4 2 4 407.8 ± 404.1 ± 945. 4 ± 760.1 ± 100 mm Precip. Precip. 0.6A 3.7A 84.3B 15.9C 292.9 ± 295.2 ± 33.33 mm A A 4.9 20.8 84

341.2 ± 366.9 ± 407.0 ± 415.1 ± 411.3 ± 405.7 ± 75 mm 3.4ABD 13.4ABC 10.4BC 44.4C 8.6BC 17.0BC 292.2 ± 299.0 ± 25 mm 0.4D 5.2D 310.3 ± 326.6 ± 366.1 ± 340.4 ± 387.0 ± 406.0 ± 50 mm 0.7AC 10.0ACD 25.8ABD 24.5ABCD 6.8BD 26.2B 284.2 ± 301.6 ± 16.67 mm 8.4C 10.3AC 278.5 ± 291.9 ± 299.2 ± 284.6 ± 292.2 ± 299.0 ± 25 mm 5.9AB 10.0A 11.2A 17.4AB 0.4A 5.2A 263.2 ± 291.9 ± 8.33 mm 1.6B 6.6A 217.0 ± 217.0 ± 217.0 ± 217.0 ± 217.0 ± 0 mm 217.0 ± 0.6 0.6 0.6 0.6 0.6 0.6

550 KCl NaCl KSCN 500 NH4SCN K2SO4 (NH4)2SO4

450

400

350 Siza (nm) Siza

300

250

200 0 10 20 30 40 50 60 70 80 90 100 Ionic Strength (I)

Figure 4.7 Size (Z-average diameter) of particles formed from thermally treated 0.1% β- lactoglobulin and 0.05% pectin as a function of salt species and ionic strength at pH 4.75

The addition of 100 mm of thiocyanate salts caused significant changes based on the specific ion. At pH 4.50, a divergence in particle size was seen towards larger sizes at ionic strength of 100 mm. Additionally, at pH 4.75, the inclusion of 100 mm thiocyanate

salts caused precipitation of heated particles, which prevented accurate size determination. 85

Thiocyanate salts had a destabilizing effect on heated particles at both pH 4.50 and 4.75 and had statistically larger size at ionic strength values of 100 mm compared to chloride and sulfate salts.

Morphological characterization of heated particles containing 0.1% BLG and 0.05% pectin with 50 mm of various salt species at pH 4.50 was determined using AFM (figure

4.8). Smaller particle size at pH 4.50 led us to investigate morphology of these particles.

Understanding size and shape of particles can give us better insight into both formation

86 and stability. In all solutions containing 50 mm salts, average particle size has increased compared to the control and there is possible flocculation of particles with chloride salts,

NH4SCN, and K2SO4. With 16.67 mm K2SO4 and (NH4)2SO4 (I=50 mm), particle shape was round and little flocculation of particles was seen (appendix B, figure 6).

87 87

Figure 4.8 AFM amplitude images of thermally treated protein-polysaccharide complexes containing 0.1% β-Lactoglobulin (w/w) and 0.05% pectin (w/w) at pH 4.50 with 50 mm a) KCl b) NaCl c) KSCN d) NH4SCN e) K2SO4 f) (NH4)2SO4

At pH 4.50, no significant differences in size of 0.1% BLG and 0.05% pectin particles were seen. As ionic strength increased there was an increase in particle size due to screening of charges, making it more difficult for the BLG and pectin to interact. This

88 led to the formation of larger, more irregular shaped particles. While there were no significant differences in size as determined by light scattering, there may be differences in particle shape determined by AFM. Additional images should be collected prior to interpretation of morphological differences. Also, more sensitive light scattering techniques could help to determine information about particle flocculation.

4.5 Conclusion

Results from this study show changes to thermally treated electrostatic complexes can result based both on ionic strength and ionic species. Differences in the turbidity profiles of thermally treated BLG-pectin complexes were seen with the addition of KCl, with concentrations greater than 75 mm causing decreases in turbidity. Morphological characterization with AFM revealed at KCl concentrations of 50 mm and greater, there was disruption in particle structure from rounded to more amorphous shapes and possible flocculation of particles. With increasing KCl concentration up to 75 mm, there was a decrease in Tagg of solutions containing both protein and pectin but Tagg increased with the 88

addition of 100 mm KCl when compared to the control.

