Copyright is owned by the Author of the thesis.  Permission is given for 
a copy to be downloaded by an individual for the purpose of research and 
private study only.  The thesis may not be reproduced elsewhere without 
the permission of the Author. 
 



 

 

 

 

Behaviour of Milk Protein Ingredients and 

Emulsions Stabilised by Milk Protein Ingredients  

In the Simulated Gastrointestinal Tract 

 

 

A thesis presented in partial fulfilment of the requirements 

for the degree of  

Master of Food Technology  

Massey University, Manawatu, New Zealand 

 

 

 

 

Xin Wang 

2017 

 

 



 

 



 
 

Abstract 

i 
 

Abstract 

Milk clotting behaviours in the stomach impact the digestion rates of protein and 

fat. A variety of milk protein products are applied as functional ingredients in many foods. 

This research was conducted to investigate the digestion behaviours of various 

commercial dairy ingredients and lipids in emulsions stabilised by these ingredients using 

a dynamic in vitro digestion model, i.e., a human gastric simulator (HGS), with a focus 

on the effect of different structures of clots formed in dairy ingredients during gastric 

digestion on hydrolysis of proteins and/or lipids.  

Skim milk powder (SMP), milk protein concentrate (MPC) 4851, MPC 4861, 

sodium caseinate, whey protein isolate (WPI) and heated (90°C, 20 min) WPI were used 

in the present study. Results showed that SMP and MPC 4851, which contained casein 

micelles, formed ball-like clots with a relatively dense network after 10 min of gastric 

digestion. These clots did not disintegrate after 220 min of digestion. MPC 4861 and 

sodium caseinate generated clots at around 40 min, and a loose, fragmented structure was 

observed at the end of the gastric digestion due to a lacking micellar structure of caseins. 

No clot was observed in WPI or heated WPI after 220 min gastric digestion, although 

aggregation occurred at around 40 min in heated WPI. These differences in coagulation 

behaviours apparently affected the rate of gastric emptying and protein hydrolysis by 

pepsin in the gastric system. In SMP and MPC 4851, the gastric emptying and hydrolysis 

of caseins was much slower than that observed in MPC 4861 and sodium caseinate. The 

most rapid gastric emptying of proteins was observed in the WPI samples both with and 

without heating. This is attributed to the formation of varied structured clots at different 

times under the gastric conditions.  

The effect of protein concentration on the gastric behaviour of these dairy 

ingredients in solution was then examined, with a particular emphasis on the structure of 

clots. SMP and MPC 4851 have been selected as model protein ingredients. Their gastric 

behaviours were investigated over a protein concentration range of 0.5-5.0% (w/w). The 

results showed that the digestion behaviour of SMP and MPC 4851 followed a similar 

pattern. The rate of pH changes in the emptied digesta during digestion was protein 

concentration dependent. With an increase in protein concentration, the decrease in pH 

slowed. The protein concentration had no apparent impact on the casein clotting time. 



 
 

Abstract 

ii 
 

Clots were formed in the first 10 min of digestion in all samples. However, in both SMP 

and MPC 4851, when protein concentration was lower than 2.0% (w/w) the clots 

consisted of small protein pieces with a loose, porous and open structure after a 220 min 

digestion. Whereas a cheese ball-like clot with a denser network was observed at the end 

of gastric digestion when the protein concentration varied from 2.0% to 5.0% (w/w). Such 

a difference in the structure apparently affected the rate of protein hydrolysis. A more 

rapid hydrolysis (P < 0.05) of the clotted protein was observed when protein 

concentration was lower than 2.0% (w/w) compared to the samples containing a higher 

proportion of protein (2.0%-5.0%, w/w). 

To study the effect of different coagulation behaviours on the digestion of oil 

droplets in oil-in-water emulsions, these dairy ingredients (with the exception of SMP) 

were used to prepare an oil-in-water emulsion (20.0% soy oil and 4.0% protein, w/w). 

They were digested under the dynamic gastric conditions using the HGS. The gastric 

digesta was emptied at 20 min intervals. Then all digesta were mixed to investigate the 

lipid digestion under the small intestinal conditions. Changes in physicochemical 

properties of emulsions, involving the particle size, the microstructure, the oil content of 

the emptied gastric digesta and the amount of free fatty acids (FFAs) released during the 

small intestine stage, were determined using an in vitro small intestinal digestion model.  

Aggregation of MPC 4851-stabilised emulsion took place after 5 min of digestion 

in the HGS with the largest size. The aggregates remained in the stomach and did not 

disappear during the whole gastric digestion. The hydrolysis of the aggregated network 

by pepsin was largely slowed by the reduced ability of the simulated gastric fluid (SGF, 

containing pepsin) to diffuse into the larger sized aggregates. MPC 4851-stabilised 

emulsion thus resulted in the slowest release of oil droplets into the small intestine. In 

comparison, MPC 4861 and sodium caseinate-stabilised emulsions aggregated in the 

stomach at approximately 40 min, forming smaller sized aggregates. These aggregates 

disintegrated at the mid and late-stages of digestion in these two emulsions. Therefore, 

MPC 4861 and sodium caseinate-stabilised emulsions had a more rapid delivery of oil 

droplets into the small intestine. In relation to the WPI-stabilised emulsions both with and 

without heating, the aggregations formed at a similar time to that which was observed in 

MPC 4861 and sodium caseinate-stabilised-emulsions; i.e., at approximately 40 min. 

However, they had the smallest sized aggregates amongst all samples and they 



 
 

Abstract 

iii 
 

disintegrated quickly with further digestion. WPI-stabilised emulsions both with and 

without heating had the fastest gastric emptying and hydrolysis by pepsin in the early and 

mid-stages of the gastric digestion process. Thus, the highest level of oil content contained 

in the emptied gastric digesta was produced from both WPI-stabilised emulsions. In the 

mixed gastric digesta, which were subjected to the small intestinal digestion, the oil 

contents contained in the different emulsion samples varied. This difference impacted the 

extent of lipid digestion by pancreatic lipase. The sample with a higher oil content 

released a greater amount of FFAs compared to the sample with a lower oil content. The 

extent of lipid digestion of different emulsion samples adhered to the following pattern: 

MPC 4851-stabilised emulsion < MPC 4861-stabilised emulsion < sodium caseinate-

stabilised emulsion, WPI-stabilised emulsions both with and without heating. 

Overall, the gastric behaviours of dairy ingredients either in solutions or 

emulsions were affected by the formation of structured clots/aggregates. The differences 

in clotting/aggregation times and their structures were greatly dependent on the 

component and structure of protein, the processing prior to digestion and the 

susceptibility to proteases. These differences in protein coagulation/aggregation 

behaviour impacted the rates of protein hydrolysis and gastric emptying. The oil content 

and protein composition of the gastric digesta transferred into small intestine and the 

extent of lipid digestion in small intestine were also affected. These results are important 

in an application perspective. They provide useful information for the design and 

development of healthier food products by allowing greater control over the manipulation 

of protein bioavailability, which subsequently provides greater control over lipid 

metabolism. 

 

 

 

 

 

 



 
 

Abstract 

iv 
 

 



 
 

Acknowledgment 

v 
 

Acknowledgment  

First and foremost, I would like to thank my supervisor Associate Professor 

Aiqian Ye, who has been supportive of my research and provided me with encouragement, 

direction, assistance, insightful comments and extensive personal and professional 

guidance throughout my Master study, and taught me a great deal about both scientific 

research and life in general. He also helped me to coordinate my project especially in 

writing this report. As my supervisor and mentor, he has taught me more than I will ever 

know. He has shown me, by his example, what a good scientist should be.  

 I would like to express my deepest appreciation to Professor Harjinder Singh for 

providing me the possibility to complete my research in Riddet Institute with a financial 

assistance. 

My special thanks go to my teammate Quanquan Lin, who gave me selfless help, 

encouragement and sharing her pearls of wisdom with me during the course of this 

research. 

Furthermore, I would also like to acknowledge with much appreciation Ms 

Maggie Zou, Ms Janiene Gilliland, Mr. Chris Hall, and Mr. Steve Glasgow, who gave me 

the permission to use all required equipment and the necessary materials to complete my 

research, as well as providing timely assistance for reagents ordering, laboratory 

induction, safety advice, and training and guidance of the use of instruments. A special 

gratitude I give to Mr. Jian Cui, who has provided me training, technical support and 

scientific suggestions in my overall practical work in the laboratory. I am especially 

indebted to Dr. Matthew Savoian, Ms Jordan Taylor and Ms Niki Minards for their 

valuable help and training in using Laser Scanning Confocal Microscopy (LSCM).  

I am grateful to Ms Ansley Te Hiwi, Ms Terri Palmer, Ms Hannah Hutchinson 

and Dr. Michael Parker for their administrative assistances. I would like to thank Mr. 

Matt Levin for his assistance in information systems. I am also thankful Mr. John Henley-

King. 

I am also immensely grateful to all the staffs and research fellows whom I have 

had pleasure to work during this project at Riddet Institute and Massey Institute of Food 



 
 

Acknowledgment 

vi 
 

Science and Technology. I also would like to express my appreciation to my friends, Nan 

Luo, Yu Cheng, Xiaoqi Sang, Zhigao Niu, Siqi Li, Lisanne Fermin, Sewuese Okubanjo, 

Geeshani Somaratne, Feng Ming Chian, Chih-Chieh Chuang and Nicole Chen for their 

encouragements and supports. 

Finally, I wish to thank my parents for their encouragement, generosity and 

financial support. I would not complete my study without them. Their love and guidance 

are with me in whatever I pursue. 



