Reducing Skim Milk Powder Dispersion Turbidity by Dissociation of Casein Micelles at Acidic and Neutral Ph: Physicochemical Properties and Possible Mechanisms

University of Tennessee, Knoxville TRACE: Tennessee Research and Creative Exchange

Masters Theses Graduate School

5-2019

Reducing skim milk powder dispersion turbidity by dissociation of casein micelles at acidic and neutral pH: Physicochemical properties and possible mechanisms

Inseob Choi University of Tennessee, [email protected]

Follow this and additional works at: https://trace.tennessee.edu/utk_gradthes

Recommended Citation Choi, Inseob, "Reducing skim milk powder dispersion turbidity by dissociation of casein micelles at acidic and neutral pH: Physicochemical properties and possible mechanisms. " Master's Thesis, University of Tennessee, 2019. https://trace.tennessee.edu/utk_gradthes/5407

This Thesis is brought to you for free and open access by the Graduate School at TRACE: Tennessee Research and Creative Exchange. It has been accepted for inclusion in Masters Theses by an authorized administrator of TRACE: Tennessee Research and Creative Exchange. For more information, please contact [email protected]. To the Graduate Council:

I am submitting herewith a thesis written by Inseob Choi entitled "Reducing skim milk powder dispersion turbidity by dissociation of casein micelles at acidic and neutral pH: Physicochemical properties and possible mechanisms." I have examined the final electronic copy of this thesis for form and content and recommend that it be accepted in partial fulfillment of the requirements for the degree of Master of Science, with a major in Food Science.

Qixin Zhong, Major Professor

We have read this thesis and recommend its acceptance:

Vermont P. Dia, Tao Wu

Accepted for the Council: Dixie L. Thompson

Vice Provost and Dean of the Graduate School

(Original signatures are on file with official studentecor r ds.) Reducing skim milk powder dispersion turbidity by dissociation of casein micelles at acidic and

neutral pH: Physicochemical properties and

possible mechanisms

A Thesis Presented for the

Master of Science

Degree

The University of Tennessee, Knoxville

Inseob Choi

May 2019

ACKNOWLEDGEMENTS

Dr. Qixin Zhong for accepting me kindly as a master’s student and providing me with research opportunity with great guidance.

Dr. Tao Wu and Dr. Vermont Dia for being my committee members.

Dr. Philipus Pangloli, Dr. John Dunlap, Dr. Michael Essington, Dr. Edward Wright, and Melanie Stewart for scheduling and assisting experiments.

Mary Lamsen, Hao Zhang, Teng Li, Shan Hong, Jennifer Vuia-Riser, Andrea Nieto,

Anyi Wang, Nan Li, Dara Smith, and Melody Fagan, Wenfei Xiong, Cuixia Sun, and Tao Wang for being my best friends.

My family, Heunggeun Choi, Boklee Kim, and Hyeongseob Choi for being my supporters.

USDA National Institute of Food and Agriculture, TEN2015-05921 for supporting the study.

iii

ABSTRACT

Casein micelles comprise 80% of dairy proteins. However, casein micelles as the source for protein fortification in beverages are limited due to its contribution to dispersion turbidity. On the other hand, protein beverages are experiencing substantial market growth and what follows is the demand for new protein ingredients. At acidic conditions, like those commonly found in lots of beverages, dispersions of skim milk powder (SMP) have high turbidity and readily precipitate.

In order to circumvent these problems, the dissociation of casein micelles is required to reduce turbidity and improve dispersion stability during storage.

Additionally, beverages with neutral pH are preferred to consumers suffering dental problems such as tooth erosion. The objectives of the present study were

1) to reduce turbidity of SMP dispersions and 2) characterize their physicochemical properties after treatment at acidic and neutral pH. To acidify

SMP dispersions to pH 2.4-3.0, citric acid or glucono delta-lactone was added, followed by subsequent heating at 60, 70, 80, or 90 °C for 2, 5, 10, 30, or 60 min.

A lower pH (i.e., pH 2.4) was more effective than pH 3.0 to reduce the dispersion turbidity (e.g., 223 vs 861 NTU) and the hydrodynamic diameter (e.g., 115 vs 265 nm) after heating at 90 °C for 10 min. At neutral pH, translucent dispersions were obtained with the addition of calcium chelators; sodium tripolyphosphate, trisodium citrate, or sodium hexametaphosphate. Higher concentrations of calcium chelators in SMP dispersions resulted in significant reduction in turbidity values compared to the control (e.g., 230 NTU vs >4000). The amount of dissolved calcium in the SMP serum phase apart from the casein micelles

increased from ~150 to more than 500 mg/L after treatments at both acidic and neutral pH, indicating the dissociation of casein micelles due to the disruption of colloidal calcium phosphate. The SMP dispersions with modified casein structures can be utilized to produce novel types of beverages.

TABLE OF CONTENTS

Chapter 1 Introduction ...... 1

1.1 Abstract ...... 2

1.2 Literature review ...... 2

1.2.1 Early studies on casein ...... 2

1.2.2 Protein as a nutrient ...... 5

1.2.3 Reduction of milk turbidity ...... 7

1.2.4 Structural changes due to heating of milk ...... 11

1.2.5 Effect of acidity on casein-whey protein complexation during heating of

milk ...... 13

1.2.6 Significance of hydrophobic interactions on casein micelle structure . 14

1.2.7 Significance of colloidal calcium phosphate (CCP) on casein micelle

structure ...... 17

1.2.8 Calcium chelating agents ...... 21

1.3 Hypotheses and objectives...... 22

References ...... 25

Chapter 2 Reducing turbidity of skim milk powder dispersions by dissolving colloidal calcium phosphate from casein micelles and weakening hydrophobic interaction through acidification and subsequent heating ...... 41

2.1 Abstract ...... 42

2.2 Materials and methods ...... 42

2.2.1 Chemicals ...... 42

2.2.2 Sample preparation ...... 43

2.2.3 Turbidity ...... 44

2.2.4 Hydrodynamic diameter and particle size distribution ...... 44

2.2.5 Fluorescence spectrometry ...... 44

2.2.6 Inductively coupled plasma-optical emission spectrometry (ICP-OES)

...... 45

2.2.7 Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-

PAGE) ...... 46

2.2.8 Determination of free thiol groups (SH) concentration ...... 46

2.2.9 Scanning transmission electron microscopy (STEM) ...... 47

2.2.10 Statistical analysis ...... 48

2.3 Results and discussion ...... 48

2.3.1 Turbidity, hydrodynamic diameter, and particle size distribution of skim

milk powder dispersions ...... 48

2.3.1.1 Effect of heating duration for samples acidified with citric acid .... 48

2.3.1.2 Effect of heating temperature for samples acidified with citric acid

...... 49

2.3.1.3 Effect of pH for samples acidified with citric acid or GDL ...... 50

2.3.2 Relative role of casein and whey protein on SMP dispersions after

acidification and heating ...... 51

2.3.3 Hydrophobic interactions between proteins in skim milk powder

dispersions ...... 52

vii

2.3.4 Calcium and phosphorous determination from serum phase of skim

milk powder dispersions ...... 56

2.3.5 Effects of acidification and heating on free thiol groups (SH)

concentration and molecular weights of proteins ...... 57

2.3.6 Structural change of casein micelles using scanning transmission

electron microscopy (STEM) ...... 58

2.4 Conclusion...... 59

References ...... 61

Appendix ...... 67

Chapter 3 Physicochemical properties of skim milk powder dispersions as affected by calcium chelating sodium tripolyphosphate, trisodium citrate, and sodium hexametaphosphate ...... 87

3.1 Abstract ...... 88

3.2 Materials and methods ...... 88

3.2.1 Chemicals ...... 88

3.2.2 Sample preparation ...... 89

3.2.3 Turbidity and hydrodynamic diameter ...... 89

3.2.4 Inductively coupled plasma-optical emission spectrometry (ICP-OES)

...... 90

3.2.5 Zeta-potential ...... 90

3.2.6 Transmission electron microscopy (TEM) ...... 91

3.2.7 Statistical analysis ...... 91

3.3 Results and discussion ...... 92

viii

3.3.1 Turbidity and hydrodynamic diameter of skim milk powder dispersions

as affected by chelators or SDS ...... 92

3.3.2 Calcium and phosphorous determination from serum phase of

dispersions after calcium chelation ...... 93

3.3.3 Zeta-potential of skim milk powder dispersions with chelators ...... 95

3.3.4 Structural changes of calcium-chelated casein micelles using

transmission electron microscopy (TEM) ...... 96

3.4 Conclusion...... 97

References ...... 99

Appendix ...... 100

Chapter 4 Conclusion and future work ...... 120

Vita ...... 123

LIST OF FIGURES

Figure 2.1. The thermal profiles of 10 mL SMP dispersions (5% w/v) in

borosilicate glass scintillation vials after heating at 60 °C (a), 70 °C (b), 80 °C

(c), or 90 °C for 10 min ...... 67

Figure 2.2. The appearance (a), turbidity and hydrodynamic diameter (b), and

particle size distributions (c) of SMP dispersions (5% w/v) after acidifying to

pH 2.4 with citric acid and heating at 90 °C for 2-60 min. The far-left sample

(a) is the unheated control at pH 6.8. Different letters above symbols indicate

significant difference in the compared parameter (P < 0.05). Error bars are

SD (n=3) ...... 68

Figure 2.3. The appearance (a), turbidity and hydrodynamic diameter (b), and

particle size distributions (c) of SMP dispersions (5% w/v) after acidifying to

pH 2.4 with citric acid and heating at 60-90 °C for 10 min. The far-left sample

(a) is the unheated control at pH 6.8. Different letters above symbols indicate

significant difference in the compared parameter (P <0.05). Error bars are

SD (n=3) ...... 69

Figure 2.4. The appearance (a), turbidity and hydrodynamic diameter (b), and

particle size distributions (c) of SMP dispersions (5% w/v) after acidifying to

pH 2.4-3.0 with citric acid and heating at 90 °C for 10 min. The far-left

sample (a) is the unheated control at pH 6.8. Different letters above symbols

indicate significant difference in the compared parameter (P < 0.05). Error

bars are SD (n=2) ...... 70

Figure 2.5. The appearance (a), turbidity and hydrodynamic diameter (b), and

particle size distributions (c) of SMP dispersions (5% w/v) after acidifying to

pH 2.5-3.0 with GDL and heating at 90 °C for 10 min. The far-left sample (a)

is the unheated control at pH 6.8 ...... 71

Figure 2.6. The effect of heating (90 °C for 10 min) on appearance (a) and

turbidity (b) after acidifying to pH 2.4-3.0 with citric acid. The far-right sample

(a) is the unheated control at pH 6.8. Different letters above bars indicate

significant difference in the compared parameter (P < 0.05). Error bars are

SD (n=3) ...... 72

Figure 2.7. Turbidity and hydrodynamic diameter of SMP dispersions (5% w/v)

after acidification to pH 2.5 with citric acid and heating at 70 °C (a) or 90 °C

(b) for 2-30 min. Different letters above symbols indicate significant

difference in the compared parameter (P < 0.05). Error bars are SD (n=3). 73

Figure 2.8. Turbidity and hydrodynamic diameter of MC dispersions (1.73% w/v)

after acidification to pH 2.5 with citric acid and heating at 70 °C (a) or 90 °C

(b) for 2-30 min. Different letters above symbols indicate significant

difference in the compared parameter (P < 0.05). Error bars are SD (n=3) . 74

Figure 2.9. Turbidity and hydrodynamic diameter of dispersions with 1.39% MC

and 0.39% WPI (w/v) after acidification to pH 2.5 with citric acid and heating

at 70 °C (a) or 90 °C (b) for 2-30 min. Different letters above symbols

indicate significant difference in the compared parameter (P < 0.05). Error

bars are SD (n=3) ...... 75

Figure 2.10. Particle size distributions of MC dispersions (1.73% w/v) after

acidification to pH 2.5 with citric acid and heating at 70 °C (a) or 90 °C (b) for

2-30 min ...... 76

Figure 2.11. Intrinsic fluorescence spectra of SMP dispersions (5% w/v) before

and after acidification to pH 2.4-3.0 with citric acid, before and after heating

at 90 °C for 2 min ...... 77

Figure 2.12. Fluorescence spectra of SMP dispersions (5% w/v) with pyrene

before and after acidification to pH 2.4-3.0 with citric acid, before and after

heating at 90 °C for 2 min ...... 79

Figure 2.13. Fluorescence spectra of SMP dispersions (5% w/v) with ANS

before and after acidification to pH 2.4-3.0 with citric acid, before and after

heating at 90 °C for 2 min ...... 81

Figure 2.14. Calcium (a) and phosphorous (b) concentrations in the serum

phase of SMP dispersions (5% w/v) before and after acidification to pH 2.4-

3.0 with citric acid, before and after heating at 90 °C for 2 min. Different

letters above bars indicate significant difference in the compared parameter

(P < 0.05). Error bars are SD (n=3) ...... 82

Figure 2.15. SDS-PAGE of SMP dispersions (5% w/v) before (a) and after (b)

heating at 90 °C for 2 min after acidification to pH 2.4, 2.7, and 3.0 (Left to

Right) with citric acid. The far-right sample is unacidified control at pH 6.8 . 83

Figure 2.16. Contents of free thiol groups in the SMP dispersions (5% w/v)

before and after acidification to pH 2.4-3.0, before and after subsequent

xii

heating at 90 °C for 2 min. Different letters above bars indicate significant

difference in the compared parameter (P < 0.05). Error bars are SD (n=2) . 84

Figure 2.17. STEM image of SMP dispersion (1% w/v) at pH 6.8 ...... 85

Figure 2.18. STEM image of SMP dispersion (1% w/v) after acidification to pH

2.4 with citric acid and heating at 90 °C for 10 min ...... 86

Figure 3.1. The appearance of 5% w/v SMP dispersions with 0-30 mM STPP

before (a) and after (b) heating at 90 °C for 5 min ...... 100

Figure 3.2. The appearance of 5% w/v SMP dispersions with 0-30 mM TSC

before (a) and after (b) heating at 90 °C for 5 min ...... 101

Figure 3.3. The appearance of 5% w/v SMP dispersions with 0-30 mM SHMP

before (a) and after (b) heating at 90 °C for 5 min...... 102

Figure 3.4. The hydrodynamic diameter of 5% w/v SMP dispersions with 0-30

mM STPP before and after heating at 90 °C for 5 min. Different letters above

bars indicate significant difference (P < 0.05). Error bars are SD (n=2) ..... 103

Figure 3.5. The hydrodynamic diameter of 5% w/v SMP dispersions with 0-30

mM TSC before and after heating at 90 °C for 5 min. Different letters above

bars indicate significant difference (P < 0.05). Error bars are SD (n=2) ..... 104

Figure 3.6. The hydrodynamic diameter of 5% w/v SMP dispersions with 0-30

mM SHMP before and after heating at 90 °C for 5 min. Different letters

above bars indicate significant difference (P < 0.05). Error bars are SD (n=2)

...... 105

xiii

Figure 3.7. The turbidity of 5% w/v SMP dispersions with 0-30 mM STPP before

and after heating at 90 °C for 5 min. Different letters above bars indicate

significant difference (P < 0.05). Error bars are SD (n=2) ...... 106

Figure 3.8. The turbidity of 5% w/v SMP dispersions with 0-30 mM TSC before

and after heating at 90 °C for 5 min. Different letters above bars indicate

significant difference (P < 0.05). Error bars are SD (n=2) ...... 107

Figure 3.9. The turbidity of 5% w/v SMP dispersions with 0-30 mM SHMP before

and after heating at 90 °C for 5 min. Different letters above bars indicate

significant difference (P < 0.05). Error bars are SD (n=2) ...... 108

Figure 3.10. The appearance of 5% w/v SMP dispersions with 0-0.35 g SDS . 109

Figure 3.11. Turbidity and hydrodynamic diameter of 5% w/v SMP dispersions

after addition of 0-0.35 g SDS. Different letters above symbols indicate

significant difference in the compared parameter (P < 0.05). Error bars are

SD (n=2) ...... 110

Figure 3.12. TEM images of 1% w/v SMP dispersion at pH 6.8 ...... 115

Figure 3.13. TEM images of 1% w/v SMP dispersion with 1 mM (a), 7 mM (b), or

30 mM (c) STPP...... 116

Figure 3.14. TEM images of 1% w/v SMP dispersion with 1 mM (a), 7 mM (b), or

30 mM (c) TSC...... 117

Figure 3.15. TEM images of 1% w/v SMP dispersion with 1 mM (a), 7 mM (b), or

30 mM (c) SHMP...... 118

Figure 3.16. TEM images of 1% w/v SMP dispersion with 0.30 g SDS ...... 119

xiv

LIST OF TABLES

Table 2.1. The wavelength (λmax) of 5% w/v SMP dispersions before and after

acidification to pH 2.4-3.0, before and after heating at 90 °C for 2 min ...... 78

Table 2.2. I3/I1 of SMP dispersions (5% w/v) with pyrene before and after

acidification to pH 2.4-3.0 with citric acid, before and after heating at 90 °C

for 2 min ...... 80

Table 3.1. Calcium (Ca) concentrations in the serum phase of 5% w/v SMP

dispersions after addition of 0-30 mM STPP, TSC, and SHMP ...... 111

Table 3.2. Phosphorous (P) concentrations in the serum phase of 5% w/v SMP

dispersions after addition of 0-30 mM STPP, TSC, and SHMP ...... 112

Table 3.3. Calcium (Ca) and phosphorous (P) concentrations in the serum

phase of 5% w/v SMP dispersions after addition of 0-0.30 g SDS...... 113

Table 3.4. Zeta-potential of 5% w/v SMP dispersions after addition of 0-30 mM

STPP, TSC, and SHMP ...... 114

CHAPTER 1

Introduction

1.1 Abstract

Casein micelles in milk are colloidal particles consisting of four types of molecules (ɑs1-, ɑs2-, β-, and κ-casein). Casein is one of the most studied food protein because of its techno-functionalities such as its thermal stability, viscosity, gelation properties, and emulsifying/foaming properties. The structure of the casein micelles is governed by a fine balance of electrostatic repulsion, attractive hydrophobic interactions, linkage of colloidal calcium phosphate, disulfide bonds, etc. Studying these bonds and interactions, therefore, is essential to understanding structural changes of casein micelles within various environments.

