Aya Osman - University of Surreyepubs.surrey.ac.uk/846301/1/Aya Osman Thesis V1.0... · casein free...

The Impact of Milk Caseins on Behavioural

Development in Rats: Exploring the Role of the Gut

Brain Axis

by

Aya Osman

BSc (Hons) MSc (Distinction)

A thesis submitted in accordance with the requirements of the

University of Surrey for the degree of Doctor of Philosophy

The Faculty of Health and Medical Sciences

University of Surrey

January 2018

© Aya Osman 2018

2

STATEMENT OF ORIGINALITY

This thesis and the work to which it refers are the results of my own efforts. Any

ideas, data, images or text resulting from the work of others (whether published or

unpublished) are fully identified as such within the work and attributed t o their

originator in the text or bibliography. This thesis has not been submitted in whole

or in part for any other academic degree or professional qualification. I agree that

the University has the right to submit my work to the plagiarism detection service

TurnitinUK for originality checks. Whether or not drafts have been so-assessed, the

University reserves the right to require an electronic version of the final document

(as submitted) for assessment as above.

Aya Osman

.……………………………………………………………………..

3

ABSTRACT

Exposure to milk caseins beyond the normal weaning age in rats (21 days) was shown to impact

the development of the opioid system, a key player in emotional regulation. Further, milk alters

the gut microbiome and evidence suggests the gut microbiota can modify brain function and

behaviour. Given such evidence, it is hypothesised exposure to milk casein and its bioactive

breakdown product, beta casomorphin-7 (BCM-7) beyond weaning age in rats results in

disruptions to brain neurochemistry and behaviour via a microbiome-gut-brain axis

mechanism. This work aimed to investigate this hypothesis using in vivo behavioural,

neurochemical, gut microbial and metabonomic studies.

The behavioural results showed that exposure to casein from postnatal day 21-26 resulted in

an increase in ‘depressive-like’ behaviour as measured by the forced swim test (FST) and that

BCM-7 is not the only casein derived product driving such behavioural impairments.

Quantitative autoradiography revealed the casein induced depressive phenotype was

concomitant with changes to oxytocin and opioid receptors in regions of the brain associated

with mood. Fluorescence in situ hybridization showed an increase in Clostridium Histolyticum

in the gut suggesting a microbiome-gut-brain axis role in the casein induced changes reported.

In line with this, antibiotic knock-down of gut microbial activity not only prevented the

development of the casein induced depressive phenotype, but also caused a casein independent

antidepressive-like effect as assessed by the FST. Lastly, metabonomic analysis revealed an

increase in gut microbial metabolites and disruptions to choline and energy metabolism in

response to prolonged casein exposure, which have been implicated in mood disorders.

4

Collectively, these findings suggest casein exposure beyond weaning age results in

neurochemical and emotional impairments via a potential microbiome-gut-brain axis

mechanism. This indicates that prolonged breastfeeding periods may cause mood disorders in

humans and highlight the need for more comprehensive guidelines regarding an appropriate

weaning age.

5

DEDICATION

I would like to dedicate this PhD thesis to my mother and father, Amal Hamza Baldo and

Abdulhalim Osman for being my source of academic inspiration.

6

ACKNOWLEDGMENTS

Firstly I would like to convey my deepest gratitude to my supervisors Dr. Alexis Bailey, Prof.

Ian Kitchen, Dr. Jonathan Swann and Prof. Debra Skene. Alexis thank you not only for

providing me with consistent guidance and support, but also for giving me the opportunity to

embark on this interesting project and helping me develop and grow as a scientist. Ian thank

you for all the light hearted laughs, words of wisdom and advice you have shared with me over

the years, not forgetting the calming uplifting words in times of panic. Jon thank you for always

being open and approachable, I appreciate all the skills you have taught me. Debra thank you

for the very useful monthly gatherings, helping me keep everything organised and on time.

A very special thank you to my friends and former colleagues Polymnia Georgiou and Panos

Zanos. Poly thank you for passing on your excellent lab and experimental skills. Panos, your

training and assistance every step of the way despite the distance was highly appreciated!

Thank you to Dr. Gemma Walton, James Butcher and Simone Zuffa, for your help with FISH

and metabonomic analysis. Further special thanks to Fani Pantouli for being by my side

throughout all the transitions we faced. To my friend Matshediso Zachariah I am grateful to

you for being a vital source of emotional support throughout this process and providing me

with endless memories and laughs. To my dear friend Shazma, a huge thank you for being there

for me night and day. Ainsley Jones and Audrey Brown, I thank you both for the important life

lessons and a great time during my internship.

Last but by no means least, a massive heartfelt thank you to my family for supporting me in

every way possible throughout this PhD journey and for putting up with me during my difficult

writing stages! Thank you for believing in me and being such a strong support network, without

every single one of you I wouldn’t have been able to achieve what I have now. This is for you!

7

List of Figures

Figure 1.1: Major bioactive compounds derived from milk. ................................................. 38

Figure 1.2: Schematic illustration of the generation of milk-derived bioactive peptides and

their physiological functionalities. ........................................................................................... 42

Figure 1.3: Diagram showing the amino acid sequence of A1 and A2 beta casein and how beta

casomorphin-7 is released from the A1 variant. ...................................................................... 45

Figure 1.4: Factors influencing the infant gut microbiota development, as well as the adult

microbiota.. .............................................................................................................................. 75

Figure 1.5: Potential pathways involved in the bidirectional communication between the gut

microbiota and the brain. ......................................................................................................... 81

Figure 2.1: Experimental timeline showing key points in the methodology of casein free vs

casein rich milk study ............................................................................................................ 113

Figure 2.2: Effects of casein free and casein rich diets on milk consumption and body weight

................................................................................................................................................ 139

Figure 2.3: Increase in depressive-like behaviour following prolonged exposure to milk

caseins beyond the normal age of weaning in rats (PND21) ................................................. 140

Figure 2.4: Region-specific alterations of [125I]-OVTA binding in rats provided with casein

free or casein rich milk from PND21-PND26. ...................................................................... 142

Figure 2.5: Region specific alterations of [3H] DAMGO specific binding in rats provided with

casein free or casein rich milk from PND21-PND25. ........................................................... 144

Figure 2.6: Down-regulation of [3H] Deltorphin -1 specific binding in the deep layer of the

somatosensory cortex of rats fed casein free milk beyond the normal age of weaning (PND21).

................................................................................................................................................ 146

8

Figure 2.7: Multivariate analysis of urinary 1H NMR spectral profiles showing the influence

of casein free diet and casein rich diet on urinary metabolites. ............................................. 148

Figure 2.8: Effect of casein free vs. casein rich milk diet from PND21-PND26 on the

composition of total bacteria within different gut regions. .................................................... 150

Figure 2.9: Effect of casein free vs. casein rich milk diets from PND21-PND26 on Clostridium

Histolyticum bacterial groups within different gut regions.................................................... 151

Figure 2.10: Effect of casein free vs. casein rich milk diets from PND21-PND26 on

Bifidobacterium bacterial groups within different gut regions. ............................................. 152

Figure 2.11: Effect of casein free vs. casein rich milk diets from PND21-PND26 on

Lactobacillus-Enterococcus bacterial groups within different gut regions. .......................... 153

Figure 3.1: Diagram showing experimental groups in the antibiotic study.. ........................ 175

Figure 3.2: Experimental timeline showing key points in the methodology for antibiotic dosing

study. ...................................................................................................................................... 178

Figure 3.3: Water consumption in rats provided with free access to bottled water treated with

or without penicillin/streptomycin from PND17 until PND26 .............................................. 179

Figure 3.4: Milk consumption in rats provided with casein free or casein rich milk treated with

or without penicillin/streptomycin from PND21 until PND26 .............................................. 180

Figure 3.5: Effects of casein free and casein rich diets with or without penicillin/streptomycin

treatment on body weight change from PND14 until PND26. .............................................. 181

Figure 3.6: Immobility times exhibited by rats fed no milk, casein free or casein rich milk

treated with or without streptomycin/penicillin from PND17 until PND26 in the forced swim

test. ......................................................................................................................................... 183

Figure 3.7: Effect of 10 days of streptomycin/penicillin treatment from PND17 on the

composition of total bacteria within the caecum of control, casein free and casein rich treated

animals. .................................................................................................................................. 185

9

Figure 3.8: Effect of antibiotic treatment from PND17 until PND26 on host urinary metabolic

profile ..................................................................................................................................... 187

Figure 3.9: Effect of casein rich or casein free milk diets from PND21 until PND26 on host

urinary metabolic profile........................................................................................................ 189

Figure 3.10: Effect of casein rich milk or control diets from PND21 until PND26 on host

urinary metabolic profile........................................................................................................ 190

Figure 3.11: Effect of microbial suppression from PND17 until PND26 on urinary metabolic

profiles of control and control AB treated rats. ..................................................................... 192


profiles of in casein rich and casein rich AB treated rats. ..................................................... 194


profiles of control and casein rich AB treated rats ................................................................ 195

Figure 3.14: Metabolic impact of microbial suppression from PND17 until PND26 on urinary

metabolic profiles of animals provided with casein free and casein free AB milk. .............. 197


metabolic profiles of animals provided with casein free AB milk and control treatment group.

................................................................................................................................................ 198

Figure 4.1: Experimental timeline showing key points in the methodology for A1/A2 and A2

milk study............................................................................................................................... 229

Figure 4.2: Effect of A2 milk versus A1/A2 milk on milk consumption from PND21- until

PND26 and weight change from PND14 until PND27. ........................................................ 229

Figure 4.3: Effect of A2 milk versus A1/A2 Milk consumption from PND21 until PND26 on

locomotor activity. ................................................................................................................. 231

Figure 4.4: Increase in depressive-like behaviour caused by consumption of both a2 and

A1/A2 milk from PND21- PND26 ........................................................................................ 232

10

Figure 4.5: PCA of frontal cortex samples from Control, A1/A2 and A2 treated animals . 234

Figure 4.6: The effect A1/A2 milk from PND21 until PND26 on rat frontal cortex metabolites

compared to control treatment: OPLS-DA plots. ................................................................. 237

Figure 4.7: The effect of A2 or A1/A2 milk types from PND21 until PND26 on rat frontal

cortex metabolites: OPLS-DA plots. .................................................................................... 238

Figure 4.8: Diagram showing the interconversion between ATP, AMP, IMP and Adenosine

................................................................................................................................................ 248

11

List of Tables

Table 1.1: Correlating a human year with rat days based on different phases of life. ..................... 28

Table 1.2: Summary of selective studies relating manipulations of weaning age and rearing

conditions to behaviour with possible underlying mechanisms in various species.. ............... 34

Table 1.3: Key reported health benefits of prolonged breastfeeding in human infants and

children .................................................................................................................................... 36

Table 1.4: The average composition of human and bovine milk.. .......................................... 40

Table 1.5: Amino acid composition of naturally occurring beta-casomorphins in human and

bovine milk. ............................................................................................................................. 44

Table 1.6: Table summarising precursosrs, sequences and affinity of endogenous opioid

peptides. ................................................................................................................................... 58

Table 1.7: Key findings related to the impact of gut microbiota on behaviour. ..................... 89

Table 2.1: Formulations of casein rich and casein free milk diets. ....................................... 109

Table 2.2: Bregma coordinates used for brain regions analysed by [125I] OVTA Receptor, [3H]

Deltorphin-1 and [3H] DAMGO respectively. ........................................................................ 127

Table 2.3: List of reagents required for FISH analysis ........................................................ 130

Table 2.4: Hybridization conditions for oligonucleotide probes used in FISH analysis. ..... 131

Table 2.5: Sample dilutions used for FISH analysis ............................................................. 133

Table 2.6: Composition of hybridisation buffers used to achieve different hybridization

stringencies for FISH analysis ............................................................................................... 134

Table 2.7: Composition of wash buffers used for FISH analysis. ........................................ 135

Table 2.8: Full 1H- NMR chemical shifts for metabolites identified in urine samples......... 149

12

Table 3.1: Summary of 1H-NMR chemical shifts for metabolites associated with alterations

following antibiotic treatment and casein free or casein rich milk treatment ........................ 199

Table 4.1: Table showing the nutritional content of A2 whole milk formula and Tesco’s

(A1/A2) whole milk formula. ................................................................................................ 224

Table 4.2: Full 1H- NMR chemical shifts for metabolites identified in frontal cortex samples

of A1/A2 treated rats .............................................................................................................. 240

13

List of Abbreviations

3-HPPA 3-hydroxyphenylpropionate

5-HT 5-hydroxytryptamine

Acb Nucleus accumbens

ACTH Adrenocorticotropic hormone

AMP Adenosine monophosphate

AMPH Amphetamine

Amy Amygdala

AOL Anterior olfactory nucleus-lateral

AOM Anterior olfactory nucleus-medial

AOV Anterior olfactory nucleus-ventral

ASD Autism spectrum disorder

ATM Antimicrobial

BBB Blood brain barrier

BCM-7 Beta casomorphin-7

BDNF Brain derived neurotropic factor

BLA Basolateral amygdala

BMA Basalmedial amygdala

CE Central nucleus of amygdala

CgCx Cinglulate cortex

CHD Coronary heart disease

CNS Central nervous system

CON Conventionalised

CPu Caudate putamen

14

CRF Corticotropin-releasing factor

CSN2 Beta casein gene

DA Dopamine

DAMGO [D-Ala2,N-Me-Phe4-Gly5-ol]-enkephalin

DELT-1 Deltorphin-1

DMA Dimethylamine

DMG Dimethylglycine

DOPr δ-opioid receptor

DPP4 Dipeptidyl peptidase 4

DSS Dextran sodium sulphate

EBU Experimental behavioural unit

ENS Enteric nervous system

EPM Elevated plus-maze

FST Forced swim test

GABA gamma-aminobutyric acid

GC Gas chromatography

GF Germ free

GI Gastrointestinal

GSH Antioxidant glutathione

GTP Guanosine-5'-triphosphate

Hip Hippocampus

HPA-axis Hypothalamic-pituitary-adrenal axis

Hyp Hypothalamus

i.c.v. Intracerebroventricular

15

i.v. Intravenous

IBS Irritable bowel disease

IDDM Insulin dependent diabetes mellitus

IMP Inosine monophosphate

KO Knockout

KOPr κ-opioid receptors

LC Liquid chromatography

LD Light dark box

LDL Low density lipoprotein

LS Lateral septum

M1/M2 Motor cortex

MOPr Mu opioid receptor

MOPr μ-opioid receptor

mRNA Messenger ribonucleic acid

MS Medial Septum

NA Noradrenaline

NFE Nitrogen free extract

NMR Nuclear magnetic resonance

NOESY Nuclear Overhauser effect spectroscopy

NSB Non-specific binding

OF Open field

OT Oxytocin

OTr Oxytocin receptor

OVTA Ornithine vasotocin

16

PCA Principal component analysis

PDYN Prodynorphin

PENK Proenkephalin

Pir Piriform cortex

PND Postnatal day

POMC Proopiomelanocortin

PrL Prelimbic cortex

PVN Paraventricular nucleus of the Hyp

RGS Regulator of G-protein signalling

S1/S2 Somatosensory cortex

SAH S-adenosylhomocystine

SAM S-adenosylmethionine

SCFAs Short chain fatty acids

SEM Standard error of the mean

SIA Stress Induced Analgesia

SIH Stress induced hyperthermia

SON Supraoptic nucleus of the Hyp

SPF Specific pathogen free

Th Thalamus

TMAO Trimethylamine-N-oxide

UMP Uridine monophosphate

UPLC Ultra-high pressure liquid chromatography

VDB Vertical limb of the diagonal band of Broca

17

Contents ABSTRACT .............................................................................................................................. 3

DEDICATION.......................................................................................................................... 5

ACKNOWLEDGMENTS ....................................................................................................... 6

List of Figures ........................................................................................................................ 7

List of Tables ........................................................................................................................ 11

List of Abbreviations ............................................................................................................ 13

Chapter 1 ................................................................................................................................ 24

General Introduction ............................................................................................................. 24

1.1 Early Life Development and Weaning ...................................................................... 25

1.1.1 Weaning in Rats ................................................................................................. 26

1.1.2 The Effect of Weaning Time on behaviour ....................................................... 28

1.1.3 Weaning in Humans and Milk Rich Diets ......................................................... 35

1.2 Milk as a Bio-messenger ........................................................................................... 37

1.2.1 Constituents of Mammalian Milk ...................................................................... 39

1.2.2 Milk Caseins and Beta-Casomorphins ............................................................... 42

1.3 The Influence of BCM-7 on Health and Behaviour .................................................. 46

1.3.1 Cardiovascular Effects ....................................................................................... 48

1.3.2 Diabetes.............................................................................................................. 49

1.3.3 Neurological disorders ....................................................................................... 51

1.4 The Effect of Weaning and Milk Rich Diets on Brain Neurochemistry ................. 55

18

1.4.1 The Opioid system ............................................................................................. 55

1.4.2 Role of the Opioid System in Emotional Regulation......................................... 59

1.4.3 Opioid Receptor Development .......................................................................... 61

1.4.4 Role of Weaning in Opioid Receptor Development .......................................... 62

1.4.5 Role of Milk Caseins in Opioid Receptor Development ................................... 64

1.4.6 The Oxytocin System ......................................................................................... 66

1.4.7 Role of the Oxytocin System in Emotional Regulation ..................................... 67

1.4.8 The Role of Weaning on Oxytocin Receptor development ............................... 69

1.4.9 The Role of Weaning on other CNS Systems .................................................... 70

1.5 The gut-Brain Axis .................................................................................................... 72

1.5.1 Gut microbiota .................................................................................................. 73

1.5.2 Gut Microbiota Metabolism ............................................................................... 77

1.5.3 Bidirectional Gut-Brain Axis .............................................................................. 79

1.5.4 Gut-Brain Axis and Emotional Behaviour ......................................................... 82

1.5.5 Gut-Brain Axis and Neurochemistry .................................................................. 90

1.6 Metabonomics ........................................................................................................... 92

1.6.1 Neurodisorders..................................................................................................... 93

1.7 Thesis Hypothesis and Aims ................................................................................. 96

Chapter 2 ................................................................................................................................ 98

Effect of Prolonged Casein Exposure in Early Postnatal Life on Behavioural

Development and the Gut-Brain Axis .................................................................................. 98

19

2.1 Introduction ............................................................................................................... 99

2.2 Methodology ........................................................................................................... 108

2.2.1 Animals and Experimental Conditions ............................................................ 108

2.2.2 Casein Free and Casein Rich Milk Diets ............................................................ 108

2.2.3 Body Weights and Milk Consumption ............................................................ 110

2.2.4 Behavioural tests .............................................................................................. 110

2.2.5 Data Analysis ................................................................................................... 113

2.2.6 Metabolomics ................................................................................................... 114

2.2.7 Tissue Dissection ............................................................................................. 119

2.2.8 Quantitative Receptor Autoradiography .......................................................... 120

2.2.9 Fluorescence in situ Hybridization (FISH) ...................................................... 128

2.3 Results ..................................................................................................................... 138

2.3.1 Body Weight and Milk Consumption .............................................................. 138

2.3.2 Behaviour ......................................................................................................... 138

2.3.3 Neurochemical Analysis .................................................................................. 141

2.3.4 Analysis of Urine Samples (Metabolomics) .................................................... 147

2.3.5. Assessment of Intestinal Microbiota (FISH) ................................................. 150

2.4 Discussion .............................................................................................................. 154

Chapter 3 .............................................................................................................................. 170

Role of the Gut Microbiota in Mediating the Depressive Phenotype Induced by Prolonged

Casein Exposure ................................................................................................................... 170

20

3.1 Introduction ............................................................................................................. 171

3.2 Methodology ........................................................................................................... 174


3.2.2 Antibiotic Treatment ........................................................................................ 174

3.2.3 Milk Diets ........................................................................................................ 176

3.2.4 Body Weights, Milk and Water Consumption ................................................. 176

3.2.5 Behavioural Testing- Forced Swim Test .......................................................... 176

3.2.6 Collection of Urine .......................................................................................... 176


3.2.8 Metabolomics of Urine Samples ...................................................................... 177

3.2.9 Fluorescence in situ Hybridization (FISH) ...................................................... 177

3.2.10 Data Analysis .................................................................................................... 178

3.3 Results ..................................................................................................................... 179

3.3.1 Water Consumption ......................................................................................... 179

3.3.2 Milk Consumption ........................................................................................... 180

3.3.3 Weight Change................................................................................................. 181

3.3.4 Behaviour ......................................................................................................... 182

3.3.5 Assessment of Intestinal Microbiota (FISH) ................................................... 184

3.3.6 Metabonomic Analysis of Urine Samples ....................................................... 186

3.3.6.2 PCA of Urine Samples: Metabolic impact of casein exposure ........................ 188

3.3.6.3 PCA of Urine Samples: Metabolic impact of microbial suppression .............. 191

21

3.3.6.4 PCA of Urine Samples: Metabolic Impact of Microbial Suppression in Casein

Rich Treated Animals ..................................................................................................... 193

3.3.6.5 PCA of Urine Samples: Metabolic Changes in Casein Free Animals Following

Microbial Suppression .................................................................................................... 196

3.4 Discussion ............................................................................................................... 200

Chapter 4 .............................................................................................................................. 217

Effect of A1 vs A2 Variant of Milk Casein Consumption at an Early Developmental Age

on Depressive-Like Behaviour and Metabolic Profile ...................................................... 217

4.1 Introduction ............................................................................................................. 218

4.2 Methodology ........................................................................................................... 223


4.2.2 A2 Milk Study.................................................................................................. 223

4.2.3 Body Weights and Milk Consumption ............................................................ 224

4.2.4 Behavioural Tests............................................................................................. 225


4.2.6 Data Analysis ................................................................................................... 225

4.2.7 Metabolomics ................................................................................................... 226

4.3 Results ..................................................................................................................... 228

4.3.1 Milk Consumption & Weight Change ............................................................. 228

4.3.2 Locomotion ...................................................................................................... 230

4.3.3 Forced Swim Test ............................................................................................ 232

4.3.4 Metabonomic Analysis of Brain samples ........................................................ 233

22

4.4 Discussion ............................................................................................................... 241

Chapter 5 .............................................................................................................................. 251

5 General Discussion .................................................................................................... 252

5.1 Behavioural Effects of Casein Exposure ............................................................. 253

5.2 Effect of Casein Exposure on Neurochemistry ................................................... 255

5.3 Behavioural Modulation by Gut Microbiota ....................................................... 256

5.4 Metabolic Impact of casein Exposure ................................................................. 257

5.5 Study Limitations ................................................................................................ 258

5.6 Future Studies ...................................................................................................... 259

5.7 Conclusion ........................................................................................................... 261

Bibliography ......................................................................................................................... 262

23

‘God does not burden a soul beyond that it can

bear’

Surah Al-Baqarah, Verse 286

24

Chapter 1

General Introduction

25

1.1 Early Life Development and Weaning

Early childhood represents a period in which the most rapid development occurs in humans.

The years from conception through to eight years of age are critical for healthy, cognitive,

emotional and physical growth of children. The cell precursors of the brain and spinal cord,

which compose the central nervous system (CNS) begin to develop early in embryonic life and

extend into the postnatal period. By two years of age the human brain reaches 80-90% of its

adult size (Clancy et al., 2007). Therefore, during this period of rapid growth and development,

the CNS is in a state of constant change and plasticity and hence a stage at which it is sensitive

to both genetic and environmental influences (see Rice and Barone 2000).

Research has shown that environmental determinants such as early nutrition, stress, levels of

stimulation and social interaction all exert a critical impact on the maturation of brain structure

and behavioural function (see Loman and Gunnar, 2010; Lucassen et al., 2013). The ability of

the CNS to change and adapt to both internal and external stimuli is an important characteristic

which allows it to generate a range of phenotypes suitable for different environments. However,

developmental plasticity can also play a role in the generation of psychiatric disorders (see

Borre et al., 2014).

The concept of early life nutrition as a key environmental factor influencing brain development

and subsequent adult behaviour is fast gaining credence (for a review see Sachser et al., 2011;

Sachser, et al., 2013; Pryce and Feldon 2003). Unlike other environmental factors, nutrition

can directly influence brain development by providing the specific molecules, which enable

genes to exert their targeted effects on brain growth and development. For example, normal

brain development requires adequate concentrations of certain nutrients such as choline, folic

acid, iron, zinc and special fats which enable genes to exert their effect on brain development

26

and are obtained through the diet (Bryan et al., 2004). Therefore factors influencing dietary

intake in early life will subsequently effect normal brain development.

One of the most important nutritional events in the early stages of life in all milk-producing

mammals is weaning. The term weaning comes from the Anglo-Saxon word “wenian” meaning

“to become accustomed to something different”. Weaning can be generally defined as a gradual

process involving a progressive reduction in rate of milk transfer from the mother to her young,

accompanied by an increasing intake of solid food by the offspring, or for the purpose of this

thesis, as the complete cessation of milk transfer from mother to young (abrupt or final wean).

After weaning the young mammals become nutritionally and behaviourally independent from

their mothers (Kikusui et al.,2008; Kikusui and Mori, 2009) . This weaning process is marked

not only by a decline in maternal investment and offspring suckling but also a cascade of

nutritional, immunological, microbiological, biochemical and psychological adjustments

(Lawrence and Lawrence 1994). These changes which occur during weaning have very

important influences on subsequent behaviours of the infants as they become adults (Kikusui

et al., 2008). Therefore any manipulations to the weaning process are also likely to have an

influence on behavioural development.

1.1.1 Weaning in Rats

The rat has long been used as the experimental species of choice in developmental research

(Clancy et al., 2007). As a result the rat macro and micro neuroanatomy, neurophysiology and

behaviour have become well mapped during development (see Raedler, Raedler and Feldhaus,

1980; Bayer et al., 1993). As with humans, milk is the only source of nutrition for rat pups in

early postnatal development (first two weeks of life in rats) (Thiels et al., 1990). Spontaneous

weaning then begins around the third week of age and continues until around 30-34 days of

27

age when the pups cease milk intake (Kikusui et al., 2009). In rats, weaning occurs in nature

with the birth of the next litter, circa 27 days. Although suckling can continue normally up until

34 days of age, the time spent by offspring suckling reduces gradually and reaches adult levels

well before, at or around 25 days of age (Thiels et al., 1990). For laboratory rats, females

generally show postpartum oestrous and become pregnant again, which results in another

delivery every 21 or 22 days. Therefore, to increase productivity, laboratory rats are generally

weaned on postnatal day (PND) 21 (Kikusui et al., 2009). It should be noted that in laboratory

settings “weaning” is defined as the day the mother is removed from her pups (Ferdman et al.,

2007; Curley et al., 2009) it reflects breakage of the bond between the dam and her offspring,

which includes the cessation of suckling and physical separation from the dam and also the

cessation of social protection (Kikusui et al., 2009).

Several key differences between the laboratory rat and human brain maturation have been

shown. Rat pups are born after a short gestation period of 22.5 days, compared to humans at

280 days. Generally the new born rat CNS is equivalent approximately to a 24-week human

foetus (Romijn et al.,1991). However the rat CNS maturation is much more accelerated in

comparison to the human and is mature by three or four weeks (Clancy et al., 2007).

Furthermore, it is estimated that the rat brain at PND1–10 equates to the 3rd trimester in humans

and rat neurodevelopment at PND7 is equivalent to that of the human brain at birth (Clancy et

al., 2007). Moreover, Quinn (2005) has provided details of age comparison between the rat

and human in relation to weaning times in both species and have shown that a 42 day-old rat

can be thought of as being equivalent to a 1 year-old human.

28

Phase of life Rat Human

1.1.2 The Effect of Weaning Time on behaviour

To date, there has been strong evidence from numerous laboratory models that experiences

occurring in early life can have long term influences on neurobiology and behaviour during

adulthood. In particular, studies in rodents show that variations in mother-infant interactions

during the postpartum period leading-up to weaning can induce changes in gene expression in

the brain leading to differences in stress reactivity, social and reproductive behaviour, cognition

and reward mediated behaviour (Curley et al., 2009). Studies looking at the neurobehavioral

consequences of manipulations to mother-infant interactions during the postnatal period have

predominantly focused on maternal separation paradigms, more specifically focusing on “early

weaning” in which mothers are removed between postnatal days 14 and 15 (early-weaned) or

between postnatal days 20 and 25 (considered as normal or standard weaning), far less is

documented on the effects of delayed weaning on offspring behavioural development.

As summarised in Table 1.2 studies in mice have shown that removal of the dam prematurely

at PND14 versus normal weaning at PND21, results in the early weaned animals being more

aggressive (Kikusui et al., 2004), showing increased anxiety-like behaviour (Kikusui et al.,

Entire Life Span

Weaning period

Pre-pubertal period

Adolescent period

Adulthood

Aged phase

Average

13.2 days

42.4 days

3.3 days

10.5 days

11.8 days

17.1 days

16.4 days

= 1 Human year

Table 1.1: Correlating a human year with rat days based on different phases of life. Taken and adapted

from (Sengupta, 2013).

29

2005; Iwata et al., 2007) and increased general activity (Laviola et al.,1997; Kikusui et

al.,2005). With respect to physiological and neural correlates, ‘early weaning’ was associated

with an elevated neuroendocrine response to mild stress (Kikusui et al., 2006), decreased brain

derived neurotrophic factor (BDNF) concentrations in the hippocampus (Kikusui et al., 2009)

and altered myelin formation in the brain during the developmental period (Kikusui et

al.,2007). Similar results were obtained in rats, where the removal of the mother at postnatal

day 15 or 16 was also found to elevate levels of anxiety like behaviour and general activity

(Kanari et al., 2005; Shimozuru et al., 2007). Moreover, male rats were found to be more

vulnerable to early weaning induced stress than their female counterparts (Kikusui et al., 2006).

Other changes seen in early-weaned female rodents include a reduction in maternal behaviour

towards their own offspring, as well as producing offspring later in life (Kikusui et al., 2006).

Together, these findings indicate that premature weaning exerts effects on later behavioural

profile.

As mentioned above, so far, relatively little is known about the effects of ‘late-weaning’

however studies summarised in Table 1.2 indicate that removal of the mother at later time

points during early development also induces changes in the behavioural profile of male and

female offspring. For example, Farshim et al., (2016) showed that delay weaning male rat pups

on PND25 resulted in increased depressive-like behaviour compared to rats normally weaned

on PND21. Moreover, a study by Cook, (1999) tested four different weaning schedules,

including abrupt weaning where pups were removed entirely from their mother at 21 days of

age, self- weaning where pups were left with their mother until 30 or 40 days of age and gradual

weaning where pups were removed from their mothers for an increasing number of hours each

day starting with day 21 and finally having the mother removed permanently on day 30 (Cook,

1999). It was found that the basal levels of corticosterone, an indicator of stress in rats, was

significantly higher in the abruptly weaned and self-weaned rats compared to the gradually

30

weaned group. These tests were repeated again once the same animals had reached 12 months

of age, and the same results were seen (Cook, 1999). Taken together, these studies suggest

weaning age has both long term and short-term consequences on behaviour.

Indeed, a recent study by Richter et al., (2016) investigated the effect of late weaning on the

behavioural profile of male and female mice in adulthood. They reported at the group level,

weaning mice on week 4 (late weaning) was associated with a reduction of anxiety-like

behaviour and increased exploratory locomotion in adulthood compared to mice weaned on

week 3. Stability of the reported behavioural profiles was also investigated by repeating

behavioural tests at intervals of eight weeks during adulthood. Behavioural stability was found

to be highly dependent on the combination of weaning age and sex in that females weaned after

3 weeks of age were characterized by a high degree of behavioural stability, while this was not

the case for females weaned at 4 weeks of age. In males, pups weaned at 4 weeks of aged

expressed a high degree of behavioural stability in adulthood while those weaned at 3 weeks

of age did not (Richter et al., 2016). Taken together these findings show weaning age and sex

impact behavioural profiles overtime.

The experience of weaning at different time points was also found to produce changes in social

interaction in mice once they reached adolescence, with interesting differences observed

between males and females (Terranova et al., 2001). In one study, varying weaning ages were

tested: early weaning where the dam was removed on PND15, normal weaning where the dam

was removed on PND20 and delayed weaning at PND25. When the dam was removed, each

litter was split into males and females. A number of social behaviours were then tested for on

PND30 including, social sniffing, social grooming and social following (Terranova et al.,

2001). Both males and females belonging to the early weaned group were much more involved

in investigative social behaviours (sniffing of novel partners etc.) whilst spending less time in

31

self-directed behaviour such as self-grooming compared to their normally weaned counter

parts. In females, increased social investigation was also found as a consequence of delayed

weaning. These results were explained by sexual dimorphism in the development of opioid

systems in the brain (Terranova et al., 2001).

The effect of weaning on exploratory behaviour was also investigated using the open field test

in two different studies. A study by Curley et al., (2009) showed only a gender difference

irrespective of weaning age, where male mice spent more time in the centre in the open filed

showing lower levels of anxiety compared to female mice (Curley et al., 2009). A study by

Ferdman et al., (2007) assessed the combination of manipulations to weaning age and social

isolation on exploratory behaviour in adult rats. It was found that in the late weaned group

(PND 30) the males showed more exploratory behaviour than the females. This effect was only

observed in the late weaned but group reared pups (Ferdman et al., 2007). Rats reared in

isolation after weaning showed a reversed effect (i.e. females showing more exploratory

behaviour). This shows that rearing conditions are another early life influence, which can have

long lasting effects on behaviour (Ferdman et al., 2007).

Altogether, these studies show that manipulations to weaning age, most markedly within the

first 30 days of postnatal life, can produce significant and persistent behavioural changes.

These changes could be attributed to maternal factors or to nutritional factors induced by

changing the length of time animals are exposed to milk in early postnatal life.

32

Weaning Time Age of Testing

days/weeks

Species (strain) Behavioural

Phenotype(s)

Possible Mechanism (s) Reference (s)

Early weaned (PND

15)

Normally weaned

(PND21)

Delayed weaned

(PND27)

>PND60 Male CD-1 mice Regular weaned- AMPH

induced hyper-reactivity

and conditioned

locomotion

Deleted weaned and

Early weaned –no

induction

Weaning age (early life

experiences) play a role

in abuse of drugs

(Adriani et al., 2002)

Early weaned (PND

21)

Normally weaned

(PND 28)

PND50 and

PND74

Male and Female

C57BL/6J and B6 mice

Sexual dimorphism,

females weaned on

PND28 less active and

less explorative

compared to males. No

difference in anxiety-like

behaviour

Oxytocin and

Vasopressin

(Curley et al., 2009)

Early weaned (PND

14)

Normal weaned (PND

21)

3-26 weeks old Male and Female balb/CA

ICR mice

Increased level of

aggression

Increased anxiety

Increased general

activity

Elevated corticosterone

response to stress in

early-weaned animals

compared to normally

weaned.

Lower 5HT1B mRNA is

hippocampus

Changes in

glucocorticoid mRNA

expression

(Kikusui et al., 2004;

Kikusui et al., 2006;

Iwata et al., 2007b;

Nakamura et al., 2008)

Early weaned (PND

15)

Normally weaned

(PND 25)

PND30 and

PND70

Male and female CD1 mice Increased locomotor

activity and altered

social play in early

weaned compared to

normally weaned.

Dopaminergic and

opioid systems

Increased sensitivity to

δ-opioid receptor

agonists

(Laviola et al., 1997;

Terranova et al., 2001)

Early weaned (PND21)

Normally weaned

(PND 30)

14 weeks of age Male and female Wistar

rats

Early weaned- more

active in open field test

Regularly weaned- more

explorative

In early weaned low

dendrite length and

density in prefrontal

cortex

(Ferdman et al., 2007)

33

(late weaned rats avoid

social interaction)

Early Weaned ( PND

17)

Normal Weaned (PND

23)

PND30 Male C57BL/6 and

DBA/2 mice

Early weaned- increased

anxiety in Elevated Plus

Maze in both strains,

increased depressive-like

behaviour in D2, no

change in BL6.

Epigenetic (Carlyle et al., 2012)

Normally weaned

(PND 20)

Milk formula fed or

food pellets given to

two groups

10.5 weeks Male ddY mice Exploratory behaviour in

milk fed mice lower than

in pellet fed.

Nutritional (Ishii et al., 2005)

Normally reared

Artificially reared with

common milk

Artificially reared with

high protein (3g

casein/100ml milk)

PND18 Female Long-Evans

Hooded rats

Increased locomotion in

normal weaned rats

compared to two other

groups. High protein rats

had smaller brains

Protein-enriched diet (Diaz et al., 1982)


Normally weaned

(PND 30)

8 weeks Male and female wistar rats Early weaned males but

not females showed

lower frequency and

duration in the open arm

of elevated plus maze

Mother-pup interaction

(Stress)

(Ito et al., 2006)

Early weaned

(injection of

bromocriptine on

PND28, 30 and 33)

Control (group kept

with mothers, injected

with saline )

PND30 and

PND60

Domestic cat (Felis Catus) Early weaned kittens

showed increased

frequency of social and

locomotive play but not

of object play compared

to control

Maternal milk

supply/interaction

(Martin and Bateson

1985)

34

Table 1.2: Summary of selective studies relating manipulations of weaning age and rearing conditions to behaviour with possible underlying

mechanisms in various species. Abbreviations: AMPH: amphetamine; 5H-T;5-hydroxytryptamine; PND: postnatal day.

Early weaned (before

PND 20)

Regular weaned (PND

20)

PND20 Female and male Guinea

pigs

Early weaned showed

depressive behaviour

Proinflammatory activity (Schneider et al., 2012)


Regular weaned (PND

30)

Weeks 4-7 Female and Male Wistar

rats

Early Weaned show

lower playful

interactions and

increased anxiety in

Elevated Plus Maze

Mother-pup interaction (Shimozuru et al.,

2007)

Early Weaning

(breastfeeding time)

Up to week 12 145 children diagnosed

with infantile autism

Early weaned (lower

duration of breast

feeding) may contribute

to aetiology of infantile

autism

Duration of breast

feeding

(Tanoue et al., 1989)

Delayed Weaning on

PND25

Control normally

weaned PND21

PND25 Male Wistar Rats Delayed Weaned

(Housed with mother

until PND25) showed

increased depressive like

behaviour compared to

weaned on PND 21

Duration of

breastfeeding

(Farshim et al., 2016)

Delayed Weaning

week 4

Control normally

weaned week 3

Weeks 20 and

28

Male and Female

C57BL/6J mice

Group level: mice

weaned week 4 less

anxious and more

explorative than W3

animals. Sexual

dimorphisim week 3

females and week 4

males higher behavioural

stability in adulthood.

Mother-pup interaction (Richter et al., 2016)

35

1.1.3 Weaning in Humans and Milk Rich Diets

Weaning can vary in onset and the length of time between species. Beliefs and practices

surrounding weaning vary dramatically from culture to culture and are often also influenced

by religious beliefs and the structure of a mother’s work activities and age amongst other

factors (Macadam 1995). For example, mothers in Zulu societies have traditionally breastfed

their infants until 12 to 18 months, at which point a new pregnancy would be anticipated.

Ancient Hebrews completed weaning at about 3 years old, and most children in traditional

societies are completely weaned between 2 and 4 years of age (see Mutch et al., 2004). Due

to such variations, the appropriate age at which weaning should occur in humans remains

unclear, however The World Health Organisation (WHO) recommends infants should be

exclusively breastfed for the first six months of life to achieve optimal growth, development

and health, after which infants should receive complementary foods in order to meet their

evolving nutritional requirements, while breastfeeding continues to two years and beyond

(WHO, 2018).

It is well established that breastfeeding is an unequalled method of providing optimal nutrition

in infancy and boasts a wide range of health benefits for the developing child. Extensive

evidence has shown that breastmilk contains a variety of bioactive agents that modify the

function of the gastrointestinal tracts and the immune system, as well as in brain development

(see Binns et al., 2016). Some of the key health benefits of breastfeeding are summarised in

Table 1.3. As a result of such findings, prolonged breastfeeding periods and milk rich diets are

widely promoted in infancy.

36

Area Health benefit Reference

Neurobehavioral McCrory and Murray, 2013;

Julvez et al., 2007;

Gastrointestinal Kramer, 2002;

Klement et al., 2004;

Metabolic Harder et al., 2005;

Pettitt et al., 1997;

Atopic Chiu et al., 2016;

Immunological Demmelmair et al., 2006;

Kwan et al., 2004;

However, despite the health benefits associated with milk consumption in early life, the

prolonged consumption of milk especially cow’s milk has come under scrutiny in recent years

for its reported links to a number of disease states including diabetes, cardiovascular disease

and neurological disorders such as autism and schizophrenia (see Bell et al., 2006) Some of

these health effects have been linked to bioactive peptides liberated from milk proteins (see

Section 1.3).

Up until the 20th century, a wet nurse was the only safe alternative to breastfeeding infants.

Since then, alternative milk sources have evolved to include the synthetic formulas used today,

which are based on bovine milk as a protein source (Stevens et al., 2009). In the western world,

the use of artificial feeding substances grew rapidly around the mid-1980’s and was

significantly influenced by advertising campaigns aimed at the public. Such campaigns

Reduced risk of diarrhoea.

Reduced risk of inflammatory

bowel disease.

Reduced risk of obesity and

type 2 diabetes.

Reduced occurrence of

Asthma and Allergic diseases.

Increased cognitive ability or

intelligence. Improved language and

motor skills. Reduced attention deficit

or hyperactivity disorders.

Improved immune defence

Reduced risk of leukaemia.

Table 1.3: Key reported health benefits of prolonged breastfeeding in human infants and children

37

promoted infant and follow on formulas for their resemblance to mothers breast milk, boasting

many of the same health benefits as breast milk as well as some additional benefits, such as

being fortified (Stevens et al., 2009). As a result, many parents make the transition from

breastfeeding to formula or cow’s milk when the infant is less than one year of age (Leung et

al., 2003). Although breastmilk remains to be the medically preferred method of infant feeding,

milk formulas and cow’s milk have provided a way of supplementing breastfeeding while also

extending an infant’s exposure to milk rich diets beyond the ‘normal’ age of weaning. It is also

important to note that in evolutionary terms, milk and dairy products are relatively new to

humans, having begun with the domestication of diary animals (Ludwig et al., 2013). Therefore

although today milk is considered to be essential for a balanced diet both in infancy and

adulthood, it was not a part of the traditional hunter-gatherer diet. In many parts of the world

such as china, milk is still a minor part of the diet (Ludwig et al., 2013).

1.2 Milk as a Bio-messenger

For centuries milk has been regarded as a mere nutritional provider. Milk proteins were

extensively studied primarily as amino-acid donors and from a nutritional angle. Although this

conventional concept of milk is still valid and accepted, accumulating evidence over the last

three decades suggests that milk proteins and their by-products act as bioactive components

and bio-messengers, which exhibit important physiological and biochemical functions.

Milk was found to contain modulators of digestive and gastrointestinal functions, hormones

and growth factors capable of influencing the development and growth of the CNS (McCrory

et al., 2013), gastrointenstinal (GI) tract (Donovan, 2006), play a role in immunoregulation

(Walker et al., 2014) and modulation of the gut microflora population (Jost et al., 2015).

Among the bioactive constituents of milk (see Figure 1.1), this thesis will focus on the

bioactive peptides derived from milk proteins and more specifically the bioactive peptides

38

derived from milk caseins. For an extensive review on the peptides encoded in and yielded by

all milk proteins the reader is directed to Shah (2000).

The identification of bioactive peptides in milk was first reported over 60 years ago, since then

knowledge of bioactive peptides has rapidly increased and peptides with various functions

including opiate, anti-hypertensive and immunomodulatory activities have been described

(Meisel et al., 2000; Korhonen Hannu et al., 2006). This wide spectrum of biological activity

predetermines the role of bioactive milk peptides in human health and disease states. Much of

the experimental data to date indicate the positive effects of milk derived peptides on health

during early infancy (Saito, 2008). However, during the last decade, a growing body of

evidence linked bioactive peptides in milk with various adverse biological responses

(Woodford, 2007) (see Section 1.3).

Figure 1.1: Major bioactive compounds derived from milk. Image adapted from (Park et al., 2015)

39

The bioactivities of milk proteins are latent within the sequence of the parent protein, and only

during proteolytic cleavage of the parent protein are the active peptide fractions released; these

peptides can be released in three ways: (i) through hydrolysis by digestive enzymes, (ii) through

hydrolysis of proteins by proteolytic microorganisms, and (iii) through the action of proteolytic

enzymes derived from microorganisms (Pihlanto-Leppälä, 2000; Korhonen et al., 2003).

Therefore, given that milk is the main source of nutrition in early postnatal life, and it is known

to release bioactive peptides which play crucial roles in the development of a number of

systems in the body including the CNS, it is perhaps no surprise that manipulations to weaning

time or periods of exposure to milk in early life are likely to result in alterations to normal brain

development and subsequent behaviour.

1.2.1 Constituents of Mammalian Milk

Milk is introduced as the main source of nutrition in early postnatal life either in the form of

mother’s breast milk or cow’s milk as a supplement/replacement for breastmilk. It is recognised

as being nutritionally balanced and as a source of bio-messengers that support the growth and

development of infants, it has therefore attracted a lot of scientific interest over the years. Milks

are complex oil -in-water emulsions containing protein, fat, lactose, vitamins and minerals, as

well as biological products such as enzymes, cells, hormones and immunoglobulins. The

composition and profiles of milk vary amongst the mammalian species depending on the

nutritional needs of the young (Miller et al.,1990). This thesis will focus on human and bovine

milk as these are the most relevant milk types during early postnatal development in humans.

Human and bovine milk differ not only in protein quantities, but also in general profile, these

differences are illustrated in Table 1.4. As shown, one of the major differences in composition

between the two milk types is the total protein content. Total protein makes up 3.3% of bovine

milk, compared to 1.22% found in human milk (Miller et al.,1990).

40

Constituents Human milk (%) Bovine milk (%)

Protein

Lactose

Fats

Ash (minerals)

Total solids

There are two major protein groups in milk: Caseins which account for around 80% of the total

protein content in bovine milk (Kamiński et al., 2007) and Whey proteins which account for

the remaining 20% (Kamiński et al., 2007). Both of these major proteins in milk are known to

release peptides with bioactive functions (see Figure 1.2). Whey protein gives rise to β-

lactoglobulin and α-lactalbumin while caseins breakdown to release beta casomorphins (see

Nongonierma et al., 2015). This thesis will focus on casein and its bioactive breakdown

products.

In human milk, the casein to whey protein ratio varies throughout the lactation period. Early in

lactation the concentration of whey proteins is higher than casein, with casein concentrations

being very low or absent, resulting in whey/casein ratio of about 90:10 in early lactation (Kunz

et al., 1990, 1992). Subsequently, casein synthesis in the mammary gland and milk casein

concentrations increase, whereas the concentration of total whey proteins decreases, resulting

in a 50:50 ratio in late lactation (Kunz et al., 1992). As a consequence, unlike bovine milk there

is no fixed ratio of whey to casein in human milk. However it is clear that with lactation time,

the compositional makeup of mother’s breastmilk changes along with the nutritional

requirements of the infant, further highlighting the need to establish an appropriate weaning

age.

1.22

7.0

3.8

0.21

12.4

3.3

4.8

3.7

0.72

12.8

Table 1.4: The average composition of human and bovine milk. The various constituents are given

as a percentage of whole milk (Miller et al.,1990).

41

Infant formulas provided to developing human babies are designed to simulate the content of

human breast milk and are often used as a replacement or supplement to breastmilk, however

they are bovine milk based and therefore differ considerably from human breastmilk. For

example, protein provides approximately 7% of the calories in human milk, whereas in cow’s

milk proteins provide 20% of the calories (see Martin et al., 2016). Furthermore, cow’s milk

contains six to seven times as much casein as does human milk (see Martin et al., 2016). The

high casein content in cow’s milk is undesirable for developing infants not only because of

bioactive breakdown products liberated from milk caseins, but also casein forms a tough, hard

to digest curd that is difficult for young infants to digest resulting in gastrointestinal

complications (Leung et al., 2003). Recently, research has focused on better matching infant

formulas to mothers breast milk, for example during the manufacturing process of cow-milk

based infant formula, skim milk powder (high in casein) is mixed with whey protein

concentrate in order to make the whey/casein ratio more similar to human milk (McMahon et

al., 2008). However as milk is the main source of nutrition in early postnatal life, a period of

rapid growth and development, more research into the health implications of prolonged

exposure to milk rich diets either in the form of mothers breastmilk or cow’s milk in infants is

required.

42

1.2.2 Milk Caseins and Beta-Casomorphins

The caseins are a family of phosphoproteins synthesised in the mammary gland and secreted

as large colloidal aggregates termed micelles, which are responsible for many of the unique

physical properties of milk (Ginger et al., 1999). Bovine milk was found to contain 4 casein

types: alpha S1 (39-46% of total casein) alpha S2 (8-11%) beta (25-35%) and kappa (8-

15%).There is also gamma casein which is a product of degradation of beta casein (Kamiński

et al., 2007). Human milk contains three classes of caseins, namely alpha- S1, beta and kappa

caseins (McMahon et al., 2008). These caseins are encoded by members of a multigene family,

namely α β and κ. Among these, the beta casein gene has been studied extensively due to the

potential influence of beta-casein variants on human health.

Milk

Casein

Whey

α caseinΒ caseinƘ casein

α-lactalbuminβ-lactoglobulin

• In vivo digestion by digestive enzymes

• In vivo digestion by microbial enzymes

• In vivo digestion by proteolytic

microorganisms

Bioactivepeptides

Cardiovascular systemAntihypertensiveAntithrombotic

AntioxidantAnti-inflammatory

Nervous systemOpioid activity

Metabolic heatlhGlucose control

HypochloestrolaemicAppetite control

Immune systemImmunomodulatory

CytomodulatoryAnti-inflammatory

Antimicrobial ActivityAntibacterial

AntiviralAntifungal

Major proteins in milk Generation of peptides Bioactivities of milk derived peptides

Figure 1.2: Schematic illustration of the generation of milk-derived bioactive peptides and their

physiological functionalities. Taken and adapted from (Mohanty et al., 2016)

43

Mutations in the bovine beta casein gene (CSN2) have led to 13 genetic variants of this gene,

out of these A1 and A2 are the most common forms in cattle breeds (Kamiński et al., 2007).

These two variants contain 209 amino acids and differ by one amino acid in their sequence

(See Figure 1.3). In position 67 of the beta casein chain, proline in variant A2 is substituted by

histidine in variant A1 (Greenberg et al., 1984). This difference is also displayed between

minor variants, which on the basis of this amino acid substitution and known digestion of beta

casein variants A1 and A2, may be classified as ‘A1 like’ and ‘A2 like’. These variants include:

A3, D and E which, like A2 contain a proline at this position; and B,C and F which, like A1

contain a histidine at amino acid residue 67 (Kamiński et al., 2007).

A2 beta casein is recognized as the original beta casein variant as it existed before a proline 67

to histidine 67 point mutation caused the appearance of A1 beta casein in some European herds

some 5000–10,000 years ago (Farrell et al., 2004). A1 beta casein has only been found in cattle

of European origin. Pure bred Asian and African cattle produce milk containing only the A2

beta casein type, although some cattle presenting phenotypically as Asian or African cattle may

produce A1 beta casein as a consequence of cross bred ancestry (Farrell et al., 2004). In human

milk, no such polymorphism was found in the beta casein gene (A2-Corporation 2006;

Greenberg et al.,1984).

Upon proteolytic cleavage in the gut, A1 beta caseins break down to release peptides with

opioid effects similar to morphine, hence given the name beta casomorphins (BCMs) including

BCM-7. BCMs can be released from both human and bovine milk and are typically 4-11 amino

acids long (see Table 1.5), all starting with a tyrosine residue in position 60 in bovine beta

casein and position 51 of human beta casein (Kamiński et al., 2007). They are yielded from A1

beta caseins by digestion in the small intestine by pepsin, leucine aminopeptidase and elastase

and are substrate for the enzyme dipeptidyl peptidase IV (DPP IV) which is a cell surface

44

protease, belonging to the prolyl oligopeptidase family. BCMs are hydrolysed by DPP IV to a

mixture of Tyr-Pro, Phe-Pro-Gly, Phe-Pro and Gly (Kreil et al., 1983).

Beta-casomorphin Amino acid composition

In contrast, proteolytic cleavage of the A2 variant of bovine beta casein does not yield BCM-7

as the proline at position 67 hinders a split at this site under normal gut conditions (See Figure

1.3) (Noni, 2008; De Noni et al., 2010; Boutrou et al., 2013). However, it should be noted that

under specific in vitro conditions relating to pH and enzyme combinations not found in the

human gut A2 beta casein can indeed release BCM-7 (Cieślińska et al., 2012).

Bovine BCM-4

Bovine BCM-5

Bovine BCM-6

Bovine BCM-7

Bovine BCM-8

Bovine BCM-11

Human BCM-7

Human BCM-8

Tyr-Pro-Phe-Pro

Tyr-Pro-Phe-Pro-Gly

Tyr-Pro-Phe-Pro-Gly-Pro

Tyr-Pro-Phe-Pro-Gly-Pro-Ile

Tyr-Pro-Phe-Pro-Gly-Pro-Ile-Pro

Tyr-Pro-Phe-Pro-Gly-Pro-Ile-Pro-Asn -Ser-Leu

Tyr-Pro-Phe-Val-Glu-Pro-Ile

Tyr-Pro-Phe-Val-Glu-Pro-Ile-Pro

Table 1.5: Amino acid composition of naturally occurring betacasomorphins (BCMs) in human

and bovine milk (Kamiński et al., 2007).

45

In human milk, beta casein is of the A2 type, with a proline at the equivalent position on the

beta casein protein chain (Jarmołowska, Sidor, Iwan, Bielikowicz, Kaczmarski, Elżbieta

Kostyra, et al., 2007). Thus, thermodynamics does not favour the yielding of BCM-7 from

human beta casein. Indeed, Wada et al., (2015) examined non-digested and in vitro-digested

human milk, and reported the presence of human BCM-9 but not human BCM-7 (Wada et al.,

2015). However, Jarmołowska et al., (2007) reported the presence of BCM-7 in colostrum (the

mothers fluid produced in the first day after delivery), but at 2 months into the lactation period,

the authors reported much lower quantities (Jarmołowska, Sidor, Iwan, Bielikowicz,

Kaczmarski, Elżbieta Kostyra, et al., 2007). Furthermore, Kost et al., (2009) were able to detect

substances with immunoreactivity of human BCM-7 in blood plasma of 3 month old breast fed

infants (Kost et al., 2009a). Human BCM-7 has a different amino acid sequence to bovine

Amino Acid Sequence of β-casein

Val – Tyr – Pro – Phe – Pro – Gly – Pro – Ile – X

Val – Tyr – Pro – Phe – Pro – Gly – Pro – Ile – Pro Val – Tyr – Pro – Phe – Pro – Gly – Pro – Ile – His

Βeta-casein A2 Βeta-casein A1

Tyr – Pro – Phe – Pro – Gly – Pro – Ile

Beta-casomorphin 7

PepsinPancreatic elastaseLeucine amino peptidase

Elastase

X= Pro in beta-casein A2His in beta-casein A1

Figure 1.3: Diagram showing the amino acid sequence of A1 and A2 beta casein and how beta

casomorphin-7 is released from the A1 variant. Taken and adapted from (Kamiński et al., 2007).

46

BCM-7,with homology in five of seven amino acids (differing amino acids at positions four

and five) (Wada et al., 2015). BCM-7 has been the focus of much research due to its ability to

act on opioid receptors found in the body and consequently eliciting a wide range of health

effects (see ul Haq et al., 2014).

1.3 The Influence of BCM-7 on Health and Behaviour

BCM-7 was first isolated as a peptide having morphine-like activity in 1979 (Brantl, 1984). It

contains the N-terminal amino acid sequence Tyr-Pro-Phe-Pro which possesses preferential μ

opioid receptor (MOPr) agonist activity (Brantl, 1984; Sun et al., 1999) and moderate potency

compared to morphine. Furthermore BCM-7 was shown to exhibit a low affinity for δ opioid

(DOPr) and almost no affinity for Ƙ-opioid receptors (KOPr) (see Section 1.4.1)

(Teschemacher et al. 1997). These pharmacodynamic characteristics of BCM-7 were

elucidated alongside other BCMs in studies using radioligand binding assays and in isolated

organ preparations from the guinea pig ileum or from the mouse vas deferens, which are

typically used test systems for MOPr and DOPr activities, respectively. In general, these studies

showed that human BCM-7 was 3-30 times less potent than bovine BCM-7, and 300-600 times

less potent than morphine (Human BCM-7 IC50 = 29.00 µmol/l, Bovine BCM-7 IC50 = 5.14

µmol/l, Normorphine = 0.05 µmol/l )(Koch et al.,1985).

As an opioid receptor ligand, BCM-7 can be expected to act as other opioids, i.e. to bind to

opioid receptors and elicit effects in all cells or tissues in which opioid receptors are known to

be expressed (Smotherman et al., 1992). The first physiological effects of BCMs take place in

the gastrointestinal tract. Indeed it has been shown that BCMs bind to opioid receptors found

in the luminal side of the gut and can affect gastrointestinal functions such as absorption,

secretion and gut motility (Mahè et al., 1989). Daniel et al., (1990) have shown that

casomorphins inhibit intestinal motility in isolated segments of rat ileum (Daniel et al., 1990).

47

Rats fed casein rich suspensions took longer in gastric emptying and gastrointestinal transit

time than rats given a whey protein suspension (Daniel et al., 1990). These effects were found

to be reversed when the opioid antagonist naloxone was administered, implicating the role of

MOP receptors in eliciting these effects. Furthermore, Barnett et al., (2014) showed that

feeding rodent’s milk containing A1 beta casein resulted in significantly delayed

gastrointestinal transit time compared with milk containing the A2 beta casein variant. Again,

this delay could be eliminated by the administration of opioid blocker naloxone suggesting a

MOPr mediated effect. This group also demonstrated a 40% up regulation of the enzyme DPP4

in the jejunum of A1- relative to A2 fed rodents (Barnett et al., 2014) suggesting BCM-7 in

mediating these gastrointestinal effects, as it is known BCM-7 is a substrate for DPP4. BCM-

7 has also been shown to induce a strong release of gastrointestinal mucin secretion in rat

jejunum through the activation of µ-opioid receptors (Zoghbi et al., 2006).

Based on findings from Kost et al., (2009) who reported detecting substances with

immunoreactivity to human BCM-7 in blood plasma of 3 month old breast fed infants. It is

feasible to suggest BCM-7 can be transferred intact from the luminal side to the blood side of

the gut where it can be transported to tissues and organs and can elicit further opioid mediated

biological and physiological responses. Furthermore, bovine BCM-7 immunoreactive material

has been detected in the plasma of new born calves following milk intake (Umbach et al.,

1985), and in new-born but not adult dogs (Singh et al., 1989). The absorption of large

macromolecules such as BCMs from the GI to the blood seems to only occur in the case of

immature neonatal mucosa (Weaver, 1992; Sun et al., 2003). New born mammals are known

to have leaky guts as they need to be able to pass large molecules through the gut wall enabling

them to absorb nutrients from mother’s milk (Zanardo et al., 2001). This could explain why

BCM-7 is detected in new born infants and other mammals but not in adults, and also suggests

infants may be exposed to higher levels of BCMs. In line with this, patients with “leaky gut”

48

syndrome have been shown to have high levels of BCM-7 in their urine compared to

individuals with normal gut walls (Mulloy et al., 2010).

Once in the blood, it has been suggested that BCM-7 can act as a causative agent in

cardiovascular disease, type 1 diabetes, sudden infant death syndrome as well as autism and

schizophrenia (see minireview by Lister et al., 2015).

1.3.1 Cardiovascular Effects

The connection between beta casein A1 consumption and heart disease was first drawn

following epidemiological studies. McLachlan (2001)noted a strong connection between milk

consumption and coronary heart disease (CHD) mortality across countries (McLachlan, 2001)

and from comparisons of the levels of antibodies to milk protein in cases of CHD and non-

affected controls (Davies et al., 1969). Later Elliott et al., (1999) published an epidemiological

study concluding that A1 beta casein was significantly and positively correlated with CHD in

20 affluent countries over a 20 year period (Elliott et al., 1999). Supporting the early

epidemiological work was the observation that casein in comparison to other milk proteins is

a potentiator of heart disease or related conditions such as hypercholesterolemia, in rabbits

(Terpstra et al., 1981) and monkeys (Barth et al., 1984). However this work did not identify

which variants of milk caseins are atherogenic. A rabbit feeding study later showed a causative

link between A1 consumption and the development and progression of arterial lesions (Tailford

et al., 2003). The proposed mechanism underlying the atherogenic properties of A1 beta casein

stem from findings that the bioactive peptide BCM-7 released from A1 beta casein has been

shown to catalyse the oxidation of low density lipoproteins (LDL) (Torreilles et al., 1995). The

uptake of oxidised LDL by endothelium bound macrophages has been widely implicated in the

pathogenesis of atherosclerosis (Torreilles et al., 1995). Studies have also shown that cow’s

milk infant formula fed infants raised antibodies to oxidised LDL at ten times the rate of infants

49

who were breast fed , highlighting a difference in the atherogenic properties of bovine BCM-7

compared to human BCM-7 (Steinerová et al., 1999).