  
 

Table of Contents 

vii 
 

Table of Contents 

 
Abstract ............................................................................................................................. i 

Acknowledgment ............................................................................................................. v 

Table of Contents .......................................................................................................... vii 

List of Tables .................................................................................................................. xi 

List of Figures ............................................................................................................... xiii 

List of Abbreviations ................................................................................................... xix 

Chapter 1: Introduction ................................................................................................. 1 

Chapter 2: Literature Review ........................................................................................ 5 

2.1 The human gastrointestinal tract ........................................................................ 5 

2.1.1 Stomach .................................................................................................... 5 

2.1.1.1 The pH and ionic strength of gastric fluid ....................................... 6 

2.1.1.2 Enzyme ............................................................................................ 7 

2.1.2 Small intestine .......................................................................................... 7 

2.1.2.1 pH and ionic strength ...................................................................... 7 

2.1.2.2 Bile salts .......................................................................................... 8 

2.1.2.3 Pancreatic lipase .............................................................................. 9 

2.2 Milk protein ..................................................................................................... 10 

2.2.1 Casein ..................................................................................................... 10 

2.2.1.1 Casein micelle structure ................................................................ 11 

2.2.1.2 The stability of casein micelles ..................................................... 13 

2.2.1.3 Enzymatic coagulation of caseins ................................................. 14 

2.2.2 Whey protein .......................................................................................... 15 

2.2.2.1 β-lactoglobulin............................................................................... 15 

2.2.2.2 α-lactalbumin ................................................................................. 16 

2.2.2.3 Heating induced denaturation of whey protein ............................. 16 

2.2.3 Milk protein products ............................................................................. 18 

2.2.3.1 Skim milk powder ......................................................................... 19 

2.2.3.2 Milk protein concentrate ............................................................... 21 

2.2.3.3 Sodium caseinate ........................................................................... 21 



  
 

Table of Contents 

viii 
 

2.2.3.4 Whey protein isolate ...................................................................... 22 

2.2.4 The digestion behaviours of milk protein during gastric digestion ....... 22 

2.3 Emulsion .......................................................................................................... 24 

2.3.1 Emulsion formation ............................................................................... 25 

2.3.2 Emulsion stability .................................................................................. 26 

2.3.2.1 Gravitational separation ................................................................ 26 

2.3.2.2 Flocculation ................................................................................... 28 

2.3.2.3 Coalescence ................................................................................... 29 

2.3.3 Protein emulsifier ................................................................................... 29 

2.3.3.1 MPC .............................................................................................. 31 

2.3.3.2 Sodium caseinate ........................................................................... 32 

2.3.3.3 Whey protein isolate ...................................................................... 33 

2.3.4 The digestion behaviours of milk protein-stabilised emulsions ............ 34 

2.3.4.1 Milk protein-stabilised emulsions in the gastric environment ...... 35 

2.3.4.2 Milk protein-stabilised emulsions in the intestinal environment .. 36 

Chapter 3: Materials and Methods ............................................................................. 39 

3.1 Materials .......................................................................................................... 39 

3.1.1 Dairy ingredients .................................................................................... 39 

3.1.2 Soybean oil ............................................................................................. 39 

3.1.3 Chemicals ............................................................................................... 39 

3.1.4 Enzymes ................................................................................................. 40 

3.1.5 Simulated gastric fluid (SGF) ................................................................ 40 

3.1.6 Simulated intestinal fluid (SIF) .............................................................. 40 

3.2 Methods ........................................................................................................... 41 

3.2.1 Preparation of protein solution (Chapter 4 and 5) ................................. 41 

3.2.2 Preparation of emulsions ........................................................................ 41 

3.2.2.1 Protein solution preparation (Chapter 6) ....................................... 41 

3.2.2.2 Emulsion preparation .................................................................... 41 

3.2.3 In vitro gastric digestion ........................................................................ 41 

3.2.3.1 Human gastric simulator (HGS) .................................................... 43 

3.2.3.2 pH measurement ............................................................................ 44 

3.2.3.3 Weight of clot ................................................................................ 44 



  
 

Table of Contents 

ix 
 

3.2.3.4 Measurement of oil content (Chapter 6) ........................................ 44 

3.2.4 In vitro intestinal digestion (Chapter 6) ................................................. 46 

3.2.4.1 Measurement of free fatty acid release .......................................... 46 

3.2.5 Particle size measurements .................................................................... 47 

3.2.6 Confocal laser scanning microscopy ...................................................... 48 

3.2.7 Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)  

 ............................................................................................................. 49 

3.2.7.1 Preparation of stock solutions ....................................................... 49 

3.2.7.2 Gel preparation .............................................................................. 50 

3.2.7.3 Sample preparation ........................................................................ 51 

3.2.7.4 Running of electrophoresis, staining and destaining ..................... 51 

3.2.8 Statistical analysis .................................................................................. 52 

Chapter 4: Behaviours of Different Milk Protein Ingredients during in Vitro Gastric 

Digestion ......................................................................................................................... 53 

4.1 Abstract ............................................................................................................ 53 

4.2 Introduction...................................................................................................... 54 

4.3 Results ............................................................................................................. 58 

4.3.1 pH profiles.............................................................................................. 58 

4.3.2 Coagulation behaviour of different milk protein ingredients ................. 59 

4.3.3 Protein hydrolysis .................................................................................. 62 

4.3.3.1 SDS-PAGE pattern of clots ........................................................... 62 

4.3.3.2 SDS-PAGE patterns of emptied digesta ........................................ 63 

4.4 Discussion ........................................................................................................ 68 

4.5 Conclusions ..................................................................................................... 73 

Chapter 5. The Dynamic Gastric Digestion Behaviours of Skim Milk Powder and 

Milk Protein Concentrate: the Influence of Protein Concentration ........................ 75 

5.1 Abstract ............................................................................................................ 75 

5.2 Introduction...................................................................................................... 76 

5.3 Results and Discussion .................................................................................... 77 

5.3.1 Skim milk powder .................................................................................. 77 

5.3.1.1 Results ........................................................................................... 77 

5.3.1.2 Discussion ..................................................................................... 88 



  
 

Table of Contents 

x 
 

5.3.2 MPC 4851 .............................................................................................. 90 

5.3.2.1 Results ........................................................................................... 90 

5.3.2.2 Discussion ................................................................................... 100 

5.4 Conclusions ................................................................................................... 104 

Chapter 6: Behaviours of Oil-in-Water Emulsion Stabilised by Different Milk 

Protein Ingredients during in Vitro Gastrointestinal Digestion ............................. 105 

6.1 Abstract .......................................................................................................... 105 

6.2 Introduction ................................................................................................... 107 

6.3 Results ........................................................................................................... 112 

6.3.1 In vitro gastric digestion of emulsions ................................................. 112 

6.3.1.1 pH profiles ................................................................................... 112 

6.3.1.2 Changes in emulsion during gastric digestion ............................. 113 

6.3.1.3 Average droplet size of emulsions in SGF .................................. 114 

6.3.1.4 Particle size distribution .............................................................. 118 

6.3.1.5 Oil content of emptied digesta .................................................... 122 

6.3.1.6 Microstructure of digestive residue in the stomach .................... 124 

6.3.1.7 Microstructure of emptied digesta ............................................... 128 

6.3.1.8 SDS-PAGE patterns of emptied digesta ...................................... 131 

6.3.2 In vitro small intestinal digestion of emulsion ..................................... 135 

6.3.2.2 Microstructure of initial mixed-digesta in SIF ............................ 135 

6.3.2.3 The change in average droplet size of emulsion during intestinal 

digestion ....................................................................................... 136 

6.3.2.4 Release of free fatty acids during small intestinal digestion ....... 139 

6.4 Discussion ...................................................................................................... 142 

6.4.1 The digestion behaviours of various dairy ingredients-stabilised 

emulsions during gastric digestion .................................................... 142 

6.4.2 The digestion behaviours of various dairy ingredients-stabilised 

emulsions during small intestinal digestion ....................................... 148 

6.5 Conclusions ................................................................................................... 151 

Chapter 7: Overall Summary and Recommendations ............................................ 153 

Reference ...................................................................................................................... 157 

 



 
 

List of Tables 

xi 
 

List of Tables 

 

Number           Page 

Table 2.1. Average characteristics of casein micelles (Fox, 2003). ............................... 12 

Table 2.2. Thermal denaturation temperature of whey proteins (De Wit, 1984). .......... 17 

Table 2.3. Food applications of skim milk powder (SMP) of different heat classes (Kelly 

& Fox, 2016) ................................................................................................. 20 

Table 3.1. The composition of dairy ingredients ............................................................ 39 

Table 6.1. The volume (d4,3) and surface (d3,2) mean diameters of original emulsions 

stabilised by different dairy ingredients (20.0% soybean oil and 4.0% protein, 

w/w)............................................................................................................. 115 

Table 6.2. The volume (d4,3) mean diameters and oil content of the mixed gastric digesta 

emptied from emulsions made with different dairy ingredients. ................ 135 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



  
 

Table of Contents 

xii 
 

 



 
 

List of Figures 

xiii 
 

List of Figures 

 

Number           Page 

Figure 2.1. The external and internal anatomy of the stomach of human (Tortora & 

Derrickson, 2008) ........................................................................................... 6 

Figure 2.2. Main components of milk proteins showing partitioning into casein and whey 

fractions (Cheison & Kulozik, 2017). .......................................................... 10 

Figure 2.3. The casein micelle schematic diagram (Walstra & Jenness, 1984). ............ 13 

Figure 2.4. Functional protein ingredients manufactured from skim milk [derived from 

Singh (2005)]................................................................................................ 19 

Figure 3.1. The schematic diagram of digestion process………………………………42 

Figure 3.2. Image of a HGS (A) and schematic illustration of a latex stomach chamber 

(B). (1) SGF; (2) plastic tubes for secretion; (3) pump; (4) latex stomach 

chamber; (5) mesh bag; (6) roller; (7) belt; (8) pulley; (9) shaft; (10) angle 

gear; (11) Lovejoy joint; (12) fan heater for temperature control. Picture 

derived from Ye et al. (2016b). .................................................................... 44 

Figure 4.1. Changes under gastric digestion in pH of the different dairy ingredient 

solutions with 3.0% (w/w) protein………………….....................................59 

Figure 4.2. Images of clots obtained from 200 g of dairy ingredient solution containing 

3.0% (w/w) protein after 220 min of in vitro gastric digestion .................... 60 

Figure 4.3. Wet weight of clots obtained from 200 g of dairy ingredient solution 

containing 3.0% (w/w) protein after 220 min of in vitro gastric digestion. 

Different lowercase letters indicate significant difference (P < 0.05) in the 

wet weight of the clots. ................................................................................ 61 

Figure 4.4. Dry weight of clots obtained from 200 g dairy ingredient solution containing 

3.0% (w/w) protein after 220 min of in vitro gastric digestion. Different 

lowercase letters indicate significant difference (P < 0.05) in the dry weight 

of the clots. ................................................................................................... 62 

Figure 4.5. SDS-PAGE patterns of clots collected from different dairy ingredients after 

220 min of gastric digestion in the HGS. M, trim milk; SM, skim milk powder; 



 
 

List of Figures 

xiv 
 

MPC, milk protein concentrate  4851; D-MPC, milk protein concentrate 4861; 

SC, sodium caseinate. .................................................................................. 63 

Figure 4.6. SDS-PAGE patterns of the emptied digesta collected from different dairy 

ingredients during 220 min of gastric digestion in the HGS. M, trim milk. 

BSA, bovine serum albumin. (A) skim milk powder; (B) MPC 4851; (C) 

MPC 4861; (D) sodium caseinate; (E) WPI; (F) WPI heated (90°C, 20 min).

 ...................................................................................................................... 66 

Figure 5.1. pH of emptied digesta obtained from the SMP samples with different protein 

concentrations (0.5-5.0%, w/w) as a function of digestion time. Blank refers 

to a control experiment carried out without addition of 

SMP……………………………………………………………………......78 

Figure 5.2 Images of the clots obtained from 200 g of SMP samples containing a different 

level of protein (0.5-5.0%, w/w) after 220 min of in vitro gastric digestion.

 ...................................................................................................................... 79 

Figure 5.3. Wet weight of the clots obtained after 220 min of in vitro gastric digestion of 

SMP samples containing different protein concentrations (0.5-5.0%, w/w). 

Different lowercase letters indicate significant difference (P < 0.05) in the 

wet weight of clots. ...................................................................................... 80 

Figure 5.4. Dry weight of the clots obtained after 220 min of in vitro gastric digestion of 

SMP samples containing different protein concentrations (0.5-5.0%, w/w). 