With the inclusion of ionic strength=50 mm, sulfate salts caused the greatest decrease in turbidity of thermally treated solutions of BLG-pectin, followed by chloride and thiocyanate. Tagg slightly increased with the addition of ionic strength of 100 mm chloride and sulfate salts but decreased compared to the control with 100 mm thiocyanate.

Particle size was smaller for thermally treated particles at pH 4.50 compared to pH 4.75.

Thiocyanate salts led to the greatest increase in particle size, which increased with

89 increasing ionic strength. Additional AFM images need to be collected prior to coming to conclusions regarding particle morphology at pH 4.50.

In conclusion, changes to thermally treated particles of BLG-pectin are cause by both ionic strength and ionic species. Increase in KCl ionic strength led to increase in particle size and deviations from round to amorphous particle shapes. Thiocyanate caused destabilizing effects to particles, leading to increased size and decreased Tagg.

CHAPTER 5. CONCLUSIONS

These studies have shown that BLG-pectin complex formation and coacervation as well as thermal transitions of electrostatic complexes are influenced by both ionic strength and specific ion effects.

A summary of conclusions follows:

• Interactions of proteins and polysaccharides begin near pH 5.5 and are

strengthened with decrease in pH as indicated by ζ-potential.

• Reverse Hofmeister effects occur at pH points below the pI in dilute BLG

solutions, with thiocyanate acting as a kosmotrope and sulfate acting as a

chaotrope under 100 mm.

• Shifts in pHc values occurred as a function of increasing ionic strength, with no

significant difference between chloride and thiocyanate salts. The ionic strength 90

equivalency of sulfate salts caused significant differences to pHc values, leading

us to believe that specific ion effects influence complex formation.

• Due to the significant differences seen in experimentally determined pHΦ values

above 25 mm between chloride, thiocyanate, and sulfate salts we expect specific

ion effects can influence the coacervation process of BLG-pectin.

• Differences in the turbidity profiles of thermally treated BLG-pectin complexes

were seen with the addition of 0-100 mm KCl, with concentrations greater than 75

mm causing decreases in turbidity.

• Morphological characterization with AFM revealed at KCl concentrations of 50

mm and greater, there was disruption in particle structure from rounded to more

amorphous shapes and possible flocculation of particles.

• Tagg slightly increased with the addition of ionic strength of 100 mm chloride and

sulfate salts but decreased compared to the control with 100 mm thiocyanate.

• Particle size was smaller for thermally treated particles at pH 4.50 compared to

pH 4.75. Thiocyanate salts led to the greatest increase in particle size, which

increased with increasing ionic strength.

Further research into phase behavior of coacervates, including light scattering and centrifugation experiments, is necessary order to confirm the influence of specific ions on

BLG-pectin complexation and coacervation. Additional AFM images need to be 91 collected prior to coming to conclusions regarding particle morphology at pH 4.50. The information gained from this research is important in understanding complexation of proteins and anionic polysaccharides and thermal transition of these complexes. pH, ionic strength and ionic species influence complex formation, which may be able to be controlled for use in potential applications in the food and pharmaceutical industries.

LIST OF REFERENCES

Arnaudov, L. N., Vries, R. De, Ippel, H., & Mierlo, C. P. M. Van. (2003). Multiple Steps during the Formation of -Lactoglobulin Fibrils, 1614–1622.

BeMiller, J. N., & Huber, K. C. (2008). Carbohydrates. In Fennema’s Food Chemistry (pp. 83–154).

Boström, M, Tavares, F. W., Finet, S., Skouri-Panet, F., Tardieu, a, & Ninham, B. W. (2005). Why forces between proteins follow different Hofmeister series for pH above and below pI. Biophysical chemistry, 117(3), 217–24.

Boström, Mathias, Parsons, D. F., Salis, A., Ninham, B. W., & Monduzzi, M. (2011). Possible origin of the inverse and direct Hofmeister series for lysozyme at low and high salt concentrations. Langmuir : the ACS journal of surfaces and colloids, 27(15), 9504–11.

Braudo, E. E., & Antonov, Y. A. (1993). Non-coulombic Complex Formation of Proteins as a Structure-Forming Factor in Food Systems. In Food Proteins, Structure, and Functionality (pp. 210–215).