1.3.2 Diabetes

Type 1 insulin dependent diabetes mellitus (IDDM) also known as childhood onset diabetes is

an autoimmune disease affecting approximately 0.2 to 0.6% of the population of developed

countries (see Chia et al., 2017). This form of diabetes is thought to result from an interaction

of several genetic and environmental factors that activate immune cells capable of destroying

the insulin-producing beta cells in the pancreatic islets. One of the environmental factors

implicated in diabetes pathogenesis is exposure to certain types of food including cow’s milk.

Studies have shown a strong correlation between early exposure to cow’s milk and the

incidence of IDDM (Scott, 1990; Virtanen et al., 1998; Muntoni et al., 2000). Elliot et al.,

(1997) were amongst the first to demonstrate milk protein induced diabetes via an opioid

receptor mediated mechanism. Non-obese diabetic mice were fed a basal diet supplemented

with either A1 or A2 beta casein. 47% of mice fed A1 beta casein diet developed autoimmune

diabetes while mice fed A2 beta casein did not develop the disease. Co-administration of the

opioid receptor antagonist naloxone attenuated the effects of A1 beta casein diet, suggesting

the diabetogenic effects of the A1 beta casein diet were at least partly mediated via opioid

receptors (Elliott et al., 1998). Two years later an association between the incidence of IDDM

in children younger than 14 years of age and the consumption of beta caseinA1 was reported

(Elliott et al., 1998; McLachlan, 2001). In addition a group in Iceland compared the incidence

of type 1 diabetes across five Nordic countries with measured levels of a number of milk

protein variants and concluded that beta casein variant A1 and not other milk proteins

examined, may contribute to observed variation in the diabetogenicity of cow’s milk

50

(Thorsdottir et al., 2000). It was proposed that the diabetogenic effect of A1 beta casein was

attributable to the release of BCM-7 during digestion (Elliott et al., 1998).

Different mechanisms underlying the diabetogenic effect of BCM-7 have been suggested, with

the immune system being a strong candidate (Chia et al., 2017). In humans, Padberg et al.,

(1999) reported that the ratio of A1 to A2 beta casein antibodies was higher in those with type

1 diabetes than in controls. Indeed BCM-7 has been shown to influence the function of immune

cells, for example two in vitro studies have shown BCM-7 alters lymphocyte proliferation via

an opioid-dependent pathway (Elitsur et al., 1991; Kayser et al., 1996). The first of these

studies showed suppressive effects of BCM-7 on lymphocyte proliferation at all concentrations

tested while the second study showed suppressive effects at low BCM-7 doses and stimulatory

effects at higher doses. However the full physiological relevance of the immunomodulatory

effects of BCM-7 in animals and humans requires further investigation.

Other proposed mechanisms for the pathogenesis of type 1 diabetes consider BCM-7

interacting with the intestinal microbiota, increasing mucosal barrier permeability and

intestinal immune response (Vaarala et al., 2008). Gut permeability abnormalities increase the

risk of intestinal immune system exposure to dietary antigens such as proteins and peptides

resulting in an altered immune activation and intestinal inflammation (Vaarala, 2008). Indeed

studies in rats have shown the intestine is highly permeable before the onset of type 1 diabetes.

(Neu et al., 2005). Intestinal myeloperoxidase activity are also higher in diabetes prone rats

compared to control highlighting a concomitant early intestinal inflammatory response (Neu et

al., 2005). This is notable because inflammatory mediators can compromise epithelial barrier

function and further effect gut permeability. Indeed, elevated intestinal myeloperoxidase

activity is reported in rodents fed A1 compared with A2 beta casein (Barnett et al., 2014).

Suggesting BCM-7 may indeed mediate its diabetogenic effects via the immune system.

51

It has also been shown in rats that A1 beta casein stimulates the production of the enzyme

DPP4 in the jejunum (Barnett et al., 2014). It is known that DPP4 degrades the gut incretin

hormones rapidly (Drucker 2006); in humans incretin hormones modulate insulin and glucose

metabolism (Holst et al., 2004). DPP4 inhibitors are now widely used in the management of

type 2 diabetes due to their effects on incretin hormones. It is therefore likely that BCM-7

induced up regulation of DPP4 could result in increased degradation of incretin hormones resulting

in altered metabolism of insulin and glucose and thus contribute to the development of diabetes.

1.3.3 Neurological disorders

A growing body of literature has linked diet/nutrition to the development of neurological

disorders (for a review see Liu et al., 2015) . Beta casein variant A1 and its bioactive breakdown

product BCM-7 have been shown to aggravate some neurological disorders. Autism Spectrum

Disorder (ASD) is a neurodevelopmental disorder characterized by abnormalities in social

interaction, communication skills and restrictive repetitive behaviours (Cieslinska et al., 2015).

The prevalence of ASD is estimated from 60 to 70 per 10,000 (Fombonne, 2009) and its

aetiology remains unknown. However there are strong links to a variety of genetic factors

(Hallmayer et al., 2011; Buie, 2013). The recent increase in prevalence of ASD suggests other

factors are contributory including dietary factors (Buie, 2013; Cieslinska et al., 2015). In recent

years, researchers have focused on the role of the opioid system in ASD and the “opioid excess”

theory was developed. Many authors have reported that autistic children, and in some cases

their mothers, displayed elevated levels of both endogenous opioid peptides and exogenous

opioid peptides such as BCMs in serum, urine, immune cells and the cerebrospinal fluid

(Cazzullo et al., 1999; Tordjman et al., 2009; Sokolov et al., 2014). Researchers have strongly

suggested that severity of autistic symptoms are correlated with the concentration of BCM-7

in the urine (Sokolov et al., 2014). The opioid theory of autism has been further supported by

52

pharmacological studies which have demonstrated that the opioid receptor antagonist naloxone

can improve the clinical symptoms of autism (Leboyer et al., 1993). According to this theory,

genetic predisposition and/or early exposure to environmental stressors may lead to functional

alterations in the gut, reduced proteolytic activity and the increased permeability of the gut

mucosa. These factors, possibly in combination with low levels of circulating peptidases and

increased blood-brain barrier (BBB) permeability, may cause hyperpeptidemia and

accumulation of opioid peptides such as BCM7 in the blood and the brain. Thus, chronically

elevated levels of exorphins in the brain may directly modulate the opioid and other

neurotransmitter systems, leading to the development of ASD (Whiteley et al., 2002). Indeed,

studies have demonstrated the presence of human or bovine BCM-7 in the blood of infants that

were breast fed or received cow’s milk based infant formulas respectively (Kost et al., 2009).

Out of these children those who displayed delayed psychomotor development expressed higher

levels of bovine BCM-7 than children with normal psychomotor development, whereas the

presence of human BCM-7 correlated with normal psychomotor development (Kost et al.,

2009). This study suggests an association between the levels of circulating bovine BCM-7 and

impaired early child development, setting the stage for autistic disorders. Furthermore, a

number of studies have reported that autistic children on a gluten-free casein free diet displayed

improved autistic symptoms, including improved gastrointestinal symptoms, concentration and

attention as reported by their parents (Lange et al. 2015), suggesting casein and its bioactive

breakdown products do indeed play a role in the exasperation of autistic like symptoms.

Similarly, Dohan (1966) proposed a hypothesis for schizophrenia pathogenesis that stressed an

interaction between genetic and dietary factors (Dohan, 1966). It was suggested that a dietary

overload of food exoprhins derived from casein and also gluten may produce schizophrenia-

like symptoms through their action on opioid receptors in certain genetically susceptible

individuals (Zioudrou et al., 1979). Indeed, peptides with opioid activity were detected in the

53

urine (Hole et al., 1979) and plasma (Schmauss et al., 1985) of patients with schizophrenia.

These peptides were found to induce prolonged behavioural abnormalities following

intracerebroventricular (i.c.v) injection into rats (Schmauss et al., 1985). Furthermore, BCM-7

injections in rats have been shown to cause other dramatic behavioural changes (Sun et al.,

1999). BCM-7 is capable of crossing the blood brain barrier and was shown to cause

“explosive” behaviour, which entailed violent running and jumping. This was followed by

withdrawn behaviour and eventually led to a state of complete inactivity (Sun et al.,

1999).These types of behaviours are often associated with mental disorders such as

schizophrenia and autism (Sun et al., 1999).

As with autism, gluten free casein free diets have been reported to improve the symptoms of

schizophrenia (Singh et al., 1976; Okusaga et al., 2013). The underlying mechanism of benefit

for such diets in improving neurological disorders remains unclear. However, recently an

epigenetic mechanism for the beneficial effects of these diets has been proposed, as it is known

epigenetic programming during early postnatal development can influence lifelong disease risk

(Bale et al., 2010; Reynolds et al., 2012) and can be modulated by nutritional factors especially

factors effecting DNA methylation.

DNA methylation is an important methylation reaction involving the addition of a methyl group

to DNA and is a major factor for epigenetic changes in gene expression (see Trivedi et al.,

2014). Emerging evidence from biochemical investigations indicate that systemic oxidative

stress also plays a strong role in ASD (Frustaci et al., 2012) and schizophrenia (Yao et al.,

2006) associated with significantly lower levels of the antioxidant glutathione (GSH) whose

synthesis is limited by cysteine availability. Brain GSH levels are reported to be decreased in

both autism (Melnyk et al., 2012) and schizophrenia (Do et al., 2000) strongly supporting the

proposal that a deficit in GSH might contribute to neurodevelopmental disorders.

54

Oxidative stress and DNA methylation changes are metabolically coupled via the

transsulfuration pathway. DNA methylation is carried out by a class of enzymes called DNA

methyltransferases, which depend on the level of S-adenosylmethionine (SAM) (Trivedi et al.,

2014) , which in turn are dependent on the action of the enzyme methionine synthase (MS) and

the redox status of the cell (Hondorp et al., 2004). SAM acts as a methyl donor for over 200

methylation reactions and is then converted to S-adenosylhomocystine (SAH) an inhibitor of

methylation reactions. In recent studies abnormal DNA methylation patterns have been

reported in schizophrenia (Jakovcevski et al., 2012; Nishioka et al., 2012) and ASD (Schanen,

2006).

Indeed, it has been shown that human and bovine BCM-7 decreases excitatory amino acid

transporter 3 (EAAT3) mediated cysteine uptake in cultured human neuronal and

gastrointestinal epithelial cells via a MOPr mediated mechanism (Trivedi et al., 2014). An

inhibition of cysteine uptake was observed following treatment of these two cell types with

either human or bovine BCM-7 for 30 mins in a concentration dependent manner with bovine

BCM-7 showing more potency in reducing cysteine uptake compared to human BCM-7. This

decrease in cysteine uptake was associated with changes in intracellular antioxidant GSH and

also the methyl donor SAM, which plays an important role in DNA methylation and subsequent

gene expression (Trivedi et al., 2014). Indeed Trivedi et al., (2014) also reported both human

and bovine BCM-7 increased genome-wide DNA methylation with a potency order similar to

their inhibition of cysteine uptake, i.e. bovine BCM-7 induced a higher increase in DNA

methylation compared to human BCM-7. At the transcriptional level, human BCM-7

significantly increased expression of nine genes including EAAT3. In contrast bovine BCM-7

decreased expression of the same nine genes assessed (Trivedi et al., 2014) . Taken together,

this study illustrates the potential of both bovine and human BCM-7 to exert antioxidant and

epigenetic changes that may be particularly important during postnatal development.

55

Furthermore, the differences in the antioxidant and epigenetic effects of human versus bovine

BCM-7 may contribute to the developmental differences observed between breastfed and

formula-fed infants (Kost et al., 2009a).

There is also strong evidence to suggest that opioid peptides in milk alter responses to stress.

For example, acute administration of BCM-7 was found to reduces maternal separation distress

responses in chicks (Panksepp et al., 1984). In contrast, it was found that chronic administration

of BCM-7 during development impairs the development of behavioural responses to stressful

stimuli (Panksepp et al., 1984). Furthermore, BCMs have been shown to cause analgesia, apnea

as well as changes in the sleep of neonatal rats (Kamiński et al., 2007). Naloxone pre-treatment

was found to reverse these effects, suggesting a MOPr mediated effect. These effects have been

shown by intracerebral, intraperitoneal and intraventicular injections (Sun et al., 1999).

1.4 The Effect of Weaning and Milk Rich Diets on Brain

Neurochemistry

1.4.1 The Opioid system

The discovery of the endogenous opioid system stems from the use of opium extracted from

the poppy seed (Papaver somniferum), which has powerful pain reliving properties as well as

a euphoric effect. Morphine, named after the Greek god Morpheus, was isolated from opium

in 1806 and remains the most widely used painkiller in contemporary medicine today. The

existence of opioid binding sites in the brain was established in 1973 and were later MOPr,

DOPr and KOPr. Since then the opioid system has been extensively explored and has proven

to be an important neurotransmitter system critically implicated in the control of mood,

motivation and pain. The opioid system also regulates numerous physiological functions,

56

including responses to stress, respiration, gastrointestinal transit, as well as endocrine and

immune functions (for an extensive review on the opioid system see Bodnar, 2016).

Soon after the discovery of the opioid receptors, a series of endogenous ligands were also

discovered. Three pro-hormone precursors provide the parent compounds from which these

endogenous ligands are derived. Preproenkephalin (PENK) is cleaved to form met-enkephalin

and leu-enkephalin which bind to DOP receptors. Dynorphin A and Dynorphin B are derived

from preprodynorphin (PDYN) and are agonists for KOP receptor. Prepro-opioimelanocortin

(POMC) is the parent compound for beta endorphin, an agonist at the MOP receptor, although

beta endorphin is capable of displaying agonist activity at all three classical opioid receptors

(McDonald et al., 2005; Corbett et al., 2009). In addition to beta endorphins, POMC precursor

also encodes non-opioid peptides, namely adrenocorticotropic hormone (ACTH), α-

melanocyte-stimulating hormone (α—MSH) and β-lipotropic pituitary hormone (β-LPH) (see

Bodnar 2016) All opioid peptides, both endogenous and exogenous, such as BCM-7, share a

common NH2-terminal Tyr-Gly-Gly-Phe signature sequence which allows them to interact

with opioid receptors (see Table 1.6). The genes encoding all three opioid receptors have been

successfully cloned (MOR-1 DOR-1 and KOPr-1, as well as the gene for the non-opioid

orphaninFQ/nocieptin receptor which is considered to be a non-opioid branch of the opioid

receptor family). All opioid receptors were found to display a topology characteristic of G-

protein coupled receptors (Corbett et al., 2009), and found to have a high homology with their

genomic organisation being virtually identical (Kieffer et al., 2002).

Opioid receptors and peptides are broadly expressed in different densities within the central

nervous system and, to a lesser extent, throughout the periphery, occupying sites within the vas

deferens, knee joint, gastrointestinal tract, heart and immune system, amongst others (Wittert

et al.,1996). In the brain, opioid receptors are expressed primarily in the cortex, limbic system

57

and brain stem. Binding sites for the three receptors overlap in most structures, but some

structure exhibit higher expression of one receptor more than others (see Le Merrer et al.,

2009). The highest MOP receptor densities are found in the amygdala (not including the central

nucleus of the amygdala), thalamus, caudate putamen, neocortex, nucleus accumbens,

amygdala, interpeduncular complex, and inferior and superior colliculi (Mansour et al., 1987).

Moderate densities of MOPr are found in the periaqueductal gray and raphe nuclei (Hawkins

et al., 1988). These brain regions have well-established roles in pain and analgesia. Other

physiological functions shown to be mediated by MOPr include respiratory and cardiovascular

functions, intestinal transit, feeding, emotional regulation, reward, thermoregulation, hormone

secretion and immune functions (Dhawan et al., 1996).

DOPr receptor distribution is more restricted compared to MOPr, with a prominent gradient of

receptor density from high levels in forebrain regions to relatively low levels in the most hind

brain regions. The highest densities of DOPr in the mouse/rat are found in the olfactory tract

(olfactory bulbs, anterior olfactory nucleus, TU, medial amygdala), neocortex, caudate

putamen, nucleus accumbens and amygdala (Mansour et al., 1987) and also highly expressed

in the striatum. In the spinal cord DOPr are present in the dorsal horn where they play a role in

mediating the analgesic effects of DOPr agonists. The functional roles of DOPr are less clearly

established compared to MOPr, however in addition to analgesia, DOPr may have a role in

gastrointestinal motility, mood and behaviour as well as in cardiovascular regulation

(Reinscheid et al., 1995).

KOPr are located predominantly in seven brain regions that are part of the stress axis (including

the claustrum, paraventricular hypothalamus, arcuate nucleus, supraoptic nucleus and pituitary)

and have also been implicated in the regulation of nociception, diuresis, feeding,

neuroendocrine and immune system functions (Dhawan et al., 1996).

58

Precursor Peptide Sequence Receptor affinity

Prepro-

Opioimelanocortin

Unknown

Pro-enkephalin

Pro-dynorphin

MOPr + DOPr

MOPr

DOPr

KOPr

β-endorphin

Endomorphin-1

Endomorphin-2

[Met5]Enkephalin

[Leu5]Enkephalin

Dynorphin A

Dynorphin B

α-neomorphin

β-neomorphin

Tyr-Gly-Gly-Phe-Met-Thr-Ser-Glu-Lys-Ser-Gln-Thr-Pro-Leu-Val-Thr-LeuPhe-Lys-Asn-Ala-Ile-Ile-Lys-Asn-Ala-Tyr-Lys-Lys-Gly-Glu

Tyr-Pro-Trp-Phe-NH2

Tyr-Pro-Phe–Phe-NH2

Tyr-Gly-Gly-Phe-Met

Tyr-Gly-Gly-Phe-Leu

Tyr-Gly-Gly-Phe-Leu-Arg-Arg-Ile-Arg-Pro-Lys-Trp-Asp-Asn-Gln

Tyr-Gly-Gly-Phe-Leu-Arg-Arg-Gln-Phe-Lys-Val-Val-Thr

Tyr-Gly-Gly-Phe-Leu-Arg-Lys-Tyr-Pro-Lys

Tyr-Gly-Gly-Phe-Leu-Arg-Lys-Tyr-Pro

Table 1.6: Table summarising precursors, sequences and affinity of endogenous opioid peptides.

59

1.4.2 Role of the Opioid System in Emotional Regulation

The role of the opioid system in emotional regulation has been elucidated using both

pharmacological knock-out (KO) studies and selective opioid receptor drugs (for an extensive

review see Lutz and Kieffer, 2013). Constitutive KO mice for MOPr, DOPr and KOPr were

created almost two decades ago (see Kieffer and Gavériaux-Ruff, 2002) and characterized for

mood states amongst other characteristics. In KO mice lacking all three opioid receptors, no

obvious developmental defects were evident and all mice were fertile (see Kieffer and

Gavériaux-Ruff, 2002). This suggests that the opioid system is not crucial for development, at

least when animals are raised under home cage conditions. This produced the notion that the

opioid system is primarily recruited under challenging situations (see Kieffer and Gavériaux-

Ruff, 2002).

Data from MOPr KO mice are in agreement that this receptor represents a key molecular

component in reward processing of both natural stimuli and drugs of abuse (see Lutz and

Kieffer, 2013). MOPr agonists have been shown to have strong rewarding properties, which is

likely to be linked to the well-known antidepressant effect of opioids (Tenore, 2008). However,

two studies reported both decreased anxiety and depressive-like behaviour in MOR KO mice,

indicating that the MOPr play a paradoxical depressant role in regulating emotional responses

(Filliol et al., 2000; Yoo et al., 2004). MOPr-mediated mechanisms of mood control are

therefore complex and may relate back to the pharmacology of opioid receptors (see Lutz and

Kieffer, 2013).

Genetic deletion of DOPr has been reported to result in a significant and consistent increase

in anxiety and depressive-like behaviours (Filliol et al., 2000), a phenotype also partly seen in

PENK KO mice (König et al., 1996). Further studies have confirmed that the endogenous

60

activity of enkhephalin/DOR system positively modulates mood states (Pradhan et al., 2011).

Unlike DOPr KO mice, KOPr KO mice did not differ from controls in behavioural models of

depressive-like behaviours (Filliol et al., 2000). However, it is now well known KOPr receptor

activity is initiated by stressors such as repeated forced swim and may play a role in stress-

induced psychopathology (Knoll et al., 2010). Taken together, these studies indicate that MOPr

DOPr and KOPr play highly distinct roles in regulating emotional responses.

Studies using selective MOPr DOPr and KOPr drugs have confirmed the differential roles of

opioid receptors in emotional regulation. The antidepressant activity of DOPr agonists is well

documented (Naidu et al., 2007; Javelot et al., 2010; Saitoh et al., 2011; Yang et al., 2011)

consistent with the increased depressive like behaviour seen in DOPr KO mice. Similar to

DOPr but less well documented acute pharmacological activation of MOPr reduces depressive-

like behaviour in some behavioural models of depression (Tejedor-Real et al., 1995; Rojas-

Corrales et al., 2004; Yang et al., 2011) but not all (Zhang et al., 2006). Studies reporting the

antidepressant like effect of MOPr agonists are in apparent contradiction with the decrease in

depressive like behaviour seen in MOPr KO mice, further confirming the paradoxical role of

MOPr in mood regulation. Systemic administration of KOPr drugs also produced mood altering

effects in classical models of depressive-like behaviours. KOPr agonists showed depressive-

like behaviour while antagonists were associated with antidepressant-like behaviours (Mague

et al., 2003; Beardsley et al., 2005). In summary, DOPr agonists and KOPr antagonists have

antidepressant-like effects, in contrast data from MOPr investigations appear more complex.

61

1.4.3 Opioid Receptor Development

Opioid receptors have been shown to have a marked differential development in the rat. The

development of the opioid system has been extensively studied using selective ligands for each

receptor subtype (see McDowell and Kitchen, 1987). MOPr were found to be present at birth

in rat forebrain, and after an initial decline in concentration during the first few days they

undergo a subsequent rapid increase (Spain et al., 1985). The number of MOPr sites continues

to rise during the first week after birth, with the most pronounced rate of increase in their levels

occurs during the second and third postnatal weeks. A similar pattern of MOPr development

is seen in the striatum. Unlike MOPr, DOPr are present in only very low levels, if at all during

the first postnatal week. The DOPr protein in the brain is not detectable before birth (McDowell

et al., 1986) and the expression of the DOPr gene is delayed (Leslie et al., 1998). From the

second through the fourth week a near linear increase in their concentration occurs with 75%

of the total DOPr population developing by the normal age of weaning at day 21 (McDowell

et al., 1986), which correlates well with the rise in MOPr sites (Spain et al., 1985; McDowell

and Kitchen, 1987). The development of KOPr differs slightly in that these receptors can be

found in the forebrain and hindbrain regions from birth, whereupon hindbrain levels rise

substantially. After postnatal day 14 the density of forebrain KOPr increases and hindbrain

levels decline. Indeed, the second postnatal week appears to be of high importance for the rat

brain maturation as the development of dopaminergic and cholinergic receptors are also

increasing at this time (Spain et al., 1985).

From studies carried out in human brains it is clear that the temporal aspects of the development

of the opioid receptors observed in rats is mirrored in human brain ontogenesis. MOPr and

KOPr were both found to be present and develop during gestation, whilst DOPr are not

detectable up to birth (Magnan et al., 1989). Considering that the opioid system shows marked

62

postnatal development in both man and rodent, it is possible that environmental early life

influences may be very important developmental determinants for this system. As milk is the

main source of nutrition during early postnatal development, and both human and bovine milk

have been shown to release opioid peptides upon digestion, manipulations to weaning time or

periods of exposure to milk in early life could possibly effect the normal development of the

opioid system as well as other developing neurochemical systems.

1.4.4 Role of Weaning in Opioid Receptor Development

The development of the opioid system has been shown to be vulnerable to changes when the

process of weaning is delayed at least in the rat (Muhammad et al., 1993; Goody et al., 2001).

A study carried out by Muhammad et al., (1993) was the first to demonstrate that weaning rat

pups from their mothers was a critical process in regulating the postnatal development of δ-

opioid receptors (Muhammad et al., 1993). Animals left with their mothers after the normal

age of weaning at day 21, showed delayed development of DOP receptor mediated Stress-

induced antinociception (SIA) (Muhammad et al., 1993).

Stress-induced analgesia in man or SIA in experimental animals, are adaptive physiological

responses that can be provoked by a number of different stressors including immobilisation,

cold, electroshock and swimming. Depending on the environmental stimuli, it can be divided

into opioid mediated and non-opioid mediated types (Bodnar, 1990). Swimming-SIA is

particularly useful as it can differentiate between opioid and non-opioid forms by varying the

duration and temperature of the swim session. It has been shown that the opioid receptor

subtype which mediates the opioid form of SIA differs in the neonate in comparison to the

adult (Jackson et al., 1989; Kitchen et al., 1990). Studies using Naloxone (MOPr antagonist)

and the DOPr selective antagonist ICI 174864 and naltrindole have shown that SIA is mediated

63

by μ-receptors in 20-day old rat pups, but by day 25 it is predominantly δ- sites which mediate

this behaviour (Muhammad et al., 1993). It was then shown that weaning plays a critical role

in this transition. Muhammad et al., (1993) showed this by testing for SIA responses in

normally weaned (PND21) and delayed weaned pups (PND25). To ensure that opioid SIA is

induced, rats were forced to swim for 3 minutes and then tested for SIA responses by the tail

immersion test on PND25. The results showed SIA to be partially reversed following pre-

treatment with the δ-selective antagonist naltrindole in pups weaned at PND21, which indicates

an active DOPr component. However, pups housed with their mother until PND25 and then

tested continued to exhibit MOPr mediated swim SIA, which was insensitive to naltrindole

(Muhammad et al., 1993). Therefore the delay of weaning is able to halt the MOPr to DOPr

transition. When the delayed weaned pups were tested again at day 30, naltrindole was able to

reversed swim SIA indicating that the DOP receptors were now mediating this response

(Muhammad et al., 1993).

The role of weaning on DOPr development has been further implicated by receptor binding

studies analysing the number of labelled DOPr sites in the weaned and delayed weaned rats

(Kitchen et al., 1995a). In this study, rats were weaned on PND21 or delayed weaned on

PND25 before being killed on PND25 (therefore considered non-weaned) (Kitchen et al.,

1995). Using the DOPr agonist, [H3] [D-Ala2 ] deltorphin I (DELT I), it was found that levels

of binding were significantly higher in weaned rats compared to their non-weaned/ delayed

weaned counter parts. This difference in binding was localized to the deep layers of the frontal

parietal cortex and the pontine nucleus (Kitchen et al., 1995). However, in-situ hybridization

experiments showed no difference in DOPr mRNA density in the regions in which binding

differences were observed between the two groups (Kitchen et al., 1995). These discrepancies

might indicate the existence of different delta receptor subtypes expressed in those regions,

which were not recognized by the cRNA probe used for the in-situ hyberdization. Also it is

64

possible that a subpopulation of delta receptors is coupled to another G-protein in those regions

forming heterodimers and that weaning stimulates this coupling process which is then

recognized by DELT 1. In either case, it is evident that the weaning stimulates the development

of a subpopulation of DOPr. Taken together with the results from the SIA experiments it is

clear that delaying weaning influences the development of the opioid system by preventing the

MOPr to DOPr switch, and preventing the development of the final 25% of DOPr in the brain.

Furthermore, a study by the same group went on to address whether loss of suckling, stress due

to maternal removal or loss of a nutritional component is the crucial determinant of weaning-

induced activation of DOPr (Goody et al., 2001). By measuring the δ-receptor component in

SIA responses, in animals provided with lactating and non-lactating surrogate mothers it was

demonstrated that a nutritional component rather than maternal presence is the critical factor

in this transition (Goody et al., 2001).

1.4.5 Role of Milk Caseins in Opioid Receptor Development

Following on from the set of experiments carried out by Goody et al., (2001) (see Section

1.4.4) the group then went on to investigate the possible mechanisms underlying the delayed

development of DOPr mediated by weaning time. It was hypothesised that this delay could be

caused either by a continuation of suckling/continued consumption of milk casein, or continual

maternal presence, both of which are associated with delayed weaning (Goody et al., 2001).To

determine the exact cause, the following groups were established and measured for the DOPr

component in SIA responses: Weaned on PND21, delayed weaned on PND25 with either their

own mother, a lactating surrogate mother or replacement with a non-lactating surrogate mother

on PND21, and finally three groups in which the dams were removed on PND21 and replaced

with either casein rich, casein spiked or casein free milk substitutes. The results showed that

replacement with a non-lactating surrogate until PND25 did not delay the development of

65

DOPr, as antagonism of SIA by naltrindole was still present as seen in rats normally weaned

on PND 21 (Goody and Kitchen, 2001). Therefore this indicated that the cause of the delayed

DOPr development was not due to a continuation of suckling behaviour by the pups, which

would have still occurred at the non-lactating nipple, indicating a nutritional component from

maternal milk. Indeed, this was demonstrated with the group provided with a casein-free milk

substitute. This group showed SIA was antagonised by naltrindole administration,

demonstrating that DOPr had developed normally. In contrast the groups provided with casein-

rich and casein spiked milk substitutes did not show reversal of SIA upon naltrindole

administration, hence indicating that DOPr development had been delayed (Goody et al.,

2001).

In summary, the results of the casein containing diets mirrored the results of the groups which

were delayed weaned with either their own mother or a lactating surrogate. Naltrindole was not

able to antagonise SIA in these groups, indicating that DOPr development was hindered

(Goody et al., 2001). In contrast, the results from the groups provided with casein-free diet,

mirrored the results of the group normally weaned on PND21 and the group that was housed

with a non-lactating surrogate until PND25. Naltrindole was able to antagonise SIA in these

groups, indicating that DOPr development occurred normally (Goody and Kitchen, 2001).

These findings strongly implicate milk casein in influencing the development of the opioid

system with focus on the DOPr. Although the mechanism underlying this effect is unknown,

some suggestions have been made including the involvement of the gut-brain axis (see Section

1.5).

66

1.4.6 The Oxytocin System

The neuropeptide oxytocin (OT) was first discovered in 1906 when it was observed that

extracts from human posterior pituitary gland were able to induce uterus contractions in a

pregnant cat. The name oxy, derives from the Greek words “ωκνξ τoκoxξ,” meaning “swift

birth” (Romano et al., 2016). OT is a 9 amino-acid neuropeptide which is synthesised in the

brain and in the periphery (Gimpl et al., 2001) and plays a key role in the modulation of

emotional and social behaviours (see Neumann and Landgraf, 2012). In the brain, the

hypothalamic-neurohypophysical system represents the major OT neurosecretory system and

consists of the paraventricular (PVN) and supraoptic (SON) nuclei (Gimpl et al., 2001). Within

the PVN, two populations of OT neurons have been identified: “magnocellular” and

“parvocellular” neurons. OT is mainly synthesised in the magnocellular neurons of the PVN

and SON (Gimpl et al., 2001). Oxytocinergic magnocellular neurons terminate in the posterior

lobe of the pituitary gland amongst other regions of the brain. Once activated, magnocellular

oxytocinergic neurons release OT from the nerve terminals to the posterior pituitary from

which OT is then released into the blood steam to elicit its effects on organs expressing OT

receptors (OTr) (Gimpl et al., 2001).

The OTr belongs to the G-protein coupled receptor family with seven transmembrane domains

(Kimura et al., 1992). OTr have been identified in several brain regions as well as peripheral

organs. In the rat CNS, OTr are located in several regions including the olfactory system,

cortex, thalamus, basal ganglia, ventromedial region of the hypothalamus, central amygdala,

brain stem and spinal cord (Gimpl et al., 2001). Interestingly the expression density of OTr has

been shown to change according to age. Tribollet et al., (1989) have shown that in male and

female rats the distribution and number of OT binding sites undergo changes during

development. From PND5 to PND16, the distribution of binding sites remained essentially

67

unchanged but differed significantly from that characteristic distribution seen in an adult. The

change-over from the "infant pattern" to the "adult pattern" was reported to occur during two

critical periods: the first change took place between PND16 and PND22, a time corresponding

to the pre-weaning period; the second change occurred after PND35 and thus coincided with

the onset of puberty. During the first transition period, binding was reduced or disappeared in

several areas intensely labelled at earlier stages, in particular, in the cingulate cortex and the

dorsal hippocampus. At the same time, binding sites appeared in the ventral hippocampus. At

puberty, high densities of OT binding sites appeared in the ventromedial hypothalamic nucleus

and the olfactory tubercle (Tribollet et al., 1989). Thus like the opioid system, the OT system

appears to also have a differential ontogenic pattern that may be susceptible to insult from

environmental stimuli during development. Indeed environmental stimuli (Bales et al., 2012)

and experiences during early development such as social, maternal care and stress are known

to shape the OT system possibly through epigenetic modification of the OTR (Carter, 2003;

Bales et al., 2011).

1.4.7 Role of the Oxytocin System in Emotional Regulation

OT is currently attracting considerable scientific attention as a result of its CNS functions. OT

regulates a variety of behaviours including maternal care and aggression, pair bonding, sexual

behaviour as well as anxiety-related behaviours and stress coping (for a review see Romano

et al., 2016).The role of OT in regulating these behaviours makes this neurotransmitter system

of the brain a promising target for treatment of numerous psychiatric illnesses such as anxiety

disorders and depression (Romano et al., 2016).

The synthesis and release of OT within the brain was found to be driven by anxiogenic, stressful

and notably social (both positive and negative) stimuli (Landgraf et al., 2004; Neumann, 2008).

For example, exposure of male rats to novelty, forced swimming, or social defeat rapidly

68

increases OT release into blood but also within the PVN and/or SON and in other limbic brain

regions, such as the central amygdala or septum (Engelmann et al., 2004; Neumann, 2007).

Similarly, increased OT release into blood (Neumann et al., 2001) and within the PVN and

central amygdala (Bosch et al., 2004) has also been found in female rats exposed to

psychosocial stress (maternal defeat by an aggressive lactating resident dam).

Once released in the brain regions involved in stress and anxiety regulation, OT exerts

anxiolytic effects and modulates neuronal functions related to physiological stress responses,

mainly at the level of the PVN and amygdala. Indeed pharmacological intervention studies

have shown that acute or chronic central administration of synthetic OT exerts an anxiolytic

effect both in female and in male rats and in mice (Ring et al., 2006; Blume et al., 2008). In

order to elucidate the role of the endogenous brain OT in regulation of anxiety like behaviour

in more detail, OTr antagonists were used to block these receptors, specifically within the

central amygdala (Neumann, 2002). However, the effects of antagonist administration on

anxiety were only observed in pregnant or lactating but not virgin female or male rats

(Neumann et al., 2000; Lonstein, 2007). This indicates that endogenous OT does not seem to

play a major role in the maintenance of a basal level of anxiety, but rather comes into play

during psychosocial or physiological activation of the system as seen in periparteum period in

females (Neumann et al., 2000; Lonstein, 2007). Behavioural data from transgenic mice

lacking OT provide further support for the anxiolytic importance of brain OT in anxiety

regulation, as female knockout mice show greater anxiety-related behaviour than their wildtype

counterparts on the elevated plus maze (Amico et al., 2004).

Due to the high degree of comorbidity between anxiety and depression disorders, common

mediators are likely to underlie both conditions. Indeed, in addition to its anxiolytic effect,

synthetic OT administration was shown to shift stress-coping in rodents towards a more active

69

coping style, after either central or peripheral administration indicating anti-depressive like

effects (for a review see Neumann and Slattery, 2016). Furthermore there is also preclinical

and clinical evidence that OT may also contribute to the improvement of other depression-

related symptoms including sexual dysfunction, sleep disturbances and anhedonia (Slattery et

al., 2010).

Given the importance of the oxytocinergic system in emotional regulation, combined with

evidence that this system shows developmental changes, it is likely weaning age or duration of

exposure to milk proteins may result in changes to the normal development of the oxytocinergic

system, thus impacting emotional behaviour.

1.4.8 The Role of Weaning on Oxytocin Receptor development

Recent evidence suggests that weaning can influence OTr levels in the ventromedial

hypothalamus which is likely to be involved in emotional behavioural regulation during early

postnatal development. Curley et al., (2009) has shown that duration of mother-pup interaction

during early postnatal development influences the expression of behaviour as well as OTrs in

mice (Curley et al., 2009). Similarly, Farshim et al., (2016) have shown that weaning rat pups

from their mothers on PND25 as opposed to the normal age of weaning (PND21) resulted in

depressive-like behaviour as shown by the forced swim test (FST). Furthermore this study

demonstrated that weaning influences stress-induced modulation of OTr density in delayed

weaned rat pups, suggesting that weaning time affects the central oxytocinergic system when

delayed weaned animals are exposed to stress. These changes in OTr density were observed in

the basal medial and basal lateral amygdala (Farshim et al., 2016), regions involved in

emotional processing (LeDoux, 2000). The stress from the forced swim test induced a marked

down-regulation of OTrs in weaned rats which was absent in non-weaned rats (Farshim et al.,

70

2016). There is evidence to suggest that the high degree of plasticity in the OTr system may

constitute a protective adaptive coping mechanism in response to various insults (e.g. stress) to

shift brain function and behaviour (Neumann et al., 2012). Therefore given the key role of OT

in the amygdala in modulating mood, it is possible that changes in the oxytocinergic stress

responses induced by weaning may represent a homeostatic stress-coping mechanism that is

found to be absent in animals that are delayed weaned. The mechanism underlying these effects

remains unclear and could possibly be linked to prolonged exposure to milk caseins.

1.4.9 The Role of Weaning on other CNS Systems

In both humans and rodents, neurotransmitter systems implicated in modulation of emotive

behaviours, such as serotonin (5-hydroxytryptamine: 5-HT), dopamine (DA) and corticotropin-

releasing factor (CRF), do not mature fully until early to late adolescences (Andersen, 2003;

Vazquez et al., 2006; Lukkes, 2009). Therefore, like the opioid and oxytocinergic systems, it

is possible that normal development of neurochemical systems involved in regulation of stress

responses, mood and anxiety states during adulthood are susceptible to disruptions by early life

factors such as prolonged exposure to milk beyond the normal age of weaning. Moreover,

clinical and preclinical studies suggest that perturbations to these neurochemical systems

during early-life appear to have long-lasting consequences on emotional behaviours in rodents

and mental health in humans (Gutman et al., 2003; Nemeroff, 2004).

Indeed studies have shown that weaning age does affect the development of the above

mentioned neurotransmitter systems. Serotonin, a monoamine neurotransmitter of the

mesolimbic system mediates its effects through at least 13 distinct G protein-coupled receptors.

5-HT receptors are separated into seven major classes (5-HT1 – 5-HT7), which were subdivided

further based on structural and operational features (Hoyer et al., 1994). It was reported that

mice weaned early on PND14 displayed increased aggression and shown to express reduced

71

levels of 5-HT1B but not 5-HT1A receptor mRNA in the hippocampus in comparison to mice

weaned on PND21 (Nakamura et al., 2008). Furthermore, mice weaned on PND16 reported

lower brain weights but higher myelin basic protein in brain extracts (Kikusui et al., 2007)

compared to mice weaned on PND30. This suggests that early weaning decreases myelin

formation and increases anxiety like behaviour (Nakamura et al., 2008).

Like serotonin, dopamine is a monoaminergic neurotransmitter of the mesolimbic system and

is involved in many reward related behaviours as well as being implicated in number of

neuropsychiatric and neurodegenerative diseases (Grace, 2016). DA carries out its effects by

binding to five distinct G-protein coupled receptors including D1A-D-D5. Early studies carried

out by Sharman et al., (1982) investigated the impact of weaning time on the monoaminergic

DA system. It was reported that piglets weaned early exhibited stereotyped belly-nosing

behaviour which was shown to be associated with reductions in the metabolism of DA in the

regions of the brain receiving DAergic inputs (Sharman et al., 1982; Mann et al., 1983). To

date no other studies have been published showing a direct link between the DAergic system

and weaning time.

Finally, CRF is a neurotransmitter involved in integrating multiple components of stress

responses via the hypothalamic-pituitary-adrenal (HPA) axis. The HPA axis is a major part of

the neuroendocrine system and is a set of feedback interactions which occur between the

hypothalamus, pituitary gland and the adrenal glands (see Hostinar et al.,2014). Systemic

stressors can act directly on pituitary cells and result in the release of CRF as well as ACTH,

which then stimulates the release of glucocorticoids with corticosterone being released in

rodents and cortisol released in humans (see Lightman, 2008) from the adrenal glands. The

release of glucocorticoids or corticosterone are the final effectors of the HPA axis. They also

act as the terminators of stress responses via a negative feed-back loop, binding to

72

glucocorticoid receptors found in the hypothalamus and pituitary gland (Pratt, 1989).

Variations in weaning patterns have been shown to be associated with altered responses to

stress as well as modifications of the HPA axis (Cook, 1999). In a study carried out by Kikusui

et al (2006) male mice weaned on PND14 showed higher expression of glucocorticoid receptor

mRNA expression in the hippocampus compared to mice weaned on PND21 (Kikusui et al.,

2006) . They also reported higher levels of corticosterone in male mice weaned on PND14

compared to animals weaned normally on PND21, with early weaned animals also exhibiting

increased anxiety-behaviour in the elevated plus maze test, which is shown to be linked to

higher levels of corticosterone secretion than mice weaned on PND21 (Kikusui et al., 2006).

Taken together, this evidence indicates that disruptions to the age of weaning can affect the

development of a number of neurotransmitter systems in the brain and alter behaviour.

However these studies have not clearly indicated whether the observed changes are due to

maternal deprivation at different time points or due to nutritional changes which occur with

weaning. As a result further investigation is required into whether alterations to weaning age

result in these behavioural changes due to alterations to periods of milk protein exposure in

early postnatal life.

1.5 The gut-Brain Axis

One of the proposed mechanisms by which weaning age or period of exposure to milk rich

diets in early postnatal life influence brain neurochemistry and ultimately behaviour, is via the

gut-brain axis. Communication between the gut and the brain, which is regarded as the gut-

brain axis is a well-known bidirectional neurohumoral communication system. There is

growing evidence linking the gut microbiota with brain function and behaviour. The high co-

morbidity between stress related psychiatric symptoms such as anxiety, with GI disorders

including irritable bowel syndrome (IBS) and inflammatory bowel disorder (IBD) is strong

73

evidence of a gut-brain axis. The reciprocal impact of the gastrointestinal tract on brain function

has been recognized since the middle of the 19th century through the pioneering work of Claude

Bernard and Ivan Pavlov amongst a few others (for a review see Cryan and Dinan, 2012). They

were amongst the first to emphasize the primacy of brain processing in the modulation of gut

function. It is now known that the gut-brain axis provides a bidirectional homeostatic route of

communication that uses neural, hormonal and immunological routes and that dysfunction of

this axis can have pathophysiological consequences (Grenham et al., 2011; Mayer, 2011).

There is also growing realization that the bi-directional communication between the digestive

tract and the brain plays a key role in maintaining brain health and stress responses. An

increasing number of studies using Germ free (GF) animals or antibiotic or probiotic

administration have highlighted the involvement of the gut microbiota in the gut-brain axis.

1.5.1 Gut microbiota

The gut microbiota consists of a complex mixture of microorganisms that inhabit the

mammalian GI tract, most of which reside in the distal gut. In humans, the distal gut contains

10 14 – 10 15 bacteria, that is, there are 10-100 times more bacteria in the gut than eukaryotic

cells in the human body (Gill et al., 2006). Furthermore, it is estimated that the gut microbiome

has in excess of 1000 species and 7000 strains, at the genomic level, the global human gene

complement is outnumbered by a factor of at least 100 when compared to the gut microbiome

(Cryan et al., 2013). The genes in the combined genomes of these bacteria are known

collectively as the intestinal microbiome.

Over the past several years, substantial advances have been made in the ability to assess the

microbiota composition. Recently developed molecular and metagenomic tools combined with

culture-based approaches have allowed for a better understanding of the structure and function

of the microbial gut community. Several bacterial phyla are represented in the gut. The two

74

most prominent phyla in the human microbiota are Firmicutes and Bacteroides, accounting for

at least 70-75% of the microbiome (Eckburg et al., 2005). Proteobacteria, Actinobacteria,

Fusobacteria, and Verrucomicrobia are also present but in reduced numbers (Eckburg et al.,

2005). It is becoming clearer how the diversity and distribution of these prominent phyla

contribute to health and disease. There seems to be a balance that confers health benefits, while

an alteration in beneficial bacteria can negatively influence the well-being of an individual

(Grenham et al., 2011).

The composition of the gut microbiota is highly host-specific and evolves throughout the

lifetime. The initial colonization of the gut was believed to be a strictly postnatal event.

However, recent studies characterizing the microbiota of meconium and fecal samples obtained

from preterm babies during the first three weeks of life, found Firmicutes was the main phylum

detected in the meconium and Proteobacteria was abundant in the feces (Rodríguez et al.,

2015). Following birth, the human intestine is rapidly colonized by an array of microbes. The

factors known to influence colonization during this period include, mode of delivery (vaginal

birth vs. assisted delivery), diet (Breast milk vs. formula milk), sanitation and antibiotic

treatment (Rodríguez et al., 2015) (see Figure 1.4). The first colonizers after birth, facultative

anaerobes, create a new environment that promotes the colonization of strict anaerobes such as

Bacteroides, Clostridium, and Bifidobacterium spp. The intestinal microbiota of a new born is

characterized by low diversity and a relative dominance of the phyla Proteobacteria and

Actinobacteria. The third stage of microbial colonization begins once weaning and introduction

of solid food commences (Mackie et al., 1999), with the microbiota becoming more diverse

and the emergence and dominance of Firmicutes and Bacteroidetes occurring (Eckburg et al.,

2005; Qin et al., 2010; Bäckhed, 2011). By the end of the first year of life, infants possess an

individually distinct microbial profile, converging toward the characteristic microbiota of an

adult. By 2-5 years of age, the microbiota fully resembles that of an adult in terms of

75

composition and diversity (Koenig et al., 2011; Yatsunenko et al., 2012). Therefore, the first 3

years of life represent a critical period when the intestinal microbiota, a vital asset for health

and neurodevelopment (Borre et al., 2014) is established and its alteration during this period has

the potential to profoundly affect host health and development (Figure 1.4).

Microbial colonization in rat pups has been less comprehensively studied compared to the

ecosystem in human infants; however the principle colonization of intestinal microflora in

rodents resembles that in humans (Umesaki et al., 2000). As with humans, the neonatal rat

microbiota is far less diverse compared to the mothers, but continues to become more complex

as time from birth increases. Bacterial groups favoured by mother’s milk such as Lactobacilli

and Bifidobacteria are the first to colonize the neonatal small intestine. After weaning,

anaerobic bacteria, such as Bacteroidaceae and Lactobacilli, become the dominant microbes.

Pregnancy Birth Infancy Adult

Maternal MicrobiotaHealthy status

Lifestyle Antibiotic

Geographical locationFamily Environment

Breastfeeding vsFormula

Diet Lifestyle

Window Of Opportunity for Microbial Modulation

Vaginal vs. C-section GeneticsComplementary foodDuration of lactation

Living environmentMedication

Short and Long Term Health Effects

Figure 1.4: Factors influencing the infant gut microbiota development, as well as the adult

microbiota. The first 3 years of life represent the most critical period for modulation of the gut

microbiota by environmental factors such as diet. Taken and adapted from (Rodríguez et al., 2015).

76

Inoue and Ushida, (2003) investigated the development of the rat intestinal microbiota from

PND18- 40. The authors showed that the main changes occur at the time of weaning, i.e.

PND21-22 with significant increase in the number of bacteria belonging to the phyla of

Firmicutes (Inoue et al., 2003).

As shown in Figure 1.4, one of the factors influencing the composition of the gut microbiome

in early postnatal life is diet. Therefore, it is possible that age of weaning and period of exposure

to milk proteins in infancy may result in changes to the gut microbial signature. Indeed Farshim

et al., (2016) have shown that delay weaning rat pups on PND25 resulted in significantly lower

Clostridium histolyticum bacterial groups compared to animals weaned normally on PND21.

Furthermore PND25 weaned pups exhibited a marked increase in this bacterial group following

exposure to the FST (Farshim et al., 2016). This study indicates that weaning age does indeed

influence the composition of the gut microbiota. Moreover, milk caseins have been shown to

have numerous interactions with bacteria throughout the GI tract. These interactions have been

shown to be either direct or indirect (Miller et al., 1990). For example the production of β-

lactamase by Enterobactercloacae spp is enhanced by casein. This results in more rapid

fermentation of carbohydrate to organic acids which then decrease the intraluminal pH of the

gut. This effect has been proposed as an important component in the initiation of neonatal

inflammatory bowel disease (Miller et al., 1990). Therefore prolonged exposure to milk caseins

in early postnatal life could result in changes to the gut microbiota through its effects on luminal

pH.

77

1.5.2 Gut Microbiota Metabolism

The Gut microbiome is involved in a variety of different essential functions which aid the host

to remain healthy. It is involved in direct chemical communication through signalling pathways

which connect many organs and tissues around the body including the brain (Cryan et al., 2011;

Nicholson et al., 2012). The microbial richness and diversity found in the gut encodes unique

metabolic functions that are absent from the host providing a metabolic extension to the

relatively limited host genome (see Ottman et al., 2012). This includes the capacity for the

metabolism of chemicals derived from endogenous mammalian metabolism, nutrients and

other xenobiotics. Products of microbial metabolism act as signalling molecules and can in turn

influence the host’s own metabolism (Tremaroli et al., 2012). These small molecule

messengers include bile acids, biogenic amines, products of choline degradation and short

chain fatty acids (SCFAs).

Signalling molecules are generated by GI bacteria in a number of ways, for example proteins

from the diet may be catabolized by microbial proteinases and peptidases in cooperation with

host proteinases (Hollister et al., 2014). The generation of free amino acids by protein

catabolism in the stomach and small intestine provide substrates for luminal conversion of

amino acids. These amino acids can be converted by decarboxylation reactions and catalyzed

by the GI bacteria to form different signalling molecules (Hollister et al., 2014). As an example,

L-histidine can be converted to the biogenic amine, histamine, by histidine decarboxylases

produced by GI bacteria. Histamine can suppress production of inflammatory cytokines by

signalling through the histamine receptors (Thomas et al., 2012). Other examples of

microbiome-mediated amino acid decarboxylation include the conversion of glutamate to g-

aminobutyric acid, tyrosine to tyramine, and phenylalanine to b-phenylethylamine (Hollister et

al., 2014). Bacterial genes also encode glutamate decarboxylases, which can generate the

78

neurotransmitter and potential immunomodulin g-aminobutyric acid in the intestine (De Biase

et al., 2012). Furthermore, complex carbohydrates such as dietary fibre can be digested and

subsequently fermented in the colon by gut microorganisms into SCFAs such as n-butyrate,

acetate and propionate. These metabolites not only act as energy sources for epithelial cells,

but have been shown to control colonic gene expression and metabolic regulation, for example

MacFabe et al., (2011) have reported that SCFAs regulate the function of the enzyme, histone

deacetylase and have neuroactive properties which can stimulate the sympathetic nervous

system and may modulate rodent social behaviour via these mechanisms (MacFabe et al.,

2011).

In addition to the generation of small signalling molecules, bacteria can produce antimicrobial

peptides which confer resilience or resistance to infection by enteric pathogens (Kamada et al.,

2013). The production of antimicrobial compounds by the GI microbiome might therefore

prevent expansion of enteric pathogens or indigenous pathobionts. The generation of

antimicrobial peptides, immunomodulatory compounds, and neurotransmitters by the

microbiome could affect the performance of the mucosal immune system and enteric nervous

system, and may therefore modulate general health, mood and behaviour of the host via these

mechanisms.

Metabolic profiling of biofluids, such as urine, plasma or faecal water, using high-resolution

spectroscopy methods is an effective tool for measuring metabolites from both microbial and

mammalian origin (see Section 1.6). When this technique is used in conjunction with animal

models of varying gut microbial states, this systems biology approach can be used to study the

metabolic cross-talk between the host and its gut microbiota and provide insight into the

biochemical influence of the gut microbiome on gut, liver and brain metabolism and

functioning.

79

1.5.3 Bidirectional Gut-Brain Axis

The idea of a bidirectional communication between the gut and the CNS has long been

recognized. More recently, this gut-brain axis has been extended to include contents of the

intestinal lumen and is now referred to as “microbiota-gut-brain axis’, highlighting the

increasingly appreciated importance of the bacteria in this communication (Luczynski et al.,

2016).

The microbiota-gut- brain axis consists of a bidirectional communication network that

monitors and integrates gut functions and links them to cognitive and emotional centres of the

brain. A number of potential mechanisms have been suggested by which microbiota can affect

brain function and consequently behaviour. One of the prime proposed mechanisms is via the

immune system, which has been shown to shape the composition of the gut microbiota, and

the microbiota in turn shapes the immune system (Dantzer et al., 2008). Some rodent infection

studies using subclinical doses of pathogenic bacteria have indicated that immune system

activation is not always necessary in causing changes in brain neurochemistry and behaviour

(Lyte et al., 2006). While other studies have shown clear evidence of the importance of an

intact and normally functioning immune system in gut-brain communication (El Aidy et al.,

2015). Another potential communication pathway is via the vagus nerve. The vagus nerve

regulates several organ functions including gut motility (Cryan et al., 2012). Approximately

80% of the nerves fibres are sensory, conveying information about the state of the body’s

organs to the CNS. Numerous studies have shown the effect of gut-microbiota on brain function

to be dependent on vagal activation (Goehler et al., 2008; Bravo et al., 2011; Bercik et al.,

2011), while others have found evidence of vagal independence, as vagotomy failed to

influence the effect of antimicrobial (ATM) treatment on brain function (Bercik et al., 2010).

80

The endocrine system too plays an important role in the ability of bacteria to alter brain function

(Lyte et al., 2006). Many studies have demonstrated changes in endocrine signalling in models

of intestinal dysbiosis (Sudo et al., 2004; Bercik et al., 2014) and also the ability to alter

intestinal microbial contents following stress (De Palma et al., 2015; Golubeva et al., 2015).

For example studies that have used maternal separation models have shown that neonatal stress

leads to long-term changes in the composition and diversity of the gut microbiota (Cryan et al.,

2011). In addition to altering the gut microbiota composition, chronic stress is known to

increase the permeability of the intestinal barrier, making it “leaky” and subsequently give

bacteria the opportunity to translocate the intestinal mucosa and directly access both immune

and neuronal cells of the enteric nervous system (Gareau et al., 2011). Other possible

mechanisms of communication include microbial metabolites some of which have neuroactive

properties (Thomas et al., 2012). Figure 1.5 illustrates the main mechanisms and pathways

involved in the bidirectional microbiota-gut-brain axis communication.

81

Hypothalamus

Pituitary

CRF

ACTH

Adrenals

Cortisol

Gut microbiota

Immunecells

Vagus nerve

Mood, Cognition,emotions

SCFAs

EntericMuscles

CirculationCytokines

Intestinal Lumen

Neurotransmitters

Tryptophan metabolism

Figure 1.5: Potential pathways involved in the bidirectional communication between the gut

microbiota and the brain. Taken and adapted from (Cryan and Dinan 2012).

A number of direct and indirect pathways exist through which the gut microbiota can interact with the

gut-brain axis. They include endocrine (cortisol), immune (cytokines) and neural (vagus nerve and

enteric nervous system) pathways. The brain is speculated to use these same pathways to influence the

composition of the gut microbiome. For example, under stress conditions the hypothalamus-pituitary-

adrenal axis regulates cortisol secretion, in turn cortisol can affect immune cells (including cytokine

secretion) both act locally in the gut and systemically. Cortisol can also alter both the permeability of

the gut and barrier function, and change gut microbiota composition. Conversely, the gut microbiota

and probiotic agents can alter the levels of circulating cytokines, and this can have a significant effect

on brain function. The vagus nerve is strongly implicated in relaying the influence of the gut microbiota

to the brain. Moreover, short-chain fatty acids (SCFAs) are neuroactive bacterial metabolites derived

from dietary fibres that can also modulate brain and behaviour.

82

1.5.4 Gut-Brain Axis and Emotional Behaviour

Alterations in the bidirectional brain-gut microbiota interactions are believed to play a role in

the pathogenesis of behavioural disorders. There is accumulating evidence from rodent

behavioural studies to suggest that manipulation of the gut microbiota results in altered

behavioural responses. Several experimental approaches have been used to alter the gut

microbiota, including antibiotics, probiotics, faecal microbial transplantation, infection with

pathogenic bacteria and Germ Free (GF) animal models. Table 1.7 provides a detailed

summary of the behavioural findings generated by experiments using these approaches to alter

the gut microbiota.

When viewed together, the reported evidence in Table 1.7 indicate that rodent behavioural

responses are impacted when the bacterial status of the gut is manipulated. Depressive-episodes

are associated with dysregulation of the HPA axis. As discussed in Section 1.4.9, the HPA

axis is the principle endocrine pathway, and activation of this system takes place in response

to a variety of physical and psychological stressors. A direct link between microbiota and HPA

reactivity is evident from the reported literature. The pioneering work of Sudo et al., (2004) in

this field was the first to demonstrate mild stress induced an exaggerated release of

corticosterone and ACTH in GF mice when compared to specific pathogen free (SPF) mice

(Sudo et al., 2004). Recent work has reproduced these findings (Crumeyrolle-Arias et al.,

2014). While several independent laboratories have reported adult GF mice display a reduced

HPA response to stress (Heijtz et al., 2011; Neufeld et al., 2011; Arseneault-Bréard et al., 2012;

Clarke et al., 2013). It is important to note that work studying the role of microbiota in

behaviour has been conducted in several mouse strains including inbred Balb/C, outbred Swiss

Webster, NMR1 and AKR mice. Genetic differences across strains also have an influence on

behaviour which may explain these discrepancies. Similarly, the use of antimicrobials to

83

significantly reduce the microbial number and diversity in healthy adult mice has also been

shown to result in decreased anxiety like behaviour (Verdú et al., 2006).

Interestingly, it has been demonstrated that reconstitution of microbiota in GF mice in early

life using microbiota obtained from specific pathogen free (SPF) mice results in

normalisation/reversal of stress responses (Heijtz et al., 2011; Clarke et al., 2013). By contrast,

conventionalisation of GF mice in adulthood failed to reverse anxiety like behaviour (Neufeld

et al., 2011). These data suggest that there is a critical window during development where

microbiota influence development of the CNS branches related to stress-related behaviours.

Furthermore, colonisation studies have demonstrated that behavioural phenotypes associated

with distinct strains of mice can be transferred to another animal via the gut microbiota. For

example when GF male Swiss Webster mice were colonised with microbiota from SPF Balb/C

mice, an increased anxiety-like phenotype was observed, reflecting the behavioural phenotype

that is common in SPF Balb/C mice (Bercik et al., 2011). In the reverse experiment, GF Balb/C

mice that received microbiota from SPF Swiss Webster mice showed a reduction in anxiety

like behaviour, similar to that observed in SPF Swiss Webster mice.

The studies reported in Table 1.7 show an increase in anxiety-like behaviour is observed as a

result of infection/infestation with pathogens (Heijtz et al., 2011; K. M. Neufeld et al., 2011;

Clarke et al., 2013). Furthermore, oral administration of probiotics in animals with normal gut

microbiota was found to have a positive impact on reducing basal or induced anxiety-like

behaviour (Desbonnet et al., 2008; Bravo et al., 2011; P Bercik et al., 2011) as well as

depressive like behaviours (Bravo et al., 2011; Arseneault-Bréard et al., 2012).

In contrast to the emergence of a rich and robust pre-clinical literature on the effects of

microbiota manipulation on behaviour, limited information is available from human studies. A

randomized, placebo-controlled study of healthy men and woman reported consumption of

84

probiotics containing lactobacillus and Bifidobacterium improved psychological distress and

anxiety compared to those taking a matched control product (Messaoudi et al., 2011). Thus

these findings demonstrate that intestinal bacteria could be used as potential therapeutic targets

in mood and anxiety disorders due to their well-documented effects on modulation of emotional

behaviour.

85

Species (Gender) Method Neurochemical changes Behavioural outcomes Reference

Germ Free versus SPF mice

Female Swiss-

Webster mice

Behaviour tested using

locomotor activity. EPM.

Serum corticosterone

levels and gene expression

levels measured.

In GF mice: Increased levels of corticosterone.

Decreased NMDA subunit NR2B mRNA expression in

central amygdala. Decreased BDNF expression in dentate

gyrus layer of hippocampus

Decreased anxiety-like

behaviour. No change

in locomotion

Male and female

Swiss-Webster

mice

Behaviour tested

using the LD

In GF mice: Increased concentrations of 5-hydoxytryptamine

and 5-hydroxyindoleacetic acid in the hippocampus. Increased

concentrations of serotonin precursor in males. Decreased

BDNF levels in hippocampus

Decreased anxiety-like

behaviour

Male NMR1 mice Behaviour tested

using OF, LD and

EPM

No neurochemical changes reported Decreased anxiety-like

behaviour. Increased

motor activity

Male Swiss

Webster mice

Behaviour tested using OF

and social interaction.