Different lowercase letters indicate significant difference (P < 0.05) in the dry 

weight of clots. ............................................................................................. 81 

Figure 5.5. The curd weight ratio after 220 min of in vitro gastric digestion of the SMP 

samples containing different protein concentrations (0.5-5.0%, w/w). 

Different lowercase letters indicate significant difference (P < 0.05) in the 

curd weight ratio (g dry matter in the clot/g protein in the initial sample). . 82 

Figure 5.6. The microstructure of the clot obtained from 200 g of SMP samples 

containing (A) 0.5%, (B) 2.0%, and (C) 5.0% (w/w) protein after 220 min of 

in vitro gastric digestion. The scale bar in all images is 50 μm. .................. 83 

Figure 5.7. The SDS-PAGE pattern under reducing conditions of the clots collected from 

the SMP samples with a different protein concentration (0.5-5.0%, w/w) after 



 
 

List of Figures 

xv 
 

220 min of gastric digestion in the HGS. M, trim milk. BSA, bovine serum 

albumin. ........................................................................................................ 84 

Figure 5.8. SDS-PAGE patterns of emptied digesta collected from the SMP samples 

containing (A) 0.5%, (B) 1.0%, (C) 2.0%, (D) 3.0%, (E) 4.0% and (F) 5.0% 

(w/w) protein during 220 min of gastric digestion in the HGS. M. trim milk. 

BSA, bovine serum albumin. ....................................................................... 87 

Figure 5.9. Changes in pH of emptied digesta obtained from MPC 4851 samples with a 

different protein concentration (0.5-5.0%, w/w) as a function of digestion 

time. Blank refers to a control experiment carried out without addition of 

MPC 4851…………………………………………………………………91 

Figure 5 10. Images of clots obtained from 200 g of MPC 4851 samples containing a 

different level of protein (0.5-5.0%, w/w) after 220 min of gastric digestion 

in the HGS………………………………………………………......……...92 

Figure 5 11. Wet weights of the clots obtained after 220 min of in vitro gastric digestion 

of the MPC 4851 samples containing different protein concentrations (0.5-

5.0%, w/w). Different lowercase letters indicate significant difference (P < 

0.05) in the wet weight of clot……………………………………………....93 

Figure 5.12. Dry weights of the clots obtained after 220 min of in vitro gastric digestion 

of MPC 4851 samples containing different protein concentrations (0.5-5.0%, 

w/w). Different lowercase letters indicate significant difference (P < 0.05) in 

the dry weight of the clot…………………………………………………...93 

Figure 5.13. The curd weight ratio of MPC 4851 samples containing different protein 

concentrations (0.5-5.0%, w/w). Different lowercase letters indicate 

significant difference (P < 0.05) in the curd weight ratio (g dry weight of the 

clot/g protein in the initial sample) ………………………………………...94 

Figure 5.14. The microstructure of the clots obtained from 200 g of MPC 4851 samples 

containing (A) 0.5%, (B) 2.0%, and (C) 5.0% (w/w) protein after 220 min of 

gastric digestion. The scale bar in all images is 50 μm………………….… 95 

Figure 5.15. The SDS-PAGE pattern of the clots collected from the MPC 4851 samples 

containing different protein concentrations (0.5-5.0% w/w) after 220 min of 

gastric digestion in the HGS. M, trim milk. BSA, bovine serum albumin…96 



 
 

List of Figures 

xvi 
 

Figure 5.16. SDS-PAGE patterns of the emptied digesta collected from the MPC 4851 

samples containing (A) 0.5%, (B) 1.0%, (C) 2.0%, (D) 3.0%, (E) 4.0% and 

(F) 5.0% (w/w) during 220 min of gastric digestion in the HGS. M. trim milk. 

BSA, bovine serum albumin……………………………………………….99 

Figure 6.1. Changes in pH of digesta emptied from different milk protein ingredients-

stabilised emulsions during 220 min of gastric digestion…………………113 

Figure 6.2. The aggregates collected from MPC 4851-stabilised emulsion after 220 min 

of gastric digestion in the HGS. ................................................................. 114 

Figure 6.3. The changes in mean diameters volume (d4,3) of emptied digesta obtained 

from different milk protein ingredients-stabilised emulsions (20.0% soybean 

oil and 4.0% protein, w/w) during 220 min of gastric digestion in the HGS. 

(A) MPC 4851-stabilised emulsion; (B) MPC 4861-stabilised emulsion; (C) 

sodium caseinate-stabilised emulsion; (D) WPI-stabilised emulsion; (E) WPI-

stabilised emulsion heated (90°C, 20 min). Different lowercase letters 

indicate significant difference (P  0.05) on the mean diameters volume (d4,3) 

of digesta emptied from emulsions between  different digestion time points 

within the same emulsifier. ........................................................................ 117 

Figure 6.4. Particle size distribution of emptied digesta obtained from 220 min gastric 

digestion of different protein stabilised emulsions (20.0% soybean oil and 4.0 % 

protein, w/w) in the HGS: (A) MPC 4851-stabilised emulsion; (B) MPC 

4861-stabilised emulsion; (C) sodium caseinate-stabilised emulsion; (D) 

WPI-stabilised emulsion; (E) WPI-stabilised emulsion heated (90°C, 20 min).

 .................................................................................................................... 122 

Figure 6.5. Oil content (g oil/100 g emptied gastric digesta) of the emptied gastric digesta 

at different digestion time points. Different capital letters indicate significant 

difference (P  0.05) on the oil content between the gastric digesta emptied 

from different dairy ingredients-stabilised emulsions within the same 

digestion time point. Different lowercase letters indicate significant 

difference (P  0.05) on the oil content of emptied digesta between different 

digestion time points within the same emulsifier. ...................................... 124 

Figure 6.6. Confocal microscopy images of digestion residues of different milk protein 

ingredients-stabilised emulsions in the stomach at different times during 



 
 

List of Figures 

xvii 
 

gastric digestion from 0 to 220 min. All samples were stained with Nile Red 

(for oil) and Fast Green (for protein). The scale bar in all images is 50 μm.

 .................................................................................................................... 127 

Figure 6.7. Confocal microscopy images of the emptied digesta from different milk 

protein ingredients-stabilised emulsions at different time points during gastric 

digestion from 0 to 220 min. Oil is stained red, protein is stained green. The 

scale bar in all images is 50 μm. ................................................................ 130 

Figure 6.8. SDS-PAGE patterns under reducing conditions of the emptied digesta 

obtained at the different time points during 220 min of gastric digestion from: 

(A) MPC 4851-stabilised emulsion; (B) MPC 4861-stabilised emulsion; (C) 

sodium caseinate-stabilised emulsion; (D) WPI-stabilised emulsion; (E) 

heated (90°C, 20 min) WPI-stabilised emulsion. BSA, bovine serum albumin; 

M, trim milk. .............................................................................................. 134 

Figure 6.9. Confocal microscopy images of initial mixed-digesta (0 min) in the SIF of 

the different milk protein ingredients-stabilised emulsions. All samples were 

stained with Nile Red (for oil) and Fast Green (for protein). The scale bar in 

all images is 50 μm. ................................................................................... 136 

Figure 6.10. Changes in the average droplet diameter (d4,3) of the mixed gastric digesta 

emptied from emulsions stabilised by different dairy ingredients during 

digestion in the SIF. ................................................................................... 137 

Figure 6.11. The changes in volume (d4,3) mean diameters (-1) and the size distributions 

(-2) of oil droplets in the emptied digesta from different milk protein 

ingredients-stabilised emulsions (20.0% soybean oil and 4.0% protein, w/w) 

during 120 min of intestinal digestion: (A) MPC 4851-stabilised emulsion; 

(B) MPC 4861-stabilised emulsion; (C) sodium caseinate-stabilised emulsion; 

(D) WPI-stabilised emulsion; (E) WPI-stabilised emulsion with heating 

(90°C, 20 min). Different lowercase letters indicate significant difference (P

0.05) on the volume (d4,3) mean diameters of oil droplets in emptied digesta 

between different digestion time points within the same emulsifier type. .139 

Figure 6.12. Amount of free fatty acids released (μmol FFA/mL mixed gastric digesta) 

from emulsion made with different dairy ingredients (titrated by 0.25 M 

NaOH) in pH-stat during 120 min intestinal digestion. ............................. 141 



 
 

List of Figures 

xviii 
 

Figure 6.13. Amount of free fatty acids (%) released from per gram oil contained in the 

mixed gastric digesta from emulsions made with different dairy ingredients 

(titrated by 0.25M NaOH) in pH-stat during 120 min intestinal digestion.

 .................................................................................................................... 141 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 
 

List of Abbreviations 

xix 
 

List of Abbreviations 

α-La:   α-lactalbumin 

β-Lg:   β-lactoglobulin 

BSA:  Bovine serum albumin 

CCP:  Colloidal calcium phosphate 

HGS:  Human gastric simulator 

MPC:  Milk protein concentrate 

WPI:  Whey protein isolate 

WPNI:  Whey protein nitrogen index 

SDS-PAGE: Sodium dodecyl sulfate-poly acrylamide electrophoresis 

SGF:  Simulated gastric fluid 

SIF:  Simulated intestinal fluid 

d4,3:  Average volume-weighted diameter 

d3,2:  Average surface-weighted diameter 

PI:  Isoelectric point 

w/w  Weight/weight 

w/v  Weight/volume 

v/v  Volume/volume 

 

 

 

 



 
 

List of Abbreviations 

xx 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 
 

Chapter 1: Introduction 

1 
 

Chapter 1: Introduction 

Milk protein is an important source of nutrients for humans through the different 

stages of life, and digestion behaviours of milk protein have been studied in both vivo and 

vitro models (Dangin et al., 2001; Mahé et al., 1995; Ye, Cui, Dalgleish, & Singh, 2016a, 

2016b, 2017). Milk protein contains two fractions, caseins and whey proteins. In milk, 

caseins exist in colloidal particles known as casein micelles, which contain thousands of 

individual protein molecules (O’Mahony & Fox, 2013). Previous studies have shown that 

the digestion behaviour of casein in gastrointestinal tract is affected by the state of casein. 

Because casein micelles are coagulated both by milk-clotting enzyme pepsin (Tam & 

Whitaker, 1972) and acidic pH (Dalgleish & Corredig, 2012), while the individual casein 

(e.g. caseinate) is only coagulated by acidic pH not by pepsin. 

Recently, a pervious study provided a novel insight into the role of food structure 

in digestion (Ye et al., 2016b). They proposed that the gastric digestion of milk protein is 

affected by the structure of the clot induced by the action of pepsin on κ-casein during 

dynamic in vitro gastric digestion. Different structures (e.g. dense or loose) of clots 

formed in unheated skim milk and skim milk that has been preheated at 90°C for 20 min, 

leading to a different rate in protein hydrolysis. The further study of digestion behaviour 

of (raw and heated) whole milk revealed that this difference in the rate of protein 

hydrolysis induced by a various structured clot under gastric conditions markedly 

impacted the rate in release of fat globules (Ye et al., 2016a). 