1.2 Literature review

1.2.1 Early studies on casein

The casein, which represents approximately 80% of total dairy protein in milk, consists of four types of molecules, ɑs1, ɑs2, β, and κ, although it once was considered as a homogeneous material by Hammersten, who developed the method of isoelectric precipitation of casein at pH 4.6 (Hammersten, 1883). Later electrophoresis studies have confirmed that the casein is a mixture of molecules

(Mellander, 1939). These casein molecules (i.e., ɑs1-, ɑs2-, β-, and κ-casein) exist as large colloidal particles with a micellar structure in milk (Fox & Brodkorb,

2008). The term “micelle” might not be appropriate because casein micelles do not adhere to the typical characteristics of micelles because their formation is

irreversible over normal time-scales (Dickinson, 1992). However, the word

“casein micelle” was coined for the first time by Beau (1921) and established as an acceptable term during the 1950s when properties of casein particles such as micelle formation and equilibria of caseins were actively studied (Pyne, 1955;

Waugh, 1958, 1961).

Colloidal stability of casein micelles has been of great interest for a long time because the techno-functionalities such as thermal stability, viscosity, gelation properties, and emulsifying/foaming properties, of casein micelles are considered paramount in industrial applications (Fox & Brodkorb, 2008; Broyard &

Gaucheron, 2015). Since the existence of κ-casein in casein molecules has been known (Waugh and von Hippel, 1956), the colloidal stability of casein micelles has attracted lots of attention from those who want to find how casein micelles are stabilized or destabilized as environmental conditions (e.g., heat, pH, pressure, addition of ethanol, etc) are changed. Being associated, the integrity of the casein micelle is influenced by casein-casein interactions or bridges between various casein molecules and colloidal calcium phosphate. For casein-casein interactions, various forces such as hydrogen bonding, hydrophobic interactions, and electrostatic interaction are involved in holding casein molecules together

(Smits & van Brouwershaven, 1980; Horne, 1998). Also, the importance of the presence of calcium phosphate in improving the integrity of the casein micellar structure has been well known (Aoki et al., 1987; Holt, 1998). Berridge (1954) already described it: “it has been known for a considerable time that the casein in milk exists in the form of micelles with calcium phosphate”.

The general properties of the four casein molecules (ɑs1-, ɑs2-, β-, and κ) have been well described (Horne, 1998). The net charges of casein molecules are known as -21.9, -17.1, -13.3, and -2.0 at pH 6.6 for ɑs1-, ɑs2-, β-, and κ, respectively (O’Connell & Flynn, 2007). Especially, the self-association of ɑs1- and ɑs2-casein, driven by hydrophobic interactions, shows similar association patterns due to their substantial negative net charges (O’Connell & Flynn, 2007).

The hydrophobic ends of ɑs-casein are intermolecularly connected, forming lengthy pole-like structures (Horne, 1998). β-casein has mobility influenced by temperature. For example, cooling at 4 °C liberates β-casein from casein micelles with reversible partitioning upon warming (Broyard and Gaucheron,

2015). However, the original micellar structure is not fully recovered after increasing the temperature because penetration of β-casein into micelles is inhibited, altering the surface structure of micelles (Dalgleish, 1998). Differences in the degree of phosphorylation of casein molecules provide them with the ability to bind a significant amount of calcium phosphate (Broyard and

Gaucheron, 2015). κ-casein interacts with ɑs- and β-casein through hydrophobic interactions, but it does not provide an anchor point for not only further extension of hydrophobic interaction chains but also colloidal calcium phosphate bridges

(Horne, 1998). Thus, κ-casein acts as “size limiter,” defining the degree of polymerization of casein molecules. These characteristics of each casein molecule govern the equilibria of casein micelles in solution depending on various environmental factors.

1.2.2 Protein as a nutrient

As both the global population and standards of living continue to increase, a higher demand on animal derived proteins is expected (Boland et al., 2013). For example, the Food and Agriculture Organization (FAO) has forecasted that the total consumption of dairy products will increase by 82% between 2000 and 2050

(FAO, 2006). This corresponds to 466 million tons of milk (Boland et al., 2013).

The intake of a high-protein diet has a variety of health benefits. Consumption of protein provides satiety by stimulating a hormone that delays the emptying process of the food from the stomach (Storr et al., 2003). Fulfilling the recommended protein intake (25-30 g/meal) can be adopted as a great strategy to prevent sarcopenia, which is the decline of muscle with age (Paddon-Jones et al., 2015). The benefit of animal-based proteins over plant-based proteins is its higher proportion of the amino acid leucine, which plays a significant role in muscle protein anabolism (Paddon-Jones et al., 2015). Particularly, milk is an animal-based food of high nutritive value. With reference to calcium, milk and milk-derived products comprise more than 70% of the bioavailable calcium in the

US diet (Huth, DiRienzo, & Miller, 2006). This means that it is not easy to meet the daily calcium intake without consuming milk or dairy products. It is reported that diet-induced weight loss strategies that pair high dairy protein intake with exercise promote fat mass loss and lean muscle mass gain (Josse et al., 2011).

Milk protein concentrates and whey protein concentrates can be used by the body to preferentially synthesize muscle protein in middle-aged men (Mitchell et al., 2015). A meal with a protein content of 30-45 g is recommended to increase

and/or maintain muscle strength and lean body mass (Loenneke et al., 2016).

Despite an increasing demand on high-protein food and beverages, the consumption of plain milk as a beverage has been continuously decreasing since the 1970s (Putnam & Gerrior, 1999).

Whey protein has been extensively studied and used for producing beverages with proteins (Gad et al., 2013; Shiby, Radhakrishna, & Bawa, 2013) because protein-added drinks reduce the soreness of athletes (Millard-Stafford et al.,

2005) and enhance their fluid retention, minimizing dehydration (Seifert, Harmon,

& DeClercq, 2006) compared to carbohydrate-only drinks. Skim milk powder

(SMP) is a source of high-quality protein (36 g/100 g) with little fat content (1.2 g/100 g) (Devine, Prince, & Bell, 1996; Eller & Reimer, 2010). Thus, SMP can be used to produce fortified foods to improve nutritional status of vulnerable children with malnutrition (Hoppe et al., 2008). When compared as the price per kg protein, SMP ($ 6.73/kg protein) is the better option than whey protein concentrates ($ 7.76/kg protein) or whey protein isolates (WPIs) ($ 8.07/kg protein) (Hoppe et al., 2008). Also, although casein or whey protein could be a good protein source itself, SMP might have more nutritional values as a combination of the two and other trace components. The body composition of rats fed with SMP exhibited lower body fat (16.5 %) than those fed with casein

(20.2 %) (Eller & Reimer, 2010). The digestion of casein and whey protein together seems to have a unique digestive property that influences plasma amino acid concentrations and profiles when SMP is digested compared to the individual protein sources (Eller & Reimer, 2010). However, use of SMP has

been limited in beverages because consumers prefer clear beverages with no precipitation (Singh, Singh, & Patil, 2001). Overall, the novel production and consumption of protein-fortified drinks with SMP could be explored if the size of the casein micelles in SMP could be reduced enough to produce dispersions with no sedimentation and limited turbidity.

1.2.3 Reduction of milk turbidity

Casein constitutes about 27 g of milk protein per liter of milk, while 7 g of whey protein exists in the same amount of milk (Creamer & MacGibbon, 1996). Casein micelles, spherical aggregates, exhibit polydisperse property with the range of diameters from 150 to 300 nm approximately (Muller-Buschbaum et al., 2007)

Although extensive research on casein has been done for two decades, there is still no consensus on the structure of the casein micelle. However, the following aspects seem to be generally agreed upon: (1) in the center of micelle, lots of water and colloidal calcium phosphate exist; (2) colloidal calcium phosphate is necessary to maintain the integrity of the micellar structure of casein; and (3) κ- casein covers the outer surface of casein micelles (Huppertz, 2016). It is known that the size of the casein micelle is influenced by casein polymorphism heterogeneity, and the proportion of κ-casein is higher in small casein micelle groups (11.0±0.7%) than large ones (7.6±0.6%) (Day et al., 2015). Small casein micelle groups showed a high proportion of both glycosylated and non- glycosylated κ-casein compared with large casein micelle groups but it appears

that the amount of glycosylated κ-casein is more important for determining the size of casein micelle (Day et al., 2015).

Several methods have been made to reduce turbidity of milks. The most studied method is high-pressure treatment. Pressure treatments on milk are divided into two groups depending on the level of pressure: high-pressure treatment for 150-

200 MPa and ultra-high-pressure treatment for more than 300 MPa. In native casein micelles, the integrity of micelles is maintained by well-balanced hydrophobic and electrostatic interactions (Orlien, Boserup, & Olsen, 2010).

However, high hydrostatic pressure treatment on skim milk can decrease the turbidity of dispersions by reducing the mean hydrodynamic diameter of casein micelles (Johnston, Austin, & Murphy, 1992; Desobray-Banon, Richard, & Hardy,

1994). The high-pressure treatment of 250 MPa showed temperature dependency on average diameter of particles and exposure of hydrophobic regions of milk proteins, but the effect of temperature was gradually decreased as hydrostatic pressure unit increased to 450 and 600 MPa (Gaucheron et al.,

1997). Below 300 MPa, high-pressure treatment at a lower temperature led to skim milk with lower turbidity (Needs et al., 2000) and smaller particle size

(Gaucheron et al., 1997) compared to the treatment at a higher temperature.

Transmission electron microscopy (TEM) images suggested that big casein micelles could be partially disaggregated under the high-pressure treatment of

200 MPa and complete disaggregation could occur under 400 and 600 MPa of treatment with the mean micelle diameters of 110 and 106 nm, respectively

(Needs et al., 2000). Although it is clear that high-pressure treatment could

disrupt hydrophobic and electrostatic interactions that support micellar structure, the mechanism of dissociation has not been clarified (Orlien, Boserup, & Olsen,

2010).

When milk is dialyzed, the size of casein micelles and the amount of calcium, inorganic phosphorous, and casein molecule in the pellet after centrifugation are in parallel relation (Ono, Furuyma, & Odagiri, 1981). It appeared that large micelles are dissociated faster under dialysis at 37 °C than at 5 °C (Ono,

Furuyma, & Odagiri, 1981). Also, free calcium concentration in dialysis buffer could highly affect casein micelle structure. When free calcium concentration was less than 1 mM in the buffer, casein partition in the pellet after centrifugation was distinctively lower than that of undialyzed milk, indicating the size of casein was reduced (Holt, Davies, & Law, 1986). By contrast, a higher level of free calcium

(e.g., 3 mM) in the buffer caused almost complete sedimentation of casein molecules after centrifugation (Holt, Davies, & Law, 1986). This suggests that free calcium ion content in milk is even involved with the conversion of serum casein to aggregated form as well as dissociation of casein micelles (Holt,

Davies, & Law, 1986). However, appreciable dissociation through dialysis requires considerable time (~72 h) and a significant reduction of colloidal calcium phosphate (Aoki et al., 1988).

At neutral pH (e.g., pH 6.7), most of the casein molecules (~95%) are involved in the formation of casein micelles while each molecule combines together (Zimet,

Rosenberg, & Livney, 2011). The alkalization of milk is not a common process in industry; however, several studies have been done to elucidate the behavior of

casein micelles above a pH of 7 (Vaia et al., 2006; Ahmad et al., 2009; Pan &

Zhong 2013). The decreased turbidity of milk with increased pH indicates the disruption of casein micelles at alkaline conditions. Vaia et al. (2006) explains that the alkaline disruption of casein micelles is not due to the collapse of the calcium phosphate bridge because the solubility of calcium phosphate is decreased as the pH is increased. Rather, weakened cohesive interactions between the hydrophobic regions of the caseins were likely responsible for the disruption of the casein micelles. On the other hand, Liu and Guo (2008) reported that casein molecules still self-assemble into casein micelles at a pH higher than

9.0 through hydrophobic interactions between portions of casein molecules. The different results of these studies might be attributed to the types of casein (ɑ-, β-, or micellar casein) used or the concentration of casein in the dispersions. Also, reduced contrast and diminished turbidity were observed in casein micelles dispersions with 6 M urea as measured by small angle neutron scattering

(SANS) (de Kruif et al., 2012) or UV/vis spectrophotometer (Huppertz, Smiddy, & de Kruif, 2007). This is possibly due to liberated proteins from casein micelles in the presence of urea (Dalgleish, Pouliot, & Paquin, 1987) due to its ability to destabilize hydrophobic interactions between proteins (Wallqvist, Covell, &

Thirumalai, 1998) or remove water molecules from proteins (Nandel et al., 1998), causing the structural change of casein micelles. However, the methods employed to reduce turbidity of milk protein dispersions described above are not practical in food applications.

1.2.4 Structural changes due to heating of milk

Casein micelles are remarkably heat stable in the absence of whey proteins

(Livney, Corredig, & Dalgleish, 2003). However, when skim milk is heated in the presence of whey protein, a heterogeneous casein-whey protein complex forms by deposition of whey protein onto casein micelles (Corredig & Dalgleish, 1996;

Dalgleish, van Mourik, & Corredig, 1997). The major mechanism of complex formation is attributed to thiol-disulfide interchange reactions (Havea, Singh, &

Creamer, 2001), particularly between β-lactoglobulin (β-lg) and κ-casein (Jang &

Swaisgood, 1990). Elevating temperature or pH is favorable for free thiol group in

Cys 121 of β-lg to initiate thiol-disulfide interchange reactions. (Otte, Zakora, &

Qvist, 2000). A large particle size (100-599 nm) was obtained after heating β-lg at a temperature as low as 45 °C (Sharma, Haque, & Wilson, 1996). Intensity- weighed particle size distribution of casein micelles did not show a significant change after heating at 80 °C without whey protein whereas monomodal distribution changed to bimodal distribution after the heating with whey protein, which suggests that whey protein might be deposited onto the casein micelles or there were large aggregations of whey protein (Tran Le et al., 2008). Tran Le et al. (2008) concluded that the increased diameter of the casein micelle was probably due to aggregated casein micelles but did not provide enough explanation in their work as why casein aggregation would result in bimodal size distribution whereas coating of whey protein onto casein would lead to monomodal distribution. Another study also described that the size of the casein micelle was maintained when whey protein depleted milk was heated, but the

addition of β-lg caused an increase in the size of casein micelles after the heating of milk (Anema & Li, 2003). Heating of milk (70-90 °C) can cause the exposure of the buried reactive thiol groups of β-lg, denaturing the native conformation of whey protein (Tran Le et al., 2007). This could lead to either the coating of whey protein on the casein micelle or aggregation of whey proteins among themselves

(Livney, Corredig, & Dalgleish, 2003). The formation of whey protein aggregates or casein-whey complexes depends on the pH of milk while heating. If milk is heated above pH 6.6, aggregation of whey protein occurs while casein micelles are partially coated by whey protein; below pH 6.6, heating milk induces casein- whey protein complexes without any whey protein aggregation (Vasbinder & de

Kruif, 2003). Tran Le et al. (2007) reported that highly hydrolyzed lecithin and hydroxylated lecithin have heat-stabilizing properties seemingly due to the exposed reactive thiol groups of the hydrophobic parts of whey protein adsorbing onto lysolecithins. This interaction between whey protein and lecithin was used to isolate whey protein from milk and prevent further interaction between whey protein and casein micelles prohibiting the heat-induced increase in hydrodynamic diameter because the initial aggregates of whey proteins, (i.e., β-lg and α-lactalbumin (α-la)), during heating function as intermediates of the casein- whey complexes that are formed in prolonged heating conditions (Corredig &

Dalgleish, 1996; Corredig & Dalgleish, 1999). However, the size increase of casein micelles due to heating was also inhibited by sodium laurate or sodium dodecyl sulfate (SDS) after 15 min of heating at 80 °C, demonstrating that the heat-stabilizing property is not a unique characteristic of lecithin (Tran Le et al.,

2007). Some studies described that at the initial stage of aggregation between casein micelles and whey proteins, hydrophobic or ionic interactions are important (Mulvihill & Donovan, 1987; Hill, 1989). The temperature of heating is not the only factor, but also the rate of heating milk is known to affect the association extent of whey protein components. For example, when milk was heated slowly, almost 80% of the denatured β-lg interacted with the casein micelles; quick heating, on the other hand, resulted in only about 50% of association of β-lg and α-la to casein micelles (Smits & van Brouwershaven,

1980; Corredig & Dalgleish, 1996). Heating β-lg for 4 h or WPI dispersions for 2 h

(pH 7.2, 9% w/w) at 68.5 °C led to soluble aggregates with the hydrodynamic diameter of 80 nm and 75 nm, respectively (Alting et al., 2002). On the whole, not only heating conditions, but also other parameters such as the pH of the dispersions influence the structural change of casein micelles while heating.

1.2.5 Effect of acidity on casein-whey protein complexation during heating of milk

Acidification affects structure of milk components and a variety of the functional characteristics of milk. The extent of association between whey proteins and casein micelles is also dependent on the pH of milk: a greater amount of denatured whey proteins interacts with casein micelles at a lower pH (Vasbinder

& de Kruif, 2003). For example, 24% of β-lg complexed with casein micelles in a model milk system with a pH of 7.3, and the amounts of associated β-lg increased as pH decreased (i.e., 44 % for pH 6.8, 76 % for pH 6.3, and 83 % for

pH 5.8) (Smits & van Brouwershaven, 1980). Corredig and Dalgleish (1996) also reported that the amount of whey proteins that were associated with casein micelles increased after heating as the pH was lowered (from 6.8 to 6.2 to 5.8).

Oldfield et al. (2000) explained that the dissociation of κ-casein from the casein micelles at elevated pH caused decreased association of whey proteins with the casein micelles after heating at 90 °C; 90 % of whey protein complexation with casein micelles at pH 6.48 and 60 % at pH 6.83. Even when heating of milk samples with slight pH differences, the change in particle size was significant (5-

10 nm increase for pH 6.7 milk vs 30-35 nm increase for pH 6.55 milk) (Anema &

Li, 2003).