BDNF levels in amygdala

measured

A significant reduction in the mRNA levels of BDNF

were found in the amygdala of GF compared to SPF mice Decreased anxiety-like

behaviour and increased

social interaction

Female SPF F344

rats Behaviour tested using

OF, social interaction.

Serum corticosterone.

(CRF) concentrations

measured

In GF: Displayed Increased CRF mRNA in

hypothalamus. Decreased GR mRNA expression in

hippocampus. Decreased dopaminergic turnover rate in

hippocampus, frontal cortex and striatum. Increased CRF

levels after the open field stress test

Increased anxiety like-

behaviour and reactivity

to stress and reduced

social interaction due to

absence of gut

microbiota in F344

Neufeld et al., 2011;

Clarke et al.,

2013;

Heijtz et al., 2011;

Arseneault-

Bréard et al.,

2012;

Crumeyrolle-

Arias et al., 2014;

86


n Re-colonisation of microbiota in GF mice

Female Swiss

Webster mice LD. No reported neurochemical changes

No change in anxiety

like behaviour

Male NMR1 mice GF mice were conventionalised

in early life using microbiota

obtained from SPF mice.

Behaviour tested using OF, LD,

EPM, locomotor activity and

compared to SPF and GF mice.

No reported neurochemical changes

Conventionalisation of

GF in early life

normalised anxiety and

motor activity

Male Swiss

Webster mice GF mice were colonised at 3

weeks of age and allowed to

grow to adulthood. They were

then assessed for anxiety using

LD


Conventionalisation of

GF mice reversed

anxiety like behaviour

Female Swiss

Webster mice

GF mice were colonised at 10

weeks of age. Assessed for

anxiety like behaviour using

EPM.


Conventionalisation of GF

mice reversed anxiety like

behaviour

Gareau et al.,

2011;

Heijtz et al.,

2011;

Clarke et al.,

2013;

Neufeld et al.,

2011;

87


n Antibiotics

Effect of Infection

Male BALB/c

and Swiss

Webster

Gf Swiss Webster BALB/c mice were

colonized by gavaging fresh cecal

contents from adult SPF BALB/c or

Swiss Webster mice. Behaviour tested

using step-down test.

Colonisation of GF swiss mice with

Balb/C microbiota increased anxiety

like behaviour compared to SPF swiss

mice. In contrast GF BALB/c mice

colonized with swiss microbiota

reduced anxiety like behaviour

Male BALB/c

mice (SPF)

Balb/c mice received mixture of

non-absorbable ATMS (neomycin,

bacitracin and pimaricin). Behaviour

tested using step-down and LD.

BDNF levels measured in

hippocampus and amygdala

BDNF levels were significantly higher in

the hippocapmus and lower in the

amydgala in ATM mice compared to

control mice.

Decreased anxiety like

behaviour and Increased

exploratory behaviour

Male CF-1 mice

CF-1 mice were orally treated with

C.Jejuni or C. rodentium or saline

solution for controls. Behaviour tested

using holeboard. and OF. The protein

c-Fos was measured as an activation

marker in the brain.

c-Fos expression detected in brain regions

associated with anxiety like behaviour

including amygdala.

Increased anxiety like

behaviour caused by

pathogenic bacteria


Bercik et al.,

2011;

Verdú et al.,

2006;

Lyte et al., 2006;

Goehler et al.,

2008;

88


n

Influence of Probiotics

Male Balb/c AKR

mice

AKR mice were injected

with parasite Trichuris

muris. Behaviour tested

using LD and step-down

test. BDNF levels in the

brain were measured

Colonisation of GF swiss mice

with Balb/C microbiota

increased anxiety like behaviour

compared to SPF swiss mice. In

contrast GF BALB/c mice

colonized with swiss microbiota

reduced anxiety like behaviour

Male AKR mice Mice with Chronic DSS

colitis. Treated with

B.Longum NCC3001 on the

3rd cycle of DSS. Behaviour

tested using the Step-down

and LD.


behaviour caused by

DSS-induced gut

inflammation

Male AKR mice Mice with Chronic DSS colitis.

Treated with B.Longum

NCC3001 on the 3rd cycle of

DSS. Behaviour tested using the

Step-down and LD. brain.



behaviour following

colitis, which was

normalised by B.longum

NCC3001 treatment

Lower levels of BDNF mRNA were found in the

hippocampus of T muris infected mice compared with

controls


Bercik et al.,

2010;

Bercik et al.,

2011;

Bercik et al.,

2011b;

89


Male Balb/c

AKR mice

AKR mice were injected with

parasite Trichuris muris, then

treated with probiotic

Bifidobacterium Longum.

Behaviour tested using LD

and step-down test. BDNF

levels in the brain were

measured

Treatment of T muris–infected mice with B longum:

normalised BDNF levels Increased anxiety like

behaviour caused by

parasite infection was

reversed by B longum

treatment

Male Sprague-

Dawley rats

Rats underwent MS, followed

by Administration of

B.infantis. Behaviour was

testing using FST.

Measurement of cytokine

concentrations in whole blood

samples, monoamine levels in

brains and peripheral HPA

axis indicators

MS rats displayed decreased NA content in the brain,

increased peripheral IL-6 release, and increased amygdala

CRF mRNA levels. Treatment with B.infantis resulted in

normalization of the immune response and restoration of basal

NA concentrations in the brainstem

MS mice showed increased

depressive-like behaviour.

Treatment with B.infantis

reversed depressive like

behaviour.

Male BALB/c mice Gavage with

L.rhamnosus (JB-1)

or without bacteria.

Anxiety and

depressive

behaviour tested

using SIH. EPM

and FST

Treatment with L.rhamnosus (JB-1) increased GABA (B1b) mRNA

cingulate and prelimbic cortex. Decreased GABA (B1b) mRNA

expression in hippocampus, amygdala and locus coeruleus

Increased ABA (Aα2) in hippocampus. Decreased GABA (Aα2)

mRNA expression in prefrontal cortex and amygdala.

Decreased anxiety and

depression like

behaviour

Bercik et al.,

2010;

Desbonnet et

al., 2008;

Bravo et al.,

2011;

Table 1.7: Key findings related to the impact of gut microbiota on behaviour. Abbreviations: ATM, antimicrobials; CON: conventionalised; CRF,

corticotrophin-releasing factor; CR, glucocorticoid receptor; 5-HT1A, 5-hydroxytryptamine 1A;EPM, elevated plus maze; DSS, dextran sodium sulphate; FST,

forced swim test; GF, germ free; GR, glucocorticoid receptor; LD, light-dark box; MS, maternal separation; NA, noradrenaline; OF, open field; SIH, stress

induced hyperthermia; SPF, specific pathogen free

90

1.5.5 Gut-Brain Axis and Neurochemistry

As discussed in Section 1.5.4, recent studies have clearly demonstrated that the gut microbiota

influences behaviour. The molecular and neurochemical correlates of these effects remain to

be fully eluded, however the microbiome has been shown to influence neurotransmission,

neuorotropic factors and formation of the BBB.

There is ample evidence to show that a number of neurotransmitter systems are influenced via

the microbiome-gut-brain axis. The neurotransmitter serotonin (5-hydroxytryptamine)

regulates mood, appetite, sleep and cognition (see section 1.4.9). Although it is principally

known for its role in the CNS, approximately 95% of the body’s serotonin is found in the GI

system, where it mediates pain perception, gut secretion and motility (Ridaura et al., 2015).

Tryptophan is both an essential amino acid and a precursor for serotonin. GF mice have been

shown to have elevated levels of plasma tryptophan compared with controls (Clarke et al.,

2013). Increased hippocampal levels of serotonin are also found in male, but not female GF

mice. Importantly, colonization with bacteria post weaning normalises tryptophan availability

but not hippocampal serotonin levels (Clarke et al., 2013). Together these findings indicate

that the microbiota affects serotonergic signalling in the CNS, perhaps by altering peripheral

tryptophan availability. Enteric nervous system (ENS) serotonergic levels are also affected by

microbial metabolites such as SCFAs (Reigstad et al., 2015), since circulating serotonin levels

are largely derived from GI sources, this agrees with studies showing lower serum levels of

serotonin in GF animals (Ridaura et al., 2015).

Gamma-aminobutyric acid (GABA) is the main inhibitory neurotransmitter in the CNS and the

dysfunction of GABAergic signalling has been linked to anxiety and depression (O’Leary et

al., 2014). It has been shown that Lactobacillus spp. and Bidobacterium spp. are capable of

91

metabolising glutamate to produce GABA in culture (Higuchi et al., 1997; Barrett et al., 2012).

Moreover, it has also been shown that provision of the probiotic, L. rhamnosus to mice, causes

reductions in anxiety- and depressive-like phenotypes, as well as inducing region specific

alterations in GABA (1b) mRNA in the brain, with increases being observed in cortical regions

(cingulate and prelimbic) and concomitant reductions in the expression in the hippocampus,

and amygdala, when compared to control fed mice (Bravo et al., 2011). Mice lacking the

GABA (1b) receptor isoform have been shown to be more resilient to both early-life stress and

chronic psychosocial stress in adulthood (O’Leary et al., 2014). These observations suggest

that gut microbes can indeed influence brain neurotransmitter levels.

Moreover, BDNF promotes the growth and development of new neurones and the survival of

existing neurones (Park et al., 2012). Altered levels of BDNF in the CNS are well known to be

associated with mood disorders in humans (Autry et al., 2012). As illustrated in Table 1.7,

microbiota influences BDNF expression in the CNS, for example in GF mice, BDNF

expression is lower in the cortex and amygdala compared with controls, these animals also

displayed increased anxiety-like behaviour (Heijtz et al., 2011). In the hippocampus, the

changes in BDNF levels documented are inconsistent. Studies report an increase in BDNF

levels in GF mice which have displayed reduced anxiety-like behaviour (Neufeld et al., 2011),

while others report a decrease in expression (Sudo et al., 2004; Gareau et al., 2011; Heijtz et

al., 2011; Clarke et al., 2013).Taken together these studies suggest gut microbes influence brain

BDNF levels.

Moreover, recent evidence suggests the gut microbiota may influence the development of the

BBB. The BBB filters the passage of water-soluble molecules from the blood to the brain. It is

formed early in foetal life to protect the foetus against harmful toxins and bacteria during the

sensitive period of brain development (Abbott et al., 2012). Braniste et al., (2014) examined

BBB development and function in GF mice. In foetal mice from SPF dams, the BBB prevents

92

the penetration of labelled immunoflobulin, however the BBB of foetal mice from GF dams

allows the immunoglobulin to diffuse into the brain. As adults, the BBB of GF mice continues

to be more permeable than that of their SPF counterparts (Braniste et al., 2014). It is proposed

that these changes in barrier function are due to tight junction abnormalities. Indeed, tight

junction protein expression is reduced in both foetal and adult GF mice. Moreover, tight

junctions in the brains of GF mice appear diffused and disorganised. Subsequent colonization

of adult GF mice with microbiota from SPF mice successfully reduced BBB permeability and

increased tight junction protein expression (Braniste et al., 2014). The effects of colonization

on BBB permeability is likely mediated by SCFAs or other metabolites produced by gut

bacteria. This study highlights the importance of the microbiota in BBB development;

increased diffusion of molecules into the developing brain could affect brain neurochemistry

and underlie the behavioural profiles of animals with altered gut microbiome status.

Altogether these studies highlight the role of the gut microbiota in modulating bodily

homeostasis and emotional behaviour. An area of growing interests in recent years concerns

the metabolic capability of the gut microbiota and ways by which their metabolic activity

influences mammalian metabolic pathways. Some metabolites released by the gut bacteria have

been implicated in disease states and mood disorders and can be detected using metabolic

profiling techniques such as metabonomics.

1.6 Metabonomics

Metabonomics has emerged as a powerful tool which allows for the study of metabolites and

illuminates the metabolic functionality of the gut microbiome and its interaction with the host.

The suffix ‘omics’ refers to the study of a set of biological molecules and covers fields such as

genomics, transcriptomics, proteomics and metabonomics. Metabonomics involves the

comprehensive characterization of metabolites and metabolism in biological systems in

93

response to drugs, lifestyle, environment, stimuli and genetic modulation with the aim of

characterizing the beneficial and adverse effects of such interactions (Nicholson et al., 1999).

In addition, metabolic profiling techniques have been successfully used in the discovery of

biomarkers of a number of neuropsychiatric and neurodevelopmental disorders such as

depression (Quinones et al., 2009; Huang et al., 2015), schizophrenia (Yang et al., 2013) and

autism spectrum disorder (Wang et al., 2016).

Metabonomic studies are conducted using high resolution analytical techniques such as 1 H

Nuclear Magnetic Resonance (NMR) spectroscopy and Mass Spectroscopy (MS). MS requires

pre-preparation of metabolic samples with the use of Gas Chromotrography (GC), Liquid

Chromatography (LC) or Ultra High-Pressure LC (UPLC). 1 H NMR-based metabolic profiling

allows for both the identification and quantification of very large numbers of metabolites,

giving a good overview of the metabolome, which is a collection of metabolites, of the host.

NMR analysis produces a spectra containing thousands of peaks that are related structurally to

metabolites. Analysis of this data is utilises mathematic modelling techniques combined with

multivariate data analysis. Metabonomics studies generally use biofluids, cells and tissue

extracts which are often easily obtained. Urine, plasma and faecal samples can be obtained

non-invasively and require simple sample preparation (Beckonert et al., 2007). In addition

there are a variety of other biofluids which may be utilized in metabonomic analysis including

cerebrospinal fluid, digestive fluids, amniotic fluid and lung aspirates as well as liver and brain

tissue samples (Lindon et al., 2000).

1.6.1 Neurodisorders

Mental health disorders such as major depressive disorder, bipolar disorder, schizophrenia and

addiction effect millions of people. According to WHO, the worldwide problem of

neuropsychiatric disorders is 13% higher than other major diseases such as cardiovascular

94

diseases and cancer (WHO, 2008). Current understanding of pathophysiology of these

disorders remains limited. This is due to the complex nature of these disorders which tend to

arise from a dynamic and complex dysregulation of gene networks, proteins as well as

metabolic alterations. In addition, present knowledge is also incomplete in predicting the risk

of individuals developing CNS disorders. Disease specific molecular finger prints known as

“biomarkers” can be gained through the study of omics data including metabolomics and can

aid to map dysregulated systems involved in disease pathogenesis. Furthermore, global

mapping of uncharacteristic pathways in psychiatric disorders can lead to the identification of

biomarkers of disease (Quinones et al., 2009; Sethi et al., 2016). Another aspect of

metabolomics as a tool for discovery of biomarkers is its ability to reflect the relationships and

interactions between metabolic state of an individual and environmental aspects (diet, lifestyle,

gut microbial activity, and genetics) under a particular set of conditions (Holmes et al., 2008;

Quinones and Kaddurah-Daouk, 2009)

In addition, the gut microbiota is exclusively responsible for several important metabolic

functions including SCFAs production, amino acid synthesis and bile acid biotransformation

(see Putignani et al., 2016) (see section 1.5.2). The gut microbiota, through metabolite

production has been shown to influence the brain, forming the microbiome-gut-brain axis and

this has been implicated as a critical factor in many physiological and pathophysiological

processes. For example metabonomics has been used to illustrate changes which occur both in

mammalian metabolic pathways, as well as those involved in gut microbiota metabolism in the

pathogenesis of ASD. Metabolomics studies have shown tryptophan metabolism is

significantly decreased in patients suffering from ASD (Boccuto et al., 2013). Furthermore

Ming et al., (2012) have reported defective amino acid metabolism, increased oxidative stress

as well as altered gut microbial metabolism in ASD (Ming et al., 2012). Indeed a high

comorbidity between gastrointestinal dysfunction and ASD has been reported (see Hsiao,

95

2014), thus highlighting the role of dietary intake and gut microbial metabolic activity in the

aetiology of neurodevelopmental disorders.

Taken together these studies indicate that metabonomics is an instrumental tool in identifying

metabolic signatures which reflect environmental factors which may influence the

development of disease state. The use of this technique in this thesis will allow for the

investigation of early life dietry factors which influence normal brain development and could

lead to the development of neuropsychiatric disorders.

96

1.7 Thesis Hypothesis and Aims

The evidence provided in this chapter demonstrates that early life represents a period of critical

development not only for the CNS but also for the gut microbiota. Environmental influences

during this critical time including diet, play an essential role in modulating the normal

development of both of these systems. Milk, either mother's milk or cow’s milk is the main

source of nutrition in early postnatal life in mammals and is known to contain a range of

bioactive compounds, however information regarding the influence of these bioactive products

in milk on brain development during early life is lacking. In recent years there has been an

increasing awareness regarding the impact of early life environmental influences on the

development of a range of neuropsychiatric disorders such as autism, schizophrenia and

depression. This thesis will aim to utilise a multidisciplinary approach to investigate the

influence of prolonged milk exposure in early postnatal life, in particular, exposure to milk

caseins which are known to release bioactive peptides, beta casomorphins, on brain

development and subsequent behaviour. Furthermore, given the growing body of evidence

suggesting diet can influence host metabolism and the development of the gut microbiome

which in turn has been linked with brain function and behaviour, the role of the gut-brain axis

in mediating any neurochemical or behavioural changes will also be investigated.

This thesis will test the hypothesis that prolonged exposure to milk casein and its bioactive

breakdown product BCM-7 beyond the normal age of weaning results in depressive-like

behaviour and induces alterations to neurotransmitter systems in the brain involved in mood

regulation via a gut-brain axis mediated mechanism. This hypothesis is evaluated using male

Wistar Albino rats in a number of experiments. It begins with the analysis of behavioural

development in rats using an animal model of emotional behaviour, the forced swim test,

following casein exposure from the normal age of weaning, PND21 until PND25 and the effect

97

of this on MOPr DOPr and OTr biding using receptor autoradiographic methods. The effect of

prolonged casein exposure on the host’s microbial composition will be assessed using

fluorescence in situ hybridization, and NMR-based metabonomic will be implemented to

investigate the impact on host metabolism. The role of the gut microbiome in mediating any

casein induced metabolic and behavioural changes will then be assessed via the use of

antibiotics to knock-down the gut microbiome before administrating casein for 5 extra days

beyond the standard age of weaning. Finally the hypothesised exposure to BCM-7 in inducing

behavioural alterations will be tested by exposing animals to milk containing the A1/A2 variant

of beta casein which produces BCM-7 against milk containing only the A2 variant, which does

not produce BCM-7.

98

Chapter 2

Effect of Prolonged Casein Exposure in

Early Postnatal Life on Behavioural

Development and the Gut-Brain Axis

99

2.1 Introduction

Several lines of research suggest that manipulations to the age of weaning in a number of

species results in both behavioural and neurochemical changes (see Table 1.2). These changes

are believed to be mediated in part by bioactive components derived from milk proteins. Due

to health benefits associated with breastfeeding (see Table 1.3), exposure to milk rich diets in

early life have generally seen an increase in recent years, either in the form of longer

breastfeeding periods or the use of infant formulas to supplement/extend exposure to milk in

infancy. However prolonged milk consumption has been linked to a number of negative health

effects such as gastrointestinal dysfunction and neuropsychiatric disorders (see Bell et al.,

2006). Considering milk is the main source of nutrition in early postnatal life, and its prolonged

consumption has been linked to negative outcomes, the health risks associated with prolonged

use of milk rich diets in early postnatal life warrants further investigation.

The opioid system which consists of MOPr, DOPr and KOPr plays a crucial role in the

regulation of emotional behaviour with all three receptors playing highly distinct roles in

mediating emotional responses (see Section 1.4.2). The DOPr system is implicated in

positively modulating mood states, as genetic deletion of DOPr in mouse models was shown

to result in both anxiety and depressive-like behaviours (Filliol et al., 2000). Furthermore,

DOPr agonists have been shown to have mood-enhancing properties (Pradhan et al., 2011).

Similar to DOPr but less documented, acute pharmacological activation of MOPr was found to

reduce depressive-like behaviour in rodent models (see Lutz and Kieffer, 2013), while other

studies have shown MOPr KO mice displayed decreased depressive-like behaviours (Filliol et

al., 2000), suggesting MOPr play a paradoxical role in regulating emotional responses.

The opioid system has been shown to have marked differential development in both rodents

and humans (see Section 1.4.3). In rats, the most marked changes occur from the second

100

through to the fourth postnatal week, with 75% of the total DOPr population developing by the

standard age of weaning in laboratory rats (21 days of age) (McDowell et al., 1986). As the

development of the opioid system occurs postnatally, early life environmental factors such as

diet may influence the development of the opioid system. As such, it has previously been shown

that weaning rat pups on PND 21, the standard age of weaning for laboratory rats (Cramer et

al., 1990), compared to those remaining with their mother until PND25, acts as a stimulus for

the activation of a population of DOPrs (Muhammad et al., 1993). Indeed, significantly higher

levels of DOPr binding were found predominantly in somatosensory cortical brain regions of

PND21 weaned rats compared to those weaned on PND25 (Kitchen et al,. 1995). Furthermore,

this weaning induced development of DOPrs can be delayed by continued housing of rat pups

with their dam, up to PND40 (Goody and Kitchen 2001; Muhammad and Kitchen 1993).

Goody and Kitchen (2001) then went on to show that the development of the DOPrs is

dependent on the loss of dietary caseins from milk, which break down to release casomorphins

exhibiting opioid activity (see Section 1.2.2) rather than any maternal separation factors.

Collectively, these studies suggest a potential protective effect of weaning on emotional

behaviour mediated by the opioid system. In line with this, Farshim et al., (2016) have shown

that 25 day-old rat pups that are maintained in a non-weaned state display a depressive

phenotype in comparison to pups weaned on PND21 which further supports this hypothesis.

Given this evidence, this chapter seeks to build on the preliminary findings of Farshim et al.,

(2016) and investigate the role of milk caseins in mediating the depressive-like phenotype

caused by delayed weaning. Furthermore autoradiographic binding techniques will be used to

investigate the effect of prolonged casein exposure on the development of both DOP and MOP

receptors as the opioid system plays a key role in mood modulation.

Moreover, recent findings have indicated that the central oxytocinergic system, which has also

been shown to interact with the opioid system (Gigliucci et al., 2014; Georgiou et al., 2015),

101

as another major player which may affect emotional state of mind. The oxytocinergic system

is widespread throughout the CNS. In rats, OTR are found in several brain regions including

the olfactory system, cortex, thalamus, hippocampus, and the amygdala, a region which is well-

known to play a key role in emotional responses (Weiskrantz 1956). Oxytocin also exerts a

wide range of physiological actions, which include stimulation of sexual and reproductive

behaviour (Carter, 1992), regulation of the HPA axis in response to stress (Viero et al., 2010),

social behaviour (Bosch et al., 2012) as well as the modulation of a range of neuropsychiatric

disorders including depression (Gimpl and Fahrenholz 2001; Neumann 2008; Curley et al.,

2009).

As described in Section 1.4.7, numerous studies have shown that in general increased OT levels

produce anxiolytic, antidepressant and pro-social effects (see Bosch and Neumann 2012).

Furthermore Tribollet et al., (1988) have demonstrated that the distribution and number of OT

binding sites undergo changes during development. In male and female rats two critical time

windows during the development period have been recognised: the third postnatal week, which

precedes weaning and after PND35. Thus like the opioid system, the OT system appears to also

develop differentially which may be susceptible to insult from environmental stimuli during

development such as weaning time. Farshim et al., (2016) demonstrated that weaning

influences stress-induced modulation of OTr density in the amygdala. They showed that at

PND25, delayed weaning in rat pups affected stress induced modulation of OTr in the

amygdala which may reflect a deficient stress coping mechanisms in response to delayed

weaning. It is hypothesised that environmental factors exert their effects on the OT system

though epigenetic modifications of the OTr (Carter, 1992; Bales et al., 2011). Given that

prolonged exposure to milk caseins may be driving the depressive-like behaviour observed in

delayed weaned rats, this research chapter will aim to investigate the effect of exposure to milk

102

casein beyond the normal age of weaning on OTr development using autoradiographic binding

techniques.

Furthermore, a growing body of literature suggests that the gut microbiota plays a crucial role

in human health and disease (see Section 1.5). In healthy individuals, the normal dominant

microbiota is relatively stable and forms a mutually beneficial relationship with its host.

However, it is becoming clear that diet can dramatically alter this delicate synergetic host-

microbiome relationship and result in negative effects to health and development in both

rodents and humans (Bercik, 2011; Cryan et al., 2012; Foster et al., 2013; Burokas et al., 2015).

Recent studies have shown that through interactions within the gut-brain axis, the gut

microbiota is able to influence brain development, function and behaviour (see Bercik 2011).

Probiotic bacteria such as Lactobacillus and Bififobacterium have been found to confer health

benefits such as improving mood states (Desbonnet et al., 2010; Messaoudi et al., 2011) whilst

infection with gastrointestinal pathogens such as Campylobacter jejuni and Escherichia coli,

during the perinatal period have been associated with anxiety like behaviour and cognitive

impairments (Bilbo et al., 2005; Goehler et al., 2008) (see Table 1.7). Such host-microbiome

interactions have been shown to be bidirectional in nature, with mood states such as anxiety or

depression also possessing the ability to affect gut microbial composition (see Section 1.5.3).

The intestinal microbiota is established in early postnatal life in all mammals. This process of

bacterial colonization occurs in parallel with a sensitive time in CNS development (Sudo et al.,

2004; Tau et al., 2010). Given that early postnatal life represents a critical time for both gut

microbiome development and neurodevelopment, it is feasible that the gut microbiota may

somehow affect and contribute to the process of neurodevelopment. Indeed evidence suggests

that the initial inoculation and subsequent development of the intestinal microbiota in early

postnatal life is crucial for healthy neurodevelopment (see Borre et al., 2014). One of the key

103

factors modulating the composition of the developing microbiome is diet, with mode of feeding

during infancy playing a key role in shaping the infant microbiome (Mackie et al., 1999).

Human milk is a source of prebiotics and lactic acid bacteria, therefore breastfed infants show

high levels of Bifidobacterium spp. In contrast, the gut microbiota of formula-fed infants has

an increased diversity and a high abundance of coliforms, Bacteroides and Clostridium

(Praveen et al., 2015).

Moreover, It is also well acknowledged that the process of weaning causes a significant shift

in gut microbial populations in both humans and other mammals (Hentges 1983; Favier, de

Vos, and Akkermans 2003; Palmer et al. 2007; Desclee de Maredsous et al. 2016; Mackie et

al., 1999). Thus weaning represents a critical and potentially hazardous control point in the

appropriate development of the gut microflora, which can influence both the developing CNS

(see Collins and Bercik 2009) and immune system (see Round and Mazmanian 2009).

A number of studies have been carried out to assess the impact of varying weaning ages on gut

microbial profiles. For example Franklin et al., (2002), have shown that pigs weaned at 24 days

of age had significantly higher levels of E.coli populations 3 days post weaning compared to

piglets weaned at 17 days of age. Furthermore, piglets weaned at 24 days of age had higher

lactobacillus populations compared to piglets weaned at 17 days of age. Another study by

Meale et al., (2017) showed that weaning age in calves influences the severity of microbial

shift caused by switching from a milk based diet to solid food. Calves which were weaned at 6

weeks old showed a sever shift in gut microbial populations, whereas calves weaned at 8 weeks

of age showed a more gradual shift in microbial diversity (Meale et al., 2017)

Other studies have gone on to explore the effect of varying post-weaning diets on microbial

colonization. For example Merrifield et al., (2013) compared two different protein sources

(soya and egg) after weaning piglets at 3 weeks of age. It was found that the two different

104

weaning diets caused a shift in urinary metabolic products. Furthermore, even after cessation

of the original weaning diet, the shifts in urinary metabolic profiles were sustained (Merrifield

et al., 2013). Similarly, Rist et al., (2014) investigated the effects of soybean vs. casein protein

sources on intestinal microbiota post-weaning. They showed that piglets fed a casein based diet

post weaning had significantly higher levels of Enterobacteriaceae and some Clostridium

species compared to piglets fed soybean protein (Rist et al., 2014). In addition, it has been

demonstrated in piglets that a decrease in dietary protein intake post weaning results in higher

levels of Lactobacilli which reflect a healthy gut microbial equilibrium (Wellock et al., 2006).

Collectively these studies suggest that manipulations to dietary factors around the age of

weaning can have a significant impact on both the composition and metabolic activity of the

gut microbiota.

Given the prolonged exposure of infants in today’s society to milk rich diets, this research

chapter will investigate the effect of prolonged exposure to milk caseins beyond the normal

age of weaning on the gut microbiota. Bacterial populations will be assessed with the use of

16S rRNA oligonucleotides probes using fluorescence in situ hybridization targeting

Lactobacilus-Enterococcus, Bifidobacterium spp, and the Clostridium histolyticum bacterial

groups.

As discussed in Section 1.5.5, the gut microbiome has been shown to be a metabolically active

organ with a diverse repertoire of metabolic capabilities (See review by Tremaroli and Bäckhed

(2012). Gut microbes are capable of producing a vast range of products, which contribute to

host health and disease (for a detailed review on gut microbial products see Nicholson et al.,

2012). The generation of these microbial products can depend on many factors including diet,

which has been shown to have a direct influence on host metabolic profiles and hence to disease

risk (Walsh et al., 2006; Wikoff et al., 2009).

105

As discussed in Section 1.5.3 and 1.5.4 an area of rapidly growing interest has been the ability

of both host and microbial metabolites to influence brain development and subsequent

behaviour via the gut-brain axis. Currently exact pathways of how alterations in gut microbiota

could affect behaviour remain to be fully elucidated. However evidence has shown that one of

the ways the gut microbiota can directly communicate with the brain is via the production of

neuroactive metabolites (Lyte, 2011; Nicholson et al., 2012; Hsiao et al., 2013).

For example, Hsiao et al., (2013) have provided strong evidence to suggest gut microbes

regulate metabolites that can alter behaviour in animals. In this study Bacteroides fragilis

(B.fraglis) was shown to correct gut permeability caused by maternal immune activation which

then leads to an altered microbial composition and ameliorates defects in anxiety-like, social,

stereotypic and sensorimotor behaviours in the offspring. Furthermore, these authors showed

that the offspring display altered serum metabolic profiles and that B.fraglis was able to

modulate levels of several of these metabolites. Another study showed that rats which were fed

a diet rich in fermentable carbohydrates revealed a strong correlation between D-lactic acid

abundance in the caecum and displaying anxiety like-behaviour and impaired memory

(Hanstock et al., 2010). Moreover, bacterial colonization of germ-free mice with Clostridium

sporogenes was found to result in a ≥ 2-fold increase in serotonin and its metabolites (owing

to bacterial metabolism of tryptophan), which in turn influences the brain and behaviour

(Wikoff et al., 2009). Human studies also suggest a link between microbial fermentation

products and behaviour. For example high faecal concentrations of propionic acid correlate

with anxiety in patients with IBS (Tana et al., 2010). In addition, carbohydrate malabsorption

which results in increased substrate availability for bacterial fermentation has been associated

with depression in females (Ledochowski et al., 2000).

As mentioned in Section 1.6, metabonomics has emerged as a powerful and useful tool which

allows for the characterization of host metabolic profiles in response to external influences

106

such as diet. Metabonomic studies are conducted using high resolution analytical techniques

such as 1 H Nuclear Magnetic Resonance (NMR) spectroscopy, combined with multivariate

data analysis. This present chapter will aim to utilise these metabonomic techniques to study

urinary metabotypes in order to illuminate the metabolic impact of environmental factors,

namely the diet (milk exposure beyond the normal age of weaning) and the gut microbiota on

the host.

Results from this study will provide and important insight into the neuropsychiatric effect of

prolonged exposure to milk caseins during the critical postnatal developmental period as well

as the potential mechanism driving this effect.

Hypothesis:

Rats exposed to casein rich milk beyond the normal age of weaning will exhibit

‘depressive-like’ behaviour due to dysregulation in the development of MOPr, DOPr

and the oxytocinergic system. Furthermore, variation in the community structure of the

gut microbiome, specifically an increase in the abundance of enteropathogenic bacteria,

due to prolonged casein exposure will contribute to this phenotype potentially via gut-

microbial derived metabolites.

Aims:

To investigate the effect of exposing rat pups to milk caseins from PND21 (the standard

age of weaning) till PND25 on the development of ‘depressive-like’ behaviour, MOP,

DOP and OT receptor systems, gut microbial profile and host metabolic phenotype.

107

Objectives:

To use casein free or casein rich milk treated rat models in order to investigate the role

of milk caseins beyond the normal age of weaning on depressive-like behaviour.

Depression will be assessed using the Forced Swim Test.

FISH analysis will be utilized to measure any changes to the gut microbiome induced

by milk diets.

NMR based metabolomics techniques will be used to characterize metabolic changes

associated with prolonged casein exposure in urine.

108

2.2 Methodology

2.2.1 Animals and Experimental Conditions

Male, Wistar albino rats were purchased from Charles River UK Limited, Kent. Three groups

of animals each containing eight male pups plus their cross-fostered mother arrived at the

Experimental Animal Unit (EBU) at the University of Surrey on PND7, where day of birth is

PND0. These animals were maintained in litters of 8 pups with their mothers at all times. Pups

were left to acclimatise to their new environment for 7 days (up until PND14), after which

point they were handled and weighed daily by the experimenter. The standard day of weaning

in laboratory rats is on PND21, a process where the mother is simply removed from the cage

(Cramer et al., 1990). All groups were weaned on PND21. The animals were placed in a 12

hour light-dark cycle (lights on 6AM - 6PM) with free access to a standard rat chow diet

(purchased from B & K Universal Ltd.) and tap water and kept at a constant temperature of

22°C ± 1 °C, humidity of 45-55%.. All experimental procedures were conducted in accordance

with the U.K Animal Scientific Procedures Act (1986).

2.2.2 Casein Free and Casein Rich Milk Diets

Three experimental groups were generated, with each group containing 8 pups caged together.

Group one were weaned by having their mothers removed at 10.00am on PND21. This group

acted as the control and were provided with free access to rat chow and water until the

completion of behavioural testing on PND26. Group two were weaned on PND21 and provided

with casein free milk formula as well as having free access to standard rat chow and water from

PND21 through till PND26. Group three were weaned on PND21 and provided with casein

rich milk formula as well as free access to standard rat chow and water from PND21 through

till PND26.

109

The groups provided with either casein free or casein rich milk formulas, had milk freshly made

up twice a day, once between 9:00-10:00 AM and then again between 5-6 PM. Casein rich milk

formula, (Welpi: Petlife International Ltd, Bury St. Edmunds UK) was made up as 1 part

formula to 2 parts boiling water. When cooled to approx. 40°C, the milk was poured into bottles

with ball bearing nozzles and placed in the appropriate cages to be consumed ad libitum. The

casein free milk formula (Special Diet Services, Cambridge UK) was custom manufactured to

provide the closest dietary match of the casein rich milk formula, however casein protein was

removed and replaced with overall 20% w/w soya protein (approximating to the level of casein

contained in the casein rich milk formula) (see Table 2.1). The casein free milk formula was

also made up as 1 part formula to 2 parts hot water then cooled before provision to pups.

Casein Rich Milk formula

Parameter analysed Units Results Fresh Weight

Moisture % 3.37

Crude Fat % 18.9

Crude Protein % 26.5

Crude Fibre % 0.9

Ash raw % 6.2

NFE (Nitrogen Free extract) % 44.13

Ingredients

Spray dried skimmed Milk, vegetable oils

and fats mix, spray dried milk albumin,

spray dried whey products

Casein Free Milk Formula

Parameter analysed Units Results Fresh Weight

Moisture % 5.64

Crude Fat % 18.9


Crude Fibre % 1.3

Ash raw % 8.78

NFE % 38.01

Ingredients

Whey powder, soya protein concentrate,

maze oil, soya oil, vitamin and mineral

mix

Table 2.1: Formulations of casein rich and casein free milk diets.

110

2.2.3 Body Weights and Milk Consumption

Animals belonging to all three treatment groups were weighed daily from PND14 until the day

of sacrifice on PND27. Milk consumption was also recorded on a daily basis from PND21 until

PND26.

2.2.4 Behavioural tests

2.2.4.1 Forced Swim Test

Behavioural testing was performed between 10am and 2pm when animals reached PND25. The

FST was used to assess depressive-like behaviours and was carried out as originally described

by Porsolt (Porsolt et al., 1977). On the day of behavioural testing the rats were allowed to

acclimatise to the testing laboratory for one hour prior to the start of experiments. The

experimental laboratory was window-less, air controlled and was kept under the same

conditions as the home-cage room. The testing room was maintained at a minimal noise level.

The rats tested on PND25 were placed in a 5-litre glass beaker (Fisher Scientific, UK) with the

dimensions of 170 mm diameter and 270 mm height. The beakers were filled to a depth of 165

mm with tap water (23°C ± 1°C). This depth was chosen as at this level the rats were unable to

escape with their hind limbs or tails unable to reach the bottom of the beakers. Four beakers

were set up for each trail and the beakers were separated using white sheets of paper to prevent

visual contact between the test subjects (see Figure 2.1). A black cloth was placed behind the

four beakers to assure a high contrast between the dark background and white rats. The

apparatus was kept clean by changing the water following each set of tests and cleaned with

111

odour-free detergent then dried. This process was carried out in a separate room to minimise

noise disturbance to other animals.

The forced swim test is run over two consecutive days. One the first day (PND 25 the rats were

exposed to a ‘pre-test’ swimming session lasting 15 minutes followed 24 hours later by a ‘test’

session lasting 6 minutes. During the first day, the swim session consisted of 15 minutes, when

the rats would have experienced “helplessness or behavioural despair” in the beakers. On the

second day, consisting of 6 minutes, following an initial period of struggling, the rats exhibited

a floating movement pattern (known as ‘immobility’). The immobility time is believed to

demonstrate “learned helplessness” or depressive-like behaviour. Antidepressants reduce the

duration of immobility (or floating time) and the latency to the first period of immobility are

both key predictors of antidepressant activity.

Both the test and pre-test sessions were conducted under low illumination levels adjusted from

normal lighting (~350 lux) to a dimmer setting of ~40 lux. This level of lighting was selected

as light has been shown to be an important factor influencing behaviour in the FST. It has been

shown that standard illumination caused hyperlocomotion in the FST (Bogdanova et al., 2013),

therefor dim lighting is used as rats prefer a dimmer environment, thereby eliminating a factor

which could induce extra stress and influence behaviour. Following either pre-test or test

sessions, rats were dried with towels and placed in a warm (30°C ± 2°C), fanned recovery

chamber for 15 minutes or until adequately dried before being returned to their home cages. At

the end of all 6-minute swim sessions, the rats were individually housed in metabolic cages to

allow for the collection of faecal and urine samples for metabolic analysis (see Section 2.2.6).

Both pre-test and test swim sessions were recorded using a Sony Handycam HDR-CX250. The

real time video recordings of each session were used for behavioural analysis. The total

immobility time, latency to the first immobility, total time spent swimming and total time spent

112

climbing were quantified from the recordings. The minimal duration of an episode of

immobility was set at 2 seconds. A rat was judged immobile when it stopped all active attempts

to escape from the beaker such as struggling, swimming, climbing, diving and remained a float,

making only the movements necessary to keep its head above the water. Swimming was

recorded when smooth movements were made around the beaker with all four paws immersed

in the water. Climbing was scored when strong up-ward movements with the four paws were

made whereby the water surface was vigorously broken. These behaviours from the ‘test’ day

on PND26 were scored by one independent observer who was blind to the identity of animals.

The FST was chosen as the most appropriate model to assess for depressive like behaviour in

this study for a number of reasons. Firstly, it is the most commonly used test for depression as

it involves the exposure of animals to stress, which was shown to have a role in the tendency

for major depression (Anderson, 2006). It has also been shown to share some of the factors that

are altered by depression in humans, including changes in food consumption and sleep

abnormalities (Yankelevitch-Yahav et al., 2015), making it a reliable model to use. Moreover,

it is easy to perform and the results are easily analysed making it a favourable model to use for

the current study. Secondly, the FST provides a ‘read out’ of depressive-like behaviour, i.e. it

does not induce depression unlike some of the other models available such as social defeat, but

instead provides a way of assessing already existing depressive-like behaviour which is what

is required in this current study. Finally, the FST is run over two days which is a relatively

short time window in comparison to other depression models. As the current study is looking

at the effect of casein exposure over only 5 days, the behavioural test selected needed to be one

which could be performed in the space of this limited time frame. The sucrose preference test

is another depression model which is reliable, relatively simple and quick to perform, however

this model was not selected for this study as it involves the consumption of sucrose which

would have been a confounding nutritional factor.

113

2.2.5 Data Analysis

Body weight differences between control, casein free and casein rich treated animals were

analysed using a one-way ANOVA .The behavioural data for the FST were also analysed

using a one-way ANOVA. Where ANOVA revealed a significant difference an LSD post-hoc

analysis was carried out. All data were expressed as means ± SEM, n = 8 per group. A p

value of less than 0.05 was accepted as significantly different. All statistical analyses were

performed on GraphPad Prism 7.

Postnatal Day

7 14 21 25 26 27

Appropriate groups

provided with no milk, casein free or casein

rich milk

Mothers removed from

all groupsForced

Swim Test

Animals go into metabolic

cages for 24 hours

Animals sacrificed, samples collected

Animals weighed daily

Animals arrive in EBU

Figure 2.1: Experimental timeline showing key points in the casein free vs. casein rich study.

114

2.2.6 Metabolomics

Following the completion of behavioural testing on PND25 and PND26, animals were placed

into individual metabolic cages for 24 hours. On PND27 urine was collected from the cages

using plastic pipettes into 2ml Eppendorf tubes and stored in -80°C for metabolic analysis.

2.2.6.1 Sample Preparation

2.2.6.1.1 Urine 1H NMR Protocol

The urine samples were defrosted from -80°C and vortex mixed. The samples were prepared

by mixing 400 µl of urine with 200 µl of phosphate buffer (pH 7.4) containing 10% D2O and

1 mM 3-trimethylsilyl-1 [2,2,3,3-2H4] propionate (TSP). Eppendorfs were centrifuged at

13,000 rpm at 4°C for 10 minutes and 550 µl of the resulting supernatant was transferred to a

5 mm internal diameter Norell® NMR tubes and stored at 4°C until 1H NMR spectroscopy.

2.2.6.2 Data Acquisition

For each sample the 1H NMR spectra were acquired as described by Beckonert et al., (2007)

using a Bruker Avance 700 MHz spectrometer (Bruker Biospin, Rheinstetten, Germany) with

a 14 T magnet, equipped with a 5 mm cryogenically cooled inverse probe. Standard one-

dimensional (1D) NMR spectroscopy has been used throughout this experiment and consisted

mainly of two steps, pulse and acquire. Briefly, one radiofrequency pulse generates a response,

the free induction decay (FID), which can then be acquired. This sequence can be repeated

numerous times to allow improvement of signal to noise ratio, thus reducing background noise.

Between consecutive pulses a delay time was used to allow complete relaxation to occur. For

each urine sample a standard 1D-Nuclear Overhauser Effect Spectroscopy (NOESY)-presat

pulse sequence (recycle delay (RD)-90°-t1-90°-Tm-90°-acquire-FID) was used with water

suppression. Spectra were acquired after 8 dummy scans followed by 128 scans and collected

115

into 64K data points using a spectral width of 12 ppm (parts per million). This involved an

acquisition time per scan of 3.79 seconds and a mixing time (tm) of 100 milliseconds.

2.2.6.3 Data Processing

The spectra were processed in TOPSPIN 3.0 (Bruker Biospin, Rheinstetten, Germany).

Following Fourier transformation, the spectra were manually phased, baseline-corrected and

calibrated to the TSP resonance at 0 ppm (δ 0.00). Spectra were then digitised using MATLAB

(Version R2012a, MathWorks Inc; Natwick, MA) scripts provided by Korrigan Sciences. In

order to carry out pattern recognition, the data were digitised into a data matrix, consisting of

n observations (samples) and k variables (chemical shifts or metabolites). Peak alignment was

applied to adjust peak shifts due to pH differences.

The spectral regions that possessed no real diagnostic information or disrupted other spectral

data were removed. The region 4.5-6.0 ppm was removed as this included the resonances of

water and urea. Once the unrequired spectral regions were removed, each observation was

normalised. Urine samples were normalised using the probalistic quotient method (Dieterle et

al., 2006). Normalisation is a mathematical correction method that is applied to the

experimental data to adjust variation (bias) that could potentially influence the data and hence

allows the data to more realistically reflect biological differences occurring within the data.

2.2.6.4 Multivariate data analysis

All multivariate analyses were performed using the MATLAB scripts provided by Korrigan

Sciences and using SIMCA-P+ 11 (Umetrics AB, Umeå Sweden) software. All figures were

generated using Adobe Illustrator CS5 (Adobe systems, Ireland). Metabolic data obtained from

1H NMR spectroscopy typically contain thousands of variables from each sample that are

highly correlated. The multi-dimensional nature of the data generated is difficult to visualise

and understand and thus requires statistical techniques that can be used to extract the necessary

116

information. Here, chemometrics methods were used due to their ability to decompose complex

multivariate data into simpler and potentially interpretable structures (Wold et al., 1984).

Depending on the aim of the study under analysis supervised or unsupervised methods can be

applied and can assist in obtaining an overview of the data, in variable selection and in-group

classification. The chemometric methods applied in this Chapter are described below.

2.2.6.5 Principal Component Analysis

The unsupervised multivariate technique used here was Principal Component Analysis (PCA),

as described by (Wold et al., 1984). PCA can be described as a method that reveals the internal

structure of a data set in a way that best explains the variance in the data. PCA offers a reduced

dimensional model that summarises the major variation in the data into a few axes, and in this

way, systemic variation is captured in a model that can be used to visualise which samples in

the data are similar or different from each other.

From this, it will be possible to determine if the treatment related variation has been detected

by the given PCA model with possible spectral loadings that then depict any `treatment' related

separation along the respective principal component(s).

In the present study, PCA was used as an initial explorative method to investigate to which

extent different treatments (casein free vs. casein rich milk) could be discriminated by the

urinary metabolite profiles. The most variance in the data is explained by the first principal

component (PC) and each PC thereafter explains less of the total variance existing in the data

than the previous component(s).

The PCA model provides a scores matrix that summarises the original variables and discovers

the direction of the data swarm that illustrates the maximum degree of variation (PC1) and

exposes information regarding relationships between the objects (e.g. groupings). Principal

component loadings plots provide information on which variable(s) contribute to the patterns

117

observed in the scores plots for a given component. When the PCA is applied to digitised NMR

spectra, the loadings correspond to spectral frequencies with each shift relating to specific

metabolites. From this point on changes in metabolites could be further observed and their

interaction with other metabolites assessed and this forms the basis of metabonomics.

The total amount of variance explained by the model is given by R2 and is a quantitative

measure of the goodness of fit that shows the degree of variation within the data set that can be

explained by the model. The R2 values have been expressed on each PCA scores plot as the

sum of the percentages of variance explained by each PC.

2.2.6.6 Orthogonal Projection to Latent Structures-Disciminant Analysis

The supervised method used in this study was Orthogonal Projection to Latent Structures-

Discriminant Analysis (OPLS-DA) as described by Trygg & Wold (2002). This method is an

extension of the widely used supervised method, Partial Least Squares-DA (PLS-DA) (Haque

et al., 2009). PLS regression is a robust and powerful algorithm that is used to relate a data

matrix containing independent variables from samples, such as spectral values (X matrix), to a

matrix containing dependent variables (e.g. treatments or responses) for those samples (Y

matrix) and thus find the relationship between the two matrices. Whilst PCA models the major

variation in the data itself, PLS derives a model that describes the correlation between X

variables and then a feature (Y variable) of interest. For this analysis a `class' vector is

constructed, with one variable of each class having a value of 1, if the sample belongs to a

particular class and 0 if not. By regression against this class vector, latent variables can be

derived which enable separation of classes from one another. In order to avoid over fitting and

ensure model reliability, cross validation methods were used.

In the present study, a PCA model was built using MATLAB. Where clear variability (i.e.

separation) was found between treatment groups, an OPLSDA model was constructed to

118

further define metabolic variation between the groups. The contribution of each variable

(metabolite) to the model was visualised by generating correlation coefficient plots. These are

colour-coded according to the significance of correlation (r2) to the \class" (e.g. casein free or

casein rich). Red indicates a highly significant metabolite whilst blue indicates low

significance. The direction of the peaks relate to the covariance of the variable multiplied by

its standard deviation (back-scaling). This generates a plot that closely resembles a 1H NMR

spectrum to aid model interpretation. The predictive performance (Q2Y) of each model was

also obtained using a 7-fold cross validation approach. This indicates how well data in the X

matrix (metabolites) can predict information in the Y matrix (e.g. whether the animal received

casein-rich or casein-free milk). The validity of these models (based on their Q2Y values) was

confirmed using permutation testing (10,000 permutations).

2.2.6.7 Metabolite Identification

There are two methods available for the identification of metabolites, targeted or explorative.

Which method is used depends largely on the scope of the study. The explorative metabolite

identification method was used in this study and resulted in the identification of metabolites

that related to the effect of diet. Metabolites were assigned according to their peak shifts and

multiplicity using Chenomx NMR Suite software version 7.5 (Chenomx Inc. Edmonton,

Canada) and via existing databases such as the Biological Magnetic Resonance Data Bank

(BMRB: http://www.bmrb.wisc.edu), and the Human Metabolome Database (HMDB)

Metabolomics Toolbox (http://www.hmdb.ca) as well as comparison with chemical shifts

provided in existing data in the literature.

http://www.bmrb.wisc.edu/

http://www.hmdb.ca/

119

2.2.6.8 Biological Interpretation

After metabolites have been identified, a relevant biological interpretation should be drawn

related to the study. The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway

database (http://www.genome.jp/kegg/pathway) is an example of a freely available online

database that provides information about numerous pathways and the metabolites that interact

in corresponding pathways. Furthermore, the HMDB provides metabolite information in the

‘MetaboCard’ and contains clinical, biochemical and chemical data. In addition to the use of

such online databases, the literature was searched to put the identified metabolites into

biological context.

2.2.7 Tissue Dissection

Following on from the metabolic caged on PND27, animals were then killed by cervical

dislocation on PND27 after acute exposure to CO2. Approximately 2ml of trunk blood was

collected into plastic lithium-heparin coated anticoagulant tubes. The samples were then

shaken by inversing and placed on ice during the collection procedure. All samples were

centrifuged using a Beckman Coulter Allegra X-15R centrifuge at 1600 g for 30 minutes for

collection of plasma (Beckonert et al. 2007). The supernatant was then transferred into 2 ml

Eppendorf tubes, labelled and secured using Parafilm and stored at -80°C until use in future

metabonomics studies.

Intact brains were dissected and rapidly frozen in isopentane at -20°C for 30 seconds and then

stored at -80°C for quantitative receptor autoradiographic experiments (see Section 2.2.8). The

carcasses were transferred into a fume hood and six regions of the intestine (stomach,

duodenum, jejunum,ileum, ceacum and colon) were dissected. Contents from these regions

were weighed (30-200 mg) into sterile 2 ml Eppendorf tubes and kept over ice and then stored

in -80°C for use with fluorescence in situ hybridization experiments (see Section 2.2.9).

http://www.genome.jp/kegg/pathway

120

2.2.8 Quantitative Receptor Autoradiography

2.2.8.1 Brain collection

Rat pups were killed by decapitation, intact brains removed as described in Section 2.2.7, and

brain sections prepared as detailed below.

2.2.8.2 Tissue Sectioning

2.2.8.2.1 Cleaning and Gelatin-Coating of Microscope Slides

Microscope slides (74 mm x 26 mm) (Marieneld, Germany) were loaded onto metal slide

holder racks in a washing tray and soaked in Decon detergent and warm water solution

overnight. The following day, all slides were rinsed under hot running water for 15 minutes

followed by a further rinse with cold distilled water. A 90:10, ethanol:hydrochloric solution

(v/v), was used to soak slides for 15-20 minutes before rinsing in distilled water for 15 minutes.

Slides were then coated by immersing into a 1% w/v gelatin/chrom-alum solution for 2 minutes

and left to dry in a dust-free environment before use.

2.2.8.2.2 Sectioning of Brain Tissue

Brains were removed from a -80°C freezer and placed into a cryostat apparatus (Zeiss Microm

505E, Hertfordshire, UK) set at -21°C and coronally aligned by fixing the cerebellum onto a

mounting stage using plastic embedding solution (O.C.T compound, BDH Chemicals, Dorset,

UK). Adjacent 20 μm thickness coronal sections were cut at 400 μm intervals and thaw-

mounted onto gelatine-coated ice-cold microscope slides to define receptor binding levels from

fore- to hind-brain regions. Brain slides were stored at -20°C in airtight containers containing

a layer of anhydrous calcium sulphate (Dreirite-BDH Chemicals, Dorset, U.K.) for a minimum

of 1 week before use.

121

2.2.8.3 Autoradiographic binding of Oxytocin Receptors

Quantitative OTr autoradiography was performed on brain sections from control, casein free

and casein rich treated animals tested for depression using the FST on PND 25 and PND26.

Sections were rinsed twice for 10 minutes in a pre-incubation buffer solution (50 mM Tris-HCl

pH 7.4) at room temperature to remove endogenous OT. Total binding was determined by

incubating the sections with 50 pM [125I]-Ornithine vasotocin analog

[d(CH2)5[Tyr(Me)2,Thr4,Orn8,[125I]Tyr9-NH2]-vasotocin] ([125I]-OVTA) (Perkin Elmer,

Boston, MA) in an incubation buffer medium (50 mM Tris-HCl, 10 mM MgCl2, 1 mM EDTA,

0.1% w/v bovine serum albumin, 0.05% w/v bacitracin; Sigma-Aldrich, Poole, UK, pH 7.4 at

room temperature). Adjacent sections were incubated with [125I]- OVTA (50 pM) for 60

minutes in the presence of 50 μm of OT ligand, (Thr4, Gly7)-oxytocin (Bachem, Germany), to

determine non-specific binding (NSB). Following the radioligand binding period, slides were

rinsed 3 times for 5 minutes in ice-cold rinse buffer solution (50 mM Tris-HCl, 10 mM MgCl2,

pH 7.4) followed by a 30 minute wash in the ice-cold rinse buffer, and a subsequent 2 second

wash in ice-cold distilled water. Slides were then dried under a stream of cool air for 2 hours

and stored in sealed containers with anhydrous calcium sulphate (Drierite) for 2 days.

2.2.8.3.1 Film Apposition

Slides were mounted onto card and secured into place using tape. Sections were apposed to

Kodak MR-1 films (Sigma-Aldrich, U.K.) in Hypercassettes with autoradiographic [14C]

microscales of known radioactive concentration (GE Healthcare Life Sciences, Amersham,

U.K.) for 3 days. Sections for all treatment groups were processed in parallel and apposed to

the same film at the same time to avoid inter-experimental variation.

122

2.2.8.3.2 Film development

Film development was conducted in a dark room using red-filter light. The films were

developed by immersing them individually one at a time into a tray containing 50% Kodak

D19 developer (Sigma-Aldrich, Poole, UK) for 3 minutes. The films were then immersed in a

second tray containing distilled water and three drops of glacial acetic acid solution for 30

seconds to stop the development reaction. This was followed by a 5-minute fixation step by

immersing the films into a third tray containing Kodak rapid fix solution (Sigma-Aldrich,

Poole, UK). Finally, the films were rinsed under cold running water for 20 minutes and left to

dry on hanging clips in a fume hood.

2.2.8.3.3 Image Analysis of Autoradiograms

Quantitative analysis of autoradiographic films were carried out using video-based,

computerised densitometry using an MCID image analyser (Imaging Research, Canada) as

previously described by Kitchen and co-workers (Kitchen et al. 1997). Optical density values

were quantified from the [14C]-microscale standards of known radioactive concentration and

cross- calibrated with [125I] (GE Healthcare, UK) and were entered into a calibration table on

MCID. The relationship between radioactivity and optical density was determined using the

MCID system. Specific binding was calculated by subtraction of NSB from total binding and

expressed as fmol/mg tissue equivalents. The analyses of brain structures were carried out by

reference to the rat brain atlas of Paxinos and Watson (Paxinos & Watson, 2006). Measures

were taken from both the right and left hemispheres for most structures, thus representing a

duplicate determination for each region. Central brain regions such as the hypothalamus and

thalamus were singularly taken using the free-hand drawing tool to outline the structures. The

following structures were analysed by sampling 5-8 times with a box-sampling tool (box size

in pixels shown in parantheses): cortex (8 × 8 mm), olfactory tubercle (6 × 6 mm) and

123

hippocampus regions (5 × 5 mm). For each film analysed, regions were always quantified at

the same level of the brain by reference to Bregma co-ordinates given in the Paxinos and

Watson rat brain atlas and as shown in Table 2.2.

2.2.8.4 Autoradiographic binding of Delta Opioid Receptors

Quantitative DOPr autoradiography was performed on brain sections from control, casein free

and casein rich treated animals tested for depression using the FST on PND25 and PND26.


pH 7.4) at room temperature to remove endogenous opioids. Total binding was determined by

incubating the sections with [3H] deltorphin -1 (7nM) in an incubation buffer medium (50 mM

Tris-HCl, 10 mM MgCl2, 1 mM EDTA, 0.1% w/v bovine serum albumin, 0.05% w/v

bacitracin; Sigma-Aldrich, Poole, UK, pH 7.4 at room temperature) for 60 minutes. Adjacent

sections were incubated with 1µM naloxone for 60 minutes to determine non-specific binding

(NSB). Following the radioligand binding period, slides were rinsed 3 times for 5 minutes in

ice-cold rinse buffer solution (50 mM Tris-HCl, 10 mM MgCl2, pH 7.4) followed by a 30

minute wash in the ice-cold rinse buffer, and a subsequent 2 second wash in ice-cold distilled

water. Slides were then dried under a stream of cool air for 2 hours and stored in sealed

containers with anhydrous calcium sulphate (Drierite) for 2 days.


Slides were mounted onto card and secured into place using tape. Sections were apposed to

Kodak MR-1 films (Sigma-Aldrich, U.K.) in Hypercassettes with autoradiographic [3H]

microscales of known radioactive concentration (GE Healthcare Life Sciences, Amersham,

U.K.) for 10 weeks. Sections for all treatment groups were processed in parallel and apposed

to the same film at the same time to avoid inter-experimental variation.

124


As described in Section 2.2.8.3.2


As described in Section 2.2.8.3.3 but using [3H] microscale instead of [14C]. Regions analysed

for the Delta opioid binding are shown in Table 2.2.

2.2.8.5 Autoradiographic binding of Mu Opioid Receptors

Quantitative MOPr autoradiography was performed on brain sections from control, casein free

and casein rich treated animals tested for depression using the FST on PND25 and PND26.


pH 7.4) at room temperature to remove endogenous opioids. Total binding was determined by

incubating the sections with [3H] DAMGO (4nM) in an incubation buffer medium (50 mM

Tris-HCl, 10 mM MgCl2, 1 mM EDTA, 0.1% w/v bovine serum albumin, 0.05% w/v

bacitracin; Sigma-Aldrich, Poole, UK, pH 7.4 at room temperature) for 60 minutes. Adjacent

sections were incubated with 1µM naloxone for 60 minutes to determine non-specific binding

(NSB). Following the radioligand binding period, slides were rinsed 3 times for 5 minutes in

ice-cold rinse buffer solution (50 mM Tris-HCl, 10 mM MgCl2, pH 7.4) followed by a 30

minute wash in the ice-cold rinse buffer, and a subsequent 2 second wash in ice-cold distilled

water. Slides were then dried under a stream of cool air for 2 hours and stored in sealed

containers with anhydrous calcium sulphate (Drierite) for 2 days.



125




As described in Section 2.2.8.3.3 but using [3H] microscale instead of [14C]. Regions analysed

for the Delta opioid binding are shown in Table 2.2.

2.2.8.6 Data Analysis for Quantitative Receptor Autoradiography

The mean (and standard error of the mean, SEM), n=8 per group specific radioligand binding

was determined for all brain structures analysed from each treatment group for OTR DOPr and

MOPr. Comparisons of OTR, DOPr and MOPr binding levels were made using a one-way

ANOVA on individual brain regions followed by LSD post-hoc analysis where applicable. A

p value of less than 0.05 was accepted as statistically significant. All statistical analyses were

performed using GraphPad Prism 7. All the values were expressed as mean ± SEM.