Collins, K. D., Neilson, G. W., & Enderby, J. E. (2007). Ions in water: characterizing the

forces that control chemical processes and biological structure. Biophysical 93

chemistry, 128(2-3), 95–104.

Cooper, C. L., Dubin, P. L., Kayitmazer, a. B., & Turksen, S. (2005). Polyelectrolyte– protein complexes. Current Opinion in Colloid & Interface Science, 10(1-2), 52–78.

Damodaran, S. (2008). Amino Acids, Peptides, and Proteins. In Fennema’s Food Chemistry (pp. 217–329).

De Kruif, C. G., Weinbreck, F., & de Vries, R. (2004). Complex coacervation of proteins and anionic polysaccharides. Current Opinion in Colloid & Interface Science, 9(5), 340–349.

Dickinson, E. (2008). Interfacial structure and stability of food emulsions as affected by protein–polysaccharide interactions. Soft Matter, 4, 932–942.

Dickinson, E., & Izgi, E. (1996). Foam stabilization by protein-polysaccharide complexes. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 113(1-2), 191–201.

Doublier, J., Garnier, C., Renarda, D., & Sanchezb, C. (2000). Protein-polysaccharide interactions. Current Opinion in Colloid & Interface Science, 5, 202–214.

Girard, M., Sanchez, C., Laneuville, S. I., Turgeon, S. L., & Gauthier, S. F. (2004). Associative phase separation of beta-lactoglobulin/pectin solutions: a kinetic study by small angle static light scattering. Colloids and surfaces. B, Biointerfaces, 35(1), 15–22.

Giroux, H. J., Houde, J., & Britten, M. (2010). Preparation of nanoparticles from denatured whey protein by pH-cycling treatment. Food Hydrocolloids, 24(4), 341– 346.

Graveland-Bikker, J. F., Fritz, G., Glatter, O., & de Kruif, C. G. (2006). Growth and structure of α-lactalbumin nanotubes. Journal of Applied Crystallography, 39(2), 180–184.

Grigsby, J. J., Blanch, H. W., & Prausnitz, J. M. (2001). Cloud-point temperatures for lysozyme in electrolyte solutions: effect of salt type, salt concentration and pH. Biophysical chemistry, 91(3), 231–43.

Guerin, D., Vuillemard, J. C., & Subirade, M. (2003). Protection of bifidobacteria encapsulated in polysaccharide-protein gel beads against gastric juice and bile. Journal of Food Protection, 66, 2076–2084.

Gunasekaran, S., Ko, S., & Xiao, L. (2007). Use of whey proteins for encapsulation and

controlled delivery applications. Journal of Food Engineering, 83(1), 31–40. 94

Hoffmann, M. a, & van Mil, P. J. (1999). Heat-induced aggregation of beta-lactoglobulin as a function of pH. Journal of agricultural and food chemistry, 47(5), 1898–905.

Jones, Owen G., Decker, E. a., & McClements, D. J. (2009). Formation of biopolymer particles by thermal treatment of β-lactoglobulin–pectin complexes. Food Hydrocolloids, 23(5), 1312–1321.

Jones, Owen G., Lesmes, U., Dubin, P., & McClements, D. J. (2010). Effect of polysaccharide charge on formation and properties of biopolymer nanoparticles created by heat treatment of β-lactoglobulin–pectin complexes. Food Hydrocolloids, 24(4), 374–383.

Jones, Owen Griffith. (2009). Fabrication of Protein-Polysaccharide Particulates through Thermal Treatment of Associative Complexes. University of Massachusetts Amherst.

Jones, Owen Griffith, & McClements, D. J. (2010a). Functional Biopolymer Particles: Design, Fabrication, and Applications. Comprehensive Reviews in Food Science and Food Safety, 9(4), 374–397.

Jones, Owen Griffith, & McClements, D. J. (2010b). Biopolymer nanoparticles from heat-treated electrostatic protein-polysaccharide complexes: factors affecting particle characteristics. Journal of food science, 75(2), N36–43.

Jourdain, L., Leser, M. E., Schmitt, C., Michel, M., & Dickinson, E. (2008). Stability of emulsions containing sodium caseinate and dextran sulfate: Relationship to complexation in solution. Food Hydrocolloids, 22(4), 647–659.

