Effects of Changing Branched-Chain Amino Acid and Insulin Levels In Vitro on Developing Pre-implantation Embryos

Effects of Changing Branched-Chain Amino Acid and Insulin Levels In Vitro on Developing Pre-

implantation Embryos

By Tristan Demuth

Biology Bsc 2013

Supervisor: Professor Tom P. Fleming

Word Count: 9954

i

Summary

If unborn babies (fetuses) are exposed to poor conditions in the uterus, they adapt

their development which can cause them to become more susceptible to diseases such as high

blood pressure and diabetes in adult life. This theory is called the ‘Developmental Theory of

Health and Disease’. The most common way that a fetus may be exposed to a poor

environment is through its mother’s diet.

My study will aid the understanding of how young embryos interact with their

environment during pregnancy. This is significant because it has been shown that if embryos

are exposed to a poor environment in the uterus, specifically during just the first few days of

pregnancy, then they are still at a greater risk of disease once they reach adult life. Therefore

it is important that the way which young embryos communicate with their environment in the

womb is better understood. A greater understanding will allow development of better medical

treatments and improve public health when women become better informed about dietary

requirements during pregnancy.

This project focussed on the effects of insulin and branched-chain amino acid levels

in the maternal environment. Branched-chain amino acids are a specific group of amino

acids, which are the building blocks of proteins. This was because, a study using mice

recently revealed that the level of both insulin and branched-chain amino acids available to

the embryo, are greatly reduced when the mother’s diet is poor (by means of a low protein

content). Therefore in this experiment, embryos were incubated in solutions containing

different levels of branched-chain amino acids and insulin. The results should show whether a

reduction in branched-chain amino acids and/or insulin is sufficient to cause the fetus to adapt

its own development. As stated above this adapted development indicates that the offspring

will be predisposed to disease once it reaches adulthood.

ii

To measure whether the embryo had changed its development, the number of cells in

the embryo were analysed. For example, an increase in a specific cell type known as

‘trophectoderm’ cells would indicate that the embryo had adapted its development.

The results from this experiment showed that reducing the levels of branched-chain

amino acids and insulin available to the early embryo is not sufficient to cause the embryo to

adapt its development. Therefore, further experiments are required to test other possible

methods that the early embryo may be using to detect a poor maternal environment.

Word Count: 399

iii

Abstract

The Developmental Origins of Health and Disease hypothesis states that challenges in

the maternal environment cause the developing fetus to undergo a predictive adaptive

response which can predispose the offspring to chronic disease in later life. The challenge

addressed here is maternal diet, specifically a low protein diet during the initial stages of

gestation, as the methods by which the maternal environment signals to the preimplantation

embryo are of great interest.

When mouse dams are fed a low protein diet during the preimplantation period it

causes a significant drop in the concentrations of branched-chain amino acids and insulin in

the maternal uterine fluid and serum. Therefore the focus of this experiment was to conclude

whether reducing the concentrations of branched-chain amino acids and insulin in vitro is

sufficient to cause the initiation of fetal programming, by culturing embryos from the 2-cell

to the late blastocyst stage in four different treatment groups. Blastocysts were analysed by

observation of developmental stage then by differential cell staining of late blastocysts

followed by fluorescence microscopy to measure cell number of trophectoderm and inner cell

mass. Statistical analysis was performed via chi-squared and Kruskal-Wallis tests.

The results from this experiment concluded that depletion of branched-chain amino

acid and insulin levels was not sufficient to initiate fetal programming by the late blastocyst

stage. A potential role for insulin levels in blastocyst nutrient sensing was observed but not a

significant one. This result means that further analysis of nutrient sensing in preimplantation

embryos is required.

Word Count: 248

iv

Table of Contents

v

Title Page……………………………………………………………………………………………………………

Summary…………………………………………………………………………………………………………….

Abstract………………………………………………………………………………………………………………

Table of Contents……………………………………………………………………………………………….

List of Abbreviations…………………………………………………………………………………………..

Acknowledgements…………………………………………………………………………………………….

1. Introduction

1.1 Developmental Origins of Health and Disease…………………………………………….

1.2 Developmental Re-programming caused by maternal under nutrition……….

1.3 Developmental Re-programming caused by maternal under nutrition

and culture conditions specifically during the preimplantation period………

1.4 Potential role for AA and insulin levels in induction of preimplantation

fetal programming………………………………………………………………………………………

1.5 Mammalian Target of Rapamycin (mTOR) regulation of cell growth and

development………………………………………………………………………………………..

1.6 Differing TE cell numbers as an indication of fetal programming…………………

1.7 Aims and objectives…………………………………………………………………………………….

2. Materials and Methods

2.1 Creation of culture media for four treatment groups………………………………….

2.2 Dissection and procuring of 2-cell stage embryos……………………………………….

2.3 Differential cell Staining

2.3.1 Differential Cell Staining Materials………………………………………………………….

2.3.2 Differential Cell Staining Protocol…………………………………………………………..

2.4 Blastocyst picture acquisition……………………………………………………………………..

2.5 Statistical analysis……………………………………………………………………………………….

3. Results

3.1 Activity of developing embryos…………………………………………………………………..

3.2 Effect of depleted branched-chain amino acids and insulin on fetal

programming…………………………………………………………………………………….....…….

i

ii

iv

v

vii

viii

1

3

4

8

10

13

13

15

16

16

17

18

19

20

21

vi

24

26

30

33

34

38

40

4. Discussion

4.1 Comparison with Original Hypothesis…………………………………………………………

4.2 Potential Methods used by the Preimplantation Embryo to

Communicate with the Maternal Environment after an Emb-LPD………………..

4.3 Evaluation of Experimental Techniques………………………………………………………..

4.4 Effects of In Vitro Culture……………………………………………………………………………..

4.5 Future Work…………………………………………………………………………………………………

4.6 Concluding Remarks …………………………………………………………………………………….

5. References……………………………………………………………………………………………………………

List of Abbreviations

ACE - Angiotensin converting enzyme

Akt – Protein Kinase B

Anti-DNP – Anti Dinitrophenyl

BSA – Bovine serum albumin

CO2 – Carbon Dioxide

DNA – Deoxyribonucleic Acid

DOHaD – Developmental Origins of Health and Disease

E3.5 – Day 3.5 of embryo development

eIF# - Eukaryotic translation initiation factor

eIF4BP1 - Eukaryotic translation initiation factor binding protein 1

Emb-LPD – Low protein diet (9% caesin) fed exclusively during the preimplantation period

GR – Glucocorticoid receptor

ICM – Inner cell mass

IGF-1 – Insulin-like growth factor 1

KSOM – K Simplex optimization media

mRNA – messenger Ribonucleic acid

mTOR – Mammalian target of rapamycin

mTORC – Mammalian target of rapamycin complex

NPD – normal protein diet

PAR – Predictive adaptive response

PCR – Polymerase chain reaction

PI – Propidium Iodide

PIC – Preimplantation initiation complex

PI3K - Phosphotidylinositide 3-kinase 

PPAR - Peroxisomal proliferator-activated receptor

PVP – Polyvinylpyrrolidone

vii

TC – Total Cells

TE – Trophectoderm

TNBS – Trinitrobenzene sulphonic acid

TSC1/2 – Tuberculosis sclerosis complex 1/2

UF – Uterine Fluid

Acknowledgements

I would first like to thank Professor Tom Fleming of the University of Southampton,

for the help and guidance he has provided throughout this project and additionally for

allowing me to use his laboratory at Southampton General Hospital to perform my

experiments. I would also like to thank Miguel Velazquez, research fellow at University of

Southampton for all the experimental training that he provided and for all the guidance given

to me throughout the course of the project.

viii

1. Introduction

1.1 The Developmental Origins of Health and Disease.

The developmental origin of health and disease (DOHaD) is a theory which states that

environmental challenges during the embryo’s early development, particularly maternal

undernutrition, trigger fetal programming events to aid fetal development. However this also

leads to an increased likelihood of several diseases in adulthood, including metabolic

syndrome, cardiovascular disease and obesity (McMillen, et al., 2008).

The DOHaD hypothesis is based on David Barker and his colleagues’ original

geographical experiments which showed that high systolic blood pressure was linked to low

birth weight (Barker, et al., 1989). From this original study, Barker proposed that challenges

in utero caused by maternal undernutrition lead to fetal programming events that allow the

fetus to successfully grow through the remaining gestational period by adapting its

metabolism to nutrient availability. However this then predisposes the offspring to suffer

from cardiovascular disease in adult life (Barker, 1993). Since then multiple epidemiological

studies have added proof of this developmental origin of disease (Kwong, et al., 2000,

Campbell, et al., 1996) and the theory has spread to include more than just increased

cardiovascular disease. For example, an embryo exposed to maternal low protein is more

likely to have increased adiposity (Watkins, et al., 2011) and increased anxiety behaviour

(Watkins, et al. 2008) once it reaches adulthood.

The DOHaD hypothesis has continued to develop since its discovery and ideas such as

the thrifty phenotype hypothesis (Hales & Barker 1992) and the predictive adaptive response

hypothesis (Gluckman, et al., 2005a) have contributed to the understanding of developmental

origins of disease. The thrifty phenotype hypothesis states that fetal malnutrition induces a

mechanism of nutritional thrift in the developing fetus, causing an immediate survival

1

advantage of the developing offspring in utero via differential organ growth (Hales & Barker

2001). Different tissues in the body have a hierarchy in relation to necessity for short term

adaptive advantages and therefore tissues such as muscle and liver show reduced growth in

response to under nutrition to preserve brain development (Wells 2011). This early altered

growth permanently affects the function of the offspring, which leads to an increased risk of

disease in adult life, for example fetal malnutrition can reduce endocrine pancreas

development that leads to less insulin production and increased insulin resistance, which

causes a predisposition towards type 2 diabetes (Hales & Barker 2001).

The predictive adaptive response (PAR) hypothesis builds on the thrifty hypothesis.