126

[125I]-OVTA receptor

Brain Region Bregma Coordinate (mm)

[3H] Deltorphin -1

Olfactory bulb - external and lateral, dorsal, medial, ventral Cingulate cortex

Nucleus accumbens shell and core Olfactory tubercle

Piriform cortex Medial septum

Vertical limb of the diagonal band of Broca Lateral septum Hippocampus

Thalamus Hypothalamus

Amygdaloid nuclei {central, basolateral and basomedial}

4.20 0.7 0.7 0.7 0.7 0.7 0.7 0.7

0.20 -4.80 -4.80 -4.80

Olfactory bulb - external and lateral, dorsal, medial, ventral Caudate putamen Cingulate cortex


Motor Cortex (Deep, superficial layers) Somatosensory cortex (Deep, superficial layers)

Piriform cortex Motor Cortex (Deep, superficial layers)

Somatosensory cortex (Deep, superficial layers) Hippocampus

Thalamus Hypothalamus

Amygdaloid nuclei { central, basolateral and basomedial}

4.20 1.20 1.20 1.20 1.20 1.20 1.20 1.20 2.30 2.30 2.30 2.30 2.30 2.30 2.30

127

[3H] DAMGO

4.20 1.20 1.20 1.20 1.20 1.20 1.20 2.30 2.30 2.30 2.30 2.30 2.30

Olfactory bulb - external and lateral, dorsal, medial, ventral Caudate putamen Cingulate cortex




Hippocampus Thalamus

Hypothalamus Amygdaloid nuclei {central, basolateral and basomedial}

Table 2.2: Bregma coordinates used for brain regions analysed by [125I] OVTA receptor, [3H] Deltorphin-1 and [3H] DAMGO respectively.

128

2.2.9 Fluorescence in situ Hybridization (FISH)

2.2.9.1 Intestinal Dissection and Contents Collection

Intestinal contents and tissue specimens from five various intestinal regions were collected

from casein free and casein rich treated animals behaviourally tested on PND25 & PND26 and

sacrificed on PND27. The five gut regions included: stomach, caecum, jejunum, Ileum and

colon. The University of Surrey Guide to the Autopsy Procedure for Rats was utilised to obtain

each sample. Prior to the sample collection process, sterile 2ml Eppendorf tubes were labelled

and placed in the fume hood over ice. All surgical equipment were sprayed with ethanol prior

to the autopsy procedure. An incision was made through the abdominal skin and muscle wall

of test subjects and the intestines were unravelled onto a dissection board. To start, the stomach

was cut around its greater curvature and the contents were collected using sterile forceps and

weighed into the corresponding Eppendorf tubes. Following the stomach, samples of intestine

of around 2-3 cm in length were taken and contents transferred to tubes from predetermined

sites, in the following way: duodenum: 1-2 cm down from the stomach; jejunum: 10 cm down

from stomach; ileum: 10 cm up from caecum and, colon: 3-4 cm from caecum. The caecum

was punctured open and the contents weighed and transferred into the corresponding tubes.

Approximately 20- 2000 mg of intestinal content from each region was collected. The

remaining carcasses and tissue remnants were placed in polythene bags for containment and

disposal by incineration. The cells were then vortexed using paraformaldehyde as described

below.

2.2.9.2 Sample Fixing

In order to avoid high amounts of fibre within the samples, around 0.15 g of intestinal contents

was suspended in 600 μl of phosphate buffer saline (PBS) solution; an extra 600 μl was added

129

to samples from the colon and caecum due to the higher fibre content in these regions. Samples

were then mixed thoroughly by vortexing for 1 minute, then aspirating with a 1 ml pipette tip.

The samples were then centrifuged at 1000 rpm for 2 minutes and the supernatant were

collected into separate 2 ml Eppendorf tube and the pellet was re-suspended in 400 μl of PBS

solution and centrifuged again for 2 minutes at 1000 rpm. This procedure was repeated once

more and each supernatant was added to the 2 ml tube with the pellet from the last spin being

discarded. The resulting supernatant was then centrifuged for 5 minutes at 13000 g. Following

this, the supernatant was discarded and the pellet was re-suspended in 750 μl of PBS solution.

Half (375 µl) of the produced solution was then added to 1125µl of 4% paraformaldehyde

(PFA, Sigma UK) solution, vortex mixed and promptly stored in the 4°C fridge for 4-8 hours.

The 4% PFA solution was prepared by transferring 2 g of PFA powder into a 50 ml Duran

bottle and adding 30 ml of sterile distilled water and 100 μl of 1 M NaOH solution, and

dissolved in a 50°C waterbath. Once the PFA powder dissolved, 100 μl of 1 M HCl solution

and 16.6 ml of 3 × PBS solution was added and made up to 50 ml with sterile distilled water.

The PFA solution was then filter-sterilised through a Millex® 0.22 μm Millipore syringe filter

into a sterile container.

The samples were removed from the refrigerator and centrifuged at 13000 rpm for 5 minutes.

The supernatant was then carefully discarded, and the pellet re-suspended in 1mL of 1 × PBS.

Next the samples were centrifuged twice at 1300rpm for 5 minutes. All the resulting

supernatants from the tubes were carefully removed using a 1ml pipette tip to ensure as much

supernatant is removed as possible from the tubes. The final pellet was then re-suspended in a

solution that is determined by the original weight of that sample. The equation was as follows

[(Original weight of sample / 150mg) x (150 µl)] of PBS solution and [(Original weight of

sample / 150mg) x (150 µl)] of Ethanol. Fixed samples were then stored in a -20°C freezer.

130

2.2.9.3 Reagents

The methods described for FISH in this chapter originated from (Daims et al., 2005). The

following reagents were prepared prior to the commencement of experimental procedures:

Reagent Concentration Storage Temperature

Tris-HCl (pH 8.0) 1M RT

NaCl 5M RT

Ethanol 50,80,96% (v/v) RT

EDTA 0.5M RT

SDS 10% RT

PBS/SDS (pH 7.2) 10% SDS RT

Lysozyme 1mg/ml -20°C

Table 2.3: List of reagents required for FISH. RT: Room Temperature

Table 2.4 lists the hybridization conditions for the oligonucleotide probes used in this

experiment. Details for the bacteria targeted by individual probes used can be found in the

relevant publications listed and further sequence databases can be found via ProbeBase (http:

//www.microbial-ecology.net/probebase/).

131

Probe

Name

Microorganism

Specificity

Sequence ( 5’ to 3’) Hybridization

pre-treatment

Hybridization

Temp (°C)

Washing

Temp (°C)

Reference

Bif164 Bifidobacterium spp. CATCCGGCATTACCACCC Lysozyme 50 50 (Langendijk et al.,

1995)

Chis 150 Clostridium histolyticum

group ( clusters I and II)

TTATGCGGTATTAATCTYC

CTTT

None

50 50 (Franks et al., 1998)

Lab158 Lactobacillus-

Enterococcus

GGTATTAGCAYCTGTTTCC

A

Lysozyme 50 50 (Harmsen et al., 1999)

EUB338 I* Most bacteria GCTGCCTCCCGTAGGAGT None 46 48 (Daims et al. 2003)

EUB338 II* Most bacteria GCAGCCACCCGTAGGTGT None 46 48 (Daims et al. 2003)

EUB338 III* Most bacteria GCTGCCACCGTAGGTGT None 46 48 (Daims et al. 2003)

Table 2.8: Hybridization conditions for oligonucleotide probes used in FISH. The dyes are mono-labelled with the indocarbocyanine dye Cy3 (MWG

Biotech, Milton Keynes, U.K). * These eubacterial probes were applied simultaneously and used in equimolar concentrations (all at 50 ng µl -1) to assess total

bacterial counts ( from here on referred to as EUBMIX).

132

2.2.9.4 Dehydration and Fixing of Samples to Slides

On the day of the experiment, the fixed samples stored in ethanol/PBS were removed from the

-20°C freezer and placed on ice. Samples were then dispersed by vortexing thoroughly. Next

the samples were diluted according to the probe and region (see table 5.2) with sterile PBS/SDS

diluent and vortex mixed. Diluted samples (20 µl) were carefully applied to each well of

Teflon- and poly-L-lysine-coated, 6-well, 10-,diameter slides ( Tekdon Inc. Myakka City, FL,

USA) with the appropriate samples. The slides were then dried for approx. 15 minutes at 46-

50°C on a desktop plate incubator. The cells requiring lysozyme treatment (See Table 2.5)

were treated at this stage and the reaction left to incubate on the bench top for 15 minutes. To

ensure sufficient permeabilisation of the cell wall to allow probes to enter the cells efficiently,

some gram-positive bacteria require enzymatic treatments prior to the hybridisation step for

example LAB158 and Bif164. Lysozyme treatment hydrolyses (1 4) linkages between

N- acetylmuramic acid and N-acetyl-D-glucosamine residues in peptidoglycan. Gram-positive

cells are susceptible to this hydrolysis as their wall has a high proportion of peptidoglycan

compared with Gram-negative bacteria which possesses a thinner peptidoglycan membrane.

Prior to treating these cells with lysozyme, an aliquot of the prepared lysozyme was defrosted

and 20 µl of lysozyme was added to each well and incubated on a desktop plate incubator for

15 minutes. After this, the slides were briefly dipped into distilled water to remove the

lysozyme.

All the slides were then immersed fully into 50, 80 and 96% (v/v) ethanol for 3 minutes each.

The dehydrating effect of the ethanol concentration gradient disintegrates cytoplasmic

membranes, which then become permeable to oligonucleotide probes. Following this, the slides

were dried for 2 minutes at 46-50°C on a desktop plate incubator.

133

Probe Region(s) Animal Group Dilution

BIF164 IIeum, Jejunum, duodenum, stomach Casein free and casein rich 1:10

Colon and caecum Casein free and casein rich 1:30

Lab158 All six regions Casein free and casein rich 1:30

Chis150 All six regions Casein free and casein rich 1:10

EUB MIX All six regions Casein free and casein rich 1:300

2.2.9.5 Hybridization of Probes to Samples

20 minutes prior to starting the hybridization step, the Boekel InSlide OutTM Slide hybridisation

oven (Progen Scientific, U.K) was turned on and set at the appropriate temperature for each

probe (see Table 4.2). A piece of absorbent paper towel was soaked with approximately 50 ml

of distilled water and placed in the bottom of the hybridisation tray, ensuring it was flat.

2.2.9.6 Preparation of Hybridisation Buffer

Table 2.6 below shows the components needed for the hybridization buffer. EUB338 (MIX)

probes was the only one requiring formamide (35%). This was done to achieve the correct

hybridisation stringency ensuring probe specificity. The volume required for all slides were

calculated accordingly.

Table 2.9: Sample dilutions used for FISH analysis

134

Volume (µl)

Component 0% Formamide 35% Formamide

5 M NaCl

1 M Tris/HCL (ph 8.0)

Formamide

Double distilled H2O

10% sodium dodecyl sulfate

The Oligonucleotide probes were thawed in the dark and 5 µl of probe was added to 50 µl of

hybridization buffer and mixed by inversing three times. A master mix containing sufficient

probe/hybridization buffer was prepared by calculating the amount needed to cover all slides.

From this mix, 50 µl of probe mixture was applied onto each well containing the dehydrated

sample. The slides were placed onto the slide holders with the samples facing upwards. The lid

of the slide tray was then sealed tightly ensuring it was airtight, this is to ensure no evaporation

of the probe/hybridization mixture occurs. The trays were then placed in the oven and left to

incubate for 4 hours.

2.2.9.7 Washing, Drying and Storage

2.2.9.7.1 Preparation of Wash Buffer

After placing the slide trays in the incubator, a sufficient amount of wash buffer was prepared

as shown below (Table 2.7) in 50 ml centrifuge tubes. EUBMIX required 35% formamide and

so EDTA was added to capture trace amounts of divalent cations, which frequently occur and

reduce the stringency of the washing step because they stabilise the DNA-RNA hydrides. The

180

20

0

799

1

180

20

350

499

1

Table 2.10: Composition of hybridisation buffers used to achieve different hybridization

stringencies for FISH analysis. Volumes given are required for 2 slides. The SDS solution was

added last to inhibit precipitation.

135

wash buffers were then placed in the corresponding temperature in a water bath (see Table

2.4).

Volume (µl)

Component 0% Formamide 35% Formamide

5 M NaCl

1 M Tris/HCL (ph 8.0)

Formamide

Double distilled H2O

10% sodium dodecyl sulfate

Immediately prior to the removal of the slides from the incubation oven, a centrifuge tube

containing 50 of distilled water was placed over ice. At this point 20 µl of 4’6-diamidino-2-

phenylindole dihydrochloride (DAPI;50 NG µl -1 ) was added to the wash buffer tubes. DAPI

was then added to the wash buffer to allow cells to be found easily on the microscope slide.

This reagent binds to the minor groove (at A-T BONDS) of double-stranded DNA, staining

the DNA molecule of the bacteria (Kapuscinski, 1995).

Once the incubation is complete, the hybridization tubes were removed from the oven and the

slides rapidly transferred into the wash buffers in the water baths. This step was carried out

quickly to reduce cooling of the hybridization buffer on the slide that may cause a reduction

in the stringency and non-specific binding of probes to non-targeted organisms. Two slides

were placed back to back into the 50 ml centrifuge tube for 10-15 minutes. Using tweezers,

the slides were then taken out the wash buffers and immersed rapidly into the ice cold

distilled water for 2-3 minutes. This step ensured all residual washing buffer was removed

9.0

1

0

40

1

0.70

1

0.5

47.8

1

Table 2.11: Composition of wash buffer for FISH analysis. Volumes given are for 2 slides only

and prepared in 50 ml centrifuge tubes.

136

which might not otherwise cause crystal formation on the slides thus hampering microscopic

visualisation of fluorescing cells. The low ionic strength of distilled water destabilises nucleic

acid duplexes, and the cold minimises dislocation of probes from the ribosomes of their target

organisms (Daims et al., 2003).

The slides were the placed on paper towels and dried using a stream of compressed air.

Antifade (5 µl) was then added to each well and a cover slide was applied by gently pressing

over the samples to carefully remove any bubbles being formed by the antifade. Slides were

transferred to slide boxes and placed int eh fridge overnight before microscopy.

2.2.9.8 Microscopic Visualisation

The stained sections were inspected in a dark room using EPI-fluorescence Eclipse E400

Nikon microscope (Nikon, U.K., Kingston-upon-Thames, and U.K) equipped with a 100 x

Plan Apochromat oil objective lens and a HBO 100-W vapor lamp. Cy3-stained cells were

viewed with immersion oil under 100 × magnification with the use of a green excitation filter

(510 nm 560 nm) and the DAPI-stained cells were examined under ultraviolet (UV) light

(330nm 380nm). Bacterial cells from fifteen random microscopic fields of view were counted

per sample and used for the calculation of the number of cells in each sample ( g intestinal

sample) -1 using this formula:

Log10 (bacteria / ml of sample) = 0.8 × A × 6732.42 × 50 × B

Where A is the average number of cells counted per field of view, B is the dilution factor

(e.g. 20 if 1 in 20 dilutions used) and 6732.42 is the magnification constant of the

microscope. 0.8 is the dilution factor which was initially used (i.e. 375 µl if sample into 150

µl PBS and 150 µl ethanol 300/375) and 50 µl was used in the calculations as 20 µl of sample

was applied to the slide and so bring the total to 1 ml.

137

2.2.9.9 Data Analysis

The log10 of bacterial counts were analysed for differences for each individual probes. The raw

data counts were tabulated and collected in Microsoft Excel, here they were adjusted to show

the true counts and averages of all 15 repetitions were determined. The final data set was then

transferred to GraphPad Prism 7 for graphing. Statistical analysis was conducted using a two-

way ANOVA followed by Bonferroni post-hoc analysis where applicable. No statistical

analysis was conducted for the Duodenum as there was no casein-rich to compare to the casein-

free samples.

138

2.3 Results

2.3.1 Body Weight and Milk Consumption

Rats provided with casein rich milk consumed significantly more milk from PND22 to PND26

compared to rats provided with casein free milk. Furthermore, rats fed casein rich milk gained

significantly more bodyweight over time compared to control and casein free treated animals

(p ˂ 0.001 Figure 2.2 A and p ˂ 0.01 Figure 2.2 B respectively).

2.3.2 Behaviour

The FST was used to assess depressive-like behaviour in rats provided with no milk (control),

casein free or casein rich milk from PND21 until PND26. The total immobility time and total

time spent climbing were scored from the ‘test’ session on PND26 (Figure 2.3 A and B

respectively). The quantification from the ‘test’ times were used in the assessment of

depressive-like behaviour as described by (Cryan et al., 2005).

Quantification of the results from the ‘test’ swim test session on PND26 showed that rats fed a

casein rich diet displayed depressive-like phenotypes as shown by a marked increase in

immobility time compared to both the control (p < 0.001) and casein free treated animals (p <

0.01). In addition casein rich treated rats spent significantly less time climbing compared to

both control (p < 0.01) and casein free animals (p < 0.005).

Taken together these results show that casein rich treated animals exhibit marked ‘depressive-

like’ behaviour when compared to control animals and animals fed casein free milk.

139

Co

ntr

ol

Casein

Fre

e

Casein

Ric

h

0

1 0

2 0

3 0

4 0

We

igh

t c

ha

ng

e (

g)

***

* *

A

B

18 20 22 24 26 280

50

100

150

Postnatal Days

Mean

Daily

Milk C

on

su

mp

tio

n p

er

Cag

e (

g)

Casein Rich Casein Free

Figure 2.2: Effects of casein free and casein rich diets on milk consumption and body weight.

(A) Milk consumption for casein free and casein rich groups. Animals provided with casein rich milk

consumed significantly more milk compared to their casein free counter parts from PND22-PND26 (B)

Weight change in grams for control, casein free and casein rich groups from PND14 to PND27. Data are

expressed as mean ± SEM (n=8 per group). Casein rich animals showed the greatest weight change

compared to control and casein free animals (**p<0.01, ***p<0.001 one way ANOVA followed by

LSD post-hoc test)

140

Co

ntr

ol

Casein

Fre

e

Casein

Ric

h

0

5 0

1 0 0

1 5 0

2 0 0

2 5 0

3 0 0

Cli

mb

ing

tim

e (

s)

* *

*

Figure 2.3: Increase in depressive-like behaviour following prolonged exposure to milk caseins

beyond the normal age of weaning in rats (pnd21).

Male Wistar Albino rats were provided with no milk, casein free or casein rich milk from PND21

through to PND26 then subjected to behavioural testing using the FST on PND25 & 26 (A) Immobility

time during the 6 minute test session ran on PND26 (B) Total time spent climbing during the 6 minute

test session ran on PND26. Casein rich milk caused an increase in immobility time and a decrease in

total time spent climbing compared to both control and casein free treatment. Data are expressed as

mean scores (± SEM) (n=8) obtained from manual video analysis by two independent observers. One-

way ANOVA followed by LSD post-hoc test was used and a p value of less than 0.05 was accepted as

significantly different. *p=0.05, **p<0.01, ***p<0.001

A

B

Co

ntr

ol

Casein

Fre

e

Casein

Ric

h

0

5 0

1 0 0

1 5 0

2 0 0

2 5 0

3 0 0

Imm

ob

ilit

y t

ime

(s

)

*

*

* *

*

141

2.3.3 Neurochemical Analysis

2.3.3.1 Quantitative Oxytocin Receptor Autoradiography

Quantitative OTr autoradiographic binding using [125I]-OVTA was performed on rat brain

sections cut at the level of the olfactory nuclei, caudate and thalamus from control, casein free

and casein rich treated animals. Figure 2.4 (A) shows reduced OTR binding in the olfactory

nuclei of casein free and casein rich treated animals compared to control (first row). The second

row of Figure 2.4 (A) shows a lower density of OTr binding in the thalamus of casein rich

treated animals compared to control group.

In order to assess the impact of casein rich vs. casein free milk on OT receptor levels, a One-

Way ANOVA for the factor ‘treatment’ was employed and revealed a significant general milk

effect in the AOV and Th of casein rich and casein free treated animals compared to control (p

< 0.05). In the Hip and CE a decrease of OTr binding was observed only in casein rich treated

animals compared to control (p < 0.05) indicating a casein effect. Furthermore, a decrease in

OTr binding was also observed in the BLA, and BMA of casein rich treated animals compared

to both control and casein free (p < 0.01 and p < 0.05 respectively) again highlighting a casein

effect on OTr binding in these regions (Figure 2.4 B). No effect of treatment was found in the

remainder of the regions analysed shown in Table 2.2.

142

AO

L

AO

M

AO

V

Cg

Cx

CP

u

Acb

Pir

Tu

MS

VD

BL

SH

ip Th

BL

A

BM

AC

EH

yp

0 .0

0 .1

0 .2

0 .3

0 .4

0 .5

B r a in R e g io n s

[1

25I]

-OV

TA

sp

ec

ific

bin

din

g

(fm

ol/

mg

tis

su

e e

qu

iva

len

t)

C o n tro l

C a s e in R ic h

C a s e in F re e

*

*

*

**

** *

*

Figure 2.4: Region-specific alterations of [125I]-OVTA binding in rats provided with casein free or

casein rich milk from PND21-PND26.

Male Wistar Albino rats were provided with either no milk, casein free or casein rich milk from PND21

through to PND26 (A) Computer-enhanced representative autoradiograms of adjacent coronal brain

sections from control, casein free and casein rich treated rats at the levels of olfactory nuclei (Bregma

4.20mm, first row) and thalamus (Bregma -2.80mm, second row). OTRs were labelled with [125I]-OVTA

(50 pm). The colour bar illustrates a pseudo-colour interpretation of black and white film images in

fmol/mg tissue equivalent. Representative images for the non-specific (50pm [125I]-OVTA in the presence

of 50 µM unlabelled oxytocin) are shown for all the treatment groups. (B) Quantitative oxytocin receptor

autoradiographic binding from control, casein free and casein rich treated animals. Data are expressed as

mean ± S.E.M (n=8 per group) specific [125I]-OVTA binding (fmol/mg tissue equivalent). * p<0.05;

**p<0.01 (one-way ANOVA followed by LSD post-hoc test). Abbreviations: Acb, nucleus accumbens;

Amy, amygdala; AOL, anterior olfactory nucleus-lateral; AOM, anterior olfactory nucleus-medial; AOV,

anterior olfactory nucleus-ventral; BLA, basal lateral amygdala; BMA, basal medial amygdala; CE,

Central Nucleus of Amygdala; CgCx, cingulate cortex; CPu, caudate-putamen; Hip, hippocampus; Hyp,

hypothalamus; LS, lateral septum; MS, medial septum; Pir, piriform cortex; Th, thalamus; Tu, olfactory

tubercle; VDB, vertical limb of the diagonal band of Broca.

B

A

AOV

BLA

143

2.3.3.2 Quantitative Mu Opioid Receptor Autoradiography

Quantitative MOPr autoradiographic binding using [3 H] DAMGO was performed on rat brain

sections cut at the level of the frontal cortex, striatum and thalamus from control, casein Free

and casein Rich treated animals. Figure 2.5 (A) shows a low density of MOP receptor binding

in the frontal cortex of casein rich treated animals compared to both control and casein free

animals (first row) followed by decreased MOPr binding in the striatum of casein free animals

compared to both control and casein rich groups (second row). The last row of Figure 2.5 (A)

shows reduced MOPr binding in the thalamus of casein rich treated animals compared to both

control and casein free treatment groups.

In order to assess the impact of casein free vs casein rich milk diets beyond the normal age of

weaning, a One-Way ANOVA was employed and revealed a significant general milk effect in

the CgCx, (p< 0.05), AcbC whole (casein free p< 0.05, casein rich p< 0.01 ), AcbC shell (p<

0.01) and the CPu ( casein free p< 0.01, casein rich p< 0.001) in both casein free and casein

rich treated animals compared to controls.

Casein rich milk alone caused a decrease in MOPr binding in M2 (p< 0.05), PrL (p< 0.05),

M1/M2 superficial (p< 0.05), M1/M2 deep (p< 0.05), Amy (p< 0.01) and Hyp (p< 0.01)

compared to control treated animals. Furthermore casein rich milk induced decreased binding

in S1/S2 deep compared to both control and casein free treatment groups (p< 0.05) indicating

a casein effect Figure 2.5 (B). No effect of treatment was found in the remainder of the regions

analysed shown in Table 2.2.

144

A

B

M2

PrL

AO

M

Cg

Cx

Tu

CP

u

Acb

C w

ho

le

Acb

C s

hell

Acb

C c

ore

M1/M

2 S

up

er f

icia

l

M1/M

2 D

eep

S1/S

2 S

up

er f

icia

l

S1/S

2 D

eep

Hip T

h

Am

y

Hyp

0

2 0

4 0

6 0

B r a in R e g io n s

[3

H] D

AM

GO

sp

ec

ific

bin

din

g

(fm

ol/

mg

tis

su

e)

C a s e in R ic h

C a s e in F re e

C o n tro l

*

*

*

* * *

**

*

* ** *

*

*

*

*

**

**

Frontal cortex

S1/S2

Figure 2.5: Region specific alterations of [3H] DAMGO specific binding in rats provided with

casein free or casein rich milk from PND21-PND25.

Male Wistar Albino rats were provided with either no milk, casein free or casein rich milk from

PND21 through to PND26 (A) Computer-enhanced representative autoradiograms of adjacent coronal

brain sections from control, casein free and casein rich treated rats at the levels of the frontal cortex

(Bregma 4.20, first row), striatum (Bregma 1.20, second row) and thalamus (Bregma -2.30, third row).

MOPrs were labelled with [3H] DAMGO (4nM). The colour bar illustrates a pseudo-colour

interpretation of black and white film images in fmol/mg tissue equivalent. (B) Quantitative oxytocin

receptor autoradiographic binding from control, casein free and casein rich treated animals. Data are

expressed as mean ± S.E.M (n=8 per group) specific [3H] DAMGO binding (fmol/mg tissue

equivalent). * p<0.05 **p<0.01 (one-way ANOVA followed by LSD post-hoc test). Abbreviations:

Acb, nucleus accumbens; Amy, amygdala; AOM, anterior olfactory nucleus-medial; CgCx, cingulate

cortex; CPu, caudate-putamen; Hip, hippocampus; Hyp, hypothalamus; M1/M2 motor coretex; PrL,

prelimpic cortex; S1/S2 somatosensory cortex; Th, thalamus; Tu, olfactory tubercle;

145

2.3.3.3 Quantitative Delta Opioid Receptor Autoradiography

Quantitative DOPr autoradiographic binding using [3H] Deltorphin -1 was performed on rat brain

sections cut at the level of the thalamus from no milk control, casein free and casein rich treated animals.

Figure 2.6 (A) shows a low density of DOPr in the somatosensory cortex of casein free treated animals

compared to both control and casein rich.

In order to assess the impact of casein free vs. casein rich milk diets beyond the normal age of weaning,

a One-Way ANOVA was employed and showed casein free milk induced decreased binding in S1/S2

deep (p< 0.05) compared to both control and casein rich treatment groups. Indicating a soy protein

effect Figure 2.6 (B). No effect of treatment was found in the remainder of the regions analysed shown

in Table 2.2

146

A

B

Acb

CP

u

Cg

Cx

Tu

Pir

AO

MP

r l

M1/M

2 s

up

er f

icia

l

M1/M

2 d

eep

S1/S

2 s

up

er f

icia

l

S1/S

2 d

eep

Hip T

h

Am

yH

yp

0

5 0

1 0 0

1 5 0

2 0 0

2 5 0

B ra in R e g io n s

[3

H]

de

lto

rp

hin

-1

Sp

ec

ific

Bin

din

g

(fm

ol/

mg

tis

su

e)

C a s e in R ic h

C a s e in F re e

C o n tro l

*

Figure 2.6: Down-regulation of [3H] Deltorphin -1 specific binding in the deep layer of the somatosensory

cortex of rats fed casein free milk beyond the normal age of weaning (PND21).

Male Wistar Albino rats were provided with either no milk, casein free or casein rich milk from PND21 through

to PND25 (A) Computer-enhanced representative autoradiograms of adjacent coronal brain sections from

control, casein free and casein rich treated rats at the levels of Thalamus (Bregma -2.30). DOPrs were labelled

with [3H] deltorphin -1 (7nM). The colour bar illustrates a pseudo-colour interpretation of black and white

film images in fmol/mg tissue equivalent. (B) Quantitative oxytocin receptor autoradiographic binding from

control, casein free and casein rich treated animals. Data are expressed as mean ± S.E.M (n=8 per group)

specific [3H] deltorphin -1 binding (fmol/mg tissue equivalent). * p<0.05 (one-way ANOVA followed by LSD

post-hoc test). Abbreviations: Acb, nucleus accumbens; Amy, amygdala; AOM, anterior olfactory nucleus-

medial; CgCx, cingulate cortex; CPu, caudate-putamen; Hip, hippocampus; Hyp, hypothalamus; M1/M2 motor

coretex; Pir, piriform cortex; PrL, prelimpic cortex S1/S2 somatosensory cortex; Th, thalamus; Tu, olfactory

tubercle.

147

2.3.4 Analysis of Urine Samples (Metabolomics)

2.3.4.1 PCA of Urine Samples

PCA was first used as an unsupervised multivariate statistical analysis method to give an

insight into how urine samples from control, casein free and casein rich treatment groups

differed from one another. The percentage of variance explained by each PC (R2) is shown for

each axis. In order to aid visualisation of any trends associated with casein free or casein rich

test groups, the two different groups are shown colour coded according to their group, with

green circles representing casein free animals and blue circles representing casein rich animals.

Following initial PCA modelling of the urine data, clear separation along PC1 can be seen

between casein free and casein rich treated animals Figure 2.6 (A). This indicates that casein

exposure from PND21 until PND26 is responsible for the largest source of variation in the

metabolic data.

2.3.4.2 1H-NMR Spectroscopy of Urine Samples

OPLS-DA was used to further investigate the effect of exposure to casein free or casein rich

milk on animal metabolic profiles. This supervised method allows for the examination of the

data in question with reference to only one source of variation as opposed to all possible

variations as was the case with the PCA approaches Figure 2.6 (B). The model had a strong

predictive ability (Q2Y = 0.27) that was found to be significant after permutation testing (p =

0.0; 10,000 permutations). From the correlation coefficients plot extracted from the OPLS-DA

model, casein rich rats were found to excrete higher amounts of hippurate and betaine than the

casein free rats and lower amounts of benzoate, glycine, acetate and alanine.

148

A

B

Figure 2.7: Multivariate analysis of urinary 1H NMR spectral profiles showing the influence of casein free diet and

casein rich diet on urinary metabolites.

Male Wistar Albino rats were provided with either no milk, casein free or casein rich milk from PND21-PND26 (n= 8 per

group) after which they were placed into individual metabolic cages to allow for the collection of urine. (A) PCA scores

plot of normalised NMR data showing separation along PC1 between casein free and casein rich groups (B) OPLS-DA

coefficient plot showing all metabolite responsible for driving the separation between the groups. Labelled peaks are the

metabolites which are increased in the respective groups (Q2Y =0.27).

149

Metabolite Chemical Shift (Multiplicity) Sample

Acetate 1.92 (s)

Alanine 3.77 (q), 1.8 (d)

Betaine 3.27 (s), 3.91 (s)

Benzoate 7.49 (t), 7.56 9t)

Glycine 3.57 (s)

Hippurate 3.97 (d), 7.64 (t), 7.84 (d)

Casein-free

Casein-free

Casein-rich

Casein-free

Casein-free

Casein-rich

Table 2.12: Full 1H- NMR chemical shifts for metabolites identified in Urine samples. Key: s,

singlet; d, doublet; t, triplet; q, quadruplet

150

2.3.5. Assessment of Intestinal Microbiota (FISH)

2.3.5.1 Assessment of Total Bacteria

Two-way ANOVA for factors ‘diet’ and ‘gut region’ showed no significant effect of either

factor in EUB338 probe counts between casein free and casein rich groups in all of the gut

regions analysed therefore no post hoc analysis was carried out (Figure 2.8).

Sto

mac

h

Jeju

num

Ileum

Cae

cum

Colo

n

7.0

7.5

8.0

8.5

9.0

9.5

10.0

Region

Lo

g10 C

ou

nt

of

To

tal B

acte

ria

Casein Rich (log values)

Casein Free (log values)

Total Bacteria (EUB338)

Figure 2.8: Effect of casein free vs. casein rich milk diets from PND21-PND26 on the composition

of total bacteria within different gut regions.

Male Wistar Albino rats were provided with either casein free or casein rich milk from PND21 through

to PND26. A mixture of EUB 338 I,II & III probe were used for the enumeration of total bacteria in

contents obtained from five different regions of the intestine from the study animals. All data expressed

as mean log10 cells (g wet weight of intestinal content)-1 , ± S.E.M (n=8 per group). Two-way ANOVA

was carried out and no significant differences were observed between the two groups, therefore no post hoc

analysis was carried out.

151

2.3.5.2 Assessment of Clostridium Histolyticum Groups

Two-way ANOVA for factors ‘diet’ and ‘gut region’ of Chis150 probe counts showed a

significant ‘diet’ [ F (1,60) = 21.56; p <0.001 ] and ‘gut region’ [ F (4,60) = 64:64; p <0.001 ]

and ‘diet’ x ‘gut-region’ interaction [ F (4,60) = 3.763; p <0.0085 ]. Post-hoc analysis revealed

casein rich animals harboured increased total numbers in the caecum (p= 0.01) and colon (p=

0.001) compared to casein free treated animals (Figure 2.9).

Sto

mac

h

Jeju

num

Ileum

Cae

cum

Colo

n4.5

5.0

5.5

6.0

6.5

7.0

7.5

8.0

8.5

Region

Lo

g10 C

ou

nt

of

Clo

str

idiu

m H

isto

lyti

cu

m



*****

Clostridium histolyticum (Chis150)

Figure 2.9: Effect of casein free vs. casein rich milk diets from PND21-PND26 on clostridium

histolyticum bacterial groups within different gut regions.


to PND26. The Chis150 probe was used for the enumeration of C.histolyticum spp. In contents obtained

from five different regions of the intestine from the study animals. All data expressed as mean log10

cells (g wet weight of intestinal content)-1 , ± S.E.M (n=8 per group). Animals provided with casein

rich milk had significantly higher levels of C.histolyticum in the caecum and colon compared to their

casein free counter parts (**p<0.01, ***p<0.001 two way ANOVA followed by Bonferroni post-hoc test).

152

2.3.5.3 Assessment of Bifidobacterium Groups


factor in Bif164 probe counts between casein free and casein rich groups in all of the gut

regions analysed, therefore no post-hoc analysis was carried out (Figure 2.10).

Sto

mac

h

Jeju

num

Ileum

Cae

cum

Colo

n

6.5

7.0

7.5

8.0

8.5

9.0

Bifidobacterium (Bif164)

Region

Lo

g10 C

ou

nt

of

Bif

ido

bacte

riu

m



Figure 2.9: Effect of casein free vs. casein rich milk diets from PND21-PND26 on Bifidobacterium

bacterial groups within different gut regions.


to PND26. The Bif164 probe was used for the enumeration of Bifidobacterium spp. In contents obtained

from five different regions of the intestine from the study animals. All data expressed as mean log10

cells (g wet weight of intestinal content)-1, ± S.E.M (n=8 per group). Two-way ANOVA was carried out

and no significant differences were observed between the two groups, therefore no post hoc analysis was

carried out.

153

2.3.5.4 Assessment of Lactobacillus-Enterococcus Groups


factor in Lab158 probe counts between casein free and casein rich groups in all of the gut

regions analysed, therefore no post-hoc analysis was carried out (Figure 2.11).

Sto

mac

h

Jeju

num

Ileum

Cae

cum

Colo

n

6.5

7.0

7.5

8.0

8.5

9.0

9.5

Lactobacillus-Enterococcus (Lab158)

Region

Lo

g10 C

ou

nt

of

Lacto

bacillu

s-E

nte

roco

ccu

s



Figure 2.10: Effect of casein free vs. casein rich milk diets from PND21-PND26 on Lactobacillus-

Enterococcus bacterial groups within different gut regions.


to PND26. The Lab158 probe was used for the enumeration of Lactobacillus-Enterococcus spp. In

contents obtained from five different regions of the intestine from the study animals. All data expressed

as mean log10 cells (g wet weight of intestinal content)-1 , ± S.E.M (n=8 per group). Two-way ANOVA

was carried out and no significant differences were observed between the two groups, therefore no post hoc

analysis was carried out.

154

2.4 Discussion

The aim of this study was to investigate whether exposure to milk caseins beyond the normal

age of weaning (PND21) would result in the development of a depressive-like phenotype in

male Wistar Albino rats. Moreover, we investigated the effects of prolonged casein exposure

in early postnatal life on the development of the opioid system, namely MOPr and DOPr as

well as OTr. Finally we investigated the effect of 5 extra days of casein exposure beyond the

normal age of weaning on gut microbial populations as well as urinary metabolites. The results

revealed rats fed casein rich milk beyond the normal age of weaning displayed a depressive-

like phenotype which was concomitant with a down regulation of OTr in the amygdala, as well

as a down regulation of MOPr in a number of brain regions associated with mood. Furthermore,

these rats also displayed higher levels of Clostridium Histolyticum (C.Histolyticum) species in

the gut compared to their casein free counter parts as well as an increase in microbial urinary

metabolites. Interestingly, we also observed a general milk effect with rats fed both casein free

and casein rich milk beyond the normal age of weaning displaying a down regulation of both

MOPr and OTr in a number of brain regions indicating a casein independent effect on brain

neurochemistry. Moreover, animals fed casein free milk but not casein rich milk displayed a

down regulation of DOPr in the deep layer of the somatosensory cortex, however these animals

did not display a depressive phenotype, suggesting that soy protein exposure in early postnatal

life also influences normal brain development but not necessarily depressive-like behaviour.

A rat model mimicking delayed weaning or weaning onto cow’s milk based formula was

established by providing rat pups with a casein rich milk formula (containing approx. 20% w/w

casein) from PND21 (the standard age of weaning) till PND26. Another group was provided

with a casein free milk formula, custom manufactured to match the nutritional composition of

the casein rich milk except casein protein was removed and replaced with approx. 20% w/w

155

soya protein), again from PND21 until PND26. The control group, which mimicked normal

weaning practices, where no milk is provided past weaning, had their mothers removed on

PND21 and were provided with no milk post-weaning.

Animals provided with casein rich milk consumed significantly more milk from PND22 till

PND26 compared to animals provided with casein free milk (see Figure 2.2 A). This is in line

with findings by Goody and Kitchen (2001) who provided casein free or casein rich milk

formula to 21 day old rat pups and reported that the casein rich group consumed significantly

greater amounts than their casein free counterparts. These observed differences in voluntary

milk consumption could be attributed the characteristic taste and smell of soya making the

casein free milk less palatable. Another potential mechanism which could be driving the

differences in milk consumption between the casein free and casein rich treated animals could

be linked to the direct effects of the milk proteins on the gut. The casein free milk contained

soy protein which is a mixture of different proteins classified into 4 categories, with β-

conglycinin and glycinin making up more than 80% of the soy protein content (Saio et al.

1969). The β-conglycinin protein is made up of three subunits, namely α, α1 and β (Tierney et

al., 1987). It has been shown that the β-subunit contains the same sequence as bovine and

human β-casomorphins (YPFV). Peptides which have a Tyr-Pro-aromatic amino acid sequence

are selective for the MOPr (Brantl et al., 1982) (see Section 1.3 ). Indeed it has been shown

that the peptides derived from the β-subunit of β-conglycinin also possess opioid activity and

were therefore named soymorphins (Ohinata et al,. 2007). A study by Kaneko et al., (2010)

found that oral administration of soymorphins in fasted rats resulted in a suppression of food

intake and small intestinal transit time. This effect was found to be inhibited by naloxone, hence

indicating that this was MOPr mediated effect. In general it has been well documented that

MOPr agonists such as those derived from milk caseins enhance food intake, while MOPr

antagonists supress food intake (see review by Yeomans et al., 2002). Soymorphins are the

156

first MOPr agonist peptides that were found to suppress food intake (Kaneko et al., 2010).

Although we did not measure food intake in this current study, it is feasible that soymorphins

found in the casein free milk could be responsible for the reduced amount of milk consumption

observed in this group, while BCM-7 derived from casein rich milk could be stimulating gut

MOPr to result in higher milk consumption.

Furthermore, the preference for casein rich milk over the casein free formula was very evident

in this study and indicates that the difference in consumption between casein free and casein

rich milk may represent reward properties of casein and its breakdown products found only in

the casein rich milk. Although previous studies on the reward (reinforcing) potential of milk

derived β-casomorphins did not demonstrate conditioned place preference following systemic

administration, while morphine administration did induce place preference in the same

experiments (Reid et al., 1994), it is still possible that casein rich milk was more rewarding

than its casein free variant leading to higher consumption.

Moreover, casein rich animals gained significantly more weight compared to control and casein

free treated animals (see Figure 2.2 B). The significant difference in weight gain between

casein rich and control animals indicates that, not surprisingly the absence of milk formula

results in lower body weights by 26 days of age which is more than likely due to the absence

of this significant nutritional source from the diet of control vs. casein rich animals. Casein rich

treated animals also gained significantly more weight compared to the casein free group. This

result can be explained by the fact that casein free animals consumed significantly less milk

then their casein rich counterparts and therefore gained less weight.

Furthermore, the nutritional composition of soy protein vs casein protein could also be driving

the difference observed in weight gain between the two groups. This is in line with a study by

Silva et al., (2015) who provided 21 day old mice with either soy-based infant formula or cow’s

157

milk-based infant formula for a period of 10 days. They measured milk consumption and

weight gain throughout the duration of the study and found that mice provided with soy-based

infant formula gained significantly less weight compared to mice who consumed the cow’s

milk based infant formula. This result was in part attributed to the differences in amino acid

composition between soy protein and casein. For example soya protein has lower levels of

methionine, lysine and proline. Moreover, it has previously been shown that soya based

formula contains anti-nutritional factors which prevent the absorption of nutrients such as

phytic acid (Lönnerdal, 1994).

It was hypothesised that consumption of milk caseins beyond the normal age of weaning would

result in depressive–like behaviour due to impairments to both the opioid system and the

oxytinergic system. Indeed, the results obtained in this chapter demonstrate that animals

provided with casein rich milk from PND 21 until PND 26 exhibited ‘depressive-like’

behaviour as evidenced by a significant increase in the total time spent immobile and a decrease

in time spent climbing in the FST compared to both control and casein free treated animals (see

Figure 2.3). These results extend previous findings from our laboratory conducted by Farshim

et al., (2016) which showed that rat pups which are maintained in a non-weaned state exhibit

depressive-like behaviour compared to rat pups weaned at the normal age of 21 days old.

Taken together, these results suggest that it is the prolonged exposure to maternal milk caseins

in the non-weaned rats that maybe responsible for the depressive phenotype observed.

It is possible to argue that the difference in weight gain between the three treatment groups

discussed above could have influenced the immobility times in the FST, especially as evidence

shows weight is indeed a factor which can influence immobility in the FST (Bogdanova et al.,

2013). However in order to address this, the total immobility time for each test animal (in

seconds) was divided by the individual body weights of animals on the day of testing ( animal

weight in grams on PND26) and these results were plotted and statistically analysed using a

158

one-way ANOVA followed by LSD post-hoc analysis. The results revealed that the significant

increase in immobility observed in the casein-rich treated animals was still present in

comparison to both casein free and control animals even after taking body weight into account.

This analysis provides assurance that the difference in weight between the three test groups did

not affect depressive-like behaviour. Indeed various methods of statistical analysis have

previously been carried out by other groups to account for variation in weight of animals used

in the FST (Brenes et al., 2008).

Previous investigations from our laboratory conducted by Kitchen et al., (1995) have

demonstrated weaning rat pups on PND21 activates a population of DOPrs which mediate

swim SIA whereas maintaining pups in a non-weaned state until PND 25 resulted in a lack of

development of these DOPrs. Moreover, it was shown that casein exposure was responsible for

mediating this effect as the removal of casein from the neonatal rat diet at the time of weaning

acted as developmental trigger for the transition of MOPr to DOPr in the mediation of swim

SIA (Goody and Kitchen 2001). This transition can be delayed up to PND40 by continued

dietary provision of casein rich milk formula. This set of experiments were able to exclude the

possibility of psychological (maternal presence) or physiological (suckling) influence upon this

transition and showed that this was driven by nutritional factors with potential candidates being

casein derived BCM7 or similar compounds. The results obtained in this chapter show that

casein consumption beyond the normal age of weaning results in a depressive phenotype. As

DOPrs are known to modulate depressive-like behaviour (Filliol et al., 2000) it is possible that

a dysfunction of DOPrs caused by casein exposure could be responsible for the depressive

phenotype observed in animals provided with casein rich milk till PND26.

In order to investigate this, receptor binding autoradiography on DOPrs was performed in rats

treated with no milk (control), casein free milk or casein rich milk from PND 21-PND 26.

Interestingly, a significant down regulation of DOPrs was observed in the deep layer of the

159

somatosensory cortex of animals treated with casein free milk compared to both control and

casein rich treated animals which is in apparent contradiction with the original hypothesis (see

Figure 2.6). Based on the DOPr KO model created by Filliol et al., (2000) the depressive

phenotype observed in casein rich treated rats was predicted to be mediated by a down-

regulation of DOPrs in the brain caused by casein and its bioactive breakdown product BCM-

7. However the results obtained revealed that casein rich treated animals displayed a depressive

phenotype which was not accompanied with a reduced DOPr density in the somatosensory

cortex, but it is instead casein free treated rats which displayed no signs of depressive-like

behaviour, who had a significant reduction of DOPrs in the somatosensory cortex. This data

suggests that DOP receptors may play a differential role in mediating depressive-like

behaviour. Consistent with this theory, a recent study by Chu Sin Chung et al., (2015) who

created a conditional knockout mouse line that lacks DOPrs in specific brain regions, showed

a low-anxiety phenotype in these mice. Indicating that DOPrs exert both anxiolytic and

anxiogenic activities at distinct levels of neural circuits. As decreased DOPr density in the

somatosensory cortex was not associated with emotional impairment in our study, it is unlikely

that the depressive phenotype observed in casein rich rats was due to DOPr dysregulation

induced by casein. Nonetheless, the dysregulation of DOPr in the casein free rats may be

responsible for the modulation of other behaviours which may be affected by casein free milk.

Indeed, as described earlier, casein free milk contains soymorphins which may affect DOP

regulation and as a result DOPr mediated behaviours (e.g. social reward, pain). This set of

results warrants further investigation in order to determine the implications of these findings.

Next, MOP receptor binding levels in animals treated with no milk, casein free or casein rich

milk from PND21-PND26 were investigated. It can be seen that both casein free and casein

rich milk appear to have caused a down-regulation of MOPr in the cingulate cortex and the

shell of the nucleus accumbens indicating a general milk effect independent of casein or soya

160

on brain neurochemistry (see Figure 2.5). Continued or intermittent agonist stimulation of a

receptor results in loss of receptor activity termed desensitization (Lefkowitz et al., 1983). This

is believed to be a homeostatic mechanism and is triggered in response to continued agonist

stimulation of a receptor. A large body of literature shows that one of the ways the MOP

receptor is regulated following long-term opioid exposure is via desensitization and/or

internalization (see Williams et al., 2013). A number of studies have shown peptide opioid

ligands such as DAMGO can cause desensitization or internalization of the receptor (Yu et al.,

1997; Whistler et al., 1998; Koch et al., 2001; Bailey et al., 2009). Whereas non peptide ligands

such as morphine cause desensitization of the receptor (Levitt et al,. 2012). In this present

study, it is likely that bioactive components liberated from both casein free and casein rich milk

with an affinity for MOPr have resulted in desensitization and subsequent down-regulation of

the receptor. Upon proteolytic cleavage in the gut, beta-caseins found in the casein rich milk

release BCM-7. BCM-7 has previously been detected in the plasma of infants suffering from

ASD, indicating BCM-7 can cross over from the luminal side of the gut into the blood (Kost et

al., 2009b). Evidence also suggests that BCM-7 may possess the ability to cross the BBB and

elicit effects directly on the brain. A study by Ermisch et al. (1983) measured the accumulation

of radioactivity in different brain regions following an intracarotid injection of [3H] BCM-5 in

rats. The results showed a low but measurable brain uptake of the peptide. Although BCM-7

differs slightly in structure to BCM-5 it is feasible that it also possesses the ability to cross the

BBB especially in early postnatal life when the BBB is not yet fully formed. It is thus

conceivable that the milk induced down regulation of MOPr may be due to reported

desensitization and internalization of the receptor by an excess of soymorphin and

casomorphins in the brain from PND21 until PND26.

Indeed, as discussed earlier, soya protein has been shown to liberate soymorphins. These

soymorphins have been shown to have a similar affinity to MOPr as bovine BCM-7. For

161

example the IC50 of soymorphin-5 is 6.0 µm (Ohinata et al., 2007) compared to 5.14 µm for

bovine BCM-7 (Koch et al., 1985). It is therefore feasible that soymorphins derived from soy

protein acted directly on MOPr found in the brain to cause desensitization or internalization of

the receptor.

Another proposed mechanism by which BCM-7 or soymorphins may be causing a reduction in

MOPr density is via epigenetic regulation. Evidence shows that BCM-7 and other 7aa-proline

containing food derived opioid peptides can inhibit cysteine uptake in cultured human neuronal

and GI epithelial cells. The decrease in cysteine uptake was accompanied by a decrease in

levels of the antioxidant glutathione (GSH) whose synthesis is limited by cysteine availability

(Trivedi et al., 2014). This may explain the downregulation of the MOPr observed in the CgCx,

CPu, This also indicates that BCM-7 can affect oxidative stress levels. Furthermore it was

shown that food derived opioid peptides increased genome wide DNA methylation in the

transcription start site region with a potency order similar to their inhibition of cysteine uptake.

Altered expression of genes involved in redox and methylation homeostasis was also observed

(Trivedi et al., 2014). DNA methylation of genes is well known to be associated with decreased

gene expression (Moore et al., 2013), it is therefore tempting to speculate that a potential

mechanism by which BCM-7 or soymorphin exposure causes a downregulation of MOPr DOPr

and OTr may be related to changes in DNA methylation and/or regulation. Taken together,

these findings illustrate the potential of BCM-7 and other food derived opioid peptides to exert

epigenetic changes, which may ultimately result in reduced levels of opioid and oxytocin

receptors observed in this current study.

Unlike casein rich fed rats, casein free treated animals showed no depressive phenotype in the

FST despite the neurochemical changes discussed. In support of this, it has been shown that

intraperitoneal administration of soymorphin-5, 6, and 7 at doses of 3-10 mg/kg resulted in

anxiolytic behaviour as demonstrated by increased time spent in the open arms of the elevated

162

plus maze (Fukuda et al., 2011). Given the high co-morbidity between anxiety and depression,

this could explain why casein free treated animals showed no signs of depressive-like

behaviour unlike their casein rich counter parts despite eliciting similar neurochemical changes

to the MOPr receptor in the CgCx and Acb of casein rich fed rats.

Lastly, MOP receptor binding results revealed a significant casein only effect in the deep layer

of the somatosensory cortex compared to control animals (see Figure 2.5). Interestingly, this

downregulation is concomitant with the depressive like phenotype observed in rats exposed to

casein rich milk. Although it is impossible to prove any causality between MOPr

downregulation and the emergence of depressive like behaviour based on this study alone, the

possibility is intriguing. Indeed MOPr have been show to play a differential role in regulating

emotional responses. MOPr KO mice have been shown to exhibit decreased anxiety and

depressive like behaviour (Filliol et al., 2000), whilst acute pharmacological activation of

MOPr was shown to reduce depressive-like behaviour in rodent models (Zhang et al., 2006).

Moreover, a recent study has shown that MOPr play differential roles in mediating mood and

behaviour depending on their location in the brain (Charbogne et al., 2017). Although a

downregulation of MOPr was present in brain regions linked to emotionality, such as the

hypothalamus and amygdala compared to control animals, no difference in MOPr density was

observed in these regions compared to casein free animals. Thus, the possibility that the

changes in these specific regions are driving the depressive phenotype seen in casein rich

animals is unlikely. However, the deep layer of the somatosensory cortex showed a significant

downregulation of MOPr in casein rich animals compared to both control and casein free.

Altogether we can conclude that exposure to either casein rich or casein free (soya rich) milk

during the sensitive early developmental period effects the endogenous opioid system in the

brain. These results suggest that casein and its break down products (most likely BCM7)

supress the development of MOPr in the deep layer of the somatosensory cortex, whereas soya

163

(most likely its breakdown products soymorphins) supress the development of DOPr also in

the deep layer of the somatosensory cortex. Interestingly this is the same region in which

(Kitchen et al.,1995) observed a down regulation of DOPr in rats which were delayed weaned

on PND25, indicating the somatosensory cortex is the main site of action of food derived opioid

peptides. The behavioural consequences of these alterations should be the focus of future

investigations. Nonetheless, as the reduction of MOPr density in the deep layer of the

somatosensory cortex of casein rich animals is concomitant with the depressive phenotype seen

in this group, we can postulate that there may be a link between MOP receptor alterations and

behaviour.

Given the impact of experiences during early postnatal development, on shaping the central

OTr system, including weaning time, combined with the key role of oxytocin in emotional

regulation, we hypothesised that dysregulation of the oxytocinergic system may be involved in

the observed expression of depressive-like behaviour in animals provided with casein rich milk

beyond the normal age of weaning. Indeed, the results obtained in this chapter revealed a

significant down-regulation of OTr receptor density in the Basalmedial Amygdala (BMA) of

casein rich treated animals compared to both casein free and control groups (see Figure 2.4).

These results extend the findings by Farshim et al., (2016) who demonstrated that delayed

weaning influences the stress-induced modulation of OTr density in the BLA amygdala. The

current results go beyond that to demonstrate that it is a nutritional component, milk caseins,

which are likely to be responsible for the alterations to OTr density in the amygdala observed

in delayed weaned rat pups reported by Farshim et al., (2016).

The amygdala is strongly implicated in the modulation of behavioural and neuroendocrine

stress responses (see LeDoux 2000) in particular with respect to the OT system (Bale et al.,

2001; Neumann et al., 2000). Evidence shows that OT release in the amygdala can cause

inhibition of the activity of neuronal populations that project to the hypothalamic and brainstem

164

areas regulating peripheral stress and fear responses respectively (Huber et al., 2005; Viviani

et al., 2011). Furthermore a number of studies have shown that exposure to the FST which is a

stressor (Abel, 1994), triggers OT release into the brain as shown by increased concentrations

of OT in a number of brain regions known to be involved in the modulation of stress

mechanisms such as the PVS, SON and the amygdala (Landgraf and Neumann 2004). Increases

in OT release in the amygdala attenuates fear responses to stressful stimuli (which is believed

to shift stress coping) via a number of different mechanisms including regulation of the HPA

axis activity (Neumann et al., 2000). Indeed studies have reported corticosterone levels to

increase during and after the Forced swim test (Wotjak et al., 1998). Therefore OT release in

the amygdala regulates active and passive stress-coping strategies under stressful paradigms

(Ebner et al., 2005).

Indeed, the antidepressive-like effects of oxytocin have been well documented. Arletti and

Bertolini (1987) demonstrated that intra-peritoneal administration of OT both acutely and

chronically decreased immobility time in the FST in mice. Furthermore, subcutaneous OT was

found to decrease the number of escape failures in the learned helplessness test (Nowakowska

et al., 2002) which is indicative of an antidepressant-like effect (Cryan et al., 2002). Moreover,

activation of the oxytocinergic system have also been reported amongst depressed patients

possibly as a homeostatic coping mechanism to counteract the depressive state (Purba et al.,

1996). It is therefore clear that oxytocinergic system is highly implicated in depressive like

responses. However, it is unclear whether such alteration in the oxytocinergic system represent

cause or consequence of depression.

In this current study there are two likely possibilities for the down-regulation of OT receptors

observed in the BLA of casein rich treated animals compared to both control and casein free

groups. Firstly, casein rich milk is somehow causing these animals to become depressed and

highly stressed, and therefore release higher levels of endogenous OT in the amygdala during

165

the FST as a coping mechanism that subsequently causes a significant down-regulation of OTr

in the casein rich animals. In line with this, there is evidence to suggest that the high degree of

plasticity in the OTr system may constitute a protective adaptive coping mechanism in response

to stress resulting in a shift of brain function and behaviour (Neumann et al., 2012). Hence,

these results indicate that casein rich animals were significantly more stressed compared to

casein free and control as evidenced by the down-regulation of OTr in the BLA. Secondly,

casein rich milk could be causing the depressive phenotype observed by dysregulating the OTr

system and hence resulting in an absence of homeostatic stress coping mechanisms compared

to casein free and control animals. Although it is not possible to prove any causal effect

between the emergence of depressive like behaviour and hampered development of the OTr in

casein rich treated rats.

Exactly how casein rich milk is dysregulating the OT system remains unclear. However, the

opioid system has been shown to interact with the oxytocinergic system in a number of studies

(see Georgiou et al., 2015). In particular, acute morphine administration was shown to decrease

hypothalamic OT release in female rodents (Haldaret al. 1978; Clarke et al. 1979). While

Kovács et al., (1987) observed increased OT immunoreactivity in some regions including the

amygdala of male mice, suggesting differential effects of acute opioid administration on OT

neurotransmission in different areas of the brain. In this current study, it is feasible that opioid

peptides derived from casein such as BCM-7 could be eliciting similar effects on the OT system

resulting in a down regulation of OTr in the basal lateral amygdala. Furthermore, recent

evidence linking the gut microbiota to the oxytocinergic system (Erdman et al., 2016) may

implicate the gut microbiota as a key mechanism underlying the effect of prolonged casein

consumption on the oxytocinergic system and behaviour. As casein rich milk exposure results

in alterations to the gut microbiome (see below), it maybe conceivable that milk caseins induce

the depressive phenotype by a mechanism implicating the oxytocin-gut-brain axis.

166

In order to investigate the role of the gut-brain axis in mediating the neurochemical and

depressive phenotype observed in casein rich treated animals, we sought to investigate the

effects of casein rich vs. casein free milk diets from PND 21-PND 26 on the composition of

several representative groups of intestinal microorganisms that have previously been

implicated in modulating neurobehavioral development. The results obtained from the FISH

analysis revealed negligible variation between casein free and casein rich treatment on total

Bacteria (Figure 2.8) Bifidobacterium (Figure 2.10) and Lactobacillus (Figure 2.11)

indicating that an additional 5 extra days of milk exposure was insufficient to increase the

abundance of these bacterial genera. In contrast, the population of bacteria belonging to the

C.histolyticum group were significantly increased in casein rich treated animals compared to

casein free (see Figure 2.9). Coupled with the increase in depressive-like behaviour observed

in these animals, it is feasible to suggest that casein exposure beyond the normal age of weaning

could be acting via the gut-brain axis to result in a depressive phenotype.

In line with this finding, Farshim et al., (2016) have previously shown that rat pups which were

delayed weaned on PND25 harboured significantly higher numbers of C.histolyticum

compared to animals weaned normally on PND21 after being exposed to the FST. These

delayed weaned animals also showed a significant increase in depressive like behaviour

(Farshim et al., 2016). Combined with the results from this current study, it is feasible to

suggest that the increase in C.histolyticum seen in non-weaned animals was caused by

prolonged exposure to milk caseins in particular rather than a general milk effect post weaning,

as animals provided with casein free milk from PND21-PND25 did not display a similar

increase in C.histolyticum.

Clostridia are recognised as a group of toxin-producing bacteria and their metabolic products

have been shown to exert systemic effects (Voth et al., 2005). Specifically they are known to

produce five toxins including alpha, beta, delta, gamma and epsilon (Hatheway, 1990). It is

167

feasible that casein exposure resulted in an increase of C.histolyticum bacterial groups in the

gut and subsequently an increase in the expression of toxic products in the bloodstream exerting

effects on the developing brain. As such, it has been shown that the faecal microbiota of patients

suffering from ASD were found to contain a higher number of C.histolyticum groups compared

to healthy children (Parracho et al., 2005) indicating an association between clostridia and

neurodevelopmental disorders. Moreover, an increase in clostridium species has previously

been shown to increase with stress, together with increased levels of circulating cytokines

(Bailey et al,. 2011). Indeed, one of the hallmarks of depression is increased levels of

circulating pro-inflammatory cytokines (Dantzer et al., 2008). To this end, the effect of the gut

microbiota on host immune system and in regulating behaviour via the gut brain axis has been

well documented (see Bercik et al., 2012). Furthermore, it was also shown that stress induced

increases in clostridia together with increasing levels of cytokines was shown to result in

translocation of bacteria and/or their products (Janeczko et al., 2008). Indeed bacterial

recolonization has previously been shown to be linked to mood disorders such as depression

(Bailey et al., 2011; Maes 2008; Collins et al., 2009). Hence, providing a possible mechanism

by which the increase in clostridia observed in casein rich treated animals could be causing the

depressive phenotype observed in this group.

Exactly how milk caseins result in increased levels of C.histolyticum bacterial groups remains

unclear. However, a number of studies have indicated that protein casein levels do indeed effect

microbial composition in the gut (Zhu et al., 2015; An et al. 2014; Rist et al., 2012).