Jung, J.-M., Savin, G., Pouzot, M., Schmitt, C., & Mezzenga, R. (2008). Structure of heat-induced beta-lactoglobulin aggregates and their complexes with sodium- dodecyl sulfate. Biomacromolecules, 9(9), 2477–86.

Kunz, W. (2009). Specific Ion Effects (p. 347).

Majhi, P. R., Ganta, R. R., Vanam, R. P., Seyrek, E., Giger, K., & Dubin, P. L. (2006). Electrostatically Driven Protein Aggregation at Low Ionic Strength B- Lactoglobulin, (6), 9150–9159.

Mattison, K. W., Brittain, I. J., & Dubin, P. L. (1995). Protein-Polyelectrolyte Phase Boundaries. Biotechnology Progress, 11(6), 632–637.

Nicolai, T., Britten, M., & Schmitt, C. (2011). β-Lactoglobulin and WPI aggregates: 95

Formation, structure and applications. Food Hydrocolloids, 25(8), 1945–1962.

Otte, J., Ipsen, R., Bauer, R., Bjerrum, M. J., & Waninge, R. (2005). Formation of amyloid-like fibrils upon limited proteolysis of bovine α-lactalbumin. International Dairy Journal, 15(3), 219–229.

Park, J. M., Muhoberac, B. B., Dubin, P. L., & Xia, J. (1992). Effects of Protein Charge Heterogeneity in Protein-Polyelectrolyte Complexation. Macromolecules, 25, 290– 295.

Perez, A., Carrara, C. R., Sa, C. C., & Rodrı, J. M. (2010). Interfacial and Foaming Characteristics of Milk Whey Protein and Polysaccharide Mixed Systems, 56(4), 1107–1117.

Picot, A., & Lacroix, C. (2004). Encapsulation of bifidobacteria in whey protein-based microcapsules and survival in simulated gastrointestinal conditions and in yoghurt. International Dairy Journal, 14, 505–515.

Ptitsyn, O. B. (1995). Molten Globule and Protein Folding. Advances in Protein Chemistry, 47, 83–229.

Qi, X. L., Holt, C., Mcnulty, D., Clarke, D. T., Brownlow, S., & Jones, G. R. (1997). Effect of temperature on the secondary structure of β-lactoglobulin at pH 6.7, as determined by CD and IR spectroscopy: a test of the molten globule hypothesis. Biochemistry, 346, 341–346.

Schmitt, C., Sanchez, C., Desobry-Banon, S., & Hardy, J. (1998). Structure and Technofunctional Properties of Protein- Polysaccharide Complexes : A Review. Critical Reviews in Food Science and Nutrition, 38(8), 689–753.

Schmitt, C., & Turgeon, S. L. (2011). Protein/polysaccharide complexes and coacervates in food systems. Advances in colloid and interface science, 167(1-2), 63–70.

Seyrek, E., Dubin, P. L., Tribet, C., & Gamble, E. a. (2003). Ionic strength dependence of protein-polyelectrolyte interactions. Biomacromolecules, 4(2), 273–82.

Smiechowski, M., & Stangret, J. (2006). Proton hydration in aqueous solution: Fourier transform infrared studies of HDO spectra. Journal of Chemical Physics, 125.

Sperber, B. L. H. M., Schols, H. a., Cohen Stuart, M. a., Norde, W., & Voragen, A. G. J. (2009). Influence of the overall charge and local charge density of pectin on the complex formation between pectin and β-lactoglobulin. Food Hydrocolloids, 23(3), 765–772. 96

Syrbe, a., Bauer, W. J., & Klostermeyer, H. (1998). Polymer Science Concepts in Dairy Systems—an Overview of Milk Protein and Food Hydrocolloid Interaction. International Dairy Journal, 8(3), 179–193.

Tiebackx, F. W. (1911). Gleichzeitige Ausflockung zweier Kolloide. Z. Chem. Ind. Kolloide, 8, 198–201.

Tokle, T., Decker, E. A., & McClements, D. J. (2012). Utilization of interfacial engineering to produce novel emulsion properties: Pre-mixed lactoferrin/β- lactoglobulin protein emulsifiers. Food Research International, 49(1), 46–52.

Tolstoguzov, V. B. (1997). Protein-Polysaccharide Interactions. In Food Proteins and Their Applications (pp. 171–198).