Lipids, not only exist in milk but also in most other types of foods, and play a 

significant role in the human diet, including providing texture, flavour, and mouth-feel. 

(Singh, Ye, & Horne, 2009). Lipids also perform many important functions in the human 

body, such as providing a concentrated source of energy (Golding & Wooster, 2010). 

Meanwhile, overconsumption of lipids has been associated with a variety of diseases and 

health conditions, e.g. artherosclerosis, hypertension, and in the development of obesity 

(Shahidi, 2006). Recently, the increasing consumer awareness about the relationships 

between human health and high-calorie diet has promoted the food industry to design and 

develop healthier foods with a focus on reducing the adsorption of calorie, and without 

diminishment of the desirable sensory qualities of food (Chung, Olson, Degner, & 



 
 

Chapter 1: Introduction 

2 
 

McClements, 2013; Le Révérend, Norton, Cox, & Spyropoulos, 2010; McClements, 

Decker, Park, & Weiss, 2009; Singh et al., 2009). In many processed foods, lipids exist 

in the form of emulsified oil droplets, such as mayonnaise, salad dressing, ice cream and 

soups (McClements, 2005). Milk proteins, as the most common emulsifiers, are applied 

to prepare food emulsion, due to their excellent emulsification properties. 

Digestion behaviour of oil-in-water emulsions in the gastrointestinal tract has 

received growing interests (Golding & Wooster, 2010; Li, Ye, Lee, & Singh, 2012; Mun, 

Decker, Park, Weiss, & McClements, 2006; Sarkar, Goh, Singh, & Singh, 2009; Sarkar, 

Horne, & Singh, 2010b; Sarkar, Ye, & Singh, 2016; Singh et al., 2009; Ye et al., 2016a). 

Moreover, many attempts have been made to modulate lipid digestion by rational design 

of the structure of emulsions, such as manipulating interfacial composition (Golding & 

Wooster, 2010; Maldonado-Valderrama et al., 2008; Singh & Ye, 2013). However, most 

studies have used a static digestion model to mimic the gastric environment, only very 

few studies used a dynamic gastric model. Theoretically, most protein-stabilised 

emulsions will undergo a substantial modification in a more real gastric environment due 

to the dynamic change in pH value (Singh & Ye, 2009). It can thus be inferred that these 

alterations in structure of foods during gastric digestion will, to some extent, have an 

influence on the lipid digestion in the small intestine. 

In the present research, an in vitro dynamic digestion model was employed to 

achieve the process of food digestion in the stomach stage. It is preferable to conduct 

investigations with in vitro assays because compared to in vivo study, they are faster, 

simpler and pose no ethical problems. Dairy protein ingredients have been applied in a 

wide variety of food products. They encompass different proteins (caseins or/and whey 

proteins), and have been processed through different methods during manufacturing 

process. Thus, the digestion behaviours of different dairy protein ingredients in the 

gastrointestinal tract might have some differences. The aim of this study was thus to 

investigate the behaviour of various of commercial milk protein ingredients in the 

stomach and lipids digestion in emulsions stabilised by these ingredients in the GI tract, 

with a focus on the effect of different structures formed in foods during dynamic gastric 

digestion on protein or/and lipid digestion. 

 



 
 

Chapter 1: Introduction 

3 
 

The main objectives of the present study were as follow: 

1) To understand the gastric digestion behaviours of different milk protein 

ingredients with a range of different structures in a dynamic stomach model (human 

gastric simulator-HGS). The commercial dairy ingredients employed in the present work 

include skim milk powder (SMP), milk protein concentrate (MPC) 4851, MPC 4861, 

sodium caseinate, whey protein isolate (WPI) and heated (90°C, 20 min) WPI.  

2) To explore the influence of protein concentration on the dynamic gastric 

digestion behaviours (e.g., protein clotting time and protein hydrolysis rate) of milk 

protein ingredients in the HGS.  SMP and MPC 4851 were selected as model protein 

ingredients. 

3) To investigate the dynamic gastric digestion behaviours of proteins and lipids 

in emulsions made with different milk protein ingredients in the HGS, and how they affect 

the subsequent lipid digestion in the small intestine. The oil-in-water emulsions were 

stabilised by MPC 4851, MPC 4861, sodium caseinate and WPI. WPI-stabilised emulsion 

was treated by heating at 90°C for 20 min prior to digestion. 

These objectives have been completed through three research chapters in this 

study. The gastric digestion behaviours of different milk protein ingredients have been 

reported in Chapter 4. The influence of protein concentration on dynamic gastric digestion 

behaviour of milk protein ingredients was studied in Chapter 5. The digestion behaviour 

of oil droplets in emulsions stabilised by different milk protein ingredients during 

gastrointestinal tract was investigated in Chapter 6. 

 

 

 

 

 

 

 

 



 
 

Chapter 1: Introduction 

4 
 

 



 
 

Chapter 2: Literature Review 

5 
 

Chapter 2: Literature Review 

This chapter covers the knowledge of milk proteins, milk protein ingredients, oil-

in-water emulsion and its stability, milk protein emulsifier, and human gastrointestinal 

tract etc. Moreover, it also reviews in vitro and in vivo gastric digestion of milk proteins 

and in vitro gastrointestinal digestion of oil-in-water emulsions based on current literature, 

with a focus on summarizing the behaviour of the milk proteins with different 

characteristics during digestion process. 

2.1 The human gastrointestinal tract 

The nutrients that exist in food are involved in different food structures. Digestion 

is the process of the disintegration of food matrices in the mouth, stomach and intestine, 

leading to the release of nutrients, which are finally adsorbed into plasma via the 

intestinal walls (Wickham, Faulks, & Mills, 2009). 

2.1.1 Stomach 

The human stomach is a “J” shaped muscular bag, which is composed of four 

principal parts; the cardia, fundus, body, and pylorus. Figure 2.1 shows the anatomy 

picture of human stomach. After ingestion of a meal, the food enters the stomach through 

the oesophagus and mixes with the digestive juice, then the chyme is propelled to the 

small intestine by the pyloric sphincter. The solid or semisolid food is ground and broken 

down to 1-2 mm sized pieces under gastric peristalsis (Thomas, 2006; van Aken, 2010). 



 
 

Chapter 2: Literature Review 

6 
 

 

Figure 2.1. The external and internal anatomy of the stomach of human (Tortora & 

Derrickson, 2008). 

2.1.1.1 The pH and ionic strength of gastric fluid 

Typically, the pH of human gastric fluid during the fasting state varies from about 

pH 1 to 3 due to the presence of hydrochloric acid. Such a strongly acidic environment 

prevents the growth of microorganisms (N’Goma, Amara, Dridi, Jannin, & Carrière, 

2012). However, the pH value may alter with the nature and the amount of ingested food, 

and there may exist significant difference between individuals (McClements, Decker, & 

Park, 2008). Generally, after taking a meal, the pH value of the gastric content increases 

to about 5.5-7. After around 60 min of gastric digestion (i.e. the gastric half-emptying 

time), the pH reduces to about 4-5, and then it drops down to the initial acidic pH with 

the further gastric emptying, when all digesta leaves the stomach (N’Goma et al., 2012). 

This acidic condition causes alterations in food structures (e.g. the aggregation of protein 

or the coalescence of the lipid droplets) and activates the digestive enzymes (Gallier et 

al., 2013; McClements, Decker, & Park, 2008).  

In the fasted state, the typical ionic strength of the gastric fluid is around 100 mM, 

in which Cl- is the dominant ion, with a concentration about 100±30 mM (Lindahl, Ungell, 



 
 

Chapter 2: Literature Review 

7 
 

Knutson, & Lennernäs, 1997). However, the ionic strength may further change with the 

ingestion of food, due to the additional ions from food (Kalantzi et al., 2006). 

2.1.1.2 Enzyme 

The gastric secretion also contains different enzymes (i.e. pepsin and gastric 

lipase).  Protein may be partially hydrolysed by pepsin in the stomach, and about 10-30% 

of lipids may be digested to free fatty acids (FFAs) and diacylglycerol by the gastric lipase 

in the stomach. However, some proteins are not susceptive to catalytic action of pepsin, 

e.g. β-lactoglobulin (β-Lg) (Mandalari, Mackie, Rigby, Wickham, & Mills, 2009).  

For better understanding of human digestion, porcine pepsin has been widely 

utilised in in vitro digestion models (Hollebeeck, Borlon, Schneider, Larondelle, & Rogez, 

2013; Li, Ye, Lee, & Singh, 2013; Tan et al., 2017; Ye et al., 2016b). The porcine pepsin 

is secreted as a catalytically inactive pepsinogen from the hog stomach mucosa. The 

pepsinogen, with a molecular weight of 40,400, can be converted to pepsin when pH <5.0. 

This conversion process is catalysed by pepsin (Brown & Ernstrom, 1988). Pepsin is able 

to induce milk coagulation and hydrolyse the protein under acidic conditions. The 

optimum pH of activity/stability of porcine pepsin is about pH 2.0. The milk-clotting 

activity of porcine pepsin was found to be inhibited in cheese making when pH is above 

6.3. Normally, pepsin may lose its milk-clotting ability at pH 6.8 (Brown & Ernstrom, 

1988).  

2.1.2 Small intestine 

The small intestine is the major region where the nutrients are digested and 

converted to an absorbable form. About 70-90% of lipid digestion occurs in the small 

intestine (Singh & Gallier, 2014). The human intestinal tract is a complex environment 

consisting of bile salts, pancreatic enzymes, co-enzymes, various salts, phospholipids, 

yeasts and various bacteria (Singh & Ye, 2013). 

2.1.2.1 pH and ionic strength 

When partially digested food products pass into small intestine, the pH undergoes 

a rapid increase from the highly acidic environment (pH 1-3) in the stomach to the neutral 

environment (pH 6-7.5) in the duodenum, due to the secretion of sodium bicarbonate. 

This neutral pH provides an optimal environment for the action of pancreatic enzymes 



 
 

Chapter 2: Literature Review 

8 
 

(Golding & Wooster, 2010; Hur, Decker, & McClements, 2009; McClements, Decker, & 

Park, 2008). The increase in pH may cause some critical changes in physicochemical 

properties of protein-stabilised emulsions, e.g. a reversal of the protein charge. Most 

protein-stabilised emulsions exist in anionic form, and will undergo aggregation when the 

pH is near the isoelectric point region (Singh & Ye, 2013).  

Besides, it has been found that the osmolality of the duodenum contents is ~ 180 

mOsm/kg during the fasted state, and the ionic strength is ~140 mM (Kalantzi et al., 2006; 

Lindahl et al., 1997). However, because of the presence of various ions and solutes in 

ingested food, postprandial osmolality and ionic strength may undergo an appreciable 

increase. For example, Kalantzi et al. (2006) has reported that, after ingestion of a 

nutrition beverage, the osmolality of the duodenum had a rapid increase to ~290 

mOsm/kg. In addition, the ionic strength in small intestine is known as a particularly 

significant factor that affects the electrostatic interactions of this system. The multivalent 

cations (e.g. Ca2+, Mg2+) may reduce the digestibility of long chain saturated fatty acids 

and bile salts by the formation of precipitations (Karupaiah & Sundram, 2007; Reid, 2004; 

Vaskonen, 2003). 