Furthermore Farshim et al., (2016) also found that C.histolyticum groups were significantly

higher in animals which were delayed weaned on PND25 following exposure to the FST

compared to animals weaned normally on PND21. Combined with the findings from this

current study, it is feasible to suggest that the increase of clostridia seen in non-weaned animals

168

was caused by exposure to milk caseins rather than a general milk effect beyond the normal

age of weaning.

The alterations observed in the gastrointestinal microbiota may also be a consequence of

changes in gut motility caused by casein and its break down products (see Section 1.3). It has

been previously shown that rats fed casein rich suspensions took longer in gastric emptying

and gastrointestinal transit time then rats given a whey protein suspension (Daniel et al.,1990),

these effects were found to be mediated by casomorphins including BCM-7 by acting on MOPr

found in the luminal side of the gut (Daniel et al., 1990). In addition BCM-7 has also been

shown to induce a strong release of gastrointestinal mucin secretion, again via MOPr mediated

pathway (Zoghbi et al., 2006). These gastrointestinal effects of casein and its breakdown

products could be altering gut conditions in casein rich treated animals to create favourable

conditions, which promote growth of C.histolyticum bacterial populations.

In order to investigate the metabolic impact of casein rich vs. casein free diets and the gut

microbiota on the host, we studied urinary metabotypes from these two treatment groups. Urine

is the preferred biofluid for the study of environmental impacts on host metabolism as it

provides a measure of the environmental inputs that are absorbed from the gut environment,

which can in turn interact with the host system and subsequently be excreted. From the

metabolic phenotyping analysis, casein rich rats were observed to excrete greater amounts of

betaine compared to casein free rats. Betaine is an endogenous metabolite of choline (Zeisel,

1981a). Increased excretion of betaine in the urine of casein rich animals indicates increased

choline metabolism induced by prolonged casein consumption. Choline is a micronutrient

found in food which is involved in methyl metabolism, transmembrane signalling and is needed

for the biosynthesis of the neurotransmitter acetylcholine which in turn plays an important role

in normal brain development (Wessler et al., 2008). Casein rich treated animals also excreted

higher amounts of hippurate, a mammalian-microbial co-metabolite originating from bacterial

169

action upon plant phenols to produce benzoate, which then becomes conjugated with glycine

in the liver to produce hippurate. Thus an increase in hippurate excretion as well as a decrease

in the excretion of glycine in casein rich urine indicates increased gut microbial activity.

Indeed, this is in line with the FISH analysis results, which showed an increase in bacteria

belonging to C.histolyticum genera in the gut of casein rich treated rats. Together these findings

indicate that consumption of milk caseins beyond the normal age of weaning may be acting via

the gut brain axis to induce the depressive phenotype observed.

In conclusion this study is the first to demonstrate that casein exposure beyond the normal age

of weaning results in depressive-like behaviour in rats. Based on previous work from our lab,

we hypothesised that this depressive phenotype would be driven by changes to the opioid

system as well as alterations to the oxytocinergic system. Whilst autoradiographic binding

revealed a significant down regulation of OTr in the amygdala of casein rich treated animals

which may correlate to the depressive phenotype observed, MOPr and DOPr data was less clear

cut. We observed region specific alteration to both MOPr and DOPr in both casein free and

casein rich treated rats. Indicating a general milk effect independent of casein on modulation

of the opioid system. However the deep layer of the somatosensory cortex appears to be a key

region in which milk protein break down products (casomorphins or soymorphins) exerts its

effects on the opioid system. These results further support the theory that early life

environmental factors can influence the normal development of the brain. Furthermore, the

findings in this chapter indicate that casein exposure alters the gut microbial and metabolic

profiles of the host and thus provides evidence that casein may be acting via the Gut-brain axis

to elicit its behavioural effects

170

Chapter 3

Role of the Gut Microbiota in Mediating

the Depressive Phenotype Induced by

Prolonged Casein Exposure

171

3.1 Introduction

Results obtained in Chapter 2 demonstrated that exposure to milk caseins beyond the normal

age of weaning resulted in a depressive-like effect on offspring behaviour and also resulted in

a shift in gut microbial composition, functionality and metabolic exchange with the host. These

findings suggest a key influence of milk casein exposure in early postnatal life on behaviour

and raise the potential for the gut microbiota to be involved in mediating this effect. With

increasing evidence for a bi-directional interplay between the gut microbiota and the host brain,

termed the gut-brain axis (Cryan and O’Mahony 2011), the role of the gut-microbiota in

mediating the depressive-like phenotype caused by prolonged casein exposure was explored.

Literature described in Chapter 1 Sections 1.5.3 to 1.5.5 provide strong evidence of a

bidirectional communication system enabling gut microbes to communicate with the CNS and

the CNS with the gut through direct and indirect pathways. These involve the endocrine (HPA

axis), immune system, autonomic nervous system and enteric nervous system forming the gut-

brain axis. Alterations in these gut-brain interactions are believed to be involved in the

pathogenesis of well-known brain-gut disorders such as IBS (see Kennedy et al., 2014) as well

as a possible mechanism in the pathogenesis of mood disorders such as depression (see Dinan

and Cryan, 2013; Naseribafrouei et al., 2014). Multiple lines of approach have been used to

interrogate the gut brain axis in animal models, including the use of GF animal models and

antibiotics (AB) to determine the effects of dysregulating the axis (see Table 1.7).

A study by Heijtz et al., 2011 showed that in comparison to conventional mice who were

growing in a SPF environment, GF mice under the same experimental conditions displayed

decreased anxiety like behaviour and increased 5-HT synthesis in the thalamus, confirming

that gut microbiota play a role in normal brain function and behaviour. However, GF animals

represent an extreme ‘microbial knock-out ‘model that lack morphological and physiological

172

characteristics associated with having an intact intestinal microbiota (see Luczynski et al.,

2016; Lundberg et al., 2016). Therefore, it is feasible to suggest this model does not provide a

meaningful comparison with ‘conventional’ animals. An alternative model to investigate the

role of the gut microbiota in health and disease is the use of antibiotics to inhibit gut flora

activity. Dosing conventionally-raised animals with antibiotics allows the gut microbial

populations and their activities to be temporarily suppressed whilst the conventional phenotype

is preserved. This animal model is morphologically, physiologically and immunologically

consistent with that of a conventional animal.

Suppression of the microbiota through oral administration of antibiotics was first described by

Van der Waaij et al., (1968). Since then a number of studies have used a combination of

antibiotics to deplete the gut microbiota and assess behaviour. For example a recent study by

Desbonnet et al., (2015) used a combination of antibiotics to treat mice from weaning age

onwards and assessed effects on behaviour. The results showed that antibiotic treatment not

only resulted in depletion of the gut microbiota but also caused a reduction in anxiety-like

behaviour as assessed by the light/dark box. These findings are in line with several other

studies, which have demonstrated antibiotic treatment results in reduced anxiety in rodent

models (Verdú et al., 2006; Bercik et al., 2011).

In this study, a combination of antibiotics was used to investigate the role of the gut microbiota

in mediating the effect of prolonged milk casein exposure on depressive-like behaviour. A

variety of antibiotics, including bacitracin, kanamycin, neomycin, streptomycin and penicillin

have been used in the past, either alone or in combination to effectively decontaminate the

digestive tract (Verdú et al., 2006). Although the exact microbial composition of antibiotic

(AB) treated animals cannot be controlled, different antibiotic dosing regiments can be used to

selectively target different groups of microbiota. Here, penicillin and streptomycin sulfate were

chosen due to their complementary bacteriocidal qualities. Penicillin G, is most effective

173

against Gram-positive bacteria (Rammelkamp et al.,1943), while streptomycin is effective

against Gram-negative bacteria (Adcock et al., 1946). Both antibiotics are also poorly absorbed

and so should not have a direct impact on host biochemistry.

This research chapter will determine the microbial and metabolic impact of antibiotic treatment

and casein rich or casein free milk exposure in rats. Results from this study will provide an

insight into the role of the gut microbiota in mediating behavioural changes induced by changes

in diet in early postnatal life.

Hypothesis:

The depressive phenotype observed in animals exposed to casein rich milk beyond the

standard age of weaning is mediated by the gut microbiota. As such, suppression of the

gut microbiota through antibiotic treatment will inhibit the depressive phenotype

observed after prolonged exposure to casein rich milk.

Aim:

Investigate the role of the gut microbiome in mediating the depressive phenotype

induced by prolonged casein exposure.

Objectives:

Use the streptomycin/penicillin treated rat model to explore the role of gut bacteria in

mediating the depressive phenotype induced by prolonged casein exposure.

Assess depression using the forced swim test.

Use fluorescence in-situ hybridisation (FISH) analysis to enumerate the intestinal

bacterial populations of study animals following AB treatment.

Characterise the urinary metabolic profiles of the study animals using 1H nuclear

magnetic resonance (NMR) spectroscopy to identify biochemical modulations induced

by bacterial attenuation, casein consumption and combinations of the two.

174

3.2 Methodology


Male, Wistar Albino rats were purchased from Charles River UK Limited, Kent. Six groups of

animals each containing eight male pups plus their cross-fostered mother arrived at the

Behavioural Research Facility (BRF) in St. Georges University of London on PND7, where

day of birth is PND0. These animals were maintained in litters of 8 pups with their mothers at

all times. Pups were left to acclimatize to their new environment for eleven days, after which

point they were handled and weighed daily by the experimenter. The standard day of weaning

in laboratory rats is on PND21, a process where the mother is simply removed from the cage

(Cramer et al., 1990). Antibiotic-treated and control groups were all weaned on PND21. All

rats were placed in a 12 hour light-dark cycle (lights on 6 AM -6 PM) with free access to a

standard rat chow diet (purchased from SDS, Kent U.K) and tap water and kept at a constant

temperature of 22°C ± 1 °C, humidity of 45-55%. All experimental procedures were conducted

in accordance with the U.K Animal Scientific Procedures Act (1986).

3.2.2 Antibiotic Treatment

3.2.2.1 Control Animals


All three groups were provided with free access to rat chow and water throughout the duration

of the study. The control group was weaned by having their mothers removed at 10:00 am on

PND21. The casein free group were also weaned on PND21 and subsequently provided with

casein free milk formula from PND21 until the end of behavioural testing on PND26. The

casein rich group were weaned on PND21 and provided with casein rich milk formula from

PND21 until PND26.

175

3.2.2.2. Antibiotic Treated Animals

Three experimental groups were generated with each group containing 8 pups caged together.

All three groups were provided with free access to rat chow and water throughout the duration

of the study, and were all weaned by having their mothers removed at 10.00am on PND21. On

PND17 4mg/ml Streptomycin and 2mg/ml Penicillin G (Purchased from Sigma Aldrich) were

added to 250ml of drinking water and provided to all three groups daily until PND21. On

PND21, the control antibiotic (control AB) group continued to be provided with bottled water

containing the antibiotic solution daily until the end of behavioural testing on PND26. The

casein free antibiotic group (casein free AB) were provided with casein free milk formula

mixed with the AB solution daily from PND21 until the end of behavioural testing on PND26.

Similarly, the casein rich antibiotic group (casein rich AB) were provided with casein rich milk

formula from PND21 until PND26, which was treated with antibiotic solution. As the AB

solution was now being provided in the milk diets for both casein rich AB and casein free AB,

both these groups were provided with normal tap water free of AB solution from PND21

onwards. The antibiotic treatment was therefore effectively switched from the drinking water

to the milk diets as pups provided with milk diets show a preference for milk over drinking

water. Thus the switch would ensure the animals continued to receive their dose of antibiotics.

AB free

n= 8 each group

Control AB

Casein free AB

Casein rich AB

Control

Casein free

Casein rich

AB treated

n= 8 each group

Figure 3.11: Diagram showing experimental groups in the antibiotic study. Antibiotic treated

animals were dosed with 4mg/ml Streptomycin and 2mg/ml Penicillin G drinking water/milk for a total

of 10 days.

176

3.2.3 Milk Diets

The groups provided with casein free or casein rich milk with or without AB solution had milk

freshly made up twice a day, once between 09:00-10:00 and again at 17:00-18:00. Milk diets

were made as described in Section 2.2.2, for the formulations of casein free and casein rich

milk diets please see Table 1.2.

3.2.4 Body Weights, Milk and Water Consumption

To monitor the general health of the animals, body weights were recorded daily from PND14

until the end of behavioural testing on PND26 and also on the day of sacrifice. To monitor

antibiotic treatment, water consumption was measured daily from PND17 until PND26.

Similarly, milk consumption was measured daily from PND21 until PND26.

3.2.5 Behavioural Testing- Forced Swim Test

Behavioural testing was performed between 10am and 2pm when animals reached PND25

using the forced swim test, which is run over two consecutive days (PND25 & PND26). The

FST was run as described in Section 2.2.3.

3.2.6 Collection of Urine

Following the completion of behavioural testing on PND25 and PND26, animals were placed

into individual metabolic cages for 24 hours with access to standard rat chow and water. On

PND27 urine was collected from the cages for metabolic analysis (see Section 3.2.7).

177


Animals were killed by cervical dislocation on PND27 after acute exposure to CO2. Intact

brains were dissected and rapidly frozen in isopentane at -20°C for 30 seconds and then stored

at -80°C for quantitative receptor autoradiographic experiments. The carcasses were then

transferred into a fume hood content from the caecum was collected and weighed (30-200 mg)

into sterile 2 ml Eppendorf tubes and kept over ice and then stored in -80°C for use with

fluorescence in situ hybridization experiments (see Section 3.2.9).

3.2.8 Metabolomics of Urine Samples

As described in Section 2.2.6

3.2.8.1 Multivariate Data Analysis

As described in Section 2.2.6.4



3.2.9 Fluorescence in situ Hybridization (FISH)

FISH analysis was carried out on caecal content collected from control, casein free, casein rich,

control antibiotic, casein free antibiotic and casein rich antibiotic groups behaviourally tested

on PND25 & PND26 and sacrificed on PND27, for the enumeration of total bacteria as

described in Section 2.2.6. Prior to the dehydration stage (Section 2.2.9.4), EUB MIX probes

were added and samples collected from non- antibiotic treated animals were diluted 1:300 using

178

sterile PBS/SDS diluent. However samples collected from antibiotic treated groups were

diluted 1:20 due to a low number of bacteria.

3.2.10 Data Analysis

The weight change, water consumption, milk consumption and behavioural data for the FST

were analysed using two-way ANOVA for factors “diet” and “AB treatment’ in GraphPad

Prism 7. Where ANOVA revealed a significant difference an LSD post-hoc analysis was

carried out. All data were expressed as means ± SEM, n = 8 per group. A p value of less than

0.05 was accepted as significantly different. All statistical analyses were performed on

GraphPad Prism 7.

Postnatal Day

7 14 17 21 25 26 27

Antibiotic treatment begins in

drinking water

Mothers removed,

casein free or casein rich milk

provided

Antibiotic treatment

switched to milk

Forced swim test


cages for 24 hours



Figure 3.12: Experimental timeline showing key points in the methodology for antibiotic

dosing study.

179

3.3 Results

3.3.1 Water Consumption

Male Wistar Albino rats were provided with free access to bottled water treated with or without

antibiotics from PND17 until the end of behavioural testing on PND26. Water consumption

was measured daily for this duration for all treatment groups. Two-way ANOVA for factors

‘diet’ and ‘AB treatment‘ showed a significant ‘diet’ only effect on water consumption [F (2,48)

= 3.63; p < 0.05]. Post-hoc analysis showed animals provided with casein rich milk drank

significantly less water than control treatment group (p < 0.05), furthermore animals provided

with casein rich milk treated with AB drank significantly less water than both control AB and

casein free AB treatment groups (p < 0.05) (Figure 3.3).

AB-fre

e

AB-tre

ated

0

50

100

150

200

Control

Casein free

Casein rich

Mean

Daily W

ate

r

Co

nsu

mp

tio

n p

er

Cag

e (

g)

* *

Figure 13.3: Water consumption in rats provided with free access to bottled water treated with

or without penicillin/streptomycin from PND17 until PND26.

Mean daily water consumption in grams for control, casein free, casein rich, control antibiotic, casein

free antibiotic and casein rich antibiotic groups from PND17 until PND26. Data are expressed as mean

± SEM (n=9 readings). Animals provided with casein rich milk drank significantly less water

than control treatment group, animals provided with casein rich milk treated with AB drank

significantly less water than both control AB and casein free AB treatment groups (*p<0.05

two-way ANOVA followed by LSD post-hoc test).

180

3.3.2 Milk Consumption

Male Wistar Albino rats were provided with casein rich or casein free milk treated with or

without antibiotics from PND21 until the end of behavioural testing on PND26 as well as being

provided with free access to bottled tap water. Milk consumption was measured daily for this

duration. Two-way ANOVA for factors ‘diet’ and AB treatment showed a significant effect

of both ‘diet’ [F (1,16) = 16.03; p < 0.001] and ‘AB treatment’ [F (1,16) = 4.742; p< 0.01] on

milk consumption. Post-hoc analysis revealed animals provided with casein free AB milk drank

significantly less milk than animals provided with casein free milk (p < 0.05), and animals

provided with casein rich AB milk (p < 0.001) (Figure 3.4).

AB-fre

e

AB-tre

ated

0

50

100

150

200

Mean

Daily M

ilk

co

nsu

mp

tio

n p

er

cag

e (

g)

Casein free

Casein rich

* ***

Figure 3.14: Milk consumption in rats provided with casein free or casein rich milk treated with

or without penicillin/streptomycin from PND21 until PND26.

Mean daily milk consumption in grams for control, casein free, casein rich, control antibiotic, casein

free antibiotic and casein rich antibiotic treated groups from PND21 until PND26. Data are expressed

as mean ± SEM (n=4 readings). Animals provided with casein free AB milk drank significantly

less milk than animals provided with casein free milk and animals provided with casein rich

AB milk (*p<0.05 , ***p<0.001. Two-way ANOVA followed by LSD post-hoc test)

181

3.3.3 Weight Change

Male Wistar Albino rats were weighed daily from PND14 until the end of behavioural testing on

PND26. Antibiotic treated groups (see Figure 1.3) were provided with water containing

penicillin/streptomycin from PND17 until PND21. On PND21 appropriate groups were provided with

casein free or casein rich milk treated with or without penicillin/streptomycin while control groups

continued with or without antibiotic treatment in bottled water. Two-way ANOVA revealed no

significant effect of diet or AB treatment on weight gain (Figure 3.5).

AB fr

ee

AB-tre

ated

0

20

40

60

80

Weig

ht

ch

an

ge (

g)

Control

casein free

casein rich

Figure 3.15: Effects of casein free and casein rich diets with or without penicillin/streptomycin

treatment on body weight change from PND14 until PND26.

Male Wistar albino rats were weighed daily from PND14 until the end of behavioural testing on PND26.

Graph shows weight change in grams for control, casein free, casein rich, control AB , casein free AB

and casein rich AB treatment groups. Data are expressed as mean ± SEM (n=8 per group). No difference

in weight change was observed between all treatment groups (two-way ANOVA followed by LSD post-

hoc analysis).

182

3.3.4 Behaviour

The FST was used to assess depressive-like behaviour on PND25 and PND26 in all the study

animals. The total immobility time was scored from the ‘test’ session on PND26. The

quantification from the ‘test’ times were then used in the assessment of depressive-like

behaviour as described by Cryan et al., (2005).

Two-way ANOVA for factors ‘diet’ and ‘AB treatment’ showed a significant interaction

between both factors [F (2,29) = 14.76; p < 0.001], a significant ‘diet’ effect [F (2,29) = 4.395; p

< 0.05] and a significant ‘AB treatment’ effect [F (1,29) = 23.27; p < 0.001]. Post-hoc analysis

showed rats fed a casein rich diet displayed an increase in depressive-like behaviour as

indicated by an increase in immobility time compared to casein free treated animals (p = 0.05).

AB treatment caused a significant decrease in immobility times in both casein rich AB

compared to conventional casein rich animals (p < 0.001) and control AB compared to

conventional control animals (p = 0.01) but not in the casein free AB treatment group.

Immobility exhibited by the casein free AB animals was significantly higher than both casein

rich AB and control AB (p < 0.001) (Figure 3.6).

183

AB fr

ee

AB tr

eate

d 0

100

200

300

400

Imm

ob

ilit

y t

ime (

s)

Control

Casein free

Casein rich

**

*********

*

Figure 3.16: Immobility times exhibited by rats fed no milk, casein free or casein rich milk

treated with or without streptomycin/penicillin from PND17 until PND 26 in the forced swim

test.

Male Wistar albino rats belonging to control, casein free, casein rich, control antibiotic, casein free

antibiotic and casein rich antibiotic treated groups were subjected to behavioural testing using the FST

on PND25 &26. Figure shows immobility time during the 6-minute test session on PND26. Data are

expressed as mean scores (± SEM) (n=8) obtained from manual video analysis by one independent

observer. Casein rich milk caused an increase in immobility time compared to casein free milk. AB

treatment caused a significant decrease in immobility times in both casein rich AB and control AB but

not casein free AB animals compared to conventional casein rich animals and conventional control

animals. Immobility exhibited by the casein free AB animals was significantly higher than both casein

rich AB and control AB. Two-way ANOVA followed by LSD post-hoc test was used and a p value equal

to or less than 0.05 was accepted as significantly different. *p =0.05, ** p ˂ 0.01, *** p ˂ 0.001.

184

3.3.5 Assessment of Intestinal Microbiota (FISH)

3.3.5.1 Assessment of Total Bacteria

Male Wistar albino rats were provided with no milk, casein free or casein rich milk from

PND21 until PND26. Antibiotic treated animals were provided with water containing 4mg/mL

streptomycin and 2mg/ml Penicillin from PND17 until PND21, after which AB treatment was

switched into their milk diets. Two-way ANOVA for the factors ‘diet’ and ‘AB treatment’ on

the Log10 of total bacteria counts showed a significant ‘AB treatment’ effect only following 10

days of AB treatment on total bacteria [F (1, 38) = 21317; p < 0.001]. Post-hoc analysis showed

antibiotic treatment significantly reduced the average number of total bacteria (determined by

EBU mix probes) contained within the caecal content in all AB treated groups compared to

non AB treated animals (p < 0.001 ) (Figure 3.7).

185

No A

BAB

No A

BAB

No A

BAB

7

8

9

10

Lo

g10 C

ou

nt

of

To

tal B

acte

ria

Control

Casein Free

Casein Rich

*** *** ***

Total Bacteria (EUB338)

Figure 3.7: Effect of 10 days of streptomycin/penicillin treatment from PND17 on the composition

of total bacteria within the caecum of control, casein free and casein rich treated animals.

Male Wistar albino rats were provided with no milk, casein free milk or casein rich milk from PND21

until PND26. Antibiotic treated animals were provided with water containing 4mg/mL streptomycin

and 2mg/ml Penicillin G from PND17 until PND21. On PND21 the control AB group continued to be

provided with antibiotic solution in the drinking water until PND26 whereas casein free AB and casein

rich AB groups were provided with antibiotic solution in the milk rather than in drinking water until

PND26. All groups were subjected to behavioral testing on PND25 & PND26. A mixture of EUB 338

I, II and III probe were used for the enumeration of total bacteria in the caecum of study animals. All

data expressed as mean log10 cells (g weight of intestinal content) -1 , S.E.M (n=8).. Two-way ANOVA

followed by Bonferroni post--hoc test, a p value of less than 0.05 was accepted as significantly different,

*** p˂0.001.

186

3.3.6 Metabonomic Analysis of Urine Samples

3.3.6.1 PCA of Urine Samples: All data

To gain an overview of how 10 days of AB treatment modulated the biochemical status of the

animal, PCA analysis was used to identify metabolic variation in urinary profiles. The

percentage of variance described by each PC (R2 ) is shown for each axis. Individual animals

are color-coded in the scores plot to indicate their class membership. Figure 3.8 A shows two

groups separating along the first PC. Animals receiving AB separated in the positive direction

of PC1 (with the exception of a casein-rich AB animal) and AB-free treated animals separated

in the negative direction of PC1 (with the exception of a control animal). To identify the

metabolites driving this separation, the loadings for PC1 were investigated. Figure 3.8 B shows

that AB treatment caused an increase in the excretion of energy metabolites citrate and glucose,

choline metabolism metabolites betaine and TMAO and the amino acid ascorbate, while

reducing the excretion of gut microbial metabolites hippurate and 4-HPPA, the amino acid

taurine, choline metabolism metabolite DMA, and also reduced the excretion of acetate.

187

A

B

Figure 3.8: Effect of antibiotic treatment from PND17 until PND26 on host urinary metabolic

profile.

PCA model (R2 = 30.1%.) constructed on the urinary metabolic profiles of all study animals measured

by 1H NMR spectroscopy. A) PCA scores plot showing separation along PC1 between AB-treated and

AB-free groups. B) Corresponding loadings plot showing the metabolic features contributing to the

separation observed in the PC1 scores plot. n= 3-4 per group Abbreviations: DMA: dimethylamine.

188

3.3.6.2 PCA of Urine Samples: Metabolic impact of casein exposure

In order to examine the effect of casein exposure on metabolic profiles, pairwise comparisons

were made between control, casein rich and casein free groups. Variation between urinary

metabolic profiles from casein free and casein rich animals was observed along PC1 (Figure

3.9 A). The corresponding loading plots shows that the excretion of citrate, succinate and

DMA was higher in the rats receiving casein free milk compared to those consuming casein-

rich milk (Figure 3.9 B) while excretion of creatinine, taurine, alanine and hippurate were

lower compared to casein rich treated animals, with citrate being the main driving factor.

PCA modelling of control versus casein rich treated animals shows clear separation along PC3

(Figure 3.10 A). The corresponding loading plots identified that the excretion of betaine,

TMAO taurine, and citrate was higher in control animals compared to rats consuming casein-

rich milk (Figure 3.10 B).

189

A

B

Figure 3.9: Effect of casein rich or casein free milk diets from PND21 until PND26 on host urinary

metabolic profile.

PCA model (R2 = 71.6%) constructed on the urinary metabolic profiles casein rich and casein free

animals measured by 1H NMR spectroscopy. A) PCA scores plot showing separation along PC3

between casein rich and casein free groups. B) Corresponding loadings plot showing the metabolic

features contributing to the separation observed in the PC3 scores plot. P shows the loadings vector

values. n =3-4 per group. Abbreviations: DMA; dimethylamine; TMAO: trimethylamine-N-oxide

190

A

B

Figure 3.17: Effect of casein rich milk or control diets from PND21 until PND26 on host urinary

metabolic profile.

PCA model (R2 = 84.5%) constructed on the urinary metabolic profiles of casein rich and control

animals measured by 1H NMR spectroscopy. A) PCA scores plot showing separation along PC3

between casein rich and control groups B) Corresponding loadings plot showing the metabolic features

contributing to the separation observed in the PC3 scores plot. P shows the loadings vector values. n=3-

4 per group Abbreviations: DMA; dimethylamine; TMAO: trimethylamine-N-oxide

191

3.3.6.3 PCA of Urine Samples: Metabolic impact of microbial suppression

To assess the metabolic impact of AB treatment, the urinary metabolic profiles of control rats

were compared with control rats that received AB using PCA analysis (Figure 3.11 A).

Separation between the two groups was observed along PC3 and the corresponding loadings

plot for PC3 identified citrate, ascorbate, betaine, alanine and taurine to be reduced with

antibiotics while DMA and 2-oxoglutarate were increased (Figure 3.11 B) with taurine and

citrate being the main driving factors.

A

192

B


profiles of control and control AB treated rats.

PCA model (R2 = 93.7%) constructed on the urinary metabolic profiles of control and control with

antibiotic animals measured by 1H NMR spectroscopy. A) PCA scores plot showing separation along

PC3 between control and control with antibiotic treatment groups. B) Corresponding PC3 loadings plot

showing the metabolic features contributing to the separation observed in the PC3 scores. n=3 per group

Abbreviations: AB: antibiotic; DMA: dimethylamine.

193

3.3.6.4 PCA of Urine Samples: Metabolic Impact of Microbial

Suppression in Casein Rich Treated Animals

AB treatment was found to ameliorate the behavioural alterations induced by casein rich milk

exposure.To assess whether AB treatment modified the metabolic changes associated with

casein-rich milk intake, the urinary metabolic profiles of the rats receiving casein rich milk

were compared with those receiving casein rich milk plus AB (Figure 3.12 A). Differences

were observed between the two groups with the profiles separating along PC1. Casein rich milk

plus antibiotics cause an increase in taurine, betaine, alanine, ascorbate, citrate and succinate

while causing a decrease in DMA excretion as well as acetate compared to casein rich animals.

(Figure 3.12 B)

To assess the metabolic impact of both casein and AB treatment combined, the urinary

metabolic profiles of rats consuming casein rich milk with AB were compared with control rats

using PCA analysis (Figure 3.13 A). Differences were observed between the two groups with

the profiles separating along PC3 with taurine and citrate excretion increasing as a result of

casein and AB treatment, while hippurate, alanine, ascorbate, DMA, arginine, 2-OG and

acetate are decreased (Figure 3.13 B).

194

A


profiles of in casein rich and casein rich AB treated rats.

PCA model (R2 = 73%) constructed on the urinary metabolic profiles of casein rich and casein rich with

antibiotic animals measured by 1H NMR spectroscopy. A) PCA scores plot showing separation along

PC1 between casein rich and casein rich AB treatment groups. B) Corresponding PC1 loadings plot

showing the metabolic features contributing to the separation observed in the PC1 scores. P shows the

loadings vector values. n=3-4 per group Abbreviations: AB: antibiotic; DMA: dimethylamine.

B

195

B

A


profiles of control and casein rich AB treated rats.

PCA model (R2 = 93.9%) constructed on the urinary metabolic profiles of casein rich with AB and

control animals measured by 1H NMR spectroscopy. A) PCA scores plot showing separation along PC4

between casein rich with AB and control treatment groups. B) Corresponding PC4 loadings plot


loadings vector values. Abbreviations: AB: antibiotic; DMA: dimethylamine; 2-OG; 2-oxoglutarate

196

3.3.6.5 PCA of Urine Samples: Metabolic Changes in Casein Free

Animals Following Microbial Suppression

The metabolic impact of AB treatment in animals provided with casein free milk was assessed

through a pair-wise PCA model comparing urinary profiles from casein free rats and casein

free AB treated rats. Figure 3.14 A shows the two groups separate along PC1. Excretion of

glucose, DMA, DMG, TMAO and taurine were increased in casein free AB animals, while

citrate, succinate, acetate, Hippurate and 3-HPPA were reduced (Figure 3.14 B).

To assess the metabolic impact of both casein free milk and AB treatment combined the urinary

metabolic profiles of rats consuming casein free milk with AB were compared with control rats

using PCA analysis (Figure 3.15 A). Differences were observed between the two groups with

the profiles separating along PC3. Unlike AB treatment in casein rich animals, AB treatment

in casein free animals increased the excretion of TMAO, leucine and ascorbate, while causing

a decrease in the excretion of creatine. Consistent with AB treatment in casein rich animals,

AB treatment in casein free animals caused an increase in the excretion of taurine while DMA

excretion was reduced (Figure 3.15 B).

Table 3.1 provides a summary of 1H-NMR chemical shifts for the metabolites identified in

urine samples from all six-treatment groups.

197

A

B


metabolic profiles of animals provided with casein free and casein free AB milk.

PCA model (R2 = 60.2%) constructed on the urinary metabolic profiles of casein free and casein free

with AB animals measured by 1H NMR spectroscopy. A) PCA scores plot showing separation along

PC1 between casein rich with AB and control treatment groups. B) Corresponding PC1 loadings plot


loadings vector values. n= 2-3 per group Abbreviations: AB: antibiotic; 3-HPPA: 3-

hydrophenylpropionate; DMA: dimethylamine; DMG: dimethylglycine; TMAO: trimethylamine-N-

oxide

198

A

B


metabolic profiles of animals provided with casein free AB milk and control treatment group.

PCA model (R2 = 100%) constructed on the urinary metabolic profiles of casein free with AB and

control animals measured by 1H NMR spectroscopy. A) PCA scores plot showing separation along PC3

between casein free with AB and control treatment groups. B) Corresponding PC3 loadings plot


loadings vector values. n= 2-3 per group Abbreviations: AB: antibiotic; DMA: dimethylamine; TMAO:

trimethylamine-N-oxide

199

Table 3.13: Summary of 1H-NMR chemical shifts for metabolites associated with alterations following antibiotic treatment and casein free or casein

rich milk treatment. (+) indicates higher excretion (positive correlation) in the treatment group (identified according to the corresponding colour); ( - ) , lower

excretion ( negative correlation) in the relevant control groups (identified in black colour) and ‘-‘ represents absent peak or no significant difference found

between the corresponding groups. Metabolites were assigned using the literature and available databases. The metabolites identified have been separated

according to metabolic pathways they are mainly involved in. Abbreviations: 3-HPPA: 3-hydrophenylpropionate; TMAO: trimethylamine-N-oxide DMA;

dimethylamine; DMG: dimethylglycine; AB: antibiotic; CR: casein rich; CF: casein free; CTRL: control.

Metabolite

Chemical Shift AB CR CTRL CTRLAB CRAB CRAB CFAB CFAB

(Multiplicity) vs

No AB

vs

CF

vs

CR

vs

CTRL

vs

CR

vs

CTRL

vs

CF

vs

CTRL

Microbial Derived

3-HPPA 2.44 (t), 3.02(t) ( - ) − − − − − ( - ) −

Acetate 1.93 (s) ( - ) − − − ( - ) ( - ) − −

Hippurate 7.8 (d), 7.6 (t), 7.5 (d) ( - ) ( + ) − − − ( - ) ( - ) −

Choline Metabolism

Betaine 3.91 (s) 3.27 (s) ( + ) − ( + ) ( - ) ( + ) − ( - ) −

TMAO 3.2 (s) ( + ) − ( + ) − − − ( + ) ( + )

DMA 2.7 (s) ( - ) ( - ) ( - ) ( + ) ( - ) ( - ) ( + ) ( - )

DMG 2.9 (s) − − − − − ( - ) ( + ) −

Energy Metabolism

2-oxoglutarate 2.44 (t) − − − ( + ) − ( - ) − −

Citrate 2.71 (d), 2.56 (d) ( + ) ( - ) ( + ) ( - ) ( + ) ( + ) ( - ) ( + )

Glucose 5.2 (d) ( + ) − − − − − ( + ) −

Succinate 2.41 (s) − ( - ) − − ( + ) − ( - ) ( - )

Other

Alanine 3.77 (q) − ( + ) ( - ) ( - ) ( + ) ( - ) − −

Arginine 3.2 (t) − − − − − ( - ) − −

Creatinine 3.0 (s) − ( + ) − ( - ) − − − ( - )

Ascorbate 4.5 (d) ( + ) − − ( - ) ( + ) ( - ) ( - ) ( + )

Leuicine 3.72 (m) − − − − − − − ( + )

Taurine 3.43 (t) 3.28 (t) ( - ) ( + ) ( + ) ( - ) ( + ) ( + ) ( + ) ( + )

200

3.4 Discussion

The aim of this study was to investigate if the gut microbiota plays a role in mediating the

depressive-like phenotype observed in animals provided with casein rich milk beyond the normal

age of weaning. To achieve this, the abundance and activity of the gut microbiota was temporarily

suppressed for 10 days from PND17 through the administration of penicillin and streptomycin in

the drinking water/ milk diets. To assess the diet-microbiota-biochemical host cascade and its

impact on behavioural development, the bacterial load was measured in the intestinal contents

using FISH analysis, biochemical alterations associated with antibiotic and casein exposure was

characterized by 1H NMR spectroscopic analysis of urine samples, and finally depressive-like

behaviour was determined using the FST. The effect of antibiotic treatment alone, independent of

casein was also investigated.

Following 10 days of antibiotic treatment the total number of bacteria present in the gut was

markedly reduced with accompanying changes in the excretion of numerous metabolites including

those derived from gut microbial metabolism. We also observed a profound antibiotic effect on

behaviour in both control and casein rich treated rats but not in casein free. These results provide

evidence for a role of the gut microbiota in mediating depressive-like behaviour.

A rat model representing a gut free of bacteria was established by dosing rat pups belonging to the

AB treatment groups (see Figure 3.1) with penicillin and streptomycin for a total of 10 days

starting on PND17. Animals belonging to the AB treatment groups were all weaned on PND21

after which point one group received no milk (control AB), casein rich milk (casein rich AB) or

201

casein free milk (casein free AB). Matched controls were established by having the same exact

three treatment groups but with no AB administered.

Water consumption data showed a significant effect of diet but not AB treatment on water

consumption. Rats provided with casein rich milk with or without antibiotics consumed

significantly less water compared to control, control AB as well as casein free AB treatment groups

(see Figure 3.3), indicating that rats preferred to drink casein rich milk over water, whether it

contained antibiotics or not. Animals provided with casein free milk treated with AB drank

significantly more water than animals provided with casein rich AB milk, indicating these animals

preferred to drink water for hydration rather than the casein free milk treated with antibiotics.

Indeed casein free AB animals consumed similar levels of water as the control group who only

had access to bottled water. Milk consumption data confirms these findings and shows a significant

AB and diet effect on milk consumption in that casein free AB animals consumed significantly

less milk compared to all other groups provided with milk diets (see Figure 3.4). Furthermore,

milk consumption data shows that animals provided with casein free milk consumed less milk than

the casein rich group although this was not statistically significant; it indicates that in general,

animals showed less of a preference towards casein free milk with or without AB. These results

are in line with findings from the previous results chapter (see Section 2.3.1) as well as findings

from Goody & Kitchen, (2001) which both showed rat pups provided with casein rich milk

consumed significantly greater amounts compared to animals provided with casein free milk.

Interestingly, unlike the significant differences in weight gain observed in results Chapter 2

following casein rich milk treatment in comparison to casein free treated animals, no significant

weight differences were observed between all 6-treatment groups in the current study, indicating

that the varying milk diets with or without antibiotics had no effect on weight gain (see Figure

202

3.5). It should be noted however that the casein free AB free treatment group displayed lower body

weights compared to all other treatment groups but this was not significantly different. Similarly,

in results Chapter 2 it was the casein free treatment group which had the lowest weight gain,

further indicating a soy protein effect on milk consumption and subsequent weight gain as

discussed in detail in Section 2.4.

The results obtained from FISH analysis showed that following 10 days of AB administration the

average number of total bacteria contained within the caecum was reduced from 10.1 log10 cells.g

in AB-free rats, to 6.7 log10 cell/g in AB treated groups. This is in line with findings from Swann

et al., (2011) who dosed 10 week old male rats for 8 days with the same daily dose of

penicillin/streptomycin and found the average total number of bacterial cells contained within

intestinal contents was reduced from 10.7 log10 cells/g in control animals to 8.4 log10 cells/g in AB

treated animals. These findings show that AB regime used in this study is capable of successfully

suppressing the gut microbial profile of rat pups.

The results obtained from the FST showed a significant diet and AB effect on immobility and more

importantly a significant diet and AB interaction effect. Animals provided with casein rich milk

without AB treatment, exhibited depressive-like behaviour as evidenced by a significant increase

in immobility time in the FST compared to animals provided with casein free milk (no AB) (see

Figure 3.6). This confirms findings from the previous results chapter (see Section 2.3.2) and

further confirm that casein exposure beyond the age of weaning results in a depressive phenotype

in rats. Interestingly, we observed a profound reduction in immobility time in AB treated control

and AB treated casein rich animals demonstrating an ‘antidepressive-like’ effect of AB treatment

at least in these two groups of animals. This is in line with behavioural data from a number of

studies including Bercik et al., (2011) who show that following 7 days of AB treatment, mice

203

exhibited more exploratory and less apprehensive like behaviour than controls in the light/dark

box indicating an anxiolytic-like effect of AB treatment. They also showed that AB treated mice

had increased levels of hippocampal BDNF levels (Bercik et al., 2011). An increase in BDNF

levels in the hippocampus has been shown to be associated with anxiolytic and antidepressant

behaviour (Deltheil et al., 2008) further supporting the anti-depressant like effect caused by AB

treatment in control and casein rich animals. Furthermore, Desbonnet et al., (2015) showed that

mice who were treated with a combination of AB from weaning age onwards had a depleted gut

microbiome and displayed reduced anxiety further supporting the finding that AB treatment in

early postnatal life results in alterations to behaviour. Thus, the antidepressive-like effect of AB

treatment reported in our study is not likely to be mediated by any other mechanism aside from its

effect as an antimicrobial agent, as similar anxiolytic changes are reported in GF mice in which

bacteria are absent from birth (Heijtz et al., 2011; Neufeld et al., 2011). Taken together, these

findings further support the link between gut microbial suppression using antibiotics with

alterations to mood.

In support of the antibiotic induced antidepressive-like behaviour reported here, the antibiotic

minocyline has been show to illicit antidepressive-like effects via the immune system. Pro-

inflammatory agents such as Interleukin-1 beta (IL-1β) and Tumour Necrosis Factor alpha (TNF-

α) are believed to be amongst the most important players in the pathogenesis of depression and

anxiety related behaviours (see Singhal et al., 2017). Indeed Majidi et al., (2016) showed that

following neonatal injection with Eschichia coli, mice went on to develop anxiety and depression

like behaviour in adulthood accompanied by increased expression of glial cell markers, IL-1β and

TNF-α in the hippocampus. Treatment with minocycline was able to attenuate Eschichia coli

induced expression of pro-inflammatory cytokines and reverse depressive-like behaviour by

204

reducing immobility time in the FST. These findings were replicated in a similar study by Zheng

et al., (2015) who also showed minocycline treatment resulted in attenuation of LPS-induced

increase of pro-inflammatory cytokines and depressive-like behaviour. In line with this, it has also

been shown that minocycline attenuates LPS-induced neuroinflammation, sickness behaviour and

anhedonia in adult mice through reduced microglial activation and decreased mRNA levels of IL-

1β in the hippocampus (Henry et al., 2008). Collectively, these studies suggest antibiotics may be

acting directly via the immune system to elicit their anti-depressive like effects. Indeed in this

present study, 10 days of AB treatment was shown to result in a reduction in total bacteria

accompanied with a reduction in gut microbial metabolites as shown by metabolomic analysis,

which could indicate a reduction in immune system activation and in turn explain the protective

effect AB treatment displayed against the development of depressive like phenotype observed in

casein rich treated animals. In future it would be useful to measure inflammatory markers in control

and casein rich treated animals with or without AB to determine whether or not the gut microbiota

is interacting with the immune system in mediating the behavioural changes observed.

As the results from the FST showed that immobility time was decreased in the control rats treated

with AB compared to control rats that did not receive AB, it suggests the antibiotic treatment,

independent of casein, reduced immobility time in the FST. It should also be noted that in general

AB treated animals were a lot more active and harder to handle than control animals and as a

personal observation from the experimenter, even displaying signs of aggression. These

observations are in line with Leclercq et al., (2017) who reported decreased anxiety like behaviour

and increased aggression as well as persistent alterations to the gut microbiota in mice following

penicillin treatment during the perinatal period.

205

One of the limitations of this study is the lack of measurement of locomotor activity in the groups

of animals exposed to the FST to exclude the possibility that the changes in immobility observed

between the groups merely represent changes in locomotion rather than mood. Nonetheless, this

is unlikely to be the case as Leclercq et al., (2017) measured locomotion using the open field test

following early life penicillin treatment in mice and reported no difference in locomotion. This is

also in line with findings from Molina-Hernández et al., (2008) who also showed minocycline

treatment did not induce hyperlocomtion and in some cases even decreased motor activity. These

findings support the hypothesis that the effects of AB treatment on the FST results are unlikely to

be driven by increased motor activity.

Interestingly, animals provided with casein free milk treated with AB did not display the same

drastic decrease in immobility time in the FST as the control and casein rich AB treated animals.

To the contrary, they displayed immobility times in the same range as AB-free animals. This

indicates that AB treatment had no effect on behaviour in casein free treated animals. It is possible

the casein free AB group did not receive the same dose of AB as control and casein rich AB

treatment groups as milk consumption data shows casein free AB treated animals consumed

significantly less milk than all the other groups provided with milk diets. As AB treatment was

switched into the milk diets from PND17, this could indeed suggest these animals received a lower

dose of AB and therefore did not display the same ‘anti-depressive’ like behaviour observed

following AB treatment. In line with this hypothesis, the study by Bercik et al., (2011) discussed

above showed that smaller doses of AB treatment failed to influence anxiety like behaviour in

mice suggesting that the changes in bacterial composition caused by AB treatment, are responsible

for the changes in anxiolytic behaviour observed in their study. Although FISH analysis data

showed that the gut microbiota of casein free AB animals was indeed suppressed to the same extent

206

as the other AB treatment groups, it could be argued that a lower dose of AB in these animals

resulted in a different compositional makeup of the gut microbiome. Indeed Swann et al., (2011)

used FISH analysis for the enumeration of several key bacterial groups following 8 days of

antibiotic treatment. They observed that the relative proportions of bacteria belonging to the

Bacteroidetes phylum were reduced, while bacteria belonging to the Firmicutes and

Actinobacteria phyla were increased, suggesting not only does antibiotic treatment suppresses the

total bacteria in the gut, but also results in shifts to the compositional makeup of the gut

microbiome. Furthermore, Leclercq et al., (2017) also reported an increase in Proteabacteria due

to increased penicillin resistance Enterobacteria family following penicillin treatment in early

postnatal life in mice, this bacterial group is a potential source of lipopolysaccharides, which are

endotoxins and can elicit a strong immune response in the host. This increase in Enterobacteria

following penicillin treatment has been reported in other studies alongside decreased anxiety like

behaviour (Candon et al., 2015; Desbonnet et al., 2015). Given this evidence it may be possible to

suggest that casein free AB animals did not receive the same dose of penicillin and thus did not

harbor the same increase in Enterobacteria thus displaying a similar behavioural profile to AB-

free treated animals. In future it would be of use to characterize the faecal microbiota of casein

free AB study animals versus control and casein rich AB groups to determine any compositional

differences in the microbiome of these animals which could in turn be linked to the discrepancies

in behaviour.

Another explanation for the lack of antidepressive-like effect of AB treatment in casein free

exposed rats may be due to soy protein exposure, which was used as the protein substitute for

casein in casein free milk. As discussed in the previous results chapter (see Section 2.4), exposure

to soy rich/casein free milk induces a number of region specific neurochemical changes to MOPr

207

in the brain as well as a down regulation of DOPr in the deep layer of the somatosensory cortex

compared to control and casein rich treatment via an unknown mechanism. As it has been shown

that peptides derived from soy protein named soymorphins possess opioid receptor activity

(Kaneko et al., 2010; Ohinata et al., 2007), it is tempting to speculate that the resistance of soya

fed animals to exhibit the ‘anti-depressant’ like effect induced by AB treatment may be due to a

gut microbiota independent effect of soymorphines, which may have contributed a protective

effect against behavioural changes induced by AB treatment. These results indicate a soy effect

on behaviour which warrants further investigation.

The excretion of urinary metabolites were studied using NMR based metabonomics in order to

investigate the metabolic impact of diet (milk casein exposure beyond the normal age of weaning)

and AB treatment in early postnatal life on the host metabolic profile. Pairwise PCA models were

built to identify metabolic variation between the different study groups. This included: the effect

of casein consumption alone on host metabolic profile, the effect of AB treatment alone on host

metabolic profiles, the effect of AB and casein rich treatment and finally the effect of AB treatment

and casein free treatment on host metabolic profiles. Loading plots were generated from each

pairwise PCA model showing the metabolites (1H NMR chemical shifts) responsible for any

separation witnessed in the PCA scores plot. The corresponding loading plots indicated that the

most significant metabolic differences induced by AB and casein treatment were related to

disruptions in the metabolism of choline and energy homeostasis as well as differences in

microbial-derived metabolites (Summarized in Table 3.1).

208

Metabolic Alterations Caused by Casein Intake

A decrease in the excretion of betaine was observed in the urine of casein rich animals compared

to the control group (see Figure 3.10). This may be related to a decrease in endogenous choline

metabolism by the host or may suggest an increase in bacterial choline degradation in the gut.

Choline is a micronutrient found in food and serves as the precursor for the biosynthesis of the

phospholipids, phosphotadylcholine and sphingomyelin which are essential constituents of

membranes and play key roles in normal brain function and development (Zeisel, 1981, 2006).

Moreover, choline is needed for the synthesis of the neurotransmitter acetylcholine which plays

an important role in influencing the structure and organization of brain regions, neurogenesis,

myelination and synapse formation (Wessler et al., 2008). Choline is metabolized within liver

mitochondria and converted into betaine by two enzymes betaine aldehyde dehydrogenase and

choline dehydrogenase. Alternatively, dietary choline can be metabolized by the gut microbiota to

trimethylamine (TMA). TMA is subsequently absorbed from the gut and oxidized in the liver to

form trimethylamine-N-oxide (TMAO) before being excreted in the urine. Interestingly, casein

rich treated animals excreted lower amounts of TMAO compared to the control group. Taken

together with the lower excretion of betaine, these observations suggest that animals receiving

casein rich milk until PND26 obtain less choline from the diet then the control animals

The reduced excretion of betaine could be indicative of disruption to the metabolism of choline.

This is consistent with the findings of Farshim et al., (2016) who reported that animals who were

delayed weaned on PND25 excreted decreased levels of betaine compared to animals who were

normally weaned on PND21 and also displayed a depressive phenotype in the FST (Farshim et al.,

2016). These observations indicate that prolonged consumption of milk caseins beyond the normal

age of weaning disrupt choline metabolism and could potentially be linked to the depressive

209

phenotype observed in casein rich treated animals. Indeed a series of studies conducted by Zheng

and colleagues have shown alterations in TMAO and betaine in depressive states in both human

and rodent models (Zheng et al., 2011, 2013; Huang et al., 2015). For example, Zheng et al.,

(2013) showed that urinary levels of betaine and TMAO were significantly decreased in patients

with major depressive disorder compared to healthy controls (Zheng et al., 2013). These findings

further reinforce that disruptions to choline metabolism may indeed be linked to the depressive

phenotype observed in casein rich treated animals observed in our study. Interestingly, data

obtained in results Chapter 2 showed that animals fed casein rich milk excreted higher amounts of

betaine compared to casein free counter parts. This is in conflict with the current results and may

requires further investigation.

Casein consumption appears to have caused alterations in metabolites associated with energy

metabolism. Urinary excretion of three tricarboxlic acid (TCA) cycle associated metabolites,

citrate, succinate and 2-oxoglutarate were shown to be excreted at lower levels in the urine of

casein rich treated animals compared to control and casein free groups, thus highlighting a casein

effect on energy metabolism (see Figures 3.9 and 3.10). Decreased levels of succinate, citrate

and 2-oxoglutarate in casein rich animals are indicative of a dysfunction in the TCA cycle induced

by casein, which may in turn be associated with the depressive phenotype observed in this group.

In line with this Zheng et al., (2010) carried out a study to look at urinary metabolic changes in

chronic unpredictable mild stress model of depression and observed decreased citrate and 2-

oxoglutarate in the urine of the depressed rats compared to control (Zheng et al., 2010).

Furthermore Farshim et al., (2016) also reported decreased levels of citrate, succinate and 2-

oxoglutarate in the urine of rats who were delayed weaned on PND25 and displayed depressive

like behaviour compared to animals weaned normally on PND21 (Farshim et al., 2016). Taken

210

together these findings suggest that exposure to milk caseins beyond the normal age of weaning

results in disruptions to the TCA cycle and this may be linked to the development of the depressive

phenotype observed. The lower levels of TCA cycle metabolites observed in casein rich animals

indicate energy deficiencies and fatigue, which is in line with the increased immobility times

observed in casein rich rats, as well as the behavioural data from chapter 2, where a significant

decrease in active swimming and climbing behaviours was observed in casein rich treated animals

during the FST. It should also be noted that energy deficiencies or fatigue are amongst the most

popular reported symptoms of depression.

Creatinine is another metabolite associated with energy metabolism, which was found to be

increased in casein rich treated rats compared only to the casein free group and not control.

Creatinine is a non-protein nitrogenous compound derived from creatine and phosphocreatine

metabolism. The creatine-phosphocreatine system acts as an energy transport system which carries

high-energy phosphates from mitochondrial production sites to energy utilization sites,

particularly in tissues with high energy demands like skeletal or cardiac muscle as well as the brain

(Wyss et al., 2000). Excretion of creatinine in the urine represents the endpoint of endogenous

energy transfer from stored adenosine triphosphate (ATP) in cells. It is becoming increasingly

evident that endogenous creatine plays a critical role in a range of cognitive functions including

emotion (see Allen, 2012). In fact, there is evidence to show that supplementation with 4% creatine

for five weeks in male rats results in increased depressive like behaviour in the FST (Allen et al.,

2010). Moreover, this is also in accordance with findings from Farshim et al. (2016) who reported

increased creatinine excretion in animals who were exposed to mothers milk for 5 extra days

compared to animals weaned normally on PND21. Together these findings further implicate casein

211

consumption in influencing energy metabolic products, which may be linked to the depressive

phenotype observed in the casein rich group

Metabolic Alterations Caused by Microbial Suppression

In order to assess the metabolic impact of AB treatment all AB treated animals were compared

against animals which did not receive AB. Consistent with the FISH results showing a suppressed

gut microbial profile in AB treated rats, PCA analysis of urine samples collected from all six

treatment groups revealed differential metabolic patterns of clustering between AB-free and AB

treated groups (See Figure 3.11). The corresponding loadings plot showed metabolites involved

in energy metabolism such as glucose and citrate were increased in AB treated animals as well as

an increase in choline metabolism as indicated by increased TMAO and betaine in AB treated

urine. Meanwhile bacterial metabolites such as hippurate and 4-HPPA were decreased in

comparison to control AB-free treated groups. These findings suggest that treatment with a

combination of penicillin/streptomycin for a period of 10 days was indeed sufficient in causing

disruption to the gut microbiome and subsequently host metabolic profile.

In order to assess the metabolic impact of AB treatment alone independent of milk type on host

metabolic profiles, a pairwise comparison between control and control AB treated animals was

performed. An increase in the excretion of the metabolite DMA was observed in control AB

animals compared to the control group. TMA can be further demethylated by the gut microbiota

to DMA. The fact that we see an increase in DMA following AB treatment is surprising, however

this metabolic function may be ubiquitous amongst the gut microbial species. Hence, AB-

associated disruption of the gut microbiota may reduce the specific functionalities associated with

212

a particular bacterial group and increase the general traits of the communities left behind.

Furthermore, we observed an increase in the excretion of 2-oxoglutarate in control AB treated

animals, which is a TCA cycle associated metabolite. An increase in this metabolite indicates

increased energy release. Indeed, this is in line with the behaviour observed in control AB treated

animals who displayed an antidepressive-like effect as shown by a drastic increase in swimming

behaviour in the FST compared to the control AB-free group.

Metabolic Alterations Caused by Microbial Suppression in Casein Rich Rats

Pair wise comparisons were made between casein rich, casein rich AB, control AB and control

treatment groups to study the effects of AB treatment in rats exposed to a casein rich milk diet.

Consistent with bacterial suppression post AB treatment in control animals and not casein free

animals, the gut microbial metabolite hippurate was excreted in lower amounts in casein rich AB

animals when compared to control (see Figure 3.13), and in the urine of control AB animals when

compared to casein rich (see Figure 3.12). Moreover, the gut microbial metabolite 3-HPPA which

arises through bacterial hydrolysis of dietary derived cholinergic acid and subsequent bacterial

reduction by caffeic acid was also decreased in control AB animals compared to casein rich

animals, together these results suggest that AB treatment independent of casein suppresses gut

microbial activity resulting in a reduction of the excretion of gut microbial metabolites which

suggests lower gut activity may be linked to the anti-depressive like effect seen in control and

casein rich AB animals.

The excretion of amino acids alanine, and taurine were increased in casein rich AB treated animals

compared to control and casein rich groups. These findings are in accordance with Swann et al.,

213

(2011) who reported similar increases in these amino acids following 8 days of

streptomycin/penicillin treatment in the drinking water. The increase in excretion of these

metabolites could be due to the loss of bacterial activity in the gut, which would typically be

responsible for the degradation of these metabolites, thus improving their bioavailability and

subsequent excretion in the urine. Indeed, increased excretion of taurine in the urine is a

characteristic feature of germ-free and acclimatizing germ-free rodents (Nicholls et al., 2003;

Claus et al., 2008). An increase in these metabolites provides further evidence for the effectiveness

of the antibiotic regime in suppressing bacterial abundance as well as activity in casein rich

exposed rats.

Moreover, the urinary extraction of betaine was seen to be greater in casein rich AB treated rats

compared to conventional casein rich animals. This indicates increased choline metabolism in

casein rich animals following AB treatment. Together these findings reinforce that AB treatment

alone, independent of casein, affects choline metabolism. This is in agreement with Swann et al.,

(2011) who reported an increase in urinary extraction of betaine and DMG in the urine following

8 days of AB treatment in rats suggestive of increased choline metabolism. Although choline can

be synthesized in the liver most choline is obtained from the diet. It is feasible that AB treatment

suppresses the microbial metabolism of dietary choline in the gut, thus increasing its availability

for the host. This could explain why products of endogenous choline metabolism are increased in

the urine following AB treatment. Consistently, DMA (a microbial metabolite of choline) was

excreted in lower amounts by the casein rich AB group compared to the AB-free casein rich group.

Interestingly, when comparing casein rich AB to the casein rich group, we observed higher

excretion of DMA in casein rich animals. This indicates that casein rich milk treatment also has

an effect on choline metabolism independent of AB treatment (as discussed above). It appears

214

casein rich milk exposure decreases choline excretion which is accompanied by a depressive-like

phenotype in the FST, while AB treatment appears to increases choline excretion while eliciting

an anti-depressive like effect in casein rich rats, therefore it is possible to suggest that disruptions

to the metabolism of choline maybe implicated in driving the behavioural differences observed

between these two treatment groups. Further studies are required to understand the importance of

choline metabolism in the diet-associated behavioural changes reported in this study.

Casein rich AB treated animals excreted higher levels of the TCA energy cycle metabolite citrate

when compared to casein rich animals. This suggests casein increases citrate excretion, which is

further exacerbated by AB treatment. This may reflect changes in energy metabolism, which is

further supported by lower glucose excretion in the casein rich animals compared to the control

AB treatment group. These findings are consistent with the increased mobility and swimming

behaviours observed in control AB and casein rich AB treated animals.

Metabolic Alterations Caused by Microbial Suppression in Casein Free Rats

Casein free AB animals excreted higher amounts of TMAO compared to control and casein free

animals (see Figures 3.14 and 3.15). This indicates that casein free milk exposure alone

independent of AB results in higher TMAO excretion, which is then further exasperated by AB

treatment. Choline can be metabolized to glycine through a strictly mammalian pathway or it can

be metabolized via bacteria in the gut to produce trimethylamine (TMA), which in turn is oxidized

in the liver to produce TMAO (al-Waiz et al., 1992). An increase in TMAO excretion in casein

free AB animals points towards an increased flux towards bacterial metabolism of choline.

Furthermore, as we did not observe an increase in betaine, an endogenous choline metabolite, in

215

the urine of casein free AB animals, it is further indication of an increase in bacterial choline

metabolism in this group, despite AB treatment. As this increase in TMAO was not observed in

control and casein rich animals following AB treatment, it suggests that the gut microbiome of

casein free AB animals may indeed have differed to the other AB groups and thus provide an

explanation for the behavioural phenotype observed in casein free AB treated animals.

Furthermore, it was observed during the running of the experiment, that casein free AB animals

were calmer and easier to handle compared to the other AB subjects and did not show any signs

of aggression.

As observed in control AB and casein rich AB we also observed a decrease in the excretion of the

gut microbial metabolites hippurate and 3-HPPA in casein free AB treated animals compared to

the casein free group. This result indicates that AB treatment was indeed successful in suppressing

the gut microbial activity in these animals independent of milk type. This is further supported by

the FISH analysis results, which showed a reduced total bacterial count in casein free AB animals

in line with the other AB treatment groups. Without enumeration of individual bacterial groups it

is not possible to tell if the gut microbiome of the casein free AB group may have differed from

the other AB treatment groups. Furthermore, in line with the other AB treatment groups, casein

free AB animals excreted higher amounts of the amino acid taurine, which is characteristic of a

reduced gut microbiome. Taken together these findings show that AB treatment in casein free

animals resulted in a similar reduction of gut microbial activity as seen in control and casein rich

AB animals, however due to the increase in TMAO it can be speculated that the gut microbiome

was more active in this treatment group compared to control AB and casein rich AB. Moreover,

the casein free AB group showed a reduction in the excretion of succinate, which is another product

of energy metabolism, compared to both casein free and control treatment groups. This may

216

suggest that energy metabolism was not increased in casein free AB treatment groups to the same

extent that it was increased in the other AB animals.