Verheul, M., Roefs, S. P. F. M., & Kruif, K. G. De. (1998). Kinetics of Heat-Induced Aggregation of B-Lactoglobulin. Journal of Agricultural and Food Chemistry, 46(3), 896–903.

Vetri, V., & Militello, V. (2005). Thermal induced conformational changes involved in the aggregation pathways of beta-lactoglobulin. Biophysical chemistry, 113(1), 83– 91.

Wang, X., Lee, J., Wang, Y.-W., & Huang, Q. (2007). Composition and rheological properties of beta-Lactoglobulin/pectin coacervates: effects of salt concentration and initial protein/polysaccharide ratio. Biomacromolecules, 8(3), 992–7.

Wang, X., Wang, Y.-W., Ruengruglikit, C., & Huang, Q. (2007). Effects of salt concentration on formation and dissociation of beta-lactoglobulin/pectin complexes. Journal of agricultural and food chemistry, 55(25), 10432–6.

Weinbreck, F, de Vries, R., Schrooyen, P., & de Kruif, C. G. (2003). Complex coacervation of whey proteins and gum arabic. Biomacromolecules.

Weinbreck, Fanny, Wientjes, R. H. W., Nieuwenhuijse, H., Robijn, G. W., & de Kruif, C. G. (2004). Rheological properties of whey protein/gum arabic coacervates. Journal of Rheology, 48(6), 1215.

Xia, J., Dubin, P. L., & Dautzenbergt, H. (1993). Light Scattering, Electrophoresis, and Turbidimetry Studies of Bovine Serum Albumin-Poly (dimethyldiallylammonium chloride) Complex. Langmuir : the ACS journal of surfaces and colloids, 9(8), 2015–2019.

Xiong, Y. L. (1992). Influence of pH and ionic environment on thermal aggregation of

whey proteins. Journal of Agricultural and Food Chemistry, 40(3), 380–384. 97

Yeo, Y., Bellas, E., Firestone, W., Langer, R., & Kohane, D. S. (2005). Complex coacervates for thermally sensitive controlled release of flavor compounds. Journal of Agriculture and Food Chemistry, 49.

Zhang, Y., & Cremer, P. S. (2006). Interactions between macromolecules and ions: The Hofmeister series. Current opinion in chemical biology, 10(6), 658–63.

Zhang, Y., & Cremer, P. S. (2009). The inverse and direct Hofmeister series for lysozyme. Proceedings of the National Academy of Sciences of the United States of America, 106(36), 15249–53.

APPENDICES

Appendix A Chapter 3: Specific Ion Effects on Electrostatic Interactions and

Coacervation of Aqueous β-Lactoglobulin and Pectin Mixtures

20 a) 18 100 mm 16 75 mm 14 50 mm 12 25 mm 10 0 mm

100-%T 100-%T 8 6 4 2 0 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 pH

40 b) 35 100 mm 30 75 mm 50 mm 25 25 mm 20 0 mm 98 100-%T 100-%T 15 10 5 0 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 pH

20 c) 18 100 mm 16 75 mm 50 mm 14 33.33 mm 12 25 mm 16.67 mm 10 8.33 mm

100-%T 100-%T 8 0 mm 6 4 2 0 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 pH

20 d) 18 100 mm 75 mm 16 50 mm 14 33.33 mm 25 mm 12 16.67 mm 10 8.33 mm 0 mm 100-%T 100-%T 8 6 4 2 99

0 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 pH

Figure 1 Effect of salt concentration (mm) on turbidity of 0.1% β-lactoglobulin solutions as a function of pH (a) Sodium Chloride (NaCl) (b) Ammonium Thiocyanate (NH4SCN) (c) Potassium Sulfate (K2SO4) (d) Ammonium Sulfate ((NH4)2SO4)