1.2.6 Significance of hydrophobic interactions on casein micelle structure

Various models of the micellar structure of casein have been proposed for several decades, however, most proposed models fall under the three categories: coat-core, submicelle, or internal structure models, which were initially proposed by Waugh and Noble (1965), Morr (1967), and Rose (1969), respectively. Each model has been further developed by younger generation such as the submicelle model (Slattery & Evard, 1973; Schmidt, 1982; Walstra,

1999) or the equilibrium thermodynamic model by Holt (2004). The submicelle model emphasizes the role of hydrophobic interactions between casein molecules as the first step for submicelle aggregation in which the variation of κ- casein contents is induced in each submicelle. Therefore, submicelles rich in κ- casein are oriented toward the exterior of the micellar surface while other

submicelles are packed in the interior of the whole micelle structure (Holt, 2004) because κ-casein has glycosylation up to 50%, providing a hydrophilic property for its C-terminal portion (Broyard & Gaucheron, 2015). The κ-casein performs as a size-limiter at the surface location of micelles by providing casein micelles with steric stabilization (Slattery, 1976). Hydrophobic regions of ɑs1-, ɑs2-, and β- caseins surround a nanocluster network inside the micelle (Horne, 2006).

Therefore, attempts to control the hydrophobic interactions have been made to manipulate the micellar structure by controlling extrinsic factors. In milk, thermal stability is a notable feature of casein; however, heating above 80 °C causes the denaturation of whey proteins because denaturation temperatures of β-lg and ɑ- la are 78 °C and 62 °C, respectively (Donovan & Mulvihill, 1987; Ryan, Zhong, &

Foegeding, 2013). Disulfide bonding between κ-casein and β-lg is the major driving force of their interaction, which explains the casein-whey protein complex during the heat treatment of milk (Cho, Singh, & Creamer, 2003); however, hydrophobic interactions might also affect the association of whey protein with casein micelles (Hill, 1989).

According to one study (Mozhaev et al., 1996), electrostatic and hydrophobic interactions are disrupted in proteins by higher pressure effects. Also, above the critical pressure, denaturation of proteins occurs as the hydrophobic effect decreases (Marques et al., 2003). This also applies to casein micelles. Based on the model suggested by Horne (1998), the association state of casein micelles is the result of a balance between an attractive hydrophobic interaction and an electrostatic repulsion. In this model, calcium phosphates make a bridge between

calcium sensitive caseins (ɑs1- and β-casein), which are highly phosphorylated

(Liu and Guo, 2008). This strengthens hydrophobic interactions between hydrophobic regions of caseins by neutralizing negatively charged phosphoserine groups. The disruption of hydrophobic interaction by high- pressure, which is involved in the stability of casein micelles, causes the dissociation of the micelles, resulting in a decreased average diameter of casein micelles (Huppertz, Fox, & Kelly, 2004; Sandra & Dalgleish, 2005). Needs et al.

(2000) suggested that the disrupting effect of high-pressure on hydrophobic bonds is transient and reversible as hydrophobic bonds are rapidly recovered after the release of pressure. As mentioned above, the hydrophobic interaction and electrostatic repulsion of casein micelles can be controlled by interactions between calcium phosphate and casein.

The equilibria between serum and micellar calcium or micellar phosphate are greatly influenced by temperature, indicating solubilization of calcium phosphate is temperature dependent. Koutina et al. (2014) confirmed that the micellar calcium concentration was lowest at 4 °C in a pH range between 4.6 and 6.6 after storage of milk samples at 4, 20, 30, or 40 °C for one day. Also, at the same temperature, the micellar calcium concentration decreased dramatically as the pH was decreased whereas the free calcium concentration increased with the decrease in pH. This explains the difference in the turbidity of milk samples after high-pressure treatment at different temperatures and why the threshold pressure for the dissociation of casein micelle is lower at pH 5.5 (150 MPa) than at pH 7.5 (350-400 MPa (Orlien, Boserup, & Olsen, 2010).

1.2.7 Significance of colloidal calcium phosphate (CCP) on casein micelle structure

Before the term “casein micelle” was used universally, casein particles were mentioned as “calcium caseinate-calcium phosphate particles” (Richardson &

Palmer, 1929; Hankinson & Palmer; 1943). The presence of calcium phosphate plays a significant role in casein micelle integrity. Although the very first submicelle model did not consider calcium phosphate as a major factor for micellar structure, Schmidt (1980) elaborated on the model by adding calcium phosphate into the already suggested assembly process of submicelles. Even though the submicelle model has been developed and elaborated, the hypothesis in the model has not been accepted by some researchers (Holt et al., 1998;

Horne, 1998). Improvement of electron microscopy has contributed to the further characterization of the casein micelle structure. McMahon and McManus (1998) suggested that submicelle model should be rethought because they failed to obtain sufficient evidence to support the submicelle model using a TEM technique. Also, no spherical submicelles were visible in the images obtained by the technique of Field Emission Scanning Electron Microscopy (FESEM)

(Dalgleish, Spagnuolo, & Goff, 2004).

However, calcium phosphate is recognized as a key factor in the integrity of micellar structure of casein, even in the submicelle model. Fine structural details of the casein micelle, characterized by Cryo-TEM technique has strengthened the model suggested by Holt (2004), showing that the detected substructures are

CCP nanoclusters without the existence of submicelles (Marchin et al., 2007).

Because of the importance of the interaction between nanoclusters of calcium phosphate and caseins in the micellar structure, studies on the dissociation of the casein micelle have focused on the disruption of calcium phosphate linkages with casein by high-pressure treatment (Huppertz & Kruif, 2007; Orlien et al., 2010;

Bravo et al., 2013) or change in the calcium-phosphate balance between colloidal and diffusible phases by the addition of calcium chelating agents

(Panouille et al., 2005; Kaliappan & Lucey, 2011; Lazzaro et al., 2017). Extrinsic factors during high-pressure processing (pressurization time, level of pressure, release rate of pressure, or temperature) and intrinsic factors of milk (pH, ionic strength, or protein concentration) have an influence on the modification of the casein micellar structure. The most notable characteristic of the casein micelle after high-pressure treatment was the decrease in its hydrodynamic diameter through the disintegration of the casein micelle. Many studies have covered the effect of high-pressure treatments on milk (Datta & Deeth, 1999; Huppertz, Kelly,

& Fox, 2002), to find an accurate mechanism as an in-situ measurement.

However, it is tough to observe how high pressure affects hydrophobic interactions or calcium phosphate linkages because the effect of pressure is not maintained after release of the pressure treatment (Anema, Lowe, & Stockmann,

2005). But the solubilization of CCP from casein micelles might occur during high-pressure treatment. Schrader, Buchheim, & Morr (1997) showed that dissolved calcium was reduced after heating of milk (118 °C /15 min), however, high-pressure treatment (400 MPa/ 5 min) increased solubilized calcium to the concentration it was before heat treatment. Also, calcium phosphate precipitates

formed during heat treatment were resolubilized after high-pressure treatment.

However, it was reported that a considerable portion of these heat-induced chemical changes was reversible during storage at room temperature of ~24 °C

(Schrader, Buchheim, & Morr, 1997).

Calcium phosphate in the casein micelle exists as CCP and is in equilibrium with diffusible (or soluble) calcium ions and phosphate ions in the serum phase of milk

(Walstra et al, 1999). The concentrations of calcium and phosphate, which are solubilized during acidification were studied, showing great solubilization of calcium and inorganic phosphate ions from pH 6.0 to pH 4.0 (Graet &

Gaucheron, 1999).

According to the original model presented by Mekmene, Graet, & Gaucheron

(2009), the equilibria of salts between the colloidal and diffusible phase can be predicted. Calcium chelating agents such as citrate, EDTA, or polyphosphate have the ability to cause casein micelles to be demineralized. Thus, calcium chelating agents or emulsifying salts have been added to dissociate casein micelles, changing the properties of a cheese system (Lucey, Johnson, & Horne,

2003; O’Mahony, Lucey, & McSweeney, 2005) or the emulsifying properties of disaggregated casein micelles (Lazzaro et al., 2017). Each calcium chelating salt has a different interaction with casein micelles, causing the calcium chelating salts-dependent functionality of processed cheeses (Mizuno & Lucey, 2005). In the study, CCP in casein micelles was chelated by trisodium citrate, increasing the dispersion of casein micelles, whereas sodium hexametaphosphate (0.1% wt/wt) made a complex with diffusible casein and casein micelles (Mizuno &

Lucey, 2005). However, the behavior of binary mixtures of calcium chelating salts on casein particles could not be anticipated based on the behavior of single calcium chelating salts (Kaliappan & Lucey, 2011). Also, some chelating salts

(e.g., sodium hexametaphosphate, monosodium phosphate, disodium phosphate, and dipotassium phosphate) at high concentrations caused the turbidity of skim milk ultrafiltrate dispersion to be increased, suggesting newly formed aggregates in the dispersion (Culler, Saricay, & Harte, 2017).

Like the structure of casein micelles, that of CCP has also not been clarified.

McGann et al. (1983) described CCP as amorphous granules with 2.5 nm for the diameter using electron microscopy. A recent model assumes bi-pyramid or tetrahedron shapes with three or four phosphoserine residues per one CCP nanocluster (Horne, Lucey, & Choi, 2007). Several attempts were made to study the molecular weight of CCP. Holt et al. (1998) suggested the shell model to describe CCP with 1.6 nm thickness of shell, assuming approximate 48 peptides stabilize CCP. They estimated that the molecular weight of CCP was 61,000 g/mol; however, too many peptides were involved in their model conflicting with

Aoki, Umeda, & Kako (1992), who argued that only 3 phosphate groups are needed to crosslink CCP with casein micelles. On the basis of results from samples treated with trypsin or papain/proteinase using size exclusion chromatography, estimated molecular weights of CCP were 7,282 g/mol and

13,940 g/mol, respectively (Choi, Horne, & Lucey, 2011).

The importance of CCP to maintain the integrity of micellar structure is in agreement with all three models: coat-core (Paquin et al., 1987), submicelle

(Walstra, 1999), and internal structure (Holt, 1992; Horne, 1998). The function of

CCP in casein micelle is to link each casein molecule as a cement (Horne, 1998), therefore CCP is regarded as cross-linking agent that can hold micellar network together (Holt, 1992). Another role of CCP as stabilizer that contributes micellar structure is neutralization of phosphoserine residues of casein by being positively charged (Schmidt, 1982). By binding to negatively charged phosphoseryl motifs, hydrophobic patches of casein molecules are allowed to interact each other due to effective reduction in electrostatic repulsion (Horne, 1998).

1.2.8 Calcium chelating agents

Most important minerals that consist of dairy system are calcium and phosphate.

Chelators refer to materials that form complexes with cations using high affinities, and therefore, can displace them (Gaucheron, 2005). Some calcium chelators

(e.g., phosphates, citrate salts, EDTA, and oxalates) can be used to enhance thermal stability of milk leading to improved shelf-life and prevent fouling of equipment surface during milk processing (Holt, 1985; Gaucheron, 2005). Casein micelles are dissociated into small clusters and release specific caseins by mechanism of chelators that govern protein-mineral equilibria, resulting in reduced diffusible calcium and depletion of CCP (de Kort et al., 2011). Therefore, decreased turbidity after the addition of calcium chelating agents has been well reported (Mizuno & Lucey, 2005; Pitkowski, Nicolai, & Durand, 2008; Culler,

Saricay, & Harte, 2017). One of the most well-known calcium chelators is EDTA.

Lin et al. (1972) reported that a small amount of EDTA addition might dissociate

soluble caseins such as β- and κ-caseins by chelating easily accessible calcium ion bridges, whereas a higher concentration of EDTA is required to dissociate

CCP and micellar structure of casein. One interesting point in their study is that hydrodynamic radii remained unchanged after the addition of EDTA. The amount of dissociated casein after EDTA addition was not dependent of pre-heating conditions of skim milk (72 °C for 30 s or 85 °C for 30 min) (Ward et al., 1997).

The extent of micellar structure modification by addition of chelating agents is dependent on calcium-binding ability of chelators (de Kort et al., 2009). It is known that the structure of chelators is highly related to their calcium-binding capacity, for instance, polyphosphates such as sodium hexametaphosphate can display higher chelating ability than orthophosphates such as disodium hydrogen phosphate (de Kort et al., 2009).

1.3 Hypotheses and objectives

The equilibrium of calcium (colloidal calcium vs soluble calcium) in milk and optimized hydrophobic interactions between casein molecules are dependent on environmental factors such as temperature and pH. There should be optimal derivatization conditions that can produce the most translucent skim milk dispersions by disrupting hydrophobic interaction and dissolving CCP from casein micelles. Heating might increase turbidity of SMP dispersions at neutral pH because thiol-disulfide exchange reaction is expected between whey proteins and κ-caseins while heating the samples. However, acidification might reduce the

turbidity of dispersions mainly through dissolution of CCP from casein micelles to serum phase of milk. The dependence of hydrophobic interaction on the temperature has been one of the puzzles in studies on casein micelle. It is expected that heat might affect hydrophobic interactions within casein micelles in a different way when condition of dispersions is not stabilized (e.g., acidified dispersions). Hence, the first hypothesis of the thesis is that acidification can reduce turbidity of SMP dispersion effectively and subsequent heating might break hydrophobic interaction between casein molecules.

Various calcium chelators can chelate calcium from CCP causing disruption of

CCP and subsequent dispersion of casein micelles. The dissociation extent of casein micelles depends on the type of calcium chelators used. Many different chelating agents (e.g., sodium hexametaphosphate, trisodium citrate, tetrasodium pyrophosphates, etc.) have been studied for their effects on casein micellar structure in dairy systems (Udabage, McKinnon, & Augustin, 2000;

Mizuno & Lucey, 2005; Mizuno & Lucey, 2007; de Kort et al., 2009; de Kort et al.,

2011). However, some undesirable aggregation or precipitation is observed before or after heat treatment of SMP dispersions containing some of these chelating agents. Sodium tripolyphosphate is the calcium chelator commonly used in meat processes (Jayasingh & Cornforth, 2003; Soeda et al., 2003) but has rarely been used in milk systems. Owing to polyphosphate structure, it is likely that sodium tripolyphosphates chelate more calcium than other chelating agents while keeping other parameters constant. Thus, the second hypothesis is that sodium tripolyphosphate can form complexes with calcium, which is stable

under the heat treatment; and each calcium chelator interacts with CCP and/or casein differently by chelating a different amount of calcium.

No studies have examined how the structure and physicochemical properties of casein micelles are changed when milk is heated at acidic pH below its isoelectric point though some papers studied the effect of heating at a narrow range of neutral pH (Anema & Li, 2003). Also, the structure of casein micelles after the addition of calcium chelator has barely been studied. The overall objectives of this study were 1) to characterize physicochemical properties of

SMP dispersions acidified by various acids after heating and 2) to characterize physical and chemical properties of neutral SMP dispersions supplemented with different amounts and types of calcium chelators. The goal of these treatments was to reduce the turbidity of SMP dispersions that can be shelf-stable. The fundamental information was studied to understand macroscopic changes of

SMP dispersions using turbidity, dynamic light scattering, fluorescence spectrometry, inductively coupled argon plasma-optical emission spectrometry, scanning transmission electron microscopy, transmission electron microscopy, zeta-potential, and SDS-PAGE.

References

Ahmad, S., Piot, M., Rousseau, F., Grongnet, J. F., & Gaucheron, F. (2009).

Physico-chemical changes in casein micelles of buffalo and cow milks as a

function of alkalinisation. Dairy Sci Technol, 89, 387-403.

Alting, A. C., de Jongh, H. H. J., Visschers, R. W., & Simons, J. F. A. (2002).

Physical and chemical interactions in cold gelation of food proteins. J Agric

Food Chem, 50, 4682-4689.

Alting, A. C., Hamer, R. J., de Kruif, C. G., & Visschers, R. W. (2000). Formation

of disulfide bonds in acid-induced gels of preheated whey protein isolate. J

Agric Food Chem, 48, 5001-5007.

Anema, S. G., & Li, Y. (2003). Effect of pH on the association of denatured whey

proteins with casein micelles in heated reconstituted skim milk. J Agric Food

Chem, 51, 1640-1646.

Anema, S. G., Lowe, E. K., & Stockmann, R. (2005). Particle size changes and

casein solubilization in high-pressure-treated skim milk. Food Hydrocolloids,

19, 257-267.

Aoki, T., Kawahara, A., Kako, Y., & Imamura, T. (1987). Role of individual milk

salt constituents in cross-linking by colloidal calcium phosphate in artificial

casein micelles. Agric Biol Chem, 51, 817-821.

Aoki, T., Umeda, T., & Kako, Y. (1992). The least number of phosphate groups

for crosslinking of casein by colloidal calcium phosphate. J Dairy Sci, 75, 971-

975.

Aoki, T., Yamada, N., Kako, Y., & Imamura, T. (1988). Dissociation during

dialysis of casein aggregates cross-linked by colloidal calcium phosphate in

bovine casein micelles. J Dairy Res, 55, 189-195.

Beau, M. (1921). Les matieres albuminoide du lait. Le Lait, 1,16-26.

Berridge, N. J. (1954). Rennin and the clotting of milk. Adv Enzymol, 15, 423-

448.

Boland, M. J., Rae, A. N., Vereijken, J. M., Meuwissen, M. P. M., Fisher, A. R. H.,

van Boekel, M. A. J. S., Rutherfurd, S. M., Gruppen, H., Moughan, P. J., &

Hendriks, W. H. (2013). The future supply of animal-derived protein for human

consumption. Trends Food Sci Technol, 29, 62-73.

Bravo, F. I., Felipe, X., Lopez-Fandino, R., & Molina, E. (2013). High-pressure

treatment of milk in industrial and pilot-scale equipments: effect of the

treatment conditions on the protein distribution in different milk fractions. Eur

Food Res Technol, 236, 499-506.

Broyard, C., & Gaucheron, F. (2015). Modifications of structures and functions of

caseins: A scientific and technological challenge. Dairy Sci Technol, 95, 831-

862.