2.1.2.2 Bile salts  

Bile salts play a significant role in both the digestion and adsorption of lipids due 

to its high surface activity (Sarkar et al., 2016). Bile salts are present in the small intestine 

and, originates from the liver through the gall bladder (Singh & Ye, 2013). Bile salt is a 

native biosurfactant (Golding & Wooster, 2010). Unlike other surfactants, it does not 

contain a hydrophobic head and hydrophilic tail group. Instead, its amphiphilic nature is 

mainly because of its flat steroidal structure, with methyl groups on the convex side and 

polar hydroxyl groups on the concave side (Euston, Baird, Campbell, & Kuhns, 2013; 

Galantini et al., 2015). Bile salts can adsorb readily at the oil-water interface in an 

emulsion and displace the initial surfactants at the oil droplet surface when is introduced 

to a simulated intestinal fluid (Hur et al., 2009). Therefore, bile salts facilitate the 

digestion of lipids by providing the accessibility of the lipase/co-lipase complexes to the 

bile-coated lipid droplets (Sarkar et al., 2016). Besides this, bile salts can facilitate the 

deformation of oil droplets under mechanical agitation, which enhances the stability 

against aggregation and transports the hydrophobic substance by forming micelles 



 
 

Chapter 2: Literature Review 

9 
 

(McClements, Decker, & Park, 2008).  It has been reported that bile salts can improve the 

digestibility of both adsorbed and unadsorbed proteins in an emulsion, e.g. β-Lg-, 

myoglobin- and bovine serum albumin (BSA)-stabilised emulsions (Gass, Vora, 

Hofmann, Gray, & Khosla, 2007).  

2.1.2.3 Pancreatic lipase 

For healthy human adults, the digestion of lipids by the gastric lipases is very 

limited in the stomach. It mainly (~70-90%) takes place in the small intestine, which is 

catalysed by pancreatic lipases (Bauer, Jakob, & Mosenthin, 2005; Mun et al., 2006; 

Singh & Ye, 2009). Pancreatic lipases can work efficiently at about pH 6.5 (Singh & Ye, 

2013), although their optimum pH is 8-9 (Patton & Carey, 1981). When the partially 

digested lipid droplets pass into the small intestine, the pancreatic lipase tends to adsorb 

at the droplet surface as a complex with co-lipase and/or bile salts (Bauer et al., 2005).  

Triglycerides and diacylglycerol are then broken down to free fatty acids (FFAs) and 2-

monoglycerides. It is reported that pancreatic lipase has no obvious specificity for the 

chain length of fatty acids, but it is preferential to cleave the positions of sn-1 and sn-3 in 

the fatty acids (Mu & Høy, 2004). This complexation reaction of pancreatic lipase at lipid 

droplet interface requires the presence of bile salts, co-lipase and calcium (Hur, Lim, 

Decker, & McClements, 2011). Its degree of binding appears to depend on the electrical 

charge of the interface and the competitive adsorption with bile salts, digestion products 

or other surfactants (McClements, Decker, & Park, 2008). The co-lipase is a kind of 

coenzyme, which is essential to the action of lipase. It interacts with lipase to form a 

stoichemetric complex that adsorbs at the oil-water interface, and thereby facilitates the 

accessibility to the lipid substrate. Co-lipase consists of a hydrophilic group that combines 

with lipase, and a hydrophobic part that connects with the interfacial layers (Bauer et al., 

2005).  

Interestingly, the influence of the presence of bile salts on the activity of 

pancreatic lipase is complex. When bile salts present in a relatively low concentration, 

they tend to solubilise the products of lipid digestion, such as free fatty acids (FFAs) and 

2-monoglycerides, and remove them from the interfacial layers. In that situation, they 

accelerate the activity of pancreatic lipids. In contrast, a relatively high concentration of 

bile salts will restrain the digestive ability of pancreatic lipase, which is mainly due to the 



 
 

Chapter 2: Literature Review 

10 
 

competitive adsorption between bile salts and lipases (Gargouri, Julien, Bois, Verger, & 

Sarda, 1983). 

2.2 Milk protein  

Milk, serving as an important protein source in the human diet, contains about 3.3% 

(w/w) protein. Milk protein is divided into two distinct groups: caseins, and whey proteins 

(serum proteins). Whey proteins are the remainder in the solution after the caseins are 

precipitated by acid or rennet (Oftedal, 2013). The main components of milk proteins are 

shown in Figure 2.2. 

 

Figure 2.2. Main components of milk proteins showing partitioning into casein and whey 

fractions (Cheison & Kulozik, 2017). 

 

2.2.1 Casein 

Caseins are abundant in milk and constitute about 80% in total milk protein. It  

can be precipitated from raw milk by adjusting pH to 4.6 at 20°C (Swaisgood, 1992). The 

caseins of bovine milk can be fractionated into four main classes: αs1-, αs2-, β-, and κ-

caseins (Varnam & Sutherland, 2001). Caseins are phosphoproteins and are generally 

considered to be very hydrophobic proteins, with the exception of β-caseins (O’Mahony 

& Fox, 2013; Swaisgood, 1992). αs-Caseins are very sensitive to calcium due to the 

existence of phosphate groups, and may precipitate at pH 7.0 when calcium ions are 



 
 

Chapter 2: Literature Review 

11 
 

present (Swaisgood, 1992).  β-Caseins have high surface hydrophobicity, but they are not 

exceptionally hydrophobic due to lacking stable secondary and tertiary structures 

(O’Mahony & Fox, 2013). β-Caseins have been reported to have a strong tendency to 

bind metal ions, i.e., calcium ions in the milk, due to their high content of phosphate 

groups (O’Mahony & Fox, 2013). κ-Casein on the other hand,  is calcium insensitive 

(Huppertz, 2013), which only contains one phosphoserine group (Varnam & Sutherland, 

2001). 

Caseins are susceptible to proteolysis due to their open and flexible structures, 

which is crucial for their nutritional function and serves as an important source of amino 

acids (O’Mahony & Fox, 2013). Caseins lack stable tertiary structures; consequently, they 

are highly thermally stable (Fox, 1981a). Besides, caseins have a strong tendency to 

associate in aqueous solutions due mainly to their high hydrophobicity (Ye, Srinivasan, 

& Singh, 2000). 

2.2.1.1 Casein micelle structure  

In milk, about 95% of caseins exist in colloidal particles known as casein micelles, 

rather than as individual molecules (O’Mahony & Fox, 2013). The structure and 

properties of casein micelles have attracted much scientific interest (Holt, 1992; Walstra, 

1990). The knowledge of the structure of casein micelles is essential to understand the 

digestion behaviours of milk in the gastrointestinal tract and nutrient delivery.  

The casein micelles are highly hydrated, possessing about 2.0-4.0 g H2O/g protein. 

The dry matter of casein micelles is composed of protein (94%) and colloidal calcium 

phosphate (CCP, 6%) (O’Mahony & Fox, 2013). The casein micelles are generally 

spherical in shape, with an average diameter of 120 nm. Some of the main properties of 

casein micelles are summarised in Table 2.1.  

 

 

 

 

 



 
 

Chapter 2: Literature Review 

12 
 

 

Table 2.1.  Average characteristics of casein micelles (Fox, 2003). 

Characteristic Value 
Diameter 130-160 nm 
Surface 8×10-10 cm2 
Volume 2.1×10-15 cm3 
Density (hydrated) 1.0632 g/cm3 
Mass 2.2×10-15g 
Water content 63% 
Hydration 3.7 g H2O/g protein 
Voluminosity 4.4 cm3/g 
Molecular weight (hydrated) 1.3×109 Da 
Molecular weight (dehydrated) 5×104 Da 
Number of peptide chains  
(MW: 30,000 Da) 

104 

Number of particles per mL milk 1014-1016 
Whole surface of particle 5×104 cm2/mL milk 
Mean free distance 240 nm 
 

Many models of the casein micelle structure proposed in the past 50 years were 

inconsistent and often contradictory, including three main categories: core-coat, internal 

structure, and submicelles (Dalgleish, 2011; McMahon & Brown, 1984; McMahon & 

Oommen, 2008; Rollema, 1992; Rose, 1969; Slattery & Evard, 1973). Numerous earlier 

models proposed that the casein micelle is made up of many smaller sub-micelles, with a 

size of about 10-15 nm, and a molecular weight about 106 Da (Figure 2.3). Morr (1967) 

first proposed this kind of model in 1967. The CCP is believed to provide the link between 

sub-micelles, affording an open, porous structure. The CCP may be removed by using 

some reagents, e.g. EDTA, oxalate, citrate, urea (˃ 5M) and, ethanol (35% at 70°C), or 

by increasing the pH to above pH 9. As a result, the casein micelles structure is disrupted 

(O’Mahony & Fox, 2013).  

 



 
 

Chapter 2: Literature Review 

13 
 

 

Figure 2.3. The casein micelle schematic diagram (Walstra & Jenness, 1984). 

 

However, the theory of the existence of sub-micelles has been controversial. Hill 

and Wake (1969) suggested that the amphiphilic structure of κ-caseins plays a major role 

in stabilising the micellar structure. The information known about the casein micelle 

structure has continued to be updated and elaborated by scientists with the development 

of analytical approaches. Walstra and Jenness (1984) and Walstra (1999) assumed that 

there is a hairy layer with a thickness of 5-10 nm, coated on the surface of micelles, which 

is composed of a hydrophilic C-terminal area of κ-caseins. The primary responsibility of 

this hairy layer is to stabilise the casein micelle through steric stabilisation. However, the 

stabilisation of casein micelles may be disrupted when the hairy layer is collapsed by 

adding ethanol, or being removed through specific proteases, e.g. rennet or pepsin (Holt 

& Horne, 1996). Holt (1992) envisaged the casein micelle structure as a crosslinked gel 

network, which is built by relatively flexible casein molecules. The CCP micro-granules 

are responsible to keep the stability of the gel-like structure, and the surface of the 

micelles is covered with a protruding hairy layer, which is comprised of the C-terminal 

end of the κ-casein. 

2.2.1.2 The stability of casein micelles  

The micelle is considered to be a very stable system.  They are stable to high 

temperature, compaction, commercial homogenisation and in the presence of high levels 

of calcium ion concentration (O’Mahony & Fox, 2013). However, casein micelle 



 
 

Chapter 2: Literature Review 

14 
 

properties can be changed under high-pressure processing, low temperature (0-5°C), and 

freezing. Moreover, in the concentration process of dairy products, casein micelles can 

be destroyed to some extent by evaporation, ultrafiltration and spray-drying (Fox & 

Brodkorb, 2008; Havea, 2006).  