This hypothesis states that the fetus uses maternal nutrition to predict nutrient availability

postnatally and therefore early adaptations and fetal programming in development are aimed

at creating a benefit in adult life, rather than just an immediate benefit as seen in the thrifty

hypothesis (Gluckman & Hanson, 2004). This predisposes the offspring to adult metabolic

disease when there is a mismatch in the predicted and the actual post natal environment and

the greater the mismatch, the greater the risk of disease (Gluckman, et al., 2005b). Tests on

mouse models have proven that offspring which are malnourished in utero and then given a

rich nutritional diet postnatally, have significantly reduced life spans (Ozanne and Hales,

2004). Gluckman proposed that PARs are the reason that developing societies changing to a

resource rich environment from an impoverished environment have greatly increased

numbers of people suffering from metabolic syndrome (Gluckman, et al., 2005a).

DOHaD is a challenge to the pre-existing theory that the risk of chronic disease is

dependent on genetics and adult lifestyle. David Barker’s theory explains how a disease such

as cardiovascular disease often associated with affluence has become most common in the

poorest areas of Britain. Challenges to the fetus during critical periods of development

increase the risk of heart disease in adult life (Barker & Martyn, 1992). DOHaD doesn’t

2

discount the importance of adult lifestyle factors such as smoking and diet, instead it works

alongside these factors.

1.2 Developmental Re-programming Caused by Maternal Malnutrition

Studies in both humans and animals have determined that maternal malnutrition, often

undernutrition, leads to alterations in fetal programming. An early human model study in

1988 observed that fetal undernutrition during conception was linked to low fetal birth weight

and therefore this undernutrition was likely to be adversely influencing embryonic

development (Wynn & Wynn, 1988). In 1996 it was shown that fetal programming caused by

undernutrition leads to long term health issues. A low protein diet fed to mothers during late

gestation led to their offspring having significantly increased blood pressure at 40 years of

age compared to the offspring of mothers fed a control diet throughout gestation (Campbell,

et al., 1996).

In 1997 a human study showed that fetal malnutrition can lead to fetal programming

that alters the development of specific tissues. Mothers with a low protein but high

carbohydrate diet during pregnancy gave birth to offspring with significantly lower skeletal

muscle tissue and once these offspring reached adult life, they were more susceptible to

coronary heart disease and type 2 diabetes (Godfrey, et al., 1997).

Additionally, rats fed a low protein diet just prior to pregnancy and then throughout

gestation, gave birth to offspring that had increased systolic blood pressure compared to

control mice (Langley & Jackson, 1994). Several different low protein concentrations were

tested (6, 9, 12% by weight) and offspring showed that there was an inverse relationship

3

between maternal protein intake and the offspring systolic blood pressure. An experiment

using a guinea pig model showed that undernutrition (85% ad libitum intake) throughout the

mother’s pregnancy led to decreased fetal birth weight and altered adult cholesterol

homeostasis (Kind, et al., 1999). Male offspring showed an exaggerated response to

cholesterol loading, taking in around 30% more than control offspring due to altered fetal

programming in utero.

Animal models using rats have determined the effects of a maternal low protein

throughout gestation, showing that fetal growth is altered differently at different stages of

pregnancy. Up to day 20 of gestation, fetal growth is actually increased but from day 20

onwards growth is retarded so that by the time the offspring reaches term, they are more

likely to be low birth weight compared to control pups (Langley-Evans, et al., 1996a) which

is known to be an indicator that the offspring will be more susceptible to disease in adulthood

(Barker, et al., 1989). This growth retardation affects overall length, as well as specific

organs such as skeletal muscle and liver

1.3 Developmental Re-programming caused by challenges to the fetal environment

specifically during the preimplantation period

Fetal programming is differentially sensitive to environmental challenges depending on

the period in gestation when the fetus is exposed to the nutritional challenge. Evidence from

the 1945 Dutch hunger winter famine has shown this to be true. Children conceived during

the famine and therefore exposed only to malnutrition during conception and early

development, developed an increased glucose tolerance in adult life (De Rooij, et al., 2006).

This phenotype was not observed in offspring who were conceived prior to the famine and so

4

Figure 1 – A) Diagrammatic representation of the mouse blastocyst including the 2 different cell lineages and the Zona pellucida. Modified from Rolstan & Rossant, 2010)B) Light Microxcope image of a mouse blastocyst. Trophectoderm (TE), Inner Cell Mass (ICM) and Blastocoel (BC) are labelled. (Modified fromMarikawa and Alarcon, 2009)

were only exposed to undernutrition during late fetal development. Therefore fetal

programming must be possible from very early stages of development. This is further proven

by an experiment using a rat model which showed that rat mothers fed a low protein diet only

during the first 7 days of pregnancy, resulted in the male offspring still having a significantly

increased risk of hypertension in adult life (Langley-Evans, et al., 1996b).

The first experimental model to test fetal malnutrition just the preimplantation period

was performed by Kwong, et al (2000) and they showed that fetal reprogramming can occur

if the challenge is presented at the

preimplantation stage alone. Mouse dams were

fed a low protein diet (9% casein), known to

cause increased blood pressure in adult offspring

when fed throughout gestation, only during the

preimplantation period and then were fed a

normal protein diet (18% casein) for the

remainder of the pregnancy. The offspring

produced still showed the increased systolic

blood pressure phenotype and had significantly

lower birthweights. Figure 1 A and B show the

differentiation between the TE and ICM cell

lineage in a mouse blastocyst and this experiment

also observed a reduction in the inner cell mass

(ICM) and trophectoderm (TE) cell numbers,

indicating a reduced mitotic index. These results

proved that fetal reprogramming was occurring

during the preimplantation period and that it

5

A)

B)

began soon after the low protein challenge. Further experimental proof that reprogramming

can occur by the blastocyst stage was shown by Watkins, et al (2008) who transferred

blastocysts collected from mothers fed a low protein diet, into mothers who were fed a

control NPD diet (prior to the transfer and afterwards for the remainder of gestation) and

observed that the offspring still produced a phenotype associated with low protein diet

induced adaptations.

The periconceptual period of development is highly vulnerable, as the maternal nutrient

environment provides metabolic information to the developing embryo that regulates its

progression through preimplantation development and can trigger fetal programming events

(Fleming, et al., 2012). There is increasing evidence, that human assisted reproductive

technologies which pose a challenge to the preimplantation embryo, may be linked to the

offspring having low birth weight (Schieve, et al., 2002) and being more susceptible to a

number of imprinting disorders such as beckwith-wiedeman syndrome (Gicquel, et al., 2003).

Furthermore in a follow up experiment, performed on 131 IVF children compared against

131 control offspring, Ceelen et al (2008) showed that IVF children had significantly

increased systolic and diastolic blood pressure and that this could not be explained by birth

weight or post natal environmental factors, suggesting that the altered phenotype was due to

the challenge to the preimplantation embryo of being exposed to an in vitro culture

environment.

The mechanistic understanding of why these phenotypes persevere from challenges

during the preimplantation period in humans is largely unclear however, but several animal

models are now providing more information. A mouse model experiment showed that

mothers fed Emb-LPD produced male and female offspring with significantly higher blood

pressure than control offspring (Watkins, et al, 2008). Watkins, et al., (2010) discovered that

this increased blood pressure was caused by an increased expression of angiotensin

6

converting enzyme (ACE) in the offspring of dams fed Emb-LPD, localised to the lungs of

males and the serum of females. ACE overexpression leads to increased angiotensin II

production which stimulates vasoconstriction and therefore causes the increase in blood

pressure in Emb-LPD offspring.

Fetal programming has also been associated with increased risk of glucose intolerance

and diabetes in later life (Phillips, 2002). An experiment using a rat model showed how dams

fed Emb-LPD produce offspring that have altered hepatic gene expression (Kwong, et al.,

2007). This altered gene expression resulted in a gender specific response whereby male

fetuses had increased expression of phosphoenolpyruvate carboxylase, a rate-limiting enzyme

in gluconeogenesis, and female fetuses displayed increased expression of 11B-hydroxysteroid

dehydrogenase type 1, which increases glucocorticoid production. These phenotypes lead to

both the male and female offspring having increased insulin resistance and males having

increased likelihood of diabetes because of the preimplantation low protein diet.

Epigenetic changes via histone

modification and DNA methylation

have also been observed to be altered

by maternal diet, summarised in figure

2. Specifically, changes to the

epigenome have been recorded in

response to Emb-LPD (Lillicrop, et al.,

2005). Effects on the regulation of the

glucocorticoid receptor (GR) and

peroxisomal proliferator-activated

receptor (PPAR) by a restricted protein

diet were detected using methylation-

7

Figure 2 – An overview of the way in which maternal diet can alter the epigenome of a developing fetus, leading to an increase risk of adult disease (Lillycrop & Burdge (2011)

sensitive PCR to measure DNA methylation and real time PCR to show mRNA expression.

The study showed that restricted protein diet offspring had significantly reduced gene

methylation and increased gene expression of both GR and PPAR compared to control pups.

The preimplantation period is a critical period of epigenetic control, as the methylation and

imprinting of fetal genes after the demethylation of parental genome begins from the

blastocyst stage, around E3.5 in mice (Reik & Walter, 2001).

1.4 Potential role for branched-chain amino acids and insulin levels in induction of

preimplantation fetal programming

When rat dams are fed a low protein diet exclusively during the periconceptional period

(Emb-LPD), the maternal environment for the preimplantation embryo is altered

metabolically by reduced plasma insulin and essential amino acid levels (Kwong et al., 2000).

Although it is currently unknown whether these changes in concentrations are nutrient

messages that the preimplantation embryo uses to stimulate fetal programming, evidence is

continually emerging that suggests a potential role for amino acids and insulin in this process.