In summary this research chapter demonstrated that AB treatment was indeed successful in

reducing both bacterial abundance and activity. It was also shown that not only does AB treatment

ameliorate the depressive-like effect induced by casein rich milk consumption, but also AB

treatment alone independent of casein induces an anti-depressive like effect as observed in control

animals. However, this antidepressive-like effect was not observed in casein free treated animals

following AB treatment, indicating casein free milk played a protective role against the

behavioural changes induced by AB treatment. Taken together, these results clearly demonstrate

AB treatment in early postnatal life effects behaviour independent of casein consumption and

highlights the need for further investigation into the use of AB in early postnatal life. Furthermore,

the metabolic effects induced by casein treatment alone, antibiotic treatment alone and casein or

casein free treatment combined with AB were explored and revealed alterations to choline

metabolism, energy metabolism and bacterial metabolism which warrant further investigation into

their association with the behavioural changes observed.

217

Chapter 4

Effect of A1 vs A2 Variant of Milk Casein

Consumption at an Early Developmental Age

on Depressive-Like Behaviour and Metabolic

Profile

218

4.1 Introduction

The studies performed in Chapters 2 and 3 provide evidence showing exposure to milk caseins

beyond the normal age of weaning results in a depressive-like phenotype as well as dysregulations

to the opioid and oxytocinergic systems in rats compared to control animals and animals provided

with casein free milk beyond the normal age of weaning on PND21. Furthermore we have shown

that the gut microbiota appears to play a role in mediating behaviour as knock-down of the gut

microbiota dramatically reduced immobility time in the FST in rats who consumed casein rich

milk as well as control treated animals at an early developmental stage. This chapter aims to

determine the genetic variant of milk casein responsible for the depressive phenotype observed

following consumption of milk caseins beyond the normal age of weaning.

As discussed in Section 1.2 – 1.3 there has been an increasing interest in recent years regarding

the consequences associated with the consumption of milk casein derived beta casomorphins for

extended periods of time on health. These include increased risk of neurological diseases such as

autism and schizophrenia, cardiovascular diseases, type 1 diabetes, gastrointestinal dysfunction

and immune/inflammation related disorders and which have been attributed to a group of peptides,

particularly BCM-7, derived from the proteolysis of beta casein, (see ul Haq et al., 2014).

Beta casein is expressed as 13 genetic variants in cows (Kamiński et al., 2007) resulting from

single nucleotide polymorphisms in the casein beta gene (CSN2) (Caroli et al., 2009). Out of

these, A1 and A2 are the most common forms in cattle breeds making up around 20% of the total

protein content in cow’s milk (Kamiński et al., 2007). Depending on the individual cow’s genetic

makeup, cows can be homozygous for one type, or heterozygous resulting in both A1 and A2 types

being expressed. According to the literature, A2 beta casein is recognized as the original beta

219

casein variant expressed in cows before the appearance of the A1 variant in some European herds

5000-10,000 years ago (Farrell et al., 2004). The appearance of the A1 variant is believed to have

arisen due to a CSN2 gene mutation which occurred in Holstein cows and was then passed on to

many other cow breeds due to breeding practices, by which Holstein cows were used to genetically

improve the production of other breeds (The A2 Milk Company, 2014

https://thea2milkcompany.com/). Over time the A1 beta casein variant became the dominant form

expressed in milk in Europe. While dairy herds in much of Asia, Africa and parts of southern

Europe remain naturally high in cows producing A2 milk, making the A1 version of the protein

common in cattle in the western world (The A2 Milk Company, 2014

https://thea2milkcompany.com/).

BCM-7 is uniquely derived from the digestion of the A1 β-casein but not the A2 β-casein type.

These two variants differ in their protein structure owing to a substitution of the amino acid at

position 67. A2 β-casein and related sub-variants contain a proline amino acid at position 67,

whereas A1 β-casein and related sub-variants contain a histidine at this same position. This

histidine within A1 is susceptible to proteolytic cleavage, which allows the preceding seven amino

acid residues to be cleaved, yielding BCM-7 (see Figure 1.2) (Jinsmaa et al., 1999). Beta casein

expressed in human, goat, sheep and buffalo milk though not strictly of the A2 type, are classed

as “A2-like”. However, evidence suggests human milk does indeed release BCM-7, which is of a

differing amino acid sequence to bovine BCM-7 (Jarmołowska, Sidor, Iwan, Bielikowicz,

Kaczmarski, Elzbieta Kostyra, et al., 2007; Kost et al., 2009; Wada and Lönnerdal, 2015).

As discussed in Section 1.3, literature shows that BCM-7 possesses opioid like activity due to the

N-terminal amino acid sequence Tyr-Pro-Phe-Pro that gives it preferential MOPr agonist activity

(Brantl et al., 1981). MOP receptors are found primarily in the central nervous system and the

220

gastrointestinal tract and play an important role in responding to stress and pain as well as in

controlling food intake (Teschemacher, 2003). The first site of action of BCM-7 following

digestion are the MOPr in the gut. Indeed when injected into animals, BCM-7 was shown to slow

gastrointestinal motility similar to the effect of morphine (Shah, 2000) and this observation was

also replicated in humans (Andiran et al., 2003; Ho et al., 2014). In a direct comparison of A1 and

A2 milk fed to Wistar rats, gastrointestinal transit time was significantly longer in the A1 group

compared to rats fed A2 beta casein. Furthermore co-administration of naloxone decreased

gastrointestinal transit time significantly in the A1 group but not in the A2 group indicating the

greater gut transit time reported in the A1 group is mediated by a MOPr dependant mechanism

(Barnett et al., 2014).

Furthermore, this group reported that rats fed A1 β-casein showed an increase in a range of markers

of intestinal inflammation compared to the A2 group and this was inhibited by naloxone treatment

(Barnett et al., 2014) again indicating a MOPr mediated increase in markers of inflammation

caused by A1 β-casein. Similarly in mice, dietary consumption of β-casein isolated from A1/A1

or A1/A2 Indian Karan Fries cows resulted in increased IL-4, IgE, IgG concentrations in the

intestinal fluid compared to groups provided with beta casein extracts from A2/A2 cows or

controls (ul Haq et al., 2014). In a second study by the same research group in which mice received

BCM-7 or saline by gavage for 15 days, similar increases in indicators of inflammation were

reported compared to saline control. Taken together these studies indicate that A1 beta casein and

its bioactive breakdown product BCM-7 are responsible for the reported gastrointestinal and

inflammatory effects.

Notably, human studies have gone on to confirm that consumption of milk containing A1/A2 beta

casein results in an increase in inflammation related biomarkers as well as a longer gut transit time

221

compared to milk containing A2/A2 beta casein (Jianqin et al., 2015). It is therefore evident

consumption of A1 derived BCM-7 results in gastrointestinal effects and is proinflammatory.

Whether the proinflammtory effects are due to a direct inflammatory response to BCM-7 or an

indirect consequence of delayed gut transit time influencing other biological processes is still

unclear.

Moreover, some studies have reported effects of BCM-7 on the CNS following perinatal

administration in animal models. Brantl et al., (1981) were amongst the first to report analgesic

activities of BCMs after i.c.v injections in rats, and that this analgesic effect can be antagonized

by naloxone indicating a MOPr mediated mechanism of action. Another study has reported that

consumption of milk containing both A1/A2 variants resulted in minor impairments to cognitive

function, which were linked to the increased serum inflammation-related markers associated with

consumption of milk containing A1/A2 relative to control milk containing only A2 beta-casein

variant. Indeed elevated levels of circulating inflammatory markers have been linked to significant

impairments in memory, attention and mood (Dantzer et al., 2008; Myers et al., 2008; Heringa et

al., 2014).

Given the health effects associated with consumption of milk containing the A1 variant of β-

casein, recently bovine milk that is free of A1 has been made available commercially in the United

Kingdom amongst other countries (The A2 Milk Company, 2014

https://thea2milkcompany.com/). Also infant formula containing casein but free of A1 is now

marketed in China and Australia. This chapter aims to compare the effect of consumption of milk

containing A1/A2 variants against milk containing only the A2 variant during early developmental

period on mood. Moreover the effect of A1/A2 vs A2/A2 milk consumption on motor function

was also assessed. Furthermore, metabonomic analysis was carried out on the frontal cortex region

222

of the brain in order to assess any metabolic changes associated with A2 versus A1/A2 milk

consumption. The frontal cortex was selected as it has been identified as a key region in the

regulation of emotions, in particular it has been linked to depression (Drevets, 2000). There is also

evidence to suggest that diet can elicit a profound effect on frontal cortex neurochemistry

(Sandoval-Salazar et al., 2016) and was thus selected as a key region for metabonomic analysis.

Hypothesis:

Rats provided with A2 milk beyond the normal age of weaning will not display a depressive

phenotype due to a lack of the bioactive peptide BCM-7 compared to rats provided with

A1/A2 milk. Moreover, neither A2 nor A1/A2 milk will have an effect on motor activity.

Due to the differing post weaning diets, metabolic profiles of A2 animals A1/A2 and

control will differ significantly.

Aim:

To investigate the variant of β-casein in milk responsible for mediating the depressive-like

phenotype observed in animals provided with casein rich milk from PND21 until PND26.

Objectives:

To provide animals with milk containing only A2 β-casein or milk containing both A1/A2

from PND21 until PND26 and assess motor activity.

Assess depressive-like behaviour using the FST

Asses for metabolic changes in the frontal cortex region of the brain using NMR based

metabolomics.

223

4.2 Methodology


Male, Wistar Albino rats were purchased from Charles River UK Limited, Kent. All rat pups and

their cross-fostered mother arrive at the Biological Research Facility (BRF) on PND7, where day

of birth is PND0. These animals were maintained in litters of 8 pups with their mothers at all

times. Pups were left to acclimatize to their new environment for 11 days, after which point they

were handled and weighed daily by the experimenter. All groups were weaned on PND21, a

process where the mother is simply removed from the cage (Cramer et al., 1990). All rats were

placed in a 12 hour light-dark cycle (lights on 6 AM – 6 PM) with free access to a standard rat

chow diet (purchased SDS, Kent UK) and tap water and kept at a constant temperature of 22°C ±

1 °C, humidity of 45-55%. All experimental procedures were conducted in accordance with the

U.K Animal Scientific Procedures Act (1986).

4.2.2 A2 Milk Study


Group one were weaned by having their mothers removed at 10.00 am on PND21. This group

acted as the control and were provided with free access to rat chow and water until the completion

of behavioural testing on PND26. Group two, the conventional cow’s milk group (A1/A2 group)

were weaned on PND21 and provided with a bottle containing Tesco’s whole milk from PND21

until PND26, as well as continued provision of rat chow and water. Group three, the A2 milk

group, were also weaned on PND21 and provided with a bottle containing A2 milk daily from

PND21 until PND26 as well as continued provision of rat chow and water.

224

The groups provided with either conventional cow’s milk or A2 milk, were provided with 100 ml

of fresh milk in bottles with ball bearing nozzles twice a day once between 9:00-10:00 AM and

then again between 5-6 PM. Milk consumption was also recorded on a daily basis. Milk bottles

were heated for 30 seconds in a microwave before being provided to the animals. Table 4.1 shows

the nutritional composition of the A2 full fat milk used and Tesco’s whole milk, which was used

to represent A1/A2 milk.

A2 Whole Milk Formula

Parameter Analysed Units Results Fresh Weight

Moisture % 87.6

Crude Fat % 3.62


Crude Fibre % 0.0

Ash raw % 0.7

NFE (Nitrogen Free extract) % 4.83

Tesco Milk Formula

Parameter Analysed Units Results Fresh Weight

Moisture % 87.5

Crude Fat % 3.57


Crude Fibre % 1.3

Ash raw % 0.7

NFE % 5.04

Table 4.14: Table showing the nutritional content of A2 whole milk formula and Tesco’s (A1/A2)

whole milk formula.

4.2.3 Body Weights and Milk Consumption

To monitor the general health of the animals, body weights were recorded daily from PND14 until

the end of behavioural testing on PND26 as well as on the day of sacrifice. Milk consumption per

cage was also recorded daily from PND21 until PND26.

225

4.2.4 Behavioural Tests

4.2.4.1 The Forced Swim Test

The forced swim test was performed between 10 am and 2 pm when all animals reached the age

of PND25. The FST is run over two consecutive days as described in Section 2.2.3.1.

4.2.4.2 Locomotor Activity

Locomotor activity of each rat was measured for 20 minutes 1 hour prior to the FST (9:00 AM)

on PND25 in locomotor chambers (40cm x 20cm x 20cm; Linton Instrumentation, Norfolk, U.K).

Each chamber had two lines of 16 photocells located at right angles to each other, projecting

horizontal infrared beams 2.5 cm apart and 6 cm above the cage floor to measure horizontal and

vertical activity, respectively. Activity was recorded as the number of sequential infrared beam

breaks, every 5 mins in the space of 20 minutes.


Animals were killed using cervical dislocation on PND27 following acute exposure to CO2. Intact

brains were dissected and rapidly frozen in isopentane at -20°C for 30 seconds and then stored at

-80°C for metabonomic analysis.

4.2.6 Data Analysis

Body weight differences, milk consumption and behavioural data from the FST were analysed

using a one-way ANOVA. Locomotion data was analysed using Two-way ANOVA for factors

‘time’ and ‘diet. Where ANOVA revealed a significant difference an LSD post-hoc analysis was

carried out. All data were expressed as means ± SEM, n = 8 per group. A p value of less than 0.05

226

was accepted as significantly different. All statistical analyses were performed on GraphPad Prism

7.

4.2.7 Metabolomics

As described in Section 2.2.6. Metabonomic analysis was conducted by Simone Zuffa at Imperial

College London.

4.2.7.1 Sample Preparation-Brain Samples

Prior to the start of the experiments, brains were removed from -80°C freezer and placed on ice.

Using a scalpel, the frontal cortex (~30 mg) was removed and weighed. The frontal cortex from

each brain sample was homogenized using 6-7 zirconium beads and 300 µl of

chloroform:methanol, which was made up in a 2:1 ratio. Samples were placed into a homogeniser

and homogenised at 5500 rpm for 20 seconds. This step was repeated twice for each sample. The

homogenate was then combined with 300 µl of demineralized water and vortexed to mix. Samples

were spun at 8000 g (rcf) for 10 minutes at 4°C to separate the organic (upper) and aqueous (lower)

phases. These phases were separated into two Eppendorf tubes using a pipette. The organic phase

was left overnight in a fume hood to allow the chloroform to evaporate and then stored at -40 °C.

The aqueous phase was dried overnight in a vacuum condenser (Speedvac) at 40 °C. Once the

samples were dried they were reconstituted in 700 µl of phosphate buffer (pH 7.4) containing 10%

D2O and 1 mM 3-trimethylsilyl-1 [2,2,3,3-2H4] propionate (TSP). Samples were vortexed

thoroughly, centrifuged at 8000 g (rcf) for 10 minutes at 4 °C and then 600 µl of the supernatant

was transferred to 5 mm internal diameter Norell® NMR tubes. The sample were run on the NMR

machine as described in Section 2.2.5.2.

227

4.2.7.2 Multivariate Data Analysis




Postnatal Day

7 14 21 25 26 27

Appropriate groups

provided with A2 or A1/A2

milk diets

Mothers removed from

all groupsForced

Swim Test


cages for 24 hours



Animals arrive in EBU

Figure 4.1: Experimental timeline showing key points in the methodology for antibiotic

dosing study.

228

4.3 Results

4.3.1 Milk Consumption & Weight Change

No differences in milk consumption were observed between rats provided with A2 milk and those

provided with A1/A2 milk from PND21 until PND26. Furthermore, no differences in weight gain

were observed between control, A2 milk and A1/A2 milk treatment groups between PND14 until

day of culling on PND27 (Figure 4.2 A, B respectively).

229

A

B

Contr

ol

A2

Milk

A1/

A2

Milk

0

20

40

60

80

Weig

ht

Ch

an

ge (

g)

21 22 23 24 250

50

100

150

200

250

Postnatal Days

Mean

Daily M

ilk

Co

nsu

mp

tio

n p

er

Cag

e (

g)

A2 Milk A1/A2 Milk

Figure 4.2: Effects of A2 milk versus A1/A2 Milk on milk consumption from PND21 until PND26 and

weight change from PND14 until PND27.

Male Wistar albino rats were provided with no milk (control), A2 milk or A1/A2 milk from PND21 until

PND26. (A) Milk consumption for A2 and A1/A2 treatment groups. (B) Weight change in grams for

control, A2 milk and A1/A2 milk from PND14 until day of culling on PND27. No difference in either milk

consumption or weight change were observed. Data are expressed as mean ± SEM (n=8 per group). One

way ANOVA followed by LSD post-hoc test.

230

4.3.2 Locomotion

Locomotion testing was carried out on PND25 1 hour prior to the FST. Animals provided with no

milk (control), A2 or A1/A2 milk were placed into the aforementioned locomotor chambers for 20

minutes and total horizontal or vertical beam breaks were measured. Two-way repeated measures

ANOVA for factors ‘time’ and ‘diet’ showed a significant ‘time’ only effect on both total

horizontal activity and total vertical activity [F (3,21) = 65.23); p< 0.01] and [F (3,18) = 15.43); p<

0.01] respectively (Figure 4.3 A, B). LSD post-hoc analysis revealed A2 milk consumption

significantly reduced horizontal activity at 10 minutes (p < 0.05) compared to the control group

and increased total vertical activity compared to the A1/A2 treatment groups (p < 0.01).

231

5 10 15 200

200

400

600

Time (min)

To

tal H

ori

zo

nta

l A

cti

vit

y

(Ho

rizo

nta

l B

eam

Bre

aks/2

0 m

in) Control

A2 Milk

A1/A2 Milk

*

B

A

5 10 15 200

50

100

150

Time (min)

To

tal R

eari

ng

Acti

vit

y

(Vert

ical B

eam

Bre

aks/2

0 m

ins) Control

A2 Milk

A1/A2 Milk **

Figure 4.3: Effect of A2 milk versus A1/A2 milk consumption from PND21 until PND26 on locomotor

activity.


PND26. (A)Total horizontal activity and (B) Total vertical activity were measured on PND25 1 hour prior

to the forced swim test in 5 min bins for 20 minutes. A decrease in total horizontal activity was seen in A2

treated animals at 5-10 minutes compared to control animals, and an increase in total rearing activity was

seen in A2 treated animals compared to A1/A2 animals at 5-10 minutes. Data are expressed as mean ± SEM

(n= 7-8 per group). Two-way ANOVA followed by LSD post-hoc test was used and a p value of less than

0.05 was accepted as significantly different. *p<0.05, **p<0.01.

232

4.3.3 Forced Swim Test

The FST was used to assess depressive-like behaviour in rats provided with no milk (control), A2

milk or A1/A2 milk from PND21 until PND26. The total immobility time was scored from the

‘test’ session on PND26 (Figure 4.4). The quantification from the ‘test’ times were then used in

the assessment of depressive-like behaviour as described by (Cryan et al., 2005). Quantification

of the results from the ‘test’ swim session showed that rats fed both A2 and A1/A2 milk displayed

an increase in depressive-like behaviour as shown by a marked increase in immobility time

compared to control treated animals (p < 0.05 and p < 0.01 respectively).

Contr

ol

A2

Milk

A1/

A2

Milk

0

100

200

300

400

Imm

ob

ilit

y t

ime (

s)

**

*

Figure 4.4: Increase in depressive-like behaviour caused by consumption of both A2 and A1/A2 Milk

from PND21- PND25.


PND26. Then subjected to behavioural testing using the FST. Immobility time during the 6 minute test

session was measured on PND26. Data are expressed as mean (± SEM) obtained from manual video analysis

by one independent observer. An increase in immobility time was seen in both A1/A2 and A2 treated

animals compared to control animals indicating an increase in depressive-like behaviour. One-way

ANOVA followed by LSD post-hoc test was used and a p value of less than 0.05 was accepted as

significantly different. *p<0.05, **p<0.01.

233

4.3.4 Metabonomic Analysis of Brain samples

4.3.4.1 PCA of Frontal Cortex: All data

PCA was first used as an unsupervised multivariate statistical analysis method to give an insight

into metabolic variation in the frontal cortex samples from the brains of control, A2 milk and

A1/A2 milk treatment groups. The percentage of variance described by each PC (R2 ) is shown for

each axis. To aid the visualisation of any trends associated with the treatment groups, individual

animals are colour coded in the scores plot to indicate their class membership. Following initial

PCA modelling of all frontal cortex data, clear separation along PC1 can be seen between the

A1/A2 profiles and those from control animals and the A2 group (Figure 4.5 A). This indicates

that A1 β-casein exposure from PND21 until PND 26 is responsible for the largest source of

variation in the metabolic data. Pair-wise PCA models (Figure 4.5 B-C) further demonstrate this

metabolic variation between A1/A2 samples and those from the control and A2 groups. No

variation was observed in the frontal cortex metabolic profiles from control and A2-treated

animals.

234

A B

C D

Figure 4.5: PCA of frontal cortex samples from Control, A1/A2 and A2 treated animals.

PCA scores plot constructed on the brain metabolic profile of: (A) All groups (R2 = 78.5%) showing

separation along PC1 between control/A2 milk and A1/A2 milk, (B) control and A1/A2 group, (R2 = 78.8%)

showing separation along PC1 between the two groups, (C) A1/A2 and A2 milk group, (R2 = 81.3%)

showing separation along PC1 between the two groups, (D) Control and A2 milk group, (R2 = 74.3%)

showing no separation between the two groups. Frontal cortex metabolomics samples were run and

analysed by Simone Zuffa at Imperial Collage London.

235

4.3.4.2 OPLSDA of Frontal Cortex

OPLS-DA was used to further investigate the effect of exposure to A1 β-casein on animal

metabolic profiles. This supervised method allows the metabolic variation specifically associated

with class membership (control, A1/A2, A2) to be identified. To identify the metabolites in the

frontal cortex that differed between control, A2 milk and A1/A2 milk treatment groups, pair-wise

OPLSDA models were constructed using milk type as the Y predictor as explained in Section

2.2.6.6. Models with strong predictive ability (Q2Y) were returned for comparisons. The pairwise

comparison between control and A1/A2 milk groups can be seen in Figure 4.6 and the pairwise

comparison between A2 and A1/A2 milk groups can be seen in Figure 4.7. Both models had a

strong predictive ability (Q2Y = 0.356 and Q2Y = 0.483 respectively), which were found to be

significant after permutation testing (p = 0.041 and p = 0.022 respectively; 10,000 permutations).

From the correlation coefficients plot extracted from the OPLS-DA model comparing control and

A1/A2, it was found A1/A2 rats had higher amounts of aspartate, N-acetyl groups, glutamine,

succinate, glutamate, gamma-amino butyric acid (GABA), acetate, alanine, lactate, ascorbate,

inosine, myo-inositol, creatine, threonine, O-phoshphocholine, taurine, O-acetylcholine, choline,

adenosine monophosphate (AMP), inosine monophosphate (IMP), nicotinurate/niacinamide,

adenosine, uridine monophosphate (UMP), phenylalanine, histidine, S-adenosylhomocysteine,

GTP, and tyrosine/tyramine than the control group rats. Comparing A2 and A1/A2 profiles

identified that A1/A2 rats had higher amounts of creatine, GABA, asparate, N-acetyl group,

glutamine, succinate, glutamate, acetate, alanine, lactate, ascorbate, myo-inositol, threonine, O-

phosphocholine, taurine, choline, O-acetylcholine, niacinamide/nicotinurate, IMP, AMP, tyrosine,

GTP and UMP compared to the A2 rats. The predictive performance of the OPLS-DA model

236

comparing control and A2 milk treatment groups was poor (Q2Y = -0.29) indicating no metabolic

differences between these samples. This was also seen in the PCA scores plot comparing these the

two groups.

237

Figure 4.6: The effect A1/A2 milk from PND21 until PND26 on rat frontal cortex metabolites

compared to control treatment: OPLS-DA Plots.

OPLS-DA model generated from frontal cortex 1H-NMR spectra from control and A1/A2 treated animals.

The OPLS-DA correlation coefficients (R2 0.723) plot illustrating all metabolites responsible for driving the

separation between the two groups. Labelled peaks are the metabolites which are increased in the respective

groups (A1/A2 milk at the top and control at the bottom) (Q2Y =0.356). For the list of metabolites and peak

shifts refer to Table 4.2. Frontal cortex metabolomics samples were run and analysed by Simone Zuffa

at Imperial College London.

238

Figure 4.7: The effect of A2 or A1/A2 Milk types from PND21 until PND26 on rat frontal cortex

metabolites: OPLS-DA plots.

OPLS-DA model generated from frontal cortex 1H-NMR spectra from A2 and A1/A2 treated animals. The

OPLS-DA correlation coefficients (R2 0.799) plot illustrating all metabolites responsible for driving the

separation between the two groups. Labelled peaks are the metabolites which are increased in the respective

groups (A2 milk at the top and A1/A2 milk at the bottom) (Q2Y =0.483). For the list of metabolites and peak

shifts refer to Table 4.2. Frontal cortex metabolomic samples were run and analysed by Simone Zuffa

at Imperial College London.

239


A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

Aspartate

N-Acetyl Group

Glutamine

Succinate

Glutamate

GABA

Acetate

Alanine

Lactate

Ascorbate

Inosine

Myo-Inositol

O-Phosphocholine

Taurine

O-Acetylcholine

Choline

AMP

1.48 (d), 3.76 (q)

2.02 (s), 2.50 (q), 2.69 (q), 4.39 (q)*

2.14 (m), 2.46 (m), 3.77 (t), 6.86 (s)

2.41 (s)

2.1 (m), 2.35 (m), 3.76 (q)

1.9 (m), 2.3 (t), 3.01 (t)

1.92 (s)

1.48 (d), 3.77 (q)

1.33 (d), 4.11 (q)

3.74 (m), 4.02 (m), 4.52 (d)

4.28 (q), 4.44 (t), 6.10 (d), 8.25 (s), 8.35 (s)

3.29 (t), 3.54 (q), 3.63 (t), 4.07 (t)

3.04 (s), 3.93 (s)

1.33 (d), 3.59 (d), 4.25 (m)

3.22 (s), 3.59 (m), 4.16 (m)

3.27 (t), 3.42 (t)

240


IMP

Nicotinurate/Niacinamide

Adenine

UMP

Phenylalanine

Histidine/Histamine

Adenosine

S-Adenosylhomocysteine

GTP*

Tyrosine/Tyramine

4.48 (q), 6.14 (d), 8.20 (s), 8.53 (s)

7.60 (q), 8.25 (m), 8.72 (q), 8.94 (d)

8.15 (s), 8.22 (s)

4.34 (t), 4.41 (t), 5.99 (m), 8.11 (d)

7.33 (d), 7.38 (t), 7.43 (t)

7.10 (s), 7.89 (s)

4.44 (q), 6.08 (d), 8.25 (s), 8.35 (s)

4.32 (m), 4.44 (t), 6.10 (d), 8.27 (s), 8.35 (s)

5.94 (d), 8.14 (s)

6.91 (d), 7.19 (d)

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk

A1/A2 milk and A2 milk

A1/A2 milk and A2 milk

A1/A2 milk

A1/A2 milk

Table 4.2: Full 1H- NMR chemical shifts for metabolites identified in frontal cortex samples of A1/A2 treated rats. Key: s, singlet; d, doublet; t, triplet;

q, quadruplet. *not confirmed. Abbreviations: GABA: gamma-amino butyric acid; AMP: adenosine monophosphate; IMP: inosine monophosphate; UMP:

uridine monophosphate; GTP: Guanine triphosphate

241

4.4 Discussion

The aim of this study was to determine the genetic variant of milk casein responsible for the

depressive-like phenotype observed following consumption of milk caseins beyond the standard

age of weaning. To achieve this, one group of animals was provided with milk containing only the

A2 variant of casein (A2 milk) while another group was provided with milk containing a mixture

of A1/A2 β-casein again from PND21-PND26 and then assessed for depression using the FST.

The effect of A2 or A1/A2 milk consumption on motor activity was also investigated. Furthermore

brain (frontal cortex) metabolic profiles were also compared between control, A2 and A1/A2

treated animals.

Interestingly, animals provided with both A2 milk and A1/A2 milk types displayed an increase in

depressive-like behaviour compared to the control group as indicated by a significant increase in

total immobility time (Figure 4.4). It was originally hypothesised that only milk containing the

A1 variant of casein (A1/A2 milk) would result in a depressive phenotype as A1 beta casein has

been shown to release the bioactive peptide BCM-7 (Kamiński et al., 2007) and a large body of

evidence has linked BCM-7 exposure with a number of negative health effects including

neurological diseases (see ul Haq et al., 2014). However, the results obtained show that A2 milk

which is devoid of the A1 variant, also elicited a depressive phenotype in animals despite the lack

of BCM-7 presence. This indicates that there are other bioactive components in milk, besides the

A2 beta casein derived casomorphins which may be driving the behavioural changes associated

with prolonged milk consumption. However as behavioural results from Chapters 2 and 3 have

shown that consumption of casein-free milk for 5 extra days beyond the normal age of weaning

did not result in a depressive-like phenotype. Taken together with the current results, this indicates

242

that the depressive-like behaviour is driven by casein but not specifically A2 beta casein derived

BCM-7 as originally hypothesised.

Indeed, the α variant of casein (α casein) is also known to contain fragments with opioid receptor

activity. These are termed α casein exorphins and they are released from α casein by pepsin

digestion (in bovine milk it is the αs1 casein which releases these peptides) (see Teschemacher,

Koch and Brantl, 1997; Nyberg et al., 2013). These α casein exorphins exhibit DOPr activity.

Human α casein was found to give rise to the opioid antagonist casoxin D which was shown to

exhibit low DOPr affinity. Similarly, peptic and tryptic digestion of bovine Ƙ casein gives rise to

casoxins A, B and C which were all found to behave as opioid receptor antagonists, exhibiting a

low affinity for both MOPr and KOPr (see Teschemacher, Koch and Brantl, 1997). Thus, it is

feasible that biologically active opioid peptides besides BCM-7 arising from proteolytic cleavage

of A1 and A2 beta casein, aswell as bioacive break down products from other varients of casein

are responsbile for driving the behavioural changes observed in the FST.

In order to assess that the behavioural changes observed in the FST were not driven by a change

in locomotor function, total horizontal and total rearing activies were assessed for 20 minutes in

control, A2 and A1/A2 treated animals 1 hour proir to the FST on PND25. The data show no

significant effect of diet on either parameter of locomotion (Figure 4.3). The first 5 minites in the

locomotion chamber is used to score exploration of a novel environment (Crawley, 2008). Upon

first exposure to a novel open field environment, rodents display marked behavioural activation in

the form of exploratory behaviour, which usually includes sniffing, walking/running and hind leg

rearing (Drai et al., 2001). In agreement, the present data showed that during the initial 5 mins

both vertical and horizontal activities were high in all rats indicating no difference in exploratory

behaviour between all three treatment groups in the first 5 minutes. This reinforces that the changes

243

observed in the 6 minute FST are unlikely to be driven by a difference in motor activity or

exploratory behaviour and is indeed indicative of a true depressive-like phenotype.

Over time, all groups showed a significant decrease in their total horizontal and vertical activity

over the 20 minute test period (time effect). This is an effect of behavioural habituation, a basic

form of learning (Thiel et al., 1998) whereby repeated or prolonged exposure to an environment

renders it less novel and thus reduces exploratory behaviour which is then replaced by

grooming/cleaning behaviour and eventually hibernation.

Interestingly, post-hoc analysis revealed that at 5-10 minute time bin, a significant decrease in

horizontal activity was observed in A2 treated animals compared to control animals. This may

indicate that A2 treated animals habituated quicker to their new environment and hence stoped

exploring the novel environment sooner. However, a significant increase in total time spent rearing

was also observed in A2 treated animals at 5-10 minute time bin compared to A1/A2 exposed rats,

this may explain the decrease in horizontal activity as A2 animals spend more time rearing (

vertical activity), perhaps exploring other aspects of their new environment. Although the FST

data suggest that consumption of both A2 milk ( no BCM-7) and A1/A2 milk (containing BCM-

7) result in depressive like behaviour, there appears to be some modulation of other behavioural

phenotypes mediated by BCM-7 exposure which require further investigation.

Metabolomic analysis was carried out on the frontal cortex, a key region in emotional regulation,

in order to assess the neurochemical effect of BCM-7 exposure through the consumption of A1/A2

milk compared to the lack of exposure to BCM-7 through the consumption of A2 milk or no milk

(control animals). A clear difference in metabolic profile was observed in animals exposed to

BCM-7 through A1/A2 milk consumption compared to control and A2 treated animals. PCA

analysis showed animals who had no exposure to BCM-7 clustered together (control and A2

244

treatment groups) while BCM-7 exposed animals ( A1/A2 treatment group) separated in metabolic

profile (Figure 4.5) OPLS-DA analysis further highlighted these metabolic differences confirming

that exposure to BCM-7 from PND21-PND26 altered the neurobiochemical profiles.

A significant increase in a number of neurotransmitters in the frontal cortex of A1/A2 treated

animals was observed in comparison to both control and A2 animals (Figure 4.6 and 4.7),

including GABA, glutamate (Glu) and glutamine (GLn) (together GLx). GABAergic and

glutamatergic neurotransmitter systems are the major inhibitory and excitatory neurotransmitters

in the mammalian central nervous system (Fonnum, 1984) and are thereby involved either directly

or indirectly in most aspects of normal brain function, including mood regulation. A large body of

evidence shows that GABAergic and glutamergic neurtotransmission are central to the

pathophysiology of depression (Hasler, 2010). Indeed a number of clinical and preclinical studies

show that major depressive disorder (MDD) and stress exposure are associated with an increase in

the abundance of glutamate in the prefrontal cortex (see Popoli et al., 2011). Although it is not

possible to determine the causal association between increased glutamate and GABA levels

observed in the frontal cortex of rats exposed to BCM-7 through A1/A2 milk ingestion and the

observed depressive-like phenotype, it is unlikely that this is driving it as the depressive phenotype

was also observed in A2 animals who were not exposed to BCM-7. Nonetheless, the data clearly

indicate a profound effect of BCM-7 exposure through A1/A2 milk on brain neurochemistry.

The current data clearly suggests the presence of BCM-7 in the A1/A2 treated animals is driving

the increase in GABAergic and glutamergic metabolites in the frontal cortex. Although it appears

no studies have measured the release of these neurotransmitters following BCM-7 treatment, this

argument is supported by evidence that shows acute and chronic morphine treatment (a MOPr

agonist similar to BCM-7 albeit of a higher potency) results in significant alterations to

245

extracellular concentrations of glutamate and GABA in a number of brain regions including the

prefrontal cortex (Hao et al., 2005; Gao et al., 2007). In the study by Gao et al., (2007) metabolic

changes in the prefrontal cortex of rats were measured following 10 days of morphine injections

using 1H NMR spectroscopy. They showed repeated morphine treatment induced a progressive

change in a number of cerebral metabolites including an increase in GABA concentration and a

decrease in Glu concentrations in the prefrontal cortex. The observed changes are believed to be

an adaptive response of cerebral metabolism to ongoing morphine exposure and perhaps the

development of morphine tolerance and dependence (Gao et al., 2007). Although this in agreement

with the higher amount of GABA metabolites observed in the current study, it is not consistent

with the lower Glu reported. This could reflect differences in potency between BCM-7 and

morphine and also the length of exposure to both agonists (only 5 days of exposure in the current

study versus 10 days of exposure in Gao et al.,( 2007) study). The metabolic pathways regulating

the synthesis and cycling of GABA and glutamate are tightly coupled. Gln is the primary precursor

for both neuronal Glu and GABA (Pascual et al., 1998). Therefore, any changes to this shared

pathway will affect levels of both associated metabolites. The current data show an increase in Gln

accompanied by a concomitant increase in both Glu and GABA. These findings indicate that

similar to the effect of morphine, sustained BCM-7 exposure from PND21- PND26 alters the Glu,

Gln and GABA content of the brain. In line with this, morphine-induced activation of MOPr has

been shown to enhance GABAergic neuronal activities in the prefrontal cortex (Mansour et al.,

1987), however this has been shown to provide an inhibitory influence on glutamatergic neurones

(Cowan et al., 1994). Thus given the opposing excitatory and inhibitory roles of both of these

neurotransmitter systems, more evidence is needed to pin-point the exact mechanisms underlying

the reported increased concentrations in the frontal cortex following BCM-7 exposure.

246

Metabonomic analysis of the prefrontal cortex also showed an increase in a number of other

metabolites including myo-inositol and taurine. Both of these metabolites are markers of astrocytic

activity and play important roles in regulating intracellular osmoloarity of astrocytes (Isaacks et

al., 1994). This suggests an increase in astocytic activity in the frontal cortex following BCM-7

exposure through ingestion of A1/A2 milk. Increased astrocytic activity has been implicated in the

regulation of glutamergic and GABAergic neurotransmission amongst other roles such as supply

of energy metabolites for synaptic function (see Miguel-Hidalgo, 2009). Therefore an increase in

astrocytic activity may be linked to the increase in GABA and glutamate neurotransmission

following exposure to BCM-7 through A1/A2 milk consumption. In support of the concept of

BCM-7 induced astrocytic activity, chronic treatment with morphine has been shown to result in

an increase in astrocytic activity as measured by glial fibrillary acidic protein (GFAP), a key

component of the astrocytes cytoskeleton, in a number of brain regions including the frontal cortex

(see Miguel-Hidalgo, 2009;Song and Zhao, 2001; Alonso et al., 2007).

The current data also shows an increase in N-acetyl groups (NAG) which has a spectrum very

similar to N-acetylaspartate (NAA) in the frontal cortex of BCM-7 exposed rats through the

ingestion of A1/A2 milk. NAA is a putative marker of neuronal activity. It is localised within

neurons and is involved in synaptic processes (see Miller, 1991). NAA levels have been shown to

be elevated in several degenerative neurological conditions and high concentrations have been

shown to behave like a neurotoxin (see Miller, 1991). A greater amount of NAA in the frontal

cortex following BCM7 exposure could indicate enhanced or disturbed neuronal activity caused

by BCM-7 exposure. This notion is supported by evidence that suggests NAA is responsible for

the elimination of metabolic water formed in large quantities especially during high neuronal

activity (Baslow, 2003). Increased neuronal activity is indeed implicated in A1/A2 treated rats as

247

indicated by an increase in a number of neurotransmitter metabolites including glutamate, GABA,

and acetylcholine. Furthermore, Deng et al., (2012) showed an elevated level of NAA

concentrations as measured by 1H NMR spectroscopy-based metabonomics in the prefrontal cortex

of monkeys treated with an escalating dose of morphine for 90 days. Together these findings

suggest BCM-7 exposure is causing increased neuronal activity in the frontal cortex and possibly

resulting in neuronal damage/dysregulation as indicated by increased levels of NAA a neuronal

marker of neural damage.

Metabonomic analysis also revealed differences in a number of energy metabolites between

control, A2 and A1/A2 treated animals. Creatine has a role in both energy homeostasis and

osmoregulation in the brain (Bothwell et al., 2001). It is a high energy compound necessary for

the transport of phosphate groups and for maintaining appropriate ATP levels (Andrus et al., 1998)

. In this study, creatine was found to be higher in the prefrontal cortex of A1/A2 treated animals

indicating that the presence of BCM-7 is causing alterations in brain energy metabolism. In

support of this, Deng et al., (2012) also reported increased prefrontal cortex levels of creatine

following chronic treatment with morphine in monkeys (Deng et al., 2012). Moreover the current

study showed an increase in adenosine, IMP and AMP in both A1/A2 and A2 treatment groups

which are all linked to energy homeostasis. ATP is rapidly broken down to adenosine, inosine or

inosine monophosphate (IMP) (see Figure 4.8).

248

Figure 4.23: Diagram showing the interconversion between ATP, AMP, IMP and Adenosine.

Abbreviations:ATP: Adenosine 5’-triphosphate, AMP: adenosine monophosphate, IMP: inosine

monophosphate.

Indeed, together these findings suggest both milk types alter energy metabolism which is likely to

be linked to the depressive phenotype observed in both A1/A2 and A2 treatment groups. This is

further supported by findings from result chapter 3 which showed casein-rich exposed rats who

displayed and increase in depressive-like behaviour showed metabolic alterations in products of

energy metabolism in urine.

Moreover, both A1/A2 and A2 treated animals showed changes in levels of the metabolite s-

adenosylhomocysteine which is involved in DNA methylation. DNA methylation is a reaction

involving the addition of a methylgroup to DNA and is a major factor in epigenetic changes and

subsequent gene expression (see Trivedi et al., 2014). DNA methylation is carried out by a class

of enzymes called DNA methyltransferases, which depend on the level of S-adenosylmethionine

(SAM) (Trivedi et al., 2014). The activity of SAM is dependent on the redox status of the cell

(Hondorp et al., 2004). SAM acts as a methyl donor for over 200 methylation reactions and is then

converted to s-adenosylhomocystine (SAH) an inhibitor of methylation reactions. Higher SAH in

both A1/A2 and A2 treated animals could indicate increased DNA methylation reflecting increased

ATP AMP Adenosine

Inosine IMP

ADP

249

SAM to SAH conversion or decreased DNA methylation as SAH is an inhibitor of this process.

Nevertheless, recent studies have shown abnormal DNA methylation patterns have been reported

in schizophrenia (Jakovcevski et al., 2012) and ASD (Schanen, 2006). The current results indicate

disrupted DNA methylation maybe a driving factor in the observed depressive like behaviour

displayed in both A1/A2 and A2 treated animals. Furthermore, bovine BCM-7 has been shown to

increase genome wide DNA methylation (Trivedi et al., 2014). At the transcriptional level, bovine

BCM-7 decreased expression of nine genes including genes involved in cysteine uptake which is

associated with intracellular antioxidant GSH and also SAM levels (Trivedi et al., 2014). As A2

treated animals who were not exposed to BCM-7 also showed changes in SAH it is possible

bioactive compounds derived from other variants of casein found in A2 milk maybe acting via a

similar mechanism to BCM-7 to induce changes in cell redox potential and DNA methylation.

Taken together these findings indicate potential antioxidant and epigenetic effects of bovine milk

consumption which may contribute to the depressive-like phenotype observed in animals provides

with A1/A2 and A2 milk from PND21 until PND26.

Whether the observed changes in brain activity and neurotransmitter levels observed in the A1/A2

group drive the alterations in behaviour remains to be elucidated. There is some evidence to

suggest an association between proteolytic products of casein such as BCM-7 and increased risk

of neurological diseases such as autism and schizophrenia. However the current data suggests it is

unlikely that BCM-7 in milk is driving the depressive-like behaviour observed as A2 milk treated

animals without BCM-7 exposure also displayed a similar depressive phenotype.

In summary, the current results provide evidence that consumption of milk containing either the

A2 variant of casein or both A1/A2 variants from PND21 until PND26 resulted in an increase in

depressive-like behaviour in comparison to control animals as assessed by the FST. This finding

250

indicates BCM-7 is not the component in milk responsible for mood regulation. At least as

assessed by the FST. However, differences in potential anxiety-like behaviour were observed

between the A2 and A1/A2 treatment groups as assessed by the open field test, indicating that the

presence of BCM-7 may indeed have an influence on aspects of behaviour which may not have

been reflected in the FST. Both A1/A2 and A2 treated animals showed changes in energy

metabolites in the frontal cortex a region of the brain key in emotional regulation, as well as

important changes in metabolites involved in epigenetic regulation which are likely driving the

depressive-like behaviour observed in these animals especially as epigenetic changes have been

implicated in other neurological disorders such as ASD and schizophrenia, indicating a BCM-7

independent mechanism of action in milk induced mood regulation. Moreover, the presence of

BCM-7 was found to induce neurochemical changes including higher abundance of a number of

neurotransmitters and energy metabolites in the brains of A1/A2 treated animals compared to

control and A2-fed rats, whether these metabolic changes are responsible for driving behavioural

effects remains to be elucidated. All together these results provide novel and intriguing insights

into the health effects of milk consumption in early postnatal life.

251

Chapter 5

General Discussion

252

5 General Discussion

The first two years of life represent a period of unique sensitivity during which the CNS undergoes

the most rapid development and is particularly plastic to environmental influences such as diet.

Milk is the main source of nutrition for infants during this sensitive period and in recent years

awareness has increased regarding the potential health risks associated with the consumption of

bioactive breakdown products of milk proteins. The mechanisms underlying the influence of

nutritional factors on brain development in early postnatal life and the subsequent effect on

behavioural outcomes are undoubtedly complex and remain to be fully understood. However, over

the last two decades an increasing number of studies have begun to show that the gut microbiota

can modulate brain development and behaviour. Such work has demonstrated the existence of the

gut-brain axis, a bi-directional communication network between microbes in the gut and the brain.

Interestingly, the opening of some critical periods in CNS development have been found to

coincide with the time of weaning in a number of species including humans. In addition, it has

been demonstrated that the process of stopping breast feeding is one of the major determinants for

the assembly of a healthy adult like gut microbiome, with post weaning diets rapidly altering the

composition of commensal bacteria (see Tognini, 2017). The impact of milk rich diets beyond the

standard age of weaning on host behavioural, microbial and metabolic profiles was investigated in

this thesis using a multidisciplinary approach. This approach was used to explore the potential role

of the gut-brain axis in modulating brain neurochemistry and behaviour and its relationship with

weaning time.

253

5.1 Behavioural Effects of Casein Exposure

The behavioural results obtained in this thesis provide clear evidence that manipulations to diet

during the weaning period in the form of prolonged exposure to milk caseins in rats results in

emotional impairments. The depressive-like phenotype which was caused as a result of

consumption of milk containing casein for 5 extra days post-weaning was observed and replicated

across all three results chapters, making it a consistent and reliable finding. Moreover, contrary to

the original hypothesis of this thesis, it was shown that A1 beta casein derived BCM-7 is not the

only casein-derived component which could be mediating behavioural alterations reported, as

animals provided with A2 milk devoid of BCM-7 also displayed increased depressive like

behaviour. Indeed, casein exists in other variants besides beta casein, which have also been shown

to produce bioactive breakdown products (Teschemacher et al., 1997; Nyberg et al., 2013). These

bioactive peptides alongside A2 beta casein derived BCM-7 are likely to be responsible for the

emotional deficits reported following prolonged exposure to milk caseins in early postnatal life.

Such findings have wide reaching implications in regards to infant feeding. In today’s society

prolonged breastfeeding is promoted as advantageous to child development with no clear

guidelines as to when a mother should completely stop breastfeeding her young. The World Health

Organization recommends that infants should be exclusively breastfed for the first 6 months of life

(WHO, 2018). The American Academy of Paediatrics recommends breastfeeding for at least 12

months (Moretti, 2012). Recently the Academy of Nutrition and Dietetics also stated that

breastfeeding with complementary foods from six until at least 12 months of age is the ideal

feeding pattern for infants (Lessen et al., 2015). However, the casein content in human milk has

no fixed ratio and has been shown to increase in proportion to whey protein during late lactation

(Kunz et al., 1992). Therefore, given the evidence presented in this thesis, prolonged breastfeeding

254

periods warrant a closer look due to casein exposure and its potential health effects to the

developing infant.

Moreover, infant formulas are often used post weaning or as a replacement/supplement to

breastfeeding and provide a means of extending or increasing exposure to milk caseins in early

postnatal life. Globally, only 38% of infants are exclusively breastfed. In the United States for

example, by 3 months of age 67% of infants rely on infant formula for some portion of their

nutrition (Moretti, 2012). Although infant formulas are intended as an effective

substitute/supplement for infant feeding, the production of formula milk identical in composition

to breast milk has not been achieved. This is despite much research into mimicking the profile of

human breast milk for normal growth and development. Cow’s milk or soymilk are the most

commonly used bases for infant formulas. Given the health implications of prolonged exposure to

bovine caseins provided in this thesis, the use of infant formulas based on bovine protein in early

postnatal life requires attention. Moreover, the studies in this thesis used soy protein as a

replacement for casein in ‘casein free’ milk, which was intended to be the control milk type.

However rats fed casein free milk (containing approx. 20% soy protein) for 5 extra days post

weaning displayed significant neurochemical changes in a number of brain regions despite not

eliciting behavioural changes as assessed by the FST. Indeed soy protein has been found to release

soymorphins, which like beta casomorphins released from casein have an affinity for opioid

receptors. These findings raise the need to consider the health effects associated with the release

of soymorphins from soy based infant formulas on brain development in early life. Collectively

the behavioural results from this thesis highlight a need for the development of safer infant

formulas, which do not contain potent bioactive protein break down products, which may elicit

and effect on brain development and subsequent behaviour in early life.

255

Furthermore, currently A2 milk is marketed and sold in the United Kingdom amongst other

developed countries, as the safer alternative to standard cow’s milk as it does not contain A1 beta

casein derived BCM-7 (Woodford, 2007; The A2 Milk Company, 2014). However, the

behavioural findings from results Chapters 2, 3 and 4 together suggest prolonged casein

consumption, independent of A1 or A2 variants, result in behavioural impairments. Over the past

two years there has been a substantial decline in cow’s milk consumption globally, this is reflected

in increasing consumption of plant based milk beverages such as coconut and almond milk, with

worldwide sales of non-dairy milk alternatives more than doubling between 2009 and 2015

(Whipp and Daneshkhu, 2016). This shift has largely been due to health risks associated with the

cow’s milk exposure such as gastrointestinal problems from lactose intolerance (Thorning et al.,

2016). The data obtained in this thesis add to the health concerns linked to dairy-milk consumption,

highlighting the potential development of neurological disorders. Together, such evidence

reinforces the push for plant based infant formulas as a safer alternative or supplement to

breastfeeding.

5.2 Effect of Casein Exposure on Neurochemistry

Neurochemical analysis using autoradiographic binding techniques in results Chapter 2 showed

significant alterations in brain neurochemistry in the form of downregulations to both OTr and

MOPr in key regions of the brain associated with mood regulation as a result of casein consumption

beyond the standard age of weaning, which were concomitant with the increase in depressive like

behaviour observed. The exact mechanisms by which casein exposure is causing the behavioural

and neurochemical changes reported still remain to be fully elucidated. However metabonomic

data from Chapter 4 revealed critical clues in the form of an increase in the metabolite SAH in the

frontal cortex of rats provided with milk rich diets from PND21 until PND26 compared to control

256

animals weaned normally on PND21. SAH is a key component involved in DNA methylation, a

major factor in epigenetic changes and subsequent gene expression (see Trivedi et al., 2014). This

finding suggests mechanisms of epigenetic gene regulation are likely to be involved in mediating

the down regulation of MOPr and OTr reported following prolonged casein exposure. Indeed, this

theory is further reinforced by recent evidence which suggests that food derived opioid peptides

such as BCM-7 can induce genome wide DNA methylation and hence are able to exert epigenetic

changes (Trivedi et al., 2014, 2015; Deth et al., 2016). This may be of particular importance during

early postnatal development when the brain is sensitive to environmental influences. Moreover,

abnormal DNA methylation patterns have been reported in neurodevelopmental disorders such as

ASD (Schanen, 2006), further supporting a casein induced epigenetic mode of action in mediating

the neurochemical and emotional impairments observed. Moreover, these findings corroborate

clinical studies showing a positive effect associated with casein-free diets on children suffering

from neurodevelopmental disorders such as autism (Whiteley et al., 2010).

5.3 Behavioural Modulation by Gut Microbiota

The gut-brain axis is a communication system which includes neural, immunological and

hormonal signalling between the gut microbiota and the brain, providing a potential route by which

intestinal bacteria and their metabolites can influence the brain and subsequent behaviour. The

literature cited in Chapter 1 (Section 1.5.4) provide strong evidence showing the ability of the gut

microbiota in influencing brain development and behaviour. Indeed results obtained in this thesis

demonstrate that the casein induced increase in depressive like behaviour was concomitant with a

significant increase in bacteria belonging to C. histolyticum group. This group of bacteria are

recognised as toxin-producing bacteria that can exert systemic effects. Moreover, results Chapter

3 demonstrated that AB induced knock-down of the gut microbiome during the weaning period

257

resulted in significant alterations to behaviour as assessed by the FST, confirming that the gut

microbiota does indeed play a key role in mediating behaviour.

These findings are fascinating and raise concerns regarding the use of AB in early postnatal life in

infants. Currently, research regarding the use of AB in infants and its effect on behavioural

development is lacking. Antibacterial agents are frequently prescribed worldwide in infants

(McCaig et al., 2002) despite a substantial number of expert guidelines advocating more limited

use due to antibiotic resistance (Finkelstein et al., 2003; Thompson et al., 2008). Furthermore,

with growing appreciation regarding the role of the gut microbiome in many aspects of human

physiology including metabolism and immune responses, knowledge of the importance of the

microbiome in human development raises new issues about antibiotic use in infants as such

exposures may disrupt the normal development of microbial ecology (Bedford et al., 2006). Unlike

the microbiota of adults which appears to be relatively stable, the microbiota of infants is

considerably more variable and more vulnerable to antibiotic perturbation. Indeed, increased risk

from childhood AB exposure to atopic dermatitis, asthma as well as inflammatory bowel diseases

has been reported. The results obtained in Chapter 3 add to this growing body of evidence and

raise concerns regarding AB exposure in infants and its potential effects on behaviour.

5.4 Metabolic Impact of casein Exposure

A systems biology approach using NMR-based metabonomics was used to highlight metabolic

variations associated with casein rich milk diets as well as AB exposure in early postnatal life. The

data obtained shows changes in a number of metabolites associated with bacterial metabolism, as

well as choline and energy metabolism following prolonged casein exposure, most of which have

been identified as biomarkers of depressive disorders. Moreover, consistent with changes in

258

microbial composition following AB treatment in Chapter 3, the use of NMR-based metabonomics

techniques demonstrated significant changes in a number of gut microbial metabolites, further

implicating the role of gut microbiota and its metabolites in mood regulation during the critical

developmental period in early life. This is consistent with studies showing an association between

gut conditions such as IBS and mood alterations (Gros et al., 2009). Whether the metabolic

pathways found to be disrupted in these studies are a direct cause of the emotional changes reported

or a product of them remains unknown. Nonetheless, these findings provide useful evidence of

potential metabolic pathways involved in the aetiology of CNS disorders, which in future may

enable the design of drugs that specifically target these pathways.

Interestingly, metabolic analysis also revealed that BCM-7 exposure through consumption of milk

containing the A1 variant of casein induced higher neuronal activity in the frontal cortex, compared

to animals provided with A2 milk containing no BCM-7 and control treated animals. This was

shown in the form of increased GABAergic and glutamergic metabolites as well as higher

expression of a marker of neuronal damage (Baslow, 2003) which has been shown to be elevated

in several degenerative neurological disorders (see Miller, 1991) in BCM-7 exposed rats. These

findings are interesting and support evidence from parents of children suffering from ASD who

have reported improved symptoms following consumption of A2 milk. This indicates that A2 milk

may indeed confer some health benefits over the standard milk type found in western countries

today.

5.5 Study Limitations

259

The FST was the only behavioural test used for depression in this thesis. Depression is a

complex phenotype, other tests such as the tail suspension test could be used to confirm

depression.

Although consumption of casein free milk did not elicit an increase in depressive-like

behaviour as assessed by the FST, the neurochemical changes induced by casein free milk

may be driving other behavioural phenotypes such as anxiety or social interaction which

were not measured.

Although no difference in depressive-like behaviour was observed between A1/A2 and A2

treated animals, differences in brain metabolic profiles between the two groups could be

driving other behavioural differences such as sociability or anxiety which were not tested.

As locomotion was not assessed prior to the FST in results Chapter 3, it is not possible to

exclude the possibility that the increase in swimming behaviour observed in the FST

following AB treatment is not driven by differences in motor activity.

5.6 Future Studies

It would be beneficial for future studies to measure the persistence of the observed

behavioural deficits following prolonged casein exposure into adulthood either after

cessation of casein consumption following delayed weaning or following continued

consumption of milk caseins into adulthood in rats, which is more reflective of milk

drinking practices in most western societies today. Especially as studies show that

environmental influences in early life can have long lasting effects into adulthood including

development of neuropsychiatric conditions.

Given the role of the immune system in mediating communication between the gut brain

axis, future studies should measure inflammatory markers following casein and AB

260

exposure in early life to shed light on potential mechanism by which the gut and the brain

could be communicating in mediating behavioural alterations.

The exact casein breakdown products driving the behavioural neurochemical metabolic

and microbial changes observed in these studies are yet to be identified. Measurement of

behaviour using a battery of behavioural tests following i.v injections of BCM-7 and other

casein derived breakdown products could help answer this question.

The findings in this thesis highlight the need for human studies to answer the effect of

prolonged casein exposure on the gut microbiome and behaviour. One proposed study

could assess behavioural and microbial differences in infants from countries like China

which typically do not use milk beyond the age of weaning compared to infants in western

countries in which milk rich diets are popular. Also, epidemiological studies assessing

durations of breast feeding and associated behavioural development would be of interest.

261

5.7 Conclusion

In summary the findings from this thesis provide strong evidence that dietary manipulation during

the sensitive developmental period in early postnatal life result in behavioural, neurochemical, gut

microbial and metabolic alterations and raise concerns regarding early life factors and their

potential harmful effects on mood disorders. In particular, it was shown that prolonged

consumption of milk caseins beyond the normal age of weaning results in emotional impairments

and has important public health implications regarding exposure to milk caseins in infancy either

in the form of prolonged breastfeeding durations or provision of cow’s milk as strong

environmental factor influencing mood disorders. These public health considerations should

include comprehensive guidelines on weaning practises, consumption of cow’s milk and other

food products such as soy which may release bioactive breakdown products, as well as exposure

to antibiotics in early postnatal life, all of which have been shown to have potential implications

in the pathogenesis of paediatric psychiatric illnesses. In utilising a multidisciplinary approach the

work described here demonstrates that brain development and subsequent mood are influenced by

changes in gut microbial profile and metabolic pathways as well as potential epigenetic alterations.

Future work should aim to investigate direct mechanisms by which the microbiome-gut-brain axis

communication influences behaviour during early life development as well as direct mechanisms

by which casein could be causing alterations to the gut microbiome and brain neurochemistry.

262

Bibliography

Abbott, N. J. and Friedman, A. (2012) ‘Overview and introduction: The blood-brain barrier in

health and disease’, Epilepsia, 53, pp. 1–6.

Abel, E. L. (1994) ‘A further analysis of physiological changes in rats in the forced swim test’,

Physiology & Behavior, 56(4), pp. 795–800.

Adcock, J. D. and Hettig, R. A. (1946) ‘Absorption, distribution and excretion of streptomycin.’,

Archives of internal medicine (Chicago, Ill. : 1908), 77, pp. 179–95.

Adriani, W. and Laviola, G. (2002) ‘Spontaneous novelty seeking and amphetamine-induced

conditioning and sensitization in adult mice: evidence of dissociation as a function of age at

weaning.’, Neuropsychopharmacology : official publication of the American College of

Neuropsychopharmacology, 27(2), pp. 225–36.

El Aidy, S., Dinan, T. G. and Cryan, J. F. (2015) ‘Gut Microbiota: The Conductor in the Orchestra

of Immune–Neuroendocrine Communication’, Clinical Therapeutics, 37(5), pp. 954–967.

al-Waiz, M., Mikov, M., Mitchell, S. C. and Smith, R. L. (1992) ‘The exogenous origin of

trimethylamine in the mouse.’, Metabolism: clinical and experimental, 41(2), pp. 135–6.

Allen, P. J. (2012) ‘Creatine metabolism and psychiatric disorders: Does creatine supplementation

have therapeutic value?’, Neuroscience and biobehavioral reviews, 36(5), pp. 1442–62.

Allen, P. J., D’Anci, K. E., Kanarek, R. B. and Renshaw, P. F. (2010) ‘Chronic creatine

supplementation alters depression-like behavior in rodents in a sex-dependent manner.’,

263

Neuropsychopharmacology : official publication of the American College of


Alonso, E., Garrido, E., Díez-Fernández, C., Pérez-García, C., Herradón, G., Ezquerra, L., Deuel,

T. F. and Alguacil, L. F. (2007) ‘Yohimbine prevents morphine-induced changes of glial fibrillary

acidic protein in brainstem and α2-adrenoceptor gene expression in hippocampus’, Neuroscience

Letters, 412(2), pp. 163–167.

Amico, J. A., Mantella, R. C., Vollmer, R. R. and Li, X. (2004) ‘Anxiety and Stress Responses in

Female Oxytocin Deficient Mice’, Journal of Neuroendocrinology, 16(4), pp. 319–324.

An, C., Kuda, T., Yazaki, T., Takahashi, H. and Kimura, B. (2014) ‘Caecal fermentation,

putrefaction and microbiotas in rats fed milk casein, soy protein or fish meal’, Applied

Microbiology and Biotechnology, 98(6), pp. 2779–2787.

Andersen, S. L. (2003) ‘Trajectories of brain development: Point of vulnerability or window of

opportunity? Trajectories of brain development: Point of vulnerability or window of opportunity?’,

Neuroscience & Biobehavioral Reviews, 27(1–2), pp. 3–18.

Anderson, M. J. (2006) Tasks and techniques : a sampling of the methodologies for the

investigation of animal learning, behavior, and cognition.