Cho, Y., Singh, H., & Creamer, L. K. (2003). Heat-induced interactions of β-

lactoglobulin A and κ-casein B in a model system. J Dairy Res, 70, 61-71.

Choi, J., Horne, D. S., & Lucey, J. A. (2011). Determination of molecular weight

of a purified fraction of colloidal calcium phosphate derived from the casein

micelles of bovine milk. J Dairy Sci, 94, 3250-3261.

Corredig, M., & Dalgleish, D. G. (1996). Effect of temperature and pH on the

interactions of whey proteins with casein micelles in skim milk. Food Res Int,

29, 49–55.

Corredig, M., & Dalgleish, D. G. (1999). The mechanisms of the heat-induced

interaction of whey proteins with casein micelles in milk. Int Dairy J, 9, 233-

236.

Creamer, L. K., & MacGibbon, A. K. H. (1996). Some recent advances in the

basic chemistry of milk proteins and lipids. Int Dairy J, 6, 539-568.

Culler, M, D., Saricay, Y., & Harte, F. M. (2017). The effect of emulsifying salts

on the turbidity of a diluted milk system with varying pH and protein

concentration. J Dairy Sci 100, 4241-4252.

Dalgleish, D. G. (1998). Casein micelles as colloids: Surface structures and

stabilities. J Dairy Sci, 81, 3013-3018.

Dalgleish, D. G., Pouliot, Y., & Paquin, P. (1987). Studies on the heat stability of

milk: Ⅱ. Association and dissociation of particles and the effect of added urea.

J Dairy Res, 54, 39-49.

Dalgleish, D. G., Spagnuolo, P.A., & Goff, D. (2004). A possible structure of the

casein micelle based on high-resolution field-emission scanning electron

microscopy. Int Dairy J, 14, 1025-1031.

Dalgleish, D. G., van Mourik, L., & Corredig, M. (1997). Heat-induced interactions

of whey proteins and casein micelles with different concentrations of

alphalactalbumin and beta-lactoglobulin. J Agric Food Chem, 45, 4806–4813.

Datta, N., & Deeth, H. C. (1999). High pressure processing of milk and dairy

products. Australian J Dairy Technol, 54, 41-48.

Day, L., Williams, R. P. W., Otter, D., & Augustin, M. A. (2015). Casein

polymorphism heterogeneity influences casein micelle size in milk of individual

cows. J Dairy Sci, 98, 3633-3644. de Kort, E., Minor, M., Snoeren, T., van Hooijdonk, T., & van der Linden, E.

(2009). Calcium-binding capacity of organic and inorganic ortho-and

polyphosphates. Dairy Sci Technol, 89, 283-299. de Kort, E., Minor, M., Snoeren, T., van Hooijdonk, T., & van der Linden, E.

(2011). Effect of calcium chelators on physical changes in casein micelles in

concentrated micellar casein solutions. Int Dairy J, 21, 907-913. de Kruif, C. G., Huppertz, T., Urban, V. S., & Petukhov, A. V. (2012). Casein

micelles and their internal structure. Adv Colloid Interface Sci, 171, 36-52.

Desobry-Banon, S., Richard, F., & Hardy, J. (1994). Study of acid and rennet

coagulation of high pressurized milk. J Dairy Sci, 77, 3267-3274.

Devine, A., Prince, R. L., & Bell, R. (1996). Nutritional effect of calcium

supplementation by skim milk powder or calcium tablets on total nutrient intake

in postmenopausal women. Am J Clin Nutr, 64, 731-737.

Dickinson, E. (1992). An introduction to food colloids. 1th ed. Oxford, UK: Oxford

Science Publications.

Donovan, M., & Mulvihill, D. M. (1987). Thermal denaturation and aggregation of

whey proteins. Irish J Food Sci Tec, 11, 87-100.

Eller, L. K., & Reimer, R. A. (2010). A high calcium, skim milk powder diet results

in a lower fat mass in male, energy-restricted, obese rats more than a low

calcium, casein, or soy protein diet. J Nutr, 140, 1234-1241.

FAO. (2006). World agriculture: Towards 2030/2050. Interim Report, Rome: Food

and Agriculture Organisation.

Fox, P. F., & Brodkorb, A. (2008). The casein micelle: Historical aspects, current

concepts and significance. Int Dairy J, 18, 677-684.

Gad, A. S., Emam, W. H., Mohamed, G. F., & Sayd, A. F. (2013). Utilization

whey in production of functional healthy beverage “whey-mango beverages”.

Am J Food Technol, 8, 133-148.

Gaucheron, F. (2005). The minerals of milk. Reprod Nutr Dev, 45, 473-483.

Gaucheron, F., Famelart, M. H., Mariette, F., Raulot, K., Michel, F., & Le Graet,

Y. (1997). Combined effects of temperature and high-pressure treatments on

physicochemical characteristics of skim milk. Food Chem, 59, 439-447.

Graet, Y. L., & Gaucheron, F. (1999). pH-Induced solubilization of minerals from

casein micelles: Influence of casein concentration and ionic strength. J Dairy

Res, 66, 215-224.

Hammersten, O. (1883). Zur Frage, ob das Casein ein Einheitlicher Stoff sei.

Zeitschrift fur Physiologische Chemie, 7, 227-273.

Hankinson, C.L., & Palmer, L. S. (1943). Rennin action in relation to the water

binding and electrokinetic properties of calcium and sodium caseinate sols. J

Dairy Sci, 26, 1043-1056.

Havea, P., Singh, H., & Creamer, L. K. (2001). Characterization of heat-induced

aggregates of β-lactoglobulin, α-lactalbumin and bovine serum albumin in a

whey protein concentrate environment. J Dairy Res, 68, 483-497.

Hill, A. R. (1989). The β-lactoglobulin-κ-casein complex. Can Inst Food Sci

Technol J, 22, 120-123.

Holt, C. (1985). The milk salts: Their secretion, concentrations and physical

chemistry. Developments in dairy chemistry - 3. Lactose and minor

constituents, ed. P. F. Fox. Elsevier Applied Science Publishers, London, UK,

pp. 143-181.

Holt, C. (1992). Structure and stability of bovine casein micelles. Advances in

protein chemistry, vol. 43, eds. C. B. Anfinsen, J. D. Edsall, F. R. Richards, &

D. S. Eisenberg. Academic Press, San Diego, pp. 63-151.

Holt, C. (1998). Casein micelle substructure and calcium phosphate interactions

studied by sephacryl column chromatography. J Dairy Sci, 81, 2994-3003.

Holt, C. (2004). An equilibrium thermodynamic model of the sequestration of

calcium phosphate by casein micelles and its application to the calculation of

the partition of salts in milk. Eur Biophys J, 33, 421-434.

Holt, C., Davies, D. T., & Law, A. J. R. (1986). Effects of colloidal calcium

phosphate content and free calcium ion concentration in the milk serum on the

dissociation of bovine casein micelles. J Dairy Res, 53, 557-572.

Holt, C., Timmins, P. A., Errington, N., & Leaver, J. (1998). A core-shell model of

calcium phosphate nanoclusters stabilized by β-casein phosphopeptides,

derived from sedimentation equilibrium and small angle X-ray and neutron

scattering measurements. Eur J Biochem, 252, 73-78.

Hoppe, C., Anderson, G. S., Jacobsen, S., Molgaard, C., Friis, H., Sangild, P. T.,

& Michaelsen, K. F. (2008). The use of whey or skimmed milk powder in

fortified blended foods for vulnerable groups. J Nutr, 138, 145S-161S.

Horne, D. S. (1998). Casein interactions: casting light on the black boxes, the

structure in dairy products. Int Dairy J, 8, 171-177.

Horne, D. S. (2006). Casein micelle structure: Models and muddles. Curr Opin

Colloid Interface Sci, 11, 148-153.

Horne, D. S., Lucey, J. A., & Choi, J. (2007). Casein interactions: does the

chemistry really matter? In Dickinson, E., Leser, M. (eds) Food Colloid: Self-

assembly and Material Science. RSC Publishing, Cambridge, UK.

Huppertz, T. (2016). Heat stability of milk. In McSweeney, P., O’Mahony, J.

(eds): Advanced Dairy Chemistry. Springer, New York, NY.

Huppertz, T., Fox, P. F., & Kelly, A. L. (2004). High pressure treatment of bovine

milk: effects on casein micelles and whey proteins. J Dairy Res, 71, 97-106.

Huppertz, T., & Kelly, A. L. (2002). Effects of high pressure on constituents and

properties of milk. Int Dairy J, 12, 561-572.

Huppertz, T., & Kruif, D. (2007). Disruption and reassociation of casein micelles

during high pressure treatment:Influence of whey proteins. J Dairy Res, 74,

194-197.

Huppertz, T., Smiddy, M. A., & de Kruif, C. G. (2007). Biocompatible micro-gel

particles from cross-linked casein micelles. Biomacromolecules, 8, 1300-1305.

Huth, P. J., DiRienzo, D. B., & Miller, G. D. (2006). Major scientific advances with

dairy foods in nutrition and health. J Dairy Sci, 89, 1207-1221.

Jang, H. D., & Swaisgood, H. E. (1990). Disulphide bond formation between

thermally denatured β-lactoglobulin and κ-casein in casein micelles. J Dairy

Sci, 73, 900-904.

Johnston, D. E., Austin, B. A., & Murphy, R. J. (1992). Effects of high pressure on

milk. Milchwissenschaft, 47, 760-763.

Josse, A. R., Atkinson, S. A., Tarnopolsky, M. A., & Phillips, S. M. (2011).

Increased consumption of dairy foods and protein during diet- and exercise-

induced weight loss promotes fat mass loss and lean mass gain in overweight

and obese premenopausal women. J Nutr, 141, 1626-1634.

Kaliappan, S., & Lucey, J. A. (2011). Influence of mixtures of calcium-chelating

salts on the physicochemical properties of casein micelles. J Dairy Sci, 94,

4255-4263.

Koutina, G., Knudsen, J. C., Andersen, U., & Skibsted, L. H. (2014). Temperature

effect on calcium and phosphorous equilibria in relation to gel formation during

acidification of skim milk. Int Dairy J, 36, 65-73.

Lazzaro, F., Saint-Jalmes, A., Violleau, F., Lopez, C., Gaucher-Delmas, M.,

Madec, M., Beaucher, E., & Gaucheron, F. (2017). Gradual disaggregation of

the casein micelle improves its emulsifying capacity and decreases the stability

of dairy emulsions. Food Hydrocolloids, 63, 189-200.

Lin, S. H. C., Leong, S. L., Dewan, R. K., Bloomfield, V. A., & Morr, C. V. (1972).

Effect of calcium ion on the structure of native bovine casein micelles.

Biochemistry, 11, 1818-1821.

Liu, Y., & Guo, R. (2008). pH-Dependent structures and properties of casein

micelles. Biophys Chem, 136, 67-73.

Livney, Y. D., Corredig, M., & Dalgleish, D. G. (2003). Influence of thermal

processing on the properties of dairy colloids. Curr Opin Colloid Interface Sci,

8, 359-364.

Loenneke, J. P., Loprinzi, P. D., Murphy, C. H., & Phillips, S. M. (2016). Per meal

dose and frequency of protein consumption is associated with lean mass and

muscle performance. Clin Nutr, 35, 1506-1511.

Lucey, J. A., Johnson, M. E., & Horne, D. S. (2003). Perspectives on the basis of

the rheology and texture properties of cheese. J Dairy Sci, 86, 2725-2743.

Marchin, S., Putaux, J., Pignon, F., & Leionil J. (2007). Effects of the

environmental factors on the casein micelle structure studied by cryo

transmission electron microscopy and small-angle x-ray scattering/ultrasmall-

angle x-ray scattering. J Chem Phys, 126, 045101.

Marques, M., Borreguero, J. M., Stanley, H. E., & Dokholyan, N. V. (2003).

Possible mechanism for cold denaturation of proteins at high pressure. Phys

Rev Lett, 91, 138103.

McGann, T. C. A., Buchheim, W., Kearney, R. D., & Richardson, T. (1983).

Composition and ultrastructure of calcium phosphate-citrate complexes in

bovine milk systems. Biochim Biophys Acta, 760, 415-420.

McMahon, D., & McManus, W. R. (1998). Rethinking casein micelle structure

using electron microscopy. J Dairy Sci, 81, 2985-2993.

Mekmene, O., Graet, Y. L., & Gaucheron, F. (2009). A model for predicting salt

equilibria in milk and mineral-enriched milks. Food Chem, 116, 233-239.

Mellander, O. (1939). Elektrophophoretische Unterschung von Casein.

Biochemische Zeitschrift, 300, 240-245.

Millard-Stafford, M., Warren, G. L., Thomas, L. M., Doyle, J. A., Snow, T., &

Hitchcock, K. (2005). Recovery from run training: Efficacy of a carbohydrate-

protein beverage?. Int J Sport Nutr Exerc Metab, 15, 610-624.

Mitchell, C. J., McGregor, R. A., D’Souza, R. F., Thorstensen, E. B., Markworth,

J. F., Fanning, A. C., Poppitt, S. D., & Cameron-Smith, D. (2015).

Consumption of milk protein or whey protein results in a similar increase in

muscle protein synthesis in middle aged man. Nutrients, 7, 8685-8699.

Mizuno, R., & Lucey, J. A. (2005). Effects of emulsifying salts on the turbidity and

calcium-phosphate-protein interactions in casein micelles. J Dairy Sci, 88,

3070-3078.

Morr, C. V. (1967). Effect of oxalate and urea upon ultracentrifugation properties

of raw and heated skim milk casein micelles. J Dairy Sci, 50, 1744-1751.

Mozhaev, V. V., Heremans, K., Frank, J., Masson, P., & Balny, C. (1996). High

pressure effects on protein structure and function. Proteins: Struct Funct

Genet, 24, 81-91.

Muller-Buschbaum, P., Gebhardt, R., Roth, S. V., Metwalli, E., & Doster, W.

(2007). Effect of calcium concentration on the structure of casein micelles in

thin films. Biophys J, 93, 960-968.

Mulvihill, D. M., Donovan, M. (1987). Whey proteins and their thermal

denaturation-a review. Ir J Food Sci Technol, 11, 43-75.

Nandel, F. S., Verma, R., Singh, B., & Jain, D. V. (1998). Mechanism of hydration

of urea and guanidinum ion: A model study of denaturation of proteins. Pure

Appl Chem, 70, 659-664.

Needs, E. C., Capellas, M., Bland, A. P., Manoj, P., Macdougal, D., & Paul, G.

(2000). Comparison of heat and pressure treatments of skim milk, fortified with

whey protein concentrate, for set yogurt preparation: effects on milk proteins

and gel structure. J Dairy Res, 67, 329-348.

O’Connell, J. E., & Flynn, C. (2007). The manufacture and applications of casein-

derived ingredients. Handbook of food products manufacturing, 557-591.

Oldfield, D. J., Singh H., Taylor, M. W., Pearce, K. N. (2000). Heat-induced

interactions of β-lactoglobulin and α-lactalbumin with the casein micelle in pH-

adjusted skim milk. Int Dairy J, 10, 509-518.

O’Mahony, J. A., Lucey, J. A., & McSweeney, P. L. H. (2005). Chymosin-

mediated proteolysis, calcium solubilization, and texture development during

the ripening of Cheddar cheese. J Dairy Sci, 88, 3101-3114.

Ono, T., Furuyama, T., & Odagiri, S. (1981). Dissociation of large and small

bovine casein micelles by dialysis. Agric Biol Chem, 45, 511-512.

Orlien, V., Boserup, L., & Olsen, K. (2010). Casein micelle dissociation in skim

milk during high-pressure treatment: Effects of pressure, pH, and temperature.

J Dairy Sci, 93, 12-18.

Otte, J., Zakora, M., & Qvist, K. B. (2000). Involvement of disulfide bonds in

bovine β-lactoglobulin B gels set thermally at various pH. J Food Sci, 65, 384-

389.

Paddon-Jones, D., Campbell, W. W., Jacques, P. F., Kritchevsky, S. B., Moore,

L. L., Rodriguez, N. R. & van Loon, L. J. C. (2015). Protein and healthy aging.

Am J Clin Nutr, 101, 1339S-1345S.

Pan, K., & Zhong, Q. (2013). Improving clarity and stability of skim milk powder

dispersions by dissociation of casein micelles at pH 11.0 and acidification with

citric acid. J Agri Food Chem, 61, 9260-9268.

Panouille, M., Benyahia, L., Durand, D., & Nicolai, T. (2005). Dynamic

mechanical properties of suspensions of micellar casein particles. J Colloid

Interface Sci, 287, 468-475.

Paquin, P., Britten, M., Laliberte, M. F., & Boulet, M. (1987). Interfacial properties

of milk casein proteins. ACS symposium Series; American Chemical Society,

677-686.

Pitkowski, A., Nicolai, T., & Durand, D. (2008). Scattering and turbidity study of

the dissociation of casein by calcium chelation. Biomacromolecules, 9, 369-

375.

Putnam, J., & Gerrior, S. (1999). Trends in the US food supply, 1970-97.

America’s eating habits: Changes and consequences, 133-160.

Pyne, G. T. (1955). The chemistry of casein. D Sci Abstracts, 17, :532-553.

Richardson, G. A., & Palmer, L. G. (1929). Rennin action in relation to

electrokinetic phenomena. J Phys Chem, 33, 557-576.

Rose, D. (1969). A proposed model of micelle structure in bovine milk. Dairy Sci

Abstr, 31, 171-175.

Ryan, K. N., Zhong, Q., & Foegeding, E. A. (2013). Use of whey protein soluble

aggregates for thermal stability-A hypothesis paper. J Food Sci, 78, R1105-

R1115.

Sandra, S., & Dalgleish, D. G. (2005). Effects of ultra-high-pressure

homogenization and heating on structural properties of casein micelles in

reconstituted skim milk powder. Int Dairy J, 15, 1095-1104.

Seifert, J., Harmon, J., & DeClercq, P. (2006). Protein added to a sports drink

improves fluid retention. Int J Sport Nutr Exerc Metab, 16, 420-429.