Insulin receptors are first expressed in the preimplantation embryo at the compaction

stage (Harvey & Kaye, 1988) and therefore from this point insulin may play a potential role

in embryo nutrient sensing. Additionally human preimplantation embryos do not produce

their own insulin or insulin-like growth factor-1 (IGF-1) and cannot as they do not produce

the necessary mRNA transcripts until later in development (Lighten, et al., 1997), but as

stated above they do produce their own receptors for both insulin and IGF-1. Therefore

insulin could be a good indicator of predicted post natal nutrient environment as the

preimplantation embryo is dependent on the supply from the maternal environment.

8

Another reason for insulin to be a potential regulator of fetal programming is that it is a

moderator of blastocyst growth and is therefore required for optimal fetal growth (Kaye &

Gardner, 1999). In vitro experiments have shown that insulin stimulates protein synthesis in

both the TE and ICM cell lineages of the developing blastocyst but increased insulin only

causes an increase to cell proliferation in the ICM cell lineage, having no effect on the TE

(Harvey & Kaye 1990). An increase in the ICM increases the pool of cells available that go

on to form the developing offspring and therefore increased insulin aids optimal fetal growth

as stated above.

The fluid within the female reproductive tract also contains free amino acids (Miller &

Schultz, 1987) alongside insulin and other nutrients. Lane & Gardner (1997) performed an

experiment that showed that amino acid signalling increased blastocyst growth and showed

the importance of amino acids in the successful development of the preimplantation embryo.

They determined that, prior to the 8 cell stage embryo, exposure to increased non-essential

amino acids greatly increased the rate of cleavage but that exposure to increased essential

amino acids had no effect. However from the 8-cell stage onwards, they discovered that

embryo exposure to increased essential amino acids not only increased the rate of cleavage

throughout the remaining stages of preimplantation development, but also increased the

number of ICM cells produced in the blastocyst. Whereas increased non-essential amino

acids at this stage no longer increased cleavage but did increase blastocoel development.

In vitro experiments have also shown that amino acid availability is important at the late

blastocyst stage to aid in the development of a mature TE lineage, which is capable of

successfully invading the maternal stroma during implantation (Martin & Sutherland, 2001).

This experiment showed the inability of mouse embryos to form a cell outgrowth on

fibronectin, an accepted in vitro model of implantation (Wartiovaara, et al., 1979), when

9

cultured in a medium lacking amino acids. The correct phenotype returned when blastocysts

were switched back into a control medium. Martin & Sutherland further proved that amino

acids were stimulating successful outgrowth via an mTOR pathway, as introduction of

rapamycin to the culture medium containing amino acids blocked the successful formation of

a cell outgrowth.

An experiment by Eckert, et al., (2012) examined methods of induction of adverse fetal

programming and the onset of early compensatory responses by the embryo when mouse

dams are fed Emb-LPD. They discovered that when dams were fed Emb-LPD, the level of

insulin and combined levels of essential and branched-chain amino acids in the maternal

serum, were significantly depleted by the blastocyst stage (E3.5). Furthermore analysis of

maternal uterine fluid (UF) highlighted a potential role in ‘nutrient signalling’ of the

branched-chain amino acids specifically (isoleucine, leucine and valine), which were all

significantly lower in Emb-LPD UF at E.3.5 and still lower by E.4.5 when many other amino

acids had returned to levels comparable in control diet UF.

1.5 Mammalian Target of Rapamycin (mTOR) regulation of cell growth and development.

Mammalian target of rapamycin (mTOR) is a protein kinase family involved in

regulating eukaryotic protein synthesis, the conversion of mRNA to protein, depending on the

availability of certain nutrients (Proud, 2002). The mTOR signal transduction pathway, is

used by the developing fetus, to allow it to adapt its own cell growth and development in

response to nutrient availability (Maloney & Rees, 2005).

Mammalian target of rapamycin complex 1 (mTORC1) is an mTOR signalling protein

kinase, that phosphorylates serine and threonine residues on target proteins via its conserved

C-terminal kinase domain (Mayer & Grummt, 2006). mTORC1 senses certain nutrient levels

10

such as amino acid, particularly leucine (Bruhat, et al., 2002) and energy levels (Asnaghi, et

al., 2004) and depending on the availability of these nutrients, mTORC1 regulates processes

involved in cell growth and development, mainly protein translation as stated above but a role

has also been discovered in protein degradation and actin organisation (Mayer, et al., 2004).

mTORC1 regulates protein translation via phosphorylation of proteins which are

repressors of translation initiation complex proteins (Hay, 2004). When there is high nutrient

availability, particularly amino acids, active mTORC1 phosphorylates and inactivates eIF-

4BP1, which is an inhibitor of the cap binding protein eIF-4E. In its active form, eIF-4E can

freely to bind to eIF-4G to help form the eIF-4F complex, and bind to the 5’ cap of target

mRNA to allow translation to occur (Proud, 2002). Active mTORC1 also phosphorylates S6

kinase 1 (S6K1), which when phosphorylated releases eIF3 which it is bound to in its basal

state. The released eIF3 is now free to bind to the 40s ribosomal subunit and bring it into

contact with the translation preinitiation complex (PIC) (Holz, et al., 2005).

Many nutrients including insulin regulate the mTORC1 activity via a Tuberculosis

sclerosis complex (TSC1 & 2) dependant pathway (Dowling, et al., 2010). Insulin binds

directly to the insulin receptor (IR) kinase which then triggers a phosphorylation cascade,

IR PI3K Akt/PKB TSC1/2, this then leads to the activation of Rheb-GTPase which

activates mTORC1 (Cheng, et al., 2010). Amino acids however, activate Rheb-GTPase via a

poorly understood TSC1 independent mechanism which still activates Rheb-GTPase to active

mTORC1, potentially via the activity of the kinase Vacuolar sorting protein 34 (Vps34)

(Proud, 2007).

Eckert, et al., (2012) implicated a role for mTOR signalling in developmental

reprogramming caused by Emb-LPD. When mouse dams were fed Emb-LPD, the level of

phosphorylated S6 was reduced as was the ratio of phosphorylated to total S6 protein,

indicating a reduction in mTORC1 signalling. As previously stated, in this experiment Emb-

11

LPD fed mothers had significantly decreased insulin and branched-chain amino acids in their

serum and UF, which are both known activators of mTOR signalling, summarised in figure 3,

and this lack of nutrient availability is likely to alter blastocyst programming via an mTORC1

dependant process.

12

1.6 Differing TE cell numbers as an indication of fetal programming

When mice mothers are fed a reduced protein diet during the weeks leading up to mating

then the developing blastocysts that form have a reduced ICM to TE ratio (Mitchell, et al.,

2009). An experiment by Eckert, et al., (2012) replicated this result showing that after mouse

mothers are fed Emb-LPD, by E3.75 the produced blastocysts have a significantly increased

number of trophectoderm lineage cells and total cell number, but they also showed that this

increased TE cell number led to a significant increase in cell outgrowth after blastocyst

hatching. Therefore the increased proliferation of trophectoderm lineage cells may be viewed

as a compensatory response being triggered by the blastocyst stage in response to maternal

LPD. Stimulated fetal growth is known to be an indicator of an increased risk of adult disease

as specified by the DOHaD hypothesis (Watkins, et al., 2008).

1.7 Aims and Hypothesis

The work of Eckert, et al., (2012) suggested a potential role of branched-chain amino

acid and insulin availability in the nutrient signalling that leads to fetal reprograming in the

preimplantation embryo. Therefore in this project I will examine using an in vitro model

with mouse blastocysts whether low levels of these nutrients, branched-chain amino acids

and insulin, are sufficient on their own to bring about the same phenotype that has been

observed in the Emb-LPD in vivo studies (Eckert, et al., 2012: Watkins, et al., 2008) or if

there is another cause behind the observed phenotypes.

The first aim is to detect whether culture media, containing different levels of insulin and

branched-chain amino acids, is sufficient to significantly alter the rate at which

13

preimplantation embryos develop from the 2-cell stage (E1.5) through to the late blastocyst

stage (E3.75). My hypothesis is that there will be no significant change, as this has been

tested for in an ‘in vivo’ study and blastocysts from mouse dams fed Emb-LPD were not at a

significantly different stage of development compared to embryos from control NPD mothers

when removed at E3.5 (Eckert, et al., 2012).

The second aim is to assess whether the 4 different treatment groups of:

1 – Normal Insulin Normal BCAA 2 – Normal Insulin Low BCAA

3 – Low Insulin Normal BCAA 4 – Low Insulin Low BCAA,

cause a significant difference in the TE:ICM ratio which has been determined to be suitable

evidence that a compensatory response has been triggered in the developing embryo by the

blastocyst stage (Eckert, et al., 2012), which suggests there will be an adult predisposition to

metabolic disease in the offspring.

My hypothesis is that there will be a significant difference, at least in treatment group

4 and potentially in treatment groups 2 and 3 because Emb-LPD experiments have

consistently shown evidence of early fetal programming (Kwong, et al., 2000). In addition

there has been evidence of significantly lower insulin and branched-chain amino acids in the

uterine fluid and serum of mothers fed an Emb-LPD (Eckert, et al., 2012).

This area of research is important, to clarify the role of maternal diet as a cause of

fetal reprogramming leading to an increased susceptibility to disease in the offspring’s later

life. With a greater understanding of this area, an effective public health guideline can be

produced to improve the diets of pregnant women (Langley-Evans, et al., 1999). This greater

understanding will aid in preventing the increased rate of adult metabolic diseases observed

in modern society (Gluckman, et al., 2005a).

14

Table 1 – Uterine fluid concentrations of free amino acids at 3.5 days of development from mice fed Normal protein diet (modified from Eckert, et al., (2012).