Andiran, F., Dayi, S. and Mete, E. (2003) ‘Cows milk consumption in constipation and anal fissure

in infants and young children.’, Journal of paediatrics and child health, 39(5), pp. 329–31.

Andrus, P. K., Fleck, T. J., Gurney, M. E. and Hall, E. D. (1998) ‘Protein oxidative damage in a

transgenic mouse model of familial amyotrophic lateral sclerosis.’, Journal of neurochemistry,

71(5), pp. 2041–8.

264

Arletti, R. and Bertolini, A. (1987) ‘Oxytocin acts as an antidepressant in two animal models of

depression.’, Life sciences, 41(14), pp. 1725–30.

Arseneault-Bréard, J., Rondeau, I., Gilbert, K., Girard, S.-A., Tompkins, T. A., Godbout, R. and

Rousseau, G. (2012) ‘Combination of Lactobacillus helveticus R0052 and Bifidobacterium

longum R0175 reduces post-myocardial infarction depression symptoms and restores intestinal

permeability in a rat model’, British Journal of Nutrition, 107(12), pp. 1793–1799.

Autry, A. E. and Monteggia, L. M. (2012) ‘Brain-Derived Neurotrophic Factor and

Neuropsychiatric Disorders’, Pharmacological Reviews, 64(2), pp. 238–258.

Bäckhed, F. (2011) ‘Programming of Host Metabolism by the Gut Microbiota’, Annals of Nutrition

and Metabolism, 58(2), pp. 44–52.

Bailey, C. P., Oldfield, S., Llorente, J., Caunt, C. J., Teschemacher, A. G., Roberts, L., McArdle,

C. A., Smith, F. L., Dewey, W. L., Kelly, E. and Henderson, G. (2009) ‘Involvement of PKC alpha

and G-protein-coupled receptor kinase 2 in agonist-selective desensitization of mu-opioid

receptors in mature brain neurons.’, British journal of pharmacology, 158(1), pp. 157–64.

Bailey, M. T., Dowd, S. E., Galley, J. D., Hufnagle, A. R., Allen, R. G. and Lyte, M. (2011)

‘Exposure to a social stressor alters the structure of the intestinal microbiota: implications for

stressor-induced immunomodulation.’, Brain, behavior, and immunity, 25(3), pp. 397–407.

Bale, T. L., Baram, T. Z., Brown, A. S., Goldstein, J. M., Insel, T. R., McCarthy, M. M., Nemeroff,

C. B., Reyes, T. M., Simerly, R. B., Susser, E. S. and Nestler, E. J. (2010) ‘Early Life Programming

and Neurodevelopmental Disorders’, Biological Psychiatry, 68(4), pp. 314–319.

Bale, T. L., Davis, A. M., Auger, A. P., Dorsa, D. M. and McCarthy, M. M. (2001) ‘CNS region-

265

specific oxytocin receptor expression: importance in regulation of anxiety and sex behavior.’, The

Journal of neuroscience : the official journal of the Society for Neuroscience, 21(7), pp. 2546–52.

Bales, K. L., Boone, E., Epperson, P., Hoffman, G. and Carter, C. S. (2011) ‘Are behavioral effects

of early experience mediated by oxytocin?’, Frontiers in Psychiatry, 2(5), p. 24.

Bales, K. L. and Perkeybile, A. M. (2012) ‘Developmental experiences and the oxytocin receptor

system’, Hormones and Behavior, 61(3), pp. 313–319.

Barnett, M. P. G., McNabb, W. C., Roy, N. C., Woodford, K. B. and Clarke, A. J. (2014) ‘Dietary

A1 β -casein affects gastrointestinal transit time, dipeptidyl peptidase-4 activity, and inflammatory

status relative to A2 β -casein in Wistar rats’, International Journal of Food Sciences and

Nutrition, 65(6), pp. 720–727.

Barnett, M. P. G., McNabb, W. C., Roy, N. C., Woodford, K. B. and Clarke, A. J. (2014) ‘Dietary

A1 β -casein affects gastrointestinal transit time, dipeptidyl peptidase-4 activity, and inflammatory

status relative to A2 β -casein in Wistar rats’, International Journal of Food Sciences and

Nutrition, 65(6), pp. 720–727.

Barrett, E., Fitzgerald, P., Dinan, T. G., Cryan, J. F., Ross, R. P., Quigley, E. M., Shanahan, F.,

Kiely, B., Fitzgerald, G. F., O’Toole, P. W. and Stanton, C. (2012) ‘Bifidobacterium breve with

α-Linolenic Acid and Linoleic Acid Alters Fatty Acid Metabolism in the Maternal Separation

Model of Irritable Bowel Syndrome’, PLoS ONE, 7(11), p. e48159.

Barth, C. A., Pfeuffer, M. and Hahn, G. (1984) ‘Influence of dietary casein or soy protein on serum

lipids and lipoproteins of monkeys (Macaca fascicularis).’, Annals of nutrition & metabolism,

28(3), pp. 137–43.

266

Baslow, M. H. (2003) ‘N-Acetylaspartate in the Vertebrate Brain: Metabolism and Function’,

Neurochemical Research, 28(6), pp. 941–953.

Bayer, S. A., Altman, J., Russo, R. J. and Zhang, X. (1993) ‘Timetables of neurogenesis in the

human brain based on experimentally determined patterns in the rat.’, Neurotoxicology, 14(1), pp.

83–144.

Beardsley, P. M., Howard, J. L., Shelton, K. L. and Carroll, F. I. (2005) ‘Differential effects of the

novel kappa opioid receptor antagonist, JDTic, on reinstatement of cocaine-seeking induced by

footshock stressors vs cocaine primes and its antidepressant-like effects in rats’,

Psychopharmacology, 183(1), pp. 118–126.

Beckonert, O., Keun, H. C., Ebbels, T. M. D., Bundy, J., Holmes, E., Lindon, J. C. and Nicholson,

J. K. (2007) ‘Metabolic profiling, metabolomic and metabonomic procedures for NMR

spectroscopy of urine, plasma, serum and tissue extracts.’, Nature protocols, 2(11), pp. 2692–703.

Bedford Russell, A. and Murch, S. (2006) ‘Could peripartum antibiotics have delayed health

consequences for the infant?’, BJOG: An International Journal of Obstetrics and Gynaecology,

113(7), pp. 758–765.

Bell, S. J., Grochoski, G. T. and Clarke, A. J. (2006) ‘Health implications of milk containing beta-

casein with the A2 genetic variant.’, Critical reviews in food science and nutrition, 46(1), pp. 93–

100.

Bercik, P. (2011) ‘The microbiota-gut-brain axis: learning from intestinal bacteria?’, Gut, 60(3),

pp. 288–289.

Bercik, P. and Collins, S. M. (2014) ‘The Effects of Inflammation, Infection and Antibiotics on

267

the Microbiota-Gut-Brain Axis’, in Advances in experimental medicine and biology, pp. 279–289.

Bercik, P., Collins, S. M. and Verdu, E. F. (2012) ‘Microbes and the gut-brain axis’,

Neurogastroenterology & Motility, 24(5), pp. 405–413.

Bercik, P., Denou, E., Collins, J., Jackson, W., Lu, J., Jury, J., Deng, Y., Blennerhassett, P., Macri,

J., McCoy, K. D., Verdu, E. F. and Collins, S. M. (2011a) ‘The intestinal microbiota affect central

levels of brain-derived neurotropic factor and behavior in mice.’, Gastroenterology, 141(2), pp.

599–609, 609–3.

Bercik, P., Denou, E., Collins, J., Jackson, W., Lu, J., Jury, J., Deng, Y., Blennerhassett, P., Macri,

J., McCoy, K. D., Verdu, E. F. and Collins, S. M. (2011b) ‘The intestinal microbiota affect central

levels of brain-derived neurotropic factor and behavior in mice.’, Gastroenterology, 141(2), pp.

599–609.

Bercik, P., Park, A. J., Sinclair, D., Khoshdel, A., Lu, J., Huang, X., Deng, Y., Blennerhassett, P.

A., Fahnestock, M., Moine, D., Berger, B., Huizinga, J. D., Kunze, W., McLean, P. G.,

Bergonzelli, G. E., Collins, S. M. and Verdu, E. F. (2011) ‘The anxiolytic effect of Bifidobacterium

longum NCC3001 involves vagal pathways for gut-brain communication.’,

Neurogastroenterology and motility : the official journal of the European Gastrointestinal Motility

Society, 23(12), pp. 1132–9.

Bercik, P., Verdu, E. F., Foster, J. A., Macri, J., Potter, M., Huang, X., Malinowski, P., Jackson,

W., Blennerhassett, P., Neufeld, K. A., Lu, J., Khan, W. I., Corthesy–Theulaz, I., Cherbut, C.,

Bergonzelli, G. E. and Collins, S. M. (2010) ‘Chronic Gastrointestinal Inflammation Induces

Anxiety-Like Behavior and Alters Central Nervous System Biochemistry in Mice’,

Gastroenterology, 139(6), pp. 2102–2112.

268

De Biase, D. and Pennacchietti, E. (2012) ‘Glutamate decarboxylase-dependent acid resistance in

orally acquired bacteria: function, distribution and biomedical implications of the gadBC operon’,

Molecular Microbiology, 86(4), pp. 770–786.

Bilbo, S. D., Levkoff, L. H., Mahoney, J. H., Watkins, L. R., Rudy, J. W. and Maier, S. F. (2005)

‘Neonatal infection induces memory impairments following an immune challenge in adulthood.’,

Behavioral neuroscience, 119(1), pp. 293–301.

Binns, C., Lee, M. and Low, W. Y. (2016) ‘The Long-Term Public Health Benefits of

Breastfeeding’, Asia Pacific Journal of Public Health, 28(1), pp. 7–14.

Blume, A., Bosch, O. J., Miklos, S., Torner, L., Wales, L., Waldherr, M. and Neumann, I. D.

(2008) ‘Oxytocin reduces anxiety via ERK1/2 activation: local effect within the rat hypothalamic

paraventricular nucleus’, European Journal of Neuroscience, 27(8), pp. 1947–1956.

Boccuto, L., Chen, C.-F., Pittman, A. R., Skinner, C. D., McCartney, H. J., Jones, K., Bochner, B.

R., Stevenson, R. E. and Schwartz, C. E. (2013) ‘Decreased tryptophan metabolism in patients

with autism spectrum disorders’, Molecular Autism, 4(1), p. 16.

Bodnar, R. J. (1990) ‘Effects of opioid peptides on peripheral stimulation and “stress”-induced

analgesia in animals’, Crit Rev Neurobiol, 6(1), pp. 39–49.

Bodnar, R. J. (2013) ‘Endogenous opiates and behaviour’, Peptides, 50, pp. 55–95.

Bogdanova, O. V, Kanekar, S., D ’anci, K. E. and Renshaw, P. F. (2013) ‘Factors influencing

behavior in the forced swim test HHS Public Access’, Physiol Behav. June, 13(118).

Borre, Y. E., O’Keeffe, G. W., Clarke, G., Stanton, C., Dinan, T. G. and Cryan, J. F. (2014)

‘Microbiota and neurodevelopmental windows: implications for brain disorders’, Trends in

269

Molecular Medicine, 20(9), pp. 509–518.

Borre, Y. E., O’Keeffe, G. W., Clarke, G., Stanton, C., Dinan, T. G. and Cryan, J. F. (2014)

‘Microbiota and neurodevelopmental windows: implications for brain disorders.’, Trends in

molecular medicine, 20(9), pp. 509–18.

Bosch, O. ., Krömer, S. ., Brunton, P. . and Neumann, I. . (2004) ‘Release of oxytocin in the

hypothalamic paraventricular nucleus, but not central amygdala or lateral septum in lactating

residents and virgin intruders during maternal defence’, Neuroscience, 124(2), pp. 439–448.

Bosch, O. J. and Neumann, I. D. (2012) ‘Both oxytocin and vasopressin are mediators of maternal

care and aggression in rodents: from central release to sites of action.’, Hormones and behavior,

61(3), pp. 293–303.

Bothwell, J. H., Rae, C., Dixon, R. M., Styles, P. and Bhakoo, K. K. (2001) ‘Hypo-osmotic

swelling-activated release of organic osmolytes in brain slices: implications for brain oedema in

vivo.’, Journal of neurochemistry, 77(6), pp. 1632–40.

Boutrou, R., Gaudichon, C., Dupont, D., Jardin, J., Airinei, G., Marsset-Baglieri, A., Benamouzig,

R., Tomé, D. and Leonil, J. (2013) ‘Sequential release of milk protein-derived bioactive peptides

in the jejunum in healthy humans.’, The American journal of clinical nutrition, 97(6), pp. 1314–

23.

Braniste, V., Al-Asmakh, M., Kowal, C., Anuar, F., Abbaspour, A., Tóth, M., Korecka, A.,

Bakocevic, N., Ng, L. G., Guan, N. L., Kundu, P., Gulyás, B., Halldin, C., Hultenby, K., Nilsson,

H., Hebert, H., Volpe, B. T., Diamond, B. and Pettersson, S. (2014) ‘The gut microbiota influences

blood-brain barrier permeability in mice.’, Science translational medicine, 6(263), p. 158.

270

Brantl, V. (1984) ‘Novel opioid peptides derived from human beta-casein: human beta-

casomorphins.’, European journal of pharmacology, 106(1), pp. 213–4.

Brantl, V., Pfeiffer, A., Herz, A., Henschen, A. and Lottspeich, F. (1982) ‘Antinociceptive

potencies of beta-casomorphin analogs as compared to their affinities towards mu and delta opiate

receptor sites in brain and periphery.’, Peptides, 3(5), pp. 793–7.

Brantl, V., Teschemacher, H., Bläsig, J., Henschen, A. and Lottspeich, F. (1981) ‘Opioid activities

of β-casomorphins’, Life Sciences, 28(17), pp. 1903–1909.

Bravo, J. A., Forsythe, P., Chew, M. V, Escaravage, E., Savignac, H. M., Dinan, T. G.,

Bienenstock, J. and Cryan, J. F. (2011) ‘Ingestion of Lactobacillus strain regulates emotional

behavior and central GABA receptor expression in a mouse via the vagus nerve.’, Proceedings of

the National Academy of Sciences of the United States of America, 108(38), pp. 16050–5.

Brenes, J. C., Rodríguez, O. and Fornaguera, J. (2008) ‘Differential effect of environment

enrichment and social isolation on depressive-like behavior, spontaneous activity and serotonin

and norepinephrine concentration in prefrontal cortex and ventral striatum’, Pharmacology

Biochemistry and Behavior, 89, pp. 85–93.

Bryan, J., Osendarp, S., Hughes, D., Calvaresi, E., Baghurst, K. and van Klinken, J.-W. (2004)

‘Nutrients for cognitive development in school-aged children’, Nutrition reviews, 62(8), pp. 295–

306.

Buie, T. (2013) ‘The Relationship of Autism and Gluten’, Clinical Therapeutics, 35(5), pp. 578–

583.

Burokas, A., Moloney, R. D., Dinan, T. G. and Cryan, J. F. (2015) ‘Microbiota Regulation of the

271

Mammalian Gut e Brain Axis’, Advances in Applied Microbiology, pp. 1–62. doi:

10.1016/bs.aambs.2015.02.001.

Candon, S., Perez-Arroyo, A., Marquet, C., Valette, F., Foray, A.-P., Pelletier, B., Milani, C.,

Ventura, M., Bach, J.-F., Chatenoud, L. and Chatenoud, L. (2015) ‘Antibiotics in Early Life Alter

the Gut Microbiome and Increase Disease Incidence in a Spontaneous Mouse Model of

Autoimmune Insulin-Dependent Diabetes’, PLOS ONE, 10(5), p. 125448.

Carlyle, B. C., Duque, A., Kitchen, R. R., Bordner, K. A., Coman, D., Doolittle, E., Papademetris,

X., Hyder, F., Taylor, J. R. and Simen, A. A. (2012) ‘Maternal separation with early weaning: A

rodent model providing novel insights into neglect associated developmental deficits’,

Development and Psychopathology, 24(4), pp. 1401–1416.

Caroli, A. M., Chessa, S. and Erhardt, G. J. (2009) ‘Invited review: Milk protein polymorphisms

in cattle: Effect on animal breeding and human nutrition’, Journal of Dairy Science, 92(11), pp.

5335–5352.

Carter, C. S. (1992) ‘Oxytocin and sexual behavior.’, Neuroscience and biobehavioral reviews,

16(2), pp. 131–44.

Carter, C. S. (2003) ‘Developmental consequences of oxytocin.’, Physiology & behavior, 79(3),

pp. 383–97.

Cazzullo, A. G., Musetti, M. C., Musetti, L., Bajo, S., Sacerdote, P. and Panerai, A. (1999) ‘Beta-

endorphin levels in peripheral blood mononuclear cells and long-term naltrexone treatment in

autistic children.’, European neuropsychopharmacology : the journal of the European College of


272

Charbogne, P., Gardon, O., Martín-García, E., Keyworth, H. L., Matsui, A., Mechling, A. E.,

Bienert, T., Nasseef, T., Robé, A., Moquin, L., Darcq, E., Ben Hamida, S., Robledo, P., Matifas,

A., Befort, K., Gavériaux-Ruff, C., Harsan, L.-A., von Elverfeldt, D., Hennig, J., et al. (2017) ‘Mu

Opioid Receptors in Gamma-Aminobutyric Acidergic Forebrain Neurons Moderate Motivation

for Heroin and Palatable Food.’, Biological psychiatry, 81(9), pp. 778–788.

Chia, J. S. J., McRae, J. L., Kukuljan, S., Woodford, K., Elliott, R. B., Swinburn, B. and Dwyer,

K. M. (2017) ‘A1 beta-casein milk protein and other environmental pre-disposing factors for type

1 diabetes’, Nutrition & Diabetes, 7(5), p. 274.

Chiu, C.-Y., Liao, S.-L., Su, K.-W., Tsai, M.-H., Hua, M.-C., Lai, S.-H., Chen, L.-C., Yao, T.-C.,

Yeh, K.-W. and Huang, J.-L. (2016) ‘Exclusive or Partial Breastfeeding for 6 Months Is

Associated With Reduced Milk Sensitization and Risk of Eczema in Early Childhood: The PATCH

Birth Cohort Study.’, Medicine, 95(15), p. 3391.

Chu Sin Chung, P., Keyworth, H. L., Martin-Garcia, E., Charbogne, P., Darcq, E., Bailey, A.,

Filliol, D., Matifas, A., Scherrer, G., Ouagazzal, A.-M., Gaveriaux-Ruff, C., Befort, K.,

Maldonado, R., Kitchen, I. and Kieffer, B. L. (2015) ‘A Novel Anxiogenic Role for the Delta

Opioid Receptor Expressed in GABAergic Forebrain Neurons’, Biological Psychiatry, 77(4), pp.

404–415.

Cieślińska, A., Kostyra, E., Kostyra, H., Oleński, K., Fiedorowicz, E. and Kamiński, S. (2012)

‘Milk from cows of different β-casein genotypes as a source of β-casomorphin-7’, International

Journal of Food Sciences and Nutrition, 63(4), pp. 426–430.

Cieslinska, A., Sienkiewicz-Szlapka, E., Wasilewska, J., Fiedorowicz, E., Chwala, B., Moszynska-

Dumara, M., Cieslinski, T., Bukalo, M. and Kostyra, E. (2015) ‘Influence of candidate

273

polymorphisms on the dipeptidyl peptidase IV and mu-opioid receptor genes expression in aspect

of the beta-casomorphin-7 modulation functions in autism’, Peptides, 65, pp. 6–11. doi:

10.1016/j.peptides.2014.11.012.

Clancy, B., Finlay, B. L., Darlington, R. B. and Anand, K. J. S. (2007) ‘Extrapolating brain

development from experimental species to humans.’, Neurotoxicology, 28(5), pp. 931–7.

Clarke, G., Grenham, S., Scully, P., Fitzgerald, P., Moloney, R. D., Shanahan, F., Dinan, T. G. and

Cryan, J. F. (2013) ‘The microbiome-gut-brain axis during early life regulates the hippocampal

serotonergic system in a sex-dependent manner.’, Molecular psychiatry, 18(6), pp. 666–73.

Clarke, G., Wood, P., Merrick, L. and Lincoln, D. W. (1979) ‘Opiate inhibition of peptide release

from the neurohumoral terminals of hypothalamic neurones.’, Nature, 282(5740), pp. 746–8.

Claus, S. P., Tsang, T. M., Wang, Y., Cloarec, O., Skordi, E., Martin, F. P., Rezzi, S., Ross, A.,

Kochhar, S., Holmes, E. and Nicholson, J. K. (2008) ‘Systemic multicompartmental effects of the

gut microbiome on mouse metabolic phenotypes’, Molecular Systems Biology, 4, p. 219.

Collins, S. M. and Bercik, P. (2009) ‘The relationship between intestinal microbiota and the central

nervous system in normal gastrointestinal function and disease.’, Gastroenterology, 136(6), pp.

2003–14.

Collins, S., Verdu, E., Denou, E. and Bercik, P. (2009) ‘The Role of Pathogenic Microbes and

Commensal Bacteria in Irritable Bowel Syndrome’, Digestive Diseases, 27(1), pp. 85–89.

Cook, C. J. (1999) ‘Patterns of weaning and adult response to stress.’, Physiology & behavior,

67(5), pp. 803–8.

Corbett, A. D., Henderson, G., McKnight, A. T. and Paterson, S. J. (2009) ‘75 years of opioid

274

research: the exciting but vain quest for the Holy Grail’, British Journal of Pharmacology,

147(S1), pp. S153–S162.

Cowan, R. L., Sesack, S. R., Van Bockstaele, E. J., Branchereau, P., Chan, J. and Pickel, V. M.

(1994) ‘Analysis of synaptic inputs and targets of physiologically characterized neurons in rat

frontal cortex: Combined in vivo intracellular recording and immunolabeling’, Synapse, 17(2), pp.

101–114.

Cramer, C. P., Thiels, E. and Alberts, J. R. (1990) ‘Weaning in rats: I. Maternal behavior’,

Developmental Psychobiology, 23(6), pp. 479–493.

Crawley, J. N. (2008) ‘Behavioral Phenotyping Strategies for Mutant Mice’, Neuron, 57(6), pp.

809–818.

Crumeyrolle-Arias, M., Jaglin, M., Bruneau, A., Vancassel, S., Cardona, A., Dauge, V., Naudon,

L. and Rabot, S. (2014) ‘Absence of the gut microbiota enhances anxiety-like behavior and

neuroendocrine response to acute stress in rats’, Psychoneuroendocrinology, 42, pp. 207–217.

Cryan, J. F. and Dinan, T. G. (2012) ‘Mind-altering microorganisms: the impact of the gut

microbiota on brain and behaviour.’, Nature reviews. Neuroscience, 13(10), pp. 701–12.

Cryan, J. F. and Kaupmann, K. (2005) ‘Don’t worry “B” happy!: a role for GABA(B) receptors in

anxiety and depression.’, Trends in pharmacological sciences, 26(1), pp. 36–43.

Cryan, J. F., Markou, A. and Lucki, I. (2002) ‘Assessing antidepressant activity in rodents: recent

developments and future needs.’, Trends in pharmacological sciences, 23(5), pp. 238–45.

Cryan, J. F. and O’Mahony, S. M. (2011) ‘The microbiome-gut-brain axis: from bowel to

behavior’, Neurogastroenterology & Motility, 23(3), pp. 187–192.

275

Curley, J. P., Jordan, E. R., Swaney, W. T., Izraelit, A., Kammel, S. and Champagne, F. A. (2009)

‘The meaning of weaning: influence of the weaning period on behavioral development in mice.’,

Developmental neuroscience, 31(4), pp. 318–31.

Daniel, H., Vohwinkel, M. and Rehner, G. (1990) ‘Effect of casein and beta-casomorphins on

gastrointestinal motility in rats.’, The Journal of nutrition, 120(3), pp. 252–7.

Dantzer, R., O’Connor, J. C., Freund, G. G., Johnson, R. W. and Kelley, K. W. (2008) ‘From

inflammation to sickness and depression: when the immune system subjugates the brain’, Nature

Reviews Neuroscience, 9(1), pp. 46–56.

Davies, D. F., Davies, J. R. and Richards, M. A. (1969) ‘Antibodies to reconstituted dried cow’s

milk protein in coronary heart disease’, Journal of Atherosclerosis Research, 9(1), pp. 103–107.

Deltheil, T., Guiard, B. P., Cerdan, J., David, D. J., Tanaka, K. F., Repérant, C., Guilloux, J.-P.,

Coudoré, F., Hen, R. and Gardier, A. M. (2008) ‘Behavioral and serotonergic consequences of

decreasing or increasing hippocampus brain-derived neurotrophic factor protein levels in mice’,

Neuropharmacology, 55(6), pp. 1006–1014.

Demmelmair, H., von Rosen, J. and Koletzko, B. (2006) ‘Long-term consequences of early

nutrition’, Early Human Development, 82(8), pp. 567–574.

Deng, Y., Bu, Q., Hu, Z., Deng, P., Yan, G., Duan, J., Hu, C., Zhou, J., Shao, X., Zhao, J., Li, Y.,

Zhu, R., Zhao, Y. and Cen, X. (2012) ‘(1) H-nuclear magnetic resonance-based metabonomic

analysis of brain in rhesus monkeys with morphine treatment and withdrawal intervention.’,

Journal of neuroscience research, 90(11), pp. 2154–62.

Desbonnet, L., Clarke, G., Traplin, A., O’Sullivan, O., Crispie, F., Moloney, R. D., Cotter, P. D.,

276

Dinan, T. G. and Cryan, J. F. (2015) ‘Gut microbiota depletion from early adolescence in mice:

Implications for brain and behaviour’, Brain, Behavior, and Immunity, 48, pp. 165–173.

Desbonnet, L., Garrett, L., Clarke, G., Bienenstock, J. and Dinan, T. G. (2008) ‘The probiotic

Bifidobacteria infantis: An assessment of potential antidepressant properties in the rat’, Journal of

Psychiatric Research, 43(2), pp. 164–174.

Desbonnet, L., Garrett, L., Clarke, G., Kiely, B., Cryan, J. F. and Dinan, T. G. (2010) ‘Effects of

the probiotic Bifidobacterium infantis in the maternal separation model of depression.’,

Neuroscience, 170(4), pp. 1179–88.

Desclee de Maredsous, C., Oozeer, R., Barbillon, P., Mary-Huard, T., Delteil, C., Blachier, F.,

Tome, D., van der Beek, E. M. and Davila, A.-M. (2016) ‘High-Protein Exposure during Gestation

or Lactation or after Weaning Has a Period-Specific Signature on Rat Pup Weight, Adiposity,

Food Intake, and Glucose Homeostasis up to 6 Weeks of Age’, Journal of Nutrition, 146(1), pp.

21–29.

Deth, R., Clarke, A., Ni, J. and Trivedi, M. (2016) ‘Clinical evaluation of glutathione

concentrations after consumption of milk containing different subtypes of beta-casein: results from

a randomized, cross-over clinical trial’, Nutrition Journal, 15(1), p. 82.

Dhawan, B. N., Cesselin, F., Raghubir, R., Reisine, T., Bradley, P. B., Portoghese, P. S. and

Hamon, M. (1996) ‘International Union of Pharmacology. XII. Classification of opioid receptors.’,

Pharmacological reviews, 48(4), pp. 567–92.

Diaz, J., Moore, E., Petracca, F., Schacher, J. and Stamper, C. (1982) ‘Artificial rearing of rat pups

with a protein-enriched formula.’, The Journal of nutrition, 112(5), pp. 841–7.

277

Dinan, T. G. and Cryan, J. F. (2013) ‘Melancholic microbes: a link between gut microbiota and

depression?’, Neurogastroenterology and motility : the official journal of the European

Gastrointestinal Motility Society, 25(9), pp. 713–9.

Dinan, T. G., Stanton, C. and Cryan, J. F. (2013) ‘Psychobiotics: A Novel Class of Psychotropic’,

Biological Psychiatry, 74(10), pp. 720–726.

Do, K. Q., Trabesinger, A. H., Kirsten-Krüger, M., Lauer, C. J., Dydak, U., Hell, D., Holsboer, F.,

Boesiger, P. and Cuénod, M. (2000) ‘Schizophrenia: glutathione deficit in cerebrospinal fluid and

prefrontal cortex in vivo.’, The European journal of neuroscience, 12(10), pp. 3721–8.

Dohan, F. C. (1966) ‘Cereals and schizophrenia data and hypothesis.’, Acta psychiatrica

Scandinavica, 42(2), pp. 125–52.

Donovan, S. M. (2006) ‘Role of human milk components in gastrointestinal development: Current

knowledge and future NEEDS’, The Journal of Pediatrics, 149(5), pp. S49–S61.

Drai, D., Kafkafi, N., Benjamini, Y., Elmer, G. and Golani, I. (2001) ‘Rats and mice share common

ethologically relevant parameters of exploratory behavior.’, Behavioural brain research, 125(1–

2), pp. 133–40.

Drevets, W. C. (2000) ‘Functional anatomical abnormalities in limbic and prefrontal cortical

structures in major depression’, in Progress in Brain Research, pp. 413–431.

Ebner, K., Wotjak, C. T., Landgraf, R. and Engelmann, M. (2005) ‘Neuroendocrine and behavioral

response to social confrontation: residents versus intruders, active versus passive coping styles.’,

Hormones and behavior, 47(1), pp. 14–21.

Eckburg, P. B., Bik, E. M., Bernstein, C. N., Purdom, E., Dethlefsen, L., Sargent, M., Gill, S. R.,

278

Nelson, K. E. and Relman, D. A. (2005) ‘Diversity of the human intestinal microbial flora.’,

Science, 308(5728), pp. 1635–8.

Elitsur, Y. and Luk, G. D. (1991) ‘Beta-casomorphin (BCM) and human colonic lamina propria

lymphocyte proliferation.’, Clinical and experimental immunology, 85(3), pp. 493–7.

Elliott, R. B., Harris, D. P., Hill, J. P., Bibby, N. J. and Wasmuth, H. E. (1999) ‘Type I (insulin-

dependent) diabetes mellitus and cow milk: casein variant consumption’, Diabetologia, 42(3), pp.

292–296.

Elliott, R., Wasmuth, H. E., Bibby, N. J. and Hill, J. P. (1998) ‘The role of beta-casein varients in

the induction of insulin-dependant diabetes in the non-obese diabetic mouse and humans’, Milk

Protein Poly, pp. 445–453.

Engelmann, M., Landgraf, R. and Wotjak, C. T. (2004) ‘The hypothalamic–neurohypophysial

system regulates the hypothalamic–pituitary–adrenal axis under stress: An old concept revisited’,

Frontiers in Neuroendocrinology, 25(3–4), pp. 132–149.

Erdman, S. E. and Poutahidis, T. (2016) ‘Microbes and Oxytocin’, in International review of

neurobiology, pp. 91–126.

Ermisch, A., Ruhle, H. ‐J, Neubert, K., Hartrodt, B. and Landgraf, R. (1983) ‘On the Blood‐Brain

Barrier to Peptides: [3H]βCasomorphin‐5 Uptake by Eighteen Brain Regions In Vivo’, Journal of

Neurochemistry, 41(5), pp. 1229–1233.

Farrell, H. M., Jimenez-Flores, R., Bleck, G. T., Brown, E. M., Butler, J. E., Creamer, L. K., Hicks,

C. L., Hollar, C. M., Ng-Kwai-Hang, K. F. and Swaisgood, H. E. (2004) ‘Nomenclature of the

Proteins of Cows’ Milk—Sixth Revision’, Journal of Dairy Science, 87(6), pp. 1641–1674.

279

Farshim, P., Walton, G., Chakrabarti, B., Givens, I., Saddy, D., Kitchen, I., R. Swann, J. and

Bailey, A. (2016) ‘Maternal Weaning Modulates Emotional Behavior and Regulates the Gut-Brain

Axis’, Scientific Reports, 6(1), p. 21958.

Favier, C. F., de Vos, W. M. and Akkermans, A. D. . (2003) ‘Development of bacterial and

bifidobacterial communities in feces of newborn babies’, Anaerobe, 9(5), pp. 219–229.

Ferdman, N., Murmu, R. P., Bock, J., Braun, K. and Leshem, M. (2007) ‘Weaning age, social

isolation, and gender, interact to determine adult explorative and social behavior, and dendritic

and spine morphology in prefrontal cortex of rats.’, Behavioural brain research, 180(2), pp. 174–

82.

Filliol, D., Ghozland, S., Chluba, J., Martin, M., Matthes, H. W., Simonin, F., Befort, K.,

Gavériaux-Ruff, C., Dierich, A., LeMeur, M., Valverde, O., Maldonado, R. and Kieffer, B. L.

(2000) ‘Mice deficient for delta- and mu-opioid receptors exhibit opposing alterations of emotional

responses.’, Nature genetics, 25(2), pp. 195–200.

Finkelstein, J. A., Stille, C., Nordin, J., Davis, R., Raebel, M. A., Roblin, D., Go, A. S., Smith, D.,

Johnson, C. C., Kleinman, K., Chan, K. A. and Platt, R. (2003) ‘Reduction in antibiotic use among

US children, 1996-2000.’, Pediatrics, 112(3 Pt 1), pp. 620–7.

Fombonne, E. (2009) ‘Epidemiology of Pervasive Developmental Disorders’, Pediatric Research,

65(6), pp. 591–598.

Fonnum, F. (1984) ‘Glutamate: a neurotransmitter in mammalian brain.’, Journal of

neurochemistry, 42(1), pp. 1–11.

Foster, J. A. and McVey Neufeld, K.-A. (2013) ‘Gut–brain axis: how the microbiome influences

280

anxiety and depression’, Trends in Neurosciences, 36(5), pp. 305–312.

Franklin, M. A., Mathew, A. G., Vickers, J. R. and Clift, R. A. (2002) ‘Characterization of

microbial populations and volatile fatty acid concentrations in the jejunum, ileum, and cecum of

pigs weaned at 17 vs 24 days of age’, Journal of Animal Science, 80(11), p. 2904.

Franks, A. H., Harmsen, H. J. M., Raangs, G. C., Jansen, G. J., Schut, F. and Welling, G. W. (1998)

‘Variations of bacterial populations in human feces measured by fluorescent in situ hybridization

with group-specific 16S rRNA-targeted oligonucleotide probes’, Applied and Environmental

Microbiology, 64(9), pp. 3336–3345.

Frustaci, A., Neri, M., Cesario, A., Adams, J. B., Domenici, E., Dalla Bernardina, B. and Bonassi,

S. (2012) ‘Oxidative stress-related biomarkers in autism: Systematic review and meta-analyses’,

Free Radical Biology and Medicine, 52(10), pp. 2128–2141.

Fukuda, S., Toh, H., Hase, K., Oshima, K., Nakanishi, Y., Yoshimura, K., Tobe, T., Clarke, J. M.,

Topping, D. L., Suzuki, T., Taylor, T. D., Itoh, K., Kikuchi, J., Morita, H., Hattori, M. and Ohno,

H. (2011) ‘Bifidobacteria can protect from enteropathogenic infection through production of

acetate’, Nature, 469(7331), pp. 543–547.

Gao, H., Xiang, Y., Sun, N., Zhu, H., Wang, Y., Liu, M., Ma, Y. and Lei, H. (2007) ‘Metabolic

changes in rat prefrontal cortex and hippocampus induced by chronic morphine treatment studied

ex vivo by high resolution 1H NMR spectroscopy’, Neurochemistry International, 50(2), pp. 386–

394.

Gareau, M. G., Wine, E., Rodrigues, D. M., Cho, J. H., Whary, M. T., Philpott, D. J., MacQueen,

G. and Sherman, P. M. (2011) ‘Bacterial infection causes stress-induced memory dysfunction in

mice’, Gut, 60(3), pp. 307–317.

281

Georgiou, P., Zanos, P., Garcia-Carmona, J.-A., Hourani, S., Kitchen, I., Kieffer, B. L., Laorden,

M.-L. and Bailey, A. (2015) ‘The oxytocin analogue carbetocin prevents priming-induced

reinstatement of morphine-seeking: Involvement of dopaminergic, noradrenergic and MOPr

systems’, European Neuropsychopharmacology, 25(12), pp. 2459–2464.

Gigliucci, V., Leonzino, M., Busnelli, M., Luchetti, A., Palladino, V. S., Dâ€TMAmato, F. R. and

Chini, B. (2014) ‘Region Specific Up-Regulation of Oxytocin Receptors in the Opioid

Oprm1âˆ’/âˆ’ Mouse Model of Autism’, Frontiers in Pediatrics, 2, p. 91.

Gill, S. R., Pop, M., Deboy, R. T., Eckburg, P. B., Turnbaugh, P. J., Samuel, B. S., Gordon, J. I.,

Relman, D. A., Fraser-Liggett, C. M. and Nelson, K. E. (2006) ‘Metagenomic analysis of the

human distal gut microbiome.’, Science, 312(5778), pp. 1355–9.

Gimpl, G. and Fahrenholz, F. (2001) ‘The oxytocin receptor system: structure, function, and

regulation.’, Physiological reviews, 81(2), pp. 629–83.

Ginger, M. R. and Grigor, M. R. (1999) ‘Comparative aspects of milk caseins.’, Comparative

biochemistry and physiology. Part B, Biochemistry & molecular biology, 124(2), pp. 133–45.

Goehler, L. E., Park, S. M., Opitz, N., Lyte, M. and Gaykema, R. P. A. (2008) ‘Campylobacter

jejuni infection increases anxiety-like behavior in the holeboard: possible anatomical substrates

for viscerosensory modulation of exploratory behavior.’, Brain, behavior, and immunity, 22(3),

pp. 354–66.

Golubeva, A. V., Crampton, S., Desbonnet, L., Edge, D., O’Sullivan, O., Lomasney, K. W.,

Zhdanov, A. V., Crispie, F., Moloney, R. D., Borre, Y. E., Cotter, P. D., Hyland, N. P., O’Halloran,

K. D., Dinan, T. G., O’Keeffe, G. W. and Cryan, J. F. (2015) ‘Prenatal stress-induced alterations

in major physiological systems correlate with gut microbiota composition in adulthood’,

282

Psychoneuroendocrinology, 60, pp. 58–74.

Goody, R. J. and Kitchen, I. (2001) ‘Influence of maternal milk on functional activation of delta-

opioid receptors in postnatal rats.’, The Journal of pharmacology and experimental therapeutics,

296(3), pp. 744–748.

Goody, R. J. and Kitchen, I. (2001) ‘Influence of maternal milk on functional activation of delta-

opioid receptors in postnatal rats.’, The Journal of pharmacology and experimental therapeutics,

296(3), pp. 744–748.

Grace, A. A. (2016) ‘Dysregulation of the dopamine system in the pathophysiology of

schizophrenia and depression’, Nature Reviews Neuroscience, 17(8), pp. 524–532.

Greenberg, R., Groves, M. and Dower, H. (1984) ‘Human beta-casin. Amino acid sequence and

identification of phosphorylation sites.’, The Journal of Biological Chemestry, 259(8), pp. 5132–

5138.

Grenham, S., Clarke, G., Cryan, J. F. and Dinan, T. G. (2011) ‘Brain-gut-microbe communication

in health and disease’, Front Physiol, 2, p. 94.

Gros, D. F., Antony, M. M., McCabe, R. E. and Swinson, R. P. (2009) ‘Frequency and severity of

the symptoms of irritable bowel syndrome across the anxiety disorders and depression’, Journal

of Anxiety Disorders, 23(2), pp. 290–296.

Gutman, D. A. and Nemeroff, C. B. (2003) ‘Persistent central nervous system effects of an adverse

early environment: clinical and preclinical studies.’, Physiology & behavior, 79(3), pp. 471–8.

Haldar, J. and Sawyer, W. H. (1978) ‘Inhibition of oxytocin release by morphine and its analogs.’,

Proceedings of the Society for Experimental Biology and Medicine. Society for Experimental

283

Biology and Medicine (New York, N.Y.), 157(3), pp. 476–80. Available at:

http://www.ncbi.nlm.nih.gov/pubmed/24854 (Accessed: 31 July 2017).

Hallmayer, J., Cleveland, S., Torres, A., Phillips, J., Cohen, B., Torigoe, T., Miller, J., Fedele, A.,

Collins, J., Smith, K., Lotspeich, L., Croen, L. A., Ozonoff, S., Lajonchere, C., Grether, J. K. and

Risch, N. (2011) ‘Genetic Heritability and Shared Environmental Factors Among Twin Pairs With

Autism’, Archives of General Psychiatry, 68(11), p. 1095.

Hanstock, T. L., Mallet, P. E. and Clayton, E. H. (2010) ‘Increased plasma d-lactic acid associated

with impaired memory in rats.’, Physiology & behavior, 101(5), pp. 653–9.

Hao, Y., Yang, J. Y., Guo, M., Wu, C. F. and Wu, M. F. (2005) ‘Morphine decreases extracellular

levels of glutamate in the anterior cingulate cortex: an in vivo microdialysis study in freely moving

rats.’, Brain research, 1040(1–2), pp. 191–6.

Haque, M. M., Chin, H. C. and Huang, H. (2009) ‘Modeling fault among motorcyclists involved

in crashes’, Accident Analysis and Prevention, 41(2), pp. 327–335.

Harder, T., Bergmann, R., Kallischnigg, G. and Plagemann, A. (2005) ‘Duration of Breastfeeding

and Risk of Overweight: A Meta-Analysis’, American Journal of Epidemiology, 162(5), pp. 397–

403.

Harmsen, H. J. M., Elfferich, P., Schut, F. and Welling, G. W. (1999) ‘A 16S rRNA-targeted probe

for detection of lactobacilli and enterococci in faecal samples by fluorescent in situ hybridization’,

Microbial Ecology in Health and Disease, 11(1), pp. 3–12.

Hasler, G. (2010) ‘Pathophysiology of depression: do we have any solid evidence of interest to

clinicians?’, World psychiatry : official journal of the World Psychiatric Association (WPA), 9(3),

284

pp. 155–61.

Hatheway, C. L. (1990) ‘Toxigenic clostridia.’, Clinical microbiology reviews, 3(1), pp. 66–98.

Hawkins, K. N., Knapp, R. J., Gehlert, D. R., Lui, G. K., Yamamura, M. S., Roeske, L. C., Hruby,

V. J. and Yamamura, H. I. (1988) ‘Quantitative autoradiography of [3H]CTOP binding to mu

opioid receptors in rat brain.’, Life sciences, 42(25), pp. 2541–51.

Heijtz, R. D., Wang, S., Anuar, F., Qian, Y., Bjorkholm, B., Samuelsson, A., Hibberd, M. L.,

Forssberg, H. and Pettersson, S. (2011) ‘Normal gut microbiota modulates brain development and

behavior’, Proceedings of the National Academy of Sciences, 108(7), pp. 3047–3052.

Henry, C. J., Huang, Y., Wynne, A., Hanke, M., Himler, J., Bailey, M. T., Sheridan, J. F. and

Godbout, J. P. (2008) ‘Minocycline attenuates lipopolysaccharide (LPS)-induced

neuroinflammation, sickness behavior, and anhedonia’, Journal of Neuroinflammation, 5(1), p. 15.

Hentges, D. J. (1983) HUMAN INTESTNL MICROFLORAIN HLTH & DISEASE.

Heringa, S. M., van den Berg, E., Reijmer, Y. D., Nijpels, G., Stehouwer, C. D. A., Schalkwijk,

C. G., Teerlink, T., Scheffer, P. G., van den Hurk, K., Kappelle, L. J., Dekker, J. M. and Biessels,

G. J. (2014) ‘Markers of low-grade inflammation and endothelial dysfunction are related to

reduced information processing speed and executive functioning in an older population – the

Hoorn Study’, Psychoneuroendocrinology, 40, pp. 108–118.

Higuchi, T., Hayashi, H. and Abe, K. (1997) ‘Exchange of glutamate and gamma-aminobutyrate

in a Lactobacillus strain.’, Journal of bacteriology, 179(10), pp. 3362–4.

Ho, C. (2013) Optimal duration of exclusive breastfeeding, Cochrane Database of Systematic

Reviews.

285

Ho, S., Woodford, K., Kukuljan, S. and Pal, S. (2014) ‘Comparative effects of A1 versus A2 beta-

casein on gastrointestinal measures: a blinded randomised cross-over pilot study.’, European

journal of clinical nutrition, 68(9), pp. 994–1000.

Hole, K., Bergslien, H., Jørgensen, H. A., Berge, O. G., Reichelt, K. L. and Trygstad, O. E. (1979)

‘A peptide-containing fraction in the urine of schizophrenic patients which stimulates opiate

receptors and inhibits dopamine uptake.’, Neuroscience, 4(12), pp. 1883–93.

Hollister, E. B., Gao, C. and Versalovic, J. (2014) ‘Compositional and functional features of the

gastrointestinal microbiome and their effects on human health.’, Gastroenterology, 146(6), pp.

1449–58.

Holmes, E., Wilson, I. D. and Nicholson, J. K. (2008) ‘Metabolic Phenotyping in Health and

Disease’, Cell, 134(5), pp. 714–717.

Holst, J. J. and Gromada, J. (2004) ‘Role of incretin hormones in the regulation of insulin secretion

in diabetic and nondiabetic humans’, AJP: Endocrinology and Metabolism, 287(2), pp. 199–206.

Hondorp, E. R. and Matthews, R. G. (2004) ‘Oxidative stress inactivates cobalamin-independent

methionine synthase (MetE) in Escherichia coli.’, PLoS biology, 2(11), p. 336.

Hostinar, C. E., Sullivan, R. M. and Gunnar, M. R. (2014) ‘Psychobiological mechanisms

underlying the social buffering of the hypothalamic–pituitary–adrenocortical axis: A review of

animal models and human studies across development.’, Psychological Bulletin, 140(1), pp. 256–

282.

Hoyer, D., Clarke, D. E., Fozard, J. R., Hartig, P. R., Martin, G. R., Mylecharane, E. J., Saxena, P.

R. and Humphrey, P. P. (1994) International Union of Pharmacology Classification of Receptors

286

for 5-Hydroxytryptamine (Serotonin), Pharmacological reviews.

Hsiao, E. Y. (2014) ‘Gastrointestinal Issues in Autism Spectrum Disorder’, Harvard Review of

Psychiatry, 22(2), pp. 104–111.

Hsiao, E. Y., McBride, S. W., Hsien, S., Sharon, G., Hyde, E. R., McCue, T., Codelli, J. A., Chow,

J., Reisman, S. E., Petrosino, J. F., Patterson, P. H. and Mazmanian, S. K. (2013) ‘Microbiota

modulate behavioral and physiological abnormalities associated with neurodevelopmental

disorders.’, Cell, 155(7), pp. 1451–63.

Huang, T.-L. and Lin, C.-C. (2015) ‘Advances in Biomarkers of Major Depressive Disorder’, in

Advances in clinical chemistry, pp. 177–204.

Huber, D., Veinante, P. and Stoop, R. (2005) ‘Vasopressin and oxytocin excite distinct neuronal

populations in the central amygdala.’, Science, 308(5719), pp. 245–8.

Inoue, R. and Ushida, K. (2003) ‘Development of the intestinal microbiota in rats and its possible

interactions with the evolution of the luminal IgA in the intestine’, FEMS Microbiology Ecology,

45(2), pp. 147–153.

Isaacks, R. E., Bender, A. S., Kim, C. Y., Prieto, N. M. and Norenberg, M. D. (1994) ‘Osmotic

regulation of myo-inositol uptake in primary astrocyte cultures.’, Neurochemical research, 19(3),

pp. 331–8.

Ishii, T., Itou, T. and Nishimura, M. (2005) ‘Comparison of growth and exploratory behavior in

mice fed an exclusively milk formula diet and mice fed a food-pellet diet post weaning’, Life

Sciences, 78(2), pp. 174–179.

Ito, A., Kikusui, T., Takeuchi, Y. and Mori, Y. (2006) ‘Effects of early weaning on anxiety and

287

autonomic responses to stress in rats’, Behavioural Brain Research, 171(1), pp. 87–93.

Iwata, E., Kikusui, T., Takeuchi, Y. and Mori, Y. (2007) ‘Fostering and environmental enrichment

ameliorate anxious behavior induced by early weaning in Balb/c mice.’, Physiology & behavior,

91(2–3), pp. 318–24.

Jackson, H. C. and Kitchen, I. (1989) ‘Swim???stress???induced antinociception in young rats’,

British Journal of Pharmacology, 96(3), pp. 617–622.

Jakovcevski, M. and Akbarian, S. (2012) ‘Epigenetic mechanisms in neurological disease’, Nature

Medicine, 18(8), pp. 1194–1204.

Janeczko, S., Atwater, D., Bogel, E., Greiter-Wilke, A., Gerold, A., Baumgart, M., Bender, H.,

McDonough, P. L., McDonough, S. P., Goldstein, R. E. and Simpson, K. W. (2008) ‘The

relationship of mucosal bacteria to duodenal histopathology, cytokine mRNA, and clinical disease

activity in cats with inflammatory bowel disease.’, Veterinary microbiology, 128(1–2), pp. 178–

93.

Jarmołowska, B., Sidor, K., Iwan, M., Bielikowicz, K., Kaczmarski, M., Kostyra, E. and Kostyra,

H. (2007) ‘Changes of beta-casomorphin content in human milk during lactation.’, Peptides,

28(10), pp. 1982–6.

Jarmołowska, B., Sidor, K., Iwan, M., Bielikowicz, K., Kaczmarski, M., Kostyra, E. and Kostyra,

H. (2007) ‘Changes of β-casomorphin content in human milk during lactation’, Peptides, 28(10),

pp. 1982–1986.

Javelot, H., Messaoudi, M., Garnier, S. and Rougeot, C. (2010) ‘Human opiorphin is a naturally

occurring antidepressant acting selectively on enkephalin-dependent delta-opioid pathways.’,

288

Journal of physiology and pharmacology : an official journal of the Polish Physiological Society,

61(3), pp. 355–62.

Jianqin, S., Leiming, X., Lu, X., Yelland, G. W., Ni, J. and Clarke, A. J. (2015) ‘Effects of milk

containing only A2 beta casein versus milk containing both A1 and A2 beta casein proteins on

gastrointestinal physiology, symptoms of discomfort, and cognitive behavior of people with self-

reported intolerance to traditional cows’ milk’, Nutrition Journal, 15(1), p. 35.

Jinsmaa, Y. and Yoshikawa, M. (1999) ‘Enzymatic release of neocasomorphin and beta-

casomorphin from bovine beta-casein.’, Peptides, 20(8), pp. 957–62.

Jost, T., Lacroix, C., Braegger, C. and Chassard, C. (2015) ‘Impact of human milk bacteria and

oligosaccharides on neonatal gut microbiota establishment and gut health’, Nutrition Reviews,

73(7), pp. 426–437.

Julvez, J., Ribas-Fitó, N., Forns, M., Garcia-Esteban, R., Torrent, M. and Sunyer, J. (2007)

‘Attention behaviour and hyperactivity at age 4 and duration of breast-feeding’, Acta Paediatrica,

96(6), pp. 842–847.

Kamada, N., Seo, S.-U., Chen, G. Y. and Núñez, G. (2013) ‘Role of the gut microbiota in immunity

and inflammatory disease’, Nature Reviews Immunology, 13(5), pp. 321–335.

Kamiński, S., Cieslińska, A. and Kostyra, E. (2007) ‘Polymorphism of bovine beta-casein and its

potential effect on human health.’, Journal of applied genetics, 48(3), pp. 189–98.

Kanari, K., Kikusui, T., Takeuchi, Y. and Mori, Y. (2005) ‘Multidimensional structure of anxiety-

related behavior in early-weaned rats.’, Behavioural brain research, 156(1), pp. 45–52.

Kaneko, K., Iwasaki, M., Yoshikawa, M. and Ohinata, K. (2010) ‘Orally administered

289

soymorphins, soy-derived opioid peptides, suppress feeding and intestinal transit via gut μ 1 -

receptor coupled to 5-HT 1A , D 2 , and GABA B systems’, American Journal of Physiology-

Gastrointestinal and Liver Physiology, 299(3), pp. G799–G805.

Kapuscinski, J. (1995) ‘DAPI: a DNA-specific fluorescent probe.’, Biotechnic & histochemistry :

official publication of the Biological Stain Commission, 70(5), pp. 220–33.

Kayser, H. and Meisel, H. (1996) ‘Stimulation of human peripheral blood lymphocytes by

bioactive peptides derived from bovine milk proteins.’, FEBS letters, 383(1–2), pp. 18–20.

Kennedy, P. J., Cryan, J. F., Dinan, T. G. and Clarke, G. (2014) ‘Irritable bowel syndrome: a

microbiome-gut-brain axis disorder?’, World journal of gastroenterology, 20(39), pp. 14105–25.

Kieffer, B. L. and Gavériaux-Ruff, C. (2002) ‘Exploring the opioid system by gene knockout’,

Progress in Neurobiology, 66(5), pp. 285–306.

Kikusui, T., Ichikawa, S. and Mori, Y. (2009) ‘Maternal deprivation by early weaning increases

corticosterone and decreases hippocampal BDNF and neurogenesis in mice.’,

Psychoneuroendocrinology, 34(5), pp. 762–72.

Kikusui, T., Isaka, Y. and Mori, Y. (2005) ‘Early weaning deprives mouse pups of maternal care

and decreases their maternal behavior in adulthood.’, Behavioural brain research, 162(2), pp. 200–

6.

Kikusui, T., Kiyokawa, Y. and Mori, Y. (2007) ‘Deprivation of mother-pup interaction by early

weaning alters myelin formation in male, but not female, ICR mice.’, Brain research, 1133(1), pp.

115–22.

Kikusui, T. and Mori, Y. (2009) ‘Behavioural and Neurochemical Consequences of Early Weaning

290

in Rodents’, Journal of Neuroendocrinology, 21(4), pp. 427–431.

Kikusui, T., Nakamura, K., Kakuma, Y. and Mori, Y. (2006) ‘Early weaning augments

neuroendocrine stress responses in mice.’, Behavioural brain research, 175(1), pp. 96–103.

Kikusui, T., Nakamura, K. and Mori, Y. (2008) ‘A review of the behavioral and neurochemical

consequences of early weaning in rodents’, Applied Animal Behaviour Science, 110(1–2), pp. 73–

83.

Kikusui, T., Takeuchi, Y. and Mori, Y. (2004) ‘Early weaning induces anxiety and aggression in

adult mice.’, Physiology & behavior, 81(1), pp. 37–42.

Kimura, T., Tanizawa, O., Mori, K., Brownstein, M. J. and Okayama, H. (1992) ‘Structure and

expression of a human oxytocin receptor’, Nature, 356(6369), pp. 526–529.

Kitchen, I., Leslie, F. M., Kelly, M., Barnes, R., Crook, T. J., Hill, R. G., Borsodi, A., Toth, G.,

Melchiorri, P. and Negri, L. (1995a) ‘Development of delta-opioid receptor subtypes and the

regulatory role of weaning: radioligand binding, autoradiography and in situ hybridization

studies.’, The Journal of pharmacology and experimental therapeutics, 275(3), pp. 1597–607.

Kitchen, I., Leslie, F. M., Kelly, M., Barnes, R., Crook, T. J., Hill, R. G., Borsodi, A., Toth, G.,

Melchiorri, P. and Negri, L. (1995b) ‘Development of delta-opioid receptor subtypes and the

regulatory role of weaning: radioligand binding, autoradiography and in situ hybridization

studies.’, The Journal of pharmacology and experimental therapeutics, 275(3), pp. 1597–607.

Kitchen, I. and Pinker, S. R. (1990) ‘Antagonism of swim-stress-induced antinociception by the

delta-opioid receptor antagonist naltrindole in adult and young rats.’, British journal of

pharmacology, 100(4), pp. 685–8.

291

Klement, E., Cohen, R. V, Boxman, J., Joseph, A. and Reif, S. (2004) ‘Breastfeeding and risk of

inflammatory bowel disease: a systematic review with meta-analysis’, The American Journal of

Clinical Nutrition, 80(5), pp. 1342–1352.

Knoll, A. T. and Carlezon, W. A. (2010) ‘Dynorphin, stress, and depression’, Brain Research,

1314, pp. 56–73.

Koch, G., Wiedemann, K. and Teschemacher, H. (1985) ‘Opioid activities of human beta-

casomorphins.’, Naunyn-Schmiedeberg’s archives of pharmacology, 331(4), pp. 351–4.

Koch, T., Schulz, S., Pfeiffer, M., Klutzny, M., Schröder, H., Kahl, E. and Höllt, V. (2001) ‘C-

terminal splice variants of the mouse mu-opioid receptor differ in morphine-induced

internalization and receptor resensitization.’, The Journal of biological chemistry, 276(33), pp.

31408–14.

Koenig, J. E., Spor, A., Scalfone, N., Fricker, A. D., Stombaugh, J., Knight, R., Angenent, L. T.

and Ley, R. E. (2011) ‘Succession of microbial consortia in the developing infant gut microbiome’,

Proceedings of the National Academy of Sciences, 108(1), pp. 4578–4585.

König, M., Zimmer, A. M., Steiner, H., Holmes, P. V., Crawley, J. N., Brownstein, M. J. and

Zimmer, A. (1996) ‘Pain responses, anxiety and aggression in mice deficient in pre-

proenkephalin’, Nature, 383(6600), pp. 535–538.

Korhonen, H. and Pihlanto, A. (2003) ‘Food-derived bioactive peptides--opportunities for

designing future foods.’, Current pharmaceutical design, 9(16), pp. 1297–308.

korhonen Hannu, A. P., Korhonen, H. and Pihlanto, A. (2006) ‘Review Bioactive Peptides :

Production and Functionality’, International Dairy Journal, 16(9), pp. 945–960.

292

Kost, N. V., Sokolov, О. Y., Kurasova, О. B., Dmitriev, A. D., Tarakanova, J. N., Gabaeva, М.

V., Zolotarev, Y. A., Dadayan, А. K., Grachev, S. A., Korneeva, Е. V., Mikheeva, I. G. and

Zozulya, А. A. (2009a) ‘β-Casomorphins-7 in infants on different type of feeding and different

levels of psychomotor development’, Peptides, 30(10), pp. 1854–1860.

Kost, N. V., Sokolov, О. Y., Kurasova, О. B., Dmitriev, A. D., Tarakanova, J. N., Gabaeva, М.

V., Zolotarev, Y. A., Dadayan, А. K., Grachev, S. A., Korneeva, Е. V., Mikheeva, I. G. and

Zozulya, А. A. (2009b) ‘β-Casomorphins-7 in infants on different type of feeding and different

levels of psychomotor development’, Peptides, 30(10), pp. 1854–1860.

Kovács, G. L., Laczi, F., Vecsernyés, M., Hódi, K., Telegdy, G. and László, F. A. (1987) ‘Limbic

oxytocin and arginine 8-vasopressin in morphine tolerance and dependence.’, Experimental brain

research, 65(2), pp. 307–11.

Kreil, G., Umbach, M., Brantl, V. and Teschemacher, H. (1983) ‘Studies on the enzymatic

degradation of beta-casomorphins.’, Life sciences, 33, pp. 137–40.

Kunz, C. and Lönnerdal, B. (1990) ‘Human-milk proteins: analysis of casein and casein subunits

by anion-exchange chromatography, gel electrophoresis, and specific staining methods.’, The

American journal of clinical nutrition, 51(1), pp. 37–46.

Kunz, C. and Lönnerdal, B. (1992) ‘Re-evaluation of the whey protein/casein ratio of human milk’,

Acta Paediatrica, 81(2), pp. 107–112.

Kwan, M. L., Buffler, P. A., Abrams, B. and Kiley, V. A. (2004) ‘Breastfeeding and the Risk of

Childhood Leukemia: A Meta-Analysis’, Public Health Reports, 119(6), pp. 521–535.

Landgraf, R. and Neumann, I. D. (2004) ‘Vasopressin and oxytocin release within the brain: a

293

dynamic concept of multiple and variable modes of neuropeptide communication’, Frontiers in

Neuroendocrinology, 25(3–4), pp. 150–176.

Langendijk, P. S., Schut, F., Jansen, G. J., Raangs, G. C., Kamphuis, G. R., Wilkinson, M. H. F.

and Welling, G. W. (1995) ‘Quantitative fluorescence in situ hybridization of Bifidobacterium spp.

with genus-specific 16S rRNA-targeted probes and its application in fecal samples’, Applied and

Environmental Microbiology, 61(8), pp. 3069–3075.

Laviola, G. and Dell’omo, G. (1997) ‘Precocious weaning and changes in social variables during

prepuberty affect cocaine reinforcing properties in adult mice.’, Psychobiology. Springer-Verlag,

25(2), pp. 163–170.

Lawrence, R. A. and Lawrence, R. M. (1994) Breastfeeding: A Guide for the Medical Profession.

Leboyer, M., Bouvard, M. P., Launay, J. M., Recasens, C., Plumet, M. H., Waller-Perotte, D.,

Tabuteau, F., Bondoux, D. and Dugas, M. (1993) ‘Opiate hypothesis in infantile autism?