Schmidt, D. G. (1980). Colloidal aspects of the caseins. Neth Milk Dairy J, 34,

42-64.

Schmidt, D. G. (1982). Association of caseins and casein micelle structure. Dev

Dairy Chem, 1, 61-86.

Schrader, K., Buchheim, W., & Morr, C. V. (1997). High pressure effects on the

colloidal calcium phosphate and the structural integrity of micellar casein in

milk. Nahrung, 41, 133-138.

Sharma, M., Haque, Z. U., & Wilson, W. W. (1996). Association tendency of β-

lactoglobulin AB purified by gel permeation chromatography as determined by

dynamic light scattering under quiescent conditions. Food Hydrocolloid, 10,

323-328.

Shiby, V. K., Radhakrishna, K., & Bawa, A. S. (2013). Development of whey-fruit-

based energy drink mixes using D-optimal mixture design. Int J Food Sci

Technol, 48, 742-48.

Singh, S., Singh, A. K., & Patil, G. R. (2001). Whey utilization for health

beverages. Indian Food Industry, 21, 38-41.

Slattery, C. W. (1976). Casein micelle structure; An examination of models. J

Dairy Sci, 59, 1547-1556.

Slattery, C. W., & Evard, R. (1973). A model for the formation and structure of

casein micelles from subunits of variable composition. Biochim Biophys Acta,

317, 529-538.

Smits, P., van Brouwershaven, J. H. (1980). Heat-induced association of β-

lactoglobulin and casein micelles. J Dairy Res, 47, 313-325.

Storr, M., Sattler, D., Hahn, A., Schusdziarra, V., & Allescher, H-D. (2003).

Endogenous CCK depresses contractile activity within the ascending

myenteric reflex pathway of rat ileum. Neuropharmacology, 44, 524-532.

Tran Le, T., El-Bakry, M., Neirynck, N., Bogus, M., Dinh Hoa, H., & Van der

Meeren, P. (2007). Hydrophilic lecithins protect milk proteins against heat-

induced aggregation. Colloids Surf B, 60, 167-173.

Tran Le, T., Saveyn, P., Dinh Hoa, H., & Van der Meeren, P. (2008).

Determination of heat-induced effects on the particle size distribution of casein

micelles by dynamic light scattering and nanoparticle tracking analysis. Int

Dairy J, 18, 1090-1096.

Vaia, B., Smiddy, M. A., Kelly, A. L, & Huppertz, T. (2006). Solvent-mediated

disruption of bovine casein micelles at alkaline pH. J Agric Food Chem, 54,

8288-8293.

Vasbinder, A. J., & de Kruif, C. G. (2003). Casein-whey protein interactions in

heated milk: the influence of pH. Int Dairy J, 13, 669-677.

Wallqvist, A., Covell, D. G., & Thirumalai, D. (1998). Hydrophobic interactions in

aqueous urea solutions with implications for the mechanism of protein

denaturation. J Am Chem Soc, 120, 427-428.

Walstra, P. (1999). Casein sub-micelles: Do they exist? Int Dairy J, 9, 189-192.

Walstra, P., Geurts, T. J., Noomen, A., Jellema, A., & van Boekel, M. A. J. S.

(1999). Dairy technology: Principles of milk properties and processes. 1th ed.

New York, NY: Marcel Dekker Inc.

Ward, B. R., Goddard, S. J., Augustin, M-A., & McKinnon, I. R. (1997). EDTA-

induced dissociation of casein micelles and its effect on foaming properties of

milk. J Dairy Res, 64, 495-504.

Waugh, D. F. (1958). The interaction of ɑs, β- and κ-caseins in micelle formation.

Faraday Society Discussion, 25, 186-192.pH

Waugh, D. F. (1961). Casein interactions and micelle formation. J Phys Chem,

65, 1793-1797.

Waugh, D. F., & Noble, R. W. (1965). Casein micelles. Formation and structure

Ⅱ. J Am Chem Soc, 87, 2246-2257.

Waugh, D. F., & von Hippel, P. H. (1956). κ-Casein and the stabilization of casein

micelles. J Am Chem Soc, 78, 4576-4582.

Zimet, P., Rosenberg, D., & Livney, Y. D. (2011). Re-assembled casein micelles

and casein nanoparticles as nano-vehicles for ω-3 polyunsaturated fatty acids.

Food Hydrocolloids, 25, 1270-1276.

CHAPTER 2

Reducing turbidity of skim milk powder dispersions by dissolving colloidal calcium phosphate from casein micelles and

weakening hydrophobic interaction through acidification and

subsequent heating

2.1 Abstract

Turbid skim milk powder (SMP) dispersions can be clarified when casein micelles are disrupted. In this chapter, dispersions with 5% w/v SMP were acidified to pH

2.4-3.0 with citric acid or glucono delta-lactone (GDL) and subsequently heated at 60-90 °C for 2-60 min. Improved clarity of SMP dispersions and decreased dimension of casein micelles were observed using turbidimeter and dynamic light scattering. Acidification increased dispersibility of casein by solubilizing colloidal

(CCP) calcium phosphate in casein micelles. Decreased hydrophobic interactions between caseins were observed with fluorescence spectroscopy.

The SMP dispersions with reduced size of casein micelles may be used to launch new types of beverage.

2.2 Materials and methods

2.2.1 Chemicals

Carnation SMP (Nestle, Solon, OH) was used with ~34.8% protein and ~1.30% calcium according to the nutrition labeling from the manufacturer. Citric acid was purchased from Fisher Scientific (Fair Lawn, NJ) and GDL was purchased from

MP Biomedicals (Solon, OH). Micellar casein (MC) was from American Casein

Company (Burlington, NJ) and whey protein isolate (WPI) was from Hilmar

Ingredients (Hilmar, CA).

2.2.2 Sample preparation

For preparation of SMP dispersion, SMP was hydrated at 5% w/v (0.500 g in

10.0 mL deionized water in 20 mL scintillation vials). After vortexing for 10 s, samples were incubated at room temperature (RT, ~21 °C) for at least 8 h for complete hydration. Citric acid and GDL were used as acidulants because these acids are commonly used to prepare edible beverages (Soccol et al., 2006) or other dairy products (Lucey et al., 1998). Citric acid (2.0 M) was dropwise added and GDL powder was added to the aqueous SMP dispersions to reach pH 2.40,

2.70, and 3.0 ± 0.02. The amount of GDL was determined in preliminary experiments. The pH of SMP dispersions was monitored using Accumet AE150 pH meter (Thermo Orion INC, Chelmsford, MA) at RT. pH of SMP dispersions with citric acid was changed instantaneously, whereas SMP dispersions with

GDL required a time to change the pH because GDL is hydrolyzed first to gluconic acid (de Kruif, 1997) to decrease pH progressively (Eshpari, Tong, &

Corredig, 2014). The SMP dispersion at pH 6.80 without acidification was used as a control. For heating samples, TC-102D circulating water bath (AMETEK

Brookfield, Middleborough, MA) was used. The heating durations were 2, 5, 10,

30, and 60 min and heating temperatures were 60, 70, 80, and 90 °C. Thermal profiles of samples are shown in Fig. 2.1. To study the behavior of MC dispersion under the same treatment, MC dispersions were also prepared with and without

WPI simulating the proximate mass ratio of casein and whey protein (4:1) in bovine milk (Zadow, 1994; Smithers et al., 1996).

2.2.3 Turbidity

Turbidity of SMP dispersions was measured using 2020 we/wi turbidimeter

(LaMotte, Chestertown, MD). Calibration was performed using AMCO primary turbidity standard (10 NTU). All the measurement was done at RT.

2.2.4 Hydrodynamic diameter and particle size distribution

Dynamic light scattering (DLS) has been used to measure the particle size distribution of casein micelles because the size range of casein micelle is within 5 nm – 5 μm, the measurement range of DLS (Tran Le et al., 2008). In this study,

Zetasizer nano-ZS (Malvern Instrument, Enigma Business Park, UK) was used for determination of particle size distribution by photon correlation spectroscopy.

The mean hydrodynamic diameter was calculated using Zetasizer Software 7.11.

One milliliter of sample was placed in a PS cuvette and the measurement was done with 173 ° backscatter angle and 25 °C of equilibrium temperature.

2.2.5 Fluorescence spectrometry

Steady state fluorescence spectra were measured using a LS 55 fluorescence spectrometer (Perkin Elmer, Waltham, MA) to study effects of acidification and/or heating on the structure of proteins in SMP dispersions. For SMP dispersions without any fluorescent probe, the excitation wavelength was 295 nm and the range of emission spectra was between 300 and 500 nm. Pyrene and 8- anilinonaphthalene-1-sulfonic acid (ANS) were used as probes binding to hydrophobic sites of proteins. Pyrene, a polycyclic aromatic hydrocarbon, shows

vibronic band in fluorescence spectra due to its significant fine structure

(Kalyanasundaram & Thomas, 1977). The excitation wavelength was set at 338 nm and 380 nm for SMP dispersions with pyrene and ANS, respectively. The emission spectra were collected from 350 to 600 nm for SMP dispersions with pyrene and from 400 to 560 nm for SMP dispersions with ANS. The concentration of probes used was 1.0×10-6 M for pyrene and 1.0×10-5 M for

ANS. Excitation and emission slit sizes were fixed at 5.0 nm for SMP dispersion without probes and SMP dispersion with pyrene and 2.5 nm for SMP dispersion with ANS. The scan rate for all the measurements was set as 240 nm/min.

2.2.6 Inductively coupled plasma-optical emission spectrometry (ICP-OES)

To investigate the effect of acidification and heating on the dissolution of CCP, acidified SMP dispersions were determined for serum calcium and phosphorous concentrations before and after heating at 90 °C for 2 min. Amicon ultra-15 centrifugal filter unit having an ultracel-10 membrane with a molecular weight cut- off of 10,000 Da (Merck Millipore Ltd, Cork, Ireland) was used to obtain the serum phase of SMP dispersions. Centrifugation was performed at 4,500 g for 3 h using Sorvall RC-5B plus centrifuge (Sorvall, Newtown, CT). Each serum was diluted 100 times with deionized water to reach certain calcium concentration

(i.e., ~5 ppm) prior to analysis. The diluted serum was analyzed using a Spectro

Ciros ICP-OES (Spectro Ciros, Mahwah, NJ). To calibrate the instrument before calcium and phosphorous analysis, ICP-OES standards were used (VHG Labs,

Manchester, NH). The detection limit for calcium and phosphorous in the method was both 0.01 mg/L.

2.2.7 Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-

PAGE)

SMP dispersions without or with heating at 90 °C for 2 min were analyzed under reducing SDS-PAGE. The sample buffer was prepared with β-mercaptoethanol and Laemmli sample buffer with a volume ratio of 5:95. Each SMP dispersion was diluted 20 times with the sample buffer. Samples were heated in boiling water for 5 min. The running buffer for gel electrophoresis was prepared with 100 mL of 10× Tris/Glycine/SDS and 900 mL of deionized water. Ten μL of sample was loaded in each well of Mini-PROTEAN TGX Gels, 4-20% (Bio-Rad

Laboratories, Hercules, CA). The gel was subjected to 400 mA for current, 200 V for voltage for 35 min of running time. After gel electrophoresis, the gel was washed with deionized water three times for 5 min. Then, the gel was stained using Bio-safe Coomassie G-250 Stain for 1 h and then rinsed with deionized water for 30 min.

2.2.8 Determination of free thiol groups (SH) concentration

Free thiol groups (SH) in SMP dispersions before and after acidification and heating were determined using the method of Beveridge et al. (1974) with slight modification. SMP dispersions were adjusted to pH 2.4, 2.7, or 3.0 using citric acid (2.0 M) and heated at 90 °C for 2 min. For control, half of samples were not

heated. Ten μL of SMP dispersions was solubilized in 1.25 mL of Tris-Glycine buffer (0.086 M tris, 0.09 M glycine, 4 mM EDTA, and 8 M urea) and then reacted with 12.6 μL of Ellman’s reagent (10 mM DTNB). Absorbance was measured at

412 nm using Evolution 201 UV-Vis spectrophotometer (Thermo Scientific,

Waltham, MA). The concentration of SH groups was determined using Beer’s law:

Abs = ε l c

Where Abs is the absorbance; ε is the molar extinction coefficient of Ellman’s reagent, 13,600 M-1cm-1; l is the path length, 1 cm; and c is the concentration of

SH groups.

2.2.9 Scanning transmission electron microscopy (STEM)

The morphology of particulates before and after acidification and after heating at

90 °C for 2 min was studied using STEM. About 50 μL of SMP dispersions (1% w/v) was placed on a piece of Parafilm. A freshly glow discharged 400 mesh copper grid with a thin carbon film was placed on the sample drop, film side down. After 1 min, excess sample was quickly removed from the grid and the grid was placed on a drop of water for 10 s. The water was removed using Kimwipes and the grid was placed on a drop of 1% uranyl acetate for 1 min. Then excess stain was quickly removed using a Kimwipes and the sample was allowed to dry at RT. The stained samples were imaged using a Zeiss Auriga (Carl Zeiss,

Thornwood, NY) operating in STEM mode at 30 kV.

2.2.10 Statistical analysis

Two or three replicates were prepared for statistical analysis. One-way analysis of variance (ANOVA) was performed using SAS enterprise guide 7.1 (SAS

Institute Inc., Cary, NC). Mean and standard deviation (SD) were calculated and differences of calculated means were considered significant with P < 0.05 using

Fisher’s least significance difference (LSD) test.

2.3 Results and discussion

2.3.1 Turbidity, hydrodynamic diameter, and particle size distribution of skim milk powder dispersions

2.3.1.1 Effect of heating duration for samples acidified with citric acid

After acidification and heat treatment with various conditions, most SMP dispersions became translucent (Figs. 2.2.a, 2.3.a, 2.4.a, and 2.5.a). All sampled showed little turbidity after heating at 90 °C (Fig. 2.2.a). When samples were heated at 90 °C, longer duration increased both turbidity and hydrodynamic diameter significantly (Fig. 2.2.b). In Figure 2.2.c, it is obvious that longer duration of heating at 90 °C resulted in rightward shift of the mode of each graph.

This is in agreement with the data obtained from turbidity and hydrodynamic diameter (Fig. 2.2.b). Also, as heating time was increased the particles with bigger size of range (1000-10000 nm) were observed. This might be, supposedly, due to more amounts of denatured whey protein at longer heating duration causing more interactions between casein micelles and whey protein.

However, casein micelles among themselves without whey protein also showed a little aggregation at high temperature (Fig. 2.10).

2.3.1.2 Effect of heating temperature for samples acidified with citric acid

When the heating duration was 10 min, samples heated at 60 °C were more turbid than other samples (Fig. 2.3.a). Samples subjected to heat for 10 min under different temperature did not show any linear function between heating temperature and turbidity/hydrodynamic diameter; however, 70 °C produced the dispersion with the lowest turbidity and the lowest hydrodynamic diameter (Fig.

2.3.b). A previous study observed that turbidity of skim milk could decrease at the initial stage (~120 s) during heating at 60 °C (Jeurnink, 1992). This might explain why heating at 70 °C could decrease turbidity and hydrodynamic diameter significantly after short period of heating (from 2 min to 5 min), however, 70 °C was not enough to denature whey protein components such as β-lg or α-la so there was no significant increase in both turbidity and hydrodynamic diameter

(Figs. 2.3.b and 2.7). When there was no whey protein in MC dispersion, aggregation occurred slightly after heating at 70 °C (Fig. 2.10.a) whereas 90 °C of heating showed marked increase in the particle size range of 1000-10000 nm

(Fig. 2.10.b). In Figure 2.3.c, samples that was subjected 60 °C of heat treatment had bigger particles while new peak in the left-side was emerged in the samples with 70 °C. This result could be used to explain why samples with 60 °C heat treatment were turbid (Fig. 2.3.a) whereas samples with 70 °C had the lowest turbidity (Fig. 2.3.b).

2.3.1.3 Effect of pH for samples acidified with citric acid or GDL

Samples acidified with citric acid were visually more turbid (Fig. 2.4.a) than samples acidified with GDL to the same pH (Fig. 2.5.a). No matter which acidulant was used to decrease the pH of SMP dispersions, samples with lower pH had less turbidity with lower hydrodynamic diameter values (Figs. 2.4.b and

2.5.b). Samples with pH 3.0 had obviously larger particles than sample with pH

2.4 and 2.7 (Fig. 2.4.c), which could clarify turbidity differences in Figs. 2.4.a and

2.4.b. However, hydrodynamic diameter in acidified samples with GDL varied relatively little but still showed smaller particles at more acidic pH (Fig. 2.5.c).

After heat treatment of acidified SMP dispersions, small portion of aggregated particle was observed in most samples. Considering results from turbidity, hydrodynamic diameter, and particle size distribution collectively, it appears that how much caseins micelles are dissociated has greater influence on the turbidity than how much casein micelles are aggregated because high amounts of bigger size particles in the range of 1000-10000 nm was not always directly related to increase in turbidity. For example, when GDL was used as acidulant, more aggregated particles were observed in the samples with more acidic pH after heating at 90 °C (Fig. 2.5.c), however, both turbidity and hydrodynamic diameter were dropped with decrease in pH (Fig. 2.5.b).

2.3.2 Relative role of casein and whey protein on SMP dispersions after acidification and heating

Heating at 90 °C confirmed that aggregation occurred with increased heating time through possible formation of casein-whey complex (Fig. 2.7). Anema and Li

(2003) showed that increase in size of casein micelle occurred when more than

80% of denatured whey protein was associated with casein micelles. In their study, amounts of associated whey protein were proportional to the increased size of casein micelle size. As the Fig. 2.8 represents, MC dispersion itself did not show increase in mean hydrodynamic diameter with longer heating duration, which suggests that there was no significant aggregation that can increase turbidity.

When MC and WPI were heated together at 70 °C or 90 °C, hydrodynamic diameter was decreased after heating at both temperatures (Fig. 2.9). This is not in agreement with results from other studies that reported that there was no increase in casein micelle size only when there was no whey protein (Anema &

Li, 2003; Tran Le et al., 2008). Decreased turbidity and hydrodynamic diameter in

MC dispersion in this study is probably because free thiol groups of β-lg in added

WPI were not reactive as that of in SMP, resulting in no significant thiol/disulfide exchange between β-lg and κ-casein.