There is no covalent bonding between CCP and proteins. The acidic pH, specific 

proteolytic enzyme, and ethanol can result in different degrees of destabilisation of casein 

micelles. Lowering the pH to casein’s isoelectric point (pI=4.6), the polymerisation and 

precipitation will take place. The CCP can be fully removed when the pH is equal to or 

lower than 4.9. For acid-induced destabilisation, most of the CCP may be removed 

without destruction of the micellar structure, because casein micelles are probably rebuilt 

by increasing the pH to 6.7 under certain conditions. Besides this, some milk-clotting 

enzymes, e.g. chymosin and pepsin, can specifically catalyse the hydrolysis of κ-casein, 

which is split into para-κ-casein and macropeptide. As a result, the micellar caseins will 

coagulate and form a curd-like gel in the presence of calcium or other divalent ions (Lucey, 

2011). These properties are significant in the digestion of milk, as both acidic conditions 

and proteases pepsin exists in the stomach.    

2.2.1.3 Enzymatic coagulation of caseins 

Milk-clotting is believed to be a complex process, involving a primary enzymatic 

hydrolysis (first stage) that causes the alteration of κ-casein and a loss of its ability to 

stabilise the rest of the caseinate complex. Second, the aggregation of the altered caseinate 

takes place, which is a non-enzymatic step. Then the aggregates of caseins further form a 

firm cross-linked milk gel, and probably curd syneresis occurs (McMahon & Brown, 

1984). 

In the first stage (i.e. enzymatic hydrolysis) of the coagulation of milk, the 

proteases cleave the κ-casein molecules that are on the surface of casein micelles into 

para-κ-casein and a macropeptide, and hence initiate the milk clotting process: 

κ-casein → para-κ-casein + macropeptide 

The hydrolysis rate is proportional to the enzyme concentration and is dependent 

on the pH (with an optimal pH at ~5.6) (Carlson, Hill, & Olson, 1987). Chymosin is 

capable of hydrolysing  κ-casein (Holt & Horne, 1996), and uniquely cleaves the Phe105-



 
 

Chapter 2: Literature Review 

15 
 

Met106 peptide bond, while other milk-clotting enzymes (e.g. pepsin) are less specific. 

However, all milk-clotting enzymes have the same general functions (Fox, 1981b). 

During the secondary step, the aggregation of altered micelles is due mainly to the 

loss of electrostatic repulsion caused by reducing the pH to its pI (pH~4.6), and the loss 

of steric repulsion of κ-caseins. The presence of calcium ions facilitates the formation of 

coagulum by connecting micelles as a bridge and inducing an isoelectric condition 

(Douglas, 2017). 

2.2.2 Whey protein 

Whey proteins occupy 20% of the total protein in bovine milk,  consisting 

principally of β-lactoglobulin (β-Lg), α-lactalbumin (α-La), bovine serum albumin (BSA), 

and immunoglobulins (Kinsella & Whitehead, 1989). Whey proteins have a relatively 

more ordered globular structure in comparison to caseins, which is stabilised by 

intramolecular disulphide bonds between cysteine residues (Dickinson, 2001). Whey 

protein is a group of acid-soluble proteins, which can be utilised over a wider pH range 

than caseins. While caseins are insoluble at their pI (Kinsella & Whitehead, 1989). β-Lg 

is applied as an emulsifier, thickener or foam stabiliser for nutritional purposes due to its 

excellent functional properties. While α-La serves as an important source of nitrogen in 

infant food because of its low allergy potential. (Dickinson, 2001; Foegeding, Davis, 

Doucet, & McGuffey, 2002; Khalloufi, Alexander, Goff, & Corredig, 2008; Ye, 2008).  

2.2.2.1 β-lactoglobulin 

β-Lactoglobulin (β-Lg) accounts for approximately 50% of whey protein, and 10% 

of total protein in milk. It is a major component of whey protein in bovine milk and tends 

to dominate the characteristics of whey protein products. β-Lg consists of 162 amino acids 

and its molecular weight is about 18.3 kDa (Hambling, McAlpine, & Sawyer, 1992). It 

exists as a dimer (MW=36.6 kDa) of two monomeric molecules linked by a non-covalent 

bond at neutral pH at room temperature (McKenzie & Sawyer, 1967). It contains one free 

sulfhydryl group at Cys121, and two disulphide (-S-S-) bonds at Cys66-Cys160 and 

Cys106-Cys119 (Walstra, Wouters, & Geurts, 2005). The secondary structure of β-Lg 

contains about 6-10% α-helix, 44-52% β-sheet, 8-10% reverse turn and 32-35% of 

random coil (Casal, Köhler, & Mantsch, 1988; Dong et al., 1996). Changes in the 

quaternary structure of β-Lg may occur when environmental conditions such as 



 
 

Chapter 2: Literature Review 

16 
 

temperature, pH and ionic strength are altered (McKenzie & Sawyer, 1967). pH can 

significantly affect the molecular conformation of β-Lg. Although at neutral pH, β-Lg 

exists as a dimer; when pH is higher than 6.8, or between 2 and 3, it dissociates into 

monomers (Sawyer, 2003). β-Lg has been reported to be resistant to digestion in the 

stomach by the action of pepsin due to its stability under acidic pH conditions (Peram, 

Loveday, Ye, & Singh, 2013). 

2.2.2.2 α-lactalbumin   

α-Lactalbumin (α-La) is the second major component in bovine whey protein. It 

is a compactly folded protein molecule of approximately spherical shape (Walstra et al., 

2005). Its isoelectric point is about pH 5.1, and it exists as a monomer at its isoelectric 

region (Nakai & Modler, 1996). α-La is resistant to enzymatic proteolysis, due to its 

compact globular structure. Numerous attempts have been made to increase its 

susceptibility to proteolysis, using methods including esterification, binding of zins ions 

to α-La, or lowering the pH to 2.0 (El-Zahar et al., 2005; Permyakov et al., 1991; Sitohy, 

Chobert, & Haertle, 2001). α-La has excellent heat stability.  Heating has no significant 

influence on the digestibility of α-La, as the protein will refold during cooling when the 

heating temperature is lower than 100°C (Schmidt & Poll, 1991; Schmidt & van 

Markwijk, 1993). 

2.2.2.3 Heating induced denaturation of whey protein 

Heat treatment of milk proteins is a common industrial processing procedure to 

ensure food safety. The three-dimensional structure of whey protein can be changed 

dramatically by extremes of temperature, whereas the highly stable caseins are not 

markedly affected by thermal treatment (Almaas et al., 2006; Barbé et al., 2013). When 

being exposed to heating above a certain temperature, whey proteins unfold, denature and 

aggregate, and form an open conformation (Brodkorb, Croguennec, Bouhallab, & Kehoe, 

2016). The thermal denaturation temperature of whey proteins is summarized in Table 

2.2. 

 

 

 



 
 

Chapter 2: Literature Review 

17 
 

Table 2.2. Thermal denaturation temperature of whey protein (De Wit, 1984). 

Protein TD (°C) Ttr(°C) 
β-lactoglobulin (β-Lg) 78 93 
α-lactalbumin (α-La)   62 68 
Bovine serum albumin (BSA) 64 70 
Immunoglobulin (Ig) 72 89 

TD represents the initial denaturation temperature. 

Ttr is the temperature at the DSC peak maximum. 

 

One of the earliest researches on denaturation and aggregation of whey protein 

was reported by Briggs and Hull (1945), who indicated that the denaturation process is 

made up of two steps: (1) protein unfolding, and (2) protein aggregation. This two-step 

process has been accepted for several years. In the first stage, the whey protein denatures 

through an unfolding step. The hydrophobic interactions between protein molecules 

increase due to the exposure of hydrophobic groups from the interior of the native 

molecule onto the molecular surface (Relkin, 1998).  Some of the free thiol groups and 

disulphide bonds in the β-Lg also become exposed during the initial unfolding step; hence, 

they are available to further form intramolecular disulphide bonds through oxidation of 

the free sulfhydryl groups or rearrangement of disulphide bonds (Chaudry & Humbert, 

1968). The denaturation of whey proteins in milk has been studied. Denatured β-Lg 

aggregates with κ-casein at the casein micelle surface via disulphide bonds in milk. This 

process varies greatly with the pH of the serum (Brodkorb et al., 2016). The association 

extent of β-Lg with casein micelle is predominantly dependent on the heating conditions. 

Corredig and Dalgleish (1996) reported that heating milk at 75°C to 90°C for 80 min 

allows all β-Lg to form complexes with casein micelles via κ-casein binding; when 

heating at 130°C for 100 s, all β-Lg denatures but only half of them associates with κ-

caseins (Oldfield, Singh, & Taylor, 2005). Under these heating conditions, all the β-Lg is 

denatured but it does not all form aggregates with the micellar κ-casein. Sava, Van der 

Plancken, Claeys, and Hendrickx (2005) indicated that the aggregation reaction of β-Lg 

is initiated when heating up to 78°C, but β-Lg only unfolds when heating is below 78°C. 

In the second stage, the main irreversible aggregation of β -Lg occurs via covalent 

linkages between thiol groups of cysteine residues. After denaturation, β-Lg  aggregates 

initially into oligomers and subsequently into soluble aggregates, finally generating a 



 
 

Chapter 2: Literature Review 

18 
 

large insoluble colloidal particle under further heating (Baussay, Le Bon, Nicolai, Durand, 

& Busnel, 2004; Pouzot, Nicolai, Visschers, & Weijers, 2005). The formation of 

aggregations can be affected by the heating temperature (Sawyer, 1968). Moreover, the 

free sulfhydryl group also plays an key role in the initial period of aggregation when ionic 

strength is low and pH is higher than 7.0 (Hoffmann & van Mil, 1997).  

By contrast, pure α-La does not aggregate under a mild thermal treatment (e.g. 

80°C, pH 6.7) (Calvo, Leaver, & Banks, 1993; Hines & Foegeding, 1993), which is 

primarily due to the absence of free thiol groups (Singh & Havea, 2003), although its 

initial denaturation temperature is lower than that of β-Lg.  

2.2.3 Milk protein products 

A wide range of milk protein products are produced worldwide and utilised as 

functional ingredients in many foods as shown in Figure 2.4. SMP, MPC and sodium 

caseinate are dairy ingredients where casein is the major protein component (Hemar, 

Tamehana, Munro, & Singh, 2001). The micellar structure of casein is retained in SMP 

and MPC, but it is destroyed during the manufacture of sodium caseinate (Kinsella & 

Morr, 1984; Mulvihill, 1992). Whey protein isolate (WPI) contains ≥90% protein, which 

is derived from the cheese whey by ion-exchange chromatographic methods (Mulvihill, 

1992). 

 



 
 

Chapter 2: Literature Review 

19 
 

 

Figure 2.4. Functional protein ingredients manufactured from skim milk (derived from 

Singh (2005)) 

2.2.3.1 Skim milk powder 

Milk powders are defined as “milk products that can be obtained by the partly 

removal of water from milk” by the Codex Standard 207-1999 (Codex Alimentarius 

Commission, 1999). SMP and whole milk powder (WMP) are two primary types of 

commercial milk powder. SMP is the most widely applied as a functional dairy ingredient 

(Singh & Creamer, 1991). The manufacture of SMP involves heat treatment (normally 

known as preheating), evaporation, and spray drying. SMPs are normally classified as 

low-, medium- and high-heat powder according to the whey protein nitrogen index 

(WPNI). SMP with different preheat treatment conditions designed to meet the specific 

needs of the food industry has gained wide acceptability in different types of food 

applications (in Table 2.3). The heating conditions play an essential role in determining 

the functional properties of milk powder. The most important effect is the induction of 

denaturation of whey protein, achieving a partially denatured whey protein, which could 

simply self-aggregate and/or associate with casein micelles via micellar κ-casein (Singh, 

2007). 