2. Materials and Methods

2.1 Creation of Culture Media for Four Treatment Groups

This experiment tested 4 different treatment groups and therefore 4 different culture

media were produced;

1 – Normal Insulin Normal BCAA 2 – Normal Insulin Low BCAA

3 – Low Insulin Normal BCAA 4 – Low Insulin Low BCAA,

These culture media consisted of KSOM medium supplemented with serum insulin (1 ng/ml)

and UF amino acid concentrations (Table 1) based on the levels recorded in the serum and UF

of mouse dams fed normal protein diet (18% casein) (Eckert, et al., 2012). ‘Normal’ level of

these nutrients was 100% the value recorded by Eckert, et al., (2012) and ‘Low’ levels were

50% of that. In the ‘Low’ BCAA media, only the branched-chain amino acid (isoleucine,

leucine and valine) concentrations were lowered to 50%, all other amino acid concentrations

remained at 100% of the observed value.

2.2. Dissection and procuring of 2-cell stage embryos

15

Embryos were collected at the 2-cell stage (E1.5) from non-superovulated MF1

mouse dams. Dams were sacrificed via cervical dislocation, the oviducts were then removed

and placed into saline solution. Next the oviducts were transferred into H6-BSA drops in a

petri dish and 2-cell embryos were flushed out under a dissection microscope. Total collected

2-cell embryos were split equally into 4 groups, 1 group per different culture media. Embryos

were then moved via mouth pipette into 30µl drops of culture media, and incubated for 66

hours at 37oC in a 5% CO2 incubator. 66 hours is the time required in vitro for embryos to

reach the late blastocyst stage from the 2-cell stage.

2.3 Differential Cell Staining

2.3.1 Differential Cell Staining Materials

Acid tyrodes – sigma, pH2.3, heated to 37oC

Anti-DNP & H6-PVP solution – Anti-DNP stock solution (1 mg/ml in distilled water) and

H6-PVP in the ratio 20.8µl of anti-DNP to 29.2µl of H6-PVP.

Bizbenzimide (Hoechst) – sigma, 2.5 mg/ml in distilled water

Ethanol - 100% pure ethanol

Guinea Pig Complement – low tox guinea pig complement, diluted 1:10 ratio with H6-BSA

H6-Bovine Serum Albumin (BSA)

H6-Polyvinylpyrrolidone (PVP)

Propidium Iodide (PI) – sigma, 1 mg/ml in distilled water, aloquat stored at -20oC

Trinitrobenzene sulphonic acid (TNBS) – Picrysulfonic acid, diluted 1:10 in PBS

2.3.2 Differential Cell Staining Protocol

16

After the 66-hour incubation period, embryos were removed from the incubator and

observed under a microscope to record what stage of development they had progressed to.

Stages recorded were morulae, early blastocyst (blast), mid blast, late blast and hatching

blast. After this only the blastocysts which were at the late blastocyst stage were used and the

differential cell staining protocol began. The first step of which was removal of the

blastocysts’ zona pellucida via acid tyrodes. Each treatment group required 1 cavity block

containing 500µl of acid tyrodes, both the cavity block and tyrodes were pre-heated to 37oC.

Blastocysts were transferred from the incubation culture media into the acid tyrodes via

mouth pipette. They were then observed under a microscope to determine when the zona

pellucida had been fully dissolved. Once the zona was removed the blastocysts were moved

via mouth pipette to another cavity block (1 per treatment group), filled with 1ml of handling

medium containing BSA (H6-BSA), for 20 minutes, to allow the blastocyst to recover before

beginning the second step.

During the second step, blastocysts were incubated for 10 minutes at room

temperature in a 50µl drop of TNBS. To start this step, one small drop (25µl) and one large

drop (50µl) of TNBS were prepared per treatment group in a ‘Cellstar’ petri dish. Blastocysts

were moved between drops via mouth pipette, first into the small drop to wash away the H6-

BSA then into the large drop for the incubation period. After incubation blastocysts were

washed through 3, 50µl drops of H6-PVP to remove the TNBS.

Next, blastocysts were incubated for 10 minutes at room temperature in an anti-DNP

with H6-PVP solution (described in 2.3.1). Once again, a small drop (8µl) and a large drop

(40µl) of anti-DNP solution were prepared per treatment group in a Cellstar petri dish.

Blastocysts were washed through the small drop via mouth pipetting, to ensure the incubation

stage was in pure anti-DNP solution. After incubation, blastocysts were washed through 3,

50µl drops of H6-PVP to remove the anti-DNP solution.

17

Figure 4 – An example fluorescence microscopy image of a diffrerenctial cell stained late blastocyst. Trophectoderm = pink, Inner cell mass = Blue

Next, blastocysts were stained with propidium iodide (PI). 1 culture drop was

prepared per treatment group containing 50µl of guinea pig complement and 4µl of PI and

Blastocysts were incubated in these drops for 10 minutes at 37oC. After this incubation

blastocysts were washed through 3, 50µl drops H6-BSA. Finally embryos were fixed for 3

hours at 4oC in 990µl of absolute ethanol and 10µl of Bizbenzimide (Hoechst) in a 4 well

plate, one well per treatment group.

2.4. Blastocyst Picture Acquisition

After blastocysts were fixed, they were mounted onto slides for observation with a

fluorescence microscope. Blastocysts were first moved from the Ethanol and Hoechst

solution into absolute ethanol, for 5 minutes, to wash them. Coverslips were then washed

with methanol in preparation and then a drop of glycerol was placed in the centre for

mounting the embryos. Blastocysts were moved

from the ethanol to the glycerol drops via

mouth pipette in groups of 3-5. The coverslip

was then placed on top of the glycerol drop so

the slide could be used under the microscope.

Coverslips were examined under a

fluorescence microscope. Using the program

MetaMorph®, embryos were searched for on

the coverslip and then an image, such as the one

in figure 4, of each individual embryo was

captured. Images were captured at 20x magnification. Cell numbers were counted manually

18

using MetaMorph®, ICM cells in blue and TE cells in red, the values were then combined to

determine the total cell number.

2.5 Statistical Analysis

146 embryos were incubated at the 2-cell stage per treatment group. From these

embryos; 140 from treatment group 1, 142 from treatment group 2, 141 from treatment group

3 and 142 from treatment group 4 were useable for the first part of the experiment (outlined

in 2.3.2). The from these embryos, 95 from treatment group 1, 104 from treatment group 2,

96 from treatment group 3 and 101 from treatment group 4 were at the late blastocyst stage

by the end of incubation and were usable for cell number counts after the differential cell

staining protocol.

The data from the first experiment determining the stage reached in development

after 66 hour incubation period was analysed using a chi-square test. The data from the

second experiment determining relative TE and ICM cell proliferation, turned out to not be

normally distributed and no easy conversion was available, therefore this data was analysed

via non-parametric ANOVA based on ranks. A Kruskal-Wallis test was first used to determine

if any of the 4 treatment groups had produced a significantly different result from the others

and then Mann-Whitney tests would be used on individual pairs of treatment groups to

determine which group had produced a significant difference in trophectoderm cell

allocation.

19

Figure 5 – The effect of 4 different culture media containing different insulin and BCAA levels on the number of embryos that reached each specific stage in development as a percentage of the total embryos first cultured. The key relates to the four different treatment groups; N = normal, L = Low, I = Insuling & BCAA = Branched chain amino acids.

3. Results

3.1 Activity of Embryos in Development

The first experiment carried out was to assess the activity of embryos developing in

response to incubation in each different culture medium. This was done by recording what

stage of development each embryo had reached by the end of the 66-hour incubation period

from the 2-cell stage. The data was analysed using a chi-squared test. Figure 5 shows the

number of embryos, per treatment group, that had reached each stage of development as a

percentage of the total number of embryos originally cultured (n = 146 per treatment group).

A chi squared test was performed to determine whether the activity of the embryos was

significantly altered by culture in different concentrations of BCAA and insulin. The test

determined that there was no significant difference in embryo activity, p ≠ 0.99 for total

embryo, total blastocyst and late blastocyst comparison. There was slightly more variation

20

Figure 6 – The effect of 4 different culture media containing different insulin and BCAA levels on the average number of cells produced in each cell lineage of the blastocyst (Trophectoderm and Inner Cell Mass) and the total cell number. The key relates to the four different treatment groups; N = normal, L = Low, I = Insuling & BCAA = Branched chain amino acids.

when comparing the number of embryos which reached the hatching blastocyst stage but p-

values still showed an non-significant relationship. Therefore embryos did not develop

significantly slower or faster in response to a depleted BCAA and/or insulin environment,

when incubated from the 2-cell stage for 66 hours.

3.2 Effect of Depleted Branched-chain Amino Acids and Insulin on Fetal Programming

The second experiment assessed how changing the levels of insulin and branched-

chain amino acids, effected the distribution of cells within the two different possible cell

lineages of the late blastocyst, the trophectoderm and the inner cell mass. To measure this,

cells in the late blastocyst were differentially stained, depending on which lineage they

belonged to and then values were recorded by imaging via fluorescence microscopy. The data

was analysed using a Kruskal-Wallis non-parametric ANOVA based on ranks. Figure 2,

shows the number of cells in the trophectoderm and inner cell mass and then the total number

of cells on average in the blastocysts when cultured in each of the four different treatment

21

Figure 7 – The effect of 4 different culture media containing different insulin and BCAA levels on ratio of Trophectoderm to inncer cell mass cells formed in blastocysts. The key relates to the four different treatment groups; N = normal, L = Low, I = Insuling & BCAA = Branched chain amino acids.

groups. The statistical tests performed showed that there was no significant difference in the

cell number of the trophectoderm lineage (p = 0.212), inner cell mass lineage (p = 0.580) or

total cells (TC) (p = 0.493) at the late blastocyst stage, between the 4 treatment groups.

Therefore the effect of lowering the insulin and branched-chain amino acid concentrations

and then the combined effect of lowering them both in the supplemented KSOM media had

no significant effect on the proliferation of cells in either of the two different lineages.

Furthermore there was no significant change to the total proliferation of cells within the entire

blastocyst by the late blastocyst stage.