Therapeutic trials with naltrexone.’, Encephale, 19(2), pp. 95–102.

Leclercq, S., Mian, F. M., Stanisz, A. M., Bindels, L. B., Cambier, E., Ben-Amram, H., Koren, O.,

Forsythe, P. and Bienenstock, J. (2017) ‘Low-dose penicillin in early life induces long-term

changes in murine gut microbiota, brain cytokines and behavior’, Nature Communications, 8, p.

15062.

Ledochowski, M., Widner, B., Sperner-Unterweger, B., Propst, T., Vogel, W. and Fuchs, D. (2000)

‘Carbohydrate malabsorption syndromes and early signs of mental depression in females.’,

Digestive diseases and sciences, 45(7), pp. 1255–9.

LeDoux, J. E. (2000) ‘Emotion circuits in the brain.’, Annual review of neuroscience, 23(1), pp.

294

155–84.

Lefkowitz, R. J., Stadel, J. M. and Caron, M. G. (1983) ‘Adenylate cyclase-coupled beta-

adrenergic receptors: structure and mechanisms of activation and desensitization.’, Annual review

of biochemistry, 52(1), pp. 159–86.

Leslie, F. M., Chen, Y. and Winzer-Serhan, U. H. (1998) ‘Opioid receptor and peptide mRNA

expression in proliferative zones of fetal rat central nervous system.’, Canadian journal of

physiology and pharmacology, 76(3), pp. 284–93.

Lessen, R. and Kavanagh, K. (2015) ‘Position of the academy of nutrition and dietetics: Promoting

and supporting breastfeeding’, Journal of the Academy of Nutrition and Dietetics, 115(3), pp. 444–

449.

Leung, A. K. and Sauve, R. S. (2003) ‘Whole cow’s milk in infancy’, Paediatrics and Child

Health, 8(7), pp. 419–421.

Levitt, E. S. and Williams, J. T. (2012) ‘Morphine desensitization and cellular tolerance are

distinguished in rat locus ceruleus neurons.’, Molecular pharmacology, 82(5), pp. 983–92.

Lightman, S. L. (2008) ‘The neuroendocrinology of stress: a never ending story.’, Journal of

neuroendocrinology, 20(6), pp. 880–4.

Lindon, J. C., Nicholson, J. K., Holmes, E. and Everett, J. R. (2000) ‘Metabonomics: Metabolic

processes studied by NMR spectroscopy of biofluids’, Concepts in Magnetic Resonance, 12(5),

pp. 289–320.

Lister, J., Fletcher, P. J., Nobrega, J. N. and Remington, G. (2015) ‘Behavioral effects of food-

derived opioid-like peptides in rodents: Implications for schizophrenia?’, Pharmacology

295

Biochemistry and Behavior, 134, pp. 70–78.

Liu, J., Zhao, S. R. and Reyes, T. (2015) ‘Neurological and Epigenetic Implications of Nutritional

Deficiencies on Psychopathology: Conceptualization and Review of Evidence.’, International

journal of molecular sciences, 16(8), pp. 18129–48.

Loman, M. M. and Gunnar, M. R. (2010) ‘Early experience and the development of stress

reactivity and regulation in children’, Neuroscience & Biobehavioral Reviews, 34(6), pp. 867–876.

Lönnerdal, B. (1994) ‘Nutritional aspects of soy formula.’, Acta paediatrica, 402, pp. 105–8.

Lonstein, J. S. (2007) ‘Regulation of anxiety during the postpartum period’, Frontiers in

Neuroendocrinology, pp. 115–141.

Lucassen, P. J., Naninck, E. F. G., van Goudoever, J. B., Fitzsimons, C., Joels, M. and Korosi, A.

(2013) ‘Perinatal programming of adult hippocampal structure and function; Emerging roles of

stress, nutrition and epigenetics’, Trends in Neurosciences, pp. 621–631.

Luczynski, P., McVey Neufeld, K.-A., Oriach, C. S., Clarke, G., Dinan, T. G. and Cryan, J. F.

(2016) ‘Growing up in a Bubble: Using Germ-Free Animals to Assess the Influence of the Gut

Microbiota on Brain and Behavior.’, The international journal of neuropsychopharmacology,

19(8), p. 20.

Ludwig, D. S. and Willett, W. C. (2013) ‘Three Daily Servings of Reduced-Fat Milk’, JAMA

Pediatrics, 167(9), p. 788.

Lukkes, J. L. (2009) ‘Consequences of post-weaning social isolation on anxiety behavior and

related neural circuits in rodents’, Frontiers in Behavioral Neuroscience, 3, p. 18.

Lundberg, R., Toft, M. F., August, B., Hansen, A. K. and Hansen, C. H. F. (2016) ‘Antibiotic-

296

treated versus germ-free rodents for microbiota transplantation studies.’, Gut microbes. Taylor &

Francis, 7(1), pp. 68–74. doi: 10.1080/19490976.2015.1127463.

Lutz, P.-E. and Kieffer, B. L. (2013) ‘Opioid receptors: distinct roles in mood disorders.’, Trends

in neurosciences, 36(3), pp. 195–206.

Lyte, M. (2011) ‘Probiotics function mechanistically as delivery vehicles for neuroactive

compounds: Microbial endocrinology in the design and use of probiotics.’, BioEssays : news and

reviews in molecular, cellular and developmental biology, 33(8), pp. 574–81.

Lyte, M., Li, W., Opitz, N., Gaykema, R. P. A. and Goehler, L. E. (2006) ‘Induction of anxiety-

like behavior in mice during the initial stages of infection with the agent of murine colonic

hyperplasia Citrobacter rodentium’, Physiology and Behavior, 89(3), pp. 350–357.

MacFabe, D. F., Cain, N. E., Boon, F., Ossenkopp, K.-P. and Cain, D. P. (2011) ‘Effects of the

enteric bacterial metabolic product propionic acid on object-directed behavior, social behavior,

cognition, and neuroinflammation in adolescent rats: Relevance to autism spectrum disorder’,

Behavioural Brain Research, 217(1), pp. 47–54.

Mackie, R. I., Sghir, A. and Gaskins, H. R. (1999) ‘Developmental microbial ecology of the

neonatal gastrointestinal tract.’, The American journal of clinical nutrition, 69(5), pp. 1035–1045.

Maes, M. (2008) ‘The cytokine hypothesis of depression: inflammation, oxidative &

nitrosative stress (IO&NS) and leaky gut as new targets for adjunctive treatments in

depression.’, Neuro endocrinology letters, 29(3), pp. 287–91.

Magnan, J. and Tiberi, M. (1989) ‘Evidence for the presence of mu- and kappa- but not of delta-

opioid sites in the human fetal brain.’, Brain research. Developmental brain research, 45(2), pp.

297

275–81.

Mague, S. D., Pliakas, A. M., Todtenkopf, M. S., Tomasiewicz, H. C., Zhang, Y., Stevens, W. C.,

Jones, R. M., Portoghese, P. S. and Carlezon, W. A. (2003) ‘Antidepressant-Like Effects of kappa

-Opioid Receptor Antagonists in the Forced Swim Test in Rats’, Journal of Pharmacology and

Experimental Therapeutics, 305(1), pp. 323–330.

Mahè, S., Tomè, D., Dumontier, A. M. and Desjeux, J. F. (1989) ‘Absorption of intact beta-

casomorphins (beta-CM) in rabbit ileum in vitro.’, Reproduction, nutrition, development, 29, pp.

725–733.

Majidi, J., Kosari-Nasab, M. and Salari, A.-A. (2016) ‘Developmental minocycline treatment

reverses the effects of neonatal immune activation on anxiety- and depression-like behaviors,

hippocampal inflammation, and HPA axis activity in adult mice’, Brain Research Bulletin, 120,

pp. 1–13.

Mann, S. P. and Sharman, D. F. (1983) ‘Changes associated with early weaning in the activity of

tyrosine hydroxylase in the caudate nucleus of the piglet.’, Comparative biochemistry and

physiology, 74(2), pp. 267–70.

Mansour, A., Khachaturian, H., Lewis, M. E., Akil, H. and Watson, S. J. (1987) ‘Autoradiographic

differentiation of mu, delta, and kappa opioid receptors in the rat forebrain and midbrain.’, The

Journal of neuroscience : the official journal of the Society for Neuroscience, 7(8), pp. 2445–64.

Martin, C., Ling, P.-R. and Blackburn, G. (2016) ‘Review of Infant Feeding: Key Features of

Breast Milk and Infant Formula’, Nutrients, 8(5), p. 279.

Martin, P. and Bateson, P. (1985) ‘The influence of experimentally manipulating a component of

298

weaning on the development of play in domestic cats’, Animal Behaviour, 33(2), pp. 511–518.

Mayer, E. A. (2011) ‘Gut feelings: the emerging biology of gut–brain communication’, Nature

Reviews Neuroscience, 12(8), pp. 453–466.

McCaig, L. F., Besser, R. E. and Hughes, J. M. (2002) ‘Trends in antimicrobial prescribing rates

for children and adolescents.’, The journal of the American Medical Association, 287(23), pp.

3096–3102.

McCrory, C. and Murray, A. (2013) ‘The Effect of Breastfeeding on Neuro-Development in

Infancy’, Maternal and Child Health Journal, 17(9), pp. 1680–1688.

McDonald, J. and Lambert, D. (2005) ‘Opioid receptors’, Continuing Education in Anaesthesia,

Critical Care & Pain, 5(1), pp. 22–25.

McDowell, J. and Kitchen, I. (1986) ‘Ontogenesis of delta-opioid receptors in rat brain using

[3H][D-Pen2,D-Pen5]enkephalin as a binding ligand.’, European journal of pharmacology,

128(3), pp. 287–9.

McDowell, J. and Kitchen, I. (1987) ‘Development of opioid systems: peptides, receptors and

pharmacology’, Brain Research Reviews, 12(4), pp. 397–421.

McLachlan, C. N. S. (2001) ‘β-casein A1, ischaemic heart disease mortality, and other illnesses’,

Medical Hypotheses, 56(2), pp. 262–272.

McMahon, D. J. and Oommen, B. S. (2008) ‘Supramolecular Structure of the Casein Micelle’,

Journal of Dairy Science, 91(5), pp. 1709–1721.

Meale, S. J., Li, S. C., Azevedo, P., Derakhshani, H., DeVries, T. J., Plaizier, J. C., Steele, M. A.

and Khafipour, E. (2017) ‘Weaning age influences the severity of gastrointestinal microbiome

299

shifts in dairy calves.’, Scientific reports, 7(1), p. 198.

Meisel, H. and FitzGerald, R. J. (2000) ‘Opioid peptides encrypted in intact milk protein

sequences.’, The British journal of nutrition, 84(1), pp. 27–31.

Melnyk, S., Fuchs, G. J., Schulz, E., Lopez, M., Kahler, S. G., Fussell, J. J., Bellando, J., Pavliv,

O., Rose, S., Seidel, L., Gaylor, D. W. and James, S. J. (2012) ‘Metabolic imbalance associated

with methylation dysregulation and oxidative damage in children with autism.’, Journal of autism

and developmental disorders, 42(3), pp. 367–77.

Le Merrer, J., Becker, J. A. J., Befort, K. and Kieffer, B. L. (2009) ‘Reward Processing by the

Opioid System in the Brain’, Physiological Reviews, 89(4), pp. 1379–1412.

Merrifield, C. A., Lewis, M. C., Claus, S. P., Pearce, J. T. M., Cloarec, O., Duncker, S.,

Heinzmann, S. S., Dumas, M.-E., Kochhar, S., Rezzi, S., Mercenier, A., Nicholson, J. K., Bailey,

M. and Holmes, E. (2013) ‘Weaning diet induces sustained metabolic phenotype shift in the pig

and influences host response to Bifidobacterium lactis NCC2818.’, Gut, 62(6), pp. 842–51.

Messaoudi, M., Lalonde, R., Violle, N., Javelot, H., Desor, D., Nejdi, A., Bisson, J.-F., Rougeot,

C., Pichelin, M., Cazaubiel, M. and Cazaubiel, J.-M. (2011) ‘Assessment of psychotropic-like

properties of a probiotic formulation (Lactobacillus helveticus R0052 and Bifidobacterium longum

R0175) in rats and human subjects.’, The British journal of nutrition, 105(5), pp. 755–64.

Miguel-Hidalgo, J. J. (2009) ‘The Role of Glial Cells in Drug Abuse.’, Current drug abuse

reviews, 2(1), pp. 76–82.

Miller, B. L. (1991) ‘A review of chemical issues in 1H NMR spectroscopy: N-acetyl-L-aspartate,

creatine and choline.’, NMR in biomedicine, 4(2), pp. 47–52.

300

Miller, M. J., Witherly, S. A. and Clark, D. A. (1990) ‘Casein: a milk protein with diverse biologic

consequences.’, Proceedings of the Society for Experimental Biology and Medicine., 195(2), pp.

143–59.

Ming, X., Stein, T. P., Barnes, V., Rhodes, N. and Guo, L. (2012) ‘Metabolic Perturbance in

Autism Spectrum Disorders: A Metabolomics Study’, Journal of Proteome Research, 11(12), pp.

5856–5862.

Mohanty, D. P., Mohapatra, S., Misra, S. and Sahu, P. S. (2016) ‘Milk derived bioactive peptides

and their impact on human health – A review’, Saudi Journal of Biological Sciences, pp. 577–583.

Molina-Hernández, M., Tellez-Alcántara, N. P., Pérez-García, J., Olivera-Lopez, J. I. and

Jaramillo-Jaimes, M. T. (2008) ‘Antidepressant-like actions of minocycline combined with several

glutamate antagonists’, Progress in Neuro-Psychopharmacology and Biological Psychiatry,

32(2), pp. 380–386.

Moore, L. D., Le, T. and Fan, G. (2013) ‘DNA methylation and its basic function.’,


Moretti, M. (2012) ‘Breastfeeding and the Use of Human Milk’, The American Academy of

Pediatrics, 129(3), pp. 827–841.

Muhammad, B. Y. and Kitchen, I. (1993) ‘Effect of delayed weaning on opioid receptor control

of swim stress-induced antinociception in the developing rat.’, British journal of pharmacology,

109(3), pp. 651–4.

Mulloy, A., Lang, R., O’Reilly, M., Sigafoos, J., Lancioni, G. and Rispoli, M. (2010) ‘Gluten-free

and casein-free diets in the treatment of autism spectrum disorders: A systematic review’,

301

Research in Autism Spectrum Disorders, 4(3), pp. 328–339.

Muntoni, S., Cocco, P., Aru, G. and Cucca, F. (2000) ‘Nutritional factors and worldwide incidence

of childhood type 1 diabetes.’, The American journal of clinical nutrition, 71(6), pp. 1525–9.

Mutch C, Baxter C, James W, Leduc D, Ponti M, Wong D, I. E. (Community P. aediatrics C.

(2004) ‘Weaning from the breast’, Paediatrics & child health, 9(4), pp. 249–253.

Myers, J. S., Pierce, J. and Pazdernik, T. (2008) ‘Neurotoxicology of chemotherapy in relation to

cytokine release, the blood-brain barrier, and cognitive impairment.’, Oncology nursing forum,

35(6), pp. 916–20.

Naidu, P. S., Lichtman, A. H., Archer, C. C., May, E. L., Harris, L. S. and Aceto, M. D. (2007)

‘NIH 11082 produces anti-depressant-like activity in the mouse tail-suspension test through a

delta-opioid receptor mechanism of action.’, European journal of pharmacology, 566(1–3), pp.

132–6.

Nakamura, K., Kikusui, T., Takeuchi, Y. and Mori, Y. (2008) ‘Changes in social instigation- and

food restriction-induced aggressive behaviors and hippocampal 5HT1B mRNA receptor

expression in male mice from early weaning.’, Behavioural brain research, 187(2), pp. 442–8.

Naseribafrouei, A., Hestad, K., Avershina, E., Sekelja, M., Linløkken, A., Wilson, R. and Rudi,

K. (2014) ‘Correlation between the human fecal microbiota and depression’,

Neurogastroenterology & Motility, 26(8), pp. 1155–1162.

Nemeroff, C. B. (2004) ‘Early-Life Adversity, CRF Dysregulation, and Vulnerability to Mood and

Anxiety Disorders.’, Psychopharmacology bulletin, 38(1), pp. 14–20.

Neu, J., Reverte, C. M., Mackey, A. D., Liboni, K., Tuhacek-Tenace, L. M., Hatch, M., Li, N.,

302

Caicedo, R. A., Schatz, D. A. and Atkinson, M. (2005) ‘Changes in intestinal morphology and

permeability in the biobreeding rat before the onset of type 1 diabetes.’, Journal of pediatric

gastroenterology and nutrition, 40(5), pp. 589–95.

Neufeld, K. M., Kang, N., Bienenstock, J. and Foster, J. A. (2011) ‘Reduced anxiety-like behavior

and central neurochemical change in germ-free mice’, Neurogastroenterology & Motility, 23(3),

pp. 55–119.

Neufeld, K. M., Kang, N., Bienenstock, J. and Foster, J. A. (2011) ‘Reduced anxiety-like behavior

and central neurochemical change in germ-free mice.’, Neurogastroenterology and motility : the

official journal of the European Gastrointestinal Motility Society, 23(3), pp. 255–64.

Neumann, I. D. (2002) ‘Involvement of the brain oxytocin system in stress coping: interactions

with the hypothalamo-pituitary-adrenal axis.’, Progress in brain research, 139, pp. 147–62.

Neumann, I. D. (2007) ‘Stimuli and consequences of dendritic release of oxytocin within the

brain’, Biochemical Society Transactions, 35(5), pp. 1252–1257.

Neumann, I. D. (2008) ‘Brain Oxytocin: A Key Regulator of Emotional and Social Behaviours in

Both Females and Males’, Journal of Neuroendocrinology, 20(6), pp. 858–865.

Neumann, I. D., Krömer, S. A., Toschi, N. and Ebner, K. (2000) ‘Brain oxytocin inhibits the

(re)activity of the hypothalamo-pituitary-adrenal axis in male rats: involvement of hypothalamic

and limbic brain regions.’, Regulatory peptides, 96(1–2), pp. 31–8.

Neumann, I. D. and Landgraf, R. (2012) ‘Balance of brain oxytocin and vasopressin: Implications

for anxiety, depression, and social behaviors’, Trends in Neurosciences, pp. 649–659.

Neumann, I. D. and Slattery, D. A. (2016) ‘Oxytocin in General Anxiety and Social Fear: A

303

Translational Approach’, Biological Psychiatry, 79(3), pp. 213–221.

Neumann, I. D., Torner, L. and Wigger, A. (2000) ‘Brain oxytocin: differential inhibition of

neuroendocrine stress responses and anxiety-related behaviour in virgin, pregnant and lactating

rats.’, Neuroscience, 95(2), pp. 567–75.

Neumann, I. D., Toschi, N., Ohl, F., Torner, L. and Krömer, S. A. (2001) ‘Maternal defence as an

emotional stressor in female rats: correlation of neuroendocrine and behavioural parameters and

involvement of brain oxytocin.’, The European journal of neuroscience, 13(5), pp. 1016–24.

Nicholls, A. W., Mortishire-Smith, R. J. and Nicholson, J. K. (2003) ‘NMR spectroscopic-based

metabonomic studies of urinary metabolite variation in acclimatizing germ-free rats.’, Chemical

research in toxicology, 16(11), pp. 1395–404.

Nicholson, J. K., Holmes, E., Kinross, J., Burcelin, R., Gibson, G., Jia, W. and Pettersson, S. (2012)

‘Host-Gut Microbiota Metabolic Interactions’, Science, 336(6086).

Nicholson, J. K., Lindon, J. C. and Holmes, E. (1999) ‘“Metabonomics”: understanding the

metabolic responses of living systems to pathophysiological stimuli via multivariate statistical

analysis of biological NMR spectroscopic data’, Xenobiotica, 29(11), pp. 1181–1189.

Nishioka, M., Bundo, M., Kasai, K. and Iwamoto, K. (2012) ‘DNA methylation in schizophrenia:

progress and challenges of epigenetic studies.’, Genome medicine, 4(12), p. 96.

Nongonierma, A. B. and FitzGerald, R. J. (2015) ‘The scientific evidence for the role of milk

protein-derived bioactive peptides in humans: A Review’, Journal of Functional Foods, pp. 640–

656.

Noni, I. De (2008) ‘Release of [beta]-casomorphins 5 and 7 during simulated gastro-intestinal

304

digestion of bovine [beta]-casein variants and milk-based infant formulas’, Food Chemistry,

110(4), pp. 897–903.

De Noni, I. and Cattaneo, S. (2010) ‘Occurrence of β-casomorphins 5 and 7 in commercial dairy

products and in their digests following in vitro simulated gastro-intestinal digestion’, Food

Chemistry, 119(2), pp. 560–566.

Nowakowska, E., Kus, K., Bobkiewicz-Kozłowska, T., Hertmanowska, H., Bobkiewicz-

kozowska, T. and Hertmanowska, H. (2002) ‘Role of neuropeptides in antidepressant and memory

improving effects of venlafaxine.’, Polish journal of pharmacology, 54(6), pp. 605–13.

Nyberg, F., Carlsson, A. and Hallberg, M. (2013) ‘Casomorphins/Hemorphins’, in Handbook of

Biologically Active Peptides, pp. 1550–1555.

O’Leary, O. F., Felice, D., Galimberti, S., Savignac, H. M., Bravo, J. A., Crowley, T., El Yacoubi,

M., Vaugeois, J.-M., Gassmann, M., Bettler, B., Dinan, T. G. and Cryan, J. F. (2014) ‘GABA B(1)

receptor subunit isoforms differentially regulate stress resilience’, Proceedings of the National

Academy of Sciences, 111(42), pp. 15232–15237.

Ohinata, K., Agui, S. and Yoshikawa, M. (2007) ‘Soymorphins, novel mu opioid peptides derived

from soy beta-conglycinin beta-subunit, have anxiolytic activities.’, Bioscience, biotechnology,

and biochemistry, 71(10), pp. 2618–21.

Okusaga, O., Yolken, R. H., Langenberg, P., Sleemi, A., Kelly, D. L., Vaswani, D., Giegling, I.,

Hartmann, A. M., Konte, B., Friedl, M., Mohyuddin, F., Groer, M. W., Rujescu, D. and Postolache,

T. T. (2013) ‘Elevated gliadin antibody levels in individuals with schizophrenia’, The World

Journal of Biological Psychiatry, 14(7), pp. 509–515.

305

Ottman, N., Smidt, H., de Vos, W. M. and Belzer, C. (2012) ‘The function of our microbiota: who

is out there and what do they do?’, Frontiers in cellular and infection microbiology, 2, p. 104.

Padberg, S., Schumm-Draeger, P.-M., Petzoldt, R., Becker†, F. and Federlin, K. (1999)

‘Wertigkeit von A1- und A2-Antikörpern gegen β-Kasein beim Typ-1-Diabetes mellitus’, DMW -

Deutsche Medizinische Wochenschrift, 124(50), pp. 1518–1521.

De Palma, G., Blennerhassett, P., Lu, J., Deng, Y., Park, A. J., Green, W., Denou, E., Silva, M. A.,

Santacruz, A., Sanz, Y., Surette, M. G., Verdu, E. F., Collins, S. M. and Bercik, P. (2015)

‘Microbiota and host determinants of behavioural phenotype in maternally separated mice’, Nature

Communications, 6, p. 7735.

Palmer, C., Bik, E. M., DiGiulio, D. B., Relman, D. A. and Brown, P. O. (2007) ‘Development of

the Human Infant Intestinal Microbiota’, PLoS Biology, 5(7), p. 177.

Panksepp, J., Normansell, L., Siviy, S., Rossi, J. and Zolovick, A. J. (1984) ‘Casomorphins reduce

separation distress in chicks’, Peptides, 5(4), pp. 829–831.

Park, H. and Poo, M. (2012) ‘Neurotrophin regulation of neural circuit development and function’,

Nature Reviews Neuroscience, 14(1), pp. 7–23.

Parracho, H. M. R. T., Bingham, M. O., Gibson, G. R. and McCartney, A. L. (2005) ‘Differences

between the gut microflora of children with autistic spectrum disorders and that of healthy

children.’, Journal of medical microbiology, 54(Pt 10), pp. 987–91.

Pascual, J. M., Carceller, F., Roda, J. M. and Cerdán, S. (1998) ‘Glutamate, glutamine, and GABA

as substrates for the neuronal and glial compartments after focal cerebral ischemia in rats.’, Stroke,

29(5), pp. 1048–57.

306

Pettitt, D. J., Forman, M. R., Hanson, R. L., Knowler, W. C. and Bennett, P. H. (1997)

‘Breastfeeding and incidence of non-insulin-dependent diabetes mellitus in Pima Indians’, The

Lancet, 350(9072), pp. 166–168.

Pihlanto-Leppälä, A. (2000) ‘Bioactive peptides derived from bovine whey proteins: opioid and

ace-inhibitory peptides’, Trends in Food Science & Technology, 11(9–10), pp. 347–356.

Popoli, M., Yan, Z., McEwen, B. S. and Sanacora, G. (2011) ‘The stressed synapse: the impact of

stress and glucocorticoids on glutamate transmission’, Nature Reviews Neuroscience, 13(1), pp.

22–37.

Pradhan, A. A., Befort, K., Nozaki, C., Gaveriaux-Ruff, C. and Kieffer, B. L. (2011) ‘The delta

opioid receptor: an evolving target for the treatment of brain disorders’, Trends Pharmacological

sciences, 32(1873–3735), pp. 581–590.

Pratt, W. B. (1989) ‘Glucocorticoid receptor structure and the initial events in signal transduction’,

Prog.Clin.Biol.Res., pp. 119–132.

Praveen, P., Jordan, F., Priami, C. and Morine, M. J. (2015) ‘The role of breast-feeding in infant

immune system: a systems perspective on the intestinal microbiome.’, Microbiome, 3(1), p. 41.

Pryce, C. R. and Feldon, J. (2003) ‘Long-term neurobehavioural impact of the postnatal

environment in rats: Manipulations, effects and mediating mechanisms’, in Neuroscience and

Biobehavioral Reviews, pp. 57–71.

Purba, J. S., Hoogendijk, W. J., Hofman, M. A. and Swaab, D. F. (1996) ‘Increased number of

vasopressin- and oxytocin-expressing neurons in the paraventricular nucleus of the hypothalamus

in depression.’, Archives of general psychiatry, 53(2), pp. 137–43.

307

Putignani, L., Del Chierico, F., Vernocchi, P., Cicala, M., Cucchiara, S. and Dallapiccola, B.

(2016) ‘Gut Microbiota Dysbiosis as Risk and Premorbid Factors of IBD and IBS Along the

Childhood–Adulthood Transition’, Inflammatory Bowel Diseases, 22(2), pp. 487–504.

Qin, J., Li, R., Raes, J., Arumugam, M., Burgdorf, K. S., Manichanh, C., Nielsen, T., Pons, N.,

Levenez, F., Yamada, T., Mende, D. R., Li, J., Xu, J., Li, S., Li, D., Cao, J., Wang, B., Liang, H.,

Zheng, H., et al. (2010) ‘A human gut microbial gene catalogue established by metagenomic

sequencing’, Nature, 464(7285), pp. 59–65.

Quinn, R. (2005) ‘Comparing rat’s to human’s age: how old is my rat in people years?’, Nutrition,

21(6), pp. 775–7.

Quinones, M. P. and Kaddurah-Daouk, R. (2009) ‘Metabolomics tools for identifying biomarkers

for neuropsychiatric diseases.’, Neurobiology of disease, 35(2), pp. 165–76.

Raedler, E., Raedler, A. and Feldhaus, S. (1980) ‘Dynamical aspects of neocortical histogenesis

in the rat.’, Anatomy and embryology, 158(3), pp. 253–69.

Rammelkamp, C. H. and Keefer, C. S. (1943) ‘THE ABSORPTION, EXCRETION, AND

DISTRIBUTION OF PENICILLIN.’, The Journal of clinical investigation, 22(3), pp. 425–37.

Reid, L. D. and Hubbell, C. L. (1994) ‘An assessment of the addiction potential of the opioid

associated with milk.’, Journal of dairy science, 77(3), pp. 672–5.

Reigstad, C. S., Salmonson, C. E., Rainey, J. F., Szurszewski, J. H., Linden, D. R., Sonnenburg, J.

L., Farrugia, G. and Kashyap, P. C. (2015) ‘Gut microbes promote colonic serotonin production

through an effect of short-chain fatty acids on enterochromaffin cells’, The FASEB Journal, 29(4),

pp. 1395–1403.

308

Reinscheid, R. K., Nothacker, H. P., Bourson, A., Ardati, A., Henningsen, R. A., Bunzow, J. R.,

Grandy, D. K., Langen, H., Monsma, F. J. and Civelli, O. (1995) ‘Orphanin FQ: a neuropeptide

that activates an opioidlike G protein-coupled receptor.’, Science, 270(5237), pp. 792–4.

Reynolds, L. P. and Caton, J. S. (2012) ‘Role of the pre- and post-natal environment in

developmental programming of health and productivity’, Molecular and Cellular Endocrinology,

354(1–2), pp. 54–59.

Rice, D. and Barone, S. (2000) ‘Critical periods of vulnerability for the developing nervous

system: evidence from humans and animal models.’, Environmental health perspectives, pp. 511–

33.

Richter, S. H., Kästner, N., Loddenkemper, D.-H., Kaiser, S. and Sachser, N. (2016) ‘A Time to

Wean? Impact of Weaning Age on Anxiety-Like Behaviour and Stability of Behavioural Traits in

Full Adulthood’, PLOS ONE, 11(12).

Ridaura, V. and Belkaid, Y. (2015) ‘Gut Microbiota: The Link to Your Second Brain’, Cell,

161(2), pp. 193–194.

Ring, R. H., Malberg, J. E., Potestio, L., Ping, J., Boikess, S., Luo, B., Schechter, L. E., Rizzo, S.,

Rahman, Z. and Rosenzweig-Lipson, S. (2006) ‘Anxiolytic-like activity of oxytocin in male mice:

behavioral and autonomic evidence, therapeutic implications’, Psychopharmacology, 185(2), pp.

218–225.

Rist, V. T. S., Eklund, M., Bauer, E., Sauer, N. and Mosenthin, R. (2012) ‘Effect of feeding level

on the composition of the intestinal microbiota in weaned piglets.’, Journal of animal science,

90(4), pp. 19–21.

309

Rist, V. T. S., Weiss, E., Sauer, N., Mosenthin, R. and Eklund, M. (2014) ‘Effect of dietary protein

supply originating from soybean meal or casein on the intestinal microbiota of piglets.’, Anaerobe,

25, pp. 72–9.

Rodríguez, J. M., Murphy, K., Stanton, C., Ross, R. P., Kober, O. I., Juge, N., Avershina, E., Rudi,

K., Narbad, A., Jenmalm, M. C., Marchesi, J. R. and Collado, M. C. (2015) ‘The composition of

the gut microbiota throughout life, with an emphasis on early life’, Microbial Ecology in Health

& Disease, 26, p. 26050. doi: 10.3402/mehd.v26.26050.

Rojas-Corrales, M. O., Berrocoso, E., Gibert-Rahola, J. and Micó, J. A. (2004) ‘Antidepressant-

Like Effect of tramadol and its Enantiomers in Reserpinized Mice: Comparativestudy with

Desipramine, Fluvoxamine, Venlafaxine and Opiates’, Journal of Psychopharmacology, 18(3),

pp. 404–411.

Romano, A., Tempesta, B., Di Bonaventura, M. V. M. and Gaetani, S. (2016) ‘From autism to

eating disorders and more: The role of oxytocin in neuropsychiatric disorders’, Frontiers in

Neuroscience, p. 497.

Romijn, H. J., Hofman, M. A. and Gramsbergen, A. (1991) ‘At what age is the developing cerebral

cortex of the rat comparable to that of the full-term newborn human baby?’, Early human

development, 26(1), pp. 61–7.

Round, J. L. and Mazmanian, S. K. (2009) ‘The gut microbiota shapes intestinal immune responses

during health and disease’, Nature Reviews Immunology, 9(5), pp. 313–323.

Sachser, N., Hennessy, M. B. and Kaiser, S. (2011) ‘Adaptive modulation of behavioural profiles

by social stress during early phases of life and adolescence’, Neuroscience and Biobehavioral

Reviews, pp. 1518–1533.

310

Sachser, N., Kaiser, S. and Hennessy, M. B. (2013) ‘Behavioural profiles are shaped by social

experience: when, how and why’, Philosophical Transactions of the Royal Society B: Biological

Sciences, 368(1618), pp. 20120344–20120344.

Saio, K., Kamiya, M. and Watanabe, T. (1969) ‘Food Processing Characteristics of Soybean 11S

and 7S Proteins’, Agricultural and Biological Chemistry, 33(9), pp. 1301–1308.

Saito, T. (2008) ‘Antihypertensive Peptides Derived from Bovine Casein and Whey Proteins’, in

Bioactive Components of Milk, pp. 295–317.

Saitoh, A., Sugiyama, A., Nemoto, T., Fujii, H., Wada, K., Oka, J.-I., Nagase, H. and Yamada, M.

(2011) ‘The novel δ opioid receptor agonist KNT-127 produces antidepressant-like and

antinociceptive effects in mice without producing convulsions’, Behavioural Brain Research,

223(2), pp. 271–279.

Sandoval-Salazar, C., Ramírez-Emiliano, J., Trejo-Bahena, A., Oviedo-Solís, C. I. and Solís-Ortiz,

M. S. (2016) ‘A high-fat diet decreases GABA concentration in the frontal cortex and

hippocampus of rats’, Biological research, 49, p. 15. doi: 10.1186/s40659-016-0075-6.

Schanen, N. C. (2006) ‘Epigenetics of autism spectrum disorders’, Human Mo, 15(2), pp. 138–

150.

Schmauss, C. and Emrich, H. M. (1985) ‘Dopamine and the action of opiates: a reevaluation of

the dopamine hypothesis of schizophrenia. With special consideration of the role of endogenous

opioids in the pathogenesis of schizophrenia.’, Biological psychiatry, 20(11), pp. 1211–31.

Schneider, R. L., Schiml, P. A., Deak, T. and Hennessy, M. B. (2012) ‘Persistent sensitization of

depressive-like behavior and thermogenic response during maternal separation in pre- and post-

311

weaning guinea pigs.’, Developmental psychobiology, 54(5), pp. 514–22.

Scott, F. W. (1990) ‘Cow milk and insulin-dependent diabetes mellitus: is there a relationship?’,

The American journal of clinical nutrition, 51(3), pp. 489–91.

Sengupta, P. (2013) ‘The Laboratory Rat: Relating Its Age With Human’s.’, International journal

of preventive medicine, 4(6), pp. 624–30.

Sethi, S. and Brietzke, E. (2016) ‘Omics-Based Biomarkers: Application of Metabolomics in

Neuropsychiatric Disorders’, International Journal of Neuropsychopharmacology, 19(3), p.

pyv096.

Shah, N. P. (2000) ‘Effects of milk-derived bioactives: an overview.’, The British journal of

nutrition, 84(1), pp. 3–10.

Sharman, D. F., Mann, S. P., Fry, J. P., Banns, H. and Stephens, D. B. (1982) ‘Cerebral dopamine

metabolism and stereotyped behaviour in early-weaned piglets’, Neuroscience, 7(8), pp. 1937–

1944.

Shimozuru, M., Kodama, Y., Iwasa, T., Kikusui, T., Takeuchi, Y. and Mori, Y. (2007) ‘Early

weaning decreases play-fighting behavior during the postweaning developmental period of Wistar

rats.’, Developmental psychobiology, 49(4), pp. 343–50.

Silva, M. de L. C., Speridião, P. da G. L., Marciano, R., Amâncio, O. M. S., de Morais, T. B. and

de Morais, M. B. (2015) ‘Effects of soy beverage and soy-based formula on growth, weight, and

fecal moisture: experimental study in rats’, Jornal de Pediatria, 91(3), pp. 306–312.

Singh, M. M. and Kay, S. R. (1976) ‘Wheat gluten as a pathogenic factor in schizophrenia.’,

Science, 191(4225), pp. 401–2.

312

Singh, M., Rosen, C. L., Chang, K. J. and Haddad, G. G. (1989) ‘Plasma β-Casomorphin-7

Immunoreactive Peptide Increases after Milk Intake in Newborn but not in Adult Dogs’, Pediatric

Research, 26(1), pp. 34–38.

Singhal, G. and Baune, B. T. (2017) ‘Microglia: An Interface between the Loss of Neuroplasticity

and Depression’, Frontiers in Cellular Neuroscience, 11, p. 270.

Slattery, D. A. and Neumann, I. D. (2010) ‘Oxytocin and Major Depressive Disorder:

Experimental and Clinical Evidence for Links to Aetiology and Possible Treatment.’,

Pharmaceuticals (Basel, Switzerland), 3(3), pp. 702–724.

Smotherman, W. P. and Robinson, S. R. (1992) ‘Prenatal experience with milk: fetal behavior and

endogenous opioid systems.’, Neuroscience and biobehavioral reviews, 16(3), pp. 351–64.

Sokolov, O., Kost, N., Andreeva, O., Korneeva, E., Meshavkin, V., Tarakanova, Y., Dadayan, A.,

Zolotarev, Y., Grachev, S., Mikheeva, I., Varlamov, O. and Zozulya, A. (2014) ‘Autistic children

display elevated urine levels of bovine casomorphin-7 immunoreactivity’, Peptides, 56, pp. 68–

71.

Song, P. and Zhao, Z. Q. (2001) ‘The involvement of glial cells in the development of morphine

tolerance.’, Neuroscience research, 39(3), pp. 281–6.

Spain, J. W., Roth, B. L. and Coscia, C. J. (1985) ‘Differential ontogeny of multiple opioid

receptors (mu, delta, and kappa).’, The Journal of neuroscience, 5(3), pp. 584–8.

Steinerová, A., Racek, J., Stožický, F., Tatzber, F. and Lapin, A. (1999) ‘Autoantibodies against

Oxidized LDL in the First Phase of Life’, Clinical Chemistry and Laboratory Medicine, 37(9), pp.

913–7.

313

Stevens, E. E., Patrick, T. E. and Pickler, R. (2009) ‘A history of infant feeding.’, The Journal of

perinatal education, 18(2), pp. 32–9.

Stuart-Macadam, P. and Dettwyler, K. A. (1995) Breastfeeding : biocultural perspectives.

Sudo, N., Chida, Y., Aiba, Y., Sonoda, J., Oyama, N., Yu, X.-N., Kubo, C. and Koga, Y. (2004)

‘Postnatal microbial colonization programs the hypothalamic-pituitary-adrenal system for stress

response in mice.’, The Journal of physiology, 558(1), pp. 263–75.

Sun, Z. and Cade, J. R. (1999) ‘A Peptide Found in Schizophrenia and Autism Causes Behavioral

Changes in Rats’, Autism, 3(1), pp. 85–95.

Sun, Z., Zhang, Z., Wang, X., Cade, R., Elmir, Z. and Fregly, M. (2003) ‘Relation of beta-

casomorphin to apnea in sudden infant death syndrome.’, Peptides, 24(6), pp. 937–43.

Swann, J. R., Tuohy, K. M., Lindfors, P., Brown, D. T., Gibson, G. R., Wilson, I. D., Sidaway, J.,

Nicholson, J. K. and Holmes, E. (2011) ‘Variation in antibiotic-induced microbial recolonization

impacts on the host metabolic phenotypes of rats.’, Journal of proteome research, 10(8), pp. 3590–

603.

Tailford, K. A., Berry, C. L., Thomas, A. C. and Campbell, J. H. (2003) ‘A casein variant in cow’s

milk is atherogenic.’, Atherosclerosis, 170(1), pp. 13–9.

Tana, C., Umesaki, Y., Imaoka, A., Handa, T., Kanazawa, M. and Fukudo, S. (2010) ‘Altered

profiles of intestinal microbiota and organic acids may be the origin of symptoms in irritable bowel

syndrome.’, Neurogastroenterology and motility, 22(5), pp. 512–9, 114–5.

Tanoue, Y. and Oda, S. (1989) ‘Weaning time of children with infantile autism.’, Journal of autism

and developmental disorders, 19(3), pp. 425–34.

314

Tau, G. Z. and Peterson, B. S. (2010) ‘Normal development of brain circuits.’,


Tejedor-Real, P., Mico, J. A., Maldonado, R., Roques, B. P. and Gibert-Rahola, J. (1995)

‘Implication of endogenous opioid system in the learned helplessness model of depression.’,

Pharmacology, biochemistry, and behavior, 52(1), pp. 145–52.

Tenore, P. L. (2008) ‘Psychotherapeutic Benefits of Opioid Agonist Therapy’, Journal of

Addictive Diseases, pp. 49–65.

Terpstra, A. H., Harkes, L. and van der Veen, F. H. (1981) ‘The effect of different proportions of

casein in semipurified diets on the concentration of serum cholesterol and the lipoprotein

composition in rabbits.’, Lipids, 16(2), pp. 114–9.

Terranova, M. L. and Laviola, G. (2001) ‘delta-Opioid modulation of social interactions in juvenile

mice weaned at different ages.’, Physiology & behavior, 73(3), pp. 393–400.

Teschemacher, H. (2003) ‘Opioid receptor ligands derived from food proteins.’, Current

pharmaceutical design, 9(16), pp. 1331–44.

Teschemacher, H., Koch, G. and Brantl, V. (1997) ‘Milk protein-derived opioid receptor ligands’,

Biopolymers, 43(2), pp. 99–117.

The A2 Milk Company (2014) The A2 Milk Company.

Thiel, C. M., Huston, J. P. and Schwarting, R. K. W. (1998) ‘Hippocampal acetylcholine and

habituation learning’, Neuroscience, 85(4), pp. 1253–1262.

Thiels, E., Alberts, J. R. and Cramer, C. P. (1990) ‘Weaning in rats: II. Pup behavior patterns’,

Developmental Psychobiology, 23(6), pp. 495–510.

315

Thomas, C. M., Hong, T., van Pijkeren, J. P., Hemarajata, P., Trinh, D. V., Hu, W., Britton, R. A.,

Kalkum, M. and Versalovic, J. (2012) ‘Histamine Derived from Probiotic Lactobacillus reuteri

Suppresses TNF via Modulation of PKA and ERK Signaling’, PLoS ONE, 7(2), p. 31951.

Thomas, R. H., Meeking, M. M., Mepham, J. R., Tichenoff, L., Possmayer, F., Liu, S. and

MacFabe, D. F. (2012) ‘The enteric bacterial metabolite propionic acid alters brain and plasma

phospholipid molecular species: further development of a rodent model of autism spectrum

disorders’, Journal of Neuroinflammation, 9(1), p. 695.

Thompson, P. L., Gilbert, R. E., Long, P. F., Saxena, S., Sharland, M. and Wong, I. C. K. (2008)

‘Has UK guidance affected general practitioner antibiotic prescribing for otitis media in children?’,

Journal of Public Health, 30(4), pp. 479–486.

Thorning, T. K., Raben, A., Tholstrup, T., Soedamah-Muthu, S. S., Givens, I. and Astrup, A.

(2016) ‘Milk and dairy products: good or bad for human health? An assessment of the totality of

scientific evidence’, Food & Nutrition Research, 60(1), p. 32527.

Thorsdottir, I., Birgisdottir, B. E., Johannsdottir, I. M., Harris, D. P., Hill, J., Steingrimsdottir, L.

and Thorsson, A. V (2000) ‘Different beta-casein fractions in Icelandic versus Scandinavian cow’s

milk may influence diabetogenicity of cow’s milk in infancy and explain low incidence of insulin-

dependent diabetes mellitus in Iceland.’, Pediatrics, 106(4), pp. 719–24.

Tierney, M. L., Bray, E. A., Allen, R. D., Ma, Y., Drong, R. F., Slightom, J. and Beachy, R. N.

(1987) ‘Isolation and characterization of a genomic clone encoding the b-subunit of b-

conglycinin’, Planta, 172(3), pp. 356–363.

Tognini, P. (2017) ‘Gut Microbiota: A Potential Regulator of Neurodevelopment’, Frontiers in

Cellular Neuroscience, 11, p. 25.

316

Tordjman, S., Anderson, G. M., Botbol, M., Brailly-Tabard, S., Perez-Diaz, F., Graignic, R.,

Carlier, M., Schmit, G., Rolland, A.-C., Bonnot, O., Trabado, S., Roubertoux, P. and Bronsard, G.

(2009) ‘Pain Reactivity and Plasma β-Endorphin in Children and Adolescents with Autistic

Disorder’, PLoS ONE, 4(8), p. 5289.

Torreilles, J. and Guérin, M. C. (1995) ‘Casein-derived peptides can promote human LDL

oxidation by a peroxidase-dependent and metal-independent process’, Comptes rendus des

seances de la Societe de biologie et de ses filiales, 189(5), pp. 933–42.

Tremaroli, V. and Bäckhed, F. (2012) ‘Functional interactions between the gut microbiota and

host metabolism’, Nature, 489(7415), pp. 242–249.

Tribollet, E., Barberis, C., Jard, S., Dubois-Dauphin, M. and Dreifuss, J. J. (1988) ‘Localization

and pharmacological characterization of high affinity binding sites for vasopressin and oxytocin

in the rat brain by light microscopic autoradiography.’, Brain research, 442(1), pp. 105–18.

Tribollet, E., Charpak, S., Schmidt, A., Dubois-Dauphin, M. and Dreifuss, J. J. (1989) ‘Appearance

and transient expression of oxytocin receptors in fetal, infant, and peripubertal rat brain studied by

autoradiography and electrophysiology.’, The Journal of neuroscience, 9(5), pp. 1764–73.

Trivedi, M. S., Hodgson, N. W., Walker, S. J., Trooskens, G., Nair, V. and Deth, R. C. (2015)

‘Epigenetic effects of casein-derived opioid peptides in SH-SY5Y human neuroblastoma cells’,

Nutrition & Metabolism, 12(1), p. 54.

Trivedi, M. S., Shah, J. S., Al-Mughairy, S., Hodgson, N. W., Simms, B., Trooskens, G. A., Van

Criekinge, W. and Deth, R. C. (2014) ‘Food-derived opioid peptides inhibit cysteine uptake with

redox and epigenetic consequences.’, The Journal of nutritional biochemistry, 25(10), pp. 1011–

8.

317

ul Haq, M. R., Kapila, R., Shandilya, U. K. and Kapila, S. (2014) ‘Impact of Milk Derived β-

Casomorphins on Physiological Functions and Trends in Research: A Review’, International

Journal of Food Properties, 17(8), pp. 1726–1741.

Umbach, M., Teschemacher, H., Praetorius, K., Hirschhäuser, R. and Bostedt, H. (1985)

‘Demonstration of a beta-casomorphin immunoreactive material in the plasma of newborn calves

after milk intake.’, Regulatory peptides, 12(3), pp. 223–30.

Umesaki, Y. and Setoyama, H. (2000) ‘Structure of the intestinal flora responsible for development

of the gut immune system in a rodent model.’, Microbes and infection, 2(11).

Vaarala, O. (2008) ‘Leaking gut in type 1 diabetes’, Current Opinion in Gastroenterology, 24(6),

pp. 701–706.

Vaarala, O., Atkinson, M. A. and Neu, J. (2008) ‘The “perfect storm” for type 1 diabetes: The

complex interplay between intestinal microbiota, gut permeability, and mucosal immunity’,

Diabetes, pp. 2555–2562.

Vazquez, D. M., Bailey, C., Dent, G. W., Okimoto, D. K., Steffek, A., López, J. F. and Levine, S.

(2006) ‘Brain corticotropin-releasing hormone (CRH) circuits in the developing rat: Effect of

maternal deprivation’, Brain Research, 1121(1), pp. 83–94.

Verdú, E. F., Bercik, P., Verma-Gandhu, M., Huang, X.-X., Blennerhassett, P., Jackson, W., Mao,

Y., Wang, L., Rochat, F. and Collins, S. M. (2006) ‘Specific probiotic therapy attenuates antibiotic

induced visceral hypersensitivity in mice.’, Gut, 55(2), pp. 182–90.

Viero, C., Shibuya, I., Kitamura, N., Verkhratsky, A., Fujihara, H., Katoh, A., Ueta, Y., Zingg, H.

H., Chvatal, A., Sykova, E. and Dayanithi, G. (2010) ‘REVIEW: Oxytocin: Crossing the bridge

318

between basic science and pharmacotherapy.’, CNS neuroscience & therapeutics, 16(5), pp. 138–

56.

Virtanen, S. M., Hyppönen, E., Läärä, E., Vähäsalo, P., Kulmala, P., Savola, K., Räsänen, L., Aro,

A., Knip, M. and Åkerblom, H. K. (1998) ‘Cow’s milk consumption, disease-associated

autoantibodies and Type 1 diabetes mellitus: a follow-up study in siblings of diabetic children’,

Diabetic Medicine, 15(9), pp. 730–738.

Viviani, D., Charlet, A., van den Burg, E., Robinet, C., Hurni, N., Abatis, M., Magara, F. and

Stoop, R. (2011) ‘Oxytocin selectively gates fear responses through distinct outputs from the

central amygdala.’, Science, 333(6038), pp. 104–7.

Voth, D. E. and Ballard, J. D. (2005) ‘Clostridium difficile Toxins: Mechanism of Action and Role

in Disease’, Clinical Microbiology Reviews, 18(2), pp. 247–263.

van der Waaij, D. and Sturm, C. A. (1968) ‘Antibiotic decontimination of the digestive tract of

mice. Technical procedures.’, Laboratory animal care, 18(1), pp. 1–10.

Wada, Y. and Lönnerdal, B. (2015) ‘Bioactive peptides released from in vitro digestion of human

milk with or without pasteurization’, Pediatric Research, 77(4), pp. 546–553.

Wagner, M., Horn, M. and Daims, H. (2003) ‘Fluorescence in situ hybridisation for the

identification and characterisation of prokaryotes’, Current Opinion in Microbiology, 6(3), pp.

302–309. doi: 10.1016/S1369-5274(03)00054-7.

Walker, W. A. and Shuba Iyengar, R. (2014) ‘Breastmilk, Microbiota and Intestinal Immune

Homeostasis’, Pediatric Research, 77(1–2), p. 220.

Walsh, M. C., Brennan, L., Malthouse, J. P. G., Roche, H. M. and Gibney, M. J. (2006) ‘Effect of

319

acute dietary standardization on the urinary, plasma, and salivary metabolomic profiles of healthy

humans.’, The American journal of clinical nutrition. American Society for Nutrition, 84(3), pp.

531–9.

Wang, H., Liang, S., Wang, M., Gao, J., Sun, C., Wang, J., Xia, W., Wu, S., Sumner, S. J., Zhang,

F., Sun, C. and Wu, L. (2016) ‘Potential serum biomarkers from a metabolomics study of autism.’,

Journal of psychiatry & neuroscience, 41(1), pp. 27–37.

Weaver, L. T. (1992) ‘Breast and gut: the relationship between lactating mammary function and

neonatal gastrointestinal function’, Proceedings of the Nutrition Society, 51(2), pp. 155–163.

WEISKRANTZ, L. (1956) ‘Behavioral changes associated with ablation of the amygdaloid

complex in monkeys.’, Journal of comparative and physiological psychology, 49(4), pp. 381–91.

Wellock, I. J., Fortomaris, P. D., Houdijk, J. G. M. and Kyriazakis, I. (2006) ‘The effect of dietary

protein supply on the performance and risk of post-weaning enteric disorders in newly weaned

pigs’, Animal Science, 82(3).

Wessler, I. and Kirkpatrick, C. J. (2008) ‘Acetylcholine beyond neurons: the non-neuronal

cholinergic system in humans.’, British journal of pharmacology, 154(8), pp. 1558–71.

Whipp, L. and Daneshkhu, S. (2016) ‘Big business identifies appetite for plant-based milk’,

Financial Times, p. https://www.ft.com/content/7df72c04-491a-11e6-8d68.

Whistler, J. L. and von Zastrow, M. (1998) ‘Morphine-activated opioid receptors elude

desensitization by beta-arrestin.’, Proceedings of the National Academy of Sciences of the United

States of America, 95(17), pp. 9914–9.

Whiteley, P., Haracopos, D., Knivsberg, A.-M., Reichelt, K. L., Parlar, S., Jacobsen, J., Seim, A.,

320

Pedersen, L., Schondel, M. and Shattock, P. (2010) ‘The ScanBrit randomised, controlled, single-

blind study of a gluten- and casein-free dietary intervention for children with autism spectrum

disorders’, Nutritional Neuroscience, 13(2), pp. 87–100.

Whiteley, P. and Shattock, P. (2002) ‘Biochemical aspects in autism spectrum disorders: updating

the opioid-excess theory and presenting new opportunities for biomedical intervention’, Expert

Opinion on Therapeutic Targets, 6(2), pp. 175–183.

WHO (2018) Exclusive breastfeeding for optimal growth, development and health of infants,

WHO. World Health Organization.

Wikoff, W. R., Anfora, A. T., Liu, J., Schultz, P. G., Lesley, S. A., Peters, E. C. and Siuzdak, G.

(2009) ‘Metabolomics analysis reveals large effects of gut microflora on mammalian blood

metabolites.’, Proceedings of the National Academy of Sciences of the United States of America,

106(10), pp. 3698–703.

Williams, J. T., Ingram, S. L., Henderson, G., Chavkin, C., von Zastrow, M., Schulz, S., Koch, T.,

Evans, C. J. and Christie, M. J. (2013) ‘Regulation of μ-opioid receptors: desensitization,

phosphorylation, internalization, and tolerance.’, Pharmacological reviews, 65(1), pp. 223–54.

Wittert, G., Hope, P. and Pyle, D. (1996) ‘Tissue Distribution of Opioid Receptor Gene Expression

in the Rat’, Biochemical and Biophysical Research Communications, 218(3), pp. 877–881.

Wold, S., Albano, C. W. J. D., Dunn III, W. J., Edlund, U., Esbensen, K., Geladi, P., Hellberg, S.,

Johansson, E., Lindberg, W. and Sjöström, M. (1984) ‘Multivariate data analysis in chemistry’, in

Chemometrics, pp. 17–95.

Woodford, K. (2007) ‘Devil in the milk. Illness, health and politics. A1 and A2 milk’, p. 237.

321

World Health Organization (2008) ‘The Global Burden of Disease: 2004 update’, 2004 Update, p.

146.

Wotjak, C. T., Ganster, J., Kohl, G., Holsboer, F., Landgraf, R. and Engelmann, M. (1998)

‘Dissociated central and peripheral release of vasopressin, but not oxytocin, in response to repeated

swim stress: new insights into the secretory capacities of peptidergic neurons.’, Neuroscience,

85(4), pp. 1209–22.

Wyss, M. and Kaddurah-Daouk, R. (2000) ‘Creatine and creatinine metabolism.’, Physiological

reviews, 80(3), pp. 1107–213.

Yang, J., Chen, T., Sun, L., Zhao, Z., Qi, X., Zhou, K., Cao, Y., Wang, X., Qiu, Y., Su, M., Zhao,

A., Wang, P., Yang, P., Wu, J., Feng, G., He, L., Jia, W. and Wan, C. (2013) ‘Potential metabolite

markers of schizophrenia.’, Molecular psychiatry, 18(1), pp. 67–78.

Yang, Q.-Z., Lu, S.-S., Tian, X.-Z., Yang, A.-M., Ge, W.-W. and Chen, Q. (2011) ‘The

antidepressant-like effect of human opiorphin via opioid-dependent pathways in mice.’,

Neuroscience letters, 489(2), pp. 131–5.

Yankelevitch-Yahav, R., Franko, M., Huly, A. and Doron, R. (2015) ‘The forced swim test as a

model of depressive-like behavior.’, Journal of visualized experiments : JoVE, (97).

Yao, J. K., Leonard, S. and Reddy, R. (2006) ‘Altered glutathione redox state in schizophrenia.’,

Disease markers, 22(1–2), pp. 83–93.

Yatsunenko, T., Rey, F. E., Manary, M. J., Trehan, I., Dominguez-Bello, M. G., Contreras, M.,

Magris, M., Hidalgo, G., Baldassano, R. N., Anokhin, A. P., Heath, A. C., Warner, B., Reeder, J.,

Kuczynski, J., Caporaso, J. G., Lozupone, C. A., Lauber, C., Clemente, J. C., Knights, D., et al.

322

(2012) ‘Human gut microbiome viewed across age and geography’, Nature, 486(7402), pp. 222–

7.

Yeomans, M. R. and Gray, R. W. (2002) ‘Opioid peptides and the control of human ingestive

behaviour.’, Neuroscience and biobehavioral reviews, 26(6), pp. 713–28.

Yoo, J.-H., Lee, S.-Y., Loh, H. H., Ho, I. K. and Jang, C.-G. (2004) ‘Altered emotional behaviors

and the expression of 5-HT1A and M1 muscarinic receptors in micro-opioid receptor knockout

mice.’, Synapse, 54(2), pp. 72–82.

Yu, Y., Zhang, L., Yin, X., Sun, H., Uhl, G. R. and Wang, J. B. (1997) ‘Mu opioid receptor

phosphorylation, desensitization, and ligand efficacy.’, The Journal of biological chemistry,

272(46), pp. 28869–74.

Zanardo, V., Nicolussi, S., Carlo, G., Marzari, F., Faggian, D., Favaro, F. and Plebani, M. (2001)

‘Beta endorphin concentrations in human milk.’, Journal of pediatric gastroenterology and

nutrition, 33(2), pp. 160–4.

Zeisel, S. H. (1981a) ‘Dietary Choline: Biochemistry, Physiology, and Pharmacology’, Annual

Review of Nutrition, 1(1), pp. 95–121.

Zeisel, S. H. (1981b) ‘Dietary Choline: Biochemistry, Physiology, and Pharmacology’, Annual

Review of Nutrition, 1(1), pp. 95–121.

Zeisel, S. H. (2006) ‘Choline: critical role during fetal development and dietary requirements in

adults.’, Annual review of nutrition, 26(1), pp. 229–50.

Zhang, H., Torregrossa, M. M., Jutkiewicz, E. M., Shi, Y.-G., Rice, K. C., Woods, J. H., Watson,

S. J. and Holden Ko, M. C. (2006) ‘Endogenous opioids upregulate brain-derived neurotrophic

323

factor mRNA through δ- and µ-opioid receptors independent of antidepressant-like effects’,

European Journal of Neuroscience, 23(4), pp. 984–994.

Zheng, L.-S., Kaneko, N. and Sawamoto, K. (2015) ‘Minocycline treatment ameliorates

interferon-alpha- induced neurogenic defects and depression-like behaviors in mice.’, Frontiers in

cellular neuroscience, 9, p. 5.

Zheng, P., Wang, Y., Chen, L., Yang, D., Meng, H., Zhou, D., Zhong, J., Lei, Y., Melgiri, N. D.

and Xie, P. (2013) ‘Identification and validation of urinary metabolite biomarkers for major

depressive disorder.’, Molecular & cellular proteomics : MCP, 12(1), pp. 207–14.

Zheng, S., Yu, M., Lu, X., Huo, T., Ge, L., Yang, J., Wu, C. and Li, F. (2010) ‘Urinary

metabonomic study on biochemical changes in chronic unpredictable mild stress model of

depression’, Clinica Chimica Acta, 411(3–4), pp. 204–209.

Zheng, S., Zhang, S., Yu, M., Tang, J., Lu, X., Wang, F., Yang, J. and Li, F. (2011) ‘An 1H NMR

and UPLC–MS-based plasma metabonomic study to investigate the biochemical changes in

chronic unpredictable mild stress model of depression’, Metabolomics, 7(3), pp. 413–423.

Zhu, Y., Lin, X., Zhao, F., Shi, X., Li, H., Li, Y., Zhu, W., Xu, X., Lu, C. and Zhou, G. (2015)

‘Meat, dairy and plant proteins alter bacterial composition of rat gut bacteria’, Scientific Reports,

5, p. 15220. doi: 10.1038/srep15220.

Zioudrou, C., Streaty, R. A. and Klee, W. A. (1979) ‘Opioid peptides derived from food proteins.

The exorphins.’, The Journal of biological chemistry, 254(7), pp. 2446–9.

Zoghbi, S., Trompette, A., Claustre, J., El Homsi, M., Garzón, J., Jourdan, G., Scoazec, J.-Y. and

Plaisancié, P. (2006) ‘beta-Casomorphin-7 regulates the secretion and expression of

324

gastrointestinal mucins through a mu-opioid pathway.’, American journal of physiology.

Gastrointestinal and liver physiology, 290(6), pp. 1105–13.

Aya Osman - University of Surreyepubs.surrey.ac.uk/846301/1/Aya Osman Thesis V1.0... · casein free...

Documents

Transcript of Aya Osman - University of Surreyepubs.surrey.ac.uk/846301/1/Aya Osman Thesis V1.0... · casein free...