 



 
 

Chapter 2: Literature Review 

20 
 

Table 2.3. Food applications of skim milk powder of different heat classes (Kelly & Fox, 
2016). 

Heat 
classification 

Heat treatments 
typically applied WPNI Functional 

properties 
Food 
applications 

Low heat 70°C for 15 s >6.0 mg/ml 
Solubility, lack 
of cooked 
flavour 

Recombined 
milk, milk 
standardisation, 
cheese making 

Medium heat 

85°C for 1 min 

1.5-6.0 mg/ml 

Foaming, 
emulsification, 
flavour, colour, 
viscosity 

Chocolate,  
ice cream, 
confectionery 

90°C for 30 s 
105°C for 30 s 
 

High heat 
90°C for 5 min 
120°C for 1 min 
135°C for 30 s 

<1.5 mg/ml 
Gelation, heat 
stability, water 
adsorption 

Recombined 
evaporated milk 
 

High-high heat  <1.5 mg/ml Colour, flavour, 
water binding 

Recombined 
evaporated milk, 
bakery 

WPNI, whey protein nitrogen index. 

In milk powder, the micellar structures of caseins largely retain their integrity 

during processing, conferring a number of unique functional properties, such as heat 

stability, curd formation and emulsification. Whey proteins may already be somewhat 

denatured by the process of preheating, depending on the heat treatment conditions (Singh, 

2007). The interaction of denatured whey proteins with κ-casein at the surface of the 

casein micelles may modify the properties of casein micelles. This interaction may impair 

the clot formation and increase the rennet coagulation times, thereby milk powder is not 

suitable for the preparation of recombined cheese milk (Singh, 2007). The denatured 

whey protein-casein micelle complexes have a high water-binding capacity (Schkoda, 

1999), which increases the gel strength and viscosity of yoghurt and other fermented milk 

products (Singh, 2007). Milk powder is inferior to caseinate in the emulsifying property 

due to the micellar structure of caseins in milk powder (Mulvihill & Murphy, 1991). 

However, Euston and Hirst (1999) suggested that the stability of milk powder-stabilised 

emulsions against creaming is superior than caseinate-stabilised emulsion systems, which 

is probably because of the formation of a weak gel network or the lower susceptibility to 

depletion flocculation in milk powder-stabilised system. Singh and Newstead (1992) 

reported that milk powder exhibits better emulsifying properties than sodium caseinate 

when pH is close to 5.2, which may be due to the dissociation of casein micelles around 

this pH value. 



 
 

Chapter 2: Literature Review 

21 
 

2.2.3.2 Milk protein concentrate 

Milk protein concentrate (MPC) was the earliest casein-based product obtained 

by membrane technique on the market (Carr & Golding, 2016). MPCs are considered an 

essential source of calcium, and are widely applied as an ingredient for milk extension in 

cheese manufacture, nutritional beverages, frozen desserts and yoghurt production 

(Huffman & Harper, 1999; Ye, 2011). MPC has protein content in the range of 

approximately 50-85%. It is processed directly through the partial removal of  lactose and 

soluble salts from skim milk using a combined ultrafiltration/diafiltration process 

(Mulvihill, 1992; O’Donnell & Butler, 1996) prior to evaporation and spray drying. In 

MPCs, casein is in a micellar form as that found in milk, casein and whey protein ratio 

remains as in milk (about 80:20) (Carr & Golding, 2016), and is similar to that in SMP 

(Singh, 2007). However, a fraction of CCP in casein micelle may be dissolved during the 

ultrafiltration and diafiltration process, which leads to the loose structure of casein 

micelles, resulting in a smaller fragmented micellar structure. Whey proteins in MPC 

remains largely in their native state, as the manufacturing process does not involve a 

preheat treatment (Singh, 2007).  

Recently, in order to improve the functional properties (e.g. solubility) the casein 

micelles in some new MPC products have been dissociated to a certain extent by 

removing the calcium content (Ye, 2011). This type of product refers to “MPC 4861” in 

the present study. MPC with a micellar structure can be produced from a native micelle 

structure and altered to a structure closer to sodium caseinate, depending on the level of 

calcium depletion (Carr & Golding, 2016). However, a significant difference between 

MPCs and caseinate is that MPCs contain phosphate, whereas caseinates contain a 

reduced level of phosphate because of the acidic precipitation and subsequent washing. 

Micellar phosphate levels can increase buffer capacity because of the formation of 

dihydrogenphosphate (Ferreira, Oliveira, & Rocha, 2003) on acidic solubilisation of CCP 

(Carr & Golding, 2016). 

2.2.3.3 Sodium caseinate 

Traditionally, caseins are separated using precipitation techniques such as rennet 

or acid precipitation (Mulvihill, 1992), which destroy the native micellar structure, and 

the products produced in this manner contain individual casein molecules. Sodium 



 
 

Chapter 2: Literature Review 

22 
 

caseinate is one of the most commonly used ingredients in foods. It is the water-soluble 

form of casein, which is prepared by solubilising acid casein with NaOH (Mulvihill & 

Ennis, 2003). The casein curd separated from milk by acidification to pH 4.6 is further 

separated through a process of dewheying, and is then washed to remove whey protein, 

minerals, lactose and residual acid. Next, the casein curd from a de-watering device is 

minced and mixed with NaOH, followed by a process of spray drying to form caseinate 

(Carr & Golding, 2016; Mulvihill & Ennis, 2003). The excellent heat stability is one of 

the most important functional benefits of the caseinates, which limits modifications to 

product properties in consequence of thermal processing (Carr & Golding, 2016). 

2.2.3.4 Whey protein isolate 

Whey is the liquid remaining after removing caseins from milk. Whey protein 

isolate (WPI) is a concentrated form of whey protein component, which is highly soluble, 

with a high level of protein concentration that is above 90% (Bansal & Bhandari, 2016). 

WPI is processed from ultrafiltration (UF) and sometimes diafiltration (DF), or 

microfiltration, or can be obtained from whey by ion-exchange (IX) chromatography (Fox 

& McSweeney, 1998). Whey protein in WPI is in its native state; thus it retains, to a great 

extent, its functional properties. WPI is widely applied in the food industries, because of 

its high protein content, excellent water-binding capacity, emulsification, gelling and 

foaming properties (Singh, 2005). 

2.2.4 The digestion behaviours of milk protein during gastric digestion 

The digestion behaviour of milk proteins has recently been the subject of many 

investigations in in vitro and in vivo (Dangin et al., 2001; Mahé et al., 1995; Ye et al., 

2016a, b, 2017). The different milk protein has a different hydrolysis rate induced by 

pepsin and can cause different gastric emptying rates. It has been reported that caseins 

caused a delayed delivery of amino acid to the small intestine in an in vivo digestion 

(Mahé et al., 1996). Thus, casein is considered as a “slow” digested protein. This “slow” 

digestion of casein can be explained by that casein micelles coagulate under gastric 

conditions, and remain for several hours in the stomach. Casein micelles have been 

reported as able to be coagulated by the milk-clotting enzyme pepsin (Tam & Whitaker, 

1972) and acidic pH (Dalgleish & Corredig, 2012). Both them exist in the stomach. 

However, caseinate is only coagulated by acid and not by pepsin, because of its existence 



 
 

Chapter 2: Literature Review 

23 
 

in the form of individual casein molecules. This digestion behaviour of casein has been 

observed in an in vivo study (Miranda & Pelissier, 1981). The gastric emptying rate and 

hydrolysis rate of a mixture of individual caseins in a rat stomach was faster than that of 

casein in the skim milk samples (Miranda & Pelissier, 1981). On the contrary, whey 

protein is considered as a “fast” digested protein. It remains soluble in the presence of 

pepsin at acidic conditions and can pass rapidly from stomach into small intestine without 

being hydrolysed by digestive enzymes. Thus, this enables a fast delivery of proteins to 

the small intestine, where it is further digested and absorbed (Boirie et al., 1997).  

Previous studies have shown that the digestion behaviour of protein in the 

gastrointestinal tract can be affected by its structure (Kitabatake & Kinekawa, 1998; 

Schmidt & van Markwijk, 1993; Tunick et al., 2016; Zeece, Huppertz, & Kelly, 2008). 

The structure of protein in food is dependent on the processing treatment and its source. 

Different proteins have different conformational properties; therefore, the nature of the 

protein may markedly affect its susceptibility to proteolysis by pepsin. Casein has a highly 

flexible, disordered conformation (Modler, 1985). It is more exposed to gastric hydrolysis 

by pepsin in the stomach (Mahé et al., 1996). In contrast, β-Lg is highly resistance to 

pepsin action due to its compact, folded tertiary structure. However, the molecular 

flexibility and the susceptibility to proteases can be modified by physical processing (e.g. 

emulsification, exposure under high-pressure, or heating) (Peram et al., 2013; Zeece et 

al., 2008). For example, heat treatment can improve the availability of enzymatic cleavage 

sites of β-Lg, and lead to an increase in the susceptibility to proteolysis by pepsin (Barbé 

et al., 2013; Li et al., 2013; Peram et al., 2013; Schmidt & van Markwijk, 1993). Miranda 

and Pelissier (1987) reported that heat treatment can accelerate gastric emptying of milk 

and increase the hydrolysis rate of caseins in an in vivo digestion study.  

Recently, studies in our lab showed the gastric digestion behaviour of unheated 

and heated skim milk (90°C for 20 min) using a dynamic digestive model-Human Gastric 

Simulator (HGS) (Ye et al., 2016b). The results provide a novel insight into the influence 

of the structured clot on the protein hydrolysis rate. Ye et al. (2016b) found that unheated 

skim milk formed a firm, dense, cheese-like clot with a porous network structure. Such a 

structure prevented the accessibility of the pepsin to the interior of the clot. In contrast, 

the heated skim milk, in which whey protein had been denatured, formed a looser, 

fragmented network-structured clot with numerous larger voids. With further digestion, 



 
 

Chapter 2: Literature Review 

24 
 

the curd became more tightened and less permeable to serum and solute with increasing 

digestion time, in particular, in unheated skim milk. The hydrolysis rate of caseins and 

whey proteins by pepsin varied from the formation of different structured clots under the 

gastric conditions. The hydrolysis of caseins in heated milk became much faster than that 

in unheated milk, which is mainly because of the formation of the clot with an open, 

fragmented structure. Such a structure increased the effective contact area with simulated 

gastric fluid (SGF, containing pepsin). Pepsin was thus prone to diffuse and act on the 

clot. Additionally, the denatured whey protein in heated milk was hydrolysed rapidly 

while β-Lg and α-La were maintained intact during the whole digestion period in 

unheated milk. Besides, in the heated milk, casein and serum proteins can be observed in 

the coagula, but the emptied digesta did not contain intact casein and whey proteins. This 

further confirmed that the digestibility of casein and β-Lg can be considerably improved 

by thermal treatment (Barbé et al., 2013).  