100

180.00 a) 100 mm 75 mm 50 mm 33.33 mm 160.00 25 mm 16.67 mm 8.33 mm 0 mm

140.00

120.00

100.00

Scattering Intensity Scattering (mHz) 80.00

60.00 4.20 4.40 4.60 4.80 5.00 5.20 5.40 5.60 5.80 6.00 pH

b) 180.00 100 mm 75 mm 50 mm 33.33 mm 160.00 25 mm 16.67 mm 8.33 mm 0 mm

140.00

120.00

100.00 100 Scattering Intensity Scattering

80.00

60.00 3.80 4.00 4.20 4.40 4.60 4.80 5.00 5.20 5.40 5.60 5.80 6.00 pH

Figure 2 Effect of salt concentration (mm) and species on light scattering of 0.1% β- lactoglobulin and 0.05% pectin solutions as a function of pH with (a) 0-100 mm of potassium sulfate (K2SO4) (b) 0-100 mm ammonium sulfate (NH4)2SO4

101

120.00 a) 100 mm 75 mm 50 mm 33.33 mm 25 mm 16.67 mm 8.33 mm 0 mm 100.00

80.00

60.00 100-%T 100-%T 40.00

20.00

0.00 3 3.25 3.5 3.75 4 4.25 4.5 4.75 5 5.25 5.5 5.75 6 pH

120.00 b) 100 mm 75 mm 50 mm 33.33 mm 25 mm 16.67 mm 8.33 mm 0 mm 100.00

80.00

60.00 100-%T 100-%T

40.00 101

20.00

0.00 3 3.25 3.5 3.75 4 4.25 4.5 4.75 5 5.25 5.5 5.75 6 pH

Figure 3 Effect of salt concentration (mm) and species on turbidity of 0.1% β- lactoglobulin and 0.05% pectin solutions as a function of pH with (a) 0-100 mm of potassium sulfate (K2SO4) (b) 0-100 mm ammonium sulfate (NH4)2SO4

102

Appendix B Chapter 4: Thermal Transitions Among β-Lactoglobulin-Pectin

Electrostatic Complexes 102

Figure 1 AFM height images of thermally treated protein-polysaccharide complexes containing 0.1% β-Lactoglobulin (w/w) and 0.05% pectin (w/w) at pH 4.50 with potassium chloride a) 0 mm b) 25 mm c) 50 mm d) 75 mm e) 100 mm

103

Figure 2 AFM height image and apparent size determination of thermally treated 0.1% BLG and 0.05% pectin particles with 0 mm KCl using Igor software

104

Figure 3 AFM height image and apparent size determination of thermally treated 0.1% BLG and 0.05% pectin particles with 50 mm KCl using Igor software

105 a) 120.000 100 mm 75 mm 50 mm 25 mm 0 mm 100.000

80.000

60.000 100-%T 100-%T 40.000

20.000

0.000 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 pH b) 120.000 100 mm 75 mm 50 mm 25 mm 0 mm 100.000

80.000

60.000 100-%T 100-%T 40.000 105 20.000

0.000 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 pH

106 c) 120.000 100 mm 75 mm 50 mm 25 mm 0 mm 100.000

80.000

60.000 100-%T 100-%T 40.000

20.000

0.000 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 pH d) 120.000 100 mm 75 mm 50 mm 33.33 mm 25 mm 16.67 mm 8.33 mm 0 mm 100.000

80.000

60.000 100-%T 100-%T

40.000 106

20.000

0.000 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 pH

107

e) 120.000 100 mm 75 mm 50 mm 33.33 mm 25 mm 16.67 mm 8.33 mm 0 mm 100.000

80.000

60.000 100-%T 100-%T 40.000

20.000

0.000 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 pH

Figure 4 Thermal transitions 0.1% (w/w) β-lactoglobulin and 0.05% (w/w) pectin solution with a) sodium chloride (NaCl) b) potassium thiocyanate (KSCN) c) ammonium thiocyanate (NH4SCN) d) potassium sulfate (K2SO4) and e) ammonium sulfate -3 -1 (NH4)2SO4 (in 10 mole·kg , mm) as a function of pH

107

a) b) 108

120 100 mm 75 mm 50 mm 25 mm 0 mm 100

60 100-%T 100-%T

0 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Temperature (°C)

Figure 5 Changes to the thermal profile of 0.1% β-Lactoglobulin (w/w) with varying concentrations of Potassium Chloride at pH 4.50 with temperature increase of 1°C per minute

108

Figure 6 AFM amplitude images of thermally treated protein-polysaccharide complexes containing 0.1% β-Lactoglobulin (w/w) and 0.05% pectin (w/w) at pH 4.50 with 16.67 mm a) K2SO4 b) (NH4)2SO4