2.3.3 Hydrophobic interactions between proteins in skim milk powder dispersions

Fluorescence spectroscopy is a useful technique to study microenvironment of polymeric molecules and casein micelles (Sahu, Kasoju, & Bora, 2008; Esmaili et al., 2011). The fluorescence intensity from proteins is mainly due to aromatic amino acid constituents such as tyrosine, phenylalanine, and tryptophan (Konev,

1957) and fluorescence spectra, in which only tryptophan influences, can be obtained under the certain range of excitation wavelength (Teale, 1960). Two tryptophan residues are present in αs1-casein at 164 and 199 positions (Alaimo,

Wickham, & Farrell, 1999) and β-casein contains one tryptophan at 143 position

(Kumosinski, Brown, & Farrell, 1993). Trp 164 is located in hydrophobic portion of casein molecules and highly involved in self-association of αs1-casein (Alaimo,

Wickham, & Farrell, 1999).

The intrinsic fluorescence emission spectra of SMP dispersions are presented in

Fig. 2.11. The intensity of fluorescence spectra and maximum wavelength of Trp residues varied among samples. Compared to the pH 6.8 SMP dispersion control, there was a significant increase in emission intensity after acidification to pH 3.0, and the increase in intensity was reduced as dispersions were acidified to a lower pH of 2.7 and 2.4. At a specific pH, heating at 90 °C for 2 min resulted in a reduction in fluorescence intensity. The wavelength corresponding to the maximum intrinsic fluorescence of tryptophan, λmax, shows high sensitivity depending on the exposure of the chromophore. Particularly, a shift of λmax to longer wavelengths (red shift) occurs when there exist positively charged or

negatively charged molecules near benzene- or pyrrole-ring of tryptophan (Vivian

& Callis, 2011). Compared to samples before heating, small red shift of about 2 nm was observed in both acidified and neutral SMP dispersions after heating

(Table 2.1). Significant red shifts (~10 nm) can occur when buried tryptophan residues are exposed to water; however, this is only allowed when both benzene and pyrrole rings of tryptophan residues are exposed to water (Vivian & Callis,

2011). Thus, the small red shift may be interpreted that heat treatment induced only a minute change in the exposure of tryptophan residues. These exposed residues might be a result of weakened hydrophobic interaction between casein molecules resulting in decreased turbidity in acidified SMP dispersions after heating (Figs. 2.6.a and 2.6.b). As SMP dispersions were acidified from pH 3.0 to

2.7 and 2.4, both λmax and fluorescence intensity decreased (Fig. 2.11 and Table

2.1). This observation could be attributed to two mechanisms: 1) hydrogen bonding and 2) cation-π interaction. First, hydrogen bonding within proteins competes with water. Below the isoelectric point (pI) of casein micelles, amino acids of casein molecules are protonated to a higher extent at a lower pH, and

+ hydrogen ions in amino group, -NH3 , have higher affinity than those in -NH2, causing strong hydrogen bond pairing (Liu & Guo, 2008). Second, cation-π interaction, (i.e., noncovalent attraction between cation and electron-rich π system) is electronically favored as aromatic side chains of tryptophan corresponding to the electron-rich π system can easily interact with protonated nitrogen-containing side chains of arginine or lysine. It has been reported that cation-π interaction usually occurs on the surface of protein, and therefore,

aromatic amino acids are less likely to be exposed to water (~20% of surfaces)

(Gallivan & Dougherty, 2000). These two mechanisms might contribute to the compact structure of casein molecules as well as decreased λmax and fluorescence intensity.

Aqueous solubility of a well-known hydrophobic probe, pyrene, has been reported as 0.135 mg/L (Mackay & Shiu, 1977) and this low solubility promotes solubilization of pyrene in hydrophobic regions that exist in aqueous medium (Liu

& Guo 2008). Being characterized by five vibronic band peaks (i.e., Band Ⅰ, Ⅱ,

Ⅲ, Ⅳ, and Ⅴ) from fluorescence spectra of monomeric pyrene in solvents, the

ratio of Band Ⅲ (~385 nm) to Band Ⅰ (~375 nm) (I3/I1) provides insights into the polarity of local environment of pyrene (Kalyanasundaram & Thomas, 1977). It is known that fluorescence intensity of Band Ⅲ and Band Ⅰ is strengthened in a

hydrophobic and polar environment, respectively (Bains et al., 2012). I3/I1 is a well-acknowledged parameter that indicates polarity of solvent in a homogeneous solvents system (Chakraborty & Basak, 2008). However, pyrene interacts with hydrophobic domains of aggregates when micelles exist in the solvent system with abrupt change in I3/I1 being a parameter for the critical micellar concentration (Kalyanasundaram & Thomas, 1977). Although casein micelles were disrupted after acidification and subsequent heating, I3/I1 for each sample appeared to be quite similar (Figure 2.12 and Table 2.2). It was expected that I3/I1 should have a higher value at acidic pH due to the dissociation of casein

micelles. However, similar I3/I1 was observed in a previous study (Liu & Guo,

2008) regardless of pH of casein micelle dispersions if the concentration of casein was higher than 0.5% (w/v). This means that caseins form micellar structure as long as their concentrations are higher than 0.5% (w/v). Therefore, our results suggest that not all the casein micelles could be dissociated despite weakened hydrophobic interaction between casein molecules and significant dissolution of CCP, which will be explained further.

ANS is another fluorescent probe for study of protein conformations due to its strong binding affinity to hydrophobic regions of protein (Berry & Creamer, 1975;

Semisotnov et a., 1991). Being dissolved in only water, fluorescence intensity of

ANS is quite low with λmax ~520 nm; however, binding of ANS to hydrophobic patches can greatly increase emission intensity while shifting λmax towards shorter wavelength, i.e., blue shift (Stryer, 1965; Turner & Brand, 1968; Liu &

Guo, 2008). The fluorescence emission spectra of SMP dispersions with ANS can be seen in Fig. 2.13. The emission intensity of ANS itself is independent on the pH of solvent (Liu & Guo, 2008). Therefore, the increased intensity after acidification of SMP dispersions can be attributed to increased binding of ANS to hydrophobic residues in casein micelles. It has been reported that ion pairing between sulfonate group in ANS and the cationic groups such as the side chains of arginine, lysine, and histidine affects ANS binding to polypeptides (Matulis &

Lovrien, 1998; Gasymov & Glasgow, 2007). This explains the sharp increase in the fluorescence emission intensity after acidification. Moreover, the intensity was further increased after heating, which is indicative of weakened hydrophobic

interaction within casein micelles, allowing more ANS binding to exposed hydrophobic patches.

2.3.4 Calcium and phosphorous determination from serum phase of skim milk powder dispersions

ICP-OES is an analytical technique that can detect chemical elements in the samples. Each element has its unique emission spectrum when atoms are excited at a high temperature (~10,000 K) and the concentration of the element in the sample can be quantified using the emission intensity (Chakraborty &

Gupta, 2009; Ghosh et al., 2013). Acidification increased (P < 0.05) both calcium and phosphorous concentrations in the serum phase of SMP dispersions, while heating had insignificant effect (P > 0.05) (Fig. 2.14). The results suggest that calcium and phosphorous were dissolved from CCP in the casein micelles after acidification. A possible mechanism is that charged amino acids in casein molecules, such as negatively charged phosphoserine residues, are neutralized during acidification that caused the diffusion of the attached CCP to the serum phase (Jaros et al., 2010). Because the structure of casein micelles cannot be well-established without CCP, these results can explain how the hydrodynamic diameter was reduced in the samples after acidification and subsequent heating, resulting in less turbid sample. Also, citric acid exists as citrate ions at pH 6.8

(pKa1=3.13, pKa2=4.76, and pKa3=6.40). Therefore, it is possible that some calcium ion was chelated by citrate due to metal-chelating property of citrate when citric acid was added to neutral SMP dispersions (Meier-Kriesche et al.,

1999). The insignificant changes in concentrations of calcium and phosphorous after heating support a greater role of hydrophobic interactions in reducing turbidity due to heating.

Decreased calcium and phosphorous concentrations in the serum phase of SMP dispersions were reported after acidification to pH 4.6 - 6.4 using GDL (Koutina &

Skibsted, 2015). It is known that some parts of calcium and all the phosphate of

CCP are solubilized at pH 5.2 and calcium is only completely solubilized at pH

3.5 (Gaucheron, 2005). However, our results suggest that there were still insolubilized calcium and phosphate at pH 3.0. Slight acidification (from pH 6.6 to

6.0) of milk protein concentrate dispersion with GDL before diafiltration also lowered total calcium concentration in the powder after spray drying (1.84 g vs.

1.59 g /100 g of powder) (Eshpari, Tong, & Corredig, 2014), which demonstrated slight acidification could also lead to significant solubilization of calcium from casein micelles.

2.3.5 Effects of acidification and heating on free thiol groups (SH) concentration and molecular weights of proteins

Comparison of molecular weight between acidified samples before and after heating could be analyzed using polyacrylamide electrophoresis (PAGE)

(Shapiro, Vinuela, & Maizel, 1967). As shown in Fig. 2.15, the thickness of band around 35 kDa was slightly reduced after acidification of SMP dispersions. Also, the increased band intensity near 80 kDa was observed after acidification.

Disulfide bonding is the major mechanism governing the formation of whey protein-casein micelle complexes due to sulfhydryl-disulfide exchange (Havea,

Singh, & Creamer, 2001). Therefore, some blocking agents are used to modify thiol groups of whey proteins so that heat-induced aggregation is prohibited

(Alting et al., 2000). In this study, free thiol groups in SMP dispersions were quantified using Ellman’s reagent. The concentration of free thiol groups decreased significantly (P < 0.05) after heating samples at pH 2.7-6.8, while the reduction at pH 2.4 was insignificant (P > 0.05) (Fig. 2.16). This indicates free thiol groups were involved with disulfide bonding between whey proteins by β-lg or between whey proteins and caseins, e.g., β-lg and κ-casein after the heating at pH 2.7-6.8. The insignificant change in free thiol group content at pH 2.4 is expected, as formation of disulfide bonds is inhibited below pH 2.5 due to limited reactivity of free thiol groups at low pH (Alting et al., 2002). Combining all the results above, however, there is still no clear evidence that why a vivid band emerged on the gel corresponding to the ~80 kDa.

2.3.6 Structural change of casein micelles using scanning transmission electron microscopy (STEM)

Being the subject of controversy, sophisticated observation technique has been required as the part of efforts to clarify the structure of casein micelles, which remains elusive (Smith & Campbell, 2007). STEM technique is a valuable tool to characterize nanostructures of materials, where the sample principle of scanning electron microscopy (SEM) applies but transmitted electrons can be obtained by

scanning a sample after focusing the electron beam (Pennycook et al., 2006;

Gee, Carey, & Auty, 2010).

Topographical features of casein micelles in SMP dispersions before and after acidification and subsequent heating are represented in Figs 2.17 and 2.18.

Intact casein micelles are displayed as darker particles compared to dissociated casein micelles with reduced sphericity. This difference in contrast mainly results from difference in the number of stained particles in the sample (McMahon &

Oommen, 2008). Intact casein micelles at pH 6.8 are comprised of a greater number of casein molecules compared to dissociated casein micelles.

Accordingly, a large number of stained casein molecules in intact casein micelles provide them with high electron density causing difference in electron scattering between intact and dissociated casein micelles. The variation in the size of intact casein micelles could be varied from 20 to 600 nm while 200 nm is an average size in the distribution (Bloomfield and Mead, 1975; de Kruif, 1998), and this variation was observed from 20 nm to 350 nm at pH 6.8 without treatment.

2.4 Conclusion

The dimension of casein micelles in 5% w/v SMP dispersions was greatly reduced, enabling the production of dispersions with lower turbidity after acidification and subsequent heating. Lower pH values were preferable to higher ones, allowing smaller hydrodynamic diameter of casein micelles regardless of the types of acidulants used. The increase in calcium and phosphorous

concentrations in the serum phase of SMP dispersion after acidification can attributed to the dissolution of CCP, which resulted in the dissociation of casein micelles. However, the subsequent reduction in turbidity after heating of acidified dispersions resulted due to the weakened hydrophobic interaction between proteins. This simple method may be used to provide protein beverages with an acidic pH.

References

Alaimo, M. H., Wickham, E. D., & Farrell, H. M. (1999). Effect of self-association

of αs1-casein and its cleavage fractions αs1-casein (136-196) and αs1-casein (1-

197), on aromatic circular dichroic spectra: comparison with predicted models.

Biochim Biophys Acta, 1431, 395-409.

Alting, A. C., Hamer, R. J., de Kruif, C. G., & Visschers, R. W. (2000). Formation

of disulfide bonds in acid-induced gels of preheated whey protein isolate. J

Agric Food Chem, 48, 5001-5007.

Alting, A. C., de Jongh, H. H. J., Visschers, R. W., & Simons, J. F. A. (2002).

Physical and chemical interactions in cold gelation of food proteins. J Agric

Food Chem, 50, 4682-4689.

Anema, S. G., & Li, Y. (2003). Effect of pH on the association of denatured whey

proteins with casein micelles in heated reconstituted skim milk. J Agric Food

Chem, 51, 1640-1646.

Bains, G. K., Kim, S. H., Sorin, E. J., & Naraynaswami, V. (2012). Extent of

pyrene excimer fluorescence emission is a reflector of distance and flexibility:

Analysis of the segment linking the LDL receptor-binding and termination

domains of apolipoprotein E3. Biochemistry, 51, 6207-6219.

Berry, G. P., & Creamer, L. K. (1975). The association of bovine β-casein. The

importance of the C-terminal region. Biochemistry, 14, 3542-3545.

Beveridge, T., Toma, S. J., & Nakai, S. (1974). Determination of SH- and SS-

groups in some food proteins using Ellman’s reagent. J Food Sci, 39, 49-51.

Bloomfield, V. A., & Mead, R. J. (1975). Structure and stability of casein micelles.

J Dairy Sci, 58, 592-601.

Chakraborty, A., & Basak, S. (2008). Effect of surfactants on casein structure: A

spectroscopic study. Colloids Surf B, 63, 83-90.

Chakraborty, A., & Gupta, T. (2009). Chemical characterization of submicron

aerosol in Kanpur region: A source apportionment study. Int J Env Ac Eng, 1,

19-27. de Kruif. (1997). Skim milk acidification. J Colloid Interface Sci, 185, 19-25. de Kruif. (1998). Supra-aggregates of casein micelles as a prelude to coagulation. J Dairy Sci, 81, 3019-3028.

Eshpari, H., Tong, P. S., & Corredig, M. (2014). Changes in the physical

properties, solubility, and heat stability of milk protein concentrates prepared

from partially acidified milk. J Dairy Sci, 97, 7394-7401.

Esmaili, M., Ghaffari, S. M., Moosavi-Movahedi, Z., Atri, M. S., Sharifizadeh, A.,

Farhadi, M., Yousefi, R., Chobert, J-M., Haertle, T., & Moosavi-Movahedi, A. A.

(2011). Beta casein-micelle as a nano vehicle for solubility enhancement of

curcumin; food industry application. LWT-Food Sci Technol, 44, 2166-2172.

Gallivan, J. P., & Dougherty, D. A. (2000). A computational study of cation-π

interactions vs salt bridges in aqueous media: Implications for protein

engineering. J Am Chem Soc, 122, 870-874.

Gasymov, O. K., & Glasgow, B. J. (2007). ANS fluorescence: Potential to

augment the identification of the external binding sites of proteins. Biochim

Biophys Acta, 1774, 403-411.

Gaucheron, F. (2005). The minerals of milk. Reprod Nutr Dev, 45, 473-483.

Gee, V. L., Carey, M. A., & Auty, M. A. E. (2010). A new cryo-STEM method for

imaging food colloids in the scanning electron microscope. Scanning Vol, 32,

150-154.

Ghosh, S., Prasanna, V. L., Sowjanya, B., Srivani, P., Alagaraja, M., & Banji, D.

(2013). Inductively coupled plasma-optical emission spectroscopy: A review.

Asian J Pharm Ana, 3, 24-33.

Havea, P., Singh, H., & Creamer, L. K. (2001). Characterization of heat-induced

aggregates of β-lactoglobulin, α-lactalbumin and bovine serum albumin in a

whey protein concentrate environment. J Dairy Res, 68, 483-497.

Jaros, D., Jacob, M., Otto, C., & Rohm, H. (2010). Excessive cross linking of

caseins by microbial transglutaminase and its impact on physical properties of

acidified milk gels. Int Dairy J, 20, 321-327.

Jeurnink, T. J. M. (1992). Change in milk on mild heating: turbidity

measurements, Neth Milk Dairy J, 46, 183-196.

Kalyanasundaram, K., & Thomas, J. K. (1977). Environmental effects on vibronic

band intensities in pyrene monomer fluorescence and their application in

studies of micellar systems. J Am Chem Soc, 99, 2039-2044.

Konev, S. V. (1957). Fluorescence spectra and spectra of action of fluorescence

in some proteins. Dokl Akad Nauk USSR, 116, 594-597.

Koutina, G., & Skibsted, L. H. (2015). Calcium and phosphorous equilibria during

acidification of skim milk at elevated temperature. Int Dairy J, 45, 1-7.

Kumonosinski, T. F., Brown, E. M., & Farrell, H. M. (1993). Three-dimensional

molecular modeling of bovine caseins: An energy-minimized β-casein

structure. J Dairy Sci, 76, 931-945.

Liu, Y., & Guo, R. (2008). pH-Dependent structures and properties of casein

micelles. Biophys Chem, 136, 67-73.

Livney, Y. D., Corredig, M., & Dalgleish, D. G. (2003). Influence of thermal

processing on the properties of dairy colloids. Curr Opin Colloid Interface Sci,

8, 359-364.

Lucey, J. A., Tamehana, M., Singh, H., & Munro, P. A. (1998). A comparison of

the formation, rheological properties and microstructure of acid skim milk gels

made with a bacterial culture or glucono-δ-lactone. Food Res Int, 31, 147-155.