The gastric digestion of whole milk has also recently been investigated in our lab 

with respect to the effect of pre-treatment on the behaviour of protein and fat globules 

during dynamic digestion (Ye et al., 2016a, 2017). It was found that the rates of protein 

hydrolysis and the release of milk fat globules from the curd into the small intestine can 

be modified by different pretreatment (e.g. homogenisation and heat treatment) prior to 

the digestion of milk. This is because homogenisation and heating treatments affect the 

formation of structured clot in the stomach. The curds produced in the homogenised milk 

and heated homogenised milk had a more crumbled and fragmented structure than that 

produced in the raw whole milk. The fat globules were embedded in the curds as they 

generated. After the formation of the curd, a further quantity of voids was found in the 

structured clots formed from the homogenised milk and heated milk, which gave rise to 

a greater rate of protein hydrolysis by pepsin. This led to a more rapid release of 

incorporated fat globules from the curds into the digesta. The formation of clots with 

different structures cause the changes in the rate of protein hydrolysis and release of milk 

fat into the digesta in the stomach (Ye et al., 2017). 

2.3 Emulsion  

An emulsion refers to an intimate dispersion of at least one immiscible liquid in 

another in the form of discrete droplets (McClements, 2005). In most foods, the droplet 



 
 

Chapter 2: Literature Review 

25 
 

size of an emulsion is usually ranged between 0.1 and 100 μm in diameter (Dickinson, 

1992; Dickinson & Stainsby, 1982; Friberg & Larsson, 1997). An emulsion consists of a 

discontinuous (dispersed or internal) phase and a continuous (external) phase. The former 

is made up of droplets, while the latter refers to the surrounding liquid. According to 

different spatial distribution of water and oil phases, emulsions can be divided classically 

into two types. An emulsion composed of oil droplets dispersed in an aqueous phase is 

referred to as an oil-in-water emulsion (e.g. milk, mayonnaise, beverages, sauce and 

cream). An emulsion made up of water droplets dispersed in an oil phase is referred to as 

a water-in-oil emulsion (e.g. butter and margarine) (McClements, 2005). 

2.3.1 Emulsion formation 

Normally, when pure water and pure oil are poured into a container, a layer of oil 

(lower density) will be rapidly separated onto the top on the layer of water (higher density), 

due to their tendency to arrangement in the most thermodynamically stable state. Make 

an emulsion requires an intense shear and the presence of emulsifiers. The intense shear 

may increase the interfacial area between oil and water phases, leading the dispersed oil 

phase into tiny oil droplets. The shear may be provided through mechanical agitation, 

such as high pressure valve homogenisers, high speed blenders and colloid mills (Singh 

& Ye, 2013). Homogenisation is known as the process of transformation of two separate 

liquid immiscible phases (usually oil and water) into an emulsion, or of reducing the 

droplet size of a premixed emulsion. An emulsion can be kinetically stable for a period 

of time when the stabilisers are present. An emulsifier or a texture modifier (e.g. 

thickening agents) can be a stabiliser to improve the stability of the emulsion. An 

emulsifier is a surfactant that can adsorb onto newly formed emulsified droplets and form 

a protective membrane against aggregation and coalescence of droplets (McClements, 

2005). Moreover, it contributes to lowering the interfacial tension and Laplace pressure 

and, hence facilitates to break the droplets into smaller ones (Walstra, 2003). 

Emulsion formations may include a single step or several consecutive steps, 

depending on the methods employed to make it, the nature of the ingredients and the 

desired droplet size. Homogenisation is conveniently classified into two categories; 

primary, and secondary homogenisation. The primary homogenisation is the process of 

converting two separated liquids into a coarse emulsion that involves some fairly large 



 
 

Chapter 2: Literature Review 

26 
 

droplets by using a homogeniser (e.g. high-speed blender). The secondary 

homogenisation is employed to reduce the droplet size of an already existing emulsion by 

using another type of homogeniser; for example, a two-stage homogeniser (Walstra, 

2003).  

2.3.2 Emulsion stability 

The term emulsion stability is known as the ability to an emulsion to resist 

modifications in its properties over time (Dickinson, 2003; McClements, 1999). For food 

emulsion systems, because they are thermodynamically labile systems, the emulsion 

breakdown will eventually take place with the passage of enough time (McClements, 

2005). Even though some emulsion systems appear fairly stable, so that the product has 

a long shelf life. In fact, the total number of droplets, the size distribution and the spatial 

arrangement are always changing. 

The destabilisation of an emulsion normally involves two aspects: physical 

instability, and chemical instability. The most important physical instability mechanisms 

include creaming, flocculation, coalescence, phase inversion, and Ostwald ripening. It 

may change the spatial distribution of oil and water phases, or lead to the structural re-

organisation of molecules (Dickinson, 1992; Dickinson & Stainsby, 1982; Walstra, 1996, 

2003). Chemical instability normally involves oxidation, or hydrolysis, and results in the 

generation of new molecules in the system (Fennema, 1996). The extent of destabilisation 

of an emulsion during its lifetime is dependent on its composition, microstructure and 

external environment (e.g. storage conditions and temperature) (McClements & Decker, 

2000). 

The stabilisation and destabilisation of protein-stabilised emulsions is driven by a 

number of factors, such as the strength and type of interactions occurring between 

droplets (e.g. van der Waals attractive forces, electrostatic interaction, steric repulsion, 

hydration forces), and these factors are dependent on the composition, structure, and 

concentration of adsorbed layer (Leman, Kinsella, & Kilara, 1989).  

2.3.2.1 Gravitational separation 

Generally, in an emulsion, the droplets in a discontinuous phase do not have the 

same density as the liquid surrounding them. Therefore, a net gravitational force has an 

essential influence on the stability of the system (Dickinson, 1992; Dickinson & Stainsby, 



 
 

Chapter 2: Literature Review 

27 
 

1982; Hunter, 1989; Walstra, 1996). Gravitational separation involves creaming and 

sedimentation. Creaming refers to the upward movement of emulsified droplets without 

an alteration in droplet size, because the density of droplets is lower than the density of 

liquid that surrounds them (McClements, 2005; Walstra, 1987). On the contrary, when 

the surrounding liquid has a lower density, the droplets have a tendency to move down, 

which is known as sedimentation. Creaming is more common than sedimentation in a 

food emulsion system. The density of water is higher than that of most liquid oils. In the 

case of oil-in-water emulsion, oil droplets tend to cream and suspend on the top in the 

system when the water goes to the bottom (McClements, 2005).  

The creaming rate of an isolated rigid spherical droplet in an ideal liquid is 

descried in Stockes’ law equation (Equation 2.1), which is determined by the balance of 

frictional force and gravitational force. This equation is applied in food emulsion to 

estimate the stability against creaming: 

= − 2 ( − )9  

Where υStokes is the velocity of creaming, g is the acceleration due to gravity, r is 

the radius of the emulsion droplet, ρ1 is the density of the continuous phase, ρ2 is the 

density of discontinuous phase and η1 is the shear viscosity of continuous phase 

(McClements, 2005).  

The creaming rate can be affected by a number of factors, such as droplet 

flocculation, droplet fluidity, electrical charge of droplets, fat crystallisation and 

Brownian motion (McClements, 2005). 

The gravitational separation is an instability problem, which accelerates the 

process of flocculation or coalescence, and eventual oiling off. A number of methods can 

be used to control gravitational separation, such as reducing the particle size, minimising 

the density difference between the suspended droplets and the liquid phase, increasing 

the droplet concentration and modifying the rheology of the continuous phase using a 

thickening agent (McClements, 2005). 

 

 



 
 

Chapter 2: Literature Review 

28 
 

2.3.2.2 Flocculation 

Flocculation is the process in which destabilised suspended droplets associated 

with each, but maintain their individual integrity (McClements, 2005; Tadros & Vincent, 

1983). In diluted emulsion, flocculation accelerates the rate of gravitational separation, 

and therefore, its shelf life is drastically reduced (Luyten, Jonkman, Kloek, & Van Vliet, 

1993; Tan, 2004). It can also lead to a great increase of the viscosity of food emulsion, 

and may even accelerate the development of a gel network (Demetriades, Coupland, & 

McClements, 1997). Although flocculation modifies the psychical properties of an 

emulsion, the droplet size may maintain unaltered and the flocs may be dispersed because 

the interaction force is weak (Walstra, 1987). 

The following mathematical model may describe the droplets flocculation rate in 

a colloidal system that includes monodisperse globose particles (Evans & Wennerström, 

1999): 

= − 12  

Where dnT/dt is the droplet flocculation rate, t is time, nT is the total amount of 

droplets per unit volume, F represents collision frequency and E represents collision 

efficiency. The factor ½ refers to the collision between two droplets leading to one droplet 

decreasing in the total amount of droplets.  

According to the above equation, the flocculation rate is dependent on the 

collision frequency between particles and the collision extent that causes aggregation. 

The collision frequency refers to the total amount of particles in contact with their 

neighbours per unit time per unit volume of emulsion. Molecular movement can induce 

the collision between particles; hence, collision is caused by Brownian motion, 

gravitational separation, or mechanical agitation (McClements, 2005). If each encounter 

between particles may result in flocculation, then the emulsion will rapidly become 

unstable. Thus, a high enough repulsive energy barrier is necessary against the droplets 

coming too close together. The likelihood of flocculation induced by droplet collision is 

known as collision efficiency. The collision efficiency is highly dependent on the height 

of the energy barrier. 



 
 

Chapter 2: Literature Review 

29 
 

In protein-stabilised emulsions, the net charge of protein is zero at its pI, is 

positive at low pH values, and is negative at high pH values. Such a change in droplet 

charge has a substantial influence on the emulsion stability against flocculation. When 

the pH is near the pI of protein, the electrostatic repulsive force is no longer strong enough 

to prevent flocculation because the net charge on the droplets is not high enough 

(Demetriades et al., 1997; Kulmyrzaev, Chanamai, & McClements, 2000).  

However, flocculation may be retarded by controlling the collision frequency. In 

addition, the droplet-droplet interactions (i.e., electrostatic interactions, steric interactions, 

hydrophobic interactions, depletion interactions, hydrodynamic interactions, biopolymer 

bridging interactions, and covalent interactions) play a crucial role on the flocculation 

rate of the emulsion system (McClements, 2005). 

2.3.2.3 Coalescence 

Coalescence is another type of droplet aggregation where two or more emulsified 

droplets irreversibly merged into a single larger droplet. Coalescence may result in the 

increase of droplet size over time, thus accelerating the creaming and sedimentation 

process. In the oil-in-water emulsion, it gradually leads to the separation of oil phase and 

the aqueous phase. Besides this, the contact area between the dispersed phase and the 

continuous phase