As well as examining just the number of cells present, I also calculated the TE:ICM

ratio and the ICM/TC proportion. These were both then statistically analysed, also with a

Kruskal-Wallis test. Figure 7 shows the TE:ICM ratio, on average, of blastocysts cultured in

the four different treatment groups. The Kruskal-Wallis test performed on this data showed

that there was no significant difference (p = 0.217) in the TE:ICM ratio of blastocysts

cultured in the different

treatment groups.

Therefore the depleted

insulin and branched-

chain amino acid levels

appear to have had no

significant effect on the

allocation of cells into

each specific lineage

during blastocyst

development.

22

Figure 8 – The effect of 4 different culture media containing different insulin and BCAA levels on the proportion of the total number of cells in a blastocyst which is made up by the inner cell mass lineage The key relates to the four different treatment groups; N = normal, L = Low, I = Insuling & BCAA = Branched chain amino acids.

To further confirm this result, the ICM/TC parameter shown in figure 8 was

statistically analysed. If one treatment group produced a significantly lower value then

incubation in that

concentration of BCAA

and insulin would have

caused the blastocysts to

respond by investing

more cells into the TE

lineage compared to the

ICM lineage during

early development. A

Kruskal-Wallis test

performed on the data

in figure 8 determined

that there was no significant difference (p = 0.217) in the ICM/TC parameter between either

of the four different treatment groups. Therefore this confirmed the results from the analysis

of the TE:ICM ratio, that culturing 2-cell embryos up to the late blastocyst stage in depleted

insulin and branched-chain amino acid levels, both individually and the combined effect, had

no impact on allocation of cells to a specific cell lineage during blastocyst development.

Overall these results show that, after statistical analysis of the data acquired, the four

different culture media that were used to test the effect of depleting insulin and branched-

chain amino acid levels, did not have a significant effect on the developing blastocyst. There

was not a significant change in the total number of cells in either the TE or ICM, or in the

total number of cells by the late blastocyst stage. Furthermore, the relative allocation of cells

to the two different possible lineages, TE and ICM, was not affected by the different culture.

23

4. Discussion

4.1 Comparison with Original Hypothesis

The result from the first half of the experiment confirms my first hypothesis stated in

the introduction, that depleting the levels of branched–chain amino acids and insulin had no

effect on the rate at which embryos developed from the 2-cell stage, over the 66-hour

incubation period. Statistical analysis via chi-squared tests proved that any variation observed

in figure 5 was almost solely due to random sampling variability, as the p-values were all

very high. This has been observed before in another study using in vitro culture. Velazquez,

et al., (2012) recently performed an experiment where embryos obtained at the 2-cell stage

were incubated in several different culture media containing gradually depleting

concentrations of branched-chain amino acids for 66-hours, by which time it embryos should

reach the late blastocyst stage. The embryos cultured in the lower levels of branched-chain

amino acids did not develop at a significantly different rate compared to those cultured at

normal levels, which is consistent with my results.

The results from the second part of my experiment disproved my hypothesis stated in

the introduction. The effects of reducing the branched-chain amino acids and/or insulin levels

by 50% in the different culture media turned out to not cause a significant effect on the

proliferation of cells in the developing blastocyst. Furthermore, there was no significant

effect on the allocation of cells to either the trophectoderm or the inner cell mass lineage

within the blastocyst. Therefore, despite the evidence from Eckert, et al., (2012) which

showed a potential link between reduced branched-chain amino acids and insulin and the

onset of preimplantation fetal reprogramming, this data instead suggests that no fetal

programming had occurred in the cultured embryos by the late blastocyst stage, in response

to the reduced nutrient levels compared to the control values (obtained from treatment group

1).

24

From looking at the distribution of data in figures 6-8, blastocysts incubated in the

two culture media which contained 50% lower levels of branched-chain amino acids, showed

no indication that fetal programming was occurring, which was then confirmed by the

statistical analysis. This was a big surprise as there is prior experimental evidence that

branched-chain amino acids specifically are known to be present at significantly lower

concentrations in maternal UF when dams are fed a low protein diet and embryos exposed to

a low protein diet undergo fetal programming. The results from this evidence could therefore

suggest that branched-chain amino acid levels are not part of the nutrient sensing mechanism

used by the preimplantation embryo when exposed to an Emb-LPD environment.

However, although not a significant difference, the data from figure 6 appears to

indicate a potential increase in the trophectoderm cell number when embryos were incubated

in the two culture media which contained 50% insulin (treatment groups 2 and 4) compared

to the control. Based on this observation I performed a Mann-Whitney test to analyse the

trophectoderm cell numbers between treatment group 2 (low insulin with normal BCAA) and

group 1 (the control containing normal insulin and normal BCAA levels). Although still not

significantly different, the result was very close as the p value came out to be, p = 0.058, just

0.008 away from a significant result. Furthermore the error bars in figure 6 are not very large,

but a large enough to conceive that this statistical test could be showing a significant result.

This is a strong indication that the depleted insulin level may have been stimulating fetal

programming and causing this slight increase in trophectoderm lineage by the late blastocyst

stage. If this were true then that would indicate that fetal programming may have occurred in

these embryos by the late blastocyst stage, as expected from my original hypothesis.

There are several points to consider about why these results came out as non-

significant despite the evidence which formed my hypothesis in the introduction, that a

reduction in branched-chain amino acids and insulin in an in vitro culture medium would be

25

sufficient to cause cause fetal reprogramming to occur by the late blastocyst stage. First of all

other factors in the initial in vivo LPD diet studies may have been causing the observed

change in blastocyst phenotypes, secondly there may have been some technical issues during

the experimental process that have led to a misleading conclusion. Furthermore the fact that a

replicate in vitro study may not be completely accurate when trying to determine the effects

of nutrient level changes in vivo, as the in vitro culture environment presents many unique

stresses of its own to the developing embryo. Finally, based on these results there are a

number of follow up studies that could be done to confirm or disprove the results from this

study that depleting the availability of branched-chain amino acids and insulin to a

developing preimplantation embryo is not sufficient to cause an onset of fetal reprogramming

by the late blastocyst stage of development.

4.2 Potential Methods Used by the Preimplantation Embryo to Communicate with the

Maternal Environment after an Emb-LPD

A study in 2000 was the first to show that when mouse dams are fed a low protein diet

selectively during the preimplantation period it is sufficient to cause fetal programming

events to occur that resulted in the offspring having an increased risk of cardiovascular

disease in later life (Kwong, et al., 2000). During this experiment it was noted that several

nutrient levels in the maternal serum changed significantly during this time in response to the

low protein diet, such as a decrease in insulin and essential amino acids and an increase in

glucose. This caused these nutrients to be proposed as potential nutrient messengers used by

the developing embryo to sense the low protein environment and adapt its development.

26

Since then, multiple research studies have focussed on the effects that a low protein

diet causes on the specific nutrient availability within the environment of the preimplantation

embryo. If this is known it can lead to an understanding as to what the developing embryo is

sensing specifically which causes it to adapt its fetal programming in an attempt to be better

suited to its environment.

My project studied the effects of lowering just branched-chain amino acid and insulin

levels on the blastocyst to undergo fetal programming because recent experimental evidence

suggested both of these nutrients may have a role in nutrient sensing by the preimplantation

embryo (Eckert, et al., 2012). However the fact that my results turned out to be non-

significant implies that an embryo low protein diet causes other significant changes to the

preimplantation uterine environment which are signalling to the preimplantation embryo and

triggering fetal programming. These could include changes to amino acid, glucose or pH

levels.

Amino Acids

An experiment performed by Eckert, et al., (2012), observed a significant decrease in

branched-chain amino acid levels in the maternal serum and also in the uterine fluid of mouse

dams fed a low protein diet exclusively during the preimplantation period of development.

However, the branched-chain amino acid levels in the maternal UF of dams fed low protein

diet, compared to control dams, were only uniquely lower than other amino acid

concentrations at E4.5, at which point mouse embryos begin the implantation process

(Stephens, et al., 1995). Before this stage, at E.3.5, other amino acids (glycine, histidine,

lysine, taurine and threonine) were also measured to be lower in the maternal UF, at trend

level ‘P < 0.1’, compared to the UF of control dams. During my experiment, depleted

27

branched-chain amino acid levels were found to have no effect on the developing embryo

when cultured to E3.75, the late blastocyst stage. This lack of response could therefore mean

that prior to E4.5, the preimplantation embryo may be responsive to amino acids, such as

those observed in the experiment by Eckert, et al., (2012) and not just the branched-chain

amino acids. At this stage the embryo may require a combined signal from multiple to amino

acids before fetal programming can be triggered. The branched-chain amino acids may play a

more significant role in nutrient sensing by E4.5, because by this point they are the only

amino acids to be at a significantly lower concentration in the UF of Emb-LPD fed dams,

compared to the UF of control dams.

The potential role of amino acids, other than just the branched-chain amino acids, in

nutrient sensing by the preimplantation embryo has been suggested by several other

experiments. One such experiment observed that both rat and mouse dams have significantly

decreased serum levels of threonine and phenylalanine in response to an Emb-LPD (Petrie, et

al., 2002). In another experiment where mouse dams were fed a low protein diet during the

preimplantation period, there was a significant reduction in the level of 6 different amino

acids (the three branched-chain amino acids and methionine, proline and threonine) in the

maternal serum (Kwong, et al., 2000). All of these results suggest that the developing

preimplantation embryo is sensitive to more amino acids than just the branched-chain amino

acids. This is a possible explanation as to why depleting the concentration of just the

branched-chain amino acids was not sufficient to trigger fetal reprogramming in my

experiment, as the embryo requires nutrient signals from multiple amino acids.