Mackay, D., & Shiu, W. Y. (1977). Aqueous solubility of polynuclear aromatic

hydrocarbons. J Chem Eng Data, 22, 399-402.

Matulis, D., & Lovrien, R. (1998). 1-Anilino-8-naphthalene sulfonate anion-protein

binding depends primarily on ion pair formation. Biophys J, 74, 422-429.

McMahon, D. J., & Oommen, B. S. (2008). Supramolecular structure of the

casein micelle. J Dairy Sci, 91, 1709-1721.

Meier-Kriesche, H. U., Finkel, K. W., Gitomer, J. J., & DuBose, T. D. (1999).

Unexpected severe hypocalcemia during continuous venovenous hemodialysis

with regional citrate anticoagulation. Am J Kidney Dis, 33, e8-1.

Pennycook, S. J., Lupini, A. R., Varela, M., Borisevich, A. Y., Peng, Y., Oxley, M.

P., van Benthem, K., & Chisholm, M. F. (2006). Scanning transmission

electron microscopy for nanostructure characterization. Scanning microscopy

for nanotechnology, Springer, New York, pp. 152-191.

Sahu, A, Kasoju, N, & Bora, U. (2008). Fluorescence study of the curcumin-

casein micelle complexation and its application as a drug nanocarrier to cancer

cells. Biomacromolecules, 9, 2905-2912.

Semisotnov, G. V., Rodionova, N. A., Razgulyaev, O. I., Uversky, V. N., Gripas,

A. F., & Gilmanshin, R. I. (1991). Study of the “molten globule” intermediate

state in protein folding by a hydrophobic fluorescent probe. Biopolymers, 31,

119-128.

Shapiro, A. L., Vinuela, E., & Maizel, J. V. (1967). Molecular weight estimation of

polypeptide chains by electrophoresis in SDS-polyacrylamide gels. Biochem

Biophys Res Commun, 28, 815-820.

Smithers, G. W., Ballard, F. J., Copeland, A. D., de Silva, K. J., Dionysius, D. A.,

Francis, G. L., Goddard, C., Grieve, P. A., McIntosh, G. H., Mitchell, I. R., &

Pearce, R. J. (1996). New opportunities from the isolation and utilization of

whey proteins. J Dairy Sci, 79, 1454-1459.

Soccol, C. R., Vandenberghe, L. P. S., Rodrigues, C., & Pandey, A. (2006). New

perspectives for citric acid production and application. Food Technol

Biotechnol, 44, 141-149.

Stryer, L. (1965). The interaction of a naphthalene dye with apomyoglobin and

apohemoglobin: A fluorescent probe of non-polar binding sites. J Mol Biol, 13,

482-495.

Teale, F. W. J. (1960). The ultraviolet fluorescence of proteins in neutral solution.

Biochem J, 76, 381-388.

Tran Le, T., Saveyn, P., Dinh Hoa, H., & Van der Meeren, P. (2008).

Determination of heat-induced effects on the particle size distribution of casein

micelles by dynamic light scattering and nanoparticle tracking analysis. Int

Dairy J, 18, 1090-1096.

Turner, D. C., & Brand, L. (1968). Quantitative estimation of protein binding site

polarity. Fluorescence of N-arylaminonaphthalenesulfonates. Biochemistry, 7,

3381-3390.

Vivian, J. T., & Callis, P. R. (2001). Mechanisms of tryptophan fluorescence shifts

in proteins. Biophys J, 80, 2093-2109.

Zadow, J. G. (1994). Utilisation of milk components: Whey. In Robinson: Modern

Dairy Technology. Springer, Boston, MA, pp. 313-373.

Appendix

A B

C D

Figure 2.1. The thermal profiles of 10 mL SMP dispersions (5% w/v) in borosilicate glass scintillation vials after heating at 60 °C (a), 70 °C (b), 80 °C (c), or 90 °C for 10 min.

Heating 2 5 10 30 60 A time (min)

B a a b

b a c b

d c c d

Figure 2.2. The appearance (a), turbidity and hydrodynamic diameter (b), and

particle size distributions (c) of SMP dispersions (5% w/v) after acidifying to pH

2.4 with citric acid and heating at 90 °C for 2-60 min. The far-left sample (a) is the

unheated control at pH 6.8. Different letters above symbols indicate significant

difference in the compared parameter (P < 0.05). Error bars are SD (n=3).

A Heating 60°C 70°C 80°C 90°C temperature

a B a b b c a b c d

Figure 2.3. The appearance (a), turbidity and hydrodynamic diameter (b), and

particle size distributions (c) of SMP dispersions (5% w/v) after acidifying to pH

2.4 with citric acid and heating at 60-90 °C for 10 min. The far-left sample (a) is

the unheated control at pH 6.8. Different letters above symbols indicate

significant difference in the compared parameter (P <0.05). Error bars are SD

(n=3).

A pH with pH 2.4 pH 2.7 pH 3.0 citric acid

a B a

b c b c

Figure 2.4. The appearance (a), turbidity and hydrodynamic diameter (b), and

particle size distributions (c) of SMP dispersions (5% w/v) after acidifying to pH

2.4-3.0 with citric acid and heating at 90 °C for 10 min. The far-left sample (a) is

the unheated control at pH 6.8. Different letters above symbols indicate

significant difference in the compared parameter (P < 0.05). Error bars are SD

(n=2).

A pH with pH 2.4 pH 3.0 pH 2.7 pH 2.5 GDL (citric acid)

Figure 2.5. The appearance (a), turbidity and hydrodynamic diameter (b), and

particle size distributions (c) of SMP dispersions (5% w/v) after acidifying to pH

2.5-3.0 with GDL and heating at 90 °C for 10 min. The far-left sample (a) is the

unheated control at pH 6.8.

A pH 2.4 pH 2.4 pH 2.7 pH 2.7 pH 3.0 pH 3.0 (before (after (before (after (before (after heating) heating) heating) heating) heating) heating)

B a

d d d

Figure 2.6. The effect of heating (90 °C for 10 min) on appearance (a) and

turbidity (b) after acidifying to pH 2.4-3.0 with citric acid. The far-right sample (a)

is the unheated control at pH 6.8. Different letters above bars indicate significant

difference in the compared parameter (P < 0.05). Error bars are SD (n=3).

A a

b b b a

b b b

b b a b

Figure 2.7. Turbidity and hydrodynamic diameter of SMP dispersions (5% w/v)

after acidification to pH 2.5 with citric acid and heating at 70 °C (a) or 90 °C (b)

for 2-30 min. Different letters above symbols indicate significant difference in the

compared parameter (P < 0.05). Error bars are SD (n=3).

A a

ab b

a c ab

a a

b c a a

Figure 2.8. Turbidity and hydrodynamic diameter of MC dispersions (1.73% w/v)

after acidification to pH 2.5 with citric acid and heating at 70 °C (a) or 90 °C (b)

for 2-30 min. Different letters above symbols indicate significant difference in the

compared parameter (P < 0.05). Error bars are SD (n=3).

A a

ab b

a c a

B a ab b a a c a

Figure 2.9. Turbidity and hydrodynamic diameter of dispersions with 1.39% MC and 0.39% WPI (w/v) after acidification to pH 2.5 with citric acid and heating at

70 °C (a) or 90 °C (b) for 2-30 min. Different letters above symbols indicate significant difference in the compared parameter (P < 0.05). Error bars are SD

(n=3).

Figure 2.10. Particle size distributions of MC dispersions (1.73% w/v) after

acidification to pH 2.5 with citric acid and heating at 70 °C (a) or 90 °C (b) for 2-

30 min.

Figure 2.11. Intrinsic fluorescence spectra of SMP dispersions (5% w/v) before and after acidification to pH 2.4-3.0 with citric acid, before and after heating at

90 °C for 2 min.

Table 2.1. The wavelength (λmax) of 5% w/v SMP dispersions before and after acidification to pH 2.4-3.0, before and after heating at 90 °C for 2 min.

pH of SMP dispersion λmax (nm)

Before heating After heating

2.4 348.5 350.5

2.7 348.5 352

3.0 353 355

6.8 359 361.5

Figure 2.12. Fluorescence spectra of SMP dispersions (5% w/v) with pyrene before and after acidification to pH 2.4-3.0 with citric acid, before and after heating at 90 °C for 2 min.

Table 2.2. I3/I1 of SMP dispersions (5% w/v) with pyrene before and after acidification to pH 2.4-3.0 with citric acid, before and after heating at 90 °C for 2 min.

pH of SMP dispersion I3/I1

Before heating After heating

2.4 1.029 1.029

2.7 1.009 1.019

3.0 1.026 1.010

6.8 1.105 1.124

Figure 2.13. Fluorescence spectra of SMP dispersions (5% w/v) with ANS before and after acidification to pH 2.4-3.0 with citric acid, before and after heating at

90 °C for 2 min.

A a ab bc cd cd d

e e

a a b b c c B

d d

Figure 2.14. Calcium (a) and phosphorous (b) concentrations in the serum phase

of SMP dispersions (5% w/v) before and after acidification to pH 2.4-3.0 with

citric acid, before and after heating at 90 °C for 2 min. Different letters above bars

indicate significant difference in the compared parameter (P < 0.05). Error bars

are SD (n=3).

A kDa B 120- 85-

50-

35-

25- 20-

Figure 2.15. SDS-PAGE of SMP dispersions (5% w/v) before (a) and after (b) heating at 90 °C for 2 min after acidification to pH 2.4, 2.7, and 3.0 (Left to Right) with citric acid. The far-right sample is unacidified control at pH 6.8.

Figure 2.16. Contents of free thiol groups in the SMP dispersions (5% w/v) before and after acidification to pH 2.4-3.0, before and after subsequent heating at

90 °C for 2 min. Different letters above bars indicate significant difference in the compared parameter (P < 0.05). Error bars are SD (n=2).

Figure 2.17. STEM image of SMP dispersion (1% w/v) at pH 6.8.

Figure 2.18. STEM image of SMP dispersion (1% w/v) after acidification to pH

2.4 with citric acid and heating at 90 °C for 10 min.

CHAPTER 3

Physicochemical properties of skim milk powder dispersions as

affected by calcium chelating sodium tripolyphosphate,

trisodium citrate, and sodium hexametaphosphate

3.1 Abstract

Calcium is essential for retaining the structure of casein micelles. Three calcium chelating agents, sodium tripolyphosphate (STPP), trisodium citrate (TSC), and sodium hexametaphosphate (SHMP), were studied at 1-30 mM for impacts on physicochemical properties of skim milk powder (SMP) dispersions. Each chelator had different influence on the turbidity, hydrodynamic diameter, and zeta-potential of SMP dispersions, with SHMP being the most effective in reducing turbidity of SMP dispersions (1074 NTU for STPP, 3483 NTU for TSC, and 431 NTU for SHMP at 1 mM). The amount of dissolved calcium was not directly related to the decreased turbidity indicating different chelating mechanisms by each chelator (146 mg/L for STPP, 185 mg/L for TSC, and 147 mg/L for SHMP at 1 mM). Addition of calcium chelators in SMP dispersions can be applied to the liquid products that require transparency before and after heat treatment.

3.2 Materials and methods

3.2.1 Chemicals

Carnation skim milk powder was from Nestle (Solon, OH). Sodium tripolyphosphate was purchased from Acros Organics (Pittsburgh, PA) and trisodium citrate and sodium hexametaphosphate were purchased from Alfa

Aesar (Ward Hill, MA). SDS was obtained from Fisher Scientific (Pittsburgh, PA).

3.2.2 Sample preparation

Approximate 5 percent (w/v) skim milk powder (SMP) dispersions with various calcium chelators; sodium tripolyphosphate (STPP), trisodium citrate (TSC), and sodium hexametaphosphate (SHMP) were prepared using 0.500 g of SMP and

10.0 mL of deionized water in 20 mL of scintillation vials. The range of calcium chelator concentration in SMP dispersions was 1, 2, 3, 5, 7, 10, 15, and 30 mM.

SMP dispersions with each chelator were vortexed for 10 s and then hydrated at least for 8 h at room temperature (RT, ~21 °C). After hydration, pH of each dispersion was adjusted by adding hydrochloric acid (1.0M) to pH 6.80 ± 0.01.

The pH of SMP dispersions was monitored using an Accumet AE150 pH meter

(Thermo Orion INC, Chelmsford, MA) at RT (~21 °C). The heat treatment was done using a TC-102D circulating water bath (AMETEK Brookfield,

Middleborough, MA) at 90 °C for 5 min. SMP dispersions with SDS (0.001, 0.005,

0.01, 0.03, 0.05, 0.10, 0.15, 0.20, 0.25, 0.30, and 0.35 g) were also prepared to study the significance of hydrophobic interactions on physical properties of SMP dispersions.

3.2.3 Turbidity and hydrodynamic diameter

A 2020 we/wi turbidimeter (LaMotte, Chestertown, MD) was used to determine turbidity of SMP dispersions at RT. Calibration was performed using AMCO primary turbidity standard (10 NTU). Dynamic light scattering (DLS) was used to measure the hydrodynamic diameter of dispersions (model Zetasizer nano-ZS,

Malvern Instrument, Enigma Business Park, UK). The backscatter angle was set as 173 ° and the measurement was done at 25 °C.

3.2.4 Inductively coupled plasma optical emission spectrometry (ICP-OES)

Calcium and phosphorous concentrations in the serum phase of SMP dispersions with various concentrations of chelators were determined. The serum phase was separated using an Amicon ultra-15 centrifugal filter unit with ultracel-

10 membrane (molecular weight cut off ~10,000 Da) (Merck Millipore Ltd, Cork,

Ireland) and centrifugation at 4,500 g for 3 h (model RC-5B plus, Sorvall,

Newtown, CT). After dilution for 100 times with deionized water, the analysis was performed using a Spectro Ciros ICP-OES instrument (Spectro Ciros, Mahwah,

NJ). ICP-OES standards were used in calibration (VHG Labs, Manchester, NH).

The detection limit for calcium and phosphorous was 0.01 mg/L.

3.2.5 Zeta-potential

The zeta-potential of SMP dispersions with various calcium chelators was determined using a zetasizer nano-ZS instrument (Malvern Instrument, Enigma

Business Park, UK). On the basis of Laser Doppler Electrophoresis, electrophoretic mobility, 푈퐸, was measured. Using Zetasizer Software 7.11, zeta- potential was derived from the Henry equation:

2휀휁 푈 = ( )푓(푘푎) 퐸 3휂

where 휀 is the dielectric permittivity of solvent, 휁 is the viscosity of solution, and

푓(푘푎) is Henry’s function (de Kort et al., 2012).

Eight hundred μL of sample was loaded in disposable folded capillary cells and the measurement was done at 25 °C with 120 s of equilibration time.

3.2.6 Transmission electron microscopy (TEM)

The morphology of SMP dispersions as affected by calcium chelators and SDS was studied using TEM. Approximate 50 μL of a SMP dispersion with 1, 7, or 30 mM STPP, TSC, or SHMP or with 0.30g of SDS was placed on a piece of

Parafilm. A freshly glow discharged 400 mesh copper grid with a thin carbon film was placed on the sample drop, film side down. Excess sample was quickly removed with a Kimwipes from the grid after 1 min and the grid was placed on a drop of water for only a few seconds. The water was removed with a Kimwipes and the grid was placed on a drop of 1% uranyl acetate for 1 min. Then excess stain was quickly removed using a Kimwipes and the sample was allowed to dry.

The stained samples were imaged on a Zeiss Libra 200MC (Carl Zeiss Inc., Fort

Lauderdale, FL) operating at 200 kV.

3.2.7 Statistical analysis

Statistical analyses were conducted using SAS enterprise guide 7.1 (SAS

Institute Inc., Cary, NC) with one-way analysis of variance (ANOVA). Fisher’s least significance difference (LSD) test was used to calculate mean ± standard

deviation (SD) from at least two replicates and differences of calculated mean values were considered significant with the level of significance of P < 0.05.

3.3 Results and discussion

3.3.1 Turbidity and hydrodynamic diameter of skim milk powder dispersions as affected by chelators or SDS

As presented in Figs. 3.1, 3.2, and 3.3, more transparent SMP dispersions were produced as the concentrations of calcium chelators were increased before heating. High concentrations (>15 mM) of all three calcium chelators increased hydrodynamic diameter before heating (Figs. 3.4, 3.5, and 3.6) whereas turbidity of SMP dispersions decreased continuously with increased calcium chelator concentration before heating (Figs. 3.7, 3.8, and 3.9). Similar trend was observed when SDS was added to SMP dispersions. Added SDS decreased the turbidity of SMP dispersions effectively (Fig. 3.10), however, hydrodynamic diameter linearly increased once it started to increase at 0.3 g of SDS (Fig. 3.11). These seemingly incompatible results (i.e., decreased turbidity with increased hydrodynamic diameter) might be due to conformational changes of casein micelles with high concentrations of calcium chelators or SDS.

Heating influenced both hydrodynamic diameter and turbidity of SMP dispersions with calcium chelators. The effect of heating on the hydrodynamic diameter of

SMP dispersions with STPP, TSC, and SHMP is represented in Figs. 3.4, 3.5, and 3.6. Decreased hydrodynamic diameter was observed after heating when

chelator concentration is less than or equal to 10 mM for SMP dispersions with

STPP (Fig. 3.4) and 3 mM for SMP dispersions with TSC (Fig. 3.5). No increase was observed in hydrodynamic diameter of SMP dispersions with SHMP, which means most heat-stable complexes between calcium and SHMP could be formed in the studied range of calcium chelator concentration. These observations are reflected in the turbidity results. For the SMP dispersions with

STPP and TSC, increased turbidity after heating could be observed at most concentrations (Figs. 3.7 and 3.8), while no significant increase was observed in the turbidity of SMP dispersions with SHMP (Fig. 3.9).

3.3.2 Calcium and phosphorous determination from serum phase of dispersions after calcium chelation

Calcium and phosphorous concentrations in the serum phase of SMP dispersions were studied using ICP-OES. As expected, SMP dispersions with all three calcium chelators showed increase in both calcium and phosphorous concentrations with increasing chelator concentration (Tables 3.1 and 3.2).