28

Insulin

In my project, there appeared to be some impact when depleting the level of insulin in

the culture media. The Mann-Whitney test performed in section 4.1 suggests there may have

been a response to the 50% decrease in insulin concentration via an increase in the number of

trophectoderm cells produced by the late blastocyst stage. Although not a significant

difference, the data in figures 6-8 appears to indicate a potential impact on fetal programming

caused by the reduction in insulin levels, as the trophectoderm cell number, total cell number

and TE:ICM ratio is consistently highest in treatment groups 2 and 4. However these results

were not significant and therefore it is not safe to assume that the insulin was directly

responsible for this variation and it was not simply due to random sampling variability.

Insulin may not be involved in nutrient signalling by the preimplantation embryo, previously

a reduction in insulin levels caused by a low protein diet has been observed in several

experiments (Kwong, et al., 2000: Eckert, et al., 2012) but, the reduction has always been

observed in the maternal serum and not in the UF, to which the blastocyst is directly exposed.

Glucose

Additionally, a low protein diet causes changes to the maternal environment that

were not examined in my project. The rodent preimplantation embryo is known to be

sensitive to glucose; a maternal hyperglycaemic state has been observed to significantly

lower the expression of facilitated glucose transporters in the developing blastocyst (Moley,

et al., 1999), in attempt by the blastocyst to adapt to a predicted high glucose environment in

adult life. Furthermore, an in vitro culture experiment proved that when rat preimplantation

embryos were incubated in a high glucose medium (17mM) it caused, irreversible inhibition

of the TE and ICM lineage growth and an increase in apotosis within the blastocyst.

29

Going back to the original paper examining a low protein diet specifically during the

preimplantation period, Kwong, et al., (2000), a significantly increased glucose content in the

maternal serum was observed by day four of gestation, in response to an Emb-LPD. This

result has since been replicated; mouse dams fed a low protein diet during the

preimplantation period were recorded as having significantly higher glucose in the maternal

serum by day 3.5 in development, around the time of blastocyst formation (Eckert, et al.,

2012). Therefore these results suggest that a hyperglycaemic environment surrounding the

developing blastocyst was likely contributing to the fetal reprogramming that was observed in

response to the low protein diet. This fetal programming led to increased likelihood of adult

diseases such as hypertension in these offspring.

However, other experiments have observed that high glucose concentrations in the

maternal environment caused by Emb-LPD have little affect once the embryo reaches later

stages of gestation (Fernandez-Twinn, et al., 2003: Kwong, et al., 2007), therefore further

experimentation is required to fully understand the effect of glucose concentration on fetal

development.

4.3 Evaluation of experimental techniques

I believe that the method I used during this experiment was well designed and suitable

to produce an accurate result. The differential cell staining protocol has been previously

tested and used on blastocysts by researchers at the University of Southampton (Velazquez, et

al., 2012) and therefore it was a reliable method to measure TE and ICM lineage and

blastocyst total cell numbers. Additionally, the concentrations of amino acids used for the

30

‘normal’ level used in the different treatment groups, were discovered by an in vivo

experiment which directly measured the concentration of every amino acid in the uterine

fluid of mouse dams which had been fed a normal protein diet (Eckert, et al., 2012).

Similarly, the concentration of insulin used for the ‘normal’ level in the different treatment

groups was measured in the same in vivo experiment (Eckert, et al., 2012), in the serum of

dams fed a normal protein diet. Furthermore a 50% decrease from the normal values was a

sufficient reduction, therefore if any change was observed it could be reliably assumed to be

due to the nutrient reduction.

However, there are several technical issues that arose during the experimental process

which potentially could have contributed to the second half of the experiment disagreeing

with my hypothesis and coming to the conclusion that lowering the concentration of

branched-chain amino acids and insulin was not sufficient to trigger fetal programming in the

developing. First of all, towards the beginning of the data collection period, when I was still

not fully experienced at performing the staining protocol, some embryos from each treatment

group were lost due to errors whilst using the mouth pipette. This improved after the first few

repeats as I became increasingly proficient at the experimental procedure and no more

embryos were lost, however there is potential that some of those lost embryos from the early

repeats could have shown a fetal programming phenotype and may have contributed to a

significant result in the end.

Additionally, for the initial repeats of the differential cell staining protocol, during the

final step blastocysts were moved directly from the H6-BSA drops in the previous step, into

the ethanol and bizbenzimide stain solution for 3 hour incubation, with no interim washing

step. A problem arose whereby some of the blastocysts were attaching to the base of the four

well plate and when these blastocysts had to be moved prior to mounting onto a microscope

slide, some of the cells in the trophectoderm layer were being ripped away from the rest of

31

the blastocyst and remaining attached to the base of the plate. This would lead to a poor

image being captured with an incorrect number of trophectoderm cells being recorded. To

resolve this issue, embryos were first placed into a washing well containing 990µl of absolute

ethanol and 10µl of bizbenzimide to completely remove any H6-BSA from the previous step,

which had been responsible for the blastocysts attaching previously. Furthermore when

embryos were then moved from this washing well into their final staining solution, some air

bubbles were blown in via the mouth pipette so that the blastocysts did not sink straight to the

bottom. These changes to the method eradicated the problem of the blastocysts attaching to

the four well plates and images taken during fluorescence microscopy became consistently

accurate. However, this problem in the initial repeats meant that cell counts from the

blastocysts who had attached during the first few repeats had to be removed from the final

statistical analysis. If these blastocysts had undergone fetal programming it may have

contributed to a significant result in the end, but this cannot be known for sure.

Finally there may have been issues with pH levels during the experimental

procedure. pH regulation is an important part of cell homeostasis for all mammalian cells

(Phillips, et al., 2000). To maintain homeostasis, mouse embryos use a HCO3-/Cl- transporter

to relieve alkalosis (Zhao, et al., 1995) and two different transporters to relieve acidosis, an

Na+/H+ antiporter (Steeves, et al., 2001) and an Na+,HCO3–/Cl– exchanger (Phillips, et al.,

2000). When mouse embryos are subjected to low pH during the preimplantation period,

often caused in vivo by an increase in urea within the maternal UF, the developing

blastocysts show a significant decrease in cell proliferation and increase in apoptosis

(Bystriansky, et al., 2012).

In my experiment the pH of the final culture media was not specifically tested, however the

pH of the KSOM media was known to be ~pH 7.4. Although unlikely, there is potential that

contamination could have caused the culture media to become a low enough pH that it could

32

have caused a challenge to the developing blastocyst which led them to have a reduced cell

proliferation that led to an insignificant result. Furthermore, during the experiment at the end

of the incubation stage, blastocysts were submitted to acid tyrodes to dissolve the zona

pellucida. If the embryos were left in too long during this step then the low pH challenge

could have caused an increase in trophectoderm apoptosis. However this is again very

unlikely as the challenge would only have been very brief and furthermore steps in the

protocol were in place to prevent embryo elongated exposure to the acid tyrodes. Only 2-3

embryos at a time were moved into the tyrodes so that it would be easy to observe when the

zona pellucida had fully dissolved

4.4 Effects of In Vitro Culture

An in vitro model was required for this experiment as it is the only possible method

that allowed me to observe the effects of specifically reducing just branched-chain amino

acid and insulin concentrations on developing blastocysts. An in vivo model can only change

the diet fed to dams and observe its effects, it cannot be used to alter specific aspects of

uterine fluid composition. Therefore the use of an in vitro model was appropriate for this

experiment. However, placing preimplantation embryos into an in vitro culture environment

is known to induce a number of cellular stresses on the embryo (Chandrakanthan, et al.,

2006). For example, in vitro culture has been seen to induce oxidative stress via stimulating

the embryo to produce a significantly increased amount of reactive oxygen species (Nasr-

Esfahani & Johnson, 1991) and also cause metabolic imbalance where the developing

embryos don’t use enough endogenous resources if cultured in vitro (Leese 2002).

33

Furthermore, experimental evidence has shown mouse embryos placed into in vitro

culture produce a reduced number of TE lineage cells compared to in vivo embryos

(Giritharan, et al., 2012). This was an issue for my experiment as for part of my results I was

trying to record an increase in TE lineage cell number to represent the onset of fetal

programming by the blastocyst stage. Therefore it may have been the case that the embryos

being incubated in culture during this experiment were repressing the proliferation of the TE

lineage and could not increase TE proliferation properly in response to the nutrient challenges

in the different treatment groups

If it were possible to alter the concentrations of branched-chain amino acids and

insulin in the uterine fluid in vivo, without changing the concentration of any other nutrient,

then the blastocysts could be observed without having to consider the potential effects caused

by the stress of being in an in vitro culture. These hypothetical in vivo conditions may have

produced a phenotype associated with fetal reprogramming as was expected from the

literature.

4.5 Future Work

From the results of my project, there are a number of follow up experiments I would

suggest to first confirm these results and secondly examine more possibilities of how a low

protein diet limited to the preimplantation period signals to the fetus to initiate fetal

programming, causing the fetus to adapt to the predicted post-natal environment. It

First of all there are several experiments possible to confirm the results of this

experiment, that reducing the concentration of branched-chain amino acids and insulin by

34

50% in the direct embryo environment is not sufficient to induce fetal programming in the

preimplantation embryo by the late blastocyst stage. In this experiment I used the

proliferation of the TE lineage, the TE:ICM ratio and the ICM/TC parameter to indicate

whether fetal programming had occurred, however there are other methods of ensuring that

fetal programming has occurred by the late blastocyst stage (E3.75). Therefore I would repeat

the experiment with the same 4 treatment groups and the same protocol for incubation, but

instead of performing differential cell staining at the end, the analysis of whether fetal

programming had been initiated would be done by either detecting the level of mTOR

signalling within the blastocysts, measuring blastocyst outgrowth formation on suitable

surface such as fibronectin and reinserting embryos into mothers fed a normal protein diet to

see if the offspring still showed an enhanced growth phenotype.

mTOR signalling

Both insulin and branched-chain amino acids are known to activate mTORC1

signalling (Bruhat, et al., 2002: Dowling, et al., 2010). When mouse dams are fed a low

protein diet specifically during the pre-implantation period, the capacity for mTOR signalling

is reduced in the blastocysts due to a significant reduction in the level of phosphorylated S6

(a downstream target of mTORC1) compared to blastocysts from dams fed a normal protein

diet (Eckert, et al., 2012). Therefore, mTORC1 activity of the blastocysts from the four

different treatment groups of my experiment would be analysed using the protocol used by

Eckert, et al., (2012). This would confirm that a reduction in insulin and/or branched-chain

amino acid levels is sufficient to cause a reduction in mTORC1 signalling and therefore is

responsible for the reduction in mTORC1 signalling observed in blastocysts in vivo when

dams are fed Emb-LPD. A reduction in mTORC1 signalling would suggest an occurrence of

35

fetal programming by the blastocysts stage, because the mTORC1 signalling pathway acts as

a sensor for maternal nutrient levels in the developing blastocyst (Eckert, et al., 2012).

Blastocyst Outgrowth Formation

A predicative adaptive response by the developing blastocyst in response to a reduced

nutrient environment is an increase in placental development to ensure that the developing

fetus can obtain sufficient nutrients throughout gestation and maintain necessary fetal growth

(Watkins, et al., 2008). When mouse dams were fed a low protein diet during the

preimplantation period, blastocysts were extracted at E3.5 and allowed to hatch, attach and

initiate spreading in an in vitro setting (Eckert, et al., 2012). The blastocysts from dams

which were fed a low protein diet displayed an increased capacity for trophoblast spreading

after attachment compared to blastocysts from dams fed a normal protein diet.

Therefore in a repeat of my experiment, to confirm whether embryos are undergoing

fetal programming in response to the depletion in BCAA and insulin levels, a set of

blastocysts from each treatment group would be cultured in media containing ‘normal’

nutrient levels for a further 96 hours after the initial 66 hour incubation, to allow outgrowth

formation (Velazquez, et al., 2012). If the embryos from a particular treatment group show a

significant increase in trophoblast outgrowth it would indicate fetal programming had

occurred by the end of the initial incubation period in response to the decreased concentration

of BCAA and/or insulin.

36

Restricted intrauterine growth

In response to a low protein diet fed exclusively during the preimplantation period,

fetal growth is accelerated later on in gestation (Watkins, et al., 2008). Mouse embryos

transferred at the blastocyst stage from dams fed Emb-LPD into dams fed a NPD, maintain

the enhanced growth phenotype later on in pregnancy (Watkins, et al., 2008), which proves

that fetal programming has occurred by the blastocyst stage. Therefore, to determine whether

depleting branched-chain amino acid and insulin levels is sufficient to trigger fetal

programming, embryos from the 4 treatment groups would be transferred after 66 hours of

incubation from the 2-cell to the late blastocyst stage, into dams which have been fed a NPD

and will continue to be fed NPD throughout the remainder of gestation. At Day 17 of

gestation, offspring would be collected, weighed and compared to determine whether the

growth rates had been significantly different between embryos from the different treatment

groups. A significantly increased weight compared to the control offspring would indicate

that fetal programming had occurred by the late blastocyst stage in response to reduced

BCAA and insulin levels.

After these three different experiments are completed, the results of my project, that

depleting branched-chain amino acids and/or insulin alone is not sufficient to cause fetal

programming by the late blastocyst stage, will either be confirmed or contradictory results

will instead indicate a role for insulin and branched-chain amino acid levels in nutrient

sensing by the preimplantation embryo.

If these proposed future experiments were to confirm the results from my project, then

additional future research should be focussed on the effect of altering the concentrations of

other nutrients, such as glucose, in vitro on triggering fetal programming. Additionally, the

37

maternal environment after an Emb-LPD should be further investigated to determine other

possible causes of the resulting fetal programming observed in developing offspring. This

will further improve scientific understanding of how the preimplantation embryo

communicates with its environment during development.

4.6 Concluding remarks

The results from this project have shown that depleting the concentration of branched-

chain amino acids and/or insulin by 50% from the values found in the UF and serum of dams

fed a normal protein diet, is not sufficient to change the rate at which embryos develop and is

also not sufficient to trigger fetal programming in preimplantation embryos by the late

blastocyst stage. A possible role for insulin was observed, but it was not a significant

relationship and so further experimentation is required to prove whether insulin truly was

having an effect of the onset of fetal programming.

These results can improve the scientific understanding of how the preimplantation

embryo communicates with its environment in utero. They indicate that branched-chain

amino acids and insulin levels may not play as significant a role in nutrient sensing by the

blastocyst stage as the current literature suggests. Therefore, the focus of research may need

to change away from assessing the effect had by branched-chain amino acid concentrations

and insulin levels, to investigating additional methods by which the preimplantation embryo

may be communicating with its environment when the mother has had low protein

environment, or other forms of poor maternal nutrition, before interactions between the

preimplantation embryo and its environment can be fully understood.

38

Ultimately a greater understanding of how maternal diet triggers fetal programming is

required before advancements in accurate preventative methods can be made to reduce the

chance of fetal programming occurring in pregnant women. Additionally this understanding

will also help prevent embryos undergoing fetal programming during in vitro culture. If the

specific nutrient imbalances which can cause embryos to trigger fetal programming are

identified, it can lead to an improvement in culture media used during assisted reproductive

technologies such as in vitro fertilisation.

39

References

Asnaghi, L., Bruno, P., Priulla, M., Nicolin, A., (2004) mTOR: a protein kinase switching between life and death Pharmacology Research 50, pp.545–549

Barker, D.J., (1993) The intrauterine origins of cardiovascular disease Acta Paediactrica 82, pp.93-99

Barker, D.J., Martyn, C.N., (1992) The Maternal and foetal origins of cardiovascular disease Journal of Epidemiol Community Heath 46(1), pp.8-11

Barker, D.J., Osmond, C., Golding, J., Kuh, D., Wadsworth, M.E., (1989) Growth in utero, blood pressure in child hood, and mortality from cardiovascular disease British Medical Journal 298, pp.564–567

Bruhat, A., Averous, J., Carraro, V., Zhong, C., Reimold, A.M., Kilberg, M.S., Fafournoux, P., (2002) Differences in the molecular mechanisms involved in the transcriptional activation of the CHOP and asparagine synthetase genes in response to amino acid deprivation or activation of the unfolded protein response Journal of Biological Chemistry 277, pp.48107-48114

Bystriansky, J., Burkus, J., Juhas, S., Fabian, D., Koppel, J., (2012) Effect of increased urea levels on mouse preimplantation embryos developed in vivo and in vitro Bulletin of the Veterinary Institute in Pulawy 56, pp.211-216

Campbell, D.M., Hall, M.H., Barker, D.J., Cross, J., Sheill, A.W., Godfrey, K.M., (1996) Diet in pregnancy and the offspring’s blood pressure 40 years later British Journal of Obstetrics and Gynaecology 103(3), pp.273-280

Ceelen, M., Van Weissenbruch, M.M., Vermeiden, J.P.W., Van Leeuwen, F.E., Delemarre-vam de Waal, H.A., (2008) Cardiometabolic Differences in Children Born After in Vitro Fertilization: Follow-up Study Journal of Clinical Endocrinology & Metabolism 93(5), 1682

Chandrakanthan, V., Li., A., Chami, O., O’Neil, C., (2006) Effects of in vitro fertilization and embryo culture on TRP53 and Bax expression in B6 mouse embryos Reproductive Biology and Endocrinology 4, 61

Cheng, Z., Tseng, Y., White, M.F., (2010) Insulin signalling meets mitochondria in metabolism Trends in Endocrinology & Metabolism 21(10), pp.589-598

De Hertogh, R., Vanderheyden, I., Pampfer, S., Robin, D., Dufrasne, E., Delcourt, J., (1991) Stimulatory and Inhibitory Effects if Glucose and Insulin on Rat Blastocyst Development In

40

Vitro Diabetes 40, pp.641-647

Dowling, R.J., Topisirovic, I., Fonseca, B.D., Sonenberg, N., (2010) Dissecting the role of mTOR: lessons from mTOR inhibitors Biochimica et Biophysica Acta 1804(3), pp.433-439

Fernandez-Twinn, D.S., Ozanne, S.E., Ekizoglou, S., Doherty, C., James, L., Gusterson, B., Hales, C.N., (2003) The maternal endocrine environment in the low-protein model in intra-uterine growth restriction British Journal of Nutrition 90, pp.815-822

Fleming, T.P., Velazquez, M.A., Eckert, J.J., Lucas, E.S., Watkins, A.J., (2012) Nutrition of females during the peri-conceptional period effects on foetal programming and health of offspring Animal Reproduction Science 130, pp.193-197

Gicquel, C., Gaston, V., Mandelbaum, J., Siffroi, J.P., Flahault, A., Le Bouc, Y., (2003) In vitro fertilization may increase the risk of Beckwith-Wiedemann syndrome related to the abnormal imprinting of the CN1OT gene American Journal of Human Genetics, 72, pp.1338–1341

Giritharan, G., Dell Piane, L., Donjacour, A., Esteban, F.J., Horcajadas, J.A., Maltepe, E., Rinaudo, P., (2012) In bitro culture of mouse embryos reduces differential gene expression between inner cell mass and trophectoderm 19(3), pp.243-252

Gluckman, P.D., Hanson, M.A., (2004) The developmental origins of the metabolic syndrome Trends in Endocrinology and Metabolism 15(4), pp.183-187

Gluckman, P.D., Hanson, M.A., Pinal, C., (2005b) The developmental origins of adult disease Msternal & Child Nutrition 1(3), pp.130-141

Gluckman, P.D., Hanson, M.A., Spencer, H.G., (2005a) Predictive Adaptive Responses and Human Evolution Trends in Ecology and Evolution 20(10), pp.527-533

Hales, C.N., Barker, D.J., (1992) Type 2 (non-insulin-dependent) diabetes mellitus: the thrifty phenotype hypothesis Diabetologia 35(7), pp.595-601