However, due to phosphorous in STPP and SHMP, net dissolved phosphorous concentration that is obtained from the serum phase of SMP dispersion with

STPP and SHMP is uncertain (Table 3.2). Calcium chelators bind to calcium ion with different affinity forming soluble or insoluble complexes (de Kort et al.,

2009). The reason that SMP had different turbidity at the same concentration but with different calcium chelators (Figs 3.7, 3.8, and 3.9) is probably because each chelator has different calcium-binding capacity. However, same mechanism is

applied to all the chelators to reduce the turbidity of SMP dispersions; 1) calcium is chelated, 2) CCP is disrupted, and then 3) casein micelles are dissociated.

Increased calcium and phosphorous concentration in the serum phase is in agreement with a previous study (Mizuno & Lucey, 2005), where casein-bound calcium and phosphorous were decreased with the addition of TSC. When low concentrations of STPP and SHMP were present in SMP dispersions (less than or equal to 3 mM for STPP and 2 mM for SHMP), calcium concentration in the serum phase was decreased compared to the control (Table. 3.1). Similar observation was obtained in a previous study (Mizuno & Lucey, 2005); casein- bound calcium was increased with small amount of SHMP addition. Considering the amount of solubilized calcium concentration, i.e., in the order of TSC > STPP

> SHMP (Table 3.1), the best calcium chelator for calcium in SMP dispersions should be TSC. However, TSC was not as effective as SHMP to reduce the turbidity of SMP dispersions. A possible explanation for these results is that different calcium chelating mechanisms by different chelators. Calcium is involved with two types of association; with inorganic phosphate (e.g., CCP) or with organic phosphate (e.g., phosphoserine residues in casein) (Gaucheron,

2005). TSC might interact with calcium that is associated with phosphoserine residues, thus SMP dispersions with TSC showed that turbid dispersions (Fig.

3.2) than other dispersions at the same concentration of chelators because disruption of CCP was not facilitated. Also, a relatively small decrease in the hydrodynamic diameter was observed in SMP dispersions with TSC, from 289.6 nm to 272.3 nm over TSC concentrations in the range 0 to 5 mM. In contrast, the

interaction between calcium in CCP and STPP or SHMP probably occurred. At low concentrations of STPP and SHMP, complexes consisting of casein molecules, calcium, and calcium chelators could be formed. These complexes containing partially dissociated casein micelles could be a plausible explanation for not only the decreased hydrodynamic diameter but also the decreased calcium concentration in the serum phase. Complexes including any amount of casein molecules are not allowed to penetrate the ultrafiltration membrane with a molecular weight cut-off of 10,000 Da after centrifugation because each casein molecule (i.e., ɑs-, β-, and κ-casein) has a molecular weight larger than 10,000

Da (von Hippel & Waugh, 1955; Waugh, 1958).

3.3.3 Zeta-potential of skim milk powder dispersions with chelators

Slight decrease in zeta-potential of SMP dispersions with increasing concentrations of STPP, TSC, and SHMP is represented in Table 3.4.

Regardless of association types of calcium, chelating of calcium from CCP or phosphoserine residues eventually could lead to more exposure of negatively charged phosphoserine residues. However, at higher concentration of chelators

(>15 mM), zeta-potential was highest. This could be interpreted that electrostatic repulsion is not strong enough to prevent some aggregation of particles when high concentration of calcium chelators exists in SMP dispersions. The higher zeta potentials at the concentrations of 15 and 30 mM are might be the reason why hydrodynamic diameter was increased at these concentrations (Figs. 3.4,

3.5, and 3.6).

3.3.4 Structural changes of calcium-chelated casein micelles using transmission electron microscopy (TEM)

Morphology of intact casein micelles in pure SMP dispersions without any addition of calcium chelator is displayed using TEM technique (Fig. 3.12).

Spherical shape of particles was observed with the diameter being around 150-

200 nm. Some aggregated form of particles with network formation was also identified having diameters of more than 400 nm. This variation of size could be associated with the mean hydrodynamic diameter of ~280 nm without calcium chelator (Fig. 3.4).

TEM images of SMP dispersions with STPP, TSC, and SHMP at the concentrations of 1, 7, and 30 mM are shown in Figs 3.13, 3.14, and 3.15.

Although hydrodynamic diameter values from dispersions were different each other, similar tendency that aggregation of dissociated particles at higher calcium chelator concentration was observed in all the images. At 1 mM of concentration,

SMP dispersions with STPP and TSC had spherical particles (Figs. 3.13 and

3.14) and its morphology is comparable to that of intact casein micelles (Fig.

3.12). However, particles in SMP dispersions with 1 mM of SHMP showed slight different structure with more dissociated and networked form (Fig. 3.15). This is related to the lowest value of hydrodynamic diameter with SHMP at 1 mM among dispersions. As concentration increased from 1 to 7 mM and further to 30 mM, more dissociated casein particles were observed, however, aggregated particles with bigger size were also emerged, which are identified by darker images with higher density. In SHMP, six negative charges are distributed evenly around the

molecules lending them the ability to crosslink caseins (de Kort et al., 2011). This characteristic could be observed in Fig. 3.15 showing network structure that is distinct from the structure of casein micelles with STPP or TSC. SMP dispersions with SHMP also exhibited darker particles indicating aggregation of dissociated casein micelles. Aggregations of casein micelles at higher concentrations of calcium chelators are in agreement with low zeta-potential at higher chelator concentrations. Similarly, huge aggregation that corresponds to the hydrodynamic diameter of ~450 nm at 0.30 g of SDS was observed in Fig. 3.16.

3.4 Conclusion

In conclusion, the turbidity of 5% w/v SMP dispersions could be effectively decreased using three types of calcium chelators: STPP, TSC, or SHMP. Higher chelator concentrations resulted in less turbid dispersions by chelating more calcium from CCP. TSC was the most effective in releasing calcium from casein micelles, while SHMP decreased the turbidity of the dispersions the most. This suggests that the mechanism of calcium chelation and possible complexes formed between calcium and chelators are slightly different for each chelator.

The hydrodynamic diameter of SMP dispersions started to decrease with the initial increase in concentration of added chelators as casein micelles are disrupted. However, at higher concentration of chelators, the aggregation of dissociated casein micelles was observed using DLS and TEM. The approach

studied in this chapter may be used to produce shelf-stable and heat-stable beverages with casein micelles.

References de Kort, E., Minor, M., Snoeren, T., van Hooijdonk, T., & van der Linden, E.

(2009). Calcium-binding capacity of organic and inorganic ortho-and

polyphosphates. Dairy Sci Technol, 89, 283-299. de Kort, E., Minor, M., Snoeren, T., van Hooijdonk, T., & van der Linden, E.

(2011). Effect of calcium chelators on physical changes in casein micelles in

concentrated micellar casein solutions. Int Dairy J, 21, 907-913. de Kort, E., Minor, M., Snoeren, T., van Hooijdonk, T., & van der Linden, E.

(2012). Effect of calcium chelators on heat coagulation and heat-induced

changes of concentrated micellar casin solutions: The role of calcium-ion

activity and micellar integrity. Int Dairy J, 26, 112-119.

Gaucheron, F. (2005). The minerals of milk. Reprod Nutr Dev, 45, 473-483.

Mizuno, R., & Lucey, J. A. (2005). Effects of emulsifying salts on the turbidity and

calcium phosphate-protein interactions in casein micelles. J Dairy Sci, 88,

3070-3078. von Hippel, P. H., & Waugh, D. F. (1955). Casein: Monomers and polymers1. J

Am Chem Soc, 77, 4311-4319.

Waugh, D. F. (1958). The interactions of ɑs- β- and κ-caseins in micelle

formation. Discuss Faraday Soc, 25, 186-192.

Appendix

A 0 mM 1 mM 2 mM 3 mM 5 mM 7 mM 10 mM 15 mM 30 mM

B 0 mM 1 mM 2 mM 3 mM 5 mM 7 mM 10 mM 15 mM 30 mM

Figure 3.1. The appearance of 5% w/v SMP dispersions with 0-30 mM STPP

before (a) and after (b) heating at 90 °C for 5 min.

100

A 0 mM 1 mM 2 mM 3 mM 5 mM 7 mM 10 mM 15 mM 30 mM

B 0 mM 1 mM 2 mM 3 mM 5 mM 7 mM 10 mM 15 mM 30 mM

Figure 3.2. The appearance of 5% w/v SMP dispersions with 0-30 mM TSC

before (a) and after (b) heating at 90 °C for 5 min.

101

A 0 mM 1 mM 2 mM 3 mM 5 mM 7 mM 10 mM 15 mM 30 mM

B 0 mM 1 mM 2 mM 3 mM 5 mM 7 mM 10 mM 15 mM 30 mM

Figure 3.3. The appearance of 5% w/v SMP dispersions with 0-30 mM SHMP

before (a) and after (b) heating at 90 °C for 5 min.

102

a b

c d e f g g j h i l k kl m m m n

Figure 3.4. The hydrodynamic diameter of 5% w/v SMP dispersions with 0-30 mM STPP before and after heating at 90 °C for 5 min. Different letters above bars indicate significant difference (P < 0.05). Error bars are SD (n=2).

103

a b c d d e f g g

j i h j k l m n n

Figure 3.5. The hydrodynamic diameter of 5% w/v SMP dispersions with 0-30 mM TSC before and after heating at 90 °C for 5 min. Different letters above bars indicate significant difference (P < 0.05). Error bars are SD (n=2).

104

b c

d de f g e fg e h h j ij i k l m

Figure 3.6. The hydrodynamic diameter of 5% w/v SMP dispersions with 0-30 mM SHMP before and after heating at 90 °C for 5 min. Different letters above bars indicate significant difference (P < 0.05). Error bars are SD (n=2).

105

c d

e f g hi k i l j m h l n

Figure 3.7. The turbidity of 5% w/v SMP dispersions with 0-30 mM STPP before and after heating at 90 °C for 5 min. Different letters above bars indicate significant difference (P < 0.05). Error bars are SD (n=2).

106

b c d e f g

h i j l k m n

Figure 3.8. The turbidity of 5% w/v SMP dispersions with 0-30 mM TSC before and after heating at 90 °C for 5 min. Different letters above bars indicate significant difference (P < 0.05). Error bars are SD (n=2).

*Data were not available at 5 mM before heating and at 7 mM after heating due to limited detectable range of the turbidimeter.

107

b c d d e e f f g f g h i j

Figure 3.9. The turbidity of 5% w/v SMP dispersions with 0-30 mM SHMP before and after heating at 90 °C for 5 min. Different letters above bars indicate significant difference (P < 0.05). Error bars are SD (n=2).

*Data were not available at 1 mM after heating due to limited detectable range of the turbidimeter.

108

SDS 0g 0.05g 0.10g 0.15g 0.20g 0.25g 0.30g 0.35g

Figure 3.10. The appearance of 5% w/v SMP dispersions with 0-0.35 g SDS.

109

a a

b b

d c de de e f d g e f g h i j j

h h

Figure 3.11. Turbidity and hydrodynamic diameter of 5% w/v SMP dispersions after addition of 0-0.35 g SDS. Different letters above symbols indicate significant difference in the compared parameter (P < 0.05). Error bars are SD (n=2).

110

Table 3.1. Calcium (Ca) concentrations in the serum phase of 5% w/v SMP dispersions after addition of 0-30 mM STPP, TSC, and SHMP.

Calcium chelator Ca concentration (mg/L)*

concentration (mM) STPP TSC SHMP

0 165.8± 0.4m 165.8± 0.4m 165.8± 0.4m

1 145.9 ± 0.1o 184.7± 0.5l 146.6 ± 1.0no

2 130.4 ± 4.2pq 206.4 ± 2.5k 145.4 ± 0.3o

3 129.4 ± 1.4q 231.7 ± 1.3j 159.6 ± 6.1mn

5 143.0 ± 0.2op 268.5 ± 5.8i 213.8 ± 0.7k

7 164.6 ± 6.4m 308.9 ± 6.9h 265.2 ± 7.4i

10 233.8 ± 4.2j 360.9 ± 1.3fg 357.2 ± 0.5g

15 371.1 ± 5.5ef 437.8 ± 3.1c 390.2 ± 24.9d

30 630.4 ± 3.4a 529.7 ± 1.7b 382.1 ± 10.9de

*Numbers are mean ± SD (n=2). Different superscript letters indicate significant difference (P < 0.05).

111

Table 3.2. Phosphorous (P) concentrations in the serum phase of 5% w/v SMP dispersions after addition of 0-30 mM STPP, TSC, and SHMP.

Calcium chelator P concentration (mg/L)*

concentration (mM) STPP TSC SHMP

0 220.9 ± 2.1n 220.9 ± 2.1n 220.9 ± 2.1n

1 241.9 ± 0.2lmn 221.6± 4.1n 257.6 ± 7.0lmn

2 274.0 ± 6.0klm 235.4 ± 5.2mn 346.2 ± 7.8ij

3 330.2 ± 4.8j 249.0 ± 2.8lmn 470.1 ± 8.1h

5 478.6 ± 1.2h 264.2 ± 1.8klmn 766.0 ± 10.0f

7 657.7 ± 25.0g 283.6 ± 0.5kl 1054.0 ± 0.1d

10 984.5 ± 9.5e 306.5 ± 0.4jk 1560.6 ± 8.3c

15 1580.5 ± 8.7c 334.2 ± 4.2ij 1965.6 ± 10.7b

30 3418.5 ± 19.6a 376.8 ± 6.5i 3411.7 ± 98.8a

*Numbers are mean ± SD (n=2). Different superscript letters indicate significant difference (P < 0.05).

112

Table 3.3. Calcium (Ca) and phosphorous (P) concentrations in the serum phase of 5% w/v SMP dispersions after addition of 0-0.30 g SDS.

SDS (g) Concentration (mg/L)*

Ca P

0 165.8 ± 0.4b 220.9 ± 2.1e

0.001 180.1 ± 1.4a 223.7 ± 3.7e

0.005 178.5 ± 1.1a 227.9 ± 3.9e

0.01 176.8 ± 7.7a 230.5 ± 11.2de

0.03 158.3 ± 3.6c 242.5 ± 7.9d

0.05 159.9 ± 0.5bc 266.4 ± 0.8c

0.10 142.9 ± 1.4d 301.5 ± 1.5b

0.30 108.9 ± 2.2e 385.6 ± 2.5a

*Numbers are mean ± SD (n=2). Different superscript letters indicate significant difference (P < 0.05).

113

Table 3.4. Zeta-potential of 5% w/v SMP dispersions after addition of 0-30 mM

STPP, TSC, and SHMP.

Calcium chelator Zeta-potential (mV)*

concentration (mM)

STPP TSC SHMP

0 -13.73 ± 1.61de -13.73 ± 1.61de -13.73 ± 1.61de

1 -14.01 ± 1.58def -12.90 ± 1.65d -15.00 ± 1.16efgh

2 -15.00 ± 0.77efgh -14.17 ± 1.01defg -15.25 ± 1.35fgh

3 -15.88 ± 1.45h -14.08 ± 1.13def -15.43 ± 1.16gh

5 -15.27 ± 0.99fgh -14.68 ± 1.25efgh -15.13 ± 1.49fgh

7 -15.60 ± 0.92h -15.70 ± 1.88h -11.40 ± 0.35c

10 -11.22 ± 0.59c -14.78 ± 0.95efgh -10.54 ± 0.72abc

15 -9.31 ± 1.57a -10.83 ± 0.78bc -10.48 ± 0.59abc

30 -9.80 ± 1.15ab -10.20 ± 0.74abc -9.57 ± 0.76ab

*Numbers are mean ± SD (n=2). Different superscript letters indicate significant difference (P < 0.05).

114

Figure 3.12. TEM images of 1% w/v SMP dispersion at pH 6.8.

115

Figure 3.13. TEM images of 1% w/v SMP dispersion with 1 mM (a), 7 mM (b), or

30 mM (c) STPP.

116

Figure 3.14. TEM images of 1% w/v SMP dispersion with 1 mM (a), 7 mM (b), or

30 mM (c) TSC.

117

Figure 3.15. TEM images of 1% w/v SMP dispersion with 1 mM (a), 7 mM (b), or

30 mM (c) SHMP.

118

Figure 3.16. TEM images of 1% w/v SMP dispersion with 0.30 g SDS.

119

CHAPTER 4

Conclusion and future work

120

In the present study, two different methods were studied to reduce turbidity of SMP dispersions (5% w/v) at acidic (≤pH 3) and neutral pH (pH 6.8). Acidification of

SMP dispersions led to solubilization of calcium and phosphorous improving clarity of SMP dispersions. Subsequent heating further decreased the turbidity of acidified

SMP dispersions. The weakened hydrophobic interactions between casein molecules in casein micelles might contributed to improved turbidity after heating of acidified SMP dispersions. At neutral conditions, TSC was more effective than

STPP and SHMP in demineralization of calcium from casein micelles. However, the turbidity of SMP dispersion was more effectively decreased by SHMP or STPP than TSC. This finding may indicate each chelator binds to different types of calcium: chelation of calcium from CCP may destabilize the structure of caseins micelles, whereas chelating calcium bound on phosphoserine residues may not.

Detailed information on the molecular weights of casein micelles, calcium chelators, and casein-chelator complexes should be obtained using analytical ultracentrifugation to study the mechanism of calcium chelation for each chelator.

Also, it is possible that the effect of calcium chelators on the turbidity of SMP dispersions might be slightly different at other pH values because the solubility of the complexes between calcium and chelators is dependent on pH. Therefore, the turbidity and hydrodynamic diameter of SMP dispersions with chelators at pH values other than 6.8 will be necessary to study the solubility of casein-chelator complexes. A higher concentration of calcium chelators in SMP dispersions increased zeta-potential, which resulted in aggregation of dissociated casein micelles. Translucent quality and the improved stability against gravitational

121

sedimentation due to reduced dimension of casein micelles may enable production of novel protein beverages using SMP.

122

VITA

Inseob Choi moved to Knoxville, Tennessee to study Food Science as a master student after he graduated from Sungkyunkwan University with a degree in Food

Science and Technology, He plans to continue his study in PhD program at the

University of Tennessee.

123