Hales, C.N., Barker, D.J., (2001) The Thrifty Phenotype Hypothesis British Medical Bulletin 60(1), pp.5-20

Harvey, M.B., Kaye, P.L, (1988) Insulin stimulates protein synthesis in compacted mouse embryos Endocrinology 122(3), 1182

Harvey, M.B., Kaye, P.L., (1990) Insulin increases the cell number of the inner cell mass and stimulates morphological development of mouse blastocysts in vitro Development 110, pp.963-967

41

Haye, N., Sonenberg, N., (2004) Upstream and downstream of mTOR Genes & Development 18, pp.1926-1945

Holz, M.K., Ballif, B.A., Gygi, S.P., Blenis, J., (2005) mTOR and S6K1 Mediate assembly if the translation preinitiation complex through dynamic protein interchange and ordered phosphorylation events Cell 123(4), pp.569-580

Kaye, P.L., Gardner, H.G., (1999) Preimplantation access to maternal insulin and albumin increases fetal growth rate in mice Human Reproduction 14(12), pp.3052-3059

Kind, K.L., Clifton, P.M., Katsman, A.I., Tsiounis, M., Robinson, J.S., Owens, J.A., (1999) Restricted fetal growth and the response to dietary cholesterol in the guinea pig. American Journal of Physiology 6(2), pp.1675-1682

Kwong, W.Y., Miller, D.J., Wilkins, A.P., Dear, M.S., Nevillewright, J., Osmond, C., Zhang, J., Fleming, T.P., (2007) Maternal low protein diet restricted to the preimplantation period induces a gender-specific change on hepatic gene expression in rat foetuses Molecular Reproduction and Development 74(1), pp.48-56

Kwong, W.Y., Wild, A.E., Roberts, P., Willis, A.C., Fleming T.P., (2000) Maternal undernutrition during the preimplantation period of rat development causes blastocyst abnormalities and programming of postnatal hypertension Development 127, pp.4195-4202

Lane, M., Gardner, D.K., (1997) Differential regulation of mouse embryo development and viability by amino acids Journal of Reproduction and Fertility 109, pp.153-164

Langley, S.C., Jackson, A.A., (1994). Increased systolic blood pressure in adult rats induced by foetal exposure to maternal low protein diet Clinical.Science 86(2), pp.217-222.

Langley-Evans, S.C., Gardner, D.S., & Jackson, A.A., (1996a) Association of disproportionate growth of fetal rats in late gestation with raised systolic blood pressure in later life Journal of Reproduction and Fertility 106, pp.307–312.

Langley-Evans, S.C., Sherman, R.C., Welham, .J., Nwagwu, M.O., Gardner D.S., Jackson, A.A., (1999) Intrauterine programming of hypertension: the role of the renin-angiotensin system Biochemical Society Transactions 27(2), pp.88-93

Langley-Evans, S.C., Welham, S.J., Sherman, R.C., Jackson, A.A., (1996b) Weanling rats exposed to maternal low-protein diets during discrete periods of gestation exhibit differing severity of hypertension Clinical Sceince 91(5), pp.607-615

Leese, H.J., (2002) Quiet please, do not disturb: a hypothesis of embryo metabolism and viability BioEssays 24(9), pp.845-849

42

Lighten, A.D., Hardy, K., Winston, R.M., Moore, G.E., (1997) Expression of mRNA for the insulin –like growth factors and their receptors in human preimplantation embryos Molecular Reproductive Develoment 47(2), pp.134-139

Lillykrop, K.A., Burdge, G.C., (2011) The effect of nutrition during early life on the epigenetic regulation of transcription and implications for human diseases Journal of Nutrigenetics and Nutrigenomics 4(5), pp.248-260

Lillycrop, K.A., Phillips, E.S., Jackson, A.A., Hanson, M.A., Burdge, G.C., (2005) Dietart protein restriction of pregnant rats induces and folic acid supplementation prevents epigenetic modification of hepatic gene expression in the offspring Journal of Nutrition 135(6), pp.1382-1386

Maloney, C.A., Rees., W.D., (2005) Gene-nutrient interactions during fetal development Reproduction 130, pp.401-410

Marikawa, Y., Alarcon, V.B., (2009) Establishment of trophectoderm and inner cell mass lineages in the mouse embryo Molecular Reproduction and Development 76(11), pp.1019-1032

Martin, P.M., Sutherland, A.E., (2001) Exogenous amino acids regulate trophectoderm differentiation in the mouse blastocyst through an mTOR-dependant pathway Developmental Biology 240, pp.182-193

Mayer, C., Grummt, I., (2006) Ribosome Biogenesis and cell growth: mTOP coordinates transcription by all three classes of nuclear RNA polymerase Oncogene 25(48), pp.6384-6391

McMillen, I.C., MacLaughlin, S.M., Muhlhausler, B.S., Gentili, S., Duffield, J.L., Morrison J.L., (2008) Developmental Origins of Health and Disease: The Role of Periconceptional and Foetal Nutrition Basic & Clinical Pharmacology & Toxicology 102(2), pp.82-89

Miller, J.G., Schultz, G.A., (1987) Amino acid content of preimplantation rabbit embryos and fluids of the reproductive tract Biology of Reproduction 36(1), pp.125-129

Mitchell, M., Schulz, S.L., Armstrong, D.T., Lane, M., (2009) Metabolic and mitochondrial dysfunction in early mouse embryos following maternal dietary protein intervention Biology of Reproduction 80(4), pp.622-630

Moley, K.H., Maggie, M., Chi, Y., Mueckler, M.M., (1998) Maternal hyperglycemia alters glucose transport and utilization in mouse preimplantation embryos Endocrinology and Metabolism 275(1), pp.38-47

43

Nasr-Esfahani, M.M., Johnson, M.H., (1991) The origin of reactive oxygen species in mouse embryos cultured in vitro Development 133, pp.551-560

Ozanne, S.E., Hales, C.N., (2004) Catch-up growth and obesity in male mice Nature 427, pp.411–412

Petrie, L., Duthie, S.J., Rees, W.D., McConnell, J.M.L., (2002) Serum concentrations of homocysteine are elevated during early pregnancy in rodent models of fetal programming British Journal of Nutrition 88, pp.471-477

Phillips, D., (2002) Endocrine programming and fetal origins of adult disease Trends Endocrinology and Metabolism 13(9), pp.363.

Phillips, K.P.,  Léveillé, M.C., Claman, P., Baltz, J.M., (2000) Intracellular pH regulation in human preimplantation embryos Human Reproduction 15(4), pp.896-904

Proud, C,G., (2002) Regulation of mammalian translation factors by nutrients European Journal of Biochemistry 269(22), pp.5338-5349

Proud, C.G., (2007) Amino acids and mTOR signalling in anabolic function Biochemical Society Transactions 35(5), pp.1187-1190

Ralston, A., Rossant, J., (2010) The genetics of induced Pluripotency Reproduction 139, pp.35-44

Schieve, L.A., Meikle, S.F., Ferre, C., Peterson, H.B., Jeng, G., Wilcox, L.S., (2002) Low and Very Low Birth Weight in Infants Conceived with Use of Assisted Reproductive Technology New England Journal of Medicine 346, pp.731-737

Steeves, C.L., Lane, M., Bavister, B.D., Phillips, K.P., Baltz, J.M., (2001) Differences in Intracellular pH Regulation by Na+/H+ Antiporter among Two-Cell Mouse Embryos Derived from Females of Different Strains Biology of Reproduction 65(1), pp.14-22

Stehpens, L.E., Sutherland, A,E., Klimanskaya, I.V., Andrieux, A., Meneses, J., Pedersen, R.A., Damsky, C.H., (1995) Deletion of beta 1 intergrins in mice results in inner cell mass failure and peri-implantation lethality Genes & Development 9, pp.1883-1895

Velazquez, M.A., Eckert, J.J., Fleming, T.P., (2012) Exposure time is critical in blastocyst adaptations to branched chain amino acid depletion Reproduction Fertility and Development 25(1), pp.194

Wartiovaara, J., Leivo, I., and Vaheri, A., (1979) Expression of the cell surface-associated glycoprotein, fibronectin, in the early mouse embryo, Developmental Biology 69(1), pp.247–257

44

Watkins, A.J., Lucas, E.S., Torrens, C., Cleal, J.K., Green, L., Osmond, C., Eckert, J.J., Gray, W.P., Hanson, M.A., Fleming, T.P., (2010) Maternal low-protein diet during mouse pre-implantation development induces vascular dysfunction and altered renin-angiotensin-system homeostasis in the offspring British Journal of Nutrition 103, pp.1762–1770

Watkins, A.J., Lucas, E.S., Wilkins, A., Cagampang, F.R., Fleming, T.P., (2011) Maternal periconceptional and gestational low protein diet affects mouse offspring growth, cardiovascular and adipose phenotype at 1 year of age PLoS One 6, e28745.

Watkins, A.J., Ursell, E., Panton, R., Papenbrock, T., Hollis, L., Cunningham, C., Wilkins, A., Perry, V.H., Sheth, B., Kwong, W.Y., Eckert, J.J., Wild, A.E., Hanson, M.A., Osmund, C., Fleming, T.P., (2008) Adaptive Responses by Mouse Early Embryos to Maternal Diet Protect Fetal Growth but Predispose to Adult Onset Disease, Biology of Reproduction 78, pp.299-306

Wells, J.C.K., (2011) The Thrifty Phenotype: An Adaptation in Growth or Metabolism? American Journal of Human Biology 23, pp.65-75

Wynn, A., Wynn, M., (1988) Nutrition around conception and the prevention of low birthweight Nutrition and Health 6(1), pp.37-52

Zhao, Y., Chauvet, P.J.P., Alper, S.L., Baltz, J.M., (1995) Expression and Function of Bicarbonate/Chloride Exchangers in the Preimplantation Mouse Embryo The Journal of Biological Chemistry 270(41), pp.24428-24434

45